4. The Proper Name Issachar

Cambridge Semitic Languages and Cultures The Historical Depth of the Tiberian Reading Tradition of Biblical Hebrew AAron D. HornkoHl THE HISTORICAL DEPTH OF THE TIBERIAN READING TRADITION OF BIBLICAL HEBREW The Historical Depth of the Tiberian Reading Tradition of Biblical Hebrew Aaron D. Hornkohl https://www.openbookpublishers.com © 2023 Aaron D. Hornkohl This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International (CC BY-NC 4.0). This license allows you to share, copy, distribute and transmit the text; to adapt the text for non-commercial purposes of the text providing attribution is made to the authors (but not in any way that suggests that they endorse you or your use of the work). Attribution should include the following information: Aaron D. Hornkohl, The Historical Depth of the Tiberian Reading Tradition of Biblical Hebrew. Cambridge Semitic Languages and Cultures 17. Cambridge, UK: Open Book Publishers, 2023, https://doi.org/10.11647/OBP.0310 Copyright and permissions for the reuse of many of the images included in this publication differ from the above. Copyright and permissions information for images is provided separately in the List of Illustrations. Further details about CC BY-NC licenses are available at, https://creativecommons.org/ licenses/by-nc/4.0/ All external links were active at the time of publication unless otherwise stated and have been archived via the Internet Archive Wayback Machine at https://archive.org/web Updated digital material and resources associated with this volume are available at https://doi.org/10.11647/OBP.0310#resources Every effort has been made to identify and contact copyright holders and any omission or error will be corrected if notification is made to the publisher. Semitic Languages and Cultures 17. ISSN (print): 2632-6906 ISSN (digital): 2632-6914 ISBN Paperback: 978-1-80064-980-4 ISBN Hardback: 978-1-80064-981-1 ISBN Digital (PDF): 978-1-80064-982-8 DOI: 10.11647/OBP.0310 Cover image: T-S AS 8.129. A leaf from a Cairo Geniza biblical codex containing Gen. 30.17–20 and showcasing Moshe Moḥe’s non-standard Tiberian pointing of the standard Tiberian pronunciation of Issachar (see within, ch. 4), courtesy of the Syndics of Cambridge University Library. Cover design: Jeevanjot Kaur Nagpal In fond memory of Michael Rand, friend and colleague CONTENTS Acknowledgments............................................................. ix Abbreviations ................................................................... xi Introduction ....................................................................... 1 Part I: Conscious Replacement ........................................ 43 1. The Tetragrammaton ................................................... 45 2. ‫ת־פנֵ י יְּ הוָ ה‬ ְּ ‫ ֵל ָראֹות ֶא‬and Similar ........................................ 55 3. Ketiv-Qere Euphemisms ................................................ 67 Part II: Linguistic Developments ...................................... 81 4. The Proper Name Issachar ........................................... 83 5. ‫ ִל ְּק ַראת‬Liqra(ʾ)ṯ ............................................................. 93 6. The 2MS Endings ........................................................ 101 7. The 2FS Endings ......................................................... 145 8. The Qere Perpetuum ‫הוא‬.............................................. ִ 161 9. The 2/3FPL Endings ................................................... 171 10. Nifalisation .............................................................. 183 11. Hifilisation ............................................................... 209 12. Pielisation ................................................................ 253 13. Hitpaelisation .......................................................... 289 14. Ṭɛrɛm Qaṭal .............................................................. 319 viii The Historical Depth of the Tiberian Reading Tradition 15. Ha-Qaṭal ................................................................... 347 16. Wayyiqṭol ................................................................. 373 17. 1st-person Wayyiqṭol ................................................ 385 18. I-y We-yiqṭol for Weqaṭal .......................................... 441 Conclusion ..................................................................... 463 References ...................................................................... 481 Index .............................................................................. 519 ACKNOWLEDGMENTS The research for the individual studies that comprise this volume was conducted over a span of years. I am grateful to multiple colleagues and students who participated in workshops, lectures, and symposia where I had the opportunity to present and refine approaches and arguments. The interaction has been invaluable. Benjamin Kantor, Ethan Jones, and Geoffrey Khan have read and commented upon large sections of this work. I appreciate their insight. Thanks to Ben Outhwaite for suggesting the cover image. I owe special thanks to my colleagues in the Hebrew and Semitics division of the Faculty of Asian and Middle Eastern Studies, Cambridge University, as well as to several post-doctoral fellows, graduate students, and undergraduates, for having supported my efforts in the preparation of this volume by taking on portions of my teaching, supervisorial, and administrative duties over some of past three years. This afforded me more time to research and write than I would otherwise have had. These include Geoffrey Khan, Michael Rand (z"l), Benjamin Kantor, Dorota Molin, Magdalen Connolly, Wiktor Gebski, Estara Arrant, and Yoav Ronel. The writing of a few chapters and most of the manuscript’s final preparation took place during the summer of 2022, while I was teaching in Middlebury College’s Summer School of Hebrew. The town’s coffee houses—Haymaker Bun Company, Lost Monarch Coffee, Royal Oak Coffee, and Little Seed Coffee Roasters—combined provided me a comfortable and inspiring x The Historical Depth of the Tiberian Reading Tradition home away from home and helped fuel the process of writing, editing, and consistency checking. A great deal of the research for this volume and the writing thereof took place during lockdowns in connection with the Covid 19 pandemic. These entailed extended periods working at home. Though I wish to acknowledge the challenges and heartbreak that many understandably associate with this time, in my own case, the home-office environment was so pleasant and conducive to working, that it has been difficult to return to a preCovid routine. For this I thank Anna, for her unique combination of ever-reliable and ever-renewing forms of love and support, as well as Yonatan, Yoel, Emily—and Moses (our English bulldog)— for helping me try to achieve a healthy work-life balance. ABBREVIATIONS 1 1st-person 2 2nd-person 3 3rd-person A Aleppo Codex b. Talmud Bavli BA Biblical Aramaic BCE Before the Common Era BDSS Biblical Dead Sea Scrolls BH Biblical Hebrew BS Ben Sira c. circa C common (gender) CBH Classical Biblical Hebrew CE Common Era ch. chapter chs chapters col. column cols columns DSS Dead Sea Scrolls DSSBA Dead Sea Scrolls Biblical Aramaic DSSBH Dead Sea Scrolls Biblical Hebrew F feminine fn. footnote intr. intransitive L Leningrad Codex LBH Late Biblical Hebrew (Esther, Daniel, Ezra–Nehemiah, Chronicles) xii The Historical Depth of the Tiberian Reading Tradition LBH+ Late Biblical Hebrew+ (Ps. 119; Job 1–2; 42.7–17; Qohelet, Esther, Daniel, Ezra–Nehemiah, Chronicles) ln. line lns lines LXX Septuagint m. Mishna M masculine MT Masoretic Textual Tradition NBDSS Non-biblical Dead Sea Scrolls PL plural QA Qumran Aramaic QH Qumran Hebrew RH Rabbinic Hebrew S singular SH Samaritan Hebrew SP Samaritan Pentateuch t. Tosefta TA Targumic Aramaic TAM Tense, Aspect, Mood TBH Transitional Biblical Hebrew TJ Targum Jonathan TO Targum Onqelos tr. transitive y. Talmud Yerushalmi INTRODUCTION This book focuses on an acknowledged dimension of the received Tiberian Masoretic biblical tradition the extent and significance of which is seldom fully appreciated: dissonance between its pronunciation and spelling arising from its composite nature. At issue are cases of linguistic disharmony wherein the written and reading components of the tradition, i.e., its consonantal text and vocalisation, diverge.1 Sometimes, such differences are explicitly signalled within the Tiberian manuscript tradition via the mechanism known as ketiv-qere and/or are noted in masoretic grammatical treatises. In many other cases, however, dissonance is not so acknowledged, and is detectable only in apparent mismatch between orthography and vowel pointing. The composite nature of the Tiberian tradition is not a novel object of enquiry; nor are apparent instances of resulting dissonance.2 Indeed, in the case of many of the individual phe1 Of course, the extant so-called consonantal text is not purely conso- nantal, as it includes numerous matres lectionis that represent vowel sounds. There is also a degree of dissonance internal to the Tiberian reading tradition itself, i.e., between vocalisation and accentuation; see M. Breuer (1980; 1981, 262); Y. Breuer (1991, 191–242; 2022); Kogut (1994); Price (2006); Revell (2015, 1–3); Habib (2021, esp. 13–14, 186– 315). 2 See Ginsberg (1934; 1937); Kahle (1959, 78–86, 100, 171–79); Barr (1981, 27, 35–36; 1984, 31; 1987, 207–22); Morag (1974); Hughes (1994); Tov (2012, 46–47); Joosten (2015); Hendel (2016, 31–32); Khan (2013a, 45–52, 68; 2013b; 2021, I:56–85); Habib (2020); Hornkohl (2020a; 2020b). © 2023 Aaron D. Hornkohl, CC BY-NC 4.0 https://doi.org/10.11647/OBP.0310.20 2 The Historical Depth of the Tiberian Reading Tradition nomena summarised in this introduction or discussed in chs 1– 18 below, scholars have previously raised the possibility of discord within the combined Tiberian written-recitation tra-dition. It is also commonplace to attribute the dissonance in question to secondary developments in the reading tradition vis-à-vis the tradition reflected in the consonantal text. Against this scholarly background, the present monograph is intended to make a pair of contributions. One is the mere collection of relevant features in a single resource. It is hoped that this will serve to improve upon the current situation, in which important discussions of Tiberian written-reading dissonance are scattered among various studies, so that the frequency of the phenomenon goes underestimated and the combined significance under-appreciated. The other innovation involves the attempt to contextualise more precisely than is often done secondary deviation of the pronunciation tradition from the ostensible earlier pronunciation reflected in the consonantal tradition. Sensing secondary development, scholars often correctly, but rather cursorily and vaguely, declare the pronunciation tradition that has been preserved in the Tiberian vocalisation anachronistic and unreliable, without plumbing its historical depth. Obviously, the pronunciation tradition predates the medieval development of the graphic symbols with which it was eventually recorded, but by how much? As is repeatedly emphasised in this study, though the Tiberian pronunciation tradition regularly preserves Iron Age features and is not immune to Byzantine and medieval developments, the regularity of meaningful affinity between its apparent Introduction 3 secondary devel-opments and acknowledged Second Temple forms of Hebrew demands that the Tiberian reading tradition be considered a product of Second Temple times. But this is not the whole story. First, because much of the Tiberian pronunciation tradition accompanies a consonantal tradition anchored in First Temple times, its linguistic testimony cannot be considered exclusively representative of the Second Temple Period. The Tiberian reading tradition may have largely crystallised in the Second Temple Period, with clear indications of drift in the direction of later norms, especially where the ambiguity of certain consonantal forms made them amenable to secondary realisations. Yet, beyond the fact that the similarity between Iron Age and Second Temple Hebrew far exceeds the difference that distinguishes them, some degree of linguistic evolution was prevented by the unambiguousness of many consonantal forms that were not amenable to secondary realisations. In other words, in the marriage of the reading and written components, the latter acted as a brake of sorts, preventing fuller development of the reading tradition in line with Second Temple linguistic conventions. Second, as is regularly stressed below, many of the secondary, characteristically late developments discussed in this study, have clear antecedents in CBH and/or Iron Age epigraphic Hebrew. This means that, while they may accurately be described as especially typical of Second Temple Hebrew, they often crop up as minority alternatives in earlier material. Thus, even in palpable cases of dissonance there is continuity between the First Temple Hebrew of the CBH consonantal tradition and of Iron Age 4 The Historical Depth of the Tiberian Reading Tradition epigraphy and Second Temple deviations in the Tiberian pronunciation tradition. 1.0. Ketiv-Qere, Qere Perpetuum, and Beyond The works that comprise the Hebrew Bible reflect diverse authors, sources, genres, locales, social groups, time periods, and secondary hands. It would be reasonable to expect substantial linguistic diversity. Yet various processes of standardisation have resulted in the levelling of a great deal of the expected diversity, so that the combined Tiberian written-reading tradition is remarkably uniform. Even so, Tiberian BH shows signs of diverse idiolects, registers, genrelects, regional dialects, sociolects, and chronolects. Another aspect of BH diversity stems from variation in the traditions in which the Hebrew Bible has been transmitted. For example, the Tiberian, Babylonian, and Samaritan traditions present different manifestations of BH, with differences ranging from pronunciation to grammar. Even within the dominant Tiberian Masoretic tradition, readers confront differences between the written and reading components of the tradition, i.e., the consonantal text and the vocalisation, respectively. In many places in the text, such dissonance is explicitly acknowledged and marked by the mechanism known as ketiv-qere. In the majority of such cases—the approximate number of which, estimated between 800 and 1500, varies depending on the manuscript and expert opinion (Yeivin 1980, 55; Ofer 2019, 92; Habib 2020, 285)—divergence between what is written (ketiv = the Aramaic passive participle ‫‘ כתיב‬written’) and what is read (qere = the Aramaic passive participle ‫קרי‬ Introduction 5 ‘read’) is indicated via vocalisation of the written form with the vowels of the form to be read, the consonants of which are given in the side or intercolumn margin. The discrepancy can involve a single letter, a whole word, or spacing between words. In other cases, the reading tradition has no parallel for a word or phrase, or, alternatively, requires the recitation of a word or words not included in the accompanying consonantal text. Sometimes, the qere specifies the meaning of a ketiv (Khan 2013a, 45–46; 2021, 33–49). In cases of consistent conflict between the written and reading components of the tradition, no marginal note signals the discrepancy between consonantal spelling and pronunciation. Rather, the vocalisation alone signals the correct reading (Khan 2021, 34). Examples include realisation of the tetragrammaton ‫ יהוה‬yhwh as ‫ יְּהֹוָ ה‬ʾăḏōnaẙ̄ ‘LORD’ (= ‫‘ ֲאד ֹנָ י‬Lord’) or ‫ יְּ הֹוִ ה‬ʾɛ̆lōhīm ‘GOD’ (= ‫ֹלהים‬ ִ ‫‘ ֱא‬God, god’); see below, ch. 1) and of ‫ירושלם‬ ̊̄ ̊̄ *yǝrūšalēm ‘Jerusalem’ (cf. ‫‘ ָש ֵ֔ ֵלם‬Salem’ Gen. 14.18) as yǝrūšalayim (see below, Introduction, §3.1). The phenomenon of consistent replacement of the ketiv with the qere is commonly known as qere perpetuum. Whatever the exact explanation for individual cases of ketiv-qere, they constitute, at their most basic level, acknowl- edged instances of divergence between the written and pronunciation traditions, wherein the latter supersedes the former for purposes of oral recitation. The ketiv-qere phenomenon is relevant to the subject of this monograph in two respects. First, many such divergences apparently reflect secondary developments in the reading tradition vis- 6 The Historical Depth of the Tiberian Reading Tradition à-vis the corresponding earlier, i.e., more original, consonantal feature, e.g., the tetragrammaton ‫ יהוה‬yhwh realised as ‫ יְּ הֹוָ ה‬or ‫יְּ הוָ ה‬ ʾăḏōnaẙ̄ ‘LORD’ or ‫ יְּ הֹוִ ה‬ʾɛ̆lōhīm ‘GOD’.3 Additionally, notwithstanding their secondary character— and despite the fact that evidence for the (inter)marginal mechanism for signalling ketiv-qere and of qere perpetuum in masoretic codices comes no earlier than medieval manuscripts—the specific forms encountered in the qere tradition are clearly not just Byzantine or medieval developments, but are rooted in antiquity. This is borne out by several pieces of evidence, be it rabbinic, textual/versional, or perceptible within the Masoretic tradition itself. First, several types of ketiv-qere are mentioned in the Talmud (Yeivin 1980, 56, §98, 58–59, §§102–4). Euphemistic qere: ‫ת"ר (=תנו רבמן) כל המקראות הכתובין בתורה לגנאי קורין אותן‬ ...‫לשבח‬ Our Sages taught: All of the scriptures that are written in the Torah in impolite language are read in language beyond reproach…’ (Megilla 25b; see below, ch. 3) Qere wela ketiv ‘read but not written’ and ketiv wela qere ‘written but not read’: ‫אמר רבי יצחק מקרא סופרים ועיטור סופרים וקריין ולא כתיבן וכתיבן‬ ...‫ולא קריין הלכה למשה מסיני‬ 3 But cf. the discussion in Hornkohl (2022), where it is emphasised that there is not always clear diachronic linguistic progression between ketiv and qere readings of more or less equal plausibility. Introduction 7 Rabbi Yitzḥak said: “The vocalisation of the scribes, and the ornamentation of the scribes, and the verses with words that are read, but not written, and those that are written, but not read, are all halakha transmitted to Moses from Sinai…” (b. Nedarim 37b) Qere perpetuum: ‫ר' אבינא רמי כתיב זה שמי לעלם וזה זכרי לדור דור אמר הקב"ה לא‬ ‫כשאני נכתב אני נקרא נכתב אני ביו"ד ה"א ונקרא אני באל"ף דל"ת‬ Rabbi Avina posed a challenge: “It is written ‫זה שמי לעלם‬ ‫‘ וזה זכרי לדור דור‬This is my name forever and this is my memorial for all generations’ (Exod. 3.15). The Holy One, blessed be he, said: ‘Not as I am written am I read. I am written with yod and heh, but I am read with ʾalef and dalet.’” (b. Pesaḥim 50a) Moreover, qere-type forms (along with ketiv-type forms) are routinely reflected in the ancient translations and non-Masoretic biblical traditions.4 An intriguing case discussed below (ch. 3, §1.1) is that of the Latin Vulgate rendering of ketiv ‫‘ שיניהם‬their urine’ versus qere ‫יהם‬ ֶ֖ ֶ ‫ימי ַרגְּ ֵל‬ ֵ֥ ֵ ‫‘ ֵמ‬water of their feet’ (Isa. 36.12b). Jerome’s rendering is urinam pedum suorum ‘urine of their feet’, which looks to be a conflation of the ketiv and qere traditions. This and other examples show that the interpretive diversity that many ketiv-qere cases reflect significantly preceded the literalisation of said diversity via the medieval masoretic ketiv-qere mechanism. As further evidence, consider the preliminary figures 4 See Gordis (1971, 55–66) for the relationship between ketiv-qere and the ancient versions. See Hornkohl (2022) for a comparison of Tiberian ketiv and qere and the combined Samaritan written and reading tradition. 8 The Historical Depth of the Tiberian Reading Tradition given by Hornkohl (2020a, 412, fn. 5), who reports approximately equal proportions of agreement with ketiv and qere among the 159 cases of MT ketiv-qere paralleled in the BDSS.5 There are also instances of inner-biblical diversity that indicate the adoption of a secondary tradition over an earlier one in the case of parallel texts. CBH Josh. 21.11–39 presents around fifty instances of the word ‫‘ ִמגְּ ָרש‬pastureland’ followed by the 3FS possessive suffix ‫ה‬-. Written ‫מגרשה‬, these show that the word was treated as a singular, presupposing a Tiberian realisation along the lines of ‫‘ ִמגְּ ָר ָשּה‬its pastureland’.6 In the LBH parallel to Josh. 21.11–39 in 1 Chron. 6.40–66, the orthography is consistently different, ‫מגרשיה‬, the added yod indicating that the noun had come to be construed as a plural, ‘its pasturelands’. Intriguingly, the vocalisation of the form in Josh. 21.11–39, i.e., ‫מגְּ ָר ֶש ָה‬, ִ is not that of the singular implied by the orthography in Joshua, but corresponds instead to the plural morphology reflected in the spelling (and vocalisation) in 1 Chron. 6.40–66, ‫יה‬ ָ ‫‘ ִמגְּ ָר ֶש‬its pasturelands’ (Barr 1984). The crucial point in the context of the present discussion is that the plural construal in question and the resulting dissonance between the written and reading compo- 5 More precisely, of the 159 cases of MT ketiv-qere paralleled in the BDSS, 70 show at least partial agreement with the qere, 72 partial agreement with the ketiv, and in 17 cases the form agrees with neither or is ambiguous. See also Kutscher (1974, 519–21). 6 This form may be attested in the phrase ‫( ְּל ַ ֵ֥מ ַען ִמגְּ ָר ָ ֶ֖שּה ָל ַ ַֽבז‬Ezek. 36.5), cf. ESV ‘that they might make its pasturelands a prey’, but the phrase is also analysable as an Aramaic-style infinitive (see below, ch. 12, §2.2, fn. 17). Introduction 9 nents of the tradition in Joshua should be dated no later than the consonantal text of the Chronicles passage (Khan 2020, I:57). Beyond demonstrating special affinity between the Tiberian pronunciation of a CBH text and the orthography and pronunciation of its LBH parallel against the pronunciation tradition ostensibly reflected by the CBH orthography, the foregoing example also draws attention to an important point regarding explicit notation: the written-reading divergence in Joshua is nowhere acknowledged in the Masoretic tradition as an instance of ketiv-qere dissonance. This highlights the necessity of moving beyond cases of ketiv-qere dissonance formally acknowledged in the Masoretic tradition in order more fully to appreciate the historical depth of the Tiberian BH linguistic tradition. To be sure—and this is of critical importance in the present connection—the extent of divergence between the Tiberian written and reading traditions exceeds instances of written-reading divergence explicitly recognised as ketivqere or qere perpetuum. Indeed, most of the studies of writtenreading divergence collected in the present volume have not traditionally been considered cases of ketiv-qere. At this point, it is worth dedicating a few lines to terminology. In several of his studies, Khan (2013b, 464; 2020, I:34) utilises the terms qere and ketiv not just for acknowledged instances of dissonance explicitly recorded as cases of ketiv-qere and qere perpetuum, but also for cases of dissonance unacknowledged in masoretic sources. This is justified, since the extent of diversity within the Tiberian tradition is not exhausted by its recognition in masoretic sources. Notwithstanding the unassailable logic Khan’s broad definitions of ketiv and qere, however, in deference 10 The Historical Depth of the Tiberian Reading Tradition to common usage and to avoid misunderstanding, the terms ketiv and qereare in the present work reserved for traditionally acknowledged cases. For their part, instances of written-reading dissonance not explicitly recognised in masoretic notations and treatises are referred to herein as differences between ‘the written and reading components of the Tiberian biblical tradition’ or, more briefly, as differences between ‘the Tiberian written (or orthographic or spelling) and reading (or pronunciation or recitation) traditions’. This terminology is not entirely satisfying. Beyond its verbosity, it is admitted that the labels suffer from a degree of inconsistency and imprecision. For one thing, the Tiberian written and reading forms are alternately treated as divergent elements of a single composite tradition and as related but separate traditions. The reader should bear in mind both the interrelatedness and the independence of the two elements. Moreover, it is clear that the written tradition (or the written component of the combined tradition) was more than just the product of scribal transmission, but presupposes its own accompanying oral realisation. From this perspective, even within the composite Tiberian written-reading tradition, the reading tradition (or the reading component of the combined tradition) is not the sole pronunciation tradition reflected. The spelling of the consonantal text also presupposes a corresponding pronunciation tradition. Further, the written tradition (or component), often referred to as the ‘consonantal text’, itself likely incorporates multiple layers, probably including material that was at one time written in (more) purely consonantal orthography and only later Introduction 11 augmented with final and internal matres lectionis.7 This obviously means that the orthographic tradition itself likely reflects various strata of oral realisations. While this level of diversity rarely has implications for the phenomena discussed throughout the monograph, where it is significant, e.g., in the case of 1stperson wayyiqṭol forms, in ch. 17, it is discussed in detail. Finally, as already noted, the extant Tiberian pronunciation tradition manifests a degree of diversity. The occasional divergence between vocalisation and accents has already been mentioned (above, fn. 1). Beyond this, diversity in the Tiberian pronunciation tradition sometimes arises from differences in opinion and realisation among representatives of the tradition (Khan 2020, 92–99). For example, see below, ch. 4, on diversity among Tiberian authorities on the graphic representation and phonetic realisation of the proper name Issachar. 2.0. The Tiberian Reading Tradition in Historical Context As is well known, it was not until the Middle Ages that the Tiberian vocalisation was definitively literised in the form of diacritics added to consonantal manuscripts. In contrast to the Tiberian consonantal tradition, which is already reflected in proto-masoretic DSS manuscripts (as one tradition among several repre7 Consider, in this connection, the orthographic disparity between Deut. 2.24–35; 3.14–4.1 as reflected in 4Q31 (4QDeutd) and in the MT. While both show final and internal matres, the Qumran rendition is consistently more defective than the MT rendition, thereby almost certainly reflecting an earlier stage in orthographic development, though there is no obvious evidence of linguistic disparity and only slight textual incongruence. 12 The Historical Depth of the Tiberian Reading Tradition sented in the Dead Sea fragments), the comparatively late written attestation of the vocalic tradition has led some to regard it with suspicion, as a largely secondary product of dubious textual, exegetical, and linguistic credibility. This scepticism arises from two considerations: first, the acknowledged oral nature of the reading tradition; second, the presumed temporal distance between textual composition and crystallisation of the reading tradition, at least as far as CBH material is concerned. All things being equal, it is reasonable to suspect that an unwritten tradition temporally far removed from its written counterpart would be more vulnerable to change than a similar written tradition, a temporally proximate oral tradition, or a temporally proximate written tradition. While such concerns cannot be dismissed, they arguably betray a degree of misunderstanding. First, it is important to bear in mind that there was never a time when the written tradition of the Hebrew Bible was unaccompanied by audible tradition. Barr (1981, 35) states: Reading traditions existed in the temple and synagogue from ancient times. Such reading traditions may well have antedated, rather than followed, the acceptance of a particular manuscript tradition as authoritative. When a more or less authoritative written text came to be accepted, it was found that no manuscript agreed entirely with the reading tradition that was already deemed to be correct. In this way Barr accounts for acknowledged instances of ketivqere dissonance. But it is equally applicable to divergences between the written and reading components of the Tiberian tradition unregistered as instances of ketiv-qere in masoretic sources. Introduction 13 As to the matter of the presumed relative vulnerability of an orally transmitted pronunciation tradition vis-à-vis a written tradition, it is illustrative to present as a corrective the Karaite view noted by Khan (2021, I:123–24): The Karaite Hebrew grammarians of the tenth and eleventh centuries were, in general, concerned with the reading tradition (qere) reflected by the Tiberian vocalization signs and showed little concern for the orthography of the written text (ketiv) (Khan 2000b; 2003; 2013b). The Kara- ite al-Qirqisānī, in his discussions of the bases of authority for the Hebrew Bible, contended that the ultimate authoritative source was the reading tradition of the people of Palestine (by which he meant Tiberias), rather than the written form of the text with orthographic inconsistencies. One of his justifications was that the reading tradition had been transmitted by the whole community (ʾumma) since the time of the prophets whereas the written orthography had been transmitted on the authority of small circles of scribes, which is, therefore, more liable to corruption or wilful change. (Khan 1990c) The textual centrality of the oral tradition among the Karaites is illustrated by, among other things, their practice of recording biblical texts in Arabic letters. Crucially, the letters are not mere transliterations of the Hebrew consonantal tradition, but transcribe the oral realisation of the biblical text (Khan 2021, I:122– 23). Similarly, as already seen, while masoretic scribes were obliged to reproduce the established consonantal text of the Hebrew Bible without changes, the definitive form of the biblical text read in public was that represented by the consonants with the vocalisation and accentuation, and—decisively—the qere when this differed from the ketiv. 14 The Historical Depth of the Tiberian Reading Tradition And what of the time span that is thought to separate the composition of the biblical texts and their final vocalisation? Even if one or more communities eventually managed to preserve an ancient oral tradition, is it reasonable to imagine that such traditions might extend back to the biblical period? In the present volume an effort is made to answer this question. In the meantime, several preliminary considerations may be raised. First, it is important to acknowledge that, as far as the relationship between the consonantal text and the vocalisation is concerned, instances of written-reading dissonance, while not rare, are far from the norm. Throughout the vast majority of the biblical text, the consonantal text and pronunciation tradition seem to be in harmony, with no reason to suspect divergence between the written and reading components of the Tiberian biblical tradition. Second, focusing on the relatively rare cases of writtenreading dissonance, it is true that points of divergence between the Tiberian tradition’s written and reading components often reflect secondary developments in the reading tradition. Significantly, however, these secondary divergences frequently correspond to developments especially characteristic of the language of Second Temple sources. The marked affinity between the Tiberian reading tradition and Second Temple Hebrew is strong evidence that the reading tradition was largely finalised in the Second Temple Period. But there is need for nuance. The reading tradition’s late crystallisation should not be taken to mean that it is uniformly comprised of Second Temple Hebrew. Beyond the fact that com- Introduction 15 monalities linking First and Second Temple Hebrew far outnumber differences that divide them, there is no reason to doubt the routine preservation of genuine Iron Age linguistic features in a tradition that acquired its final shape in the post-exilic period. Finally, it is here emphasised that many cases of dissonance between the Tiberian consonantal and vocalisation traditions, though secondary and relatively late, are not in fact Second Temple innovations. Rather, they frequently constitute minority Iron Age developments whose distinctive Second Temple character relates to late proliferation. Indeed, it was precisely on the basis of such Second Temple proliferation that their use was extended within the biblical reading tradition to pre-Second Temple material. In other words, the anachronistic character of the recitation tradition’s deviations from the pronunciation implied by the consonantal text frequently lies not in the nature of the deviation— many of which are attested in early material—but in the extension of such secondary features, often to the point of their standardisation. It is this standardisation, rather than mere occurrence, that is diagnostic of Second Temple crystallisation. If the arguments in this volume prove compelling, then the Tiberian reading tradition must be deemed a linguistic artefact of considerable historical depth. The analogy of depth can be understood in two ways, i.e., the linguistic tradition both extends deeply into history and comprises multiple layers of material (Hornkohl 2020b, 228–29). Indeed, its most obvious secondary features, in the form of divergences from the written tradition— which, again, it must be emphasised, are comparatively few— reflect dates no later than the Second Temple Period and, in many 16 The Historical Depth of the Tiberian Reading Tradition cases, represent secondary developments already attested in the CBH consonantal tradition and/or Iron Age epigraphy. This, in turn, demands a broad scholarly reassessment of the ramifications of the reading tradition’s antiquity for exegetical, textual, and linguistic research. No longer can the Tiberian vocalisation be summarily dismissed as hopelessly anachronistic, with little to no connection to the earliest linguistic forms of the biblical texts. Rather, it merits serious consideration, even in its most obviously secondary and most conspicuously late features. 3.0. Examples Before turning to the eighteen individual studies that make up the bulk of this monograph, it will be helpful to prime the reader with brief summaries of known cases of dissonance between the written and reading components of the Tiberian biblical tradition, most of which have been discussed elsewhere. In the following cases, the Tiberian reading tradition is characterised by the standardisation of a secondary development known from postexilic sources. Even so, in some cases, the secondary feature has roots in CBH and/or Iron Age epigraphy. 3.1. The Toponym ‘Jerusalem’: ‫רּוש ַלם‬ ָ ְּ‫ י‬versus ‫רּוש ַליִ ם‬ ָ ְּ‫י‬8 The accepted Tiberian pronunciations of the toponym ‘Jeruså̄ lem’—namely, contextual ‫רּוש ַלם‬ ָ ְּ‫ י‬yǝrūšalayim (pausal ‫רּוש ָלם‬ ָ ְּ‫ י‬yǝrū- ̊̄ yim) ̊̄ ̊̄ ̊̄ (pausal šala and contextual directional ‫רּוש ַ ְּל ָמה‬ ָ ְּ‫ י‬yǝrūšalayma ̊̄ yma ̊̄ ̊̄ directional ‫רּוש ָ ְּל ָמה‬ ָ ְּ‫ י‬yǝrūšala )—conflict with the dominant spellings of the name in the written component of the Tiberian 8 Hornkohl (2013a, 91–95). Introduction 17 biblical tradition, namely ‫ ירושלם‬and ‫( ירושלמה‬which spellings occur in all but five of 643 cases). The orthography does not ̊̄ reflect the triphthong in the ending -ayim (pausal -ayim) or the ̊̄ ̊̄ This mismatch diphthong in the ending -ayma ̊̄ (pausal -ayma ). has resulted in the unique situation of two vowels being marked between the last two consonants of the word: ‫רּוש ַלם‬ ָ ְּ‫( י‬pausal ‫רּוש ָלם‬ ָ ְּ‫ )י‬or ‫רּוש ַל ְָּמה‬ ָ ְּ‫( י‬pausal ‫רּוש ַל ְָּמה‬ ָ ְּ‫)י‬. A similar strategy is employed in the Babylonian tradition, though it not infrequently shows just a single vowel between the lamed and mem. Yeivin (1985, 1088–89) attributes such incomplete vocalisations in the most ancient stratum of the tradition and in the composite vocalisation to no more than a lack of rigour on the part of punctuators, whereas he entertains the possibility that the frequency of such vocalisations in the tradition’s intermediate stratum reflects a different phonological realisation. Aside from dominating in the Tiberian and Babylonian written traditions, the spelling ‫ ירושלם‬is also found in the earliest epigraphic attestation of the city’s name, in an inscription from Khirbet Beit Lehi (5.2), which dates to the late sixth century BCE. And such spellings persist in Second Temple documents and literature. The realisation represented by the spelling might have been expected to yield something along the lines of Tiberian ̊̄ ‫רּוש ֵלם‬ ָ ְּ‫* *י‬yǝrūšalēm. Similar realisations with monophthongs in the final syllable are found in BA ‫רּוש ֶלם‬ ְ ְ‫י‬, TA ‫ ָלם‬-/‫רּוש ַלם‬ ְ ְ‫י‬, Syriac ܶ ܺ ܿ ܶ ܺܽ ‫ܐܘܪܫܠܡ‬/‫ܐܘܪܫܠܡ‬, Greek Ιερουσαλημ, and Latin Hierusalem (HALOT 437a). Consider also the form of the toponym ‫‘ ָש ֵלם‬Salem’ (Gen. 14.18; Ps. 76.3). 18 The Historical Depth of the Tiberian Reading Tradition However, against the view that the Tiberian reading tradi̊̄ tion’s pronunciation yǝrūšalayim is a medieval innovation, spellings presupposing the diphthongal ending, in the form of ‫ירושלים‬ and ‫ירושלימה‬, appear five times in the Tiberian written tradition (Jer. 26.18; Est. 2.6; 1 Chron. 3.5; 2 Chron. 25.1; 32.9) and are common in non-Tiberian biblical and post-biblical sources, e.g., DSS biblical and non-biblical material, coins from the Second Temple Period, and rabbinic literature. The overall distribution of the spelling ‫ ירושלים‬in ancient Hebrew sources, including the combined Tiberian written-reading tradition, indicates that a realisation along the lines of ̊̄ yǝrūšalayim represented a Second Temple convention that was standardised in the Tiberian reading tradition despite the dominant orthography. This is consistent with the view that the Tiberian reading tradition took its essential shape in the Second Temple Period. Evidence is insufficient to substantiate whether or not the sort of pronunciation preserved in the Tiberian reading tradition predates the Second Temple Period. 3.2. Univerbalisation of the Infinitive Construct with Prefixed -‫ל‬9 In the Tiberian tradition, the phonetic realisation of the qal IIbgdkpt construct infinitive varies depending on whether or not the form is preceded by a prefixed preposition and on the identity of the preposition. Blau (2010, 213–14) explains as follows: The construct infinitive is frequently governed by prepositions, especially by ‫ל‬. Originally this ‫ ְּל‬had a fully preposi9 Hornkohl (2020a, 230–57). Introduction 19 tional meaning, as, e.g., ‘in order to’ (e.g., ‫את‬ ֹ ֵ֥ ‫הוה ִל ְּר‬ ֵָ֔ ְּ‫וַ יֵ ֵֶּ֣רד י‬ ‫ת־ה ִ ֶ֖עיר‬ ָ ‫‘ ֶא‬and the Lord came down to see the town’ Gen 11:5); later the ‫ ל‬became a part of the infinitive…. This is reflected both by the form and by the syntactic usage of the preposition. Formally, the ‫ ל‬became integrated into the infinitive. In some forms of the qal infinitive, the ‫ ל‬appears to be in close internal juncture: the šwa that begins the in- finitive behaves as a genuine quiescent šwa, and subse- quent ‫ב‬, ‫ג‬, ‫ד‬, ‫כ‬, ‫פ‬, ‫ ת‬letters are vocalized as stops, e.g., ‫ִלנְּ פֹל‬ ‘to fall’, as opposed to simple ‫ נְּ פֹל‬and ‫בנְּ פֹל‬/‫ֹל‬ ִ ‫‘ ִכנְּ פ‬when fall- ing’. In Rabbinic Hebrew the univerbalization of the infinitive with ‫ ל‬is even more progressed: the ‫ ל‬is always attached to the infinitive, even after other prepositions, and the infinitive is totally remodelled after the prefixtense…. The special vocalization of the construct infinitive in Biblical Hebrew after ‫ל‬, corresponding to the vocalization of the prefix-tense… is undoubtedly in the line of Rabbinic Hebrew (and may even reflect the impact of Rabbinic Hebrew on the Masoretes) (see also Blau 2010, 115). However, several lines of argumentation converge to show that the apparent distinction between the Tiberian written and reading traditions is not as neat and tidy as a mere dichotomy of BH versus RH. Rather, pre-rabbinic evidence, including some from the Tiberian written tradition itself, shows that the process of univerbalisation that is attested in the reading tradition and that culminated in RH, was also earlier very much underway. Significant pieces of evidence include: 1. apparent DSS transitional forms, e.g., ‫* *לגוע‬liggoaʿ ‘to touch’ (4Q53 f2–5i.5; cf. BH ‫לנְּ ג ַֹע‬/‫ת‬ ִ ‫ ָלגַ ַע‬and RH ‫) ִליגַ ע‬, which ‫נ‬ was secondarily corrected to ‫* *ל גוע‬lingoaʿ, and ‫* לשול‬liššol ‘to clear away’ (1QM 10.1–2; cf. BH ‫*לנְּ ש ֹל‬ ִ and RH ‫ישל‬ ַ ‫—)*ל‬ ִ 20 The Historical Depth of the Tiberian Reading Tradition the apparent assimilation of n in these forms was possible only after the vowel following n had shortened to zero; 2. the distinction in preposition vocalisation, -‫ל‬, on the one hand, versus -‫ ב‬and -‫כ‬, on the other, in qal I-y and II-w/y verbs, e.g., ‫‘ ְּב ֶל ֶדת‬when bearing’ versus ‫‘ ָל ֶל ֶדת‬to bear’ and ‫‘ ְּבבֹוא‬in coming’ and ‫‘ ְּכבֹוא‬after coming’ versus ‫‘ ָלבֹוא‬to come (in the Tiberian as well as Babylonian traditions, and with parallels in the Samaritan tradition); 3. the overall rarity of infinitives construct without a preceding preposition in all biblical consonantal traditions and the dominance of infinitives with -‫ ל‬in late material, e.g., Tiberian LBH, BA, DSS Hebrew, the Hebrew of BS, and RH; 4. the predominantly late character of structures involving an infinitive with -‫ ל‬preceded by another preposition; 5. the substitution in late material of infinitives with preceding -‫ ל‬for CBH infinitives without preceding -‫ל‬. It has been argued that the Tiberian phonological realisation of qal II-bgdkpt construct infinitives is a rabbinic or later anachronism alien to older BH phonology. Against this claim, phonological, morphological, and syntactic evidence may be adduced to demonstrate that the univerbalisation of the infinitive construct with -‫ ל‬was underway in the linguistic stratum reflected in classical biblical consonantal material. The corresponding CBH reading tradition may indeed reflect a later stratum, perhaps vaguely contemporaneous with the combined Tiberian LBH written-reading tradition, but the difference more of degree than essence, since both strata lie at points on the same developmental line, which culminated in RH. Introduction 21 3.3. ‫ יֹום ַה ִש ִשי‬versus ‫‘ ַהיֹום ַה ִש ִשי‬The Sixth Day’ and Similar10 BH norms of noun-attribute concord typically involve agreement in gender, number, and definiteness. However, exceptions, especially in terms of agreement in definiteness, have long been known. Further complicating matters is the apparent dissonance between the written (consonantal) and reading (vocalic) components of the Tiberian biblical tradition, especially in poetry (Ley 1891; Lambert 1898; GKC §126h; Barr 1989, 310–12, 325–33). In poetic compositions in the Hebrew Bible, when the sequence [noun+article+adjective] is preceded by a clitic preposition, e.g., -‫ב‬, -‫כ‬, or -‫ל‬, the double-article DETERMINED NOUN+ DETERMINED ADJECTIVE formulation dominates; but when the noun has no attached preposition, the construction occasionally has a single-article ANARTHROUS NOUN+DETERMINED ADJECTIVE formulation. Conspicuous in this connection—even outside of poetry— are expressions comprising the noun ‫‘ יֹום‬day’ and an attributive ordinal numeral. In the Tiberian biblical tradition, when this combination is preceded by a clitic preposition, it consistently comes in the symmetrical, double-article formulation DETERMINED NOUN+DETERMINED ORDINAL (of the 126 occurrences, 125 involve -‫ב‬, one -‫)ל‬. Conversely, on eight occasions when there is no preceding clitic preposition, an alternative, asymmetric, single-article ANARTHROUS NOUN+DETERMINED ORDINAL syntagm obtains. The incongruity is especially conspicuous in the local discord 10 Hornkohl (2020b). 22 The Historical Depth of the Tiberian Reading Tradition among the three relevant cases in (1), which occur in successive verses. (1) ‫יעי‬ ִִ֔ ‫אכ ֶ֖תֹו ֲא ֶשר ָע ָ ָׂ֑שה וַ יִ ְּשב ֹ֙ת בַּ יֹּ֣ום הַּ ְּׁש ִב‬ ְּ ‫יעי ְּמ ַל‬ ִִ֔ ‫ים בַּ יֹּ֣ום הַּ ְּׁש ִב‬ ֙ ‫ֹלה‬ ִ ‫וַ יְּ ַכַ֤ל ֱא‬ ...‫יעי וַ ַיְּק ֵ ֶ֖דש א ָֹׂ֑תֹו‬ ִִ֔ ‫ים אֶ ת־יֹּ֣ום הַּ ְּׁש ִב‬ ֙ ‫ֹלה‬ ִ ‫אכ ֶ֖תֹו ֲא ֶ ֵ֥שר ָע ָ ַֽשה׃ וַ יְּ ָ ַ֤ב ֶרְך ֱא‬ ְּ ‫ל־מ ַל‬ ְּ ‫ִמ ָכ‬ ‘And on the seventh day God finished his work that he had done, and he rested on the seventh day from all his work that he had done. So God blessed the seventh day and made it holy…’ (Gen. 2.2–3a) Consider also the diversity between the three cases in (2): (2) ‫אר ִמ ָב ֵת ֶיכָׂ֑ם ִכי‬ ֹ ֶ֖ ‫אשֹון ַת ְּש ִ ֵ֥ביתּו ְּש‬ ִ֔ ‫ֹאכלּו ַ ַ֚אְך בַּ יֹּ֣ום הָ ִר‬ ֵֵ֔ ‫ים ַמצֹות ת‬ ֙ ‫וש ְּב ַ ַ֤עת יָ ִמ‬ ִ ‫אשן עַּ ד־יֹּ֥ום‬ ֹׁ֖ ‫וא ִמיִ ְּש ָר ֵֵ֔אל ִמיֹּ֥ום הָ ִר‬ ֙ ‫׀ ָכל־א ֵֹכל ָח ֵֵ֗מץ וְּ נִ ְּכ ְּר ָָ֞תה ַה ֶנ ֶַ֤פש ַה ִה‬ ‫הַּ ְּׁש ִב ִ ִֽעי׃‬ ‘Seven days you shall eat unleavened bread. On the first day you shall remove leaven out of your houses, for if anyone eats what is leavened, from the first day until the seventh day, that person shall be cut off from Israel.’ (Exod. 12.15) Broadly speaking, there are two approaches to explaining the clash between single- and double-article ‫יֹום‬+ordinal constructions in the Tiberian biblical tradition. According to the first approach, they are to be viewed as abbreviations of common phrasal constructions in which the initial article has been deleted, perhaps under vernacular pressure. This is in line with S. R. Driver’s ([1892] 1998, §209) observation on such RH cases as ‫‘ כנסת הגדולה‬the great synagogue’ (m. ʿEruvin 10.10) and ‫יצר הרע‬ ‘evil inclination’ (m. ʾAvot 2.11) that “the usage appears to have arisen in connexion with familiar words, which were felt to be sufficiently definite in themselves without the addition of the ar- Introduction 23 ticle.” Parade Masoretic BH examples of single-article constructions include ‫ימי‬ ִ֖ ִ ִ‫‘ ָח ֵ ֵ֥צר ַה ְפּנ‬inner court’ (Ezek. 40.28), ‫יעית‬ ִ ֵ֔ ‫ִב ְּשנַ ֙ת ָ ַֽה ְּר ִב‬ ‫‘ ִליהֹויָ ִ ֵָ֥קים‬in the fourth year of Jehoiakim’ (Jer. 46.2), and ‫ְּב ֶ ֵ֥ד ֶרְך‬ ‫טֹּובה וְּ ַהיְּ ָש ָ ַֽרה‬ ֶ֖ ָ ‫‘ ַה‬in the good and right way’ (1 Sam. 12.23). While some such ‘pseudo-construct’ expressions are likely genuine vestiges that reflect a linguistic stage before the standardisation of determination agreement (Borg 2000), others (like the three preceding examples) are probably secondary results of construal as fixed compounds, whether the resulting nouns were deemed common (lexicalisation) or proper (onymisation) (Moshavi and Rothstein 2018, 116, fn. 54). Single-article ‫יֹום‬+ordinal constructions are arguably to be explained differently (GKC §126w, fn. 9). Several pieces of evidence may be cited in support of the view that, in this case, an archaic single-article construction was secondarily supplanted by a double-article alternative. First, within Tiberian BH, the complementary distribution of single- and double-article ‫יֹום‬+ordinal constructions is suspiciously suppletive. The double-article alternative obtains only where a cliticised preposition permits its articulation before ‫יֹום‬, or, in the absence of such a preposition— crucially—in acknowledged late contexts: LBH Dan. 10.12 and Neh. 8.18 and NBDSS 4Q216 7.12 = Jub. 2.21 and 4Q284 f2ii 3–4; f3.2. Further evidence of the Second Temple character of the symmetrical DETERMINED NOUN+DETERMINED ORDINAL construction comes from Aramaic and Syriac. Not only do the Targums and the Peshiṭta, respectively, rather consistently present doublearticle constructions composed of DETERMINED NOUN+DETER- 24 The Historical Depth of the Tiberian Reading Tradition MINED ORDINAL—including, notably, in most of their renderings of the eight cases of Masoretic CBH single-article formulation— but this agreement is routine in those languages outside of biblical translations, too. It is possible that convergence with Aramaic contributed to the process of movement from single- to doublearticle ‫יֹום‬+ordinal structures, though the process may well have begun within Hebrew in connection to the standard norm of adjectival agreement. If double-article ‫יֹום‬+ordinal structures are indeed secondary in ancient Hebrew, then this explains the suppletion in Tiberian CBH. The single-article construction was preserved only where the consonantal text was not amenable to double-article vocalisation. On the basis of the consistency of single-article ‫יֹום‬+ordinal when ‫ יֹום‬is preceded by -‫ ב‬or -‫ל‬, it stands to reason that BH at one time knew structures of the type ‫*ביֹום‬ ְּ +‫ה ִש ִשי‬, ַ in accord with the type ‫יֹום ַה ִש ִשי‬. If so, at least some portion of the extant cases of the type ‫ ַביֹום ַה ִש ִשי‬must be due to secondary reinterpretation, which has led to the current dissonance between the vocalisation implied by the consonantal tradition and the Tiberian vocalisation. As already noted, the recognition of dissonance is not new (Lambert 1895; GKC §126h; Sperber 1966, 603; Barr 1989, 310– 12, 325–33; Borg 2000, 31, 33; JM §138b). It is commonly hypothesised that the consistent double-article syntax of expressions of the type ‫ ַביֹום ַה ִש ִשי‬is due to secondary recasting in line with both standard BH noun-adjective concord and post-exilic consonantal evidence of the double-article structure ‫יֹום‬+ordinal. Borg (2000, 33) goes so far as to speculate that all biblical and Introduction 25 DSS ‫יֹום‬+ordinal expressions with cliticised prepositions were originally single-article constructions. This seems extreme, given the occurrence of consonantally unambiguous double-article constructions in LBH and the DSS. A plausible hypothesis in light of the evidence is that Second Temple Hebrew was characterised by genuine cases of the type ‫ ַביום הששי‬as well as persistence of the type ‫* ְּביום הששי‬. Barr’s (1989, 330) comments on early poetry have broader application: [A]lthough we cannot assume that every ‘article’ marked upon a preposition b, k, or l in early poetry was ‘really’ there, it is unwise scepticism to suppose that none of them were really there or that only those marked with the consonantal h can be taken as actual.… Though the reading tradition was not always ‘right’, this is not an adequate reason for supposing that in this respect it was always wrong…. The use of the article was in a process of change during— perhaps one should even say ‘throughout’—the biblical pe- riod; and I have said nothing of the post-biblical usage, which certainly deserves to be taken into consideration here as well. This could mean that some of the reconstitution of patterns in the later reading tradition was in continuity with processes that were taking place during biblical times; it could even mean that some of this reconstitution was already under way within the formation of the Bible. The Second Temple consonantal evidence adduced above for ‫ היום הששי‬gives a latest possible date for the development of the syntax reflected in masoretic vocalisations of the type ‫ַביֹום‬ ‫ה ִש ִשי‬. ַ Significantly, however, establishing an earliest possible 26 The Historical Depth of the Tiberian Reading Tradition date is precluded by a frustrating lack of evidence. One might speculate that, with a larger sample size of CBH cases without clitic prepositions, sporadic CBH cases of the type ‫ היום הששי‬might conceivably have occurred. Irrespective of this eventuality, a scenario can be imagined in which doubly-determined ‫ַביֹום ַה ִש ִשי‬ structures developed without double-article ‫ היום הששי‬ever having enjoyed widespread currency. Indeed, this is the most straightforward reading of the evidence, since double-article ‫היום‬ ‫ הששי‬is very rarely attested in any phase of ancient Hebrew. Indeed, it is not beyond the realm of possibility that doubly-determined expressions with clitic prepositions, like ‫ביֹום ַה ִש ִשי‬, ַ preceded and influenced the development of doubly-determined cases without clitic prepositions, like ‫היום הששי‬. If suppletive syntax could take hold in the Tiberian reading tradition, why not earlier? One cannot discount the possibility that the double-article structure ‫ ַביום הששי‬developed in Iron Age Hebrew, coexisting with single-article ‫יום הששי‬, and that the Tiberian reading tradition merely standardised the double marking where possible. In sum, while single-article constructions without prepositions of the type ‫ יֹום ַה ִש ִשי‬likely predate double-article ‫ַהיֹום ַה ִש ִשי‬ alternatives, the Tiberian vocalisation of double-article expressions with prepositions, as in ‫ ַביֹום ַה ִש ִשי‬, are likely secondary in some CBH contexts, but are in line with unequivocal LBH and DSS Hebrew consonantal evidence. A dearth of evidence precludes determining when the double-article formulation was coined. It was certainly established by Second Temple times; it may well have arisen earlier. Introduction 27 3.4. The 3MPL Gentilic: ‫ִים‬ִ versus ‫ִיִ ים‬ִ 11 The typical Tiberian BH MPL gentilic ending is generally the same as that characteristic of MPL substantives, i.e., ‫ ִִים‬- -īm. It seems clear in the case of 3MPL gentilics that this is due to secondary syncope of an earlier phonetic realisation with consonantal y, e.g., -iy(y)im/-i:im/-iʾim/-īm < -iyyim.12 In view of the consistently defective spelling of plural -im in Iron Age Hebrew inscriptional sources (Gogel 1998, 61–73), the yod in such forms as the Arad letters’ ‫‘ כתים‬Kittites’ is almost certainly consonantal, i.e., kittiy(y)im. A similar picture emerges from cognate inscriptions, with spellings like Phoenician ‫ דננים‬danuniy(y)im and Ugaritic /ʾugrtym/ ʾugaritiy(y)im ‘Ugarites’. Turning to Second Temple sources, the DSS present orthographic evidence consistent with both the continued consonantal realisation of y (or some reflex thereof) and contraction to simple -im. Forms spelled with double yod outnumber those with a single yod by counts of 23:18 in the BDSS and 11:3 in the NBDSS (for details, see Hornkohl 2018a, 89, fn. 51). While the phonetic values of the relevant spellings cannot be determined with certainty, it is reasonable to assume that they reflect a variety of pronunciations, presumably a continuum from geminated or singleton consonantal realisation, through hiatus, glottal epenthesis, and/or extended i-vowel, to complete contraction to -im (Reymond 2014, 120–22; cf. Qimron 1986, 24; 2018, 95–97). Codex 11 Hornkohl (2018, 86–91). 12 The gemination of y in such cases may itself be secondary, though early (Suchard 2019, 59 and fn. 8). 28 The Historical Depth of the Tiberian Reading Tradition Kaufmann of the Mishna, material from BS, and the Samaritan reading tradition, in all of which contracted MPL gentilic dominates, furnish confirmatory evidence of the late proliferation of syncope. Coming to the relevant form in the Tiberian reading tradition, we find that it is with very few exceptions syncopated to -īm, corresponding to the standard MPL suffix on non-gentilic substantives, -īm. Given the evident incidence of syncopated realisations of MPL gentilic ‫ים‬- in the DSS, BS, the Samaritan biblical reading tradition, and RH, it is clear that the Tiberian reading tradition presents a phonetic realisation in line with late Second Temple practices. But might such a syncopated realisation date to even earlier? There is evidence, albeit ambiguous and/or limited, suggesting that it might. The Tiberian consonantal tradition presents a single potential case of contracted 3MPL gentilic ending. Consider example (3): (3) ‫ּול ֻא ִ ַֽמים׃‬ ְּ ‫טּושים‬ ֶ֖ ִ ‫ּול‬ ְּ ‫שּורם‬ ּ֥ ִ ַּ‫ּוב ֵני ְּד ָ ֵ֔דן ָהי֛ ּו א‬ ְּ ‫ת־ד ָ ָׂ֑דן‬ ְּ ‫ת־ש ָ ֶ֖בא וְּ ֶא‬ ְּ ‫וְּ יָ ְּק ָשן יָ ֵ֔ ַלד ֶא‬ ‘And Jokshan fathered Sheba and Dedan. And the sons of Dedan were Asshurim and Letushim and Leummim. (Gen. 25.3) While identification of the form ‫שּורם‬ ִ ‫ ַא‬as a gentilic with syncopated -īm ending arguably suits the genealogical context, it may be otherwise explained (Kiel 2000, 204). More promising, but still questionable evidence for syncope comes from Iron Age Hebrew epigraphy. In contrast to the routine consonantal y in the Arad Letters’ ‫ כתים‬kittiy(y)im ‘Kittites’ comes potential evidence of contraction -iy(y)im > -im in the Introduction 29 form ‫אדמם‬, presumably ʾadomim ‘Edomites’ (Arad 3.12). Though the context is broken, mention of Edom elsewhere in the corpus, most explicitly in Arad 24.20 (see also 21.5; 40.10, 15) lends support to this interpretation. Intriguingly, the main argument raised in objection to the reading of a MPL gentilic here is the otherwise unattested contracted realisation of the MPL gentilic ending in the ancient Hebrew epigraphic corpus (see Gogel 1998, 182, fn. 217, and the works cited there). The most secure supporting evidence for the early contraction of the MPL gentilic ending is found in the relatively frequent Phoenician reference to ‫‘ צדנם‬Sidonians’, which goes as far back as the 8th century BCE.13 In its consistent presentation of a syncopated MPL gentilic ending, the Tiberian reading tradition reflects standardisation of a secondary development. Though secondary, the development in question is not only well represented in Second Temple consonantal sources, but apparently sporadically evidenced in even earlier written material. The contraction -im < -iy(y)im is presumably an early vernacular phenomenon, only sporadically preserved in early sources, that came to dominate in certain Second Temple traditions, including the Tiberian reading tradition. 13 KAI 31.1 (8th cent BCE); 13.1–2 (5th cent BCE); 14.1–2, 13–15, 18, 20 (5th cent BCE); Gibson 1971–1982, no. 29 (3x) (400 BCE). 30 The Historical Depth of the Tiberian Reading Tradition 3.5. The 3MS Possessive Suffix on Singulars and Similar: ‫ה‬- versus ‫ֹו‬-14 In all traditions of BH, the dominant 3MS possessive (nominal) suffix for singular nouns and similar is ‫ֹו‬-. In the Tiberian tradition, the written and reading components agree on this morphology in 7710 of 7765 cases (Andersen and Forbes 1986, 183, 323). In the 55 exceptions, the written tradition presents ‫ה‬-. Sometimes this is the ketiv and the accompanying qere calls for ‫ֹו‬-. On other occasions, the standard vocalisation is simply imposed upon the anomalous orthography in the form of ‫ִ ֹה‬-. Either way, these appear to be instances of phonological dissonance between the written and reading components of the Tiberian biblical tradition. The spelling ‫ה‬- dominates for the 3MS possessive suffix in ancient Hebrew epigraphy (Gogel 1993, 155–56). It is generally thought to have developed to reflect realisations of the type -ahū, -ihū, or -uhū. Yet, given the propensity for marking final long vowels in ancient Hebrew inscriptions, it is not impossible that -ahū had already shifted to -ō (via elision of heh and monophthongisation of -aw) (Zevit 1980, 17, no. 23). Another possibility is that ‫ה‬- in the inscriptions and the Bible was meant to reflect something along the lines of -ēh, which is the standard Aramaic parallel (Young 1993, 105–6, 126). Assuming BH 3MS ‫ה‬- reflected some realisation other than standard -ō, there is strong evidence that the dissonance on this point between the Tiberian tradition’s written and reading components is early. In other words, though ‫ה‬- is clearly archaic and 14 See Hornkohl (2012, 67–69). Introduction 31 was probably not originally meant to represent -ō, there are strong indications that 3MS -ō is itself quite ancient. Not only is it the dominant form throughout the combined Tiberian writtenreading tradition;15 it is also attested as a minority form in Iron Age Hebrew epigraphy (‫‘ ושלחו‬and send [MS] it!’ Arad 13.4 [ver- bal]; ‫‘ בו‬in him’ Ketef Ḥinnom 1.11). Moreover, Tiberian 3ms ‫ה‬- is sometimes paralleled in the BDSS by ‫ו‬- (e.g., ), while in SH, it is consistently paralleled by ‫ו‬- -u. Ancient transcriptional evidence also reflects -o—the Secunda has -ω (Brønno 1943, 362) and Jerome has -o.16 While the difference between the majority Iron Age epigraphic orthography ‫ה‬- and the majority biblical spelling ‫ו‬- must 15 The orthography ‫ה‬- pointed with ḥolam is common in the Tiberian biblical tradition in other categories as well, especially proper nouns, like ‫‘ ְשֹלמֹה‬Solomon’, ‫‘ ַפּ ְרעֹה‬Pharaoh’, ‫‘ ִשֹלה‬Shiloh’, ‫‘ ׂשֹוכֹה‬Socoh’, and ‫גִ ֹלה‬ ‘Gilo’, and the III-y qal infinitive absolute forms. In contrast to the spelling of 3MS ‫ה‬-, which largely gave way to ‫ו‬-, the spelling of such proper names and toponyms with ‫ִ ֹה‬- persists throughout all chronolects of Hebrew. 16 I am grateful to my friend and colleague, Benjamin Kantor (f.c.), for supplying the following data from his forthcoming book: brucho || BHS ‫‘ ְּברּוחֹו‬in his spirit’ (Ps. 32.2); dercho || BHS ‫‘ ַד ְּר ָׂ֑כֹו‬his way’ (Prov. 8.22); baaphpho || BHS ‫אַפֹו‬ ָׂ֑ ‫‘ ְּב‬in his nose/nostrils’ (Isa. 2.22); mnuatho || BHS ‫‘ ְּמנֻ ָח ֶ֖תֹו‬his residence/resting place’ (Isa. 11.10); cadeso || BHS ‫ָק ְּד ָׂ֑שֹו‬ ‘his holiness’ (Isa. 63.10); chullo || BHS ‫‘ ֻכלָּׂ֑ ה‬all of it [MS]’ (Ezek. 11.15); aphpho || BHS ‫אַפֹו‬ ֵ֔ ‘his anger’ (Amos. 1.11); masio || BHS ‫ה־ש ֵ֔חֹו‬ ֵ ‫‘ ַמ‬what his meditation [is]’ (Amos. 4.13); messio || comments on ‫‘ ְּמ ִשיחֹו‬his Mes- siah’ (Amos. 4.13); baemunatho || BHS ‫‘ ֶב ֱאמּונָ ֵ֥תֹו‬by his faith’ (Hab. 2.4); iado || BHS ‫‘ ִמיָ ֶ֖דֹו‬from his hand’ (Hab. 3.4). Note that the Tiberian form in Ezek. 11.15 ends in heh: ‫כלָּׂ֑ ה‬. ֻ 32 The Historical Depth of the Tiberian Reading Tradition be explained (by a Second Temple orthographic revision?) and while there is no certainty that First and Second Temple spellings with ‫ו‬- were necessarily read with an o-vowel, the combination of the unanimous testimony of the ancient transcriptions and the Masoretic Tiberian and Babylonian reading traditions makes an o-vowel the most likely candidate (against Samaritan -u < -hu). In this case, then, the antiquity of the Tiberian reading tradition’s -ō where the written tradition has ‫ה‬- seems to be vouchsafed by robust Second Temple evidence. Assuming that the minority epigraphic and dominant Masoretic spellings ‫ו‬- also represent -ō, the phonology in question can be traced all the way back to First Temple times. Alternatively, the realisation was -aw, for which -ō is a later reflex. 3.6. The 3MS Possessive Suffix on Plurals and Similar: ‫ו‬- versus ‫ִיו‬ָ 17 In the Tiberian biblical tradition, the standard 3MS possessive suf̊̄ [ɔːv]. Such a fix on plural nouns is written ‫יו‬-, but realised as -aw written-reading corelation is counterintuitive, but sufficiently established that a number of words without the 3ms suffix that end ̊̄ [ɔːv], have also acquired spellings with ‫יו‬-, e.g., MT ketiv in -aw and qere ‫‘ יַ ְּח ָדיו‬together’, MT qere ‫‘ ְּס ָתיו‬winter/autumn, rainy sea- son’, MT qere ‫‘ ָענָ יו‬humble’, DSS ‫‘ עישיו‬Esau’, DSS ‫‘ תיו‬hook’, RH ‫‘ עכשיו‬now’. Two general explanations have been offered for the unex̊̄ [ɔːv]. One is pected presence of a yod in a suffix pronounced -aw that it was added secondarily as a grammatical mater lectionis to 17 See Hornkohl (2020, 257–73). Introduction 33 indicate plurality. The other is that it is not secondary, but reflects an oral realisation different from the one preserved in the Tiberian pronunciation tradition. Specifically, it is thought that it represented triphthongal -ayu or -eyu in contrast to the diph̊̄ [ɔːv]. Given the not-infrethongal Tiberian pronunciation -aw quent occurrence in the Tiberian written tradition of ‫ו‬- without yod in cases where the combination of a plural with 3MS suffix is expected, along with the dominant use of ‫ו‬- alone in such cases in Iron Age Hebrew epigraphic sources, the view that attributes the dissonance between the written and reading components to diversity in pronunciations of the 3MS suffix is arguably the more compelling of the two. Crucially, however, no matter which explanation is adopted, both presuppose the relative antiquity of the form preserved in the reading tradition vis-à-vis the standard orthography. For whether the orthography ‫יו‬- is due to secondary addition of a grammatical mater or reflects genuine phonology with consonantal yod, the extant historical evidence points to the antiquity of the spelling ‫ו‬- and of a realisation consistent therewith, whether -aw (> -o?) or -ew, with inscriptional evidence from Gezer (ninth-tenth century BCE), Yavne Yam (=Meṣad Ḥashavyahu; late seventh century BCE), and Lachish (early sixth century BCE). If so, this constitutes a rather rare situation in which the reading component of the Tiberian biblical tradition may preserve a feature older than that reflected in the corresponding written component. But there is more to the story. The spelling ‫יו‬- is also known form ancient Hebrew epigraphy, specifically from the mid-sev- 34 The Historical Depth of the Tiberian Reading Tradition enth-century BCE Ketef Ḥinnom silver inscriptions. If so, then the spelling ‫יו‬-, apparently representative of a triphthongal realisation, might constitute an ancient minority feature, which was standardised in the Tiberian written tradition. By contrast, an apparently majority ancient spelling-pronunciation tradition underlies the dominant Tiberian pronunciation, which is also preserved in a minority of spellings in the MT. Later, the co-occurrence of the spelling ‫יו‬- and the realisation -aw/-av led to the extension of the use of written ‫יו‬- to other instance of realisations of -aw/-av, even where there was no 3MS suffix. If the above discussion is correct, the dominant 3MS traditions of both the written and reading components of the Tiberian biblical tradition are authentically old, but the normal situation, according to which the reading tradition reflects the standardisation of an ancient minority feature in line with Second Temple conventions, has been reversed. For in this case, it is the written form ‫יו‬- that is the minority form in unambiguously dated early material, becoming common only in Second Temple sources. Against this, apparently diphthongal ‫ו‬- is the majority Iron Age form and is preserved in the Tiberian reading tradition. 3.7. Attenuation of a to i Narrowly interpreted, the Tiberian Hebrew a > i vowel shift traditionally termed ‘attenuation’ is a case of dissimilation operative when there are two consecutive closed syllables with /a/ vowels, the second of which is stressed: C1aC2C3áC4 > C1iC2C3áC4. Wellknown examples include ‫‘ ִמגְּ ָדל‬tower’ (< magdal), ‫‘ ִמ ְּריָ ם‬Miriam’ (< maryam), and ‫‘ ִש ְּב ָעה‬seven (M)’ (< šabʿat). The process is said Introduction 35 to be blocked if C2 = C3 (i.e., if the syllable is closed by gemina- tion), e.g., ‫‘ ַמ ָתנָ ה‬gift’, ‫‘ ַמ ַסע‬journey’; if C1 = C3 or C2 = C4 (i.e., in the case of reduplication), e.g., ‫‘ גַ ְּלגַ ל‬wheel’ (but cf. ‫‘ גִ ְּלגָ ל‬Gilgal’); and by the presence of a guttural or, sometimes, /r/ or /l/, e.g., ‫‘ ַמ ְּעגָ ל‬circle’, ‫‘ * ַמ ְּר ַבד‬carpet, tapestry’, ‫‘ * ַמ ְּל ָמד‬prod, ox goad’. Once these cases are accounted for, there are very few exceptions (Koller 2013; see also Sivan and Qimron 1995, 20–26). Broader interpretations of attenuation that lump together various other sorts of shifts a > i under the same heading are today largely rejected (Blake 1950; Lambdin 1985; Koller 2013). Because attenuation seems to be largely absent from the Greek and Latin transcriptions, as well as from SH, and because it is far less extensive in the Babylonian biblical pronunciation tradition than in Tiberian Hebrew, its extensiveness in the Tiberian biblical tradition is widely regarded as a very late development (Blau 2010, 132, §3.5.7.6.13; Koller 2013; Hendel 2016, 32). Indeed, since Jerome still has Magdal in his Latin translation of the Bible (c. 400 ce), Rendsburg (2013, 108) dates the shift to sometime between 400 and 850 CE. The frequent exceptions to attenuation are also taken by some as evidence that the shift was late and never completed (Blau 2010, 132, §3.5.7.6.13). There seems little doubt that from the perspective of the extent of attenuation a to i, the Tiberian biblical pronunciation tradition reflects greater innovation than what is seen in the pronunciation evidence of the LXX, Origen’s Hexapla, Jerome, and the Samaritan and Babylonian reading traditions (see Khan 2020, I:66–67). But does this necessarily entail the view that the sound shift began post-400 CE, i.e., that it was unknown in earlier He- 36 The Historical Depth of the Tiberian Reading Tradition brew? In light of the historical precedence seen in other linguistic features that became standard in the Tiberian reading tradition, it seems worth entertaining the possibility that in the case of attenuation, too, a relatively early feature of limited extension was eventually regularised in Tiberian pronunciation. Indeed, there are sporadic signs of a > i attenuation in preTiberian Hebrew sources. In his discussion of the Second Column of Origen’s Hexapla (i.e., the Secunda, c. 250 CE18), Brønno (1943, 284–85) lists the forms μισγαβ || MT ‫‘ ִמ ְּש ָ ַֽגב־‬stronghold’ (Ps. 46.8, 12) and μισχνωθαμ || MT ‫‘ ִמ ְּש ְּכנ ָֹתם‬their dwellings’ (Ps. 49.12). Consider also the burial epitaph ‫[‘ [שלום ע]ל מישכבך‬peace] upon your resting’ (CIJ 1414), dated by Tal (2008, 162, no. 23) to the third century CE. In all of the above cases, however, it is possible that the preceding sibilant triggered the shift a > i. Conversely, no such conditioning factor applies in the case of the Greek Φυλῆς Μιγδαληνων ‘tribe of the Migdalenes’ from the Hellenistic–Roman Periods of what is modern day Syria (Waddington 1870, no. 2483; Burke 2007, 34, 52).19 Whatever the language of the people group in question—presumably, a Hebrew or Aramaic dialect—Trombley (2014, 359–61) dates the arrival of the Migdalenoi to no later than the third century CE, to which period he also dates the relevant inscription. 18 Kantor (2017, 9–17) argues for a late Roman date, i.e., 150–225 CE (“mid-to-late second or early third century CE”) for the compilation of the pre-Secunda, on which source Origen is thought to have based the Second Column of the Hexapla. 19 I owe this citation to Jan Joosten. Introduction 37 Consider also the spelling ‫‘ מירים‬Miriam’ in a burial inscription from Beth Shearim that Mazar (1973, 54, 197–98) dates to the third-century CE (Tal 2012, 187, no. 5, fn. 13, dates it more generally to “Pre-352,” because “This is the year in which Beth She‘arim was destroyed”; see also Tal 2012, 38, §7.5.1). The plene form representing an i-vowel in the first syllable is especially striking in contrast to the Greek form Μαριαμένη with a-vowel in another inscription in the same chamber, evidently referring to the same person (Mazar 1973, 197). Though admittedly meagre, the foregoing come as indisputable evidence of a pre-400 CE a > i shift consistent with Tiberian attenuation representing various times and locales in preTiberian Hebrew. Though they do not prove the antiquity of attenuation’s extensiveness as reflected in the Tiberian tradition, they at least show that Tiberian pronunciation standardised a feature sporadically documented in late antiquity. What is more, given the limited, fragmentary, and equivocal state of the extant relevant data from the period, it is likely that the historical picture remains somewhat obfuscated. One should bear in mind, among other considerations, that though plene spellings with yod unambiguously represent an i-vowel, defective spellings do not unequivocally reflect a. It is thus not unreasonable to speculate that results of the a > i shift in question were more common in various types of Hebrew and Aramaic far earlier than the Masoretic tradition crystallised and, therefore, that the apparent innovation that Tiberian Hebrew exhibits might rather be a case of the preservation and standardisation of a relatively early second- 38 The Historical Depth of the Tiberian Reading Tradition ary development, perhaps especially characteristic of specific types of Hebrew or Aramaic. 4.0. Structure of the Monograph Like the seven cases summarised above, the vocalic realisations treated in the body of this monograph must be regarded as departures from the pronunciation tradition reconstructable on the basis of the consonantal text. In this sense, the extant Tiberian vocalisations are secondary and relatively late. This, however, is only part of the picture. In all cases, the realisations attested in the pronunciation tradition are themselves characterised by substantial historical depth. Their innovation in no case postdates the Second Temple Period, as is clear from their attestation in the combined Tiberian LBH written and reading tradition, DSS Hebrew, SH, the Hebrew of BS, Tannaitic RH, and forms of Second Temple Aramaic. What is more, in several instances, CBH and/or Iron Age epigraphic material shows that the relevant secondary feature had already developed as a minority alternative prior to Second Temple times. In such cases, the Tiberian reading tradition engages in what may be characterised as the late extension of an otherwise early peripheral feature. This is consonant with the reading tradition’s profile as one that crystallised during Second Temple times, simultaneously absorbing late features and preserving genuine Iron Age traits. The monograph is divided into two parts. The shorter Part I focuses on what may be considered conscious, theologically motivated developments. In such cases, certain phenomena the oral realisation of which had come for various reasons to be deemed Introduction 39 problematic were substituted in the pronunciation tradition, though not in the consonantal text, with more acceptable alternatives. Such examples serve as a useful introduction into the conceptual domain of written-reading dissonance in the Tiberian biblical tradition. They differ in kind, however, from many of the features discussed in Part II. These seem to reflect written-reading dissonance that resulted from developments within Hebrew that had greater effect on the pronunciation tradition than on the orthographic tradition. Crucially, whatever the character of the development—whether motivated by concerns of propriety or driven by unconscious linguistic evolution—all the features listed below are similarly characterised by a degree of mismatch between their written representation and their oral realisation. This is most often due to secondary development—again, either deliberate or unconscious—in the Hebrew preserved in the reading tradition. In a few cases, conversely, it seems that the written and reading components of the Tiberian biblical tradition present alternatives of more or less equal antiquity that became fused in the combined written-reading tradition. The structure of the monograph is as follows: Part I: Conscious Replacement ▪ ch. 1: The Tetragrammaton ▪ ch. 2: ‫ת־פנֵ י יְּהוָ ה‬ ְּ ‫ ֵל ָראֹות ֶא‬and Similar ▪ ch. 3: Ketiv-Qere Euphemisms Part II: Linguistic Development • phonology ▪ ch. 4: The Proper Name Issachar 40 The Historical Depth of the Tiberian Reading Tradition ▪ • • • ch. 5: ‫ לקר(א)ת‬liqra(ʾ)ṯ pronominal morphology ▪ ch. 6: The 2MS Endings ▪ ch. 7: The 2FS Endings ▪ ch. 8: The Qere Perpetuum ‫ִהוא‬ ▪ ch. 9: The 2/3FPL Endings verbal stem morphology ▪ ch. 10: Nifalisation ▪ ch. 11: Hifilisation ▪ ch. 12: Pielisation ▪ ch. 13: Hitpaelisation verbal morphosyntax ▪ ch. 14: Ṭɛrɛm Qaṭal ▪ ch. 15: Ha-qaṭal ▪ ch. 16: Wayyiqṭol ▪ ch. 17: 1st-person Wayyiqṭol ▪ ch. 18: I-y We-yiqṭol for Weqaṭal In some of the cases discussed, the notion of divergent pronunciation traditions—one embodied in the Tiberian vocalisation, the other underlying the Tiberian written text—is uncontroversial or, at the very least, represents a commonly suggested scholarly option, e.g., those discussed in chs 1–3 and 10–13. In other cases, such an explanation has been only rarely proposed and alternative accounts are far more frequently suggested in the literature. For example, according to a common approach to the Tiberian 2MS endings ‫ ָת‬- and ‫ָך‬- in ch. 6, there is no written-reading Introduction 41 dissonance. Rather, both components of the tradition are thought to reflect vowel-final endings, with the routine lack of a final mater attributed to an anomalous (though now standard) orthographic convention. Likewise, explanations for the qere perpetuum ‫ ִהוא‬in the Tiberian Pentateuch in ch. 8 typically hang on the move from defective to plene orthography and similarity in letter shape. Though the rather implausible prospect of an epicene 3CS form has also been raised, the possibility that the spelling and vocalisation might both correctly reflect divergent realisations of the 3FS independent subject pronoun has been rarely entertained. Notwithstanding the existence of plausible and accepted alternative explanations in the case of some of the phenomena discussed in the studies below, the approach here is intentionally programmatic. That is, a conscious effort is made to explore the suitability and ramifications of the view that phonetic dissonance plays a determinative role in all of the relevant features and, as such, is a reality that should routinely be taken into consideration in biblical studies, whether linguistic, exegetical, textual, or literary. The study closes with a conclusion that summarises results, highlights meaningful trends, and discusses ramifications and potential avenues of future study. PART I: CONSCIOUS REPLACEMENT 1. THE TETRAGRAMMATON The routine spelling ‫ יהוה‬in both biblical and extra-biblical sources implies an originally phonetic realisation along the lines of *yahwɛ. Additionally, the contraction ‫* חיהוה‬ḥa(y)-yahwɛ (< ‫* חי יהוה‬ḥay yahwɛ) in Iron Age epigraphy (Arad 21.5; Lachish 3.9; cf. Lachish 6.12; 12.3) presupposes that the form ‫ יהוה‬was realised with an initial consonant identical to that with which ‫חי‬ ends (Suriano 2013, 752). Whatever the exact ancient pronunciation of the divine name, by the time that the medieval Tiberian Masoretic reading tradition was textualised in the form of vowel points, any phonetic pronunciation had long been eclipsed by alternative realisations: 1. Usually, the phonetic realisation is that of the dedicated plural-of-majesty + 1CS possessive suffix ‫ ֲאד ֹנָ י‬ʾădōnaẙ̄ ‘my Lord’,1 resulting in such consonant-vowel combinations 1 The trifold division of labour of forms of the noun ‫‘ ָאדֹון‬lord, master’ with 1CS possessive suffixes is itself a result of secondary development. Almost without exception, possessed forms of ‫ ָאדֹון‬are plurals of majesty, whether the referent is human or divine: thus ‫‘ ֲאד ֹנֶ יָך‬your (MS) lord/Lord’ (22x), ‫‘ ֲאד ֹנַ יִ ְך‬your (FS) lord/Lord’ (2x), ‫‘ ֲאד ֹנָ יו‬his lord/Lord’ (42x; ketiv ‫ אדנו‬with no yod 1x), ‫יה‬ ָ ֶ‫‘ ֲאד ֹנ‬her lord/Lord’, ‫‘ ֲאד ֹנֵ ינּו‬our lord/Lord’ (11x; ‫ ֲאד ֹנֵ נּו‬with no yod 1x), ‫יכם‬ ֶ ֵ‫‘ ֲאד ֹנ‬your (MPL) lord/Lord’ (11x), ‫יהם‬ ֶ ֵ‫‘ ֲאד ֹנ‬their (MPL) lord(s)/Lord’ (11x). This points to a single early 1CS form ‫ ֲאד ֹנַ י‬in the sense of ‘my lord/lords/Lord’. If so, the current Tiberian trichotomy of ‫‘ ֲאד ֹנִ י‬my (human) lord’, ‫‘ ֲאד ֹנָ י‬my (divine) Lord’, and ‫‘ ֲאד ֹנַ י‬my (human) lords’ is secondary, having added a special singular form for human referents and a special pausal-like form for reference to the Israelite © 2023 Aaron D. Hornkohl, CC BY-NC 4.0 https://doi.org/10.11647/OBP.0310.01 46 The Historical Depth of the Tiberian Reading Tradition as ‫( יְּ הֹוָ ָֹ֨ה‬L Gen. 3.14) and ‫הוָ֧ה‬ ָ ְּ‫( י‬L Gen. 3.13), both pro̊̄ 2 nounced ʾădōnay. 2. Alternatively, when preceded or followed by the word ‫ ֲאד ֹנָ י‬ʾădōnaẙ̄ ‘my Lord’, the realisation is that of ‫ֹלהים‬ ִ ‫ֱא‬ ʾĕlōhīm ‘god’, e.g., ‫( יֱ הוִ ֙ה‬L Gen. 15.2) or ‫הוה‬ ֵ֗ ִ ְּ‫( י‬L Deut. 3.24), both pronounced ʾĕlōhīm.3 deity. Cf. the lone instance of preservation of the non-divine plural ‫ֲאד ֵ֗ ַֹני‬ ʾădōnāy ‘my lords’ (Gen. 19.2). In the Samaritan reading tradition, pho- nological processes have resulted in the levelling of any distinction between forms of ‫ אדון‬with 1CS suffixes that refer to humans—Tiberian ‫‘ ֲאד ֹנִ י‬my (human) lord’ and ‫ ֲאד ֵ֗ ַֹני‬ʾădōnāy ‘my lords’ are both realised as ̊̄ ‫ אדני‬adanni. The form ‫ אדני‬in reference to the deity in the Samaritan ̊̄ ni, ̊̄ i.e., with no gemination. The tradition is generally realised as ada Samaritan realisation of the tetragrammaton is šēmå. 2 JM (§16f fn. 1) opines that the vocalisation ‫( יְ הוָ ה‬lacking ḥolam) com- mon in L (as opposed to the rarer ‫יְ הֹוָ ה‬, with ḥolam) is based on Aramaic šǝma ̊̄ ‘the name’, also known from the Targumic reading tradition and similar to the Samaritan. However, beyond the fact that the realisation ʾĕlōhīm is also often represented by forms lacking an explicit ḥolam vowel sign, e.g., ‫( יֱ הוִ ה‬L Gen. 15.2), certain features in the Masoretic vocalisation (also noted in JM §16f) show that šǝma ̊̄ cannot have been the Tiberian realisation. For example, the vocalisation of the preposi̊̄ tions -‫ ֵמ‬, -‫ ַב‬, and -‫ ַל‬presuppose a following ă-vowel, as in ʾădōnay, whereas šǝma ̊̄ would have required preceding -‫ ִמ‬, -‫ ִב‬, and -‫ ִל‬, respectively. 3 According to Khan (2013b, 464), the vocalisation of ‫ יְ הוה‬with simple shewa (as opposed to the composite shewas in ‫ ֲאד ֹנָ י‬and ‫ֹלהים‬ ִ ‫)א‬ ֱ “is a vestige of a primitive stage of the development of Tiberian vocalization, in which a shewa rather than a ḥaṭeph sign was written on the ʾalef.” Cf. the vocalisation of ‫ יהוה‬with composite shewa in accord with the vocal- 1. The Tetragrammaton 47 In other words, according to the medieval Tiberian tradition, the written form ‫ יהוה‬is consistently to be read with the consonants and vowels of an alternative divine epithet.4 1.0. Second Temple Evidence But this medieval convention has far earlier roots. Against the suspicion that substitutive readings for ‫ יהוה‬such as ʾădōnaẙ̄ and ʾĕlōhīm should be chalked up to rabbinic or medieval hypersensitivity to sacrilege, it should be noted that the practice of reserving special treatment for the divine name was already widespread in the last centuries before the Common Era and may extend more deeply into history. In some DSS Hebrew and Aramaic manuscripts, e.g., 11Q5 (=11QPsa); 1QpHab; 4Q243 (=4QpsDana ar), and in certain some Greek manuscripts, as well, e.g., 8ḤevXII gr (Roberts 1951, 173–75; Vasileiadis 2014), the name is distinguished from the surrounding words via the use of old Canaanite script. In some Aramaic DSS, the name is replaced by dots (see, e.g., 4Q196 f18.15). Presumably reflecting special reverence for the name (Yeivin 1980, 59, §103), such strategies had the practical effect of reminding readers to avoid pronouncing it as written. Consider, e.g., hwhy in Ps. 151 as preserved in 11QPsa 28.6, 11 (underlined below in lns 3, 8 of Figure 1). isation of ʾădōnaẙ̄ in some Babylonian manuscripts (Yeivin 1985, II:912; Khan 2013a, 44). 4 Readers unfamiliar with the convention of pronouncing ‫ יהוה‬with the vowels of ‫אד ֹנָ י‬, ֲ inadvertently coined on the basis of the written-reading combination ‫ יְּ הֹוָ ה‬the hitherto unknown divine name yĕhōva,̊̄ i.e., ‘Jehovah’. 48 The Historical Depth of the Tiberian Reading Tradition Figure 1: 11QPsa (11Q5) 28.3–14. Image used by permission of the Israel Antiquities Authority Similarly, as already noted, in the Samaritan and Aramaic reading traditions, God’s name was replaced with the Aramaic word ‫ שמא‬šǝma ̊̄ ‘the name’. In the Greek, Syriac, and Latin versions, it was replaced with words meaning ‘Lord’—κύριος, ‫ܡܪܝܐ‬, and Dominus, respectively—an approach commonly perpetuated in modern Bible translations. And in some cases where the Masoretic Bible vocalises ‫ יהוה‬as ʾădōnaẙ̄ ‘my LORD’ or ʾĕlōhīm ‘GOD’, a parallel DSS text has the actual consonants of the replacement form, e.g., ‫יְּהוֶ֖ה‬ ָ (MT Deut. 32.27) || ֵ֗‫אדנֵ֗ י‬ ֵ֗ (1Q5 f16–19.9), ‫ֲאד ָֹנָ֤י‬ ‫( יְ הוִ ה‬MT Isa. 50.5) || ‫( אדוני אלוהים‬1QIsaa 42.6). 1. The Tetragrammaton 49 2.0. Iron Age Epigraphy and the Classical Biblical Hebrew Written Tradition But at least two questions remain: (1) does the convention of reference to the Israelite deity as ‫‘ ֲאד ֹנָ י‬my Lord’ predate the Second Temple Period and, if so, by how much? (2) Does the convention of replacing the original pronunciation of ‫ יהוה‬with that of ‫ֲאד ֹנָ י‬ predate the Second Temple Period and, if so, by how much? On the first question, epigraphic evidence seems clear. In the admittedly meagre corpus of Iron Age Hebrew inscriptions, referents designated by the forms ‫ אדני‬and ‫ יהוה‬are kept rigidly distinct, the former consistently referring to a human and never substituting for the latter. The Aramaic equivalent to ‫אדני‬, i.e., ‫מרא‬, does, however occur in the fifth-century documents from Elephantine. Moreover, ‫ אדני‬appears in reference to the Israelite deity as a minority form throughout the Tiberian consonantal tradition, including in acknowledged CBH texts in the Pentateuch, Former Prophets, and Latter Prophets. Excluding sequences of ‫אדני יהוה‬ and ‫( יהוה אדני‬where ‫ אדני‬was originally in apposition to *yahwɛ), MT instances in which ‫ אדני‬refers to the Israelite deity total some 133 cases (against more than 6800 cases of ‫)יהוה‬. In books where the ‫ אדני‬occurs, it normally makes up a small minority of references to the Israelite deity. See Table 1. There may be a diachronic factor in the above distribution, as the statistical outliers are the post-exilic compositions of Daniel (where cases of ‫ אדני‬in reference to the Israelite deity outnumber those of ‫ ;יהוה‬Daniel also has instances of ‫ מרא‬in reference to the Israelite deity: Dan. 50 The Historical Depth of the Tiberian Reading Tradition 2.47; 5.23) and Lamentations (where ‫ אדני‬comes in over a third of the cases). Table 1: ‫ אדני‬for ‫ יהוה‬in the written component of the Tiberian biblical tradition Instances of ‫אדני‬ 11 14 2 47 23 4 8 2 1 1 1 6 5 2 5 1 1 1 Book Dan. Lam. Neh. Ps. Isa. Amos Gen. Mal. Job Ezra Mic. Exod. Ezek. Judg. Kgs Zech. Josh. Num. Instances % Instances Instances % Book ‫אדני‬ ‫אדני‬ of ‫יהוה‬ of ‫אדני‬ of ‫יהוה‬ 8 57.9 0 Lev. 311 0 32 30.44 0 Deut. 550 0 17 10.53 0 Sam. 473 0 695 6.33 0 Jer. 726 0 450 4.86 0 Hos. 46 0 81 4.71 0 Joel 33 0 165 4.62 0 Obad. 7 0 46 4.17 0 Jon. 26 0 32 3.03 0 Nah. 13 0 37 2.63 0 Hab. 13 0 40 2.44 0 Zeph. 34 0 398 1.49 0 Hag. 35 0 434 1.14 0 Prov. 87 0 175 1.13 0 Song 0 0 534 .93 0 Qoh. 0 0 133 .75 0 Est. 0 0 224 0.44 0 Chron. 559 0 396 0.25 One is inclined to question the authenticity of CBH cases of ‫ אדני‬for ‫יהוה‬. However, when it comes to the fourteen occurrences in the Pentateuch (Gen. 18.3, 27, 30, 31, 32; 19.18; 20.4; Exod. 4.10, 13; 5.22; 15.17; 34.9, 9; Num. 14.17), it is worth noting that the composite Samaritan written-reading tradition agrees with the Tiberian consonantal tradition on thirteen; the exception is Exod. 15.17 where MT ‫(‘ ִמ ְּק ָ ָ֕דש אֲדנָ ֹׁ֖י כֹונְּ נֵ֥ ּו יָ ֶ ַֽדיָך‬the) sanctuary, O LORD, that your hands established’ || SP ‫מקדש יהוה כוננו ידך‬ maqdɑš šēmɑ̊ kūnēnu yēdɑk, which also in 4Q14 6.41 reads ‫מקדש‬ ‫יהוה כוננו ידך‬. 1. The Tetragrammaton 51 Beyond the Pentateuch, MT Isaiah’s ‫ אדני‬is regularly paralleled by the same in 1QIsaa; of the 23 MT instances, 1QIsaa reads ‫ אדוני‬in seventeen of them (Isa 3.18; 4.4; 6.1, 8; 7.20; 8.7 [erasure of ‫ ;]יהוה‬9.16; 10.12; 11.11; 21.6, 8; 29.13; 30.20; 37.24; 38.14, 16; 49.14). MT ‫ אדני‬is also paralleled by the same in other DSS Isaiah material (MT Isa. 3.17 || 4Q56 3i.12; MT Isa. 21.16 || 4Q55 f10–11i+12–14.35; MT Isa. 38.16 || 1Q8 16.4; MT Isa. 49.14 || 4Q58 4.23) and elsewhere (MT Amos 9.1 || Mur88 8.7; MT Ps. 2.4 || 11Q7 f1–2.3; MT Ps. 35.17 || 4Q83 f6.3; MT 38.16 || 4Q83 f9ii.2; MT Ps. 38.23 || 4Q83 f9ii.5; MT Ps. 54.6 || 4Q83 f11–12.8; MT Ps. 66.18 || 4Q83 f14ii.30; MT Ps. 86.5 || 1Q10 f1.1; MT Ps. 89.50 || 4Q87 f8.1; MT Ps. 89.51 || 4Q87 f8.2; MT Lam. 1.15 || 4Q111 3.6). On the above evidence, the interchange of ‫ אדני‬and ‫יהוה‬ dates back to at least the late Second Temple Period. The fact that the Tiberian Torah and the SP agree on ‫ אדני‬as nomenclature for the Israelite deity points to a convention that had become rooted before the separation of the proto-Tiberian and proto-Samaritan traditions (see Kartveit 2009; Pummer 2012; Kantor 2020, 108–9 for background). Regarding the antiquity of the avoidance of the pronunciation of ‫יהוה‬, unambiguous information is much harder to come by, since it is difficult to reconstruct the pronunciation that originally accompanied the Tiberian consonantal text, before it became wedded to the Tiberian reading tradition. In other words, assuming that the graphic sequence ‫ יהוה‬was originally pronounced along the lines of *yahwɛ, does the Tiberian written tradition give any hint as to avoidance of this pronunciation in 52 The Historical Depth of the Tiberian Reading Tradition accord with what is seen in the Tiberian reading tradition and the other Second Temple traditions listed above? Schniedewind (2004, 32) notes that Chronicles “often replaces the sacred four letter name of God in its source (known from the books of Samuel and Kings) with the more generic Elohim (which translates simply as ‘God’)” (see Japhet 2009, 24, fn. 64). Japhet (2009, 24–30) disagrees with this assessment of Chronicles, but makes a similar claim about Qohelet and the Elohistic Psalter (on the latter see also Ben-Dov 2010, 81–82, 87–88, 101–4; Suriano 2013, 752). The latter, encompassing Pss 42–83 and showing no signs of LBH, are apparently classical works evincing reticence to overuse of the tetragrammaton.5 Suriano (2013, 752) sees even earlier avoidance of ‫ יהוה‬in the preference for ‫ אלהים‬in the E source of the Pentateuch, though this is considered a separate issue by Japhet (2009, 29, fn. 85). 3.0. Conclusion Given the extant evidence, it is not entirely clear how long the supposed realisation *yahwɛ persisted. However, avoidance of the name dates as far back as the composition of CBH texts (the Elohistic Psalter, if not the putative E source of the Pentateuch). Further, the pronunciation of the tetragrammaton as ʾădōnaẙ̄ reflected in the medieval Tiberian vocalisation signs clearly preserves pre-medieval sensitivities characteristic of multiple Second Temple biblical traditions, wherein early use of the pluralis majestatis epithet ‫‘ אדני‬my Lord’ for the Israelite deity was 5 In the Elohistic Psalter the counts of divine epithets are ‫ אלהים‬245 times, ‫ יהוה‬45 times, and ‫ אדני‬23 times. 1. The Tetragrammaton 53 extended and became standard, even where ‫ יהוה‬was still written. Indeed, the graphic form of name of the Israelite deity ‫ יהוה‬was ̊̄ that so identified with pronunciations along the lines of ʾădōnay, the writing of ‫ אדני‬itself came to be proscribed in Second Temple texts (Japhet 2009, 16–19; cf. 31, fn. 96). 2. ‫ת־פנֵ י יְּ הוָ ה‬ ְּ ‫ ֵל ָראֹות ֶא‬AND SIMILAR Eleven times in the Tiberian biblical tradition readers encounter an expression composed of a form of the nifʿal ‫ נִ ְר ָאה‬and the phrase ‫יְּ הוָ ה‬/‫ֹלהים‬ ִ ‫ ְפּנֵ י ֱא‬, with or without an intervening direct object marker or preposition. Standard renderings include ‘appear be- fore the face of God/the LORD’ and ‘appear in God’s/the LORD’s presence’. It has been claimed, however, that in all such cases the consonantal spelling was actually intended to represent a form of the qal verb ‫ר ָאה‬,ָ with the meaning ‘see God’s/the LORD’s face’, and that the form was only secondarily interpreted as nifʿal out of concerns for theological propriety (BDB 816b, 908a). Such changes were presumably made both in deference to a general aversion to anthropomorphising the Israelite deity and for the sake of theological harmony in adherence to the prohibition against seeing the divine visage, which employs qal ‫‘ ָר ָאה‬see’, in (1). (1) ‫ֹתי ֶאת־‬ ֙ ִ ‫ וַ ֲה ִסר‬...‫תּוכל ִל ְּׁר ּ֣את אֶ ת־פָ נָ ָ֑י ִ ֛כי ַֽל ֹא־יִ ְּׁר ַּאּ֥נִ י ָה ָא ָ ֶ֖דם וָ ָ ַֽחי׃‬ ֶ֖ ַ ‫אמר ֵ֥ל ֹא‬ ֶ ֹ ‫וַ ָ֕י‬ ‫לא י ֵָר ִֽאּו׃‬ ּ֥ ‫ת־אח ָ ָֹׂ֑רי ּופָ נַּ ֹׁ֖י‬ ֲ ‫ית ֶא‬ ָ ‫ַכ ֵ֔ ִפי וְּ ָר ִ ֶ֖א‬ ‘And he said, “You cannot see my face, because no mortal will see me and live…. And I will remove my hand, and you will see my back, but my face will not be seen.”’ (Exod. 33.20, 23). By avoiding the qal form in other verses, readers might be helped to avoid the misconception that God’s face could be seen. © 2023 Aaron D. Hornkohl, CC BY-NC 4.0 https://doi.org/10.11647/OBP.0310.02 56 The Historical Depth of the Tiberian Reading Tradition Instances where qal in ‫יְּ הוָ ה‬/‫ֹלהים‬ ִ ‫ל־)פנֵ י ֱא‬ ְּ ‫א‬/‫ת־‬ ֶ ‫(א‬ ֶ ‫‘ ָר ָאה‬see the God’s/the LORD’s face’ are thought to have been reinterpreted as nifʿal due to theological concern may be contrasted with cases in which ‫ל־)פנֵ י‬ ְּ ‫א‬/‫ת־‬ ֶ ‫(א‬ ֶ ‫‘ ָר ָאה‬see X’s face’ has no divine referent and was maintained.1 There are even comparable cases in which qal ‫ ָר ָאה‬is preserved with the face of a divine referent as object.2 The current chapter examines cases of apparent substitution for qal, attempting to determine whether the hypothesis of secondary development is equally applicable to all of them. It then seeks to gauge the antiquity of the reinterpretation. 1.0. Unambiguous Cases of Dissonance Evidence of morphological mismatch involving both orthography and vocalisation suggest that at least some cases of nifʿal ‫*נִ ְּר ָאה‬ ‫יְּהוָ ה‬/‫ֹלהים‬ ִ ‫ל־)פנֵ י ֱא‬ ְּ ‫ ֶא‬/‫(את־‬ ֶ are secondary reworkings of original formulations with qal ‫ר ָאה‬.ָ The most conspicuous cases of mismatch between the written and reading components of the Tiberian biblical tradition are reproduced in examples (2)–(4). (2) ‫ֱֹלהיך‬ ֶ ִ֔ ‫ת־פנֵי יְּׁ הוָ ּ֣ה א‬ ְּׁ ֶ‫ת־א ְּר ְּצ ֵָ֔ך ַב ֲע ַֹֽל ְּת ֵָ֗ך לֵ ָראֹות א‬ ַ ‫יש ֶ ַֽא‬ ֙ ‫מד ִא‬ ֹ ֵ֥ ‫וְּ לֹא־יַ ְּח‬... ‫ָשֹלֵ֥ ש ְּפ ָע ִ ֶ֖מים ַב ָש ָנַֽה׃‬ ‘…and no one shall covet your land, when you go up to appear before the face of the LORD your God three times ̊̄ in the year.’ (Exod. 34.24; SP ‫ להראות‬lērraʾot; Greek ὀφθῆναί; Vulgate et apparente; TO ‫ ;לאתחזאה‬Syriac ‫)ܠܡܬܚܙܝܘ‬ 1 Gen. 31.2, 5; 32.21; 43.3, 5; 44.23, 26; 46.30; 48.11; Exod. 10.28, 28; 34.35; 2 Sam. 3.13, 13; 14.24, 32; 2 Kgs 25.19 (|| Jer. 52.25); Jer. 52.25 || (2 Kgs 25.19); Est. 1.14; Dan. 1.10. 2 Gen. 32.31; 33.10; Judg. 6.22; Jer. 18.17; Job 33.26. 2. ‫ת־פנֵ י יְּ הוָ ה‬ ְּ ‫ ֵל ָראֹות ֶא‬and Similar (3) 57 ...‫ֱֹלהיך ַב ָמ ֶ֖קֹום ֲא ֶשר יִ ְּב ָ ָׂ֑חר‬ ֶ ִ֔ ‫ת־פנֵי יְּׁ הוָ ּ֣ה א‬ ְּׁ ֶ‫ְּבבֹוא ָכל־יִ ְּש ָר ֵֵ֗אל לֵ ָראֹות א‬ ‘When all Israel comes to appear before the face of the Lord your God at the place that he will choose,…’ (Deut. ̊̄ 31.11; SP ‫ להראות‬lērraʾot; Greek ὀφθῆναί; Vulgate ut appareant; TO ‫ ;לאתחזאה‬Syriac ‫)ܠܡܬܚܙܝܘ‬ (4) ‫מס ֲח ֵצ ָ ַֽרי׃‬ ֹ ֵ֥ ‫י־ב ֵ ֵ֥קש ֛ז ֹאת ִמיֶ ְּד ֶכֶ֖ם ְּר‬ ִ ‫ִכי ָת ֵ֔בֹאּו לֵ ָראֹׁ֖ ֹות פָ נָ ָ֑י ִמ‬ ‘When you come to appear before me, who has required of you this trampling of my courts? (Isa. 1.12; 1QIsaa ‫;לראות‬ Greek ὀφθῆναί; Vulgate ante conspectum meum; TJ ‫;לאתחזאה‬ Syriac ‫)ܠܡܚܙܐ‬ In all of the above, an infinitive construct with transparently qal spelling (i.e., lacking the heh of the corresponding nifʿal infini- tive) is realised as nifʿal in the pronunciation tradition. While syncope of heh is common in certain environments in ancient Hebrew, the nifʿal infinitive is not one of them. For example, un- ambiguous nifʿal infinitive construct forms of ‫ נִ ְּר ָאה‬come ten times in the Hebrew Bible, consistently with the expected heh, even following a cliticised preposition.3 The three exceptional cases in (2)–(4) above, where the infinitives are read as nifʿal despite ap- parent qal orthography, all make reference to the deity’s face/ presence. The exclusive connection between the mixed qal-nifʿal form ‫ ֵל ָראֹות‬and contexts including reference to the divine face/ presence is unlikely to be random. In all instances, ancient versional evidence agrees with the Tiberian reading tradition on the meaning ‘appear’. This extends to the Samaritan written tradition, which has the unambiguous 3 Lev. 13.7, 14; Deut. 31.11; Judg. 13.21; 1 Sam. 3.21; 2 Sam. 17.17; 1 Kgs 18.2; Isa. 1.12; Ezek. 21.29; Mal. 3.2. 58 The Historical Depth of the Tiberian Reading Tradition nifʿal theological lectio facilior ‫ להראות‬in both of the Pentateuchal instances. It is of interest that in the parallel to (4) in the Great Isaiah Scroll, the spelling ‫( לראות‬1QIsaa 1.14) is also most straightforwardly analysed as a qal infinitive. The form in the Peshiṭta is also a match for that represented by the Tiberian written tradition, while the Greek, Latin, and TJ reflect the same understanding as the Tiberian reading tradition. Syntactically, it is worth pointing out that, in the case of a variety of verbs, ‫ ֶאת־ ְּפנֵ י‬is synonymous with ‫ל ְּפנֵ י‬, ִ ‫ ֶאל־ ְּפנֵ י‬, and ‫ִעם‬ ‫פנֵ י‬, ְּ meaning ‘before, in the presence of’.4 The particle ‫ ֵאת‬in such cases is most plausibly analysed as the comitative preposition ‫ֵאת‬ ‘with’. If so, in cases (2) and (3), the nifʿal realisation in the Tiberian recitation tradition also involves the reinterpretation of the originally accusative/direct object particle ‫ ֶאת‬as the homony- mous preposition ‫‘ ֵאת‬with’. In (4), the presumed original syntax of qal infinitive ‫* ִל ְראֹות‬ followed by ‫‘ ָפּ ָנָ֑י‬my face, presence’ with no intervening preposition or particle is within the bounds of acceptable BH usage.5 The grammaticality of the same formulation with nifʿal is more difficult to gauge. On the one hand, phrases with ‫ ְפּנֵ י‬have two char4 Gen. 19.13, 27; 27.30; 33.18; 43.34; Exod. 10.11; 32.11 (?); Lev. 4.6, 17; 10.4; 1 Sam. 2.11, 17, 18; 22.4; 1 Kgs 12.6; 13.6 (2x?); 2 Kgs 13.4 (?); 16.14; Jer. 26.19 (?); Zech. 7.2 (?); 8.21, 22 (?); Ps. 16.11; 21.7; 140.14; Job 2.7; Prov. 17.24; Est. 1.10; Dan. 9.13 (?); 2 Chron. 33.12 (?). 5 See, e.g., Gen. 32.21; 33.10, 10; 43.3, 5; 44.23, 26; 48.11; Exod. 10.28, 29; 2 Sam. 14.32; Job 33.26, all with qal ‫‘ ָר ָאה‬see’ preceding ‫‘ ָפנִ ים‬face’ with no intervening particle. 2. ‫ת־פנֵ י יְּ הוָ ה‬ ְּ ‫ ֵל ָראֹות ֶא‬and Similar 59 acteristics common for so-called accusatives of place, in that (a) they begin construct phrases and (b) they begin with a bilabial (GKC §118g). Also, in some LBH texts, ‫ ָפּנִ ים‬functions as a locative adverbial in the sense ‘before, toward, in front, eastward’ (see BDB 815, §6). For example, consider (5). (5) ‫רּואל׃‬ ַֽ ֵ ְּ‫אתם א ָֹת ֙ם ְּבסֹוף ַה ֵ֔ ַנ ַחל ְּׁפנֵ ֹׁ֖י ִמ ְּד ַ ֵ֥בר י‬ ַ֤ ֶ ‫ּומ ָצ‬ ְּ ... ‘…You will find them at the end of the valley, east of/in front of the wilderness of Jeruel.’ (2 Chron. 2.16) The syntax of qal ‫ ָר ָאה ְּפנֵ י‬is clearly acceptable, that of nifʿal ‫נִ ְּר ָאה‬ ‫ ְּפנֵ י‬questionable.6 Since the orthography in (4) is transparently qal, syntactic considerations there only confirm the secondary status of the nifʿal recasting. But questionable syntax may be a more decisive factor in the assessment of ambiguous cases. Before proceeding to more ambiguous cases of possible qal > nifʿal shift, it is worth examining potentially related cases in- volving qal and hifʿil. Example (6) presents an apparent instance of the shift qal > hifʿil. (6) ‫יכם ַב ֶ ֵ֗ד ֶרְך ָל ֵ֥תּור ָל ֶכ֛ם ָמ ֶ֖קֹום ַ ַֽל ֲח ַֹֽנ ְּת ֶכָׂ֑ם ָב ֵאש ׀ ֵ֗ ַליְּ ָלה לַּ ְּׁר ִֽא ְּׁתכֶ ם‬ ֶֶ֜ ֵ‫ַהה ָֹ֨ ֵֹלְך ִל ְּפנ‬ ‫כּו־בּה‬ ֵָ֔ ‫ַב ֶ ֙ד ֶר ְ֙ך ֲא ֶשר ֵ ַֽת ְּל‬ ‘…who went before you in the way to seek you out a place to pitch your tents, in fire by night and in the cloud by day, to show you by what way you should go.’ (Deut. 1.33; SP ‫ ;להראתכם‬4Q35 f2–4.26: ‫ ̇;להראות‬Greek δεικνύων ὑμῖν; Vulgate ostendens vobis; TO ‫ ;לאחזיותכון‬Syriac ‫)ܚܘܝܟܘܢ‬ 6 All four instances in which nifʿal ‫ נִ ְּר ָאה‬precedes ‫ ָפנִ ים‬with no interven- ing particle are among those identified as potential cases of revocalised qal forms: Exod. 25.15; 34.20; Isa. 1.12; Ps. 42.3. All are discussed in the present study. 60 The Historical Depth of the Tiberian Reading Tradition According to the hifʿil realisation in the Tiberian reading tradition, explicit in the orthography of SP and 4Q35, the Tiberian written component’s ‫ לראתכם‬is to be understood as the causative ‘make you (MPL) see, show you (MPL)’, which interpretation is supported by the versions. The ostensible qal ‫*ל ְּרא ְֹּת ֶכם‬ ִ ‘for your (MPL) seeing, for you (MPL) to see’ would presumably have re- ferred to the purpose of providing light on the road at night. If this is indeed a case of recasting, the motivation would seem to be to forestall misunderstanding, lest readers conclude that God could be seen. Example (7) exhibits a potential hifʿil > qal shift. (7) ‫ידם׃‬ ַֽ ָ ‫א־פ ִנ֛ים אֶ ְּׁר ֵאֹׁ֖ם ְּביֵ֥ ֹום ֵא‬ ָ ֹ ‫אֹויָׂ֑ב ָ֧עֹ ֶרף וְּ ַֽל‬ ֵ ‫־ק ִ ֵ֥דים ֲא ִפ ֵיצֶ֖ם ִל ְּפ ֵני‬ ָ ‫ְּכ ַֽר ַּוח‬ ‘I will scatter them before their enemies like dust blowing in front of a burning east wind. (My) back and not (my) face I will show them on the day of disaster.’ (Jer. 18.17; Greek δείξω αὐτοῖς; Vulgate ostendam eis; TJ ‫ ;אחזינון‬Syriac ݁ ‫)ܐܚܘܐ ܐܢܘܢ‬ Here, were it not for the vocalisation, the most straightforward reading would arguably be as hifʿil ‫*א ְּר ֵאם‬. ַ This not only fits the ellipses ‘knape and not face I will show them’, but has the support of the versions and modern translations. The Tiberian reading tradition’s qal may betray aversion to the notion that God might actually show his face. But the resulting phraseology, presumably entailing adverbial accusatives, gives the awkward ‘(with) knape and not (with) face I will see them’. 2. ‫ת־פנֵ י יְּ הוָ ה‬ ְּ ‫ ֵל ָראֹות ֶא‬and Similar 61 2.0. Ambiguous Cases Whereas cases (2)–(4) above present unequivocal cases of dissonance between a written qal and a nifʿal in the pronunciation tradition, other cases of mismatch are not so readily apparent. Consider (8). (8) ‫ֱֹלהים׃‬ ִֽ ִ ‫ים ְּל ֵ ֵ֪אל ָ ֵ֥חי ָמ ַ ֵ֥תי ָא ָׂ֑בֹוא וְּׁ ְ֝אֵ ָר ֶֶ֗אה ְּׁפנֵ ּ֣י א‬ ֮ ‫אֹלה‬ ִ ‫ָצ ְּמ ָָ֬אה נַ ְּפ ִָֹ֨שי ׀ ֵל‬ ‘My soul thirsts for God, for the living God. When shall I come that I might appear before God?’ (Ps. 42.3; Greek ὀφθήσομαι; Vulgate et parebo; Targum ‫;ואחמי זיו שכינתא דיהוה‬ ݁ ̈ ‫ܘܐܚܙܐ‬ Syriac ‫ܐܦܝܟ‬ ) The lack of a preposition or particle after the verb makes it possible that consonantal ‫ ואראה‬represents an original qal, ‫*וְ ֶא ְר ֶאה‬ ‘that I may see’. Additionally, while the Greek and Latin show theological concern like that ostensibly behind the Tiberian vocalisation, the Targum and Syriac support a qal ‘see’ reading (though the Targum mitigates by replacing ‘face’ with ‘glory of the presence of the LORD’). Other ambiguous cases include (9) and (10). (9) ‫מֹוע ֙ד חֹ ֶדש‬ ֵ ‫יתָך ְּל‬ ִֵ֗ ִ‫ֹאכל ַמ ֶ֜צֹות ַ ַֽכ ֲא ֶשר ִצּו‬ ַָֹ֨ ‫ים ת‬ ֩ ‫יָמ‬ ִ ‫מר֒ ִש ְּב ַעת‬ ֹ ‫צֹות ִת ְּש‬ ֮ ‫ת־חג ַה ַמ‬ ַ ‫ֶא‬ ‫את ִמ ִמ ְּצ ָ ָׂ֑ריִ ם וְּׁ לא־י ֵָראּ֥ ּו פָ נַּ ֹׁ֖י ֵרי ָ ַָֽקם׃‬ ָ ‫י־בֹו יָ ָצ‬ ֶ֖ ‫ָ ַֽה ָא ִֵ֔ביב ִכ‬ ‘You shall keep the Feast of Unleavened Bread. As I commanded you, you shall eat unleavened bread for seven days at the appointed time in the month of Abib, for in it you came out of Egypt. None shall appear before me emptẙ̄ Greek ὀφθήσῃ; Vulhanded.’ (Exod. 23.15; SP ‫ יראו‬yirraʾu; gate apparebis; TO ‫ ;יתחזון‬Syriac ‫)ܬܬܚܙܘܢ‬ 62 The Historical Depth of the Tiberian Reading Tradition (10) ‫ם־ל ֹא ִת ְּפ ֶ ֶ֖דה וַ ֲע ַר ְּפ ָׂ֑תֹו כֹל ְּב ַ֤כֹור ָבנֶ֙ ֙יָך ִת ְּפ ֶ ֵ֔דה וְּׁ ִֽלא־‬ ֵ֥ ‫מֹור ִת ְּפ ֶדה ְּב ֵֶ֔שה וְּ ִא‬ ֙ ‫ֶּופ ֶַ֤טר ֲח‬ ‫י ֵָראּ֥ ּו פָ נַּ ֹׁ֖י ֵרי ָ ַָֽקם׃‬ ‘The firstborn of a donkey you shall redeem with a lamb, or if you will not redeem it you shall break its neck. All the firstborn of your sons you shall redeem. And none shall appear before me empty-handed.’ (Exod. 34.20; SP ‫יראו‬ ̊̄ Greek ὀφθήσῃ; Vulgate apparebis; TO ‫ ;יתחזון‬Syriac yirraʾu; ‫)ܬܬܚܙܘܢ‬ In both, the verb can easily be read as a qal. The lack of any particle or preposition between the verb and ‫‘ ָפ ַנֶ֖י‬my face, pres- ence’ makes a nifʿal reading in the sense ‘will (not) appear’ questionable. Also, the shift in referent from 2nd- to 3rd-person is jarring. Why not continue each verse with ‫‘ *וְּ לֹא ֵת ָר ֶאה‬and you will not be seen, appear’, if that is the intended meaning? The ancient versions universally translate ‘appear before’, as if ‫ ָפנַ י‬were equivalent to ‫ ְּל ָפנַ י‬and ‫ ֶאת־ ָפנַ י‬or ‫ ָפנַ י‬were an accusative of place (see above). Some modern translations deftly sidestep part of the problem via impersonal rendering, e.g., ‘And none shall appear before me empty-handed’. Yet, this does not resolve the problem of the lack of a preposition or particle. In both cases it seems more likely that the verbs are either impersonal qal forms, ‫*יִ ְּראּו‬ ‘(none) will see’, or nifʿal forms with ‫‘ ָפ ַנֶ֖י‬my face’ as subject, i.e., ‘my face will not be seen in vain’. Cf. the clear instance where ‫‘ ָפנִ ים‬face’ serves as subject of nifʿal ‫ נִ ְּר ָאה‬in example (11) (though, in that instance, too, a qal reading is possible). (11) ‫ת־אח ָ ָֹׂ֑רי ּופָ נַּ ֹׁ֖י ֵ֥ל ֹא י ֵָר ִֽאּו׃‬ ֲ ‫ית ֶא‬ ָ ‫ת־כ ֵ֔ ִפי וְּ ָר ִ ֶ֖א‬ ַ ‫ֹתי ֶא‬ ֙ ִ ‫וַ ֲה ִסר‬ ‘“And I will remove my hand, and you will see my back, but my face will not be seen.”’ (Exod. 33.23) 2. ‫ת־פנֵ י יְּ הוָ ה‬ ְּ ‫ ֵל ָראֹות ֶא‬and Similar 63 Now, consider (12)–(14). (12) ‫ֹלהי‬ ֵ֥ ֵ ‫הוֶ֖ה ֱא‬ ָ ְּ‫האָ ּ֥דן ׀ י‬ ִֽ ָ ‫ת־פנֵ ֵ֛י‬ ְּׁ ֶ‫ּור ִ֔ך א‬ ְּׁ ּ֣‫ָשֹלֵ֥ ש ְּפ ָע ִ ֶ֖מים ַב ָש ָנָׂ֑ה י ֵָראֶ ה כָ ל־ ְּׁזכ‬ ‫יִ ְּש ָר ֵ ַֽאל׃‬ ‘Three times in the year will all your males appear before the Lord, the LORD God of Israel.’ (Exod. 34.23; SP ‫יראה‬ ̊̄ Greek ὀφθήσεται; Vulgate apparebit; TO ‫ ;יתחזון‬Syriac yirraʾi; ‫)ܢܬܚܙܐ‬ (13) ‫ֱֹלהיך‬ ֶ ֶ֗ ‫ת־פנֵ ּ֣י ׀ יְּׁ הוָ ּ֣ה א‬ ְּׁ ֶ‫כּור ָ֜ך א‬ ְּׁ ‫ָשלֹוש ְּפ ָע ִמים ׀ ַב ָש ָ֡ ָנה י ֵָר ֶֶ֨אה כָ ל־ ְּׁז‬ ...‫ּוב ַחג ַה ֻס ָׂ֑כֹות‬ ְּ ‫ּוב ַ ֵ֥חג ַה ָש ֻב ֶ֖עֹות‬ ְּ ‫קֹום ֲא ֶשר יִ ְּב ֵָ֔חר ְּב ַ ָ֧חג ַה ַמ ֛צֹות‬ ֙ ‫ַב ָמ‬ ‘Three times a year will all your males appear before the LORD your God at the place that he will choose: at the Feast of Unleavened Bread, at the Feast of Weeks, and at the Feast ̊̄ Greek ὀφθήσεται; of Booths…’ (Deut 16.16a; SP ‫ יראה‬yirraʾi; Vulgate apparebit; TO ‫ ;יתחזון‬Syriac ‫)ܢܬܚܙܐ‬ (14) ‫ת־פנֵ ּ֥י יְּׁ הוָ ֹׁ֖ה ֵרי ָ ַָֽקם׃‬ ְּׁ ֶ‫וְּ ָ֧ל ֹא י ֵָר ֶאֵ֛ה א‬... ‘…and they shall not appear before the LORD emptẙ̄ Greek ὀφθήσῃ; Vulhanded.’ (Deut. 16.16b; SP ‫ יראו‬yirraʾu; gate apparebit; TO ‫ ;יתחזון‬Syriac ‫)ܬܬܚܙܐ‬ In cases (12)–(14), the fact that the sequence ‫ת־פנֵ י‬ ְּ ‫ ֶא‬can be taken as a prepositional phrase in the sense of ‘before, in the presence of’ legitimises the nifʿal reading of the verbal form ‫ יֵ ָר ֶ ֛אה‬in the meaning ‘will appear’. This is the understanding in the versions. Admittedly, however, the nifʿal reading is no more grammatically felicitous than qal ‫‘ *יִ ְּר ֶאה‬will see’ would be, in which case the ensuing ‫ ֶאת‬would be construed as the marker of the definite accusative/direct object. Another equivocal case is presented in (15). 64 The Historical Depth of the Tiberian Reading Tradition (15) ‫ישּה ַעד יִ גָ ֵ ַ֤מל ַהנַ֙ ַע ֙ר וַ ֲה ִבא ִֵֹ֗תיו וְּׁ נִ ְּׁראָ ה אֶ ת־‬ ֵָ֗ ‫י־א ְּמ ָרה ְּל ִא‬ ָ ‫וְּ ַח ָנֶ֖ה ל ֹא ָע ָל ָָׂ֑תה ִ ַֽכ‬ ‫ד־עֹולם׃‬ ַֽ ָ ‫ְּׁפנֵ ּ֣י יְּׁ ה ָוִ֔ה וְּ ָי ֵַ֥שב ָ ֶ֖שם ַע‬ ‘But Hannah did not go up, for she said to her husband, “As soon as the child is weaned, I will bring him, so that he may appear in the presence of the LORD and dwell there forever.” (1 Sam. 1.22; Greek ὀφθήσεται; Vulgate appareat; TJ ‫ ;ויתחזי‬Syriac ‫)ܘܢܬܚܙܐ‬ Here, the graphic unit ‫ ונראה‬has three contextually defensible analyses: (1) 3MS nifʿal weqaṭal ‫‘ וְּ נִ ְּר ָאה‬and he will appear’, as in the Tiberian reading tradition; (2) 1CPL nifʿal we-yiqṭol ‫‘ *וְּ נֵ ָר ֶאה‬that we may appear’; (3) 1CPL qal we-yiqṭol ‫‘ *וְּ נִ ְּר ֶאה‬that we may see’.7 Thus, while the possibility that an original qal was recast as a nifʿal exists here, the 3rd-person nifʿal option is at least as fitting as the two 1st-person plural options, one of which is, in any case, also nifʿal. Unsurprisingly, the ancient versions agree with the Ti- berian reading tradition on the meaning ‘appear’. 3.0. The Antiquity of the Interpretation Having identified cases in which developments in the Tiberian pronunciation tradition either likely or possibly constitute secondary shifts to avoid a theological difficulty, the most relevant question for this study is: when did the purported qal > nifʿal (or qal > hifʿil) recasting take place? Its secondary nature in some of the aforecited cases seems beyond question. Yet, what should be emphasised is that, even where secondary, the nifʿal reinterpre7 The ensuing spelling ‫ וישב‬is also contextually ambiguous: weqaṭal ‫וְּ יָ ָשב‬ ‘and he will dwell’ or we-yiqṭol ‫‘ וְּ יֵ ֵשב‬that he might dwell’. See below, ch. 18. 2. ‫ת־פנֵ י יְּ הוָ ה‬ ְּ ‫ ֵל ָראֹות ֶא‬and Similar 65 tation cannot be explained as Byzantine- or medieval-period intervention. Rather, it is firmly rooted in the Second Temple Period—when Hebrew was, crucially, still a living vernacular. That this is so is evidenced by the widespread agreement among the ancient translations and the consistent Samaritan nifʿal reading—to the point that the latter has unambiguous consonantal nifʿal infinitives (with heh) in Exod. 34.24 and Deut. 31.11, i.e., examples (2) and (3), respectively, above. The agreement between the Tiberian and Samaritan traditions on this point likely dates to a period before the two respective proto-traditions had diverged, i.e., no later than the second century BCE, and probably earlier. The DSS support for the Tiberian reading tradition’s hifʿil form at Deut. 1.33 in example (6) also comes as evidence of the antiquity of discomfort with qal readings potentially understandable as indications that God could be seen. What is more, from the perspective of the Tiberian consonantal tradition, in several cases, a nifʿal reading must be considered at least as felicitous as a qal reading, if not more so. This applies to the case of 1 Sam. 1.22 in example (15) above. It is also true of example (16). (16) ‫ל־פנֵ ֹׁ֖י הָ אָ ּ֥דן ׀ יְּׁ הוָ ִֽה׃‬ ְּׁ ֶ‫ָשֹלֵ֥ ש ְּפ ָע ִ ֶ֖מים ַב ָש ָנָׂ֑ה י ֵָראֶ ה ָכל־זְּ כ ְּּור ֵָ֔ך א‬ ‘Three times in the year will all your males appear to the ̊̄ Greek Lord, the LORD.’ (Exod. 23.17; SP ‫ יראה‬yirraʾi; ὀφθήσεται; Vulgate apparebit; TO ‫ ;יתחזון‬Syriac ‫)ܢܬܚܙܐ‬ Unless the preposition ‫ ֶאל‬here is due to corruption,8 it would seem to furnish consonantal support for an original nifʿal reading, 8 The collocation ‫ ָר ָאה ֶאל‬is uncommon, occurring only in Ezek. 43.3, where ‫)?( ַעל < ֶאל‬. In the MT the construction ‫ל־פנֵ י‬ ְּ ‫ ֶא‬often involves a 66 The Historical Depth of the Tiberian Reading Tradition since the ostensible qal ‫ל־פּנֵ י‬ ְ ‫*יִ ְר ֶאה ֶא‬, while perhaps not impossible, is far less expected than ‫ל־פּנֵ י‬ ְ ‫‘ יֵ ָר ֶאה ֶא‬appear to/before the face/in the presence of’. 4.0. Conclusion In sum, in the case of the expressions in question, the Tiberian biblical tradition presents several cases of probable mismatch between its written and reading components. In these cases, the vocalisation in the reading component almost certainly reflects the theologically motivated replacement of qal ‘seeing God’s face’ with nifʿal ‘appearing before God’. A few other morphological shifts may also be part of the same strategy. Though secondary, the ancient Hebrew and translational evidence substantiates the profound historical depth of the nifʿal interpretive tradition for ‘appearing before God’. This interpretation dates back to at least the Second Temple Period, as is clear from the unequivocal hifʿil spelling in a DSS version of Deut. 1.33 shown above in example (6). In other cases, the consonantal form is ambiguous. In any of them, the form may well represent an original qal; however, the apparently genuine nifʿal in Exod. 23.17 means that several may alternatively constitute genuine nifʿals. motion verb, e.g., Lev. 9.5; 14.53; 16.2; 17.8; Ezek. 44.4; Neh. 2.13; 2 Chron. 19.2. More comparable to the case in Exod. 23.17 are Lev. 6.7; Num. 20.10; Ezek. 41.4, 12, 15, 25; 42.2, 3, 7, 10, 10, 13; 45.7, 7; 48.21; Job 2.5; 13.15. The occurrence of ‫ ֵאת‬in SP Exod. 23.17 is unsurprising given that version’s harmonistic tendencies in the case of both content and grammar. 3. KETIV-QERE EUPHEMISMS A rather rare type of ketiv-qere mismatch involves the evidently euphemistic replacement of a written form deemed inappropriate for public reading with a more acceptable alternative (Ofer 2019, 98–99; see also Yeivin 1980, 56; Cohen 2007, 264–71). Words deemed impolite or vulgar may refer to objects, notions, or actions, often involving such ‘unmentionables’ as excreta, shameful infirmities, and rape, but can also extend to potential theological misunderstandings. Euphemistic ketiv-qere instances are mentioned explicitly in the Talmud (b. Megilla 25b): ‫ כגון ישגלנה‬,‫כל המקראות הכתובין בתורה לגנאי קורין אותן לשבח‬ ‫ לאכל את חוריהם ולשתות‬,‫ חריונים דביונים‬,‫ עפולים טחורים‬,‫ישכבנה‬ ‫ למחראות‬,‫את מימי שיניהם לאכול את צואתם ולשתות את מימי רגליהם‬ .‫למוצאות‬ All of the scriptures that are written in the Torah in impolite language are read in language beyond reproach, such as ‫‘ ישגלנה‬ravish her’ is read ‫‘ ישכבנה‬lie with her’ (Deut. 28.30); ‫‘ בעפלים‬with haemorrhoids’ is read ‫‘ בטחורים‬with tumours’ (Deut. 28.27); ‫‘ חריונים‬dove-droppings’ is read ‫( דביונים‬2 Kgs 6.25); ‫לאכול את חוריהם ולשתות את מימי שיניהם‬ ‘to eat their excrement and drink their urine’ is read ‫לאכול‬ ‫‘ את צואתם ולשתות את מימי רגליהם‬to eat their excrement and drink the water of their legs’ (2 Kgs 18.27); ‫‘ למחראות‬latrines’ is read ‫‘ למוצאות‬toilets’ (2 Kgs 10.27). (Ofer 2019, 98) © 2023 Aaron D. Hornkohl, CC BY-NC 4.0 https://doi.org/10.11647/OBP.0310.03 68 The Historical Depth of the Tiberian Reading Tradition 1.0. Euphemistic Ketiv-Qere Cases in the Tiberian Tradition and Other Ancient Witnesses 1.1. Excreta The terms written but not pronounced are ‫‘ *חרא‬faeces’, ‫*שין‬ ‘urine’, and ‫‘ *מחראה‬latrine’. They are replaced in the reading tradition with the respective synonyms ‫ ִדב‬or ‫צ ָֹאה‬, ‫ימי ַרגְּ ַליִ ם‬ ֵ ‫*מ‬, ֵ and ‫*מֹוצ ָאה‬. ָ In (1) the ketiv ‫ חרי‬is read aloud as qere ‫דב‬.ִ (1) ‫מֹור ִב ְּשמ ִֹנים ֵֶ֔כ ֶסף‬ ֙ ‫אש־ח‬ ֲ ֹ ‫ָׂ֑יה ַעד ֱהיַ֤ ֹות ר‬ ָ ‫וַ יְּ ִָֹ֨הי ָר ָ ַ֤עב גָ דֹול֙ ְּבש ֹ ְּמ ֵ֔רֹון וְּ ִה ֵנֶ֖ה ָצ ִרים ָע ֶל‬ ‫ה־כ ֶסף׃‬ ַֽ ָ ‫יֹונֹׁ֖ים ַב ֲח ִמ ָש‬ ִ )Q( ‫) ִדב‬K( ‫וְּ ֛ר ֹ ַבע ַה ַ ֵ֥קב חרי‬ ‘And there was a great famine in Samaria, as they besieged it, until a donkey’s head was sold for eighty shekels of sil- ver, and the fourth part of a kab of dove’s dung for five shekels of silver.’ (A 2 Kgs 6.25; Greek κόπρου περιστερῶν; Vulgate stercoris columbarum; TJ ‫ ;זיבל מפקת יוניא‬Syriac ‫ܚܪܝ‬ ̈ ‫)ܝܘܢܐ‬ Cohen (2007, 265) observes a difference between L and A regarding this ketiv-qere. In L, it applies to the entire graphic string ‫חרייונים‬, read as ‫יֹונֶ֖ים‬ ִ ‫ ִ;ד ְּב‬no space separates the two words in either the internal text or the marginal note and a shewa is written beneath the ‫ ר‬in the internal text, i.e., ‫ח ְּריי ֹו֯ ִנֶ֖ים‬. ִ 1 By contrast, in A the ketiv-qere is restricted to the elements ‫ חרי‬and ‫ ִ;דב‬a space separates the words ‫ חרי‬and ‫ יונים‬in the internal text, the marginal note 1 Indeed, the singular ‫ דביון‬is found in the Hebrew of Saadia Gaon (see ‫ ויקרא‬,‫יוצרות לשבתות השנה‬, ln. 19, accessed via the Maʾagarim website of the Academy of the Hebrew Language). 3. Ketiv-Qere Euphemisms 69 has only ‫דב‬, and no shewa is written beneath the ‫ ר‬in the internal text, i.e., ‫יֹונֶ֖ים‬ ִ ֯‫חרי‬. ִ A’s testimony is preferable, with ‫ ִדב‬the substi- tute for ‫חרי‬, and ‫יֹונֶ֖ים‬ ִ ‘doves’ serving as the nomen rectum in a construct formation. The lexeme ‫*דב‬ ִ is a hapax legomenon in BH. It is thought to be an Aramaism or dialectal form related to Hebrew ‫‘ זָ ב‬flow’ (Cohen 2007, 266, cites Rashi and Qimḥi). Since the ketiv and qere forms are synonyms, the testimony of the ancient versions is rather opaque with regard to the identity of the term being translated, i.e., the ketiv or the qere, though TJ’s explanatory gloss is reminiscent of the qere’s circumlocution. Examples (2)–(5) deal with parallel verses that include both ‫ *חר(א)ים‬and ‫*שינים‬.2 According to the qere, they are to be read aloud, respectively, as ‫צֹואה‬ ָ and ‫ימי ַרגְּ ַליִ ם‬ ֵ ‫*מ‬. ֵ 3 2 HALOT (1479) notes that the two terms also occur together in Ugaritic. Intriguingly, the written and reading components of the Tiberian tradi- tion consistently agree on a verbal form related to ketiv ‫‘ *שין‬urine’. Six occurrences of the word ‫‘ ַמ ְּש ִתין‬urinator’ (1 Sam. 25.22, 34; 1 Kgs 14.10; 16.11; 21.21; 2 Kgs 9.8) come in BH. Thought to be a Gt-stem participle (BDB 1010; HALOT 1479), the form was reanalysed as a hifʿil of ‫שת"ן‬, from which the noun ‫‘ ֶש ֶתן‬urine’, first attested in Talmudic Hebrew (t. Bekhorot 7.5 [44b]), was secondarily derived. 3 Cf. the development in select English translations of 2 Kgs 18.27, which testify to the shifting acceptability of English terms for excreta: ‘…toordis… pisse’ (Wycliffe, 1380s) ‘…donge… stale’ (Coverdale, 1535) ‘…doung… pisse’ (KJV, 1611) ‘…vilest excretions’ (Webster’s KJV Revision, 1833) ‘…dung… urine’ (RSV, 1946) ‘…excrement… urine’ (NIV, 1978) 70 (2) The Historical Depth of the Tiberian Reading Tradition ‫ת־ה ְּד ָב ִרים‬ ַ ‫ב־ש ֵ ֵ֗קה ַה ָֹ֨ ַעל ֲאד ֶֹנַ֤יָך וְּ ֵא ֶ ֙ל ֙יָך ְּש ָל ַחנִ י ֲאד ֵ֔ ִֹני ְּל ַד ֵ ֶ֖בר ֶא‬ ָ ‫יהם ַר‬ ֶֶ֜ ‫אמר ֲא ֵל‬ ֶ ֹ ‫וַ ָֹ֨י‬ )K( ‫ל־הח ֵָֹ֔מה ֶל ֱאכֹל ֶאת חריהם‬ ַ ‫ים ַע‬ ֙ ‫ל־ה ֲאנָ ִֵ֗שים ַה ַֹֽי ְּש ִב‬ ָ ‫ָה ֵ ָׂ֑א ֶלה ֲהל ֹא ַע‬ ...)Q( ‫צֹואָ ֶ֗ ָתם‬ ‘But the Rabshakeh said to them, “Has my master sent me to speak these words to your master and to you, and not to the men sitting on the wall, who are doomed to eat their own dung…”’ (2 Kgs 18.27a; Greek κόπρον αὐτῶν; Vulgate stercora sua; TJ ‫ ;מפקתהון‬Syriac ‫)ܬܒܬܗܘܢ‬ (3) ‫) ִע ָמ ֶ ַֽכם׃‬Q( ‫יהֹׁ֖ם‬ ֶ ֵ‫) ֵ ִֽמימֵ י ַּרגְּׁ ל‬K( ‫וְּ ִל ְּש ֛תֹות ֶאת־שיניהם‬... ‘“…and to drink their own urine with you?”’. (2 Kgs 18.27b; Greek οὖρον αὐτῶν; Vulgate urinam suam; TJ ‫מימי‬ ݂ܵ ‫ ;רגליהון‬Syriac ‫)ܬܝܢܝܗܘܢ‬ (4) ‫ת־ה ְּד ָב ִרים ָה ֵ ָׂ֑א ֶלה‬ ַ ‫ב־ש ֵ ֵ֗קה ַה ֶָֹ֨אל ֲאד ֶֹנַ֤יָך וְּ ֵא ֶ ֙ל ֙יָך ְּש ָל ַחנִ י ֲאד ֵ֔ ִֹני ְּל ַד ֵ ֶ֖בר ֶא‬ ָ ‫אמר ַר‬ ֶ ֹ ‫וַ י‬ ‫) צֹואָ ֶ֗ ָתם‬K( ‫חֹומה ֶל ֱאכֹל ֶאת־חראיהם‬ ֵָ֔ ‫ל־ה‬ ַ ‫ים ַע‬ ֙ ‫ל־ה ֲאנָ ִֵ֗שים ַה ַֹֽי ְּש ִב‬ ָ ‫ֲהל ֹא ַע‬ ...)Q( ‘But the Rabshakeh said, “Has my master sent me to speak these words to your master and to you, and not to the men sitting on the wall, who are doomed to eat their own dung…”’ (Isa. 36.12a; 1QIsaa 29.19 ‫ ;חריהמה‬Greek κόπρον; Vulgate stercora sua; TJ ‫ ;מפקתהון‬Syriac ‫)ܬܒܬܗܘܢ‬ (5) ‫) ִע ָמ ֶ ַֽכם׃‬Q( ‫יהֹׁ֖ם‬ ֶ ֵ‫ימּ֥י ַּרגְּׁ ל‬ ֵ ֵ‫) מ‬K( ‫וְּ ִל ְּש ֛תֹות ֶאת־שיניהם‬... ‘“…and drink their own urine with you?”’ (Isa. 36.12b: 1QIsaa 29.19 ‫ ;שיניהמה‬Greek οὖρον; Vulgate urinam pedum ݂ܵ suorum; TJ ‫ ;מימי רגליהון‬Syriac ‫)ܬܝܢܝܗܘܢ‬ The qere lexeme ‫ צ ָֹאה‬is variously analysed as reflecting the roots ‫‘ צו"א‬be foul’, ‫‘ יצ"א‬exit’, and ‫‘ וצ"א‬pollute’ (BDB 844a; HALOT 992a). Beyond the qere usages under discussion, it is attested in BH at Isa. 4.4; 28.9; and Prov. 30.12, where it possibly has the 3. Ketiv-Qere Euphemisms 71 more general sense of ‘filth’. It may be related to the lexeme ‫ֵצ ָאה‬ ‘excrement’ (Deut. 23.14; Ezek. 4.12). The lexeme ‫ צ ָֹאה‬is common in RH for reference to ‘excrement’.4 Among the ancient versions, TJ’s rendering might be evidence of an etymological translation of the qere, but this is not the only explanation. It is significant that 1QIsaa explicitly agrees with the ketiv. The qere parallel for ‫ שיניהם‬in ‫יהם‬ ֶ֖ ֶ ‫ימי ַרגְּ ֵל‬ ֵ֥ ֵ ‫ ֵמ‬is not found else- where in BH. It is a common term for ‘urine’ in RH.5 1QIsaa reflects the ketiv, TJ the qere. The Latin rendering in Isa. 36.12b, urinam pedum suorum ‘urine of their feet’, is noteworthy because it seems to reflect a conflation of the respective ketiv and qere traditions, ‘their urine’ and ‘water of their feet’. Other ancient renderings furnish arguably ambiguous evidence of the term being translated. In (6), the ketiv noun ‫מחראות‬, denoting a place for defecation is read as ‫מֹוצאֹות‬, ָ apparently representing a common nominal pattern of the ‫ יצ"א‬root. (6) )K( ‫ת־בית ַה ֵַ֔ב ַעל וַ יְּ ִש ֻ ֵ֥מהּו למחראות‬ ֵ ‫צּו ֶא‬ ֙ ‫ַוֵַּֽ֣יִ ְּת ֵ֔צּו ֵ ֶ֖את ַמ ְּצ ַבת ַה ָ ָׂ֑ב ַעל ַוֵַּֽ֣יִ ְּת‬ ‫ד־היַֽ ֹום׃‬ ַ ‫) ַע‬Q( ‫ְּׁל ִֽמֹוצָ אֹׁ֖ ֹות‬ ‘And they demolished the pillar of Baal, and demolished the house of Baal, and made it into a latrine to this day.’ (2 Kgs 10.27; Greek: εἰς λυτρῶνας; Vulgate: latrinas; TJ: ‫לבית‬ ‫ ;מפקת אנש‬Syriac: ‫)ܒܝܬ ܡܚܪܝܐ‬ 4 E.g., m. Berakhot 3.5; Shabbat 16.7; ʿAvoda Zara 4.5; ʾAvot 3.3; Ḥullin 3.5; Kelim 10.2; Miqwaʾot. 9.2, 4; Makhshirin 5.6. 5 E.g., m. Shabbat 9.5; Bava Batra 2.1; ʿEduyot 5.1, 4; Kelim 1.3; Toho- rot 4.5; Miqwaʾot. 10.6; Nidda 4.3; 9.6, 7; Makhshirin. 6.5, 6; Zavim 5.7. 72 The Historical Depth of the Tiberian Reading Tradition The MS form ‫מֹוצא‬ ָ ‘place/time of going out, utterance, source’ is common in the Bible, while the FS ‫*מֹוצ ָאה‬ ָ occurs only here and in Mic. 5.1, where it may mean ‘origins’ or ‘goings out = activities’. The form in 2 Kgs 10.27 is possibly a homonym that derives from ‫‘ צו"א‬be foul’ or ‫‘ וצ"א‬pollute’ (see above). The lexical tradition reflected in the ancient versions is not sufficiently clear to identify the source word—though, again, TJ’s circumlocution ‫בית‬ ‫ מפקת אנש‬looks to be a calque of the qere—on the assumption that ‫*מֹוצ ָאה‬ ָ here means, or was understood to mean, ‘place of excre- tion’ or ‘outhouse’. 1.2. Shameful Infirmities Six times in the Tiberian tradition, the ketiv plural ‫ עפלים‬is replaced by the qere ‫חֹורים‬ ִ ‫ט‬. ְּ These are given in (7)–(12). (7) ‫ּוב ָ ָׂ֑ח ֶרס‬ ֶ ‫ּובגָ ָ ֶ֖רב‬ ַ )Q( ‫) ּובַּ ְּׁטח ִ ִ֔רים‬K( ‫הוה ִב ְּש ִ ַ֤חין ִמ ְּצ ַ ֙ריִ ֙ם ובעפלים‬ ֶ֜ ָ ְּ‫יַ ְּכ ָָֹ֨כה י‬ ‫א־תּוכל ְּל ֵה ָר ֵ ַֽפא׃‬ ֶ֖ ַ ֹ ‫ֲא ֶ ֵ֥שר ל‬ ‘The LORD will strike you with the boils of Egypt, and with tumours and scabs and itch, of which you cannot be ̊̄ lǝm; ̊̄ healed.’ (Deut. 28.27; SP ‫ ובעפלים‬wbafa Greek ἐν ταῖς ἕδραις; Vulgate et parte corporis per quam stercora digeruntur; TO ‫ ;ובטחורין‬Syriac ‫)ܘܒܛܚܘܪܐ‬ (8) ‫) בַּ ְּׁטח ִ ִ֔רים‬K( ‫דֹודים וַ יְּ ִש ֵ ָׂ֑מם וַ יַ ְֵַּ֤֣ך א ָֹת ֙ם בעפלים‬ ֶ֖ ִ ‫ל־ה ַא ְּש‬ ָ ‫הו֛ה ֶא‬ ָ ְּ‫וַ ִת ְּכ ַ ָ֧בד יַ ד־י‬ ‫יה׃‬ ָ ‫בּול‬ ַֽ ֶ ְּ‫ת־א ְּש ֶ֖דֹוד וְּ ֶאת־ג‬ ַ ‫) ֶא‬Q( ‘The hand of the LORD was heavy against the people of Ashdod, and he terrified and afflicted them with tumours, both Ashdod and its territory.’ (1 Sam. 5.6; Greek εἰς τὰς ναῦς; Latin in secretiori parte natium; TJ ‫ ;בטחורין‬Syriac ̈ ‫)ܒܚܘܛܪܝܗܘܢ‬ 3. Ketiv-Qere Euphemisms (9) 73 ‫אד וַ יַ ְ֙ך ֶאת־‬ ֹ ֵ֔ ‫דֹולה ְּמ‬ ָ ְּ‫הּומ ֙ה ג‬ ָ ‫יר ְּמ‬ ֙ ‫הוַ֤ה ׀ ָב ִע‬ ָ ְּ‫וַ יְּ ִָ֞הי ַא ֲח ֵרי ׀ ֵה ַסבּו א ֵֹ֗תֹו וַ ְּת ִָֹ֨הי יַ ד־י‬ ‫)׃‬Q( ‫) ְּׁטח ִ ִֽרים‬K( ‫ַאנְּ ֵשי ָה ֵ֔ ִעיר ִמ ָק ֶ֖טֹן וְּ ַעד־גָ ָׂ֑דֹול וַ יִ ָש ְּת ֵ֥רּו ָל ֶ ֶ֖הם עפלים‬ ‘But after they had brought it around, the hand of the LORD was against the city, causing a very great panic, and he afflicted the men of the city, both young and old, so that tumours broke out on them.’ (1 Sam. 5.9; 4Q51 5b–c.6: ‫ ;ב]עפלים‬Greek ἕδρας; Latin extales; TJ ‫ ;בטחוריא‬Syriac ̈ ݁ ‫)ܛܚܘܪܝܗܘ‬ ‫ܢ‬ (10) ‫) וַ ַ ֛ת ַעל ַ ַֽשוְּ ַ ֵ֥עת‬Q( ‫) בַּ ְּׁטח ִ ָ֑רים‬K( ‫א־מתּו ֻה ֶ֖כּו בעפלים‬ ֵֵ֔ ֹ ‫ים ֲא ֶשר ל‬ ֙ ‫וְּ ָ ַֽה ֲאנָ ִש‬ ‫ָה ִ ֶ֖עיר ַה ָש ָ ַֽמיִ ם׃‬ ‘The men who did not die were struck with tumours, and the cry of the city went up to heaven.’ (1 Sam. 5.12; Greek εἰς τὰς ἕδρας; Latin in secretiori parte natium; TJ ‫ ;בטחוריא‬Syr̈ ݁ ‫)ܒܛܚܘܪܝܗܘ‬ iac ‫ܢ‬ (11) ‫אמ ֵ֗רּו ִמ ְּס ַפ ֙ר ַס ְּר ֵני ְּפ ִל ְּש ִֵ֔תים ֲח ִמ ָש ֙ה‬ ְּ ֹ ‫אמ ֵ֗רּו ָמה ָה ָא ָש ֮ם ֲא ֶשר נָ ִשיב לֹו֒ וַ י‬ ְּ ֹ ‫וַ י‬ ‫י־מגֵ ָ ֵ֥פה ַא ַ ֛חת ְּל ֻכ ָלֶ֖ם‬ ַ ‫) זָ ֵָ֔הב וַ ֲח ִמ ָ ֶ֖שה ַע ְּכ ְּב ֵרי זָ ָ ָׂ֑הב ִ ַֽכ‬Q( ‫) ְּׁטח ֵרּ֣י‬K( ‫עפלי‬ ‫יכם׃‬ ַֽ ֶ ֵ‫ּול ַס ְּרנ‬ ְּ ‘And they said, “What is the guilt offering that we shall re- turn to him?” They answered, “Five golden tumours and five golden mice, according to the number of the lords of the Philistines, for the same plague was on all of you and on your lords.’ (1 Sam. 6.4; 4Q51 6a–b.13 ‫;עפלי‬ ֯ Greek ἕδρας; Latin —; TJ ‫חֹורי‬ ֵ ‫;ט‬ ְּ Syriac ‫)ܛܚܘܪܝܢ‬ 74 The Historical Depth of the Tiberian Reading Tradition ִ ‫יכם ַה ַמ ְּש ִח‬ ֵֶ֗ ‫) וְּ ַצ ְּל ֵמי ַע ְּכ ְּב ֵר‬Q( ‫יכם‬ ֶ ָ֜ ‫) ְּׁטח ֵר‬K( ‫ית ֩ם ַצ ְּל ֵָֹ֨מי עפליכם‬ ֶ ‫וַ ֲע ִש‬ (12) ‫ית ֙ם‬ ‫ּומ ַ ֵ֥על‬ ֵ ‫יכם‬ ֵֶ֔ ‫ת־יָדֹו ֵ ַֽמ ֲע ֵל‬ ֙ ‫אּולי יָ ֵ ַָ֤קל ֶא‬ ַ ֵ֗ ‫אֹלהי יִ ְּש ָר ֵ ֶ֖אל ָכ ָׂ֑בֹוד‬ ֵ֥ ֵ ‫ת־ה ֵָ֔א ֶרץ ּונְּ ַת ֶ ֛תם ֵל‬ ָ ‫ֶא‬ ‫ּומ ַ ֵ֥על ַא ְּר ְּצ ֶ ַֽכם׃‬ ֵ ‫ֹלה ֶיכֶ֖ם‬ ֵ ‫ֱא‬ ‘So you must make images of your tumours and images of your mice that ravage the land, and give glory to the God of Israel. Perhaps he will lighten his hand from off you and your gods and your land.’ (1 Sam. 6.5; 4Q51 6a–b.14: ‫ ;העפ] ֯ל[י]ם‬Greek: —; Latin: anos; Targum: ‫חֹוריכֹון‬ ֵ ‫;ט‬ ְּ Syriac: ̈ ‫)ܛܚܘܪܝܟܘܢ‬ The matter is complicated by apparent textual divergence in the Samuel narrative, as well as by a lack of semantic certainty regarding the meaning of the ketiv and qere terms. Suffice it to say that, whatever its meaning, ketiv ‫‘ עפלים‬tumours, haemorrhoids’ was deemed inappropriate for public reading and was replaced in the reading tradition with qere ‫חֹורים‬ ִ ‫‘ ְּט‬tumours, haemorrhoids’. As is their wont, TO and TJ agree with the qere. Where extant, 4QSama (5Q51) preserves the ketiv. Whether the ketiv, qere, or another reading lies behind the other ancient witnesses cannot be determined with anything approaching certainty. Interestingly, the qere ‫חֹורים‬ ִ ‫ ְּט‬is shared by the written and reading components of the Tiberian tradition in two instances in the Samuel narrative; see (13) and (14), neither paralleled in DSS Samuel material and one without a parallel in the Greek. ָ ‫הוֶ֖ה ֶא‬ ָ ְּ‫ת־א ֵ֥רֹון י‬ ֲ ‫וַ יָ ִ ֛שמּו ֶא‬ (13) ‫ל־ה ֲעגָ ָלָׂ֑ה וְּ ֵאת ָה ַא ְּר ֵ֗ ַגז וְּ ֵא ֙ת ַע ְּכ ְּב ֵרי ַהזָ ֵָ֔הב וְּ ֵ ֶ֖את ַצ ְּל ֵ ֵ֥מי‬ ‫יהם׃‬ ִֽ ֶ ‫ְּׁטח ֵר‬ ‘And they put the ark of the LORD on the cart and the box with the golden mice and the images of their tumours.’ (1 Sam. 6.11; Greek —; Latin anorum; TJ ‫ ;טחוריהון‬Syriac ̈ ‫)ܛܚܘܪܝܗܘܢ‬ 3. Ketiv-Qere Euphemisms 75 (14) ‫יהוָׂ֑ה ְּל ַא ְּש ָֹ֨דֹוד ֶא ֶָ֜חד ְּל ַע ָזַ֤ה‬ ָ ‫וְּ ֵ֙א ֶל ֙ה ְּׁטח ֵרּ֣י ַהזָ ֵָ֔הב ֲא ֶָֹ֨שר ֵה ִ ָ֧שיבּו ְּפ ִל ְּש ִ ֛תים ָא ָ ֶ֖שם ַ ַֽל‬ ‫ֶא ָח ֙ד ְּל ַא ְּש ְּקלֹון ֶא ֵָ֔חד ְּל ַגֵ֥ת ֶא ָ ֶ֖חד ְּל ֶע ְּק ֵ֥רֹון ֶא ָ ַֽחד׃‬ ‘These are the golden tumours that the Philistines returned as a guilt offering to the LORD: one for Ashdod, one for Gaza, one for Ashkelon, one for Gath, one for Ekron.’ (1 ̈ Sam. 6.17; Greek ἕδραι; Latin ani; TJ ‫ ;טחורי‬Syriac ‫)ܛܚܘܪܐ‬ 1.3. Rape Four times in the Tiberian biblical tradition, the ketiv has a verb with the root ‫ שג"ל‬in a context of wartime rape. On all occasions the qere calls for a verb with root ‫‘ שכ"ב‬lie (down)’. (15) ‫א־ת ֵשב‬ ֵ ֹ ‫) ַ ֵ֥ביִ ת ִת ְּב ֶנֶ֖ה וְּ ל‬Q( ‫) יִ ְּׁשכָ ִ֔ ֶבנָה‬K( ‫ִא ָשה ְּת ָא ֵ ֵ֗רש וְּ ִ ַ֤איש ַא ֵח ֙ר ישגלנה‬ ‫ָׂ֑בֹו ֶ ֵ֥כ ֶרם ִת ַ ֶ֖טּע וְּ ֵ֥ל ֹא ְּת ַח ְּל ֶ ַֽלנּו׃‬ ‘You shall betroth a wife, but another man shall ravish her. You shall build a house, but you shall not dwell in it. You shall plant a vineyard, but you shall not enjoy its fruit.’ (Deut. 28.30; 4Q30 f50.3 ‫ ;[ישג] ֯לנ֯ ה‬SP ‫ ישכב עמה‬yiškåb imma; Greek ἕξει αὐτήν; Latin: dormiat cum ea; TO ‫ ;ישכבינה‬Syriac ‫)ܢܣܒܝܗ‬ (16) ‫) ִתשָ ַּ ִֽכ ְּׁבנָה‬K( ‫יהם תשגלנה‬ ֶ֖ ֶ ‫יהם ּונְּ ֵש‬ ֵֶ֔ ‫סּו ָ ַֽב ֵת‬ ֙ ‫יהם יִ ַ֙ש‬ ָׂ֑ ֶ ֵ‫יהם יְּ ֻר ְּטּ ֶ֖שּו ְּל ֵ ַֽעינ‬ ֵ֥ ֶ ‫וְּ ע ְֹּל ֵל‬ ‫)׃‬Q( ‘Their infants will be dashed in pieces before their eyes; their houses will be plundered and their wives ravished.’ (Isa. 13.16; 1QIsaa 11.24 ‫;ת ֯ש ֯כבנה‬ ̇ 1Q8 6a–b.2 ]‫;תש[◦◦נה‬ ֯ 4Q55 f8.13 ]‫ ;תשג] ֯ל[נה‬Greek ἕξουσι; Latin violabuntur; TJ ̈ ݁ ‫ ;ישתכבן‬Syriac ‫)ܢܨܛܥܪܢ‬ 76 The Historical Depth of the Tiberian Reading Tradition (17) ‫ים‬ ֙ ‫ל־ד ָר ִכ‬ ְּ ‫) ַע‬Q( ‫כ ְּׁב ְּׁת‬ ַּ ִ֔ ֻׁ‫) ש‬K( ‫ל־ש ָפ ִֶ֜ים ְּּור ִֵ֗אי ֵאיפ ֹ֙ה ל ֹא שגלת‬ ְּ ‫י־ע ָֹ֨ ַיניִ ְך ַע‬ ֵ ‫ְּש ִ ַֽא‬ ‫ּוב ָר ָע ֵ ַֽתְך׃‬ ְּ ‫נּותיִ ְך‬ ֶ֖ ַ ְּ‫יפי ֵֶ֔א ֶרץ ִבז‬ ִ ‫יָ ַש ְּב ְּת ָל ֵֶ֔הם ַכ ֲע ָר ִ ֶ֖בי ַב ִמ ְּד ָ ָׂ֑בר וַ ַת ֲח ִנ‬ ‘Lift up your eyes to the bare heights, and see! Where have you not been ravished? By the waysides you have sat awaiting lovers like an Arab in the wilderness. You have polluted the land with your vile whoredom.’ (Jer. 3.2; Greek ἐξεφύρθης; Latin prostrata sis; TJ ‫אתחברת‬ ݁ ‫ ;ליך למפלח לטעותא‬Syriac ‫)ܐܬܛܢܦܬܝ‬ (18) ‫סּו ַה ָב ִֵ֔תים‬ ֙ ‫ה וְּ נִ ְּל ְּכ ָדה ָה ֵ֗ ִעיר וְּ נָ ַ֙ש‬ ֒ ‫רּוש ַל ֮ם ַל ִמ ְּל ָח ָמ‬ ָ ְּ‫ל־הגֹויִ ֵ֥ם ׀ ֶ ַֽאל־י‬ ַ ‫ת־כ‬ ָ ‫וְּ ָא ַס ְּפ ִָֹ֨תי ֶא‬ ‫גֹולה וְּ ֶי ֶתר ָה ֵ֔ ָעם‬ ָ ֵ֔ ‫יר ַב‬ ֙ ‫) וְּ יָ ָָ֞צא ֲח ִ ַ֤צי ָה ִע‬Q( ‫) ִתשָ ַּכ ְָּׁ֑בנָה‬K( ‫וְּ ַהנָ ִ ֶ֖שים תשגלנה‬ ‫ן־ה ִ ַֽעיר׃‬ ָ ‫ֵ֥ל ֹא יִ ָכ ֵ ֶ֖רת ִמ‬ ‘For I will gather all the nations against Jerusalem to battle, and the city shall be taken and the houses shall be plundered and the women shall be raped. Half of the city shall go out into exile, but the rest of the people shall not be cut off from the city.’ (Zech. 14.2; Greek μολυνθήσονται; Latin ̈ violabuntur; TJ ‫ ;ישתכבן‬Syriac ‫)ܢܨܛܥܪܢ‬ The euphemistic employment of ‫‘ ָש ַכב‬lie (down)’ in reference to sexual relations is common throughout BH (and is matched by euphemistic renderings in the ancient versions). This usage was also extended to cases of ketiv ‫‘ שג"ל‬rape’. The change could not be effected, however, without certain grammatical modifications. First, in reference to sex, ‫ ָש ַכב‬normally takes one of the comitative prepositions ‫ ִעם‬or ‫ ֵאת‬both ‘with’ (Orlinsky 1944). On seven occasions one encounters ‫ ָש ַכב‬with a form of -‫—אֹת‬apparently the definite accusative/direct object marker—but in six of the seven, the vocalisation alone indicates that the particle is not 3. Ketiv-Qere Euphemisms 77 the preposition ‫‘ ֵאת‬with’.6 Second, the verb ‫ ָש ַכב‬nowhere in BH bears an object suffix except where it is read as the qere for presumably qal ketiv ‫שגל‬, as in (15) above. Third, BH lacks a nifʿal ‫ נִ ְּש ַכב‬except where it is read instead of apparently nifʿal ‫*נשגל‬, in examples (16) and (18) above. Finally, and of crucial significance, unambiguous consonantal nifʿal ‫ *נשכב‬is first attested in material in the NBDSS7 and persists in RH. Relatedly, no passive qal or puʿʿal cognate of ‫ ָש ַכב‬is known from ancient Hebrew beyond that in the qere of (17) above.8 All of the above point to the likely secondary development of -‫ ָש ַכב אֹת‬, perhaps in the early Second Temple Period (cf. -‫ ָש ַכב אֹות‬with mater waw in Ezekiel) (Beuken 2004, 663). In other words, the expression -‫ ָש ַכב אֹת‬is itself a case of ketiv-qere mismatch unacknowledged within the Masoretic tradition and is in line with the ‫שג"ל‬-‫ שכ"ב‬correspondence under discussion. 1.4. Potential Misunderstanding Cohen (2007, 269–71) lists a final instance of euphemistic ketivqere, as seen in (19). (19) )Q( ‫) ְּׁבעֵ ִינ ָ֑י‬K( ‫הוֶ֖ה בעוני‬ ָ ְּ‫אּול֛י יִ ְּר ֶ ֵ֥אה י‬ ַ ְּ‫ו‬ ‘It may be that the LORD will look upon my eye…’ (2 Sam. 16.12; Greek ταπεινώσει μου; Latin adflictionem meam; TJ ‫ ;דמעת עיני‬Syriac ‫)ܒܫܘܥܒܕܝ‬ 6 -‫אֹת‬: Gen. 34.2; Lev. 15.18, 24; Num. 5.13, 19; 2 Sam. 13.14; -‫אֹות‬: Ezek. 23.8. 7 4Q270 f5.19; 4Q271 f3.12. 8 Ancient Hebrew attests no piʿʿel. 78 The Historical Depth of the Tiberian Reading Tradition The ketiv is doubly problematic, in that the context calls for a word meaning ‘suffering, misfortune, plight’, whereas, on the one hand, ‫ ָעֹון‬presupposes an element of guilt not evident from the context and, on the other, it does not generally denote mere suffering. Some modern commentators assume that the text should reflect ‫‘ ֳענִ י‬suffering’ or ‫‘ ָענְּ יִ י‬my suffering’ (BDB 730b; cf. the Greek, Latin, and Syriac). Cohen (2007, 269–70, fn. 29) posits a semantic shift, whereby the meaning of ‫ ָעֹון‬developed from ‘sin, guilt’ through ‘punishment’ to ‘trouble, suffering, torment, an- guish’. Even if the proposed semantic shift is valid, the ketiv remains contextually difficult, given the standard force of ‫עֹון‬. ָ The qere ‫ ְּב ֵע ִינָׂ֑י‬can be taken either elliptically, for ‘tear of the eye’ (cf. the Targum), or metonymically, with ‘eye’ standing for the entire self (Cohen 2007, 270–71). 2.0. Diachronic Considerations Given the obvious euphemistic status of the qere forms discussed above, there seems no need to prove their secondary status. Even so, the regular apparent agreement of the ketiv with the DSS (where extant) and the ancient versions is evidence of the primacy of the ketiv tradition (though many of the individual renderings of the ancient translations leave room for doubt). Against the general agreement of the other ancient versions with the ketiv, the Targums regularly accord with the qere tradition. Sometimes, the qere and the Targums both resort to terms common in RH, as in the case of ‫ צואה‬and ‫מימי רגלים‬. The Vulgate’s urinam pedum also seems partially influenced by the rabbinic idiom. 3. Ketiv-Qere Euphemisms 79 However, it is also important to point out non- or pre-rabbinic evidence for qere forms. For example, the qere form ‫חֹורים‬ ִ ‫ְּט‬ used in place of ketiv ‫ עפלים‬is not restricted to the Tiberian reading tradition, but appears twice in the Tiberian consonantal tradition, as well. Also, qere ‫ שכ"ב‬for ketiv ‫ שג"ל‬finds support in the combined Samaritan biblical written and reading tradition, the BDSS, and is in line with both general biblical euphemistic use of ‫ ָש ַכב‬in relation to sex as well as with an apparently secondary usage according to which the verb came to be used transitively. This latter development, manifested in the verb’s use with the accusative/direct object particle, with object suffixes, and in the appearance of cognate qal internal passive or nifʿal verbs, is clearly one rooted in the Second Temple Period, its initial stages seen in the orthography of exilic or post-exilic biblical passages and DSS Hebrew. 3.0. Conclusion While the euphemistic qere alternatives for public reading are secondary and reflect relatively late sociolinguistic concerns, where clear evidence exists, it shows that the readings are in the main Second Temple developments, no later than Tannaitic Hebrew, and are sometimes validated by DSS and, albeit rarely, even Tiberian CBH written evidence. PART II: LINGUISTIC DEVELOPMENTS 4. THE PROPER NAME ISSACHAR In the case of the proper name Issachar, the relationship between orthography and phonetic realisation is famously anomalous.1 Put simply, the name’s pronunciation according to most biblical reading traditions is at odds with the dominant Hebrew spelling. The mismatch is blatant in the standard Tiberian qere perpetuum ‫שכר‬ ָ ‫יִשּׂ‬, ָ wherein readers are consistently instructed to ignore the ̊̄ r, ̊̄ 2 as form’s third consonant in favour of the articulation yiśśåḵå if the form were written ‫*יִשּׂ ָכר‬. ָ 3 The dissonance in question is evidently a result of secondary phonological development. It seems to be a case of gemina- 1 Early awareness of variation in the name’s pronunciation is evidenced in Mishaʾel ben ʿUzziʾel’s tenth- or eleventh-century Judaeo-Arabic Kitāb Al-Khilaf ‘Book of Differences’ (Hebrew Sefer ha-Ḥillufim), which focuses on points of dispute in the respective biblical reading traditions of the leading Masoretes Ben Asher and Ben Naftali. The first difference that Ben Uzziʾel cites is that of the name Issachar (see the edition by Lipschütz 1964; 1965). 2 For ancient realisations of ‫ ש‬ś, especially its Second Temple phonetic identity with ‫ ס‬s, see Khan (2020, I:62–65, fn. 59, 234–36). 3 The vocalisation of ‫שכר‬ ָ ‫יִשּׂ‬ ָ is consistent in the extant cases in the A. In about one-third of the cases in L (14 of 43), the dagesh is missing from the ‫ש‬: Gen. 46.13; Num. 10.15; 34.26; 1 Kgs 4.17; Ezek. 48.25–26; 1 Chron. 2.1; 6.47, 57; 7.1, 5; 12.41; 26.5; 27.18. A rafe is marked over the second ‫ ש‬once in L (Exod. 35.23), never in the extant portions of A. Yeivin (1985, 1090) lists several graphic representations of the name’s vocalisation in the Babylonian tradition, all of which correspond to the accepted Tiberian convention. © 2023 Aaron D. Hornkohl, CC BY-NC 4.0 https://doi.org/10.11647/OBP.0310.04 84 The Historical Depth of the Tiberian Reading Tradition tion due to anticipatory assimilation of the first of two originally distinct sibilants—likely śś < šś—possibly reflecting the contrac- tion of an earlier compound, such as ‫‘ *יֵ ש ָׂש ָכר‬there is a wage’ or ‫*איש ָׂש ָכר‬ ִ ‘man of wage’.4 Aharon ben Moshe ben Asher’s repre̊̄ r̊̄ by means sentation of the standard Tiberian realisation yiśśåḵå of the accepted (but highly irregular) consonant-diacritic combination ‫שכר‬ ָ ‫יִשּׂ‬, ָ was not the only possibility. Another early Masorete, Moshe Moḥe, opted for the alternative graphic representation ‫( יִ ְׂש ָׂש ָכר‬see the image on the front cover), which in Tiberian Hebrew would have had the same phonetic value as Ben ̊̄ r̊̄ (see below on the alternative Tiberian reAsher’s ‫שכר‬ ָ ‫יִשּׂ‬ ָ yiśśåḵå alisation given by Ben Naftali).5 4 See BDB sub. voc. and HALOT sub. voc. for these and other sugges- tions. Ancient interpretations can be found in Jerome’s commentaries: unde et issachar, qui interpretatur: est merces, ex uirtutibus nomen accepit ‘Whence is also Issachar, which is interpreted: there is a wage, has taken the name from manliness’; unde et issachar interpretatur: est merces; et sachar μέθυσμα, id est ebrietas, ceteri que ebrios; soli lxx mercenarios transtulerunt ‘Whence is also Issachar interpreted: there is a wage; and sachar as μέθυσμα, that is, intoxication, others also as drunken ones; only the Seventy have translated it as those hired for wages’ (on Isaiah, lib. 6, 14.24–25; Migne 1844–1855, XXIV, col. 227); et de issachar legimus, quod supposuerit humerum suum ad laborandum, et uir agricola sit ‘And from/about Issachar we read, that he placed his upper arm to work, and was a farmer/agricultural man’ (on Hosea, lib. 2, 6.9; Migne 1844– 1855, XXV, col. 871); issachar enim interpretatur merces ut significetur pretium proditoris ‘For Issachar is interpreted as wage so as to signify the price of a traitor’ (on Matthew, lib. 1, 10.4; Migne 1844–1855, XXVI, col. 63). 5 In the Tiberian pronunciation tradition, shewa on the second of two identical consonants was silent after a short vowel, e.g., ‫ ִהנְּ נִ י‬hinnī (Khan 4. The Proper Name Issachar 85 The question that the present study seeks to answer involves the antiquity of the dissonance between the Tiberian written and reading traditions, specifically, how far back the reading tradition reflected by the Masoretic vocalisation signs reaches. 1.0. Diversity in Antiquity 1.1. Double-sibilant Realisations The first thing to note is that, while converging lines of evidence point to the early emergence of a phonetic realisation similar to what was to become standard in the Tiberian tradition, there are also traditions reminiscent of the Tiberian orthography, i.e., that reflect the pronunciation of two distinct sibilants. 1.1.1. The Samaritan Tradition For example, though the Samaritan Hebrew consonantal spelling is identical to that of Masoretic Hebrew, the Samaritan phonetic ̊̄ kår. ̊̄ realisation is yåšīšå As Samaritan Hebrew preserves just one phoneme represented by the grapheme ‫ש‬, namely š, the quality of the sibilant is unsurprising. The Samaritan realisation of a vowel between the two sibilants is, however, unique among pronunciation traditions. The vowel in question not improbably developed from an earlier shewa, as Samaritan Hebrew routinely parallels Tiberian shewa with a full vowel, long in open syllables 2013, 100; 2020, I:352–53; cf. Ofer 2018, 196). The Maʾagarim website of the Academy of the Hebrew Language’s Historical Dictionary Project lists a number of variant spellings in agreement with the standard received pronunciation. 86 The Historical Depth of the Tiberian Reading Tradition (Ben-Ḥayyim 2000, 53–55). If so, this seems to have been an alternative to the gemination due to assimilation known from other traditions, one that allowed for the preservation of the distinct realisation of once-adjacent sibilants. 1.1.2. The Tiberian Tradition according to Ben Naftali Possible evidence of a pre-assimilation realisation may also be reflected in the alternative Tiberian vocalisation advocated by ̊̄ r, ̊̄ which shows the sequence Ben Naftali, namely, ‫ יִ ְש ָׂש ָכר‬yišśåḵå of two distinct sibilants, i.e., šś (Khan 2020, I:94). Such a realisation might be characterised as purist and/or etymological, possi̊̄ r̊̄ bly an attempt to combat the perceived ‘lax’ or ‘slurred’ yiśśåḵå pronunciation recorded by Ben Asher and eventually accepted as the standard. Khan (2020, I:103), however, emphasises the possibility that the realisation advocated by Ben Naphtali rather represents an innovative attempt at orthoepy, and that it may not preserve a genuinely archaic pronunciation. 1.1.3. The Temple Scroll (11QTa = 11Q19) Similar purist and/or etymological tendencies may also have factored in the writing of the name in the Temple Scroll from Qumran (11QTa = 11Q19). In all five of its occurrences in this manuscript (and nowhere else in the Dead Sea Scrolls), the name is written as two separate words, i.e., ‫יש שכר‬. These are reproduced 4. The Proper Name Issachar 87 in examples (1)–(4) (note that the final example includes two tokens).6 (1) ‫( יש שכר‬11Q19 24.15) (2) ‫( יש שכר‬11Q19 39.13) (3) [‫שכר‬ ֵ֗ ‫( יש‬11Q19 41.4) (4) ‫( יש שכר לבני יש שכר‬11Q19 44.16) This written representation may be an early example of orthoepic effort, that is, the attempt to promote correct enunciation, presumably in the face of the perceived threat of hurried and/or lax articulation.7 Alternatively—or, additionally—the word spacing possibly reflects exegetical concerns. Crucially, if the division of the name into distinct graphic words reflects a realisation like yiš sakar, it comes as indirect early evidence of alternative realiså̄ r. ̊̄ tions to Ben Asher’s Tiberian ‫שכר‬ ָ ‫יִשּׂ‬ ָ yiśśåḵå 6 The images in examples (1)–(4) are from the Temple Scroll, 11Q19, Qumran, late 1st century BCE–early 1st century CE, reproduced here with permission of the Shrine of the Book, The Israel Museum, Jerusalem. 7 On the orthoepic character of the Tiberian vocalisation system see Khan (2018b). 88 The Historical Depth of the Tiberian Reading Tradition 1.2. Single-sibilant Realisations 1.2.1. The Versions But additional direct early evidence is also available. First, in contrast to the double-sibilant realisations in Samaritan ̊̄ kår, ̊̄ ̊̄ r, ̊̄ and 11QT’s ‫יש שכר‬, other ancient yåšīšå Ben Naftali’s yišśåḵå traditions agree on forms of the name with a single sibilant sound. Thus, Greek has Ισσαχαρ, Latin Isachar, TA ‫יִשָשכָר‬, and Syriac ‫ܐܝܣܟܪ‬. Clearly, this evidence points to the relatively early emergence and diffusion of a realisation (or realisations) in which the presumably original sequence of discrete sibilants indicated by the dominant spelling ‫ יששכר‬and preserved in a minority of traditions (like Samaritan, Ben Naftali, and the Temple Scroll) was realised as a one sibilant, whether geminate or singleton. 1.2.2. 4Q522: Apocryphon of Joshua Second, and of more immediate relevance to the possibly orthoepic motivation for the Temple Scroll’s explicit representation of the name Issachar as two discrete graphic words, is the form of the name as presented in an apparent allusion to Josh. 17.11 found in the Apocryphon of Joshua (4Q522 f8.3), where the name is written ‫ישכר‬. See example (5). (5) ‫ ]וישכר את בית שן ואשר א[ת‬-- [ (4Q522 f8.3)8 8 Image used by permission of the Israel Antiquities Authority. 4. The Proper Name Issachar 89 Cf. the Masoretic version in example (6). ...‫יה‬ ָ ‫נֹות‬ ֶ ‫ית־ש ָאן ּו ְּּ֠ב‬ ְּ ‫ּוב ָא ֵֵ֗שר ֵב‬ ְּ ‫שכ ּ֣ר‬ ָ ָ‫וַ יְּ ִָֹ֨הי ִל ְּמנַ ֶֶ֜שה ְּׁביִ ש‬ (6) ‘And it was: to Manasseh were assigned within Issachar’s and Asher’s territories Beth Shean and its villages (Josh. 17.11) Under different circumstances, the spelling ‫ ישכר‬for Issachar— unique in the DSS—might be considered a mere scribal lapse, the accidental graphic omission of a repeated consonant with no phonetic import. However, given the aforementioned versional evidence, which demonstrates the existence in antiquity of singlesibilant realisations, the DSS ‫ ישכר‬orthography has the look of a phonetic spelling along the lines of [jisːakar] (< yiśśakar)—apparently confirming the antiquity of the type of phonetic realiså̄ r. ̊̄ tion also preserved in Tiberian ‫שכר‬ ָ ‫יִשּׂ‬ ָ yiśśåḵå To DSS ‫ ישכר‬one may add later spellings of this type, such as those that crop up in texts from the Cairo Geniza, where a minority of forms with single-sibilant spellings evidently reflect phonetic realisations. While the single-sibilant realisation (with or without gemination) became conventional in most Hebrew (and foreign) pronunciation traditions, the classical double-sibilant orthographic tradition was successfully preserved. In Jewish Hebrew and Aramaic pronunciation traditions, this led to mismatch, first, between the written and reading traditions and, eventually, between the consonants and vocalic diacritics that combine to make up the written Masoretic tradition. 90 The Historical Depth of the Tiberian Reading Tradition 2.0. Historical Considerations The historical depth of single-sibilant realisations is unclear. Judging by 11QT’s author’s apparent call for a realisation of the type yišśakar via the spacing in ‫—יש שכר‬possibly in the face of the yiśśakar-type realisation underlying 4Q522’s ‫—ישכר‬the single-sibilant pronunciation goes back to the late Second Temple Period, at the latest. Since such realisations were sufficiently widespread to achieve representation in the LXX and at Qumran (as well as in the later Latin, Aramaic, and Syriac traditions), they may well have emerged earlier. It is likewise difficult to assess the extent of the penetration of the single-sibilant realisations. The five cases of ‫ יש שכר‬in the Temple Scroll and the lone case of ‫ ישכר‬in 4Q522 are transparent enough, but what of the more standard DSS orthography ‫יששכר‬, which comes five times in the biblical DSS and twice in non-biblical material?9 Does their double-sibilant spelling indicate a corresponding double-sibilant realisation, or should 11QT’s ‫יש שכר‬ be construed as evidence that ‫ יששכר‬is mere historical spelling for what had already come to be pronounced as yiśśakar or yissakar? Is there significance to the fact that classical doublesibilant spellings characterise DSS biblical material, while six of the eight forms in non-biblical texts (including the ‫ יש שכר‬cases from the Temple Scroll and ‫ ישכר‬from 4Q522) have unconventional orthographies? There seems no getting around the ambiguity of the DSS spelling ‫יששכר‬. It could conceivably have been 9 BDSS: 4Q1 f17–18.1 = MT Exod. 1.3; 4Q11 f1+39.6 = MT Exod. 1.3; 4Q13 f1.4 = MT Exod. 1.3; 4Q27 f3ii+5.1 = MT Num. 13.7; Mas1c faii+b.2 = MT Deut. 33.18. NBDSS: 4Q484 f1.1; 11Q20 6.14. 4. The Proper Name Issachar 91 used by writers and scribes to reflect diverse phonetic realisations and may have been subject to various articulations on the parts of readers. 3.0. Conclusion Be that as it may, the available evidence is plausibly interpreted as indicating relatively ancient dissonance between the standard double-sibilant Tiberian Hebrew orthography ‫ יששכר‬and singlesibilant oral articulations, of which the Tiberian reading tradi̊̄ r̊̄ is a well-known representative. In this case tion’s ‫שכר‬ ָ ‫יִשּׂ‬ ָ yiśśåḵå of divergence between the written and reading components of the Tiberian tradition, both are shown to reflect comparatively old realisations. The admittedly secondary single-sibilant articulation dates to no later than Hellenistic times, though there is arguable indirect evidence that it emerged and diffused earlier. 5. ‫ ִל ְק ַראת‬LIQRA(ʾ)Ṯ In view of its semantic link to ‫ ָק ָרה‬I (= ‫ ָק ָרא‬II) ‘meet, befall, hap- pen’, combined with consistent orthography with the radical ʾalef characteristic of ‫ ָק ָרא‬I (= ‫ ָק ָרה‬II) ‘call, read’, the Masoretic BH infinitive-cum-preposition ‫‘ ִל ְק ַראת‬to meet; opposite, toward’ is anomalous. Cf. the expected infinitive construct of ‫ ָק ָרא‬I in ‫)קרֹא‬ ְ ‫(ל‬ ִ and of ‫ ָק ָרה‬I in ‫)קרֹות‬ ְ ‫*(ל‬. ִ 1 It also, arguably, furnishes an especially instructive glimpse of dissonance between the written and reading components of the Tiberian tradition of BH. In this case, not only can divergence of the spelling and vocalisation traditions be substantiated, but there is evidence that each reflects a realisation of profound historical depth, with roots stretching back to the Iron Age. If so, the disharmony, though evidently secondary, reflects truly ancient diversity. When one takes seriously the testimony of the individual components, the ostensible ‘problem’ inherent in a scriptural tradition composed of discordant elements proves invaluable in tracing the phonological development of the specific form in question as well as characterising the historicity of the components of the tradition. 1 A clear case of conflation in the MT is ‫תי ְּקר ֹאות‬ ֙ ִ ‫‘ ְּל ִב ְּל‬not to call’ (Judg. 8.1). The merger of ‫ קר"א‬and ‫ קר"ה‬is more advanced in RH, where in- finitival forms such as ‫(י)קר ֹ(א)ות‬ ְּ ‫‘ ִל‬to read, call, recite (the Shemaʿ)’ outnumber those of ‫ל ְּקרֹוא‬. ִ The expected III-y infinitive of ‫ ָק ָרה‬I in ‫)קרֹות‬ ְ ‫*(ל‬ ִ may be attested in 4Q179 f1i.3, but seems otherwise undocumented until piyyuṭ in the Byzantine Period. © 2023 Aaron D. Hornkohl, CC BY-NC 4.0 https://doi.org/10.11647/OBP.0310.05 94 The Historical Depth of the Tiberian Reading Tradition 1.0. The Tiberian Spelling and Similar Traditions The Masoretic spelling—which is also the dominant orthography in the DSS, the Samaritan written tradition, and RH—is ‫לקראת‬. As ʾalef is traditionally grouped with heh, waw, and yod as matres lectionis, in the case of phonetic realisations of ‫ לקראת‬along the lines of Tiberian liqraṯ, it is sometimes assumed that the otherwise otiose ʾalef serves as a mater lectionis for a. There is widespread agreement, however, that quiescent ʾalef in the Masoretic written tradition is nearly always etymological and that only at a relatively late date, under the influence of Aramaic spelling conventions, became a pure mater for a-vowels (GKC §7e; Andersen and Forbes 1986, 32, 49; Ariel 2013, 942). The use of ʾalef as a mater for a is comparatively common in the DSS (Reymond 2014, 43–47). While the exact Iron Age pronunciation of the consonantal form ‫לקראת‬, including whether it was realised with or without an audible ʾalef, must remain conjectural, the consistency of the spelling with ʾalef in Masoretic and other sources can be interpreted as evidence of the erstwhile existence of a matching pronunciation characterised by a word-internal glottal stop. How long such a pronunciation endured is difficult to determine given the available evidence. 2.0. The Tiberian Pronunciation and Similar Traditions Conspicuous in the Tiberian phonetic realisation ‫ ִל ְּק ַראת‬liqraṯ is syncope of the ʾalef consistently preserved in the orthographic tradition. Similar pronunciations are known from the Babylonian 5. ‫ ִל ְּק ַראת‬Liqra(ʾ)ṯ 95 biblical recitation tradition (Yeivin 1985, 258–59, 1133–34) and from RH. The Samaritan realisation alqēˈrat̊̄ appears to result from normal phonological processes that resolve syllable-initial consonant clusters, presupposing a form along the lines of lqrat. Given SH’s penchant for eliding gutturals, it is no surprise that the ʾalef goes unpronounced here, as in the Tiberian and similar traditions. Interestingly, however, the ultimate stress in the form alqēˈrat̊̄ may constitute indirect evidence of a formerly realised glottal stop, since ultimate stress in the Samaritan recitation tradition is restricted to words with a guttural second or third radical (Ben-Ḥayyim 2000, §§4.4.2–3). On the assumption that the ʾalef in the standard Tiberian spelling ‫ לקראת‬represents historical etymology, the form is arguably best explained as an infinitive in the (lǝ)qiṭla ̊̄ (< PS qaṭlatu [or (lǝ)qåṭla ̊̄ < quṭlatu]) nominal pattern, primarily associated with semantically stative verbs, e.g., ‫אֹותם‬ ָׂ֑ ָ ‫‘ ְּל ַא ֲה ָבה‬to love them’ (Deut. 10.15), ‫‘ ְּל ַא ְּש ָ ֵ֥מה ָ ַֽבּה‬to become guilty thereby’ (Lev. 5.26), ‫‘ ַא ֲח ֵ ֶ֖רי זִ ְּקנָ ָ ָׂ֑תּה‬after becoming old’ (Gen. 24.36), ‫ה־בּה‬ ַֽ ָ ‫‘ ְּל ָט ְּמ ָא‬to become unclean thereby’ (Lev. 15.32), ‫‘ ְּליִ ְּר ָאה א ִֵֹ֗תי‬to fear me’ (Deut. 4.10), ‫אכה‬ ֶ֖ ָ ‫ל־ה ְּמ ָל‬ ַ ‫‘ ְּל ָק ְּר ָ ֵ֥בה ֶא‬to approach to the work’ (Exod. 36.2), ‫‘ ְּל ִר ְּב ָעה א ֵָֹ֔תּה‬to lie with it’ (Lev. 20.16), ‫‘ ְּל ָ ַֽר ֳח ָק ֙ה‬to be far’ (Ezek. 8.6), ‫אֹותם‬ ֵָ֔ ‫ּומ ִשנְּ ָאתֹו‬ ִ ‘and from his hating them’. In the specific case of ‫ל ְּק ַראת‬, ִ preservation of the final tav is explained in line with consistent construct status. The vowel pattern is explained as follows: qarʾat > qarat, due to weakening of the ʾalef; qarat > qǝrat (liqrat), due to reduction of vowel distant from primary stress in construct state (GKC §§19k, 45d; Bauer and Leander 1922: 96 The Historical Depth of the Tiberian Reading Tradition Nachträge und Verbesserung (Schluß.), p. II, n. to p. 425, ln. 8ff).2 The use of the feminine infinitival form permitted semantic disambiguation: ‫ ִל ְּק ַראת‬is ‘to meet; opposite, toward’, whereas ‫)קרֹא‬ ְּ ‫(ל‬ ִ is ‘to call, read’. Again, the expected infinitive for ‫ ָק ָרה‬I is ‫)קרֹות‬ ְ ‫*(ל‬. ִ 3.0. Ambiguous Traditions Jerome’s transcription lacerath for ‫( ִל ְּק ַראת־‬Amos 4.12) is ambiguous. Brønno (1970) concluded that gutturals were preserved in Jerome’s Hebrew. They are often reflected by helping vowels, e.g., ‫ּול ֻא ִ ַֽמים‬ ְּ et Loommim ‘and Leummim’ (Gen. 25.3), or Latin h, e.g., ‫ ָ ַֽה ַא ְּש ִ ַֽר ֵא ִ ָׂ֑לי‬Asrihelitarum ‘the Asrielites’ (Num. 26.31), ‫ְּר ָא ָיַֽה‬ Rahaia ‘Reaiah’ (Ezra 2.47). However, the lack of any represen- tation of ʾalef here cannot be taken as unequivocal evidence of pronunciation without a glottal stop in light of such transcriptions as ‫ ִפ ְּר ָָֹ֨אם‬Pharam ‘Piram’ (Josh. 10.3) and ‫ וְּ ַת ְּר ֲא ָ ַֽלה‬et Tharala ‘Taralah’ (Josh. 18.27). The a-vowel after -‫ ל‬and the e-vowel after ‫ ק‬are both conventional in the transcription of his Hebrew tradition (Yuditsky 2013, 807–8, 821). 4.0. The Antiquity of the Realisation without ʾalef Assuming the validity of the explanations proffered above, two principal questions remain unanswered. First, how long did a pronunciation of ‫ לקראת‬preserving the glottal stop persist in Hebrew? While the extant reading traditions unanimously disregard it, it is legitimate, given its consistent orthographic representa2 This paragraph is an expanded and corrected revision of Hornkohl (2013a, 124, fn. 50). 5. ‫ ִל ְּק ַראת‬Liqra(ʾ)ṯ 97 tion, to wonder whether and to what extent it may have continued to be realised. There is no definitive answer to this question. The second question is: what is the historical depth of the phonetic realisation with syncopation of the glottal stop? 4.1. Second Temple Evidence For purposes of answering the second question, the available information is clearer. As has been noted, notwithstanding the received pronunciations of ‫ ִל ְּק ַראת‬without a glottal stop, the form is regularly written with an ʾalef. There are, however, exceptions. As early as the Second Temple Period, minority spellings without ʾalef in the DSS apparently reflect phonetic realisations with an elided glottal stop: ‫( לקרת‬1QIsaa 12.10 || ‫ ִל ְּק ַראת‬MT Isa. 14.9); ‫( לקרת‬4Q481a f2.4 || ‫אתֹו‬ ֵ֔ ‫ ִל ְּק ָר‬MT 2 Kgs 2.15); perhaps also ‫בנו‬ ̇ ‫ק ֯רת‬ ̇ ]‫‘ ל‬to] meet his son’ (4Q200 f5.1 || Tobit 11.10); ‫‘ ל[קר]תנו‬ag[ain]st us’ (4Q504 f1– 2Riii.13). Though the exact realisation of these forms is unknown, the omission of ʾalef comes as evidence of pronunciation without a glottal stop. 4.2. Iron Age Evidence But there is even earlier evidence of a realisation without the word-medial glottal stop. The Siloam inscription, which dates to ca. 700 BCE, includes the sentence ·‫נקבה·הכו·החצבמ·אש‬/‫ובימ·ה‬ ‫‘ לקרת·רעו·גרזנ על·[ג]רזנ‬And on the day of the / breach, the hewers struck, each man to meet (or toward) his partner, pick-axe against [pi]ck-axe’ (KAI 189, lns 3–4). As in the case of the DSS examples above, there is no certainty regarding every phonetic 98 The Historical Depth of the Tiberian Reading Tradition detail. Yet, the absence of ʾalef in an official inscription comes as compelling testimony in favour of an Iron Age pronunciation of ‫ לקר(א)ת‬without a glottal stop, not dissimilar from that preserved in the Tiberian reading tradition. According to one approach, no glottal stop (i.e., ʾalef) was lost in the Siloam inscription’s ‫לקרת‬. Rather, the spelling reflects a realisation along the lines of liqrot. If so, the ʾalef in the Tiberian and other traditions is to be considered secondary. So reason, among others, Aḥituv, Garr, and Fassberg (2016, 61), thought it is not clear whether they believe that the ʾalef in question was ever pronounced as a glottal stop in the many traditions of BH and extra-biblical Hebrew in which it appears. By contrast, the view propounded here is that the spelling with ʾalef ‫ לקראת‬is historical, i.e., reflects an ancient realisation with a medial glottal stop, and that the Siloam inscription’s ‫לקרת‬ is an early manifestation of the secondary syncope of the glottal stop seen in the Tiberian reading tradition and similar pronunciations and in minority spellings in the DSS. The syncope in question was presumably due to lax realisation, possibly connected with vernacular Hebrew, but which at some point came to be recognised as standard despite the spelling convention with ʾalef (which may have come to be considered a mater). It should also be noted that phonetic erosion is comparatively more common in the case of function words that have undergone grammaticalisation (Heine 1993, 106), which here seems to have involved the shift from lexical infinitive to preposition.3 3 I am grateful to Geoffrey Khan for calling my attention to this perspec- tive. 5. ‫ ִל ְּק ַראת‬Liqra(ʾ)ṯ 99 5.0. Conclusion Giving due consideration to both orthographic and recitation traditions, the spelling-pronunciation mismatch of the infinitiveturned-preposition ‫ ִל ְּק ַראת‬appears to reflect the intersection of divergent written and reading traditions. The written traditions bear witness almost exclusively to a pronunciation that up to some point preserved a word-medial glottal stop. Occasionally, written material omits the ʾalef and, in so doing, furnishes early (Iron Age, Second Temple) evidence of realisations in which the presumed word-medial glottal stop had become syncopated. This is the dominant sort of pronunciation preserved in most of the extant reading traditions (Tiberian, Babylonian, RH; Jerome’s Latin transcriptions and the Samaritan form with ultimate stress are possible, though by no means certain, exceptions). While it is not known when pronunciations without the glottal stop came to dominate in Hebrew, it is evident from minority Second Temple and Iron Age occurrences of ‫( לקרת‬without ʾalef) that such realisations were attested long before the medieval textualisation of the Tiberian reading tradition. It is not impossible that the ubiquity of pronunciations without the glottal stop in the extant reading traditions somewhat obscures a degree of variation in the word’s realisation. Perhaps in antiquity, forms with and without glottal stops could be heard. Be that as it may, it is plausible to conclude on the basis of the earliest cases of ‫ לקרת‬that any potential anachronism with regard to this form in the Tiberian reading and similar traditions does not apply to the phenomenon of syncope of the glottal stop, but only to the extent of the syncope. In other words, while the 100 The Historical Depth of the Tiberian Reading Tradition pronunciation without glottal stop is likely secondary, and while its standardisation may be late, early evidence confirms the deep historical roots of the feature eventually made standard. 6. THE 2MS ENDINGS In the combined Masoretic biblical written-reading tradition, word-final qameṣ (-a)̊̄ typically co-occurs with a mater heh, i.e., ‫ ִָה‬-. This norm applies across a variety of categories, including (a) substantives with the feminine singular ending, e.g., ‫ִא ָשה‬ ‘woman’, ‫דֹולה‬ ָ ְּ‫‘ ג‬great, large’; (b) the qaṭal 3FS verbal ending, e.g., ‫‘ נָ ְּתנָ ה‬she gave’; (c) the qaṭal of 3MS III-y forms, e.g., ‫‘ ָהיָ ה‬he was’; and (d) adverbial endings of various sorts, e.g., ‫‘ ַע ָתה‬now’, ‫ַא ְּר ָצה‬ ‘to the land/ground’. Conspicuously exceptional in this regard are the 2MS nominal (i.e., object/possessive) suffix ‫ָך‬- and the 2MS verbal ending ‫ ָת‬-, both of which routinely end in -a,̊̄ but—anomalously—employ defective word-final orthography, regularly eschewing the respective plene alternatives ‫ ָכה‬- and ‫ ָתה‬-.1 Correspondingly, note that the zero vocalisation of the -t of the 2FS qaṭal verbal ending and the 2FS independent subject pronoun ‫ ַא ְּת‬are regularly represented by a written shewa sign, i.e., ‫ ְּת‬-, as are all voiceless final kafs, e.g., ‫ְך‬-; no other final voiceless consonants are so treated. 1 Important discussions include Sievers (1901, §207); Kahle (1921; 1947, 95–102; 1959, 171–77); Torczyner et al. (1938, 37, 41, 51, 55); Tur-Sinai ([1940] 1987, 37–42); Cross and Freedman (1952, 53, no. 51, 65–67); Yalon (1952, ‫ ;)יז–טו‬Ben-Ḥayyim (1954); Kutscher (1963, 264– 66; 1974, 446–47; 1982, 32–35, §46); Steiner (1979); Zevit (1980, 31– 32); Blau (1982; 2010, 169–71); Barr (1989b, 114–27); Sáenz-Badillos (1993, 185); Gogel (1997, 155–64); Hutton (2013b, 966–67); Khan (2013a, 48–49; 2013b, 307; 2020, I:90); Aḥituv, Garr, and Fassberg (2016, 61); Qimron (2018, 265–67). © 2023 Aaron D. Hornkohl, CC BY-NC 4.0 https://doi.org/10.11647/OBP.0310.06 102 The Historical Depth of the Tiberian Reading Tradition It may be that this glaring mismatch between spelling and pronunciation is a simple idiosyncrasy of the Masoretic tradition, whereby normal spelling practices do not apply in the vast majority of cases of 2MS afformatives. In other words, standard ‫ך‬and ‫ת‬- and much rarer ‫כה‬- and ‫תה‬- may be thought, no matter their orthographic differences, to reflect -ḵa ̊̄ and -ta,̊̄ respectively (e.g., Koller 2021, 18). The point of departure adopted here, conversely, is that the contrast between the spellings ‫ך‬- and ‫ת‬-, on the one hand, and realisations -ḵa ̊̄ and -ta,̊̄ on the other, is linguistically significant, reflecting the merger of distinct pronunciation traditions: the written tradition with consonant-final forms and the reading tradition with vowel-final forms. To put this in context, leaving aside instances of orthographic-vocalic disparity traditionally acknowledged via the ketiv-qere mechanism—including cases of qere perpetuum, e.g., ‫ יהוה‬realised as ʾăḏōnaẙ̄ or ʾɛ̆lōhīm (see above, Introduction, §1.0, and ch. 1)—the 2MS endings arguably represent one of the most common categories of phonic divergence between the written and reading components that comprise the combined Masoretic biblical tradition. If so, they constitute a major case of unacknowledged ketiv-qere dissonance within the Masoretic tradition.2 This chapter seeks to explore this phonetic variety in ancient He- 2 Note on terminology: throughout this chapter, ‫כה‬-/‫ך‬- are referred to as 2MS nominal endings, ‫תה‬-/‫ת‬- as 2MS verbal endings; the orthographies ‫ך‬-/‫ת‬- are termed short spellings, ‫כה‬-/ ‫תה‬- long spellings; the phonetic realisations -k/-t are designated consonant-final, -ka/-ta vowelfinal; there is no presumption that the spellings ‫ך‬-/‫ת‬- were originally intended as defective for vowel-final pronunciations. 6. The 2MS Endings 103 brew morphology across dialects, chronolects, registers, and traditions as well as to plumb the historical depth of the variants.3 One further preliminary remark: though the cases of the nominal ‫ך‬- -ka ̊̄ and verbal ‫ת‬- -ta ̊̄ endings are similar, it is not assumed here that their respective orthographic and phonetic development proceeded in lockstep. Each ending merits its own study sensitive to similarities and differences, and potential analogous treatment. As such, in the present chapter a conscious effort is made to separate the relevant arguments and data. 1.0. Comparative Semitic Perspective One similarity between the verbal and nominal afformatives in question is that there is general agreement that both evolved from early Semitic forms that ended in some form of a-vowel, most likely long, though some argue for a short or anceps vowel. In Comparative Semitic perspective, there is general agreement that the Proto-Semitic 2MS object/possessive suffix from which the various ancient Hebrew forms developed was some form of -ka, the precise quantity of the vowel of which remains unclear (Ben-Ḥayyim 1954, 15–18; Moscati 1964, 109, §13.23; 3 Other notable cases in which word-final qameṣ (-a)̊̄ goes orthograph- ically unmarked in the Tiberian tradition include: (a) the 3FS ob- ject/possessive suffix ‫ ִֶי ָה‬-, e.g., ‫יה‬ ָ ‫‘ ֵא ֶל‬to her’, not ‫יהה‬ ָ ‫;*א ֶל‬ ֵ (b) ‫‘ נַ ֲע ָר‬girl’, eight times the qere perpetuum in Genesis (24.14, 16, 28, 55, 57; 34.3, 3, 12)—note also the 13 cases of qere ‫ נַ ֲע ָרה‬for ketiv ‫ נער‬in Deuteronomy (22.15, 15, 16, 20, 21, 23, 24, 25, 26, 26, 27, 28, 29) against a single case of written-reading agreement in ‫‘ ַהנַ ֲע ָ ָ֔רה‬the girl’ (Deut. 22.19); (c) the 2FPL/3FPL verbal endings (see below, ch. 9); (d) ‫‘ ַע ָת‬now’ (Ezek. 23.43; Ps. 74.6). 104 The Historical Depth of the Tiberian Reading Tradition Blau 2010, 169, §4.2.3.3.1; Huehnergard 2019, 54). The development of the consonant-final -k realisation is debated, with explanations involving variable (anceps) vowel length, Aramaic contact, and vernacular (perhaps RH) influence being suggested as factors (see Ben-Ḥayyim 1954, 18, 59–64; Steiner 1979, 158– 61; cf. Blau 1982). The verbal ending, likewise, is thought to descend from a vowel-final proto-Semitic form with -a, -ā, or -å̄ ̆ (Ben-Ḥayyim 1954, 15–18; Moscati, 1964, 138, §16.41; Blau 2010, 55, §1.18.1n, 209, §4.3.3.4.3; Huehnergard 2019, 53). As is made clear below, a number of factors complicate tracing the history of the two endings, including: widespread dissonance between the written and reading components of the Masoretic tradition; evidence of majority spellings and realisations side by side with minority alternants in both components of that tradition; evidence for the various options in other traditions of biblical material; and confirmation of contrasting orthography and/or phonology in extra-biblical material. 2.0. Diversity within the Tiberian Tradition Turning to Tiberian BH, another point of similarity between the 2MS verbal and nominal afformatives is that both exhibit some degree of diversity within Masoretic Hebrew. The nominal suffix has the short spelling in around 92 percent of its 2850 tokens. Likewise, the verbal suffix is written short in nearly 92 percent of its 1800+ occurrences. This means that in both cases, not insignificant minorities of the two afformatives are written long, i.e., with heh. 6. The 2MS Endings 105 A salient difference in this connection is that in some 200 instances of the nominal suffix, the spelling and vocalisation agree on consonant-final realisation. These consist of the 2MS pausal forms ‫בְך‬, ָ ‫לְך‬, ָ ‫א ָֹתְך‬, ‫א ָתְך‬, ִ and ‫ע ָמְך‬. ִ As many pausal forms are thought to safeguard archaic phonology (Fassberg 2013, 54; cf. Blau 2010, §§3.5.8.8n, 3.5.12.2.5n, 3.5.13.4), in the case of the 2MS nominal suffix, they purport to preserve vocalic evidence of the consonant-final realisation that underlies the written tradition’s standard ‫ך‬- orthography. Similar consonant-final forms of the verbal ending are not known in the pronunciation tradition, though instances of the apparently masculine independent pronoun ‫ את‬realised as ʾatta ̊̄ are attested in rare cases of ketiv-qere (1 Sam. 24.19; Ps. 6.4; Job 1.10; Qoh. 7.22; Neh. 9.6) and in still rarer cases of written-reading agreement on ‫( ַא ְּת‬Num. 11.15; Deut. 5.27; Ezek. 28.14). One way of looking at the apparent mix of vowel- and consonant-final forms in both the orthographic and recitation traditions is to see that the respective minority form in each corroborates the other’s majority form Khan (2013a, 48–49). Table 1: Majority and minority forms of the 2MS nominal suffix spelling pronunciation majority ‫ך‬- -ḵa minority ‫כה‬- -aḵ Thus, as can be seen in Table 1, the pausal realisation -aḵ agrees with the majority written short spelling, whereas the long spel- 106 The Historical Depth of the Tiberian Reading Tradition ling with heh agrees with the majority vowel-final pronunciation -ḵa. Likewise, in the case of the verbal ending, as seen in Table 2, the minority long spelling with heh comes as apparently early confirmation of the majority vowel-final realisation -ta and the evidence for a 2MS independent subject pronoun ʾat supports the consonant-final short spelling with ‫ת‬-. Table 2: Majority and minority forms of the 2MS verbal ending spelling pronunciation majority ‫ת‬- -ta minority ‫תה‬- 2MS ‫ את‬ʾat (ketiv) Accounts of the distribution of the minority long spellings vary in terms of explanatory power and comprehensiveness. James Barr (1989b, 114–19) judiciously discusses several of them. In the case of the 39 cases in which Tiberian 2MS written ‫ ָכה‬- and the realisation -ka ̊̄ coincide as the nominal suffix, proposed factors favouring the long spelling include graphic word length, a root consonant -k adjacent to the suffix, accumulation due to attraction, or some combination thereof (see §10.1 for citations). Prosodic factors may also be at work, as one-third of the long spellings occur with a major disjunctive accent. Of course, in most of these categories, ‫כה‬- forms nevertheless comprise a small minority of the total. The long verbal ending is especially common in certain weak verbs, e.g., ‫( נתן‬64 of the 147 total cases of 2MS qaṭal forms end in ‫תה‬-), III-y, hollow, geminate, and hifʿil I-n (Barr 1989b, 6. The 2MS Endings 107 116–19, 124–24). It also exhibits a possible prosodic connection: in 19 cases ‫תה‬- correlates with a major disjunctive accent. Be that as it may, in most of these categories, the long spellings remain the minority option (see §10.2 for citations). Interestingly, as far as accepted theories on diachrony and linguistic periodisation go, there seems to be no discernible chronological trend (Barr 1989b, 119). A possible exception in the case of the nominal suffix is apparent evidence of the late standardisation of spellings without heh, which emerges from comparisons of CBH passages with LBH parallels—though Barr (1989b, 119, 123–24) also notes the preservation of residual long spellings of the nominal suffix in the “higher and more solemn style” of prayers and divine speeches. 3.0. Kahle’s View At this point, it is worthwhile to cite the forceful opinion of Paul Kahle (1921; 1947, 95–102; 1959, 174–77) on the subject. Kahle famously opined that the consonant-final pronunciations were original in BH, regarding the Tiberian -ka ̊̄ and -ta ̊̄ realisations as Masoretic innovations of the Islamic Period imported from Qurʾānic Arabic. 4.0. Diversity beyond the Tiberian Tradition Kahle based his view not just on the dominant spellings in the Tiberian tradition, but on evidence from beyond that tradition as well. Consonant-final forms of the nominal suffix are found in transcriptional evidence in Greek (Ben-Ḥayyim 1954, 22–27; Kahle 1959, 171; Yuditsky 2016, 106; 2017, 104–6, §3.1.1.2.3) 108 The Historical Depth of the Tiberian Reading Tradition and Latin (Ben-Ḥayyim 1954, 22–27; Kahle 1959, 171–72; Yuditsky 2016, 106), the written and reading components of the Samaritan biblical tradition (Ben-Ḥayyim 1954, 37–39; 2000, 228– 29, §§3.2.2–3.2.2.0; Kahle 1959, 172–73),4 non-biblical manuscripts with Palestinian vowel pointing (Ben-Ḥayyim 1954, 27– 29; Kahle 1959, 173–74), RH (Breuer 2013, 736), and piyyuṭ manuscripts with Palestinian pointing (Kahle 1959, 172–73; see also Ben-Ḥayyim 1954, 29–32). Conversely, the Babylonian biblical reading tradition mirrors the Tiberian with -ka ̊̄ and, in the case of certain particles in pause, -ak̊̄ (Ben-Ḥayyim 1954, 32–37; Yeivin 1985, 749; cf. 421). Turning to the 2MS verbal suffix, a consonant-final -t pronunciation is found in the Greek and Latin transcriptions (BenḤayyim 1954, 43–46; Kahle 1959, 178; Yuditsky 2016, 109–10; 2017, 112–13, §3.2.1.1) and Palestinian liturgical texts (though not piyyuṭ proper) (Kahle, 1959, 178–79). Not surprisingly, these traditions also tend to favour the consonant-final form of the related 2MS independent subject pronoun, ‫ את‬ʾat (Yuditsky 2016, 4 The Samaritan tradition, like its Tiberian counterpart, is composite, comprising written and pronunciation components. Of the 39 cases of plene 2MS ‫כה‬- found in the Tiberian Pentateuch (see below, §10.1.1), the Samaritan written tradition has ‫כה‬- just seven times (see §10.2.1). For its part, the Samaritan reading tradition shows even greater preference for the -k realisation at the expense of -ka, even occasionally contradicting the spelling ‫כה‬- in the Samaritan consonantal tradition, e.g., in the case of ‫ איכה‬and ‫( יככה‬see §10.2.2). In the Samaritan reading tradition, ̊̄ ‘your coming’ (Gen. 10.30; 13.10; 25.18) this leaves only ‫ באכה‬baka with the 2MS -ka suffix, though Ben-Ḥayyim (2000, 228, §3.2.2) observes that the suffix was often otherwise interpreted. 6. The 2MS Endings 109 109–10; 2017, §3.1.1). By contrast, joining the Tiberian reading tradition with a vowel-final ending are the Babylonian biblical reading tradition (Yeivin 1985, 427), the Samaritan reading tradition (Ben-Ḥayyim 2000, 103, §2.0.13),5 and the RH written and reading tradition.6 In all four of the traditions just mentioned, the dominant form of the 2MS independent subject pronoun is also afinal (Yeivin 1985, 1103; Ben-Ḥayyim 2000, 225–26, §§3.1– 3.1.2; Breuer 2013, 735; but see below).7 Significantly, in the case of the reading components of the Samaritan biblical tradition and of RH, the a-final verbal ending conflicts with the characteristic consonant-final nominal suffix. 5 From a purely arithmetic perspective, the Samaritan written tradition, with some 49 cases of ‫תה‬-, is broadly comparable to the written tradition of the Tiberian Pentateuch, with some 44. However, the two frequently diverge on details. Nearly all of the differences appear to arise from levelling within the Samaritan tradition: on the one hand, in Samaritan, the 2MS qaṭal form of ‫ נתן‬is consistently (all 49 times) written ‫ (ו)נתתה‬in the absence of an object suffix, whereas spelling varies in the MT (30 cases if ‫ [ו]נתתה‬out of 49 potential cases); on the other, in contrast to the MT, no other Samaritan’s verb’s 2MS qaṭal form is written plene. 6 Regarding RH, in Codex Kaufmann, the vocalisation is nearly always vowel-final and the spelling is ‫תה‬- in 103 of 144 cases. 7 Samaritan Hebrew knows no remnant of the 2MS independent subject pronoun ‫את‬. In the Tiberian tradition of RH, ‫ ַא ְת‬comprises a sizable minority, e.g., in Codex Kaufmann, the 2MS independent subject pronoun is usually ‫א ָתה‬, ַ but 2MS ‫ ַא ְת‬comes in 23 of 138 cases. On the Babylonian RH tradition see Yeivin (1985, 1103 and fn. 6). 110 The Historical Depth of the Tiberian Reading Tradition Important, but somewhat complicated evidence may also be adduced from the DSS and from Iron Age epigraphy, each of which corpora is treated in detail below (§§5.0; 7.0). 5.0. The Dead Sea Scrolls: General Picture In the case of both the nominal and verbal afformatives, DSS biblical texts and non-biblical material exhibit divergent tendencies. See Table 3 for a summary of the incidence of the two spellings of the nominal suffix. Table 3: Short and long spellings of the 2MS nominal suffix in the DSS ‫ך‬- ‫כה‬- BDSS 1050 800 NBDSS 650 2000 In the BDSS, both the short and long spellings of the nominal suffix are common, short outnumbering long by a margin approaching 1050 to 800.8 The NBDSS present a different picture. Here, overall, for each instance of the 650 cases of the spelling without heh, there are more than three instances of the spelling with heh.9 8 Based on Abegg et al. (2009a). Accordance (v. 13.1.4) searches of He- brew material returned the following counts: ‫ך‬- 1050; ‫כה‬- 792. 9 The totals are approximately, ‫ך‬- 650 and ‫כה‬- 2000. The figures are based on Accordance (v. 13.1.4) searches using Abegg (1999–2009) and excluding probable Aramaic material. Though uncertainty about the language of composition, broken cases, and ambiguity, inter alia, make precise counts elusive, the picture painted is sufficiently indicative for the purposes of this study. 6. The 2MS Endings 111 Turning to the verbal ending, consult Table 4 for incidence of alternative spellings. Table 4: Short and long spellings of the 2MS verbal ending in the DSS ‫ת‬- ‫תה‬- BDSS 180 160 NBDSS 40 493 In biblical material, the spelling with heh is common, but not dominant. Against some 180 cases of the short spelling, there are around 160 occurrences of the long spelling.10 Conversely, nonbiblical material displays overwhelming affinity for the form of the verbal suffix with heh. Indeed, the long spelling, with 493 occurrences, is twelve times as common as the short one, with just 40. The broad statistical picture just painted is simplistic. Drilling down reveals complexities that merit discussion. 5.1. Nominal Suffix 5.1.1. Biblical Material As indicated above, in the BDSS both ‫ך‬- and ‫כה‬- are common, the ratio approximately 5 to 4 (1050 to 800). There is, however, striking disparity in the relative concentrations of the two options among the scrolls. See Tables 5–7. 10 An Accordance search of the Hebrew material in Abegg et al. (2009) returned figures of 262 for ‫ת‬- and 165 for ‫תה‬-. However, broken endings make 80 apparent cases of ‫ת‬- ambiguous; the same is true for a few cases of ‫תה‬-. 112 The Historical Depth of the Tiberian Reading Tradition Table 5: BDSS Mss with high incidence of ‫ך‬- (minimum ten cases) ‫ך‬- ‫כה‬- Manuscript (cont’d) Manuscript a 1QIsa 1–27 37 2 4QIsa (4Q56) 27 0 101 18 4QSam (4Q51) b 1QIsa (1Q8) 109 0 a ‫ך‬- ‫כה‬- a b n 1QPs (1Q10) 11 0 4QDeut (4Q41) 48 0 1QPhyl (1Q13) 17 0 4QJera (4Q70) 10 0 a 4QGen–Exod (4Q1) l 4QpaleoGen–Exod (4Q11) c 4QExod (4Q14) m 4QpaleoExod (4Q22) b 4QDeut (4Q29) c 4QDeut (4Q30) 13 18 15 38 16 66 0 0 0 1 0 0 c 4QJer (4Q72) 12 0 b 14 0 c 20 0 4QPs (4Q84) 4QPs (4Q85) g 4QPs (4Q89) 17 0 c 34 0 g 22 4 r 4QPhyl (4Q130) 4QPhyl (4Q134) e 4QDeut (4Q32) 23 0 4QPhyl (4Q145) 10 0 4QDeutf (4Q33) 25 0 5QDeut (5Q1) 17 0 g 4QDeut (4Q34) 13 0 8QPhyl (8Q3) 64 1 i 11 0 8QMez (8Q4) 20 0 0 a 16 0 4QDeut (4Q36) r 4QpaleoDeut (4Q45) 33 11QpaleoLev (11Q1) TOTALS 877 26 Table 6: BDSS Mss with high incidence of ‫כה‬- (minimum ten cases) Manuscript a 1QIsa 28–54 g 4QLev (4Q27) k 4QDeut (4Q38a) c 4QSam (4Q53) c 4QIsa (4Q57) g 4QXII (4Q82) a 4QPs (4Q83) ‫ך‬- ‫כה‬- Manuscript (cont’d) ‫ך‬- ‫כה‬- a 0 24 b 1 23 j 0 37 k 0 23 l 0 10 m 0 21 20 212 4QPhyl (4Q128) 0 0 0 0 23 4QPhyl (4Q129) 10 4QPhyl (4Q137) 11 4QPhyl (4Q138) 13 4QPhyl (4Q139) 4 12 4QPhyl (4Q140) 0 a 24 11QPs (11Q5) TOTALS 6 232 31 675 Table 7: BDSS Mss with mixed use of ‫ך‬- and ‫כה‬- (minimum ten cases) Manuscript j 4QDeut (4Q37) ‫ך‬- ‫כה‬- Manuscript (cont’d) 8 7 h 4QPhyl (4Q135) TOTALS ‫ך‬- ‫כה‬4 6 12 13 6. The 2MS Endings 113 Most scrolls show a discernible predilection for one form or the other. This includes a marked difference between the two halves of 1QIsaa, cols 1–27 (see above, Table 5) and cols 28–54 (see above, Table 6; see Kutscher 1974, 564–66; Abegg 2010, 40–41). In two-thirds of the manuscripts listed above (29 of 45), ‫ך‬- is the preferred variant. Nearly 600 of the approximately 800 occurrences of ‫כה‬- in the BDSS are found in the selection of material comprised of the two large scrolls 1QIsaa (‫ך‬- 121; ‫כה‬- 230) and 11QPsa (11Q5) (‫ך‬- 6; ‫כה‬- 232), along with the phylacteries from Cave 4, 4QPhyla–4QPhyls (4Q128–4Q146) (‫ך‬- 91; ‫כה‬- 164).11 Among texts with ten or more cases of the nominal suffix, only 4QDeutj (4Q37) and 4QPhylh (4Q135) show truly mixed usage, with no obvious preference for short or long spelling. 5.1.2. Non-biblical Material NBDSS material presents a different picture. Here, overall, for each instance of ‫ך‬-, there are more than three instances of ‫כה‬-. See Tables 8–10. Table 8: NBDSS Mss with high incidence of ‫ך‬- (minimum ten cases) Manuscript ‫ך‬- ‫כה‬- Manuscript ‫ך‬- ‫כה‬- CD 1QLitPrb (1Q34bis) 20 14 67 12 1 2 20 146 1 4 4Q Narrative and… (4Q372) 13 11 0 0 4Q Non-Canonical Pss B (4Q381) 4Q Barki Nafshid (4Q437) 0 11QapocrPs (11Q11) TOTALS However, certain individual phylacteries in this group show a decided preference for ‫ך‬-, as does 8QMezuzah (8Q4). 114 The Historical Depth of the Tiberian Reading Tradition Table 9: NBDSS Mss with high incidence of ‫כה‬- (minimum ten cases) Manuscript 1QS ‫ך‬- ‫כה‬- Manuscript 1 28 4Q Instructionb (4Q416) ‫ך‬3 ‫כה‬98 1QSb (1Q28b) 1QM (1Q33) 0 4 74 4Q Instructionc (4Q417) 98 4Q Instructiond (4Q418) 1 6 55 192 1 0 12 27 0 1 16 21 1QHa (1QHa) 1QInstruction (1Q26) 158 409 4Q Instructione (4Q418a) 0 11 4Q Instructiong (4Q423) 11 4QHa (4Q427) 10 4QHb (4Q428) 1QHymns (1Q36) 4QRPa (4Q158) 0 0 4QJubd (4Q219) 4QpapJubh (4Q223–224) 1 1 11 4Q Narrative Work… (4Q460) 0 11 4QMa (4Q491) 3 4QBera (4Q286) 4QBerb (4Q287) 0 0 16 4QapocrLam B (4Q501) 10 4QpapRitMar (4Q502) 0 1 11 11 4QRPb (4Q364) 4QPEnosh (4Q369) 0 1 21 QpapPrQuot (4Q503) 22 4QDibHama (4Q504) 0 4 21 91 4QapocrMosesa (4Q375) 4QapocrJoshuaa (4Q378) 0 15 4QPrFêtesb (4Q508) 10 12 4QpapPrFêtesc (4Q509) 4Q pap paraKings… (4Q382) 1 4QRitPur A (4Q414) 0 1 2 9 52 24 4QpapRitPur B (4Q512) 14 4QBeat (4Q525) 0 4 28 30 4Q Instructiona (4Q415) 12 11QTa (11Q19) TOTALS 0 13 10 1 138 205 1644 Table 10: NBDSS Mss with mixed usage of ‫ך‬- and ‫כה‬- (minimum ten cases) Manuscript ‫ך‬- ‫כה‬- Manuscript ‫ך‬- ‫כה‬a 4QTest (4Q175) 7 4 4QapocrJoshua (4Q378) 10 12 4QTanh (4Q176) 5 8 TOTALS 22 24 Most texts strongly favour one option over the other, though cooccurrence of the two within a single text and/or line is not uncommon. The overall preference for ‫כה‬- in the NBDSS is apparent, particularly in comparison to the preference for ‫ך‬- in BDSS material. Beyond this, however, it is difficult to discern meaningful usage patterns. Focusing on texts with mixed usage (Table 10)— two of the three include verbatim biblical citations, but the ‫ך‬- 6. The 2MS Endings 115 and ‫כה‬- spellings occur in biblical as well as non-biblical material, with no obvious correlation.12 5.2. Verbal Ending 5.2.1. Biblical Material The 2MS qaṭal spelling ‫תה‬- is common, but not dominant in BDSS material. Against some 180 cases of ‫ת‬-, come around 160 occurrences of ‫תה‬-. Tables 11–13 tally manuscripts with at least five total cases. Table 11: BDSS Mss with high incidence of ‫ת‬- (minimum five cases) ‫ת‬- ‫תה‬- Manuscript Manuscript a 1QIsa ‫ת‬- ‫תה‬- b 0 6 m 0 6 0 10 12 73 4QPhyl (4Q129) (1–27 8 24) 4QPhyl (4Q140) a (28–54 4 49) 11QPs (11Q5) TOTALS 12 95 Table 12: BDSS Mss with high incidence of ‫תה‬- (minimum five cases) Manuscript ‫ת‬- ‫תה‬- Manuscript 1QIsab (1Q8) 13 1 4QDeutg (4Q34) l 4QpaleoGen–Exod (4Q11) f 4QExod–Lev (4Q17) m 4QpaleoExod (4Q22) f 4QDeut (4Q33) 5 0 r 5 0 13 0 4QPs (4Q85) 6 0 6 1 5 1 90 6 11 1 13 4QpaleoDeut (4Q45) c c 17 2 4QPhyl (4Q130) 9 0 5QDeut (5Q1) TOTALS 12 ‫ת‬- ‫תה‬- In 4QTest (4Q175) all eleven forms parallel MT forms with ‫ך‬-; in 4QTanh (4Q176) the six forms that parallel MT ‫ך‬- split evenly between ‫ך‬- and ‫כה‬- (these latter totals exclude instances of MT 2FS suffixes, in some cases of which 4QTanh (4Q176) has ‫כה‬- or apparently 2MS ‫ך‬-. 13 = MT Exod. 12.44. 116 The Historical Depth of the Tiberian Reading Tradition Table 13: BDSS Mss with mixed use of ‫ת‬- and ‫תה‬- (minimum five cases) Manuscript n 4QDeut (4Q41) ‫ת‬- ‫תה‬- Manuscript 4 a 4 4QSam (4Q51) TOTALS ‫ת‬- ‫תה‬2 5 6 9 By dint of its length, the Great Isaiah Scroll often skews statistical presentations of DSS material. Such is the case here, as 1QIsaa accounts for just under half of the cases of both 2MS qaṭal forms in general (85 of 180) and 2MS ‫תה‬- spellings specifically (73 of 160).14 Similar outliers characterised by the use of ‫תה‬- are 11QPsa and many of the Cave 4 phylacteries. If 1QIsaa, 4QPhyla–s, and 11QPsa are excluded from consideration, the ‫ת‬- to ‫תה‬- ratio is about 150 to 50 (compared to 180 to 160, as above). As is evident from the tables, most manuscripts show strong preference for one form or the other, with only a few manuscripts exhibiting mixed usage. It is interesting to compare the preferences for qaṭal ‫ת‬- versus ‫תה‬- in Tables 11–13 with preferences for ‫ך‬- versus ‫כה‬- above, §5.1.1, in Tables 5–7. Most scrolls that prefer ‫ת‬- also prefer ‫ך‬- and most that prefer ‫תה‬- also prefer ‫כה‬-. For instance, 1QIsab is strongly partial to ‫ת‬- and ‫ך‬-, whereas 11QPsa is strongly disposed to ‫תה‬- and ‫כה‬-. Yet, there are a few surprises. For example, while 1QIsaa exhibits high incidence of both ‫תה‬and ‫כה‬-, the dominance of ‫תה‬- over ‫ת‬- (73 to 12) is far more pronounced than that of ‫כה‬- over ‫ך‬- (230 to 120). Moreover, the striking difference between the two halves of 1QIsaa concerning 14 While there is some disparity in the use of qaṭal ‫ת‬- versus ‫תה‬- between the two halves of the scroll (Table 11), they are far more similar in their usage of the 2MS verbal ending than in the case of the variants of the 2MS nominal suffix (§5.1.1 and Tables 5–6). 6. The 2MS Endings 117 ‫כה‬- and ‫ך‬- (cols 1–27: 18 versus 101; cols 28–54: 212 versus 20) obtains in the case of ‫תה‬- and ‫ת‬- only in the second half of the scroll (cols 1–27: 24 versus 8; cols 28–54: 49 versus 4). While cols 28–54 show striking preferences for both ‫תה‬- and ‫כה‬-, cols 1–27 prefer ‫תה‬- to ‫ת‬- (24 versus 8) but not ‫כה‬- over ‫ך‬-, the latter far more prevalent than the former (‫כה‬- 18 versus ‫ך‬- 101). Though involving far smaller numbers, a similar situation obtains in the case of 4QSama (4Q51), where ‫ך‬- is far more common than ‫כה‬- (37 to 2), but ‫ת‬- is less frequent than ‫תה‬- (2 to 5). Such differences are reminiscent of the situation in the Samaritan reading tradition and RH, all confirming the importance of independent analysis of the 2MS nominal and verbal morphology. 5.2.2. Non-biblical Material DSS non-biblical material displays overwhelming affinity for 2MS qaṭal forms ending in ‫תה‬-. Indeed, ‫תה‬-, with 493 occurrences, is more than twelve times as common as ‫ת‬-, with just 40. Table 14: NBDSS Mss with high incidence of ‫ת‬- (minimum five cases) Manuscript ‫ת‬- ‫תה‬- Manuscript ‫ת‬- ‫תה‬- 4Q Non-Canonical Pss B (4Q381) 4 1 TOTALS 4 1 118 The Historical Depth of the Tiberian Reading Tradition Table 15: NBDSS Mss with high incidence of ‫תה‬- (minimum five cases) ‫ת‬- ‫תה‬- Manuscript Manuscript ‫ת‬- ‫תה‬c 1QpHab 0 5 4Q Barki Nafshi (4Q436) 0 17 1QM (1Q33) 0 22 4QDibHama (4Q504) a 1 30 b 1QHa (1QH ) 2 159 4QPrFêtes (4Q508) 1 4 c 1QDM (1Q22) 0 6 4QpapPrFêtes (4Q509) 0 14 a 0 7 4QShirb (4Q511) 0 7 4QRP (4Q364) 1 7 4QBeat (4Q525) 0 6 4Q pap paraKings et al. (4Q382) 0 5 5Q Rule (5Q13) 0 6 4QD (4Q266) b d 4Q Instruction (4Q418) b 4QH (4Q428) a 1 14 11QT (11Q19) 1 71 0 6 TOTALS 7 386 Table 16: NBDSS Mss with mixed use of ‫ת‬- and ‫תה‬- (minimum five cases) Manuscript d 4Q Barki Nafshi (4Q437) ‫ת‬- ‫תה‬- Manuscript ‫ת‬- ‫תה‬- 3 3 5 TOTALS 5 With so few cases of ‫ת‬- in the NBDSS, one wonders about the possibility of conditioning factors, e.g., conventional spellings associated with biblical passages. For example, ‫ושמחת בחגך אתה‬ ‫‘ ובנך‬and you will rejoice in your festival, you and your son’ (4Q366 f4i.10) is an exact orthographic match for the same expression in MT Deut. 16.14. Additionally, the lone unambiguous case of ‫ת‬- in the Temple Scroll ‫‘ ובערת הרע מקרבכה‬and you will purge the evil from among you’ (11Q19 54.17–18) comes seven times in MT Deuteronomy with a ‫ ָת‬- ending (and a 2MS ‫ָך‬- pronominal suffix) (but there may also be other factors at work in this example; see below). Yet, biblical citation or allusion is certainly no guarantee of a ‫ת‬- spelling. Consider ‫מוצא שפתיכה תשמור כאשר נדרתה נדבה‬ ‫‘ בפיכה לעשות כאשר נדרתה‬What passes your lips take care, as you have vowed a freewill offering with your mouth, to do as you have vowed’ (11Q19 54.13) || ‫ית ַכ ֲא ֶָֹ֨שר נָ ַ ֶ֜ד ְּר ָת‬ ָ ‫מר וְּ ָע ִ ָׂ֑ש‬ ֹ ‫מֹוצא ְּש ָפ ֶ ֶ֖תיָך ִת ְּש‬ ֵ֥ ָ 6. The 2MS Endings 119 ‫ֹלה ֙יָך נְּ ָד ֵָ֔בה ֲא ֶ ֵ֥שר ִד ַ ֶ֖ב ְּר ָת ְּב ִ ַֽפיָך‬ ֶ֙ ‫יהוַ֤ה ֱא‬ ָ ‫‘ ַל‬What passes your lips take care of and do, as you have vowed to the LORD your God a freewill offering that you have spoken with your mouth’ (MT Deut. 23.24), where MT ‫ ָת‬- is consistently paralleled by ‫תה‬- (and ‫ָך‬- by ‫כה‬-). By the same token, MT ‫תה‬- occasionally parallels DSS ‫ת‬-, as in ‫ונת]ת‬ [‫הכספ‬ ̇ ‘and you will give the money’ (4Q364 f32.4) || ‫וְּ נָ ַת ָתה ַה ֶָ֡כ ֶסף‬ (Deut. 14.26), despite the fact that the ‫תה‬- ending is dominant in the MT in the case of 2MS qaṭal ‫נָ ַתן‬.15 Concerns of space might have influenced spelling. The lone ‫ת‬- ending in the Temple Scroll’s ‫‘ ובערת‬and you will purge’ (11Q19 54.17) is line-final; as are a few—but not many—other cases of the short spelling (4Q435 f2i.5; 4Q437 f2i.12; 4Q438 f4ii.2). Even so, line-final ‫תה‬- spellings are not uncommon. It is reasonable to ponder the possible effects of prosodic and phonological factors, but it is difficult to assess them given the limitations and ambiguities of the available data. 5.3. DSS Summary Most scrolls show a discernible predilection for one form or the other. And usually, texts that prefer the short or long nominal spelling also prefer the corresponding length of verbal ending. However, this is not true of the two halves of the Great Isaiah Scroll, where each half prefers either short nominal and long verbal 2MS morphology or vice versa. In the biblical material, a few individual scrolls, including 1QIsaa, 11QPsa, and the phylacteries 15 Though, in the present case, it might be argued that the first heh in a sequence of two consecutive hehs has simply dropped out. 120 The Historical Depth of the Tiberian Reading Tradition from Cave 4, contain three-quarters of the long nominal spelling and nearly 70 percent of the long verbal spelling. The regularity of the long spellings in some BDSS material is indisputable evidence that a vowel-final realisation similar to Tiberian -ka and -ta was in common use in the late Second Temple Period. It thus seems gratuitous to attribute the Tiberian -ka and -ta realisations to eighth-century CE Arabic influence.16 Conversely, short spellings are ambiguous. One option is to view them as straightforward evidence of consonant-final realisations.17 This is probably legitimate in a great many, if not most cases. However, caution is in order. The co-occurrence of the two spellings in the same text, and even in the same line,18 arguably 16 Pace Kahle (1959, 174–77), who maintains that the -ka suffix re- flected in DSS orthography “was lost for centuries and was reintroduced with great regularity by the Tiberian Masoretes,… and has therefore to be regarded as an innovation of the eighth century” (175), under the influence of Qurʾānic Arabic and the orthography of DSS manuscripts. Kahle could not have known the extent to which his formulation “a certain number of Hebrew manuscripts from the Dead Sea Caves in which an ending ‫ ה‬appears” (1959, 176) represented a gross underrepresentation of the frequency of ‫כה‬- and -ka there, nor of the existence of Iron Age epigraphic evidence for -ka (see below). However, the limited data did not prevent other scholars from proposing sounder approaches, e.g., Cross and Freedman (1952, 67); Ben-Ḥayyim (1954). 17 This was obviously Kahle’s view (1921, 234–35; 1959, 171–77). Khan (2013b, 307) seems to imply that the DSS ‫ך‬- and ‫כה‬- spellings represent distinct phonetic realisations. Kutscher (1974, 446–47), Reymond (2014, 35–36, 39, 156, 226), and Qimron (2018, 265–66) all to varying degrees view the issue as purely orthographic. 18 Outside of 1QIsaa, where co-incidence of 2MS ‫ך‬- and ‫כה‬- in a single line is encountered 45 times, intralinear co-occurrence is chiefly, but 6. The 2MS Endings 121 points to the possibility that some cases of the short spellings are defective representations of vowel-final realisations under the influence of classical biblical spelling practices.19 Given the strong evidence for the Second Temple Period coexistence of consonant-final and vowel-final variants of the 2MS sufformatives in sources representative of registers both formal and vernacular, the most prudent hypothesis would seem to be that DSS short spellings reflect both consonant- and vowelfinal realisations. The one possible exception is the short spelling of the verbal ending in the non-biblical scrolls, the rarity of which might indicate that this is consistently conservative spelling for a vowel-final pronunciation. A plausible reading of the evidence is that the DSS mixture of forms reflects both competing archaic and contemporary spelling practices as well as opposing diachronic, dialectal, and registerial phonetic realisations. not exclusively, limited to phylacteries: 4QPhylf (4Q133) f1.1 (|| MT Exod. 13.11–12); 4QPhylg (4Q134) f1.20 (erasure) (|| MT Deut. 5.14), 24 (suspended heh) (|| Deut. 5.16); 4QPhylh (4Q135) f1.11 (|| Deut. 6.2–3); 8Q3Phyl (8Q3) f1–11i.22 (|| Exod. 13.15–16); 11QPsa (11Q5) 20.12 (|| Ps. 139.20–21); XHev/SePhyl (XHev/Se5) f1.7 (|| Exod. 13.15–16). 19 This is in line with Barr’s (1989b, 123) observation regarding appar- ent Second Temple levelling of the perhaps once more prevalent ‫תה‬and ‫כה‬- to ‫ת‬- and ‫ך‬-, respectively, in the Masoretic consonantal tradition. In the precious few cases of diachronically separated parallel passages, there is a tendency to replace the former with the latter according to late scribal norms. See Barr (1989b, 125) on broader textual possibilities. 122 The Historical Depth of the Tiberian Reading Tradition 6.0. Aramaic Given the influential role attributed to Aramaic in several theories of the development of the Hebrew 2MS object/possessive suffix, it is fitting to focus briefly on the situation in Aramaic itself. Beginning with the nominal suffix, in Masoretic BA, the spelling ‫ך‬- (with 99 cases) occurs to the total exclusion of ‫כה‬- and it coincides consistently with consonant-final vocalisation (though there are 18 cases of ketiv-qere dissonance involving vocalic realisation before the suffix). In the Aramaic of the DSS, there is greater variety, but consonant-final spellings still dominate. Thus, in biblical DSS Aramaic material, the counts are ‫ך‬- 11 versus ‫כה‬- 3,20 while in nonbiblical DSS Aramaic the totals are ‫ך‬- 200 versus ‫כה‬- 40. In the Genesis Apocryphon alone, the totals are ‫ך‬- 74 versus ‫כה‬- 1 (‫מנכה‬ ̇ ‘from you’ 1Q20 20.26 and one erasure in }‫‘ לכ{א‬to you’ 1Q20 5.9). Of course, while ‫כה‬- is phonetically transparent, ‫ך‬- may con- ceivably represent a consonant- or vowel-final realisation. Even so, it is clear that neither Aramaic, in general, nor BA and QA, specifically, are uniform regarding the realisation of the 2MS object/possessive suffix. Elsewhere in Aramaic of the Judaean Desert, in Syriac, and in later Aramaic dialects consonant-final forms dominate. 20 The three cases of disparity between Masoretic BA and DSSBA all come in the same scroll, 4QDanb (4Q113), which preserves only these three cases: ‫‘ ֲא ֗בּוך‬your father’ (Dan. 5.11) || ‫( אבוכה‬4Q113 f1–4.3); ketiv ‫ עליך‬qere ‫‘ ֲע ָ֔ ָלך‬about you’ (Dan. 5.16) || ‫( עליכה‬4Q113 f1–4.14); ‫ֱא ָל ָ֗הך‬ ‘your God’ (Dan. 6.21) || ‫( אלהכה‬4Q113 f7ii–8.18). 6. The 2MS Endings 123 In the case of the verbal suffix in Aramaic, variety ensues. Masoretic BA shows the following pattern of incidence: ‫ ְּת‬- 6, ‫ ָת‬16, ‫ ָתה‬- 3. The related 2MS independent subject pronoun likewise shows deviation from uniformity: according to the ketiv, it is ‫אנת‬ 1, ‫ אנתה‬14; according to the qere, ‫ ַאנְּ ְּת‬all 15 times. In DSSBA, all six 2MS suffix conjugation forms end in ‫ת‬-, but the 2MS independent subject pronoun is thrice ‫ אנת‬and twice ‫אנתה‬. And in nonbiblical QA, vowel-final forms of both the 2MS verbal ending and the 2MS independent subject pronoun prevail—verbal ending: ‫ת‬15; ‫תא‬-/‫תה‬- 23; pronoun: 0 ‫ ;אנת‬26 ‫אנתה‬. Short spellings are standard in Aramaic documents from elsewhere in the Judaean Desert, as well as in Syriac and later forms of Aramaic. As in the case of the 2MS nominal suffix, it seems that early diversity eventually gave way to later preference for short spellings, whatever their phonetic realisation. 7.0. Iron Age Inscriptions A fundamental question involves the historical depth of the Hebrew vowel-final -ka and -ta realisations. The earliest unequivocal attestation usually proffered consists of the dominant DSS long spellings. As noted above, this firmly anchors vowel-final pronunciations like those of the Tiberian reading tradition in the Second Temple Period. The affinity between the Tiberian pronunciation tradition and Second Temple written evidence is not a coincidence, as there are many salient commonalities between the Tiberian vocalisation and Second Temple Hebrew material (LBH, DSS), where both appear to deviate from the linguistic testimony of the Masoretic written tradition of CBH material. Cru- 124 The Historical Depth of the Tiberian Reading Tradition cially, though, in many cases where it seems that the Tiberian reading tradition reflects relatively late secondary standardisation of a feature, the feature itself proves to have far earlier roots. This also applies to the 2MS afformatives under discussion here, as is evident from Iron Age inscriptional material. Regarding the nominal suffix—in Iron Age Hebrew epigraphy, the short spelling ‫ך‬- dominates. In view of the normal use of mater heh to mark final -a (Cross and Freedman 1952, 57; Zevit 1980, 14–15, 24–25, 31–32; Gogel 1997, 59; Hutton 2013b, 966– 67), this spelling is probably generally indicative of the consonant-final -k realisation of the 2MS nominal suffix. There are, however, a minority of inscriptional forms bearing ‫כה‬-: (1) ‫לשנותכה‬ ‘to change/recount to you’ (Ḥorvat ʿUzza Literary Text ln. 2; Hutton 2013, 967b; cf. Aḥituv 2008, 173–74) (2) ‫זרעתיכה‬ ‘your arms’ (Ḥorvat ʿUzza Literary Text ln. 11; Hutton 2013b, 967;21 cf. Aḥituv 2008, 173–74) (3) ‫קב[ר]כה‬ ‘your tom[b]’ (Ḥorvat ʿUzza Literary Text ln. 12; Aḥituv 2008, 173–7722) (4) ‫וקברכה‬ ‘and your tomb’ (Ḥorvat ʿUzza Literary Text ln. 13) 21 This is Cross’s reading according to Beit-Arieh (1993, 64); cf. Beit- Arieh (1993, 61). 22 This is Lemaire’s (1995) reading according to Aḥituv (2008, 176). 6. The 2MS Endings (5) 125 ‫אלהיכה‬ ‘your God’ (Khirbet Beit Lei Cave Inscription 1.1; see Gogel 1997, 158; Aḥituv 2008, 233) In this connection, mention should also be made of the orthography of ‫ יברך‬in the following instances (6) ‫ ישרמך‬/ ‫ך יהוה‬/‫יבר‬ ‘may Yhwh bless you (?), keep you’ (Ketef Ḥinnom 1.14) (7) ‫שרמך‬/ ̇‫ י‬/ ‫ י̇ הוה‬/ ‫יברך‬ ‘may YHWH bless you (?), keep you’ (Ketef Ḥinnom 2.5) (8) ‫רך וישמרך‬/‫ יב‬.‫ ולאשרתה‬/ ‫הוה תמן‬/‫ לי‬.‫ברכתך‬ ‘I have blessed you to YHWH of Teman and to his Ashera. May he bless you (?) and keep you’ (Kuntillet ʿAjrud 2.4– 7) If the forms written ‫ יברך‬are to be interpreted as including an object suffix, as in MT ֵ֥‫יְּב ֶר ְּכָך‬ ָ in ‫הוֶ֖ה וְּ יִ ְּש ְּמ ֶ ַֽרָך‬ ָ ְּ‫‘ יְּ ָב ֶר ְּכָךֵ֥ י‬may the LORD bless you and keep you’ (MT Num. 6.24)—and not as simple ‫יְּ ָב ֵרְך‬ in a cataphorically elliptical ‫‘ יברך יהוה וישמרך‬may YHWH bless and keep you’, with no 2MS suffix on the first verb—then the omission from ‫ יברך‬of the expected ‫ך‬- suffix is most plausibly explained as a result of assimilatory gemination,23 which process presupposes 23 An alternative explanation, namely, that the omission is due to scribal lapse, seems implausible, given that it assumes the mistake was made all three times the phrase was written in two separate corpora. Further, note that in only one case is the end of the word line-final. 126 The Historical Depth of the Tiberian Reading Tradition a vowel-final form.24 This is far from certain, however, and there are alternative views. In summary, Iron Age Hebrew epigraphy presents up to eight cases of the vowel-final nominal suffix -ka, the most secure of which is example (4) above. Though not the majority spelling or, probably, the majority pronunciation, the inscriptional long spellings confirm the antiquity of the relevant spelling and pronunciation in the DSS and of the standard Tiberian pronunciation. Turning to the verbal ending—as is often the case, Iron Age Hebrew epigraphic material is important as pristine evidence, but problematic due to the phonetic ambiguity of its orthography— even the most plene Hebrew spelling is characterised by partial vocalic ambiguity, and the spelling in Iron Age epigraphy is more defective than in most Hebrew writing. Be that as it may, the epigraphic evidence, though somewhat ambiguous, is sufficiently transparent to confirm the antiquity of a vowel-final realisation. 24 For Aḥituv (2008, 53) the writing of a single ‫ך‬- might be a labour- and/or space-saving strategy, whereby it serves double duty, like the yod in ‫‘ חיהוה‬as surely as YHWH lives’ (Arad 21.5) and ‫‘ וכיאמר‬and be- cause (my lord) says’ (Lachish 3.8–9). In ‫ כיאמר‬kyʾmr for ‫ כי יאמר‬ky yʾmr, -‫ כ‬is defective for ‫ כי‬and cliticised to the following word beginning with consonantal yod. In the cases of ‫ חיהוה‬ḥyhwh for ‫ חי יהוה‬ḥy yhwh and ‫יברך‬ ybrk for ‫ יברכך‬ybrkk(a), the relevant double-duty letters presumably sig- nal geminated consonants. Cf. Aḥituv, who postulates two phonetic options without gemination: yĕḇārĕḵǝḵa or yĕḇārĕḵa. The first is arguably a poor candidate for double-duty spelling with ‫ך‬-, because the k consonants are separated by a reduced vowel. The second goes one step further, assuming gemination followed by degemination (and a fricative ḵ!). Regardless, both assume a -ka realisation of the 2MS suffix. 6. The 2MS Endings 127 The extant inscriptional corpus includes twelve relevant examples. Some of the cases provide unequivocal evidence of ‫ת‬- as the spelling of the 2MS qaṭal sufformative. While this spelling is phonetically ambiguous, in light of the routine usage of matres lectionis for final vowel sounds in the corpus (Gogel 1997, 59; Hutton 2013b, 965), they are commonly taken as evidence of a consonant-final -t phonetic realisation. Consider examples (9)– (14): (9) ‫סבת מחר‬/‫וה‬ ‘and you will make the rounds tomorrow’ (Arad 2.5–6) (10) ‫ת להמ‬/‫ואמ·עוד·חמצ·ונת‬ ‘and if there is still vinegar, you will give (it) to them’ (Arad 2.7–8) (11) ‫אתמ·בצק‬ ֯ /‫וצררת‬ ‘and you shall bind them’ (Arad 3.5–6) (12) ‫ שמנ‬1·‫ת·משמ‬/‫ולקח‬ ‘and you will take therefrom 1 (unit of) oil’ (Arad 17.3–4) (13) ‫והש]בת את [בגד ע]בדכ‬ ‘[and] you [will retu]rn the [garment of] your [se]rvant’ (Yavne Yam 14) (14) ]‫[ת‬/‫כ·א[ת]·הספרמ כזא‬/‫כלב·[כי]·שלחת אל עבד‬/‫מי·עבדכ‬ ‘Who is your servant (but) a dog [that] you have sent to your servant the letters like this? (Lachish 5.3–6)25 25 Against the spelling ‫ שלחתה‬reconstructed by some scholars, see Dobbs-Allsopp et al. (2005, 320–21); but cf. Gogel (1998, 83, 86). As- 128 The Historical Depth of the Tiberian Reading Tradition In other cases, the spelling ‫תה‬- appears. This spelling is also often ambiguous and, as such, is variously interpreted. Consider the alternative renderings in examples (15)–(20): (15) ‫לפני֯ כ‬ ֯ ‫כתבתה‬/]‫[ ו‬ ‘and you will write before you’ – or – ‘and you will write it before you’ (Arad 7.5–6) (16) [ ‫והנ·ידעתה‬ ‘And behold, you knew/know…’ – or – ‘And behold, you knew/know it…’ (Arad 40.9) (17) ‫דבר·אשר לא·ידעתה‬... ‘anything that you do not know’ – or – ‘anything that you do not know it’ (Lachish 2.6) (18) ‫לע ֯ב ֯ד ֯כ ֯אמש‬ ֯ ‫שלחתה אדני‬/‫[נא] ֯א ֯ת ֯אז֯ נ֯ ֯עבדכ·לספר·אשר‬/‫ועת·הפקח‬ ‘And now, please open the ear of your servant to the letter that you sent, my lord, to your servant yesterday’ – or – ‘And now, please open the ear of your servant to the letter that you sent it, my lord, to your servant yesterday’ (Lachish 3.4–6) (19) ‫אשר שלח[ת]ה ֯אלי‬/·‫כתבתי על הדלת ככל‬ ‘I have written upon the door according to all that you have sent to me’ – or – ‘I have written upon the door according to all that you have sent it to me’ (Lachish 4.3–4) suming a correct reconstruction as above, a 2MS qaṭal form with heh would be strong evidence of plene spelling of -ta. 6. The 2MS Endings 129 (20) ‫פר‬ ֯ ‫רא·ס‬ ֯ ‫ק‬/· ֯ ‫אמר·אדני֯ ·לא·ידעתה‬ ֯ ‫וכי‬ ‘and because my lord said, “You do not know (how) to read a letter”’ – or – ‘and because my lord said, “You do not understand it. Call a scribe!’” (Lachish 3.8–9)26 Scholars are divided on the interpretation of such forms: are they reflections of a vowel-final 2MS qaṭal ending -ta (as in the Tiberian reading tradition) or consonant-final -t with a 3MS or 3FS object suffix? Persuaded by the unambiguous cases of ‫ת‬- in examples (9)–(14), above, some scholars take all cases of 2MS ‫תה‬in the relevant corpus as incorporating an object suffix (Parunak 1978, 28 [on Arad]; Cross 1985, 43–46; Dobbs-Allsopp et al. 2005, 23, 73, 307, 311; Rollston 2006, 62, fn. 42; Hutton 2013b, 967–68). But as Dobbs-Allsopp et al. (2005, 23, 73, 307, 311) repeatedly make clear, these judgments are based on a balance of probability, not certainty. In other words, because the Arad and Lachish evince unequivocal cases of ‫ת‬-, it is reasoned that ambiguous ‫תה‬- should be regarded as -t + object suffix. But this seems to assume a degree of orthographic and phonological consistency arguably foreign to Iron Age epigraphic Hebrew. Consider the presentatives ‘behold’ ‫( הן‬Arad 21.3; 40.9) and ‫( הנה‬Arad 24.18; Jerusalem 2.2; Lachish 6.5, 10). Or, perhaps more relevant, consider forms of the 1CS qaṭal: most cases end with ‫י‬- (Arad 16.4; 24.18; 60.1; 88.1; Lachish 3.12; 4.3; 12.4; Yavne Yam 11), but several end with ‫ת‬- (Kuntillet Ajrud 18.1; Meṣad Ḥashavyahu 8; Murabbaʿat 1.1). It was clearly not impossible for scribes (or a 26 Similarly, some render the words ‫פר‬ ֯ ‫רא·ס‬ ֯ ‫·ק‬ ֯ ‫‘ לא·ידעתה‬you do not un- derstand it—call a scribe!’ (see below). 130 The Historical Depth of the Tiberian Reading Tradition single scribe) to utilise orthographic and/or phonetic variants that differed in terms of final spelling and/or phonetic realisation. While examples (15)–(16) are truly ambiguous, in (17)– (20) there are linguistic factors that appear to favour interpretation of the spelling ‫תה‬- as plene for a vowel-final -ta realisation with no pronominal object suffix. In the case of examples (17)–(19), the pertinent considerations are grammatical and pragmatic. Wholesale interpretation of the long spelling ‫תה‬- as -t + pronominal suffix entails positing three cases of relativising ‫ אשר‬followed by 2MS qaṭal and a resumptive accusative object pronoun. As Holmstedt (2008, 5, 13– 14) shows, such structures are rare in BH—the combination ‫ֲא ֶש ֹר‬ +2MS finite verb+(‫ )את‬resumptive accusative pronominal suffix comes in, e.g., Gen. 45.4; Lev. 23.2, 4; Num. 34.13; Deut. 33.8; Josh. 2.10. Neither are they the preferred structure in inscriptional Hebrew. The formulation ֯‫‘ ועת ככל אשר·שלח·אדנ֯ י‬and now according to all that my lord sent’ (Lachish 4.2) is a 3rd-person parallel for examples (18) and (19) above, but shows no resumptive accusative pronominal suffix after ‫ ;אשר‬cf. ‫(‘ שלחה‬your serv- ant) has sent it’ (Lachish 3.21). Also relevant is ‫אש ֯ר‬ ֯ ·‫ככל·האתת‬ ֯ ‫אדני‬/‫נתנ‬ ֯ ‘according to all the signs that my lord gave’ (Lachish 4.12)—again with no resumptive accusative. It would certainly be surprising for such a rare grammatical structure to appear twice in the limited corpus presented by the Lachish letters. Further, it is altogether suspicious that the purported instances are limited to 2MS cases of ‫תה‬- that are amenable to alternative readings. 6. The 2MS Endings 131 Holmstedt (2008, 5, 13–14) provides an explanation for the rarity of the structure discussed above as well as an argument for why the assumed cases thereof in the Lachish letters are best explained otherwise. He applies Keenan and Comrie’s (1977, 66) Noun Phrase Accessibility Hierarchy (NPAH) Subj > DO > IO > Oblique > Gen > Obj of Comparison Holmstedt (2008, 6) explains that it strongly predicts the positions in which a language may use resumptive pronouns, i.e., first and more often for less accessible positions farther to the right on the hierarchy. He (2008, 14, fn. 12) elaborates as follows: There are many examples of RC [relative clause] resumption in the Hebrew Bible and, as the NPAH leads us to expect, the great majority are in the genitive/NP-internal and oblique (object of preposition) positions within the RC. Resumption in the object position occurs less frequently and its use is highly constrained: it is used (1) to disambiguate verbal semantics in cases when a verb taking an accusative or oblique complement results in distinct meanings, or (2) to signal that the object carried focus pragmatics within the RC. In agreement with Holmstedt, neither of the verbs in examples (17)–(19) requires semantic disambiguation based on meaning differences with accusative versus oblique complements. Nor does either seem a good candidate for argument focus. There is thus no grammatical or pragmatic motivation for resumption of the accusative after relativising ‫ אשר‬in examples (17)–(19). Turning to example (20)—again, the conviction that ‫תה‬must include a pronominal suffix seems to have led a number of 132 The Historical Depth of the Tiberian Reading Tradition scholars to render ‫פר‬ ֯ ‫ ֯קרא· ֯ס‬/·‫ לא·ידעתה‬as ‘you don’t understand it—call a scribe!’ (Lachish 3.8–9). While the consonantal string ‫ ספר‬is ambiguous, representing something along the lines of Tiberian ‫‘ ס ֵֹפר‬scribe’ or ‫‘ ֵס ֶפר‬letter’, Schniedewind (2000b, 160) is correct to problematise the semantic elasticity assumed for the verb ‫ ידע‬by those who render it ‘understand’ (pace Cross 1985, 43–46; Rollston 2006, 62, fn. 42). In this case, too, then, the long spelling ‫תה‬- seems merely to indicate a vowel-final 2MS realisation -ta. This means that the spelling of 2MS qaṭal ‫תה‬- in four of the six ostensibly equivocal cases listed above is more likely to represent -ta with no object suffix than -t with an object suffix. This supports the theory of probable phonetic variety in 2MS verbal morphology in Iron Age Hebrew epigraphy (in agreement with Zevit 1980, 19, 28; Gogel 1998, 83–88; Schniedewind 2000b, 160; Holmstedt 2008, 13–14; Aḥituv, Garr, and Fassberg 2016, 61), similar to that characteristic of various other forms of ancient Hebrew, including the combined written and reading Masoretic tradition.27 27 Zevit (1980, 31–32) and Rainey (Aharoni 1981, 22) raise the possi- bility that the distinction between 2MS verbal ‫ת‬- and ‫תה‬- is somehow related to the well-known stress distinction between qaṭal and weqaṭal in Tiberian BH. The proposal, however, has not been well received (see, e.g., Pardee 1985, 69; Gogel 1998, 83–84; Hutton 2013b, 967–68). First, early qaṭaltá would be expected to result in Tiberian qǝṭalta ̊̄ ̊̄́ (as in the 2M/FPL forms); the preservation of a full vowel in the antepenultimate syllable is evidence that the rules that resulted in the distinction ̊̄ ̊̄ and qǝṭaltɛ̊̄́m were no longer operative when between Tiberian qaṭálta ̊̄ ̊̄ ̊̄́ came into being. Second, given the BH stress distinction, one wǝqaṭalta ̊̄ would expect ‫תה‬- to coincide with the stress in wǝqaṭalta ,̊̄ ̊̄́ but the pro- 6. The 2MS Endings 133 8.0. Conclusion: Historical Depth of 2MS ‫ָך‬- and ‫ ָת‬in the Tiberian Reading Tradition In summary, though the Tiberian vocalisation tradition’s dominant vowel-final 2MS -a ̊̄ nominal and verbal endings likely differ from the prevailing consonant-final endings that the Masoretic consonantal spellings are probably intended to represent, there is substantial evidence indicating that vowel-final 2MS morphology was in use in the Second Temple Period. There is also evidence, albeit arguable, of minority vowel-final 2MS morphology in First Temple sources, including apparently pre-exilic biblical consonantal material and, of special importance, Iron Age Hebrew epigraphy. Vowel-final 2MS morphology thus qualifies as a departure of the Tiberian reading tradition from its written counterpart involving the secondary standardisation of an early minority linguistic feature. posal is for the opposite. Third, there appear to be inscriptional qaṭal forms ending in ‫ת‬- and weqaṭal forms ending in ‫תה‬-, so the most that can be said is that there is a preference for distinct spellings, not full consistency. Fourth, even if the spelling distinctions are generally characteristic, there is no certainty that they represent phonological distinctions. Finally, the Second Temple crystallisation of the Tiberian reading tradition provides a context for the secondary development of disambiguating stress, as there is mounting evidence that the proto-Tiberian reading tradition included the implementation of orthoepic strategies to preserve the precise realisation and safeguard understanding of the biblical text (Khan 2018b; 2020, I:99–105). 134 The Historical Depth of the Tiberian Reading Tradition 9.0. Appendix: Further Consideration of Complexities While the significance of the early attestation of the long spellings ‫כה‬- and ‫תה‬- is obvious, the import of the ‫ך‬- and ‫ת‬- spellings— whether merely orthographic or phonetic—is ambiguous in many Hebrew traditions. In this section, the discussion centres on various complicating considerations regarding the nominal suffix. For example, central to Kahle’s (1921; 1947, 99–100; 1959, 175–76) argument for the secondary nature of Tiberian ‫ָך‬- was the view that the prevailing ‫ך‬- spelling of the Masoretic consonantal tradition represents dominance of a classical, high-register -k realisation. Yet, in other corpora -k is considered representative of the vernacular and/or due to late Aramaic influence. Consider the words of Cross and Freedman (1952, 66–67): The longer form of the suffix was native to old Hebrew, and survived in elevated speech and literary works. The shorter form developed in the popular speech at a very early date (with the dropping of the final å̄,̆ which is to be regarded as anceps). The present Massoretic [sic] text represents a mixture of these forms, both of which have been extended throughout the Bible. The short form is preserved in the orthography, the long form in the vocalization. The orthography was standardized, clearly on the basis of manuscripts in which the short form predominated. The vocalization, however, was based on manuscripts in which the long form was common. It is a testament to the complexity of the problem that Cross and Freedman are compelled to make several counterintuitive claims. First, in this connection they consider the Masoretic consonantal 6. The 2MS Endings 135 tradition, with the spelling ‫ך‬-, more innovative than the Tiberian reading tradition, which preserved the -ka ̊̄ of “elevated speech and literary works.” Such a view runs counter to common scholarly attitude regarding the diachronic relationship between the Masoretic written and reading traditions, whereby the reading tradition is generally considered the more evolved of the two. Second, they argue that in this case it is the consonantal tradition that reflects the form associated with “popular speech,” the vocalisation reflecting a conservative manuscript tradition. Again, while not impossible, this is at odds with the usual linking of the Tiberian reading tradition to Second Temple vernacular conventions, especially as seen in RH. Khan (2020, I:90) responds to Kahle’s privileging of Pales- tinian material, discussing the ‘vernacular’ or ‘popular’ character of multiple Second Temple traditions, including in connection to the 2MS -k variant: The distinctive features of Palestinian pronunciation, which are particularly discernible in the non-biblical manuscripts with Palestinian pronunciation, have close parallels with what is known about the vowel system of Jewish Palestinian Aramaic [Fassberg 1991, 28–57]. Unlike Tiberian and Babylonian, the Palestinian biblical reading is unlikely to be a direct descendant of the proto-Masoretic reading, but rather it has its roots in other traditions of reading that were current in Palestine in antiquity. The Greek transcription in Origen’s Hexapla (the middle of the third century C.E.) reflects a reading that has even more evidence of influence from the Aramaic vernacular, especially in the pronominal suffixes, such as the 2ms suffix -akh, e.g. σεμαχ ‘your name’ (Tiberian ‫ ִִׁ֝ש ְּמ ֵָ֗ך‬Psa. 31.4) [Brønno 1943, 110, 196–200]. This is also a feature of the 136 The Historical Depth of the Tiberian Reading Tradition Samaritan tradition, e.g. yēdåk ‘your hand’ (Tiberian ‫)יָ ְּדָך‬ [Ben-Ḥayyim 2000, 228]. Some of these features, such as the Aramaic type of pronominal suffixes, appear in medieval non-biblical texts with Palestinian vocalization. In the second half of the first millennium, however, it appears that the popular biblical reading converged to a greater extent with the prestigious Tiberian tradition. As a result, the Aramaic type of suffixes were eliminated in the biblical reading [Yahalom, 1997, Introduction].28 If the orthography ‫ך‬- and the realisation -k are early, then perhaps even in RH they might be considered a retention rather than an innovation. After all, despite its overall late character, RH is thought to preserve individual archaisms (Pérez-Fernández 1999, 7–9; cf. Cook 2017, 5 and fn. 3). Most scholars, however, attribute RH -k to late Aramaic influence (Ben-Ḥayyim 1954, 62– 64; Kutscher 1963, 264–66; Sáenz-Badillos 1993, 185; even Pérez-Fernández 1999, 5). For his part, Breuer (2013, 736) sees the conditioned distribution of RH -k (after consonants) and -ka (after vowels) in contrast to -ak alone in Aramaic as evidence that RH -k is a secondary development, but not one of Aramaic origin. The difficulty in definitively characterising the use of -k and -ka in the DSS should now be evident. It was proposed above that the spelling ‫ך‬- should sometimes be considered a retention. But what of the realisation -k? Is it to be considered an archaic phonetic retention, in line with the classical BH realisation presumed to underly MT ‫ך‬- and assumed by some to be preserved in Tibe̊̄ Or is it rather to be deemed an innovation due rian pausal -aḵ? to contact with Aramaic and/or the influence of a late vernacular 28 See also Blau (2010, 171, §4.2.3.3.5). 6. The 2MS Endings 137 in the line of RH? Is only the DSS spelling ‫כה‬- to be considered innovative and popular, but the -ka realisation it surely reflects conservative and prestigious? The intersection of various considerations to do with orthography, phonology, chronolect, dialect, register, and transmission within various traditions complicates the discussion. Bauer and Leander (1922, 30) and Cross and Freedman (1952, 66) consider the widespread reduction of -ka to -k a very early phenomenon. Steiner (1979, 162 and fn. 9) agrees that it “must be dated to a time when Hebrew was still a living language,” but that the evidence for Aramaic influence adduced by BenḤayyim [1954] and Kutscher [1963] makes it difficult to accept the suggestion of Bauer and Leander (1922, p. 30) that the development in question had already taken place during the Biblical period, in a dialect different from the one which formed the basis of the Masoretic vocalisation. However, given (a) the regularity of final -a marked by ‫ה‬- in the case of non-2MS morphology in both Iron Age inscriptional Hebrew and all traditions of BH, (b) the regular absence of ‫ה‬- in cases of the 2MS suffix in Iron Age inscriptional Hebrew and the Masoretic consonantal tradition, and (c) the usual affinity between Iron Age inscriptional Hebrew and the Masoretic consonantal tradition, a relatively strong case can be made for routine Iron Age realisation as -k. Indeed, in both the inscriptions and the Masoretic consonantal text, it is the ‫כה‬- spelling—the only unequivocal evidence for the -ka realisation—that constitutes the decisive minority. 138 The Historical Depth of the Tiberian Reading Tradition It is likely that, as time passed, the original variation became further complicated, whether due to dialectal, registerial, or mixed factors. Aramaic was almost certainly a factor, both for ‫ך‬- and -k (RH, the DSS, the transcriptions) and thereagainst (BH reading traditions, the DSS).29 Second Temple vernacular registers, such as that later documented as Tannaitic RH, must also have played a role, again, both for ‫ך‬- and -k (RH, the transcriptions) and against them (the DSS).30 So, too, if Steiner is correct about pausal forms, elevated reading practices must also have played a part (in BH, RH, and the DSS). From this perspective, it is interesting that among the DSS the ‫ך‬- spelling, while overall the minority form, is comparatively more common in biblical than in non-biblical material, though, as Qimron (2018, 266) notes, ‫כה‬- occurs in DSS biblical material “even where other phases of Hebrew use the apocopated form, e.g., with the prepositions ‫לכה‬, ‫בכה‬, ‫אתכה‬, ‫ עמכה‬in pausal position….” Whatever pronunciation DSS ‫ך‬- represents, adherence to classical spelling 29 While bilingual readers may have conflated Hebrew and Aramaic suf- fixes, the more careful among them may have made an effort to prevent the penetration of Aramaic features into the classical Hebrew tradition (Ben-Ḥayyim 1954, 61). 30 Here it seems fitting to acknowledge Schniedewind’s (1999; 2000a; 2021) theory of Qumran Hebrew as an anti-language; cf. Tigchelaar (2018). It may also be worth considering in this connection two Qumran compositions the Hebrew of which is often considered uniquely representative of contemporary vernacular traits. In the Copper Scroll (3Q15; on the language of which, see Wolters 2013), all three cases of the 2MS suffix are ‫ך‬-; in 4QMMT (comprising 4Q394–397, 4Q399; on the language of which, see Yuditsky 2013a), there are four cases of ‫ך‬- (all in 4Q399) and five of ‫כה‬- (all in 4Q397). 6. The 2MS Endings 139 norms seems to have been more common in biblical than in nonbiblical sources. As for the DSS ‫כה‬- spelling and -ka realisation— the regularity of the orthography is clearly a late phenomenon, but as the related phonetic realisation tallies with the minority Iron Age inscriptional orthography, there seems no reason to doubt a genetic link between the two involving -ka, which until the late Second Temple Period, seemingly by chance, enjoyed only sporadic orthographic representation. Circling back to the combined Tiberian written and reading tradition, it is possible to summarise. To begin with, if the -ka affinity between First and Second Temple extra-biblical material (inscriptions and DSS) is organic, then -ka ̊̄ in the Tiberian reading tradition likely also has genuinely old roots—even if anti-Aramaic and anti-vernacular concerns may have contributed to its preservation. Second, while RH -k is probably rightly considered a late vernacular feature, this does not mean that Tiberian consonantal ‫ך‬- and its presumed -k realisation are not, along with the Tiberian reading tradition’s -ka,̊̄ authentic Iron Age phenomena.31 31 In an Iron Age Hebrew dialect with 2MS -k, it is not clear how related and complementary morphology would be realised. For example, forms similar to the Tiberian reading tradition’s 2MS independent subject pronoun ʾatta ̊̄ are the norm in the Masoretic consonantal text (where 2MS ‫‘ ַא ְת‬you’ [Num. 11.15; Deut. 5.27; Ezek. 28.14] and cases of ketiv ‫את‬ read as qere ‫] ַא ָתה‬1 Sam. 24.19; Ps. 6.4; Job 1.10; Qoh. 7.22; Neh. 9.6] are rare), the Babylonian reading tradition (Yeivin 1985, 1103), the DSS (Qimron 2018, 260), Ben Sira, SH (Ben-Ḥayyim 2000, 225–26, §§3.0– 3.1, 3.1.2), the Secunda [normally αθθα, just once αθ] (Yuditsky 2013b, 811), Jerome (attha, ath); RH, though ‫ ַא ְת‬occurs in a sizeable minority of cases (Breuer 2013, 735). Obviously, users of some forms of Hebrew 140 The Historical Depth of the Tiberian Reading Tradition In terms of detailing the merger and explaining things as they now stand, Barr’s (1989b, 123–25) view is an attractive place to begin. The Hebrew Bible’s oldest material probably exhibited greater spelling (and, thus, phonological) variety, i.e., a larger number of cases of ‫כה‬-. But early Second Temple scribes, copying and composing during the period of LBH, standardised the spelling ‫ך‬-, leaving only a tiny remnant of ‫כה‬- (a spelling that certain factors helped to preserve). This standardisation may well have been influenced by a dialect and/or register in which the use of -k had largely pushed out that of -ka, whether due to convergence with Aramaic, diffusion of liturgical or vernacular apocope, or some combination of these. Crucially, however, the scribal process responsible for depiction of the 2MS suffix in the Masoretic written tradition did not dictate a matching realisation in the tolerated a difference in the realisation of a vowel-final 2MS independent pronoun and a consonant-final 2MS object/possessive suffix. (My thanks to Ben Kantor for the forms from Jerome.) Finally, it is also interesting to consider the 2FS object/possessive suffix. As is well known, in the Tiberian reading tradition the pausal form of the 2MS suffix is identical to that of the 2FS suffix in the case of certain particles. This evinces toleration of a certain degree of ambiguity, also characteristic of various forms of the corresponding Samaritan suffixes. On the assumption that the Masoretic consonantal text regularly reflects 2MS -ak, the standard Tiberian 2FS -ēk would have been sufficient for gender disambiguation; other 2FS alternatives include RH and Aramaic ‫יך‬- -ik and the variously represented ‫כי‬- -ki, which is sporadic in the Tiberian written and reading traditions and rare in DSS orthography—though Qimron (2018, 267–68) posits -ki as the majority (defectively spelled) DSS realisation—but well attested in Aramaic dialects (including the Syriac written tradition) and Deir ʿAlla. 6. The 2MS Endings 141 proto-Tiberian reading tradition. Here, too, there was a process of levelling, but in this case -ka ̊̄ became the standard (except in pause in the case of a few forms)—perhaps out of resistance to the very factors that led the expansion of ‫ך‬- and -k in the written tradition. Of course, much of this proposal is conjecture, neither verifiable nor falsifiable, but it arguably fits the facts and is somewhat reminiscent of other cases of dissonance between the Masoretic written and pronunciation traditions examined in this monograph. At any rate, the picture painted by the combined evidence is one of diversity as far back as the evidence goes, extending back into the late Second Temple Period and beyond. The consonantal-vocalic dissonance in the combined written-reading Masoretic tradition concerning ‫ָך‬- appears to be the artificial result of the merging of divergent pronunciation traditions. The anachronism lies not in the spelling ‫ך‬- for -k or the realisation -ka ̊̄ reflected in ‫כה‬-—as each of the respective orthographies and realisations reliably represents a genuine First Temple variant— but in the standardisation of one or the other in each component of the tradition. The BDSS evidence points to the conclusion that -k and -ka were contemporary options for the realisation of the 2MS object/possessive suffix in the late Second Temple Period. What this all means for the literary Hebrew of the early Second Temple Period, to say nothing of the Iron Age, has been a matter of some controversy. Kahle (1959, 174–77) downplayed the historical relevance of the DSS spelling ‫כה‬- for the question of the dissonance between the Tiberian vocalisation and conso- 142 The Historical Depth of the Tiberian Reading Tradition nantal text. Barr (1989b, 117–18) seems to imply that Kutscher (1982, 32–35, §46), or his followers, were guilty of overstating the importance of DSS ‫כה‬-: the discovery of Qumran texts with -ka written plene as ‫כה‬-, many times, was hailed as proof that the ancient form had been, as the Masoretic tradition had it, -eka or the like. This, however, was to claim too much: the Qumran texts which so spell prove only that in Qumran times some people thought that this was the pronunciation, they do not prove that it had always and universally been so. Indeed, the very fact of these writings at Qumran could be taken as an indication that opinion on the matter was divided and that efforts were then being made to induce the community to use the pronunciation -eka or the like. Though Kutscher’s (1982, 32, §46) proclamation of the defeat of Kahle’s hypothesis—“The discovery of the DSS… sounded the death knell of this theory”—can be interpreted as a simplistic rejection of Kahle’s evidence and arguments, Kutscher’s earlier (1974, 446–47) discussion in the context of 1QIsaa shows his awareness of the possibility of multiple realisations at Qumran and in Second Temple Hebrew more generally. From this perspective, it now seems superfluous to insist on Islamic Period Arabic influence on Tiberian -ḵa.̊̄ On the other hand, Kutscher’s insistence that -k realisations were due to “the influence of the substandard (= Rab. Hebr.) on the standard” suggests that he considered -ka the standard, classical, biblical form, which may not do justice to the complexity of the situation. 6. The 2MS Endings 143 10.0. Citations 10.1. Tiberian Biblical Tradition 10.1.1. Cases of the Plene 2MS Nominal Suffix In order of frequency, the 39 instances according to L involve yiqṭol forms of ‫‘ ִה ָכה‬strike’ (8x: Deut. 28.22, 27, 28, 35; 2 Sam 2.22; Isa. 10.24; Jer. 40.15; Ps. 121.6), the infinitival expression ‫ ב ֲֹא ָכה‬lit. ‘your coming’ (6x: Gen. 10.19, 19, 30; 13.10; 25.18; 1 Kgs 18.46), the prepositional forms ‫‘ ְל ָכה‬to, for you’ (3x: Gen. 27.37; 2 Sam. 18.22; Isa. 3.6) and ‫‘ ְב ָכה‬on you, in you, because of you’ (3x: Exod. 7.29; 2 Sam. 22.30; Ps. 141.8), the direct object particle ‫‘ א ְֹת ָכה‬you’ (2x: Exod. 29.35; Num. 22.33), yiqṭol forms of ‫‘ ֵב ַרך‬bless’ (2x: Gen. 27.7; Ps. 145.10), the preposition ‫‘ ְכמֹו‬like’ (2x: Exod. 15.11, 11), and single cases of ‫‘ ַא ֶי ָכה‬where are you?’ (Gen. 3.9), ‫‘ ִהנְ ָכָ֤ה‬here you (see)’ (2 Kgs 7.2), ‫‘ ָי ְָ֣ד ָָ֔כה‬your hand’ (Exod. 13.16), ‫‘ כ ֶֹח ָכה‬strength’ (Prov. 24.10), ‫‘ ַכ ֶפּ ָכה‬your hand’ (Ps. 139.5), ‫‘ יִ ְמ ָצ ֶא ָכה‬they could (not) find you’ (1 Kgs 18.10), ‫‘ ֲאנַ ְס ָכה‬I will test you’ (Qoh. 2.1), ‫ִתנְ ְצ ֶר ָכה‬ ‘(understanding) will guard you’ (Prov. 2.11), ‫‘ ִע ְמ ָכה‬with you’ (1 Sam 1.26), ‫ּוכה‬ ָ ‫‘ יַ ֲענ‬they will (not) answer you’ (Jer. 7.27), ‫(‘ יַ ַע ָצ ְר ָכִ֖ה‬the rain) will (not) stop you’ (1 Kgs 18.44), the infinitive construct ‫אֹות ָכִ֖ה‬ ְ ‫(‘ ַה ְר‬in order to) show you’ (Ezek. 40.4), ‫‘ ְב ִש ְמ ָָ֜כה‬in your name’ (Jer. 29.25). This list differs slightly from Barr’s (1989b, 116, 127), in that his includes two cases of ‫‘ ֵח ְל ָכה‬helpless’ (Ps. 10.8, 14), despite his own doubts on their relevance (Barr 1989b, 115; cf. also BDB, 319; HALOT 319), and excludes ‫(‘ ַא ֶכ ָכה‬Why) should I strike you?’ (2 Sam. 2.22). 10.1.2. Cases of the Plene 2MS Verbal Ending In order of frequency, the most salient categories are: ‫ – (וְ )נָ ַת ָתה‬64x (Gen. 3.12; 15.3; Exod. 21.23; 25.12; 26.32, 33; 27.5; 28.14, 24, 25, 27; 29.12, 20; 30.6, 36; 40.5, 6; Num. 3.9, 48; 7.5; 27.20; 31.29, 30; Deut. 11.29; 14.25, 26; 15.17; 26.10, 12, 15; Josh. 15.19; 17.14; Judg. 1.15; 1 Sam. 1.11; 1 Kgs 8.36, 36, 40, 48; 9.13; Jer. 29.26; Ezek. 4.1, 2, 2, 3, 9; 43.19, 20; Ps. 4.8; 18.41; 21.3, 5; 39.6; 60.6; Ezra 9.13; Neh. 9.15, 20, 36, 37; 2 Chron. 6.25, 27, 30, 31, 38; 20.10); III-y – 32x ‫ִ֖יתה‬ ָ ‫( ֻצֵּו‬Gen. 45.19); ‫יתה‬ ָ ‫( (וְ ָ)ר ִא‬Num. 27.13; 2 Sam. 18.21; Ps. 10.14; 35.22; Lam. 3.59, 60); ‫יתה‬ ֵ֥ ָ ִ‫)צּו‬ ִ ְ‫( (ו‬Num. 27.19; Jer. 32.23; Ps. 119.4; Lam. 1.10); ‫ֵ֥יתה‬ ָ ִ‫( וְ ָהי‬Judg. 11.6; 2 Sam. 10.11), ‫יתה‬ ָ ‫( ָפּ ִ ָ֤צ‬Judg. 11.36); ‫יתה‬ ָ ‫( ָע ִׂש‬1 Sam. 14.43; 15.6; 24.19, 20; 2 Sam. 3.24; 12.21; 16.10; Ezek. 35.11); ‫יתה‬ ָ ‫)ה ִכ‬ ִ ְ‫( (ו‬1 Sam. 15.3; 2 Kgs 9.7; Jer. 5.3); ‫יתה‬ ָ ‫( גָ ָ֜ ִל‬2 Sam. 7.27); ‫( ָב ִָ֔נ ָתה‬1 Kgs 9.3); ‫יתה‬ ָ֤ ָ ‫( וְ ִה ְש ִק‬Jer. 25.15); ‫( וְ ָחיִ ִָ֖תה‬Jer. 38.17); ‫יתה‬ ָ ‫( נִ ְד ֵָ֔מ‬Obad. 5); ‫יתה‬ ָ ‫( ָפּ ִ ִ֖ד‬Ps. 31.6); ‫יתה‬ ָ ‫( ִה ְר ִא‬Ps. 60.5); strong verbs – 22x ‫( נִ ְכ ַ ִ֖ס ְפ ָתה‬Gen. 31.30); ‫( וְ ִהזְ ַה ְר ָתה‬Exod. 18.20); ‫( וְ יָ ַש ְב ָתה‬Deut. 17.14); 144 The Historical Depth of the Tiberian Reading Tradition ‫( וְ ָח ַפ ְר ָתה‬Deut. 23.14); ‫( זָ ַקנְ ָתה‬Josh. 13.1); ‫( וְ ָא ַס ְפ ָ ֵ֥תה‬Judg. 18.25); ‫( וְ ַה ֲח ַר ְמ ָָּ֞תה‬1 Sam. 15.18); ‫( יָ ַ ָ֔ד ְע ָתה‬2 Sam. 2.26); ‫( ָח ַש ְב ָתה‬2 Sam. 14.13); ‫( ִה ְת ַח ַננְ ָתה‬1 Kgs 9.3); ‫וְ ָנ ַפ ְל ָָ֔תה‬ (2 Kgs 14.10); ‫( נָ ַ֗ט ְש ָתה‬Isa. 2.6); ‫( וְ ָש ַמ ְט ָ֗תה‬Jer. 17.4); ‫( זָ ָ֔ ַע ְמ ָתה‬Zech. 1.12); ‫ָב ַג ְָ֣ד ָתה‬ (Mal. 2.14); ‫כֹוננְ ָתה‬ ָ (Ps. 8.4); ‫( ֶה ֱע ַ ַ֪מ ְד ָתה‬Ps. 30.8); ‫( ָס ַ ַ֪פ ְר ָָּ֫תה‬Ps. 56.9); ‫( ִה ְר ַע ְש ָתה‬Ps. 60.4); ‫( ִא ַ ֵ֥מ ְצ ָתה‬Ps. 80.16); ‫( ִמ ַג ְָ֣ר ָתה‬Ps. 89.45); ‫( ִה ְס ַכנְ ָתה‬Ps. 139.3); II-w/y – 15x ‫( ַג ְֵָ֥֣ר ָתה‬Gen. 21.23); ‫ּומ ְל ָתה‬ ַ (Exod. 12.44); ‫( ַה ֵע ָ֤ד ֹ ָתה‬Exod. 19.23); ‫( וְ ֵה ַמ ָתה‬Num. 14.15; 1 Sam. 15.3); ‫אתה‬ ָ ‫( ָ ִ֖ב‬2 Sam. 3.7); ‫( וְ ַנ ְִ֖ס ָתה‬2 Kgs 9.3); ‫ֹותה‬ ָ ‫( ֲה ִרימ‬Isa. 37.23); ‫( ַר ְצ ָתה‬Jer. 12.5); ‫( וַ ֲה ִכינ ָֹתה‬Ezek. 4.3); ‫( וָ ַמ ָתה‬Ezek. 28.8); ‫אתה‬ ָ ‫( ֻה ָב‬Ezek. 40.4); ‫ַש ָתה‬ (Ps. 8.7); ‫( ֱֱ֝ה ִב ֗שֹ ָתה‬Ps. 53.6); ‫( ַ ֵ֥בנְ ָתה‬Ps. 139.2); geminate – 6x ‫( ֲה ֵר ֹ֨עֹ ָתה‬Exod. 5.22); ‫( וְ ַקצ ָ ִֹ֖תה‬Deut. 25.12); ‫( וְ ֵה ַפ ְר ָתה‬2 Sam. 15.34); ‫( ַס ֵ֥כֹ ָתה‬Ps. 140.8; Lam. 3.43, 44); hifʿil I-n – 4x ‫( ִה ַ ָ֑ג ְָ֣ד ָתה‬Judg. 14.16); ‫יתה‬ ָ ‫)ה ִכ‬ ִ ְ‫( (ו‬1 Sam. 15.3; 2 Kgs 9.7; Jer. 5.3); II/III-ʾ – 4x ‫אתה‬ ָ ‫( נָ ָׂש‬Num. 14.19); ‫( ָמ ַא ְס ָתה‬Judg. 9.38); ‫( ָמ ַא ְס ָתה‬1 Sam. 15.26); ‫( ֵ ֵ֭נ ַא ְר ָתה‬Ps. 89.40); III-t – 1x ‫( ֱִ֝ה ְצ ַ֗מ ָתה‬73.27); miscellaneous – 1x ‫( ַ ֵ֥ת ָתה‬2 Sam. 22.41). Groves–Wheeler (1991–2010, v. 4.14) counts 148, but mistakenly tags as 2MS the 3FS ‫‘ וְ ָחיְ ָתה‬and (your soul) will live’ (Jer. 38.17). Barr (1989b, 116, 125–27) lists 146, omitting ‫ִ֖יתה‬ ָ ‫( ֻצֵּו‬Gen. 45.19) and ‫( נָ ַ ָ֤ת ָתה‬Neh. 9.15), while including ketiv ‫ שת‬qere ‫( ַש ָתה‬Ps. 90.8). 10.2. Samaritan Biblical Tradition 10.2.1. Cases of the Plene 2MS Nominal Suffix in the Written Tradition ̊̄ ‘your coming’ (Gen. 10.30; 13.10; 25.18); ‫ איכה‬īka ‘how!’ (Gen. 3.9); ‫ באכה‬baka ‫ יככה‬yikkåk ‘(the LORD) will strike you’ (Deut. 28.22, 27, 35). 10.2.2. Cases of the Plene 2MS Nominal Suffix in the Reading Tradition ‫ יככה‬yikkåk ‘(the LORD) will strike you’ (Deut. 28.22, 27, 35; cf. ‫ יכך‬SP Deut. 28.28 || MT ‫ ;)יַ ְכ ָכה‬MT ‫ באכה‬is twice entirely unparalleled in SP Gen. 10.19; SP ‫ || ואברכך‬MT ‫‘ וַ ֲא ָב ֶר ְכ ָכה‬that I may bless you’ (Gen. 27.7); SP ‫ּול ָכה || ולך‬ ְ ‘and for you’ (Gen. 27.37); SP ‫ || ובך‬MT ‫ּוב ָ ֵ֥כה‬ ְ ‘and on you’ (Exod. 7.29); SP ‫‘ ידיך‬your hands’ || MT ‫‘ יָ ְָ֣ד ָָ֔כה‬your hand’ (Exod. 13.16); SP ‫ || כמוך‬MT ‫‘ ָכמ ָֹכה‬like you’ (Exod. 15.11 [2x]); SP ‫ || אתך‬MT ‫‘ א ָֹת ָכה‬you’ (Exod. 29.35; Num. 22.33). Though SP ‫( איכה‬Gen. 3.9) has ‫כה‬-, the realisation īka is identical to that of rhetorical ‫איכה‬ || MT ‫יכה‬ ֵ֥ ָ ‫‘ ֵא‬how?’ (Deut. 1.12; see Ben-Ḥayyim 2000, 319, §6.3.7). 7. THE 2FS ENDINGS A degree of diversity characterises ancient Hebrew 2FS morphology. Specifically, the 2FS independent subject pronoun, the 2FS suffix conjugation ending, and the 2FS nominal (ob- ject/possessive) suffix all exhibit both majority consonant-final forms, namely, standard ‫א ְּת‬, ַ ‫ ְּת‬-, ‫ְך‬-, and their respective minority vowel-final alternants, ‫אתי‬, ‫תי‬-, ‫כי‬- (Hornkohl 2013, 112–19). The present chapter focuses on dissonance between the written and reading components of the Tiberian biblical tradition involving the realisation of such 2FS morphological forms. 1.0. The Combined Tiberian Biblical Tradition Examining the written and reading components of the Tiberian biblical tradition in terms of 2FS morphology, one encounters slight deviation within broad uniformity. Consider Table 1. Table 1: 2FS morphological variety in the MT1 pronoun (‫א ְּת‬, ַ ‫*א ִתי‬, ַ ‫)אתי‬ harmony -C 50 -CV 0 dissonance ketiv -CV, qere -V 7 verbal ending (‫ ְּת‬-, ‫ ִתי‬-, ‫תי‬-) 199 6 17 nominal suffix (‫ְך‬-, ‫ ִכי‬-, ‫כי‬-) 1545 11 5 Table 1 demonstrates that in the case of all of the categories of 2FS morphology under discussion, the dominant scenario is one of written-reading agreement on consonant-final morphology, 1 For detailed reference lists, see below, §5.1. Cf. the comparable, but slightly different figures given in Hornkohl (2013, 114). © 2023 Aaron D. Hornkohl, CC BY-NC 4.0 https://doi.org/10.11647/OBP.0310.07 146 The Historical Depth of the Tiberian Reading Tradition i.e., ‫א ְּת‬, ַ ‫ ְּת‬-, and ‫ְך‬-. Instances of written-reading dissonance, in the form of ketiv-qere mismatches, occur in all categories, though with very different relative frequencies. The incidence of verbal ‫ אתי‬in place of ‫ ַא ְּת‬and of verbal ending ‫תי‬- in place of ‫ ְּת‬- is relatively high in comparison to that of the nominal suffix ‫כי‬- in place of ‫ְך‬-. Interestingly, when it comes to both the verbal ending and the nominal suffix, the ketiv forms are not the sole evidence of vowel-final 2FS morphology. They are confirmed by cases of apparent vowel-final 2FS morphology where the written and reading components of the tradition agree. While the vowel-final occurrences of the nominal suffix ‫ ִכי‬- are unambiguous, those of the verbal ‫ ִתי‬- merit note. In all seven of these cases, it is possible that the preservation of vowel-final forms in the reading component of the tradition owes to their having been interpreted as cases of 1CS morphology.2 Also relevant are 2FS suffix conjugation forms with object suffixes; a majority of these have an -i- linking vowel before the suffix, which is sometimes represented by a mater yod in the tradition’s corresponding written component (see Hornkohl 2013a, 112, fn. 17, for detail). 2.0. Beyond the Tiberian Tradition 2.1. Biblical Hebrew Material Non-Tiberian biblical material also presents dedicated 2FS morphology. In the traditions represented by this material, vowel- 2 Cf. the Syriac and TJ at Judg. 5.7, 7; Jer. 2.20, 20; the Greek, Syriac, and TJ at Ezek. 16.50; and the Vulgate at Mic. 4.13. 7. The 2FS Endings 147 final endings dominate to the near exclusion of consonant-final forms—the latter of which are, however, occasionally attested. The Samaritan tradition displays its own mixture of forms and traditions (see Ben-Ḥayyim 2000, 107–8, 225–26, 228). Table 2: 2FS morphological variety in the SP3 pronoun (‫ אתי‬atti/åtti) verbal ending (‫ת‬- -ti, ‫תי‬- -ti) harmony dissonance -C -CV written -C, reading -CV 0 7 0 0 5 6 nominal suffix (‫(י)ך‬- -k, ‫כי‬- -ki) 54 1 0 The independent subject pronoun is written ‫ אתי‬and realised atti/åtti, i.e., both the written and reading components of the tradition attesting vowel-final morphology.4 According to the written component of the Samaritan tradition, the verbal ending varies between consonant-final ‫ת‬- and vowel-final ‫תי‬-, but in the reading component it is consistently vowel-final -ti. Conversely, the 2FS nominal suffix is written ‫(י)ך‬- and pronounced with no final vowel, despite written-reading agreement on vowel-final morphology in a single case of ‫כי‬- -ki: ‫ מלכי‬ma ̊̄ līki ‘what troubles you (FS)?’ (Gen. 21.17).5 The scrolls from the Judaean Desert also exhibit variety when it comes to the relevant 2FS forms. 3 For detailed reference lists, see below, §5.2. Cf. the comparable, but slightly different figures in Hornkohl (2013, 118, fn. 28). 4 The apparent exception ‫ ואת‬wit (Num. 5.20) is analysed as a demon- strative (Ben-Ḥayyim 2000, 226, §3.1.3, 237–38, §3.3.1.3). 5 Similar to Aramaic and RH, SH routinely distinguishes between the 2MS and 2FS nominal suffixes via the quality of the vowel that links the noun to the suffix (Ben-Ḥayyim 2000, 228–29, §§3.2.2–3.2.2.1). 148 The Historical Depth of the Tiberian Reading Tradition Table 3: 2FS morphological variety in the BDSS6 1QIsaa Other BDSS -C -CV -C -CV pronoun (‫את‬, ‫)אתי‬ 0 3 7 0 verbal ending (‫ת‬-, ‫תי‬-) 12 18 23 2 nominal suffix (‫ך‬-, ‫כי‬-) 217 27 179 1 While the independent subject pronoun is written ‫ אתי‬in the Great Isaiah Scroll (against ‫ ַא ְּת‬in MT Isaiah), other biblical scrolls present ‫את‬: the latter include parallels to cases of Tiberian writtenreading agreement on ‫א ְּת‬, ַ parallels to Tiberian qere forms against ketiv ‫אתי‬, and parallels to Tiberian forms that graphically resemble ‫אתי‬. Likewise with the verbal ending: 1QIsaa, which accounts for 30 of the 45 extant cases, has 12 instances of ‫ת‬- and 18 of ‫תי‬(all ‫ ְּת‬- in the MT); in the rest of the biblical scrolls, there are 23 instances of ‫ת‬- and just two of ‫תי‬- (all but one of which parallel ‫ ְּת‬in the MT, the exception a ketiv-qere discrepancy where the DSS = ketiv). In the case of the nominal suffix, the biblical scrolls show 395 cases of ‫ך‬- and 28 cases of ‫כי‬-. Again, however, there is a distinction between 1QIsaa and the other biblical scrolls. In 1QIsaa, cases of ‫ך‬- outnumber those of ‫כי‬- by a margin of 216 to 27; in the rest of the biblical scrolls, the counts are ‫ך‬- 179, ‫כי‬- 1 (the single case of ‫כי‬- in 4Q84 is parallel to ‫כי‬- in the corresponding Tiberian text: Ps. 116.19; however, the five remaining instances of ‫כי‬- in MT Ps. 103.3–5 are paralleled by ‫ך‬- in 4Q84). 6 For detailed reference lists, see below, §5.3. Cf. the comparable, but slightly different figures in Hornkohl (2013, 118, fn. 27). 7. The 2FS Endings 149 Precious few examples come in Greek and Latin transcriptional material.7 The lone extant case of the verbal ending is vowel-final: Jerome’s carathi || MT ‫‘ וְּ ָק ָ ֵ֥ראת‬and you will call’ (Isa. 7.14). There is more substantial evidence for the 2FS nominal suffix, all of it indicating consonant-final morphology: Theodotion’s Ἐλωαίχ ‘your God’ || MT ‫ֹלהיָך‬ ַֽ ֶ ‫‘ ֱא‬your (MS) God’ (Mic. 6.8); Je- rome’s semmathech || MT ‫‘ ַצ ָמ ֵ ָ֧תְך‬your veil’ (Isa. 47.2); Jerome’s gebulaic || MT ‫בּולָׂ֑יִ ְך‬ ָ ְּ‫‘ ג‬your borders’ (Ezek. 27.4); Jerome’s bonaich || MT ‫‘ ב ָ֕ ַֹניִ ְך‬your builders’ (Ezek. 27.4). Transcriptions of the 2FS independent pronoun are evidently unattested. 2.2. Extra-biblical Hebrew Material Iron Age epigraphy is entirely lacking in 2FS morphology. The same is true of BS. In the NBDSS, the picture is similar to that of the BDSS, excluding 1QIsaa (see above, §2.1). Table 4: 2FS morphological variety in the NBDSS8 -C -CV pronoun (‫את‬, ‫)אתי‬ 0 1 verbal ending (‫ת‬-, ‫תי‬-) 2 0 nominal suffix (‫ך‬-, ‫כי‬-) 39 6 Summarising Table 4, the single fragmentary instance of the 2FS independent subject pronoun appears to be vowel-final. The two consonant-final suffix conjugation endings come in a biblical citation where they are also consonant-final. Relatively more data 7 My thanks to Ben Kantor for the citations. 8 For detailed reference lists, see below, §5.4. Cf. the comparable, but slightly different figures in Hornkohl (2013, 118, fn. 28). 150 The Historical Depth of the Tiberian Reading Tradition are available regarding the 2FS nominal suffix: ‫ך‬- outnumbers ‫כי‬by a margin of 39 to six. RH, for its part, is more informative on Second Temple 2FS morphology. In Codex Kaufmann of the Mishna, the 2FS independent pronoun, the verbal ending of the suffix conjugation, and the nominal suffix are consistently consonant-final.9 2.3. Aramaic Aramaic 2FS morphology is summarised in Table 5. Table 5: 2FS morphology in select Aramaic dialects pronoun verb ending nominal suffix — — — BA — DSSA — ‫כי‬- TA ‫את‬/‫אנת‬ ‫ת‬- ‫ך‬- (‫כי‬-) Syriac ‫ ܐܢܬܝ‬ʾat ‫ܬܝ‬- (‫ܬ‬-) -t ‫ܟܝ‬- -k BA has no relevant forms, and DSSA has only ‫כי‬- forms of the 2FS nominal suffix. In TA, the forms in all three categories are generally consonant-final, with a small minority of ‫כי‬- nominal suffixes. Syriac’s written-reading dissonance is well known. The written component reflects ancient vowel-final 2FS morphology in all three categories, but the final vowel goes unpronounced in the reading tradition (and is unrepresented in a minority of cases of the verbal ending). 9 The apparent 2FS ending ‫ ָת‬- in m. Nedarim 10.4b is evidently an error on the part of the vocaliser. As in SH, the 2MS and 2FS nominal suffixes are frequently distinguished by an i-vowel before the latter, often indicated in the spelling by a mater yod. 7. The 2FS Endings 151 2.4. Realisation of 2FS Morphology in the Dead Sea Scrolls The orthographic evidence adduced above concerning the oral realisation of 2FS morphology in the DSS is partially ambiguous. On the one hand, it is reasonable to hypothesise that the mater in forms ending in yod reflects the vowel-final realisation -i (cf., however, the situation in Syriac mentioned above, §2.3). On the other hand, forms ending in ‫ת‬- or ‫ך‬- are variously understood by scholars. Hornkohl (2013, 112) favours assuming “the correspondence of the written and pronunciation traditions, i.e., that orthographic forms ending in a consonant were indeed pronounced without a final vowel.” Against the background of wide- spread gender confusion, Kutscher (1974, 213) raises the possibility that no final vowel was pronounced on the relevant 2FS (and 2MS) forms. At the other extreme, Qimron (2018, 154– 55, 259–60 and fn. 11, 265, 267–68) argues on the basis of mixed usage in single texts or lines that all the relevant 2FS categories consistently ended in some shade of i-vowel (perhaps e), no matter their spelling, in which case consonant-final orthography is merely defective. In light of the statistics given above (§§2.1–2), a nuanced view may be the most plausible. Qimron’s view of consistent vowel-final realisations seems most tenable in the specific cases of the subject pronoun and verbal ending in 1QIsaa. The dominance of consonant-final forms of the independent subject pronoun and verbal ending outside 1QIsaa support the view that consonant-final realisations were the norm in most of the DSS. Regarding the nominal suffix—as vowel-final spellings are rare 152 The Historical Depth of the Tiberian Reading Tradition throughout the DSS, including 1QIsaa, it would appear as though consonant-final realisations were the norm. Though the patterns of phonetic realisation suggested above cannot be absolutely confirmed, they do find support in extant Hebrew pronunciation traditions. 1QIsaa patterns like the combined written-reading tradition of SH, with vowel-final independent subject pronoun ‫ אתי‬ʾatti and verbal ending ‫תי‬- -ti paired with consonant-final nominal suffix ‫ך‬- -k. Throughout the rest of the DSS, the norm would seem to be ‫ את‬ʾat, ‫ת‬- -t, and ‫ך‬- -k, which is in line with the testimony of the combined Tiberian writtenreading tradition. 3.0. Diachronic Considerations The written-reading dissonance concerning 2FS morphology differs from many other situations of dissonance discussed in the present work. First, apparent instances are relatively rare. Second, in contrast to cases in which the reading tradition diverges from the written tradition in agreement with late propagation of an early minority form—e.g., vowel-final 2MS morphology (ch. 6)—in this instance, the consonant-final alternant standardised in the reading tradition appears also to have been the dominant option in the written tradition. More than anything, then, in this case, the departure of the reading component from its written counterpart can be described as one of levelling, whereby minority irregular forms, especially the independent pronoun and the verbal ending, were regularised. Verbal forms that escaped regularisation were evidently read as 1CS forms. When it comes to the nominal suffix, genre is determinative: ketiv ‫כי‬- is normalised to 7. The 2FS Endings 153 qere ‫ְך‬- in prose, but the written and reading components of the Tiberian tradition agree on ‫ ִכי‬- wherever it appears in poetry. On the assumption that the written tradition’s heterogene- ity reflects an earlier linguistic reality than the reading tradition’s more homogenous presentation of 2FS morphology, there is very little information that might aid in dating the latter’s deviation from the former. If the DSS spellings are to be taken at face value—i.e., apparently consonant-final spellings are not in large measure defective and apparently vowel-final spellings are not merely graphic morphological indicators (historical spelling, as in Syriac)—then, with the notable exception of 1QIsaa, they seem to indicate a standardisation of consonant-final 2FS morphology more advanced than what is seen in the written component of the Tiberian tradition, but consistent with the Tiberian reading component. In other words, when it contradicts its written counterpart, the reading component of the Tiberian biblical tradition is more or less in agreement with the normalisation of consonantfinal 2FS morphology dominant in most of the DSS. Of course, it is important to point out that the Tiberian reading tradition’s divergence from the written tradition is not particularly frequent, radical, or innovative. Unless the dominant consonant-final 2FS spellings characteristic of the written tradition are regularly defective, the written tradition itself testifies to the hegemony of the same consonant-final realisations that the reading tradition further standardised. Thus, while the written and reading components of the Tiberian tradition offer ‘windows’ on the chronological development of the spelling and realisation of 2FS morphology, there is relatively little diachronic change to 154 The Historical Depth of the Tiberian Reading Tradition speak of. In the vast majority of cases, the images seen through the two windows are identical; in a minority, the window afforded by the reading component reveals the advance of regularisation, the effects of which are, however, already widespread in the corresponding written component. Finally, it is also important to bear in mind that other factors may have contributed to morphological diversity, e.g., especially, but not exclusively, genre. 4.0. Conclusion In the case of 2FS morphology, the reading component of the Tiberian biblical tradition is rarely out of tune with the corresponding written component. On the view that the tradition of oral realisation was largely fixed by Second Temple times, one might expect that it maintains First Temple conventions while at the same time implementing Second Temple innovations. The innovation in this case was the further expansion of consonant-final 2FS morphology already standard in the written component of the Tiberian biblical tradition. In this way, the Tiberian reading tradition diverges from the corresponding written tradition, but only marginally, and in so doing merely continues the developmental journey already largely accomplished in the written tradition along the same trajectory. 5.0. Citations 5.1. Tiberian Biblical Tradition The following list includes only vowel-final cases of the relevant 2FS morphology, excluding cases of the standard consonant-final forms on which the written 7. The 2FS Endings 155 and reading components of the Tiberian biblical tradition agree. Pronoun— ‫אתי‬: Judg. 17.2; 1 Kgs 14.2; 2 Kgs 4.16, 23; 8.1; Jer. 4.30; Ezek. 36.13. Verbal ending—‫ ִתי‬-: Judg. 5.7, 7; Jer. 2.20, 20; Ezek. 16.50; Mic. 4.13; ketiv ‫תי‬- || qere ‫ ְּת‬-: Jer. 2.33; 3.4, 5; 4.19; 31.21; 46.11; Ezek. 16.13, 18, 22, 31, 31, 43, 43, 47, 51; Ruth 3.3, 4. Nominal suffix—‫ ִכי‬-: Jer. 11.15; Ps. 103.3, 3, 4, 4, 5; 116.7, 7, 19; 135.9; 137.6; ketiv ‫כי‬- || qere ‫ְך‬-: 2 Kgs 4.2, 3, 7, 7; Song 2.13. 5.2. Samaritan Pentateuch Pronoun—‫ אתי‬atti/åtti: Gen. 12.11, 13; 24.23, 47, 60; 39.9. Verbal ending—‫ת‬- -ti: Gen. 16.11, 11; 27.12 (|| MT 1cs ‫אתי‬ ֵ֥ ִ ‫‘ וְּ ֵה ֵב‬and I will bring); Num. 5.19, 20, 20; ‫תי‬- -ti: Gen. 3.13; 16.8; 18.15; 30.15 (|| MT infinitive construct [?] ‫‘ וְּ לָ ַ ָ֕ק ַחת‬and to take’); Num. 22.29. Nominal suffix—‫(י)ך‬- -k: Gen. 3.16, 16, 16, 16, 16; 12.12, 12, 13, 13; 16.6, 6, 6, 9, 10, 11, 11; 20.16, 16, 16; 21.18; 24.14, 17, 23, 43, 45, 60; 25.23, 23; 30.2, 14, 15, 15, 15; 35.17; 38.11, 13, 16, 18; 39.9; Exod. 2.7, 7, 9; Num. 5.19, 19, 20, 20, 20, 21, 21, 21, 21, 22; 22.29; Deut. 33.8; ‫כי‬-: Gen. 21.17. 5.3. Biblical Dead Sea Scrolls In the following lists, the parallel MT form is consonant-final unless otherwise specified. Pronoun—‫את‬: 1Q8 22.22 || MT Isa. 51.10; 2Q17 f1.5 || MT Ruth 3.16; 4Q107 f2ii.7 || MT ‫ ִא ִ ַ֤תי‬Song 4.8; 4Q107 f2ii.7 || MT ‫ ִא ִ ֶ֖תי‬Song 4.8; 6Q4 f15.2 || MT ketiv ‫ אתי‬qere ‫ ַא ְּת‬2 Kgs 8.1; Mur88 17.19 || MT Nah. 3.11; Mur88 17.20 || MT Nah. 3.11; ‫אתי‬: 1QIsaa 42.24 || MT Isa. 51.9; 1QIsaa 42.25 || MT Isa. 51.10; 1QIsaa 42.28 || MT Isa. 51.12. Verbal ending—‫ת‬-: 1Q1 f2.3 || MT Gen. 3.13; 1QIsaa 14.16 || MT Isa. 17.10; 1QIsaa 23.9 || MT Isa. 29.4; 1QIsaa 41.20 || MT Isa. 49.21; 1QIsaa 43.6 || MT Isa. 51.17; 1QIsaa 47.7 || MT Isa. 57.8; 1QIsaa 47.8 || MT Isa. 57.8; 1QIsaa 47.8 || MT Isa. 57.8; 1QIsaa 47.9 || MT Isa. 57.10; 1QIsaa 47.9 || MT Isa. 57.10; 1QIsaa 47.10 || MT Isa. 57.10; 1QIsaa 47.9 || MT Isa. 57.11; 1QIsaa 50.13 || MT Isa. 62.3; 1Q8 20.19 || MT Isa. 47.6; 1Q8 26.8 || MT Isa. 60.5; 1Q8 26.23 || MT Isa. 60.16; 1Q8 26.23 || MT Isa. 60.16; 1Q8 26.27 || MT Isa. 60.18; 1Q8 27.1 || MT Isa. 62.3; 1Q8 27.7 || MT Isa. 62.8; 2Q16 f1ii–4i.8 || MT Ruth 2.19; 2Q16 f5ii–6i.6 || MT Ruth 3.2; 2Q16 f6ii–7.3 || MT Ruth 3.4; 2Q16 f6ii–7.3 || MT Ruth 3.4; 4Q51 f102ii+103–106i.43 || MT 2 Sam. 14.2; 4Q51 f102ii+103–106i.44 || MT 2 Sam. 14.3; 4Q55 f9.4 || MT Isa. 17.10; 4Q56 f8–9.3 || MT Isa. 17.10; 4Q58 11.15 || MT Isa. 57.10; 4Q58 11.15 || MT Isa. 57.10; 4Q58 11.16 || MT Isa. 57.11; 4Q62a f2.4 || MT Isa. 57.8; 4Q72 f34ii+36–43.18 || MT Jer. 31.4; 4Q106 f2ii.14 || MT Song 7.7; Mur88 21.5 || MT Zeph. 3.11; ‫תי‬-: 1QIsaa 14.15 || MT Isa. 7.10; 1QIsaa 17.4 || MT Isa. 22.2; 1QIsaa 39.25 || MT Isa. 47.6; 1QIsaa 39.25 || MT Isa. 47.6; 1QIsaa 39.26 || MT Isa. 47.7; 1QIsaa 39.26 || MT Isa. 47.7; 1QIsaa 39.30 || MT Isa. 47.10; 1QIsaa 40.1 || MT Isa. 47.12; 1QIsaa 40.4 || MT Isa. 47.15; 1QIsaa 41.24 || MT 156 The Historical Depth of the Tiberian Reading Tradition Isa. 49.23; 1QIsaa 43.5 || MT Isa. 51.17; 1QIsaa 43.6 || MT Isa. 51.17; } ̇‫יגעת{י‬ 1QIsaa 47.9 || MT Isa. 57.10; 1QIsaa 47.10 || MT Isa. 57.11; 1QIsaa 47.10 || MT Isa. 57.11; 1QIsaa 49.19 || MT Isa. 60.16; 1QIsaa 49.19 || MT Isa. 60.16; 1QIsaa 50.20 || MT Isa. 62.8; 1Q8 20.20 || MT Isa. 47.7; 4Q72 f47–48ii+51–54.11 || MT ketiv ‫ הלכתי‬qere ‫ ָה ָל ְָּׂ֑כ ְּת‬Jer. 31.21. Nominal suffix—‫(י)ך‬-: 1QIsaa 1.25 || MT Isa. 1.22; 1QIsaa 1.25 || MT Isa. 1.22; 1QIsaa 1.28 || MT Isa. 1.25; 1QIsaa 1.29 || MT Isa. 1.25; 1QIsaa 1.29 || MT Isa. 1.25; 1QIsaa 1.29 || MT Isa. 1.26; 1QIsaa 2.1 || MT Isa. 1.26; 1QIsaa 4.2 || MT Isa. 3.25; 1QIsaa 4.3 || MT Isa. 3.25; 1QIsaa 10.16 || MT Isa. 10.30; 1QIsaa 11.11 || MT Isa. 12.6; 1QIsaa 13.1 || MT Isa. 14.29; 1QIsaa 13.3 || MT Isa. 14.30; 1QIsaa 13.3 || MT Isa. 14.30; 1QIsaa 13.19 || MT Isa. 16.3; 1QIsaa 13.20 || MT Isa. 16.3; 1QIsaa 13.26 || MT Isa. 16.9; 1QIsaa 13.26 || MT Isa. 16.9; 1QIsaa 14.16 || MT Isa. 17.10; 1QIsaa 14.16 || MT Isa. 17.10; 1QIsaa 14.17 || MT Isa. 17.11; 1QIsaa 14.17 || MT Isa. 17.11; 1QIsaa 14.17 || MT Isa. 17.11; 1QIsaa 17.4 || MT Isa. 22.1; 1QIsaa 17.5 || MT Isa. 22.2; 1QIsaa 17.6 || MT Isa. 22.3; 1QIsaa 17.6 || MT Isa. 22.3; 1QIsaa 17.10 || MT Isa. 22.7; 1QIsaa 18.6 || MT Isa. 23.2; 1QIsaa 18.14 || MT Isa. 23.10; 1QIsaa 18.18 || MT Isa. 23.12; 1QIsaa 18.21 || MT Isa. 23.14; 1QIsaa 20.14 || MT Isa. 26.2; 1QIsaa 23.8 || MT Isa. 29.3; 1QIsaa 23.9 || MT Isa. 29.3; 1QIsaa 23.9 || MT Isa. 29.3; 1QIsaa 23.10 || MT Isa. 29.4; 1QIsaa 23.10 || MT Isa. 29.4; 1QIsaa 23.10 || MT Isa. 29.4; 1QIsaa 23.11 || MT Isa. 29.5; 1QIsaa 27.27 || MT Isa. 33.23; 1QIsaa 33.8 || MT Isa. 40.9; 1QIsaa 38.5 || MT Isa. 44.27; 1QIsaa 38.21 || MT Isa. 45.14; 1QIsaa 38.21 || MT Isa. 45.14; 1QIsaa 39.21 || MT Isa. 47.1; 1QIsaa 39.21 || MT Isa. 47.2; 1QIsaa 39.22 || MT Isa. 47.2; 1QIsaa 39.22 || MT Isa. 47.3; 1QIsaa 39.22 || MT Isa. 47.3; 1QIsaa 39.24 || MT Isa. 47.5; 1QIsaa 39.25 || MT Isa. 47.6; 1QIsaa 39.25 || MT Isa. 47.6; 1QIsaa 39.28 || MT Isa. 47.9; 1QIsaa 39.29 || MT Isa. 47.9; 1QIsaa 39.29 || MT Isa. 47.9; 1QIsaa 39.29 || MT Isa. 47.9; 1QIsaa 39.30 || MT Isa. 47.10; 1QIsaa 39.30 || MT Isa. 47.10; 1QIsaa 39.30 || MT Isa. 47.10; 1QIsaa 39.30 || MT Isa. 47.10; 1QIsaa 39.30 || MT Isa. 47.10; 1QIsaa 39.31 || MT Isa. 47.11; 1QIsaa 39.31 || MT Isa. 47.11; 1QIsaa 40.1 || MT Isa. 47.11; 1QIsaa 40.1 || MT Isa. 47.12; 1QIsaa 40.1 || MT Isa. 47.12; 1QIsaa 40.2 || MT Isa. 47.12; 1QIsaa 40.2 || MT Isa. 47.13; 1QIsaa 40.2 || MT Isa. 47.13; 1QIsaa 40.4 || MT Isa. 47.15; 1QIsaa 40.4 || MT Isa. 47.15; 1QIsaa 40.4 || MT Isa. 47.15; 1QIsaa 40.5 || MT Isa. 47.15; 1QIsaa 41.15 || MT Isa. 49.16; 1QIsaa 41.15 || MT Isa. 49.16; 1QIsaa 41.16 || MT Isa. 49.17; 1QIsaa 41.16 || MT Isa. 49.17; 1QIsaa 41.16 || MT Isa. 49.17; 1QIsaa 41.16 || MT Isa. 49.17; 1QIsaa 41.16 || MT Isa. 49.18; 1QIsaa 41.18 || MT Isa. 49.19; 1QIsaa 41.18 || MT Isa. 49.19; 1QIsaa 41.18 || MT Isa. 49.19; 1QIsaa 41.19 || MT Isa. 49.19; 1QIsaa 41.19 || MT Isa. 49.20; 1QIsaa 41.19 || MT Isa. 49.20; 1QIsaa 41.20 || MT Isa. 49.21; 1QIsaa 41.23 || MT Isa. 49.22; 1QIsaa 41.23 || MT Isa. 49.22; 1QIsaa 41.23 || MT Isa. 49.23; 1QIsaa 41.23 || MT Isa. 49.23; 1QIsaa 41.24 || MT Isa. 49.23; 1QIsaa 41.24 || MT Isa. 49.23; 1QIsaa 41.26 || MT Isa. 49.25; 1QIsaa 41.27 || MT Isa. 49.25; 1QIsaa 41.27 || MT Isa. 49.26; 1QIsaa 41.28 || MT Isa. 49.26; 1QIsaa 43.6 || MT Isa. 51.18; 1QIsaa 43.8 || MT Isa. 51.19; 1QIsaa 43.8 7. The 2FS Endings 157 || MT Isa. 51.20; 1QIsaa 43.9 || MT Isa. 51.20; 1QIsaa 43.10 || MT Isa. 51.22; 1QIsaa 43.10 || MT Isa. 51.22; 1QIsaa 43.11 || MT Isa. 51.22; 1QIsaa 43.12 || MT Isa. 51.23; 1QIsaa 43.12 || MT Isa. 51.23; 1QIsaa 43.13 || MT Isa. 51.23; 1QIsaa 43.14 || MT Isa. 52.1; 1QIsaa 43.15 || MT Isa. 52.1; 1QIsaa 43.16 || MT Isa. 52.2; 1QIsaa 43.22 || MT Isa. 52.7; 1QIsaa 43.22 || MT Isa. 52.8; 1QIsaa 44.25 || MT Isa. 54.2; 1QIsaa 44.25 || MT Isa. 54.2; 1QIsaa 44.25 || MT Isa. 54.2; 1QIsaa 44.26 || MT Isa. 54.3; 1QIsaa 44.28 || MT Isa. 54.4; 1QIsaa 44.28 || MT Isa. 54.4; 1QIsaa 45.1 || MT Isa. 54.5; 1QIsaa 45.3 || MT Isa. 54.6; 1QIsaa 45.3 || MT Isa. 54.6; 1QIsaa 45.4 || MT Isa. 54.7; 1QIsaa 45.4 || MT Isa. 54.7; 1QIsaa 45.5 || MT Isa. 54.8; 1QIsaa 45.5 || MT Isa. 54.8; 1QIsaa 45.7 || MT Isa. 54.9; 1QIsaa 45.7 || MT Isa. 54.9; 1QIsaa 45.8 || MT Isa. 54.10; 1QIsaa 45.10 || MT Isa. 54.11; 1QIsaa 45.10 || MT Isa. 54.11; 1QIsaa 45.11 || MT Isa. 54.12; 1QIsaa 45.11 || MT Isa. 54.12; 1QIsaa 45.12 || MT Isa. 54.12; 1QIsaa 45.12 || MT Isa. 54.13; 1QIsaa 45.14 || MT Isa. 54.14; 1QIsaa 45.14 || MT Isa. 54.15; 1QIsaa 45.14 || MT Isa. 54.15; 1QIsaa 45.16 || MT Isa. 54.17; 1QIsaa 47.7 || MT Isa. 57.8; 1QIsaa 47.8 || MT Isa. 57.9; 1QIsaa 47.8 || MT Isa. 57.9; 1QIsaa 47.9 || MT Isa. 57.10; 1QIsaa 47.9 || MT Isa. 57.10; 1QIsaa 47.11 || MT Isa. 57.12; 1QIsaa 47.12 || MT Isa. 57.12; 1QIsaa 47.12 || MT Isa. 57.12; 1QIsaa 47.12 || MT Isa. 57.12; 1QIsaa 47.12 || MT Isa. 57.13; 1QIsaa 47.12 || MT Isa. 57.13; 1QIsaa 47.12 || MT Isa. 57.13; 1QIsaa 49.6 || MT Isa. 60.1; 1QIsaa 49.6 || MT Isa. 60.2; 1QIsaa 49.7 || MT Isa. 60.2; 1QIsaa 49.7 || MT Isa. 60.3; 1QIsaa 49.7 || MT Isa. 60.3; 1QIsaa 49.7 || MT Isa. 60.4; 1QIsaa 49.8 || MT Isa. 60.4; 1QIsaa 49.8 || MT Isa. 60.4; 1QIsaa 49.8 || MT Isa. 60.4; 1QIsaa 49.8 || MT Isa. 60.5; 1QIsaa 49.9 || MT Isa. 60.5; 1QIsaa 49.9 || MT Isa. 60.5; 1QIsaa 49.9 || MT Isa. 60.6; 1QIsaa 49.10 || MT Isa. 60.7; 1QIsaa 49.10 || MT Isa. 60.7; 1QIsaa 49.12 || MT Isa. 60.9; 1QIsaa 49.13 || MT Isa. 60.9; 1QIsaa 49.13 || MT Isa. 60.10; 1QIsaa 49.13 || MT Isa. 60.10; 1QIsaa 49.13 || MT Isa. 60.10; 1QIsaa 49.14 || MT Isa. 60.10; 1QIsaa 49.14 || MT Isa. 60.11; 1QIsaa 49.14 || MT Isa. 60.11; 1QIsaa 49.15 || MT Isa. 60.13; 1QIsaa 49.16 || MT Isa. 60.13; 1QIsaa 49.17 || MT Isa. 60.14; 1QIsaa 49.17 || MT Isa. 60.14; 1QIsaa 49.17 || MT Isa. 60.14; 1QIsaa 49.17 || MT Isa. 60.14; 1QIsaa 49.17 || MT Isa. 60.14; 1QIsaa 49.18 || MT Isa. 60.15; 1QIsaa 49.18 || MT Isa. 60.15; 1QIsaa 49.19 || MT Isa. 60.16; 1QIsaa 49.19 || MT Isa. 60.16; 1QIsaa 49.21 || MT Isa. 60.17; 1QIsaa 49.21 || MT Isa. 60.17; 1QIsaa 49.21 || MT Isa. 60.18; 1QIsaa 49.21 || MT Isa. 60.18; 1QIsaa 49.22 || MT Isa. 60.18; 1QIsaa 49.22 || MT Isa. 60.18; 1QIsaa 49.22 || MT Isa. 60.19; 1QIsaa 49.23 || MT Isa. 60.19; 1QIsaa 49.23 || MT Isa. 60.19; 1QIsaa 49.23 || MT Isa. 60.19; 1QIsaa 49.23 || MT Isa. 60.19; 1QIsaa 49.23 || MT Isa. 60.20; 1QIsaa 49.23 || MT Isa. 60.20; 1QIsaa 49.24 || MT Isa. 60.20; 1QIsaa 49.24 || MT Isa. 60.20; 1QIsaa 49.24 || MT Isa. 60.21; 1QIsaa 50.12 || MT Isa. 62.2; 1QIsaa 50.12 || MT Isa. 62.2; 1QIsaa 50.14 || MT Isa. 62.4; 1QIsaa 50.15 || MT Isa. 62.4; 1QIsaa 50.15 || MT Isa. 62.4; 1QIsaa 50.16 || MT Isa. 62.5; 1QIsaa 50.16 || MT Isa. 62.5; 1QIsaa 50.16 || MT Isa. 62.5; 1QIsaa 50.16 || MT Isa. 62.6; 1QIsaa 50.19 || MT Isa. 62.8; 1QIsaa 50.20 || MT Isa. 62.8; 1QIsaa 50.20 || MT Isa. 62.8; 158 The Historical Depth of the Tiberian Reading Tradition 1QIsaa 50.22 || MT Isa. 62.9; 1QIsaa 50.24 || MT Isa. 62.11; 1QIsaa 53.23 || MT Isa. 66.9; 1Q8 9a.5 || MT Isa. 23.2; 1Q8 17.11 || MT Isa. 41.14; 1Q8 17.12 || MT Isa. 41.15; 1Q8 18.7 || MT Isa. 43.6; 1Q8 18.7 || MT Isa. 43.6; 1Q8 19.9 || MT Isa. 44.27; 1Q8 20.14 || MT Isa. 47.1; 1Q8 20.19 || MT Isa. 47.6; 1Q8 20.20 || MT Isa. 47.7; 1Q8 20.22 || MT Isa. 47.9; 1Q8 20.23 || MT Isa. 47.9; 1Q8 20.23 || MT Isa. 47.9; 1Q8 20.24 || MT Isa. 47.10; 1Q8 20.24 || MT Isa. 47.10; 1Q8 20.25 || MT Isa. 47.11; 1Q8 23.1 || MT Isa. 52.7; 1Q8 23.1 || MT Isa. 52.8; 1Q8 23.29 || MT Isa. 54.3; 1Q8 23.31 || MT Isa. 54.4; 1Q8 23.32 || MT Isa. 54.5; 1Q8 26.4 || MT Isa. 60.1; 1Q8 26.5 || MT Isa. 60.2; 1Q8 26.5 || MT Isa. 60.2; 1Q8 26.6 || MT Isa. 60.3; 1Q8 26.6 || MT Isa. 60.3; 1Q8 26.6 || MT Isa. 60.4; 1Q8 26.7 || MT Isa. 60.4; 1Q8 26.7 || MT Isa. 60.4; 1Q8 26.7 || MT Isa. 60.4; 1Q8 26.8 || MT Isa. 60.5; 1Q8 26.8 || MT Isa. 60.5; 1Q8 26.9 || MT Isa. 60.5; 1Q8 26.9 || MT Isa. 60.6; 1Q8 26.11 || MT Isa. 60.7; 1Q8 26.11 || MT Isa. 60.7; 1Q8 26.13 || MT Isa. 60.9; 1Q8 26.14 || MT Isa. 60.9; 1Q8 26.14 || MT Isa. 60.9; 1Q8 26.15 || MT Isa. 60.10; 1Q8 26.15 || MT Isa. 60.10; 1Q8 26.15 || MT Isa. 60.10; 1Q8 26.16 || MT Isa. 60.10; 1Q8 26.16 || MT Isa. 60.11; 1Q8 26.17 || MT Isa. 60.11; 1Q8 26.18 || MT Isa. 60.12; 1Q8 26.20 || MT Isa. 60.14; 1Q8 26.21 || MT Isa. 60.14; 1Q8 26.21 || MT Isa. 60.14; 1Q8 26.21 || MT Isa. 60.14; 1Q8 26.22 || MT Isa. 60.15; 1Q8 26.22 || MT Isa. 60.15; 1Q8 26.24 || MT Isa. 60.16; 1Q8 26.24 || MT Isa. 60.16; 1Q8 26.26 || MT Isa. 60.17; 1Q8 26.26 || MT Isa. 60.17; 1Q8 26.27 || MT Isa. 60.18; 1Q8 26.27 || MT Isa. 60.18; 1Q8 26.28 || MT Isa. 60.18; 1Q8 26.28 || MT Isa. 60.18; 1Q8 26.28 || MT Isa. 60.19; 1Q8 26.29 || MT Isa. 60.19; 1Q8 26.29 || MT Isa. 60.19; 1Q8 26.30 || MT Isa. 60.20; 1Q8 26.30 || MT Isa. 60.20; 1Q8 27.1 || MT Isa. 62.2; 1Q8 27.2 || MT Isa. 62.4; 1Q8 27.2 || MT Isa. 62.4; 1Q8 27.2 || MT Isa. 62.4; 1Q8 27.3 || MT Isa. 62.4; 1Q8 27.3 || MT Isa. 62.4; 1Q8 27.3 || MT Isa. 62.4; 1Q8 27.4 || MT Isa. 62.5; 1Q8 27.4 || MT Isa. 62.5; 1Q8 27.4 || MT Isa. 62.5; 1Q8 27.4 || MT Isa. 62.5; 1Q8 27.4 || MT Isa. 62.6; 1Q8 27.6 || MT Isa. 62.8; 1Q8 27.7 || MT Isa. 62.8; 1Q8 27.9 || MT Isa. 62.11; 1Q8 27.10 || MT Isa. 62.12; 1Q8 28.19 || MT Isa. 66.9; 2Q13 f9ii–12.4 || MT Jer. 48.28; 2Q13 f9ii–12.8 || MT Jer. 48.32; 2Q13 f9ii–12.9 || MT Jer. 48.32; 2Q14 f1.2 || MT Ps. 103.4; 2Q16 f5ii–6i.2 || MT Ruth 2.22; 2Q16 f5ii–6i.5 || MT Ruth 3.1; 2Q16 f5ii–6i.8 || MT Ruth 3.3; 2Q16 f5ii–6i.8 || MT Ruth 3.3; 2Q17 f1.1 || MT Ruth 3.13; 4Q13 f3i– 4.6 || MT Exod. 2.7; 4Q13 f3i–4.7 || MT Exod. 2.7; 4Q51 2a–d.4 || MT 1 Sam. 1.23; 4Q51 2a–d.5 || MT 1 Sam. 1.23; 4Q53 f2–5i.17 || MT 2 Sam 14.18; 4Q53 f2–5i.18 || MT 2 Sam. 14.19; 4Q56 f8–9.4 || MT Isa. 17.11; 4Q57 f9ii+11+12i+52.14 || MT Isa. 23.10; 4Q57 f41–42.2 || MT Isa. 54.8; 4Q57 f44–47.4 || MT Isa. 54.12; 4Q57 f44–47.7 || MT Isa. 54.15; 4Q57 f44–47.8 || MT Isa. 54.17; 4Q58 2.20 || MT Isa. 47.3; 4Q58 3.2 || MT Isa. 47.9; 4Q58 8.24 || MT Isa. 54.2; 4Q58 8.24 || MT Isa. 54.2; 4Q58 9.7 || MT Isa. 54.8; 4Q58 9.9 || MT Isa. 54.9; 4Q58 11.14 || MT Isa. 57.9; 4Q58 11.16 || MT 57.11; 4Q58 11.17 || MT 57.12; 4Q58 11.18 || MT 57.12; 4Q58 11.18 || MT 57.13; 4Q58 11.18 || MT 57.13; 4Q58 11.18 || MT 57.13; 4Q59 f17–18i+19.5 || MT Isa. 7. The 2FS Endings 159 12.6; 4Q60 f3–6.7 || MT Isa. 1.22; 4Q62a f2.2 || MT Isa. 57.6; 4Q64 f1–5.6 || MT Isa. 29.3; 4Q64 f1–5.6 || MT Isa. 29.4; 4Q66 f1–3.1 || MT Isa. 60.20; 4Q68 f1.4 || MT Isa. 14.30; 4Q69a f1.2 || MT Isa. 54.11; 4Q69a f1.3 || MT Isa. 54.12; 4Q72 f1ii.4 || MT Jer. 4.14; 4Q72 f19–21.8 || MT Jer. 22.21; 4Q72 f19–21.9 || MT Jer. 22.22; 4Q72 f47–48ii+51–54.10 || MT Jer. 31.21; 4Q77 f3.1 || MT Zeph. 3.19; 4Q78 f24–29+48.4 || MT Amos 3.11; 4Q78 f24–29+48.4 || MT Amos 3.11; 4Q82 f3ii+4ii+5–7.11 || MT Hos. 2.22; 4Q84 f15iii+20–22.15 || MT ‫כי‬- Ps. 103.3; 4Q84 f15iii+20–22.16 || MT ‫כי‬- Ps. 103.3; 4Q84 f15iv+21ii– 24.1 || MT ‫כי‬- Ps. 103.4; 4Q84 f15iv+21ii–24.2 || MT ‫כי‬- Ps. 103.4; 4Q84 f15iv+21ii–24.3 || MT Ps. 103.5; 4Q84 f15iv+21ii–24.4 || MT ‫כי‬- Ps. 103.5; 4Q85 f12.5 || MT Ps. 45.11; 4Q86 2.1 || MT Ps. 147.13; 4Q86 2.1 || MT Ps. 147.13; 4Q86 2.1 || MT Ps. 147.13; 4Q86 2.2 || MT Ps. 147.14; 4Q105 f4.5 || MT Ruth 1.15; 4Q106 f2i+3–5.8 || MT Song 4.1; 4Q106 f2i+3–5.8 || MT Song 4.1; 4Q106 f2i+3–5.9 || MT Song 4.2; 4Q106 f2i+3–5.11 || MT Song 4.3; 4Q106 f2i+3–5.11 || MT Song 4.3; 4Q106 f2ii.10 || MT Song 7.4; 4Q106 f2ii.11 || MT Song 7.5; 4Q106 f2ii.13 || MT Song 7.6; 4Q107 f1.2 || MT Song 2.10; 4Q107 f1.2 || MT Song 2.10; 4Q107 f1.6 || MT Song 2.13; 4Q107 f1.9 || MT Song 2.14; 4Q107 f1.9 || MT Song 2.14; 4Q107 f2ii.2 || MT Song 4.1; 4Q107 f2ii.3 || MT Song 4.2; 4Q107 f2ii.5 || MT Song 4.3; 4Q107 f2ii.5 || MT Song 4.3; 4Q107 f2ii.6 || MT Song 4.3; 4Q107 f2ii.6 || MT Song 4.3; 4Q107 f2ii.10 || MT Song 4.9; 4Q107 f2ii.11 || MT Song 4.9; 4Q107 f2ii.11 || MT Song 4.10; 4Q107 f2ii.12 || MT Song 4.10; 4Q107 f2ii.13 || MT Song 4.10; 4Q107 f2ii.14 || MT Song 4.11; 5Q6 f1iv.2 || MT Lam. 4.21; 5Q6 f1iv.4 || MT Lam 4.22; 11Q4 f3b+6.2 || MT Ezek. 5.12; 11Q5 3.8 || MT Ps. 122.2; 11Q5 3.12 || MT Ps. 122.6; 11Q5 3.12 || MT Ps. 122.7; 11Q5 3.12 || MT Ps. 122.7; 11Q5 3.13 || MT Ps. 122.8; 11Q5 14.9 || MT Ps. 135.2; 11Q5 21.1 || MT Ps. 137.9. ‫כי‬-: 1QIsaa 1.25 || Isa. 1.23; 1QIsaa 17.4 || Isa. 22.1; 1QIsaa 33.8 || Isa. 40.9; 1QIsaa 38.22 || Isa. 45.14; 1QIsaa 38.22 || Isa. 45.14; 1QIsaa 38.22 || Isa. 45.14; 1QIsaa 39.26 || Isa. 47.7; 1QIsaa 41.15 || Isa. 49.15; 1QIsaa 41.17 || Isa. 49.18; 1QIsaa 41.28 || Isa. 49.26; 1QIsaa 43.7 || Isa. 51.19; 1QIsaa 43.7 || Isa. 51.19; 1QIsaa 43.12 || Isa. 51.23; 1QIsaa 44.24 || Isa. 54.2; 1QIsaa 45.1 || Isa. 54.5; 1QIsaa 45.1 || Isa. 54.5; 1QIsaa 45.5 || Isa. 54.8; 1QIsaa 45.9 || Isa. 54.10; 1QIsaa 45.12 || Isa. 54.13; 1QIsaa 49.15 || Isa. 60.12; 1QIsaa 50.12 || Isa. 62.2; 1QIsaa 50.13 || Isa. 62.3; 1QIsaa 50.14 || Isa. 62.4; 1QIsaa 50.15 || Isa. 62.4; 1QIsaa 50.16 || Isa. 62.5; 1QIsaa 50.25 || Isa. 62.12; 4Q84 f28i.18 || MT ‫כי‬- Ps. 116.19. 5.4. Non-Biblical Dead Sea Scrolls In the following lists, the parallel MT form is consonant-final unless otherwise specified. Pronoun—‫אתי‬: 4Q223–224 f2ii.11 || Jub. 35.17. Verbal ending—‫ת‬: 4Q169 f3–4ii.10 || Nah. 3.5; 4Q169 f3–4ii.11 || Nah. 3.5. Nominal suffix— ‫ך‬-: 1QM 12.14, 14, 14, 14, 14; 19.6, 6, 6; 4Q168 f1.4 || Mic. 4.10; 4Q169 f3– 4ii.10 || Nah. 3.5; 4Q169 f3–4ii.11 || Nah. 3.5; 4Q169 f3–4ii.11 || Nah. 3.5; 160 The Historical Depth of the Tiberian Reading Tradition 4Q169 f3–4iii.1 || Nah. 3.6; 4Q169 f3–4iii.1 || Nah. 3.6; 4Q169 f3–4iii.1 || Nah. 3.6; 4Q169 f3–4iii.2 || Nah. 3.7; 4Q169 f3–4iii.2 || Nah. 3.7; 4Q169 f3– 4iii.6 || Nah. 3.7; 4Q169 f5.3 || Nah. 3.14; 4Q176 f1–2ii.5 || Isa.49.16; 4Q176 f8–11.6 || Isa. 54.4; 4Q176 f8–11.6 || Isa. 54.5; 4Q176 f8–11.8 || Isa. 54.6; 4Q176 f8–11.8 || Isa. 54.6; 4Q176 f8–11.9 || Isa. 54.7; 4Q176 f8–11.9 || Isa. 54.7; 4Q176 f8–11.10 || Isa. 54.8; 4Q176 f8–11.10 || Isa. 54.8; 4Q176 f8–11.11 || Isa. 54.9; 4Q385a f17a–eii.4, 5, 7; 4Q415 f2ii.2, 5, 7; 4Q492 f1.6, 6, 7; 4Q522 f22–26.5 || Ps. 122.7; ‫כי‬-: 4Q161 f5–6.7 || Isa. 10.30; 4Q176 f8–11.6 || Isa. 54.4; 4Q176 f8–11.7 || Isa. 54.5; 4Q176 f8–11.12 || Isa. 54.10; 4Q176 f50.1; 4Q223– 224 f2i.47 || Jub. 35.8. 8. THE QERE PERPETUUM ‫ִהוא‬ In the majority of sources that represent ancient Hebrew traditions, the 3FS independent subject pronoun is written with medial yod, e.g., DSS )‫היא(ה‬. Likewise, in extant pronunciation traditions, it is realised with a corresponding i-vowel, e.g., standard Tiberian (non-Pentateuchal) BH and RH ‫היא‬, ִ SH ī. The written component of the Tiberian tradition of the Pentateuch, exhibiting the spelling ‫הוא‬, is an outlier. Whereas the combined Tiberian written-reading tradition in the MT Prophets and Writings routinely exhibits the unified consonantal-vocalic form ‫( ִהיא‬in 282 of 286 cases), in the Torah such unity is rare (just 18 of 212 cases).1 Instead of ‫היא‬, ִ the anomalous graphic spelling-vocalic combination ‫ ִהוא‬is normative in the Tiberian Pentateuch. 1.0. The Tiberian Tradition On four occasions in the Hebrew Bible, readers are explicitly instructed via the (inter)marginal ketiv-qere mechanism to read 3FS ‫ ִהיא‬instead of apparently 3MS written ‫( הוא‬Deut. 13.16; 1 Kgs 17.15; Isa. 30.33; Job 31.11). In five additional cases, the ketivqere gives the opposite instruction, that is, to read 3MS ‫ הּוא‬for the apparent 3FS spelling ‫( היא‬1 Kgs 17.15; Ps. 73.16; Job 31.11; Qoh. 5.8; 1 Chron. 29.16).2 Finally, in 192 instances in the Pentateuch 1 The figures given here are representative, but scholars differ on their counts. Throughout the MT, written-reading agreement on ‫ ִהיא‬obtains in about 300 out of 500 instances. 2 Thus, 1 Kgs 17.15 and Job 31.11 each involve both changes. © 2023 Aaron D. Hornkohl, CC BY-NC 4.0 https://doi.org/10.11647/OBP.0310.08 162 The Historical Depth of the Tiberian Reading Tradition and once in the Prophets, the written form ‫ הוא‬is vocalised with ḥiriq to signal the qere perpetuum ‫הוא‬. ִ 3 As already noted, in the Tiberian Pentateuch, the orthography and vocalisation agree on the realisation of 3FS ‫ ִהיא‬just 18 times in 210 cases (see §5.1 for citations). Scholarly explanations for the routine written-reading mismatch in the Tiberian Pentateuch vary from the graphic to the linguistic. According to one widely accepted version of the graphic approach, the Tiberian Torah ultimately goes back to a manuscript characterised by defective spelling, where both the 3MS and 3FS independent subject pronouns were originally written ‫( הא‬cf. the 3MS forms in Arad 18.10, 12; Kuntillet Ajrud 9.1; Lachish 21.5; Meshaʿ [KAI 181] 6, 21; Deir ʿAlla [KAI 312] 1). Into this form in a manuscript of the proto-Masoretic tradition, so it is claimed, a scribe mechanically inserted mater waw, not realising that ‫ הא‬often represented the 3FS independent pronoun (GKC §32l). In a variation of the same approach, the scribe attempted to distinguish the two pronouns, but wrote waw and yod so similarly (a practice common in the DSS), that later copyists, unable to discern any difference, reproduced waw on all occasions. Even later copyists, loathe out of respect for the manuscript to modify the apparent 3FS ‫ הוא‬spellings, left them uncorrected (Cross 1998, 222–23; JM §39c). Neither explanation accounts for the Masoretic Pentateuch’s 18 exceptions in which the written and reading traditions agree on 3FS ‫( ִהיא‬Fassberg 2012, 171–72). 3 Rendsburg (1982, 353) gives the figure 120, which is repeated by Fassberg (2012, 171). 8. The Qere Perpetuum ‫ִהוא‬ 163 A well-known linguistic proposal is that the Hebrew of the Tiberian Torah preserves an epicene 3CS pronoun ‫ ה(ו)א‬hū (Green 1872, 96; Lambert 1946, 34, fn. 3; Rendsburg 1982; Tropper 2001; Morgenstern 2007, 49–50). The spelling in the Tiberian Pentateuch would thus preserve an old feature that is out of line with the corresponding Pentateuchal recitation tradition as well as with the combined written-reading tradition of the rest of the Masoretic Bible. According to recent versions of this approach, the explanation for the epicene pronoun in the Pentateuch is Hittite or Hurrian influence (Rendsburg 1982) or a single 3CS oblique pronoun ‫[ הוא‬huʾā] (<*huʾat̄̆ ) (as opposed to distinct 3MS and 3FS nominative pronouns) (Tropper 2001). The problems with approaches of this sort are that (a) the alleged feature is not known outside the written component of the Tiberian tradition as preserved in the Pentateuch; (b) the Semitic languages commonly distinguish 3MS and 3FS pronouns; and, perhaps most decisively, (c) Tiberian BH grammar, e.g., the verbal system, pronominal suffixes, including that reflected in the written component of the tradition in the Pentateuch, consistently reflects gender distinction in the 3rd-person singular. The current chapter takes as its jumping-off point a different sort of linguistic hypothesis. As suggested by Cohen (2007, 113–15) and buttressed by Fassberg (2012), the ‫ הוא‬spelling common to the 3MS and 3FS independent subject pronouns in the written component of the Tiberian tradition reflects distinct morphological forms, namely 3MS *huwa or *hūw and 3FS *hiwa or *hīw, which in the corresponding Pentateuchal reading tradition, 164 The Historical Depth of the Tiberian Reading Tradition and the Masoretic biblical reading tradition more generally, shortened to hū and hī, respectively (see further below, §3.0). 2.0. Non-Tiberian Biblical and Extra-biblical Evidence Beyond the Tiberian biblical written and pronunciation evidence, it is instructive to consider additional ancient Hebrew evidence. The rather opaque inscriptional 3MS form ‫ הא‬has already been cited. The quality of its medial vocalisation is uncertain, as is the presence, quality, and quantity of a final vowel (though final long vowels are generally thought to have been marked in ancient inscriptional Hebrew). No 3FS form is attested in the extant epigraphic corpus. Babylonian Torah manuscripts know the same phenomenon seen in the Tiberian Pentateuch. Yeivin (1985, 1103) notes the written-reading mismatch in a vocalised Babylonian manuscript at Deut. 11.10. In DSS Hebrew, alongside the more standard spellings ‫הוא‬ and ‫ היא‬come ‫ הואה‬and ‫היאה‬, respectively (Qimron 1986, 57–58; 2018, 261–62; Reymond 2014, 158). The two sets of forms occur in both biblical and non-biblical manuscripts, the former more frequently than the latter. Crucially, where the written component of the Tiberian biblical tradition has 3FS ‫הוא‬, corresponding DSS manuscripts usually have ‫( היא‬or ‫ הי‬or ‫)היאה‬, showing agreement with the qere perpetuum of the recitation tradition (see §5.2 8. The Qere Perpetuum ‫ִהוא‬ 165 for citations).4 A minority of BDSS manuscripts appear to match the Tiberian written tradition with 3FS ‫( הוא‬see §5.2 for citations; but cf. Reymond 2014, 158). The combined written-reading tradition of the SP furnishes important information. The written component of the tradition, as evidenced in the Shechem Synagogue Ms 6 (C), consistently has ‫ היא‬against Tiberian written 3FS ‫( הוא‬Ben-Ḥayyim 2000, 226, §3.1.4). This is in agreement with the Samaritan pronunciation tradition, according to which ‫ היא‬is realised as ī. In BS manuscripts from antiquity and the Middle Ages, 3FS ‫ היא‬is consistently distinguished from 3MS ‫הוא‬. The same is true for the Tannaitic RH tradition of Codex Kaufmann of the Mishna, where the form is ‫היא‬. ִ Most of the evidence cited in this section shows Second Temple unanimity regarding a realisation of the 3FS independent subject pronoun in line with the standard non-Pentateuchal Tiberian orthography ‫היא‬. According to a straightforward reading of the data, the Tiberian reading tradition of the Torah joins in with the combined Tiberian written and reading tradition of the rest of the Bible and with various Second Temple traditions on pronunciation resembling hī, including hiʾā, hiyā, and ī. 4 This assumes that the relevant editor has correctly distinguished waw and yod in texts where the distinction can be anywhere from minimal to non-existent. 166 The Historical Depth of the Tiberian Reading Tradition 3.0. A Linguistic Explanation for 3FS ‫ הוא‬in the Written Component of the Tiberian Torah Both internal and external evidence militate against the theory that apparently 3FS ‫ הוא‬in the written component of the Tiberian Torah reflects an epicene 3CS pronoun. Beyond the fact that the Semitic languages, in general, and ancient Hebrew, more specifically, routinely distinguish gender in the 3rd-person singular, the Tiberian written tradition of the Torah reflects gender distinction in 3rd-person singular morphology, including pronominal suffixes and the verbal system. An epicene 3rd-person singular independent pronoun would thus from multiple perspectives be exceptional. Explanations based on the graphic similarity of waw and yod are also probably to be rejected, since they fail to account for the generally correct distinction between waw and yod in other words in the Tiberian Torah and leave a number of cases of standard ‫ ִהיא‬unexplained. If the 3FS ‫ הוא‬spelling is not to be attributed to graphic factors, a different sort of the linguistic explanation must be sought. As mentioned above, Cohen (2007, 113–13) has proposed an intriguing alternative. In his view, development of the standard Tiberian 3FS independent subject pronoun ‫ ִהיא‬may be schematised as follows (Cohen 2007, 114–15): 1 2 3 4 5 *hiʾa-tu > *hiʾat > *hiʾa > *hiwa > *hiya > It is worth quoting Cohen in full: 6 *hiy 7 > hī 8. The Qere Perpetuum ‫ִהוא‬ 167 According to this hypothesis, it appears that the ketiv and the qere before us—‫הוא‬/‫—היא‬are ִ in fact nothing but different forms of the same 3FS pronoun, testifying to different stages of development in the form of this pronoun (stage 4 *hiwa [=*‫]הוַ א‬ ִ and final stage 7 hī [=‫)]היא‬, ִ and it is not impossible that these two forms, which were a sort of doublet in Hebrew, served contemporaneously in two parallel linguistic traditions. (Cohen 2007, 115, my translation) This approach has the advantage of making sense of the otherwise anomalous 3FS spelling ‫הוא‬. Moreover, it is not incompatible with the minority DSS spelling ‫היאה‬, which can be viewed as the retention of a comparatively archaic form (Qimron 1986, 57–58; 2018, 261–62; cf. Kutscher 1974, 433–34). In allowing for the contemporaneity of the two pronunciations, it also comprehends diversity both within and beyond the Torah. Finally, the typologically later hī realisation in the Tiberian reading component of the Torah is consistent with the combined written-reading tradition in the rest of the Hebrew Bible, apparently reflecting standardisation of a Second Temple feature with early roots as a minority form. Yet, Cohen’s approach is not without problems. Fassberg (2012, 175, fn. 13) observes that the conjectured development from stage 3 *hiʾa to stage 4 *hiwa is unexpected, a y glide being expected contiguous to an i-vowel, as in Arabic َ‫ ِهي‬hiya. If *hiwa or *hīw (Fassberg 2012, 177) are behind the spelling of 3FS ‫הוא‬ in the Tiberian Torah, then one must assume that the unexpected shift of -iʾa to -iw(a) was motivated by analogical pressure from the more common corresponding 3MS form, where the development *huʾa to *huwa is expected. 168 The Historical Depth of the Tiberian Reading Tradition Fassberg (2012, 177) also entertains the possibility that 3FS ‫ הוא‬in the Tiberian written tradition of the Pentateuch reflects the realisation *hū, apparently not as an original epicene pronoun, but as a result of phonetic neutralisation, presumably along the lines of *hiwa > *hiw > hū. In any case, it may be that Cohen’s proposed scheme should be reordered and modified to allow for parallel developments, i.e., 5a *hiy 1 2 3 6a  hī 4 7a  ī 7b *hiʾa-tu  *hiʾat  *hiʾa  *hiya *hū 5b 6b *hiwa  *hiw 7c hī According to this revised scheme, the Tiberian reading tradition reflects stage 6a, the DSS stages 3, 4, and/or 6a, the Samaritan reading tradition 7a, and the Tiberian written tradition of the Torah 5b, 6b, or 7b (with the passage from stage 4 to 5b due to the aforementioned analogy to 3MS *huʾa > *huwa). It is also not impossible that the 3FS pronunciation hī in the Tiberian Torah in 7c (= 6a) could have developed naturally from *hiw. While the diphthong iw is expected to resolve to ū, the alternative development to ī is not unknown (Blau 2010, 97, §3.4.3.3).5 5 It is worth noting that according to the approaches adopted here, the earliest form included a glottal stop, the orthographic representation of which persisted despite its eventual elision. Also, the early form begin- 8. The Qere Perpetuum ‫ִהוא‬ 169 4.0. Conclusion On the assumption that the spelling of 3FS ‫ הוא‬in the Tiberian Pentateuch represents a linguistic reality different from ‫ ִהיא‬of the Tiberian reading tradition, it would not be surprising that it preserves an authentically old variant pronunciation, nor that it should be replaced in the reading tradition by a rival ancient form that became common in Second Temple Hebrew. As a conservative linguistic tradition, the Tiberian recitation component preserves genuine Iron Age features. But as a tradition that crystallised in the Second Temple Period, it was also subject to the standardisation of certain Second Temple conventions. 5.0. Citations 5.1. Tiberian Biblical Tradition 3FS ‫הוא‬: Gen. 2.12; 3.12, 20; 4.22; 7.2; 10.11, 12; 12.14, 18, 19; 14.7, 8; 17.14; 19.20, 38; 20.2, 3, 5, 5, 12; 21.22; 22.20, 24; 23.2, 15, 19; 24.44; 25.21; 26.7, 9, 9, 12, 12; 27.38; 29.2, 9, 25; 32.19; 34.14; 35.6, 19, 20, 22, 27; 37.32; 38.1, 14, 16, 21, 25; 43.32; 47.6, 17, 18; 48.7; Exod. 3.8; 8.15; 12.15, 19; 22.26, 26; 31.13, 14, 14, 17; Lev. 2.6, 15; 5.12; 6.2, 10, 18, 22; 7.20, 21, 27; 10.12, 13, 17; 11.6, 6, 26; 13.4, 8, 11, 20, 22, 23, 25, 25, 26, 28, 28, 28, 42, 52, 55, 57; 14.44; 15.3, 23, 25; 17.11, 14; 18.7, 8, 11, 12, 13, 14, 15, 16, 17, 22; 19.8, 20; 20.6, 14, 21; 22.3, 12; 23.3, 30, 36; 25.10, 11, 12, 33; 27.4; Num. 5.6, 13, 14, 18, 28, 31; 8.4; 9.13; 13.18, 19, 20, 27, 32; 14.8, 41; 15.25, 30, 31; 18.19; 19.9, 13, 20; 21.16, 26; 22.4; 32.4; 33.36; Deut. 1.9, 16, 18; 2.20, 34; 3.4, 8, 11, 12, 18, 21, 23; 4.6, 14; 5.5; 9.19, 20; 10.1, 8, 10; 11.10; 14.28; 17.5; 20.20; 21.3, 4, 6; 22.18, 24; 24.4; 29.21, 26; 30.11, 11, 12, 13; Isa. 39.1. 3FS ‫היא‬: Gen. 14.2; 19.20; 20.5; 26.7; 38.25; 40.10; Exod. 1.16; Lev. 5.11; 11.39; 13.6, 10, 21; 16.31; 20.17, 18; 21.9; Num. 5.13, 14. ning with h may well have arisen due to lenition of more archaic š, as in east Semitic. 170 The Historical Depth of the Tiberian Reading Tradition 5.2. Biblical Dead Sea Scrolls 3FS ‫היא‬: 1Q3 f3–4.2 || Lev. 20.11; 1Q13 f23–25.5 || Deut. 11.10; 2Q12 f1.5 || Deut. 10.10; 4Q1 f5.3 || Gen. 35.19; 4Q6 f1.13 || Gen. 48.7; 4Q22 25.7 (2x) || Exod. 22.26 (2x); 4Q22 37.7 || Exod. 31.14; 4Q23 f4.5 || Lev. 14.44; 4Q23 f34ii+44–50.22 || Num. 5.6; 4Q24 f9i+10–17.20 || Lev. 22.12; 4Q24 f9ii+11ii+18–20i.2 || Lev. 23.3; 4Q25 f5.2 || Lev. 5.12; 4Q26b f1.2 || Lev. 7.20; 4Q26b f1.4 || Lev. 7.21; 4Q27 f3ii+5.7 || Num. 13.18; 4Q29 f1–2i+3.16 || Deut. 30.11; 4Q29 f1–2i+3.17 || Deut. 30.13; 4Q30 f12–15.3 || Deut. 11.10; 4Q31 1.15 || Deut. 2.34; 4Q31 2.12 || Deut. 3.23; 4Q33 f17–19.1 || Deut. 21.4; 4Q35 f1.8 || Deut. 1.9; 4Q37 1.6 || Deut. 5.5; 4Q38 f2.9 || Deut. 11.10; 4Q40 f1–3.5 || Deut. 3.21; 4Q41 2.10 || Deut. 5.5; 4Q134 f1.11 || Deut. 5.5; 4Q138 f1.26 || Deut. 11.10; 8Q4 f1.28 || Deut. 11.10; 11Q1 4.7 || Lev. 25.33; XQ3 1.12 || Deut. 5.5. 3FS ‫הוא‬: Mas1b 3.21 (addition) || Lev. 10.17; Mas1b 4.9 || Lev. 11.6; 4Q26 f4.16 || Lev. 17.11; 8Q3 f26–29.19 (2x) || Deut. 11.10. 9. THE 2/3FPL ENDINGS Ancient Hebrew sources exhibit diversity in 2/3FPL morphology, specifically in the endings of 2/3FPL prefix conjugation forms and of FPL imperatives.1 1.0. The Combined Tiberian Biblical Tradition In the majority of cases of 2/3FPL prefix conjugation (way)yiqṭol forms and of FPL imperatival forms, the written and reading components of the Tiberian biblical tradition agree on a vowel-final ending written and vocalised ‫נָ ה‬-. In far fewer cases, they agree on consonant-final endings, such as ‫ ִַן‬- or ‫ ִֶן‬-. In the remaining cases, the orthography and vocalisation diverge, resulting in the graphic representation ָ ‫ן‬- (Andersen and Forbes 1986, 180; Barr 1989b, 127–31).2 See Table 1. 1 Excluded from this discussion are forms of the infinitive construct with 2/3FPL afformatives. While these vary between vowel- and consonantfinal endings, there are no cases of dissonance between the written and reading components of the Tiberian biblical tradition: ‫ ִָן‬-: Gen. 30.38; 2 Sam. 20.3; Ezek. 1.9, 12, 17; 42.12; ‫ ִָנָ ה‬-: Jer. 8.7; Job 39.2; Ruth 1.19, 19. 2 For a succinct discussion of the relevant ancient Hebrew FPL endings in a broader Semitic context, as well as bibliography, see Blau (2010, 203–4, §4.3.3.1.2n). © 2023 Aaron D. Hornkohl, CC BY-NC 4.0 https://doi.org/10.11647/OBP.0310.09 172 The Historical Depth of the Tiberian Reading Tradition Table 1: 2/3FPL endings according to the written and reading components of the Tiberian biblical tradition (see §5.1 for references) ‫נָ ה‬- ‫ ִַן‬-/‫ ִֶן‬- ָ ‫ן‬- prefix conjugation 295 1 37 imperative 17 2 3 In terms of the prefix conjugation, written-reading divergence resulting in the graphic representation ָ ‫ן‬- occurs in 37 of 333 cases. When it comes to the imperative, ָ ‫ן‬- occurs in 3 of 22 cases. The incidence of mismatch between the written and reading components of the Tiberian biblical tradition is not evenly distributed throughout the biblical text. For the 2/3FPL prefix conjugation, see Table 2. Table 2: Distribution of 2/3FPL prefix conjugation forms in Tiberian BH Genesis Exodus Leviticus Numbers Deuteronomy Joshua Judges Samuel Kings Isaiah Jeremiah Ezekiel Hosea Joel Amos ‫נָ ה‬- ‫ ֶֶן‬- ָ ‫ן‬- 15 7 10 11 1 3 5 15 8 37 29 58 4 1 3 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 12 11 0 1 2 0 0 3 0 0 0 7 0 0 0 Obadiah Jonah Micah Zechariah Malachi Psalms Job Proverbs Ruth Song of Songs Lamentations Esther Daniel Nehemiah Chronicles TOTALS ‫נָ ה‬- ‫ ֶֶן‬- ָ ‫ן‬- 1 0 4 9 1 20 12 10 16 1 3 2 4 1 4 295 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 37 As can be seen in the table, instances of Tiberian written and reading dissonance reflected in the consonant-vowel combina- 9. The 2/3FPL Endings 173 tion ָ ‫ן‬- congregate appreciably in the Pentateuch, where, indeed, they account for more than a third of the cases (especially in Genesis and Exodus). In Samuel, one-sixth of the 18 cases show ָ ‫ן‬-, while Ezekiel, with far more 2/3FPL prefix conjugation forms than any other book, has an incidence of just over one in ten. Turning to FPL imperatival forms, consult Table 3. Table 3: Distribution of FPL imperatival forms in Tiberian BH Genesis Exodus Samuel Isaiah ‫נָ ה‬- ‫ ֶֶן‬-/‫ ִַן‬- ָ ‫ן‬- 1 0 1 3 1 1 0 0 0 0 0 0 Jeremiah Ruth Song of Songs TOTALS ‫נָ ה‬- ‫ ֶֶן‬-/‫ ִַן‬- ָ ‫ן‬- 6 4 2 17 0 0 0 2 0 3 0 3 Though a dearth of data precludes certainty, a few tentative observations may be ventured. First, the variety of forms in Genesis and Exodus is consistent with what was seen above in conjunction with the prefix conjugation. The lack of any consonantvowel mismatch may be due to the rarity of the forms. Second, the dominance of vowel-final orthography and realisation throughout the rest of the Bible also tallies with the distribution of the prefix conjugation. The outlier is Ruth, where, similar to the case of Ezekiel noted above with regard to the prefix conjugation, a relatively high concentration is characterised by a degree of diversity. Focusing on the Torah, the variation does not appear to be a function of putative source. On the basis of the division into sources found in Friedman (1997, 246–55), the principal reconstructed documents, i.e., J, E, and P, are all characterised by the use of both ‫נה‬- and ‫ן‬-. Indeed, in four places in the Tiberian tradition, twice in the Pentateuch, a verse contains at least one in- 174 The Historical Depth of the Tiberian Reading Tradition stance of each alternant: Gen. 30.38; 37.7; 1 Sam. 18.7; Ezek. 16.55. Also, no phonological or prosodic factor governing the preference for one or the other alternants is apparent. Andersen and Forbes (1986, 180–81) and Barr (1989, 130– 31) agree that the difference between ‫נָ ה‬- and ָ ‫ן‬- is not to be regarded as merely orthographic, but as reflecting diverse pronunciations, the one vowel-final and the other consonant-final. If so, then the consonant-vowel combination ָ ‫ן‬- represents mismatch in the combined written-reading tradition. Since orthographic ‫נה‬cases and -na ̊̄ realisations are the norm, it is reasonable to consider the apparent dissonance in cases of ָ ‫ן‬- a result of the secondary extension of the majority realisation that resulted in the levelling of several non-conforming cases, though their orthography was left unchanged. Admittedly, this is not the only logical explanation. It may simply be that the written and reading components differed in this regard from a very early date, each with a slightly different constellation of forms. The choice between these two approaches is informed via examination of non-biblical and non-Tiberian material. 2.0. Beyond the Tiberian Biblical Tradition 2.1. Non-Tiberian Biblical Material In non-Tiberian biblical material, dedicated FPL morphology is common. Vowel-final endings dominate to the near exclusion of consonant-final forms, which are, however, occasionally attested. The Samaritan tradition exhibits its own internal diversity. First, parallel to the 26 cases of Tiberian 2/3FPL prefix conjugation forms ending in ָ ‫ן‬-, and against the one case with ‫ ִֶין‬- (Gen. 9. The 2/3FPL Endings 175 49.26), the SP generally has ‫נה‬-. Two of the exceptions, along with seven other forms, end in ‫הן‬-.3 Thus, according to the Samaritan written tradition, vowel-final forms outnumber consonant-final forms by a margin of 64 to nine (see §5.2.1 for citations; this compares to the ratio of 44 to 27 in the Tiberian written tradition). Vowel-final forms are even more dominant in the Samaritan reading tradition, where the endings are either -na or, more commonly, -inna (see §5.2.2 for citations).4 Samaritan FPL imperatives present written and oral forms consistent with those found in the Tiberian written tradition— ̊̄ ‫ שמען‬šēˈmān and ‫ האזינה‬azīna (Gen. 4.23); ‫ קראן‬qēˈrīn (Exod. 2.20)—i.e., with no mismatch between the two components of the Samaritan tradition (see §§5.1–2). Turning to material from the Judaean Desert, and focusing on the 2/3FPL prefix conjugation, BDSS material preserves forms ending in both ‫נה‬- and ‫ן‬-, with the former far more common than the latter. Indeed, of the 73 BDSS cases of prefix conjugation forms with a dedicated 2/3FPL ending, just two have ‫ן‬-, one of which parallels ָ ‫ן‬- in the MT. Overall, where the BDSS preserve forms parallel to those in the MT, agreement between the two on the 2/3FPL ending is the norm; see Table 4. 3 SP Exod. 1.10 has ‫ תקראנו‬tiqrānnu against Tiberian ‫ת ְּק ֶ ַ֤ראנָ ה‬. ִ 4 On SH -inna Ben-Ḥayyim (2000, 105) explains as follows: Since the 2nd and 3rd fem. pl. were generally expressed in post-BH by means of 2nd and 3rd masc. pl. forms, the feminine endings may have become somewhat obscure, the doubling of the nun resulted in this case from analogy to forms with object suffixes. 176 The Historical Depth of the Tiberian Reading Tradition Table 4: 2/3FPL prefix conjugation endings in the BDSS and the MT (see §5.3.1 for citations) MT ‫נָ ה‬- MT ָ ‫ן‬- MT Other BDSS ‫נה‬- 66 3 2 BDSS ‫ן‬- 1 1 BDSS Other 2 The BDSS preserve just five FPL imperatival forms, all orthographic matches for the ‫נָ ה‬- forms in the relevant Tiberian parallels. Jerome’s Latin transcriptions of BH include a single case of a 3FPL wayyiqṭol form. The Tiberian ‫‘ וַ יֵ ַ ֶ֖ח ְּמנָ ה‬and they mated’ (MT 30.38) is transcribed iaamena (Kantor 2020, 118–19).5 2.2. Extra-biblical Hebrew Material In the nature of things, no relevant 2/3FPL forms appear in the fragmentary corpus of Iron Age epigraphy. Later extra-biblical material is characterised by replacement of dedicated verbal 2/3FPL morphology with 2/3CPL < 2/3MPL morphology (Qimron 2018, 159–60). Thus, for example, the Hebrew of BS lacks any dedicated 2/3FPL morphology.6 Where the relevant dedicated verbal 2/3FPL morphology is preserved in late extra-biblical Hebrew material, often in citation of the Bible or allusion thereto, it nearly always has vowel-final morphology. 5 See Kantor (2020, 118–22) on the omission of any representation of the waw at the beginning of the transcription of this wayyiqṭol. 6 See, by way of example, ‫‘ עיני אל יראו מעשיו‬the eyes of God will see his deeds’ (SirA 6r.29 [Sir. 15.19]); ‫‘ על כן ענו לו בנות ⟦ ⟧ ויכנוהו ברבבה‬for this reason the young women sang to him among ten thousand’ (?; SirB 16v.11 [Sir. 47.6]). 9. The 2/3FPL Endings 177 When it comes to non-biblical material from the Judaean Desert (including that categorised as rewritten Bible), FPL ‫נה‬dominates to the total exclusion of ‫ן‬-. This is true of both the 2/3FPL prefix conjugation and the FPL imperative (see §5.3.2 for citations). Given the shift in RH from dedicated 2/3FPL morphology to 2/3CPL morphology, the Mishna (as represented by Codex Kaufmann) exhibits very few relevant cases. Of the mere nine, eight come in biblical citations, all with ‫נָ ה‬- in both sources (see §5.4 for citations). In another case, the (unvocalised) phrase ‫עד שתכהין‬ ‫‘ עיניו‬before his eyes darken’ (m. Peʾa 8.9) is part of an interlinear addition. The three FPL forms that end in ‫ן‬- in m. Ketubbot 4.11 are in Aramaic. The Mishna also includes five FPL imperative forms, all ending with ‫נָ ה‬-, four of which are direct biblical quotations, with the fifth (m. Nedarim 9.10a) an explicit allusion (see §5.4 for citations). 2.3. Aramaic Material Though it is of questionable relevance, FPL prefix conjugation morphology in BA, DSSA, TA, and Syriac is consistently consonant-final. The FPL imperative is unattested in BA and DSSA, is consonant-final in Syriac, and varies in TA, e.g., ‫‘ שמעא‬listen!’ (Gen 4.23); ‫‘ קרין‬call!’ (Exod. 2.20); ‫‘ איזילנא תובנא‬go, return!’ (Ruth 1.8). 3.0. Diachronic Considerations Based on the non-Tiberian and extra-biblical data surveyed above, it is reasonable to hypothesise that the diversity seen in 178 The Historical Depth of the Tiberian Reading Tradition ancient Hebrew sources, especially in the orthography of the Tiberian written tradition in the Pentateuch, is representative of early diversity, whereby FPL morphology in both the prefix conjugation and the imperative was alternatively vowel- or consonant-final. Even the Tiberian reading tradition preserves a degree of diversity in the form of rare consonant-final FPL imperatives, which are, again, limited to the Pentateuch. Be that as it may, it is difficult to ignore the fact that, by and large, the Tiberian pronunciation tradition patterns like Second Temple Hebrew sources when it comes to FPL verbal morphology, standardising the vowel-final alternant reflected in the majority ‫נה‬- spelling, even where the orthography ‫ן‬- most likely reflects an original consonant-final ending. Whether differentiation between Hebrew and Aramaic FPL morphology played any role in the late standardisation of vowelfinal FPL verbal morphology is unclear. While the Tiberian reading tradition both diverges from the apparently early diversity preserved in the written tradition and shows close affinity to Second Temple sources in its levelling of FPL verbal morphology, it is worth emphasising that the specific form that became the standard is not itself an exclusively late feature, but is already common, if not dominant, in the earliest Hebrew evidence. This scenario is in line with the view that the recitation component of the Tiberian biblical tradition crystallised in the Second Temple Period, extending certain late conventions, but at the same time preserves minority Iron Age features. 9. The 2/3FPL Endings 179 4.0. Conclusion The reading component of the Tiberian biblical tradition shows not infrequent dissonance in comparison to the corresponding written component in the case of 2/3FPL verbal endings. In accord with the supposition that the reading component’s development was largely complete by the Second Temple Period, it should come as no surprise that it exhibits both affinity with the corresponding written component, via use of a feature well attested therein, and simultaneously diverges therefrom in agreement with Second Temple material in the standardisation of vowel-final 2/3FPL verbal morphology. 5.0. Citations 5.1. Tiberian Biblical Tradition Prefix conjugation—‫נָ ה‬-: Gen. 3.7; 24.61, 61; 30.38; 31.14; 37.7; 41.2, 3, 4, 7, 18, 20, 21, 53, 54; Exod. 1.10; 2.16, 16, 16, 18; 8.5, 7; Lev. 4.2, 13, 22, 27; 5.17; 7.30; 10.19; 23.15, 17, 17; Num. 27.1, 2; 35.11, 13, 14, 15; 36.3, 4, 6, 6, 11; Deut. 1.44; Josh 17.4; 21.42; 24.7; Judg. 5.26, 29; 7.11; 11.40; 15.14; 1 Sam 3.11; 4.20; 6.12; 7.14; 9.3, 12, 12; 10.7; 14.27; 18.6, 7; 2 Sam. 1.20, 20; 2.7; 20.3; 1 Kgs 3.16, 16, 22; 10.7; 2 Kgs 2.24, 24; 21.12; 22.20; Isa. 3.16, 16, 16; 5.15; 11.7; 13.7, 16, 18; 16.2; 17.2, 7; 27.11; 28.3; 29.18; 30.21; 32.3, 3, 10; 33.17, 17, 20; 35.5, 5; 41.22; 42.9; 44.7, 26; 47.9; 48.3; 49.15, 22; 54.10; 60.4, 8; 65.17, 17; 66.14; Jer. 4.7; 9.16, 16, 17, 17, 17; 14.17, 17; 18.21; 19.3; 24.2, 3, 8; 29.6, 17; 31.29, 30; 32.4; 33.13; 34.3; 44.6, 25, 25, 25; 48.6, 9; 49.2, 13; 50.20; Ezek. 1.24, 25; 6.6, 6; 7.17, 17, 27; 12.20; 13.11, 18, 18, 19, 19, 19, 23, 23; 16.50, 50, 52, 55; 17.23; 18.2, 24; 21.12; 22.14; 23.3, 4, 4, 40, 48, 49; 26.6, 10; 30.7, 17, 18, 25; 31.5, 5, 12; 32.16, 16; 33.13, 16; 34.5, 5, 5, 8, 14, 14, 19, 19, 22; 35.9, 10; 36.10, 38; 37.3; Hos. 4.13, 13, 14, 14; Joel 4.18; Amos 4.3; 8.13; 9.13; Obad. 1.13; Mic. 2.12; 6.1; 7.10, 16; Zech. 1.17; 4.9; 5.9; 6.7; 8.9, 13; 11.9; 14.2, 12; Mal. 1.5; Ps. 17.2; 31.19; 35.10; 37.15, 17; 45.16, 16; 48.12; 51.10; 65.13; 66.7; 69.24; 71.23; 75.11; 78.64; 81.7; 92.12; 97.8; 119.171; 130.2; Job 5.12, 18; 11.20; 17.5, 16; 20.10; 27.4, 15; 39.2, 3, 3, 3; Prov. 5.3; 6.27, 28; 10.27; 23.16, 26; 24.2; 27.20, 20; 30.15; Ruth 1.7, 9, 9, 10, 11, 13, 13, 14, 14, 19, 19, 20, 21; 4.14, 17, 17; Song 4.11; Lam 2.20; 4.1, 17; 180 The Historical Depth of the Tiberian Reading Tradition Est. 1.18; 4.4; Dan. 8.8, 22, 22; 12.7; Neh. 12.40; 1 Chron. 7.15; 2 Chron. 9.6, 21; 34.28. ‫ ִֶן‬-: Gen. 49.26. ָ ‫ן‬-: Gen. 19.33, 35, 36; 26.35; 27.1; 30.38, 39; 33.6, 6; 37.7; 41.24, 36; Exod. 1.17, 18, 18, 19; 2.19; 15.20; 25.27; 26.3; 27.2; 28.21, 21; Num. 25.2; Deut. 21.15; 31.21; 1 Sam. 18.7; 25.43; 2 Sam. 13.18; Ezek. 3.20; 7.4, 9; 16.55, 55; 29.12; 34.10; Zech. 13.7. Imperative—‫נָ ה‬-: Gen. 4.23; 2 Sam. 1.24; Isa. 32.9, 9, 9; Jer. 9.19, 19; 49.3, 3, 3, 3; Ruth 1.8, 8, 11, 12; Song 3.11, 11. ‫ ִַן‬-: Gen. 4.23; ‫ ִֶן‬-: Exod. 2.20. ָ ‫ן‬-: Ruth 1.9, 12, 20. 5.2. Samaritan Tradition 5.2.1. Samaritan Written Tradition Prefix conjugation—‫נה‬-: Gen. 3.7; 19.33, 35, 36; 24.61, 61; 26.35; 27.1; 30.38, 39*, 39; 31.14; 33.6; 37.7, 7; 41.2, 3, 4, 7, 18, 20, 24, 36, 53, 54; 49.26; Exod. 1.17, 17, 18, 19; 2.16, 16, 16, 18, 19; 8.5, 7; 15.20; 25.27; 26.3, 3*; 27.2; 28.21, 21; Lev. 7.30; 10.19; 23.15, 17, 17; Num. 14.45; 25.2; 27.1, 2; 35.11, 13, 14, 15; 36.3, 4, 6, 6, 11; Deut. 21.15; 31.21. ‫הן‬-: Gen. 30.38; 33.6; 41.21; Lev. 4.2, 13, 22, 27; 5.17; Deut. 1.44. Imperative—‫ן‬-: Gen. 4.23; Exod. 2.20; ‫נה‬-: Gen. 4.23. 5.2.2. Samaritan Reading Tradition Prefix conjugation— -na: Gen. 3.7; Num. 25.2; Deut. 31.21. -inna: Gen. 19.33, 35, 36; 24.61, 61; 26.35; 27.1; 30.38, 38, 39*, 39; 31.14; 33.6, 6; 37.7, 7; 41.2, 3, 4, 7, 18, 20, 21, 24, 36, 53, 54; 49.26; Exod. 1.17, 17, 18, 19; 2.16, 16, 16, 18, 19; 8.5, 7; 15.20; 25.27; 26.3, 3*; 27.2; 28.21, 21; Lev 4.2, 13, 22, 27; 5.17; 7.30; 10.19; 23.15, 17, 17; Num. 14.45; 27.1, 2; 35.11, 13, 14, 15; 36.3, 4, 6, 6, 11; Deut. 1.44; 21.15. Imperative— -ān: Gen. 4.23; -na: Gen. 4.23; -īn: Exod. 2.20 5.3. Dead Sea Scrolls 5.3.1. Biblical Dead Sea Scrolls Prefix conjugation—DSS ‫נה‬- || MT ‫נָ ה‬-: 1QIsaa 3.22 || MT Isa. 3.16; 1QIsaa 3.23 || MT Isa. 3.16; 1QIsaa 3.24 || MT Isa. 3.16; 1QIsaa 5.2 || MT Isa. 5.15; 1QIsaa 10.25 || MT Isa. 11.7; 1QIsaa 11.16 || MT Isa. 13.7; 1QIsaa 11.24 || MT Isa. 13.16; 1QIsaa 11.26 || MT Isa. 13.18; 1QIsaa 13.18 || MT Isa. 16.2; 1QIsaa 14.4 || MT Isa. 17.2; 1QIsaa 14.12 || MT Isa. 17.7; 1QIsaa 20.17 || MT Isa. 26.6; 1QIsaa 21.22 || MT Isa. 27.11; 1QIsaa 22.1 || MT Isa. 28.3; 1QIsaa 23.29 || MT Isa. 29.18; 1QIsaa 25.2 || MT Isa. 30.21; 1QIsaa 26.11 || MT Isa. 32.3; 1QIsaa 26.12 || MT Isa. 32.3; 1QIsaa 26.20 || MT Isa. 32.10; 1QIsaa 27.19b || MT Isa. 9. The 2/3FPL Endings 181 33.17b; 1QIsaa 27.23 || MT Isa. 33.20; 1QIsaa 28.21 || MT Isa. 35.5; 1QIsaa 28.22 || MT Isa. 35.5; 1QIsaa 35.19 || MT Isa. 42.9; 1QIsaa 37.13 || MT Isa. 44.7; 1QIsaa 38.4 || MT Isa. 44.26; 1QIsaa 39.28 || MT Isa. 47.9; 1QIsaa 40.9 || MT Isa. 48.3; 1QIsaa 41.15 || MT Isa. 49.15; 1QIsaa 41.23 || MT Isa. 49.22; 1QIsaa 45.8 || MT Isa. 54.10; 1QIsaa 49.8 || MT Isa. 60.4; 1QIsaa 49.11 || MT Isa. 60.8; 1QIsaa 52.27 (2x) || MT Isa. 65.17 (2x); 1QIsaa 54.1 || MT Isa. 66.14; 1Q8 5b.9 || MT Isa. 13.7; 1Q8 6c–d.9 || MT Isa. 16.2; 1Q8 14.2 || MT Isa. 35.5; 1Q8 19.8 || MT Isa. 44.26; 1 Q8 20.22 || MT Isa. 47.9; ‫ תנשינה‬1Q8 26.7 || ‫ ֵת ָא ַ ַֽמנָ ה‬MT Isa. 60.4; 1Q8 26.12 || MT Isa. 60.8; 1Q8 28.1 || MT Isa. 65.17; 1Q8 28.24 || MT Isa. 66.14; 4Q3 f1ii.8 || MT Gen. 41.2; 4Q3 f1ii.10 || MT Gen. 41.4; 4Q5 f4i– 5.10 || MT Gen. 41.3; 4Q25 f2.5 || MT Lev. 4.2; 4Q51 9e–i.7 || MT 1 Sam. 10.7; 4Q55 f8.7 || Isa. 13.7; 4Q57 f6.4 || MT Isa. 11.7; 4Q57 f44–47.1 || MT Isa. 54.10; 4Q58 3.2 || MT Isa. 47.9; 4Q58 4.24 || MT Isa. 49.15; 4Q70 f29.8 || MT Jer. 18.21; 4Q78 f18–20.12 || MT Joel 4.18; 4Q94 f5–6.3 || MT Ps. 97.8; 11Q5 14.1 || MT Ps. 119.171; 4Q104 f1.12 || MT Ruth 1.9; 4Q107 f2ii.13 || MT Song 4.11; Mur88 8.3 || MT Amos 8.13; Mur 88 8.32 || MT Amos 9.13; Mur88 9.21 || MT Obad. 13; Mur88 12.32 || MT Mic. 2.12; Mur88 15.29 || MT Mic. 7.10; Mur88 15.38 || MT Mic. 7.16. DSS ‫נה‬- || MT ָ ‫ן‬-: 4Q13 f2.5 || MT Exod. 1.19; 4Q14 6.43 || MT Exod. 15.20; 4Q22 28.6 || MT Exod. 25.27. DSS ‫ן‬- || MT ָ ‫ן‬-: 4Q13 f2.3 || MT Exod. 1.17. DSS ‫ן‬- || MT ‫נָ ה‬-: 4Q3 f1ii.13 || ‫נָ ה‬- MT Gen 41.7. DSS ‫נה‬- || MT Other: ‫ תשפלנה‬1QIsaa 2.19 || ‫ ָש ֵֵ֔פל‬MT Isa. 2.11; ‫ תנשינה‬1QIsaa 53.28 || ‫ ִתנָ ֵֵ֔שאּו‬MT Isa. 66.12. DSS Other || MT ‫נָ ה‬-: ‫ תחזיון‬1QIsaa 27.19a || ‫ֶת ֱח ֶזינָ ה‬ MT Isa. 33.17a; ‫ תקראון‬1QIsaa 34.28 || ‫ ִת ְּק ֶ ָׂ֑רינָ ה‬MT Isa. 41.22. Imperative— DSS ‫נה‬- || MT ‫נָ ה‬-: 1QIsaa 26.19 (3x) || MT Isa. 32.9 (3x); 4Q104 f1.10 (2x) || MT Ruth 1.8 (2x). 5.3.2. Non-biblical Dead Sea Scrolls Prefix conjugation: CD 19.8; 1QM 8.1; 1QHa 15.14; 4Q171 f1–2ii.16 || MT Ps. 37.15; 4Q176 f8–11.12 || MT Isa. 54.10; 4Q268 f1.1; 4Q364 f8ii.2 || MT Gen. 37.8; 4Q365 f6b.6 || MT Exod. 15.21; 4Q378 f3ii+4.11; 4Q433a f2.4, 4; 4Q437 f2i.3 || MT Ps. 37.15; 4Q481 f2.2; 11Q19 21.13. Imperative: 1QM 12.13, 15, 15, 15; 19.5, 7, 7, 7; 4Q365 f6aii+6c.6; 4Q492 f1.7, 7. 5.4. Mishna Prefix conjugation: Nedarim 3.11 (2x) || MT 1 Sam. 1.20 (2x); Soṭa 1.6 || MT Ezek. 23.48; Soṭa 9.9 (2x) || MT Hos. 4.14 (2x); ʿArayot 1.22 (2x) || MT Num. 35.14 (2x); Makkot 2.4 || MT Num. 35.13. Imperative: Taʿanit 4.8 (2x) || MT Song 3.11 (2x); Moʿed Qaṭan 3.9 || MT Jer. 9.19; Nedarim 9.10 (2x) || 2 Sam. 1.24. 10. NIFALISATION A well-known example of ancient Hebrew historical development involves the realignment of verbal stems. Over time, many Gstem (qal) verbs were replaced by synonymous cognates in other stems (binyanim). The present chapter focuses specifically on the shift from G- to N-stem (nifʿal). This process, which is here termed nifalisation, was neither wholesale nor haphazard. Rather, it was limited chiefly to originally qal verbs with stative, medio-passive, reflexive, or more broadly intransitive semantics, including qal internal passive forms. The process often resulted in suppletive paradigms, sometimes with only vestigial qal representation. The phenomenon of nifalisation is especially characteristic of Second Temple chronolects—such as LBH, DSS Hebrew, SH, the Hebrew of BS, and RH—though the extent and specific manifestations in each varies. Since a large portion of the Tiberian biblical reading tradition’s crystallisation took place in the Second Temple Period, it is not surprising that nifalisation is also detectable in the Tiberian vocalisation of classical biblical material, specifically in deviations of the Tiberian reading tradition from the consonantal text. Even so, it must be emphasised that Tiberian vocalisation also preserves evidence of resistance to nifalisation and that shifts from qal to nifʿal are not exclusively late, but extend back into presumably early Tiberian consonantal biblical material. © 2023 Aaron D. Hornkohl, CC BY-NC 4.0 https://doi.org/10.11647/OBP.0310.10 184 The Historical Depth of the Tiberian Reading Tradition 1.0. Second Temple Evidence 1.1. Tiberian Late Biblical Hebrew The shift away from medio-passive qal and qal internal passive is seen in developments that characterise Tiberian LBH as contrasted with Tiberian CBH. 1.1.1. ‫‘ כש"ל‬stumble’ Consider the example of apparent suppletion involving qal ‫ָכ ַשל‬ and nifʿal ‫‘ יִ ָכ ֵשל‬stumble’. On the surface, BH seems to exhibit an indiscriminate mixture of qal and nifʿal, e.g., (1) ‫ם־יְּהּודה ִע ָ ַֽמם׃‬ ֶ֖ ָ ַ‫כ ְּׁשלּו ַב ֲע ֵ֔ ָֹונם כָ ַּשּ֥ל ג‬ ִֽ ָ ִ‫וְּ יִ ְּש ָר ֵאל וְּ ֶא ְּפ ַ ֵ֗ריִ ם י‬... ‘…Israel and Ephraim stumble in their guilt; Judah has also stumbled with them.’ (Hos. 5.5) (2) ‫ל־ה ָ ֵ֥עם‬ ָ ‫הוה ִהנְּ ִ ֵ֥ני נ ֵ ֹ֛תן ֶא‬ ֵ֔ ָ ְּ‫ ָל ֵֵ֗כן ַֹ֚כה ָא ַמר י‬... ‫יְּהוה׃‬ ַֽ ָ ‫ת־פ ַק ְּד ִ ֵ֥תים יִ כָ ְּׁשלֹׁ֖ ּו ָא ַ ֵ֥מר‬ ְּ ‫ְּב ֵע‬ ...‫ַה ֶ ֶ֖זה ִמ ְּכש ִ ָֹׂ֑לים וְּׁ ָכ ְּּׁ֣שלּו ּ֠ ָבם‬ ‘“…at the time that I punish them, they will stumble,” says the LORD. …“Behold, I will lay before this people stumbling blocks and they will stumble against them…”’ (Jer. 6.15, 21) Upon closer inspection, however, a situation of suppletion emerges in CBH. Forms are vocalised as nifʿal unless the consonantal spelling is not amenable, in which case qal forms are preserved. It is only in LBH that that the written tradition ‘catches up with’ the vocalisation and one encounters a comparative proliferation of consonantally unambiguous nifʿal forms, e.g., ‫‘ וְ נִ ְכ ָשלּו‬and they will fail’ (Dan. 11.14; see also Dan. 11.19, 33) and ‫ּוב ִה ָכ ְש ָ֔ ָלם‬ ְ ‘and when they stumble’ (Dan. 11.34). This trend continues in QH and RH (see below, §2.1.1; see further Khan 2020, I:58). 10. Nifalisation 185 1.1.2. Qal Internal Passive > Nifʿal Another LBH manifestation of nifalisation is replacement of qal internal passive with nifʿal.1 A useful example involves forms of the qal internal passive ‫‘ יֻ ַלד‬be born’. These appear throughout the Bible—Torah, Prophets, Writings2—but are rare in LBH (where the sole case, in 1 Chron. 1.19, was likely imported from Gen. 10.25). Conversely, consonantally unambiguous nifʿal alter- 1 A succinct account of the disappearance of the qal internal passive is given by Fassberg (2001, 254): One finds in the literature two related explanations for the disappearance of the Qal internal passive. The first is phonetic: at a certain stage, Hebrew phonology no longer tolerated a short vowel (in this case u in *qutal) in an open pretonic syllable. The u-vowel, which was the marker of the passive, could be maintained only in a closed syllable; the closing of the syllable was accomplished by secondarily geminating the following consonant. The resulting form with geminated second radical became identical to the Puʿʿal and hereafter was interpreted as Puʿʿal. In the case of the imperfect, forms like ‫ יֻ ַתן‬and ‫ יֻ ַקח‬were reanalyzed as Hofʿal forms with regressive assimilation of the first radical: *yuntan > yuttan and *yulqaḥ> *yuqqaḥ. The second reason is morpho-semantic: Nifʿal, which may have been originally reflexive in Hebrew, began to take on a passive meaning as well, thus rendering the Qal internal passive redundant. See Fassberg (2001, 254) for bibliographical references. 2 Gen. 4.26; 6.1; 10.21, 25; 24.15; 35.26; 36.5; 41.50; 46.22, 27; 50.23; Judg. 18.29; 2 Sam. 3.5; 21.20, 22; Isa. 9.5; Jer. 20.14–15; 22.26; Ps. 87.4–6; 90.2; Job. 5.7; Ruth 4.17; 1 Chron. 1.19. 186 The Historical Depth of the Tiberian Reading Tradition natives, like qaṭal ‫נֹולד‬ ַ and infinitival ‫הּוָ ֵלד‬, ִ preponderate conspicuously in LBH.3 Consider the parallels: (3a) ‫ ֵ ֛א ֶלה י ְֻּׁׁל ּ֥דּו ְּל ָדִוֶ֖ד ְּב ֶח ְּב ַֽרֹון׃‬... ‘These were born to David in Hebron’ (2 Sam. 3.5) (3b) ...‫ִש ָש ֙ה נִֽ ֹולַּ ד־לֹו ְּב ֶח ְּב ֵ֔רֹון‬ ‘Six were born to him in Hebron’ (1 Chron. 3.4) and (4a) ‫ם־הּוא י ַֻּׁלּ֥ד ְּל ָה ָר ָ ַֽפה׃‬ ֶ֖ ַ‫וְּ ג‬... And he, too, was born to the Rapha (2 Sam. 21.20) (4b) ‫נֹולּ֥ד ְּל ָה ָר ָ ַֽפא׃‬ ַּ ‫ם־הּוא‬ ֶ֖ ַ‫וְּ ג‬... And he, too, was born to the Rapha (1 Chron. 20.6) Likewise, while unambiguous spellings of both qal internal passive ‫ יֻ ַתן‬and nifʿal ‫‘ יִ נָ ֵתן‬will be given (3MS)’ come in CBH texts, LBH texts have only nifʿal forms, the qal internal passive forms having fallen away. Indeed, more generally in the late corpus consisting of Qohelet, Esther, Daniel, Ezra, Nehemiah, and Chronicles, Hughes (1994, 76, fn. 20) counts just four cases of the qal internal passive, ‫‘ י ָּוק ִשים‬are snared’ (Qoh. 9.12), ‫‘ ֻא ְכלֵ֥ ּו‬have been consumed’ (Neh. 2.3, 13), and the aforementioned inherited ‫‘ יֻ ַלִ֖ד‬were born’ (1 Chron. 1.19 || Gen. 10.25). See further Reymond (2016, 1138); Qimron (2018, 221). 3 Beyond the infinitival forms in Gen. 21.5 and Hos. 2.5, occurrences of finite and infinitive forms are limited to LBH: Qoh. 4.14; 7.1; 1 Chron. 2.3, 9; 3.1, 4; 20.6; 26.6. Not unrelated are the nufʿal forms parallel to more classical alternatives in 1 Chron. 3.5 || 2 Sam. 5.14 and 1 Chron. 20.8 || 2 Sam. 21.22. 10. Nifalisation 187 1.2. Dead Sea Scrolls Hebrew 1.2.1. Late Nifalisation DSS Hebrew shows continuity of the LBH features listed above, most notably, consonantally unambiguous forms, such as ‫נכשלו‬ ‘stumbled (3MP)’ (CD 2.17; 4Q266 f2ii.17) and ‫נולדה‬ ̇ ‘she was born’ (4Q215 f13.4; see also 11Q19 40.6). It also furnishes the earliest unequivocal consonantal evidence of the nifʿal morphology for the Tiberian suppletive verb ‫נִ גַ ש‬-‫‘ יִ גַ ש‬approach’, in the form ‫‘ בהנגשו‬when he approaches’ (4Q512 f40–41.2) (see below, §§1.3.6; 2.1.2). 1.2.2. Qal Internal Passive > Nifʿal Additionally, nifʿal ‫‘ ינתן‬will be given (3MS)’ is employed to the exclusion of qal internal passive ‫יתן‬. Indeed, the NBDSS present no clear-cut cases of the qal internal passive.4 Reymond (2016, 1139–40) lists many DSS Hebrew alternatives for MT qal internal passive forms. Qimron (2018, 222) observes that DSS Hebrew develops a nifʿal ‫‘ *נשלם‬be fulfilled, completed’ (infinitival forms at 1QS 10.6; 4Q256 19.5; 4Q270 f3ii.21; 4Q385 f11i.3) corresponding to MT stative qal ‫* ָש ֵלם‬. 4 According to the tagging in Abegg’s (1999–2009) QUMRAN Accord- ance module, ‫( תו̇ גע‬4Q417 f1i.23) is qal internal passive, but Qimron (2020, II:148) reads the form as ‫‘ תיֿ גע‬do (not) touch’. 188 The Historical Depth of the Tiberian Reading Tradition 1.3. Samaritan Hebrew5 As a biblical tradition characterised by pervasive Second Temple linguistic evolution, it is no surprise that SH also documents the shift in question. Indeed, though transmitting a literary tradition likely rooted in the Iron Age, SH presents a relatively advanced stage of nifalisation compared to other ancient Hebrew traditions. Yet, the Samaritan picture is complicated by several factors. First, like the Tiberian biblical tradition, the linguistic testimony of the SP is composite. It comprises related, but partially independent written and reading components. Crucially, as regards both nifalisation and other linguistic developments, the two components of the tradition present historically distinct stages. Second, while SH both confirms and exceeds the nifalisation seen in several other Second Temple Hebrew traditions, it also evinces qal forms reminiscent of pre-Tiberian Hebrew. Brushing aside cases of local divergence in which SH nifʿal forms differ from Tiberian qal counterparts due to textual and/or interpretive factors not representative of broader trends, more pervasive Samaritan nifalisation manifests in several ways. 1.3.1. Comprehensive Nifalisation First, there are Tiberian qal verbs with forms amenable to nifʿal recasting that are consistently read as nifʿal in SH. These are the broadly stative, reflexive, intransitive, and weakly transitive verbs in the following list. 5 For a study focused on nifalisation in the Samaritan biblical tradition, see Hornkohl (2022). 10. Nifalisation 189 ‫‘ ָד ַבק‬cling’, ‫‘ ָדגָ ה‬multiply’, ‫‘ ָחגַ ר‬gird’, ‫‘ ָחזַ ק‬be/become strong’, ‫‘ ָכ ַבד‬be/become heavy’, ‫‘ ָכ ָבה‬go out, be extin- guished (of fire)’, *‫‘ ָמ ַעט‬be small’, ‫‘ ָס ַחר‬travel about, engage in trade’, *‫‘ ָס ַרח‬hang over (of a covering)’, *‫‘ ָע ַבט‬give/take collateral (for a loan)’, ‫‘ ָצנַ ף‬wrap one’s head (with a turban)’, ‫‘ ָק ַצר‬be/become short’, ‫‘ ָר ֵעב‬be hungry’, ‫‘ ָש ַכח‬forget’, ‫‘ ָש ַכר‬become drunk’, ‫‘ ָש ַקע‬die down, be extinguished (of fire)’ Many such verbs are rarely attested, but a few of the more frequent have conspicuously suppletive paradigms in SH. These include the parallels to Tiberian ‫‘ ָחזַ ק‬be/become strong’, ‫ָס ַחר‬ ‘traffic, travel about, engage in trade’, and ‫‘ ָש ַכח‬forget’. Forms amenable to reinterpretation—especially in the prefix conjugation—are realised as nifʿal, whereas other forms—in the suffix conjugation, participles, imperatives, infinitives—remain qal. In these cases, no perceptible semantic shift accompanies the morphological shift. Such realignments often tally with late Aramaic use of Dt-stem forms, as seen in the Targums and/or Syriac. ‫חז"ק‬ ̊̄ Consider the suppletive relationship of SH G-stem ‫ חזק‬azåq (5) ̊̄ and N-stem ‫ ויחזק‬wiyyazåq (6), which occur in successive verses: 190 (5) The Historical Depth of the Tiberian Reading Tradition ̊̄ ( ‫וכל הארצות באו מצרימה לשבר אל יוסף כי חזק‬ )‫ חָ ַּז ּ֥ק‬MT || azåq ‫הרעב בכל הארץ׃‬ ‘And all the nations came to Egypt to buy food from Joseph, because the famine was severe in all the land.’ (Gen. 41.57; see also Gen. 47.20; Exod. 19.19; Deut. 12.23; 31.6, 7, 23) (6) ‫והרעב היה על פני כל הארץ ויפתח יוסף את כל אשר בהם בר וישביר‬ ̊̄ ( ‫למצרים ויחזק‬ ‫ח ַּז ּ֥ק) הרעב בארץ מצרים׃‬ ֱ ‫ ַּו ֶי‬MT || wiyyazåq ‘And the famine had spread over all the land and Joseph opened everything in which there was grain and he sold to Egypt and the famine was severe in the land of Egypt.’ (Gen. 41.56; see also Exod. 7.13, 22; 8.15; 9.35; 12.33; Deut. 11.8) 1.3.2. Partial Formal Nifalisation In the case of the verb ‫‘ מאן‬refuse’, SH presents a suppletive par- adigm composed of nifʿal prefix conjugation and piʿel B suffix conjugation, participle, and infinitive (see below, ch. 12, §2.1). 1.3.3. Partial Nifalisation for Grammatical/Semantic Disambiguation In other cases where the Tiberian tradition makes do with qal forms with varying valency and/or semantics, SH seems to exploit nifalisation for purposes of grammatical and/or semantic disambiguation. Consider the case of the SH counterpart to Tiberian ‫‘ ָד ַבק‬cling’ in examples (7)–(9) (see Hornkohl 2021a, 6–7). 10. Nifalisation (7) 191 ̊̄ ( ‫על כן יעזב איש את אביו ואת אמו ודבק‬ )‫ וְּׁ ָד ַּבּ֣ק‬MT || wdabǝq ‫באשתו והיה משניהם לבשר אחד׃‬ ‘Therefore a man will leave is father and his mother and cling to his wife and it will become from them one flesh.’ (Gen. 2.24; Deut. 28.60/616) (8) ̊̄ ( ‫ותדבק‬ ‫ ו ִַּת ְּׁד ַּבּ֣ק) נפשו בדינה בת יעקב ויאהב את‬MT || wtiddabǝq ‫הנערה וידבר אל לב הנערה׃‬ ‘And his soul was drawn to Dina the daughter of Jacob and he loved the girl and he spoke tenderly to her.’ (Gen. 34.2; see also Num. 36.7, 9; Deut. 10.20; 11.22; 13.5, 18) The passages cited in examples (7) and (8) represent suppletion similar to that discussed above: morphologically ambiguous yiqṭol forms originally in qal could be recast as nifʿal, while qaṭal forms preserve qal morphology, because their orthography leaves no room for nifʿal analysis. (9) ̊̄ MT || wlēddabēqa ( ‫לאהבה את יהוה אלהיך לשמע בקולו ולדבקה‬ ...‫ּול ָד ְּׁבקָ ה) בו‬ ְּׁ ‘loving the LORD your God, obeying his voice and holding fast to him,...’ (Deut. 30.20a) Example (9) demonstrates that nifalisation could affect even forms ill-suited to nifʿal analysis, such as the infinitive ‫ולדבקה‬ ̊̄ wlēddabēqa, whose original qal form is preserved in MT ‫ּול ָד ְּב ָקה‬. ְּ 6 The distinction in number between the verb form in the two traditions entails different subject referents. The SP’s singular verb refers across the verse boundary to the singular subject ‫‘ כל חלי‬every illness’ in the previous verse. 192 The Historical Depth of the Tiberian Reading Tradition (10) ‫הנה נא מצא עבדך חן בעיניך ותגדל חסדך אשר עשית עמדי לחיות את‬ ̊̄ MT || tidbaqinni ( ‫נפשי ואנכי לא אוכל להמלט ההרה פן תדבקני‬ ‫ִת ְּׁדבָ ַּ ּ֥קנִ י) הרעה ומתי׃‬ ‘Behold, your servant has found favour in your sight, and you have shown me great kindness in saving my life. But I cannot escape to the hills, lest the disaster overtake me and I die.’ (Gen. 19.19) Finally, example (10) testifies to the fact that the shift from G- to N-stem in the case of this verb is not one of mere formal suppletion, but was also evidently exploited for morphosemantic disambiguation. Here, the sole prefix conjugation form of ‫ דבק‬that retains qal morphology is strongly transitive (taking an object suffix) and semantically dynamic (‘to overtake’ rather than just ‘cling to’). The rest of the SH prefix conjugation forms of this verb, i.e., those mentioned in (8) and (9), all take objects with -‫ב‬ and have stative semantics. Similar morphosemantic disambiguation obtains in the cases of the SH equivalents of Tiberian qal ‫‘ ָע ַבט‬take collat- eral/lend, give collateral/borrow’, ‫‘ ָכ ַבד‬be/become heavy’, ‫ָחגַ ר‬ ‘gird’, and ‫‘ ָק ַצר‬be/become short’ (see Hornkohl 2021, 5–6). 1.3.4. Nifalisation Resulting in Nifʿal B Alongside its standard nifʿal, SH has a second N-stem (Ben-Ḥay- yim 2000, 117–18). The so-called nifʿal B is a hybrid that incorporates components of the N- and Dt-stems. It has both nifʿal orthography and the middle radical gemination characteristic of hitpaʿʿel, thus partially resembling RH nitpaʿʿal (see below, §1.5). The resemblance is not total, because crucial to the reinterpretation of qal forms as nifʿal B was the routine assimilation of the -t- 10. Nifalisation 193 infix in some Second Temple Aramaic and Hebrew dialects, such as SA, Jewish Palestinian Aramaic, Jewish Babylonian Aramaic, and RH, according to which hitpaʿʿel/nitpaʿʿel > hippaʿʿel/nippaʿʿel (Ben-Ḥayyim 2000, 117–18; Bar-Asher 2016, 209–10). An original qal form was not amenable to reinterpretation as a hitpaʿʿel/nitpaʿʿel due to the mismatch involving the absence or presence of infix -t-. Conversely, the nifʿal B realisation of original qal forms faced no such obstacle, as the -t- infix had assimilated, resulting in a form with geminated first and second radicals. Originally qal prefix forms and the like could easily be pronounced as Nifʿal B forms. The Tiberian counterparts of these SH nifʿal B forms con- sistently show qal morphology, whereas in SH their paradigms are suppletive: qal is read where necessary, nifʿal B where possible. Again, the Targums also sometimes resort to dedicated middle Dt morphology. Relevant Tiberian verbs with Samaritan nifʿal B parallels include qal ‫‘ גָ ַבר‬prevail’ and ‫‘ ָק ָשה‬be hard, severe’, and both qal ‫‘ ָכ ָלה‬finish (intr.)’ and puʿal ‫‘ ֻכ ָלה‬be finished’, in which all prefix conjugation forms were levelled to nifʿal B (Hornkohl 2022, 7–9). Consider the Samaritan equivalents to qal suffix conjugation ‫ גָ ְּב ֶ֖רּו‬and prefix conjugation ‫ וַ יִ גְּ ְּב ֵ֥רּו‬in examples (11) and (12). (11) ‫ ג ְָּׁב ֹׁ֖רּו) המים ויכסו‬MT || gēbēru( ‫חמש עשרה אמה מלמעלה גברו‬ ‫ההרים׃‬ ‘The waters prevailed above the mountains, covering them fifteen cubits deep.’ (Gen. 7.20; see also Gen. 7.19; 49.26) ̊̄ ( ‫ויגברו‬ (12) ‫ וַּיִ גְּׁ ְּׁב ּ֥רּו) המים על הארץ חמשים ומאת‬MT || wyiggåbbaru ‫יום׃‬ ‘And the waters prevailed on the earth 150 days.’ (Gen. 7.24; see also Gen. 7.18) 194 The Historical Depth of the Tiberian Reading Tradition In contrast to the G-stem paradigmatic consistency in the Tiberian tradition, the SH verb has a suppletive paradigm. Nifʿal is read where possible, qal where consonantal form precludes nifʿal analysis. TO resorts to dedicated medio-passive Dt-stem verbs in select cases, e.g., MT ‫רּו‬ ֙ ‫( ָ ַֽג ְּב‬Gen. 49.26) || TO ‫ ;יתוספן‬MT ‫ וְּ גָ ַ ֵ֥בר‬...‫וְּ גָ ַבר‬ (Exod. 17.11) || TO ‫ מתגברין‬...‫מתגברין‬. 1.3.5. Qal Internal Passive > Nifʿal A phenomenon partially related to nifalisation is the well-known replacement of the qal internal passive with alternatives, a process more pronounced in SH than in Tiberian Hebrew. Tiberian qal passive ‫ גֻ נַ ב‬is twice paralleled by orthographic nifʿal alternatives, not just in the reading component of the Samaritan tradition, but in the written component, as well (Gen. 40.15; Exod. 22.6).7 Nifalisation, however, is not the usual SH alternative to Tiberian qal internal passive. Among the more common strategies are the qal passive participle (parallel to Tiberian paʿul), the 3MPL qal impersonal, and active interpretation. 1.3.6. Conditioned Qal Preservations Despite the comparatively advanced stage of nifalisation it displays, SH also exhibits conditioned, and possibly secondary, qal forms parallel to Tiberian nifʿal forms. These are suggestive of pre-Tiberian Hebrew. For example, the Tiberian verb ‫ נִ גַ ש‬is fa7 Interestingly, while the Samaritan written tradition has apparently nifʿal ‫( נגנבתי‬Gen. 40.15) and ‫( ונגנב‬Exod. 22.6) against the Tiberian qal internal passives ‫ גֻ ֵ֔ ַנ ְּב ִתי‬and ‫וְּ גֻ ַנֶ֖ב‬, respectively, the Samaritan reading trå̄ dition differentiates between nifʿal niggɑnåbti and nifʿal B wniggɑnnɑb. 10. Nifalisation 195 mously suppletive: nifʿal wherever the consonantal text allows, i.e., suffix conjugation (‫ )נִ גַ ש‬and participle (‫ ;)נִ גָ ִשים‬qal where con- sonantal form precluded nifʿal recasting, i.e., prefix conjugation (‫)יִ גַ ש‬, infinitive construct (‫)גֶ ֶשת‬, imperative (‫גש‬/ ַ ‫( )גֶ ש־‬see below, §2.1.2). For its part, the Samaritan verb is uniformly qal, includ̊̄ ing suffix conjugation (nagåš) and participle (nēgǝš). On the one hand, a unified ‫*נָ גַ ש‬-‫ יִ גַ ש‬qal paradigm, as in SH, is precisely what has been hypothesised for pre-Tiberian Hebrew. On the other, it must be emphasised that the apparent Samaritan preservation of qal is conditioned, since Samaritan I-n consonantal forms are not amenable to nifʿal phonology. This is true not just of the prefix conjugation, where—as in Tiberian Hebrew— only those I-n forms that preserve a first radical nun are eligible for nifʿal realisation, but also of the suffix conjugation and certain forms of the participle.8 This is because—unlike in Tiberian Hebrew—1st-radical gemination applies throughout the Samaritan nifʿal paradigm, which would yield such forms as prefix conjugå̄ ̊̄ tion *yinnagåš, suffix conjugation *ninnagåš, and verbal participle ̊̄ *ninnagǝš, none of which suit their respective consonantal spell- ings, i.e., ‫יגש‬, ‫נגש‬, and ‫נגש‬.9 8 In SH this secondary gemination applies only to participles with verbal semantics; participles with nominal semantics preserve the inherited morphology without gemination (Ben-Ḥayyim 2000, 193). 9 Other weak roots for which SH regularly has qal against Tiberian nifʿal include ‫מו"ל‬/‫‘ נמ"ל‬circumcise’; ‫פו"ץ‬/‫‘ נפ"ץ‬scatter’; ‫סב"ב‬/‫‘ נס"ב‬surround’; ‫בו"ך‬/‫‘ נב"ך‬be confused’; ‫מו"ג‬/‫‘ נמ"ג‬melt’; ‫מס"ס‬/‫‘ נמ"ס‬melt’; ‫מק"ק‬/‫נמ"ק‬ ‘rot’; ‫חת"ת‬/‫‘ נח"ת‬be dismayed’. 196 The Historical Depth of the Tiberian Reading Tradition 1.4. Ben Sira Despite unmistakable indications of the late linguistic milieu that it represents, the language of BS is remarkably classical. In terms of the phenomenon of nifalisation here under discussion, however, BS shows unmistakable affinities with other late Hebrew corpora. 1.4.1. Qal > Nifʿal in the Case of Medio-passive Semantics First, several Tiberian medio-passive qal verbs find nifʿal alternatives in BS. These include ‫‘ נדבק‬cling’ (SirB 3v.14) (Dihi 2004, 162–65), ‫‘ נדעך‬go out (of fire), be extinguished, uprooted’ (Mas1h 2.5; SirB 10r.7), and ‫‘ נחכם‬be wise’ (SirB 7v.13; SirC 4v.3; SirD 1v.9; SirD 1v.10) (Dihi 2004, 162–65), though BS’s classical penchant is displayed in the continued use of qal ‫ דבק‬and ‫חכם‬.10 1.4.2. Qal Internal Passive > Nifʿal Second, despite the classical mien of BS’s Hebrew, the corpus attests to only highly equivocal cases of potential qal internal passive forms (Reymond 2016, 1142–50). Moreover, some of the more common BH qal internal passive forms go unused in BS in favour of nifʿal alternatives, such as ‫‘ נלקח‬was taken (MS)’ (SirB 13v.18; 17v.13; 19r.4) and ‫‘ ינתן‬will be given (MS)’ (SirA 6r.28 || SirB 2v.1 [margin]; SirC 6r.3). 10 In Tiberian BH the verbs in question are almost exclusively qal, the lone exception being ‫‘ נִ ְּד ֲע ֵ֥כּו‬they dry up, disappear’ (Job 6.17). 10. Nifalisation 197 1.5. Rabbinic Hebrew RH is well known for several processes subsumed in this study under the heading nifalisation. 1.5.1. Qal > Nifʿal in the Case of Stative and Mediopassive Semantics It has already been mentioned that RH joins LBH and DSS Hebrew in the attestation of consonantally unambiguous nifʿal infinitive ‫ּוב ִה ָכ ְשלֹו‬ ְ ‘and when he stumbles’ (m. ʾAvot 4.19), matching the nifʿal vocalisation of MT ‫ּוב ָכ ְש ֗לֹו‬ ִ (Prov. 24.17), in opposition to its qal consonantal orthography.11 Additional cases of RH nifʿal || MT qal include ‫‘ אבד‬be/become lost, die’, ‫‘ ארך‬be/become long’, and ‫‘ חסר‬lack’ (Bendavid 1967–1971, II:483). 1.5.2. Qal > Nitpaʿʿal Especially typical of RH is replacement of medio-passive qal with nitpaʿʿal (often in conjunction with movement of active qal > piʿʿel; see below, ch. 12, §1.5). This is evident in such verbs as ‫‘ נתחמץ‬become leavened’ (m. Ṭevul Yom 3.4), ‫‘ נתמלא‬become full’ (e.g., m. Yoma 5.1 || MT Isa. 6.4), ‫‘ נתרחק‬be distant, avoid’ (m. Sanhedrin 3.4; m. ʾAvot 2.9), and ‫‘ נשתתק‬be mute’ (m. Giṭṭin 7.1). These contrast with the Tiberian consonantal tradition, which 11 It is worth noting that such authentic nifalisations in reliable Mishna manuscripts are often, due to a biblicising tendency, replaced in printed editions with qal forms. For example, the Eshkol (2000) version of the Mishna reads ‫ּוב ָכ ְּשלֹו‬ ִ in m. ʾAvot 4.19 in agreement with MT Prov. 24.17 and against Kaufmann’s ‫ּוב ִה ָכ ְשלֹו‬. ְ I am grateful to Geoffrey Khan for reminding me of this matter. 198 The Historical Depth of the Tiberian Reading Tradition prefers qal forms for the relevant semantic values. Turning to the qal internal passive—aside from biblical allusions, it is generally absent from RH (Sharvit 2004, 45; Reymond 2016, 1141, fn. 37).12 2.0. The Tiberian Reading Tradition of Classical Biblical Hebrew Texts Since the Tiberian reading tradition crystallised in the Second Temple Period, it is not surprising that nifalisation is also detectable in the oral realisation (vocalisation) of classical, i.e., ostensibly First Temple biblical material, specifically in secondary deviations in the Tiberian pronunciation tradition from the pronunciation implied by the written tradition. 2.1. Partial Nifalisation of Intransitive Verbs 2.1.1. ‫‘ כש"ל‬stumble’ A clear case involves the aforementioned shift of qal ‫ > ָכ ַשל‬nifʿal ‫‘ נִ ְכ ַשל‬stumble’ (§§1.1.1; 1.2.1). As noted above, consonantally un- ambiguous nifʿal forms, especially in the suffix conjugation, have a conspicuously late distribution. Yet, nifʿal vocalisation is not restricted to LBH, but is routine in CBH, too. This is because, unlike their suffix conjugation counterparts, the ambiguous conso- 12 Biblical allusions include the phrase )‫ן(־מיִ ֙ם‬ ַ֙ ‫‘ וְּ ִ ַ֤כי יֻ ַת‬but if water is put’ (Lev. 11.38) in m. Makhshirin (e.g., 1.1, 2 [4x], 3, etc.) and ‫יריִ ם‬ ֛ ַ ‫ַתנָ֧ ּור וְּ ִכ‬ ‫‘ יֻ ָ ֶ֖תץ‬and oven or stove will be smashed’ (Lev. 11.35) in m. ʿAvoda Zara 3.9. Beyond such allusions, the sole possible case in MS Kaufmann is ‫יּולד‬ ַ ‫( ַה‬m. Bekhorot 1.2), but the reading is doubtful (see Maʾagarim s.v.). 10. Nifalisation 199 nantal prefix conjugation form, initially qal—‫ *יִ ְכ ַשל‬or ‫—*יִ ְכש ֹל‬ was amenable to reanalysis as nifʿal—‫—יִ ָכ ֵשל‬in line with Second Temple linguistic trends, as manifested in the LBH written tradition, DSS Hebrew, and RH. It is noteworthy that the nifʿal reanalysis extended even to consonantal forms ill-suited to reanalysis, e.g., the infinitive construct ‫‘ ובכשלו‬and when he stumbles’ (Prov. 24.17), which, despite lacking the consonantal heh characteristic of a nifʿal infinitive construct, is vocalised as nifʿal ‫ ּוָ֜ ִב ָכ ְש ֗לֹו‬rather than qal ‫ּוב ָכ ְשלֹו‬ ְ *. The nifʿal morphology matches not just the aforementioned LBH consonantal nifʿal forms, including infinitival ‫ּוב ִה ָכ ְש ָ֔ ָלם‬ ְ (Dan. 11.34), but also DSS Hebrew ‫( נכשלו‬CD 2.17; 4Q266 f2ii.17), and—pointedly—RH ‫ּוב ִה ָכ ְשלֹו‬ ְ (m. ʾAvot 4.19), which is a citation of MT ‫‘ ּוָ֜ ִב ָכ ְש ֗לֹו‬and when he stumbles’ (Prov. 24.17), with orthography updated to match nifʿal pronunciation. 2.1.2. ‫‘ נג"ש‬approach’ Likewise, the aforementioned suppletion between qal prefix conjugation ‫( יִ ַגֵ֥ש‬Exod. 24.14), infinitive construct ‫( ִמ ֶג ֵֶ֥שת‬Exod. 34.30), and imperative ‫גש‬/‫ש־‬ ַ ֶ‫( ג‬2 Sam. 1.15; Gen. 19.9), on the one hand, and nifʿal suffix conjugation ‫( נִ ַגֵ֥ש‬Exod. 33.7) and participle ‫( ַהנִ גָ ִ ֵ֥שים‬Exod. 19.22), on the other, is probably due to reanalysis where allowed by the written forms (see above, §§1.2.1; 1.3.6). Significantly, the earliest unambiguous consonantal evidence matching the nifʿal vocalisation is found in Second Temple DSS Hebrew: ‫‘ בהנגשו‬when he approaches’ (4Q512 f40–41.2). 2.2. Qal Internal Passive > Nifʿal Similarly, in the Tiberian reading tradition, the replacement of qal internal passive with nifʿal nearly always occurs except where 200 The Historical Depth of the Tiberian Reading Tradition spelling precludes it (Böttcher 1866–1868, I:98–105; Barth 1890; Lambert 1900; Blake 1901, 53–54; Ginsburg 1929; 1934; 1936 Williams 1977; Hughes 1994, 71–76; Sivan 2009, 50–51; Reymond 2016).13 Consider the matter of qal infinitives absolute with cognate nifʿal finite forms in the so-called tautological construction. In several cases of qal-nifʿal mismatch, the consonantally ambiguous nifʿal finite form possibly conceals a qal passive, e.g., ‫‘ ָסקֹול יִ ָס ֵקל‬he/it will surely be stoned’ (Exod. 19.13; 21.28); ‫ָענֹוש‬ ‫( יֵ ָע ֗ ֵנש‬Exod. 21.22); ‫( גָ ֵֹ֥נב יִ גָ ֵנִ֖ב‬Exod. 22.11); ‫(‘ ָט ֵ֥ר ֹף יִ ָטּ ֵ ִ֖רף‬if) it is torn in pieces’ (Exod. 22.12). The special affinity concerning nifalisation between the Tiberian reading tradition and Second Temple consonantal traditions is borne out in the data. In Table 1, consider the earliest consonantal evidence for each of seven14 qal internal passive qaṭal 13 For the analysis of qal internal passive forms as hofʿal and puʿʿal forms as part of the processes of hifilisation and pielisation, see below, chs 11 and 12. 14 Williams includes the ketiv verb ‫*שגל‬, whose reconstructed oral reali- sation can only be conjecture. Rare in the Bible, the verb is even rarer in post-biblical material. On the relative antiquity of the qere, see above, ch. 3, §1.3. Excluded from Williams’s list is nifʿal ‫‘ נִ ְּק ַבר‬be buried’. This may be due to the D-stem passive classification of ‫‘ ֻק ַ ֵ֥בר‬was (were) buried’ (Gen. 25.10). Since D-stem ‫*ק ֵבר‬ ִ ‘bury en masse’ (Num. 33.4; 1 Kgs 11.15; Jer. 14.16; Ezek. 39.14–15; Hos. 9.6) has pluractional semantics, which are arguably lacking in the context in question, the form is more likely to be a qal internal passive (see below, ch. 12, §3.0, fn. 18). Moreover, the absence of any consonantally unambiguous biblical evidence for nifʿal ‫‘ נִ ְּק ַבר‬be buried’—for which all representative forms are in the prefix conjugation—coupled with the fact that unambiguous consonantal evi- dence of nifʿal ‫‘ נִ ְּק ַבר‬be buried’ is not extant until RH (m. Moʿed Qaṭan 10. Nifalisation 201 forms with corresponding nifʿal yiqṭol forms as listed by Williams (1977, 49). Table 1: Earliest unambiguous consonantal evidence of nifʿal morphology of suppletive Tiberian verbs with qal internal passive qaṭal forms and nifʿal yiqṭol forms Verb and Gloss Second Temple Reference ‫‘ ָד ָחה‬push, drive’ BS 13.21 ‫‘ ָה ַרג‬kill’ ‫‘ ָח ַצב‬hew’ 43Q372 f3.12; Tannaitic Hebrew (Mishna) Tannaitic Hebrew (Mekhilta deRabbi Ishmael, Mekhilta Devarim) ‫‘ ָט ַרף‬pluck’ Bar Kokhva (XHev/Se30 f1R.7), Tannaitic Hebrew (Mekhilta deRabbi Ishmael, Mekhilta deRabbi Shimon ben Yokhai) ‫‘ ָמ ַרט‬polish’ ‫‘ ָש ַרף‬burn’ Tannaitic Hebrew (Mishna, Sifra, Tosefta) Tannaitic Hebrew (Mishna, Mekhilta deRabbi Ishmael, Sifra, Seder Olam Rabba, Sifre Bemidbar) ‫‘ ָש ַטף‬rinse’ Tannaitic Hebrew (Sifra, Sifre Bemidbar, Sifre Devarim) 2.3. Nippaʿʿel/Hippaʿʿel (< Nitpaʿʿel/Hitpaʿʿel) < Nifʿal There is one further affinity between the Tiberian and Samaritan reading traditions worthy of emphasis in this connection: the occurrence of nifʿal B, that is the N-stem pattern with geminated middle radical common in SH and late Aramaic dialects (see above, §1.3.4), which is not unrelated to RH’s characteristic nitpaʿʿal (above, §1.5.2). Tiberian vocalisations of this sort are relatively rare. In the case of some Masoretic forms, the vocalisa3.9; m. Bekhorot 1.6; m. Temura 7.4–6), entails the possibility that many, if not all, of the apparent nifʿal forms conceal original qal internal passives. 202 The Historical Depth of the Tiberian Reading Tradition tion reflects a nitpaʿʿel/nifʿal B analysis (with gemination in first and middle radical), though the spelling is amenable to simple nifʿal interpretation, e.g., ‫‘ וְ ִתנַ ֵ ִ֖שּׂא‬and (his kingdom) will be exalted’ (Num. 24.7); ‫‘ וְ נִ ַכ ֵ ֵ֥פּר‬and (the blood guilt) will be atoned for’ (Deut. 21.8); ‫‘ וְ ִנָּ֣וַ ְסרּו‬and (all women) should take warning’ (Ezek. 23.48); ‫(‘ ִת ַכ ֶסה‬hatred) will be covered’ (Prov. 26.26); ‫(‘ יִ נַ ְשּׂאּו‬and the sons of the violent of your people) will rise up’ (Dan. 11.14); ‫‘ וַ יִ נַ ֵשּׂא‬so he was exalted’ (2 Chron. 32.23); several of these come in exilic or post-exilic material. In a few cases, however, suffix conjugation forms in texts from no earlier than the Exile cannot be read as nifʿal, and are more plausibly interpreted as hitpaʿʿel forms with assimilated tav: ‫‘ ִהנַ ְבאּו‬they prophesied’ (Jer. 23.13); ‫‘ וְ ִהנֶ ָ ָ֑ח ְמ ִתי‬and I will be satisfied’ (Ezek. 5.13); ‫אתי‬ ִ ‫‘ וְ ִהנַ ֵ ִ֖ב‬and I prophesied’ (Ezek. 37.10). Clearly, these probable consonantal hitpaʿʿel forms with assimilated tav lend credence to the vocalisation of the preceding apparently nitpaʿʿel forms (see below, ch. 13, §2.1). 3.0. Iron Age Epigraphy and the Tiberian Classical Biblical Hebrew Written Tradition Though many nifʿal readings of otherwise ambiguous consonantal forms are probably secondary, a crucial consideration is that the use of nifʿal and, therefore, the potential for nifalisation, were not restricted to post-exilic times. In other words, while the association between nifalisation and Second Temple Hebrew is meaningful, it is not exclusive. There are also indications of early nifalisation, specifically in classical consonantal evidence. 10. Nifalisation 203 3.1. Early Nifʿal Usage Especially important in this connection are early nifʿal forms that are primary derivations rather than instances of secondary nifalisation of originally qal forms. From Iron Age inscriptions, consider the nifʿal imperative ‫‘ השמר‬take care!’ (Lachish 3.21) and the infinitive ]‫‘ להנ[קב‬to be he[wn]’ (Siloam 1.2).15 While the for- mer is analysable as semantically middle, the latter would seem to be medio-passive. Turning to BH, in the case of many common orthographically unequivocal nifʿal verbs, qal counterparts are rare or even non-existent. Thus, ‫‘ נִ ְּפ ַרד‬separate (intr.)’ has consistent nifʿal spelling and vocalisation throughout BH. Likewise, though a vestige of qal ‫‘ ָש ַאר‬remain’ (1 Sam. 16.11) is once attested in CBH, the synonymous nifʿal ‫ נִ ְּש ַאר‬is unambiguously represented in all biblical chronolects.16 3.2. Qal Internal Passive > Nifʿal The same holds true for the qal internal passive’s replacement by nifʿal. There is ample early unambiguous consonantal evidence of 15 N-stem ‫‘ נאנח‬groan’ occurs in the eighth-century Deir Alla inscription (see KAI 312 B.12). 16 It is worth noting that such distributions of medio-passive, reflexive, and/or intransitive nifʿal forms with rare or unattested qal cognate synonyms are common. Limiting the discussion to verbs found in MT Gen- esis, cases of verbs with unambiguous nifʿal consonantal forms in the Bible include ‫‘ נֵ אֹות‬be willing’, ‫‘ נִ ְּב ַהל‬fear’, ‫נֹותר‬ ַ ‘remain’, ‫‘ נֶ ְּח ָבא‬hide’, ‫נִ ְּכ ַמר‬ ‘be hot’, ‫‘ נִ ְּכ ַסף‬yearn’, ‫‘ נִ ְּלוָ ה‬join’, ‫‘ נִ ְּמ ַלט‬take refuge’, ‫‘ נִ ְּס ַתר‬hide’, ‫‘ נִ ְּפ ָלא‬be wonderful’, ‫‘ נִ ְּש ַבע‬swear’, ‫‘ נִ ְּש ַחת‬be destroyed’, ‫‘ נִ ְּש ַמד‬be destroyed’, ‫נִ ְּש ַען‬ ‘lean’. In many of these cases, the corresponding transitive form is hifʿil. 204 The Historical Depth of the Tiberian Reading Tradition nifʿal semantically equivalent to qal internal passive, e.g., ‫נִ ְל ָ ְָ֑קח‬ ‘(the Ark of God) has been taken’ (1 Sam. 1.4, etc.; cf. ‫)ל ַקח‬, ֻ ‫יִ ָנ ֵֶ֥תן‬ ‘(straw) will (not) be given’ (Exod. 5.18; 2 Sam. 21.6 ketiv; cf. ‫)יֻ ַתן‬. In light of this evidence, the nifʿal’s eclipsing of qal internal passive should be seen as a process that was already underway in the Iron Age, only reaching its conclusion in the Second Temple Period. Given the antiquity of nifʿal’s association with middle and medio-passive semantics, along with the gradual pace of language change, it stands to reason that cognate qal internal passive and nifʿal forms might have coexisted over an extended period of time. Hughes (1994, 74–75) has sought to discern semantic and syntactic differences in CBH, before the qal internal passive fell out of use. He argues that in some cases the nifʿal serves as an intransitive against the strictly passive force of the qal internal passive, but the pervasiveness of this distinction is questionable. As such, the possible co-occurrence of qal internal passive and passive nifʿal forms, even in close proximity, should not be dismissed. Consider examples (13). (13) ‫ּומת ַת ַחת יָ ָׂ֑דֹו נ ָֹׁ֖קם יִ נ ֵ ִָֽקם׃ ַ ֵ֥אְך‬ ֶ֖ ֵ ‫תֹו ַב ֵֵ֔ש ֶבט‬ ֙ ‫ת־א ָמ‬ ֲ ‫ת־ע ְּב ֶ֜דֹו ַ֤אֹו ֶא‬ ַ ‫וְּ ִ ַֽכי־יַ ֶכ ֩ה ִָֹ֨איש ֶא‬ ‫מד ל ֹא י ִ֔ ַֻּׁקם ִ ֵ֥כי ַכ ְּס ֶ֖פֹו ַֽהּוא׃‬ ֹ ָׂ֑ ‫יֹומיִ ם יַ ֲע‬ ֶ֖ ַ ‫ִאם־י֛ ֹום ֵ֥אֹו‬ ‘When a man strikes his slave, male or female, with a rod and the slave dies under his hand, he shall be avenged. But if the slave survives a day or two, he is not to be avenged, for the slave is his money.’ (Exod. 21.20–21) There seems no reason to doubt the authenticity of the stem diversity between the qal infinitive absolute and nifʿal finite cognate in the tautological construction ‫‘ נָ ִ֖קֹם יִ נָ ְֵקם‬he should surely be 10. Nifalisation 205 avenged’ (Exod. 21.20) or between the aforementioned nifʿal and the following verse’s qal passive ‫יֻקם‬ ָ֔ ַ ‘he will (not) be avenged’ (Exod. 21.21). A similar consideration applies to the contrasting cognate forms in bold in example (14). (14) ‫ה־לְך ְּב ֶ ֶ֖ט ֶרם אֶ לָ ַּ ּ֣קח‬ ָ ֵ֔ ‫יש ֙ע ְּש ַאל֙ ָמה ֶ ַֽא ֱע ֶש‬ ָ ‫ל־א ִל‬ ֱ ‫וַ יְּ ִהי ְּכ ָע ְּב ָ ֵ֗רם וְּ ֵָֹ֨א ִל ֶָ֜יהּו ָא ַ ַ֤מר ֶא‬ ‫ית ִל ְּש ָׂ֑אֹול‬ ָ ‫אמר ִה ְּק ִש‬ ֶ ֹ ‫רּוחָךֶ֖ ֵא ָ ַֽלי׃ וַ ֶ֖י‬ ֲ ‫י־ש ַ ֵ֥נֵּ֣יִ ם ְּב‬ ְּ ‫י־נ֛א ִ ַֽפ‬ ָ ‫יה‬ ִ ‫ישע ִ ַֽו‬ ֵָ֔ ‫אמר ֱא ִל‬ ֶ ֹ ‫ֵמ ִע ָ ָׂ֑מְך וַ י‬ ‫יִה ֶיַֽה׃‬ ְּ ‫ם־איִ ן ֵ֥ל ֹא‬ ֶ֖ ַ ‫י־לָך ֵֵ֔כן וְּ ִא‬ ְּ ‫ם־ת ְּר ֶָֹ֨אה א ִֶֹ֜תי לֻׁ ָ ָּ֤קח ֵ ַֽמ ִא ָת ְ֙ך יְּ ִ ַֽה‬ ִ ‫ִא‬ ‘When they had crossed, Elijah said to Elisha, “Ask what I shall do for you, before I am taken from you.” And Elisha said, “Please let there be a double portion of your spirit on me.” And he said, “You have asked a hard thing; yet, if you see me being taken from you, it shall be so for you, but if you do not see me, it shall not be so.”’ (2 Kgs 2.9–10) The morphological diversity of the neighbouring nifʿal ‫‘ ֶא ָל ְַקח‬I am taken’ (2 Kgs 2.9) and qal passive participle ‫‘ ֻל ָ ָ֤קח‬being taken’ (2 Kgs 2.10) indicates the chronological coexistence of the two forms. Similar stem diversity may also be original in cases such as qal passive ‫( יֻ ַ ִ֖תן‬Num. 26.54) and nearby nifʿal ‫ נִ ַ ָ֤תן‬eight verses later (Num. 26.62)—though the total absence of qaṭal ‫ *נֻ ַתן‬raises suspicions. While many cases of qaṭal ‫ נִ ַ ָ֤תן‬may not involve dissonance between the consonants and vocalisation, at least some probably reflect original ‫ *נֻ ַתן‬reread as nifʿal. Finally, consider the preservation of qal internal passive ‫יֻ ַתן־‬ ‘let there be given’ in the qere of 2 Sam. 21.6 against the apparently synonymous nifʿal ‫ ינתן‬in the ketiv. Hughes (1994, 76) opines: 206 The Historical Depth of the Tiberian Reading Tradition In this instance it seems likely that the Qere has preserved the original reading, providing an interesting contrast to the normal pattern of revocalisation. Here, the process of replacing qal passive forms by niphal forms has affected the consonantal text, but has not affected the Masoretic reading tradition. This may be correct. Yet, it bears emphasising that the shift to nifʿal in the written tradition allegedly responsible for the ketivqere dissonance may well reflect truly ancient diversity in the combined Tiberian written and reading tradition. In other words, given evidence for the coexistence of the qal internal passive and nifʿal, this may be a genuine instance of early textual fluctuation. 3.3. Early Nifalisation of Participial Forms Returning to the previously discussed qal > nifʿal shifts ‫נֹולד > יֻ ַלד‬ ַ ‘be born’ and ‫‘ נִ ְכ ַשל > ָכ ַשל‬stumble’—while unambiguous consonantal evidence of N-stem finite and infinitival verbal forms is limited chiefly to late material, the relevant N-stem participles— with consonantally unambiguous forms—are attested in CBH sources. It may be relevant that forms such as ‫נֹול ִדים‬ ָ ‫‘ ַה‬the ones born’ (Gen. 48.5; see also Gen. 21.3; 1 Kgs 13.2) and ‫‘ נִ ְכ ָש ִלים‬feeble ones’ (1 Sam. 2.4) have nominalised adjectival, rather than truly eventive semantics. Such substantival and descriptive participle functions, conveying characteristics rather than actions, perhaps proved fertile ground for the initial nifʿal encroachment into semantic values formerly belonging to qal.17 Even so, the 17 I am grateful to my friend and colleague Geoffrey Khan for a helpful conversation on this point. Not unrelatedly, Khan (2020, I:80) raises the possibility that the nufʿal < nifʿal shift in the realisation of Chronicles’ 10. Nifalisation 207 Iron Age epigraphic and CBH usage of unambiguous consonantal nifʿal forms with eventive and actional semantics (see above, earlier in this section) confirms that the transparent middle marking of intransitive, medio-passive, and passive verbs via nifalisation is not exclusively late, but can legitimately be characterised as an Iron Age process the effects of which became most perceptible in Second Temple Hebrew. 4.0. Conclusion It has often been claimed that secondary developments in the reading component of the Tiberian tradition that was wedded to the CBH written component are due to anachronistic, post-biblical impositions of RH onto BH (Lambert 1900; Ginsberg 1929; 1934; 1936; see also Blau 2010, 213–14), “[b]ut the discoveries of the Qumran texts and subsequent research on Second Temple Hebrew show that many of the later features underlying the vocalisation existed already in the Second Temple period” (Joosten 2015, 30). In the specific case of nifalisation, affinities between the Tiberian reading tradition, on the one hand, and the LBH written tradition, DSS Hebrew, SH, the Hebrew of BS, and RH, on the other, demonstrate that the linguistic development in question had taken place long before the Masoretes engaged in the preservation and transmission of the tradition in the Middle Ages. Jeremy Hughes discussed the Tiberian secondary vocalisation shift from qal internal passive to nifʿal in a study entitled ‫נּולדּו‬ ְּ ‘were born’ (1 Chron. 3.5; 20.8) reflects an interpretive distinction according to which nufʿal was considered more eventive than nifʿal in the case of the root ‫יל"ד‬. 208 The Historical Depth of the Tiberian Reading Tradition “Post-Biblical Features of Biblical Hebrew Vocalisation.” Notwithstanding the provocative title, Hughes (1994, 75–76) offers a remarkably nuanced summary on the relevant process of nifalisation: First, it represents a continuation of a process which had begun in classical biblical Hebrew, where the niphal conjugation replaced the qal passive conjugation as the normal syntactic passive of most verbs. Secondly, this process was also continued in late biblical Hebrew, where the niphal conjugation replaced the qal passive conjugation as the normal syntactic passive of all verbs. [emphasis in the original] The most revealing element in Hughes’s summary is the pronounced continuity between the Tiberian reading tradition and both CBH and LBH. Given the already advanced stage of the shift in LBH, there is arguably no reason to class the Tiberian reading tradition’s penchant for nifalisation a ‘post-biblical’ feature of vocalisation. Rather, this proclivity for nifʿal seems very much in line with LBH conventions, though it also preserves features lost in more representative forms of Second Temple Hebrew, like LBH, DSS Hebrew, SH, BS’s Hebrew, and RH. This all points to the plausibility of a theory whereby the Tiberian reading tradition crystallised around the time that the LBH texts were being written. If so, it may be expected to preserve a great deal of authentic First Temple detail along with evidence of secondary development rooted in Second Temple linguistic drift. 11. HIFILISATION As part of the broad morphosemantic shift in ancient Hebrew away from the G-stem in favour of morphology perceived to have greater semantic transparency, a number of qal verbs shifted to hifʿil. The phenomenon is variously manifested: (a) certain apparently qal verbs with ambiguous forms analysable as hifʿil—espe- cially certain morphologically weak and semantically stative verbs—secondarily developed unambiguous hifʿil forms; (b) hifilisation affected qal consonantal forms amenable to hifʿil pro- nunciation, resulting in suppletive qal-hifʿil paradigms—including the occasional hifʿil vocalic realisation of consonantal forms ill-suited to hifʿil reinterpretation; (c) hifilisation was exploited for purposes of semantic and/or grammatical disambiguation. Individual examples of the phenomenon were noticed early on by the likes of S. D. Luzzato (1827–1828, 125) and F. Böttcher (1866–1868, II:279–80, 436). Yalon’s (1971, 43–54) treatment remains an excellent source of examples, discussion, and bibliography. 1.0. Second Temple Evidence 1.1. Tiberian Late Biblical Hebrew Hifʿil forms are by no means rare in Tiberian CBH and there is abundant morphological continuity between CBH and LBH. Even so, LBH reveals unmistakable signs of the advancement of the process of hifilisation vis-à-vis CBH. © 2023 Aaron D. Hornkohl, CC BY-NC 4.0 https://doi.org/10.11647/OBP.0310.11 210 The Historical Depth of the Tiberian Reading Tradition 1.1.1. Hifʿil Innovations in Late Biblical Hebrew This is especially clear in the case of qal verbs that are joined or replaced in LBH by hifʿil synonyms (Moreshet 1996).1 qal ‫ > זָ נַ ח‬hifʿil ‫‘ ִהזְ נִ ַיח‬reject’ The only remarkable aspect of the qal’s distribution is that it is absent from LBH (Hos. 8.3, 5; Zech. 10.6; Ps. 43.2; 44.10, 24; 60.3, 12; 74.1; 77.8; 88.15; 89.39; 108.12; Lam. 2.7; 3.17, 31), while the hifʿil form occurs only in LBH (1 Chron. 28.9; 2 Chron. 11.14; 29.19).2 qal ‫ > ָל ַעג‬hifʿil ‫‘ ִה ְל ִעיג‬mock’ The qal (2 Kgs 19.21; Isa. 37.22; Jer. 20.7; Ps. 2.4; 59.9; 80.7; Job 9.23; 11.3; 22.19; Prov. 1.26; 17.5; 30.17) occurs alongside the hifʿil (Ps 22.8; Job 21.3) in CBH texts and/or diachronically ambiguous material, but LBH proper knows only the hifʿil alter- native (Neh. 2.19; 3.33; 2 Chron. 30.10), with no obvious difference in meaning from the qal. qal ‫ > ָבזָ ה‬hifʿil ‫‘ ִה ְבזָ ה‬despise’ The qal occurs throughout CBH and LBH (Gen. 25.34; Num. 15.31; 1 Sam. 2.30; 10.27; 17.42; 2 Sam. 6.16; 12.9, 10; Isa. 49.7; Ezek. 16.59; 17.16, 18, 19; 22.8; Mal. 1.6; Ps. 22.25; 51.19; 1 Cf. Yalon (1971, 43–54), who argues that many of the apparent hifʿil prefix conjugation forms are actually of the qal stative yaqṭel pattern. 2 Excluded from this discussion is the form ‫(‘ וְּ ֶה ֶאזְּ ִניחּו‬canals) will become foul’ (Isa. 19.6) on the grounds that it represents a separate lexeme. Cf. ‫( והאז֯ נ֯ [יחו‬4Q56 f10–13.11) || ‫( והזניחו‬1QIsaa 15.10). 11. Hifilisation 211 69.34; 73.20; 102.18; Prov. 14.2; 15.20; 19.16; Est. 3.6; Neh. 2.19; 1 Chron. 15.29; 2 Chron. 36.16), whereas the apparently synonymous hifʿil infinitive ‫ ְל ַה ְבזֵ֥ ֹות‬comes in BH only in Esther (1.17).3 qal ‫ > ָר ַעד‬hifʿil ‫‘ ִה ְר ִעיד‬tremble’ No derivation is common in BH, but the distribution pattern reflects LBH preference for hifʿil (Dan. 10.11; Ezra 10.9) over qal (Ps. 104.32). qal ‫ > ָׂש ַחק‬hifʿil ‫‘ ִה ְׂש ִחיק‬laugh’ If the hifʿil in 2 Chron. 30.10 has the meaning ‘laugh’, then this comes in place of the CBH qal form with that meaning. 1.1.2. Qal > Hifʿil Movement in the Case of Stative and Inchoative Verbs Another result of hifilisation is the shift from qal to hifʿil in the case of verbs with stative or inchoative semantics. The alternation of qal ‫ ָצ ַלח‬and hifʿil ‫‘ ִה ְּצ ִליח‬succeed, prosper (intr.)’ is illuminating in this connection. Observe Table 1. 3 The shift of transitive semantics from qal to hifʿil evidently opened the door to the innovation of middle semantics for the qal, as in ‫וַ יִ ֶבז‬ ‫‘ ְּב ֵע ֵ֗ ָיניו ִל ְּשֹלַ֤ ח יָ ֙ד ְּב ָמ ְּר ֳד ַכי ְּל ַב ֵ֔דֹו‬but it was disdainful in his eyes to send his hand against Mordechai alone’ (Est. 3.6). 212 The Historical Depth of the Tiberian Reading Tradition Table 1: Qal and hifʿil of ‫ צל"ח‬in the MT (see §5.1 for citations) qal+‫רּוח‬ ַ intr. trans. qal+‫רּוח‬ ַ hifʿil qal hifʿil intr. trans. hifʿil qal hifʿil Gen. 0 6 0 1 Ezek. 0 0 5 0 Num. 0 0 1 0 Amos 0 0 1 0 Deut. 0 1 0 0 Ps. 0 2 1 1 Josh. 0 1 0 0 Prov. 0 0 0 1 Judg. 3 1 0 0 Dan. 0 0 1 4 Sam. 5 0 1 0 Neh. 0 2 0 0 Kgs 0 0 0 2 Chron. 0 1 0 12 Isa. 0 2 2 0 LBH 0 3 1 16 Jer. 0 0 5 3 TOTALS 8 16 17 24 Excluding from consideration the specific qal idiom ‫רּוח‬ ַ ‫ָצ ְּל ָחה‬ ‫‘ יהוה ַעל‬the spirit of the LORD came over’ along with transitive usages of hifʿil ‫ה ְּצ ִל ַיח‬, ִ one is left with apparently synonymous qal and hifʿil forms vying for the intransitive sense of ‘succeed, prosper’. It would seem that the process of hifilisation began rather early, since both the qal and the hifʿil are attested in CBH material (as well as in texts of ambiguous date), and was quite advanced by the Second Temple Period, as LBH shows preference for hifʿil over qal by a margin of 16 to 1. Similar encroachment of hifʿil verbs into the stative or intransitive semantic domains originally occupied by qal include the following: qal ‫*ש ֵמן‬ ָ > hifʿil ‫‘ ִה ְש ִמין‬become fat’ The classical, semantically predictable combination of stative qal (Deut. 32.15, 15; Jer. 5.28) and transitive hifʿil (Isa. 6.10) contrasts with the late stative hifʿil in LBH (Neh. 9.25). 11. Hifilisation 213 qal ‫ > ָר ַשע‬hifʿil ‫‘ ִה ְר ִש ַיע‬be wicked, commit wickedness’ Stative/intransitive qal and transitive hifʿil ‫‘ ִה ְר ִש ַיע‬condemn’ rep- resent a typical classical combination. Occasionally, the hifʿil seems to intrude into the semantic space originally occupied by the qal, with most of these in LBH (Ps. 106.6; Job 34.12; Dan. 9.5 [cf. 9.15]; 11.32; 12.10; Neh. 9.33; 2 Chron. 20.35). qal ‫ > גָ ַדל‬hifʿil ‫גדיל‬ ִ ‫‘ ִה‬grow, become great’ Common in CBH are stative qal ‫‘ גָ ַדל‬grow, become great’ and transitive hifʿil ‫‘ ִהגְ ִדיל‬magnify’. While the poetic idiom ‫‘ ִהגְ ִדיל ַעל‬act arrogantly against, taunt’ is common, hifʿil forms with no direct or indirect object, whether interpreted as ‘act arrogantly’ or ‘grow, become great’ are restricted to later material (Lam. 1.9; Dan. 8.4, 8, 11, 25). 1.1.3. Hifilisation of Qal II-y Verbs ‫בי"ן‬ A different manifestation of hifilisation particularly (though not exclusively) characteristic of Tiberian LBH has resulted from the formal identity of the prefix conjugation forms of qal and hifʿil II- y verbs, e.g., ‫‘ יָ ִבין‬he understands, will understand’. Consider, in Table 2, the distribution of unequivocal qal forms, ambiguous qal/hifʿil, and unequivocal hifʿil forms. 214 The Historical Depth of the Tiberian Reading Tradition Table 2: Qal and hifʿil of ‫ בי"ן‬in the MT (see §5.1 citations) qal ambiguous hifʿil qal ambiguous hifʿil Deut. 1 1 0 Job 0 13 2 Sam. 0 2 0 Prov. 1 13 9 Kgs 0 0 2 Dan. 3 7 11 Isa. 0 7 5 Ezra 0 1 1 Jer. 1 1 0 Neh. 0 2 6 Hos. 0 2 0 Chron. 0 0 9 Mic. 0 0 1 TOTALS 10 57 55 Unambiguous qal forms are rare in the MT, while unambiguous hifʿil forms are over five times as common. What is more, an argument can be made that, in view of the complete absence of unambiguous qal forms and the frequency of unambiguous hifʿil forms in certain texts, some of the ambiguous forms, especially those in Isaiah and Job, should be considered probable cases of hifʿil. While the few qal forms are distributed throughout all historical phases of biblical literature, and while there are no grounds for characterising the hifʿil as distinctively late, it seems significant that early unequivocal qal forms are limited to poetry. A plausible supposition is that rather early on in the history of BH, analysis of original qal ‫ יָ ִבין‬and the like as hifʿil led to the secondary development of forms like ‫ ֵה ִבין‬and ‫ל ָה ִבין‬, ְּ which are certainly the norm in LBH, but may already have been dominant in CBH, too (Nöldeke 1904, 34–47; Blau 2010, 255, §4.3.8.7.2.8; cf. Bergsträsser 1918–1929, II:153, §28t). ‫זי"ד‬ The case of forms of the root ‫‘ זי"ד‬act arrogantly’ is similar. There are unequivocally qal forms (Exod. 18.11; Jer. 50.29) and forms amenable to both qal and hifʿil analysis (Exod. 21.14; Deut. 1.43; 11. Hifilisation 215 17.13; 18.20), with unequivocally hifʿil forms limited to LBH (Neh. 9.10, 16, 29).4 Unambiguous hifʿil forms are also attested in the NBDSS, BS, and RH. ‫שי"ם‬ Likewise, hifʿil analysis of the ambiguous prefix conjugation of qal ‫יָ ִׂשים‬-‫ ָׂשם‬-‫‘ ָׂשם‬put’ led in the BH written tradition to rare unambiguous hifʿil forms, such as suffix conjugation ‫יהּו‬ ֙ ‫‘ וַ ֲה ִ ַֽשמ ִֹ֙ת‬and I will make him’ (Ezek. 14.8), imperative ‫ימי‬ ִ ‫‘ ָה ִש‬set (FS)’ (Ezek. 21.21), participle ‫‘ ִֵׁ֝מ ִֵ֗שים‬someone (MS) who regards’ (Job. 4.20). The hifʿil form is known also from BS (SirA 4v.22 || Sir. 11.30), and RH (Sifre Devarim; Tosefta; Yerushalmi; Bavli). This has been cited as the reason for the secondary development of qal ‫( יָ שּום‬Blau 2010, 255, §4.3.8.7.2.8). For the potentially hofʿal qere ‫יּושם‬ ַ֤ ַ ַ‫ ו‬for ketiv qal passive ‫( ויישם‬Gen. 24.33)—the latter a match for the qal passive ‫ֵ֥ישם‬ ֶ ִ‫( וַ י‬Gen. 50.26) (Blau 2010, 97, §3.4.3.3, see below §2.0). ‫קי"א‬, ‫רי"ב‬, and ‫לי"ץ‬ Clear qal, hifʿil, and equivocal derivations of ‫‘ קי"א‬vomit’, ‫רי"ב‬ ‘quarrel’, and ‫‘ לי"ץ‬scoff’ also seem to compete in the Tiberian written tradition. For ‫ קי"א‬unambiguous hifʿil forms come in Prov- erbs (23.8) and the Mishna (Para 9.3). In the case of ‫ רי"ב‬and ‫לי"ץ‬, it may be significant that the apparently earliest unambiguous hifʿil morphology is limited to participles with nominal seman- tics, while the more transparently verbal forms ‫(‘ ֱה ִל ֻיצנִ י‬the inso4 Excluded here on semantic grounds is the morphologically ambiguous ‫‘ וַ יָ ֵֵּ֥֣זֶ ד‬and (Jacob) cooked’ (Gen. 25.29). 216 The Historical Depth of the Tiberian Reading Tradition lent) have derided me’ (Ps. 119.51), ‫‘ ולהליץ‬and to deride’ (4Q184 f1.2), and ‫‘ להריב‬to contend’ (4Q390 f2i.6) all come in acknowledged late material. Leaving behind hollow roots, similar distributional patterns are known for other verbs. Consider ‫‘ נח"י‬lead, guide’ in Table 3. ‫נח"י‬ Table 3: Qal and hifʿil of ‫ נח"י‬in the MT (see §5.1 for citations) qal ambiguous hifʿil qal ambiguous hifʿil Gen. 1 0 1 Isa. 2 1 0 Exod. 4 0 0 Ps. 6 12 0 Num. 0 1 0 Job 0 3 0 Deut. 0 1 0 Prov. 0 3 0 Sam. 0 1 0 Neh. 0 0 2 Kgs 0 2 0 TOTALS 13 24 3 While the evidence arguably reflects a state of early mixed usage, the only LBH forms, both infinitives, are unequivocally hifʿil. Hifʿil infinitives are also attested in the NBDSS (1QS 9.18 || 4Q256 18.1 || 4Q259 3.16) and in the Tiberian reading tradition’s pointing of the ostensibly qal infinitive in Exod. 13.21. Sig- nificantly, three of the four hifʿil cases in the Tiberian Torah have consonantal forms more fitting for qal (Exod. 13.21) or equally suitable to qal and hifʿil analyses (Num. 23.7; Deut. 32.12). ‫יס"ף‬ Another interesting case is that of qal ‫ יָ ַסף‬versus hifʿil ‫הֹוסיף‬ ִ ‘add, repeat’. See Table 4. 11. Hifilisation 217 Table 4: Qal and hifʿil of ‫ יס"ף‬in the MT (see §5.1 for citations) qal ambiguous hifʿil qal ambiguous hifʿil Gen. 2 12 0 Nah. 0 Exod. 0 7 1 Lev. 7 3 1 Zeph. 0 Num. 3 5 0 Ps. Deut. 4 8 3 PENT. 16 35 Josh. 0 Judg. 0 1 1 0 11 18 0 1 7 Job 0 5 6 5 Prov. 0 6 7 0 2 Ruth 0 0 1 2 8 3 Qoh. 0 0 5 Sam. 5 17 9 Lam. 0 0 3 Kgs 2 4 7 Est. 0 1 0 F. PROPH. 9 L. PROPH. 8 29 21 Dan. 0 1 0 Isa. 6 4 10 Ezra 0 0 1 Jer. 2 0 1 Neh. 0 0 1 Ezek. 0 3 0 Chron. 1 2 8 Hos. 0 2 1 WRITINGS 2 14 39 Joel 0 1 0 LBH+ 1 4 15 Amos 0 0 4 TOTALS 35 89 83 Jon. 0 0 1 A CBH situation of mixed usage, with apparent qal dominance in the Pentateuch and apparent hifʿil dominance in the Prophets and Writings, gives way to striking hifʿil supremacy in LBH. See below, §2.0, on the Tiberian reading tradition. ‫יל"ד‬ Related to the late extension of hifʿil was exploitation of C-stem morphology for disambiguating distinct nuances originally subsumed within the qal, for example the use of qal ‫ יָ ַלד‬for the pro- creative act associated with both mother ‘bear’ and father ‘beget, 218 The Historical Depth of the Tiberian Reading Tradition sire’ (Driver 1882, 209; Joüon 1920, 359; Hendel 2000, 38–425). Consider Table 5. Table 5: Qal and hifʿil masculine finite verbs and active participles of ‫ יל"ד‬in the MT (see §5.1 for citations) Gen. Num. Deut. Judg. Kgs Isa. Jer. Ezek. Hos. Zech. qal 12 0 1 0 0 3 2 0 1 2 hifʿil 42 2 2 1 1 4 2 2 0 0 Ps. Job Prov. Ruth Qoh. Dan. Neh. Chron. TOTALS qal 1 1 4 0 0 1 0 7 35 hifʿil 0 1 0 9 2 0 4 83 154 Again, the figures appear to indicate that hifilisation was well underway already in CBH, but that it was not until LBH that qal 5 Hendel (2000) focuses on this issue in a discussion of the dating of Pentateuchal sources. On the one hand, he argues that “the complementary distribution of yālad (Qal) for ‘beget’ in the J source and hôlîd (Hiphil) for ‘beget’ in the P source is attributable to a diachronic devel- opment in Classical Hebrew Biblical” (Hendel 2000, 42), i.e., not diachronic development between CBH and LBH. On the other hand, he dates P to the exilic or early Persian Period (Hendel 2000, 46). Hendel’s figures differ from those given above, because he focuses on genealogies, whereas the figures here are mechanical, including metaphorical usages. For example, one of the cases of qal in Jeremiah should probable be considered a counterexample of the semantics ‘father, sire’ for qal ‫יָ ַלד‬. Consider the verse ‫לּו־נא ְּּור ֵ֔אּו ִאם־י ֵלֹׁ֖ד ז ָָכ ָ֑ר‬ ָ ‫‘ ַש ֲא‬Ask now, and see, can a man bear a child?’ (Jer. 30.6). While technical genealogical usage of qal ‫‘ יָ ַלד‬father, sire’ is still found in LBH, the form had become especially associated with female agency prior to LBH. 11. Hifilisation 219 ‫‘ יָ ַלד‬beget’ was effectively supplanted. Outside of LBH proper and Qohelet, the figures are qal 27, hifʿil 66. In LBH proper and Qohelet combined, they are qal 8, hifʿil 90. Moreover, six of the eight LBH qal cases come in texts borrowed from the Pentateuch (1 Chron. 1.10, 11, 13, 18, 18, 20 || Gen. 10.8, 13, 15, 24, 24, 26). In sum, the picture that emerges from the Tiberian LBH written tradition involves a trend in favour of forms that either can or must be read as hifʿil replacing one of mixed qal-hifʿil or dominant qal morphology. 1.2. Dead Sea Scrolls Hebrew 1.2.1. The Biblical Dead Sea Scrolls The BDSS show relatively little evidence of hifilisation beyond that also exhibited in the Tiberian written tradition. Where the BDSS have parallels to the MT involving the verbs discussed above, §1.1, they show nearly the same distribution of morphology, whether qal, ambiguous, or hifʿil, with mixed usage in CBH material and hifʿil concentration in LBH. The lone exception in this regard is the verb represented by qal ‫ יָ ַסף‬and hifʿil ‫הֹוסיף‬. ִ In the case of this verb, there are several instances in CBH material in which an unequivocal DSS hifʿil parallels a MT qal or ambiguous form: (1) ‫וסיפו‬ ̇ ̇‫‘ י‬they will (not) continue’ (4Q30 f24.2) || MT ‫יֹוספּו‬ ִ ‘they will (not) continue’ (Deut. 13.12) (2) ‫‘ יוסיף‬will add’ (4Q35 f1.9) || MT ‫‘ י ֵ ָֹ֧סף‬will add, is adding’ (Deut. 1.1) 220 The Historical Depth of the Tiberian Reading Tradition A few such cases centre on Deut. 5.25:6 (3) ‫מו]סי̇ פים‬ ̇ ‘(we) co]ntinue’ (4Q37 3.7) || MT ‫‘ ִאם־י ְֹּס ִפים‬if (we) continue’ (Deut. 5.25) (4) ]‫‘ כי מ[וסיפים‬if (we) c[ontinue’ (4Q129 f1R.13) || MT ‫ִאם־‬ ‫‘ י ְֹּס ִפים‬if (we) continue’ (Deut. 5.25) (5) ‫מו]סיפים‬ ֯ ‘(we) co]ntinue’ (4Q135 f1.4) || MT ‫‘ ִאם־י ְֹּס ִפים‬if (we) continue’ (Deut. 5.25) (6) ‫יספים‬/‫מ‬ ̇ ‫‘ אם‬if (we) continue’ (4Q137 f1.30–31) || MT ‫ִאם־‬ ‫‘ י ְֹּס ִפים‬if (we) continue’ (Deut. 5.25) Though textual factors should also be considered, these cases of qal > hifʿil movement in acknowledged Second Temple scribal products tally with the process of hifilisation described above, in general, and in the case of the root ‫יס"ף‬, more specifically. 1.2.2. The Non-biblical Dead Sea Scrolls Less anchored within the biblical text, NBDSS material exhibits more pronounced effects of hifilisation than the BDSS. This is manifest in (a) the use of hifʿil verbs with biblical distribution limited to LBH (‫זנ"ח‬, ‫לע"ג‬, ‫)בז"י‬, (b) the replacement of sta- tive/intransitive qal verbs with hifʿil cognates, as in LBH (‫צל"ח‬, ‫)רש"ע‬, (c) the employment of unambiguous hifʿil forms of origi- nally qal verbs with ambiguous prefix conjugation forms (‫בי"ן‬, ‫זי"ד‬, ‫רי"ב‬, ‫לי"ץ‬, ‫נח"י‬, ‫יס"ף‬, ‫)שי"ר‬, and (d) exploitation of morpholog- 6 In examples (3)–(6), the potential sequences of both ‫י‬-‫ מ‬in ‫ אם יספים‬and ‫ו‬-‫מ‬-‫ מ‬in ‫ אם מוסיפים‬would have been vulnerable to graphic and/or phonetic corruption. 11. Hifilisation 221 ical distinction between qal and hifʿil for semantic differentiation (‫)יל"ד‬. Table 6 provides a quantitative summary. Table 6: Frequency of qal and hifʿil of select diachronically significant verbs in the NBDSS (see §5.2 for citations) qal ambiguous hifʿil qal ambiguous hifʿil (a) ‫זנ"ח‬ 0 2 3 (c) ‫בי"ן‬ 3 36 69 ‫לע"ג‬ 0 0 1 ‫זי"ד‬ 1 2 2 ‫בז"י‬ 9 2 1 ‫רי"ב‬ 4 11 1 0 1 0 ‫לי"ץ‬ 1 0 2 ‫שח"ק‬ 2 7 0 ‫נח"י‬ 1 0 3 ‫צל"ח‬ 1 4 1 ‫יס"ף‬ 2 7 29 ‫רש"ע‬ 1 2 10 (d) ‫יל"ד‬ 0 0 7 ‫גד"ל‬ 1 1 0 (b) ‫רע"ד‬ Sometimes, the NBDSS fail to exhibit clear-cut cases of the diagnostically late hifʿil verbs (‫רע"ד‬, ‫שח"ק‬, ‫ )גד"ל‬or appear to favour the more classical alternative (‫בז"י‬, ‫)רי"ב‬. In other cases, the char- acteristically late hifʿil usage is conspicuously dominant (‫רש"ע‬, ‫בי"ן‬, ‫יס"ף‬, ‫)יל"ד‬. 1.3. Samaritan Hebrew A scriptural corpus embodying related but semi-independent written and reading components, the Samaritan biblical tradition has roots extending at least as far back as the Iron Age, but at the same time shows clear signs of late development. Morphological shifts from G- to C-stem in the Samaritan tradition, though noted, have not generally been discussed as part of a grammatical trend. Indeed, they go unmentioned in Ben-Ḥayyim’s discussion of regular stem shifts (2000, 222–24, §§2.15.4–7), relegated to a few examples in a paragraph that begins “Other alternations between 222 The Historical Depth of the Tiberian Reading Tradition stems do not display general tendencies, but each individual verb must be explained separately, so that discussion of them belongs in a lexicon, not a grammar” (Ben-Ḥayyim 2000, 224, §2.15.8). As the ensuing discussion demonstrates, the applicability of this statement to hifilisation in the Samaritan tradition may be questioned, as the phenomenon is both more pervasive in SH than Ben-Ḥayyim implies and exhibits affinities to the same process in other Second Temple traditions. The relevant verbs may be divided into several categories. 1.3.1. Hifilisation of ‫‘ שי"ר‬sing’ and ‫‘ טמ"ן‬hide’ First are those verbs for which qal is standard in both Tiberian and Samaritan Hebrew, but which have undergone partial hifilisation in the latter, sometimes in line with trends seen in other manifestations of Second Temple Hebrew. An illustrative example is the Samaritan counterpart to Tiberian ‫‘ ָשר‬sing’. It has an unambiguous qal imperative (Exod. 15.21), ambiguous yiqṭol forms (Exod. 15.1a; Num. 21.17), and an unambiguous and synonymous hifʿil imperative according to the combined testimonies ̊̄ of the written and reading tradition: ‫ אשירו‬ašīru ‘sing (PL)!’ || MT ‫‘ ָא ִ ַ֤ש ָירה‬I would sing’ (Exod. 15.1b), which has also been tentatively read, with causative force, in the NBDSS: ‫מלי֯ הם במשקל תכן‬ ‫‘ וישרם כחלילים‬their words by weight he apportioned and caused them to sing like flutes’ (4Q434 f1i.9). Similarly, while Tiberian qal ‫ ָט ַמן‬is twice paralleled by its Samaritan qal counterparts, in the prefix conjugation (Exod. 2.12) and the passive participle (Deut. 33.19), on another occasion, MT qal ‫מן‬ ֹ ַ֤ ‫ || וַ יִ ְּט‬SP ‫ ויטמן‬wyåṭmǝn ‘and he hid (tr.)’ (Gen. 35.4), 11. Hifilisation 223 with no obvious distinction in meaning separating the qal and the hifʿil (also in the Masada BS material, RH, and the Tiberian CBH reading tradition; see below, §§1.4–5; 2.0). 1.3.2. Hifilisation of ‫‘ יל"ד‬bear (a child); beget, father, sire’ In the case of ‫יל"ד‬, like Tiberian Hebrew, SH generally distinguishes between qal ‫‘ ילד‬bear (a child)’ and hifʿil ‫‘ הוליד‬beget, fa- ther, sire’. On occasions where the MT presents a qal form that denotes ‘beget, father, sire’, SH does not tolerate the polysemy of the qal. Instead, the same morphosemantic shift observed above with regard to ‫‘ יל"ד‬father, sire’ in Tiberian BH (§1.1) and the NBDSS (§1.2.2) also obtains in SH, albeit inconsistently. On three occasions where the MT has qal ‫ יָ ַלד‬in the meaning ‘beget, father, sire’, the combined written-reading Samaritan tradition resorts to a hifʿil instead: Gen. 6.4; 10.8; 22.23. Hifilisation is not, however, the preferred Samaritan solution to the problem in the case of ‫יל"ד‬. A more common strategy for distinguishing the male procreative act from the female act denoted by the qal is the reading of forms that refer to the male as piʿʿel (see ch. 12, §1.3.1). 1.3.3. Hifilisation of ‫‘ יס"ף‬add, repeat, do again’ In one further case of partial hifilisation relative to the Tiberian tradition, the combined Samaritan written and reading tradition testifies to increased use of unequivocal hifʿil forms of ‫יס"ף‬. There is one case in which an unambiguous MT qal || SP hifʿil and 14 cases in which an MT form of ambiguous stem || SP unambiguous plene hifʿil. The opposite situation obtains just twice (see §5.3 for citations). Indeed, the situation in SH is one of orderly, if compli- 224 The Historical Depth of the Tiberian Reading Tradition cated, suppletion: all 3rd-person qaṭal forms and all participles are qal;7 all 1st- and 2nd-person qaṭal forms are piʿʿel; all yiqṭol forms and infinitives are hifʿil.8 1.3.4. Extensive Hifilisation More extensive shifts are also known. Consider the Tiberian qal verb ‫‘ ָר ָצה‬accept, be pleased, make amends for’. On six occasions, most involving consonantally ambiguous yiqṭol forms, the SP has a hifʿil (Gen. 33.10; Lev. 26.34, 41, 43, 43; Deut. 33.11), and on five more occasions, a nifʿal in the MT is paralleled by a passive hifʿil in the SP (Lev. 7.18; 19.7; 22.23, 25, 27).9 The Samaritan treatment of the Tiberian qal verbs ‫‘ ָח ַבש‬wrap, saddle’ and ‫ַיָקד‬ ‘light, kindle’ can also be analysed as one of wholesale hifilisation.10 7 ̊̄ Some apparent SP qal qaṭal forms of ‫ יסף‬yɑsəf, especially those parallel to Tiberian weqaṭal forms, are arguably interpretable as secondary hifʿil yiqṭol forms 8 According to Ben-Ḥayyim (1977, 123, 193), pronunciation of the yiqṭol forms reflects derivation from both ‫ יס"ף‬and ‫סו"ף‬. 9 In the remaining three cases, all consonantally unambiguous, the MT and SP agree on a nifʿal (Lev. 1.4), hifʿil (Lev. 26.34), and passive qal participle (Deut. 33.24). 10 See Ben-Ḥayyim (2000, 224, §2.15.8) on ‫יק"ד‬. Regarding ‫חב"ש‬: one form is unambiguously hifʿil according to the reading component of the Samaritan tradition, while the remaining three are analysable as either piʿʿel or hifʿil (Ben-Ḥayyim 123, §§2.2.1.2.2–3). 11. Hifilisation 225 1.3.5. Hifilisation and Levelling In other cases of apparent wholesale hifilisation, the result may be due partially to grammatical harmonisation, whereby an aberrant form was regularised in conformity with the majority. For example, in the MT ‫ צל"ח‬is normally represented by hifʿil forms whether the sense is transitive ‘cause to prosper’ (Gen. 24.21, 40, 42, 56; 39.3, 23; Deut. 28.29) or intransitive ‘succeed’ (Gen. 39.2), and these are all paralleled in the SP by hifʿil forms; on the one occasion where the MT has a qal intransitive, the SP reads it as a hifʿil (Num. 14.41). Likewise, the MT’s internal qal-hifʿil diversity in (7), is paralleled in the SP by hifʿil consistency (8): (7) ...‫עֹוד הַּ ְּׁצ ִפינֹו‬ ֮ ‫ֹלשה יְּ ָר ִ ַֽחים׃ וְּ לֹא־יָ ְּכ ָלה‬ ֵ֥ ָ ‫וַּ ִִֽת ְּׁצ ְּׁפנֵ ֹׁ֖הּו ְּש‬... ‘…and she hid him (qal) three months. And she could no longer hide him (hifʿil)…’ (Exod. 2.2–3) (8) ...‫ותצפנהו שלשה ירחים׃ ולא יכלה עוד הצפנהו‬... …wtåṣfīnēʾu ̊̄ yēˈrīm. wla ̊̄ yaka ̊̄ la ̊̄ ūd åṣfīnēʾu… šēlaša While this may well be due to the Samaritan version’s penchant for levelling, and though the orthography of ֒‫ ַה ְּצ ִפינֹו‬prevented harmonisation in favour of qal, the hifilisation in question is consistent with that seen in other Second Temple chronolects, such as BS and RH (see below, §§1.4–5). Similar situations of grammatical levelling arguably took place with ‫‘ נז"י‬sprinkle’, ‫רפ"י‬ ‘leave, slacken’, and ‫‘ שמ"ט‬drop, release’. 1.3.6. Hifilisation in the Case of Rare Verbs Finally, there are rarely occurring verbs in the Pentateuch that are qal in the MT and hifʿil in the SP, some representative of 226 The Historical Depth of the Tiberian Reading Tradition broader hifilisation patterns: ‫‘ חש"ך‬be/become dark’ (Exod. 10.15; cf. RH, CBH), ‫יע"ץ‬/‫‘ עו"ץ‬advise’ (Exod. 18.19; Num. 24.14; cf. Aramaic C-stem ‫)אמליך‬, ‫‘ נפ"ח‬blow’ (Gen. 2.7; cf. BH), ‫עק"ד‬ ‘bind’ (Gen. 22.9), ‫‘ צפ"י‬observe’ (Gen. 31.49). 1.3.7. Hifilisation Resulting in Suppletion Qal-hifʿil suppletion is comparatively more common in SH than in the Tiberian Torah. The suppletive paradigm of ‫ נח"י‬characteristic of the MT (§§1.1.3; 2.1) is also found in the SP. Consider also consistently qal Tiberian ‫‘ ָחנַ ן‬show mercy’—in the SP, conversely, it is generally qal where required by consonantal spelling (Gen. 33.5, 11), but otherwise hifʿil (Gen. 43.29; Exod. 33.19, 19; Num. 6.25; Deut. 7.2; 28.50), including a hifʿil reading in oppo̊̄ ̊̄ sition to qal spelling: ‫ וחנתי את אשר אחן‬wɑʾinti it ēšɑr ɑʾən || MT ‫ת־א ֶשר ָא ֵ֔חֹן‬ ֲ ‫ֹתי ֶא‬ ֙ ִ ‫‘ וְּ ַחנ‬and I will be gracious to whom I will be gracious’ (Exod. 33.19a). Various suppletive patterns obtain in the case of ‫‘ בא"ש‬stink’ (qal Exod. 7.18; hifʿil Exod. 7.21; 8.10; 16.20), ‫‘ גמ"ל‬repay, bear (fruit)’ (qal Gen. 50.5, 17; hifʿil Num. 17.23; Deut. 32.6), ‫‘ הד"ף < דו"ף‬thrust’ (qal Num. 35.20; hifʿil Num. 35.22; Deut. 6.19; 9.4), ‫‘ נג"ש‬oppress’ (qal Exod. 3.7; 5.6, 10, 13, 14; hifʿil Deut. 15.2, 3), ‫‘ סג"ר‬close’ (qal Gen. 19.6, 10; 14.3; hifʿil Gen. 2.21; 7.16), ‫‘ ער"ך‬arrange’ (qal Exod. 40.4; Lev. 1.7, 8; 6.5; hifʿil11 Gen. 22.9; Exod. 27.21; 40.23; Lev. 1.12; 24.3, 4, 8), and 11 ̊̄ Ben-Ḥayyim (1977, 217) analyses the SH forms ‫ ויערכו‬wyarrēku (Gen. 14.8) and ‫ ערכתי‬ʿarrikti (Num. 23.4) as piʿʿel. The former is alternatively analysable as hifʿil, which is indeed the analysis given in Ben-Ḥayyim (2000, 375a, cf. 375b). 11. Hifilisation 227 ‫‘ רמ"ש‬crawl’ (qal Gen. 1.21, 26, 28, 30; 7.8, 14, 21; 8.17, 19; Lev. 11.44, 46; Deut. 4.18; hifʿil Gen. 9.2; Lev. 20.25). 1.3.8. Hifilisation and Semantic Disambiguation Finally, SH seems to exploit hifilisation for purposes of distinguishing semantic nuance. ‫‘ רח"ק‬distance’ In the case of ‫‘ רח"ק‬distance’, the MT and SP agree on qal forms in the context of distance with no movement (Deut. 12.21; 14.24) and on hifʿil forms when agency and movement are involved (Gen. 21.16; 44.4; Exod. 8.24, 24; 33.7). Mismatch between MT qal and SP hifʿil obtains in the case of the metaphorical MT ‫ִמ ְּד ַבר־‬ ̊̄ ‫‘ ֶ ֶ֖ש ֶקר ִת ְּר ָ ָׂ֑חק‬keep far from a false charge’|| SP ‫ תרחק‬tɑˈrēq ‘distance yourself (?)’ (Exod. 23.7), where there is agency, but the matter of stasis versus movement is ambiguous. ‫‘ שב"ר‬buy/sell food’ SH also uses morphology to distinguish distinct senses of ‫שב"ר‬ ‘buy and sell food’ left indistinct in Tiberian Hebrew. Whereas the MT is content with a qal verb ‫ ָש ַבר‬meaning both ‘buy food’ (Gen. 41.57; 42.2, 3, 5, 7, 10; 43.2, 4, 20, 22; 44.25; 47.14; Deut. 2.6) and ‘sell food’ (Gen. 41.56), it also has a hifʿil form meaning ‘sell food’ (Gen. 42.6; Deut. 2.28). SH more strictly observes the morphosemantic distinction, reading Joseph’s action in ‫וַ יִ ְּשבֹר‬ 228 The Historical Depth of the Tiberian Reading Tradition ‘and he sold (grain to Egypt)’ (Gen. 41.56) as hifʿil ‫וישביר‬ wyašbǝr.12 1.4. Ben Sira Moreshet (1996) lists a number of verbs in BS that reflect hifilisation. Those relevant to Tiberian BH include: • • ‫‘ הדמיע‬weep’ (SirA 5r.19 || Sir. 12.16) ‫‘ הטמין‬hide (tr.)’ (Mas1h 3.17 || SirB 11r.7 || Sir. 41.15; see above, §1.3.1) • ‫ספיק‬/‫‘ הש‬be sufficient’ (Mas1h 5.4 || SirB 12r.9 || Sir. 42.17) • ‫‘ העריך‬arrange’ (SirB 9r.3 || Sir. 39.17; SirB 19v.12 || Sir. 50.18) • ‫‘ הרעיד‬tremble’ (SirB 8v.15 || Sir. 38.25; see above, §§1.1.1; 1.2.2) • ‫‘ השים‬put’ (SirA 4v.22 || Sir. 11.30; see above, §1.1.3)13 To Moreshet’s list may be added: • • 12 ‫‘ האריח‬travel’ (Mas1h 5.23 || SirB 12v.7 || Sir. 43.6) ‫‘ הגביר‬prevail’ (SirB 9v.7 || Sir. 39.34) It is unclear why the same qal-hifʿil mismatch between MT and SP occurs in MT ‫א ֶכל ִת ְּׁש ְּׁב ֧רּו ֵ ַֽמ ִא ָ ֛תם ַב ֶכ ֶֶ֖סף‬ ֹ ‘food you will buy from them for ̊̄ money’ (Deut. 2.6) || SP ‫ אכל תשבירו מאתם‬akal tašbīru miyyētimma ̊̄ ‘food you will buy (?) from them for money’, unless it is due to afkasəf local ‘contamination’ from ‫א ֶכל ַב ֶכ ֶַ֤סף ַת ְּש ִב ֵ ֙רנִ ֙י‬ ֹ ‘food for money sell to me’ (Deut. 2.28), which has a hifʿil in both the MT and SP, or the hifʿil has a nuance of ‘actively trade’. 13 He also lists ‫‘ הזיף‬reprove’ (SirA 4r.25 || SirB 1v.12 || Sir. 11.7), which seems to reflect hifilisation relative to RH and Aramaic G-stem ‫נזף‬. 11. Hifilisation • 229 ‫‘ הזיד‬act arrogantly’ (SirA 1r.8 || Sir. 3.16; see above, §§1.1.3; 1.2.2) • • ‫‘ החריף‬reprove, stir up the wind’ (Mas1h 6.10 || Sir. 43.16) ‫‘ הלעיג‬mock’ (SirB 4v.4 || Sir. 31.22; see above, §§1.1.1; 1.2.2) • ‫‘ העריך‬arrange’ (SirB 9r.3 || Sir. 39.17; SirB 19v.12 || Sir. 50.18; cf. above, §1.3.7) • ‫‘ הצפין‬hide’ (SirA 1v.12 || Sir. 4.23; SirC 2a.3 || Sir. 20.31; SirC 2a.4 || Sir. 20.31; SirB 11r.7 || Sir. 41.15; SirB 11r.7 || Sir. 41.15) • • • ‫‘ הקנה‬buy’ (SirB 7v.2 || Sir. 37.11) ‫‘ הקשיח‬become hard’ (SirB 3r.4 || Sir. 30.12) ‫‘ השעה‬look’ (SirB 13v.11 || Sir. 44.8) Several of the above are variants with non-hifʿil counterparts. In a few cases, the semantics of the hifʿil may be argued to differ from those of the qal,14 but the general trend is clear. Beyond these, BS’s Hebrew sides with Second Temple Hebrew on additional hifilisation trends, e.g., • consistent hifʿil treatment of ‫—בי"ן‬all clearcut forms (Mas1h 5.11 || Sir. 42.21; SirA 1v.2 || Sir. 4.11; SirA 3v.18 || Sir. 10.1; SirA 4v.5 || Sir. 11.15; SirB 7r.1 || Sir. 36.24; SirB 7v.7 || Sir. 37.13; SirB 8r.10 || Sir. 38.4; SirB 12r.15 || Sir. 42.21); 14 In context, ‫ העריך‬can be understood in its classical meaning of ‘esti- mate’, whereas ‫ הרעיד‬is open to a causative interpretation. 230 • The Historical Depth of the Tiberian Reading Tradition exclusive use of hifʿil ‫ הוליד‬rather than qal ‫ ילד‬in the sense of ‘father, sire’ (Mas1h 3.10 || Sir. 41.9; SirA 4v.26 || Sir. 11.33; SirB 10v.18 || Sir. 41.9; SirB 10v.18 || Sir. 41.9; • dominance of hifʿil ‫ הוסיף‬to the exclusion of qal ‫( יסף‬SirA 1r.16 || Sir. 3.27; SirA 1v.25 || Sir. 5.5; SirB 8r.5 || Sir. 37.31; SirB 13r.12 || Sir. 43.27 [?]; SirC 2r.7 || Sir. 5.5; SirC 1b.10 || Sir. 3.27; SirD 1v.20 || Sir. 37.31); • comparatively frequent incidence of intransitive ‫הצליח‬ (Mas1h 2.25 || Sir. 41.1; SirA 3v.11 || Sir. 9.12; SirB 8v.1 || Sir. 38.13; SirB 9r.4 || Sir. 39.18; SirB 10v.8 || Sir. 41.1; though possible cases of the qal are also attested: SirA 3r.18 || Sir. 8.10; SirA 4v.7 || Sir. 11.17; SirB 8v.2 || Sir. 38.14; SirB 13r.11 || Sir. 43.26). 1.5. Rabbinic Hebrew Moreshet (1996) divides his lists of RH hifʿil innovations into several categories. Given below are those with greatest relevance to BH. 1.5.1. RH Hifʿil || MT Transitive Qal ‫‘ טמ"ן‬hide’ The BH hifʿil ‘hide (tr.)’ is rare (2 Kgs 7.8, 8), but becomes common in RH, though the qal is still frequent, especially as a participle. ‫‘ מש"ך‬draw, extend’ In BH the qal is normally transitive, with nifʿal serving for intransitive (Isa. 13.22; Ezek. 12.25, 28), though the qal can also be 11. Hifilisation 231 intransitive (Judg. 20.37; Job 21.33; Neh. 9.30 [?]); the same is generally true in RH, but a transitive hifʿil has also appeared. ‫‘ נש"א‬raise (a signal flare)’ In Tannaitic Hebrew, the qal is common and the hifʿil is normally causative (‘marry off, allow to marry’), but one also finds it used for the raising of a signal flare (m. Rosh haShana 2.2, 3; t. Rosh haShana 1.17), for which cf. the qal forms in Jer. 6.1 (BH has hifʿil forms in Lev. 22.16; 2 Sam. 17.13). ‫‘ שמ"ט‬unfasten, remove, cancel (debt, oath)’ Qal in BH (on the apparent hifʿil in Deut. 15.3, see §§1.3.5); in RH the qal continues in literal senses (‘unfasten, remove’; cf. its nifʿal passive/intransitive), while the hifʿil is reserved for cancellation of debts (m. Sheviʿit 10.1–3) and oaths (m. Shevuʿot. 7.8) and for letting fields lie fallow (Sifra, BaHar, parasha 2, ch. 3 [p. 107, col. 3]). 1.5.2. RH Hifʿil || MT Intransitive Qal ‫‘ חכ"ם‬be/become wise’ In BH the qal is stative ‘be wise’ (e.g., Deut. 32.29; Prov. 23.15) and inchoative ‘become wise’ (e.g., Prov. 6.6; 9.9; 19.20), the only hifʿil being causative (Ps. 19.8); in RH, the hifʿil can be inchoative (m. Bava Batra 10.8; m. ʾAvot 2.5). 232 The Historical Depth of the Tiberian Reading Tradition ‫‘ ספ"ד‬mourn’ The BH qal ‘mourn’ never takes a direct object (2 Sam. 3.31; internal object in Gen. 50.10; it takes -‫ל‬, e.g., Gen. 23.2, or ‫על‬, e.g., 2 Sam. 11.26), though nifʿal is clearly passive (Jer. 16.4; 25.33); RH also has an intransitive qal (m. Yevamot 16.5) and passive nifʿal (m. Shabbat 23.4), but adds a hifʿil either transitive (m. Moʿed Qaṭan 1.5) or intransitive (m. Megilla 3.3). ‫‘ תמ"ה‬be surprised, astonished, wonder’ The predominantly BH qal intransitive ‘be surprised, astonished, wonder’ persists in RH, but is joined by a synonymous hifʿil (Mekhilta deRabbi Ishmael, Sifre Devarim, Mekhilta deRabbi Shimʿon ben Yoḥai). 1.5.3. RH Hifʿil || MT Transitive and Intransitive Qal ‫‘ טב"ל‬immerse’ In BH the qal is usually transitive ‘immerse’ (e.g., Gen. 37.31), with a nifʿal intransitive (Josh. 3.15), though an intransitive/ reflexive qal (2 Kgs 5.14) is also attested; RH knows qal transitives (e.g., m. Shabbat 5.1) and intransitives (e.g., m. Shabbat 6.1), as well as a hifʿil transitive (e.g., m. Shabbat 2.7). ‫‘ רח"ץ‬wash, rinse’ BH qal forms dominate, with both transitive (e.g., Gen. 18.4) and intransitive/reflexive (e.g., Exod. 2.5) meanings of ‘wash, rinse’ (there are also rare qal passive [Ezek. 16.4; Prov. 30.12] and hitpaʿʿel forms [Job 9.30; Dan. 3.28]); the RH qal is typically in- 11. Hifilisation 233 transitive/reflexive (e.g., m. Sheviʿit 8.11) or transitive/reflexive with body parts (e.g., m. Yoma 8.1), while the hifʿil functions in both of the latter senses (e.g., Sifra, Nedava, parasha 11, ch. 1 [p 10, col. 4]; ʾEmor, parasha 4, ch. 2 [p. 96, col. 4]) and more prototypically transitive senses (e.g., m. Shabbat 9.3). 1.5.4. RH Hifʿil || Rare BH Qal ‫‘ חמ"ץ‬ferment, be/become leavened’ BH form knows the intransitive qal ‘ferment, be(come) leavened’ (Exod. 12.34, 39; Hos. 7.4); in RH both the qal and hifʿil can have intransitive meaning (e.g., respectively, Mekhilta deRabbi Ishmaʿel, Paskha, parasha 14 [p. 49]; m. Terumot 3.1). ‫‘ טע"ן‬load’ BH has the transitive qal hapax meaning ‘load (a beast of burden’ (Gen. 45.17); in RH cf. the qal (e.g., m. Bava Qama 9.1) and the synonymous hifʿil (e.g., Sifre Devarim, 343 [p. 396]). ‫‘ כח"ש‬become thin’ The sole BH qal comes in the intransitive sense ‘become thin’ (Ps. 109.24); this sense occurs in RH in the hifʿil (e.g., t. Bava Qama 3.5, 5), as well as in qal (e.g., t. Bava Qama 7.17). ‫פׂש"ע‬/‫‘ פס"ע‬step, march’ The BH qal hapax means ‘step, march’ (Isa. 27.4); in RH the root is normally ‫פס"ע‬, with the qal continuing and the innovation of a synonymous hifʿil (e.g., y. Berakhot 1.1). 234 The Historical Depth of the Tiberian Reading Tradition ‫‘ רח"ש‬express’ Assuming that the BH usage in Ps. 45.2 means ‘express’, RH exhibits persistence of the qal (e.g., y. Berakhot 2.1) and innovation of a synonymous hifʿil e.g., (y. Berakhot 4.1). ‫‘ שח"ר‬darken’ A BH hapax qal meaning ‘darken (intr.)’ (Job. 30.30); cf. RH hifʿil (e.g., m. Negaʿim 1.5, 5) and hofʿal (m. Sukkot 4.9). 1.5.5. RH Hifʿil Innovations Moreshet also lists hifʿil RH root innovations: ‫‘ הגדיש‬heap, stack’, ‫‘ הגליד‬form a crust, scab’, ‫‘ החזיר‬return (intr.), repeat’, ‫‘ הכמין‬hide (tr.)’, ‫‘ המתין‬wait’, ‫‘ הסדיר‬arrange’.15 1.5.6. RH Hifilisation Features in Common with Other Second Temple Hebrew Types RH also exhibits the following Second Temple Hebrew hifilisation tendencies discussed above: • • strong preference for hifʿil ‫;בי"ן‬ occurrences of hifʿil ‫( שי"ם‬t. Giṭṭin 7.13; Sifre Devarim 315; y. Sanhedrin 1.1; frequently in the BT); 15 From this list, several roots cited by Moreshet have been omitted due either to absence of the hifʿil form from the authoritative RH manu- scripts cited on the Maʿagarim site of the Academy of the Hebrew Lan- guage, e.g., ‫‘ חז"ם‬prune’, ‫‘ חל"י‬become ill’, ‫‘ טר"ד‬disturb, drive away’, ‫פנ"י‬ in the passive sense ‘free, empty’, ‫‘ שט"ן‬accuse’, or to semantic remote- ness relative to the BH qal, e.g., ‫פס"ק‬/‫‘ פׂש"ק‬cease’. 11. Hifilisation • 235 strong preference for hifʿil ‫ הוליד‬over qal ‫ ילד‬with masculine subjects, in the sense of ‘father, sire’; • • dominance of hifʿil ‫ הוסיף‬to the near exclusion of qal ‫;יסף‬ dominance of hifʿil intransitive ‫הצליח‬. 2.0. The Tiberian Reading Tradition of Classical Biblical Hebrew Texts When it comes to hifilisation, like other traditions rooted in the biblical text, the Tiberian reading component generally adheres closely to the parallel orthographic component. This is not surprising, as (a) the two are related components of a composite tradition and (b) development of each component was to some degree influenced and constrained by its association with the other. Even so, apparent cases of dissonance occur, some centring on hifilisation. In the case of CBH material, the reading component of the composite Tiberian tradition reflects a linguistic stage more chronologically advanced than the written component. In LBH material, the two components exhibit greater correspondence. This is consistent with the view that a significant degree of the crystallisation of the Tiberian reading tradition took place during the Second Temple Period. 2.1. ‫‘ נח"י‬lead, guide’ The root ‫‘ נח"י‬lead, guide’ is represented in Tiberian BH by a paradigm that is largely suppletive. Consider Table 7. 236 The Historical Depth of the Tiberian Reading Tradition Table 7: Qal and hifʿil forms of ‫ נח"י‬according to the Tiberian reading tradition (see §5.4 for citations) qal hifʿil suffix conjugation 8 2 imperative 4 0 infinitive construct 0 2 prefix conjugation 0 17 When it comes to the suffix conjugation and the imperative, the dominant morphology is qal. Against this background, it is telling that there are no qal prefix conjugation forms in the 17 potential cases. This is even more suspicious when one considers the fact that one of the infinitive construct forms realised according to the reading tradition as a hifʿil has the orthography of a qal, namely, ‫‘ ַלנְּ ח ָֹתם‬to guide them’ (Exod. 13.21). Lacking the expected heh of a hifʿil infinitive, it seems likely that the consonants presuppose qal ‫*לנְּ ח ָֹתם‬, ִ in line with the aforementioned qal suffix conjugation and imperative forms. Interestingly, the only other infinitive construct with this root is the unambiguous hifʿil ‫ְּל ַהנְּ ח ָֹתם‬ ‘to guide them’ (Neh. 9.19) in an LBH allusion to this very verse. It is also to be noted that one of the two unequivocally hifʿil suffix conjugation forms (Neh. 9.12) comes in LBH (on the other, see below, §3.0). According to a plausible reading of the data, early stem diversity characterised verbs with the root ‫נח"י‬. This is to say, the process of hifilisation was underway well before the era of LBH. Yet it was by no means complete. If so, however, why according to the reading tradition are qal forms restricted to imperatives and qaṭal forms? Surely, given the apparent early incidence of qal imperatives and suffix conjugation forms, one might expect at least some incidence of qal infinitives and prefix conju- 11. Hifilisation 237 gation forms, rather than consistent hifʿil vocalisation. Here, again, the reading tradition appears to have extended an ancient feature in line with Second Temple preference for the C-stem. Where hifʿil could be read without undue deviation from the consonantal orthography, i.e., in yiqṭol forms, it was so read. The hifʿil analysis was extended even in opposition to the consonantal spelling of infinitival ‫‘ ַלנְּ ח ָֹתם‬to guide them’ (Exod. 13.21), because this was considered close enough phonetically to the expected ‫חֹותם‬ ָ ְּ‫*ל ַהנ‬. ְּ 2.2. ‫‘ טמ"ן‬hide, bury’ Next, consider Tiberian verbal representatives of the root ‫טמ"ן‬. Most evidence points to an active-middle stem arrangement involving qal ‫‘ ָט ַמן‬hide, bury (tr.)’ (21x) (with passive participle ‫ָטמּון‬ ‘hidden’ [7x]) and nifʿal ‫‘ *נִ ְּט ַמן‬hide (intr.), bury oneself’ (1x). In a single verse in the book of Kings, however, one encounters two cases of hifʿil ‫*ה ְּט ִמין‬ ִ ‘hide (tr.)’ (2 Kgs 7.8), with no apparent semantic difference from the qal. Since the orthography in both forms—‫—ויטמנו‬is ambiguous as far as stem identity goes, it may be that the hifʿil vocalisation here reflects ‘drift’ toward Second Temple morphology (as seen in SH, BS, and RH; see above, §§1.3–5). It must be noted, though, that other consonantally ambiguous forms, all wayyiqṭol (Gen. 35.4; Exod. 2.12; Josh. 2.6; Jer. 13.5), are read as qal. 2.3. Hofʿal of II-w/y Verbs as Evidence of Hifilisation While the Tiberian reading tradition is opaque with regard to the analysis of finite II-y yiqṭol verbal forms, i.e., whether they are 238 The Historical Depth of the Tiberian Reading Tradition qal or hifʿil, this is not the case with hofʿal forms. Based on regular sound changes (for which see Blau 2010, 97, §3.4.3.3), for the verb ‫‘ ָשם‬put’, the expected qal passive wayyiqṭol form is ‫ישם‬ ֶ ִ‫‘ וַ י‬and it was put’ (Gen. 50.25). This is precisely the orthography one finds in the ketiv ‫( ויישם‬Gen. 24.33), but the corresponding qere ‫יּושם‬ ַ֤ ַ ַ‫‘ ו‬and it was put’ is a hofʿal. This reflects two diachronic developments: the well-known decline of the qal internal passive (see ch. 10, §§2.2; 3.2) and, since hofʿal represents the internal passive of hifʿil, hifilisation. In other words, a realisation such as qere ‫יּושם‬ ַ֤ ַ ַ‫ ו‬implies the existence of hifʿil ‫ה ִשים‬, ֵ as seen occasionally in the Tiberian written tradition (Ezek. 14.8; 21.21; Job 4.20) and more commonly in late antique extra-biblical Hebrew (Ezekiel; Job, see above §1.1.3; BS, see above, §1.4; RH, see above, §1.5.6). 2.4. The Preservation of Archaic Hifʿil-like Qal Forms While the preceding paragraphs detail departures of the Tiberian reading tradition from the pronunciation tradition implied by the consonantal text in line with Second Temple linguistic developments, it is important, for the sake of balance, to highlight conservatism, even archaism, in the reading tradition. One relevant phenomenon involves qal verbs with prefix conjugation forms in the yaqṭel pattern (Yalon 1971). Consider, for example, forms representative of the root ‫גנ"ן‬: the suffix conjugation form ‫נֹותי‬ ִ ַ‫וְּ ג‬ ‘and I will defend’ (2 Kgs 19.34 || Isa. 37.5; 2 Kgs 20.6 || Isa. 38.6) and the infinitive absolute ‫‘ גָ נֵ֥ ֹון‬protecting’ (Isa. 31.5b) are unambiguously qal, whereas the prefix conjugation ‫‘ יָ גֵ ן‬will protect (3MS)’ (Isa. 31.5a; Zech. 9.15; 12.8) is alternatively qal yaqṭel or hifʿil. Since there are no unambiguous hifʿil forms in BH, and 11. Hifilisation 239 since the qal infinitive absolute occurs alongside the equivocal prefix conjugation in the same verse (Isa. 31.5), the verb is plausibly analysed as uniformly qal in BH (Blau 2010, 222–23, §4.3.5.2.3.2). This contrasts with orthographically unequivocal RH hifʿil forms, such as ‫( הגן‬e.g., ʿAravit, fourth blessing, ln. 4), ‫( מגין‬e.g., Mekhilta deRabbi Ishmaʿel, BeḤodesh [Yitro], parasha 1 [p. 204]), ‫( להגן‬e.g., y. Pesaḥim 7.12 [p. 35b]).16 In a similar way Yalon (1971, 46–47) explains such forms as ‫ וַ יַ ְּד ְּרכּו‬in ‫ת־לשֹונָ ֙ם ַק ְּש ָתם ֵֶ֔ש ֶקר‬ ְּ ‫‘ ַוֵַּֽ֣יַ ְּד ְּר ַ֤כּו ֶא‬they bend their tongue like their bow for deceit’ (Jer. 9.2; otherwise ‫ ָד ַרְך ֶק ֶשת‬consistently qal); ‫ וַ יַ ְּד ְּבקּו‬in ‫יהם ַב ִמ ְּל ָח ָ ַֽמה‬ ֶ֖ ֶ ‫ם־ה ָמה ַא ֲח ֵר‬ ֛ ֵ ַ‫‘ ַוֵַּֽ֣יַ ְּד ְּב ֵ֥קּו ג‬they too pursued them in the battle’ (1 Sam. 14.22) and ‫וַ יַ ְּד ְּבקּו ְּפ ִל ְּש ִֵ֔תים ַא ֲח ֵ ֵ֥רי ָש ֶ֖אּול‬ ‫‘ וְּ ַא ֲח ֵרי ָב ָנָׂ֑יו‬and the Philistines pursued Saul and his sons’ (1 Chron. 10.2), and even ‫ וַ יַ ְּד ִביקּו‬in ‫י־דן‬ ַֽ ָ ֵ‫ת־בנ‬ ְּ ‫‘ וַ יַ ְּד ִ ֶ֖ביקּו ֶא‬they overtook the peo- ple of Dan’ (Judg. 18.22)—the latter on the assumption that the ī so reminiscent of hifʿil results from a lengthening of the original short i vowel of the qal yaqṭel pattern.17 It is from qal forms with yaqṭel prefix conjugation forms, opines Yalon, that many unambiguous hifʿil forms developed. Basing himself partially on the likes of Barth (1889; 1891, 117, 147, 119–20, 136, 285–86, 305), Böttcher (1866–1868, II:436), and Brockelmann (1908–1913, 16 Perhaps also in 4Q403 f1i.25; 4Q405 f3ii.17 (see the Maʾagarim web- site of the Academy of the Hebrew Language), but these are also interpreted as instances of the noun ‫‘ ָמגֵ ן‬shield’ (Abegg’s 1999–2009 QUMRAN module for Accordance). 17 Cf. the causative hifʿil in ‫י־איש ֵכן ִה ְּׁד ַּב ְּּׁ֣ק ִתי‬ ִֵ֗ ֵ‫ל־מ ְּתנ‬ ָ ‫כ ֲא ֶשר֩ יִ ְּׁד ֶַּ֨בק ָה ֵאזֶ֜ ֹור ֶא‬... ַ ...‫ל־בית יִ ְּש ָר ֵֶ֜אל‬ ֵָֹ֨ ‫ת־כ‬ ָ ‫…‘ ּ֠ ֵא ַלי ֶא‬as a loincloth clings to a man’s waist, so I have made the whole house of Israel… cling to me’ (Jer. 13.11; cf. Deut. 28.21; Ezek. 3.26; 29.4). 240 The Historical Depth of the Tiberian Reading Tradition I:548),18 Yalon argues for the preservation of qal yaqṭel and/or related infinitival or imperatival forms representing such roots as, ‫אב"ד‬, ‫אב"ל‬, ‫אח"ז‬, ‫אט"ם‬, ‫אכ"ל‬, ‫אמ"ר‬, ‫אס"ף‬, ‫אצ"ל‬, ‫אמ"ץ‬, ‫בי"ן‬, ‫גל"ל‬, ‫גנ"ן‬, ‫הל"ל‬, ‫הפ"ך‬, ‫זי"ד‬, ‫זל"ל‬, ‫זק"ן‬, ‫חל"ל‬, ‫חל"ק‬, ‫טמ"ן‬, ‫יס"ף‬, ‫יצ"ב‬, ‫יש"ר‬, ‫ית"ר‬, ‫כת"ת‬, ‫לע"ג‬, ‫מר"י‬, ‫מת"ק‬, ‫נג"ש‬, ‫נט"י‬, ‫נח"י‬, ‫נס"ך‬, ‫נפ"ל‬, ‫נש"י‬, ‫נש"ך‬, ‫סת"ר‬, ‫עז"ר‬, ‫עמ"ד‬, ‫עש"ר‬, ‫עת"ק‬, ‫צב"י‬, ‫צל"ל‬, ‫צפ"ן‬, ‫קה"ל‬, ‫קר"ב‬, ‫רו"ח‬, ‫רצ"ץ‬, ‫שי"ם‬, ‫שמ"ד‬, ‫שמ"ט‬, ‫שמ"ע‬, ‫שק"י‬. Many of these have apparently suppletive qal-hifʿil paradigms, on the basis of which it may be postulated that unequivocal hifʿil forms secondarily arose. An illustrative case showcasing the combination of conservation and development that characterises the Tiberian reading tradition centres on qal and hifʿil forms of ‫( יס"ף‬Huehnergard 2005). Nearly full qal and hifʿil paradigms can be adduced, with no obvious semantic distinction between the two stems. Table 8: The paradigms qal ‫ יָ ַסף‬and hifʿil ‫הֹוסיף‬ ִ Suffix conjugation qal ‫יָ ַסף‬ hifʿil ‫הֹוסיף‬ ִ Active participle ‫י ֵֹסף‬ ‫מֹוסיף‬ ִ )‫י ֵֹסף (וַ י ֶֹסף‬ )‫יֹוסיף (וַ י ֶֹסף‬ ִ ‫*סף‬ ֵ ‫*הֹוסף‬ ֵ ‫ל ְּספֹות‬/‫פֹות‬ ִ ‫)ס ֶפת > ְּס‬ ֶ ‫*(ל‬ ָ ‫)הֹוסיף‬ ִ ‫(ל‬ ְּ Prefix conjugation Imperative Infinitive construct The assumption of synonymous qal and hifʿil paradigms resolves certain grammatical problems, such as what must otherwise be explained as the rather frequent use of jussive forms where indic18 Yalon (1971, 43) also adduces opinions among Jewish interpreters, such as Ibn Janaḥ, Rashi, and Samuel David Luzzatto. Cf. Bergsträsser (1918–1929, II: 80, 82, 127), who for many of the forms suggested by Barth rejects a qal yaqṭel explanation, adopting instead the view that the vocalisation is simply wrong. 11. Hifilisation 241 ative alternatives are expected (e.g., Gen. 4.2; Lev. 5.16, 24; 37.31; Num. 5.7; 22.19; Deut. 13.1; 18.16) and the apparent use of the 3rd-person jussive where the participle is expected (Isa. 29.14; 38.5). It entails the assumption that the qal I-y infinitive construct ‫ לספת‬in the Meshaʿ Stele (KAI 181.21) was realised as if it were a III-y form in the combined Tiberian written-reading tradition. Such a situation of parallel paradigms presumably evolved from an original qal, whose yaqṭel < PS yaqtil prefix conjugation spurred the secondary formation of unambiguous hifʿil forms. The diachronic character of the process is manifest in the distribution of unequivocal consonantal qal and hifʿil forms as well as forms with matres or vocalisations that unambiguously identify the binyan. Table 9: Distribution of qal and hifʿil forms of ‫ יס"ף‬according to the various layers of the Tiberian biblical tradition unequivocal consonantal Pentateuch Prophets (Former (Latter Writings (non-LBH (LBH+ TOTALS prefix conjugation vocalisation qal hifʿil qal 13 15 9 6 1 0 1 29 1 3 3 0 7 1 6 11 8 3 0 3 0 0 0 11 hifʿil defective plene 11 4 11 36 6 18 5 18 3 30 3 22 0 8 25 70 ambiguous jussive/wayyiqṭol 4 1 1) 0) 6 5) 1) 11 When it comes to the distribution of forms of qal ‫ יָ ַסף‬and hifʿil ‫הֹוסיף‬, ִ the various Masoretic corpora exhibit conspicuous differences that appear to have diachronic significance. Thus, in MT LBH+, there is virtually no dissonance between the three types of evidence: hifʿil morphology predominates to the near exclusion 242 The Historical Depth of the Tiberian Reading Tradition of qal in unequivocal consonantal forms; vocalisation of yiqṭol is exclusively hifʿil; and hifʿil prefix conjugation vocalisation is consistently matched by exclusively plene hifʿil orthography.19 The morphological harmony among consonantal text, vocalisation, and matres lectionis in Persian Period material tallies with other evidence confirming a special affinity between the Tiberian vocalisation and the period in which LBH+ texts were composed. The rest of the MT is characterised by more or less conflicting totals. Consider the Pentateuch: unequivocal consonantal forms are nearly all qal—with the problematic ‫הֹוסיף‬ ֵ֥ ִ ‫( ְּל‬Lev. 19.25) the single arguable exception20—but yiqṭol vocalisation is di- vided—eight qal and fifteen hifʿil. Intriguingly, however, only four of the fifteen yiqṭol forms with indisputable hifʿil vocalisation have equally unambiguous plene hifʿil spelling. This situation obviously contrasts with the one described above for LBH+ texts. Whereas there is consonantal, vocalic, and orthographic har19 The relevant distribution in the non-LBH+ Writings seems similar, but the dearth of unequivocal consonantal forms precludes certainty. 20 In the passage’s context of harvesting, ‘gather’ is at least as apposite as ‘add’. Vulgate congregantes reflects the former; LXX πρόσθεμα, TO ‫אֹוס ָפא‬ ָ ‫ל‬, ְּ and the Syr ‫ ܘܢܘܣܦܘܢ‬the latter. The Samaritan evidence is varied. The ST has ‫‘ למכנשה‬gather’ against the SAP’s ‫‘ ليضاعف‬multiply’. For the meaning ‘gather’ one expects qal ‫ לאסף‬in Samaritan as well as Tibe- rian Hebrew; indeed, the hifʿil is otherwise unknown. Also, the Samari- tan pronunciation līsǝf reflects neither ‫ לאסף‬nor ‫להאסיף‬, but seemingly ‫‘ להסיף‬bring to an end’. Cf. MT ‫ֹאס ָ֞פּון‬ ִ ‫ || ת‬SP ‫ תוסיפון‬tūsīfon (Exod. 5.7), where, again, the context is amenable to both ‘continue’ and ‘gather’. Similar cases of possible conflation occur within the Tiberian tradition: ‫ אס"ף‬and ‫) סו"ף‬Jer. 8.13; Zeph. 1.2), ‫ אס"ף‬and ‫( יס"ף‬1 Sam. 18.29; 2 Sam. 6.1); see Ben-Ḥayyim (2000, 143, 213). 11. Hifilisation 243 mony in LBH+, striking dissonance obtains in the Pentateuch. Unambiguous qal consonantal forms and the rare incidence of plene orthography with mater yod to signal hifʿil morphology con- trast with rather common hifʿil vocalisation. The complexity of the combined Tiberian written-reading tradition in the Pentateuch is further manifested in the preservation of archaic qal yaqṭel prefix conjugation morphology, according to which forms like non-jussive ‫ י ֵֹסף‬are to be analysed as cases of the indicative qal yaqṭel prefix conjugation, not as short jussive hifʿil forms. Apparently occupying a sort of intermediate position between the Pentateuch and LBH+, the books of the Prophets exhibit significant discord between preservation of qal in the case of unequivocal consonantal forms and development of hifʿil yiqṭol, but noticeably greater affinity than in the Pentateuch between hifʿil vocalisation and plene orthography in the prefix conjugation. A further point of contrast with the Pentateuch is the infrequency in the Prophets of archaic qal yaqṭel vocalisations. Focusing on the relationship between the vocalisation and the orthographic tradition regarding hifilisation of qal ‫יָ ַסף‬, the statistics constitute arguable evidence of linguistically significant development in orthographic practice within the MT. Concentrating on yiqṭol forms where a long i vowel might be expected, we find that explicit hifʿil spellings constitute a minority in the Pentateuch, come in three-quarters of the cases in the Prophets, and are the norm in the Writings, including LBH+, where hifʿil orthography is employed to the total exclusion of potential qal spellings. Crucially, the plene percentages reflect various degrees 244 The Historical Depth of the Tiberian Reading Tradition of agreement between the orthographic and vocalisation components of the combined Tiberian tradition. Whenever the various constituent texts were composed, the written form of the Masoretic Pentateuch seems to reflect a stage in orthographic development in which the spelling of (way)yiqṭol was largely still amenable to qal morphology. Beyond the Pentateuch, there is a strong and increasing tendency to utilise (way)yiqṭol spellings exclusive to hifʿil. It is reasonable to assume that such spellings in LBH accurately reflect the post-exilic hifʿil usage common to Second Temple Hebrew material noted above. How to account for the high degree of hifʿil yiqṭol forms in CBH outside the Pentateuch is a more complicated question. It may be, of course, that the relatively high incidence of hifʿil spellings in non-Pentateuchal CBH is due partially to the anachronistic application of late linguistic conventions to this material—an enterprise from which the Pentateuch was (partially) exempted due to its relatively early compilation and/or special venerated status. A reasonable hypothesis for historical development might run as follows. An early situation of dominant qal morphology gradually gave way to one of increased hifʿil usage due in part to hifʿil-like qal yaqṭel forms. This second stage was characterised by the continued use of both consonantally unambiguous and ambiguous qal forms as well as by an increase in consonantally and orthographically unambiguous hifʿil forms. Depending on the realisation and spelling of ambiguous forms, various manifestations of suppletion might have obtained. 11. Hifilisation 245 Intriguingly, the sorts of suppletion encountered in the Masoretic corpora described above show a certain diachronic progression. The clearest situations are in LBH+ and the Pentateuch: whereas LBH+ texts show virtually no suppletion—hifʿil dominant according to all components of the tradition—much of the suppletion in the Pentateuch seems to be secondary—qal dominant both consonantally and orthographically, hifʿil restricted chiefly—though not exclusively—to vocalisation. The nature of the suppletion in the Prophets is more difficult to interpret. It may be largely organic—there being a mix of unambiguous qal consonantal forms together with hifʿil forms on which vocalisation and spelling with mater yod agree. Alternatively, of course, the greater use of mater yod for unequivocal hifʿil spelling in the Prophets vis-à-vis the Pentateuch may be due to a secondary spelling revision that impacted non-Torah CBH material more than the Torah. Limited support for such a theory emerges from the fact that, in comparison to the Pentateuch, the Prophets show increased incidence of plene spelling with both yod and waw in the relevant (way)yiqṭol forms of ‫ יָ ַסף‬and ‫הֹוסיף‬. ִ What is clear is that, whatever its origin, there is more in the way of qal-hifʿil suppletion to deal with in the Prophets than in either the Pentateuch or LBH+. 3.0. The Tiberian Classical Biblical Hebrew Written Tradition The foregoing sections have focused mainly on the secondary and late character of hifilisation in various ancient Hebrew corpora and traditions. Such a characterisation is correct, but also poten- 246 The Historical Depth of the Tiberian Reading Tradition tially misleading, as it is not the whole story. It must be emphasised that no historical phase of Hebrew—biblical or extrabiblical—is devoid of consonantally unambiguous hifʿil forms. Second, while many of the instances of hifilisation discussed above represent innovations restricted to Second Temple times, in several cases hifʿil harbingers—sometimes, but not always, minority forms—predate the post-exilic period. This is true of hifʿil forms of such roots as ‫בי"ן‬, ‫יל"ד‬, ‫יס"ף‬, ‫לי"ץ‬, ‫נח"י‬, ‫צל"ח‬, and ‫רי"ב‬, all of which, to varying degrees, show hifʿil distribution earlier than LBH (see §5.1 for citations). Indeed, in some cases, like that of ‫הֹוליד‬ ִ ‘father, sire’, hifʿil usage is dominant throughout all historical stages of ancient Hebrew according to the consonantal tradition. In the case of ‫ לי"ץ‬and ‫רי"ב‬, whose hifʿil verbal forms are limited to demonstrably late material, it may be that hifilisation began in participial forms with nominal or adjectival semantics, since these are the only relevant hifʿil forms that crop up in preLBH material (for a similar phenomenon in the process of nifalisation, see above, ch. 10, §3.0). The case of qal ‫ יָ ַסף‬versus hifʿil ‫הֹוסיף‬ ִ exemplifies several important points. First, though the vocalisation in the Pentateuch and the Prophets is probably somewhat anachronistic—involving the hifʿil reinterpretation of a number of apparently original qal forms in line with Second Temple tendencies unambiguously evidenced in late consonantal evidence—in no part of the Hebrew Bible, including those parts considered the most ancient, is the vocalisation tradition the lone witness to hifilisation of ‫יס"ף‬. Second, in its use of unambiguous plene hifʿil spellings for ‫יס"ף‬, specifically, and for hifʿil forms, more generally, the ortho- 11. Hifilisation 247 graphic tradition itself evinces several chronological windows on the hifilisation process—considerably less advanced in the Torah, nearly complete in LBH, and at an intermediate stage in the Prophets. Seen from a different perspective, since orthographic evidence for the hifilisation of ‫ יס"ף‬comes substantially earlier than the advent of the Tiberian vocalisation signs, it is clear that the hifilisation shift reflected in the medieval Tiberian reading tradition significantly predates medieval times, extending back to the Second and First Temple Periods. 4.0. Conclusions With regard to the process of hifilisation, the historical depth of the Tiberian vocalisation tradition finds confirmation in unequivocal hifʿil evidence found in MT LBH+, the biblical and non-biblical DSS, the SP, BS, RH, and, to some extent, the Tiberian consonantal tradition of different sections of the Hebrew Bible. The combined evidence shows clearly that the qal > hifʿil shift reflected in the vocalisation of the Tiberian reading tradition had already by Second Temple times profoundly impacted morphology, so that apparent cases of dissonance between the written component of the Tiberian biblical tradition and its reading counterpart should be considered differences in degree rather than kind. Clearly, hifilisation began early on in ancient Hebrew, and scholars are afforded a series of snapshots in the process by the orthographic tradition of various parts of the Hebrew Bible, by the Tiberian reading tradition, and by other Second Temple biblical traditions and extra-biblical material. 248 The Historical Depth of the Tiberian Reading Tradition 5.0. Citations 5.1. The Tiberian Biblical Tradition Table 1 ‫צל"ח‬: qal+‫—רּוח‬Judg. ַ 14.6, 19; 15.14; 1 Sam. 10.6, 10; 11.6; 16.13; 18.10; qal—Num. 14.41; 2 Sam. 19.18; Isa. 53.10; 54.17; Jer. 12.1; 13.7, 10; 22.30, 30; Ezek. 15.4; 16.13; 17.9, 10, 15; Amos 5.6; Ps 45.5; Dan 11.27; transitive hifʿil—Gen. 24.21, 40, 42, 56; 39.3, 23; Deut. 28.29; Josh. 1.8; Judg. 18.5; Isa. 48.15; 55.11; Ps. 37.7; 118.25; Neh. 1.11; 2.20; 2 Chron. 26.5; intransitive hifʿil—Gen. 39.2; 1 Kgs 22.12 (|| 2 Chron. 18.11), 15 (|| 2 Chron. 18.14); Jer. 2.37; 5.28; 32.5; Ps. 1.3; Prov. 28.13; Dan. 8.12, 24, 25; 11.36; 1 Chron. 22.11, 13; 29.23; 2 Chron. 7.11; 13.12; 14.6; 18.11 (|| 1 Kgs 22.12), 14 (|| 1 Kgs 22.15); 20.20; 24.20; 31.21; 32.30. Table 2 ‫בי"ן‬: qal—Deut. 32.7; Jer. 49.7; Ps. 5.2; 50.22; 94.8; 139.2; Prov. 23.1; Dan. 9.2, 23; 10.1; ambiguous—Deut. 32.9; 1 Sam. 3.8; 2 Sam. 12.19; Isa. 6.9, 10; 28.9; 32.4; 40.14; 43.10; 44.18; Jer. 9.11; Hos. 4.14; 14.10; Ps. 19.13; 28.5; 49.21; 58.10; 73.17; 82.5; 92.7; 94.7; Job 6.30; 9.11; 13.1; 14.21; 15.9; 18.2; 23.5, 8; 32.8, 9; 36.29; 38.20; 42.3; Prov. 2.5, 9; 7.7; 14.15; 19.25; 20.24; 21.29 qere; 23.1; 24.12; 28.5, 5; 29.7, 19; Dan. 9.22; 11.30, 37, 37; 12.8, 10, 10; Ezra 8.15; Neh. 8.8; 13.7; hifʿil—1 Kgs 3.9, 11; Isa. 28.19; 29.16; 40.21; 56.11; 57.1; Mic. 4.12; Ps. 32.9; 33.15; 119.27, 34, 73, 125, 130, 144, 169; Job 6.24; 28.23; Prov. 1.2, 6; 8.9; 14.8; 17.10, 24; 28.2, 7, 11; Dan. 1.4, 17; 8.5, 16, 17, 23, 27; 9.23; 10.11, 12, 14; Ezra 8.16; Neh. 8.2, 3, 7, 9, 12; 10.29; 1 Chron. 15.22; 25.7, 8; 27.32; 28.9; 2 Chron. 11.23; 26.5; 34.12; 35.3 qere. Table 3 ‫נח"י‬: qal—Gen. 24.27; Exod. 13.17, 21 (‫;)לנְּ ח ָֹתם‬ ַ 15.13; 32.34; Isa. 7.2; 58.11; Ps. 5.9; 27.11; 60.11; 77.21; 108.11; 139.24; ambiguous—Num. 23.7; Deut. 32.12; 1 Sam. 22.4; 1 Kgs 10.26; 2 Kgs 18.11; Isa. 57.18; Ps. 23.3; 31.4; 43.3; 61.3; 67.5; 73.24; 78.14, 53, 72; 107.30; 139.10; 143.10; Job 12.23; 31.18; 38.32; Prov. 6.22; 11.3; 18.16; hifʿil—Gen. 24.48; Neh. 9.12, 19. 11. Hifilisation 249 Table 4 ‫יס"ף‬: qal—Gen. 8.12; 38.26; Lev. 22.14; 26.18, 21; 27.13, 15, 19, 27; Num. 11.25; 32.14, 15; Deut. 5.22, 25; 19.9; 20.8; Judg. 8.28; 13.21; 1 Sam. 7.13; 12.19; 15.35; 27.4; 2 Sam. 2.28; 2 Kgs 6.23; 19.30; Isa. 26.15; 29.1, 19; 30.1; 37.31; Jer. 7.21; 45.3; 2 Chron. 9.6; ambiguous—Gen. 4.2, 12; 8.10, 21, 21; 18.29; 25.1; 30.24; 37.5, 8; 38.5; 44.23; Exod. 5.7; 8.25; 9.28, 34; 10.28, 29; 11.6; Lev. 5.16, 24; 27.31; Num. 5.7; 22.15, 19, 25, 26; Deut. 1.11; 3.26; 4.2; 13.1, 12; 17.16; 18.16; 19.20; Judg. 3.12; 4.1; 9.37; 10.6; 11.14; 13.1; 20.22, 28; 1 Sam 3.6, 8, 21; 9.8; 14.44; 18.29; 19.8, 21; 20.17; 23.4; 2 Sam. 2.22; 3.34; 5.22; 12.8; 18.22; 24.1, 3; 1 Kgs 16.33; 19.2; 20.10; 2 Kgs 6.31; Isa. 7.10; 8.5; 29.14; 38.5; Ezek. 5.16; 23.14; 36.12; Hos. 9.15; 13.2; Joel 2.2; Zeph. 3.11; Ps. 115.14; Job 27.1; 29.1; 36.1; 40.32; 42.10; Prov. 1.5; 9.9; 10.22; 19.19; 23.28; 30.6; Est. 8.3; Dan. 10.18; 1 Chron. 21.3; 2 Chron. 28.22; hifʿil—Exod. 14.13; Lev. 19.25; Deut. 25.3, 3; 28.68; Josh. 7.12; 23.13; Judg. 2.21; 10.13; 20.23; 1 Sam. 3.17; 20.13; 25.22; 2 Sam. 3.9, 35; 7.10, 20; 14.10; 19.14; 1 Kgs 2.23; 10.7; 12.11, 14; 2 Kgs 20.6; 21.8; 24.7; Isa. 1.5, 13; 10.20; 11.11; 23.12; 24.20; 47.1, 5; 51.22; 52.1; Jer. 31.12; Hos. 1.6; Amos 5.2; 7.8, 13; 8.2; Jon. 2.5; Nah. 2.1; Ps 10.18; 41.9; 61.7; 71.14; 77.8; 78.17; 120.3; Job 17.9; 20.9; 34.32, 37; 38.11; 40.5; Prov. 3.2; 9.11; 10.27; 16.21, 23; 19.4; 23.35; Ruth 1.17; Qoh. 1.16, 18; 2.9; 3.14; Lam. 4.15, 16, 22; Dan. 10.18; Ezra 10.10; Neh. 13.18; 1 Chron. 14.13; 17.9, 18; 22.14; 2 Chron. 10.11, 14; 28.13; 33.8. Table 5 masculine ‫יל"ד‬: qal—Gen. 4.18, 18, 18; 10.8, 13, 15, 24, 24, 26; 20.17; 22.23; 25.3; Deut. 32.18; Isa 49.21; 65.23; Jer. 17.11; Hos. 9.16; Zech. 13.3, 3; Ps. 7.15; Job 38.29; Prov. 23.22; 27.1; 1 Chron. 1.10, 11, 13, 18, 20; 2.48; hifʿil— Gen. 5.3, 4, 6, 7, 9, 10, 12, 13, 15, 16, 18, 19, 21, 22, 25, 26, 28, 30, 32; 6.10; 11.10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27; 17.20; 25.19; 48.6; Num. 26.29, 58; Deut. 4.25; 28.41; Judg. 11.1; 2 Kgs 20.18; Isa. 39.7; 45.10; 55.10; 66.9; Jer. 16.3; 29.6; Ezek. 18.10, 14; Job 38.28; Ruth 4.18, 19, 19, 20, 20, 21, 21, 22, 22; Qoh. 5.13; 6.3; Neh. 12.10, 10, 11, 11; 1 Chron. 1.34; 2.10, 10, 11, 11, 12, 12, 13, 18, 20, 20, 22, 36, 36, 37, 37, 38, 38, 39, 39, 40, 40, 41, 41, 44, 44, 46; 4.2, 2, 8, 11, 12, 14, 14; 5.30, 30, 31, 31, 32, 32, 33, 33, 34, 34, 35, 35, 36, 37, 37, 38, 38, 39, 39, 40, 40; 7.32; 8.1, 7, 8 9, 11, 32, 33, 33, 33, 34, 36, 36, 36, 37; 9.38, 39, 39, 39, 40, 42, 42, 42, 43; 14.3; 2 Chron. 11.21; 13.21; 24.3. 5.2. NBDSS ‫זנ"ח‬: ambiguous—1QHa 8.36; 4Q381 f46a+b.6 (‫ ;)]?[ תזנזח‬hifʿil—1QHa 17.7, 11; 4Q460 f9i.7. ‫לע"ג‬: hifʿil—1QpHab 4.2. ‫בז"י‬: qal—CD 7.18; 1QpHab 4.2; 250 The Historical Depth of the Tiberian Reading Tradition 1QHa 12.23; 13.22; 15.26; 4Q365 f6aii+6c.1; 4Q396 f1–2iii.10; 4Q397 f6_13.9; 4Q434 f1i.2; 4Q437 f1.2; 4Q508 f21.2; ambiguous—1QpHab 4.5; 4Q285 f3.4; hifʿil—CD 9.4. ‫רע"ד‬: ambiguous—1QHa 11.36. ‫שח"ק‬: qal—4Q266 f10ii.12; 4Q269 f11ii+15.1; ambiguous—1QS 7.14; 1QpHab 4.4, 6; 4Q171 f1–2ii.12; 4Q259 1.13; 4Q380 f3.2; 4Q434 f7b.3. ‫צל"ח‬: qal—4Q416 f8.1; ambiguous— 1Q27 f1ii.5; 4Q219 2.29; 4Q221 f1.7; 4Q299 f2.1; hifʿil—CD 13.21; 11Q19 58.21. ‫רש"ע‬: qal—CD 20.29; ambiguous—1QS 4.24; 1QHa 5.33; hifʿil—CD 20.26; 1QS 1.25; 1QM 1.2; 1Q34bis f3ii.4; 4Q174 f1–3ii.3 (|| Dan. 12.10); 4Q184 f1.3; 4Q266 f3ii.6; 4Q267 f2.2; f3.3; 4Q387 f3.6. ‫גד"ל‬: qal—4Q216 6.9 (= Jub. 2.10); ambiguous—4Q364 f18.2 (|| Num. 14.17). ‫בי"ן‬: qal—CD 1.1; 4Q268 f1.9; 4Q413 f1–2.4; ambiguous—CD 1.8, 10; 13.8; 1QS 11.22; 1QHa 8.13; 9.39; 20.30, 36; 22.30; 2Q27 f1.4; 4Q169 f3–4iii.4; 4Q256 23.1; 4Q264 f1.10; 4Q266 f2i.5, 14; f9ii.18; f9iii.5; 4Q268 f1.8; 4Q298 f3–4ii.9; 4Q372 f8.6; 4Q377 f2ii.2; 4Q381 f1.2; f31.5; f45a+b.1; f76–77.8; 4Q382 f15.2; 4Q390 f1.6; f2i.7; 4Q397 f14–21.10; 4Q401 f16.4; 4Q418 f46.1; f77.3; f189.2; 4Q418a f8.2; 4Q421 f1aii–b.14; 4Q424 f3.2; hifʿil—CD 2.14; 8.12; 13.5; 19.24; 1QS 3.13; 4.22; 6.15; 1QSa 1.5; 1QHa 4.33; 5.13, 14, 30; 10.20; 18.23; 19.31; fC3.4; 1Q34bis f3ii.3, 4; 4Q249a f1.2; 4Q267 f1.6; 4Q270 f2ii.21; 4Q298 f1–2i.2; 4Q299 f34.3; 4Q302 f2ii.2; 4Q303 f1.1; 4Q372 f2.5; f3.3; f8.4; 4Q379 f22i.4; 4Q381 f45a+b.1; f47.3; f49.2; f85.1; 4Q387 fA.4; 4Q398 f14–17ii.4; 4Q402 f4.14; 4Q408 f3+3a.7; 4Q415 f11.5, 6; 4Q416 f4.3; 4Q417 f1i.1, 14, 18; f1ii.10; 4Q418 f2+2a–c.7, 8; f17.2; f81+81a.15; f102a+b.3; f122i.5; f123ii.4, 5; f158.4; f176.3; f205.2; f221.2, 3; f227.1; f273.1; 4Q418a f7.2; 4Q423 f7.7; 4Q428 f10.6; 4Q443 f2.8; 4Q504 f1–2Rii.17; 4Q509 f4.4; f12i–13.3; 4Q525 f6ii.2; f14ii.18; 5Q13 f1.9. ‫זי"ד‬: qal—4Q514 f1i.7; ambiguous—4Q364 f13a– b.2; 11Q19 56.11; hifʿil—4Q171 f3–10iv.15; 4Q511 f68.4. ‫רי"ב‬: qal—1QSa 1.13; 4Q176 f1–2i.2; 4Q299 f62.2; 4Q417 f2i.14; ambiguous—1QS 4.23; 1QHa 17.23; 25.15; 1Q36 f2.1; f10.1; 4Q175 1.15; 4Q185 f4ii.3; 4Q251 f4–7i.2; 4Q299 f59.2, 7; 4Q418 f81+81a.7; hifʿil—4Q390 f2i.6. ‫לי"ץ‬: qal—4Q468i f1.1; hifʿil—1QpHab 8.6; 4Q184 f1.2. ‫נח"י‬: qal—4Q408 f3+3a.7; hifʿil—1QS 9.18; 4Q256 18.1; 4Q259 3.16. ‫יס"ף‬: qal—4Q252 1.19, 20; ambiguous—4Q252 1.16; 4Q416 f2ii.10; 4Q417 f2i.18, 20; 4Q418 f137.2; f199.2; PAM43685 f48.2; hifʿil—1QS 2.11; 6.14; 1QpHab 6.1; 8.12; 11.15; 1QHa 9.37; 1Q14 f8–10.7; 4Q265 f4ii.3; 4Q266 f6iv.8; 4Q286 f7i.8; 4Q298 f3–4ii.5, 6, 7, 8; 4Q299 f6ii.18; f30.5; 4Q416 f2iii.6; f2iv.7; 4Q418 f81+81a.17; f162.3; f221.3; 4Q420 f2.3; 4Q436 f1a+bi.2; 4Q502 f3.1; 4Q503 f15–16.10; 4Q525 f1.3; 11Q19 54.6; 56.18; 61.11; ‫שי"ר‬: hifʿil—4Q427 f7.18 (Qimron 2010, I:102, fn. 18). ‫יל"ד‬: hifʿil—1QSa 2.11; 4Q180 f1.5; 4Q225 f2i.8; f2ii.11; 4Q226 f7.3; 4Q338 2.1; 3.1. 11. Hifilisation 251 5.3. Samaritan Hebrew ‫יס"ף‬: MT qal || SP plene hifʿil—Deut. 20.8; MT ambiguous || SP plene hifʿil— Gen. 8.21, 21; 37.5, 8; 44.23; Exod. 5.7; 9.28; Deut. 3.26; 4.2; 13.1, 12; 17.16; 18.16; 19.20; MT plene hifʿil || SP ambiguous—Deut. 25.3, 3. 5.4. The Tiberian Reading Tradition ‫נח"י‬: qal—Gen. 24.27; Exod. 13.17; 15.13; 32.34; Isa. 7.2; 58.11; Ps. 5.9; 27.11; 60.11; 77.21; 108.11; 139.24; hifʿil—Gen. 24.48; Exod. 13.21; Num. 23.7; Deut. 32.12; 1 Sam. 22.4; 1 Kgs 10.26; 2 Kgs 18.11; Isa. 57.18; Ps 23.3; 31.4; 43.3; 61.3; 67.5; 73.24; 78.14, 53, 72; 107.30; 139.10; 143.10; Job 12.23; 31.18; 38.32; Prov. 6.22; 11.3; 18.16; Neh. 9.12, 19. 12. PIELISATION Complementing the shifts from G-stem to N-stem (nifalisation, ch. 10) and from G-stem to C-stem (hifilisation, ch. 11), part of ancient Hebrew’s long, gradual, and partial shift away from qal involved shifts from G-stem to D-stem, i.e., pielisation. Due to the orthographic identity of most qal and piʿʿel forms in all but their respective active and passive participial forms, it can be difficult to detect qal > piʿʿel shifts, especially in ancient corpora without a recorded reading tradition. Even so, evidence for pielisation across ancient Hebrew corpora and traditions, both biblical and extra-biblical, is extensive, especially when comparing late antique Hebrew with earlier material. The present chapter utilises as a springboard Fassberg’s (2001) survey of Qumran, BS, the Samaritan reading tradition, Tannaitic and Amoraic RH, and Paytanic Hebrew, for which he collects examples from various ancient Hebrew traditions and cites numerous expert opinions. An effort is made here to augment previous studies by pointing out evidence hitherto unnoticed. Conspicuously absent from several previous studies of pielisation in ancient Hebrew is a discussion of the trend as a sign of distinction between Tiberian CBH and LBH. If, however, scholars find substantial evidence of G- to D-stem movement in Second Temple material, it is also reasonable to expect at least a hint of this in Tiberian LBH when compared to CBH. © 2023 Aaron D. Hornkohl, CC BY-NC 4.0 https://doi.org/10.11647/OBP.0310.12 254 The Historical Depth of the Tiberian Reading Tradition 1.0. Second Temple Evidence 1.1. Late Biblical Hebrew In various ways and to varying degrees, use of the following piʿʿel verbs manifests LBH pielisation: 1.1.1. ‫‘ ֵב ֵרר‬purify’ In Tiberian BH, the piʿʿel comes in Dan. 11.35. Elsewhere, synon- ymous qal (Ezek. 20.38; Eccl. 3.18) and hifʿil (Jer. 4.11; 51.11) forms and middle/passive forms in nifʿal (2 Sam. 22.27, 27; Isa. 52.11; Ps. 18.27) and hitpaʿʿel (Ps. 18.27) occur. Significantly, probable piʿʿel forms are found in the NBDSS (1QS 1.12; 4.20; 1QHa 7.23; 4Q369 f1ii.5) with likely cases in RH (m. ʿEruvin 4.5; m. Tamid 2.5). The verb has a D-stem Aramaic cognate. 1.1.2. ‫‘ גֵ ַאל‬defile’ Most occurrences of verbs with this root are late in Tiberian BH. Nifʿal forms come in historically questionable Zephaniah (3.1) as well as transitional or early post-exilic texts (Isa. 59.3; Lam. 4.14). The piʿʿel (Mal. 1.7), puʿʿal (Mal. 1.7, 12; Ezra 2.62; Neh. 7.64), and hitpaʿʿel (Dan. 1.8, 8) are more characteristic of LBH proper, and apparently come in the NBDSS (see below, §1.2.1), as well. There is also a late noun ‫ *ג ַֹאל‬in Neh. 13.29. 1.1.3. ‫‘ ִח ֵקר‬investigate’ Qal ‫‘ ָח ַקר‬search, investigate, explore’ occurs 22 times in the He- brew Bible, while the nifʿal ‫(‘ (לֹא) נֶ ְח ַקר‬un)explored, (un)measured, (im)measurable’ comes five times; the piʿʿel appears only 12. Pielisation 255 in Qoh. 12.9. It has also been identified in the NBDSS at 4Q420 f1aii–b.3 (see below, §1.2.2). 1.1.4. ‫‘ ִט ֵלל‬cover, overshadow’ ‫‘ ִט ֵלל‬cover, overshadow’ (Neh. 3.15) is evidently a borrowing from Aramaic, where the verb is also commonly in the D-stem; apparent Hebrew cognates include qal ‫‘ ָצ ַלל‬become dark’ (Neh. 13.19) and hifʿil ‫‘ ֵה ַצל‬provide shade’ (Ezek. 31.3). 1.1.5. ‫‘ ִמ ֵעט‬be few, become few’ The stative meaning in Qoh. 12.3 is elsewhere covered in BH by the qal ‫( ָמ ַעט‬cf. esp. Isa. 21.17; Jer. 29.6; 30.19; see also Exod. 12.4; Lev. 25.16; Ps. 107.39; Prov. 13.11; Neh. 9.32); cf. the common RH puʿʿal participle ‫‘ ממועט‬small, few’ (e.g., m. Peʾa 8.4). 1.1.6. ‫‘ נִ ֵסך‬pour out (a libation)’ Against the piʿʿel in 1 Chron. 11.18, there occur throughout Tiberian BH apparently synonymous forms in qal (Exod. 30.9; Isa. 29.10; 30.1; 40.19; 44.10; Hos. 9.4; Ps. 2.6) and hifʿil (Gen. 35.14; Num. 28.7; 2 Sam. 23.16; 2 Kgs 16.13; Jer. 7.18; 19.13; 32.29; 44.17, 18, 19, 19, 25; Ezek. 20.28; Ps. 16.4), with a qal internal or hofʿal passive (Exod. 25.29; 37.16). In the Mishna, the piʿʿel occurs to the exclusion of qal or hifʿil (e.g., m. ʿAvoda Zara 5.6; m. Zevaḥim 13.6). The Targumic cognate is also D-stem. 1.1.7. ‫‘ נִ ֵתץ‬tear down, break down’ Piʿʿel forms are found mainly, but not exclusively, in late texts (Deut. 12.3; Ezek. 16.39; 2 Chron. 31.1; 33.3; 34.4, 7; 36.19); 256 The Historical Depth of the Tiberian Reading Tradition however, consonantally unambiguous piʿʿel forms are found only in Chronicles (2 Chron. 31.1; 34.4, 7; 36.19). Synonymous qal forms are the norm in BH, with some thirty cases (e.g., Exod. 34.13). Passives are vocalised as either nifʿal (Jer. 4.26; Nah. 1.6) or qal internal passive/puʿʿal (Judg. 6.28). The piʿʿel is also known from SH (Lev. 14.45 || MT qal). 1.1.8. Related Phenomena Disappearance of Qal Internal Passive Additionally, one indirect result of pielisation in LBH (and other late antique Hebrew sources) noted by Fassberg (2001, 252–55) is the disappearance of the qal internal passive. While accepting the reality of phonetic and morphosemantic factors, Fassberg opines that the shift of *quṭal forms to quṭṭal was facilitated by broader movement from qal to piʿʿel. Increased Usage of Puʿʿal Participles A not unrelated development in exilic and post-exilic Hebrew was increased usage of puʿʿal participles in place of various classical alternatives. Focusing principally on the linguistic periodisation of Ezekiel vis-à-vis the Priestly Source of the Pentateuch, Hurvitz (1982, 27–30, 35–39; 1983) calls attention to the late distribution of such terms as ‫‘ ְּמ ֻח ָלל‬defiled’, ‫‘ ְּמט ָֹהר‬purified’, ‫‘ ְּמ ֻט ָמא‬defiled’, ‫‘ ְּמ ֻק ָדש‬sanctified’, and ‫‘ ְּמ ֻר ַבע‬square’. While rejecting the extreme position that such terms were necessarily coined in exilic or post-exilic times, Hurvitz (1982, 29–30) argues that their historical usage follows a clearcut chronological sequence, accord- 12. Pielisation 257 ing to which the puʿʿal participles dominate in the late period. More diagnostically characteristic of LBH proper are: • ‫‘ ְּמ ֻל ָב ִשים‬dressed’ (1 Kgs 22.10 || 2 Chron. 18.9; Ezra 3.10; 2 Chron. 5.12) – qal ‫‘ ָל ַבש‬wear’ comes over 60 times in the Hebrew Bible, joined by a causative hifʿil 30+ times, with no piʿʿel; use of the puʿʿal continues in RH (e.g., t. Shabbat 8.17).1 • ‫‘ ְּמפ ָ ֵֹ֔ר ֶצת‬broken down’ (Neh. 1.3; see also the ketiv ‫המפרוצים‬, qere ‫רּוצים‬ ִֵ֔ ‫ ֵהם ְּפ‬Neh. 2.13) – in place of the expected ‫רּוצה‬ ָ ‫ְּפ‬ (Prov. 25.28; Neh. 2.13 qere; 4.1; 2 Chron. 32.5; and cf. the standard qal form—42 times—against zero piʿʿel forms). • ‫‘ ְּמפ ָ ָֹׂ֑רש‬made distinct’ (Neh. 8.8) – ‫( פ ַ ֵֹ֔רש‬Num. 15.34) is an- alysable as a qal internal passive on the basis of ‫‘ ִל ְּפ ֵ֥ר ֹש‬to clarify’ (Lev. 24.12); the piʿʿel is also attested in SH (§1.3.1) and RH (§1.5), the puʿʿal in the NBDSS (§1.2.2) and RH (§1.5). 1.2. Dead Sea Scrolls Fassberg (2001, 245–46) collects examples of various categories of piʿʿel replacements of qal cited by scholars, e.g., • ‫‘ )ונהלכה=( ונאלכה‬that we may walk’ (1QIsaa 2.10) || ‫וְּ נֵ ְּל ָ ֶ֖כה‬ (MT Isa. 2.3) 1 Possibly also to be read in 11Q17 9.7, but the context is broken and the reading questionable (see the Maʾagarim website). 258 • The Historical Depth of the Tiberian Reading Tradition ‫(‘ ישח{{ו}}קו‬they) mock’ (1QpHab 4.6)2 || ‫( יִ ְּש ֵָ֔חק‬MT Hab. 1.10)3 • ‫‘ הממכרת‬who sells, i.e., betrays’ (4Q169 f3–4ii.7) || ‫ַהמ ֶֹכ ֶַ֤רת‬ (MT Nah. 3.4; cf. §1.3.1, below) A Tiberian BH qal internal passive is twice replaced with puʿʿal in 1QIsaa: • ‫‘ וממורט‬polished’ (1QIsaa 14.25) || ‫ּומֹורט‬ ֵ֔ ָ (MT Isa. 18.2) • ‫( וממרט‬1QIsaa 15.1) || ‫ּומֹורט‬ ֵ֔ ָ (MT Isa. 18.7) A puʿʿal participle comes where a qal passive participle is expected in the case of: • ‫‘ מגולי אזן‬with opened ears’ (1QM 10.11); cf. ‫‘ גָ ָלה אֹזֶ ן‬open the ear’ (e.g., 1 Sam. 9.15) and ‫‘ ּוגְּ לֵ֥ ּוי ֵע ָינֵַּֽ֣יִ ם‬and with opened eyes’ (Num. 24.4, 16) (see below, §3.3) To Fassberg’s list of BDSS qal > piʿʿel shifts, the following may be added. 1.2.1. The Biblical Dead Sea Scrolls ‫‘ ִד ֶבר‬speak’ Though piʿʿel morphology prevails for this verb in the MT, significant qal vestiges remain (see below, §3.1, for details). Most cases of MT qal forms of ‫ דב"ר‬are paralleled by qal forms in the BDSS, with the following as a notable exception. 2 The waw is marked for erasure by dots above and below. 3 The citation of a parallel in MT Lam. 1.7 in Fassberg (2001, 245) is evidently an error. 12. Pielisation • 259 ‫(‘ הול{{ו}}ך צדקות וידבר מישרים‬he) walks righteously and speaks uprightly’ (1QIsaa 27.16) || ‫יש ִ ָׂ֑רים‬ ָ ‫ה ֵֹלְך ְּצ ָד ֵ֔קֹות וְּ ד ֵ ֶֹ֖בר ֵ ַֽמ‬ (MT Isa. 33.15) Here, whether due to textual or linguistic factors, or to a combination of these and/or other factors, 1QIsaa presents what is most reasonably interpreted as a piʿʿel prefix conjugation form, which arguably contemporises the language at the expense of the poetry (cf. the preceding participle). ‫‘ ִרנֵ ן‬sing’ The MT knows common qal and piʿʿel forms, as well as rarer hifʿil and puʿʿal ones, with no obvious difference in meaning (though there may well have been one). What is more, both the qal and the piʿʿel persist in late biblical traditions. Crucially, however, in late non-biblical corpora, especially the NBDSS, but RH as well, piʿʿel usage dominates markedly over that of qal. Thus, the following example from the Great Isaiah Scroll, may be part of a broad qal > piʿʿel shift. • ‫‘ ירננו‬they will sing; (1QIsaa 52.20) || ‫( יָ ֶ֖ר ֹנּו‬MT Isa. 65.14) Local Shifts Piʿʿel > Qal in the Biblical Dead Sea Scrolls In the interests of comprehensiveness, it may be noted that stem change between the MT and the BDSS involving qal and piʿʿel does not always reflect the direction exemplified above, i.e., qal > piʿʿel. Select cases of the reverse are also known, e.g., • ‫‘ ַוֵַּֽ֣יְּ ַס ְּק ֵ֗ ֵלהּו‬and he destoned it’ (MT Isa. 5.2) || ‫( ויסקולהו‬1QIsaa 4.13) • ‫‘ ַס ְּקלּו‬destone!’ (MT Isa. 62.10) || ‫( סקולו‬1QIsaa 50.23) 260 • • • The Historical Depth of the Tiberian Reading Tradition ‫‘ ְּל ַב ֵ ֶ֖צר‬to fortify’ (MT Isa. 22.10) || ‫( לבצור‬1QIsaa 17.14) ‫‘ יְּ ַש ֵ ֶ֖בר‬he shatters’ (MT Isa. 38.13) || ‫( ישבור‬1QIsaa 32.5) ‫‘ ֲא ַש ֵֵ֔בר‬I will break in pieces’ (MT Isa. 45.2) || ‫( אשבור‬1QIsaa 38.8) • ‫‘ ַו ְַּֽת ַח ְּש ֵ ַֽבהּו‬that you consider him’ (MT Ps. 144.3) || ‫ותחושבהו‬ (1Q5 23.15) • ‫‘ ְּל ַל ֵ ֵ֔קט‬to glean’ (MT Ruth 2.23) || ‫( ללקוט‬2Q16 f5ii–6i.2) In these cases, it may be that the DSS text preserves an ancient qal form that secondarily shifted to piʿʿel in the Tiberian reading tradition, presumably for purposes of semantic disambiguation, e.g., qal ‫‘ ָס ַקל‬stone (to death)’ versus piʿʿel ‘destone (a field, road); throw stones’, qal ‫‘ ָב ַצר‬harvest grapes, trim vines’ versus piʿʿel ‫*ב ֵצר‬ ִ ‘fortify’, qal ‫‘ ָש ַבר‬break’ versus piʿʿel ‫‘ ִש ַבר‬shatter, break into pieces’, qal ‫‘ ָח ַשב‬think’ versus piʿʿel ‫‘ ִח ַשב‬consider, calculate’. On the other hand, since the qal form in these cases is often the more common alternative, it may be that the technical piʿʿel lectio difficilior was inadvertently replaced bwithy the better-known form. In the specific case of ‫( ְּל ַל ֵ ֵ֔קט‬MT Ruth 2.23) || ‫( ללקוט‬2Q16 f5ii– 6i.2), there also seem to be local textual factors at work. In the MT generally and in MT Ruth more specifically there is a mix of qal ‫ ָל ַקט‬and piʿʿel ‫ל ֵקט‬, ִ the qal with perhaps a more generic sense of ‘gathering’, the piʿʿel referring specifically to ‘gathering by the less fortunate at harvest time’. As MT Ruth has both qal and piʿʿel infinitive construct forms, it may be that the tradition preserved in 2Q16 reflects secondary harmonisation of the original piʿʿel in Ruth’s actions in Ruth 2.23 to match the qal of Boaz’s instructions in Ruth 2.8. Whatever the case may be, the difference between these examples and the examples of the qal > piʿʿel shift dis- 12. Pielisation 261 cussed above is that while the latter appear to be part of a broad trend, the former seem to be more case-specific in nature. 1.2.2. The Non-biblical Dead Sea Scrolls Fassberg focuses chiefly on acknowledged differences between Tiberian BH and DSSBH, but also observes the following probable instance of qal > piʿʿel movement in the NBDSS: • ‫‘ מרוגלת‬attached’ (1QM 5.13); cf. RH ‫‘ רגול‬bound’ (m. Shabbat 5.3) To this list it is possible to add further examples. ‫‘ גאל‬defile’ Alongside piʿʿel ]‫‘ מגאלי֯ [ם‬defiling’ (4Q513 f13.3; perhaps also 4Q274 f1i.6; 4Q284a f1.7; 11Q19 47.13) and puʿʿal ‫‘( יגאולו‬that) are desecrated’ (CD 12.16) the hitpaʿʿel also occurs (1QM 9.8; 4Q379 f3i.5); for the Tiberian biblical distribution, see above, §1.1.2. ‫‘ דחק‬charge’ ‫‘ ומדחק‬and charging’ (4Q223–224 f2iv.13) is clearly piʿʿel. Verbs with this root are rare in the MT, occurring only in qal in the sense ‘press’; the Aramaic G-stem serves in a similar meaning in the Targums, though D-stem forms are comparatively more common in the Jerusalem Targum (i.e., Targum Jerusalem). ‫‘ חבא‬hide’ In Tiberian BH, the transitive form is hifʿil (6x), while the middle (reflexive/intransitive) sense is typically encoded with nifʿal 262 The Historical Depth of the Tiberian Reading Tradition (16x) or hitpaʿʿel (10x). A hofʿal passive is known (Isa. 42.22), as is a possible puʿʿal or qal internal passive in MT ‫‘ ִֻׁ֝ח ְּב ֵ֗אּו‬are hidden’ (Job 24.4). The NBDSS have the clearcut puʿʿal participle ‫מחובאים‬ ‘hidden (things)’ (1QHa 16.7, 19; see also, perhaps ‫ וחבא‬1QS 4.6). ‫‘ חקר‬seek, investigate’ ‫‘ מחקר‬seeking’ (4Q420 f1aii–b.3) may be a piʿʿel participle in line with the LBH piʿʿel form seen above (§1.1.3), but the syntax may just as well point to a nominal form or to an Aramaic-style infinitive. ‫‘ נדב‬commit’ Tiberian BH shows qal and hitpaʿʿel usage (the latter with specifically late semantics in LBH; see Hurvitz 2014, 179–81), one or both of which are also evidenced in SH, RH, and BS; RH and the NBDSS also add nifʿal forms. Against the MT’s transitive qal, the NBDSS passive puʿʿal form ‫‘ המנודבים‬those who are committed’ (4Q501 f1.3) seems indicative of pielisation. ‫‘ סכך‬confine’ The context of ‫‘ וכמסככה‬and like a confined (woman)’ (4Q179 f2.7) arguably indicates a puʿʿal participle. In Tiberian BH, the relevant forms are qal and hifʿil (though qal yaqṭel morphology may also be conjectured for some prefix conjugation forms), not piʿʿel or puʿʿal. Piʿʿel forms are common in RH, especially in the context of the sukkah (e.g., m. Sukkah 1.4). 12. Pielisation 263 ‫‘ פחד‬fear’ In the MT, against 22 qal cases come just two cases of piʿʿel. In both Isa. 51.13 and Prov. 28.14, the piʿʿel occurs with the adver- bial ‫‘ ָת ִמיד‬always’, once with ‫ל־היֵ֗ ֹום‬ ַ ‫‘ ָכ‬all day’ (Isa. 51.13). It is conceivable that the biblical piʿʿel began with a more intensive (pluractional/iterative) meaning than the qal, but that the two forms eventually became virtual synonyms.4 An active participle with no accompanying pluractional/iterative adverb comes in 4Q381 f31.8 (see also 1QS 4.2; 4Q510 f1.4; 4Q511 f8.4; f48– 49+51.25); see also on BS (see below, §1.4.3). ‫‘ פרש‬clarify’ Tiberian BH attests active qal (Lev. 24.12) and passive qal (or puʿʿal) (Num. 15.34) and nifʿal (Ezek. 34.12), with the only ex- plicit puʿʿal in LBH Neh. 8.8 (see above, §1.1.8). Like LBH, the NBDSS have explicitly puʿʿal ‫‘ מפורשים‬made distinct’ (4Q177 f1– 4.11); cf. the piʿʿel in SH (see below, §1.3.1) and the piʿʿel and puʿʿal in RH (see below, §1.5). 4 Modern Hebrew knows a quasi-suppletive paradigm not dissimilar from the paradigm in Tiberian BH (see https://hebrew-academy.org.il/2011/07/08/‫ומפחד‬-‫פוחד‬/). 5 In several of the potential NBDSS examples, the consonantal form is ambiguous, i.e., is analysable as qal or piʿʿel, and some take the meaning of the piʿʿel to be causative (as in early Paytanic Hebrew, on which see Rand 2006, 190). 264 The Historical Depth of the Tiberian Reading Tradition ‫‘ פרש‬spread (a net)’ In the Hebrew of the NBDSS, one encounters ‫‘ מפרשי רשת‬netspreaders’ (1QHa 21.24 || 4Q427 f11.2 || 4Q428 f13.7–8 [?]). In Tiberian BH, cases of qal ‫‘ ָפּ ַרׂש‬spread’ outnumber cases of the synonymous piʿʿel by a margin of 54 to 9, though it is important to note that this applies to all biblical chronolects and that the piʿʿel is absent from LBH proper. However, collocations involving ‫ פר"ׂש‬and ‫ ֶר ֶשת‬come nine times in BH, always employing a qal verb (Ezek. 12.13; 17.20; 19.8; 32.3; Hos. 5.1; 7.12; Ps. 140.6; Prov. 29.5; Lam. 1.13), which makes the NBDSS shift to the piʿʿel in this collocation especially conspicuous. It may be significant that the qal > piʿʿel shift applies specifically to cases of the active participle with substantival (nominal/adjectival) semantics, a category that excludes the biblical tokens. ‫‘ רחץ‬wash, bathe’ Against the single NBDSS case of puʿʿal ‫‘ מרחצים‬washing, rinsing (tr.)’ (11Q19 34.10), in Tiberian BH the verb is consistently qal, whether reflexive, e.g., ‫אר‬ ֹ ֵ֔ ְּ‫ל־הי‬ ַ ‫‘ ִל ְּרחֹץ ַע‬to bathe by the Nile’ (Exod. 2.5), weakly transitive, e.g., ‫‘ וְּ ַר ֲח ֶ֖צּו ַרגְּ ֵל ֶיכָׂ֑ם‬so you (MPL) may wash your (MPL) feet’ (Gen. 18.4), or strongly transitive, e.g., ‫וְּ ָר ַח ְּצ ָ ֵ֥ת א ָ ֶֹ֖תם‬ ‘and you (MS) will wash them (i.e., Aaron and his sons)’ (Exod. 29.4) (the apparent puʿʿal forms in Ezek. 16.4 and Prov. 30.12 should arguably be analysed as qal internal passives). This is generally the case in the NBDSS, too. However, compare Tiberian BH ‫ּוכ ָר ֵ֔ ָעיו‬ ְּ ‫בֹו‬ ֙ ‫‘ וְּ ָר ַח ְּצ ָ ַ֤ת ִק ְּר‬and you must wash its entrails and its legs’ (Exod. 29.17; see also Lev. 1.9, 13; 8.21; 9.14; Isa. 4.4) with NBDSS ‫‘ ומרחצים את הקרבים ואת הכרעים‬and washing the entrails and 12. Pielisation 265 the legs’ (11Q19 34.10–11). The piʿʿel also occurs in Amoraic Hebrew (y. Shabbat 9.3). ‫‘ שנא‬hate’ Tiberian BH forms of ‫ ׂשנ"א‬reflect a basically qal paradigm: ‫ׂשנֵ א‬ָ ‫ׂש ֹנֵ א‬-‫יִ ְׂשנָ א‬-‫)ׂשנ ֹא‬ ְ ‫(ל‬/‫ה‬ ִ ‫ׂשנְ ָא‬-‫א‬ ִ ֹ ‫ׂשנ‬-‫נּוא‬ ָ ‫*ׂש‬, ָ verbal passive ‫*נִ ְׂשנָ א‬. The excep- tion is the piʿʿel participle with substantival semantics ‫‘ ְמ ַׂשנֵ א‬enemy’, which appears 15 times throughout biblical literature. Of particular interest is the term used for a less-favoured wife, viz. the qal passive participle ‫נּואה‬ ָ ‫( ְׂש‬Gen. 29.31, 33; Deut. 21.15, 15, 16,17; 2 Sam. 5.8; Isa. 60.15; Prov. 30.23). Against this contextual background, one may consider the NBDSS puʿʿal participle ‫‘ משונאה‬unloved, despised, hated’ (4Q179 f1ii.3). Though the context is broken, it appears that the MT passive qal participle has been replaced in the Qumran text with a puʿʿal participle. Cf. BS for a yiqṭol form of the puʿʿal (see below, §1.4.4). 1.3. Samaritan Hebrew While the Samaritan written tradition largely resembles its Tiberian counterpart when it comes to the distribution and semantics of verbal stems, the Samaritan reading tradition exhibits systematic deviations away from the qal in favour of nifʿal (see above, ch. 10), hifʿil (see above, ch. 11), and piʿʿel. Indeed, in comparison not just to the written and reading components of the Tiberian biblical tradition, but to recognised Second Temple Hebrew biblical and non-biblical corpora, the Samaritan reading tradition exhibits an advanced stage of pielisation. This manifests in two main ways: wholesale or partial movement to standard D-stem 266 The Historical Depth of the Tiberian Reading Tradition piʿʿel/puʿʿal, with expected gemination of the middle radical (§1.3.1); wholesale or partial movement to piʿel B/puʿal B, i.e., Dstem with singleton middle radical (§1.3.2). A potentially related phenomenon is the development of qal B prefix conjugation forms, whose patterns resemble that of piʿel B (§1.3.3). Given the extensiveness of pielisation and related shifts in the Samaritan reading tradition, no attempt at exhaustiveness is made in the following treatment. 1.3.1. Qal > Piʿʿel ‫‘ מכר‬sell’: Comprehensive Shift Relative to the Tiberian biblical tradition, the SP shows comprehensive G- to D-stem shifts in the case of the verbs ‫‘ גלל‬roll’, ‫חנך‬ ‘dedicate, educate’, ‫‘ מכר‬sell’, ‫‘ ענש‬punish’, ‫‘ פרש‬explain’, and ‫קרע‬ ‘tear’. As the most common of these, ‫‘ מכר‬sell’ serves as a useful example. The dominant Tiberian active-passive qal-nifʿal arrangement is mirrored in the SP by an active-passive arrangement consisting of piʿʿel-nifʿal B (i.e., nitpaʿʿel with assimilated tav), e.g., piʿʿel ‫ מכרתם‬makkertimma ‘you (MPL) sold’ (Gen. 45.4) and nifʿal B ‫ ונמכר‬wnimmakkar ‘then he must be sold’ (Exod. 22.2). The Samaritan D-stem extends even to active participles without the characteristic preformative -‫מ‬, as in ‫ מכר‬makkǝr ‘is selling (MS)’ || MT ‫( מ ֵֹכֶ֖ר‬Lev. 25.16). For historical context, it is worth noting that a D-stem form of ‫ מכ"ר‬occurs in the NBDSS: ‫‘ הממכרת‬who sells, i.e., betrays’ (4Q169 f3–4ii.7) || ‫( ַהמ ֶֹכ ֶָ֤רת‬MT Nah. 3.4). It may also be relevant that the Aramaic equivalent ‫‘ זבן‬sell’ is also D-stem (cf. G-stem ‫‘ זבן‬buy’). 12. Pielisation 267 ‫‘ דבר‬speak’: Unification of a Mixed Paradigm In other cases of apparent Samaritan pielisation vis-à-vis qal use in the MT, the SP presents a unified piʿʿel conjugation against a mixed Tiberian paradigm. The Tiberian arrangement sometimes involves a semantic distinction between G- and D-stem, as in the case of ‫‘ זרה‬winnow’, ‫‘ לקט‬collect, gather, glean’, and ‫‘ קבץ‬gather, collect, assemble’. An alternative Tiberian arrangement is that of dominant piʿʿel morphology with vestigial qal forms, as in the well-known case of ‫‘ דבר‬speak’. In this case, against the MT’s 1000+ piʿʿel forms and forty apparently synonymous qal parti- cipial (active and passive) and infinitival forms (and nifʿal passives), the Samaritan paradigm is comprehensively piʿʿel, including piʿʿel active participles without the characteristic prefix -‫מ‬, e.g., ‫ דברות‬dabbērot ‘speak (FPL)’ (Num. 27.7; see also Gen. 16.13; Exod. 6.29; Num. 32.27; 36.5; Deut. 5.1; 15.9) (see below, §3.1).6 ‫‘ משח‬anoint’: Formal and Semantic Suppletion Finally, Samaritan pielisation can result in suppletive paradigms, whether formal or semantic/grammatical. Consider the case of ‫‘ משח‬anoint’. Against a consistently qal Tiberian paradigm (with corresponding nifʿal medio-passive), the SP preserves qal mor6 In the case of MT hitpaʿʿel ‫ || ִמ ַד ֵבר‬SP piʿʿel ‫ מדבר‬amdabbər ‘[the voice] speaking’ (Num. 7.89; cf. Ezek. 2.2; 43.6), the Samaritan D-stem is likely more original, with the Tiberian tradition exhibiting a secondary shift to hitpaʿʿel as part of the broad Second Temple trend of avoiding anthropomorphisms of the deity (see, especially, the Targums; Ben-Ḥayyim 2000, 218, fn. 189; see below, ch. 13, §2.2.4). 268 The Historical Depth of the Tiberian Reading Tradition phology in the qaṭal (13x, e.g., ‫ משחת‬māšatta Gen. 31.13), passive ̊̄ rather qaṭūl ̊̄ participle (5x, e.g., ‫ משחים‬mɑ̊̄ˈšīm Lev. 2.4, with qaṭīl morphology), and infinitive construct (‫ משחו‬mɑ̊̄šāʾu Lev. 7.36). In six of seven cases of the yiqṭol, conversely, a piʿʿel form obtains (e.g., ‫ תמשח‬tēmašša Exod. 30.30). The distinction between the dominant piʿʿel yiqṭol forms and the lone qal yiqṭol exception ‫ימשח‬ yimša (Lev. 16.32) may be explicable in terms of pluractionality—all cases of the piʿʿel involve multiple objects,7 whereas the verb in Lev. 16.32 has a single object. Beyond Samaritan Hebrew, D-stem ‫ משח‬is not known from ancient Hebrew. However, the relevant Aramaic form is D-stem ‫( רבי‬e.g., TO Num. 35.25).8 ‫‘ בכה‬weep; mourn’: Semantic/Grammatical Suppletion Semantic and/or grammatical suppletion obtains when different cognate stems have diverse semantics and/or valency. Especially illustrative is the case of ‫‘ בכה‬weep; mourn’. In the Tiberian BH tradition, qal morphology is nearly exclusive (112x), with just two piʿʿel participle exceptions (Jer. 31.15; Ezek. 8.14). Rare Dstem forms in the face of far more common G-stem morphology are known from Tannaitic RH, QA, and Syriac (Maʾagarim, s.v.; CAL, s.v.). For its part, SH is characterised by a complex situation of suppletion involving qal, piʿʿel, and qal B forms (see below, 7 This includes Lev. 8.12, where, notwithstanding the singular gram- matical object in the immediate context, it is clear from Lev. 8.10–11 that multiple objects are anointed. 8 Formal suppletion occurs in the case of ‫‘ גרש‬drive away, divorce’ (ves- tigial qal use in Tiberian), ‫‘ יסף‬add, do again’ (partial qal > hifʿil shift in Tiberian), ‫‘ נטש‬allow, leave, forsake’, ‫‘ שלח‬send’. 12. Pielisation 269 §1.3.3). The suppletion appears generally to involve both grammatical and formal factors. All infinitives construct are piʿʿel (Gen. 23.2; 43.30), and other than the infinitive at Gen. 43.30, piʿʿel forms consistently take a direct object, i.e., have the mean- ing ‘mourn (trans.)’ (6x: Gen. 23.2; 37.35; 50.3; Lev. 10.6; Num. 20.29; Deut. 21.13; 34.8). For their part, intransitives are characterised by formal suppletion: qal suffix conjugation forms (2x: Gen. 45.14; Num. 11.18) and active participles (3x: Exod. 2.6; Num. 11.10; 25.6) and qal B prefix conjugation forms (16x: Gen. 21.16; 27.38; 29.11; 33.4; 42.24; 43.30; 45.14, 15; 46.29; 50.1, 17; Num. 11.4, 13, 20; 14.1; Deut. 1.45).9 ‫‘ ילד‬bear (a child); beget, father, sire’ SH, like Tiberian Hebrew, generally distinguishes between qal ‫ילד‬ ‘bear (a child)’ and hifʿil ‫‘ הוליד‬beget, father, sire’. On occasions where the MT presents a qal form that denotes ‘beget, father, sire’, SH does not tolerate the polysemy of the qal. In a few in- stances, disambiguation is achieved via hifilisation of verbs that refer to the act of the male (see ch. 11, §1.3.2), but this is far less common than the alternative strategy, namely, pielisation. On nine occasions, the SP has piʿʿel ‫ ילד‬yallǝd ‘he fathered’ parallel to MT qal ‫‘ יָ ַלד‬he bore, i.e., fathered’ (Gen. 4.18, 18, 18; 10.13, 15, 24, 24, 26; 25.3) and on one occasion piʿʿel ‫ ילד‬yallǝd ‘he fathered’ parallel to MT qal passive ‫‘ יֻ ַלד‬was born (M)’ (Gen. 10.21). This approach achieves the formal disambiguation of distinct semantic values that would otherwise be subsumed under the same 9 ‫‘ חשב‬consider, calculate’, ‫‘ ילד‬beget, sire, father; midwife’, ‫‘ עבד‬work, serve; worship’, ‫‘ עבר‬pass, cross’, and ‫‘ פרע‬let loose, go wild’. 270 The Historical Depth of the Tiberian Reading Tradition form, but it also results in a piʿʿel form with two distinct meanings separated by gender: masculine ‘beget, father, sire’, feminine (ac- tive participle) ‘serve as midwife’ (Gen. 35.17; 38.28; Exod. 1.15, 17, 18, 19, 19, 20, 21). Clear contextual and formal differences evidently made the association of such diverse semantic values with piʿʿel more tolerable than the original association of diverse meanings with the qal. 1.3.2. Qal > Piʿel B Alongside the standard D-stem, SH knows a less frequent, though by no means rare, D-stem form without middle radical gemination, which Ben-Ḥayyim (2000, 113–15, §§2.1.3.5–7) labels piʿel B. Though most of the relevant verbs are II-guttural, the frequency in this stem of select non-II-guttural verbs—namely, ‫כבד‬ ‘honour’, ‫‘ כפר‬atone’, and ‫‘ ספר‬tell, recount’—confirms the heuristic validity of the binyan.10 10 Cf. Tiberian Hebrew, where, due to the rarity of non-II-guttural D- stem forms with singleton middle radicals, it is more parsimonious to include II-guttural D-stem forms in the standard piʿʿel category and to account phonologically for the lack of gemination. In his discussion of D-stem forms without gemination, Ben-Ḥayyim (2000, 114, §2.1.3.6–7) adduces parallels from Babylonian RH, TO, and Babylonian and Tiberian BH. The examples of ‫ כפר‬with peh rafa, all from the Sifra, are compelling (Yeivin 1985, 515). Of the alleged Tiberian BH examples, ‫ְּמ ָל ְּש ָ֬ ִני‬ ‘slanderer’ (Ps. 101.5 qere) seems pertinent, but the additional examples listed by Ben-Ḥayyim, viz. ‫‘ ְּמ ָא ְּס ָפיו‬its (M) gatherers’ (Isa. 62.9) and ‫ְּת ָר ְּצחּו‬ ‘you (MPL) murder’ (Ps. 62.4), are variants that bear more conventional vocalisation in L and A: ‫ ְּמ ַא ְּס ָפ ֙יו‬and ‫ת ָר ְּצ ֵ֪חּו‬, ְּ respectively. 12. Pielisation 271 The mixed nature of the piʿel B template is most evident in the morphology of the active participle, which forms occur both with and without the standard prefix -‫מ‬, e.g., ‫ מצחק‬amṣɑ̊̄ʾəq ‘play, ̊̄ ‘ask, borrow’. Indeed, on the basis of joke, mock’ versus ‫ שאל‬šɑʾəl examples like the latter, a reasonable hypothesis is that some IIguttural piʿel B verbs began as qal statives with PS qaṭil morphology. The broader process of pielisation and the more restricted simplification of piʿʿel’s originally geminate middle radical seem ̊̄ ‘ask, to have converged, with the result that statives like ‫ שאל‬šɑʾəl ̊̄ borrow’ and ‫ אהב‬ɑʾǝb ‘love’ could be analysed as piʿel B.11 This was facilitated by the fact that the standard Samaritan piʿʿel par- ticiple requires no prefix -‫מ‬. On this basis, piʿel B prefix forms in ̊̄ ye̊̄qaṭǝl could be secondarily generated. It should be noted, though, that Ben-Ḥayyim (2000, 109, §2.1.1.7) accounts for gen̊̄ eration of the very similar qal B prefix conjugation pattern ye̊̄qaṭål on the basis of purely phonological shifts to the standard qal ̊̄ yiqṭål template, i.e., ye̊̄qaṭål < *yiqaṭål < *yiqṭål (see below, §1.3.3)—which could conceivably equally apply to the piʿel B prefix conjugation, too. Alternatively—or complementarily—the broad process of pielisation may have been a significant factor in ̊̄ and ye̊̄qaṭål ̊̄ prefix conjugathe secondary development of yeqaṭǝl tion forms. 11 Cf. the remnants of stative pronunciation of these verbs in the Tibe- rian tradition, e.g., ‫‘ ָא ֵ ַֽהב‬he loved’ (Gen. 27.9), ‫יהּו‬ ֙ ‫‘ ְּש ִא ְּל ִ֙ת‬I asked him’ (Judg. 13.6). 272 The Historical Depth of the Tiberian Reading Tradition ‫‘ גאל‬redeem’: Comprehensive Shift As a comprehensive shift from qal to piʿel B, consider the case of ‫‘ גאל‬redeem’. The Tiberian biblical paradigm is qal-nifʿal. SH preserves the nifʿal (Lev. 25.30, 49, 54; 27.20, 27, 28, 33), but all MT qal forms are paralleled by piʿel B forms in the SP (28x), e.g., ̊̄ MT ‫‘ וְּ גָ ַא ְּל ִ ַ֤תי‬and I will redeem’ || SP ‫ וגאלתי‬wgɑʾilti (Exod. 6.6), MT ̊̄ (Lev. 25.33). Signifi֙‫‘ יִ גְּ ַאל‬will redeem (3MS)’ || SP ‫ יגאל‬yēgɑʾəl cantly, this includes the participle (13x), e.g., MT ‫‘ ַהג ֵָֹֹ֨אל‬the re̊̄ ‘the redeeming (king)’ (Gen. deeming (angel)’ || SP ‫ הגאל‬aggɑʾəl 48.16). The latter are clear evidence of the qal > piʿel B shift. The Samaritan pielisation of this verb seems unique, as the D-stem is otherwise unattested in late antique Hebrew and Aramaic traditions, though the corresponding Aramaic ‫ פר"ק‬has occasional Dstem derivations (see CAL, s.v.).12 ‫‘ אחר‬tarry, delay, stay’: Unification of a Mixed Paradigm In other cases, consistent Samaritan piʿel B morphology parallels mixed G-/D-stem morphology in the MT, e.g., ‫‘ אחר‬tarry, delay, stay’. Most of the 16 cases in the MT are piʿʿel. Qal exceptions are ‫‘ וָ ֵא ַ ֶ֖חר‬and I remained’ (Gen. 32.5) and ketiv ‫ וייחר‬qere ‫ֹוחר‬ ֶ ָ֕‫‘ וַ י‬but he exceeded (the deadline)’ (2 Sam. 20.5). In the Samaritan tradi- tion, all parallels to Tiberian piʿʿel forms and the single qal excep- tion are piʿel B.13 12 Similar cases are ‫‘ געל‬loathe, detest’, ‫‘ מאס‬reject’, ‫ מהר‬II ‘pay a bride price’, ‫‘ מחץ‬strike, shatter, crush’, ‫‘ נאף‬commit adultery’, ‫‘ פעל‬do, make’, ‫‘ צעק‬cry out’, and ‫‘ שאב‬draw, pull’. 13 Similar cases include ‫‘ אחז‬take, grasp, seize; possess’, ‫‘ לחך‬lick’, ‫לחץ‬ ‘press’, ‫‘ נאץ‬spurn, despise’, ‫‘ פקח‬open (eyes)’, and ‫‘ צחק‬laugh, play, per- 12. Pielisation 273 1.3.3. Qal > Qal B ̊̄ In SH, certain verbs have prefix conjugation forms with a yēqaṭål ̊̄ pattern of the piʿel B (seen pattern, not dissimilar from the yēqaṭǝl above, §1.3.2). Ben-Ḥayyim (2000, 109, §2.1.1.7) groups such forms under the label qal B. Though the yiqṭol pattern of strong verbs of this type can be explained as a result of sound shifts in the standard qal prefix conjugation pattern—namely, yiqṭål > ̊̄ (Ben-Ḥayyim 2000, 109, §2.1.1.7)—its similar*yiqaṭål > yēqaṭål ̊̄ ity to the piʿel B pattern (yēqaṭǝl) and, for that matter, to the standard piʿʿel pattern (yēqaṭṭǝl), may also be attributed, even if partially, to the overall expansion of D-stem and D-stem-like vocalism. It is to be noted that qal B forms are limited almost exclusively to verbs III-r and III-y (< III-ʾ).14 The most common verb ̊̄ ̊̄ r. is ‫ זכר‬zakår ‘remember’ with prefix conjugation ‫ יזכר‬yēzɑkɑ Against the contention that this (along with other III-r forms) might be more parsimoniously classified as piʿel B, attributing the shift of ǝ > å of the middle radical to the following r, one need form, revel, jest, mock’. In most of the above, the Tiberian morphological diversity is semantically and/or grammatically explicable, though there are some cases, e.g., ‫‘ אחר‬tarry, delay, stay’ and ‫‘ לחך‬lick’, where there is no obvious semantic or grammatical difference between the MT qal and piʿʿel alternatives. The relevant verbs, with example forms, are ‫‘ בטא‬speak rashly’ yēbēṭɑ, ̊̄ r, ‫‘ דקר‬pierce’ wyēdɑqɑ ̊̄ r, ‫‘ בכה‬weep’ wyēbēki, wyēbēku, ‫‘ בקר‬seek’ yēbɑqɑ 14 ̊̄ r, ‫‘ פדה‬redeem’ tēfēdi, ‫‘ פנה‬turn’ wyɑfɑ ̊̄ nu, ̊̄ ‫‘ זכר‬remember’ wyēzɑkɑ ̊̄ ̊̄ ̊̄ wnēfɑna, ‫‘ פצל‬peel’ wyēfɑṣɑl, ‫‘ פצר‬urge, press’ wyēfɑṣɑr, ‫‘ פשה‬spread’ tēfēši, yēfēši, ‫‘ ראה‬see’ wyēre, wyērēʾu, wtēre, ‫‘ רעה‬shepherd, pasture, feed’ yēˈrū. 274 The Historical Depth of the Tiberian Reading Tradition ̊̄ only compare piʿel B ‫ ויספר‬wyēsɑfər, which occasions no such shift.15 Likewise, in the case of III-y (and similar) verbs, though it may be tempting to view apparent qal B forms, such as ‫ויבך‬ wyēbēki, as mere piʿʿel allomorphs, the existence of genuine piʿʿel ‫ ויבך‬wyēbakki militates against this. So, too, does the apparent morphosemantic distinction between the forms of ‫בכה‬, viz. intransitive qal/qal B ‘cry, weep’ and transitive piʿʿel ‫‘ בכה‬mourn’ (see above, §1.3.1). In sum, notwithstanding the apparent validity of the classification of qal B forms as a G-stem subcategory primarily reflecting processes of phonetic resyllabification, in a tradition characterised by various manifestations of pielisation, it is plausible to hypothesise that the morphological shift to D-stem was favourable to parallel phonetic developments. 1.4. Ben Sira According to Fassberg (2001, 246), Ben-Ḥayyim (1958, 238) gives two examples of qal > piʿʿel shift in BS, both from the medieval MS B from the Cairo Geniza. One involves the substitution of puʿʿal participle ‫( משואל‬SirB 16r.11 = Sir. 46.13) for the MT qal passive participle ‫‘ ָשאּול‬borrowed’ (1 Sam. 1.28). The other is 15 Perhaps relevant is Ben-Ḥayyim’s (2000, 113, §2.1.3.4) contention made regarding the unexpected å, rather than ǝ/ē, vocalisation after the second radical in certain piʿʿel prefix conjugation forms: It is likely that in SH the identity of the second radical in the perfect and the imperfect is considered an obligatory feature, and so the vowel characteristic of the perfect was transferred to the imperfect in the few verbs preserving the original a-vowel. 12. Pielisation 275 ‫( ניבע‬SirB 20r.8 = Sir. 50.27), which Ben-Ḥayyim interprets as a piʿʿel with the meaning ‘poured forth’ (cf. the qal in Prov. 18.4). To these may be added further examples of qal > piʿʿel movement. 1.4.1. ‫‘ יאש‬be weary, despair’ All but one of the MT’s six forms are nifʿal intransitives in the sense ‘become weary, despair’. The sole exception is the late transitive piʿʿel infinitive in Qoh. 2.20. BS’s ‫‘ ֯מי֯ ו֯ ֯אש‬hopeless’ (SirB 17r.18 = Sir. 47.23) is in line with the MT’s late piʿʿel usage and seems to take the place of more classical intransitive nifʿal. 1.4.2. ‫‘ עטף‬cover, be enveloped’ The rare and poetic verb in the MT is qal ‫‘ ָע ַטף‬cover, be envel- oped’ (Ps. 65. 14; 73.6; Job 23.9). In one BS MS it comes as the puʿʿal participle ‫‘ במעוטף‬in being covered’ (SirB 1v.3 = Sir. 11.4). 1.4.3. ‫‘ פחד‬fear’ In the MT, the dominant form is qal (22x), which is joined by a factitive hifʿil (Job 4.14) and a piʿʿel (Isa. 51.13; Prov. 28.14) limited to contexts of pluractionality/iterativity—note the use of the adverbials ‫ל־היֵ֗ ֹום‬ ַ ‫‘ ָת ִֶ֜מיד ָכ‬always, all day’ (Isa. 51.13) and ‫‘ ָת ִ ָׂ֑מיד‬always’ (Prov. 28.14). BS material twice exhibits similar pluractional/iterative examples in usages similar to Prov. 28.14 (SirB 7v.5 || SirD 1r.19 = Sir. 37.12). In the Masada MS, however, we confront the case of ] [ ‫טוב רע איש מטוב אשה‬ ‫חרפה‬ ̇ ‫‘ ובת מפחדת מכול‬It is better to harmed by a man than to be treated well by a woman, [ ] and a daughter who fears is better 276 The Historical Depth of the Tiberian Reading Tradition than any reproach’ (Mas1h 4.25 = Sir. 42.14). While the adjectival use is not dissimilar from the pluractional/iterative biblical use, the lack of an explicit adverbial signalling such is conspicuous (cf. the active participle with adverbial in Prov. 28.14). This is comparable to the less restrictive use of the piʿʿel in the DSS. 1.4.4. ‫‘ שנא‬hate’ Tiberian BH knows the piʿʿel stem for this verb, but only in the active participle form, where it has the nominal semantics of ‘enemy’. Like the NBDSS, which attest a puʿʿal participle (see above, §1.2.2), BS also knows a puʿʿal, but it is the prefix conjugation ‫‘ ישונא‬is [3MS] hated’ (SirA 3v.18 = Sir. 9.18). 1.5. Rabbinic Hebrew Fassberg (2001, 247–49) provides a brief, but illuminating discussion of pielisation in Tannaitic and Amoraic Hebrew, acknowledging various scholarly opinions on whether or not qal and piʿʿel forms are genuine synonyms or not (Yalon 1937; 1964; Ben-Ḥayyim 1958; Kutscher 1972). From Ben-Ḥayyim (1958, 236) he lists ‫‘ בזה‬despise’, ‫‘ דין‬judge’, ‫‘ זנה‬fornicate’, ‫‘ חסך‬spare’, ‫‘ יעץ‬advise’, ‫‘ מחה‬wipe out; try to prevent’, ‫‘ מתח‬stretch’, ‫עבר‬ ‘pass’, ‫‘ עקר‬uproot’, ‫‘ צוח‬cry out’, and ‫‘ רקם‬form’.16 He also cites studies by Ben-Ḥayyim (1958, 235–36), Kutscher (1969, 64–65), and Elitzur (1987, 84–87) on the relevance of qiṭṭūl-pattern verbal nouns, such as ‫‘ איבול‬mourning’, ‫‘ איסור‬prohibition’, ‫‘ בירורין‬ar16 Fassberg (2001, 247, fn. 25) also refers to Bendavid (1967–1971, I:376, II:482–83) for additional examples, though one must be cautious regarding the supposed semantic synonymy of some of the verbs listed. 12. Pielisation 277 bitration’, ‫‘ גידול‬growth’, ‫‘ גירומים‬extra, bonus’, ‫‘ חיבוט‬beating’, ‫‘ חיסום‬sharp edge’, ‫‘ ישוב‬settlement’, ‫‘ כיבושים‬admonition; conquest’, ‫‘ לימודים‬taught, disciple; teaching’, ‫‘ פיקודים‬charge, trust, account; (book of) Numbers’, ‫ציבורים‬/‫‘ ציבור‬community, public’, and ‫‘ שיפולי‬bottom of’. Illustrative examples from Tannaitic Hebrew include midrashic treatments of biblical passages in which RH piʿʿel verbs (a) replace qal verbs (b), e.g., (1a) ‫רבי עקיבא אומר מוכר הוא אם רצה ליעד מיעד‬ ‘Rabbi Akiba says, “the father sells her, and if the master wishes to designate (espouse) her, he may do so”’ (Mekhilta, Mishpatim, parasha 3, ed. Horowitz-Rabin 257.7); cf. (1b) ‫דּה וְּ ֶה ְּפ ָ ָׂ֑דּה ְּל ַ ֵ֥עם נָ ְּכ ִ ֛רי לֹא־‬ ֹׁ֖ ָ ָ‫) יְּׁ ע‬Q( ‫) לֵ֥ ֹו‬K( ‫֛יה ֲא ֶשר־לא‬ ָ ‫ם־ר ָ֞ ָעה ְּב ֵע ֵינָ֧י ֲאד ֶֹנ‬ ָ ‫ִא‬ ‫דֹו־בּה׃‬ ַֽ ָ ְּ‫יִ ְּמ ֵ֥ש ֹל ְּל ָמ ְּכ ָ ֶ֖רּה ְּב ִבג‬ ‘if she is displeasing in the eyes of her master who designated her for himself, then he shall let her be redeemed. He shall have no right to sell her to a foreign people, since he has broken faith with her’ (Exod. 21.8) (2a) ‫ וכן הוא‬...‫ומי שינה במי אם ישראל קלקלו במקום או המקום שינה בהם‬ ‫אומר כי אני לו שניתי‬ ‘And who changed his attitude toward whom? Did Israel rebuff God, or did God change his attitude toward Israel?... and thus he says, “I have not changed”’ (Sifre Devarim, Haʾazinu, pisqa ‫שו‬, ed. Finkelstein 330.16–17); cf. (2b) ‫ָ֑יתי‬ ִ ‫הוֶ֖ה ל ֹא שָ ִנ‬ ָ ְּ‫ִ ֛כי ֲא ִ ֵ֥ני י‬ ‘For I, the LORD, I have not changed’ (Mal. 3.6a) From Amoraic Hebrew, consider: 278 The Historical Depth of the Tiberian Reading Tradition (3a) ‫אין לי אלא בזמן שביזה דברי תורה‬ ‘I know that this applies only when he despised the teaching of the Law’ (y. Sanhedrin 27d 10.4); cf. (3b) ‫ִ ַ֤כי ְּד ַבר־יְּ הוָ ֙ה בָ ָזִ֔ה‬ ‘for the word of the LORD he despised’ (Num. 15.31a) (4a) ‫שימר יעקב אבינו את השבת‬ ‘Jacob, our father, kept the Sabbath”’ (Genesis Rabba 945.4); cf. (4b) ‫ת־ה ַש ָ ָׂ֑בת‬ ַ ‫וְּׁ שָ ְּׁמ ּ֥רּו ְּב ֵנַֽי־יִ ְּש ָר ֵ ֶ֖אל ֶא‬ ‘And the children of Israel will keep the Sabbath’ (Exod. 31.16a) To the verbs listed in Fassberg’s article, one may add the following from the discussions above: ‫‘ בכה‬weep; mourn’ (§§1.3.1; 1.3.3), ‫‘ לבש‬dress, wear’ (§1.5), ‫‘ מעט‬be few’ (§1.5), ‫סכך‬ ‘confine’ (§1.2.2), ‫‘ פרש‬clarify’ (§§1.2.2; 1.3.1). For the sake of precision, it is worth noting that contemporary with the process of pielisation seen in RH specifically and in Second Temple Hebrew more generally, RH saw the disappearance of the puʿʿal in all but adjectival (i.e., participial) forms (Breuer 2013, 737–38). In verbal usage, it was largely replaced especially by hitpaʿʿel/nitpaʿʿal. 2.0. The Tiberian Reading Tradition of Classical Biblical Hebrew Texts As a form of Hebrew rooted in the Iron Age but orally transmitted by later generations, it might be expected that the reading component of the Tiberian biblical tradition of early texts should exhibit a degree of drift from G- to D-stem where the consonantal 12. Pielisation 279 text was amenable to such. And, indeed, there is evidence of limited pielisation in CBH texts in line with that documented above from Second Temple sources, especially LBH consonantal evidence. 2.1. ‫‘ ֵמ ֵאן‬refuse’ Consider the verb ‫‘ ֵמ ֵאן‬refuse’. The verb comes 46 times in the Bible, where there is usually no reason to question its piʿʿel morphology, e.g., ‫אן‬ ֙ ֵ ‫‘ וַ יְּ ָמ‬but he refused’ (Gen. 37.35). On the five occasions when its participle occurs, however, the consonantal spelling conflicts with piʿʿel analysis. In four of the five, the for- mulation is ‫ם־מ ֵ ֵ֥אן ַא ָ ֶ֖תה‬ ָ ‫‘ וְּ ִא‬and if you (MS) refuse’ (Exod 7.27; 9.2; 10.4; Jer. 38.21), leading some to suggest that the expected prefix -‫ מ‬of the piʿʿel participle was elided between two other identical sounds (GKC §52s). Beyond the fact that just such a -‫ מ‬is preserved in the similar string ֒‫ם־מ ִמ ִתים ַא ֶת ֮ם א ִֹתי‬ ְּ ‫‘ ִא‬if you put me to death’ (Jer. 26.15), the form ‫[‘ ַ ַֽה ֵמ ֲא ִנים‬this people] who refuse’ (Jer. 13.10) cannot be so explained. Since it is not until RH that one finds unequivocal piʿʿel consonantal forms, e.g., ‫( ממאנים‬m. Yevamot 13.1, 1, 1, 4, 5; m. Ketubbot 11.6; m. ʿEduyot 6.1), it seems worth entertaining the possibility that the Tiberian realisation of this verb reflects some degree of mixture of First Temple qal stative and Second Temple piʿʿel morphology. It is also worth noting that the Aramaic translational equivalent ‫ סרב‬is commonly paʿʿel. While suffix and prefix conjugation forms such as ‫ ֵמ ֵאן‬and ‫ יְּ ָמ ֵאן‬would on this view represent secondary vocalisations, since the original qal form may well have had stative morphology, the 280 MS The Historical Depth of the Tiberian Reading Tradition participle and infinitive absolute form ‫( ָמ ֵאן‬Exod. 22.16) can be viewed as instances of preservation. In SH this particular verb reflects a shift qal > nifʿal in the prefix conjugation (see above, ch. 10, §1.3.2) and a qal > piʿel B shift in the suffix conjugation. In the latter, the Samaritan lack of a requirement for prefix -‫ מ‬on D-stem participles facilitated the reinterpretation of this and other apparent qal stative forms as piʿel B (e.g., ‫אהב‬, ‫שאל‬, §1.3.2). If a qal > piʿʿel shift did occur in the case of this verb in the Tiberian tradition, notwithstanding the fact that the earliest unambiguous piʿʿel evidence is from the Mishna, it is conceivable that it took place early in the Second Temple Period, i.e., during the LBH period, though this is conjectural, because the LBH texts present no participles of this verb. It is also possible that the shift began earlier than LBH. 2.2. ‫‘ גֵ ֵרש‬drive out/away, expel, divorce’ A similar example is ‫‘ *גֵ ֵרש‬drive out/away, expel, divorce’. All consonantal forms amenable to piʿʿel analysis in the MT—suffix conjugation, prefix conjugation, imperative, infinitives—are so read (≈35x), with passives in puʿʿal, but qal forms obtain in the case of participles, both active, ‫[‘ ג ֵֹרש‬Behold, I am about to] drive out’ (Exod. 34.11), and passive, ‫רּושה‬ ֵ֥ ָ ְּ‫‘ ג‬divorced (FS)’ (Lev. 21.7, 14; 22.13; Num. 30.10; Ezek. 44.22).17 Unambiguous consonantal evidence of D-stem ‫ גרש‬comes in the DSS and RH in piʿʿel ‫מגרש‬ 17 Likewise, in ‫‘ ְּל ַ ֵ֥מ ַען ִמגְּ ָר ָ ֶ֖שּה ָל ַ ַֽבז‬to cast it out for a prey’ (Ezek. 36.5 KJV) the apparent Aramaic-style infinitive was not amenable to piʿʿel realisation. Many, however, take ‫ ִמגְּ ָר ָ ֶ֖שּה‬here as a noun (cf. NIV, ESV, NET). 12. Pielisation 281 ‘man divorcing’ (CD 13.17; m. Yevamot 3.7; 4.8; etc.) and puʿʿal ‫‘ מגורשת‬woman being divorced’ (m. Giṭṭin 7.4, 5; etc.). RH also shows pielisation of this verb in the verbal noun ‫( גירושין‬m. Yevamot 3.8; t. Yevamot 13.5). Finally, the Aramaic equivalents for biblical ‫גרש‬, namely ‫תרד‬, ‫תרך‬, and ‫ שלח‬commonly occur in Dstem. Again, it would seem that a once unified qal paradigm was secondarily made suppletive under the influence of Second Temple morphological sensibilities, though a dearth of diagnostic forms in LBH makes it difficult to determine with precision when the shift began. 2.3. ‫‘ ִכ ֵבס‬wash, launder’ Consider also the verb ‫‘ ִכ ֵבס‬wash, launder’. Most active and pas- sive forms in the MT are piʿʿel and puʿʿal, respectively. The exception is the qal participle nomen agentis that occurs in the toponym ‫כֹובס‬ ַֽ ֵ ‫‘ ְּש ֵ ֵ֥דה‬Washer’s Field’ (2 Kgs 18.17 || Isa. 36.2; Isa 7.3). In this case, the earliest clearcut consonantal evidence for piʿʿel morphology is in post-exilic ‫‘ ְּמ ַכ ְּב ִ ַֽסים‬launderers’ (Mal. 3.2), which becomes common in RH as the verbal participle alongside nominal qal ‫ ;כובס‬cf. puʿʿal ‫( מכובסין‬m. Miqwaʾot 10.4). Note also the postbiblical Hebrew knows two verbal nouns, i.e., ‫( כיבוס‬CD 11.22; 4Q271 f5i.15; m. Zevaḥim 7.1; etc.) and ‫( כביסה‬m. Miqwaʾot 8.1; t. Bava Metsiaʿ 11.2), with respective patterns typical or piʿʿel and qal. The Aramaic equivalents ‫ חור‬and ‫ צבע‬are also D-stem. It may well be that a significant number of biblical forms prior to LBH were originally qal, but were secondarily read as piʿʿel where possible, in line with Second Temple convention. 282 The Historical Depth of the Tiberian Reading Tradition 3.0. The Tiberian Classical Biblical Hebrew Written Tradition In the preceding section, the emphasis was on apparently late instances of pielisation within the Tiberian reading tradition. While it may be that the qal > piʿʿel shifts discussed began prior to Second Temple times, the evidence of unambiguous consonantal piʿʿel and puʿʿal forms seems indicative of a late development in line with post-biblical conventions. For a proper characterisation of pielisation within the reading component of the Tiberian tradition, however, one must take into consideration relevant developments rooted in the written component of the tradition as found in CBH texts. From the perspective of these, it becomes clear that the drift from qal to piʿʿel seen above is, rather than a complete innovation, the continuation of an ancient process. First of all, while Iron Age Hebrew and cognate epigraphy lack piʿʿel participles and puʿʿal forms in general, there is ample unambiguous biblical consonantal evidence of the use of D-stem morphology in the orthographic tradition of CBH texts. Among verbs with unequivocal classical piʿʿel/puʿʿal attestation, some have no cognates in other stems, e.g., ‫‘ ִב ֵקש‬seek, request’ (63x); others exhibit well-established semantic specialisation of the piʿʿel form vis-à-vis the relevant qal, e.g., pluractional ‫*ק ֵבר‬ ִ ‘bury en masse’ (Num. 33.4; 1 Kgs 11.15; Jer. 14.16; Ezek. 39.14–15; 12. Pielisation 283 Hos. 9.6) versus qal ‫‘ ָק ַבר‬bury’/nifʿal ‫‘ *נִ ְּק ַבר‬be buried’.18 Clearly, D-stem morphology was an early option in ancient Hebrew. Second, even when it comes to the drift from qal to piʿʿel— which, it was argued above (§2.0), resulted in the partial replacement of original G-stem morphology with D-stem morphology in line with Second Temple Hebrew trends—not all of the evidence is late. Rather, certain cases of early, well-stablished qal-piʿʿel suppletion responsible for apparently synonymous G- and D-stem forms seem to indicate the reality of early pielisation. It is to examples of this latter category that the discussion now turns. The early evidence of pielisation that they furnish shows that later results of pielisation, though secondary, were very much in a line of linguistic evolution long since initiated. 3.1. ‫‘ דבר‬speak’ Extremely common in BH, ‫ ִד ֶבר‬occurs in piʿʿel in all forms, making it clear that its D-stem morphology—which continues into Second Temple traditions—is of ancient pedigree. Puʿʿal forms, including a participle, also occur (Ps 87.3; Song 8.8). Alongside these, however, there occur vestigial qal forms: active participle ‫‘ ד ֵֹבר‬speaker, speaking’ (39x), passive participle ‫‘ ָד ֻ ֵ֥בר‬spoken’ 18 Given the proposed morphosemantic distinction, the form ‫‘ ֻק ַ ֵ֥בר‬was (were) buried’ (Gen. 25.10) is to be analysed as a qal passive. Despite the reference to two corpses, the event here arguably involves Abraham’s burial, Sarah having previously been buried (qal) in Gen. 23. In the absence of consonantally unambiguous biblical evidence for nifʿal ‫‘ נִ ְּק ַבר‬be buried’—for which all representative forms are in the prefix conjugation—it is possible that many, if not all, of the apparent nifʿal forms conceal original qal internal passives (see above, ch. 10, §2.2). 284 The Historical Depth of the Tiberian Reading Tradition (Prov. 25.11), and infinitive construct ‫‘ ְּב ָד ְּב ֶ ֵ֗רָך‬when you speak’ (Ps. 51.6). Since there is no obvious semantic distinction between ostensible qal ‫*ד ַבר‬ ָ and piʿʿel ‫ד ֶבר‬,ִ the particular instance of pielisation in question may well have been driven by broader cognitive processes to do with morphosemantics, such as the D-stem’s perceived active iconicity vis-à-vis qal’s perceived opacity. Whatever the case may be, given the widespread nature of unambiguous piʿʿel morphology in CBH orthography, it this verb underwent a process of pielisation, it must have occurred relatively early on in the history of CBH. Even so, in light of the fact that the qal participle ‫( ד ֵֹבר‬39x) is as common in the MT as the piʿʿel participle ‫( ְּמ ַד ֵבר‬39x), while piʿʿel use persists and qal use decreases in postexilic Hebrew (see Table 1), the Tiberian reading tradition’s wholesale pielisation of prefix and suffix conjugation may arguably be characterised as anachronistic, influenced by Second Temple linguistic trends (note that the Aramaic equivalent ‫ מלל‬is also D-stem). 12. Pielisation 285 Table 1: Frequency of qal and piʿʿel participles of ‫‘ דב"ר‬speak’ in select ancient Hebrew corpora19 qal ‫ד(ו)בר‬ piʿʿel ‫מדבר‬ MT 39 39 MT LBH 2 8 NBDSS 2 6 Ben Sira 2 1 Mishna 1 23 SP 0 18 3.2. ‫‘ ברך‬bless’ Like the verb ‫ד ֶבר‬,ִ so too Tiberian ‫‘ ֶב ַרְך‬bless’ appears at some point rather early on in its history to have undergone secondary pielisation, which eventually produced a predominantly D-stem paradigm with significant G-stem residue. Classical orthographic evidence of pielisation is seen in participles in piʿʿel (Gen. 12.3; 27.29; Num. 24.9; Isa. 66.3; Prov. 27.14) and puʿʿal (Num. 22.6; Deut. 33.13; Ps. 37.22; 113.2; Job 1.21; 1 Chron. 17.27), as well as in hitpaʿʿel forms (Gen. 22.18; 26.4; Deut. 29.18; Isa. 65.16; Jer. 4.2; Ps. 72.17). Evidence of G-stem morphology comes primarily in the form of the qal passive participle ‫‘ ָברּוְך‬blessed’ (71x) and in nifʿal forms (Gen. 12.3; 18.18; 28.14). The dominance of the qal passive participle over the puʿʿal participle may be con19 As far as can be determined given the extant data, the MT and BDSS agree on the distribution and frequency of qal and piʿʿel participles of ‫‘ דב"ר‬speak’. Regarding the SP—while there is no difference between the Samaritan and Tiberian orthographic traditions when it comes to participles of the verb in question, all Samaritan forms, whether with or without a prefix -‫מ‬, are analysed as D-stem (see above, §1.3.1). 286 The Historical Depth of the Tiberian Reading Tradition strued as evidence of an early qal verb ‘bless’,20 while the absence of any qal active participle arguably signifies very early pielisation of this verb. From this perspective, the piʿʿel dominance outside of the passive participle seems less artificial and anachronistic than does piʿʿel dominance in the case of ‫ ִד ֶבר‬above. Another difference between this case and that of ‫ד ֶבר‬,ִ discussed in the foregoing section, is that the specific arrangement of a prevailing piʿʿel paradigm with qal dominance restricted to the passive participle ‫ ָברּוְך‬vis-à-vis ‫ ְּמב ַֹרְך‬persists in post-exilic Hebrew, where the puʿʿal participle never gains ascendency. This, however, is possibly due at least in part to the conservative contexts in which the forms are used, e.g., blessings, prayers, and other forms of liturgy. Regardless, if the verb in question was subject to pielisation, it is clear that the CBH written tradition reflects a time when the process was well advanced. 3.3. ‫‘ גלה‬uncover, reveal’ In the meaning ‘uncover, reveal’, the D-stem enjoys overall numerical superiority in the Tiberian tradition (piʿʿel 56x, puʿʿal 2x),21 as well as in Second Temple extra-biblical sources (see 20 Assuming the early existence of qal ‫‘ ברך‬bless’, its shift to piʿʿel may have resulted from a perceived need to distinguish it from qal ‫‘ ָב ַרְך‬kneel’ (Ps. 95.6; 2 Chron. 6.13; related hifʿil ‫ ִה ְּב ִריְך‬at Gen. 24.11). 21 Piʿʿel: Lev. 18.6, 7, 7, 8, 9, 10, 11, 12, 13, 14, 15, 15, 16, 17, 17, 18, 19; 20.11, 17, 18, 18, 19, 20, 21; Num. 22.31; Deut. 23.1; 27.20; Isa. 16.3; 22.8; 26.21; 47.2, 2; 57.8; Jer. 11.20; 20.12; 33.6; 49.10; Ezek. 16.37; 22.10; 23.10, 18, 18; Hos. 2.12; Mic. 1.6; Nah. 3.5; Ps. 98.2; 119.18; Job 12.22; 20.27; 41.5; Prov. 11.13; 25.9; Ruth 3.4, 7; Lam. 2.14; 4.22. Puʿʿal: Nah. 2.8; Prov. 27.5. 12. Pielisation 287 above, §2.0).22 Yet, the evidence of G-stem morphology is not rare, especially in qal expressions with the nouns ‫‘ אֹזֶ ן‬ear’, ‫ַעיִ ן‬ ‘eye’, and ‫‘ סֹוד‬secret’ (21x).23 Passive nifʿal forms, also presupposing an active qal form, are even more numerous (32x).24 Thus, in the sense ‘uncover, reveal’, D-stem active and passive morphology (58x) is just slightly more common the G-/N-stem morphology (53x). What is more, while unequivocal G-stem morphology is unambiguously evidenced—in forms such as the active participle ‫ת־אזְּ נִ ֙י‬ ָ ‫‘ וְּ ֵאין־ג ֶֹלַ֤ה ֶא‬there was no one to tell me’ (1 Sam. 22.8, 8) and the passive participle ‫( גָ לּוי‬Num. 24.4, 16)—are relatively common throughout the Bible, unambiguous orthographic evidence for D-stem active and passive is rather restricted (Job 12.22; Prov. 11.13; 27.5). Notwithstanding the rather narrow list of expressions employing qal ‫גל"י‬, the rather broader use of the nifʿal arouses the suspicion that certain orthographically ambiguous forms vocalised as piʿʿel might originally have been read as qal. Mismatches occur in the case of the nouns ‫( ַעיִ ן‬qal 2x; nifʿal 3x; piʿʿel 2x), ‫‘ סֹוד‬secret’ (qal 2x; piʿʿel 2x), ‫‘ ֵע ְּרוָ ה‬nakedness’ (piʿʿel 24x; nifʿal 4x). While there is no reason to doubt the original authenticity of some or even many D-stem cases, there are grounds 22 I am grateful to my Middlebury student, Rachel Kaufman, for her question on the mixed stem morphology of ‫גל"י‬. 23 Qal: Num. 24.4, 16; 1 Sam. 9.15; 20.2, 12–13; 22.8, 8, 17; 2 Sam. 7.27; Jer. 32.11, 14; Amos 3.7; Job 33.16; 36.10, 15; Prov. 20.19; Ruth 4.4; Est. 3.14; 8.13; 1 Chron. 17.25. 24 Nifʿal: Gen. 35.7; Exod. 20.26; Deut. 29.28; 1 Sam. 2.27, 27; 3.7, 21; 14.8, 11; 2 Sam 6.20, 20, 20; 22.16; Isa. 22.14; 23.1; 38.12 (?); 40.5; 47.3; 49.9; 53.1; 56.1; Jer. 13.22; Ezek. 13.14; 16.36, 57; 21.29; 23.29; Hos. 7.1; Ps. 18.16; Job 38.17; Prov. 26.26; Dan. 10.1. 288 The Historical Depth of the Tiberian Reading Tradition for suspecting a degree of movement form qal to piʿʿel in the case of this verb, a process from which qal active and passive participles were exempted due to their orthographic intransigence. 4.0. Conclusion Based on the foregoing survey of examples of pielisation in ancient Hebrew, the following summary may be sketched. First, the shift qal > piʿʿel is unambiguously documented throughout the history of Hebrew, from CBH texts associated with the Iron Age on. Second, when compared to Tiberian CBH, Second Temple Hebrew—represented by Tiberian LBH, the DSS, the Samaritan biblical reading tradition, BS, and RH—exhibits a comparatively advanced stage in the pielisation process. However, a distinction should be drawn between the orthographic component of Tiberian CBH and the corresponding reading component, the latter sometimes showing evidence of secondary variance from the former in favour of piʿʿel morphology in line with late linguistic conventions. Though such secondary dissonance between the written and recitation components of the Tiberian biblical tradition inevitably entails the positing of a mixed tradition characterised by a degree of linguistic anachronism, the pielisation of a specific verb may not represent deviation from the normal path of BH grammatical development, but a typologically more advanced stage on a shared path. 13. HITPAELISATION Along with the fairly common processes of nifalisation (ch. 10), hifilisation (ch. 11), and pielisation (ch. 12)—all generally involving movement away from the qal verbal stem—hitpaelisation is also a known phenomenon. It differs, however, from the three aforementioned processes, in that it rather rarely manifests in the hitpaʿʿel revocalisation of qal orthographic forms. This must be due, at least in part, to the consonantal difference between qal and hitpaʿʿel, i.e., only with difficulty would original qal orthog- raphy lend itself to hitpaʿʿel realisation. More frequently, hitpaʿʿel/ nitpaʿʿal replaces passive or reflexive nifʿal or passive puʿʿal, especially in the case of finite forms. Other stems are also occasionally affected. In these cases, too, revocalisation often required special measures, especially the assimilation of hitpaʿʿel/ nitpaʿʿal’s characteristic t-infix. 1.0. Second Temple Evidence 1.1. Tiberian Late Biblical Hebrew 1.1.1. Movement to Hitpaʿʿel Broad indication of the diachronic significance of hitpaelisation may be gleaned from Baden’s (2010, 39, fn. 18) acceptance of conclusion reached by Bean (1976, 149–53), namely, that the later books of the Hebrew Bible witness increased hitpaʿʿel usage in comparison to earlier books. But Bean’s statistics must be considered no more than impressionistic, because his methodology © 2023 Aaron D. Hornkohl, CC BY-NC 4.0 https://doi.org/10.11647/OBP.0310.13 290 The Historical Depth of the Tiberian Reading Tradition has no means of ruling out the possibility that differences in content are responsible for the apparent increase, i.e., that further hitpaʿʿels might possibly have been used in CBH given the same subject matter. What is needed is a more detailed, word-by-word study that applies Hurvitz’s (2014, 9–11) three-pronged strategy for identification of diagnostically late linguistic features, namely, (1) late distribution, (2) classical opposition, and (3) extrabiblical confirmation. Such an approach is applied to a series of Tiberian LBH hitpaʿʿel forms below. ‫‘ ִה ְּת ָב ֵאש‬stink, be odious’ The root is represented by qal (Exod. 7.18, 21; 8.10; 16.20; Isa. 50.2) and nifʿal (1 Sam. 13.4; 2 Sam. 10.6; 16.21) forms. The qal consistently refer to physical smells, the nifʿal to the metaphorical ‘you have become a stench’. The hitpaʿʿel comes just once in Tiberian Hebrew, in the late parallel to the nifʿal in 2 Sam. 10.6 found in 1 Chron. 19.6; see examples (1)–(2). (1) ...‫אּו ְּב ֵני ַע ֵ֔מֹון ִ ֵ֥כי נִ ְּׁבאֲשֹׁ֖ ּו ְּב ָדִוָׂ֑ד‬ ֙ ‫וַ יִ ְּר‬ ‘When the Ammonites saw that they had become a stench to David…’ (2 Sam. 10.6) (2) ...‫ם־דִוָׂ֑יד‬ ָ ‫אּו ְּב ֵני ַע ֵ֔מֹון ִ ֵ֥כי ִ ִֽה ְּׁתבָ אֲשֹׁ֖ ּו ִע‬ ֙ ‫וַ יִ ְּר‬ ‘When the Ammonites saw that they had become a stench to David…’ (1 Chron. 19.6) While hitpaʿʿel ‫ התבאש‬is not again documented in Hebrew sources until piyyuṭ, the Targumic equivalent of both N-stem ‫ נבאש‬and Dt-stem ‫ התבאש‬in Tiberian BH is Aramaic Dt-stem ‫אתגרי‬. 13. Hitpaelisation 291 ‫‘ ִה ְּתגָ ֵאל‬defile’ All derivations of the root ‫ גאל‬II ‘defile’ are late, including piʿʿel (Mal. 1.7), puʿʿal (Mal 1.7, 12; Ezra 2.62; Neh. 7.64), nifʿal (Isa. 59.3; Zeph. 3.1; Lam. 4.14), and hitpaʿʿel (Dan. 1.8, 8). The hitpaʿʿel ‘become defiled’ is also known from NBDSS texts (1QM 9.8; 4Q379 f3i.5). The classical equivalents are derivations of ‫גע"ל‬ ‘abhor’, for the hitpaʿʿel of ‫ גא"ל‬II evidently nifʿal ‫‘ נִ גְּ ַעל‬be defiled’ (2 Sam. 1.21). ‫‘ ִה ְּתגַ ֵדל‬magnify oneself’ In classical texts, the hifʿil expression ‫ ִהגְּ ִדיל ַעל‬is sometimes used in the antagonistic sense ‘to raise oneself against’ (Ezek. 35.13; Zeph. 2.8, 10; Ps. 35.26; 41.10). Twice in LBH, the phrase with hitpaʿʿel ‫ ִה ְּתגַ ֵדל ַעל‬comes in the same meaning (Dan. 11.36–371). Cf. also RH: ‫‘ ר' ָצדֹוק אֹו' ַאל ַת ֲע ֵשם ֲע ָט ָרה ְּל ִה ְּתגַ ֵדל ָב ֶהן‬R. Sadoq says, “Do not make [Torah teachings] a crown with which to glorify yourself…’ (m. ʾAvot 4.5). Interestingly, the Targumic equivalent of C-stem ‫ ִהגְּ ִדיל ַעל‬is t-stem ‫ ;אתררב‬Syriac ‫ ܐܬܬܪܝܡ‬is also t-stem. ‫‘ ִה ְּת ַח ֵבר‬join, associate’ The qal has the basic sense of ‘join, associate’, and can refer to people (Gen. 14.3; Hos. 4.17; Ps. 94.20) or objects (Exod. 26.3, 3; 28.7; 39.4; Deut. 18.11 [?]; Ezek. 1.9, 11; Ps 58.6 [?]; 94.20). The hitpaʿʿel refers only to human alliances (Dan. 11.6, 23; 2 Chron. 20.35, 37). The hitpaʿʿel also occurs in reference to human 1 Possibly also in Isa. 10.15, but the context does not involve a ruler raising himself up. 292 The Historical Depth of the Tiberian Reading Tradition association in BS (SirA 5r.23–25 || Sir. 13.2), NBDSS material (4Q374 f1a–b.3; 4Q416 f2iii.21); and RH (m. ʾAvot 1.7). The Mishna also includes an example of non-human association (m. Ṭohorot 9.1). In reference to human association, the Targums also utilise Dt-stem forms, e.g., ‫ אתכנש‬and ‫ ;אתחבר‬so, too, occasionally the Peshiṭta. ‫‘ ִה ְּתנַ ֵדב‬freely offer (cultic)’ In the cultic sense of ‘freely offer’, the relevant CBH usages involve transitive qal with ‫רּוח‬ ַ ‘spirit’ or ‫‘ ֵלב‬heart’ as subject, e.g., ‫יש ֲא ֶשר יִ ְּד ֶבנּו ִל ֵ֔בֹו‬ ֙ ‫ל־א‬ ִ ‫‘ ָכ‬every man whose heart moves him’ (Exod. 25.2), ‫רּוחֹו א ֵֹ֗תֹו‬ ֶ֜ ‫‘ וְּ ָ֡כֹל ֲא ֶשר֩ נָ ְּד ָָֹ֨בה‬and every one whose spirit moved him’ (Exod. 35.21), ‫ם‬ ֒ ‫ל־איש וְּ ִא ֵָ֗שה ֲא ֶָֹ֨שר נָ ַדב ִל ָב ֮ם א ָֹת‬ ִ ‫‘ ָכ‬every man or woman whose heart moved them’ (Exod. 35.29). In LBH, the early transitive qal expression gives way to an apparently reflexive hitpaʿʿel involving the freewill offering of sacrifices or service (Ezra 1.6; 2.68; 3.5; Neh. 11.2; 1 Chron. 29.5–6, 9, 14, 17; 2 Chron. 17.16).2 The Dt-stem form is also common in Qumran writings (1QS 5.1, 6, 8, 10, 21–22; 6.13; 1Q14 f8–10.7; 1Q31 f1.1; 4Q256 9.1, 5; 4Q258 1.1, 5; 2.1–2; 4Q368 f10i.6; 4Q433a f2.5) and in RH (m. Sheqalim 4.1; 5.6; m. Zevaḥim 10.8, 8; m. Menaḥot 12.3, 4, 4, 4, 5, 5, 5, 5, 5; 13.3; m. Keritot 6.3, 3; m. Meʿila 3.6, 6; m. Middot 3.8). The Targums also frequently resort 2 Hitpaʿʿel forms also occur in Judg. 5.2, 9, but these are in a military, rather than cultic context. In other words, the late aspect of ‫ ִה ְּתנַ ֵדב‬is not merely its Dt-stem morphology, but its cultic semantics and use in place of qal ‫נָ ַדב‬. 13. Hitpaelisation 293 to Dt-stem forms, whether of ‫ נד"ב‬or ‫( רע"י‬the latter even in the case of two of the CBH qal usages). ‫‘ ִה ְּת ַענָ ה‬fast’ Classical cases of ‫ ִה ְּת ַענָ ה‬have the general sense of ‘humble oneself, afflict oneself, suffer affliction’ (Gen. 16.9; 1 Kgs 2.26; Ps. 107.17). It is possible that in LBH the sense narrows to ‘fast’ (Dan. 10.12; Ezra 8.21), in line with post-biblical sources (DSSH, RH; see BDB 726b; Qimron 1980, 250; Hurvitz 2014, 242). Clearly, only in the specific meaning ‘fast’ can ‫ ִה ְּת ַענָ ה‬be considered especially characteristic of post-exilic Hebrew. ‫‘ ִה ְּת ָפ ֵעם‬disturb’ Nifʿal forms in the sense ‘be disturbed’ occur in CBH (Gen. 41.8), LBH (Dan. 2.3), and poetic material of less certain diachronic linguistic profile (Ps. 77.5). Nifʿal forms are also preserved in the BDSS (4Q3 f1ii.15 || MT Gen. 41.8) and in SH (Gen. 41.8). Against the nifʿal ‫רּוחי‬ ִֵ֔ ‫‘ וַ ִת ָפ ֶעם‬and my spirit was troubled’ (Dan. 2.3), one nearby encounters hitpaʿʿel ‫רּוחֹו‬ ֵ֔ ‫‘ וַ ִת ְּת ָפ ֶעם‬and his spirit was troubled’ (Dan. 2.1). While further Hebrew examples of ‫ ִה ְּת ָפ ֵעם‬go undocumented until the time of piyyuṭ, making them non-diagnostic as far as ancient periodisation goes, TA and, to a lesser extent, Syriac resort to t-stem forms in their renderings of both Tiberian ‫ נִ ְּפ ַעם‬and ‫ה ְּת ָפ ֵעם‬. ִ ‫‘ ִה ְּש ַת ֵכ ַח‬forget’ Throughout the Tiberian biblical tradition, the standard passive of ‫‘ ָש ַכח‬forget’ is nifʿal ‫‘ נִ ְּש ַכח‬be forgotten’ (Gen. 41.30; Deut. 294 The Historical Depth of the Tiberian Reading Tradition 31.21; Isa. 23.15–16; 65.16; Jer. 20.11; 23.40; 50.5; Ps. 9.19; 31.13; Job 28.4; Qoh. 2.16; 9.5). Qohelet, widely considered late on the basis of its post-exilic linguistic profile (Schoors 1992– 2004; Hurvitz 2007; see Hornkohl 2013b, 321, for further bibliography), includes two of the classical nifʿal cases, but also the only Tiberian biblical example of hitpaʿʿel (Qoh. 8.10), apparently with the same meaning as its more common nifʿal counterpart. The hitpaʿʿel also appears in Tannaitic sources (Mekhilta deRabbi Ishmaʿel; Sifre Devarim; Tosefta) and Amoraic Hebrew (Yerushal- mi; Bavli). Finally, the Aramaic and Syriac equivalents to both Tiberian nifʿal ‫ נִ ְּש ַכח‬and hitpaʿʿel ‫ ִה ְּש ַת ֵכ ַח‬are commonly t-stem verbs. 1.1.2. Hippaʿʿel < Hitpaʿʿel On relatively rare occasions, Tiberian Hebrew evinces forms of the type hippaʿʿel < hitpaʿʿel. In these cases, suffix conjugation forms in texts from no earlier than the Exile can be read only as hitpaʿʿel forms with assimilated tav: ‫‘ ִהנַ ְבאּו‬they prophesied’ (Jer. 23.13); ‫‘ וְ ִהנֶ ָ ָ֑ח ְמ ִתי‬and I will be satisfied’ (Ezek. 5.13); ‫אתי‬ ִ ‫‘ וְ ִהנַ ֵ ִ֖ב‬and I prophesied’ (Ezek. 37.10). These unambiguous consonantal hitpaʿʿel forms with assimilated tav are not especially important in their own right, as the hitpaʿʿel forms of both ‫ נב"א‬and ‫ נח"ם‬are well attested throughout the Tiberian biblical corpus, from CBH to LBH. Their significance in the context of the phenomenon of hitpaelisation is as evidence of the door opened via assimilation of the infix tav for the apparent secondary development in the Tiberian reading tradition of consonantal nifʿal forms into nip- 13. Hitpaelisation 295 paʿʿel [< nitpaʿʿel] forms (see §2.0 below; the development is especially characteristic of SH, §1.3). 1.2. Dead Sea Scrolls Hebrew There is limited evidence of hitpaelisation in the Hebrew of the Dead Sea Scrolls, in both biblical and non-biblical material. 1.2.1. The Biblical Dead Sea Scrolls In the BDSS, a possible manifestation of hitpaelisation involves textual variation in which various MT forms are paralleled in Qumran texts by synonymous t-stem alternatives. Consider the following cases: (3) ‫‘ ̇מ ֯ת[נבא‬prophesying’ (4Q51 9e–i.13) || MT ‫( נִ ָ ָׂ֑בא‬1 Sam. 10.11) (4) ‫‘ יתי̇ סד‬its foundation will be laid’ (1QIsaa 38.6) || MT ‫ִתּוָ ֵ ַֽסד‬ ‘your foundation will be laid’ (Isa. 44.28) (5) ‫(‘ תתמוטינה‬and the hills) will shake’ (1QIsaa 45.8) || MT ‫מּוטנָ ה‬ ָׂ֑ ֶ ‫( ְּת‬Isa. 54.10)3 (6) ‫‘ ויתגרשו‬and (its waters) are tossed up (with mire and dirt)’ (1QIsaa 47.20) || MT ‫‘ וַ יִ גְּ ְּר ֵ֥שּו‬and (its waters) tossed up (mire and dirt)’ (Isa. 57.20) (7) ‫(‘ תתנחמו‬and in Jerusalem) you will be comforted’ (1QIsaa 53.29) || MT ‫(‘ ְּתנֻ ָ ַֽחמּו‬and in Jerusalem) you will be comforted’ (Isa. 66.13) 3 Cf. ‫(‘ התמוטטה‬the earth) shook’ (1QIsaa 19.18) || MT ‫מֹוט ָ ֶ֖טה‬ ְּ ‫( ִ ַֽה ְּת‬Isa. 24.19). 296 The Historical Depth of the Tiberian Reading Tradition The frequency of hitpaelisation in 1QIsaa in particular—representing shifts from qal (5)–(6), nifʿal (4), and puʿʿal (7)—seems to have diachronic significance. Despite its basis in CBH, 1QIsaa is renowned for its degree of linguistic contemporisation (Kutscher 1974, 77–89; Abegg 2010, 25; Fassberg 2013; Muraoka 2013; cf. Young 2013). The t-stem forms ‫( נב"א‬3) and ‫( נח"ם‬7) are known from CBH, and that of ‫( מו"ט‬5) occurs elsewhere in MT Isaiah and 1QIsaa (see fn. 4), so that it might stem more from stylistic harmonisation than linguistic convention, but the t-stem form of ‫יס"ד‬ (4) is unknown in Tiberian BH, being unique in Hebrew until it resurfaces in early medieval poetry, and the earliest documentation of t-stem ‫ גר"ש‬comes in the BDSS (6) and NBDSS (1QHa 10.14; 11.16–17), it next appearing in the meaning ‘be divorced’ in RH (m. Yevamot 14.1; m. Nedarim 9.9; m. Giṭṭin 6.2) and in Amoraic sources (Yerushalmi; Bavli). The evidence as such does not confirm the late character of hitpaelisation in the BDSS, but it is in line with such a theory.4 1.2.2. The Non-biblical Dead Sea Scrolls It has already been noted that the NBDSS exhibit diagnostically late hitpaʿʿel forms known from Tiberian LBH—‫‘ התגאל‬be defiled’ (2x(, ‫‘ התחבר‬join, associate’ (2x; also in BS, RH), and ‫‘ התנדב‬freely 4 There are also a few cases of apparent BDSS shifts away from hitpaʿʿel in comparison to the MT. Thus, ‫והנחלתם‬ ̇ ‘and you will bequeath’ (4Q24 f27–28.2) || MT ‫( וְּ ִה ְּתנַ ֲח ְּל ֶָֹ֨תם‬Lev. 25.46); ‫‘ הרגזכה‬your raging’ (1QIsaa 31.7) || MT ֵ֥‫( ִ ַֽה ְּת ַרגֶ זְּ ָך‬Isa. 37.28); ‫‘ יכסו‬they will (not) cover’ (1QIsaa 48.17) || MT ‫‘ יִ ְּת ַכ ֶ֖סּו‬they will not cover themselves’ (Isa. 59.6). Rather than re- flecting a broad shift away from hitpaʿʿel, these cases seem to stem from local exegetical differences and/or difficulties. 13. Hitpaelisation 297 offer (cultic)’ (17x; also in RH)—and from the BDSS—‫התגרש‬ ‘storm, be tossed up (waves)’. But this does not remotely reflect the degree of hitpaelisation encountered in the NBDSS. Indeed, many hitpaʿʿel forms unknown from BH are documented in the NBDSS, sometimes also appearing other Second Temple Hebrew material. Here they are listed in order of frequency in the NBDSS with notation of additional corpora in which they occur, if relevant: ‫‘ ה(ת)דשן‬become fat, savour’ (10x; BS); ‫‘ התיסר‬be chastised’ (8x; RH2); ‫‘ הטמא‬become defiled, unclean’ (6x; BDSS, SH, RH, Tiberian reading tradition [see below, §2.0]); ‫‘ התאחר‬be delayed’ (6x; BS (cf. below, ‫‘ השתלם ;)התקדם‬be rewarded’ (5x; RH); ‫התיחד‬ ‘unite (intr.)’ (4x; RH); ‫‘ הזכה‬be cleansed, considered innocent’ (?; 4x; BDSS, Tiberian reading tradition [see below, §2.0]; Amoraic Hebrew; ‫‘ התרגש‬storm, be tossed up (waves)’ (4x; Amoraic Hebrew); ‫‘ השתלח‬be sent’ (3x; RH); ‫‘ התבהל‬be eager, pass quickly’ (3x); ‫‘ התפזר‬be scattered’ (3x; RH); ‫‘ התרמה‬be cheated’ (3x); ‫התקדם‬ ‘go/be early’ (2x; cf. above, ‫‘ התבעה ;)התאחר‬inquire (of prophetic dreams)’ (?) (2x); ‫ התענה‬if in the meaning ‘fast’ (2x; LBH, BDSS, RH); ‫‘ התקלה‬be put to shame’ (2x); ‫‘ התארמל‬become a widow’ (RH); ‫‘ התפתה‬be fooled, deceived’ (BS); ‫‘ התקרע‬be torn asunder’ (RH); ‫‘ התרשע‬condemn oneself, be condemned’; ‫‘ התפרר‬break (intr.), be shattered’; ‫‘ התאמן‬trust’ (?); ‫‘ התאנח‬sigh, groan’ (BS, Amoraic Hebrew); ‫‘ התישר‬be right’ (?); ‫‘ התכבס‬be washed’ (RH); ‫‘ התמלא‬be filled’ (RH—different semantics in MT Job 16.10); ‫‘ התנסה‬be tested’ (?) (RH); ‫‘ התעכל‬be consumed’ (?); ‫‘ התעצל‬hesitate, be sluggish’ (RH); ‫‘ התפחד‬fear, tremble’; ‫‘ הצטרף‬be refined’ 298 The Historical Depth of the Tiberian Reading Tradition (RH). Consider the following NBDSS example with ‫‘ התקדם‬be/go early’ and ‫‘ התאחר‬be/go late’. (8) ‫דה כולה‬ ̇ ̇‫העבו‬ ̇ ‫יתקדם או יתאחר ולא ישביתו את‬ ‘let him go earlier or later so that they need not stop the whole service…’ (CD 11.23) (9) ‫י־הּוא׃‬ ַֽ ‫ל־ה ָש ַמיִ ם ִל‬ ַ ‫ִמי ִה ְִּׁ֭ק ִדימַּ נִ י וַ ֲא ַש ֵלָׂ֑ם ַ ֶ֖ת ַחת ָכ‬ ‘Who has preceded me, that I should repay him? Whatever is under the whole heaven is mine.’ (Job 41.11) (10) ‫ם־ל ָבן‬ ָ ‫ֹאמ ֵ֔רּון ַ ַֽלאד ִֹנֶ֖י ְּל ֵע ָ ָׂ֑שו ַ֤כֹה ָא ַמ ֙ר ַע ְּב ְּדָך יַ ֲע ֵ֔קֹב ִע‬ ְּ ‫מר כֹה ת‬ ֹ ֵ֔ ‫וַ יְּ ַצַ֤ו א ָֹת ֙ם ֵלא‬ ‫ד־ע ָתה׃‬ ַֽ ָ ‫ֵ֔ ַג ְּר ִתי וָאֵ ַּחֹׁ֖ר ַע‬ ‘And he commanded them, “Thus you shall say to my lord Esau: Thus says your servant Jacob, ‘I have sojourned with Laban and delayed until now.’” The Tiberian hifʿil and qal forms are matched by DSS hitpaʿʿel forms in approximately the same meanings. Consider also the case of ‫‘ הצטרף‬be refined’. Parallel to Ti- berian nifʿal ‫פּו‬ ֙ ‫‘ וְּ יִ ָ ַֽצ ְּר‬and will be refined’ (Dan. 12.10), 4Q Eschatological Commentary A presents hitpaʿʿel ‫‘ ויצטרפו‬and they will be refined’ (4Q174 f1–3ii.4). Though in RH the hitpaʿʿel generally has the sense ‘join’, the meaning ‘be refined’ also occasionally surfaces, e.g., ‫ישיִ ֳצט ְּרפּו בכבשן‬ ֶ ‫‘ ִמ‬after they are fired in a furnace’ (m. Kelim 4.4–5);5 ‫‘ מישיצטרפו בכבשן‬after they are fired in a furnace’ (t. Kelim Bava Batra). 5 In Codex Kaufmann, an interlinear ṭet has been placed above the ap- parently nifʿal form ‫ישיִ ֳצט ְּרפּו‬ ֶ ‫ ִמ‬between the tsade and the resh (Beer 1968, 447b). The vocalisation also corresponds to that of the hitpaʿʿel rather than a nifʿal—what appears to be a ḥaṭef qameṣ below the tsade is in reality a shewa beneath the ṣade and a qameṣ below the supralinear ṭet. 13. Hitpaelisation 299 1.3. Samaritan Hebrew 1.3.1. Nifʿal B = Nippaʿʿel (< Nitpaʿʿel) < Nifʿal As has already been discussed above (ch. 10, §1.3.4), from a synchronic perspective, SH has a second N-stem alongside its standard nifʿal (Ben-Ḥayyim 2000, 117–18). This so-called nifʿal B is in reality a result of hitpaelisation, since it is a hybrid that incorporates components of the N- and Dt-stems. It consists of secondary hitpaʿʿel/nitpaʿʿel pronunciation imposed on originally nifʿal orthography, with gemination of both the first and middle radicals—the former in line with assimilation of the t-infix especially common in some late Aramaic dialects (Ben-Ḥayyim 2000, 117– 18; Bar-Asher 2016, 209–10) and the latter characteristic of the Dt-stem pattern. 1.3.2. Samaritan Nifʿal B || Tiberian Qal Above in ch. 10, §1.3.4, the focus was on shifts nifʿal B < qal. Relevant Tiberian qal verbs with SH nifʿal B parallels include (in order of frequency) ‫‘ נָ ַחל‬inherit’ (6x), ‫‘ ָק ַדש‬be holy’ (5x), ‫‘ ָכ ָלה‬finish (intr.)’ (3x) (along with puʿal ‫‘ ֻכ ָלה‬be finished’), ‫‘ גָ ַבר‬prevail’ (2x), ‫‘ ָק ָשה‬be hard, severe’ (2x), with single instances of ‫‘ יָ ֵרא‬fear’, ‫‘ ָלוָ ה‬borrow’, ‫‘ ָמ ַכר‬sell’, ‫‘ ָרגַ ז‬tremble with emotion’, ‫‘ ָת ַמּה‬be astonished’.6 In these cases, qal morphology is preserved in the case of suffix conjugation forms, whereas prefix conjugation forms have 6 Certain individual cases may represent local interpretive peculiarities, rather than broad shifts in verbal morphology. 300 The Historical Depth of the Tiberian Reading Tradition secondary nifʿal B realisations (Hornkohl 2022, 7–9). Compare (11) and (12), repeated from ch. 10, §1.3.4. (11) ‫ ג ְָּׁב ֹׁ֖רּו) המים ויכסו‬MT || gēbēru( ‫חמש עשרה אמה מלמעלה גברו‬ ‫ההרים׃‬ ‘The waters prevailed above the mountains, covering them fifteen cubits deep.’ (Gen. 7.20; see also Gen. 7.19; 49.26) ̊̄ ( ‫ויגברו‬ (12) ‫ וַּיִ גְּׁ ְּׁב ּ֥רּו) המים על הארץ חמשים ומאת‬MT || wyiggåbbaru ‫יום׃‬ ‘And the waters prevailed on the earth 150 days.’ (Gen. 7.24; see also Gen. 7.18) The Tiberian form is qal in both (11) and (12), whereas the SH form is qal in (11), where required by the orthography, but nifʿal B in (12), where the spelling is amenable to nifʿal B realisation. 1.3.3. Samaritan Nifʿal B || Tiberian Nifʿal The hitpaelisation inherent in the SH proliferation of nifʿal B goes beyond shifts nifʿal B < qal. Indeed, far more common is corre- spondence between Samaritan nifʿal B and Tiberian nifʿal, which occur in the case of the following Tiberian nifʿal forms (listed here in order of frequency of Samaritan nifʿal B forms): ‫‘ נמכר‬be sold’ (10x), ‫‘ נטמא‬become unclean’ (9x; BDSS, NBDSS, RH, Tiberian reading tradition), ‫‘ נפרד‬separate (intr.), be separated’ (7x), ‫*נחם‬ ‘be comforted, regret, relent’, ‫‘ *נמלט‬escape’ (5x), ‫‘ נקדש‬be sanctified’ (4x), ‫‘ נאות‬consent’ (3x; ?), ‫‘ נברך‬bless’ (3x), ‫‘ נלוה‬join’ (3x), ‫‘ נשמד‬be destroyed’ (3x), ‫‘ *נבנה‬be built’ (2x), ‫‘ נכבש‬be subdued’ (2x),7 ‫‘ נצל‬survive, escape’ (2x), ‫‘ *נקה‬be released, freed’ (2x), 7 Note that the Samaritan reading tradition is consistent in its reading of Dt-stem forms in Gen. 12.3; 18.18; 22.18; 26.4; 28.14; Deut. 29.18, 13. Hitpaelisation 301 ‫‘ *נקרע‬tear (intr.)’ (2x), ‫‘ *נגר‬ruminate, chew the cud’, ‫‘ *ניסד‬be founded’, ‫‘ *נצה‬fight’, ‫‘ *נקש‬become ensnared’, ‫‘ *נסכר‬be closed’, ‫‘ נעלם‬be hidden’, ‫‘ *נענש‬be punished’, ‫‘ נפתח‬be opened’, ‫נקבץ‬ ‘gather (intr.)’, ‫‘ נקרב‬approach’, ‫‘ נשם‬be desolate’. Nifʿal B passives are particularly common when the corresponding active form is in piʿʿel. Consider the case of ‫‘ נמכר‬be sold’. The Tiberian active-pas- sive qal-nifʿal combination is paralleled by a piʿʿel-nifʿal B combination according to the Samaritan reading tradition (on the piʿʿel, see above, ch. 12, §1.3.1). Thus, (13) ‫אם זרחה השמש עליו דם לו שלם ישלם אם אין לו ונמכר‬ ‫ וְּׁ נִ ְּׁמ ַּכֹׁ֖ר) בגנבתו׃‬MT || wnimmakkɑr( ‘but if the sun has risen on him, there shall be bloodguilt for him. He shall surely pay. If he has nothing, then he shall be sold for his theft.’ (Exod. 22.2; see also Lev. 25.39, 47–48; 27.27) (14) ‫ ִתמָ כֵ ר) לצמיתת כי לי הארץ כי‬MT || timmakkɑr( ‫והארץ לא תמכר‬ ‫גרים ותושבים אתם עמדי׃‬ ‘The land shall not be sold in perpetuity, for the land is mine. For you are strangers and sojourners with me.’ (Lev. 25.23, 42; 27.28; Deut. 15.12) whereas the Samaritan written tradition and Tiberian tradition show a mixture of Dt- and N-stem forms. 302 The Historical Depth of the Tiberian Reading Tradition (15) ‫ ִה ָמ ְּּׁ֣כרֹו) לו עד שנת‬MT || immakkɑr( ‫וחשב עם קנהו משנת המכר‬ ‫היובל והיה כסף ממכרו במספר שנים כימי שכיר יהיה עמו׃‬ ‘He shall calculate with his buyer from the year of (his) being sold to him until the year of jubilee, and the price of his sale shall vary with the number of years. The time he was with his owner shall be rated as the time of a hired worker.’ (Lev. 25.50) The double gemination—of first and second radical—is clear evidence of the hitpaʿʿel/nitpaʿʿel derivation of these forms, showing an advanced stage of hitpaelisation in the Samaritan reading tradition. 1.4. Ben Sira Hitpaelisation in BS is evident in the occurrence of several hitpaʿʿel forms already mentioned as characteristic of • • LBH: ‫‘ התחבר‬join, associate’ (§1.1.1, above); DSS Hebrew: ‫‘ ה(ת)דשן‬become fat, savour’; ‫‘ התאחר‬delay (intr.), be delayed’ (3x); ‫‘ התפתה‬be seduced’ (2x); ‫התאנח‬ ‘sigh, groan’ (3x; Amoraic Hebrew; §1.2.2, above) • SH: ‫‘ התירא‬fear’ (see §1.3, above). BS also presents the first documentation of certain hitpaʿʿel forms (presented here in order of frequency): ‫‘ התנצב‬stand’ (8x); ‫‘ התסייד‬become intimate, take counsel with’ (7x); ‫‘ התעבר‬neglect, pass’ (5x; RH); ‫‘ התחנג‬take delight’ (2x); ‫התמרמר‬/‫‘ התמרר‬be bitter’ (2x); ‫‘ התרחק‬distance oneself, move away’ (2x; RH); ‫‘ התגר‬trade’ (2x); ‫‘ התחרש‬be deaf’ (?); ‫‘ התישן‬grow old’ (RH); ‫‘ התלבש‬wear’ (RH); ‫‘ התלעב‬mock’ (?); ‫‘ התנבל‬become a fool’ (RH2); ‫‘ התנוה‬brag’ (RH1); ‫‘ התעלה‬go up’ (?); ‫‘ התעסק‬exploit’ (MT Gen. 26.2, RH); 13. Hitpaelisation 303 ‫‘ התעשר‬become rich’ (RH2); ‫‘ התפחז‬be reckless’; ‫‘ התקצר‬be short’ (RH); ‫‘ התקרב‬come near, approach’ (RH); ‫‘ התרטש‬break down’; ‫‘ התשעה‬be looked upon’. Several of these are characteristic of RH, whether Tannaitic, Amoraic, or both. 1.5. Rabbinic Hebrew RH, consisting of Tannaitic Hebrew and Amoraic Hebrew, has in common with other Second Temple Hebrew chronolects the use of many Dt-stem/Nt-stem forms unknown from Tiberian CBH. The following list focuses on the Mishna (no attempt is made in the following lists to distinguish between hitpaʿʿel and nitpaʿʿal, i.e., all forms are listed as hitpaʿʿel): • LBH: ‫‘ התענה‬fast’ (19x; NBDSS; BS); ‫‘ התנדב‬freely offer (cultic)’ (19x; NBDSS), ‫‘ התחבר‬join, associate’ (2x; BS), and ‫‘ התגדל‬magnify yourself’ (LBH); • NBDSS: ‫‘ הטמא‬become defiled, unclean’ (167x; SH, Tibe- rian reading tradition [see below, §2.0]); ‫‘ השתלח‬be sent’ (14x); ‫‘ התיחד‬unite (intr.)’ (10x); ‫‘ התמלא‬be filled’ (4x— different semantics in MT Job 16.10); ‫‘ התפזר‬be scattered’ (2x; RH); ‫‘ התכבס‬be washed’; ‫‘ התעצל‬hesitate, be sluggish’; ‫‘ הצטרף‬be refined’; • BS: ‫‘ התעסק‬exploit’ (7x; MT Gen. 26.2); ‫‘ התירא‬fear’ (2x; SH); ‫‘ התלבש‬wear’; ‫‘ התרחק‬distance oneself, move away’. RH, generally, and the Mishna, more specifically, also manifest hitpaelisation via the innovation of many hitpaʿʿel forms unattested in earlier classical or contemporary Second Temple sources. In the following list, forms are presented in order of frequency, with cognate BH and BA forms noted where relevant: 304 The Historical Depth of the Tiberian Reading Tradition ‫‘ הצטרף‬to be joined’ (137x); ‫‘ התייבם‬enter into levirate marriage (said of a woman)’ (35x); ‫‘ התכוון‬intend’ (35x); ‫‘ התקיים‬persist, continue’ (25x; cf. LBH piʿʿel); ‫‘ הסתאב‬become blemished’ (23x); ‫‘ התעשר‬be tithed’ (20x); ‫‘ השתמש‬use’ (17x; cf. BA paʿʿel); ‫התחלק‬ ‘be divided, distributed’ (12x; cf. BH reflexive hitpaʿʿel with direct object in Josh. 18.5 || nifʿal; puʿʿal); ‫‘ התקבל‬receive, accept’ (12x; cf. LBH piʿʿel); ‫‘ השתכל‬look at’ (11x; cf. BA hitpʿʿal); ‫‘ השתתף‬partner, form a partnership’ (11x); ‫‘ התייחד‬be alone (with)’ (10x); ‫‘ התחייב‬be liable’ (8x; cf. LBH piʿʿel); ‫‘ התפרנס‬be provided for, make a living’ (7x); ‫‘ התארש‬become betrothed’ (6x; || BH puʿʿal); ‫‘ מעט‬become diminished’ (6x; || BH qal); ‫‘ התרגם‬be translated (5x; cf. BA puʿʿal); ‫‘ התארח‬be hosted’ (4x); ‫‘ התכנס‬gather (intr.)’ (4x); ‫‘ הסתלק‬move, leave, avoid’ (4x; cf. TA Dt, BA qal); ‫‘ הצטער‬suffer pain, distress’ (4x); ‫‘ הטפל‬attend to, take care of’ (3x); ‫‘ הטרף‬be shaken, torn away’ (3x; || BH qal internal passive; nifʿal); ‫הסתפר‬ ‘have one’s hair cut’ (3x); ‫‘ השתעבד‬be enslaved’ (3x); ‫‘ השתער‬be measured’ (3x; cf. BH qal); ‫‘ התבייש‬be ashamed’ (3x; BH qal); ‫‘ התבשל‬be cooked’ (3x; || BH puʿʿal); ‫‘ התגרש‬be/get divorced (in reference to the wife)’ (3x; cf. BH qal passive participle); ‫התחלל‬ ‘be profaned, deconsecrated’ (3x; || BH nifʿal; puʿʿal); ‫התקשט‬ ‘adorn oneself’ (3x); ‫‘ השתדל‬make an effort, try’ (2x); ‫‘ התבער‬be removed’ (2x); ‫‘ התגייר‬convert to Judaism’ (2x); ‫‘ התאבק‬wallow’; ‫‘ התחרר‬be freed’ (2x); ‫‘ התייאש‬despair, give up hope’ (2x || BH nifʿal; transitive piʿʿel in LBH); ‫‘ הסתפג‬dry oneself’ (2x); ‫‘ הזדייג‬form pairs’; ‫‘ הזדייף‬be falsified’; ‫‘ הסתכר‬earn a profit’; ‫‘ הצטרך‬need’; ‫‘ השתבר‬be broken’ (|| BH nifʿal); ‫‘ השתלש‬be divisible by three’ (cf. BH puʿʿal with different semantics); ‫‘ התאכל‬be digested’ (≈ BH nifʿal); ‫‘ התחכך‬rub up against’; ‫‘ התיישב‬become stable’; ‫התלבן‬ 13. Hitpaelisation 305 ‘become white, be bleached’ (the form in MT Dan. 12.10 is often rendered as a reflexive); ‫‘ התמעך‬be pressed’ (|| BH qal internal passive; qal passive participle); ‫‘ התמצה‬drain, be drained’ (|| BH nifʿal); ‫‘ התמרח‬be rubbed, smeared’; ‫‘ התנונה‬waste away’; ‫התעבר‬ ‘be intercalated’; ‫‘ התעטש‬sneeze’; ‫‘ התעכב‬be delayed’; ‫‘ התעמל‬be kneaded’; ‫‘ התקנב‬be trimmed’. Finally, it should be noted that one of the acknowledged results of hitpaelisation in RH was the replacement of puʿʿal hitpaʿʿel/nitpaʿʿal. Generally speaking, only puʿʿal participles persisted, whereas finite forms gave way to hitpaʿʿel/nitpaʿʿal alterna- tives e.g., BH ‫‘ ֻב ַשל‬be cooked’ (Lev 6.21, 21) versus RH ‫‘ נִ ְּית ַב ֵשל‬be cooked’ (m. Terumot 10.12; m. Maʿaser Sheni 2.1; m. ʿOrla 2.7, 16– 17; m. Nederim 6.6; m. Ḥullin 7.4–5; see https://hebrew-acad- emy.org.il/2018/07/24/‫ונתפעל‬-‫התפעל‬-‫על‬-‫נתבקשנו‬-‫או‬-‫התבקשנו‬/). 2.0. The Tiberian Reading Tradition of Classical Biblical Hebrew Texts The Tiberian reading tradition only occasionally deviates from the morphology reflected by the corresponding written tradition in favour of secondary hitpaʿʿel/nitpaʿʿel morphology. In so doing, it joins with the Second Temple chronolects discussed above in terms of hitpaelisation. 2.1. Nippaʿʿel (< Nitpaʿʿel) < Nifʿal Similar to the Tiberian Hebrew written tradition of exilic texts with hippaʿʿel < hitpaʿʿel forms (see above, §1.1.2), the Tiberian reading tradition occasionally interprets apparently original nifʿal orthographic forms as cases of nippaʿʿel (< nitpaʿʿel). Tiberian vo- 306 The Historical Depth of the Tiberian Reading Tradition calisations of this sort are relatively rare (see also ch. 10, §2.3): ‫‘ וְ ִתנַ ֵ ִ֖שּׂא‬and (his kingdom) will be exalted’ (Num. 24.7); ‫‘ וְ נִ ַכ ֵ ֵ֥פּר‬and (the blood guilt) will be atoned for’ (Deut. 21.8); ‫‘ וְ ִנָּ֣וַ ְסרּו‬and (all women) should take warning’ (Ezek. 23.48); ‫(‘ ִת ַכ ֶסה‬hatred) will be covered’ (Prov. 26.26); ‫(‘ יִ נַ ְשּׂאּו‬and the sons of the violent of your people) will rise up’ (Dan. 11.14); ‫‘ וַ יִ נַ ֵשּׂא‬so he was exalted’ (2 Chron. 32.23); several, but not all, of these come in exilic or post-exilic material. 2.2. I-alveolar Verbs 2.2.1. ‫‘ טמ"א‬become unclean, defile oneself’ Baden’s (2010, 38–39) discusses the case of the nifʿal and hitpaʿʿel of ‫טמ"א‬, both meaning ‘become unclean, defile oneself’. This ap- pears to be a clear case of secondary suppletion, in which the originally nifʿal form was reinterpreted as hitpaʿʿel where permitted by the consonantal spelling. Thus all 18 nifʿal forms are either suffix conjugation forms (16x: Lev. 11.43; 18.24; Num. 5.13–14, 14, 20, 27–29; Jer. 2.23; Ezek. 20.43; 23.7, 13, 30; Hos. 5.3; 6.10) or participles (2x: Ezek. 20.30–31). By contrast, all 15 hitpaʿʿel forms are in the prefix conjugation (Lev. 11.24, 43; 18.24, 30; 21.1, 3–4, 11; Num. 6.7; Ezek. 14.11; 20.7, 18; 37.23; 44.25; Hos. 9.4). Note that the two forms often come in the same context, or even the same verse, e.g., 13. Hitpaelisation 307 (16) ‫תם‬ ֹׁ֖ ֶ ֵ‫ל־ה ֶ ֶ֖ש ֶרץ ַהש ֵ ָֹׂ֑רץ וְּ ַ֤ל ֹא ִ ִֽתטַּ ְּׁמאּו ָב ֵֶ֔הם וְּׁ נִ ְּׁטמ‬ ַ ‫יכם ְּב ָכ‬ ֵֶ֔ ‫צּו ֶאת־נַ ְּפש ֹ ֵת‬ ֙ ‫ל־ת ַש ְּק‬ ְּ ‫ַא‬ ‫ָ ַֽבם׃‬ ‘You shall not make yourselves detestable with any swarming thing that swarms, and you shall not defile yourselves with them, and become unclean through them.’ (Lev. 11.43; see also Lev. 18.24) (17) ‫יהם ַאל־‬ ֶ֖ ֶ ‫ת־מ ְּש ְּפ ֵט‬ ִ ‫ל־ת ֵ֔ ֵלכּו וְּ ֶא‬ ֵ ‫יכ ֙ם ַא‬ ֶ ‫ֹות‬ ֵ ‫חּוקי ֲא ַֽב‬ ַ֤ ֵ ‫יה ֙ם ַב ִמ ְּד ֵָ֔בר ְּב‬ ֶ ֵ‫ל־בנ‬ ְּ ‫וָ א ַ ַֹ֤מר ֶא‬ ‫ל־בית יִ ְּש ָר ֵֵ֗אל ַ֤כֹה ָא ַמ ֙ר‬ ֵ ‫מר ׀ ֶא‬ ֹ ‫ ָל ֵָ֞כן ֱא‬...‫יהם ַאל־ ִתטַּ ָ ִֽמאּו׃‬ ֶ֖ ֶ ‫לּול‬ ֵ ִ‫ּובג‬ ְּ ‫מרּו‬ ֹ ָׂ֑ ‫ִת ְּש‬ ‫יהם ַא ֶ ֵ֥תם ז ִ ַֹֽנים׃‬ ֶ֖ ֶ ‫קּוצ‬ ֵ ‫ֹות ֶיכֶ֖ם ַא ֶתם נִ ְּׁט ְּׁמ ִאָ֑ים וְּ ַא ֲח ֵ ֵ֥רי ִש‬ ֵ ‫יְּהוה ַה ְּב ֶ ֵ֥ד ֶרְך ֲא ַֽב‬ ֵ֔ ִ ‫ֲאד ָֹני‬ ‫יכ ֙ם ַעד־‬ ֶ ‫ּול‬ ֵ ‫יכם ָב ֵֶ֜אש ַא ֶת ֩ם נִ ְּׁט ְּׁמ ִ ֶ֨ ָּ֤אים ְּל ָכל־גִ ַֽל‬ ֶָֹ֨ ֵ‫יכם ְּ ַֽב ַה ֲע ִביר֩ ְּבנ‬ ֶָ֡ ‫ּוב ְּש ֵאת ַמ ְּת ַֹֽנ ֵת‬ ִ ‫ם־א ָד ֵ ֶ֖רש ָל ֶ ַֽכם׃‬ ִ ‫הוה ִא‬ ֵ֔ ִ ְּ‫י־אנִ י נְּ ֻא ֙ם ֲאד ָֹני י‬ ֵָ֗ ‫ַהיֵ֔ ֹום וַ ֲא ִנ֛י ִא ָד ֵ ֵ֥רש ָל ֶכֶ֖ם ֵבית יִ ְּש ָר ֵ ָׂ֑אל ַח‬ ‘And I said to their children in the wilderness, “Do not walk in the statutes of your fathers, nor keep their rules, nor defile yourselves with their idols.”… Therefore say to the house of Israel, Thus says the Lord GOD: “Will you defile yourselves after the manner of your fathers and go whoring after their detestable things? When you present your gifts and offer up your children in fire, you defile yourselves with all your idols to this day. And shall I be inquired of by you, O house of Israel? As I live, declares the Lord GOD, I will not be inquired of by you.”’ (Ezek. 20.18, 30–31) Though translations sometimes appear to reflect a semantic distinction between the nifʿal and hitpaʿʿel forms, e.g., (14), any distinction between the two is in reality merely formal, both capable of a range of middle semantics covering passive and reflexive force, e.g., (15). The suppletion is an example of partial hitpaelisation made where allowed by the orthography. Note that in 308 The Historical Depth of the Tiberian Reading Tradition SH, all forms are nifʿal B (§1.3.3). Hitpaelisation of this verb is also documented in the NBDSS (§1.2.2) and RH (§1.5). 2.2.2. ‫‘ זכ"י‬be cleansed, cleanse yourself’ Active verbs with this root occur in qal, in the sense ‘acquit, be justified’ (Mic. 6.11; Ps. 51.6; Job 15.14; 25.4), and piʿʿel, in the sense ‘keep/make pure’ (Ps. 73.13; 119.9; Prov. 20.9). The imperatival form ‫( ִהזַ ֵ֔כּו‬Isa. 1.16) is orthographically ambiguous, theoretically presupposing nifʿal ‫*הזָ כו‬ ִ or its traditionally hitpaʿʿel morphology. This is the only apparently hitpaʿʿel form of a I-z root in BH, so it is impossible to tell whether the full assimilation of the root-initial z is routine. By way of comparison, root-initial ṣ does not assimilate, but undergoes metathesis. The morphological ambiguity of the NBDSS occurrences of this verb (1QS 3.4; 8.18; 4Q257 3.6; 5Q13 f4.2) make them unhelpful. Metathesis takes place in NBDSS [‫‘ להזד‬to…?’ (5Q13 f1.12) and in RH ‫ְּל ִהיזְּ ַדיֵ יף‬ ‘be falsified’ (m. Giṭṭin 2.4) and ‫ּומיזְּ ַדּוְּ וגִ ין‬ ִ ‘and (they) would form pairs’ (m. Sanhedrin 5.5). It seems possible that the biblical or- thography ‫( הזכו‬Isa. 1.13) reflects a nifʿal form that was secondarily read as a hitpaʿʿel. 2.2.3. ‫‘ דכ"א‬be crushed’ The verb with transitive semantics is piʿʿel (Isa. 3.15; 53.10; Ps. 72.4; 89.11; 94.5; 143.3; Job 4.19; 6.9; 19.2; Prov. 22.22; Lam. 3.34). The corresponding passive puʿʿal comes four times (Isa. 19.10; 53.5; Jer. 44.10; Job 22.9). An unequivocal nifʿal form comes in ‫‘ נִ ְּד ָכ ִ ַֽאים‬ones being crushed’ (Isa. 57.15). Ambiguous or- thographic forms vocalised as hitpaʿʿel/nitpaʿʿal come in the case 13. Hitpaelisation 309 of ‫‘ וְּ יִ ֵַַּֽ֣ד ְּכ ֵ֥אּו‬and they are crushed’ (Job 5.4) and ‫‘ וְּ יִ ַד ָ ַֽכאּו‬and they are crushed’ (Job 34.25). On semantic grounds, Baden (2010, 38) assumes an original nifʿal secondarily read as hitpaʿʿel, but the reg- ularity of piʿʿel and puʿʿal forms may point to the authenticity of the t-stem morphology. Baden (2010, 40–43) also notes that initial-alveolar and initial-affricate forms are disproportionately underrepresented in terms of nifʿal morphology, suggesting that such forms were disproportionately reinterpreted as hitpaʿʿel forms.8 2.2.4. ‫‘ דב"ר‬speak (divine)’ On three occasions in Tiberian BH one encounters the hitpaʿʿel active participle ‫מ ַד ֵבר‬: ִ (18) ‫ת־ה ֶ֜קֹול ִמ ַּד ֵבּ֣ר ֵא ֵ֗ ָליו ֵמ ַ ַ֤על‬ ַ ‫מֹוע ֮ד ְּל ַד ֵבר ִאתֹו֒ וַ יִ ְּש ַָֹ֨מע ֶא‬ ֵ ‫א ֶהל‬ ֹ ‫ּוב ָֹ֨ב ֹא מ ֶֶֹ֜שה ֶאל־‬ ְּ ‫ל־אר ֹן ָה ֵע ֻ ֵ֔דת ִמ ֵ ֶ֖בין ְּש ֵני ַה ְּכ ֻר ִ ָׂ֑בים וַ יְּ ַד ֵ ֶ֖בר ֵא ָ ַֽליו׃‬ ֲ ‫ַה ַכ ָֹ֨ ֹפ ֶר ֙ת ֲא ֶש ֙ר ַע‬ ‘And when Moses went into the tent of meeting to speak with the LORD, he heard the voice speaking to him from above the mercy seat that was on the ark of the testimony, from between the two cherubim; and it spoke to him.’ (Num. 7.89) (19) ‫ל־רגְּ ָלָׂ֑י וָ ֶא ְּש ַָ֕מע ֵ ֶ֖את ִמ ַּד ֵבּ֥ר ֵא ָ ַֽלי׃‬ ַ ‫וַ ָ ָ֧תבֹא ִבי ֵ֗ר ַּוח ַ ַֽכ ֲא ֶש ֙ר ִד ֶבר ֵא ֵ֔ ַלי וַ ַת ֲע ִמ ֵ ֶ֖דנִ י ַע‬ ‘And the Spirit entered into me as he spoke to me and [the spirit] set me on my feet, and I heard him speaking to me.’ (Ezek. 2.2) 8 Citing the likes of Yellin (1924), Bergsträsser (1918–1929, II:§16d), and Siebesma (1991, 169), Baden (2010, 39, fn. 17) also lists the roots ‫בר"ר‬, ‫גא"ל‬, and ‫ כס"י‬as mixing nifʿal and hitpaʿʿel morphology. But the suppletion in these cases is not as consistent as in those discussed above. 310 The Historical Depth of the Tiberian Reading Tradition (20) ‫וָ ֶא ְּש ַ ֛מע ִמ ַּד ֵבּ֥ר ֵא ַלֶ֖י ֵמ ַה ָ ָׂ֑ביִ ת וְּ ִָ֕איש ָה ָיֵ֥ה ע ֵ ֶֹ֖מד ֶא ְּצ ִ ַֽלי׃‬ ‘I heard one speaking to me out of the temple, while the man was standing beside me.’ (Ezek. 43.6) The apparently secondary use of hitpaʿʿel is restricted to originally piʿʿel participles, as this consonantal form is amenable to hitpaelisation due to the assimilation of the infix -t- to the following dental d. Notably, it is restricted to contexts of divine speech. This was evidently one strategy among many employed as part of a broad Second Temple effort to avoid anthropomorphism of the deity. Ben-Ḥayyim (2000, 218, §2.14.18, fn. 198) notes that such techniques are especially characteristic of the Targums. Indeed, observe that in the Aramaic rendering of Targums Onqelos in (21), Dt-stem participles correspond to both the participle and a finite verbal form in the MT: (21) ‫וכד עליל משה למשכן זמנא למללא עמיה ושמע ית קלא דמתמלל עימיה‬ ‫מעלוי כפורתא דעל ארונא דסהדותא מבין תרין כרוביא ומתמלל עמיה׃‬ ‘And when Moses would go into the tent of meeting to speak with the LORD, and he would heard the voice speaking to him from above the mercy seat that was on the ark of the testimony, from between the two cherubim; and it would speak to him.’ (TO Num. 7.89) For further evidence of the Targumic distinction between the Dstem for human speech and the Dt-stem for divine speech, see 13. Hitpaelisation 311 (22) ‫ֹלהים ֶפן־‬ ֶ֖ ִ ‫־א ָ ֵ֥תה ִע ָ ֶ֖מנּו וְּ נִ ְּש ָ ָׂ֑מ ָעה וְּ ַאל־יְּׁ ַּד ֵבּ֥ר ִע ָ ֛מנּו ֱא‬ ַ ‫רּו ֶאל־מ ֵֶֹ֔שה ַּדבֵ ר‬ ֙ ‫אמ‬ ְּ ֹ ‫וַ ַֽי‬ MT ‫נָ ַֽמּות׃‬ TO ‫ואמרו למשה מליל את עימנא ונקביל ולא יתמלל עמנא מן־קדם יוי‬ ‫דלמא נמות׃‬ ‘And they said to Moses, “You speak to us, and we will listen; but do not let God speak to us, lest we die.”’ (Ex- od. 20.19) (23) ‫הו֛ה ֵא ַלֶ֖י ַה ָלָׂ֑יְּ ָלה‬ ָ ְּ‫ל־ש ֵ֔אּול ֶ ַ֚ה ֶרף וְּ ַא ִג ָידה ְּל ֵָ֔ך ֵא ֩ת ֲא ֶָֹ֨שר ִד ֶב֧ר י‬ ָ ‫מּואל֙ ֶא‬ ֵ ‫אמר ְּש‬ ֶ ֹ ‫וַ ַ֤י‬ MT ‫) לֶ֖ ֹו ַּד ֵ ִֽבר׃ ס‬Q( ‫אמר‬ ֶ ֹ ‫) וַ ֵ֥י‬K( ‫ויאמרו‬ TO ‫ואמר שמואל לשאול אוריך ואחוי לך ית דאתמלל מן קדם יי עמי בליליא‬ ‫ואמר ליה מליל׃‬ ‘Then Samuel said to Saul, “Stop! I will tell you what which the LORD spoke to me this night.” And he said to him, “Speak.”’ (1 Sam.15.16; see also TJ Ezek. 2.2)9 The Targums, thus, reflect a tradition similar to that reflected in the Tiberian reading tradition. The same is true of RH (Tannaitic and Amoraic sources). Conversely, other Second Temple Hebrew sources show no sign of this distinction. In the relevant passage, the SP has the more expected—and original—piʿʿel form ‫ מדבר‬amdabbər ‘[the voice] speaking’ (Num. 7.89). Likewise, the Peshiṭta has D-stem forms parallel to the MT hitpaʿʿel forms. Neither the Old Greek nor the Vulgate show special forms corresponding to the MT’s hitpaʿʿels. The use of dedicated Dt-stem verbs for divine speech is thus a feature specific to Jewish interpretive traditions. It dates to at least the Tannaitic period, prior 9 For Dt-stem forms of ‫‘ מל"ל‬speak’ more generally in reference to divine speech, see in TO Gen 16.13; Exod. 33.9; TJ Jer. 9.11; Ezek. 1.3, 28; 13.7; 22.28; Hab. 2.1; Targum Song 1.2; 2.5. 312 The Historical Depth of the Tiberian Reading Tradition if the reading component of the Tiberian biblical tradition had already crystallised by then. 3.0. The Tiberian Classical Biblical Hebrew Written Tradition 3.1. Northwest Semitic Inscriptions No hitpaʿʿel forms occur in the limited corpus of Iron Age Hebrew epigraphy (Gogel 1998, 119). However, t-stem forms are found in the wider Northwest Semitic repertoire, specifically, in the Moabite of the Meshaʿ Stele, where one finds repeated occurrences of the hifteʿel form ‫‘ הלתחם‬fight’ (KAI 181 1.11, 15, 19, 32–33). Clearly, t-stem forms semantically parallel to BH nifʿal forms were extant in Iron Age sources. 3.2. Synonymy between Hitpaʿʿel and Other Stems Yet, it would be misleading to suggest that synonymy between hitpaʿʿel and other stems is an exclusively late phenomenon. Consider the following examples, which may be considered more broadly representative. 3.2.1. ‫‘ נִ ְּב ַרְך || ִה ְּת ָב ֵרְך‬be blessed, bless oneself’ Whatever the exact meaning of the hitpaʿʿel (Gen. 22.18; 26.4; Deut. 29.18; Isa. 65.16; Jer. 4.2; Ps. 72.17) and nifʿal (Gen 12.3; 18.18; 28.14), their appearance in nearly parallel contexts in Genesis would seem to demonstrate early semantic overlap. 13. Hitpaelisation 313 3.2.2. ‫‘ נֶ ְּח ָבא || ִה ְּת ַח ֵבא‬hide (intr.) In both Tiberian CBH and LBH, the hitpaʿʿel (Gen. 3.8; 1 Sam. 13.6; 14.11, 22; 23.23; 2 Kgs 11.3; Job 38.30; 1 Chron. 21.20; 2 Chron. 22.9, 12) and nifʿal (Gen. 3.10; 31.27; Josh. 2.16; 10.16– 17, 27; Judg. 9.5; 1 Sam. 10.22; 19.2; 2 Sam. 17.9; Amos 9.3; Job 5.21; 29.8, 10; Dan. 10.7; 2 Chron. 18.24) forms appear with identical semantics. Indeed, they occur separated by a single verse in the same story in Gen. 3.8 and 10. 3.2.3. ‫ה ְּתיַ ֵצב‬/ ִ ‫‘ נִ ַצב‬position oneself, stand’ The connection between the hitpaʿʿel ‫יַצב‬ ֵ ‫ ִה ְּת‬and the nifʿal ‫ נִ ַצב‬is not merely one of semantic synonymy, but of partial suppletion. In Tiberian BH the hitpaʿʿel occurs primarily as a prefix conjugation form, imperative, or infinitive construct. It occurs just twice as a suffix conjugation form, specifically in LBH. The nifʿal, conversely, occurs only as a participle and suffix conjugation form, the latter outside of LBH. Given this sort of mutual exclusivity, it is not surprising that the two forms should occur with similar semantics in close proximity, e.g., ‫‘ וְּ נִ ַצ ְּב ָ ֵ֥ת‬and you will stand’ (Exod. 34.2) and ‫‘ וַ יִ ְּתיַ ֵ ֵ֥צב‬and he stood’ (Exod. 34.5). Consider also the hitpaʿʿel forms in Num. 22.22; 23.3, 15 versus the nifʿal forms in Num. 22.23, 31, 34; 23.6, 17. Finally, nearly parallel uses involve the nifʿal ‫‘ וְּ נִ ַצ ְּב ָ ֵ֥ת‬and you will stand’ (Exod. 7.15; see also 5.20) and the hitpaʿʿel ‫‘ וְּ ִה ְּתיַ ֵצ ֙ב‬and stand’ (Exod. 8.20; see also 9.13). Clearly, the above is strong evidence of early hitpaʿʿel-nifʿal correspondence. 314 The Historical Depth of the Tiberian Reading Tradition 3.2.4. ‫‘ נָ כֹון || ִה ְּתכֹונֵ ן‬be established’ There is arguable semantic overlap between the hitpolel and the nifʿal, but the most striking feature of the hitpolel is the consonantal evidence it provides for the hippaʿʿel < hitpaʿʿel (nippaʿʿel < nitpaʿʿel), or, more specifically, hippolel < hitpolel (nippolel < nitpolel), shift more evident in the pronunciation component of the Tiberian reading tradition and other Second Temple traditions (i.e., SH). Indeed, in three of the four hitpolel instances, the t has assimilated: ‫כֹונֵּ֣ן‬ ֶ֖ ֵ ‫‘ וְּ ִת‬be established, rebuilt (FS)’ (Num. 21.27); ‫כֹונָׂ֑נִ י‬ ָ ‫‘ ִת‬you (FS) will be (re)established’ (Isa. 54.1); ‫כֹונָׂ֑נּו‬ ָ ִ‫‘ וְּ י‬and they (M) make ready’ (Ps. 59.5); cf. ‫כֹונֵַּֽ֣ן‬ ָ ‫‘ יִ ְּת‬it (M) is established’ (Prov. 24.3). Note that the relevant consonantal forms are unambiguously hippolel/nippolel < hitpolel/nitpolel, as evidenced by reduplication of the n. This is strong evidence that the apparently secondary vocalisation development seen above in §2.1 is in line with developments already seen in the Tiberian written tradition.10 3.2.5. ‫‘ נִ ָבא || ִה ְּתנַ ֵבא‬prophesy’ So apparently interchangeable are the hitpaʿʿel and nifʿal of ‫נב"א‬ that they both come throughout BH, frequently appearing in close proximity, including on four occasions within a single verse: ‫‘ נִ ְּב ִֵ֔אים‬prophesying (MPL)’ and ‫‘ וַ ִי ְַּֽתנַ ְּב ֶ֖אּו‬and they prophesied’ (1 Sam. 19.20); ‫‘ נִ ְּב ִאים‬prophesying (MPL)’ and ‫‘ ִ ַֽמ ְּתנַ ְּב ִ ֵ֥אים‬prophesying (MPL)’ (Jer. 14.14); ‫‘ ָה ָיַ֤ה ִמ ְּתנַ ֵב ֙א‬would prophesy (MS)’ and ‫וַ יִ נָ ֵָ֞בא‬ ‘and he prophesied’ (Jer. 26.20); ‫‘ ַה ִ ַֽמ ְּתנַ ְּב ֶ֖אֹות‬who are prophesying 10 Consider also hippolel/nippolel ‫רֹומם‬ ֵָ֔ ‫‘ ֵ ַֽא‬I will exalt myself’ (Isa 33.10) versus hitpolel/nitpolel ‫רֹומם‬ ַ֤ ֵ ‫‘ וְּ יִ ְּת‬and he will exalt himself’ (Dan 11.36). 13. Hitpaelisation 315 (FPL)’ and ‫‘ וְּ ִהנָ ֵ ֶ֖בא‬and prophesy! (MS)’ (Ezek. 26.20). In the case of these verbs, semantic correspondence between hitpaʿʿel and nifʿal seems to have deep historical roots. 3.2.6. ‫‘ נִ ַחם || ִה ְּתנַ ֵחם‬be comforted; regret, change one’s mind’ Hitpaʿʿel forms, usually in the sense ‘take comfort, be comforted’, (7x) are rarer than nifʿal (48x), usually ‘repent, regret’. The hitpaʿʿel occasionally has the meaning more commonly associated with the nifʿal, e.g., ...‫ן־א ָ ֶ֖דם וְּׁ יִ ְּׁתנ ֶָחָ֑ם‬ ָ ‫ּוב‬ ֶ ‫יכ ֵֵ֔זב‬ ַ ‫‘ ל ֹא ִ ֵ֥איש ֵאל֙ ִ ַֽו‬God is not a man that he should like, nor a human that he should change his mind’ (Num. 23.19); cf. ‫וְּ גַ ֙ם ֵנ ַצח יִ ְּש ָר ֵֵ֔אל ֵ֥ל ֹא יְּ ַש ֵ ֶ֖קר וְּ ל ֹא יִ נ ֵָחָ֑ם ִכי ֵ֥ל ֹא‬ ‫ָחם׃‬ ִֽ ֵ ‫‘ ָא ָ ֛דם ֶ֖הּוא ְּׁל ִהנ‬And also the Glory of Israel does not lie and does not change his mind, for he is not a man, that he should change his mind”’ (1 Sam. 15.29). The reverse semantic shift, that of nifʿal bearing the sense more typically associated with hitpaʿʿel, also occurs. Consider the following verses about Judah from consecutive chapters: (24) ‫ל־ב ִנ֛י‬ ְּ ‫י־א ֵ ָ֧רד ֶא‬ ֵ ‫אמר ִ ַֽכ‬ ֶ ֹ ‫ַּחם וַ ָ֕י‬ ֵ ִ֔ ‫אן ְּׁל ִה ְּׁתנ‬ ֙ ֵ ‫ל־בנ ֶָֹ֜תיו ְּלנַ ֲח ֵ֗מֹו וַ יְּ ָמ‬ ְּ ‫ל־ב ָֹ֨ ָניו וְּ ָכ‬ ָ ‫וַ יָ ֻקמּו֩ ָכ‬ ‫א ָלה וַ יֵ ְֵּ֥בךְּ א ֶֹ֖תֹו ָא ִ ַֽביו׃‬ ֹ ָׂ֑ ‫ָא ֵ ֶ֖בל ְּש‬ ‘All his sons and daughters stood by him to console him, but he refused to be consoled. “No,” he said, “I will go to the grave mourning my son.”’ (Gen. 37.35) (25) ‫אנֹו‬ ֙ ֹ ‫יְּהּודה וַ ֶַ֜י ַעל ַעל־ ַֹֽגזֲ ֵזַ֤י צ‬ ֵ֗ ָ ‫הּודה וַּיִ נָ ּ֣חֶ ם‬ ָׂ֑ ָ ְּ‫בּו ַהיָ ִֵ֔מים וַ ָ ֶ֖ת ָמת ַבת־ש ַּוע ֵ ַֽא ֶשת־י‬ ֙ ‫וַ יִ ְּר‬ ‫ירה ֵר ֵ ֵ֥עהּו ָה ֲע ֻד ָל ִ ֶ֖מי ִת ְּמ ָנ ַָֽתה׃‬ ֛ ָ ‫ֵ֗הּוא וְּ ִח‬ ‘After some time Judah’s wife, the daughter of Shua, died. After Judah was consoled, he left for Timnah to visit his sheepshearers, along with his friend Hirah the Adullamite.’ (Gen. 38.12) 316 The Historical Depth of the Tiberian Reading Tradition One further piece of evidence for morphosemantic overlap between hitpaʿʿel and nifʿal can be found in the form ‫‘ וְּ ִהנֶ ָ ָׂ֑ח ְּמ ִתי‬and I will satisfy myself’ (Ezek. 5.13). It represents the development hippaʿʿel < hitpaʿʿel, which in SH came to be identified as nifʿal B and is related to RH nitpaʿʿal. 3.2.7. ‫‘ נִ ְּס ַתר || ִה ְּס ַת ֵתר‬hide (intr.)’ There appears to be little to no semantic difference between hitpaʿʿel ‫ ִה ְּס ַת ֵתר‬and nifʿal ‫ נִ ְּס ַתר‬when in reference to a human sub- ject (the nifʿal is more common overall, and with non-human subjects, but cf. Isa. 29.14). For synonymous usage, compare (26) ‫דֹות‬ ֙ ‫מר ֲהלֹוא ָ ּ֠דוִ ד ִמ ְּׁסתַּ ֶ֨ ֵתר ִע ָ ַ֤מנּו ַב ְּמ ָצ‬ ֹ ָׂ֑ ‫ל־ש ֵ֔אּול ַהגִ ְּב ָ ֶ֖ע ָתה ֵלא‬ ָ ‫ים ֶא‬ ֙ ‫וַ יַ ֲעלַ֤ ּו זִ ִפ‬ ‫ימֹון׃‬ ַֽ ‫ימין ַהיְּ ִש‬ ֵ֥ ִ ‫ַב ֵ֔חֹ ְּר ָשה ְּבגִ ְּב ַע ֙ת ַ ַֽה ֲח ִכ ֵ֔ ָילה ֲא ֶ ֶ֖שר ִמ‬ ‘Then the Ziphites went up to Saul at Gibeah, saying, “Is not David hiding among us in the strongholds at Horesh, on the hill of Hachilah, which is south of Jeshimon?”’ (1 Sam. 23.19; see also 26.1) (27) ‫) ַה ֶל ֶֶ֖חם ֶל ֱא ַֽכֹול׃‬Q( ‫) ֶאל־‬K( ‫תר ָדִוֶ֖ד ַב ָש ֶ ָׂ֑דה וַ יְּ ִהי ַה ֵ֔חֹ ֶדש וַ יֵ ֶָ֧שב ַה ֶ ֛מ ֶלְך על‬ ּ֥ ֵ ָ‫וַּיִ ס‬ ‘And David hid in the field. And when the new moon came, the king sat down to eat food.’ (1 Sam. 20.24; see also 20.5, 19) 3.2.8. ‫‘ נִ ְּק ַבץ || ִה ְּת ַק ֵבץ‬gather (intr.)’ In reference to humans, the hitpaʿʿel and nifʿal are largely synon- ymous regarding the meaning ‘gather (intr.)’, though the nifʿal apparently has passive semantics as well. Cf. ‫יה‬ ָ ‫ִ ִֽה ְּׁתקַּ ְּׁבצּו ּובֹאּו ָע ֵ֔ ֶל‬ ‫‘ וְּ ֶ֖קּומּו ַל ִמ ְּל ָח ָ ַֽמה׃‬gather and come against it and rise for war’ (Jer. 49.14) and ‫אּו ֵה ָא ְּספּו ִמ ָס ִֵ֔ביב‬ ֙ ֹ ‫‘ ִהקָ ְּׁבצָּ֤ ּו וָ ֙ב‬gather and come, assemble 13. Hitpaelisation 317 around’ (Ezek. 39.17). Even more convincing as examples of semantic synonymy are the nifʿal and hitpaʿʿel in consecutive verses in nifʿal ‫‘ וַ יִ ָק ְּבצּו‬and they gathered (intr.)’ (1 Sam. 7.6) followed by hitpaʿʿel ‫(‘ ִה ְּת ַק ְּב ַ֤צּו‬the Israelites) gathered (intr.)’ (1 Sam. 7.7). 3.3. Evidence of Hitpaʿʿel-Nifʿal Merger Discussed above, in §2.1, was the reinterpretation of nifʿal forms as hitpaʿʿel/nitpaʿʿel forms with assimilated t-infix. Emphasised were the secondary nature of the vocalism and its agreement with trends characteristic of late Aramaic and Hebrew sources. In a few cases, however, suffix conjugation forms can be read only as t-stem forms with assimilated infix -t-: ‫כֹונֵּ֣ן‬ ֶ֖ ֵ ‫‘ וְּ ִת‬be established, rebuilt (FS)’ (Num. 21.27); ‫רֹומם‬ ֵָ֔ ‫‘ ֵ ַֽא‬I will exalt myself’ (Isa. 33.10); ‫כֹונָׂ֑נִ י‬ ָ ‫‘ ִת‬you (FS) will be (re)established’ (Isa. 54.1); ‫‘ ִהנַ ְבאּו‬they prophesied’ (Jer. 23.13); ‫‘ וְ ִהנֶ ָ ָ֑ח ְמ ִתי‬and I will be satisfied’ (Ezek. 5.13); ‫אתי‬ ִ ‫‘ וְ ִהנַ ֵ ִ֖ב‬and I prophesied’ (Ezek. 37.10). ‫כֹונָׂ֑נּו‬ ָ ִ‫‘ וְּ י‬and they (M) make ready’ (Ps. 59.5). Clearly, these unambiguous consonantal t-stem forms with assimilated tav lend credence to the vo- calisation of the apparently hippaʿʿel/nippaʿʿel < hitpaʿʿel/nitpaʿʿel forms seen above. 4.0. Conclusion Probably as a result of factors external (contact with Aramaic) and internal (growing use of hitpaʿʿel as a medio-passive, not just a reflexive), hitpaelisation is a characteristic of Second Temple Hebrew as reflected in multiple sources and traditions (§1.0). A number of apparent cases of dissonance between the reading and written components of the Tiberian biblical tradition involve sec- 318 The Historical Depth of the Tiberian Reading Tradition ondary hitpaʿʿel/nitpaʿʿel analysis of forms originally in other stems, especially, nifʿal (§2.0). As seen in §3.0, however, the secondary vocalic deviations find precedents in several features seen in First Temple sources, including the use of t-stem forms in Iron Age Semitic epigraphy (§3.1); not infrequent synonymy between t-stem and N-stem, including cases of suppletion (§3.2); and evidence of the N- and t-stem merger in the case of nippaʿʿel/ nippolel < nitpaʿʿel/nitpolel shifts. 14. ṬƐRƐM QAṬAL The temporal particle ‫)ט ֶרם‬ ֶ ‫מ‬/ ִ ‫(ב‬ ְּ ‘before’ comes 56 times in BH. Occasionally followed by a noun or infinitive,1 it most frequently—52 times—precedes a finite verb or verbal clause (see below). In 48 of these 52 cases, the finite verbal form in question is in the prefix conjugation yiqṭol. The focus of this chapter is the minority syntactic structure of ‫ ֶט ֶרם‬followed by the suffix conju- gation, i.e., ṭɛrɛm qaṭal. 1.0. The Majority Syntax: Ṭɛrɛm Yiqṭol It is opportune to begin with a brief discussion of the dominant syntactic structure, ‫ ֶט ֶרם‬followed by the prefix conjugation, i.e., ṭɛrɛm yiqṭol. 1.1. Ṭɛrɛm Yiqṭol with Expected Yiqṭol Semantics In some 27 cases of ṭɛrɛm yiqṭol, the prefix conjugation may be construed to have a TAM value consistent with its standard semantic range: (1) future or modal (i.e., prescriptive), (2) generic/ stative present, of (3) habitual past:2 1 Noun: ‫‘ ְּב ֶ ֵ֥ט ֶרם ֶ֖בֹ ֶקר‬before morning’ (Isa. 17.14); ‫‘ ְּב ֶט ֶרם ַ ֵ֔קיִ ץ‬before summer’ (Isa. 28.4); infinitive: ‫‘ ְּב ֶ֙ט ֶר ֙ם ֶל ֶדת ֵ֔חֹק‬before a decree takes effect’ (Zeph. 2.2a); ‫ּום־א ֶבן‬ ֛ ֶ ‫‘ ִמ ֶ ָ֧טּ ֶרם ַֽש‬before the placing of a stone’ (Hag. 2.15). 2 The TAM semantics of some cases of ṭɛrɛm yiqṭol, especially in poetry, are debatable. © 2023 Aaron D. Hornkohl, CC BY-NC 4.0 https://doi.org/10.11647/OBP.0310.14 320 The Historical Depth of the Tiberian Reading Tradition (1) ...‫ֲלה ַה ָב ֶָ֜מ ָתה ֶל ֱא ֵ֗כֹל‬ ֶ ֶ֨ ‫ְּכב ֲֹא ֶכם ָה ִעיר ֵכן ִת ְּמ ְּצאּון א ָֹ֡תֹו ְּׁבטֶ ֶרם ַּיע‬ ‘As soon as you enter the city you will find him, before he goes up to the high place to eat…’ (1 Sam. 9.13; additional future/modal cases include Gen. 27.4; 45.28; Lev. 14.36; Deut. 31.21; 2 Kgs 2.9; Isa. 7.16; 8.4; 65.24; 66.7 [?], 7 [?]; Jer. 13.16, 16; 38.10; Ps. 39.14; 58.10 [?]; Zeph. 2.2b, 2c; Prov. 30.7; Job 10.21) (2) ‫י־חיֹות ֵֵ֔הנָ ה ְּׁבטֶ ֶרם תָ ב֧ ֹוא ֲא ֵל ֶ ֛הן ַה ְּמיַ ֶל ֶֶ֖דת‬ ָ ‫ָ֧ל ֹא ַכנָ ִ ֛שים ַה ִמ ְּצ ִר ֶֹ֖ית ָ ַֽה ִע ְּב ִר ָֹׂ֑ית ִ ַֽכ‬ ‫וְּ יָ ָ ַֽלדּו׃‬ ‘“Because Hebrew women are not like Egyptian women, for they are vigorous and before the midwife comes to them, they give birth.”’ (Exod. 1.19; additional generic present cases include Exod. 9.30; 10.7; Isa. 42.9 [?]; Prov. 18.13) (3) ...‫ת־ה ֵח ֶל ֒ב ָּובא׀ ַנ ַער ַהכ ֵֵֹ֗הן וְּ ָא ַמ ֙ר ָל ִאיש ַהז ֵֵֹ֔ב ַח‬ ַ ‫גַ ם֘ ְּׁבטֶ ֶרם֘ י ְַּּׁק ִט ּ֣רּון ֶא‬ ‘Moreover, before they could burn the fat, the priest’s servant would come and say to the one sacrificing…’ (1 Sam. 2.15; Ruth 3.14) None of these usages of the prefix conjugation after ‫ ֶט ֶרם‬is unexpected or surprising, given that the yiqṭol form regularly encodes such semantic values even in the absence of ‫ט ֶרם‬. ֶ 1.2. Ṭɛrɛm Yiqṭol with Unexpected Yiqṭol Semantics In some 21 instances of ṭɛrɛm yiqṭol, however, the yiqṭol form in question appears to represent a completive eventuality temporally anterior to speech time, i.e., perfective past. In such cases, ancient and modern translations routinely (though not exclusively) resort to preterite or pluperfect renderings. Some scholars have thus concluded that the prefix conjugation in the ṭɛrɛm yiqṭol 14. Ṭɛrɛm Qaṭal 321 structure has otherwise anomalous perfective past semantics (Waltke and O’Connor 1990, 497–98, §31.1.1d, 501, §31.1.1f, 513–14, §31.6.3). To account for this, some even opine that the prefix conjugation in question is a vestige of short preterite yiqṭol (< PS yaqtul) (Arnold and Choi 2003, 60). Yet, while the eventualities depicted in the relevant cases of ṭɛrɛm yiqṭol are indeed anterior to the moment of speech (i.e., past tense) and are in context aspectually completive (i.e., perfective), where a morphological distinction is perceptible, they consistently exhibit forms consistent with long yiqṭol (< yaqtulu/a), rather than short yiqṭol (< yaqtul) morphology expected for preterite semantics (Williams 1976, 30–31, §167).3 If so, notwithstanding the propensity for perfective past glossing in translations, the usage is unlikely to consist of a genuinely perfective past yiqṭol, whether short or long. Rather, it is most plausibly explained in light of yiqṭol’s rather common reference to relative future (Hendel 1996, 159–60; JM, 342, §113j and fn. 21; Cook 2012, 262–63; van der Merwe, Naudé, and Kroeze 2017, 161, §19.3.2, 462–63, §41.8).4 In past tense narrative context, a yiqṭol form can be used to express the prospective or posterior past, i.e., future-in-the-past. Consider the bolded yiqṭol forms in examples (4)–(5): 3 Observe the long III-y forms in Gen 2.5a; 24.45; 37.18; 1 Sam. 3.3, 7b; Jer. 47.1; Ezek. 16.57; Ps. 119.67. 4 On the notion of relative tense in BH, see Goldfajn (1998); Cohen (2013, 33–34 et passim). 322 (4) The Historical Depth of the Tiberian Reading Tradition ...‫ל־ה ָא ָ ֵ֔דם ִל ְּר ֶ֖אֹות ַמה־יִ ְּׁק ָרא־לָׂ֑ ֹו‬ ָ ‫וַ יָ ֵב ֙א ֶא‬ ‘and [God] brought [each animal] to the man to see what he would call it’ (Gen 2.19) (5) ...‫ת־ח ְּליֵ֔ ֹו ֲא ֶ ֵ֥שר יָמֹׁ֖ ּות ָׂ֑בֹו‬ ָ ‫יש ֙ע ָח ָלה ֶא‬ ָ ‫ֶו ֱַֽא ִל‬ ‘And Elisha became ill with the illness from which he would die…’ (2 Kgs 13.14a) The same future-in-the-past sense of yiqṭol can occur after the particle )‫(א ֶשר‬ ֲ ‫עד‬, ַ as in (6)–(7), the latter of which includes a second example of the prefix conjugation for relative future in a subordinate clause after the particle ‫‘ ַמה‬what’. (6) ...‫איְּ ֵָ֔ביו‬ ֹ ַֽ ‫גֹוי‬ ֙ ‫וַ יִ ָֹ֨ד ֹם ַה ֶֶ֜ש ֶמש וְּ יָ ֵר ַח ָע ֵָ֗מד עַּ ד־יִ ּ֥קם‬ ‘And the sun stood still, and the moon stopped, until the nation could take vengeance upon its foes…’ (Josh. 10.13) (7) ‫יה ַב ֵֵ֔צל‬ ָ֙ ‫ן־ה ֵ֔ ִעיר וַ יֵ ֶֶ֖שב ִמ ֶק ֶדם ָל ִ ָׂ֑עיר וַ יַ ַעש֩ ָֹ֨לֹו ֶָ֜שם ֻס ֵָ֗כה וַ יֵ ֶַ֤שב ַת ְּח ֶ֙ת‬ ָ ‫וַ יֵ ֵצַ֤א יֹונָ ֙ה ִמ‬ ‫עַּ ַ֚ד א ֲֶשּ֣ר יִ ְּׁר ִֶ֔אה מַּ ה־יִ ְּׁהיֶ ֹׁ֖ה ָב ִ ַֽעיר׃‬ ‘Jonah went out of the city and sat to the east of the city and made a booth for himself there. He sat under it in the shade, till he should see what would become of the city. In (4)–(7) above, the relevant yiqṭol forms encode perfective eventualities anterior (i.e., past) in relation to speech time, but posterior (i.e., future) relative to narrative reference time, or, in Reichenbachian terms, R<E<S (see Cohen 2013, 151–53). This would seem to be the same meaning that obtains in yiqṭol following ‫‘ ֶט ֶרם‬before’, as in (8). 14. Ṭɛrɛm Qaṭal (8) 323 ...‫יֹוסף יֻ ַל ֙ד ְּש ֵני ָב ֵ֔ ִנים ְּב ֶ ֵ֥ט ֶרם תָ בֹׁ֖ ֹוא ְּש ַנת ָה ָר ָ ָׂ֑עב‬ ַ֤ ֵ ‫ּול‬ ְּ ‘And to Joseph were born two sons before the year of the famine would come…’ (Gen. 41.50; additional relative future/prospective past cases include Gen. 2.5, 5; 19.4; 24.46; 27.33; 37.18; Exod. 12.34; Num. 11.33; Josh. 2.8; 3.1; Judg. 14.18; 1 Sam. 3.3, 7b; 2 Kgs 6.32; Isa. 48.5; Jer. 1.5, 5; 47.1; Ezek. 16.57; Ps. 119.67) In (4)–(8) above, the eventualities are past from the perspective of speech time and are most naturally given to completive interpretations, but yiqṭol is employed due to the relative future force in a subordinate clause. Yiqṭol dominates after ‫ֶט ֶרם‬ to the near exclusion of qaṭal, evidently because within narrative context, the standard relative future/prospective past force of the verbal form after ‫ ֶט ֶרם‬routinely (though not always; see below) overrides the call for explicit encoding of perfective past semantics, which are contextually inferred.5 Significantly, a relative future/prospective account of ṭɛrɛm yiqṭol not only explains the otherwise anomalous use of yiqṭol in reference to perfective past eventualities, as in example (8), but is consistent with yiqṭol for future/modal, generic present, and past habitual force, as in examples (1)–(3), above. In all cases, the relationship between the eventuality conveyed by the prefix 5 While the most natural rendering of relative future yiqṭol in many lan- guages, including after ‫ ֶט ֶרם‬and ‫עד‬, ַ is by means of a perfective past form, this is by no means universal. For example, JM (342, §113j and fn. 21) note that Jerome favoured a subjunctive alternative in the Vulgate. Whatever the case may be, analysis of BH verbal semantics should seek maximal Hebrew-internal semantic consistency. 324 The Historical Depth of the Tiberian Reading Tradition conjugation following ‫ ֶט ֶרם‬is posterior (i.e., future) relative to the contextual reference time of the verb in the main clause, while other TAM values must be contextually construed. A relative future/prospective past explanation for cases of ṭɛrɛm yiqṭol where the prefix conjugation refers to a perfective past eventuality also justifies the explicit use of morphologically long yiqṭol (< yaqtulu/a), against the claim of some (see above) that the form in question derived from archaic preterite short yiqṭol (< yaqtul) the original length distinction of which was lost. 2.0. The Minority Syntax: Ṭɛrɛm Qaṭal On four occasions in Masoretic BH a verb in a ‫ ֶט ֶרם‬construction referring to a perfective past eventuality comes in the qaṭal rather than yiqṭol pattern: Gen. 24.15; 1 Sam. 3.7a; Ps. 90.2; Prov. 8.25. Before a detailed treatment of each of these passages, it is opportune to take a step back for perspective on ‫ ֶט ֶרם‬constructions within and beyond BH. 2.1. Diachronic Considerations First, it is worth noting that the four exceptional examples of ṭɛrɛm qaṭal in BH do not congregate in any one portion of Scripture. Two are in narrative sections generally regarded as CBH (Genesis and Samuel), one is in poetry (Psalms), and one comes in Wisdom literature (Proverbs). 2.1.1. Tiberian Late Biblical Hebrew None comes in LBH. Indeed, no Masoretic verbal construction employing ‫—ט ֶרם‬with ֶ qaṭal or yiqṭol—is to be found in LBH. 14. Ṭɛrɛm Qaṭal 325 2.1.2. Rabbinic Hebrew The particle ‫ ֶט ֶרם‬is also absent from Tannaitic literature. 2.1.3. The Dead Sea Scrolls More helpful are the data from the DSS. While in the BDSS verb forms after ‫ ֶט ֶרם‬match their Masoretic counterparts, in the NBDSS there is no trace of ṭɛrɛm yiqṭol where the verb refers to a perfective past eventuality, against seven apparent cases of perfective past ṭɛrɛm qaṭal. Assuming the correctness of the readings, examples (9)–(15) appear to be instances of ṭɛrɛm qaṭal, though several are also interpretable as ṭɛrɛm + infinitive construct. (9) ‫ובטרם נוסדו ידע את מעשיהם‬ ‘before they were established, he knew their deeds’ (CD 2.7–8) (10) ‫]עו֯ ֯ל ֯ם לשפוט בם ֯את כול מעשיך בטרם‬ ֯ ‫הכ[ינותה מקדם‬ ֯ ‫ואלה אשר‬ ‫בראתם‬ ‘And it is these which you pre[pared from ancient] eternity to judge, all your works before you created them’ (1QHa 5.24–25) (11) ‫פעול]תו הכינותה בטרם בראתו‬ ֯ ‫ואדעה כי בידך י֯ ֯צר כול רוח [וכול‬ ‘But I know that in your hand is the inclination of every spirit [and all] his [acts] you had prepared before you created him’ (1QHa 7.21–22) (12) ‫ובטרם בראתם ידעתה {כול} מעשיהם‬ ‘and before You created them You knew {all} their works’ (1QHa 9.9) 326 The Historical Depth of the Tiberian Reading Tradition (13) ‫טרם הייתם‬ ‘before you (MPL) were (?)’ (4Q176 f22.3)6 (14) ‫פעולות[יהם‬ ֯ ‫כטרם בראם הכין‬7 ‘Before he created them, he established [their] workings’ (4Q180 f1.2) (15) ‫מחשב[ותיהם‬ ֯ ‫בטרם בראם ידע‬ ‘before he created them, he knew [their] design[s]’ (4Q180 f2–4ii.10) 2.1.4. Ben Sira To these examples should be added one from the concluding poem of BS, preserved in 11QPsa (11Q5). (16) ‫אני נער בטרם תעיתי ובקשתיה‬ ‘I was a youth before I wandered and I found her.’ (11Q5 21.11 = Sir. 51.13) These are striking evidence of a late preference for ṭɛrɛm qaṭal over ṭɛrɛm yiqṭol, perhaps to be explained—along with Hendel (1996, 160, fn. 36)—as due to “the loss of the relative future (as with the whole relative tense system) in LBH, where ʾaz, ṭerem, and ʿad in the past frame are consistently followed by the Pf.”8 6 Cf. infinitival ‫‘ בטרם היותם‬before they were (lit. before their being)’ (1QHa 9.30). 7 Cf. infinitival ‫הבראם‬ ̇ ‫‘ בטרם‬before their creation (lit. before their being created)’ (4Q215a f1ii.9) 8 The comparison with ‫ ַעד‬+ verb in past contexts is apposite, but the relevance of ‫ ָאז‬+ verb is questionable. Notwithstanding approaches that lump together constructions composed of the particles ‫אז‬, ָ ‫ט ֶרם‬, ֶ and 14. Ṭɛrɛm Qaṭal 327 Whereas in past contexts the particle )‫(א ֶשר‬ ֲ ‫ ַעד‬is not followed by yiqṭol in LBH (except where paralleled in SBH9), it is followed by qaṭal.10 If Hendel is correct, then it is possible that Second Temple Aramaic played a role in the post-exilic substitution of qaṭal for yiqṭol after ṭɛrɛm. The typical Targumic equivalent of BH ṭɛrɛm yiqṭol is ‫ עד לא‬followed by the suffix conjugation.11 The Syriac equivalents are ‫ܥܕܟܝܠ ܠ ܐ‬, consistently followed by the suffix conjugation, and ‫ܥܕܠ ܐ‬, followed by prefix or suffix conjugation.12 In both structures, a particle meaning ‘until’ precedes a negated verb, equivalent in English to ‘as long as not…’; cf. Latin necdum ‫ ַעד‬followed by a verb referring to the perfective past (e.g., Hendel 1996; Arnold and Choi 2003, 60), it is best to distinguish cases of relative future yiqṭol after ‫ ֶט ֶרם‬and ‫ ַעד‬from the past-tense use of yiqṭol after ‫ָאז‬ (JM, 341–42, §113i–j; Cook 2012, 262), which, despite notable attempts at elucidation (Bergsträsser §7c, g; Rundgren 1961, 97–101; Rabinowitz 1984; Waltke and O’Connor 1990, §31.6.3; Hendel 1996, 160), remains enigmatic. Also, while Hendel (1996, 160, fn. 36) is broadly correct on the LBH loss of relative-future yiqṭol after ‫עד‬, ַ Cohen (2013, 151–53) identifies a few examples. 9 2 Chron. 21.10 (|| 2 Kgs 8.22). 10 Dan. 11.36; 2 Chron. 9.6; 36.21. 11 See TO to Gen. 2.5, 5; 19.4; 24.45; 41.50; Exod. 12.34; Num. 11.33; TJ to Josh. 2.8; 1 Sam. 3.3, 7b; Jer. 1.5, 5. ‫ עד לא‬+ prefix conjugation and ‫ עד לא‬+ infinitive are also attested. In BH ‫ ַעד לֹא‬+ qaṭal occurs only here in Prov. 8.25 and in Deutero-Isaiah’s Isa. 47.7, where the corresponding text in 1QIsaa 39.26 reads ‫ עוד לוא‬rather than ‫עד לֹא‬. ַ 12 ‫ܥܕܟܝܠ ܠ ܐ‬: Gen. 2.5, 5; 19.4; 24.15, 45; 1 Sam. 3.3, 7, 7; ‫ ܥܕܠ ܐ‬+ qṭl: Num. 11.33; Ps. 119.67; ‫ ܥܕܠ ܐ‬+ yqṭl: Gen. 37.18; 2 Kgs 6.32; Isa. 48.5; Jer. 1.5, 5; 47.1; Ps. 90.2; Prov. 8.24, 26. 328 The Historical Depth of the Tiberian Reading Tradition followed by a past-tense verb. Similar Hebrew ‫ עד לא‬constructions come in the NBDSS and other late sources.13 The CBH ṭɛrɛm yiqṭol structure, by contrast, has no negative component, but can be analysed with the basic semantic value of ‘before’. When fol- lowed by yiqṭol, the force is prospective, i.e., relative future ‘before he would come’; when followed by qaṭal, the force is retrospective, i.e., absolute past ‘before he came’. It is entirely possible that the diminished relative future use of yiqṭol, combined with the influence of Aramaic and Aramaic-like conjunctions including a negative and followed by suffix conjugation forms, were factors in the replacement of classical perfective past ṭɛrɛm yiqṭol with ṭɛrɛm qaṭal. As we shall see, however, the evidence is also consistent with the hypothesis of inner-Hebrew development already at work in CBH. Evidence for some sort of logical connection between ‫ֶט ֶרם‬ and Aramaic/Hebrew ‫ עד לא‬and similar negative conjunctions may be gleaned from the apparent synonymy of the three ‫ֶט ֶרם‬ structures in Zeph. 2.2: (17) ‫ף־יְּהוה‬ ֵ֔ ָ ‫רֹון ַא‬ ֙ ‫יכם ֲח‬ ֵֶ֗ ‫מץ ָע ַבר יָׂ֑ ֹום ְּׁב ֶט ֶּ֣רם ׀ לא־יָבּ֣ ֹוא ֲע ֵל‬ ֹ ֶ֖ ‫ְּׁבטֶ ֶרם ֶל ּ֣ ֶדת ֵ֔חֹק ְּכ‬ ‫הוַֽה׃‬ ָ ְּ‫יכם יֶ֖ ֹום ַאף־י‬ ֵֶ֔ ‫ְּׁבטֶ ֶרם לא־יָבּ֣ ֹוא ֲע ֵל‬ ‘before the delivery of the decree, like chaff the day has passed, when the burning anger of the LORD does not yet come upon you, when the day of the anger of the LORD does not yet come upon you.’ (Zeph. 2.2) 13 CD 10.10 (with yiqṭol); 4Q300 f1aii–b.2; Mas1h 2.7 (|| Sira 40.17; cf. SirB 10r.8). Significantly, other alternatives, also employing the suffix rather than prefix conjugation, likewise appear in late corpora, e.g., ‫ֲע ֶ ֶ֖דן‬ ‫( ל ֹא‬Qoh. 4.3), ‫עדיין לא‬/‫( אדיין‬m. Yadayim 4.4), ‫( קודם עד שלא‬y. Berakhot). 14. Ṭɛrɛm Qaṭal 329 Here the initial ‫ ֶט ֶרם‬structure containing an infinitive construct has approximately the same meaning as the two subsequent ‫ֶט ֶרם‬ constructions with negated yiqṭol forms. These all have absolute future, rather than past, semantics, but the crucial point is that the standard future-oriented ṭɛrɛm yiqṭol construction with no negative connotation or particle, probably with the force ‘before X will/does’, has acquired negative morphology and semantics, apparently with the revised force ‘when X does not yet’.14 In light of the evidence, it would seem that the particle ‫ֶט ֶרם‬ had become somewhat obsolete in Second Temple Hebrew and that when late writers employed it, they were more prone than their predecessors to opt for qaṭal over yiqṭol in reference to perfective past eventualities. Be that as it may, on the surface, the ostensible diachronic shift from ṭɛrɛm yiqṭol to ṭɛrɛm qaṭal discernible when comparing BH (whether Masoretic or DSS) to the Hebrew of the NBDSS finds no confirmation in perceptible dia- 14 Similar phenomena are known in Hebrew and crosslinguistically. For example, ‫‘ עֹוד‬while’ versus ‫‘ ַעד‬until’ in Hebrew (cf. Job 1.16–18); post- classical ancient Hebrew ‫‘ עד לא‬not yet’ parallels Modern Hebrew ‫עוד‬ ‫ ;לא‬in vernacular Italian, the construction finché non ‘until’ is routinely shortened to its logical opposite finché ‘as long as’. French avant qu’il ne vienne ‘before he comes’ seems to include a superfluous negative parti- cle. It has been suggested that ‘before’, with a basic sense of ‘when still not’, is inherently negative. Relatedly, in English ‘before’ licenses negative polarity items, e.g., ‘before they saw anyone’. I am grateful to Ambjörn Sjörs for noting many of the above points. See Hetterle (2015, 131–51)—kindly referred to me by Christian Locatell—for crosslinguistic perspective on the intersection of tense, sspect, and negation in adverbial clauses. 330 The Historical Depth of the Tiberian Reading Tradition chronic distribution within the MT, in that LBH exhibits no cases of ṭɛrɛm qaṭal (or of ‫ ֶט ֶרם‬more generally). Even so, there may be evidence, albeit both limited and arguable, of the shift in question in cases of apparent dissonance between the written (consonantal) and reading (vocalisation) components of the Tiberian biblical tradition, the latter showing slight drift towards the purported Second Temple convention. Crucially, whereas in nearly all instances of perfective past ṭɛrɛm yiqṭol, the consonantal text allows for no reading other than that of a prefix conjugation, in a tiny minority of cases, orthographic ambiguity allows for a secondary ṭɛrɛm qaṭal reading. But such reanalysis accounts for only a portion of the ṭɛrɛm qaṭal exceptions; it would seem that others are genuine classical outliers. 2.2. Secondary ṭɛrɛm qaṭal in the Tiberian Reading Tradition In two cases of ṭɛrɛm qaṭal in the Masoretic Hebrew Bible, a compelling argument can be made that the qaṭal forms reflected in the reading tradition are secondary. Both cases involve I-y qal verbs, the consonantal forms of which may well have been intended to represent more standard yiqṭol alternatives. 2.2.1. 1 Sam. 3.7 (18) ‫הוַֽה׃‬ ָ ְּ‫הוָׂ֑ה וְּ ֶט ֵֶ֛רם יִ ג ֶָלּ֥ה ֵא ָלֶ֖יו ְּד ַבר־י‬ ָ ְּ‫מּואל ֶט ֶֹׁ֖רם י ַָּדּ֣ע ֶאת־י‬ ֵָ֕ ‫ּוש‬ ְּ ‘Now Samuel did not yet know the LORD and the word of the LORD would yet be revealed to him.’ (1 Sam. 3.7) This well-known example helpfully presents two instances of ‫ ֶט ֶרם‬+ verb: the anomalous ṭɛrɛm qaṭal in the first half of the 14. Ṭɛrɛm Qaṭal 331 verse and the more common ṭɛrɛm yiqṭol in its second half. The grammatical mismatch is conspicuous. The accepted—and arguably most compelling—explanation for the instance of ṭɛrɛm qaṭal assumes secondary divergence of the recitation tradition from the tradition presupposed by the consonantal text, presumably under the influence of Second Temple Hebrew. As has been proposed by many (e.g., Driver 1890, 34), it is likely that the consonants ‫ ידע‬here were originally intended to represent a yiqṭol form expected to yield Tiberian ‫יֵ ַדע‬, but were read— presumably in line with later grammar, like that of the NBDSS Hebrew cases cited above in (9)–(15)—as qaṭal ‫יָ ַדע‬. Certainly, the conjectural yiqṭol ‫ יֵ ַדע‬is a better match than qaṭal ‫ יָ ַדע‬for the accompanying yiqṭol ‫ יִ גָ ֶלה‬later in the verse, as well as for the majority of other cases of ṭɛrɛm yiqṭol in reference to perfective past eventualities. An important consideration relevant to this example is that the proposed modification to the oral realisation would have been facilitated by the graphic identity of the I-y qal qaṭal and yiqṭol consonantal forms, in this case ‫ ַיָדע‬and ‫יֵ ַדע‬, respectively, so that the change would have occasioned no violence to the consonantal text. This is broadly characteristic of other cases of dissonance between the written and reading components of the Tiberian tradition—secondary linguistic features standard in Second Temple Hebrew supplanted their First Temple counterparts where the ambiguity of the consonantal tradition made it amenable to substitute realisations. Indeed, not even was an explicit marking of ketiv-qere necessary. 332 The Historical Depth of the Tiberian Reading Tradition A comparable phenomenon took place more generally in the case of I-y qal wayyiqṭol forms in the Samaritan reading tradition, where Tiberian wayyiqṭol forms, like ‫‘ וַ יֵ ֵֶּ֣רד‬and he went down’ (Deut. 26.5), were re-analysed as perfective conjunctive ̊̄ waw+qaṭal forms, like ‫ וירד‬wyaråd. So pervasive was the penetration of qaṭal morphology, that it was applied even to feminine I-y qal forms, e.g., MT ‫‘ וַ ֵת ֶלד‬and she gave birth’ (Gen. 4.1) || SP ̊̄ ‫ ותלד‬wtalåd (Khan 2021, 331; cf. Ben-Ḥayyim 2000, 170, 173).15 It is reasonable to assume that the Samaritan reading of original I-y qal forms in ṭɛrɛm yiqṭol might also have been along the lines of ṭɛrɛm yiqṭol, but this must remain conjecture, as the Pentateuch presents no cases of perfective past ṭɛrɛm yiqṭol with a I-y qal verb (likewise for perfective past I-y qal yiqṭol following ‫ ָאז‬and ‫ַעד‬ ]‫)[א ֶשר‬. ֲ In light of the morphological mismatch between ‫ֶ ֶ֖ט ֶרם יָ ַדע‬ and ‫ ֶ ֛ט ֶרם יִ גָ ֶלֵ֥ה‬in 1 Sam. 3.7, a local explanation for the anomalous use of the characteristically late ṭɛrɛm qaṭal structure predicated on the Tiberian reading tradition’s secondary divergence from the written tradition seems persuasive. Given this, one is primed for similar explanations in the case of the remaining tokens of ṭɛrɛm qaṭal. However, while a similar explanation might hold for one other case, and while all could conceivably be chalked up to textual fluidity in the consonantal tradition, the possible authen15 Ben-Ḥayyim (2000, 173) accepts this explanation for 3MS and 3MPL wayyiqṭol forms, but not for 2MS and 3FS wayyiqṭol forms, which he sees ̊̄ as yiqṭol forms with an a-vowel preformative reflecting original yafʿul; ̊̄ cf. Khan (2021, 331), who sees SP forms like ‫ ותלד‬wtalåd as secondary forms that developed on the analogy of qaṭal for purposes of distinguish̊̄ ̊̄́ ing preterite yiqṭol (e.g., wtåråd) from non-preterite yiqṭol (e.g., téråd). 14. Ṭɛrɛm Qaṭal 333 ticity of one or more of the remaining three cases tallies with early evidence of other secondary vocalisation features that represent standardisations of early minority options. In other words, the fact that a single case of ṭɛrɛm qaṭal is compellingly explained as a late secondary vocalic deviation from the presumed recitation of the written tradition in line with Second Temple conventions does not mean that all similar structures should be so explained. 2.2.2. Ps. 90.2 Another case of ṭɛrɛm qaṭal occurs in the poetry of Ps. 90.2: (19) ‫ֹולם ַא ָ ֵ֥תה ֵ ַֽאל׃‬ ָ ֵ֗ ‫ד־ע‬ ֶ֜ ‫עֹולם ַע‬ ֵ֥ ָ ‫ֹולל ֶא ֶרץ וְּ ֵת ֵ ָׂ֑בל ּוַֽ ֵמ‬ ַֽ ֵ ‫ֻׁלדּו וַ ְּתח‬ ָ ֶ֗ ‫ְּׁב ֶט ֶָּ֤רם׀ ָָֹ֨ה ִ ַ֤רים י‬ ‘Before the mountains were brought forth, or ever you had formed the earth and the world, from everlasting to everlasting you are God.’ (Ps. 90.2) The form ‫ יֻ ֵ֗ ָלדּו‬appears to be a puʿʿal form of the suffix conjugation, internal passive of either piʿʿel or—more likely from a semantic perspective—qal.16 The qal internal passive is itself the focus of a well-known case of divergence between the Tiberian consonantal and reading traditions (ch. 10, §§1.1.2; 2.2; 3.2). Even if the middle-radical doubling in such forms can be explained as organic secondary gemination for preservation of the characteristically passive short u vowel, it is suspicious that such qal passives are preserved chiefly where reinterpretation as alternative passive 16 Since the piʿʿel form is used exclusively in BH as a substantive in the meaning ‘midwife’: Gen. 35.17, 28; Exod. 1.15–21. 334 The Historical Depth of the Tiberian Reading Tradition patterns (suffix conjugation puʿʿal and prefix conjugation hofʿal/ hufʿal) was possible,17 but are otherwise realised as nifʿals. Thus, 1. ostensibly puʿʿal suffix conjugation ‫ גֻ נַ ב‬pairs with nifʿal prefix conjugation ‫( יִ גָ נֵ ב‬rather than ‫;)*יֻ גְּ נַ ב‬ 2. qal passive participle (or ostensibly puʿʿal participle without the expecting preformative -‫ ֻא ָכל )מ‬corresponds to puʿʿal (i.e., piʿʿel internal passive) prefix conjugation ‫יְּ ֻא ַכל‬ (rather than ‫יָ ֳא ַכל‬/‫;)*יֻ ְּא ַכל‬18 and 3. ostensibly hofʿal prefix conjugation ‫ יֻ ַתן‬parallels nifʿal suffix conjugation ‫( נִ ַתן‬rather than ‫)*נֻ ַתן‬. The problem is not the authenticity of alternatives for the qal internal passive, since, for example, consonantally unambiguous nifʿal forms are sometimes documented alongside apparent qal passives in classically-worded texts (ch. 10, §3.0).19 The issue is rather the near total absence of qal passive forms where the consonantal text permitted an alternative reading—a situation difficult to interpret as anything other than systemic dissonance in realisation between the pronunciation tradition presupposed by the consonantal orthography and that of the recitation tradition. 17 Exceptions include qal internal passive participles, e.g., ‫‘ ֵא ֶינֵ֥נּו ֻא ָ ַֽכל‬was not being consumed’ (Exod. 3.2); ‫יּולד‬ ַֽ ָ ‫‘ ַל ַנ ֵַ֥ער ַה‬to the child being born’ (Judg. 13.8); ‫ם־ת ְּר ֶ֙אה א ִֶֹ֜תי ֻל ָ ַ֤קח ֵ ַֽמ ִא ָת ְ֙ך‬ ִ ‫‘ ִא‬if you see me being taken from you’ (2 Kgs 2.10). 18 BH knows know piʿʿel ‫;א ֵכל‬ ִ cf. piʿʿel ‫ אכל‬in the Samaritan reading tra- dition and piʿʿel ‫אכל‬/‫ עכל‬in Amoraic Hebrew, as well as puʿʿal ‫ אכל‬in Tannaitic Hebrew. 19 Consider the nifʿal ‫( יִ נָ ֵ ַָֽקם‬Exod. 21.20) and the qal passive (apparently hofʿal) ‫( יֻ ַ ֵ֔קם‬Exod. 21.21) both ‘will be avenged’ in successive verses. 14. Ṭɛrɛm Qaṭal 335 The structure ‫ ְּב ֶ ַ֤ט ֶרם׀ ָָֹ֨ה ִ ַ֤רים יֻ ֵ֗ ָלדּו‬in Ps. 90.2 presents opposing diachronic tendencies. On the one hand, as noted above, the ṭɛrɛm qaṭal syntagm appears to have late affinities. On the other hand, qal passive ‫ יֻ ַלד‬is characteristically classical. Note that in terms of unambiguous consonantal spellings, forms of qal internal passive ‫( יֻ ַלד‬qaṭal) are confined chiefly to CBH, whereas forms of nifʿal ‫נֹולד‬ ַ (qaṭal, participle, infinitive construct), though docu- mented in CBH, appreciably accumulate in LBH.20 Orthographically, the relevant yiqṭol forms, e.g., ‫יולד‬, are generally ambig- uous, but are consistently vocalised as nifʿal.21 The lone exception is the subject of a ketiv-qere mismatch in 2 Sam. 3.2. (20) ...‫וַּיִ ּו ְָּׁל ֧דּו [וילדו] ְּל ָדִו֛ד ָב ִנֶ֖ים ְּב ֶח ְּב ָׂ֑רֹון‬ ‘And sons were born to David at Hebron…’ (2 Sam. 3.2) It is likely here that the ketiv ‫ וילדו‬reflects an original qal internal passive wayyiqṭol, along the lines of ‫ *וַ יֻ ְּלדּו‬wayyullǝdū,22 and that the synonymous qere ‫ וַ יִ ּוָ ְּל ָ֧דּו‬is a secondary linguistic update in line 20 Qal internal passive ‫ יֻ ַלד‬qaṭal: Gen. 4.26; 6.1; 10.21, 25; 24.15; 35.26; 36.5; 41.50; 46.22, 27; 50.23; Judg. 18.29; 2 Sam. 3.5; 21.20, 22; Isa. 9.5; Jer. 20.14–15; 22.26; Ps. 87.4–6; 90.2; Job 5.7; Ruth 4.17; 1 Chron. 1.19; nifʿal ‫נֹולד‬ ַ qaṭal, participle, infinitive construct: Gen. 21.3, 5; 48.5; 1 Kgs 13.2; Hos. 2.5; Ps. 22.32; Qoh. 4.14; 7.1; Ezra 10.3; 1 Chron. 2.3, 9; 3.1, 4–5; 7.21; 20.6, 8; 22.9; 26.6. 21 The dominant spelling with waw certainly facilitated nifʿal reinterpre- tation. However, even in the case of a I-y qal internal passive yiqṭol, the spelling with waw is expected, e.g., ‫יולד‬, as in ‫ יוכל‬and ‫תוקד‬, resulting from contraction of the diphthong uw, i.e., yūlad < yuwlad. 22 The lack of the expected mater waw, though rare, is more common in forms with suffixes, e.g., the plural here. 336 The Historical Depth of the Tiberian Reading Tradition with the Second Temple preference for nifʿal over qal internal passive in the case of the prefix conjugation. How does this shed light on the spelling ‫ ילדו‬in Ps. 90.2 in example (19)? Obviously, as spelled, it was not amenable to simple re-analysis as a nifʿal yiqṭol, i.e., without resorting to overt signalling of a ketiv-qere mismatch. So, then, why was the ketivqere mechanism left unexploited here? A plausible explanation is that the spelling ‫ ילדו‬in Ps. 90.2, following as it does the particle ‫ט ֶרם‬, ֶ was originally intended as a yiqṭol form. However, unlike in 1 Sam. 3.2, where the wayyiqṭol form could not be reanalysed as a conjunctive we+qaṭal form, the ṭɛrɛm yiqṭol structure ...‫ְּב ֶ ַ֤ט ֶרם‬ ‫ יֻ ֵ֗ ָלדּו‬in Ps. 90.2 was ripe for easy reanalysis, as both the prefix and suffix conjugation of the relevant qal internal passive verb could be written ‫ילדו‬. Original ṭɛrɛm yiqṭol was simply reinterpreted as ṭɛrɛm qaṭal. The phrase ‫ ְּב ֶ ַ֤ט ֶרם׀ ָָֹ֨ה ִ ַ֤רים יֻ ֵ֗ ָלדּו‬in Ps. 90.2 thus represents both secondary development—replacing classical ṭɛrɛm yiqṭol with ṭɛrɛm qaṭal—and classical preservation—the incidental persistence of characteristically classical qal internal passive ‫ יֻ ַלד‬in the face of the encroachment of nifʿal yiqṭol ‫ יִ ּוָ ֵלד‬or qaṭal ‫נֹולד‬. ַ To summarise: while the form ‫ יֻ ֵ֗ ָלדּו‬as realised according to the Tiberian recitation tradition is analysable as a qaṭal form in the characteristically late ṭɛrɛm qaṭal syntagm, its spelling may well represent that of a yiqṭol form in the classic ṭɛrɛm yiqṭol structure. Regardless of the validity of the arguments laid out above, two further factors may have contributed to the ṭɛrɛm qaṭal rather than ṭɛrɛm yiqṭol construction. First, the context is poetic. Though the poetry-prose linguistic distinction in ancient Hebrew is sometimes abused, it may help to explain the deviation from the stand- 14. Ṭɛrɛm Qaṭal 337 ard ṭɛrɛm yiqṭol construction here. Second, it is important to note that the syntagm employed in Ps. 90.2 is not precisely ṭɛrɛm + verb, but ṭɛrɛm + X + verb. The interruption of the syntagm due to the intervening constituent ‫ ָָֹ֨ה ִ ַ֤רים‬may have facilitated variation in the ensuing verbal form. Both factors—non-prose genre and interruption of the syntagm—also apply to the case discussed below, §2.3.1. 2.3. Original Ṭɛrɛm Qaṭal in the Tiberian Reading Tradition While evidence for the late secondary character of the two forms above may be compelling, there is no reason to reject the possibility of the non-secondary use of ṭɛrɛm qaṭal in BH. Indeed, despite the decidedly minority status of the two following biblical examples, and notwithstanding the fact that unambiguous extrabiblical evidence for ṭɛrɛm qaṭal is limited to Second Temple sources (the NBDSS), there seems no reason a priori to question the authenticity of the cases below or of the formulaic diversity they represent. 2.3.1. Prov. 8.25 (21) ‫חֹול ְּל ִתי׃‬ ַֽ ָ ‫ְּׁב ֶט ֶּ֣רם ָה ִרים הָ ְּׁט ָבָ֑עּו ִל ְּפ ֵנֶ֖י גְּ ָבעֹות‬ ‘before the mountains were settled in place, before the hills, I was given birth…’ (Prov. 8.25) Here, as in Ps. 90.2 (see above, §2.2.2), the noun ‫‘ ָה ִ ַ֤רים‬mountains’ follows ‫ ְּב ֶט ֶרם‬and precedes a passive verb denoting their origin. As has already been suggested, it is possible that the interrupted nature of the ṭɛrɛm + verb structure facilitated the use of qaṭal 338 The Historical Depth of the Tiberian Reading Tradition rather than yiqṭol. The literary and notional similarities between Ps. 90.2 and Prov. 8.25 are also evident. Whatever the case may be, accepting the text as is, ‫ ָה ְּט ָ ָׂ֑בעּו‬clearly cannot be analysed as anything other than a form of the suffix conjugation, i.e., there are no grounds for claiming that the ṭɛrɛm qaṭal structure here results from dissonance between the written and reading components of the Tiberian tradition. There are several factors that may have contributed to the use of a non-standard syntactic structure here. Beyond the interrupted nature of the syntagm, there is also the question of genre. Wisdom literature, though different from biblical poetry, nevertheless exhibits its own non-prose traits. One noted feature, probably due in part to its pan-national Ancient Near Eastern character, is its affinity for forms redolent of Aramaic (Hornkohl 2013a, 17). Indeed, in the Hebrew Bible there are four contexts in which Aramaisms are expected: LBH, due to language contact during and after the Exile; poetry, due to, inter alia, the need for lexical variation between common and rarer words (the B-words often being characteristic of Aramaic); stories set in foreign contexts or featuring foreigners, in which Aramaic forms are employed for ‘style switching’; and Wisdom literature (Stadel 2013). Regarding the specific construction under examination here, it is of crucial importance to point out that the language of Prov. 8 is replete with non-standard forms, a few especially characteristic of Aramaic.23 Of special interest here is ‫חּוצֹות‬ ָׂ֑ ְּ‫‘ ַעד־ל ֹא ָ ָ֭ע ָשה ֶא ֶרץ ו‬be- 23 E.g., ‫‘ ִמ ְּפ ָתח‬opening’ (v. 6), ‫‘ א ַֹרח‬way’ (v. 20), ‫‘ ִמ ְּפ ָעל‬deed’ (v. 22), ‫ַעד‬ ‫‘ לֹא ָע ָשה‬before he had made’ (v. 26), ‫‘ ָאמֹון‬craftsman’ (v. 30). The exclu- 14. Ṭɛrɛm Qaṭal 339 fore he had made earth and fields’ in the immediately following v. 26, since ‫ עד לא‬+ the suffix conjugation is a common Targumic rendering of BH perfective past ṭɛrɛm yiqṭol (see above, §2.1). Finally, there is the factor of grammatical attraction. In the immediate literary context, comprised of vv. 22–26, each verse begins with the structure X qaṭal, where X is either subject or adverbial. There are therefore multiple factors potentially contributing here to the choice of the suffix conjugation rather than the prefix conjugation after ṭɛrɛm, but little justification for doubting the textual authenticity of the ṭɛrɛm qaṭal syntagm. 2.3.2. Gen. 24.15 The only remaining case of ṭɛrɛm qaṭal in the Hebrew Bible comes in Gen. 24.15. (22) ...‫י־הּוא טֶ ֶרם ֮ ִכ ָל ּ֣ה ְּל ַד ֵבר֒ וְּ ִה ֵנָ֧ה ִר ְּב ָָקה י ֵֵֹ֗צאת‬ ֵ֗ ‫ַוֵַּֽ֣יְּ ִה‬ ‘And he was—before he finished speaking, and here Re- bekah… was coming out’ (Gen. 24.15) This instance comes in the narrator’s description of Abraham’s servant’s search for a wife for Isaac. Complicating any explanation of the minority construction here is the near-parallel verse with the majority ṭɛrɛm yiqṭol construction in the 1st-person account later in the chapter. sive use of ‫‘ ֲאנִ י‬I’, though not limited to Aramaic-like Hebrew, can also be interpreted as fitting Aramaic patterns. 340 The Historical Depth of the Tiberian Reading Tradition (23) ...‫את‬ ֙ ‫ל־ל ִֵ֗בי וְּ ִהנֵ֙ ה ִר ְּב ָ ַָ֤קה י ֵֹצ‬ ִ ‫לה ְּל ַד ֵבר ֶא‬ ֶ ָ֜ ַּ‫ֲאנִ י֩ טֶ ֶרם אֲכ‬ ‘Before I would finish speaking in my heart, and here was Rebekah coming out…’ (Gen. 24.45) This case of ṭɛrɛm qaṭal shows some similarity to that in 1 Sam. 3.7 (above, §2.2.1), in that there is internal inconsistency with an instance of ṭɛrɛm yiqṭol in the same context. And, indeed, it has been suggested that the qaṭal form ‫ ִכ ָלה‬in Gen. 24.15 should be considered an error for ‫( יְּ ַכ ֶלה‬GKC 1910, §107c). There are also, however, differences between 1 Sam. 3.7 and Gen. 24.15. Because the crux in 1 Sam. 3.7 involves a I-y qal verb, the purported shift from yiqṭol to qaṭal there is limited to vocalic realisation. In Gen. 24.15, conversely, the written and reading components of the Tiberian tradition agree on ṭɛrɛm qaṭal. What is more, while the evidence of the Ancient Versions is, as is generally the case, opaque with regard to verbal form in this verse, the combined Samaritan consonantal and recitation tradition joins the MT in exhibiting the mismatch between ṭɛrɛm qaṭal in Gen. 24.15 and ṭɛrɛm yiqṭol in Gen. 24.45—this despite the Samaritan tradition’s well-known harmonistic penchant. If ‫ֶט ֶר ֮ם ִכ ָלה‬ in Gen. 24.15 is a secondary development, it must be one of considerable depth, predating the divergence of the proto-Masoretic and proto-Samaritan traditions. Assuming the genuineness of the structure in Gen. 24.15, it is reasonable to ask if such a non-standard use can be explained. Cook (2012, 262, fn. 96) argues that the difference centres on the foregoing use of ‫וַ יְּ ִהי‬: In this case, the discourse ‫ …וַ יְּ ִהי‬sets the narrative deictic center in the past (Cpos1) and the qatal in the past context shifts the time back one step further (CRF) to express a past- 14. Ṭɛrɛm Qaṭal 341 in-the-past (past perfect): [CRF < Cpos1 < S]. The participle, expressing a progressive event, is then indicated as intersecting the past perfect action by the adverbial ‫ט ֶרם‬. ֶ Even if Cook’s rendering of ‫י־הּוא‬ ֵ֗ ‫ֵַּֽ֣יְּה‬ ִ ‫ ַו‬as ‘It happened’ is acceptable,24 the claim that temporal ordering of pluperfect ‫ ֶט ֶר ֮ם ִכ ָלה‬relative to simple past ‫ ַוֵַּֽ֣יְּ ִהי־‬is responsible for ṭɛrɛm qaṭal is puzzling. The temporal ordering of ‫ ַוֵַּֽ֣יְּ ִהי־‬and ֒‫ ִכ ָלה ְּל ַד ֵבר‬is irrelevant to the narrative; the emphasis is rather on the order of ‫את‬ ֙ ‫ ִר ְּב ָ ַָ֤קה י ֵֹצ‬and ֒‫ ִכ ָלה ְּל ַד ֵבר‬: while the progressive aspect of the former precludes use of the pluperfect, the ordering is clear: ‘before he finished speaking… and here Rebekah was coming out’, which could be paraphrased as ‘before he finished praying, Rebekah had already appeared’. BH ‫‘ ֶט ֶרם‬before’ explicitly signals the situation prior to the ensuing verb, whether yiqṭol or qaṭal. It also bears noting that no other biblical or extra-biblical cases of ṭɛrɛm qaṭal are conditioned by a preceding ‫יְּהי‬ ִ ַ‫ו‬. It thus seems that there is nothing peculiar to the syntax of Gen. 24.15 that requires ṭɛrɛm qaṭal instead of ṭɛrɛm yiqṭol. Turning to another line of argumentation, in three separate publications Alexander Rofé (1976; 1981; 1990) has argued, on the basis of a series of non-standard, especially Aramaic, linguistic usages, that Genesis 24 is a post-exilic composition. Though ṭɛrɛm qaṭal is not among the Aramaisms he lists, given the construction’s comparative frequency in late extra-biblical sources, as well as the late distribution of synonymous Hebrew and Aramaic constructions employing the suffix conjugation, an argu- 24 Cf. Driver (1892, §165 Obs) on the Masoretic accentuation, which the English glossing in (21) is intended to reflect. 342 The Historical Depth of the Tiberian Reading Tradition ment involving the chapter’s late provenance might neatly account here for ṭɛrɛm qaṭal, which could then be seen as an anachronistic deviation from the standard classicism ṭɛrɛm yiqṭol later in the same chapter. Gary Rendsburg (2002; 2006) is sensitive to the non-standard linguistic features detected by Rofé, but interprets them differently. Since it is specifically the accumulation of diagnostically late Aramaisms, not the mere concentration of Aramaic(-like) features, that demonstrates post-exilic provenance (Hurvitz 1968; 2003), Rendsburg argues for a literary rather than diachronic explanation for these in Genesis 24—namely that the writer engaged in style switching, intentionally employing foreign-sounding phraseology to reflect the story’s foreign setting. Rendsburg does not list ‫ ֶט ֶר ֮ם ִכ ָלה‬as a non-standard linguistic feature requiring explanation, but in light of the foregoing discussion, in which both diachronic and foreign factors have been mentioned, perhaps the syntagm bears reinvestigation. For if either Rofé or Rendsburg is correct, the construction in question, like the three cases of ṭɛrɛm qaṭal already discussed, could perhaps be considered a conditioned exception to the ṭɛrɛm yiqṭol norm—though the mismatch between vv. 15 and 45 is, admittedly, left unexplained. While the considerations above might help to explain the appearance of ṭɛrɛm qaṭal in Gen. 24.15, it is perhaps preferable here simply to accept the possibility of early grammatical diversity, in which case ‫ ֶט ֶר ֮ם ִכ ָלה‬is to be viewed as an early forerunner of the more prevalent use of ṭɛrɛm qaṭal in the NBDSS (see further, below). 14. Ṭɛrɛm Qaṭal 343 3.0. Methodological Considerations In BH, the use of relative future ṭɛrɛm yiqṭol is far more common than the use of absolute past ṭɛrɛm qaṭal. What is more, it seems that one or more cases of ṭɛrɛm qaṭal can be explained as either false positives or conditioned deviations from classical standards. Admittedly, though, the philological issues cited above as factors contributing to the use of qaṭal rather than yiqṭol after ṭɛrɛm are more convincing in some cases than others. The purported shift from ‫*ט ֶרם יֵ ַדע‬ ֶ to ‫ ֶט ֶרם יָ ַדע‬in 1 Sam. 3.7a (above, §2.2.1) is arguably the most compelling. Some of the other arguments ostensibly explaining the use of ṭɛrɛm qaṭal for ṭɛrɛm yiqṭol sound like special pleading. Of course, in the interests of grammatical consistency— i.e., ṭɛrɛm uniformly followed by yiqṭol—some might favour wholesale textual emendation of ṭɛrɛm qaṭal cases. In light of the extrabiblical (NBDSS) and extra-Masoretic (Samaritan) evidence for ṭɛrɛm qaṭal, however, this seems gratuitous. Notwithstanding the repetition of patterns inherent to language, expectation of complete formulaic uniformity is unrealistic. For all their regularity, languages are non-static human products, prone to irregularity. Or, as Sapir (1921, 39) put it, “Unfortunately, or luckily, no language is tyrannically consistent. All grammars leak.” There is no reason to expect that this should apply any less to an ancient language, like BH, representing diverse chronolects, dialects, registers, and genres and transmitted in various traditions, both written and oral, or even to a single unified component variety of BH. Even in the case of a modern homogenous language variety, one expects general linguistic regularity sprinkled with irregularity. Crosslinguistic tendencies may help to explain certain 344 The Historical Depth of the Tiberian Reading Tradition phenomena, but philological approaches may also be relevant. Bringing all these considerations to bear on non-standard Tiberian ṭɛrɛm qaṭal against the backdrop of standard ṭɛrɛm yiqṭol, it is reasonable to conclude that certain cases of ṭɛrɛm qaṭal result from late, secondary discord between the written and reading traditions, while in other cases the two traditions agree on the early authenticity of the syntagm. But if any early cases of ṭɛrɛm qaṭal are genuine, even if they might be contextually conditioned, these constitute precedent for potential later secondary shifts from ṭɛrɛm yiqṭol to ṭɛrɛm qaṭal. In other words, while ṭɛrɛm qaṭal ‫ ֶט ֶרם יָ ַדע‬in 1 Sam. 3.7a is almost certainly the result of secondary reinterpretation of original ṭɛrɛm yiqṭol ‫*ט ֶרם יֵ ַדע‬ ֶ in line with broader Second Temple trends, the early documentation of ṭɛrɛm qaṭal means that any case of late reinterpretation was not completely out of step with classical norms. As frequently obtains in such cases of dissonance between the written and reading components of the Tiberian biblical tradition, a feature especially characteristic of Second Temple Hebrew is foreshadowed by minority classical usage. Thus, if the apparently slight difference in extent of usage of ṭɛrɛm qaṭal between the Tiberian written and reading tradition is explicable as a result of secondary drift of the reading tradition in the direction of Second Temple linguistic convention, the shift does not involve wholly anachronistic innovation, but a slight extension in the use of a minority feature already documented in CBH. Indeed, given the plausible authenticity or one or more of the four cases of ṭɛrɛm qaṭal in the MT, it is not impossible, despite indications to the contrary, that all are authentic. 14. Ṭɛrɛm Qaṭal 345 It is worth making one final point that also tallies with preexilic linguistic diversity. The purported early co-occurrence of majority ṭɛrɛm yiqṭol, encoding relative future, and minority ṭɛrɛm qaṭal, encoding absolute past, is reminiscent of other CBH alternations between yiqṭol and qaṭal. Perhaps most relevant is the relative past usage of qaṭal for retrospective future (or future perfect, futurum exactum) versus the absolute future force of yiqṭol in parallel contexts. Compare the past-within-future qaṭal usages with similar future yiqṭol usages in the following examples. (24a) ...‫תים ֵָ֔שם‬ ּ֣ ִ ‫ים ֲא ֶשר ִה ַּד ְּׁח‬ ֙ ‫ל־ה ְּמק ַֹ֤מֹות ַהנִ ְּש ָא ִר‬ ַ ‫ב ָכ‬... ְּ ‘…in all the places where I have driven them…’ (Jer. 8.3; cf. Jer. 29.14, 18; 32.37; 46.28) (24b) ‫יחּ֥ם ָ ַֽשם׃‬ ֵ ‫ל־ה ְּמק ֶֹ֖מֹות ֲא ֶ ַֽשר־אַּ ִד‬ ַ ‫ב ָכ‬... ְּ ‘…in all the places where I shall drive them.’ (Jer. 24.9) (25a) ‫־לְך׃‬ ַֽ ָ ‫ל־ה ָ ֵ֥א ֶרץ ַהטּ ָ ֶֹ֖בה ֲא ֶ ֵ֥שר נָ ִֽתַּ ן‬ ָ ‫ֹלהיָך ַע‬ ֵֶ֔ ‫ּוב ַר ְּכ ָ֙ת ֶאת־ה' ֱא‬ ַֽ ֵ ‫וְּ ָא ַכ ְּל ָ ֶ֖ת וְּ ָש ָ ָׂ֑ב ְּע ָת‬ ‘And you shall eat and be full, and you shall bless the LORD your God for the good land he has given you.’ (Deut. 8.10) (25b) ...‫ל־ה ֵָ֗א ֶרץ ֲא ֶ֙שר יִ ֵת֧ן ֛ה' ָל ֶכֶ֖ם‬ ָ ‫י־תבֹאּו ֶא‬ ָ ‫וְּ ָה ָָ֞יה ִ ַֽכ‬ ‘And when you come to the land that the LORD will give you…’ (Exod. 12.25) (26a) ‫אֹו־עז ַ ַֽב ַמ ֲח ֶנָׂ֑ה ַ֚אֹו ֲא ֶשר‬ ֶ֖ ֵ ‫אֹו־כ ֶ֛שב‬ ֶ ‫יִש ֶַ֜חט ֵ֥שֹור‬ ְּ ‫יש ִמ ֵבית יִ ְּש ָר ֵֵ֔אל ֲא ֶ֙שר‬ ֙ ‫ִ ֵ֥איש ִא‬ ...‫מֹוע ֘ד ל ֹא ה ֱִביאֹו‬ ֵ ‫א ֶהל‬ ֹ ‫ל־פ ַתח‬ ֶ ֶ֜ ‫יִ ְּש ֵַ֔חט ִמ ֶ֖חּוץ ַ ַֽל ַמ ֲח ֶנַֽה׃ וְּ ֶא‬ ‘If any one of the house of Israel kills an ox or a lamb or a goat in the camp, or kills it outside the camp, and to the entrance of the tent of meeting has not brought it…’ (Lev 17.3–4) 346 The Historical Depth of the Tiberian Reading Tradition (26b) ‫מֹוע ֙ד‬ ֵ ‫א ֶהל‬ ֹ ַ֤ ‫ל־פ ַתח‬ ֶ ֶ֜ ‫אֹו־ז ַַֽבח׃ וְּ ֶא‬ ָ ‫ ֲא ֶשר־יַ ֲע ֶ ֵ֥לה ע ָֹלֶ֖ה‬...‫יש ִמ ֵבית יִ ְּש ָר ֵֵ֔אל‬ ֙ ‫ִ ֵ֥איש ִא‬ ...‫יאנּו‬ ִֶ֔ ‫ל ֹא יְּׁ ִב‬ ‘Any one of the house of Israel… who offers a burnt offering or sacrifice and to the entrance of the tent of meeting does not bring it…’ (Lev. 17.8–9) In cases such as these, involving the intersection of diverse speech, event, and reference times, BH users could opt for temporal encoding that centred on absolute tense posterior to speech time (i.e., absolute future yiqṭol) or retrospective relative tense (i.e., relative past and perfect qaṭal). A similar choice seems to have developed for verbs following ‫ט ֶרם‬, ֶ though in early sources, a relative future, prospective past yiqṭol seems to have dominated the absolute past option qaṭal, the latter becoming more common only in later sources. 4.0. Conclusion The use the qaṭal form following ‫ ֶט ֶרם‬is rare in BH, but is comparatively more common in DSS Hebrew. While one or more cases in BH may stem from the secondary recasting of I-y qal yiqṭol forms as qaṭal, other cases are not so readily explained. These latter may well be early grammatical deviations from the norm, akin to other subordinate structures in which absolute past qaṭal and relative future yiqṭol forms interchange. If any biblical ṭɛrɛm qaṭal instances are original, this calls into question—though does not entirely invalidate—the supposedly secondary character of other cases of ṭɛrɛm qaṭal. In any case, on the assumption that some cases of ṭɛrɛm qaṭal are secondary, it is clear that such rein- terpretations are in line with early minority usage. 15. HA-QAṬAL It is well known that in BH the definite article -‫ ַה‬is commonly prefixed to participles as a relativising particle.1 Indeed, with participles -‫ ַה‬is a far more common relativiser than ‫א ֶשר‬. ֲ 2 Only irregularly does relativising -‫ ַה‬occur with finite verbs, specifically the suffix conjugation. Most of the biblical cases of ha-qaṭal appear to be late, secondary, or both. 1.0. Relativising -‫ ַה‬with qaṭal in the Tiberian Biblical Tradition 1.1. Post-classical Biblical Hebrew While relativising -‫ ַה‬+ participle is found throughout the Hebrew Bible, a peripheral post-classical feature involves extension of the definite article’s relativising role to finite verbs, specifically 1 GKC (§116o); JM (§138c(2)); Williams (1970, §539); Holmstedt (2016, 69–73). Cf. WO (§19.7b), who reject the classification of -‫ ַה‬with participles as relativising on the grounds that participles can have a relativising function without -‫ ַה‬. Of course, on this logic, neither does ‫ֲא ֶשר‬ qualify as a relativiser, since qaṭal and yiqṭol forms can also be subordinated in asyndetic relative clauses with no need of an explicit relative particle. The potential for asyndetic relative clauses in no way negates the relativising function of either ‫ ֲא ֶשר‬or -‫ ַה‬. 2 There are over 1600 cases of -‫ ַה‬+ (active or passive) participle. Even if more purely adjectival participles are excluded in such a way as to leave only verbal participles, these dominate the mere 36 cases of ‫ ֲא ֶשר‬+ (active or passive) participle. © 2023 Aaron D. Hornkohl, CC BY-NC 4.0 https://doi.org/10.11647/OBP.0310.15 348 The Historical Depth of the Tiberian Reading Tradition qaṭal forms.3 Consider the acknowledged cases of -‫ ַה‬+ qaṭal from TBH and LBH compositions in examples (1)–(12).4 (1) ‫אּולים׃‬ ַֽ ִ ‫י־ים ֶ ֶ֖ד ֶרְך ַל ֲע ֵ֥ ֹבר ְּג‬ ֵָ֔ ‫־ה ֙יא ַה ַמ ֲח ֶר ֶבת ֵָ֔ים ֵ ֶ֖מי ְּתהֹום ַר ָ ָׂ֑בה ַּהשָ ָמה ַ ַֽמ ֲע ַמ ֵק‬ ִ ‫אַת‬ ְּ ‫ֲה ַ֤לֹוא‬ ‘Are you not she, who dries up the sea, the waters of the great deep, who made the depths of the sea a way for the passing of the redeemed?’ (Isa. 51.10) (2) ...‫יְּהוֶ֖ה ֵמ ַעל ַע ָׂ֑מֹו‬ ָ ‫מר ַה ְּב ֵ ָ֧דל יַ ְּב ִד ַיל֛נִ י‬ ֹ ֵ֔ ‫ן־הנֵ ֵָ֗כר הַּ נִ ְּׁלוָ ָּ֤ה ֶאל־יְּ הוָ ֙ה ֵלא‬ ַ ‫אמר ֶב‬ ַ ֹ ‫וְּ אַל־י‬ ‘And let not the foreigner who has joined himself to the LORD say “The LORD will surely separate me from his people.”…’ (Isa. 56.3) (3) ...‫ללָ ה ֲא ֶשר֩ ָהיְּ ָָֹ֨תה ֲחזָ ָ ַָ֤קה ַביָ ֙ם‬ ָ ֶ֗ ֻׁ‫נֹוש ֶבת ִמיַ ִ ָׂ֑מים ָה ִעיר הַּ ה‬ ֶ֖ ֶ ‫אָב ְּד ְּת‬ ֵַ֔ ‫ ֵאיְך‬... ‘…How you have perished, you who were inhabited from the seas, O city which was praised, who was mighty on the sea…’ (Ezek. 26.17) (4) ...֒‫את הַּ ָבּ֣אָה ָע ָליו‬ ֮ ֹ ‫ל־ה ָר ָעה ַהז‬ ָ ‫ַוֵַּֽ֣יִ ְּש ְּמ ָ֞עּו ְּשֹל ֶשת ׀ ֵר ֵעי ִאיֵ֗ ֹוב ֵאת ָכ‬ ‘And Job’s three friends heard about all this calamity that had come upon him…’ (Job 2.11) 3 GKC (§138i–k); Lambert (1931, §295); JM (§138c(2)); Williams (1970, §539); WO (§19.7c); Holmstedt (2016, 69–73). 4 The linguistic periodisation of most of the verses in the lists presented in §§1.1 and 1.2 is uncontroversial. On the post-CBH status of Isaiah 40–66 see Paul (2012) and Arentsen (2020) (cf. Rooker 1996); on that of the narrative framework of Job see Hurvitz (1974) and Joosten (2014) (cf. Young 2009). Ruth’s date of composition is debated; while it contains several non-standard features, a few with late affinities, most of these can be attributed to factors other than late provenance, and the composition’s overall linguistic style is classical. Whatever the case may be, its periodisation, whether early or late, does not materially affect the present argument. 15. Ha-Qaṭal (5) 349 ‫אַח ֵ ֛רי‬ ֲ ‫אש ַצר ַה ֶ ָׂ֑מ ֶלְך ָחזָ֞ ֹון נִ ְּראַָ֤ה ֵא ַ ֙לי ֲא ִני ָדנִ ֵֵ֔יאל‬ ַ ‫ִב ְּש ַנת ָש ֵ֔לֹוש ְּל ַמ ְּל ֶ֖כּות ֵב ְּל‬ ‫הַּ נִ ְּׁראָּ֥ה ֵא ַלֶ֖י ַב ְּת ִח ָ ַֽלה׃‬ ‘In the third year of the reign of King Belshazzar, a vision appeared to me, Daniel, after that which had appeared to me previously.’ (Dan 8.1) (6) ‫רּומת ֵבית־‬ ַ ‫ת־ה ֵכ ִ ָׂ֑לים ְּת‬ ַ ‫ת־הזָ ָ ֶ֖הב וְּ ֶא‬ ַ ‫ת־ה ֶ ֵ֥כ ֶסף וְּ ֶא‬ ַ ‫(ואשקולה) וָ ֶא ְּש ֳק ָלה ָל ֵֶ֔הם ֶא‬ ‫ֹלהינּו הַּ הֵ ִרימּו ַה ֶ֙מ ֶל ְ֙ך וְּ י ֲֹע ָציו וְּ ָש ָ ֵ֔ריו וְּ ָכל־יִ ְּש ָר ֵ ֶ֖אל ַהנִ ְּמ ָצ ִ ַֽאים׃‬ ֵֵ֗ ‫ֱא‬ ‘And I weighed out to them the silver and the gold and the vessels, the offering for the house of our God that the king and his counsellors and his lords and all Israel there present had offered.’ (Ezra 8.25) (7) ...‫יֹות יָ ב ֹ ֙א ְּל ִע ִתים ְּמזֻ ָמ ֵ֔ ִנים‬ ֙ ‫וְּ כֹל ׀ ֲא ֶשר ֶב ָע ֵ ֵ֗רינּו הַּ ה ִִ֞שיב נָ ִ ַ֤שים נָ ְּכ ִר‬... ‘…and let all in our cities who have taken foreign wives come at appointed times…”’ (Ezra 10.14) (8) ...‫שיבּו נָ ִשים נָ ְּכ ִריָׂ֑ ֹות‬ ֹׁ֖ ִ ‫וַ יְּ ַכלּו ַב ֵ֔כֹל ֲאנָ ִָ֕שים הַּ ה‬ ‘And they came to the end of all the men who had married foreign women….’ (Ezra 10.17) (9) ...‫רּויָׂ֑ה‬ ָ ‫ן־צ‬ ְּ ‫ן־נר וְּ יֹואֶָ֖ב ֶב‬ ֵ ֵ֔ ‫אַב ֵנר ֶב‬ ְּ ְּ‫ן־קיש ו‬ ִֵ֔ ‫מּואל ָהר ֶֹא ֙ה וְּ ָשאּול ֶב‬ ַ֤ ֵ ‫וְּ ָֹ֨ ֹכל ַּ ִֽה ִה ְּׁק ִ ָ֜דיש ְּש‬ ‘And all that Samuel the seer and Saul the son of Kish and Abner the son of Ner and Joab the son of Zeruiah had dedicated…’ (1 Chron. 26.28) (10) ‫ב־לְך׃‬ ַֽ ָ ‫יתי ְּב ִש ְּמ ָ ֶ֖חה ְּל ִ ַֽה ְּתנַ ֶד‬ ִ ‫וְּ ַע ֵָ֗תה ַע ְּמ ָ֙ך הַּ נִ ְּׁמ ְּׁצאּו־ ֵ֔ ֹפה ָר ִ ֵ֥א‬... ‘…and now your people, who have been found here, I have seen, joyously offering freely to you.’ (1 Chron. 29.17) (11) ...‫יד ִמ ִק ְּר ַית יְּ ָע ִ ֵ֔רים ַּ ִֽבהֵ ִ ּ֥כין לֶ֖ ֹו ָדִוָׂ֑יד‬ ֙ ִ‫ים ֶה ֱע ָלַ֤ה ָדו‬ ֙ ‫ֹלה‬ ִ ‫ֲא ֵָ֗בל ֲא ַ֤רֹון ָה ֱא‬ ‘But David brought up the ark of God from Kiriath-jearim wherein David had prepared for it…’ (2 Chron. 1.4) 350 The Historical Depth of the Tiberian Reading Tradition (12) ...‫ֹלהים ָל ָ ָׂ֑עם‬ ֶ֖ ִ ‫ל־ה ֵ֔ ָעם ַ ֛על הַּ הֵ ִ ּ֥כין ָה ֱא‬ ָ ‫הּו וְּ ָכ‬ ֙ ‫ק‬ ָ֙‫וַ יִ ְּש ַ ַ֤מח יְּ ִחזְּ ִ י‬ ‘And Hezekiah and all the people rejoiced over what God had prepared for the people…’ (2 Chron. 29.36) In a few cases above, the written tradition is ambiguous, possibly reflecting relativising -‫ ַה‬prefixed to a participle. In these instances, it is not unreasonable to entertain the possibility that the -‫ ַה‬+ qaṭal syntagm reflected in the reading tradition is due to secondary reinterpretation. In the case of the II-w/y qal forms in examples (1) and (4)—‫ ַה ָ ֙ש ָמ ֙ה‬and ‫—ה ָבאָה‬this ַ would involve no more than a shift from ultimate stress in the relevant FS participles to penultimate stress in the 3FS qaṭal forms. In the 3MS III-y nifʿal forms in examples (2) and (5)—‫ ַהנִ ְּלָוַ֤ה‬and ‫—הנִ ְּראֵָ֥ה‬it ַ presup- poses a shift from the MS participle’s expected segol to the qaṭal’s qameṣ in the final syllable. Even so, in the majority of the cases— eight of twelve: (3), (6)–(12)—the written tradition’s consonantal form and the vocalisation tradition unambiguously agree in their testimony regarding a -‫ ַה‬+ qaṭal sequence—the forms ‫ה ֻה ֵ֗ ָל ָלה‬, ַ ‫ימּו‬ ֙ ‫ה ֵה ִ ֙ר‬, ַ ‫הה ִָֹ֞שיב‬, ַ ‫הה ִ ֶֹ֖שיבּו‬, ַ ‫ה ִה ְּק ִ ֶ֜דיש‬, ַֽ ַ ‫הנִ ְּמ ְּצאּו‬, ַ and ‫ ַ ַֽב ֵה ִ ֵ֥כין‬cannot be read as anything other than qaṭal forms prefixed with relativising -‫ ַה‬. Though such frequent agreement between the LBH written tradition and the Tiberian vocalisation does not guarantee the authenticity of the reading tradition’s -‫ ַה‬+ qaṭal interpretation in the four aforementioned consonantally ambiguous forms, it is clear that the explicit understanding of equivocal structures as relativising -‫ ַה‬+ qaṭal sequences in no way contradicts, but in- 15. Ha-Qaṭal 351 deed lines up with the linguistic character of the written tradition as witnessed in consonantal evidence.5 1.2. Classical Biblical Hebrew Of course, the phenomenon of relativising -‫ ַה‬prefixed to qaṭal forms is not limited in the Masoretic tradition to post-classical texts, but also shows up in apparently pre-exilic CBH material; see examples (13)–(20). (13) ...‫ה־נא וְּ ֶא ְּר ֵֶ֔אה ַה ְּכ ַצ ֲע ָק ָ ֛תּה הַּ ָבּ֥אָה ֵא ַלֶ֖י ָעשּו ׀ ָכ ָלָׂ֑ה‬ ָ ‫ֵ ַֽא ֲר ָד‬ ‘I will go down to see whether they have done altogether as the outcry that has come to me…’ (Gen. 18.21) (14) ‫ם־בנָ֧ ֹו הַּ נִֽ ֹולַּ ד־ל֛ ֹו ֲא ֶשר־יָ ְּל ָדה־לֵ֥ ֹו ָש ָ ֶ֖רה יִ ְּצ ָ ַֽחק׃‬ ְּ ‫ת־ש‬ ֶ ‫אַב ָר ֶָ֜הם ֶ ַֽא‬ ְּ ‫וַ יִ ְּק ָ ָֹ֨רא‬ ‘Abraham called the name of his son who was born to him, whom Sarah bore him, Isaac.’ (Gen. 21.3)6 (15) ‫יְּמה ִש ְּב ִ ַֽעים׃‬ ָ ‫ל־ה ֶנ ֶָ֧פש ְּל ֵ ַֽבית־יַ ֲע ֛קֹב הַּ ָבּ֥אָה ִמ ְּצ ַ ֶ֖ר‬ ַ ‫ ָכ‬... ‘…All the persons of the house of Jacob who came to Egypt were seventy.’ (Gen. 46.27) 5 The form ‫ ַהנִ ְּמ ָצַ֤א‬in ...‫הוָׂ֑ה‬ ָ ְּ‫אֹוצר ֵבית־י‬ ַ ‫תֹו ֲא ָב ֵ֔ ִנים נָ ְּתנֶ֖ ּו ְּל‬ ֙ ‫‘ וְּׁ הַּ נִ ְּׁמ ָצָּ֤א ִא‬And those with whom precious stones were found gave them to the treasury of the house of the LORD…’ (1 Chron. 29.8) is ambiguous. Here it is considered a participle; cf. JM (§145d). 6 The qaṭal analysis of the verbal form in ‫ֹולד־‬ ַ ַֽ‫( ַהנ‬Gen. 3.21) is arguable. Though its Tiberian vocalisation with pataḥ is characteristic of the nifʿal suffix conjugation, the form is alternatively analysable as a participle, with pataḥ rather than the expected qameṣ due to the closed, unstressed status of the syllable before maqqef. See WO (§19.7d), who cite JM (§145e), though the latter do not list the verse in question. Cf. Bauer and Leander (1922, §32e). 352 The Historical Depth of the Tiberian Reading Tradition (16) ‫ל־ק ִצ ָ֞ ֵיני אַנְּ ֵ ַ֤שי ַה ִמ ְּל ָח ָמ ֙ה‬ ְּ ‫אמר ֶא‬ ֶ ֹ ‫ל־איש יִ ְּש ָר ֵֵ֗אל ַוּ֠י‬ ִ ‫ל־כ‬ ָ ‫הֹוש ַע ֶא‬ ֶֻ֜ ְּ‫ וַ יִ ְּק ָ ָֹ֨רא י‬... ...‫הֶ הָ ְּׁלכּ֣ ּוא ִא ֵ֔תֹו‬ ‘And Joshua summoned all the men of Israel and he said to the chiefs of the men of war who had gone with him…’ (Josh. 10.24) (17) ‫ֹלהי יִ ְּש ָר ֵֵ֔אל הַּ נִ ְּׁראָּ֥ה ֵא ָלֶ֖יו‬ ֵ ‫מה ִ ַֽכי־נָ ָטה ְּל ָב ֵ֗בֹו ֵמ ִ ַ֤עם יְּהוָ ֙ה ֱא‬ ֹ ָׂ֑ ‫יְּהוֶ֖ה ִב ְּשֹל‬ ָ ‫אַנֵּ֣ף‬ ֵ֥ ַ ‫וַ יִ ְּת‬ ‫ַפ ֲע ָ ַֽמיִ ם׃‬ ‘And the LORD was angry with Solomon, because his heart had turned away from the LORD, the God of Israel, who had appeared to him twice’ (1 Kgs 11.9) (18) ...‫מֹוא ִביָ ַ֤ה ַכ ָל ָת ּ֙ה ִע ֵָ֔מּה הַּ ָשֹׁ֖בָ ה ִמ ְּש ֵדי מֹואָָׂ֑ב‬ ֲ ‫וַ ָת ָשב נָ ֳע ִֵ֗מי וְּ ָֹ֨רּות ַה‬ ‘So Naomi returned, and Ruth the Moabite her daughter-inlaw with her, who returned from the country of Moab…’ (Ruth 1.22) (19) ‫ֹוא ִביָ ֙ה ִֵ֔היא הַּ ָשּ֥בָ ה ִ ַֽעם־נָ ֳע ִ ֶ֖מי ִמ ְּש ֵ ֵ֥דה מֹואַָֽב׃‬ ֲ ‫נַ ֲע ָ ַ֤רה ַֽמ‬... ‘She is the young Moabite woman, who came back with Naomi from the country of Moab.’ (Ruth 2.6) (20) ‫ימ ֶלְך ָמ ְּכ ָרה נָ ֳע ִֵ֔מי הַּ ָשֹׁ֖בָ ה ִמ ְּש ֵ ֵ֥דה מֹואַָֽב׃‬ ָׂ֑ ֶ ‫אָחינּו ֶל ֱא ִל‬ ֶ֖ ִ ‫ ֶח ְּל ַק ֙ת ַה ָש ֶ ֵ֔דה ֲא ֶ ֵ֥שר ְּל‬... ‘…Naomi, who has come back from the country of Moab, hereby offers for sale the parcel of land that belonged to our relative Elimelech.’ (Ruth 4.3) Additional cases are sometimes cited, but are excluded here.7 7 Some cite ‫ ַה ִד ֵבר‬in ...‫ים יִ ְַּֽהיּו ְּל ֵ֔ר ַּוח וְּׁ הַּ ִד ֵבֹׁ֖ר ֵאין ָב ֶ ָׂ֑הם‬ ֙ ‫יא‬ ִ ‫‘ וְּ ַהנְּ ִב‬and the prophets will become wind; and the divine word is not in them…’ (Jer. 5.13) as a case of relativising -‫ ַה‬with qaṭal, but according to the pronunciation tradition, this is a noun (Steiner 1992; Hornkohl 2013a, 294–27). JM (§145d, fn. 5) suggest the relevance of ostensibly corrupt cases in 1 Chron. 12.24 and 2 Chron. 15.11, in both of which the relativising -‫ ַה‬is 15. Ha-Qaṭal 353 1.3. Diachrony within the Masoretic Tradition There is a degree of similarity between early and late material in terms of the use of relativising -‫ ַה‬with qaṭal. However, the similarity is somewhat superficial and must not be allowed to mask significant differences. 1.3.1. Frequency and Diachronic Development First, it should be noted that the relatively smaller TBH/LBH corpus exhibits a greater proportional incidence of relativising -‫ַה‬ with qaṭal than the much more extensive CBH corpus (a discrepancy that becomes even more pronounced if Ruth, here categorised as CBH, is assigned to the post-exilic category). 1.3.2. Ambiguous Consonantal Forms and the Case for Dissonance Second, as mentioned above, eight of the twelve cases of relativising -‫ ַה‬with qaṭal in post-classical biblical material involve consonantally unambiguous qaṭal forms. By contrast, among the CBH cases just one of eight cases—example (16) above, ‫( ֶה ָה ְּלכּוא‬Josh. 10.24)—has a consonantally unambiguous qaṭal form. Put differently, nearly all of the apparently classical cases of relativising -‫ ַה‬+ qaṭal, along with a few of the later ones, involve consonantal forms amenable to analysis as participles. missing. There is also one apparent CBH case of relativising -‫ ַה‬attached to a preposition: ... ָ‫ליה‬ ֶ ָ֜ ָ‫ת־ה ָֹ֨שֹוק וְּׁ הֶ ע‬ ַ ‫‘ וַ יָ ֵֶּ֣רם ּ֠ ַה ַטּ ָבח ֶא‬So the cook took up the leg and what was on it…’ (1 Sam. 9.24). 354 The Historical Depth of the Tiberian Reading Tradition As observed above, only penultimate syllable stress distinguishes the 3FS II-w/y qal qaṭal forms—‫ ַה ָש ָמה‬, ‫ ַה ָב ָאה‬, and ‫— ַה ָש ָבה‬ from FS participles, the latter with ultimate stress, i.e., ‫ ַה ָש ָמה‬, ‫ ַה ָב ָאה‬, and ‫ ַה ָש ָבה‬. The distinction between qaṭal and participle is perceptible in contrasting examples, e.g., (21) versus (22). (21) ...֒‫את הַּ ָבּ֣אָה ָע ָליו‬ ֮ ֹ ‫ל־ה ָר ָעה ַהז‬ ָ ‫ַוֵַּֽ֣יִ ְּש ְּמ ָ֞עּו ְּשֹל ֶשת ׀ ֵר ֵעי ִאיֵ֗ ֹוב ֵאת ָכ‬ ‘And Job’s three friends heard about all this calamity that had come upon him…’ (Job 2.11) (22) ...‫ּוכ ֵל ָא ֙ה‬ ְּ ‫יתָך ְּכ ָר ֵ ַ֤חל׀‬ ֵֶ֗ ‫ל־ב‬ ֵ ‫ת־ה ִא ֶָ֜שה הַּ בָ ָאּ֣ה ֶא‬ ָ ‫הוה ֶ ַֽא‬ ָ ֙ ְּ‫יִ ֵתן֩ י‬... ‘…May the LORD make the woman who is coming into your house like Rachael and like Leah…’ (Ruth 4.11) In the case of the 3MS nifʿal qaṭal forms—‫הנִ ְּלוָ ה‬, ַ ‫הנִ ְּראָה‬, ַ ‫נֹולד‬ ַ ‫—ה‬ ַ differentiation from the corresponding MS participial forms lies in the final vowel alone, the respective participles being ‫ ַהנִ ְּלוֶ ה‬, ‫ ַהנִ ְּר ֶאה‬, ‫נֹולד‬ ָ ‫ ַה‬. For contrastive examples, see (23) and (24). (23) ‫אַח ֵ ֛רי‬ ֲ ‫אש ַצר ַה ֶ ָׂ֑מ ֶלְך ָחזָ֞ ֹון נִ ְּראַָ֤ה ֵא ַ ֙לי ֲא ִני ָדנִ ֵֵ֔יאל‬ ַ ‫ִב ְּש ַנת ָש ֵ֔לֹוש ְּל ַמ ְּל ֶ֖כּות ֵב ְּל‬ ‫הַּ נִ ְּׁראָּ֥ה ֵא ַלֶ֖י ַב ְּת ִח ָ ַֽלה׃‬ ‘In the third year of the reign of King Belshazzar, a vision appeared to me, Daniel, after that which had appeared to me previously.’ (Dan 8.1) (24) ‫ה־שם ִמזְּ ֵֵ֔ב ַח ָל ֵאל֙ הַּ נִ ְּׁר ֶאּ֣ה ֵא ֵ֔ ֶליָך ְּב ָב ְּר ֲח ֵָ֔ך‬ ָ ‫ב־שם וַ ֲע ֵש‬ ָׂ֑ ָ ‫ית־אל וְּ ֶש‬ ֶ֖ ֵ ‫קּום ֲע ֵ ֵ֥לה ֵ ַֽב‬... ֛ ‫ִמ ְּפ ֵנֶ֖י ֵע ָ ֵ֥שו ָא ִ ַֽחיָך׃‬ ‘…Arise, go up to Bethel and dwell there. Make an altar there to the God who appeared to you when you fled from your brother Esau.’ (Gen. 35.1) The salient difference between the incidence of relativising -‫ ַה‬+ qaṭal in CBH, on the one hand, and post-classical BH, on the 15. Ha-Qaṭal 355 other, can be formulated thus: while in the post-classical texts most instances of relativising -‫ ַה‬with qaṭal involve explicit agreement between unambiguous forms in the written (consonantal) and reading (vocalisation) traditions, in the more classical material the consonantal ambiguity that attaches to most of the relevant forms leaves room for a claim of dissonance between the written and reading traditions. It is certainly suspicious that such a large proportion of classical relativising -‫ ַה‬+ qaṭal cases have consonantal forms amenable to interpretation as the far more common relativising -‫ ַה‬+ participle sequence. This possibility should be seen in the light of a long list of other features in which it has been argued that the reading tradition of classical texts deviates from that of the written tradition in line with late tendencies on which the written and reading traditions of Second Temple texts agree. If a significant proportion of the apparently early cases of relativising -‫ ַה‬with qaṭal are indeed due to dissonance between the written and reading traditions, then this would be another in such a series of features in terms of which the reading tradition wedded to classical biblical material resembles the combined written-reading tradition of late material. Such a situation is most readily explained by the theory that the reading tradition of CBH material, though reliably preserving much in the way of distinctively classical features, nevertheless drifted in the direction of post-classical Hebrew until crystallisation in the Second Temple Period, i.e., approximately when the LBH material was composed. This means that, on occasion, the vocalisation of CBH texts anachronistically departs from the phonic realisation intended according to the written tradition 356 The Historical Depth of the Tiberian Reading Tradition in favour a post-classical standard. Such a hypothesis—which, again, applies in the case of a number of features discussed in the present monograph and elsewhere—accounts for the obvious disparity between Masoretic CBH and post-classical BH when it comes to the incidence of relativising -‫ ַה‬+ qaṭal: in post-classical material there is widespread agreement between the written and reading traditions involving consonantally unambiguous forms, while in CBH the dearth of consonantally unambiguous forms regularly leaves the reading tradition’s testimony regarding -‫ ַה‬+ qaṭal without corroborating testimony from the written tradition. 1.3.3. Versional Evidence Given the ambiguity of the Tiberian CBH evidence due to the possibility of dissonance between its written and reading components, it is reasonable to solicit aid from other ancient textual witnesses . Upon examination, however, it becomes apparent that these provide only general and limited evidence. The DSS evidence is fragmentary and ambiguous. The Samaritan written tradition is accompanied by a reading tradition, but the latter does not discern between the qaṭal and participle forms of the relevant verbs. The evidence from the rest of the versions is nearly complete, but ambiguous in its own way, since, as observed below, -‫ ַה‬+ qaṭal appears in contexts where the more frequent -‫ ַה‬+ participle can also be used and with similar semantic force. Thus, depending on the context, one might expect similar translations for the two. Table 1 (facing page) gives the equivalents of MT cases of -‫ ַה‬+ qaṭal in the BDSS, the SP, the Peshiṭta, the principal traditional relevant Targums, the Greek, and the Vulgate. Table 1: Versional equivalents of MT -ַ‫ ה‬+ qaṭal MT DSS* SP Peshiṭta Targums Greek Vulgate * 357 ‫( השמה‬1QIsaa 42.25); ‫( הנלוא‬1QIsaa 46.13); ‫( הנ̇ לוה‬1Q8 24.18). 15. Ha-Qaṭal Gen. 18.21 ‫ה‬ ַ ‫ה ָ ָּ֥ב ָא‬ — ‫ הבאה‬abˈbā ‫ܕܥܠܬ‬ ‫ת‬ ַ ‫דעל‬ τὴν ἐρχομένην qui venit ݂ Gen. 21.3 ‫הנּֽ ֹול ַד‬ — ‫ הנול ַד‬annuwwal̄̊ ǝd ‫ܕܐܬܝܠܕ‬ ‫ְד ִא ִת ֵילי ַד‬ τοῦ γενομένου — Gen. 46.27 ‫ה‬ ַ ‫ה ָ ָּ֥ב ָא‬ — ‫ הבאה‬abˈbā ‫ܕܥܠ‬ ‫א‬ ַ ‫ְדע ָל‬ αἱ εἰσελθοῦσαι quae ingressa Josh. 10.24 ‫א‬ ַ ‫ֶה ָה ְל ֣כּו‬ — — ‫ܕܡܗܠܟܝܢ‬ ‫ד ְאזלַּו‬ τοὺς συμπορευομένους qui… erant 1 Kgs 11.9 ‫הנִ ְר ָ ָּ֥אה‬ — — ‫ܕܐܬܓܠܝ‬ ‫ְד ִאתגְ ִלַי‬ τοῦ ὀφθέντος qui apparuerat Isa. 51.10 ‫ה‬ ַָּׂ֙ ‫ה ָ ָּׂ֙ש ָמ‬ ‫ה‬ ַ ‫השמ‬ — ‫ܥܒܕܬܝ‬ ‫יתַי‬ ִ ִ‫שו‬ ἡ θεῖσα qui posuisti ݂ ...‫ܕ‬ Isa. 56.3 ‫ה‬ ַ ‫א הנִ ְלָ ָ֤ו‬ ַ ‫ה ;הנלו‬ ַ ‫הנלו‬ ̇ — ‫݁ܕܡܬܠܘܐ‬ ‫ְד ִמתֹוס ַף‬ ὁ προσκείμενος qui adheret Ezek. 26.17 ‫ה‬ ַ ‫ה ֻה ָּ֗ ָל ָל‬ — — ‫ܥܫܝܢܬܐ‬ ‫א‬ ַ ‫חת‬ ָ ‫ְמשב‬ ἡ ἐπαινεστὴ inclita Job 2.11 ‫ה‬ ַ ‫ה ָ ֣ב ָא‬ — — ‫ܕܐܬܬ‬ ‫ת‬ ַ ‫דאת‬ τὰ ἐπελθόντα accidisset Ruth 1.22 ‫ה ָ ָּׁ֖ש ָבה‬ — — ‫ܕܐܨܛܒܝ݂ ܬ ܠܡܗܦܟ‬ ‫ת‬ ַ ‫דתב‬ ἐπιστρέφουσα ac reversa est ݂ Ruth 2.6 ‫ה ָ ָּ֥ש ָבה‬ — — ‫ܕܐܬܬ‬ ‫ת‬ ַ ‫דתב‬ ἡ ἀποστραφεῖσα quae venit Ruth 4.3 ‫ה ָ ָּׁ֖ש ָבה‬ — — — ‫ת‬ ַ ‫דתב‬ τῇ ἐπιστρεφούσῃ quae reversa est Dan. 8.1 ‫הנִ ְר ָ ָּ֥אה‬ — — ‫ ܕܐܬܚܙܝ‬...‫ܚܙܘܐ‬ — τὸ ἰδεῖν quod videram Ezra 8.25 ‫ימּו‬ ַָּׂ֙ ‫ה ֵה ִ ָּׂ֙ר‬ — — ‫ܕܝܗܒ‬ — ἃ ὕψωσεν quae obtulerat ݂ Ezra 10.14 ‫הה ִִֹׁ֞שיב‬ — — ‫ܕܐܘܬܒܘ‬ — ὃς ἐκάθισεν qui duxerunt Ezra 10.17 ‫הה ִ ָֹּׁ֖שיבּו‬ — — ‫ܕܐܘܬܒܘ‬ — οἳ ἐκάθισαν qui duxerant 1 Chron. 26.28 ‫ש‬ ַ ‫ּֽה ִה ְק ִ ִּ֜די‬ — — ‫݁ܕܩܕܫ‬ ‫ש‬ ַ ‫א דאקדי‬ ַ ‫הקדיש‬ τῶν ἁγίων haec… sanctificavit 1 Chron. 29.17 ‫הנִ ְמ ְצאַּו‬ — — ‫ܕܐܫܬܟܚ‬ ‫דאשתכחַו‬ τὸν εὑρεθέντα qui… reppertus est ݁ ‫ܐܝܟܐ‬ ݁ 2 Chron. 1.4 ‫ּֽב ֵה ִ ָּ֥כין‬ — — ‫ܕܬܩܢ ܗܘܐ‬ ‫כד אתקיַן‬ ὅτι ἡτοίμασεν in locum quem paraverat ‫ܕܫܠܡ ݂ܬ ܥܒܝܕܬܐ ܕܒܝܬܐ‬ ‫דאתקיַן‬ τὸ ἡτοιμακέναι quod ministerium (Domini) esset expletum 2 Chron. 29.36 ‫ה ֵה ִ ָּ֥כיַן‬ — — ݂ 358 The Historical Depth of the Tiberian Reading Tradition The first thing that can be seen is that, despite sporadic cases of non-equivalence—Gen. 21.3 in the Vulgate, Ruth 4.3 in the Peshiṭta—little to no textual doubt attaches to any of the cases. In other words, based on versional evidence, there is no widespread lack of equivalence interpretable as evidence for the frequent late insertion of relativising -‫ ַה‬+ qaṭal in the Masoretic tradition. Rather, in the majority of cases for which there is evidence, it would seem that the copyist or translator had at their disposal a consonantal text similar, if not identical, to the Tiberian consonantal text. It is not obvious, however, that the relevant -‫ ַה‬+ verb syntagm was necessarily interpreted as -‫ ַה‬+ qaṭal. In order to attempt to gain some clarity on this, it is useful to compare versional treatment of the -‫ ַה‬+ qaṭal syntagm with treatment of the far more common -‫ ַה‬+ participle alternative. In light of the latter syntagm’s semantic flexibility, it is unsurprising that ren- derings are by and large contextual. This is to say, a given version’s translation of a specific instance is generally in line with the semantics of the context. It is important to emphasise, however, that the semantic ambiguity that attaches to a number of forms can occasion diversity among the translations. Be that as it may, renderings tend to fall on a continuum ranging from forms that denote the general present semantics of enduring characteristics (25), through those that convey imperfective past semantics for attendant, but not necessarily permanent, circumstances of varying persistence (26)–(27), to those expressing perfective past semantics for transitory unitary events (26). 15. Ha-Qaṭal 359 (25) ...‫הה ֵלְֹׁ֖ך ִק ְּד ַמת ַא ָׂ֑שּור‬ ִֽ ַּ ‫ישי ִח ֶ ֵ֔ד ֶקל ֵ֥הּוא‬ ֙ ִ ‫( וְּ ֵָֹ֨שם ַהנָ ָ ַ֤הר ַה ְּש ִל‬DSS: ‫ההלך‬ ̇ 4Q2 ̊̄ f1ii.1; SP ‫ ההלך‬aːlǝk) ‘…The name of the third river was Tigris—this is the one that flows east of Assyria…’ (Gen. 2.14) ݁ .‫ ݁ܗܘ ܕܐܙܠ ܠܘܩܒܠ ܐܬܘܪ‬.‫ܘܫܡܗ ܕܢܗܪܐ ܕܬܠܬܐ ܕܩܠܬ‬... ‫ינחא ַד ֲאתּור‬ ָ ‫גלת הּוא ְּׁמהַּ לֵ יך ְּל ַמ ִד‬ ָ ‫ית ָאה ִד‬ ָ ‫הרא ְּת ִל‬ ָ ַ‫וְּ שֹום נ‬... …καὶ ὁ ποταμὸς ὁ τρίτος Τίγρις· οὗτος ὁ πορευόμενος κατέναντι Ἀσσυρίων. …nomen vero fluminis tertii Tigris ipse vadit contra Assyrios In the case of the MT’s active participle for a permanent characteristic in (25), all Semitic equivalents are active participles, the Greek is a present participle, and the Latin is a present-tense finite form. (26) ‫ת־ה ֱאמ ִ ֵֹ֔רי הַּ י ֵשֹׁ֖ב ְּב ַ ַֽח ְּצ ֵ֥צֹן ָת ָ ַֽמר׃‬ ָ ‫ל־ש ֵ ֶ֖דה ָה ֲע ָמ ֵל ִ ָָׂ֑קי וְּ גַ ֙ם ֶא‬ ְּ ‫ת־כ‬ ָ ‫ וַ יַ ָ֕כּו ֶ ַֽא‬... (SP ‫ הישב‬ayyēšǝb) ‘…and they defeated all the country of the Amalekites, and also the Amorites who dwelt in Hazazon-tamar.’ (Gen. 14.7) ‫ܒܥܝܢ ܓܕ܀‬ ‫ܕܝܬܒܝܢ‬ ̈ ̈ ‫ ܘܐܦ ܠ ܐܡܘܪܝܐ‬.‫ܕܥܡܠܩܝܐ‬ ‫ܘܚܪܒܘ ܟܘܠ ܪܫܢܐ‬... ‫מֹור ָאה ְּׁדיָתֵ יב ְּב ֵעין־גַ ִדי׃‬ ָ ‫קלי ֲע ָמ ְּל ָק ָאה וְּ ַאף יָ ת ֲא‬ ֵ ‫ּומחֹו יָ ת כֹל ַח‬... …καὶ κατέκοψαν πάντας τοὺς ἄρχοντας Αμαληκ καὶ τοὺς Αμορραίους τοὺς κατοικοῦντας ἐν Ασασανθαμαρ. …et percusserunt omnem regionem Amalechitarum et Amorreum qui habitabat in Asasonthamar Like the MT active participle with enduring past relevance in (26), the SP, Peshiṭta, and Targum use active participles, the Greek a present participle, and the Vulgate an imperfect past form. 360 The Historical Depth of the Tiberian Reading Tradition ̊̄ (27) ‫אן־ּוב ָ ֶָ֖קר וְּ א ָֹה ִ ַֽלים׃‬ ָ ֹ ‫ת־א ְּב ָ ָׂ֑רם ָה ָיֵ֥ה צ‬ ַ ‫ם־ל ֵ֔לֹוט הַּ ה ֵלְֹׁ֖ך ֶא‬ ְּ ַ‫( וְּ ג‬SP ‫ ההלך‬aːlǝk) ‘And Lot, who went/was travelling with Abram, also had flocks and herds and tents’ (Gen. 13.5) ̈ ‫ܘܡܫܟܢܐ‬ ̈ .‫ ܐܝܬ ܗܘܘ ܥܢܐ ܘܬܘܪܐ‬.‫ܥܡ ܐܒܪܡ‬ ‫ܕܐܙܠ‬ ‫ܘܐܦ ܠܠܘܛ‬ .‫ܣܘܓܐܐ ܕܛܒ‬ ‫שכנִ ין׃‬ ְּ ‫ּומ‬ ַ ‫תֹורין‬ ִ ְּ‫ברם ֲהוֹו ָען ו‬ ָ ‫וְּ ַאף ְּללֹוט ְּׁדאָ זֵיל ִעים ַא‬ καὶ Λωτ τῷ συμπορευομένῳ μετὰ Αβραμ ἦν πρόβατα καὶ βόες καὶ σκηναί. sed et Loth qui erat cum Abram fuerunt greges ovium et armenta et tabernacula The MT’s active participle is semantically ambiguous, conceivably referring either to the initial point of Lot’s accompaniment of Abram or to its continuation. The versions diverge: the Syriac suffix conjugation form seems to indicate a perfective past reading, while the Targum’s active participle, the Greek’s present par- ticiple, and the Latin’s imperfect appear to reflect imperfective interpretations. ̊̄ (28) ‫יהוֶ֖ה הַּ נִ ְּׁר ֶאּ֥ה ֵא ָ ַֽליו׃‬ ָ ‫וַ יִ ֶַ֤בן ָש ֙ם ִמזְּ ֵֵ֔ב ַח ַל‬... (SP ‫ הנראה‬annirraʾi) ‘…And he built there an altar to the LORD who had appeared to him.’ (Gen. 12.7) .‫ܥܠܘܗܝ‬ ݁ ‫ܘܒܢܐ‬... ‫ܬܡܢ ܡܕܒܚܐ ܠܡܪܝܐ ܕܐܬܓܠܝ‬ ‫דב ָחא קדם יוי ְּׁד ִאתגְּׁ ִלי ֵליה׃‬ ְּ ‫ּובנָ א ַת ָמן ַמ‬... …καὶ ᾠκοδόμησεν ἐκεῖ Αβραμ θυσιαστήριον κυρίῳ τῷ ὀφθέντι αὐτῷ. …qui aedificavit ibi altare Domino qui apparuerat ei In (28) the MT’s nifʿal participle seems to refer to a unitary past event. The versions likewise resort to various forms indicating perfective past tense semantics: the suffix conjugation in Syriac 15. Ha-Qaṭal 361 and Aramaic, an aorist participle in Greek, and the pluperfect in Latin. Of course, versional treatment of the -‫ ַה‬+ participle syntagm is not without exegetical and stylistic variation. Even so, the foregoing examples may be considered broadly representative of common equivalencies. In the nature of things, the much rarer -‫ ַה‬+ qaṭal syntagm that is the focus of this chapter has a far narrower semantic range. The versions, unsurprisingly, then, commonly resort to strategies consistent with past-tense interpretation. This is especially evident in the Peshiṭta, the Targums, and the Vulgate, which overwhelmingly opt for indicative forms with past-tense TAM semantics. Overall, the Greek renderings show a slightly greater degree of variation, mixing in comparatively more in the way of equivalencies arguably consistent with the reading of participles rather than qaṭal forms. The problem is that, as already mentioned, the common -‫ ַה‬+ participle syntagm had such a broad semantic range and was given to such a variety of translation strategies, that it is difficult on the basis of translations to reconstruct a Vorlage’s specific syntagm. Even so, it is intriguing that in the translations of clear-cut consonantal qaṭal forms in LBH material, there is near-unanimous past-tense translation. By contrast, cases of ostensible divergence between qaṭal and participle analysis nearly always involve a consonantally ambiguous form. Thus, the fact that the Tiberian reading tradition’s ‫‘ ַה ְּכ ַצ ֲע ָק ָ ֛תּה הַּ ָבּ֥אָה ֵא ַלֶ֖י‬whether… as the outcry that has come to me’ (Gen. 18.21) is paralleled by suffix conjugation forms in the Syriac and Aramaic, but by a Greek present 362 The Historical Depth of the Tiberian Reading Tradition participle and a Latin indicative present,8 may well indicate divergent analyses of consonantal ‫הבאה‬.9 Or not. Consider the apparently unequivocal qaṭal form in ‫ל־ק ִצ ָ֞ ֵיני אַנְּ ֵ ַ֤שי ַה ִמ ְּל ָח ָמ ֙ה הֶ הָ ְּׁלכּ֣ ּוא ִא ֵ֔תֹו‬ ְּ ‫אמר ֶא‬ ֶ ֹ ‫‘ ַוּ֠י‬and he said to the chiefs of the men of war who had gone with him’ (Josh. 10.24): in this case, TJ renders with a suffix conjugation, but the Peshiṭta has an active participle, the Greek a present participle, and the Vulgate the bland imperfective erant ‘were’. The point is that, given both the semantic range of the -‫ ַה‬+ participle syntagm and stylistic freedom of choice on the part of translators, their renderings equivalent to MT -‫ ַה‬+ qaṭal cases must be considered rather shaky evidence for the reconstruction of translator analysis of the forms in question. 2.0. Relativising -‫ ַה‬with qaṭal beyond the Tiberian Biblical Tradition The relativising -‫ ַה‬+ qaṭal syntagm is rather peripheral in the Tiberian biblical tradition. It is evidently even rarer outside of 8 Assuming that the e-vowel in venit is short. I take this opportunity to thank my friend and colleague, Ben Kantor, for his help in making sense of the Greek and Latin evidence. 9 While the Tiberian reading tradition draws a clear distinction between 3FS qaṭal ‫ ָב ָאה‬and FS participle ‫ ָב ָאה‬, this is by no means universal. They are read identically in the Samaritan tradition. Likewise, in Modern Hebrew, penultimate stress is standard in both the 3FS qaṭal and the FS participle, except when the latter is used adjectivally, e.g., ‫השנה הבאה‬ ‘next year’. It may be that some ancient exegetes recognised a single underdifferentiated II-w/y qal 3FS qaṭal/FS participle form, which they interpreted according to context. 15. Ha-Qaṭal 363 Masoretic BH—though, admittedly, many potential cases are left ambiguous due to the lack of an explicit reading tradition. Even so, the complete absence or rarity of unambiguous consonantal forms has significance. Codex Kaufmann of the Mishna presents at least one apparently certain case, and possibly an additional instance. (29) ...‫>שנָ תֹו‬ ַֽ ְּ ‫<מ‬ ִ ‫אֹומ׳ ִמ ֵשם ר׳ ֵמ ִאיר ָכל הַּ שָ כַּ ח ָד ָבר ֶא ָחד ִמ‬ ֵ ‫דֹוס ִתי ִב ְּר׳ יַ נַ יִ י‬ ְּ ‫ר׳‬ ‘R. Dosti son of R. Yannai in the name of R. Meir says, “Whoever forgets a single thing from what he has learned…”’ (ʾAvot 3.8) (30) ‫עֹולין לֹו ִמן‬ ִ ‫לֹושים יֹום ֵאין‬ ִ ‫ִמי ֶשנָ זַ ר וְּ הּוא ֵב⟦י⟧ן ַה ְּק ָברֹות ֲא ִפילּו הּוא ָשם ְּש‬ ‫ּומ ִביא‬ ֵ ‫עֹולין לֹו ִמן ַה ִמינְּ יָ ן‬ ִ ‫יכנַ ס‬ ְּ ִ‫טּומ ָאה היָצָ א וְּ נ‬ ְּ ‫ַה ִמינְּ יָ ין וְּ ֵאינּו ֵמ ִביא ָק ְּור ָבן‬ ...‫טּומ ָאה‬ ְּ ‫{ו}ר ַבן‬ ְּ ‫ָק‬ ‘He who vowed to be a Nazirite while in a graveyard, even if he was there for thirty days—they do not count for him toward the number [of days owing under the vow] and he does not bring an offering for his uncleanness [for being in the graveyard]. He who went out and re-entered [the graveyard]—they count for him toward the number [of required days] and he brings an offering for uncleanness.’ (Nazir 3.5) Neither case in the Mishna is entirely unambiguous, since the two apparent qal 3MS qaṭal forms could conceivably have been vocalised as such, but intended as qal participles.10 Moreover, the ap- 10 This is far more likely in the case of ‫ שכח‬than in that of ‫יצא‬, since in Codex Kaufmann the participle ‫(י)ח‬ ַ ‫ ש ֵֹכ‬is never written with a mater waw (see m. Peʾa 6.11; m. Shabbat 7.1) and the stative-like participle form 364 The Historical Depth of the Tiberian Reading Tradition parent article in (30) has been crossed out. The resulting subjectless verbal forms in (30), while acceptable in Rabbinic style as a type of conditional, i.e., ‘if he went out and re-entered’, can also be read as a headless relative clause parallel to ‫ִמי ֶשנָ זַ ר‬ ‫‘ וְּ הּוא‬he who vowed to be a Nazirite while he was…’ 3.0. Discussion and Ramifications 3.1. Development At some point in the history of ancient Hebrew a rather marginal syntagm consisting of relativising -‫ ַה‬+ qaṭal arose. JM (§145d, fn. 5) suggests alternative developmental scenarios for such a structure: This phenomenon may have had its origin in the 3rd pers. sg. of the perfect in cases where the form was similar to that of the participle, e.g. ‫ ַה ָבא‬and ‫הנִ ְּמ ָצא‬, ַ and then it may have spread to the 3rd pers. pl. (and the 3rd fem. sg….). The evolution may have continued, but our texts do not show it. Alternatively, the phenomenon may have originated in a fairly common structure in which an indeterminate noun is qualified by a participial phrase with the definite article…, as in Jdg 16.27 ‫ים ִאיש‬ ֙ ‫ל־ה ֵ֗ ָגג ִכ ְּשֹלַ֤ ֶשת ֲא ָל ִפ‬ ַ ‫וְּ ַע‬ ‫ וְּ ִא ֵָ֔שה ָהר ִ ֶֹ֖אים ִב ְּש ֵ֥חֹוק ִש ְּמ ַֽשֹון‬and on the roof there were about three thousand men and women watching Samson’s show. While JM raises these scenarios as mutually exclusive alternatives, both could conceivably have factored into the development of relativising -‫ ַה‬+ qaṭal. Two further explanations JM (§145d, ‫ ָש ֵכ ַח‬also occurs (see m. Miqvaʾot 4.1, 1, 1), whereas the MS participle ‫יֹוצא‬ ֵ is consistently (over 200 times) spelled plene. 15. Ha-Qaṭal 365 fn. 5)—probably rightly—reject. Andersen (2000, 53), proposed that qaṭal with relativising -‫ ַה‬represents the preservation of qaṭal’s archaic use as a verbal noun (cf. the Akkadian form vari- ously called ‘stative’, ‘verbal adjective’, ‘permansive’). However, the fact that consonantally unambiguous cases of -‫ ַה‬+ qaṭal occur with relative frequency only in LBH militates against the approach. Also, the proposed combination of a pre-classical use of qaṭal with the decidedly classical definite article seems improbable. Representing a different tack, Lambert (1931, §295 fn. 3) suggested that relativising -‫ ַה‬with qaṭal is the Hebrew cognate of the Akkadian relativiser ša. Cf. the Akkadian-Hebrew š-h interchange in the 3rd-person independent pronouns, šafʿel versus hifʿil, and locative-directional -iš versus ‫ ִָה‬-.11 The hypothesis does not enjoy wide support. 3.2. Historical Depth, Anachronism, and Preservation While the mechanism for the emergence of relativising -‫ ַה‬+ qaṭal may be satisfactorily explained, its chronology remains murky. A compelling accumulation of unequivocal consonantal evidence shows that writers had recourse thereto in the exilic and postexilic periods. The majority of -‫ ַה‬+ qaṭal forms in TBH and LBH are consonantally unambiguous. While ambiguous structures in contemporary sources vocalised and/or accented as cases of -‫ ַה‬+ qaṭal may be analysed as secondary reinterpretations of -‫ ַה‬+ par11 More broadly comparable is the analogous development between Proto Indo-European and Greek represented by such Latin-Greek correspondences as sex versus héks ‘six’, sub versus hypó ‘below’, super versus hypér, and salis versus hálas ‘salt’. 366 The Historical Depth of the Tiberian Reading Tradition ticiple, there is no proof that the vocalisation deviates from the intended written form in such cases. On the contrary, the fact that the LBH written tradition lines up with the Tiberian vocalisation tradition in many cases in which the vocalisation tradition as at odds with the CBH written tradition points to special affinity between the written and reading traditions of late Masoretic biblical material. The real question regards the extent of vocalic authenticity versus secondary analysis in CBH texts, where the majority of the apparent cases of relativising -‫ ַה‬+ qaṭal involve ambiguous consonantal spellings. As noted above, a degree of dissonance between CBH consonantal material and the Tiberian reading tradition with which it has been combined is known from analyses of numerous features. In such cases, the vocalisation anachronistically reflects Second Temple standards, often in contravention of the written tradition. This may well be the situation of the majority of the apparent CBH cases of relativising -‫ ַה‬+ qaṭal. Indeed, one scholarly approach views all relativising -‫ ַה‬+ verb syntagms as cases of -‫ ַה‬+ participle, unless the consonantal form unambiguously reflects -‫ ַה‬+ qaṭal, no matter what the vowels and accents of the reading tradition indicate (e.g., GKC §138i–k). It is important to note, however, that while anachronistic from the perspective of CBH norms as indicated by the orthographic tradition, the phenomenon is, by dint of its documentation in the late consonantal and vocalisation traditions, clearly biblical. Indeed, since the phenomenon is not characteristic of QH or RH, nor of Aramaic, it can only with difficulty be regarded as a post-biblical feature retrojected into BH. Rather, it tallies 15. Ha-Qaṭal 367 uniquely with Hebrew literary conventions of the Persian, and perhaps Hellenistic Periods, and not with later Byzantine, much less medieval norms. But the extent of the potential linguistic anachronism in question must be characterised with appropriate nuance. Beyond the fact that, overall, diachronic dissonances of this type are detectable in only a small minority of instances in BH, it is often the case that classical consonantal material presents authentic forerunners of diagnostically late features eventually to become more standard in later phases of the language, such as those reflected in the LBH written tradition and the Tiberian reading tradition. Again, such may be the case here. One could regard the Tiberian vocalisation of TBH and LBH -‫ ַה‬+ qaṭal cases as genuine, but doubt the authenticity of the vocalisation in apparent CBH cases. While most of the apparently early cases of qaṭal with relativising -‫ ַה‬involve consonantally ambiguous forms, ‫‘ ֶה ָה ְּלכּוא‬who had gone’ (Josh. 10.24) is the notable exception. The consonantal form, though displaying a non-standard spelling (with final ʾalef) more typical of the DSS, can be read only as a qaṭal form. Possibly the only consonantally unequivocal classical case of qaṭal with relativising -‫ ַה‬, it merits brief discussion. In view of parallels in the ancient versions, no real textual doubt attaches to the form. Moreover, neither the immediate nor the surrounding context raises suspicion that the form is a product of late intervention. Finally—and of profound methodological importance—though the syntagm itself is characteristically late, one should resist the impulse to prejudge it as exclusively so. Other characteristically 368 The Historical Depth of the Tiberian Reading Tradition late features are found sporadically in classical texts. While there may be various reasons to speculate on the secondary status of some such forms, it bears pointing out that no characteristically late linguistic feature went overnight from non-use to common use. Late currency often began with rare early usage. Logic, then, dictates entertaining the possibility of sporadic classical distribution followed by later characteristic usage. Consider, for example, such characteristically late features as ‫‘ ַמ ְּלכּות‬kingdom, reign, rule’ (classical attestations in Num. 24.7; 1 Sam. 20.31; 1 Kgs 2.12; Hurvitz 2014, 165–70; cf. Young, Rezetko, and Ehrensvärd 2008, II:84–85); words sharing the root ‫‘ של"ט‬rule’ (classical at- testation of ‫‘ ַש ִליט‬ruler’ in Gen. 42.6; Hurvitz 2014, 228–36; cf. Joosten 2019, 33–35); and ‫‘ נְּ ָכ ִסים‬possessions’ (classical attesta- tion in Josh. 22.8; Hurvitz 2013, 330; cf. Schoors 1992–2004, II:257–58). Similarly, it seems likely that the comparatively late proliferation of qaṭal with relativising -‫ ַה‬was a development with (albeit rare) classical roots. But once this is admitted as a possibility, it carries with it the potential that any number of the consonantally ambiguous forms construed in the reading tradition as qaṭal forms are correctly vocalised—not just in late texts, but in early ones, too (in agreement with Holmstedt 2016, 71). The argument can also be approached from another angle. Along with the apparently early consonantal evidence for relativising -‫ ַה‬+ qaṭal, there is evidence of nuance within the vocalisation of those CBH forms amenable to analysis as instances of -‫ ַה‬+ qaṭal. In other words, not every case interpretable as -‫ ַה‬+ qaṭal 15. Ha-Qaṭal 369 was so read. Consider the contrast between examples (31) and (32), which consist of successive verses: (31) ‫ל־נ ֶֶ֖פש‬ ֶ ‫יְּמ ֙ה י ְֹּצ ֵאי יְּ ֵר ֵ֔כֹו ִמ ְּל ַ ֶ֖בד נְּ ֵשי ְּבנֵ י־יַ ֲע ָׂ֑קֹב ָכ‬ ָ ‫ל־הנֶ ֶפש הַּ בָ ֶָ֨אה ְּליַ ֲע ַ֤קֹב ִמ ְּצ ַ ֙ר‬ ַ ּ֠ ‫ָכ‬ ‫ִש ִ ֵ֥שים וָ ֵ ַֽשש׃‬ ‘All the persons belonging to Jacob who came into Egypt, who were his own descendants, not including Jacob’s sons’ wives, were sixty-six persons in all.’ (Gen. 46.26) (32) ‫ל־ה ֶנ ֶָ֧פש ְּל ֵ ַֽבית־יַ ֲע ֛קֹב הַּ ָבּ֥אָ ה‬ ַ ‫יֹוסף ֲא ֶשר־יֻ ַלד־לֵ֥ ֹו ְּב ִמ ְּצ ַ ֶ֖ריִ ם ֶנ ֶפש ְּש ָ ָׂ֑נֵּ֣יִ ם ָכ‬ ֛ ֵ ‫ּוב ֵנֵ֥י‬ ְּ ‫ִמ ְּצ ַ ֶ֖ריְּ ָמה ִש ְּב ִ ַֽעים׃ פ‬ ‘And the sons of Joseph, who were born to him in Egypt, were two. All the persons of the house of Jacob who came into Egypt were seventy.’ (Gen. 46.27) Both instances of ‫ הבאה‬refer semantically to past events, but they are distinguished in the reading tradition: in (31) the form is accented as -‫ ַה‬+ participle and in (32) it is accented as -‫ ַה‬+ qaṭal. As each was conceivably given to either understanding, it is clear that the reading tradition cannot be accused of wholesale rebranding of -‫ ַה‬+ participle as -‫ ַה‬+ qaṭal wherever possible. A similar argument can be made regarding the vocalisation of ‫ הנראה‬as -‫ ַה‬+ participle in examples (33) and (34), but as -‫ ַה‬+ qaṭal in (35). (33) ‫יהוֶ֖ה הַּ נִ ְּׁר ֶאּ֥ה ֵא ָ ַֽליו׃‬ ָ ‫וַ יִ ֶַ֤בן ָש ֙ם ִמזְּ ֵֵ֔ב ַח ַל‬... ‘…So he built there an altar to the LORD, who had appeared to him.’ (Gen. 12.7) 370 The Historical Depth of the Tiberian Reading Tradition (34) ‫ה־שם ִמזְּ ֵֵ֔ב ַח ָל ֵאל֙ הַּ נִ ְּׁר ֶאּ֣ה ֵא ֵ֔ ֶליָך ְּב ָב ְּר ֲח ֵָ֔ך‬ ָ ‫ב־שם וַ ֲע ֵש‬ ָׂ֑ ָ ‫ית־אל וְּ ֶש‬ ֶ֖ ֵ ‫ ֛קּום ֲע ֵ ֵ֥לה ֵ ַֽב‬... ‫ִמ ְּפ ֵנֶ֖י ֵע ָ ֵ֥שו ָא ִ ַֽחיָך׃‬ ‘…“Arise, go up to Bethel and dwell there. Make an altar there to the God who appeared to you when you fled from your brother Esau.”’ (Gen. 46.27) (35) ‫ֹלהי יִ ְּש ָר ֵֵ֔אל הַּ נִ ְּׁראָּ֥ה ֵא ָלֶ֖יו‬ ֵ ‫מה ִ ַֽכי־נָ ָטה ְּל ָב ֵ֗בֹו ֵמ ִ ַ֤עם יְּהוָ ֙ה ֱא‬ ֹ ָׂ֑ ‫יְּהוֶ֖ה ִב ְּשֹל‬ ָ ‫אַנֵּ֣ף‬ ֵ֥ ַ ‫וַ יִ ְּת‬ ‫ַפ ֲע ָ ַֽמיִ ם׃‬ ‘And the LORD was angry with Solomon, because his heart had turned away from the LORD, the God of Israel, who had appeared to him twice’ (1 Kgs 11.9) While such variation within the Tiberian reading tradition might be chalked up to inconsistency in the application of late norms to early texts, it might just as well reflect some degree of genuine preservation. Even so, the infrequency in CBH material of consonantally unambiguous qaṭal forms with relativising -‫ ַה‬should be accorded due weight. There is one further perspective that merits consideration. Though, as mentioned, relativising -‫ ַה‬+ qaṭal apparently fails to persist in any meaningful way in QH or RH, the Samaritan reading tradition exhibits a phenomenon worthy of consideration in this connection. The Samaritan equivalents of Tiberian qal, piʿʿel, and nifʿal all have MS participles identical to the respective 3MS qaṭal forms (Ben-Ḥayyim 2000, §§2.12.2, 6, 9–10). This not infrequently results in cases of relativising -‫ ַה‬prefixed to forms identical to the Samaritan suffix conjugation, and this not just in places where the MT has relativising -‫ ַה‬with a form pointed as qaṭal. Perhaps the most striking come in D-stem, e.g., 15. Ha-Qaṭal 371 (36) MT: ...‫יה ַא ָ ֶ֖תה ֵאל ֳר ִ ָׂ֑אי‬ ָ ‫וַ ִת ְּק ָ ַ֤רא ֵשם־יְּהוָ ֙ה הַּ ד ֵבּ֣ר ֵא ֵ֔ ֶל‬ SP ...‫) אליה אתה אל ראה‬addabbǝr( ‫ותקרא שם יהוה הדבר‬ ‘So she called the name of the LORD who spoke to her, “You are a God of seeing”…’ (Gen. 46.27) (37) MT: ...‫ד־ה ָ ָׂ֑ע ֶרב‬ ָ ‫האכֵ ל ִמנִ ְּב ָל ֵָ֔תּה יְּ ַכ ֵ ֵ֥בס ְּבגָ ָ ֶ֖דיו וְּ ָט ֵמא ַע‬ ִֽ ָ ְּׁ‫ו‬ ̊̄ SP ...‫) מנבלתה יכבס בגדיו וטמא עד הערב‬wakkǝl ( ‫והאכל‬ ‘and whoever eats of its carcass shall wash his clothes and be unclean until the evening…’ (Lev. 11.40) Similar congruence between participle and qaṭal forms is noticeable in the case of, e.g., qal ‫( ַהש ֵ ֶֹ֖מ ַע‬MT Gen. 21.6) || qal ‫השמע‬ aššāma (SP Gen. 21.6); nifʿal ‫( ַהנִ ְּר ֶ ֵ֥אה‬MT Gen. 12.7) || nifʿal ‫הנראה‬ annirraʾi̊̄ (SP Gen. 12.7); qal ‫( ַהד ֵֹבר‬MT Gen 16.13) || piʿʿel ‫הדבר‬ addabbǝr (SP Gen 16.13).12 It is not clear whether or how the broader Samaritan tendency to discard the distinction between participial and qaṭal forms might be related to the extension in the Tiberian tradition of relativising -‫ ַה‬to the qaṭal form, but whether these were related or separate processes, the result was similar: late traditions in which relativising -‫ ַה‬could be prefixed to forms indistinguishable from qaṭal. 4.0. Conclusion To summarise: the combined Tiberian written-reading tradition in LBH texts and the Tiberian reading tradition wedded to CBH material constitute clear Second Temple evidence of authentic, if peripheral, use of the relativising -‫ ַה‬+ qaṭal syntagm. Most of the 12 These are cited on the basis of Tal and Florentin 2010 (written tradi- tion) and Ben-Ḥayyim 1977 (reading tradition). 372 The Historical Depth of the Tiberian Reading Tradition CBH cases of the syntagm are consonantally ambiguous, but the single exception looks to be a genuine forerunner of a feature later to become more widespread. As such, it arguably validates the vocalisation of one or more of the ambiguous CBH and LBH cases pointed as relativising -‫ ַה‬+ qaṭal. Either way, with regard to the feature under discussion, there is no disputing that the vocalisation and accentuation of the Tiberian reading tradition line up with LBH consonantal evidence, thus reflecting a date no later than the Persian or early Hellenistic Period, and potentially preserve evidence of the rare Iron Age usage of the same feature. If the Tiberian reading tradition departs from the CBH written tradition on this matter, it does so only by retrojecting onto the written tradition a more advanced stage of a process already seen to be underway therein and that is evidenced more explicitly in the combined LBH written and reading tradition. Of course, it is not impossible that the syntagm was as common, or nearly so, in CBH as it was in LBH, and that its preserved documentation is misleading. But, again, the ambiguity of the majority of the CBH cases of relativising -‫ ַה‬+ qaṭal, in conjunction with the comparative frequency with which unequivocal cases are found in the relatively more limited LBH corpus, arouses the suspicion that at least a portion of the CBH instances are secondary. 16. WAYYIQṬOL One of the defining characteristics of Masoretic BH is the wayyiqṭol verbal form. Especially common in narrative, it typically encodes perfective past semantics. The Tiberian biblical tradition distinguishes it from the consonantally homographic volitive weyiqṭol by means of gemination of the verbal preformative (or a compensatory vowel shift in the 1CS form).1 However, converging lines of evidence relevant to the development of wayyiqṭol have recently led to the hypothesis that Iron Age waw-yiqṭol was a polysemous syntagm and that its differentiation into mainly preterite wayyiqṭol and chiefly jussive/purpose we-yiqṭol was secondary and relatively late. If so, Masoretic wayyiqṭol may well represent an extremely pervasive instance of dissonance between the consonantal tradition of early biblical material and the recitation tradition embodied in the accompanying vocalisation. The present chapter deals with wayyiqṭol in general, especially evidence for (a) the early underdifferentiation of narrative (preterite) and modal waw-yiqṭol, (b) the late secondary differentiation into geminated wayyiqṭol and non-geminated we-yiqṭol, and (c) the historical depth of the semantic distinction between the two. In order to lay the groundwork for reviewing a recent 1 Notwithstanding the modern convention of transcribing shewa as e/ǝ, in the Tiberian pronunciation the chief distinction between wayyiqṭol and we-yiqṭol was one of gemination, not vowel quality. This is clear from evidence showing that the default realisation of shewa in Tiberian BH was as short a, identical to the realisation of pataḥ (Kantor 2020, 59, 66–91; Khan 2020, I:305; 2021, 332). © 2023 Aaron D. Hornkohl, CC BY-NC 4.0 https://doi.org/10.11647/OBP.0310.16 374 The Historical Depth of the Tiberian Reading Tradition proposal by Khan (2021), the discussion first centres on three strands of evidence on which Khan builds, namely: secondary developments in proto-Masoretic Hebrew, transcriptional evidence for the phonetic realisation of preterite and modal waw-yiqṭol forms in antiquity, and non-preterite wayyiqṭol semantics. The subsequent chapter (ch. 17) focuses specifically on 1stperson forms. Striking diachronic patterns involving 1st-person wayyiqṭol morphological alternatives—manifest in both the consonantal and vocalisation traditions—not only come as arguable confirmation of the general correctness of (a), (b), and (c) above, but allow for greater precision in the relative periodisation of the Masoretic written and reading traditions with respect to the wayyiqṭol form. 1.0. Supporting Evidence The following subsections summarise research into three lines of evidence fundamental to the view that the Iron Age situation of semantically undifferentiated waw-yiqṭol gave way in the Second Temple Period to one in which perfective past wayyiqṭol and volitive/purpose we-yiqṭol were secondarily differentiated.2 2 Limitations of space preclude exhaustiveness in citation of the volu- minous bibliography related to wayyiqṭol. Smith (1991) remains an oftcited resource, with more recent references in Bloch (2007); Robar (2013; 2015, 78–112; 2021); Gzella (2018); Kantor (2020); and Khan (2021). 16. Wayyiqṭol 375 1.1. Semantic Gemination, i.e., Semantic Dagesh ‘Semantic dagesh’ refers to secondary gemination in one of the ancient Hebrew recitation traditions for purposes of disambiguating perceived homophones, i.e., to divide a word considered polysemous into morphologically distinct lexemes. Khan (2018, 341–47; 2020, I:524–30) collects numerous examples of ‘semantic dagesh’ from biblical (Tiberian, Babylonian, Samaritan) and non-biblical (rabbinic) traditions. Examples from Tiberian Hebrew include ‫‘ ֲא ִביר‬powerful (divine)’ versus ‫‘ ַא ִביר‬powerful (human)’, ‫‘ ֲע ָצ ִבים‬toils’ versus ‫‘ ֲע ַצ ִבים‬idols’, and, probably, ‫‘ ִה ְּר ִעים‬make thunder (divine)’ versus ‫‘ ִה ְּר ִעים‬vex, irritate (human)’. “The gem- ination in these pairs of forms most likely originates in existing variant morphological patterns that have been exploited to avoid homophony” (Khan 2020, I:525). While his 2021 article represents Khan’s first attempt at a comprehensive account of wayyiqṭol’s development incorporating the notion of semantic gemination, he first raised the possibility in 1991 (Khan 1991, 241; 2013, 43; 2021, 330; Kantor 2020, I:104, fn. 23). 1.2. Transcriptional Evidence In a detailed survey of Greek and Latin transcriptional evidence relevant to the development of wayyiqṭol, Kantor adduces compelling evidence of historical evolution in the form’s phonetic realisation. In the late Second Temple Period, writes Kantor (2020, 99–100), The conjunction waw was usually pronounced identically before a preterite yiqṭol and non-preterite yiqṭol form, namely, with no full vowel or following gemination. Nev- 376 The Historical Depth of the Tiberian Reading Tradition ertheless, the conjunction waw was also frequently pronounced distinctly before a preterite yiqṭol form, being vocalised with a full vowel and (probably) gemination…. Subsequently, in the early Byzantine Period, “The conjunction waw was always pronounced distinctly before a preterite yiqṭol form (as opposed to before a non-preterite yiqṭol), being vocalised with a full vowel and (probably) gemination….” Extrapolating back from the diachronic trajectory, Kantor argues that in Iron Age BH “the conjunction waw was pronounced identically before a preterite yiqṭol and non-preterite yiqṭol form, probably with the original etymological */a/ vowel,” meaning “that up to some point in the Second Temple Period, yiqṭol in the sequence *wyiqṭol was a polysemous form, indicating either past or non-past (usually jussive) semantics according to context.” Significantly, Kantor (2020, 104–5) follows Khan (1991, 241; 2013, 43) in positing secondary semantic disambiguation of previously undifferentiated waw-yiqṭol into preterite wayyiqṭol and non-preterite we-yiqṭol as the most plausible explanation for gemination in Masoretic wayyiqṭol (see above, §1.1). Admittedly, one cannot totally exclude the possibility that the Tiberian reading tradition reflects an Iron Age realisation that already distinguished past waw-yiqṭol (> wayyiqṭol) from nonpast waw-yiqṭol (> we-yiqṭol) by gemination. But several lines of argumentation combine to suggest otherwise: (a) the absence of any such distinction in the Samaritan reading tradition,3 (b) the 3 For an alternative means of distinguishing preterite waw-yiqṭol in the Samaritan reading tradition, i.e., the replacement of waw-yiqṭol with 16. Wayyiqṭol 377 partial but increasing use of the distinction in the period of the Greek and Latin transcriptions, (c) a degree of disagreement between the Tiberian and Babylonian vocalisation traditions, and (d) the broad reality in the Masoretic biblical tradition of multiple cases of dissonance involving early consonantal orthography vocalised according to a characteristically later reading tradition. Such considerations are arguable evidence that the disambiguation in question took place after the Samaritan and Jewish traditions had diverged, was in the process of taking hold at the time the transcriptions were made, and had become solidly established before the division of the Masoretic Tiberian, Babylonian, and Palestinian branches. 1.3. Non-preterite Wayyiqṭol Robar (2013; 2015, 78–112) builds a multi-pronged argument against wayyiqṭol’s consensus preterite classification. She sees wayyiqṭol as a narrative present of unspecified time reference that takes its TAM semantics from the context. While Khan’s (2021) theory differs from Robar’s at important points, he cites her work favourably and agrees that certain wayyiqṭol semantic values are incompatible with core preterite semantics. He proposes a broader realis value that allows for greater semantic flexibility, which, crucially, he explains as a result of the form’s fused preterite-modal parentage. waw-qaṭal in the case of I-y qal verbs, see Khan (2021, 331). See also below, ch. 18, esp. §1.3. 378 The Historical Depth of the Tiberian Reading Tradition 2.0. The Development of Wayyiqṭol Synthesising the aforementioned studies and additional research, Khan’s (2021, 319–40) discussion appears in a paradigm-shifting study that employs Construction Grammar to explain the development of wayyiqṭol by means of the recognised mechanisms of reanalysis and schematisation. Khan seeks to improve upon existing accounts of wayyiqṭol’s development in line with its semantic range, pragmatics, and status as the sole standard remnant of archaic preterite short yiqṭol (< PS yaqtul). Khan argues that preterite yiqṭol’s preservation almost exclusively after waw is due to syntactic and semantic similarity to a “discourse dependent” (Khan 2021, 320ff.) modal short yiqṭol in a (normally) purpose/result waw-yiqṭol construction, which made preterite waw-yiqṭol ripe for reanalysis. In this way, the short yiqṭol’s originally distinct preterite and modal purpose/result semantics became fused in a semantically undifferentiated waw-yiqṭol construction.4 The core semantics of the resulting waw-yiqṭol had effectively been reduced to a “common denomi- nator” of temporal posteriority relative to preceding context (Khan 2021, 326), which was further schematised to one of broader “topical cognitive relevance” (Khan 2021, 340). Later, in some Second Temple traditions, the realis (preterite) and irrealis (volitive, often purpose/result) senses of wawyiqṭol were disambiguated via gemination of the preforma-tive in 4 Khan (2021, 319, fn. 13) explicitly sidesteps the question of whether the ancient Hebrew preterite and volitional short yiqṭol values are themselves reflexes of a single (Huehnergard 1988) or distinct PS yaqtul forms (Hetzron 1969; Rainey 1986). 16. Wayyiqṭol 379 realis (mostly preterite) waw-yiqṭol, resulting in a new distinction between realis (mainly preterite) wayyiqṭol and irrealis (volitional, often purpose/result) we-yiqṭol. Khan sees the frequent LBH conflation of 1st-person realis and irrealis waw-yiqṭol strings, i.e., both represented by ‫ואקטלה‬/‫ונקטלה‬, along with sporadic CBH conflation, as confirmation that the relevant realis–irrealis fusion “had already taken place in CBH” (Khan 2021, 321–22, 327; for detailed discussion of 1st-person forms, see ch. 17, below). Khan thus conceives of a convergence of the wayyiqṭol and directive-volitive paradigms earlier and more pervasive than what is usually envisioned. It was not merely due to late analogy with cohortative ‫ אקטלה‬that classical ‫ ואקטל‬shifted to ‫ ;ואקטלה‬rather, the antecedents of wayyiqṭol ‫ וָ ֶא ְּק ְּט ָלה‬and cohortative ‫וְּ ֶא ְּק ְּט ָלה‬, though originally conveying distinct preterite and modal senses, respectively, fused in pre-Tiberian CBH in a semantically undifferentiated waw-yiqṭol structure broadly associated with temporal consecution—only to be disambiguated anew via Second Temple gemination of realis (mostly preterite) waw-yiqṭol > wayyiqṭol. Beyond elegantly explaining the nearly exclusive clause-initial preservation of preterite yiqṭol after waw, Khan’s proposed Iron Age preterite-volitive/purpose waw-yiqṭol fusion helps to illuminate wayyiqṭol’s semantic range: by acknowledging its mixed preterite-modal parentage, the form is revealed to have genetics consistent with non-past and/or non-perfective semantics, such meanings reflecting the archaic tenseless, aspect-free character of wayyiqṭol’s volitive/purpose waw-yiqṭol ancestor. The earlier preterite and volitive semantics, however, gave way in pre-Tiberian BH to a broader sense of temporal consecution and discourse de- 380 The Historical Depth of the Tiberian Reading Tradition pendency. According to this analysis, the old preterite and nonpreterite values did not persist, but had to be inferred from context. 3.0. Pre-Tiberian Waw-Yiqṭol There remains the not trivial matter of how the pre-Tiberian BH verb system ‘worked’ given a semantically undifferentiated wawyiqṭol form, i.e., whether and how users disambiguated preterite and volitive/purpose senses of a waw-yiqṭol emptied of all but the barest of semantic values (temporal posteriority > discourse dependency). The first thing to acknowledge is the “pathway of purpose > result clause > discourse dependent” (Khan 2021, 324). Next, Khan (2021, 326) observes an important correlation: “In the at- tested corpus of Biblical Hebrew… purpose and result clauses with jussives have future main clauses, whereas past wayyiqṭol is generally preceded by a past clause.” In other words, preceding context must commonly have sufficed to disambiguate the past versus volitive/purpose/result semantics of waw-yiqṭol forms. However, Khan (2021, 328) also notes the ambiguity of a wayyiqṭol given to result interpretation, e.g., (1) ...‫ָל ָ ַ֤מה ָא ַ֙מ ְּר ָ֙ת ֲאחֹ ִתי ִֵ֔הוא וָאֶ ַּ ּ֥קח א ָ ֹ֛תּה ִ ֶ֖לי ְּל ִא ָ ָׂ֑שה‬ ‘Why did you say “She is my sister,” so that I took her for my wife?...’ (Gen. 12.19) Despite following preterite ‫‘ ָא ַ֙מ ְּר ָ֙ת‬you (MS) said’, interpretation of ‫‘ וָ ֶא ַ ֵ֥קח‬so that I took’ as heading a pseudo-subordinate result clause is contextually defensible. Indeed, the bare semantic value of temporal consecution combined with the universally attested 16. Wayyiqṭol 381 grammaticalisation pathway of purpose > result arguably make a dependent reading more attractive than one of merely sequential preterites—though both are stops along the same trajectory, i.e., it is a series of straightforward cognitive steps from ‘he went to the store that he might buy cereal’ through ‘he went… with the result that he bought cereal’ to ‘he went… and he bought cereal’. The question then arises as to why in this (or any) cases a bare ‫ ויקטל‬should have been interpreted one way or the other, i.e., as irrealis purpose/result we-yiqṭol ‫ וְּ ֶא ַקח‬or as realis preterite wayyiqṭol ‫וָ ֶא ַ ֵ֥קח‬. In this case, the preceding perfective past qaṭal seems to have influenced the realisation of the following waw-yiqṭol as a realis preterite form notwithstanding the appropriateness in context of a volitive-result reading. It is also possible that the wayyiqṭol realisation was influenced by the appearance of short (‫)ואקח‬, rather than lengthened (‫ )ואקחה‬1st-person morphology. In the case of 1st-person forms in the Hebrew of the Masoretic Torah, only four wayyiqṭol forms have lengthened pseudo-cohortative morphology (Gen. 24.48; Deut. 1.16, 18); likewise, in the same corpus, just two we-yiqṭol forms eligible for cohortative marking lack the characteristic suffixed heh (Exod. 24.7; Deut. 10.2). The mismatch between the Tiberian realis interpretation and the probable volitional-purpose pre-Tiberian sense suggests that the synchronic semantic range of Tiberian wayyiqṭol must extend beyond that of consecutive perfective past eventualities, though by dint of the regularity of such a semantic value, it can certainly be considered synchronically prototypical. Notwithstanding the import of the preceding example, it would be misleading to say that the Tiberian realisation of waw- 382 The Historical Depth of the Tiberian Reading Tradition yiqṭol forms mechanically follows the TAM of the foregoing verbal form. Consider example (2), in which a future-oriented purpose we-yiqṭol follows perfective past forms: (2) ‫א ֶכל֙ ֵ֔ ָלמֹו‬ ֹ ָֹ֨ ‫י־ב ְּק ֵ֥שּו‬ ִ ‫בי ֵה ָמה ִר ֵ֔מּונִ י כ ֲֹה ַנֵ֥י ּוזְּ ֵק ַנֶ֖י ָב ִעיר גָ ָוָׂ֑עּו ִ ַֽכ‬ ֙ ַ ‫אתי ַ ַֽל ְּמ ַא ֲה‬ ִ ‫ָק ָ ַ֤ר‬ ‫ָשיבּו ֶאת־נַ ְּפ ָ ַֽשם׃ ס‬ ֹׁ֖ ִ ‫וְּׁ י‬ ‘I called for my lovers, but they deceived me: my priests and my elders perished in the city, while they sought for themselves food that they might revive their souls.’ (Lam. 1.19) Here, though the broader context shows that ‫ וְּ יָ ִ ֶ֖שיבּו‬refers to an unrealised purpose rather than a realised achievement, the immediately preceding verbs all reference perfective past eventualities. Again, given the notional proximity of purpose, result, and simple sequential readings, it is easy to imagine the form ‫וישיבו‬ being realised as wayyiqṭol ‫‘ וַ יָ ִשיבּו‬and they revived’. This, however, would have contradicted the force of the indictment, since the search for revival was unsuccessful. To summarise: a major factor in inferring a pre-Tiberian waw-yiqṭol’s TAM reference was the narrow context of TAM values in the closely preceding clause(s). Yet, examples like (2) (cf. also Lev. 9.6 (?); Num. 23.9; 1 Sam. 12.3; 1 Kgs 13.33 (?); 2 Kgs 19.25; see JM, §116e; Joosten 2012, 154–55) demonstrate that the tradition was also sensitive to the text’s internal logic. 4.0. Wayyiqṭol’s Secondary Status and Historical Depth The lack of a geminated wayyiqṭol in the Samaritan reading tradition and the only partial evidence for gemination in the Greek 16. Wayyiqṭol 383 and Latin transcriptional material reflect a Second Temple linguistic milieu in which disambiguation of preterite and modal waw-yiqṭol via gemination in the former had not yet become entrenched. If so, then Masoretic wayyiqṭol conceivably represents a secondary and relatively late development in line with the reading tradition’s known adoption of certain linguistic features especially characteristic of Second Temple Hebrew. However, the innovation of ‘semantic dagesh’ also tallies with what Khan (2021, 330–31) describes as “a general Second Temple development in the proto-Masoretic reading tradition involving the introduction of strategies to increase care in pronunciation and clarity of interpretation” (see also Khan 2020, I:73– 85). Despite the secondary and late character of the Masoretic differentiation of wayyiqṭol and we-yiqṭol, there is in general no reason to doubt the historical depth of the interpretive tradition that the distinction reflects. In other words, while the distinction in phonetic realisation between preterite and modal waw-yiqṭol forms appears to be a relatively late proto-Masoretic innovation, it bears witness to earlier consciousness of waw-yiqṭol polysemy as well as, presumably, an incipient interpretive tradition (or traditions) on the basis of which gemination was added to forms construed as realis. While in most cases of preterite and purpose waw-yiqṭol there would have been no danger of misunderstanding, instances such (1) and (2) above are exceptions where, for purposes of interpretation, morphological disambiguation representative of semantic distinction proves semantically determinative. Whatever the antiquity of the phonological disambiguation, it seems clear that it reflects a gradually increasing discomfort 384 The Historical Depth of the Tiberian Reading Tradition with the perceived semantic ambiguity between preterite wawyiqṭol and modal waw-yiqṭol that eventually developed into the fully crystalised Tiberian tradition of semantic gemination to distinguish wayyiqṭol from we-yiqṭol. The phonological distinction goes back to the period of the transcriptions, at the latest. The discomfort with underdifferentiation between preterite and modal forms may have begun earlier. Certainly, the early and frequent morphological distinction between 1st-person preterite wayyiqṭol forms and cohortative we-yiqṭol forms (see below, ch. 17) suggests recognition of a semantic distinction within CBH. 17. 1ST-PERSON WAYYIQṬOL The morphology of the 1st-person wayyiqṭol within the combined Masoretic written-reading tradition is characterised by complex diversity.1 It also represents an area of dissonance between the tradition’s written and reading components. Thankfully, evidence from alternative biblical traditions (the BDSS and the SP) and extra-biblical sources (Iron Age epigraphy, the NBDSS, and BS) sheds light on matters. Not surprisingly, 1st-person forms comprise a small minority of the total number of occurrences of what is BH’s main narrative TAM form, accounting for just under 700 of the more than 15,000 instances, or less than 5 percent. While in the vast majority of cases across all traditions and sources, eligible 2nd- and 3rd-person wayyiqṭol forms preserve short yiqṭol (< PS yaqtul; cf. Akkadian iprus) morphology,2 the 1st-person wayyiqṭol presents in all three of the relevant morphological templates, which, for convenience, are referred to throughout the present chapter with both descriptives and prototypical forms: 1 Among the relevant studies, see S. R. Driver ([1892] 1998, §72); Ungnad (1907, 58 fn. 1); Bergsträsser (1918–1927, II:§5f); Kutscher (1974, 326–27; Rainey (1986, 13–14); Talshir (1986; 1987); Revell (1988, 423); Qimron (1997, 177; 2008, 153–54); Bloch (2007); Hornkohl (2013a, 159–71); Gzella (2018, 29–35); Khan (2021, 319–40); Sjörs (2021). 2 For various scholarly approaches to exceptions among 2nd- and 3rd- person wayyiqṭol forms and further bibliography, see Bloch (2007), Hornkohl (2013a, 171–80), and Gzella (2018). © 2023 Aaron D. Hornkohl, CC BY-NC 4.0 https://doi.org/10.11647/OBP.0310.17 386 The Historical Depth of the Tiberian Reading Tradition 1. short jussive-like ‫וָ ַָּ֫א ַעׂש‬/‫וָ ָָּ֫א ֵעד‬/‫ < *וָ ָָּ֫א ָקם‬PS yaqtul; 2. long yiqṭol-like ‫וָ ַא ֲע ֶשה‬/‫וָ ָא ִעיד‬/‫ < וָ ָאקּום‬PS yaqtulu or yaqtula; 3. lengthened pseudo-cohortative ‫וָ ֶא ְּק ְּט ָלה‬/‫(י)דה‬ ָ ‫וָ ָא ִע‬/‫(ּו)מה‬ ָ ‫*וָ ָא ֻק‬ < PS yaqtula or yaqtulan(na).3 Table 1: Short, long, and pseudo-cohortative 1st-person wayyiqṭol forms in the Tiberian tradition4 Strong III-y hifʿil qal II-w/y 1cs ‫ וָ ֶא ְּש ְּל ָחה‬,‫ וָ ֶא ֱע ֶשה וָ ֶא ְּש ַלח‬,‫ וָ ָא ִע ָידה וָ ַא ַעש‬,‫ וָ ָא ִעיד‬,‫קּומה וָ ָא ֵעד‬ ָ ‫ וָ ָא‬,‫ וָ ָאקּום‬,‫*וָ ָא ָקם‬ 3ms ‫וַ יִ ְּש ַלח‬ ‫וַ יַ ַעש‬ ‫וָ יָ ַעד‬ ‫וַ יָ ָקם‬ 1cpl ‫ וַ נִ ְּש ְּל ָחה‬,‫ וַ נַ ֲע ֶשה וַ נִ ְּש ַלח‬,‫ *וָ נָ ִע ָידה וַ נַ ַעש‬,‫ *וָ נָ ִעיד‬,‫קּומה *וָ נָ ֵעד‬ ָ ָ‫ וַ נ‬,‫ וַ נָ קּום‬,‫*וַ נָ ָקם‬ The orthographic distinction between the short (‫ואעש‬, ‫ואעד‬, ‫)ואקם‬ and long (‫ואעשה‬, ‫ואעיד‬, ‫ )ואקום‬templates is possible only with cer- tain hifʿil and weak verb forms (especially III-y and II-w/y) qal forms. The pseudo-cohortative template is possible in all but IIIy verbs.5 The variation between short and longer forms also ap3 The reconstructed forms are based on the analogy of documented forms; see below. For various opinions on the derivation of the pseudocohortative morphology see, among others, Rainey (1986, 4, 8–10); JM (§§114a–f, 116a–c); Bloch (2007, 143); Blau (2010, §4.3.3.3.4 and the note there); Dallaire (2014, 108–11); Khan (2021, 322–23); Sjörs (2021). 4 For the sake of convenient comparison, the table includes both docu- mented and reconstructed forms. Of the latter, some are less contentious than others. For example, 1CPL ‫ *וַ נָ ָקם‬is based on qere ‫‘ וַ ָנ ַָ֤שב ֻכ ָ ֙ל ֙נּו‬and we all returned’ (Neh. 4.9). For the grounds for other reconstructed forms, e.g., 1cs ‫‘ *וָ ָא ָקם‬and I arose’, see below, §2.0. 5 This is the case in the Masoretic reading tradition. Some scholars hold that this is not necessarily characteristic of other traditions of ancient Hebrew, including, theoretically, the Masoretic written tradition (Bergsträsser 1918, II:§5f; Revell 1988, 423; Bloch 2007, 150, fn. 35, 155). See below, §1.4.2, fn. 11. 17. 1st-person Wayyiqṭol 387 plies to other weak verb types, e.g., contextual 3MS ‫אמר‬ ֶ ֹ ‫ וַ י‬versus 1CS ‫( וָ א ַֹמר‬but consistently 1CPL ‫אמר‬ ֶ ֹ ‫)וַ נ‬, contextual 3MS ‫ וַ יֵ ֶלְך‬versus 1CS ‫( וָ ֵא ֵלְך‬but consistently 1CPL ‫)וַ נֵ ֶלְך‬, where the distinction is one of stress and vocalisation (see below, §2.0). While the evidence has been variously interpreted (Talshir 1986; 1987; Bloch 2007; Hornkohl 2013a; Gzella 2018), the respective distributions of the short, long, and pseudo-cohortative alternants in ancient Hebrew sources seems to indicate that an early situation characterised by the dominance of short forms in all persons gave way to situations in which short morphology continued to reign in 2nd- and 3rd-person forms, but was commonly replaced by long and/or pseudo-cohortative morphology in the 1st-person. While short, long, and pseudo-cohortative 1st-person wayyiqṭol forms seem to have coexisted throughout the history of ancient Hebrew, specific usage patterns involving the prevalence of one or more forms are especially characteristic of certain compositions and corpora. Surveying the data across the various biblical traditions and extra-biblical sources, a perceptible, if somewhat fuzzy, diachronic pattern emerges. Even so, though historical change proves to be the main factor, diachrony does not explain all. Sporadic outliers to the general typological trends suggest the relevance of additional factors.6 Even the significance of certain distribution patterns apparently governed by diachrony merit 6 For critical discussion of several phonological, prosodic, and textual explanations see Bloch (2007), Hornkohl (2013a, 174–78), and Gzella (2018, 31–35). See Robar (2013, 36–39; 2015, 178–81) for explanations related to pragmatics and discourse. 388 The Historical Depth of the Tiberian Reading Tradition scrutiny, as they may be deceptive. Be that as it may, as shown below (§1.0), the general statistical picture is sufficiently clear to warrant starting from a diachronic comparison of distribution between corpora and then moving to a more granular analysis of individual compositions and or forms together with consideration of complementary or contradictory conditioning factors. 1.0. The Masoretic Written (Consonantal) Tradition 1.1. Short III-y (‫ )ואעש‬and Pseudo-cohortative (‫ואקטלה‬, ‫ואע(י)דה‬, ‫ )ואק(ו)מה‬Forms The clearest point of departure is a comparison focusing on the respective distributions of short versus long III-y (‫ ואעש‬versus ‫ )ואעשה‬forms and pseudo-cohortative versus non-pseudo-cohortative forms (‫ואקטלה‬, ‫ואעידה‬, ‫ ואקומה‬versus ‫ואקטל‬, ‫ואע(י)ד‬, ‫)ואק(ו)ם‬ in the combined Masoretic written-reading biblical tradition and in relevant non-Masoretic biblical and extra-biblical material. Significantly, in the case of such forms the Tiberian written and reading traditions are in near total harmony (with the exception of a few instances of ketiv-qere; see below, §2.2.2). Tables 2 and 3 give the raw numbers and percentages across representative corpora in various biblical traditions and extra-biblical sources. Table 2: Incidence of short 1st-person III-y wayyiqṭol (‫ )ואעש‬forms across representative ancient Hebrew corpora (see §4.0 for citations) MT SP BS Non-LBH+ LBH+ BDSS NBDSS Writings 18/21 28/66 6/13 7/25 3/10 1/11 1/22 0/2 (85.7%) (42.4%) (46.2%) (28%) (30%) (9.1%) (4.5%) (0%) Torah Proph. 17. 1st-person Wayyiqṭol 389 Table 3: Incidence of pseudo-cohortative 1st-person wayyiqṭol (‫ואקטלה‬, ‫ואע(י)דה‬, ‫ )ואק(ו)מה‬forms across representative ancient Hebrew corpora (see §4.0 for citations) MT BDSS NBDSS SP BS Torah Proph. Non-LBH+ LBH+ Writings 4/105 19/254 8/26 69/127 21/55 23/31 34/106 4/7 (3.8%) (7.5%) (30.8) (53.9%) (38.2%) (73.3%) (32.4%) (57.1%) Chart 1 visually displays the incidence of short 1st-person III-y (‫ )ואעש‬and pseudo-cohortative 1st-person (‫ואקטלה‬, ‫ואעידה‬, ‫)ואקומה‬ forms in representative ancient Hebrew biblical traditions and extra-biblical sources as percentages of potential cases. Chart 1: Percentages of short 1st-person III-y (e.g., ‫ )ואעש‬and pseudocohortative 1st-person (‫ואקטלה‬, ‫ואעידה‬, ‫ )ואקומה‬forms across representative ancient Hebrew traditions as percentage of potential cases 100 90 short III-y ‫ואעש‬ 85.7 ps-cohort ‫ואקטלה‬ 80 73.3 70 60 57.1 53.9 50 42.4 40 46.2 38.2 30.8 30 28 32.4 30 20 10 3.8 9.1 7.5 4.5 0 0 MT Torah MT Prophets MT NonLBH+ Writings MT LBH+ BDSS NBDSS SamPent BS Short (‫ )ואעש‬forms dominate in the Tiberian Torah, where pseudo-cohortative forms are rare. Conversely, in the BDSS, NBDSS, the SP, and BS, short III-y forms are relatively infrequent. In the MT pseudo-cohortative (‫ )ואקטלה‬forms appear to be somewhat more characteristic of poetic than of prose texts outside of 390 The Historical Depth of the Tiberian Reading Tradition LBH+ (occurring in about a third of the potential cases in nonLBH+ parts of the Writings), but gain ascendancy only in LBH. They are also variously typical of other late corpora, e.g., the BDSS, NBDSS, the SP, and BS, in which, proportionally, they are between eight and eighteen times as common as in the written tradition of the Tiberian Torah. See below, §1.4, for discussion of the situation in Masoretic CBH outside the Torah, i.e., in the Prophets and Writings. The apparent diachronic significance of the variations in use of the short and pseudo-cohortative patterns discussed above finds support in Iron Age epigraphy. Though the limited corpus of Hebrew inscriptions is devoid of 1st-person wayyiqṭol forms, the Meshaʿ Stele, written in the related Canaanite dialect of Moab, contains several. Here III-y 1st-person wayyiqṭol forms are consistently short, e.g., ‫‘ ואעש‬and I made’ (lns 3, 9), ‫‘ וארא‬and I saw’ (ln. 7), ‫‘ ואבן‬and I built’ (ln. 9), ‫‘ ואשב‬and I captured’ (ln. 12). At the same time, forms eligible for pseudo-cohortative morphology show no indication thereof, e.g., ‫‘ ואהרג‬and I killed’ (lns 11, 16), ‫‘ ואהלך‬and I went’ (lns 14–15), ‫‘ ואקח‬and I took’ (lns 17, 19– 20), ‫‘ ואסחב‬and I dragged’ (ln. 18), ‫‘ ואמר‬and I said’ (ln. 24), ‫ואשא‬ ‘and I carried’ (ln. 30), and ‫‘ וארד‬and I descended’ (ln. 31). And to forestall the suggestion that a final a might be realised, but not orthographically represented (i.e., spelled defectively), it is critical to note the apparent marking of final a in such forms as ‫בללה‬ ‘at night’ (ln. 15) and ‫‘ בנה‬he built’ (ln. 18). Such spellings lead one to expect that similar orthography would have been employed in the case of pseudo-cohortative wayyiqṭol morphology, had it been in use. To summarise: evidence from several biblical traditions (MT, the BDSS, SP) and extra-biblical sources (the Meshaʿ Stele, 17. 1st-person Wayyiqṭol 391 the NBDSS, BS) converges to depict two diachronic trends involving 1st-person wayyiqṭol forms, namely, (a) a decline over time in the short III-y pattern in favour of the long pattern, e.g., ‫ואעש‬ > ‫ואעשה‬, and (b) increased usage in the later period of the pseudo-cohortative pattern in the case of other wayyiqṭol forms, e.g., ‫ואשלחה > ואשלח‬, ‫ואע(י)דה > ועעד‬, ‫ואק(ו)מה > ואקם‬. 1.2. Long III-y (‫)ואעשה‬, Hifʿil (‫ )ואעיד‬and Qal II-w/y (‫ )ואקום‬Forms Because the respective alternants of III-y and pseudo-cohortative 1st-person wayyiqṭol forms involve vowel-final versus consonantfinal realisations, the distinctions are orthographically transparent, e.g., ‫ ואעש‬versus ‫ ואעשה‬and ‫ ואשלח‬versus ‫ואשלחה‬, ‫ ואעד‬versus ‫ואע(י)דה‬, ‫ ואקם‬versus ‫ואק(ו)מה‬. More complex is the situation of the long alternatives to short forms in a number of weak verbal patterns, especially, qal II-w/y qal, and in hifʿil. See Table 4. Table 4: Short and long 1st-person wayyiqṭol forms in the Tiberian tradition 1CS 3MS 1CPL III-y ‫ וָ ֶא ֱע ֶשה‬,‫וָ ַא ַעש‬ ‫וַ יַ ַעש‬ ‫ וַ נַ ֲע ֶשה‬,‫וַ נַ ַעש‬ hifʿil ‫ וָ ָא ִעיד‬,‫וָ ָא ֵעד‬ ‫וָ יָ ַעד‬ ‫ *וָ נָ ִעיד‬,‫*וָ נָ ֵעד‬ II-w/y ‫ וָ ָאקּום‬,‫*וָ ָא ָקם‬ ‫וַ יָ ָקם‬ ‫ וַ נָ קּום‬,‫*וַ נָ ָקם‬ 1.2.1. Short versus Long III-y Morphology: ‫ ואעש‬versus ‫ואעשה‬ Thanks to their orthographic transparency, the most straight-forward evidence again involves III-y verbs, where long and short forms are distinguished by the presence and absence, respectively, of word-final mater heh. Table 5, an inverse of Table 2 392 The Historical Depth of the Tiberian Reading Tradition above, gives the relevant statistics, while Chart 2 presents a visual comparison of long and pseudo-cohortative forms. Table 5: Incidence of long 1st-person III-y forms (e.g., ‫ )ואעשה‬across representative ancient Hebrew traditions MT SP BS Torah Proph. Non-LBH+ LBH+ BDSS NBDSS Writings 3/21 38/66 7/13 18/25 7/10 10/11 21/22 2/2 (14.3%) (57.6%) (53.8%) (72%) (70%) (90.9%) (95.5%) (100%) Chart 2: Percentages of long 1st-person III-y (e.g., ‫ )ואעשה‬and pseudocohortative 1st-person (‫ואקטלה‬, ‫ואע(י)דה‬, ‫ )ואק(ו)מה‬forms across representative ancient Hebrew traditions as percentage of potential cases 100 long III-y ‫ואעשה‬ 95.5 ps-cohort ‫ואקטלה‬ 100 90.1 90 80 72 73.3 70 70 57.6 60 57.1 53.9 53.8 50 37 40 32.4 30.8 30 20 10 14.3 3.8 7.5 0 MT Torah MT Prophets MT NonLBH+ Writings MT LBH+ BDSS NBDSS SamPent BS As noted above, short forms (‫ )ואעש‬dominate long forms (‫)ואעשה‬ in the Tiberian Torah. Conversely, in a phenomenon crucially limited to 1st-person forms, the long III-y pattern (‫ )ואעשה‬substantially outnumbers the short pattern (‫ )ואעש‬in late material: Tiberian LBH+, the BDSS and NBDSS, the SP, and BS—the same corpora that witness regular usage of pseudo-cohortative ‫ואקטלה‬ morphology. Notably, long forms also occur in the majority of cases in the MT Prophets and the non-LBH+ Writings (see below, 17. 1st-person Wayyiqṭol 393 §1.4). In the former there is no corresponding high frequency of tokens of the pseudo-cohortative pattern, while in the latter the increase is significant, but less than in LBH+ proper; these facts are discussed in detail below, §1.4. 1.2.2. Short versus Long Hifʿil and Qal II-w/y Morphology: ‫ ואעד‬versus ‫ ואעיד‬and ‫ ואקם‬versus ‫ואקום‬ Turning to additional verb classes in which a distinction between short and long wayyiqṭol forms obtains, namely hifʿil and II-w/y qal, one confronts a degree of orthographic ambiguity. While plene spellings such as ‫ ואעיד‬and ‫ ואקום‬likely reflect long morphology, the corresponding spellings ‫ ואעד‬and ‫ ואקם‬are ambiguous. Theoretically, the latter spellings might have been intended to reflect short morphology, but could conceivably be defective representations of long morphology (but see below, §1.3.1). Nor does treatment of such forms in the reading tradition resolve the matter. Many forms written like ‫ ואקם‬and ‫ ואעד‬are realised with long morphology—‫ וָ ָא ֻקם‬and ‫—וָ ָא ִעד‬but there are significant exceptions (see below, §2.0). One must proceed with caution. Even so, it is difficult to ignore the striking distribution patterns. Significantly, a trend similar to that witnessed in the case of 1st-person III-y wayyiqṭol forms (‫ ואעש‬versus ‫ )ואעשה‬also obtains in the case of 1st-person hifʿil (‫ ואעד‬versus ‫ )ואעיד‬and II-w/y qal (‫ ואקם‬versus ‫ )ואקום‬wayyiqṭol forms. Table 6 lists the relevant data for the written (consonantal) component of the Tiberian biblical tradition and for several other representative ancient Hebrew corpora. 394 The Historical Depth of the Tiberian Reading Tradition Table 6: Incidence of long 1st-person III-y (‫)ואעשה‬, hifʿil (‫)ואעיד‬, and IIw/y (‫ )ואקום‬wayyiqṭol forms: number of long forms out of number of combined short, long, and pseudo-cohortative forms (percentage long; for citations, see §4.0) MT Proph. NonLBH+ Writings LBH+ BDSS III-y 3/21 38/66 (14.3%) (57.6%) 7/13 (53.8%) 18/25 (72%) 7/10 10/11 21/22 2/2 (70%) (90.9%) (95.5%) (100%) long 1/12 14/33 (8.3%) (42.4%) — 9/21 (42.9%) 3/33 (9.1%) — 10/21 2/2 (47.6%) (100%) long + 1/12 17/33 ps-cohor (8.3%) (51.5%) — 19/21 2/2 5/5 13/13 2/2 (90.4%) (100%) (100%) (100%) (100%) hifʿil Verb Class TOTALS qal ps-cohor Torah 0/12 (0%) BS 10/13 2/2 (76.9%) (100%) 3/5 (60%) 3/13 (23.1%) 0/6 (0%) 9/15 (60%) 1/3 (33.3) II-w/y ps-cohor 0/6 (0%) 1/15 (6.7%) 2/3 (66.7%) 7/21 1/3 3/3 (33.3%) (33.3%) (100%) II-w/y long + ps-cohor 0/6 (0%) 10/15 (66.7%) 4/39 61/114 (10.3%) (53.5%) long + 4/39 65/114 ps-cohor (10.3%) (57%) 0/3 (0%) SP 2/5 (40%) II-w/y long long 14/21 (66.7%) 0/2 (0%) NBDSS 0/3 (0%) — 4/5 (80%) — 1/5 (20%) — 3/3 (100%) 21/21 1/3 3/3 5/5 (100%) (33.3%) (100%) (100%) — 8/16 (50%) 41/67 7/15 12/19 35/40 4/4 (61.2%) (46.7%) (63.2%) (87.5%) (100%) 10/16 (62.5%) 58/67 10/15 18/19 39/40 4/4 (86.6%) (66.7%) (94.7%) (97.5%) (100%) Visual comparisons of the incidence of long and pseudo-cohortative wayyiqṭol morphology in the representative corpora are presented, respectively, in charts 3 and 4. 17. 1st-person Wayyiqṭol 395 Chart 3: Percentages of long 1st-person III-y (‫)ואעשה‬, hifʿil (‫)ואעיד‬, and II-w/y (‫ )ואקום‬wayyiqṭol forms in representative ancient Hebrew corpora 63.2 70 33.3 40 42.9 46.7 50 42.4 50 61.2 66.7 72 58.3 60 53.5 60 57.6 70 40 100 total long 100 100 long II-w/y ‫ואקום‬ 80 80 76.9 90 87.5 long hifʿil ‫ואעיד‬ 95.5 long III-y ‫ואעשה‬ 90.9 100 30 8.3 10.3 10 14.3 20 0 0 0 0 MT Torah MT Prophets MT NonLBH+ Writings MT LBH+ BDSS NBDSS SamPent BS Though limited sample sizes and/or the fragmentary nature of some corpora leave conspicuous gaps in the data, trends in the use of long III-y, hifʿil, qal II-w/y and in long plus pseudo-cohortative 1st-person wayyiqṭol forms are broadly discernible. The Tiberian Torah reflects classical infrequency of long and pseudocohortative forms and the representative Second Temple corpora exhibit noticeable concentrations of both. Also, it is important to point out that where long morphology does not obtain in Second Temple corpora, more often than not the text resorts to pseudocohortative, rather than short morphology. In this way, between them, long and pseudo-cohortative morphology largely crowd out short morphology in late material. 396 The Historical Depth of the Tiberian Reading Tradition Chart 4: Percentages of long III-y (‫)ואעשה‬, long + pseudo-cohortative hifʿil (‫ואעיד‬, ‫ )ואע(י)דה‬and qal II-w/y (‫ואקום‬, ‫)ואקומה‬, and total long + wayyiqṭol (‫ואעשה‬, ‫ואעיד‬, ‫ואע(י)דה‬, ‫ואקום‬, ‫ )ואקומה‬forms in representative ancient Hebrew corpora 100 100 100 97.5 95.5 100 100 94.7 90.9 100 100 100 86.6 hifʿil long+ps-cohort qal long+ps-cohort 66.7 70 53.8 51.5 57 57.6 60 72 total long+cohort 62.5 66.7 70 90.4 80 III-y long 100 90 100 100 50 33.3 40 30 8.3 10.3 10 14.3 20 0 0 MT Torah MT Prophets MT non- MT LBH+ BDSS NBDSS SP BS LBH+ Writings Perhaps surprising is the status of the Tiberian Prophets and non-LBH+ Writings, both broadly classified as CBH. Different from the situation of the pseudo-cohortative discussed above, where such forms are conspicuously lacking from the Masoretic Pentateuch, Prophets, and, to a lesser extent, the non-LBH+ Writings, when it comes to long forms, the Prophets and nonLBH+ Writings show concentrations similar to those of acknowledged Second Temple material. This matter is discussed in detail below, §1.4. 1.3. Anticipating Potential Objections Before proceeding, however, it is worth considering some potential objections. 17. 1st-person Wayyiqṭol 397 1.3.1. Spelling Variation versus Linguistic Variation First, focusing on hifʿil and II-w/y qal 1st-person wayyiqṭol morphology, and excluding III-y forms, it is reasonable to question the linguistic significance of the distinction between apparently short and long spellings. This doubt applies to all representative ancient Hebrew traditions. Beginning with the MT, do the Torah’s typical short spellings, like ‫ואקם‬, and long spellings, like ‫ואקום‬, elsewhere in the Bible reflect a genuine morphological difference, or are they merely divergent orthographic representations of the same form? After all, though a spelling like ‫ ואקום‬with mater waw almost certainly represents a form along the lines of the Tiberian ̊̄ qů̄ ̊̄ ̊̄́m, the Masoretic Torah’s spelling without long-pattern wå-ʾå waw, ‫ואקם‬, is ambiguous: conceivably defective for the same long ̊̄ qů̄ ̊̄ ̊̄́m realisation or representing something more akin to wå-ʾå ̊̄ qåm, ̊̄̊̄́ *wå-ʾå as in the corresponding Tiberian 3MS, 3FS, and 2MS forms. Given the notoriously variable character of spelling in the Tiberian written tradition (Barr 1989; cf. Andersen and Forbes 2013), is it reasonable to interpret this spelling discrepancy in linguistic terms? The view espoused here is that 1st-person wayyiqṭol spelling practices that distinguish the Tiberian Torah from the rest of the Bible have linguistic, not just orthographic, import. Three lines of argumentation may be cited in support of this view. First, plene wayyiqṭol spelling in the consonantal components of the Tiberian tradition outside the Torah and in the SP is limited to 1st-person forms, while the relevant 2nd- and 3rd-person wayyiqṭol forms preserve short orthography. 398 The Historical Depth of the Tiberian Reading Tradition Second, the dominant plene spelling of relevant standard yiqṭol (< PS yaqtulu/a) forms in all persons—‫אקום‬, ‫תקום‬, ‫—יקום‬ throughout the Tiberian and Samaritan written traditions makes it clear that long orthography was an option. If the prominent distinction in spelling between 1st-person wayyiqṭol forms in the Torah (‫ )ואקם‬and in the rest of the Bible (‫ )ואקום‬were merely a function of divergent orthographic policies, one might reasonably expect the regular incidence of defective standard yiqṭol (< PS yaqtulu/a) forms in the Torah and/or long 2nd- and 3rd-person wayyiqṭol spellings beyond the Torah. The fact that 1st-person wayyiqṭol forms in the Torah pattern orthographically like their 2nd- and 3rd-person counterparts and not like 1st-, 2nd-, and 3rdperson yiqṭol forms, while in the rest of the Bible 1st-person forms depart from the short morphology typical of 2nd- and 3rd-person wayyiqṭol forms in favour of the plene spelling characteristic of standard yiqṭol (< PS yaqtulu/a) forms suggests a morphological change in 1st-person forms, specifically a shift from the short template (‫ואעד‬, ‫ )ואקם‬in the Torah to the standard long template (‫ואעיד‬, ‫ )ואקום‬in the rest of the Bible. Finally, the distribution of short and long III-y forms in the Torah—predominantly short (‫—)ואעש‬and beyond—mixed, but predominantly long (‫—)ואעשה‬supports the linguistic significance of analogous distribution patterns in the case of hifʿil and II-w/y qal forms. The foregoing arguments apply to 1st-person wayyiqṭol morphology outside the MT as well. In the BDSS, the NBDSS, the SP, and BS there is a marked spelling difference between III-y, hifʿil, and qal II-w/y wayyiqṭol forms in the 1st person (‫ואעשה‬, ‫ואעיד‬, ‫)ואקום‬, on the one hand, and 2nd and 3rd person (‫ויעש‬, ‫ויעד‬, 17. 1st-person Wayyiqṭol 399 ‫)ויקם‬, on the other. At the same time, there is striking orthographic similarity between 1st-person III-y, hifʿil, and II-w/y qal wayyiqṭol (‫ואעשה‬, ‫ואעיד‬, ‫ )ואקום‬and yiqṭol III-y, hifʿil, and II-w/y qal forms in all relevant persons (‫יעשה‬, ‫יעיד‬, ‫( )יקום‬see further Hornkohl 2013a, 171–80). To summarise: in all the cited representative sources and traditions of ancient Hebrew, there is compelling evidence that the once-strong association unifying 1st-person wayyiqṭol morphology with 2nd- and 3rd-person wayyiqṭol morphology shifted in the Second Temple Period to one linking 1st-person wayyiqṭol morphology and standard, i.e., long, yiqṭol (< PS yaqtulu/a) morphology (or cohortative morphology; see below, §1.4). This new association is regularly manifest in the long spelling of hifʿil and II-w/y qal wayyiqṭol morphology unique to 1st-person forms. 1.3.2. Group versus Individual Distribution Patterns In the interests of clarity and convenience, the presentation of data to this point has been according to corpus, rather than individual composition. Yet, it is fair to ask whether the corporate statistical profiles are representative of the individual constituent works. MT Torah All books in the Masoretic Torah show strong preferences for short (‫ואעש‬, ‫ואעד‬, ‫ )ואקם‬1st-person wayyiqṭol forms, to the near total exclusion of long and pseudo-cohortative morphology, which justifies their combined treatment in this study. See Table 7. 400 The Historical Depth of the Tiberian Reading Tradition Table 7: Long (‫ואעשה‬, ‫ואעיד‬, ‫ )ואקום‬and pseudo-cohortative (‫ואקטלה‬, ‫ואע(י)דה‬, ‫ )ואק(ו)מה‬1st-person wayyiqṭol forms in the Tiberian Torah long IIIy Genesis 1/4 Exodus 0/2 Leviticus — Numbers 0/2 Deut.eroy 2/13 Torah 3/21 long hifʿil 0/3 0/1 0/2 1/2 0/4 1/12 long qal IIw/y 0/2 — 0/1 — 0/2 0/5 total long 1/9 0/3 0/3 1/4 2/19 4/38 pseudo-cohortative 3/42 0/8 0/8 1/6 0/41 4/105 MT Prophets It was noted above that the books of the Former and Latter Prophets resemble those of the Pentateuch in terms of relatively low incidence of pseudo-cohortative (‫ואקטלה‬, ‫ואע(י)דה‬, ‫)ואק(ו)מה‬ forms, but show comparatively high incidence of long (‫ואעשה‬, ‫ואעיד‬, ‫ )ואקום‬forms. There is, however, variation within the distribution. Samuel and Judges are outliers of a sort. Though pseudocohortative forms represent minorities in the two books, between them they account for a disproportionately high number of the cases in the Prophets as a whole (12 of 13). When it comes to long forms, Kings favours long III-y (‫ )ואעשה‬forms, but not long hifʿil (‫ )ואעיד‬and II-w/y qal (‫)ואקום‬ forms, whereas Samuel shows strong preference for ‫ואעשה‬, ‫ואעיד‬, and ‫ ואקום‬forms. Indeed, the counts of long morphology in Samuel alone are largely responsible for the difference in incidence of long forms between the Former and Latter Prophets. Excluding the outlier Samuel, the books of the Prophets, Former and Latter alike, are broadly similar in terms of incidence of long forms, making up from about one-third to one-half of the potential cases—far higher than in the books of the Masoretic Torah, sim- 17. 1st-person Wayyiqṭol 401 ilar to the non-LBH+ Writings, but lower than in LBH+. See Table 8. Table 8: Long (‫ואעשה‬, ‫ואעיד‬, ‫ )ואקום‬and pseudo-cohortative (‫ואקטלה‬, ‫ואע(י)דה‬, ‫ )ואק(ו)מה‬1st-person wayyiqṭol forms in the Tiberian Prophets long IIIy Joshua 2/3 Judges 1/2 Samuel 8/8 Kings 2/2 Isaiah7 1/2 Jeremiah 6/11 Ezekiel 11/22 The XII 6/14 F. Proph. 13/15 L. Proph. 24/49 Prophets 37/64 long hifʿil 1/4 2/3 3/3 0/2 1/1 2/6 2/6 3/5 6/12 8/18 14/30 long qal IIw/y — — 2/2 0/3 1/1 0/1 4/4 2/3 2/5 7/9 9/14 total long 3/7 3/5 13/13 2/7 3/4 8/18 17/32 11/22 21/32 39/76 60/108 pseudo-cohortative 1/20 5/14 7/25 0/17 0/13 2/53 3/68 1/44 13/76 6/178 19/254 MT Writings Because the Writings include LBH material together with compositions of likely classical or unknown provenance, it seems judicious to segregate LBH+ and non-LBH+ material. And, indeed, when one filters out the LBH+ figures from those of the rest of the Writings, two distinctive patterns emerge. In terms of long (‫ואעשה‬, ‫ואעיד‬, ‫ )ואקום‬forms, the non-LBH+ material shows an incidence broadly comparable to that of the Former and Latter 7 Given the relatively small numbers of relevant forms in Isaiah, it is perhaps not surprising that no component of the book presents a distinctive concentration of long or pseudo-cohortative forms. Long forms come in 1/1 and 2/2 potential cases in Isa. 1–39 and 40–55, respectively, but not in Isa. 56–66 (in one potential case). MT Isaiah contains no pseudo-cohortative forms. 402 The Historical Depth of the Tiberian Reading Tradition Prophets. The relatively high incidence of pseudo-cohortative (‫ואקטלה‬, ‫ואע(י)דה‬, ‫ )ואק(ו)מה‬forms in the non-LBH+ Writings, mainly Psalms (6/14 cases outside of Ps. 119), but also Job’s po- etry (2/11 cases), is possibly genre-driven, as poetic style may have favoured the relatively early use of forms not (yet) characteristic of contemporary non-poetic style. See Table 9. Table 9: Long (‫ואעשה‬, ‫ואעיד‬, ‫ )ואקום‬and pseudo-cohortative (‫ואקטלה‬, ‫ואע(י)דה‬, ‫ )ואק(ו)מה‬1st-person wayyiqṭol forms in the Tiberian Writings Psalms (Ps 119 Job (Job narrative Proverbs Qohelet Ruth Esther Daniel Ezra Nehemiah Chronicles Writings Non-LBH+ LBH+ long IIIy 3/7 — 1/2 long hifʿil — — — long qal IIw/y — — 1/1 total long 3/7 — 2/3 pseudo-cohortative 12/21 6/7) 6/15 — — — — 4/4) 3/4 2/2 — — 6/7 1/1 5/11 3/3 24/37 7/13 17/24 — — — — — 0/1 8/8 1/2 9/11 — 9/11 — — — — 1/1 1/1 10/10 2/2 15/15 1/1 14/14 3/4 2/2 — — 7/8 2/3 23/29 6/7 48/63 8/14 40/49 — 1/1 — — 10/18 17/22 31/69 0/7 77/153 8/25 69/128 For their part, the LBH+ works present 1st-person wayyiqṭol usage profiles unlike those of any other books or corpora in the MT. They consistently display clear preferences for long (‫ואעשה‬, ‫ואקום‬, ‫ )ואעיד‬morphology and in all but one case have marked accumulations of pseudo-cohortative (‫ואקטלה‬, ‫ואע(י)דה‬, ‫ )ואק(ו)מה‬morphology. Long forms comprise the majority in every LBH+ composition—Qohelet (2/2), Daniel (7/8), Ezra (2/3), Nehemiah (23/29), and Chronicles (6/7). Pseudo-cohortative forms 17. 1st-person Wayyiqṭol 403 make up sizeable proportions of the relevant cases in Ps. 119 (6/7), Job 1–2 and 42.7–17 (4/4), Qohelet (1/1), Daniel (10/18), Ezra (17/22), and Nehemiah (31/69). Chronicles is an outlier when it comes to pseudo-cohortative 1st-person wayyiqṭol morphology, completely eschewing forms of this type (in seven potential cases).8 While long forms are common in both the Prophets and the Writings, pseudo-cohortative forms dominate only in LBH+ material. In summary: the non-LBH+ Writings join the books of the Prophets in rather common use of long 1st-person wayyiqṭol morphology, but show a stronger inclination to pseudo-cohortative morphology, possibly due to poetic style. The LBH+ material shows strong preference for long morphology throughout and, excluding Chronicles, far higher incidence of pseudo-cohortative forms than any non-LBH+ Masoretic book except for Psalms. Chronicles resembles LBH+ material in its preference for long 8 This may be a result of Chronicles’ preference for long morphology, which is similar to that of MT Samuel, but perhaps more self-consciously systematic. Despite one clear-cut short form—‫‘ ואגד‬and I have said’ (MT 1 Chron. 17.10) || ‫‘ והגיד‬and (the Lord) says’ (MT 2 Sam. 7.11)—the Chronicler’s predilection for long morphology is such that he leaves unchanged long forms in his sources—‫‘ ואהיה‬and I was’ (MT 1 Chron. 17.5 = MT 2 Sam. 7.6; MT 1 Chron. 17.8 = MT 2 Sam. 7.9); ‫‘ ואבנה‬and I built’ (MT 2 Chron. 6.10 = MT 1 Kgs 8.20)—but, in the interest of consistency, levels divergent morphology, whether pseudo- cohortative, ‫‘ ואכרית‬and I cut off’ (MT 1 Chron. 17.8) || ‫( ואכרתה‬MT 2 Sam. 7.9), or short, ‫‘ ואקום‬and I arose’ (MT 2 Chron. 6.10) || ‫( ואקם‬MT 1 Kgs 8.20); ‫‘ ואשים‬and I placed’ (MT 2 Chron. 6.11) || ‫( ואשם‬MT 1 Kgs 8.21). 404 The Historical Depth of the Tiberian Reading Tradition forms, but, perhaps due to this preference, includes no pseudocohortative forms. The Biblical Dead Sea Scrolls Several upshots of the fragmentary character of the BDSS mean that care must be taken in interpreting the distribution of 1stperson wayyiqṭol variants. Considerations include the infrequency or total non-preservation of certain forms, the potential skewing of the broader picture due to the idiosyncrasies of betterpreserved manuscripts, and the arbitrary nature of the specific forms preserved. Thus, while pseudo-cohortative morphology is fairly well represented in the BDSS, relatively few cases that might showcase a distinction between short and long morphology are extant, especially with regard to hifʿil and II-w/y qal forms. See Table 10. Table 10: Long (‫ואעשה‬, ‫ואעיד‬, ‫ )ואקום‬and pseudo-cohortative (‫ואקטלה‬, ‫ואע(י)דה‬, ‫ )ואק(ו)מה‬1st-person wayyiqṭol in the BDSS: Select scrolls and totals 1QIsaa 1Q8 4Q51 4Q70 4Q80 11Q5 BDSS long IIIy — — 1/1 1/1 1/1 — 7/10 long hifʿil 1/1 — — — — — — long qal IIw/y — — — 0/1 — — 0/2 total long 1/1 — 1/1 1/2 1/1 — 7/12 pseudo-cohortative 6/12 0/2 3/3 0/2 2/2 5/5 21/55 Beginning with pseudo-cohortative forms, it must be asked whether their apparently high incidence is due largely to the fact that they are especially frequent in the largest scroll, 1QIsaa, which accounts for over 25 percent of BDSS material (Abegg 17. 1st-person Wayyiqṭol 405 2010, 25), but whose linguistic profile is rather exceptional within the broader corpus (Tov 2012, 100–10; Young 2013; Reymond 2014, 11; Rezetko and Young 2014, 138–39; Hornkohl 2016a, 1020). Likewise, the prevalence of pseudo-cohortative forms in the biblical component of 11QPsalmsa (11Q5) is at least partially due to the chance preservation there of relevant sections of Ps. 119, which also in the MT exhibits an accumulation of pseudo-cohortative forms. Similarly, two of the three pseudo-cohortative forms (as well as the single long III-y form) in 4QSamuela (4Q51) are also found in MT Samuel. In light of these considerations, it is worth entertaining the possibility that the concentration of pseudo-cohortative forms in the BDSS, rather than being broadly representative, is to some extent an accident born of their fragmentary state and the capricious nature of their preservation. Even so, a strong argument that long and pseudo-cohortative forms are more characteristic of the BDSS than of the Tiberian written tradition can be sustained if, upon examination of parallel cases, one perceives a consistent pattern of difference. As things stand, in most instances (49 out of some 67 unambiguous cases), the MT and the BDSS textual versions agree on form. The remaining 18 may be sorted as in Table 11. Table 11: Instances of variation in 1st-person wayyiqṭol: MT versus BDSS Total MT short || BDSS long MT long || BDSS short MT non-ps-cohort. || BDSS ps-cohort. MT ps.-cohort. || BDSS non-ps-cohort. 2 1 13 2 Total excluding 1QIsaa 1 0 7 2 406 The Historical Depth of the Tiberian Reading Tradition When the MT and the BDSS differ with regard to 1st-person wayyiqṭol morphology, it is more common for the MT to exhibit short (‫ואעש‬, ‫ואעד‬, ‫ )ואקם‬or non-pseudo-cohortative (‫ואקטל‬, ‫ואע(י)ד‬, ‫ )ואק(ו)ם‬morphology than for the BDSS to do so. The relative incidence of BDSS pseudo-cohortative (‫ואקטלה‬, ‫ואע(י)דה‬, ‫)ואק(ו)מה‬ morphology is especially striking. And, crucially, this remains true even if one corrects for such skewing factors as 1QIsaa’s disproportionate size and atypical linguistic profile and if one excludes LBH+ 1st-person wayyiqṭol forms (which are pseudocohortative in both the MT and the BDSS). Though the vagaries of fragmentation preclude certainty, the comparative accumulation of pseudo-cohortative forms in the BDSS is arguable evidence of a direction of change from the shorter forms preserved in the MT to longer forms in the BDSS. This is consistent with BDSS treatment of other linguistic features, which more closely conforms to Second Temple conventions than does the MT (Hornkohl 2016a). The Non-Biblical Dead Sea Scrolls Large gaps in the evidence rule out a complete picture. However, among the extant cases of the 1st-person wayyiqṭol, short (‫ואעש‬, ‫ואעד‬, ‫ )ואקם‬forms are extremely rare and long (‫ואעשה‬, ‫ואעיד‬, ‫)ואקום‬ and pseudo-cohortative (‫ואקטלה‬, ‫ואע(י)דה‬, ‫ )ואק(ו)מה‬forms are far more common, though not necessarily in the same texts. See Table 12. The Thanksgiving Scroll (1QHa), which offers the greatest number of examples by far, uses pseudo-cohortative forms wherever possible and long morphology in III-y forms. The Apocryphon of Jeremiah Ca (4Q385a) and Cd (4Q389) also exhibit concentrations of pseudo-cohortative morphology, but are too 17. 1st-person Wayyiqṭol 407 broken to sustain more extensive conclusions. The highly fragmentary 4QReworked Pentateuchb (4Q364) appears to prefer long forms—two of three candidates, all short in the MT9—but shows low incidence of pseudo-cohortative forms (just one of six). A similar pattern of long, but not pseudo-cohortative, morphology might also characterise 4QPseudo-Ezekiele (4Q391), but cases are too few to draw firm conclusions, a situation typical of other scrolls as well. In sum, though severely obscured by fragmentation, the apparently high incidence of long and pseudo-cohortative 1st-person wayyiqṭol morphology in the NBDSS is consistent with broader Second Temple trends. Table 12: Long (‫ואעשה‬, ‫ואעיד‬, ‫ )ואקום‬and pseudo-cohortative (‫ואקטלה‬, ‫ואע(י)דה‬, ‫ )ואק(ו)מה‬1st-person wayyiqṭol in the NBDSS: Select scrolls and totals 1QHa 4Q364 4Q385a 4Q389 4Q391 NBDSS long IIIy 7/7 1/2 — — 2/2 10/11 long hifʿil — 1/1 — 1/1 — 2/6 long qal IIw/y — — — — — 0/3 total long 7/7 1/2 — 1/1 2/2 12/20 pseudo-cohortative 6/6 1/6 4/4 3/4 0/1 23/31 Samaritan Pentateuch The Samaritan written tradition displays strong proclivity for long 1st-person wayyiqṭol morphology. In contrast to the rarity of forms such as ‫ואעשה‬, ‫ואעיד‬, and ‫ ואקום‬in the MT (3/21 III-y, 1/12 hifʿil, 9 ‫‘ ונעלה‬and we ascended’ (4Q364 f24a–c.15) || ‫( וַ ֵ֔ ַנ ַעל‬MT Deut. 3.1); ‫‘ ואשליך‬and I cast’ (4Q364 f26fbii+e.1) || ‫( ָו ַַֽא ְּש ִל ְ֙ך‬MT Deut. 9.21); but ‫‘ ואר] ֯א‬and I saw’ (4Q364 f26bi.6) = ‫( וָ ֵֵ֗א ֶרא‬MT Deut. 9.16). 408 The Historical Depth of the Tiberian Reading Tradition 0/5 II-w/y qal, 4/38 total), they are the rule in the SP (21/22 IIIy, 10/10 hifʿil, 4/5 II-w/y qal, 35/37 total). See Table 13. Table 13: Long (‫ואעשה‬, ‫ואעיד‬, ‫ )ואקום‬and pseudo-cohortative (‫ואקטלה‬, ‫ואע(י)דה‬, ‫ )ואק(ו)מה‬1st-person wayyiqṭol in the SP (figures of long out of total short and long forms; figures in brackets represent the total of long and pseudo-cohortative forms out of total short, long, and pseudo-cohortative forms) long III-y long hifʿil long qal II-w/y total long pseudocohortative SP Gen. 4/4 3/3 2/2 9/9 1/42 SP Exod. 3/3 1/1 — 4/4 3/8 — 3/3 0/1 3/4 1/8 1/2 2/2 — 3/4 1/6 1/1 (4/4) 2/2 (3/3) 16/16 (20/20) 28/42 10/10 (13/13) 4/5 (5/6) 35/37 (39/41) 34/106 SP Lev. SP Num. SP Deut. 13/13 SP 21/22 When it comes to pseudo-cohortative 1st-person wayyiqṭol (‫ואקטלה‬, ‫ואע(י)דה‬, ‫ )ואק(ו)מה‬morphology, however, the Samaritan situation is more complex. Overall, the proportion of 34 of 106 cases is far higher than MT Torah’s of 4 of 105. However, in the books of the Tetrateuch (Genesis–Numbers) the totals in the two traditions are comparable—Samaritan 6 of 64 versus Tiberian 4 of 64—with little in the way of disharmony between the two.10 In Deuteronomy, conversely, the SP has pseudo-cohortative forms in 28/42 cases, against a total absence of pseudo-cohortative forms in the 41 MT cases. The uniqueness of SP Deuteronomy 10 SP pseudo-cohortative || MT non-pseudo-cohortative: Exod. 3.8, 17; 6.5; Lev. 26.13. SP non-pseudo-cohortative || MT pseudo-cohortative: Gen. 41.11; 43.21. 17. 1st-person Wayyiqṭol 409 is particularly striking when its 1st-person way-yiqṭol profile is compared to that of SP Genesis, which has a comparable number 1st-person wayyiqṭol cases, but a far lower incidence of pseudocohortative morphology (1/42). While it may be tempting to hypothesise sweeping linguistic, compositional, and/or text-critical explanations for the innerSamaritan diversity between the SP Tetrateuch and SP Deuteronomy, their differential treatment of 1st-person wayyiqṭol forms turns out to be casual. When the specific verbs that obtain as nonpseudo-cohortative and pseudo-cohortative 1st-person wayyiqṭol forms are analysed, there emerges striking consistency in treatment throughout the SP. With just two exceptions, individual verbs take one pattern or the other, not both. See Table 14 (p. 420). Table 14 lists the 49 verbs that account for the 106 potential cases of pseudo-cohortative 1st-person wayyiqṭol morphology in the SP. The 72 tokens of non-pseudo-cohortative morphology (‫ואקטל‬, ‫ואע(י)ד‬, ‫ )ואק(ו)ם‬in the SP represent 32 different verbs, while the 34 tokens of pseudo-cohortative (‫ואקטלה‬, ‫ואע(י)דה‬, ‫ )ואק(ו)מה‬morphology represent 19 different verbs. Crucially, only two verbs present both non-pseudo-cohortative and pseudocohortative alternants—‫( אמר‬18 non-pseudo-cohortative cases in Genesis [11], Exodus [2], Leviticus [2], and Deuteronomy [3]; two pseudo-cohortative cases, in Exodus and Deuteronomy) and ‫( שים‬one non-pseudo-cohortative case in Genesis, one pseudo-cohortative case in Deuteronomy). Thus, despite the surface-level statistical profiles, there is virtually no basis for claiming a distinction in 1st-person wayyiqṭol morphology between SP Deuter- 410 The Historical Depth of the Tiberian Reading Tradition onomy and the rest of the SP. SP Deuteronomy’s apparently exceptional character vis-à-vis the SP Tetrateuch results merely from Deuteronomy’s use of a number of verbs unused elsewhere in the Torah. Those that appear in Deuteronomy and elsewhere either share the preservation of non-pseudo-cohortative morphology or, more rarely, present with pseudo-cohortative morphology in both the Samaritan Tetrateuch and Deuteronomy. Only among verbs exclusive to Deuteronomy is there a noticeable concentration of pseudo-cohortative morphology. Presumably, were these to appear in SP Genesis–Numbers, an analogous percentage would also have pseudo-cohortative morphology. See Table 14 (following page). Sjörs (2021a, 20–25) notes that pseudo-cohortative lengthening in the SP is used with a limited number of semantic classes of verbal lexemes, including motion verbs and verbs of appropriation. Crucially, Sjörs (2021b) observes no such semantic correlation in LBH, where the extent of lengthened 1st-person wayyiqṭol morphology demands a more comprehensive explanation. Stepping back for a broader perspective on Samaritan 1stperson wayyiqṭol morphology in comparison with other sources and traditions, the SP joins LBH+ and the DSS in displaying an overwhelming preference for long ( ‫ואעשה‬, ‫ואעיד‬, ‫ )ואקום‬forms and shows incidence of pseudo-cohortative (‫ואקטלה‬, ‫ואע(י)דה‬, ‫ )ואק(ו)מה‬forms between that characteristic of Tiberian CBH (Torah, Prophets, non-LBH+ Writings) and what obtains in Tiberian LBH+ and the NBDSS. The diachronically advanced stage of Samaritan 1st-person wayyiqṭol morphology relative to that 17. 1st-person Wayyiqṭol 411 in the Tiberian Torah is consistent with the broad linguistic profiles of the two traditions (Ben-Ḥayyim 2000, 3–4). Table 14: Alphabetical list of non-pseudo-cohortative and pseudo-cohortative 1st-person wayyiqṭol verbs in the SP non-pseudocohortative # 1CS 1CPL 1 ‫אחר‬ 2 ‫אכל‬ 3 ‫אמר‬ ‫אמר‬ 4 ‫בוא‬ 5 ‫ברך‬ 6 7 8 9 10 11 12 ‫הבדיל‬ ‫הביא‬ ‫הוליך‬ ‫הפקיד‬ 13 14 ‫חבא‬ 15 16 ‫חשך‬ 17 ‫יקץ‬/‫קוץ‬ pseudocohortative # 1CS 1CP 1 2 ‫אמר‬ ‫דבר‬ ‫הגיד‬ 3 4 ‫הקריב‬ ‫השיב‬ ‫השם‬/ ‫השה‬ ‫חלם‬ 5 ‫השליך‬ 6 ‫התחנן‬ 7 ‫התנפל‬ 8 ‫התפלל‬ 9 ‫זכר‬ ‫החרים‬ ‫הלך‬ non-pseudocohortative # 1CS 1CPL 18 ‫ירא‬ 19 ‫ישב‬ ‫ישב‬ 20 ‫לקח‬ ‫לקח‬ 21 22 23 24 ‫נשא‬ ‫נתן‬ 25 26 27 28 29 ‫קדד‬ ‫קלל‬ ‫קרא‬ 30 ‫שאל‬ 31 ‫שים‬ 32 ‫שחט‬ ‫סוב‬/‫סבב‬ ‫ספר‬ ‫פתח‬ ‫צעק‬ pseudocohortative # 1CS 1CP 10 ‫ירד‬ 11 ‫כתת‬ 12 ‫נסע‬ 13 ‫עבר‬ 14 ‫פסל‬ 15 ‫קרב‬ 16 ‫שים‬ 17 ‫שלח‬ 18 ‫שרץ‬ 19 ‫תפש‬ Ben Sira Of the relatively few relevant forms preserved in manuscripts of BS, all potentially long cases are long (‫ואפנה‬, ‫וארים ;ואצפה‬, ‫)ואביט‬, while four of seven potentially pseudo-cohortative cases are 412 The Historical Depth of the Tiberian Reading Tradition pseudo-cohortative (‫ואהללה‬, ‫ואברכה‬, ‫)ואשחקה‬. Two of the three non-pseudo-cohortative are long (‫וארים‬, ‫)ואביט‬. Only one strong form is left unlengthened (‫)ואתפלל‬. Thus, the extant BS 1st-person wayyiqṭol forms pattern like those of other Second Temple sources, with strong inclination for long and pseudo-cohortative 1st-person morphology. See Table 15. Table 15: Long and pseudo-cohortative 1st-person wayyiqṭol in Ben Sira MS long long long II- total pseudoIII-y hifʿil w/y qal long cohortative SirB 2/2 2/2 — 4/4 3/6 11Q5 — — — — 1/1 2/2 — 4/4 4/7 TOTALS 2/2 Conclusion Drilling down beneath the surface-level statistical profiles of 1stperson wayyiqṭol morphology across ancient Hebrew sources and traditions, one finds broad support for the hypotheses suggested by the corporate surveys in §§1.1–2 above. Indeed, far from contradicting the postulated diachronic contours, the details of a granular analysis of individual compositions validates distinguishing among the CBH of the Torah, the CBH of the Prophets and non-LBH+ Writings, and the late chronolects reflected in MT LBH+, the BDSS and NBDSS, SH, and BS. 17. 1st-person Wayyiqṭol 413 1.4. 1st-person Wayyiqṭol Morphology and Historical Depth in the Masoretic Written Tradition 1.4.1. Short III-y (‫ )ואעש‬and Pseudo-cohortative (‫ואקטלה‬, ‫ואע(י)דה‬, ‫ )ואק(ו)מה‬Forms The Meshaʿ Stele’s exclusive use of short III-y 1st-person wayyiqṭol (‫ )ואעש‬forms and lack of pseudo-cohortative wayyiqṭol (‫ואקטלה‬, ‫ואע(י)דה‬, ‫ )ואק(ו)מה‬forms (see above, §1.1) tally with the Masoretic Torah’s preference for short 1st-person morphology. Likewise, the striking affinity for long and pseudo-cohortative 1st-person wayyiqṭol forms among late non-Tiberian biblical traditions—the BDSS, the SP—and extra-biblical sources—the NBDSS, BS—is strong evidence of the historical authenticity of the Masoretic LBH+ preference for long and pseudo-cohortative wayyiqṭol morphology. Since the morphological shifts away from short forms seen thus far are not confined to the Tiberian reading and/or written tradition, but—even after probing beneath the surface-level statistical profiles—prove to be characteristic of late biblical and extra-biblical corpora more generally, there are no grounds for attributing the expanded use of long and pseudo-cohortative morphology to medieval or even Byzantine scribal intervention, much less to anachronistic medieval vocalisation (but see below, §2.0). Despite the Tiberian consonantal tradition’s status as a product of scribal transmission, necessarily entailing the possibility of textual fluidity, the shift from short 1st-person wayyiqṭol forms in the Tiberian Torah to long and pseudo-cohortative alternatives in Masoretic LBH+ is broadly consistent with patterns 414 The Historical Depth of the Tiberian Reading Tradition seen in early and late non-Masoretic sources. The crystallisation of Masoretic 1st-person wayyiqṭol morphology plausibly dates to Second Temple times, though, relative to contemporary sources, it must be considered conservative by dint of its comparative preservation of short morphology. 1.4.2. Long III-y (‫)ואעשה‬, Hifʿil (‫ )ואעיד‬and Qal II-w/y (‫ )ואקום‬Forms The argument advanced to this point is consistent with, but does not exhaust the evidence. The data sustain more far-reaching conclusions. Not only are long 1st-person wayyiqṭol forms— ‫ואעשה‬, ‫ואעיד‬, ‫—ואקום‬the norm in Tiberian LBH+ and other late written traditions; they are also common in what is generally considered CBH material outside the Pentateuch, e.g., the MT Prophets and non-LBH+ Writings, where their incidence is closer to that seen in MT LBH+ than to that in the MT Torah. For the sake of convenience, Chart 3 is reproduced below as Chart 5. Against the background of the associations already established—i.e., classical short, on the one hand, and late long and pseudo-cohortative, on the other—how are the specific profiles of the MT Prophets and non-LBH+ Writings—involving the apparently early distribution of long, but not pseudo-cohortative forms—to be explained? 17. 1st-person Wayyiqṭol 415 Chart 5: Percentages of long 1st-person III-y (‫)ואעשה‬, hifʿil (‫)ואעיד‬, and IIw/y (‫ )ואקום‬wayyiqṭol forms in representative ancient Hebrew corpora 60 53.9 33.3 33.3 40 100 70 42.9 50 42.4 50 61.2 66.7 58.3 60 53 56.3 57.6 70 100 100 total long 80 76.9 long II-w/y ‫ואקום‬ 87.5 long hifʿil ‫ואעיד‬ 72 80 long III-y ‫ואעשה‬ 95.5 90 90.9 100 30 8.3 10.5 10 14.3 20 0 0 0 0 MT Torah MT Prophets MT NonLBH+ Writings MT LBH+ BDSS NBDSS SamPent BS Since long orthographic forms (‫ואעשה‬, ‫ואעיד‬, ‫ )ואקום‬are ab- sent from the Torah’s written tradition, but common in the rest of the MT—again, not just in LBH+, but outside the Pentateuch more generally—one might venture the hypothesis that long forms were not originally characteristic of any CBH material and pin responsibility for the difference between the CBH of the Torah (where short forms dominate) and CBH outside the Torah (where long forms are quite standard) on late scribes. These copyists—it seems reasonable to conjecture—might have preserved the ancient orthographic integrity of the venerated Torah more strictly than that of the rest of CBH, which was allowed to ‘drift’ in the direction of LBH+. In this way, 1st-person wayyiqṭol forms in the MT Torah could have been kept pristinely short, while elsewhere in CBH they were updated under the influence of later morphological trends. The theory, while attractive, is contradicted by the data. 416 The Historical Depth of the Tiberian Reading Tradition Key in this connection is the unambiguous orthographic evidence of long 1st-person III-y (‫ )ואעשה‬and pseudo-cohortative (‫ואקטלה‬, ‫ואע(י)דה‬, ‫ )ואק(ו)מה‬forms, the incidence of which is compared in Chart 6. Chart 6: Incidence of long 1st-person III-y (‫ )ואעשה‬and pseudo-cohortative 1st-person (‫ואקטלה‬, ‫ואע(י)דה‬, ‫ )ואק(ו)מה‬forms across representative ancient Hebrew traditions as percentage of potential cases long ‫ואעשה‬ 100 ps-cohort ‫ואקטלה‬ 90.9 95.5 100 90 80 72 73.3 70 70 57.6 60 58.3 57.1 53.9 50 37 40 32.4 30.8 30 20 10 14.3 3.8 7.5 0 MT Torah MT Prophets MT Non- MT LBH+ BDSS NBDSS SamPent BS LBH+ Writings Generally speaking, frequency of long (‫ )ואעשה‬forms positively correlates with frequency of pseudo-cohortative (‫ואקטלה‬, ‫ואע(י)דה‬, ‫ )ואק(ו)מה‬forms. Thus, both largely lack in the MT Torah, but are common in MT LBH+ and in other late corpora, biblical and extra-biblical alike. The glaring exception is the MT Prophets, where long forms are frequent (57.6 percent), whereas pseudocohortative forms are rare (7.5 percent). Returning to the speculative hypothesis proffered above, i.e., that 1st-person wayyiqṭol forms may have been more or less uniformly short throughout CBH and that only outside the Torah underwent contemporisation in line with late linguistic customs—on this assumption, it 17. 1st-person Wayyiqṭol 417 would be reasonable to expect a marked increase in both long IIIy forms and pseudo-cohortative forms in CBH outside the Torah. For if late scribes felt free to append final heh to originally short 1st-person III-y wayyiqṭol forms according to Second Temple convention, i.e., changing ‫ ואעש‬to ‫ואעשה‬, then it is reasonable to expect that they would also have felt free to do the same where necessary to expand the use of pseudo-cohortative forms, changing ‫ ואקטל‬to ‫ואקטלה‬, etc., since these were no less characteristic of Second Temple Hebrew. Crucially, this state of affairs does not obtain. Against the norm in the MT Torah, and similar to MT LBH+ and other late corpora, the MT Prophets show an affinity for long 1st-person III-y wayyiqṭol (‫ )ואעשה‬forms. Yet, similar to the MT Torah and against convention in MT LBH+ and other late texts, pseudocohortative (‫ואקטלה‬, ‫ואע(י)דה‬, ‫ )ואק(ו)מה‬forms are largely absent from the CBH of the Prophets. From the admittedly narrow perspective of 1st-person wayyiqṭol forms, then, the written tradition of the MT Prophets is that of neither the MT Torah nor MT LBH+, but reflects some sort of typologically transitional phase between Pentateuchal CBH and LBH+. This leaves us with a tantalising prospect, namely, that of a tri-valent 1st-person wayyiqṭol historical typology: 1. nearly uniformly short (‫ואעש‬, ‫ואעד‬, ‫ואקם‬, ‫ )ואקטל‬morphology in the CBH of the Torah; 2. commonly long (‫ואעשה‬, ‫ואעיד‬, ‫ואקום‬, ‫ )ואקטל‬but rarely pseudo-cohortative morphology in the CBH of the Prophets; 418 The Historical Depth of the Tiberian Reading Tradition 3. commonly long (‫ואעשה‬, ‫ואעיד‬, ‫ )ואקום‬and commonly pseudo-cohortative (‫ואקטלה‬, ‫ואע(י)דה‬, ‫ )ואק(ו)מה‬morphology in LBH+.11 A note on the MT non-LBH+ Writings: their incidence of long (‫ואעשה‬, ‫ואעיד‬, ‫ )ואקום‬forms is similar to that of the MT Prophets, but Psalms especially shows a comparatively high incidence of pseudo-cohortative (‫ואקטלה‬, ‫ואע(י)דה‬, ‫ )ואק(ו)מה‬forms. Given the uncertainty inherent in the linguistic periodisation of poetry, it is difficult to determine whether this relative frequency of pseudocohortative forms is a function of chronolect, poetic genre, another factor or factors, or some combination thereof. It bears explicit acknowledgment at this point that the proposed chronological interpretation of the typology is at odds with certain views current in biblical studies, not least those that see the Torah and other CBH biblical material as products of the postexilic period and/or that reject language as reliable diachronic indicators. The position advocated here is not that alternative evidence should be deprivileged in favour of orthographic and linguistic evidence, but that the latter should receive due attention and be integrated with evidence gleaned from other approaches. 11 The specific distribution patterns seem to militate against the theory (mentioned above, fn. 5) that III-y forms could take pseudo-cohortative morphology in CBH. The general lack of pseudo-cohortative morphology in the reading tradition of the Masoretic Torah and the Prophets suggests that the final ‫ ה‬on III-y forms in those corpora reflects long rather than pseudo-cohortative morphology. This does not apply to LBH+, where pseudo-cohortative forms are plentiful. 17. 1st-person Wayyiqṭol 419 To summarise provisionally, whatever the chronological significance of the typological divisions proposed above, the Tiberian consonantal text reflects a linguistic tradition of considerable historical depth. This is true in terms of both antiquity (i.e., the extent of its reach into the past) and stratification (i.e., the number of linguistic phases to which it bears witness). 1.4.3. 1st-person Wayyiqṭol Morphology and the Linguistic Periodisation of Ancient Hebrew Most discussions of ancient Hebrew diachrony distinguish LBH from CBH (Hornkohl 2013b; Hurvitz 2013). Pre-classical poetic ABH (Mandell 2013) and an intermediate category between CBH and LBH termed TBH also have proponents (Hornkohl 2013a; 2016b). Certain aspects of 1st-person wayyiqṭol morphological diversity are consistent with such a paradigm, especially, the high frequency of short (‫ואעש‬, ‫ואעד‬, ‫ )ואקם‬morphology in the written tradition of the Tiberian Torah and the Meshaʿ Stele, on the one hand, and, on the other, the rarity of short morphology and concomitant accumulation of pseudo-cohortative (‫ואקטלה‬, ‫ואע(י)דה‬, ‫ )וא(ו)קמה‬morphology in Tiberian LBH+ and other traditions and sources that reflect Second Temple Hebrew. Yet the proposed typology also challenges at least one component of the regnant diachronic linguistic paradigm. In the distributions of the 1st-person wayyiqṭol morphological variants in the Tiberian written tradition one confronts a situation that calls for greater nuance than that which typically characterises diachronic discussions. This is because, as noted above (§1.4.2), the three-stage diachronic division of material based on distribution 420 The Historical Depth of the Tiberian Reading Tradition of 1st-person wayyiqṭol morphology is on the surface consistent with neither the ABH-CBH-LBH paradigm nor the CBH-TBH-LBH arrangement, but calls for finer shading within what is conventionally termed CBH. Preliminarily, two explanations suggest themselves. One option is that the Torah’s written linguistic tradition is typologically older than that of the rest of CBH, in which case there may be some justification to distinguishing between CBH1 and CBH2, both typologically prior to LBH (see Elitzur 2015; 2018a; 2018b; 2019; 2022). Alternatively, it is possible to envision a scenario in which original CBH short 1st-person wayyiqṭol morphological dominance gave way to secondary diversity when material outside the Torah was contemporised—not according to LBH, but in line with norms typologically transitional between those of the MT Torah and LBH proper, that is, of a period when long (‫ואעשה‬, ‫ואעיד‬, ‫ )ואקום‬forms were in wide use, but pseudo-cohortative (‫ואקטלה‬, ‫ואע(י)דה‬, ‫ )ואק(ו)מה‬were not. In this case, what appears to be CBH2 would be a result of the updating of CBH in line with TBH conventions. It bears repeating that the similarity between the CBH of the MT Prophets and MT LBH+ involving the incidence of long III-y morphology (‫ )ואעשה‬combined with their difference in regard to pseudo-cohortative (‫ואקטלה‬, ‫ואע(י)דה‬, ‫ )ואק(ו)מה‬forms militates against the view that the potentially secondary status of long 1st-person wayyiqṭol morphology in the CBH of the MT Prophets is due to levelling in line with LBH+ standards, since one should reasonably expect this to have resulted in relatively high incidence of both long and pseudo-cohortative 1st-person wayyiqṭol morphology. 17. 1st-person Wayyiqṭol 421 Pending the examination of more data with these scenarios in mind, they remain conjectural. And, of course, they are not mutually exclusive. Either way, from the perspective of the MT distribution of 1st-person wayyiqṭol forms, it seems necessary to reckon with the reality of some sort of multivalent division of CBH, whether it involves an organic distinction between CBH1 and CBH2 or the artificial creation of CBH2 due the secondary drift of some authentic CBH material in the direction of TBH. Rounding out this examination of 1st-person wayyiqṭol diachrony in the Tiberian written tradition, it is opportune to discuss a few sundry matters. Non-characteristic Diachronic Usages First, though short and pseudo-cohortative forms are charac-teristic, respectively, of classical and post-classical forms of ancient Hebrew, there is no reason to expect that they should be exclusively restricted to the corpora they characterise. According to more nuanced renditions of the dominant diachronic paradigm, many classical features remained available to late writers and copyists, even if the latter may often have opted for contemporary alternatives. By the same token, exceptional pseudo-cohortative forms in apparently classical texts do not necessarily indicate late composition or textual drift, since there is no logical impediment to the early development of a feature whose later expansion makes it characteristically post-classical. The plausibility of diachronically distinct concentrations of 1st-person wayyiqṭol morphology does not preclude the sporadic use of atypical forms at any given stage. 422 The Historical Depth of the Tiberian Reading Tradition The Problem of Archaic Heterogeneity From the perspective of Hetzron’s (1976) principle of archaic heterogeneity the situation is somewhat complex. At first glance, the claim of early short morphological unity among 1st-, 2nd-, and 3rd-person wayyiqṭol forms may appear to contravene expectations. Would it not be more appropriate to posit early wayyiqṭol heterogeneity, e.g., a paradigm consisting of pseudo-cohortative 1st-person forms and short 2nd- and 3rd-person forms, which was later levelled via analogical processes to a uniformly short paradigm, with a few pseudo-cohortative leftovers? While such an approach may seem logical from the narrow perspective of wayyiqṭol morphology, it is neither borne out in the data nor more theoretically attractive than an alternative view. The diachronic pattern of change for the wayyiqṭol paradigm cannot be described as homogenising, since the morphological distinction between 1st-person forms and 2nd- and 3rdperson forms gradually increases, rather than decreases, with time. Moreover, methodologically, early wayyiqṭol paradigmatic heterogeneity is a priori no more compelling a possibility than early heterogeneity viewed from a broader perspective, namely one that includes both the wayyiqṭol paradigm and that of the directive-volitive forms, i.e., the cohortative (‫)אקומה‬, imperative (‫)קום‬, and jussive (‫)יקם‬. Indeed, bringing into consideration this latter paradigm, especially the presumed link between the cohortative (‫ )אקומה‬and the 1st-person wayyiqṭol (‫( )ואקומה > ואקם‬see Hornkohl 2013a, 165–70; Khan 2021, 321–27; see below), it is reasonable to argue that the archaic heterogeneity eventually homogenised was that between the 1st-person wayyiqṭol and cohor- 17. 1st-person Wayyiqṭol 423 tative forms. But from the perspective of the narrow confines of the wayyiqṭol paradigm, this merging of 1st-person wayyiqṭol and cohortative morphology had the effect of increasing, rather than decreasing, heterogeneity. The Relevance of a Recent Proposal Ch. 16, above, focused mainly on Khan’s (2021, 319–40) recently propounded theory of the genesis of ancient Hebrew wayyiqṭol (see especially §§1.0–3.0). It is now opportune to assess his approach in light of what has been said here about the distribution of 1st-person wayyiqṭol forms in ancient Hebrew sources. Crucially for the present discussion, Khan speculates on the timing of the reanalysis he proposes. Some scholars have, indeed, already expressed the view that there was a convergence between the wayyiqṭol form and the modal system during the period of Late Biblical Hebrew [e.g., Bergsträsser 1918–1929, II:§5d; Talshir 1986]. I would like to argue that this had taken place already in Classical Biblical Hebrew…. The most obvious structural manifestation [of the reanalysis of the narrative yiqṭol as a schematised extension of a jussive] is the occurrence of the cohortative jussive form of first person in wayyiqṭol forms. These become particularly frequent in Late Biblical Hebrew (Cohen 2013, 121– 13), but are found sporadically already in the Pentateuch in Classical Biblical Hebrew. (Khan 2021, 321–22, 327) A few brief observations are in order. First, Khan’s proposal arguably conceives of a more profound and pervasive convergence of the wayyiqṭol and directive-volitive paradigms than is usually envisioned. According to Khan, it was not merely by late 424 The Historical Depth of the Tiberian Reading Tradition analogy with cohortative ‫ אקטלה‬that classical ‫ ואקטל‬shifted to ‫ואקטלה‬. Rather, the antecedents of wayyiqṭol ‫ וָ ֶא ְּק ְּט ָלה‬and cohortative (purpose/result) ‫ וְּ ֶא ְּק ְּט ָלה‬had already fused in Iron Age Hebrew. Originally conveying distinct preterite and modal senses, they had merged into a semantically underdifferentiated w-yiqṭol structure broadly associated with temporal consecution. Only later were they re-differentiated via gemination of realis (mostly preterite) waw-yiqṭol > wayyiqṭol in the Second Temple Period. Second, while it is clear that the frequent use of pseudocohortative 1st-person wayyiqṭol morphology is distinctive of LBH+ and other late corpora, Khan’s theory is consistent with a distribution that is not exclusively late. In other words, at the very least, it allows for the early 1st-person wayyiqṭol morphological variety acknowledged above. The dominance of short 1stperson wayyiqṭol morphology especially characteristic of the MT Torah written tradition can be interpreted as a stage in the convergence of preterite waw-yiqṭol and dependent volitional wawyiqṭol where a morphological distinction between the two yiqṭol forms was still largely preserved in the 1st person. Even so, there is no reason to deny the authenticity of sporadic pseudo-cohortative morphology in the Torah and the Former Prophets.12 Khan (2021, 327, 337–38) notes the LBH proliferation of long and pseudo-cohortative forms, providing a theoretical 12 Qimron (2018, 169) also sees the CBH pseudo-cohortative forms as authentically ancient, but claims that their apparent early infrequency is the product of “an illusion created by the defective spelling of the early Biblical books,” i.e., that verbs could be realised with final -a without final mater heh. 17. 1st-person Wayyiqṭol 425 mechanism for well-rehearsed hypotheses concerning late influence of the cohortative on the 1st-person wayyiqṭol and the late partial merger of long and cohortative morphology in wayyiqṭol and yiqṭol more generally. Attempting to flesh out Khan’s argument: on the assumption of early contrast between a uniformly short preterite paradigm (‫ואקם‬, ‫ותקם‬, ‫ )ויקם‬and a mixed modal paradigm (cohortative ‫ואקומה‬, imperative ‫וקום‬, and jussive ‫)ויקם‬, the similarity and narrative frequency of 3rd-person forms (both short) would make them the logical starting point for reanalysis. Convergence of the respective 1st- and 2nd-person forms, which were dissimilar and far less frequent, might be expected to lag. And, at least in the case of the 1st-person, this is exactly what one encounters. Not until the Persian Period does the convergence apparently begun in CBH become common in 1st-person wayyiqṭol forms.13 When it comes to the distribution of long 1st-person wayyiqṭol (‫ואעשה‬, ‫ואעיד‬, ‫ )ואקום‬morphology, Khan (2021, 337–38) notes the differences between the CBH of the Torah and Former Prophets and between the written and reading traditions of CBH material (on the latter, see below, §2.0). He describes the shift as gradual, attributing it to the “merging in function of the cohorta13 The matter of 2nd-person forms lies beyond the scope of this chapter. The BH equivalent of the 3rd-person dependent jussive (purpose/result) ‫ ויקם‬is the imperative, e.g., ‫( וקום‬JM, §116f). Narrative 2nd-person ‫וקום‬ never arose in Hebrew (unless this is behind the late penchant for the infinitive absolute replacing a finite verb (?)). It is not clear whether the expected alterative, 2nd-person dependent jussive (purpose/result) ‫ותקם‬, ever developed. Perhaps the infrequency of 2nd-person narrative forms hindered the expected effects of convergence. 426 The Historical Depth of the Tiberian Reading Tradition tive with the long [i.e., long] yiqṭol form” (Khan 2021, 337). This seems consistent with the position elaborated in Hornkohl (2013a, 165–70), where it is hypothesised that, in addition to late cohortative influence on wayyiqṭol, both the preterite and volitive short yiqṭol forms were subject to constant analogical pressure exerted by the standard yiqṭol (< PS yaqtulu), the semantics of which also included both past (mainly habitual) and modal shades. Whatever the case may be, any proposal for explaining the expanded use of long and pseudo-cohortative 1st-person wayyiqṭol morphology must successfully account for the disparities in their respective CBH frequencies. In the Tiberian written tradition of the Pentateuch, pointedly, long III-y morphology (3 of 21 cases) is comparatively more common than pseudo-cohortative morphology (4 of 105 cases). What is more, long III-y morphology is commonplace in CBH outside the Torah, while it is not until post-exilic Hebrew that pseudo-cohortative morphology becomes frequent. From a perspective of historical depth, Khan’s theory of wayyiqṭol development substantially preposes the starting point for convergence of the three yiqṭol templates employed in wayyiqṭol morphology. 2.0. The Masoretic Reading Tradition We are now in position to investigate the matter of dissonance between the written and reading components of the Tiberian biblical tradition as it manifests in 1st-person wayyiqṭol forms and to evaluate its historical significance. 17. 1st-person Wayyiqṭol 427 2.1. Dissonance and Secondary Character At issue is whether spelling and vocalisation are in harmony as regards short and long 1st-person wayyiqṭol morphology in the case of 1cs and 1cpl hifʿil and II-w/y qal forms. Table 16 compares the Tiberian written (consonantal) and reading (vocalic) traditions in terms of short 1st-person wayyiqṭol morphology. Table 16: Short 1st-person wayyiqṭol morphology according to written (orthographic) and reading (vocalisation) traditions: number of short forms out of total short and long (percentage) (for citations, see §4.0) Torah Orth. Voc. Prophets Orth. Voc. hifʿil 11/12 8/12 (91.7%) (66.7%) 18/31 4/31 (58.1%) (12.9%) qal II-w/y 6/6 (100%) 5/14 (35.7%) hifʿil + qal II-w/y 17/18 9/16 (94.4%) (56.3%) 23/45 (51.1%) 1/4 (25%) Non-LBH+ Writings Orth. Voc. LBH+ Orth. Voc. — — 2/11 (18.2%) 1/11 (9.1%) 0/10 (0%) 0/1 (0%) 0/1 (0%) 0/14 (66.7%) 1/9 (1.1%) 4/41 (9.8%) 0/1 (0%) 0/1 (0%) 2/25 (8.0%) 2/20 (10.0%) Of the 78 cases of hifʿil and qal II-w/y 1st-person wayyiqṭol forms throughout the MT, in just 15 is the vocalisation consistent with short morphology (9 in the Torah, 4 in the Prophets, 2 in LBH+). In the MT Torah the orthography nearly always reflects short morphology—16 of 17 cases, the sole exception the questionably relevant ‫( וַ נַ ִשים‬Num. 21.30). According to the Pentateuch’s vocalisation, by contrast, short morphology comes in just 8 of 15 cases.14 In the Prophets, too, one encounters dissonance: according to the spelling tradition, just over half of the instances (23 of 14 Here and throughout forms with invariable wayyiqṭol vocalic realisa- tion regardless of their orthography, such as qal ‫בֹוא‬, are excluded from the counts. 428 The Historical Depth of the Tiberian Reading Tradition 45) reflect short morphology, but that proportion drops to under ten percent (4 of 41) according to the reading tradition. The nonLBH+ Writings present just one relevant example, both traditions exhibiting long morphology. Only in LBH+ does one encounter relative harmony between the orthography and vocalisation when it comes to 1st-person wayyiqṭol morphology: short forms are rare according to both traditions. Two related observations are in order. 2.1.1. The Diachronic Significance of Dissonance in Classical Biblical Hebrew Material First, the most plausible explanation for the frequent mismatch between long vocalisation and short orthography in 1st-person wayyiqṭol (‫וָ ָא ִעד‬, ‫ )וָ ָא ֻקם‬forms throughout CBH texts is that a comparatively late reading tradition characterised by long 1st-person wayyiqṭol morphology was secondarily imposed upon a written tradition in which the spelling of many such forms reflected earlier short morphology. Since the Tiberian reading tradition coincides at salient points with post-exilic written tradition, it is reasonable to see the vocalisation as a product of Second Temple times. This means that the Tiberian reading tradition presents a stage in the development of 1st-person wayyiqṭol forms more historically advanced than that discernible in the written tradition to which it has been textually wedded. 17. 1st-person Wayyiqṭol 429 2.1.2. The Diachronic Significance of Harmony in Late Biblical Hebrew+ Material Second, the regular written-reading agreement found in LBH+ material is no accident, but rather results from historical proximity. In other words, the fact that the 1st-person wayyiqṭol morphology applied by the reading tradition throughout the MT corresponds so closely to the LBH+ written tradition (e.g., ‫וָ ָא ִעיד‬, ‫ )וָ ָאקּום‬is because, though semi-independent, they are related products of the same period. 2.2. A Need for Nuance But while the foregoing narrative is true as far as it goes, there is more to the story. Indeed, such a broad-strokes account is something of a distortion. Nuance is required. 2.2.1. The Antiquity of Long 1st-person Wayyiqṭol Morphology First, while the extent of the use of long (‫וָ ָא ִעד‬, ‫ )וָ ָא ֻקם‬forms in the reading tradition is more in line with the LBH+ written tradition than with the CBH written tradition, as has already been noted regarding the written tradition, the phenomenon itself—namely, the likely orthographic representation of long (‫ואעיד‬, ‫ )ואקום‬1stperson wayyiqṭol morphology—predates LBH+. This is clear from the particular constellation of long III-y (‫ )ואעשה‬and pseudocohortative (‫ואקטלה‬, ‫ואע(י)דה‬, ‫ )ואק(ו)מה‬1st-person wayyiqṭol morphology in Tiberian CBH outside the Torah, where—like LBH+, but unlike the Torah—long forms diffused, but—like the Torah, but unlike LBH+—pseudo-cohortative forms did not. 430 The Historical Depth of the Tiberian Reading Tradition A similar situation emerges from an examination of the morphological variety of hifʿil and II-w/y qal wayyiqṭol forms tabulated above in Table 16 (above, p. 428). Note that though long morphology’s eclipsing of short morphology in both the LBH+ written and reading traditions is especially striking (2 of 25 and 2 or 20 cases, respectively), the shift was by no means unprecedented. The extensive replacement of short with long morphology in the vocalisation of the Prophets (just 4 of 41 short) is merely the continuation of a trend already well established in the written tradition of the same material (23 of 45 short). The consistency of long vocalisation in the Prophets is probably partially secondary and anachronistic, but it is merely an extrapolation of a trend already begun, just less advanced, in the corpus’s orthography. It is in the Torah, with orthography predominantly indicative of short 1st-person wayyiqṭol morphology, that the partial deviation in favour of long morphology appears especially anachronistic (the reader is once again reminded that the linguistic significance of the long hifʿil and wayyiqṭol spellings has been demonstrated above, in §1.3.1).15 In sum, the incidence of 1stperson wayyiqṭol morphological dissonance between the written and reading components of the Tiberian tradition increases as 15 Cf. Khan (2021), who presents different explanations for long 1st- person wayyiqṭol morphology in the Masoretic written and reading traditions. In the case of the former he seems to envision a gradual process of organic convergence (337), while he attributes the latter to ‘topdown’ imposition unlikely rooted in vernacular usage (339). 17. 1st-person Wayyiqṭol 431 one moves back in time from LBH+ through the Prophets to the Torah. 2.2.2. Dissonance and Diversity within the Tiberian Reading Tradition This leads to a second important observation. Considering the hypothesis that the Tiberian reading tradition is a Second Temple oral realisation that was applied to contemporary texts and retrojected onto earlier material, it would be reasonable to suppose that it might exhibit greater uniformity, or, at the very least, that it would deviate toward late conventions wherever the written tradition was amenable thereto. Reality, however, proves more complex. Despite its clear Second Temple affinities at certain striking points, the Tiberian pronunciation tradition, like the consonantal tradition, is multifarious, routinely preserving features especially characteristic of early material in the face of the influence of later linguistic convention. Focusing on 1st-person wayyiqṭol morphology, this is manifest in linguistic diversity within the Tiberian reading tradition. 1CS versus 1CPL Forms Consider the differential treatment of singular and plural 1st-person wayyiqṭol forms in the Torah (see Khan 2021, 338–39). See Table 17. Whereas 1CS forms often—in 6 of 8 potential cases— combine short spelling with long phonology, in the 1CPL, spelling consistently matches phonology, so that the classical template is preserved except where long spelling obtains. 432 The Historical Depth of the Tiberian Reading Tradition Table 17: 1st-person short and long hifʿil and qal II-w/y wayyiqṭol morphology in the Masoretic reading tradition of the Torah Short Long Invariable Singular ‫אֹולְך‬ ֵ֥ ֵ ָ‫( ו‬Lev 26.13; Deut. 29.4) Plural ‫ד־לֹו‬ ֵ֔ ֶ‫( וַ נַ֙ ג‬Gen. 43.7; Gen. 44.24) ‫( וַ ָנ ֵֶ֥שב‬Gen. 43.21) ‫( וַ נַ ְּק ֵ ָ֞רב‬Num. 31.50) ‫( וַ ָנ ֵָ֥סב‬Deut. 2.1) ‫( וַ נַ ֲח ֵרם‬Deut. 2.34; 3.6) ‫( וָ ָא ִשם‬Gen. 24.47; Deut. 10.5) ‫( וָ ָא ִ ֵ֥בא‬Exod 19.4) ‫( וָ ָא ֻ ֶָ֖קץ‬Lev. 20.23) ‫( וָ ַא ְּב ִ ֵ֥דל‬Lev 20.26) ‫( ָו ַַֽא ְּש ִל ְ֙ך‬Deut 9.21) ‫( וַ נַ ִשים‬Num. 21.30) ‫( וָ ָא ֵ֥ב ֹא‬Gen. 24.42) ‫( וַ נָ ָ֕ב ֹא‬Deut. 1.19) The Sporadic Preservation of Short 1CS Forms But even in the case of 1CS wayyiqṭol forms: though hifʿil and IIw/y qal forms are routinely pointed long where written (and presumably intended) short, in a minority of cases, typologically early short vocalisation is preserved. Several of these might be conditioned, but it is intriguing that all occur in the reading tradition of CBH texts.16 Conversely, the LBH+ reading tradition is very much in sync with the parallel written tradition, strongly preferring long and pseudo-cohortative forms at the expense of short ones. In LBH+, the spelling of 1CS forms nearly always reflects long or pseudo-cohortative morphology (in 38 of 39 cases of hifʿil 16 Four such cases involve the specific verb ‫אֹולְך‬ ֵ ָ‫ו‬, behind whose short form there may well stand phonological factors—perhaps an original diphthong in the first syllable favoured preservation of short morphology in the second (cf. ‫ ואוליך‬in the SP). The preservation of another short form may be attributed to euphony in ‫( וָ ָא ֵ ֶ֖עד ֵע ִ ָׂ֑דים‬Jer. 32.10); cf. ‫וָ ָא ָ֕ ִעיד‬ (Neh. 13.15). That leaves only ‫ֹתֹו ָד ֵָ֔בר‬ ֙ ‫( וָ ָא ֵ ַ֤שב א‬Josh. 14.7), which contrasts with LBH ‫אֹותם ָד ֵָ֗בר‬ ֶָ֜ ‫( וָ ָא ִָֹ֨שיב‬Neh. 2.20). 17. 1st-person Wayyiqṭol 433 and II-w/y qal combined), the sole exception being ‫( וָ ַא ִגד‬1 Chron. 17.10). When it comes to the mere three relevant 1CPL cases, the two traditions once agree on short morphology (‫ וַ ֹנ ֶשב‬Ezra 10.2), once agree on long morphology (‫ וַ נַ ֲע ִמיד‬Neh. 4.3), and once clash (‫ ונשוב‬ketiv ‫ וַ ָנ ָָ֤שב‬qere Neh. 4.9). These exceptional instances of mismatch between orthography and vocalisation in Tiberian LBH+ are doubly important, evincing both the continued independence of the written and reading traditions as well as their close congruence. Indeed, their potential divergence makes their consistent agreement all the more striking. Ketiv-Qere Mismatches A final note on the six relevant instances of ketiv-qere dissonance: these are cases where the disparity occasioned by merging divergent written and reading traditions could not be resolved except by explicit emendation of the written form. See Table 18. Table 18: Ketiv-qere cases involving 1st-person wayyiqṭol forms in Codex Leningrad ketiv ‫ואראה‬ qere ‫וָ ֵא ֶרא‬ Reference Josh. 7.21 Description: ketiv || qere long || short ‫וארב‬ ‫וָ ַא ְּר ֶב ֙ה‬ Josh. 24.3 short || long ‫ואבאה‬ ‫וָ ָא ִביא‬ Josh. 24.8 pseudo-cohort. || long ‫ואוצאה‬ ‫וָ ֲא ַצֶּוַ֤ה‬ Ezra 8.17 pseudo-cohort. || long ‫ואשקולה‬ ‫וָ ֶא ְּש ֳק ָלה‬ Ezra 8.25 pseudo-cohort. || pseudo-cohortative ‫ונשוב‬ ‫וַ ָנ ַָ֤שב‬ Neh. 4.9 long || short Beyond confirming the independence of two related traditions, these do not materially alter the picture drawn to this point. Intriguingly—and contrary to what might be expected, but consistent with what was said above—there is no unambiguous correlation between the ketiv and classical short morphology or 434 The Historical Depth of the Tiberian Reading Tradition between the qere and later long or pseudo-cohortative morphology. This is a further indication that, despite crystallising in the Second Temple Period, the Tiberian reading tradition—including, but not limited to, explicit qere instances—manifests profound historical depth and intricacy, even preserving individual Iron Age phenomena in the face of the standardisation of others. 3.0. Conclusion A detailed study of 1st-person wayyiqṭol forms in the Tiberian written and reading traditions yields typologically rich results. Having established that orthographic variation in the written tradition is as at least partially indicative of typological shifts in morphology, it can be plausibly maintained that the Tiberian written tradition testifies to three typological strata of 1st-person wayyiqṭol development in chronologically suggestive concen-trations. Dissonance between the Tiberian CBH written and reading traditions shows that the reading tradition is typologically later, akin to other Second Temple traditions, including the LBH+ written tradition. However, the Tiberian reading tradition is itself typologically diverse: the relevant vocalisation in CBH is not identical to that in LBH+; 1CS and 1CPL forms receive different treatment in CBH; and there is no clear pattern to ketiv-qere divergence. The extent of long morphology in the reading tradition of CBH material seems more characteristic of the Tiberian written tradition of LBH+ and other late material than of the written tradition of CBH texts. Yet the frequency of long forms in the 17. 1st-person Wayyiqṭol 435 written CBH tradition outside the Torah shows that, in this respect, the reading tradition merely extended and standardised a feature that had diffused prior to LBH+ times. The regularity of the reading tradition’s use of long morphology appears to be anachronistic for the earliest parts of the Bible, but evidence of its initial appearance points to the Iron Age. Common usage of long and pseudo-cohortative morphology can be dated no earlier or later than the LBH+ compositions, and, given the incidence of long morphology in the CBH of the Prophets, its diffusion may well have begun centuries earlier. Long and pseudo-cohortative 1st-person wayyiqṭol morphology joins many other features of the Tiberian reading tradition that deviate from the reading tradition reflected in the consonantal in their early minority incidence followed by later standardisation. 4.0. Citations Table 2 MT Torah—short: Gen. 24.46; 31.10; 41.22; Exod. 6.3; 9.15; Num. 13.33; 23.4; Deut. 2.1, 8, 33; 3.1, 1, 18; 9.15, 16; 10.3, 3, 5; long: Gen. 24.48; Deut. 1.16, 18; MT Prophets—short: Josh. 7.21 (qere); 24.3 (ketiv); Judg. 18.4; Isa. 64.5; Jer. 3.8; 11.5; 15.6; 20.7; 35.10; Ezek. 1.4, 15, 27; 11.16; 12.7; 20.9, 22; 23.13; 24.18; 43.8; 44.4; Hos. 13.7; Zech. 2.1, 5; 4.4, 11, 12; 5.9; 6.4; long: Josh. 7.21 (ketiv); 9.24; 24.3 (qere); Judg. 12.3; 1 Sam. 10.14; 13.12; 26.21; 2 Sam. 7.6, 9; 11.23; 12.22; 22.24; 1 Kgs 8.20; 11.39; Isa. 6.1; Jer. 13.2; 25.17; 31.26; 32.9, 13; 44.17; Ezek. 1.1, 28; 2.9; 8.2, 7, 10; 10.1, 9; 11.1; 16.8; 20.14; Hos. 11.4; Amos 4.10; Zech. 5.1; 6.1; 11.7, 7; Non-LBH+—short: Ps. 18.24; 38.15; 69.12; 73.14; Job 30.9; Prov. 7.7; long: Ps. 69.11, 21; 102.8; Job 7.20; Prov. 8.30, 30; 24.32; MT LBH+—short: Dan. 10.5; Neh. 1.4; 2.11, 13, 15, 15; 4.8; long: Qoh. 4.1, 7; Dan. 8.2, 2, 3, 27; 9.4; 10.8; Ezra 8.15, 17 (qere); Neh. 1.4; 3.38; 7.2; 12.31; 13.25; 1 Chron. 17.5, 8; 2 Chron. 6.10; BDSS—short: 4Q31 2.4 (|| Deut. 3.18); Mur2 f1i.3 (|| Deut. 10.3a); 5/6Hev1b f6–7.10 (|| Ps. 18.24); long: 1QIsaa 51.19 (|| Isa. 64.5 short); 4Q51 f42a.1 (|| 1 Sam. 26.21); 4Q70 f21–22i.3 (|| Jer. 13.2); 4Q73 f2.10 (|| Ezek. 11.1); 4Q80 f14–15.2 (|| Zech. 5.9 short); 436 The Historical Depth of the Tiberian Reading Tradition 4Q112 f14.12 (|| Dan. 8.2); 4Q114 1.7 (|| Dan. 10.8); NBDSS—short: 4Q364 f26bi.6 (|| MT Deut. 9.16); long: 1QHa 10.10, 12, 16, 17; 11.8; 14.27; 16.28; 4Q364 f24a–c.15 (|| MT Deut. 3.1 short); 4Q391 f9.3 (?); f65.4 (?);17 SP—short: Num. 13.33 (|| MT short); long: Gen. 24.46, 48 (|| MT long); 31.10; 41.22; Exod. 6.3; 9.15, 19+ (|| MT Exod. 9.15 short); Num. 23.4; Deut. 1.16 (|| MT long), 18 (|| MT long); 2.1, 8, 33; 3.1, 1, 18; 9.15, 16; 10.3, 3, 5; 18 BS—long: SirB 20v.2 (|| Sir. 51.7), 2 (|| Sir. 51.7). Table 3 MT Torah—pseudo-cohortative: Gen. 32.4; 41.11; 43.21; Num. 8.19; MT Prophets—pseudo-cohortative: Josh. 24.8 (ketiv); Judg. 6.9, 10; 10.12; 12.3, 3; 1 Sam. 2.28; 28.15; 2 Sam. 4.10; 7.9; 12.8, 8; 22.24; Jer. 11.18; 32.9; Ezek. 3.3; 9.8; 16.11; Zech. 11.13; MT non-LBH+ Writings—pseudo-cohortative: Ps. 3.6; 7.5; 69.12, 21; 73.16; 90.10; Job 19.20; 29.17; MT LBH+—pseudocohortative: Ps. 119.55, 59, 106, 131, 147, 158; Job 1.15, 16, 17, 19; Qoh. 1.17; Dan. 8.13, 15, 17; 9.3, 4, 4; 10.16, 16, 19; 12.8; Ezra 7.28; 8.15, 16, 17 (ketiv), 17, 23, 23, 24, 25, 26, 28, 31; 9.3, 3, 5, 5, 6; Neh. 1.4; 2.1, 6, 9, 13; 5.7, 7, 8, 13; 6.3, 8, 11, 12; 7.5; 12.31; 13.7, 8, 9, 9, 10, 11, 11, 13, 17, 17, 19, 19, 21, 21, 22, 30;19 BDSS pseudo-cohortative: 1QIsaa 6.2 (|| Isa. 6.8 unlengthened), 5 (|| Isa. 6.11 unlengthened); 34.12 (|| Isa. 41.9 unlengthened); 40.10 (|| Isa. 48.5 long); 42.8 (|| Isa. 50.7 unlengthened); 51.20 (|| Isa. 64.5 short); 4Q13 f3ii+5–6i.8 (|| Exod. 3.17 unlengthened); 4Q51 3a–e.25 (|| 1 Sam. 2.28 pseudo-cohortative), 9e–i.16 (|| 1 Sam. 10.14 long), f61ii+63–64a–b+65–67.3 (|| 2 Sam. 4.10 pseudo-cohortative); 4Q80 f8–13.19 (|| Zech. 4.4 unlengthened), f14–15.2 (erasure || Zech. 5.9), 2 (|| Zech. 5.9), 4 (|| Zech. 5.10 unlengthened); 4Q83 f19ii–20.31 (|| Ps. 69.12 pseudo-cohortative); 4Q113 f16– 18i+19.5 (|| Dan. 8.3 unlengthened); 11Q5 9.1 (|| Ps. 119.59 pseudo-cohortative); 11.2 (|| Ps. 119.106 pseudo-cohortative); 12.4 (|| Ps. 119.131 pseudocohortative); 13.9 (|| Ps. 119.158 pseudo-cohortative); 20.2 (|| 139.11 unlengthened); NBDSS pseudo-cohortative: 1QHa 12.37; 14.9, 10; 15.23; 17.9, 10; 1Q49 f1.1; 4Q364 f26bi.8; 4Q385 f2.9; 4Q385a f1a–bii.1, 6, 7, f15i.5; 17 The two final ambiguous citations were excluded from the totals in Hornkohl (2013a, 160). 18 The slight difference between the totals here and in Hornkohl (2013a, 160) is due to the inclusion here of SP Exod. 9.19+ (|| MT Exod. 9.15), which was excluded there. 19 Hornkohl (2013a, 162) mentions the cases in Ps. 119 and Job 1, but does not count them in the relevant table’s LBH totals. 17. 1st-person Wayyiqṭol 437 4Q387 f1.7; 4Q389 f2.4, 5, f6.1; 4Q390 f1.6, 6; 4Q437 f2ii.13; 4Q504 f1– 2rv.17; 11Q5 28.5; 11Q19 65.8;20 SP pseudo-cohortative: Gen. 32.6; Exod. 3.8, 17; 6.5; Lev. 26.13; Num. 8.19 (= MT); Deut. 1.19, 19, 43; 2.1, 7+ (MT —), 8, 8, 13, 26, 34, 34; 3.4, 6, 23; 9.15, 17, 18, 20, 21, 21, 21, 25, 26, 26; 10.3, 5, 5; 22.14;21 BS—pseudo-cohortative: SirB 20v.3 (|| Sir. 51.8), 20v.11 (|| Sir. 51.12), 11 (|| Sir. 51.12); 11Q5 21.15 (|| Sir. 51.18) (?); unlengthened: SirB 20v.5 (|| Sir. 51.9); 21r.12 (|| Sir. 51.14) (?); 21r.17 (Sir. 51.19) (?). 22 Table 6 III-y—short and long: see Table 2, above; hifʿil: MT Torah—short: Gen. 43.7, 21; 44.24; Exod. 19.4; Lev. 20.26; 26.13; Num. 31.5; Deut. 2.34; 3.6; 9.21; 29.4; long: Num. 21.30 (?);23 MT Prophets—short: Josh. 14.7; 24.3, 10; Judg. 6.9; 1 Kgs 2.42; 18.13; Jer. 5.7; 32.10; 35.4; 42.21; Ezek. 28.18; 31.15; 39.23, 24; Amos 2.10; Zech. 11.8; long: Josh. 24.6; Judg. 2.1; 6.8; 1 Sam. 10.18; 12.1; 15.20; Isa. 48.5; Jer. 2.7; 11.8; Ezek. 16.50; 36.19; Amos 2.9, 11; Zech. 11.13; pseudo-cohortative: Josh. 24.8; Judg. 10.12; 2 Sam. 7.9; MT LBH+—short: Ezra 10.2; 1 Chron. 17.10; long: Neh. 2.18, 20; 4.3, 7, 7; 6.4; 7.1; 13.15; 1 Chron. 17.8; pseudo-cohortative: Ps. 119.59; Ezra 8.17, 24; Neh. 6.12; 12.31; 13.8, 9, 13, 21, 30; BDSS—pseudo-cohortative: 1QIsaa 40.10 (|| long MT Isa. 48.5); 11Q5 9.1 (|| MT Ps. 119.59); NBDSS—long: 4Q364 f26bii+e.1 (|| short MT Deut. 9.21); 4Q389 f2.2; pseudo-cohortative: 1QHa 17.9; 4Q387 f1.7; 4Q389 f6.1; SP—long: Gen. 43.7, 21; 44.24; Exod. 19.4; Lev. 18.25 (|| qal MT); 20 The slight difference between the totals here and in Hornkohl (2013a, 162) is due to the inclusion here of the (admittedly ambiguous) case in 11Q5 28.5. 21 Since the present citation list is identical to that in Hornkohl (2013a, 162), the difference between the respective tallies is apparently due to an arithmetic error in the latter. 22 The apparent pseudo-cohortative case in 11Q5 21.15 (|| Sir. 51.18) and the apparent unlengthened cases in SirB 21r.12 (|| Sir. 51.14) and 17 (Sir. 51.19) are ambiguous, e.g., is waw-yiqṭol better analysed as wayyiqṭol or we-yiqṭol or should apparently pseudo-cohortative 1CS ‫ואקטלה‬ be interpreted as standard wayyiqṭol with a FS object suffix? 23 On the problematic ‫( וַ נַ ִשים‬Num. 21.30) see Bloch (2007, 149–50); Hornkohl (2013a, 160–61, fn. 5). 438 The Historical Depth of the Tiberian Reading Tradition Lev. 20.26; 26.13; Num. 21.30 (|| long MT; ?); 31.50; Deut. 29.4;24 pseudocohortative: Deut. 2.34; 3.6; 9.21; BS—long: SirB 20v.5 (|| Sir. 51.9); SirB 21r.17 (|| Sir. 51.19); II-w/y: MT Torah—short: Gen. 24.27, 42; Lev. 20.23; Deut. 1.19; 2.1; 10.5; MT Prophets—short: 1 Kgs 3.21; 8.20, 21; Jer. 13.2; Zech. 6.1; long: 1 Sam. 10.14; 28.21; Isa. 51.6; Ezek. 3.15, 23; 8.10; 16.8; Zech. 5.1; Mal. 1.3; pseudo-cohortative: Judg. 12.3; MT non-LBH+—long: Job 38.10; pseudo-cohortative: Ps. 69.21; 90.10; MT LBH+—long: Dan. 8.27; Ezra 8.32; Neh. 2.9, 11, 12, 15, 15, 15; 4.8, 9 (ketiv); 13.7, 25; 2 Chron. 6.10, 11; pseudo-cohortative: Ezra 8.15, 17, 23; Neh. 5.7; 13.7, 11, 17; BDSS— short: 4Q56 f36.2 (|| long MT Isa. 51.16); 4Q70 f21–22i.3 (|| short MT Jer. 13.2); pseudo-cohortative: 4Q51 9e–i.16 (|| long MT 1 Sam. 10.14); NBDSS— pseudo-cohortative: 1QHa 12.37; 4Q504 f1–2Rv.17; 11Q5 28.5; SP—short: Lev. 20.23 ( || short MT); long: Gen. 24.42 (|| short MT), 47 (|| short MT); Deut. 1.19 (|| short MT); 2.1 (|| short MT); pseudo-cohortative: Deut. 10.5 (|| short MT). Table 16 Torah: hifʿil—written and reading short: ‫( וַ נַ֙ גֶ ד־‬Gen. 43.7); ‫( וַ ָנ ֵֶ֥שב‬Gen. 43.21); ‫( וַ נַ֙ גֶ ד־‬Gen. 44.24); ‫אֹולְך‬ ֵ֥ ֵ ָ‫( ו‬Lev. 26.13); ‫( וַ נַ ְּק ֵ ָ֞רב‬Num. 31.50); ‫( ַוַֽנַ ֲח ֵר ֙ם‬Deut. 2.34); ‫( וַ נַ ֲח ֵרם‬Deut. 3.6); ‫אֹולְך‬ ֵ֥ ֵ ָ‫( ו‬Deut. 29.4); written short, reading long: ‫( וָ ָא ִ ֵ֥בא‬Exod 19.4); ‫( וָ ַא ְּב ִ ֵ֥דל‬Lev 20.26); ‫( ָו ַַֽא ְּש ִל ְ֙ך‬Deut. 9.21); written and reading long: ‫וַ נַ ִשים‬ (Num. 21.30); qal II-w/y—written and reading short: ‫( וַ ָנ ֵָ֥סב‬Deut. 2.1); written short, reading long: ‫( וָ ָא ִ ַ֤שם‬Gen. 24.47); ‫( וָ ָא ֻ ֶָ֖קץ‬Lev 20.23); ‫( ָו ַָֽא ִש ֙ם‬Duet. 10.5); invariable: ‫( וָ ָא ֵ֥ב ֹא‬Gen. 24.42); ‫( וַ נָ ָ֕ב ֹא‬Deut. 1.19); Prophets: hifʿil—written and reading short: ‫שב‬ ַ֤ ֵ ‫( וָ ָא‬Josh. 14.7); ‫אֹולְך‬ ֵ֥ ֵ ָ‫( ו‬Josh. 24.3); ‫( וָ ָא ֵ ֶ֖עד‬Jer. 32.10); ‫אֹולְך‬ ֵ ָֹ֨ ָ‫ו‬ (Amos 2.10); written short, reading long: ‫( וָ ַא ִ ֵ֥צל‬Josh. 24.10); ‫( וָ ַא ִ ַ֤צל‬Judg. 6.9); ‫( וָ ָא ִ ַ֤עד‬1 Kgs 2.42); ‫( וָ ַא ְּח ִב ֩א‬1 Kgs 18.13); ‫( וָ ַא ְּש ִ ַ֤ב ַע‬Jer. 5.7); ‫( וָ ָא ִ ַ֤בא‬Jer. 35.4); ‫וָ ַא ִ ֵ֥גד‬ (Jer. 42.21); ‫ַֽאֹוצא־‬ ִ ‫( ָו‬Ezek. 28.18); ‫( וָ ַא ְּק ִ ַ֤דר‬Ezek. 31.15); ‫( וָ ַא ְּס ִ ֵ֥תר‬Ezek. 39.23); ‫( וָ ַא ְּס ִ ֵ֥תר‬Ezek. 39.24); ‫( וָ ַא ְּכ ִ ֛חד‬Zech. 11.8); written and reading long: ‫ַֽאֹוציא‬ ַ֤ ִ ‫( ָו‬Josh. 24.6); ‫( וָ ָא ִ ַ֤ביא‬Judg. 2.1); ‫( וָ א ִ ֵֹ֥ציא‬Judg. 6.8); ‫( וָ ַא ִ ַ֤ציל‬1 Sam. 10.18); ‫( וָ ַא ְּמ ִ ֵ֥ליְך‬1 Sam. 12.1); ‫( וָ ָא ִֵ֗ביא‬1 Sam. 15.20); ‫( וָ ַא ִגַ֤יד‬Isa. 48.5); ‫( וָ ָא ִ ַ֤ביא‬Jer. 2.7); ‫( וָ ָא ִָֹ֨ביא‬Jer. 11.8); ‫( וָ ָא ִ ֵ֥סיר‬Ezek. 16.50); ‫( וָ ָא ִ ַ֤פיץ‬Ezek. 36.19); ‫( וָ ַא ְּש ִ ַ֤מיד‬Amos 2.9); ‫( וָ ָא ִ ַָ֤קים‬Amos 2.11); ‫( וָ ַא ְּש ִ ֵ֥ליְך‬Zech. 11.13); written pseudo-cohortative, reading long (ketiv-qere): ‫וָ ָא ִביא‬ (Josh. 24.8); qal II-w/y—written short, reading long: ‫( וָ ָא ֻ ֵָ֥קם‬1 Kgs 3.21); ‫( וָ ָא ֻ ָ֡קם‬1 Kgs 8.20); ‫( וָ ָא ִָֹ֨שם‬1 Kgs 8.21); ‫( וָ ָא ִ ֶ֖שם‬Jer. 13.2); ‫( וָ ָא ֵֻ֗שב‬Zech. 6.1); written and reading long: ‫( וָ ָא ִ ַ֤שים‬1 Sam. 28.21); ‫( וָ ָא ִ ַ֤שים‬Isa. 51.16); ‫קּום‬ ֮ ‫( וָ ָא‬Ezek. 3.23); ‫וָ ָא ָ֕שּוב‬ (Zech. 5.1); ‫( וָ ָא ִ ַ֤שים‬Mal. 1.3); written long, invariable vocalisation: ‫( וַ נָ ֶ֖בֹוא‬1 Sam. 10.14); ‫( וָ ָא ָֹ֨בֹוא‬Ezek. 3.15); ‫בֹוא‬ ֮ ‫( וָ ָא‬Ezek. 8.10); ‫( וָ ָא ָֹ֨בֹוא‬Ezek. 16.8); Non-LBH+ 24 The total and citation list in Hornkohl (2013a, 160, 163 fn. 17) ex- clude the cases in Lev. 18.25 and Num. 21.30. 17. 1st-person Wayyiqṭol 439 Writings: qal II-w/y—written and reading long: ‫( ָו ַָֽא ִֵ֗שים‬Job 38.10); LBH+: hifʿil—written and reading short: ‫( וַ ֹ֛נ ֶשב‬Ezra 10.2); written short, reading long: ‫( וָ ַא ִגד‬1 Chron. 17.10); written and reading long: ‫( וָ ַא ָֹ֨ ִגיד‬Neh. 2.18); ‫( וָ ָא ִָֹ֨שיב‬Neh. 2.20); ‫( וַ נַ ֲע ִָֹ֨מיד‬Neh. 4.3); ‫( ָו ַַֽא ֲע ִָ֞מיד‬Neh. 4.7); ‫( ָו ַַֽא ֲע ִ ַ֤מיד‬Neh. 4.7); ‫( וָ ָא ִ ֵ֥שיב‬Neh. 6.4); ‫( וָ ַא ֲע ִ ֶ֖מיד‬Neh. 7.1); ‫( וָ ָא ָ֕ ִעיד‬Neh. 13.15); ‫( וָ ַא ְּכ ִ ֵ֥רית‬1 Chron. 17.8); qal II-w/y— written and reading long: ‫( וָ ָא ָ֕קּום‬Dan. 8.27); ‫( וָ ָאקּום‬Neh. 2.12); ‫( וָ ָא ֵ֗שּוב‬Neh. 2.15); ‫( וָ ָא ַֽשּוב‬Neh. 2.15); ‫( וָ ָא ֵ֗קּום‬Neh. 4.8); ‫( ונשוב‬Neh. 4.9 ketiv); ‫( וָ ָא ִ ַ֤ריב‬Neh. 13.25); ‫( וָ ָא ָ֡קּום‬2 Chron. 6.10); ‫( וָ ָא ִ ֵ֥שים‬2 Chron. 6.11); written long, reading short: ‫וַ ָנ ַָ֤שב‬ (Neh. 4.9 qere); written long, invariable vocalisation: ‫( וַ נָ ֶ֖בֹוא‬Ezra 8.32); ‫ָו ַָֽא ֵ֗בֹוא‬ (Neh. 2.9); ‫( וָ ָא ֶ֖בֹוא‬Neh. 2.11); ‫( וָ ָא ֛בֹוא‬Neh. 2.15); ‫( וָ ָא ֶ֖בֹוא‬Neh. 13.7). 18. I-Y WE-YIQṬOL FOR WEQAṬAL By and large in Tiberian BH prose, there is a clearcut functional difference between we-yiqṭol and weqaṭal forms. Whereas the former are used fairly exclusively in 1st- and 3rd-person for what Bybee et al. (1994, 179) call ‘speaker-oriented modality’, i.e., directives indicating the speaker’s will,1 the latter have much broader future force, including indicative meaning and both ‘speaker-oriented’ and ‘agent-oriented modality’ (see Bybee et al. 1994, 176–81; Shulman 1996, 180; Verstraete 2007, 32–35; Cook 2012, 247–48; Dallaire 2014, 39; Hornkohl 2018, 31–32; 2021, 378–80, 383–86). In a well-known functional subcategory of the modality signalled by we-yiqṭol, the structure serves to encode final, e.g., purpose and result, clauses. Though real-world purposes and results (and speaker-oriented modality, more generally) can also be communicated via weqaṭal, the latter much less transparently expresses these meanings. In sum, then, in BH prose we-yiqṭol normally has jussive semantics, whether subordinated to a previous (normally directive volitional) verb (1) or merely coordinate with a previous jussive (2). 1 The parallel 2nd-person form is not we-tiqṭol, but the imperative u-qṭol (JM §116f; cf. Lambdin 1973, 119, §107c; Muraoka 1997). © 2023 Aaron D. Hornkohl, CC BY-NC 4.0 https://doi.org/10.11647/OBP.0310.18 442 (1) (2) The Historical Depth of the Tiberian Reading Tradition ‫ּובעֹוף‬ ְּ ‫מּותנּו וְּׁ יִ ְּׁרדּו ִב ְּד ָֹ֨ ַגת ַה ֶָ֜ים‬ ָׂ֑ ֵ ‫ֹלהים ַנ ֲַֽע ֶ ֵ֥שה ָא ָ ֛דם ְּב ַצ ְּל ֵ ֶ֖מנּו ִכ ְּד‬ ִֵ֔ ‫אמר ֱא‬ ֶ ֹ ‫וַ י‬ ‫ל־ה ָ ַֽא ֶרץ׃‬ ָ ‫ל־ה ֶ ֶ֖ר ֶמש ָ ַֽהר ֵ ֵֹ֥מש ַע‬ ָ ‫ּוב ָכ‬ ְּ ‫ל־ה ֵָ֔א ֶרץ‬ ָ ‫ּוב ָכ‬ ְּ ‫ּוב ְּב ֵה ָמ ֙ה‬ ַ ‫ַה ָש ֵַ֗מיִ ם‬ ‘Then God said, “Let us make humankind in our image, after our likeness, so they may rule over the fish of the sea and the birds of the air, over the cattle, and over all the earth, and over all the creatures that move on the earth.”’ (Gen. 1.26) ‫ֶ֖ית ִל ְּק ַ ֵ֥הל ַע ִ ַֽמים׃‬ ָ ִ‫א ְּת ֵָ֔ך וְּׁ י ְַּּׁפ ְּׁרךֹׁ֖ וְּׁ י ְַּּׁר ֶבָ֑ך וְּ ָהי‬ ֹ ַֽ ‫וְּ ֵ ַ֤אל ַש ַ ֙די יְּ ָב ֵרְך‬ ‘God Almighty bless you and make you fruitful and multiply you, that you may become a company of peoples.’ (Gen. 28.3) By contrast, in order to express more generic futurity and/or the speaker-oriented modality of what convention says should or must happen, rather than we-yiqṭol, weqaṭal is the norm, e.g., (3) ‫הוֶ֖ה נִ ְּק ָרא ָע ֶלָׂ֑יָך וְּׁ יָ ִֽ ְָּֽׁראֹׁ֖ ּו ִמ ֶ ַֽמךָ ׃‬ ָ ְּ‫ל־ע ֵמי ָה ֵָ֔א ֶרץ ִ ֛כי ֵ ֵ֥שם י‬ ַ ‫אּו ָכ‬ ֙ ‫וְּ ָר‬ ‘And all the peoples of the earth shall see that you are called by the name of the LORD, and they shall be afraid of you.’ (Deut. 28.10) Similarly, the weqaṭal ‫ֶ֖ית‬ ָ ִ‫ וְּ ָהי‬in example (2), though perhaps contextually interpretable as purposive (as in the gloss), is formally unspecified for anything more than just futurity, meaning that it can just as well be taken as ‘and you will become’. In most forms of Second Temple Hebrew, the CBH TAM system, with its pragmatically distinct pairs of conversive and non-conversive perfective past forms (wayyiqṭol and qaṭal) and habitual/future forms (weqaṭal and yiqṭol), persists.2 In all forms 2 See Rabin (1958, 155; 1972, 371–73; 1976, 1015–16 fn. 2) on the rare attestation of conversive forms in Talmudic narrative. 18. I-y We-yiqṭol for Weqaṭal 443 of post-exilic Hebrew, however, the system witnesses at least some degree of erosion and, in certain cases, has been nearly or even totally eclipsed. For purposes of the present discussion, a crucial development is the use of the so-called non-conversive forms preceded by the simple conjunction -‫ ו‬with the semantic values they have without the preceding conjunction, i.e., we-qaṭal for perfective past (just like qaṭal) and we-yiqṭol for future (just like yiqṭol). 1.0. Second Temple Evidence 1.1. Late Biblical Hebrew The LBH verbal system, in general, and the use of yiqṭol, more specifically, largely adhere to CBH norms (Cohen 2013, 151–92). Even so, a significant departure from CBH convention is the use of we-yiqṭol for temporally ‘sequential’ eventualities (Cohen 2013, 151, 171–73). Consider example (4): (4) ‫יהם וְּׁ ִ ִֽי ְּׁת ַּ ִֽפ ְּׁללּו‬ ֵֶ֗ ‫א־ש ִמי ֲע ֵל‬ ְּ ‫ם־א ַש ַ ֵ֥לח ֶ ֶ֖ד ֶבר ְּב ַע ִ ַֽמי׃ וְּׁ יִ כָ נְּׁ ֶ֨עּו ַע ִֶ֜מי ֲא ֶ ָ֧שר ִ ַֽנ ְֵּּ֣ק ָר‬ ֲ ‫וְּ ִא‬... ‫ן־ה ָש ֵַ֔מיִם וְּׁ אֶ ְּׁסלַּ ח‬ ַ ‫ִ ִֽויבַּ ְּׁקשּ֣ ּו ָפ ֵ֔ ַני וְּׁ י ָֻׁשֹׁ֖בּו ִמ ַד ְּר ֵכ ֶיהם ָה ָר ִ ָׂ֑עים וַ ֲאנִ ֙י ֶא ְּש ַמע ִמ‬ ‫ת־א ְּר ָ ַֽצם׃‬ ַ ‫אתם וְּׁ אֶ ְּׁר ָפֹׁ֖א ֶא‬ ֵָ֔ ‫ְּל ַח ָטּ‬ ‘…and if I send pestilence against my people, 14 and my people who are called by my name humble themselves, and pray and seek my face and turn from their wicked ways, then I will hear from heaven and will forgive their sin and heal their land.’ (2 Chron. 7.13b–14) The passage presents a complex conditional clause that consists of a compound protasis and a compound apodosis. In both halves of the clause we-yiqṭol constructions comprise all but the first 444 The Historical Depth of the Tiberian Reading Tradition verb. In CBH, these would almost certainly have been weqaṭal forms. A classic diachronic parallel may be seen in example (5): (5) ‫־לי ֶל ֶ֛חם‬ ֵ֥ ִ ‫הֹולְך וְּׁ נָ ִֽתַּ ן‬ ֵ ֵ֔ ‫ּושמָ ַּרנִ י ַב ֶ ַ֤ד ֶרְך ַהזֶ ֙ה ֲא ֶשר ָאנ ִֹכי‬ ְּׁ ‫ֹלהים ִע ָמ ִ ֵ֗די‬ ִֶ֜ ‫ ִאם־יִ ְּה ֶָֹ֨יה ֱא‬... ‫אֹלהים׃‬ ַֽ ִ ‫הו֛ה ִ ֶ֖לי ֵל‬ ָ ְּ‫ל־בית ָא ִ ָׂ֑בי וְּׁ הָ יָ ֧ה י‬ ֵ ‫ וְּׁ שַּ ְּׁב ִ ּ֥תי ְּב ָשלֶ֖ ֹום ֶא‬21 ‫ּובגֶ ד ִל ְּל ַֽבֹש׃‬ ֵ֥ ֶ ‫ֶל ֱא ֶ֖כֹל‬ ‫ן־לי ַע ֵ ֶ֖שר‬ ִ ֵ֔ ‫ֹלהים וְּ כֹל֙ ֲא ֶשר ִת ֶת‬ ָׂ֑ ִ ‫תי ַמ ֵצ ֵָ֔בה יִ ְּה ֶיֶ֖ה ֵבית ֱא‬ ֙ ִ ‫ר־ש ְּמ‬ ֙ ַ ‫ וְּ ָה ֶא ֶבן ַה ֵ֗ז ֹאת ֲא ֶש‬22 ‫ֲא ַע ְּש ֶ ֵ֥רנּו ָ ַֽלְך׃‬ ‘…If God is with me and keeps me in this way that I go, and gives me bread to eat and clothing to wear, and I return to my father’s house in peace, then the LORD will be my God, and this stone, which I have set up for a pillar, will be God’s house. And of all that you give me I will give a full tenth to you.”’ (Gen. 28.20b–22) Here, all conditions save the initial one after ‫‘ ִאם‬if’ are weqaṭal, as is the first verb of the apodosis, ‫‘ וְּ ָה ָיָ֧ה‬then (the LORD) shall be’. These leaves just three non-weqaṭal verbs, which form is precluded due to preverbal elements preventing clause-initial position. Such sequential uses of we-yiqṭol, while constituting a noticeable departure from CBH norms, are relatively rare throughout most of the LBH corpus. Indeed, to the series of six such forms in 2 Chron. 7.14 in example (4) above, Cohen (2013, 172, fn. 42) adds cases in Est. 1.19; Neh. 6.13; 8.15; Dan. 12.4, 10; 2 Chron. 2.15; 14.6.3 Significantly, in his discussion of the LBH verbal system, Cohen (2013, 15) expressly omits Qohelet. While this is under- 3 Cohen (2013, 172 fn. 42) also lists we-yiqṭol cases in Dan. 1.12–13; 1 Chron. 13.2; 2 Chron. 12.8, but these are better seen as having classical purposive semantics. 18. I-y We-yiqṭol for Weqaṭal 445 standable insofar as Qohelet’s verbal system differs markedly from that of the core LBH works—Esther, Daniel, Ezra–Nehemiah, and Chronicles—nevertheless, the language of Qohelet is widely regarded as reflecting a late chronolect (Delitzsch 1877, 190–99; Driver 1898, 474–75; Hurvitz 1990; 2007; Schoors 1992–2004; Seow 1996). Further, when it comes to the matter of non-conversive we-qaṭal and we-yiqṭol forms, Qohelet appears to be farther along the developmental continuum than any other biblical book. In Qohelet, perfective past we-qaṭal routinely comes where one expects wayyiqṭol in CBH,4 whereas future/habitual we-yiqṭol is nearly as common as future/habitual weqaṭal.5 1.2. Dead Sea Scrolls Hebrew 1.2.1. The Biblical Dead Sea Scrolls As should be expected, classical usage of we-yiqṭol is the norm in the BDSS. Even so, in some Qumran renditions of biblical texts a drift from future/imperfective weqaṭal to future/imperfective we4 There are only three cases of wayyiqṭol in the book—1.17; 4.1, 7— against 31 cases of perfective past we-qaṭal: 1.13, 16; 2.5, 9, 9, 11, 12, 13, 14, 15, 15, 17, 18, 20; 3.22; 4.1, 4, 7; 5.13, 13, 18; 8.10, 15, 17; 9.14, 14, 14, 15, 15, 16; 12.9 (?), 9 (?). 5 Schoors (1992–2004, I:86–89) provides a corrective for extreme views, listing 15 cases of classical weqaṭal in the book, to which Qoh. 1.5, 5; 8.10; and 10.3 should be added. Future/habitual we-yiqṭol comes around 13 times: 1.18; 2.19; 6.12; 7.7; 8.10; 12.4, 4, 5, 5, 5, 6, 6, 7. The occurrences in 11.8–9 are passably classical jussives. The unique genre of Qohelet may also have contributed to its rare use of conversive verbal forms. 446 The Historical Depth of the Tiberian Reading Tradition yiqṭol is evident (Muraoka 2000, 210–11; Qimron 2018, 369, fn. 2). Kutscher (1974, 357–58) lists many examples from 1QIsaa, e.g., (6): (6) ...‫שבתותי ויבחורו‬ ̇ ‫כיא כוה אמר יהוה לסריסים אשר ישמורו את‬ ...‫ּובח ֲֹׁ֖רּו‬ ִֽ ָ ‫תֹותי‬ ֵַ֔ ‫ת־ש ְּב‬ ַ ‫רּו ֶא‬ ֙ ‫ים ֲא ֶ ַ֤שר יִ ְּש ְּמ‬ ֙ ‫יס‬ ִ ‫הוה ַל ָ ַֽס ִר‬ ֵ֗ ָ ְּ‫ִכי־כֹה ׀ ָא ַמר י‬ ‘Thus says the LORD to the eunuchs who keep my sabbaths and choose…’ (1QIsaa 46.14–15 || MT Isa. 56.4) The Great Isaiah Scroll is renowned among DSS biblical material for its frequent departures from classical norms, but other examples of DSS biblical material also present cases of we-yiqṭol parallel to weqaṭal in the MT: ‫‘ ויהיו‬and they will be’ (4Q7 f2.3) || ‫( וְּ ָהיַ֤ ּו‬MT Gen. 1.14) ‫‘ ויקמו‬and there will arise’ (4Q9 f3–4.2) || ‫( ְּו ָּ֠קמּו‬MT Gen. 41.30) ‫‘ ויהיה‬and it will be’ (XHev/Se5 f1.5) || ‫( וְּ ָה ָָ֞יה‬MT Exod. 13.14) ‫‘ ואקבצם‬and I will gather them’ (4Q72 f44–50.7) || ‫ים‬ ֮ ‫וְּ ִק ַב ְּצ ִת‬ (MT Jer. 31.8) ‫‘ ו֯ ירעשו‬and (the heavens and the earth) will shake’ (4Q78 f18–20.9) || ‫( וְּ ָר ֲע ֶ֖שּו‬MT Joel 4.16) ‫‘ ויהיו‬and they will be’ (4Q76 4.4) || ‫( וְּ ָהיּו‬MT Mal. 3.17) ‫‘ וי̇ [חנני‬and he will have mercy on me’ (4Q98a f2ii.2) || ‫וְּ ָח ֵנָׂ֑נִ י‬ (MT Ps. 30.11)6 6 It is, of course, possible that one or more of these cases reflect an interpretive rather than a linguistic difference, i.e., purposive/result semantics instead of more broadly future force. 18. I-y We-yiqṭol for Weqaṭal 447 1.2.2. The Non-biblical Dead Sea Scrolls Like the BDSS, the NBDSS by and large demonstrate adherence to the classical norms of the so-called conversive wayyiqṭol and weqaṭal. Yet, it is widely acknowledged that the NBDSS deviate from classical norms much more frequently than the BDSS. This is very clear in the case of use of we-yiqṭol where CBH would opt for weqaṭal (Smith 1991, 59; Muraoka 2000, 210–11; Qimron 2018, 369). An example of Rewritten Bible (or Reworked Scripture), The Temple Scroll (11QTa = 11Q19), with up to 60 cases showcases this usage, both where it cites biblical passages and where it presents independent material (Hornkohl 2021b, 147– 49, esp. fn. 53; a lower figure is reported by Smith 1991, 59). From Temple Scroll biblical material, consider: (7) ‫ויכבס בגדיו ורחץ [במים‬... ‫וְּׁ ִכ ֶב֧ס ְּבגָ ָ ֛דיו וְּ ָר ַ ֵ֥חץ ַב ַ ֶ֖מיִ ם‬... ‘And he will wash his clothes and bathe in water’ (11QTa 51.3 || MT Num. 19.19b) In (7), against the series of two weqaṭal forms in MT Num. 19.19b, 11QTa has an apparently synonymous combination of we-yiqṭol and weqaṭal forms. Further examples from Rewritten Bible texts include: ‫‘ וידבר‬and he will speak’ (4Q175 1.6) || ‫( וְּ ִד ֶבר‬MT Deut. 18.18) ‫‘ ויסוקלוני‬and they will stone me (4Q365 7i.3) || ‫ּוס ָק ֻ ַֽלנִ י‬ ְּ (MT Exod. 17.4) ‫‘ וידבר‬and he will speak’ (11QT 6.15) || ‫( וְּ ִד ֶ ֵ֥בר‬MT Deut. 20.2) 448 The Historical Depth of the Tiberian Reading Tradition ‫‘ וימת‬and (the man) will die (11QT 56.11) || ‫ּומ ֙ת‬ ֵ (MT Deut. 17.12) Such material also furnishes cases without biblical parallels, including: ‫ ויתן מדמו באצבעו על‬...] [‫ויקח הפר השני אשר לעם ויכפר ̇ב ̊ו‬... ...‫קרנות ̇ה[מזבח‬ ‘Then he will take the second bull, the one for the people, and he will make atonement with it [ ]… and he will put some of its blood with his finger on the horns of the altar’ (11QTa 16.14–16) ‫ועשיתה על פי התורה אשר יגידו לכה ועל פי הדבר אשר יואמרו לכה‬ ‫מספר התורה ויגידו לכה באמת‬ ‘and you must act according to the law that they proclaim to you and according to the word that they say to you from the book of the Law and they shall tell to you in truth’ (11QTa 56.3–4; cf. MT Deut. 17.9) ‫ואלוהים ֯אמר לא ידור רוחי באדם לעולם ויחתכו ימיהם מאה ועשרים‬ ‫̇שנה‬ ‘..and God said, “My spirit shall not dwell with man for- ever, and their days shall be determined to be one hundred and twenty years…”’ (4Q252 1.1; cf. Gen. 6.3) 1.3. Samaritan Hebrew Like its Tiberian counterpart, the Samaritan tradition combines a relatively early (primarily consonantal) written component with a comparatively later pronunciation component (that includes consonants and vowels). In general, the Tiberian and Samaritan traditions employ weqaṭal and we-yiqṭol similarly. Divergences 18. I-y We-yiqṭol for Weqaṭal 449 are often explicable as interpretive differences, where one tradition or the other has a more nuanced purposive/result we-yiqṭol in place of a less semantically specialised weqaṭal form or vice versa. Consider, by way of example: ‫) פקדים על הארץ ויחמש‬wyafqəd( ‫) פרעה ויפקד‬wyāš( ‫ויעש‬ ̊̄ SP ‫) את ארץ מצרים בשבע שני השבע׃‬wyɑ̊məš ( (8) MT ‫ת־א ֶרץ ִמ ְּצ ַ ֵ֔ריִם ְּב ֶ ֶ֖ש ַבע ְּש ֵנֵ֥י‬ ֶ ‫ל־ה ָ ָׂ֑א ֶרץ וְּׁ ִחמֵ ש ֶא‬ ָ ‫ַּיע ֲֶשּ֣ה ַפ ְּר ֵ֔עֹה וְּׁ י ְַּּׁפ ֵ ּ֥קד ְּפ ִק ִ ֶ֖דים ַע‬ ‫ַה ָש ָ ַֽבע׃‬ ‘Let Pharaoh do [this] and appoint overseers over the land and take one-fifth of the land of Egypt during the seven plentiful years.’ (Gen. 41.34) In (8), the MT, Joseph’s advice to Pharaoh is conveyed in a varied series of verb forms, consisting of a morphologically long yiqṭol, a morphologically short we-yiqṭol, and a weqaṭal, all apparently with 3rd-person directive force. The SP, conversely, uses a series of we-yiqṭol forms (some morphologically short). If SH ‫ויחמש‬ ̊̄ wyɑməš for MT ‫‘ וְ ִח ֵמש‬and let him take one-fifth of’ is secondary, it seems to have less to do with post-classical we-yiqṭol’s eclipsing of weqaṭal than with the perception that classical we-yiqṭol better suited the context than weqaṭal. There is, however, one relevant systematic change. Where the MT has a weqaṭal form of a I-y qal verb the SP written tradition (like its Tiberian counterpart) is frequently ambiguous, but the SP reading tradition consistently records we-yiqṭol. Though some of the following could conceivably be attributed to interpretive differences, their sheer number shows the broad nature of the shift. 450 The Historical Depth of the Tiberian Reading Tradition ‫ וידעו‬wyiddāʾu ‘and (Egypt) will know’ || ‫( וְּ יָ ְּד ַ֤עּו‬MT Exod. 7.5; see also Exod. 14.4, 18; 29.46; Num. 14.31) ‫ ויצא‬wyiṣṣɑ ‘and (the people) will go out’ || ‫( וְּ יָ ָָֹ֨צא‬MT Exod. 16.4; see also Exod. 17.6; 21.2; 34.34; Lev. 14.3, 38; 16.18, 24; 25.28, 33, 41, 54; Num. 34.4, 9; Deut. 21.2; 23.11) ‫ ויצא‬wyiṣṣɑ ‘and (water) will come out’ || ‫( וְּ יָ ְּצ ֵ֥אּו‬MT Exod. 17.6) ‫ וישב‬wyiššɑb ‘and he will dwell’ || ‫( וְּ יָ ַ ֛שב‬MT Lev. 14.8; Num. 32.17; 35.25) ‫ ויירש‬wyīrɑš ‘and he will possess’ || ‫( וְּ יָ ַרש‬MT Num. 27.11; see also Deut. 3.20; ‫ ויוסיפו‬wyūsīfu ‘and (the officers) will continue’ || ‫( וְּ יָ ְּספּו‬MT Deut. 20.8) ‫ וייראו‬wyīrāʾu ‘and they should fear’ || ‫אּו‬ ֙ ‫( וְּ ָ ַֽי ְֵּּ֣ר‬MT Deut. 28.10; 31.12)7 Another indication that the Samaritan I-y qal weqaṭal to weyiqṭol shift is part of a broad linguistic change is the corresponding Samaritan shift of I-y qal wayyiqṭol (Samaritan w-yiqtol) to we̊̄ r ‘and (the LORD) formed’ || ֩‫יצר‬ qaṭal, e.g., ‫ ויצר‬wyɑṣɑ ֶ ִ‫( וַ י‬MT Gen. 2.7) (Ben-Ḥayyim 2000, 173, §2.9.8), a shift that even affected 7 Also possibly relevant is the case of ‫ וילדו‬wyēlēdu || ‫( וְּ יָ ְּל ֵ֥דּו‬MT Gen. 31.8, 8; see also Exod. 1.19; Deut. 21.15); but see Ben-Ḥayyim (2000, 139, §2.4.3) on the ambiguity of the form. Perhaps also in the case of ‫ויסף‬ ̊̄ ‘and he will add’ || ‫( וְּ יָ ַ ַ֤סף‬MT Lev. 22.14; see also Lev. 27.13, 15, wyɑsəf 19, 27; Num. 32.15); see Ben-Ḥayyim (2000, 139, §2.4.2; above, ch. 11, ̊̄ d ‘and (the §§1.3; 2.4). The shift does not obtain in the case of ‫ וירד‬wyɑrɑ hail) will fall’ || ‫( וְּ יָ ַ ָ֧רד‬MT Exod. 9.19; see also Exod. 11.8; Num. 16.30; ̊̄ q ‘and he will pour’ || ‫( וְּ יָ ַצַ֤ק‬MT Lev. 2.1; see 34.11, 11, 12); ‫ ויצק‬wyɑṣɑ also Lev. 14.15). 18. I-y We-yiqṭol for Weqaṭal 451 ̊̄ k wtɑšɑ ̊̄ b ‘and she went and sat’ || 3FS forms, e.g., ‫ ותלך ותשב‬wtɑlɑ ‫( וַ ֵת ֶלְך֩ וַ ֵָֹ֨ת ֶשב‬Gen. 21.16), which have developed a secondary a–a realisation apparently inherited from the related qaṭal form (Khan 2021, 331; cf. Ben-Ḥayyim 2000, 173, §2.9.8). Together, both of these departures from classical norms that focus on I-y qal verbs—in comparison not just to Tiberian Hebrew, but to most Samaritan verb classes, too—exhibit the penetration of later features into the reading tradition where the written tradition was amenable to the shift. 1.4. Ben Sira Notwithstanding the book’s relatively late provenance, the language of BS—so far as it can be assessed given the extant textual sources—is remarkably classical. Post-classical roots and lexemes abound (Dihi 2004), but the grammar, while not devoid of postclassicisms, is an impressive imitation of CBH. The poetic nature of the material doubtless contributes to its classical mien. Indeed, the poetic nature of BS makes it difficult to detect diagnostically post-classical instances of we-yiqṭol. In an exhaustive discussion, van Peursen (2004, 166–79) surveys we-yiqṭol forms throughout BS’s multiple witnesses and finds CBH parallels for nearly all of them. Arguable exceptions, perhaps indicating the adoption of post-classical conventions, occur in conditional clauses: (9) ‫אם יסור מאחרי אשליכנו ואסגירנו לשדדים׃‬ ‘If he goes astray after this, I will cast him away and hand him over to robbers.’ (SirA 1v.8 = Sir. 4.19b) 452 The Historical Depth of the Tiberian Reading Tradition (10) ‫אם טוב תדיע למי תטיב ויהי תקוה לטובתך׃‬ ‘If you do good, know to whom you are doing it, and there will be hope for the good that you do.’ (SirA 4v.28–29 = Sir. 12.1) (11) ‫וגם אם ישמע לך ויהלך בנחת׃ תן לבך להתירא ממנו‬ ‘And even if he shows regard for you and walks peacefully, commit your heart to being in fear of him.’ (SirA 5r.9 = Sir. 12.11) (12) ‫אם שלך ייטיב דבריו עמך וירששך ולא יכאב לו‬ ‘If you have any possessions, he will speak pleasant words to you, and he will make you poor and it will not grieve him (SirA 5r.27–28 = Sir. 13.5) According to CBH syntactic norms, in place of the above we-yiqṭol usages, one would expect weqaṭal forms, whether encoding an ancillary condition in a compound protasis or beginning a conditional apodosis (bare, clause-initial yiqṭol would also be possible for the latter). 1.5. Rabbinic Hebrew Entirely lacking weqaṭal (and wayyiqṭol) except in biblical citation, RH has regular recourse to we-yiqṭol (in addition to other alternatives) where BH has weqaṭal (Bendavid 1967–1971, II:559–60). Consider the following contrastive pairs of BH and (Tannaitic and Amoraic) RH examples: (13a) ...‫וְּ לֹא־יִ ְּש ֵ֥אּו ָעֹוֶ֖ ן ו ֵָמָ֑תּו‬... ‘…lest they bear guilt and die…’ (Exod. 28.43) (13b) ‫אבל אנו לא נחטא ונמות‬ ‘but we will not sin and die…’ (Sifre Bemidbar 10.33) 18. I-y We-yiqṭol for Weqaṭal 453 (14a) ‫ָעֹלַ֤ ה נַ ֲע ֶל ֙ה וְּׁ י ַָּר ְּּׁ֣שנּו א ֵָֹ֔תּה‬ ‘Let us go up at once and occupy it’ (Num. 13.30) (14b) ‫ונלך ונירש את ארץ ישראל‬ ‘…but we will go and inherit the land of Israel.’ (Sifre Bemidbar 10.33) (15a) ‫אמר ָפ ַנֵ֥י יֵ ֵלֶ֖כּו ַּוהֲנִ ּ֥ח ִתי ָ ַֽלְך׃‬ ָׂ֑ ַ ֹ ‫וַ י‬ ‘And he said, “My presence will go with you, and I will give you rest.”’ (Exod. 33.14) (15b) ‫המתן לי עד שיעברו פנים של זעם ואניח לך‬ ‘Wait for me until the face of anger passes and I will give you rest.’ (b. Berakhot 7.1) (16a) ...‫וְּ ז ֹאת ׀ ֲעשּו ָל ֵֶ֗הם וְּׁ חָ יּו וְּ ל ֹא יָ ֵֻ֔מתּו‬ ‘but deal thus with them and they will live/so that they may live and not die…’ (Num. 4.19) (16b) ‫בני בקש עליו רחמים ויחיה‬ ‘my son, request mercy form him and he will live/so that he may live’ (b. Berakhot 34.2) 2.0. The Tiberian Reading Tradition of Classical Biblical Hebrew Texts We now turn to the Tiberian reading tradition of CBH material, where a limited degree of the weqaṭal to we-yiqṭol shift has been detected (Joosten 2017, 30–33). At issue here are a relatively small number of I-y qal verbal forms where weqaṭal morphology has arguably been secondarily updated with we-yiqṭol vocalisation. All cases involve we-yiqṭol forms of the verb ‫‘ יָ ֵרא‬fear’, most instances the repeated phraseology ‫‘ יִ ְּש ְּמעּו וְּ יִ ָראּו‬they will hear and 454 The Historical Depth of the Tiberian Reading Tradition fear’, where it is argued that the original weqaṭal reading was along the lines of ‫יִ ְּש ְּמעּו וְּ יָ ְּראּו‬. (17) ...‫ל־ה ָ ֶ֖עם יִ ְּש ְּמעּו וְּׁ יִ ָרָ֑אּו‬ ָ ‫ ;וְּ ָכ‬cf. 2Q11 f1.2 ‫ ;ויראו‬SP ‫ וייראו‬wyīrāʾu ‘And all the people will hear and will fear…’ (Deut. 17.13) (18) ...‫ ;וְּ ַהנִ ְּש ָא ִ ֶ֖רים יִ ְּש ְּמעּו וְּׁ יִ ָרָ֑אּו‬cf. SP ‫ וייראו‬wyīrāʾu ‘And the rest will hear and will fear…’ (Deut. 19.20) (19) ‫וְּ ָכל־יִ ְּש ָר ֵ ֶ֖אל יִ ְּש ְּמ ֵ֥עּו וְּׁ יִ ָ ִֽראּו׃‬...; cf. SP ‫ וייראו‬wyīrāʾu ‘And all Israel will hear and will fear…’ (Deut. 21.21) There is at least a modicum of subjectivity in this assessment. Could the meaning here not be something like ‘they will hear so that they fear’, rather than ‘they will hear and fear’? True, weyiqṭol with final semantics is especially common after volitional forms—short/clause initial jussive yiqṭol, imperative, cohortative—and the X-yiqṭol order in the cases cited make it unlikely that the ‫ יִ ְּש ְּמעּו‬forms that precede ‫ וְּ יִ ָראּו‬are jussive. Even so, final we-yiqṭol sometimes follows non-volitional forms/clauses, e.g., Interrogative with agent-oriented yiqṭol (20) ...‫אמ ַ֤רּו ֵא ָל ֙יו ַמה־נַּ ּ֣עֲשֶ ה ֵ֔ ָלְך וְּׁ יִ ְּׁש ּ֥תק ַהיָ ֶ֖ם ֵ ַֽמ ָע ֵלָׂ֑ינּו‬ ְּ ֹ ‫וַ י‬ ‘And they said to him: “What shall we do to you, that the sea may quiet down for us?”…’ (Jon. 1.11) Conditional future yiqṭol (21) ...‫יה ָב ְּה ַש ָמ ֙ה ֵמ ֵֶ֔הם‬ ָ ‫ת־ש ְּבת ֵֶֹ֗ת‬ ַ ‫וְּ ָה ָא ֶרץ֩ תֵ עָ ֵזֶ֨ב ֵמ ֶֶ֜הם וְּׁ ִת ֶּ֣רץ ֶא‬ ‘But the land shall be abandoned by them and enjoy its Sabbaths while it lies desolate without them…’ (Lev. 26.43) 18. I-y We-yiqṭol for Weqaṭal 455 Simple past qaṭal (22) ‫א ֶכל֙ ֵ֔ ָלמֹו‬ ֹ ָֹ֨ ‫בי ֵה ָמה ִר ֵ֔מּונִ י כ ֲֹה ַנֵ֥י ּוזְּ ֵק ַנֶ֖י ָב ִעיר גָ ָוָׂ֑עּו ִ ַֽכי־ ִב ְּׁקשּ֥ ּו‬ ֙ ַ ‫אתי ַ ַֽל ְּמ ַא ֲה‬ ִ ‫ָק ָ ַ֤ר‬ ‫ָשיבּו ֶאת־נַ ְּפ ָ ַֽשם׃ ס‬ ֹׁ֖ ִ ‫וְּׁ י‬ ‘“I called to my lovers, but they deceived me; my priests and elders perished in the city, for they sought food to revive their strength.’ (Lam. 1.19) Past habitual yiqṭol (23) ‫א־שב יָ ָר ְּב ָ ֶ֖עם ִמ ַד ְּרכֹו ָה ָר ָ ָׂ֑עה ַוּ֠יָ ָשב וַ ֶַ֜י ַעש ִמ ְּק ַ֤צֹות ָה ָע ֙ם כ ֲֹה ֵני‬ ֵ֥ ָ ֹ ‫ַא ַח ֙ר ַה ָד ָבר ַה ֵֶ֔זה ַֽל‬ ‫יהי כ ֲֹה ֵנֵ֥י ָב ַֽמֹות׃‬ ֹׁ֖ ִ ִ‫ָב ֵ֔מֹות ֶ ַֽה ָח ֵפ ֙ץ יְּׁ מַּ ֵל ּ֣א ֶאת־יָ ֵ֔דֹו ו‬ ‘After this thing Jeroboam did not turn from his evil way, but made priests for the high places again from among all the people. Any who wished, he would ordain that they be priests of the high places.’ (1 Kgs 13.33) Nominal clause (24) ...‫ן־א ָ ֶ֖דם וְּ יִ ְּתנֶ ָ ָׂ֑חם‬ ָ ‫ּוב‬ ֶ ‫יכ ֵֵ֔זב‬ ַ ‫ל ֹא ִ ֵ֥איש ֵאל֙ ִ ַֽו‬ ‘God is not man, that he should lie, or a son of man, that he should change his mind.…’ (Num. 23.19) One might also compare to Deut. 31.12–13, where the yiqṭol-weqaṭal form of v. 12 (22) is paralleled in v. 13 (23) by a weqaṭal-infinitive construct sequence. (25) ...‫יכם‬ ֵֶ֔ ‫הוה ֱא ַֹֽל ֵה‬ ָ ְּ‫ּול ַּמּ֣עַּ ן יִ ְּׁל ְּׁמ ֶ֗דּו וְּׁ יָ ִֽ ְָּֽׁראּו ֶאת־י‬ ְּׁ ‫ל ַָֹ֨מ ַען יִ ְּש ְּמ ֶ֜עּו‬... ְּ ; SP ‫וייראו‬ wyīrāʾu ‘…that they may hear and that they may learn to fear the LORD your God…’ (Deut. 31.12) 456 The Historical Depth of the Tiberian Reading Tradition (26) ...‫ֹלה ֶיכָׂ֑ם‬ ֵ ‫הוה ֱא‬ ָ ְּ‫עּו וְּׁ ָל ְּּׁ֣מ ִ֔דּו ְּׁליִ ְּׁר ָאֹׁ֖ה ֶאת־י‬ ֙ ‫יִ ְּש ְּמ‬... ‘(And their children who have not heard) will hear and will learn to fear the LORD your God…’ (Deut. 31.13) In this pair of verses, explicit final forms—‫ּול ַמ ַען יִ ְּל ְּמ ֵ֗דּו‬ ְּ in v. 12 and ‫ ְּליִ ְּר ָ ֶ֖אה‬in v. 13—are paralleled by weqaṭal forms—‫ וְּ ָל ְּמ ֵ֔דּו‬in v. 13 and ‫אּו‬ ֙ ‫ וְּ ָ ַֽי ְֵּּ֣ר‬in v. 12 (while ‫אּו‬ ֙ ‫ וְּ ָ ַֽי ְֵּּ֣ר‬is orthographically ambiguous, ‫וְּ ָל ְּמ ֵ֔דּו‬ is an unequivocal weqaṭal). The point is that even in cases where a finite form can be interpreted as having final semantics, MT Deuteronomy is content with a weqaṭal (though, as we shall see, ̊̄ here). the Samaritan tradition has ‫ וייראו‬wyīraʾu Something in the way of circumstantial evidence may be gleaned from the ancient Hebrew and foreign language textual witnesses—though, given the semantic range of weqaṭal and allowing for orthographic ambiguity, most of their renderings cannot be considered probative regarding the identity of the form translated. The Aramaic and Syriac yiqṭol forms are opaque. The Vulgate reads one future and two subjunctives. The relevant BDSS form in 2Q11 f1.2 (=MT Deut. 17.13), written ‫ויראו‬, is equivocal. By contrast, the Samaritan forms, which are spelled with mater yod, are consistently and transparently yiqṭol according to both the written and reading components of the tradition, i.e., ‫ וייראו‬wyīrāʾu—in line with the Tiberian reading tradition. Yet this is also the case at Deut. 31.12, example (22), against the Tiberian tradition. The foregoing facts are subject to various interpretations. Arguably, one of the more compelling is that a form intended to be read as weqaṭal ‫ וְּ יָ ְּראּו‬was secondarily reinterpreted in the Tiberian vocalisation tradition as we-yiqṭol ‫ וְּ יִ ָראּו‬in line with trends 18. I-y We-yiqṭol for Weqaṭal 457 seen to varying degrees in Second Temple Hebrew sources. This is not surprising, as various scholars have highlighted features within the Tiberian pronunciation tradition that indicate that, while preserving Iron Age features and not immune to Byzantine and medieval developments, it substantially crystallised in the Second Temple Period. 3.0. The Tiberian Classical Biblical Hebrew Written Tradition The obvious implication of all this is that, when it comes to internal Tiberian written-reading deviations such as these, the Tiberian reading tradition should be regarded as temporally removed from the pronunciation tradition implied by the consonantal text. This is borne out in numerous pieces of evidence, as seen throughout this monograph. Yet, as has also often been emphasised, it is not the whole story. Frequently, the Tiberian consonantal tradition itself bears witness to the very secondary features adopted that have become characteristic of the reading tradition. Consider an example relevant to the issue under examination here: 458 The Historical Depth of the Tiberian Reading Tradition (27) ‫י־ע ֵָ֔שו ַהי ְֹּש ִ ֶ֖בים ְּב ֵש ִ ָׂ֑עיר‬ ֵ ֵ‫מר֒ ַא ֶתם ַֽעֹ ְּב ִ ֵ֗רים ִבגְּ בּול֙ ֲא ֵח ֶיכם ְּבנ‬ ֹ ‫ת־ה ָע ֮ם ַצו ֵלא‬ ָ ‫וְּ ֶא‬ ‫אד׃‬ ֹ ַֽ ‫יראּ֣ ּו ִמ ֵֶ֔כם וְּ נִ ְּש ַמ ְּר ֶ ֶ֖תם ְּמ‬ ְּׁ ‫וְּׁ ִ ִֽי‬ MT DSS ‫ואת[ העם צו לאמר אתם עברים בגבול אחיכם בני עשו הישב]ים בשעיר‬ ‫]ם ונשמרתם מאד‬ ֯ ‫ויראו ֯מ[כ‬ SP ‫ואת העם צוי לאמר אתם עברים בגבול אחיכם בני עשו היושבים בשעיר‬ ‫) מכם ונשמרתם מאד׃‬wyīrāʾu( ‫וייראו‬ ‘And command the people, “You are about to pass through the territory of your brothers, the people of Esau, who live in Seir; and they will be afraid of you. So be very careful.’ (Deut. 2.4 || 4Q35 f56.9 || SP) Here the orthographically unambiguous Tiberian we-yiqṭol form ‫ וְ יִ ְיראּו‬is arguably less felicitous than weqaṭal ‫וְּ יָ ְּראּו‬, since the meaning is not purposive ‘you are crossing into their territory… so that they fear you’, but one of mere succession, one event leading to the next. Crucially, though, given the mater yod, the written and reading components of the Tiberian tradition are in harmony here; similar harmony characterises the written and reading components of the SP at this point. For purposes of contrast, one may compare the BDSS text 4Q35 f56.9, which has the more ambiguous spelling ‫ויראו‬, perhaps (but not certainly) reflecting a weqaṭal form. If the MT form here is secondary, it shows that the yiqṭol morphology has penetrated into not only that layer of the reading tradition reflected in the medieval vocalisation signs, but also into that reflected by the matres lectionis, which were presumably added earlier on, probably in the Second Temple Period. Similarly, and of more immediate relevance, in a fourth occurrence of the ‫ יִ ְּש ְּמעּו וְּ יִ ָראּו‬formula, in Deut. 13.12, the text reads: 18. I-y We-yiqṭol for Weqaṭal 459 (28) ...‫וְּ ָכל־יִ ְּש ָר ֵֵ֔אל יִ ְּש ְּמ ֶ֖עּו וְּׁ ִ ִֽי ָראָ֑ ּון‬ ‘And all Israel will hear and will fear…’ (Deut. 13.12) This case differs from the rest in that the we-yiqṭol form ends with paragogic nun. While qaṭal forms with paragogic nun are not unknown in the MT (there are three of them: Deut. 8.3, 16; Isa. 26.16), they are more than one-hundred times less frequent than yiqṭol forms with the same suffix. In this case, again, there is harmony between the Tiberian written and reading traditions. Either the we-yiqṭol form here with paragogic nun is original or the historical depth of the secondary we-yiqṭol analysis in the Tiberian tradition extends beyond the levels of vocalism reflected in niqqud and matres to consonantal realisation. 4.0. Conclusion This leads us back to the three other cases of ‫יִ ְּש ְּמעּו וְּ יִ ָראּו‬. If the apparently problematic we-yiqṭol readings of ‫ וייראו‬and ‫ ויראון‬are rooted in the written tradition, then perhaps cases in which weyiqṭol ‫ ויראו‬has been seen as a secondary vocalisation are not deviations from the ostensible pronunciation underlying the written tradition, but reliably conserve it. There are at least three ways to interpret the evidence: 1. We-yiqṭol in place of weqaṭal is strictly late, in which case all supposed forms—whether in the written or reading tradition—must be explained as late. This could mean anything from the late composition of the entire surrounding text, through the insertion of a late gloss, to a corruption, to the secondary 460 The Historical Depth of the Tiberian Reading Tradition updating of the consonantal text by means of addition of a mater or paragogic nun. This option seems extreme. 2. On the other extreme, on the basis of the consonantal evidence of we-yiqṭol for weqaṭal, one might adopt the view that all cases of suspected interchange are acceptable CBH, so that no secondary process in line with late Hebrew trends need be entertained, except for the notion that such early instances are authentic forerunners in the vein of what would later become more established convention. 3. There is also a preferable middle path between these extremes. This involves allowing for both the early agreement of the Tiberian written and reading traditions on characteristically late features and the deviation of the reading component from the typologically earlier profile of its written counterpart in line with Second Temple developments. Whether this is analysed as the early original use of a characteristically late feature secondarily extended within the reading tradition or as a process of secondary development within the written tradition, the implication is the same: less remoteness between the written and reading components, which, even in the case of apparent secondary developments, should be seen as largely overlapping on the historical continuum. Similarly, in the case of we-yiqṭol for weqaṭal, it is possible that a certain number of I-y qal forms vocalised as we-yiqṭol began as weqaṭal forms, so that there is a degree of dissonance on this point between the written and reading components of the Tiberian tradition. But given the consonantal testimony regarding the feature, this dissonance should not be interpreted as a chasm be- 18. I-y We-yiqṭol for Weqaṭal 461 tween the two. Obviously, linguistic continuity typifies the relationship between the written and reading traditions when it comes to the vast majority of linguistic features. But even in dissonance there is continuity. The distinction between the written and reading components is one of degree, not essence, characterised by drift along a continuum within a continually recited tradition, rather than a clean break and restart within the tradition. CONCLUSION This collection of research has presented twenty-five cases of dissonance between the written and reading components of the Tiberian reading tradition—seven in the Introduction and eighteen in the subsequent chapters. The argument has been twofold. 1.0. The Secondary and Late Character of Tiberian Written-reading Dissonance First, it has been argued that the relevant cases of dissonance reflect relatively late, secondary developments of the Tiberian pronunciation tradition in line with Second Temple linguistic trends vis-à-vis its orthographic counterpart. This carries with it the implication that the pronunciation tradition, despite marked conservatism regularly safeguarding genuine Iron Age features, in large part crystallised in the Second Temple Period. It therefore occasionally manifests contemporary phenomena anachronistic for First Temple texts. 2.0. The Antiquity of Secondary Features in the Reading Tradition Second, despite the late character of the pronunciation features involved in these cases of dissonance, it has been maintained that they do not derive from medieval or Byzantine Period developments, but are rooted in Second Temple linguistic conventions. To be sure, they often appear to continue evolutionary processes already documented in pre-exilic material, whether biblical or © 2023 Aaron D. Hornkohl, CC BY-NC 4.0 https://doi.org/10.11647/OBP.0310.19 464 The Historical Depth of the Tiberian Reading Tradition epigraphic. Notwithstanding the medieval origin of the Tiberian vowel signs, the fact that the secondary features of the Tiberian pronunciation tradition reflect Second Temple linguistic developments strongly suggests that the tradition’s primary features— i.e., the ones on which there is consensus between the written and reading components of the tradition—are even older. This all points to a reading tradition which, in the main, is a remarkably ancient and conservative linguistic artefact. It is readily admitted here that the individual arguments made in the case of the features discussed in this volume are unlikely to have equal cogency. It is, however, hoped that even if certain explanations have been rejected, the combined evidence and argumentation will have been sufficient to convince even the sceptic of the major prongs of the argument. If one accepts the reality of written-reading dissonance, the secondary nature of vocalic developments in line with Second Temple conventions, and a degree of continuity between such developments and minority Iron Age features, the resulting acknowledgement of the historical antiquity of the Tiberian reading tradition should affect its perceived value for exegetical, textual, literary, and linguistic research. Allowing for the historically composite nature of the Tiberian vocalisation tradition, there is no reason to disfavour its testimony in contrast to traditions characterised by earlier written attestation. The combined evidence points to an ancient interpretive tradition that largely coalesced in the post-exilic period. The vast majority of the tradition seems reliably to preserve Iron Age features, whereas the small minority that must be considered anachronistic reflects linguistic and interpretive Conclusion 465 trends that need be dated no later than the Second Temple Period. In the rest of this concluding section, an attempt is made to summarise findings with regard to the principal corpora cited as representative of First and Second Temple Hebrew and to highlight certain ancillary ramifications of the research. 3.0. Linguistic Affinity between Second Temple Chronolects and the Tiberian Reading Tradition 3.1. Tiberian Late Biblical Hebrew Though some scholars reject the diachronic import of the CBH/LBH distinction, there is no doubt that the core LBH books exhibit linguistic profiles especially marked by features characteristic of other Second Temple sources in concentrations not found in acknowledged CBH material. The significance of LBH in the present connection centres on features common to both LBH and the Tiberian reading tradition in which both differ from the Tiberian written tradition. Such features discussed in this volume include spelling of the toponym ‫ ירושלים‬reflecting diphthongisation (Introduction, §3.1); univerbalisation of the proposition -‫ ל‬and the infinitive construct (Introduction, §3.2); constructions of the type ‫ היום השישי‬instead of ‫( יום השישי‬Introduction, §3.3); the nifalisation of originally qal ‫כ ַשל‬-‫ל‬ ָ ‫( *יִ ְּכ ַש‬ch. 10, §§1.1.1; 2.1.1); the shift from qal internal pas- sive to nifʿal (ch. 10, §§1.1.2; 2.2); hifilisation of the originally qal form ‫( נָ ָחה‬see ch. 11, §1.1.3; 2.1); hitpaelisation of forms with 466 The Historical Depth of the Tiberian Reading Tradition assimilated t (see ch. 13, §§1.1.2; 2.1); relativising ha-+qaṭal (ch. 15, §§1.1; 2.0); long yiqṭol (yaqtulu) morphology in 1st-person wayyiqṭol forms, especially II-w/y qal and hifʿil forms (ch. 17, §2.1). 3.2. The Dead Sea Scrolls While the designations QH or DSSH might be understood to indicate a sort of monolithic Hebrew in use in the Judaean Desert at the turn of the epoch, the diversity of Hebrew types there has long been acknowledged (Morag 1988). At the very least, it is necessary to distinguish between BDSS Hebrew and NBDSS Hebrew (see above, ch. 6, §9.0; ch. 17, §1.1), though even this dichotomy is problematic (Hornkohl 2021b, 134, fn. 19). 3.2.1. The Biblical Dead Sea Scrolls Among the BDSS, it is well known that the Hebrew of 1QIsaa stands out against the Hebrew of the rest of the manuscripts that reflect material eventually canonised as Jewish Scripture (Tov 2012, 100–10; Young 2013; Reymond 2014, 11; Rezetko and Young 2014, 138–39; Hornkohl 2016a, 1020). Despite 1QIsaa’s biblical content and style, its linguistic character—which has been described as ‘contemporised’ and ‘popular’—includes many features that stray from the classical norms reflected in MT Isaiah and 1QIsab in favour of acknowledged Second Temple alternatives. For this reason, it might be expected that 1QIsaa would share many features with the reading component of the Tiberian biblical tradition. And, indeed, just such a scenario obtains. Consider the following list of affinities: the spelling ‫ אדני‬for ‫יהוה‬, like Conclusion 467 the Tiberian qere perpetuum ʾăḏōnaẙ̄ (ch. 1, §1.0); agreement with the Tiberian qere perpetuum ‫ שכ"ב‬for ‫( שג"ל‬ch. 3, §1.3); the spelling ‫ || לקרת‬MT ‫( ִל ְּק ָראת‬ch. 5, §4.1); 2MS ‫כה‬- || MT ‫ָך‬- (1QIsaa 28–54 only; ch. 6, §5.1.1); 2MS ‫תה‬- || MT ‫ ָת‬- (ch. 6, §5.2.1); pielisation (ch. 12, §1.2.1); hitpaelisation (ch. 13, §1.2.1); I-y qal weyiqṭol for wayyiqṭol (ch. 18, §1.2.1). More generally, the BDSS often show affinity with the Tiberian reading tradition in terms of agreement with qere over ketiv (Introduction, §1.0 and fn. 5); realisation of ‫( יששכר‬ch. 4, §2.0 [?]); 2MS ‫כה‬- || MT ‫ָך‬- (1QIsaa 28–54 only; ch. 6, §§5.1.1; 9.0); 2MS ‫תה‬- || MT ‫ ָת‬- (ch. 6, §5.2.1); 2/3FPL endings written ‫נה‬- || MT ָ ‫ן‬- (ch. 9, §2.1); hifilisation, specifically of ‫( יס"ף‬ch. 11, §1.2.1); pielisation (ch. 12, §1.2.1); hitpaelisation (ch. 13, §1.2.1); long II-w/y qal and hifʿil 1st-person wayyiqṭol forms (ch. 17, §1.2.2); I-y we-yiqṭol for wayyiqṭol (ch. 18, §1.2.1). Notwithstanding the foregoing lists of features in which BDSS material appears to side with the Tiberian reading tradition against the Tiberian written tradition, it should be emphasised that—with the notable exception of 1QIsaa—the linguistic profile of the BDSS is largely consistent with standard BH as reflected in the combined Tiberian written-reading tradition. From this perspective, there is a marked difference between the linguistic profile of the BDSS and that of the NBDSS, which are evidently more representative—than even 1QIsaa—of contemporary Second Temple language usage. 468 The Historical Depth of the Tiberian Reading Tradition 3.2.2. The Non-biblical Dead Sea Scrolls Despite a pronounced degree of resemblance between DSSH and Tiberian BH against RH, the NBDSS exhibit far greater departure from BH than do the BDSS. This should not be surprising, since the BDSS represent copies of already traditional First Temple texts, while the NBDSS appear to be Second Temple compositions. It should come as no surprise, then, that the NBDSS share many features with the Tiberian reading tradition, including constructions of the type ‫ היום השישי‬instead of ‫( יום השישי‬Introduction, §3.3); realisation of ‫( יששכר‬ch. 4, §2.0 [?]); 2MS ‫כה‬- || MT ‫ָך‬(1QIsaa 28–54 only; ch. 6, §5.1.2); 2MS ‫תה‬- || MT ‫ ָת‬- (ch. 6, §5.2.2); nifalisation, especially replacement of qal internal passive with nifʿal (ch. 10, §1.2.2); hifilisation (ch. 11, §§1.1.3; 1.2.2); pielisation (ch. 12, §§1.0; 1.2.2); hitpaelisation (ch. 13, §1.2.2); past tense ṭɛrɛm qaṭal for ṭɛrɛm yiqṭol (ch. 14, §2.1.3); long II-w/y qal and hifʿil 1st-person wayyiqṭol forms (ch. 17, §1.2.2); I-y we-yiqṭol for wayyiqṭol (ch. 18, §1.2.2). 3.3. Samaritan Hebrew Like the combined Tiberian biblical written-reading tradition, the Samaritan tradition is composite, comprising a written component that, in view of its orthography, appears to reflect a somewhat later crystallisation than that of the Tiberian Torah, together with a significantly later pronunciation component. The pronunciation tradition, though not lacking in classical features, is strikingly replete with late linguistic features, especially typical of Second Temple Hebrew and Aramaic, but also including even later elements. Characteristic Second Temple linguistic features Conclusion 469 common to both SH and the Tiberian reading tradition include univerbalisation of the proposition -‫ ל‬and the infinitive construct (Introduction, §3.2); syncopation of the 3MPL gentilic ending -īm < -iyyim (Introduction, §3.4); consistent replacement of the tetragrammaton with an alternative form (ch. 1, §§1.0; 2.0); nifʿal analysis of ‫ת־פנֵ י יְּ הוָ ה‬ ְּ ‫ ֵל ָראֹות ֶא‬and similar (ch. 2, §§1.0; 2.0); euphemistic ‫ שכ"ב‬for ‫( שג"ל‬ch. 3, §§1.3; 2.0); 2MS ‫תה‬- || MT ‫ ָת‬- (ch. 6, §§4.0; 5.2.1); ‫ היא‬ī || Tiberian qere perpetuum ‫ ִהוא‬in the Torah (ch. 8, §2.0); 2/3FPL endings written ‫נה‬- || MT ָ ‫ן‬- (ch. 9, §2.1); nifalisation (ch. 10, §1.3); hifilisation (ch. 11, §1.3); pielisation (ch. 12, §1.3); hitpaelisation (ch. 13, §1.3); ha-+qaṭal (ch. 15, §§3.2); long II-w/y qal and hifʿil 1st-person wayyiqṭol forms (ch. 17, §§1.2.2; 1.3); I-y we-yiqṭol for wayyiqṭol (ch. 18, 1.3). 3.4. Ben Sira Due partially to its wisdom genre, partially to its poetic style, and partially to the archaising predilections of its author, the linguistic profile of BS is a mixture of classical, even archaic, features, especially in terms of vocabulary. Even so, there is no mistaking the book’s inclusion of diagnostically late features, lexical as well as grammatical, in both its Second Temple and medieval manuscript evidence. Diachronically significant late features common to BS and the Tiberian reading tradition include the following: univerbalisation of the proposition -‫ ל‬and the infinitive construct (Introduction, §3.2); syncopation of the 3MPL gentilic ending -īm < -iyyim (Introduction, §3.4); ‫ || היא‬Tiberian qere perpetuum ‫ִהוא‬ in the Torah (ch. 8, §2.0); nifalisation (ch. 10, §1.4); hifilisation (ch. 11, §1.4); pielisation (ch. 12, §1.4); hitpaelisation (ch. 13, 470 The Historical Depth of the Tiberian Reading Tradition §1.4); past tense ṭɛrɛm qaṭal for ṭɛrɛm yiqṭol (ch. 14, §2.1.4); long II-w/y qal and hifʿil 1st-person wayyiqṭol forms (ch. 17, §1.3.1); I- y we-yiqṭol for wayyiqṭol (ch. 18, §1.4). 3.5. Rabbinic Hebrew It has been argued that in the cases of written-reading dissonance in the combined Tiberian biblical tradition, the Masoretes were influenced in secondary pronunciations by RH (see, e.g., Blau 2018, 115, §3.5.6.3.7n, 213–14, §§4.3.4.2.2–4.3.4.2.2n). While it is difficult definitively to disprove such a notion, several considerations combine to show that such an extreme view is unwarranted. First, if RH influenced the Masoretes, it did so very sparingly, since in most distinguishing features, BH and RH remain distinct. Second, as has already been indicated, since in its departures from the Tiberian written tradition, the Tiberian reading tradition resembles not just RH, but several late traditions and corpora, including the combined Tiberian LBH written-reading tradition, there is no reason to insist specifically on RH influence on the Tiberian reading component. Finally, as emphasised below, secondary features standardised in the Tiberian pronunciation tradition often find precedent in minority features in the Tiberian CBH written tradition and/or in Iron Age epigraphy. This implies that many characteristic Second Temple Hebrew features constitute standardisations of earlier features no matter the Second Temple tradition or corpus in which their extension took place, including the Tiberian reading tradition and RH. Even so, it would be misleading to deny the reality of significant diachronic affinity between RH and the Tiberian pronun- Conclusion 471 ciation tradition, though this should not necessarily be considered a result artificial RH influence on the Masoretes. Salient features discussed in this volume include univerbalisation of the proposition -‫ ל‬and the infinitive construct (Introduction, §3.2); syncopation of the 3MPL gentilic ending -īm < -iyyim (Introduction, §3.4); qere euphemisms (§§1.1; 1.3); the vocalisation ‫( ִל ְּק ָראת‬ch. 5, §§1.0; 2.0); 2MS ‫תה‬- || MT ‫ ָת‬- (ch. 6, §4.0); pielisation (ch. 12, §1.5); ‫ || ִהיא‬Tiberian qere perpetuum ‫ ִהוא‬in the Torah (ch. 8, §§1.0; 2.0); 2/3FPL endings written ‫נה‬- || MT ָ ‫ן‬- (ch. 9, §2.2); nifalisation (ch. 10, §1.5); hifilisation (ch. 11, §1.5); pielisation (ch. 12, §1.5); hitpaelisation (ch. 13, §1.5); I-y qal we-yiqṭol for wayyiqṭol (ch. 18, §1.5). 4.0. Iron Age Epigraphy and the Classical Biblical Hebrew Written Tradition 4.1. Iron Age Epigraphy It has been argued that all of the linguistic features discussed in this volume are secondary pronunciation features vis-à-vis the relevant written tradition alternative. Occasionally, however, there is evidence of the pronunciation feature as a minority Iron Age epigraphic alternative. This occurs in the case of syncopation of the 3MPL gentilic ending -īm < -iyyim (Introduction, §3.4); 3MS ̊̄ for polythongal ‫יו‬- (Introducpossessive suffix on plurals ‫ו‬- -aw tion, §3.6); the spelling ‫ לקרת‬liqrat [?] || MT ‫( ִל ְּק ָראת‬ch. 5, §4.2); 2MS ‫כה‬- || MT ‫ָך‬- ch. 6, §7.0); 2MS ‫תה‬- || MT ‫ ָת‬- (ch. 6, §7.0); nifalisation (ch. 10, §3.1); hitpaelisation (ch. 13, §3.1). 472 The Historical Depth of the Tiberian Reading Tradition 4.2. The Tiberian Classical Biblical Hebrew Written Tradition The late, secondary features which the Tiberian reading tradition standardised as divergences from the corresponding written tradition also sometimes appear as minority features in the Tiberian CBH written tradition. Consider the following cases discussed in this volume: univerbalisation of the proposition -‫ ל‬and the infinitive construct (Introduction, §3.2); ‫ אדני‬for ‫( יהוה‬ch. 1, §2.0; 2MS ‫כה‬- || MT ‫ָך‬- (ch. 6, §2.0); 2MS ‫תה‬- || MT ‫ ָת‬- (ch. 6, §2.0); nifalisation (ch. 10, §3.0); hifilisation (ch. 11, §3.0); pielisation (ch. 12, §3.0); hitpaelisation (ch. 13, §3.0); past tense ṭɛrɛm qaṭal for ṭɛrɛm yiqṭol (ch. 14, §§2.3; 4.0); ha-+qaṭal (ch. 15, §§1.2; 3.2); long IIw/y qal and hifʿil 1st-person wayyiqṭol forms (ch. 17, §2.2.1); I-y qal we-yiqṭol for wayyiqṭol (ch. 18, §3.0). 5.0. Further Ramifications of the Study Various combinations of data gathered in the foregoing studies support a number of hypotheses, each of which merits further investigation. 5.1. Diachronic Diversity within Classical Biblical Hebrew: The Torah versus the Rest The data pertinent to several features discussed in this volume are interpretable as evidence of diachronic development within Tiberian CBH, especially, between the Torah and the rest of the CBH corpus. However such a linguistic disparity is most convincingly explained—whether as evidence of the actual linguistic antiquity of the Tiberian Pentateuchal traditions vis-à-vis the Conclusion 473 traditions in other CBH material or as a result of early consolidation and careful preservation of the Torah’s linguistic profile relative to other CBH texts1—it is clear that in terms of select features, the Pentateuch is characterised by striking linguistic conservatism. Such features include 3FS ‫הוא‬, which, it has been argued, may well reflect an early phonetic reality standardised as ‫ ִהיא‬in the rest of the Hebrew Bible (ch. 8, §3.0), but as ‫ הוא‬in the Torah (ch. 8, §2.0); hifilisation of certain qal II-y verbs, most notably ‫‘ יס"ף‬add, continue’ (ch. 11, §§1.1.3; 2.4), the preservation of archaic hifʿil-like qal forms (ch. 11, §2.4), and hifilisation in general (ch. 11, §3.0); short rather than long or pseudo-cohortative 1st-person wayyiqṭol forms (ch. 17, §1.4.3). Scholars who accept a diachronic distinction between CBH and LBH do not generally attempt finer gradations. Though Hornkohl (2013a; 2016) has argued for the heuristic value of TBH, CBH is generally considered a single broad chronolect that includes regional, social, and genre diversity. More rarely, it is suggested that CBH can usefully be divided into chronological phases, i.e., CBH1 and CBH2 (Elitzur 2015; 2018a; 2018b; 2019; 2022). A previous study lending support to such an approach is Hornkohl’s (2013a, 83–91) analysis of proper names ending in the theophoric element )‫יָ ה(ּו‬-. There it is observed, inter alia, that “The books of the Torah and Joshua present no examples of names with either ending, apparently reflecting a time before the use of such names was prevalent” and “To be sure, the Pentateuch has only two names containing any form of the tetragram1 See above, ch. 17, §§1.4.2–3, on the need for a nuanced approach to complex data. 474 The Historical Depth of the Tiberian Reading Tradition maton, in both cases a prefix: ‫יְּהֹוש ַע‬ ֻ ‘Joshua’ and ‫יֹוכ ֶבד‬ ֶ ‘Jochabed’” (Hornkohl 2013a, 86 and fn. 35). It would seem that the onomastic tradition preserved in the Pentateuch is consistent with premonarchical times. The linguistic conservatism that distinguishes the language of the Torah from that of the rest of CBH may similarly be construed as evidence of the preservation of genuine linguistic antiquity within the tradition. Alternatively, it may be that the classical linguistic profile of the Torah was kept especially pristine, whereas the formerly more classical profile of other CBH material was allowed to drift in the direction of LBH, though it never reached the level of concentration of late features characteristic of the acknowledged LBH books. Whatever the explanation, there is a palpable difference between the CBH of the Torah and that of the Prophets and Writings. 5.2. Suppletion and Orthographic Constraints on Linguistic Development within the Tiberian Reading Tradition In the above treatments on movement between verbal stems (chs 10–13), suppletive paradigms are highlighted as a common result of linguistic evolution and the resultant written-reading dissonance. Again and again, some or even most of a given verb’s orthographic forms amenable to secondary interpretation shifted binyanim, whereas other instances were excluded from the shift because their written forms were unsuitable to the new stem. One of the clearest examples is the well-known case of nifʿal-qal ‫נִ גַ ש‬- ‫‘ יִ גַ ש‬approach’, whose principal Tiberian biblical forms are given below in Table 1 (see also above, ch. 10, §2.1.2). Conclusion 475 Table 1: Tiberian biblical forms of the suppletive nifʿal-qal verb ‫נִ גַ ש‬-‫יִ גַ ש‬ ‘approach’ nifʿal ‫נִ גַ ש‬ qal — participle ‫נִ גָ ש‬ — imperative — ‫גְּ שּו‬/‫גֹשּו‬/‫ג ִֹשי‬/‫גְּ ָשה־‬/‫גֶ ש־‬/‫גַ ש‬ prefix conjugation — ‫יִ גַ ש‬ infinitive construct — -‫גִ ְּשת‬/‫(ל)גֶ ֶשת‬ ָ suffix conjugation It is assumed that the verb was originally consistently G-stem (as it remains in SH; see above, ch. 10, §1.3.6) and was refashioned as nifʿal where possible in line with its intransitive semantics, for which nifʿal morphology was considered a better fit. The consistently suppletive biblical paradigm invites scrutiny. One question involves the extent to which the unambiguous qal spellings effectively prevented more extensive qal > nifʿal evolution. In other words, does the Tiberian biblical suppletion reflect genuine language use? Or is it an artificial arrangement relevant specifically to the Hebrew Bible’s written-reading dissonance? There is no definitive answer, but it is striking that the NBDSS attest the nifʿal infinitive construct ‫‘ בהנגשו‬when he ap- proaches’ (4Q512 f40–41.2; see above, ch. 10, §1.2.1). This may indicate that nifalisation of the verb in question was more extensive than indicated by Tiberian BH, i.e., where not anchored by unambiguous qal orthography, Second Temple Hebrew exhibited greater or even full nifalisation of this verb. Even so, as Hornkohl (2021a, 14–15) observes, “ancient Hebrew sources never present the prefix conjugation ‫*ינגש‬, the existence of which would con- firm the verb’s wholesale niphalisation.” In other cases, it seems clearer that suppletion in the combined Tiberian written-reading tradition reflects an artificial sit- 476 The Historical Depth of the Tiberian Reading Tradition uation unrepresentative of any genuine chronolect. Consider the case of the suppletive piʿʿel-qal verb ‫מ ֵאן‬-‫ן‬ ֵ ‫מ ֵא‬. ָ In this instance, the entire paradigm is piʿʿel except for the active participle, which is qal, and the infinitive absolute, which is equally analysable as piʿʿel or qal. Table 2: Tiberian biblical forms of the suppletive piʿʿel-qal verb ‫מ ֵאן‬-‫ן‬ ֵ ‫ָמ ֵא‬ ‘refuse’ suffix conjugation piʿʿel ‫ֵמ ֵאן‬ qal — prefix conjugation ‫יְּ ָמ ֵאן‬ — — ‫ה ֵמ ֲאנִ ים‬/‫ן‬ ַ ‫ָמ ֵא‬ participle infinitive absolute ‫ָמ ֵאן‬ In this case, all biblical spellings are interpretable as qal, while the pronunciation tradition reflects a shift to piʿʿel where permitted by the orthography. It should also be noted that, on the as̊̄ morphology, the extant sumption of originally qal stative qaṭēl vocalisations of the MS participle and the infinitive absolute, both ‫מ ֵאן‬, ָ can be considered faithful preservations of ancient morphology (the vocalisation of the MPL participle ‫ה ֵמ ֲאנִ ים‬, ַ by contrast, is appropriate for neither G- nor D-stem). Clearly, the suffix and prefix conjugation spellings might well also reflect original qal forms. But if the forms of the written component of the Tiberian biblical tradition point to original qal morphology, SH and RH confirm the pielisation seen in the pronunciation component of the Tiberian biblical tradition (ch. 12, §2.1). Again, the question may be asked: does the Tiberian biblical suppletion reflect an authentic linguistic situation or is it an artificial combination of diachronic snapshots? While in any given case of linguistic evolu- Conclusion 477 tion there must be intermediate stages of development characterised by mixed usage, it is not clear that the Tiberian biblical suppletion should be so explained. Since there is no unequivocal orthographic evidence of piʿʿel ‫ ֵמ ֵאן‬until the Mishna, it may well be that D-stem analysis of the verb is entirely foreign to the Tiberian BH written tradition. But this remains unverifiable, since Tiberian LBH lacks participial forms that might unambiguously (dis)confirm the antiquity of the process of pielisation. Even beyond BH, biblical orthography seems partially to have anchored ancient Hebrew and prevented fuller evolution. Even in post-biblical Hebrew, where it might be expected that biblical spelling relics would no longer influence language use, the biblical linguistic tradition still exerts force. Consider the very early pielisation of ‫‘ ִד ֶבר‬speak’, which left only a small residue of qal infinitival and active and passive participial forms (ch. 12, §3.1). While one might expect that beyond BH, such residual qal forms would be completely eclipsed, use of the active participle continues in BS, the NBDSS, Tannaitic RH, and Amoraic RH, despite the extensive pielisation of the verb in all of these traditions. Indeed, the active and passive participles continue to be used in Modern Hebrew. Evidently, the existence of clearcut archaisms in the Tiberian written tradition and the prestige of the mixed Tiberian written-reading tradition resulted in the conservation of linguistic relics that would probably otherwise have been levelled in forms of post-biblical ancient Hebrew. 478 The Historical Depth of the Tiberian Reading Tradition 5.3. Diversity within the Tiberian Reading Tradition Not unrelated to the topic of the preceding section, it might be assumed that the Tiberian reading tradition would exhibit uniformity wherever possible. That is, outside of ancient orthographic forms not amenable to secondary reclothing, it would be reasonable to expect a homogenous and level reading tradition. But such consistency does not obtain. Consider the case of 1stperson wayyiqṭol forms in the Tiberian Torah (ch. 17, §2.2.2). In view of the prevalence of short spellings of 1st-person forms in the Torah, 1CS and 1CPL might be vocalised similarly. But such is not the case. 1CPL forms are vocalised with short morphology in accord with their orthography, whereas in the case of 1CS forms long vocalisation is regularly imposed upon short orthography. Similar diversity with the Tiberian reading tradition is noticeable in the case of 2MS and 2/3FPL endings (chs 6 and 9). Against the backdrop of standard vowel-final morphology, the Tiberian pronunciation tradition also testifies to minority consonant-final realisations. The above diversity indicates that the Tiberian pronunciation was not simply a monolithic tradition mechanically wedded to the corresponding written tradition. Rather, each component of the tradition itself reflected a complex and varied linguistic reality, each component influenced the other, and their merger resulted in a layered and multifarious combination of great variety and depth.2 2 See Khan (2020, I:69–85) for a balanced discussion of heterogeneity within the Tiberian reading tradition, including different perspectives on diachrony. Conclusion 479 5.4. Majority and Minority Features in Classical Biblical Hebrew A major thrust of the present volume involves the claim that many late secondary departures of the Tiberian reading tradition find precedent in minority CBH features. In other words, rare CBH features at some point became dominant in the Tiberian tradition and were standardised at the expense of earlier dominant features. It is worth stating explicitly the corollary of this statement, namely, that by dint of including minority features among majority features, CBH was inclusive of a great deal of diversity. As an example, consider the case of standard CBH past tense ṭɛrɛm yiqṭol versus minority CBH past tense ṭɛrɛm qaṭal (ch. 14). One, perhaps two, of the exceptional past tense ṭɛrɛm qaṭal cases are explicable as secondary revocalisations. But the other two are evidently genuine. And their genuineness calls into question the necessity of explaining away the cases that can be attributed to secondary processes (see above, ch. 14, §3.0). It is admittedly tempting to formulate a theory capable of accounting for all non-standard features, but some allowance must be made for simple synchronic linguistic variety attributable to no factor beyond human inconsistency. REFERENCES Primary Sources Ben Sira Abegg, Martin G. 2009. Ben Sira Electronic Database (BENSIRAC/BENSIRA-M). Accordance Modules. Altamonte Springs, FL: OakTree Software, Inc. Dead Sea Scrolls Qimron, Elisha. 2020. Ha-Ḥiburim ha-ʿIvriyim mi-Qumran: Mahadura Meshulevet. 3 vols. Tel-Aviv: Open Access. Biblical Texts Abegg, Martin G., Jr. 2009a. Dead Sea Scrolls Bible Canonical/Manuscript Order (DSSB-C/DSSB-M). Accordance Modules. Altamonte Springs, FL: OakTree Software, Inc. Non-biblical Texts Abegg, Martin G., Jr. 2001. Non-Biblical Dead Sea Scrolls from Qumran (QUMRAN). Accordance Module. Altamonte Springs, FL: OakTree Software, Inc. Greek Bible (LXX) Taylor, Bernard A., and Dale M. Wheeler (eds.). 2008. Greek Septuagint Database. Accordance Module. Altamonte 482 The Historical Depth of the Tiberian Reading Tradition Springs, FL: OakTree Software, Inc. Based on Ralfs, Alfred (ed.). 2006. Septuaginta. Revised by Robert Hanhart. Stuttgart: Deutsche Bibelgesellschaft. Inscriptions CIJ = Frey, Jean Baptiste. 1936–1952. Corpus Inscriptionum Judaicarum. 2 vols. Rome and Vatican City: Pontifical Institute. Mazar, Benjamin. 1973. Beth Sheʿarim: Report on the Excavations during 1936–1940. Vol. 1, Catacombs 1–4. Jerusalem: Massada Press on behalf of the Israel Exploration Society and the Institute of Archaeology, Hebrew University. Waddington, William Henry. 1870. Inscriptions grecques et latines de la Syrie recueilles et expliquées. Paris: Librarie de Firmin Didot Frères, Fils et Cie, Imprimeurs de l’Institut de France. Hebrew Bible Masoretic Tradition, Edition BHS = Elliger, Karl, and Wilhelm Rudolph. 1990. Biblia Hebraica Stuttgartensia. 4th edition. Stuttgart: Deutsche Bibelgesellschaft, 1990. Masoretic Tradition, Digital Edition The J. Alan Groves Center for Advanced Biblical Research. 1991– 2016. Groves-Wheeler Westminster Hebrew Morphology. Altamonte Springs, FL: OakTree Software, Inc. References 483 Masoretic Tradition, Images Aleppo Codex: Bar Hama, Adon. 2002. Digitalisation of The Aleppo Codex. https://www.mgketer.org/mikra/1/1 Leningrad Codex: Zuckerman, Bruce E. 1990. Leningrad Codex Images. Included in BibleWorks 10. West Semitic Research Group. Kitāb al-Khilaf Lipschütz, Lazar. 1965. Kitāb al-Khilaf: Mishael Ben Uzziel’s Treatise on the Differences between Ben Asher and Ben Naphtali. Publications of the Hebrew University Bible Project 2. Jerusalem: Magnes Press. Latin Jerome’s Commentaries Gryson, Roger, and Paul-Augustin Deproost. 1994–1999. Commentaires de Jerome sur le prophete Isaie. 5 vols. Aus der Geschichte der lateinischen Bibel 23. Freiburg: Verlag Herder. Migne, Jacques-Paul (ed.). 1844–1855. Patrologi. Cursus Completus, Series Prima (Patrologia Latina). 221 vols. Paris: Migne. Vulgate Fabbri, Marco V. 2008. Tagged Latin Vulgate (VULG-T). Accordance Modules. Altamonte Springs, FL: OakTree Software, Inc. Based on Weber, Robert (ed.). 1994. Biblia Sacra Iuxta Vulgatem. 4th edition. Stuttgart: Deutsche Bibelgesellschaft. 484 The Historical Depth of the Tiberian Reading Tradition Rabbinic Sources Maʾagarim Database of the Academy of the Hebrew Language. https://maagarim.hebrew-academy.org.il/Pages/PMain.aspx Mishna (Codex Kaufmann Images) Beer, Georg (ed.). 1968. Faksimile-Ausgabe des Mischnacodex Kaufmann A 50 mit Genehmigung der Ungarischen Akademie der Wissenschaften in Budapest. Jerusalem: no publisher. Mishna (Edition) Eshkol. 2000. Six Divisions of the Mishna (Shisha Sidre Mishna) (MISHNA). Accordance module version 2.0. Eshkol Edition. Accordance Module. Silver Lake, FL: OakTree Software, Inc. Electronic text used by permission of D.B.S., Jerusalem, Israel. Samaritan Sources Samaritan Pentateuch, Oral Tradition Ben-Ḥayyim, Ze’ev. 1977. The Words of the Pentateuch. Vol. 4 of The Literary and Oral Tradition of Hebrew and Aramaic amongst the Samaritans. Jerusalem: The Academy of the Hebrew Language. [Hebrew] References 485 Samaritan Pentateuch, Written Tradition Tal, Abraham. 2010. Samaritan Pentateuch (grammatically tagged) (SAMAR-T). Accordance module version 3.1. Silver Lake, FL: OakTree Software, Inc. Tal, Abraham, and Moshe Florentin (eds). 2010. The Pentateuch: The Samaritan Version and the Masoretic Version. Tel-Aviv: The Haim Rubin Tel-Aviv University Press. Samaritan Targum Tal, Abraham. 2010. Samaritan Targum (SAMTARG). Accordance Module. Altamonte Springs, FL: OakTree Software, Inc. Semitic Inscriptions Dobbs-Allsopp, Frederick W., Jimmy J. M. Roberts, Choon-Leong Seow, and Richard E. Whitaker. 2005. Hebrew Inscriptions: Texts from the Biblical Period of the Monarchy. New Haven, CT: Yale University Press. Gibson, John C. L. 1971–1982. Textbook of Syrian Semitic Inscriptions. 3 volumes. Oxford: Clarendon. KAI = Donner, Herbert and Wolfgang Röllig. 1968–2002. Kanaanäische und Aramäische Inschriften. 4 volumes. Wiesbaden: Harrassowitz. Mazar, Benjamin. 1973. Beth Sheʿarim: Report on the Excavations during 1936–1940. Vol. 1, Catacombs 1–4. Jerusalem: Massada Press on behalf of the Israel Exploration Society and the Institute of Archaeology, Hebrew University. Torczyner, Harry, Lankester Harding, Alkin Lewis, and J. L. Starkey. 1938. The Lachish Letters. Lachish I; Wellcome Ar- 486 The Historical Depth of the Tiberian Reading Tradition chaeological Research Expedition to the Near East Publications 1. Oxford: Oxford University Press. Tur-Sinai, Naphtali H. (1940) 1987. Teʿudot Laḵish: Miḵtavim miyme Yirmeyahu ha-Navi. Jerusalem: The Bialik Institute and the Israel Exploration Society. Syriac (Peshiṭta) Abegg, Martin, and Jerome Lund (eds.). 2010. Syriac Peshitta (Old Testament) (PESHOT-T). Accordance Module. Altamonte Springs, FL: OakTree Software, Inc. Targums (Jewish) Jonathan Cook, Edward M. (ed.). 2008. Targum (TARG-T). Accordance Modules. Altamonte Springs, FL: OakTree Software, Inc. Onqelos Cook, Edward M. (ed.). 2008. Targum (TARG-T). Accordance Modules. Altamonte Springs, FL: OakTree Software, Inc. Jerusalem (= Pseudo-Jonathan) Cook, Edward M., Stephen W. Marler, and Leeor Gottlieb (eds). 2014. Aramaic Targums (Pseudo-Jonathan) (TARG3-T). Accordance Module. Altamonte Springs, FL: OakTree Software, Inc. References 487 Secondary Sources Abegg, Martin G., Jr. 2010. ‘Linguistic Profile of the Isaiah Scrolls’. In Qumran Cave 1.II – The Isaiah Scrolls Part 2: Introduction, Commentary, and Textual Variants, edited by Eugene Ulrich and Peter Flint, 25–41. Discoveries in the Judaean Desert 32. Oxford: Clarendon. Aharoni, Yochanan. 1981. Arad Inscriptions. Judean Desert Studies. Jerusalem: Israel Exploration Society. Aḥituv, Shmuel. 2008. Echoes from the Past: Hebrew and Cognate Inscriptions from the Biblical Period. Carta Handbook. Jerusalem: Carta. Aḥituv, Shmuel, W. Randall Garr, and Steven E. Fassberg. 2016. ‘Epigraphic Hebrew’. In A Handbook of Biblical Hebrew, 2 vols., ed. W. Randall Garr and Steven E. Fassberg, I:55–68, II:36–42. Winona Lake, IN: Eisenbrauns. Andersen, Francis I., and A. Dean Forbes. 1986. Spelling in the Hebrew Bible. Biblia et Orientalia. Rome: Biblical Institute Press. ———. 2013. ‘Matres Lectionis: Biblical Hebrew’. In Encyclopedia of Hebrew Language and Linguistics, edited by Geoffrey Khan et al., II:607–11. Leiden: Brill. Andersen, T. David. 2000. ‘The Evolution of the Hebrew Verbal System’. Zeitschrift für Althebraistik 13:1–66. Arentsen, Niek. 2020. ‘Leshon Yishʿayhu ha-Sheni (Yish. 40–66) u-Miqomah be-Toldot ha-Lashon ha-ʿIvrit’. PhD Disserta- tion, The Hebrew University of Jerusalem. Arnold, Bill T., and John H. Choi. 2003. A Guide to Biblical Hebrew Syntax. Cambridge: Cambridge University Press. 488 The Historical Depth of the Tiberian Reading Tradition Baden, Joel S. 2010. ‘Hithpael and Niphal in Biblical Hebrew: Semantic and Morphological Overlap’. Vetus Testamentum 60:33–44. Bar-Asher, Moshe. 2016. ‘The Contribution of Tannaitic Hebrew to Understanding Biblical Hebrew’. In A Handbook of Bibli- cal Hebrew—Volume 1: Periods, Corpora, and Reading Traditions, edited by W. Randall Garr and Steven E. Fassberg, 203–14. Winona Lake, IN: Eisenbrauns. Barr, James. 1981. ‘A New Look at Kethibh-Qere’. Oudtestamentische Studiën 21:19–37. ———. 1984 ‘Migraš in the Old Testament.’ Journal of Semitic Studies 29:15–31. ———. 1987. Comparative Philology and the Text of the Old Testament, with Additions and Corrections. Winona Lake, IN: Eisenbrauns. ———. 1989a. ‘“Determination” and the Definite Article in Biblical Hebrew’. Journal of Semitic Studies 34:307–33. ———. 1989b. The Variable Spellings of the Hebrew Bible. Schweich Lectures 1986. Oxford: Published for the British Academy by Oxford University Press. Barth, Jakob. ‘Vergleichende Studien: III. Das i-Imperfect im Nordsemitischen’. Zeitschrift der Deutschen Mor- genländischen Gesellschaft 43:177–91. ———. 1890. ‘Das passive Qal und seine Participien’. Jubelschrift zum Siebzigsten Geburtstag des Dr. Israel Hildesheimer, 145– 53. Berlin: Engel. ———. 1891. Die Nominalbildung in den semitischen Sprachen. 2 vols in 1. Leipzig: J. C. Hinrichs. References 489 Bauer, Hans, and Pontus Leander. 1922. Historische Grammatik der hebräischen Sprache des alten Testamentes. Halle: Niemeyer. BDB = Brown, Francis, Samuel R. Driver, and Charles A. Briggs. 1906. The Brown-Driver-Briggs Hebrew and English Lexicon. Boston: Houghton, Mifflin and Company. Bean, Albert Frederick. 1976. ‘A Phenomenological Study of the Hithpa‘el Verbal Stem in the Hebrew Old Testament’. PhD dissertation, The Southern Baptist Theological Seminary. Beit-Arieh, Itzhaq. 1993. ‘A Literary Ostracon from Horvat ‘Uza’. Tel Aviv 20/1:55–65. Bendavid, Abba. 1967–1971 Leshon ha-Miqra u-Lshon Ḥaḵamim. 2 vols. Tel-Aviv: Dvir. Ben-Dov, Jonathan. 2010. ‘The Elohistic Psalter and the Writing of Divine Names at Qumran’. In The Dead Sea Scrolls and Contemporary Culture: Proceedings of the Israel Museum Conference, edited by Adolfo D. Roitman, Lawrence H. Schiffman, and Shani Tzoref, 79–104. Studies on the Texts of the Desert of Judah 93. Leiden: Brill. Ben-Ḥayyim, Zeʾev. 1954. Studies in the Traditions of the Hebrew Language. Madrid: Instituto Arias Montano. ———. 1958. ‘The Samaritan Tradition and Its Relationship to the Tradition of the Dead Sea Scrolls’. Leshonenu 22:223– 45. [Hebrew] ———. 1977. ʿIvrit va-ʿAramit Nusaḥ Šomron, vol. 4: Mile Tora. Jerusalem: The Academy of the Hebrew Language. ———. 2000. A Grammar of Samaritan Hebrew. Jerusalem: Magnes and Winona Lake, IN: Eisenbrauns. 490 The Historical Depth of the Tiberian Reading Tradition Bergsträsser, Gotthelf. 1918–1929. Hebräische Grammatik. 2 vols. Leipzig: Vogel. Beuken, W. 2004. ‘‫’ש ַכב‬. ָ In Theological Dictionary of the Old Testament, XVI:659–71. Edited by G. Johannes Botterweck, Helmer Ringgren, and Heinz-Josef Fabry. Translated by Douglas W. Scott. Grand Rapids, MI: Eerdmans. Blake, Frank R. 1901. ‘The Internal Passive in Semitic’. Journal of the American Oriental Society 22:45–54. ———. 1950. ‘The Apparent Interchange between a and i in Hebrew’. Journal of Near Eastern Studies 9:76–83. Blau, Joshua. 1982. ‘Remarks on the Development of Some Pronominal Suffixes in Hebrew’. Hebrew Annual Review 6:61– 67. ———. 2010. Phonology and Morphology of Biblical Hebrew. Linguistic Studies in Ancient West Semitic 2. Winona Lake, IN: Eisenbrauns. Bloch, Yigal. 2007. ‘From Linguistics to Text-Criticism and Back: Wayyiqṭōl Constructions with Long Prefixed Verbal Forms in Biblical Hebrew’. Hebrew Studies 48:141–70. Borg, Alexander. 2000. ‘Some Observations on the ‫ יום הששי‬Syndrome on the Hebrew of the Dead Sea Scrolls’. Pages 26– 39 in Diggers at the Well, edited by Takamitsu Muraoka and John F. Elwolde. Leiden: Brill. Böttcher, J. Freidrich. 1866–1868. Ausführliches Lehrbuch der hebräischen Sprache. 2 vols. Leipzig: Barth. Breuer, Mordekhai. 1980. ‘Le-veruran shel Sugyot be-ṭaʿame ha- Miqra u-v-Niqudo, 3: Ḥilufe ha-Niqud ben Ben-Asher le- Ven Naftali’. Leshonenu 44:243–62. References 491 ———. 1981. ‘KĔṮIV, QĔRÉ AND Ṭĕ'IM (written, read and chanted)’. Leshonenu 41:260–69. Breuer, Yochanan. 1991. ‘Maḥloqet niqud u-ṭʿamim ba-ḥaluqat pesuqim’. In Sefer Yovel for R. Mordechai Breuer: A Collection of Articles in Jewish Studies. 2 vol. Edited by Mosheh BarAsher, 191–242. Jerusalem: Academon. ———. 2013. ‘Morphology: Rabbinic Hebrew’. In Encyclopedia of Hebrew Language and Linguistics, edited by Geoffrey Khan et al., II:735–41. Leiden: Brill. ———. 2022. ‘Dissonance between Masoretic Vocalisation and Cantillation in Biblical Verse Division’. In Studies in the Masoretic Tradition of the Hebrew Bible, edited by Daniel J. Crowther, Aaron D. Hornkohl, and Geoffrey Khan, 243– 88.Cambridge Semitic Languages and Cultures 15. Cambridge: Open Book Publishers and University of Cambridge. Brockelmann, Carl. 1908–1913. Grundriss der vergleichenden Grammatik der semitischen Sprachen. 2 vols. Berlin: Reuther & Reichard. Brønno, Einar. 1943. Studien über hebräische Morphologie und Vokalismus: Auf Grundlage der mercatischen Fragmente der zweiten Kolumne der Hexapla des Origenes. Abhandlungen für die Kunde des Morgenlandes 28. Leipzig: F. A. Brockhaus. ———. 1970. Die Aussprache der hebräischen Laryngale nach Zeugnissen des Hieronymus. Aarhus: Universitetsforlaget. Burke, Aaron A. 2007. ‘Magdalūma, Migdālîm, Magdoloi, and Majādīl: The Historical Geography and Archaeology of the 492 The Historical Depth of the Tiberian Reading Tradition Magdalu (Migdāl)’. Bulletin of the American Schools of Oriental Research 346:29–57. Bybee, Joan L., Revere D. Perkins, and William Pagliuca. 1994. The Evolution of Grammar: Tense, Aspect, and Modality in the Languages of the World. Chicago: University of Chicago Press. CAL = Comprehensive Aramaic Lexicon. https://cal.huc.edu. Cohen, Maimon. 2007. The Kethiḇ and Qeri System in the Biblical Text: A Linguistic Analysis of the Various Traditions Based on the Manuscript ‘Keter Aram Tsova’. Publications of the Perry Foundation for Biblical Research in the Hebrew University of Jerusalem. Jerusalem: The Hebrew University Magnes Press. Cohen, Ohad. 2013. The Verbal Tense System in Late Biblical Hebrew Prose. Harvard Semitic Studies 63. Winona Lake, IN: Eisenbrauns. Cook, Edward M. 2017. ‘Linguistic Contact and the Genesis of Mishnaic Hebrew’. The Edward Ullendorff Lectures in Semitic Philology 4. Cambridge: Faculty of Asian and Middle Eastern Studies, University of Cambridge. Cook, John A. 2012. Time and the Biblical Hebrew Verb: The Expression of Tense, Aspect, and Modality in Biblical Hebrew. Linguistic Studies in Ancient West Semitic 7. Winona Lake, IN: Eisenbrauns. Cross, Frank Moore. 1985. ‘A Literate Soldier: Lachish Letter III’. In Biblical and Related Studies Presented to Samuel Iwry, edited by Ann Kort and Scott Morchauser, 41–47. Winona Lake, IN: Eisenbrauns. References 493 ———. 1998. ‘The Stabilization of the Canon of the Hebrew Bible’. In From Epic to Canon: History and Literature in Ancient Israel, 219–29. Baltimore: Johns Hopkins University Press. Cross, Frank Moore, and David Noel Freedman. 1952. Early Hebrew Orthography: A Study of the Epigraphic Evidence. American Oriental Series 36. New Haven, CT: American Oriental Society. Dallaire, Hélène. 2014. The Syntax of Volitives in Biblical Hebrew and Amarna Canaanite Prose. Linguistic Studies in Ancient West Semitic 9. Winona Lake, IN: Eisenbrauns. Delitzsch, Franz. 1877. Commentary on the Song of Songs and Ecclesiastes. Translated by M. G. Easton. Edinburgh: T. & T. Clark. Dihi, Haim. 2004. ‘Morphological and Lexical Innovations in the Book of Ben Sira’. PhD dissertaion, Ben Gurion University of the Negev. [Hebrew] Driver, Godfrey Rolles. 1965. The Judaean Scrolls: The Problems and a Solution. Oxford: Basil Blackwell. Driver, Samuel Rolles. 1882. ‘On Some Alleged Linguistic Affinities of the Elohist’. Journal of Philology 11:201–36. ———. 1890. Notes on the Hebrew Text and the Topography of the Books of Samuel. Oxford: Clarendon. ———. 1898 An Introduction to the Literature of the Old Testament. Revised edition. Edinburgh: T. & T. Clark. ———. (1892) 1998. A Treatise on the Use of the Tenses in Hebrew and Some Other Syntactical Questions. 3rd edition. London: Oxford University Press. Reprint with an Introductory Essay by W. Randall Garr. The Biblical Resource Series. Grand 494 The Historical Depth of the Tiberian Reading Tradition Rapids, MI and Cambridge, UK: William B. Eerdmans Publishing Company and Livonia, MI: Dove Booksellers. Elitzur, Yoel. 1987. ‘The Stem Qittul in Mishnaic Hebrew Accord- ing to Cod. Kaufmann’. Language Studies 2–3:67–93. [Hebrew] ———. 2015. ‘The Divine Name ADNY in the Hebrew Bible: Surprising Findings’. Liber Annuus 65:87–106. ———. 2018a. ‘Diachrony in Standard Biblical Hebrew: The Pentateuch vis-à-vis the Prophets/Writings’. Journal of Northwest Semitic Languages 44:81–101. ———. 2018b. ‘The Interface between Language and Realia in the Preexilic Books of the Bible’. Hebrew Studies 59:129–47. ———. 2019. ‘The Names of God and the Dating of the Biblical Corpus’. In The Believer and the Modern Study of the Bible, edited by Tova Ganzel, Yehudah Brandes, and Chayuta Deutsch, 428–42. Boston: Academic Studies Press. ———. 2022. ‘Emergence and Disappearance of Words and Expressions in Pre-Exilic Biblical Hebrew’. Revue Biblique 129:481–504. Fassberg, Steven E. 1991. A Grammar of the Palestinian Targum Fragments from the Cairo Genizah. Harvard Semitic Studies 38. Atlanta: Scholars Press. ———. 2001. ‘The Movement of Qal to Piʿel in Hebrew and the Disappearance of the Qal Internal Passive’. Hebrew Studies 42:243–55. ———. 2012. ‘The Kethiv/Qere ‫הוא‬, ִ Diachrony, and Dialectology’. In Diachrony in Biblical Hebrew, edited by Cynthia L. References 495 Miller-Naudé and Ziony Zevit, 171–80. Linguistic Studies in Ancient West Semitic 8. Winona Lake, IN: Eisenbrauns. ———. 2013a. ‘Pausal Forms’. In Encyclopedia of Hebrew Language and Linguistics, edited by Geoffrey Khan et al., III:54– 55. Leiden: Brill. ———. 2013b. ‘Dead Sea Scrolls: Linguistic Features’. In Encyclopedia of Hebrew Language and Linguistics, edited by Geoffrey Khan et al., I:663–69. Leiden: Brill. Friedman, Richard Elliott. 1997. Who Wrote the Bible? 2nd edition. San Francisco: HarperSanFrancisco. Geiger, Gregor. 2016. ‘Wayyiqtol-Langformen der Verben III.h’. Liber Annuus 66:36–67. Ginsberg, Harold L. 1929. ‘Studies on the Biblical Hebrew Verb’. American Journal of Semitic Languages and Literatures 46:53– 56. ———. 1934. ‘Mi-baʿad la-Masoret’. Tarbiz 5:208–23. ———. 1936. ‘Nosafot le-“Mi-baʿad la-Masoret”’. Tarbiz 7:543. GKC = Emil Kautsch (ed.). 1909. Gesenius’ Hebrew Grammar. Translated by Arthur E. Cowley. Oxford: Clarendon. Gogel, Sandra L. 1998. A Grammar of Epigraphic Hebrew. Atlanta: Scholars Press. Goldfajn, Tal. 1998. Word Order and Time in Biblical Hebrew Narrative. Oxford Theology and Religion Monographs. Oxford: Clarendon. Green, William Henry. 1872. A Grammar of the Hebrew Language. 3rd edition. New York: John Wiley & Son. Gzella, Holger. 2018. ‘Untypical Wayyiqṭol Forms in Hebrew and Early Linguistic Diversity’. In The Unfolding of Your Word 496 The Historical Depth of the Tiberian Reading Tradition Gives Light: Studies on Biblical Hebrew in Honor of George L. Klein, edited by Ethan C. Jones, 21–37. University Park, PA: Eisenbrauns. Habib, Joseph. 2020. ‘Qere and Ketiv in the Exegesis of the Karaites and Saadya Gaon’. In Studies in Semitic Vocalisation and Reading Traditions, edited by Aaron D. Hornkohl and Geoffrey Khan, 281–330. Cambridge Semitic Languages and Cultures 3. Cambridge: Open Book Publishers and University of Cambridge Faculty of Asian and Middle Eastern Studies. doi.org/10.11647/OBP.0207.08 ———. 2021. ‘Accents, Pausal Forms and Qere/Ketiv in the Bible Translations and Commentaries of Saadya Gaon and the Karaites of Jerusalem’. PhD dissertation, University of Cambridge. HALOT = Baumgartner, Walter, and J.J. Stamm, et al. 1994– 2000. The Hebrew and Aramaic Lexicon of the Old Testament. 5 vols. Translated and edited by M. E. J. Richardson. Leiden: Brill. Heine, Bernd. 1993. Auxiliaries: Cognitive Forces and Grammaticalization. Oxford: Oxford University Press. Hendel, Ronald. 1996. ‘In the Margins of the Hebrew Verbal System: Situation, Tense, Aspect, Mood’. Zeitschrift für Althebraistik 9:152–81. ———. 2000. ‘“Begetting” and “Being Born” in the Pentateuch: Notes on Historical Linguistics and Source Criticism’. Vetus Testamentum 50:38–46. ———. 2016. Steps to a New Edition of the Hebrew Bible. TextCritical Studies 10. Atlanta: SBL Press. References 497 Hetterle, Katja. 2015. Adverbial Clauses in Cross-linguistic Perspective. Trends in Linguistic Studies and Monographs 289. Berlin: Mouton de Gruyter. Hetzron, Robert. 1969. ‘The Evidence for Perfect *yˈaqtul and Jussive *yaqtˈul in Proto-Semitic’. Journal of Semitic Studies 14:1–21. ———. 1976. ‘Two Principles of Genetic Reconstruction’. Lingua 38:89–108. Holmstedt, Robert. 2008. ‘The Relative Clause in Canaanite Epigraphic Texts’. Journal of Northwest Semitic Languages 34/2:1–34. ———. 2016. The Relative Clause in Biblical Hebrew. Linguistic Studies in Ancient West Semitic 10. Winona Lake, IN: Eisenbrauns. Hornkohl, Aaron D. 2013a. Ancient Hebrew Periodization and the Language of the Book of Jeremiah: The Case for a Sixth-century Date of Composition. Studies in Semitic Languages and Linguistics 74. Leiden and Boston: Brill. ———. 2013b. ‘Biblical Hebrew: Periodisation’. In Encyclopedia of Hebrew Language and Linguistics, edited by Geoffrey Khan et al., I:315–25. Leiden: Brill. ———. 2016a. ‘Hebrew Diachrony and the Linguistic Periodisation of Biblical Texts: Observations from the Perspective of Reworked Pentateuchal Material’. Journal for Semitics 25/2:1004–63. ———. 2016b. ‘Transitional Biblical Hebrew’. In A Handbook of Biblical Hebrew, edited by W. Randall Garr and Steven E. Fassberg, I:32–42. Winona Lake, IN: Eisenbrauns. 498 The Historical Depth of the Tiberian Reading Tradition ———. 2018. ‘Diachronic Exceptions in the Comparison of Tiberian and Qumran Hebrew: The Preservation of Early Linguistic Features in Dead Sea Scrolls Biblical Hebrew’. In The Reconfiguration of Hebrew in the Hellenistic Period: Proceedings of the Seventh International Symposium on the Hebrew of the Dead Sea Scrolls and Ben Sira at Strasbourg University, June 2014, edited by Jan Joosten Daniel Machiela, and Jean-Sébastien Rey, 61–92. Studies on the Texts of the Desert of Judah 124. Leiden: Brill. ———. 2020a. ‘The Hebrew of the Dead Sea Scrolls and the Tiberian Reading Tradition: Shared Departures from the Masoretic Written Tradition’. Dead Sea Discoveries 27:410–25. ———. 2020b. ‘Discord between the Tiberian Written and Reading Traditions: Two Case Studies’. In Studies in Semitic Vocalisation and Reading Traditions, edited by Aaron D. Hornkohl and Geoffrey Khan, 227–80. Cambridge Semitic Languages and Cultures 3. Cambridge: Open Book Publishers and University of Cambridge Faculty of Asian and Middle Eastern Studies. doi.org/10.11647/OBP.0207.07 ———. 2021a. ‘Niphalisation in Ancient Hebrew: A Perspective from the Samaritan Tradition’. Journal for Semitics 30/2:1– 17. ———. 2021b. ‘The Linguistic Profile of Select Reworked Bible Material vis-à-vis Masoretic Hebrew and Some Ramifications Thereof’. In Hebrew Texts and Language of the Second Temple Period: Proceedings of an Eighth Symposium on the Hebrew of the Dead Sea Scrolls and Ben Sira, edited by Steven References 499 E. Fassberg, 127–52. Studies on the Texts of the Desert of Judah 134. Leiden: Brill. ———. 2022. ‘Tiberian ketiv-qere and the Combined Samaritan Written-Reading Tradition: Points of Contact and Contrast’. In Studies in the Masoretic Tradition of the Hebrew Bible, edited by Daniel J. Crowther, Aaron D. Hornkohl, and Geoffrey Khan, 115–62. Cambridge Semitic Languages and Cultures 15. Cambridge: Open Book Publishers and University of Cambridge. Huehnergard, John. 1988. ‘The Early Hebrew Prefix Conjugations’. Hebrew Studies 29:19–23. ———. 2005. ‘Hebrew Verbs I-w/y and a Proto-Semitic Sound Rule’. In Memoriae Igor M. Diakonoff, edited by L. Kogan, N. Koslova, S. Loesov, and S. Tishchenko, 457–74. Babel und Bibel 2. Winona Lake, IN: Eisenbrauns. ———. 2019. ‘Proto-Semitic’. In The Semitic Languages, 2nd edition, edited by John Huehnergard and Na‘ama Pat-El, 49– 79. London: Routledge. Hughes, Jeremy. 1994. ‘Post-Biblical Features of Biblical Hebrew Vocalization’. In Language, Theology, and the Bible: Essays in Honour of James Barr, edited by Samuel E. Balentine and John Barton 67–80. Oxford: Clarendon. Hurvitz, Avi. 1974. ‘The Date of the Prose-Tale of Job Linguistically Reconsidered’. Harvard Theological Review 67:17–34. ———. 1982. A Linguistic Study of the Relationship Between the Priestly Source and the Book of Ezekiel: A New Approach to an Old Problem. Cahiers de la Revue Biblique 20. Paris: Gabalda. 500 The Historical Depth of the Tiberian Reading Tradition ———. 1983. ‘The Language of the Priestly Source and Its His- torical Setting: The Case for an Early Date’. Proceedings of the Eighth World Congress of Jewish Studies, Jerusalem, August 16–21, 1981: Panel Sessions: Bible Studies and Hebrew Lan- guage, 83–94. Jerusalem: World Union of Jewish Studies. ———. 1990. Review of Daniel C. Fredericks, Qoheleth’s Language: Re-evaluating Its Nature and Date, Ancient Near Eastern Texts and Studies 3. New York: Mellen, 1988, in Hebrew Studies 31:144–54. ———. 2007. ‘The Language of Qoheleth and Its Historical Setting within Biblical Hebrew’. The Language of Qohelet in its Context: Essays in Honour of Prof. A. Schoors on the Occasion of his Seventieth Birthday, edited by A. Berlejung and P. Van Hecke, 23–34. Leuven: Peeters. ———. 2013. ‘Biblical Hebrew, Late’. In Encyclopedia of Hebrew Language and Linguistics, edited by Geoffrey Khan et al., I:329–38. Leiden: Brill. ———. 2014. A Concise Lexicon of Late Biblical Hebrew: Linguistic Innovations in the Writings of the Second Temple Period. Supplements to Vetus Testamentum 160. Leiden: Brill. Hutton, Jeremy. 2013a. ‘Epigraphic Hebrew’. Encyclopedia of Hebrew Language and Linguistics, edited by Geoffrey Khan et al. I:835–42. Leiden: Brill. ———. 2013b. ‘Orthography: Epigraphy’. Encyclopedia of Hebrew Language and Linguistics, edited by Geoffrey Khan et al. II:964–71. Leiden: Brill. References 501 Japhet, Sara. 2009. The Ideology of the Book of Chronicles and Its Place in Biblical Thought. Translated by Anna Barber. Winona Lake, IN: Eisenbrauns. JM = Joüon, Paul, and Takamitsu Muraoka. 2008. A Grammar of Biblical Hebrew. Subsidia Biblica 27. Rome: Pontifical Biblical Institute Press. Joosten, Jan. 2014. ‘Linguistic Clues as to the Date of the Book of Job: A Mediating Position’. In Interested Readers: Festschrift for David Clines, edited by James K. Aitken, Jeremy M. S. Clines, and Christl M. Maier, 347–57. Atlanta: Society of Biblical Literature. ———. 2015. ‘The Tiberian Vocalization and the Hebrew of the Second Temple Period’. In Hebrew of the Late Second Temple Period: Proceedings of a Sixth International Symposium on the Hebrew of the Dead Sea Scrolls and Ben Sira, edited by Eibert Tigchelaar and Pierre Van Hecke, 25–36. Studies on the Texts of the Desert of Judah 114. Leiden: Brill. ———. 2019. ‘The Linguistic Dating of the Joseph Story’. Hebrew Bible and Ancient Israel 8/1:24–43. Joüon, Paul. 1920. ‘Études de morphologie hébraïque’. Biblica 1/3:353–71. Kahle, Paul. 1921. ‘Die überlieferte Aussprache des Hebräischen und die Punktation der Masoreten’. Zeitschrift für die alttes- tamentliche Wissenschaft 39:230–39. ———. 1947. The Cairo Geniza. Schweich Lectures 1941. London: Oxford University Press. ———. 1959. The Cairo Geniza. Oxford: Blackwell. 502 The Historical Depth of the Tiberian Reading Tradition Kantor, Benjamin. 2017. ‘The Second Column (Secunda) of Origen’s Hexapla in Light of Greek Pronunciation’. PhD dissertation, University of Texas. ———. 2020. ‘The Development of the Hebrew wayyiqṭol (‘waw Consecutive’) Verbal Form in Light of Greek and Latin Transcriptions of Hebrew’. In Studies in Semitic Vocalisation and Reading Traditions, edited by Aaron D. Hornkohl and Geoffrey Khan, 55–132. Cambridge Semitic Language and Cultures 3. Cambridge: Open Book Publishers and University of Cambridge Faculty of Asian and Middle Eastern Studies. doi.org/10.11647/OBP.0207 Kartveit, Magnar. 2009. The Origin of the Samaritans. Leiden: Brill. Kawashima, Robert S. 2010. ‘“Orphaned” Converted Tense Forms in Classical Biblical Hebrew Prose’. Journal of Semitic Stud- ies 40/1:11–35. Keenan, Edward L., and Bernard Comrie. 1977. ‘Noun Phrase Accessibility and Universal Grammar’. Linguistic Inquiry 8:63– 99. Khan, Geoffrey. 1991. ‘Morphological Markers of Individuation in Semitic Languages and Their Function in the Semitic Tense System’. In Proceedings of the Fifth International Hamito-Semitic Congress: Band 2, Cushitic, Egyptian, Omotic, Semitic, edited by Hans G. Mukarovsky, 235–44. Vienna: Afro-Pub. ———. 2013a. A Short Introduction to the Tiberian Masoretic Bible and Its Reading Tradition. Piscataway, NJ: Gorgias Press. ———. 2013b. ‘Biblical Hebrew: Linguistic Background of the Masoretic Text’. In Encyclopedia of Hebrew Language and References 503 Linguistics, edited by Geoffrey Khan et al. I:304–15. Leiden: Brill. ———. 2018a. ‘How Was the Dageš in Biblical Hebrew ‫ ָב ִתים‬Pronounced and Why Is It There?’. Journal of Semitic Studies 63/2:323–51. ———. 2018b. ‘Orthoepy in the Tiberian Reading Tradition of the Hebrew Bible and Its Historical Roots in the Second Temple Period’. Vetus Testamentum 68:1–24. ———. 2020. The Tiberian Pronunciation Tradition of Biblical Hebrew: Including a Critical Edition and English Translation of the Sections on Consonants and Vowels in the Masoretic Treatise Hidāyat al-Qāriʾ ‘Guide for the Reader’. 2 vols. Cambridge Semitic Languages and Cultures 1. Cambridge: Open Book Publishers and University of Cambridge Faculty of Asian and Middle Eastern Studies. doi.org/10.11647/OBP.0163 ———. 2021. ‘The Coding of Discourse Dependency in Biblical Hebrew Consecutive weqaṭal and wayyiqṭol’. In New Perspectives in Biblical and Rabbinic Hebrew, edited by Aaron D. Hornkohl and Geoffrey Khan, 299–354. Cambridge Semitic Languages and Cultures 7. Cambridge: Open Book Publishers and University of Cambridge Faculty of Asian and Middle Eastern Studies. doi.org/10.11647/OBP.0250.12 Kiel, Yehuda. 2000. Sefer Bereshit: Parashiyot Vayera–Vayyishlaḥ. Daʿat Miqra. Jerusalem: Mosad ha-Rav Kook. Kogut, Simcha. 1994. Ha-Miqra ben Ṭaʿamim le-Farshanut. Jerusalem: Magnes Press. 504 The Historical Depth of the Tiberian Reading Tradition Koller, Aaron. 2013. ‘Attenuation’. In Encyclopedia of Hebrew Language and Linguistics, edited by Geoffrey Khan et al. I:231– 32. Leiden: Brill. Kutscher, Eduard Y. 1963. ‘Leshon Ḥazal’. In Sefer Ḥanokh Yalon: Qovets maʾamarim. Edited by Shaul Liberman, Shraga Abramson, Yeḥezkel Kutscher, and Shaul Esh. Jerusalem: Kiryat Sefer. ———. 1969. ‘Studies in the Grammar of Mishnaic Hebrew (According to MS Kaufmann)’. Bar-Ilan 2:51–77. ———. 1972. ‘Some Problems of the Lexicography of Mishnaic Hebrew and Its Comparison with Biblical Hebrew’. In Ar- chive of the New Dictionary of Rabbinical Literature, I:29–82. Ramat-Gan: Bar-Ilan University. [Hebrew] ———. 1974. Language and Linguistic Background of the Isaiah Scroll (1QIsaa). Studies on the Texts of the Desert of Judah 6. Leiden: Brill. ———. 1982. A History of the Hebrew Language. Jerusalem: Magnes and Leiden: Brill. Lambdin, Thomas O. 1973. Introduction to Biblical Hebrew. London: Darton, Longman and Todd. ———. 1985. ‘Philippi’s Law reconsidered’. Biblical Studies Presented to Samuel Iwry, edited by Ann Kort and Scott Morschauser, 135–45. Winona Lake, IN: Eisenbrauns. Lambert, Mayer. 1895. ‘Le mot ‫ יום‬suivi des nombres ordinaux’. Revue des études juives 31:279–81. ———. 1898. ‘L’article dans la poésie hébräique’. Revue des études juives 37:203–9. References 505 ———. 1900. ‘L’emploi du nifal en hébreu’. Revue des études juives 41:196–214. ———. 1931. Traité de Grammaire Hébraïque. Paris: Presses Universitaires de France. Lemaire, André. 1995. ‘Epigraphie palestinienne: Nouveaux documents II: Decennie 1985–1995’. Henoch 17:221–22. Ley, Julius. 1891. ‘Über den Gebrauch des Artikels in der rhythmischen Poesie der Hebräer’. In Neue Jahrbücher fur Philolo- gie und Paedagogik (also known as Jahrbücher für classische Philologie), 2te Abteilung, vol. 144, 341–51. Lipschütz, Lazar. 1964. ‘Kitāb al-Khilaf, the Book of the Ḥillufim: Mishael Ben Uzziel’s Treatise on the Differences between Ben Asher and Ben Napthali’. Textus 4:2–29. Mandell, Alice. 2013. ‘Biblical Hebrew, Archaic’. In Encyclopedia of Hebrew Language and Linguistics, edited by Geoffrey Khan et al., I:325–29. Leiden: Brill. Morag, Shelomo. 1974. ‘Ha-Masoret ha-Ṭavranit shel Leshon haMiqra: Homogeniyut ve-Heṭerogeniyut’. Peraqim: Sefer haShana shel Meḵon Shoqen, 105–44. ———. 1988. ‘Qumran Hebrew: Typological Observations’. Vetus Testamentum 38:148–64. Moreshet, Menaḥem. 1996. ‘Hifʿil lelo Hevdel min-ha-Qal biLshon Ḥazal (be-Hashvaʾa li-Lshon ha-Miqra)’. In Qovets Maʾamarim bi-Lshon Ḥazal 2, edited by Moshe Bar-Asher, 263–95. Jerusalem: Magnes. Morgenstern, Matthew. 2007. ‘The System of Independent Pronouns at Qumran and the History of Hebrew in the Second Temple Period.’ Shaʿarei Lashon: Studies in Hebrew, Aramaic, 506 The Historical Depth of the Tiberian Reading Tradition and Jewish Languages Presented to Moshe Bar-Asher, edited by Aharon Maman, Steven E. Fassberg, and Yochanan Breuer, I:44–63. Jerusalem: Bialik Institute. [Hebrew] Moscati, Sabatino. 1964. An Introduction to the Comparative Grammar of the Semitic Languages: Phonology and Morphology. Porta Linguarum Orientalium 6. Wiesbaden: Otto Harrassowitz. Moshavi, Adina, and Susan Rothstein. ‘Indefinite Numerical Construct Phrases in Biblical Hebrew’. Journal of Semitic Studies 63:99–123. Muraoka, Takamitsu. 1997. ‘The Alleged Final Function of the Biblical Hebrew Syntagm <Waw + a Volitive Verb Form>’. In Narrative Syntax and the Hebrew Bible: Papers of the Tilburg Conference 1996, edited by Ellen van Wolde, 229–41. Biblical Interpretation Series 29. Leiden: Brill. ———. 2000. ‘An Approach to the Morphosyntax and Syntax of Qumran Hebrew’. In Diggers at the Well: Proceedings of a Third International Symposium on the Hebrew of the Dead Sea Scrolls and Ben Sira, edited by Takamitsu Muraoka and John F. Elwolde, 193–214. Leiden: Brill. ———. 2013. ‘Isaiah Scroll (1QIsaa)’. In Encyclopedia of Hebrew Language and Linguistics, edited by Geoffrey Khan et al., II:343–48. Leiden: Brill. Nöldeke, Theodor. 1904. Beiträge zur semitischen Sprachwissenschaft. Strassburg: K. J. Trübner. Ofer, Yosef. 2019. The Masora on Scripture and Its Methods. Fontes et Subsidia ad Bibliam pertinentes 7. Berlin: Walter de Gruyter. References 507 Orlinsky, Harry. 1944. ‘The Hebrew Root ŠKB’. Journal of Biblical Literature 63/1:19–44. Parunak, H. van Dyke. 1978. ‘The Orthography of the Arad Ostraca’. Bulletin of the American Schools of Oriental Research 230:25–31. Pat-El, Na’ama. 2009. ‘The Development of the Semitic Definite Article: A Syntactic Approach’. Journal of Semitic Studies 54/1:19–50. Paul, Shalom. 2012. ‘Signs of Late Biblical Hebrew in Isaiah 40– 66’. In Diachony in Biblical Hebrew, edited by Cynthia Miller-Naudé and Ziony Zevit, 293–99. Linguistic Studies in Ancient West Semitic 8. Winona Lake, IN: Eisenbrauns. Pérez Fernández, Miguel. 1997. An Introductory Grammar of Rabbinic Hebrew. Leiden: Brill. Price, James D. 2006. ‘Exegesis and Pausal Forms with NonPausal Accents in the Hebrew Bible’. Presented at the Southeastern Regional Meeting of the Evangelical Theological Society, 18 March. http://www.jamesdprice.com/images/Pausal_Forms_ETS_paper.pdf Pummer, Reinhard. 2012. Review of The Origin of the Samaritans, by Magnar Kartveit. Journal of Hebrew Scriptures 12:592– 99. Qimron, Elisha. 1980. ‘Le-Milonah shel Megilat ha-Miqdash’. Shnaton la-Miqra u-l-Ḥeqer ha-Mizraḥ ha-Qadum 4, edited by Moshe Weinfeld, 239–62. Jerusalem: Neuman. ———. 1986. The Hebrew of the Dead Sea Scrolls. Atlanta: Scholars Press. 508 The Historical Depth of the Tiberian Reading Tradition ———. 1997. ‘A New Approach to the Use of Forms of the Imperfect without Personal Endings’. In The Hebrew of the Dead Sea Scrolls and Ben Sira: Proceedings of a Symposium Held at Leiden University, December 1995, edited by Takamitsu Muraoka and John Elwolde, 174–81. Studies on the Texts of the Desert of Judah 26. Leiden: Brill. ———. 2008. ‘The Type ‫ וָ ֶא ְּבנֶ ה‬in the Hebrew of the Dead Sea Scrolls’. In Conservatism and Innovation in the Hebrew Language of the Hellenistic Period: Proceedings of a Fourth International Symposium on the Hebrew of the Dead Sea Scrolls and Ben Sira, edited by Jan Joosten and Jean-Sébastien Rey, 149–54. Studies on the Tests of the Desert of Judah 73. Leiden and Boston: Brill. ———. 2018. A Grammar of the Hebrew of the Dead Sea Scrolls. Jerusalem: Ben-Zvi. Rabin, Chaim. 1958. ‘The Historical Background of Qumran He- brew’. In Aspects of the Dead Sea Scrolls, edited by Chaim Rabin and Yigael Yadin, 144–61. Scripta Hierosolymitana 4. Jerusalem: Magnes. ———. 1972. ‘Ha-Reqaʿ ha-Hisṭori šel ha-ʿIvrit šel Qumran’. Qoveṣ Ma⁠ʾamarim bi-Lšon Ḥazal 1:355–82. ———. 1976. ‘Hebrew and Aramaic in the First Century’. In The Jewish People in the First Century: Historical Geography, Political History, Social, Cultural and Religious Life and Institutions, edited by Shemuel Safrai and Menahem Stern, 1007– 1039. Assen: Van Gorcum. References 509 Rabinowitz, Isaac. 1984. ‘ʾAz Followed by Imperfect Verb-Form in Preterite Contexts: A Redactional Device in Biblical Hebrew’. Vetus Testamentum 34:53–62. Rainey, Anson F. 1986. ‘The Ancient Hebrew Prefix Conjugation in the Light of Amarnah Canaanite’. Hebrew Studies 27:4– 19. Rand, Michael. 2006. Introduction to the Grammar of Hebrew Poetry in Byzantine Palestine. Gorgias Dissertations 22, Language and Linguistics. Piscataway, NJ: Gorgias Press. Rendsburg, Gary. 1982. ‘A New Look at Pentateuchal HWʾ’. Biblica 63:351–69. ———. 2002. ‘Some False Leads in the Identification of Late Biblical Hebrew Texts: The Cases of Genesis 24 and 1 Samuel 2:27–36’. Journal of Biblical Literature 121:23–46. ———. 2006. ‘Aramaic-Like Features in the Pentateuch’. Hebrew Studies 47:163–76. ———. 2013. ‘Phonology: Biblical Hebrew’. In Encyclopedia of Hebrew Language and Linguistics, edited by Geoffrey Khan et al., III:100–9. Leiden: Brill. Reymond, Eric. 2014. Qumran Hebrew: An Overview of Orthography, Phonology, and Morphology. Society of Biblical Literature Resources for Biblical Study 76. Atlanta: SBL. ———. 2016. ‘The Passive Qal in the Hebrew of the Second Temple Period, especially as Found in the Wisdom of Ben Sira’. In Sibyls, Scriptures, and Scrolls: John Collins at Seventy, edited by Joel Baden, Hindy Najman, and Eibert J. C. Tigchelaar, 1110–27. Supplements to the Journal for the Study of Judaism 175. Leiden: Brill. 510 The Historical Depth of the Tiberian Reading Tradition Revell, E. John. 1988. ‘First Person Imperfect Forms with Waw Consecutive’. Vetus Testamentum 38:419–26. ———. 2015. The Pausal System: Divisions in the Hebrew Biblical Text as Marked by Voweling and Stress Position. Edited by Raymond de Hoop and Paul Sanders. Sheffield: Sheffield Phoenix. Rezetko, Robert, and Ian Young. 2014. Historical Linguistics and Biblical Hebrew: Steps toward an Integrated Approach. Ancient Near Eastern Monographs 9. Atlanta: SBL. Robar, Elizabeth. 2013. ‘Wayyiqtol: An Unlikely Preterite’. Journal of Semitic Studies 58:21–42. ———. 2015. The Verb and the Paragraph in Biblical Hebrew: A Cognitive-Linguistic Approach. Studies in Semitic Languages and Linguistics 78. Leiden: Brill. ———. 2021. ‘The Rise of wayyiqṭol’. In New Perspectives in Biblical and Rabbinic Hebrew, edited by Aaron D. Hornkohl and Geoffrey Khan, 241–74. Cambridge Semitic Languages and Cultures 7. Open Book Publishers and University of Cambridge Faculty of Asian and Middle Eastern Studies. doi.org/10.11647/OBP.0250.10 Roberts, Bleddyn Jones. 1951. The Old Testament Text and Versions: The Hebrew Text in Transmission and the History of the Ancient Versions. Cardiff: University of Wales Press. Rofé, Alexander. 1976. ‘Sipur ʿEruse Rivqa (Bereshit 24): Meḥqar Sifruti-Hisṭori’. Eshel Beer-Sheva 1:42–67. ———. 1981. ‘La Composizione di Gen. 24’. Biblica et Orientalia 23:161–65. References 511 ———. 1990. ‘An Enquiry into the Betrothal of Rebekah’. In Die Hebräische Bibel und zweifache Nachgeschichte: Festschrift für Rolf Rendtorff zum 65 Geburtstag, edited by E. Blum, C. Macholz, and E. W. Stegemann, 27–39. Neukirchen: Neukirchener Verlag. Rollston, Christopher. 2006. ‘Scribal Education in Ancient Israel: The Old Hebrew Epigraphic Evidence’. Bulletin of the Amer- ican Schools of Oriental Research 344:47–74. Rooker, Mark F. 1996. ‘Dating Isaiah 40–66: What Does the Linguistic Evidence Say?’ Westminster Theological Journal 58:303–12. Rundgren, Frithiof. 1961. Das althebräische Verbum: Abiss der Aspektlehre. Stockholm: Almqvist & Wiksell. Sáenz-Badillos, Angel. 1993. A History of the Hebrew Language. Translated by John Elwolde. Cambridge: Cambridge University Press. Sapir, Edward. 1921. Language: An Introduction to the Study of Speech. New York: Harcourt, Brace. Sarfatti, Gad Ben Ami. 1989. ‘Definiteness in Noun-Adjective Phrases in Rabbinic Hebrew’. In Studies in the Hebrew Language and the Talumdic Literature, edited by Menachem Zvi Kaddari and Shimon Sharvit, 153–67. Ramat-Gan: Bar-Ilan University Press. Schniedewind, William M. 1999. ‘Qumran Hebrew as an Antilanguage’. Journal of Biblical Literature 118:235–52. ———. 2000a. ‘Linguistic Ideology in Qumran Hebrew’. In Diggers at the Well: Proceedings of a Third International Symposium on the Hebrew of the Dead Sea Scrolls and Ben Sira, 512 The Historical Depth of the Tiberian Reading Tradition edited by Takamitsu Muraoka and John E. Elwolde, 245– 55. Studies on the Texts of the Desert of Judah 36. Leiden: Brill. ———. 2000b. ‘Sociolinguistic Reflections on the Letter of a “Literate” Soldier (Lachish 3)’. Zeitschrift für Hebraïstik 13/2:157–67. ———. 2004. How the Bible Became a Book: The Textualization of Ancient Israel. Cambridge: Cambridge University Press. ———. 2021. ‘Language and Group Identity in the Dead Sea Scrolls: The Case for an “Essene Hebrew”’. In Hebrew Texts and Language of the Second Temple Period: Proceedings of an Eighth Symposium on the Hebrew of the Dead Sea Scrolls and Ben Sira, edited by Steven E. Fassberg, 280–91. Studies on the Texts of the Desert of Judah 134. Leiden: Brill. Schoors, Antoon. 1992–2004. The Preacher Sought to Find Pleasing Words: A Study in the Language of Qoheleth. 2 vols. Orientalia Lovaniensia Analecta 41, 143. Leuven: Departement Orientalistiek. Seow, Choon L. 1996 ‘Linguistic Evidence and the Dating of Qohelet’. Journal of Biblical Literature 115:643–66. Sharvit, Shimon. 2004. Peraqim be-Toldot ha-Lashon ha-ʿIvrit: HaḤaṭiva ha-Qlasit—Yeḥida 3: Leshon Ḥazal. Tel-Aviv: The Open University of Israel. Shulman, Ahouva. 1996. ‘The Use of Modal Verb Forms in Biblical Hebrew Prose’. PhD dissertation, University of Toronto. Siebesma, Pieter A. 1991. The Function of the Niphal in Biblical Hebrew. Studia Semitica Neerlandica 29. Assen/Maastricht: VanGorcum. References 513 Sievers, Eduard, von. 1901. Metrische Studien. Abhandlungen der Philologisch-historischen klasse der Königl. sächsischen Gesellschaft der Wissenschaften 21. Leipzig: B. G. Teubner. Sivan, Daniel. 2009. ‘The Internal Passive of G-Stems in Northwest Semitic Languages’. In Mas’at Aharon: Linguistic Studies Presented to Aron Dotan, edited by Moshe Bar-Asher and Chaim E. Cohen, 47–56. Jerusalem: Bialik. Sivan, Daniel, and Elisha Qimron. 1995. ‘Alternations of pataḥ and ḥiriq in Closed Unstressed Syllables in Biblical Hebrew and the Question of the Law of Attenuation’. Leshonenu 59:7–38. [Hebrew] Sjörs, Ambjörn. 2021a ‘Notes on the Lengthened Imperfect Consecutive in the Samaritan Pentateuch’. Journal of Semitic Studies 66:17–26. ———. 2021b. ‘Notes on the Lengthened Imperfect Consecutive in Late Biblical Hebrew’. In New Perspectives in Biblical and Rabbinic Hebrew, edited by Aaron D. Hornkohl and Geoffrey Khan, 275–98. Cambridge Semitic Languages and Cultures 7. Cambridge: Open Book Publishers and University of Cambridge Faculty of Asian and Middle Eastern Studies. doi.org/10.11647/OBP.0250.12 Smith, Mark S. 1991. The Origins and Development of the WawConsecutive: Northwest Semitic Evidence from Ugarit to Qumran. Harvard Semitic Studies 39. Atlanta: Scholars. Sperber, Alexander. 1966. A Historical Grammar of Biblical Hebrew: A Presentation of Problems with Suggestions to Their Solution. Leiden: Brill. 514 The Historical Depth of the Tiberian Reading Tradition Stadel, Christian. 2013. ‘Aramaic Influence on Biblical Hebrew’. In Encyclopedia of Hebrew Language and Linguistics, edited by Geoffrey Khan et al., I:162–65. Leiden: Brill. Steiner, Richard C. 1979. ‘From Proto-Hebrew to Mishnaic Hebrew: The History of ‫ ְִָך‬- and ‫ ִָּה‬-’. Hebrew Annual Review 3:157–74 ———. 1992. ‘A Colloquialism in Jer. 5:13 from the Ancestor of Mishnaic Hebrew’. Journal of Semitic Studies 37:11–26. Suchard, Benjamin D. 2019. ‘Sound Changes in the (Pre-) Masoretic Reading Tradition and the Original Pronunciation of Biblical Aramaic’. Studia Orientalia Electronica 7:52– 65. doi.org/10.23993/store.74104 Suriano, Matthew J. ‘Tetragrammaton’. In Encyclopedia of Hebrew Language and Linguistics, edited by Geoffrey Khan et al., III:751–55. Leiden: Brill. Tal, Ilan. 2008. Lexicon of Jewish Names in Late Antiquity—Part III: The Western Diaspora 330 BCE–650 CE. Texts and Studies in Ancient Judaism 126. Tübingen: Mohr Siebeck. ———. 2012. Lexicon of Jewish Names in Late Antiquity—Part II: Palestine 200–650. Texts and Studies in Ancient Judaism 148. Tübingen: Mohr Siebeck. Talshir, David. 1986. ‘ʿAl Yiḥude Taḥbir be-Lašon ha-Miqra haMeʾuḥeret’. In Proceedings of the Ninth World Congress of Jewish Studies, part 4, vol. 1, 5–8. Jerusalem: The World Union of Jewish Studies. [Hebrew] ———. 1987. ‘Hitpatḥut Maʿareḵet he-ʿAtid ha-Mehupaḵ be-Ziqa ʾel ha-Maʿareḵet ha-Modalit’. Tarbiz 56:585–91. [Hebrew] References 515 Tigchelaar, Eibert J. C. 2018. ‘Sociolinguistics and the Misleading Use of the Concept of Anti-Language for Qumran Hebrew’. In The Dead Sea Scrolls and the Study of the Humanities: Method, Theory, and Meaning—Proceedings of the Eighth Meeting of the International Organization form Qumran Studies, edited by Pieter B. Hartog, 195–206. Studies on the Texts of the Desert of Judah 125. Leiden: Brill. Tov, Emanuel. 2012. Textual Criticism of the Hebrew Bible. 3rd revised and expanded edition. Minneapolis: Fortress. Trombley, Frank R. 2014. Hellenic Religion and Christanization c. 370–529. Volume 2. Leiden: Brill. Tropper, Josef. 2001. ‘Das genusindifferente hebräische Pronomen HWʾ im Pentateuch aus sprachvergleichender Sicht’. Zeitschrift für die Althebraïstik 14:159–72. Ungnad, Arthur. 1907. ‘Zum hebräischen Verbalsystem’. In Beiträge zur Assyriologie und semitischen Sprachwissenschaft, edited by Franz Delitzsch and Paul Haupt, 55–62. Leipzig: Hinrichs. Van der Merwe, Christo H. J., Jacobus A. Naudé, and Jan H. Kroeze. 2017. A Biblical Hebrew Reference Grammar. 2nd edition. London: T&T Clark. van Peursen, Wido Th. 2004. The Verbal System in the Hebrew Text of Ben Sira. Studies in Semitic Languages and Linguistics 41. Leiden: Brill. Vasileiadis, Pavlos D. 2014. ‘Aspects of Rendering the Sacred Tetragrammaton in Greek’. Open doi.org/10.2478/opth-2014-0006 Theology 1:56–88. 516 The Historical Depth of the Tiberian Reading Tradition Verstraete, Jean-Christophe. 2007. Rethinking the Coordinate-Subordinate Dichotomy: Interpersonal Grammar and the Analysis of Adverbial Clauses in English. Topics in English Linguistics 55. Berlin: Mouton. WO = Waltke, Bruce K., and Michael O’Connor. 1990. An Introduction to Biblical Hebrew Syntax. Winona Lake, IN: Eisenbrauns. Williams, Ronald J. 1970. ‘The Passive Qal Theme in Hebrew’. In Essays on the Ancient Semitic World, edited by J. W. Wevers and D. B. Redford, 43–50. Toronto: University of Toronto Press. ———. 1976. Hebrew Syntax: An Outline. 2nd edition. Toronto: University of Toronto Press. Wise, Michael O., Martin G. Abegg, Jr., and Edward M. Cook. 2005. The Dead Sea Scrolls: A New Translation. San Francisco: Harper. Wolters, Al. 2013. ‘Copper Scroll (3Q15)’. In Encyclopedia of Hebrew Language and Linguistics, edited by Geoffrey Khan et al., I:621–22. Leiden: Brill. Yahalom, Joseph. 1997. Palestinian Vocalised Piyyut Manuscripts in the Cambridge Genizah Collections. Cambridge University Library Genizah Series 7. Cambridge: Cambridge University Press. Yalon, Ḥanoch. 1937. Review of M. H. Segal, Grammar of Mishnaic Hebrew [Hebrew edition]. Kirjath Sefer 13:306. [Hebrew] References 517 ———. 1952. Shisha Sidre Mishna: Mefurashim bi-yde Ḥanoch Albek u-Menuḳadim bi-yde Ḥanoch Yalon. Vol. 1. Jerusalem: Bialik Institute & Tel-Aviv: Dvir. ———. 1964. Introduction to the Vocalization of the Mishna. Jerusalem: Bialik. [Hebrew] ———. 1971. Pirqe Lashon. Jerusalem: Bialik Institute. Yeivin, Israel. 1980. Introduction to the Tiberian Masorah. Edited and translated by E. J. Revell. Masoretic Studies 5. Missoula, MT: Scholars Press. ———. 1985. Masoret ha-Lashon ha-ʿIvrit ha-Mishtaqefet ba-Niqud ha-Bavli. Jerusalem: The Academy of the Hebrew Language. Yellin, D. 1924. ‘The Hippa‘el-Nif‘al Conjugation in Hebrew and Aramaic, and the Assimilation of ‫ ת‬in the Hitpa‘el Conjugation’. Journal of the Palestine Oriental Society 4:85–106. Young, Ian. 2009. ‘Is the Prose Tale of Job in Late Biblical Hebrew?’ Vetus Testamentum 59:606–29. ———. 2013. ‘“Loose” Language in 1QIsaa’. In Keter Shem Tov: Essays on the Dead Sea Scrolls in Memory of Alan Crown, edited by Shani Tzoref and Ian Young, 89–112. Perspectives on Hebrew Scriptures and its Contexts 20. Piscataway, NJ: Gorgias Press. Young, Ian, Robert Rezetko, and Martin Ehrensvärd. 2008. Linguistic Dating of Biblical Texts. 2 vols. London: Equinox. Yuditsky, Alexey (Eliyahu). 2013a. ‘Miqṣat Maʿaśe ha-Torah (4QMMT)’. In Encyclopedia of Hebrew Language and Linguistics, edited by Geoffrey Khan et al., II:644–46. Leiden: Brill. 518 The Historical Depth of the Tiberian Reading Tradition ———. 2013b. ‘Transcription into Greek and Latin: Pre-Masoretic Period’. In Encyclopedia of Hebrew Language and Lin- guistics, edited by Geoffrey Khan et al., III:803–22. Leiden: Brill. ———. 2016. ‘Jerome’s Transcriptions of Hebrew: Clarifying Versions and Relations to Hebrew Traditions’. Leshonenu 76/1–2:121–36. [Hebrew] ———. 2017. A Grammar of the Hebrew of Origen’s Transcriptions. Jerusalem: Academy of the Hebrew Language. [Hebrew] Zevit, Ziony. 1980. Matres Lectionis in Ancient Hebrew Epigraphs. American Schools of Oriental Research Monograph Series 2. Cambridge, MA: American Schools of Oriental Research. INDEX 1QHa (Thanksgiving Scroll, adjective, -ival 21, 24, 206, 246, 276, 278, 347 406 a 1QIsa (Great Isaiah Scroll), adverbial, 275 48, 51, 57–58, 70–71, 75, agreement, 21, 23–24 97, 112–13, 115–16, 119– anachronism, -istic 2, 15–16, 20, 142, 148–49, 151–53, 20, 99, 141, 207, 244, 246, 155–59, 180–81, 210, 257– 284, 286, 288, 342, 344, 60, 295–96, 327, 404, 406, 365–67, 413, 430, 435, 446, 466–68 463–64 b 1QIsa , 112, 115–16, 466 analogy, -ical 15, 103, 167–68, 175, 332, 365, 379, 386, 4QMMT, 138 e 4QPseudo-Ezekiel (4Q391), 400, 422 anthropomorphism, 267, 310 407 4QReworked Pentateuch b (4Q364), 407 a 4QSam (4Q51), 74, 112, 116– 17, 405 a 11QPs (11Q5), 48, 112–13, 115–16, 121, 326, 405 11QTa (11Q19 = Temple Scroll), 86–88, 90, 118–19, 447 accents, 11, 367 accumulation, 106, 342, 365, 402, 405–6, 419 accusative, 58, 62–63, 76, 79, 130–31 anti-Aramaic, 139 anti-language, 138 anti-vernacular, 139 aorist, 361 apocope, 140 Apocryphon of Jeremiah Ca (4Q385a), 406 Apocryphon of Jeremiah Cd (4Q389), 406 Apocryphon of Joshua (4Q522), 88, 90, 160 apodosis, 445–46, 452 Arabic, 13, 83, 107, 120, 142, 167 520 The Historical Depth of the Tiberian Reading Tradition Arad, 27–28, 31, 45, 126–29, 162 Arad letters, 27–28 Aramaic, 4, 8, 23, 36–38, 46– 49, 89–90, 94, 104, 110, article, definite, 21–26, 278, 347, 364–65, 375 assimilation, 20, 84, 86, 185, 192, 289, 294, 299, 308, 310 122–23, 134–38, 140, 147, attenuation, 34–37 150, 177–78, 189, 193, 201, Babylonian, 4, 17, 20, 35, 47, 226, 228, 254–55, 261–62, 83, 94, 99, 108–9, 135, 139, 266, 268, 272, 279, 281, 164, 270, 375, 377 294, 299, 317, 327–28, BDSS, 8, 27, 31, 79, 90, 110– 338–39, 341–42, 361–62, 16, 119–20, 138, 141, 148– 366, 456 49, 164–65, 175–76, 219– Biblical, 17, 20, 122–23, 20, 247, 258–59, 285, 295– 150, 177, 303–4 97, 356, 385, 389–90, 392, biblical DSS, 122–23 398, 404–6, 413, 435–37, Dead Sea Scrolls, 150, 177 445, 447, 458, 466–68 Jewish Babylonian, 193 Ben Asher, 83–84, 86–87 Jewish Palestinian, 135, 193 Ben Naftali, 83–84, 86, 88 Judaean Desert, 122 Ben Sira, 28, 149, 165, 176, non-biblical DSS, 122 196, 228–29, 247, 263, 265, Qumran, 122–23, 268 274–76, 296, 302, 385, Samaritan, 193 389–90, 392, 398, 411, 413, Second Temple, 468 436–38, 451, 469, 477 Targumic, 15, 17, 29–31, Ben ʿUzziʾel, 83 46, 88, 150, 177, 255, Beth Shearim, 37 290–91, 293, 304, 310, bgdkpt, 18, 20 327, 339 burial epitaph, 36 Aramaism, 69, 338, 341–42 Byzantine Period, 2, 6, 65, 93, archaic heterogeneity, 424 367, 376, 413, 457, 463 archaism, 30 Cairo Geniza, 89, 274 Canaanite script, 47 Index Chronicles, 9, 52, 186, 206, 256, 404–5, 445 chronolect, 4, 103, 137, 183, 521 convention, 3, 18, 34, 41, 47, 49, 51, 83, 94, 98, 135, 154, 169, 178, 208, 244, 281–82, 203, 225, 264, 303, 305, 288, 296, 330, 333, 344, 343, 412, 418, 445, 473, 367, 373, 406, 418, 420–31, 476 442–43, 451, 461, 463–64 classical opposition, 290 convergence, 24, 140, 425–26 clitic, 21, 23, 25–26, 57, 126 Copper Scroll (3Q15), 138 Codex Aleppo, 68–69, 83, 270 copyist, 162, 358, 415, 421 Codex Kaufmann, 28, 109, core LBH (corpus), 445, 465 150, 165, 177, 197–98, 298, corruption, 13, 65, 220, 459 363–64 dagesh, semantic, 375, 383 Codex Leningrad, 46, 68, 83, 143, 270 Daniel, 49, 186, 349, 354, 402, 445 coins, 18 definiteness, 21 complement, 131 Deir ʿAlla, 140, 162, 203 complementary distribution, deity, 46, 49, 51–52, 55, 267, 23, 218 310 conflation, 7, 71, 93, 242, 379 determination, 23 conjunction, 443 Deuteronomy, 408–10, 456 consonantal text, 1–2, 4, 9–10, diacritics, 11, 89 13–15, 24, 38–39, 51, 137, dialect(al), 4, 36, 69, 103, 139–40, 142, 183, 195, 206, 121–22, 137–40, 150, 193, 238, 242, 279, 330–31, 334, 201, 299, 343, 390 358, 419, 457, 460 contemporisation, 296, 417, 420, 466 contextual form, 16, 270, 387 contraction, 27–29, 45, 84, 335 diphthong(al, -isation), 17–18, 33–34, 168, 335, 432, 465 direct object, 63, 79 direct object marker, 55, 76, 143 directional heh, 16, 365 522 The Historical Depth of the Tiberian Reading Tradition disambiguation, 96, 131, 133, 140, 190, 192, 209, 217, 260, 269, 375–77, 383 verbal, 102–3, 133, 148–52, 179 epigraphy, 3–4, 16–17, 28–33, discourse, 340, 387 38, 45, 49, 110, 120, 124, discourse dependency, -dent 126, 129, 132–33, 149, 164, 378–80 dissimilation, 34 diversity, 4, 7–9, 11, 22, 33, 93, 104, 123, 141, 145, 154, 167, 171, 173–74, 177–78, 176, 207, 282, 312, 318, 385, 390, 464, 470–71 Esther, 172, 186, 211, 445 euphemism, -istic, 6, 39, 67– 68, 76–79, 469 204–6, 225, 236, 273, 337, event time, 346 342, 345, 358, 385, 409, excreta, 67, 69 419–20, 431, 466, 473, Exile, 202, 294, 338 478–79 Exodus, 173 divine referent, 56 DSS, 11, 18–20, 25–28, 32, 38, extra-biblical confirmation, 290 47–48, 51, 65–66, 74, 78– Ezekiel, 173 79, 89–90, 94, 97–98, 110– Ezra, 50, 96, 143, 186, 211, 11, 116, 119–23, 126, 136, 214, 248–49, 254, 291, 349, 138–42, 148, 151–53, 161– 402, 433, 435–37, 439, 445 62, 167–68, 180, 183, 260, focus, argument, 131 276, 280, 298, 325, 346, Former Prophets, 49, 400, 356, 367, 410 Elephantine, 49 424–25 gemination, 27, 35, 46, 84, 86, elision, 30 89, 125–26, 192, 195, 202, Elohistic Psalter, 52 266, 270, 299, 302, 333, ending, 17, 27–29, 101, 103–6, 373, 375–76, 379–80, 382– 109, 111, 116–21, 123, 126, 129, 133, 145, 147–50, 155, 159, 171, 175, 177–78, 469, 471, 473 83, 424 gender, 21, 140, 151, 163, 166, 270 genealogy, 218 Index 523 Genesis, 173, 203, 312, 324 gloss, 69, 442, 459 Genesis Apocryphon (1Q20), glottal stop, 94–99, 168 122 genre, 4, 152, 154, 337–38, grammatical treatises, 1 Greek, 17, 35–37, 47–48, 56– 343, 402, 418, 445, 469, 65, 68, 70–78, 88, 107–8, 473 135, 146, 149, 311, 356, gentilic, 27–29, 469, 471 359–62, 365, 375, 377, 382 Gezer Calendar, 33 guttural, 95–96 gizra hapax legomenon, 69, 233–34 geminate, 88, 106, 144 harmonisation, 225, 260, 296 I-y, 20, 40, 241, 330–32, Hebrew 335, 340, 346, 377, 441, ABH, 419–20 449–51, 453, 460, 467– Amoraic, 253, 265, 276, 72 I-n, 106, 144, 195 II-guttural, 270–71 II-w/y (hollow), 20, 106, 144, 216, 237, 350, 354, 277, 294, 296–97, 302–3, 311, 334, 452, 477 BDSS, 261, 293, 295, 297, 300, 325, 329, 412, 446, 456, 466 362, 386, 391, 393, 395, Ben Sira, 20, 38, 139, 183, 397–400, 404, 408, 427, 196, 207–8, 215, 223, 430, 432–33, 438–39, 225, 237–38, 253, 262, 466–70, 472 288, 292, 297, 303, 326, II-y, 213, 237, 473 III-ʾ, 144, 273 III-y, 31, 93, 101, 106, 143, 412, 451 Biblical, 4, 9, 19–21, 23–24, 27, 58, 69–71, 76–77, 93, 241, 273–74, 321, 350, 98, 104, 107, 130, 132, 386, 388–93, 395, 397– 136–38, 161, 163, 172– 400, 406–8, 413, 417–18, 73, 175–76, 184, 196, 420, 426, 429, 437 203, 207, 211, 214–15, III-r, 273 glide, 167 223, 226, 228, 230–35, 238–39, 246, 254–56, 524 The Historical Depth of the Tiberian Reading Tradition 258, 261–65, 268, 270, extra-biblical, 45, 98, 104, 276, 283, 288, 290, 296– 139, 176–77, 238, 246– 97, 303–5, 308–9, 312– 47, 253, 341, 385, 387– 14, 319, 321, 323–24, 90, 413, 416 327, 329, 333–34, 337, First Temple, 3, 32, 133, 339, 341, 343, 346–47, 139, 141, 154, 198, 208, 354, 356, 363, 366–67, 247, 279, 318, 331, 463, 373, 376, 379, 385, 425, 465, 468 441, 452, 467, 468, 470, inscriptions, 164 475, 477 Iron Age, 282, 312, 425 CBH, 3, 8–9, 12, 16, 20, 24, LBH, 8–9, 20, 23, 25–26, 26, 38, 49–50, 52, 79, 38, 52, 59, 107, 123, 107, 123, 184, 186, 198, 140, 183–87, 197–98, 203–4, 206–14, 217–19, 207–20, 235–36, 241, 223, 226, 235, 244–45, 244, 246–47, 253–54, 253, 279, 282, 284, 284, 256–57, 262–64, 279–81, 286, 288, 290, 292–94, 288, 290–94, 296–97, 296, 303, 313, 324, 328, 302–4, 313, 324, 326–27, 335, 344–45, 348, 351, 330, 335, 338, 348, 350, 353–56, 366–68, 370–72, 353, 355, 361, 365–67, 379, 384, 390, 396, 410, 371–72, 379, 390, 392, 412, 414–21, 424–25, 410, 419–20, 424, 427, 428–29, 432, 434, 442– 432, 435–38, 443–44, 45, 447, 451–53, 460, 465, 470, 473, 477 465, 470, 472–73, 479 LBH+, 217, 241–43, 245 1 247, 390, 392, 393, 401– 2 CBH , 420–21, 473 3, 406, 410, 412–20, 424, DSS, 20, 26, 38, 47, 79, 428–36, 438–39 CBH , 420–21, 473 164, 183, 187, 197, 199, NBDSS, 292, 295, 300, 303, 207–8, 219, 288, 293, 308, 325, 329, 331, 337, 302, 346, 367, 466 342–43, 412, 466, 477 Index non-biblical, 174 525 Samaritan, 31, 35, 38, 85, non-LBH, 418 95, 109, 139–40, 147, non-LBH+, 401, 403 150, 152, 161, 183, 188– non-Tiberian (biblical), 18, 90, 192–95, 201, 207–8, 146, 174, 413 Paytanic (piyyuṭ), 93, 108, 253, 263, 290, 293 222–23, 226–27, 237, 251, 256–57, 262–63, 265, 268–70, 272–74, post-biblical, 281, 477 280, 288, 293, 295, 297, post-exilic, 256, 284, 286, 299–303, 308, 314, 316, 293, 426, 443 pre-Tiberian, 36–37, 188, 194–95, 380 Qumran, 11, 86–87, 90, 331, 334, 340, 343, 370– 71, 412, 448–49, 469, 475–76 Second Temple, 3, 14–15, 138, 142, 184, 207, 253, 25, 123, 142, 169, 178, 265, 295, 366, 370, 445, 188, 202, 207–8, 222, 466 229, 234, 238, 244, 265, Rabbinic, 19–20, 22, 28, 32, 38, 71, 77–78, 93–95, 99, 104, 108–9, 117, 135–40, 147, 150, 161, 165, 177, 183–84, 192– 93, 197–201, 207–8, 215, 223, 225–26, 228, 230– 34, 237–39, 247, 253–55, 257, 259, 261–63, 268, 270, 277–81, 288, 291– 93, 296–98, 300, 302–3, 305, 308, 311, 316, 366, 370, 452, 468, 470–71, 476, 477 278, 283, 286, 288, 297, 303, 311, 317, 329, 331, 337, 344, 372, 383, 417, 419, 442, 457, 465, 468, 470, 475 Tannaitic, 38, 79, 138, 165, 231, 253, 268, 276–77, 294, 303, 311, 325, 334, 452, 477 TBH, 348, 353, 365, 367, 419–21, 473 Hebrew letter ʾalef, 7, 46, 93–99, 367 heh, 7, 30, 57, 65, 94, 101, 104, 106–7, 110–11, 119, 526 The Historical Depth of the Tiberian Reading Tradition 121, 124, 128, 199, 236, 269, 289, 465, 467–69, 381, 392, 417, 424 471–73 kaf, 101 historical etymology, 95 ṣade, 298 homonym, 58 tav, 95, 202, 266, 294, 317 homophone, 375 ṭet, 298 idiosyncrasy, 102 waw, 77, 94, 162, 165–66, impersonal, 62, 194 176, 245, 258, 332, 335, infirmity, 67 363, 373–76, 378–79, infix, 193, 289, 294, 299, 310, 382–83, 397 yod, 7–8, 27, 32–33, 37, 45, 317 inscription, 17, 27, 30, 33–34, 94, 126, 146, 150–51, 36–37, 49, 97–98, 124, 161–62, 165–66, 243, 126–27, 130, 132, 137, 139, 245, 456, 458 164, 203, 391 Hebrew vowel ḥaṭef, 46 interrogative, 454 Iron Age, 2–3, 15–16, 26–28, haṭef qameṣ, 298 31, 33–34, 38, 45, 49, 93– ḥiriq, 162 94, 97–99, 110, 120, 123– ḥolam, 31, 46 24, 126, 129, 132–33, 137, pataḥ, 351, 375 139, 141, 149, 169, 176, qameṣ, 101, 103, 298, 350, 178, 188, 203–4, 207, 221, 351 278, 288, 318, 372–74, 376, segol, 350 379, 385, 390, 424, 434–35, shewa, 19, 46, 68, 84–85, 457, 463–64, 470–71 101, 298, 373 Hellenistic period, 36, 91, 367, 372 Iron Age epigraphy, 16, 31, 45, 110, 126, 149, 176, 385, 390, 470 Hexapla, 35–36, 135 Isaiah, 401, 466 hifilisation, 200, 209, 211–13, Islamic Period, 107, 142 218–20, 222–29, 234–36, 238, 243, 245–247, 253, Jerome, 7, 31, 35, 84, 96, 99, 139, 149, 176, 323 Index Job, 403 narrative framework, 348, 403 Judaean Desert, 123, 147, 175, 177, 466 Karaite, 13 Ketef Ḥinnom, 31, 34, 125 ketiv, 1, 4–10, 12–13, 30, 32, 527 lectio difficilior, 260 lectio facilior, 58 letter shape, 41 levelling, 4, 46, 109, 121, 141, 152, 174, 178, 193, 225, 420, 422, 477 lexeme, 69–71, 210, 375, 410, 451 39, 45, 67–69, 71–72, 74– lexicalisation, 23 79, 102–3, 105–6, 122–23, lexicon, -cal, 72, 98, 222, 338, 139, 144–46, 148, 152, 155, 469 161, 167, 200, 204–6, 215, literalisation, 7 238, 257, 272, 331, 335–36, liturgy, 108, 140, 286 388, 433–36, 438, 467 LXX, 35, 90, 242 Khirbet Beit Lehi, 17 main clause, 324 Kings, 52, 172, 237, 400 majority (orthography, Kitāb Al-Khilaf (= Sefer haḤillufim), 83 pronunciation), 140 majority (orthography, Kuntillet Ajrud, 129, 162 pronunciation, structure), 4, Lachish, 33, 45, 126–32, 162, 14, 31, 34, 102, 104–6, 126, 203 131, 145–46, 154, 161, 171, Lamentations, 50 174, 178, 225, 331, 339, language contact, 104, 136, 345, 350, 402, 479 317, 338 late distribution, 198, 256, 290, 341 Latin, 7, 17, 35, 48, 58, 61, maqqef, 351 Masada, 223, 275 Masoretic Text, 8, 11, 32, 34, 36, 48–51, 65, 90, 93, 97, 71–78, 88, 90, 96, 99, 108, 109, 115, 118–19, 121, 125, 149, 176, 327, 359–62, 365, 136, 144–45, 148–49, 155– 375, 377, 383 59, 161, 175–76, 180–81, Latter Prophets, 49, 401–2 187, 190–94, 197, 199, 203, 528 The Historical Depth of the Tiberian Reading Tradition 212, 214, 216–20, 222–28, 246, 319, 333, 337, 339, 230–32, 241–43, 247, 251, 344–46, 435, 464, 470–72, 256–63, 265–67, 269, 272– 478–79 75, 280–81, 284–85, 293, Mishna, 28, 150, 165, 177, 295–97, 300–3, 305, 310– 181, 197, 215, 255, 280, 11, 330, 332, 340, 344, 356, 292, 303, 363, 477 359, 362, 370–72, 389–90, modality, 441 392, 397–99, 401–3, 405–8, agent-oriented, 441, 454 412, 414–18, 420–21, 424, speaker-oriented, 441–42 427, 429, 435–38, 446–50, monophthong(isation), 17, 30 456, 458–59, 466–69, 471– morphology, -ical, 8, 20, 40, 72 mater lectionis, 1, 11, 32–33, 56, 66, 103, 117, 119, 132– 33, 137, 139, 145–54, 163, 41, 77, 94, 98, 101, 124, 166, 171, 174, 178, 187, 127, 146, 150–51, 162, 189, 191–93, 195, 199, 205, 241–43, 245, 335, 363, 391, 209, 215, 217, 219, 221, 397, 424, 456, 458–60 227, 236–37, 241–44, 247, medieval period, 2, 6–7, 18, 258, 262, 267–68, 271–74, 45, 47, 52, 65, 99, 136, 247, 279, 281–83, 285, 287–88, 274, 296, 367, 413, 457–58, 292, 299, 305, 308–9, 321, 463–64, 469 329, 332, 374–75, 381, 383, Meṣad Ḥashavyahu, 33, 129 Meshaʿ Stele, 162, 241, 312, 390, 413, 419 385–87, 390, 392–95, 397– 400, 402–4, 406–10, 412– 13, 415, 417–22, 424–25, Migdalenoi, 36 427–34, 453, 458, 466, minority (orthography, 475–76, 478 pronunciation, structure), 3, consonant-final, 102, 104–9, 15, 31–32, 34, 38, 49, 88– 120–22, 124, 127, 129, 89, 97–99, 104–7, 109, 124, 133, 140, 145, 147, 149, 133, 137–39, 145, 150, 152, 150–55, 159, 171, 174– 154, 165, 167, 178, 243, 75, 177–78, 391, 478 Index 529 verbal, 176–79 oblique, 131, 163 vowel-final, 41, 102, 104, onymisation, 23 106, 109, 120–23, 126, Origen, 35–36, 135 129–30, 132–33, 140, orthography, -ic, 1, 8–11, 13, 145–47, 149–54, 171, 17–18, 27, 30–33, 41, 56– 173–76, 178–79, 391, 57, 59, 60, 79, 83, 85, 89– 478 91, 93–94, 96, 99, 101–5, Moshe Moḥe, 84 118, 120, 125–26, 129–30, Nehemiah, 172, 186, 402, 445 134, 136–37, 139–40, 151, neutralisation, 168 162, 165, 168, 171, 173–74, nifalisation, 183, 185, 188, 176, 178, 191–92, 194, 197, 190–91, 194, 196–98, 200, 199, 225, 235–38, 242–44, 202–3, 207–8, 246, 253, 247, 253, 282, 284–85, 289, 466, 468–69, 471–72, 287–300, 305, 307–8, 330, 475 334, 367, 377, 386, 390–91, nominal clause, 455 393, 397–99, 415–16, 418, non-LBH+ Writings, 392, 396, 427–30, 433–34, 456, 463, 401–3, 410, 412, 414, 428 non-Masoretic (traditions, sources), 7, 388, 414 noun, 8, 21, 23–24, 30, 45, 69, 71, 147, 239, 254, 280–81, 319, 337, 352, 364 468, 474–78 defective, 11, 27, 37, 41, 101–2, 121, 126, 151, 153, 162, 241, 393, 397– 98, 424 plene, 37, 41, 101, 108–9, common, 22–23 126, 128, 130, 142, 223, proper, 11, 23, 31, 83, 473 241–43, 245–46, 251, number, 21, 191 364, 393, 397–98 numeral, ordinal, 21 paragogic nun, 459–60 object, 1, 55–56, 58, 76, 101, parsimony, 270, 273 103, 109, 122, 125, 130–31, 175, 192, 213, 232, 268–69, 304, 437 pausal form, 16–17, 45, 105, 136, 138, 140 530 The Historical Depth of the Tiberian Reading Tradition Pentateuch (Torah), 6, 41, 49– 52, 67, 108–9, 155, 161–69, polysemy, -emous, 223, 269, 375–76, 383 173, 178, 185, 216–17, 219, post-classicism, 451 225–26, 241–47, 256, 291, pragmatic(s), 130–31, 378, 332, 389–90, 392, 395–400, 408, 410–20, 423–27, 429– 31, 435–38, 468–69, 471– 74, 478 Persian Period, 218, 242, 367, 372, 427 Peshiṭta, 23, 58, 292, 311, 358–59, 361–62 Phoenician, 27, 29 phonology, -ical, 17, 20, 28, 30, 32–33, 39, 46, 83, 93, 387, 442 preformative, 266, 332, 334, 373, 378 preposition, 18–21, 23, 25–26, 46, 55, 57–58, 61–62, 65, 76–77, 93, 98–99, 131, 138, 143, 353 Pre-Secunda, 36 pronoun epicene, 41, 163, 166, 168 independent subject, 41, 95, 104–5, 119, 129, 133, 101, 105–6, 108–9, 123, 137, 140, 174, 185, 195, 139–40, 145, 147–52, 271, 383, 387, 431–32 161–63, 165–66 phylactery, 113, 116, 119, 121 pronunciation, 1–5, 9–11, 13– pielisation, 200, 253–54, 256, 16, 18, 27, 33–35, 37–40, 262, 265, 267, 269, 271–72, 45, 49, 51–53, 57, 61, 64, 274, 276, 278–79, 281–85, 83–86, 89–90, 94, 96–100, 288–89, 467–69, 471–72, 102, 105–8, 121, 123, 126, 476–77 138, 141–42, 151–52, 161, plural, 8, 27, 32–33, 45–46, 64, 72, 335, 431 plural of majesty (pluralis majestatis), 45, 52 poetry, 21, 25, 153, 214, 259, 164–65, 167–69, 174, 178, 198–99, 209, 224, 238, 242, 271, 299, 314, 334, 352, 373, 383, 431, 448, 457, 459, 463, 468, 470–71, 476, 296, 319, 324, 333, 336, 478 338, 402, 418 Palestinian, 135 Index Prophets, 161–62, 185, 217, 241, 243, 245–47, 390, 392, 396, 400, 403, 410, 412, 531 331, 335–36, 386, 388, 433–35, 438, 467, 469, 471 qere perpetuum, 4–6, 9, 40–41, 414, 416–18, 420, 427, 83, 102–3, 161–62, 164, 430–31, 435–38, 474 467, 469, 471 prose, 153, 336, 338, 389, 441 prosody, 107, 119, 174, 387 Qohelet, 52, 186, 219, 294, 402, 444 protasis, 443, 452 rabbinic literature, 6, 18 Proto Indo-European, 365 radical, 93, 95, 153, 185, 192, proto-Masoretic, 135, 162, 340, 374, 384 195, 201, 266, 270–71, 273–74, 302, 333 proto-Samaritan, 51, 340 rafe, 83 Proto-Semitic verbal form rape, 67, 75–76 qaṭil, 271 reference time, 322, 324, 346 yaqtul, 321, 324, 378, 385– register, 4, 103, 121, 134, 86 137–38, 140, 343 yaqtula, 386, 398–99 register, 121, 138 yaqtulan(na), 386 Reichenbach, 322 yaqtulu, 321, 324, 386, 398– relative clause, asyndetic, 347 99, 426, 466 relative tense, 321, 326, 346 proto-Tiberian, 51, 133, 141 result clause, 380 Proverbs, 172, 215, 324 Reworked Scripture, 447 Ps. 119, 403, 405 Rewritten Bible, 447 Psalms, 324, 402–3, 418 Ruth, 173 qere, 1, 4–10, 12–13, 30, 32, Samaritan Pentateuch, 50–51, 39–41, 67–72, 74–75, 77– 56, 57, 59–63, 65–66, 72, 79, 83, 102–3, 105, 122–23, 75, 144, 147, 165, 175, 188, 139, 144–46, 148, 153, 155, 191, 222–28, 242, 247, 251, 161, 167, 200, 205–6, 215, 266–67, 269, 272, 285, 311, 238, 248, 257, 270, 272, 332, 356, 359, 385, 389–90, 392, 397–99, 408–10, 413, 532 The Historical Depth of the Tiberian Reading Tradition 432, 436–38, 449, 454–55, cohortative, 384 458 conditional, 454 Samuel, 173, 324 directive, 441, 449 scribe(s), 7, 13, 91, 129–30, directive-volitive, 379, 423 132, 140, 162, 415, 417 Second Temple Period, 3, 14– eventive, 206–7 factitive, 275 15, 18, 38, 49, 51, 65–66, final, 441, 454, 456 77, 79, 90, 97, 120–21, 123, future, 41, 319, 320–24, 133, 139, 141, 178–79, 183, 327, 329, 345, 380, 382, 198, 204, 235, 280, 366, 441–43, 445–46, 454, 372, 374–75, 424, 434, 456 457–59, 463, 465 habitual, 426, 442, 445 Secunda, 31, 36, 139 habitual past, 319, 455 semantic shift, 78, 189, 315 imperfective, 358, 360, 362, semantic(s), 131, 183, 190, 445 198, 211, 218, 229, 246, imperfective past, 358 262, 264–65, 268, 292, 297, inchoative, 211, 231 303–4, 308, 313, 319, 323, indicative, 110, 124, 241, 329, 358, 360–61, 373–74, 243, 282, 361–62, 434, 376–80, 426, 446, 475 441 absolute future, 329, 345– 46 absolute past, 328, 343, 345–46 intransitive, 183, 188, 203– 4, 207, 212–13, 220, 225, 230–33, 235, 248, 261, 269, 274–75, 475 actional, 207 irrealis, 378–79, 381 active, 194, 281, 287, 301 iterative, -ity, 263, 275 causative, 60, 222, 229, jussive, 240–41, 243, 373, 231, 239, 257, 263 cohortative, 379, 399, 403, 409–10, 413, 416, 418, 424, 426, 434, 436–37 376, 386, 425, 441, 445, 454 Index medio-passive, 183–84, 194, 196–97, 203–4, 207, 267, 317 middle, 203, 254, 307 modal, 319–20, 323, 373– 533 purpose, 60, 373–74, 378– 83, 424–25, 441–42, 444 realis, 379–81, 381, 383, 424 reflexive, 183, 185, 188, 74, 377–79, 383–84, 203, 232–33, 261, 264, 423–26 289, 292, 304–5, 307, nominal, 195, 206, 215, 246, 264, 276 non-past, 379 317 relative future, 321–24, 326–28, 343, 345–46 non-perfective, 376, 379 relative past, 345–46 non-preterite, 374–77, 380 result, 378, 380–82, 441 passive, 185, 204, 207, simple past, 455 231–32, 234, 254, 265, stative, 95, 183, 187–88, 267, 281, 287, 289, 293, 192, 209–13, 220, 231, 301, 307, 316, 337 255, 271, 279–80, 319, past tense, 321, 327–28, 363, 365, 476 360–61, 468, 470, 472, subjunctive, 456 479 substantival, 264–65 perfective, 321 perfective past, 320–21, 323–25, 327–32, 339, temporal posteriority, 378, 380 transitive, 188, 192, 203, 358, 360, 373–74, 381– 211–13, 225, 230–33, 82, 442–43, 445 248, 261–62, 264, 274– permansive, 365 75, 292, 304, 308 pluperfect, 320, 341, 361 verbal, 195 pluractional(ity), 200, 263, volitive/volitional, 375–76, 268, 275 preterite, 320–21, 324, 332, 373–81, 383–84, 424–26 378–81, 422–23, 426, 441 weakly transitive, 264 sibilant, 36, 85, 88–91 534 The Historical Depth of the Tiberian Reading Tradition Siloam inscription, 97–98, 203 subordinate clause, 322–23 singular, 8, 45, 68, 101, 163, suffix, 28, 32, 34, 77, 101, 166, 191, 268, 431 104–11, 113, 116, 118, 120, sound shift, 35 122–26, 130–31, 134–35, source (Documentary 137–38, 140, 145, 147–50, Hypothesis), 173 155–56, 159, 192, 195, 236, E, 52, 173 280, 327–28, 333–34, 338– J, 173, 218 40, 459 P, 173, 218, 256 nominal, 146–52 speech time, 320, 322–23, 346 object, 79, 129, 130, 132, spelling, 1, 5, 8, 10, 16–18, 27, 140–41, 145–46, 175 31–34, 37, 41, 45, 55, 57– possessive, 8, 32, 45, 103, 58, 64, 66, 83, 85, 88–90, 93–95, 97–99, 102, 104–11, 113, 115–16, 118–24, 126– 122, 140, 141, 145, 473 pronominal, 135 suppletion, -tive, 23–24, 26, 30, 132–38, 140–41, 150– 183–84, 187, 189–95, 199, 51, 153, 161–64, 166–67, 209, 224, 226, 235, 240, 169, 178, 184, 186, 195, 244–45, 263, 267–68, 281, 200, 202–3, 226, 237, 242– 283, 306–7, 309, 313, 318, 46, 279, 300, 306, 335–36, 474–76 366, 367, 390, 393, 397–99, syncope, 27–29, 57, 94, 98–99 424, 427–28, 430–32, 458, synonym(-y, -ous), 68–69, 203, 465–66, 471, 475–78 standardisation, 4, 15–16, 23, 210, 263, 276, 312–13, 317–18, 328 29, 34, 37, 100, 107, 124, syntactic structure, 319, 338 133, 140–41, 153, 167, 169, syntagm, 21, 335–39, 342, 178–79, 434–35 stress, 95, 99, 132, 350, 354, 362, 387 ultimate, 95 style switching, 338, 342 344, 350, 358, 361–62, 364, 366–67, 371–73 syntax, -actic, 19–20, 24–26, 58–59, 204, 208, 262, 319, 338, 341, 378, 452 Index Syriac, 17, 23, 48, 56–57, 59– 535 translation(s), 7, 24, 35, 48, 65, 68, 70–78, 88, 90, 122– 60, 62, 65, 69, 71, 78, 307, 23, 140, 146, 150–51, 153, 320–21, 356, 358, 361 177, 189, 268, 291, 293–94, translator, 358, 362 327, 360–61, 456 transmission, 10, 137, 207, Talmud, 6, 67, 442 415 Bavli, 215, 234, 294, 296 triphthong(al), 17, 33–34 Yerushalmi, 215, 294, 296 Ugaritic, 27, 69 TAM (Tense-Aspect-Mood), 319, 324, 361, 377, 382, 385, 442 Targum(s), 17, 23, 46, 61, 74, univerbalisation, 18–20, 465, 469, 471–72 valency, 190, 268 verb, 55, 61–62, 66, 75, 77, 78, 189, 193, 255, 261, 267, 79, 109, 125, 130–32, 150, 310–11, 327, 356, 359–61 187, 190–92, 194, 200, 219, Jerusalem (i.e., Pseudo- 222, 224, 227, 238–39, Jonathan), 261 Jonathan, 57–58, 60, 64, 254–55, 258, 264, 268, 272–73, 275–76, 279–81, 68–77, 146, 311, 327, 284–85, 288, 308, 319, 362 324–27, 330, 332, 336–37, Onqelos, 56–57, 59, 61–63, 340–41, 358, 366, 380, 425, 65, 72, 74–75, 194, 242, 441, 444, 449, 453, 474–77 268, 270, 310–11, 327 weak, 386 tetragrammaton, 5–6, 39, 45– 46, 52, 469, 474 Tetrateuch, 408–10 verbal adjective, 365 verbal form active participle, 69, 189– textual fluidity, 332, 413 90, 195, 199, 206, 215, textualisation, 45 224, 230, 240–41, 246, toponym, 16–17, 281, 465 253, 259, 262–72, 274, transcription, 31, 35, 96, 99, 276, 279, 280–87, 305–6, 107–8, 135, 138, 149, 176, 374–75, 377, 383–84 309–10, 313, 335, 341, 536 The Historical Depth of the Tiberian Reading Tradition 347, 350, 354–56, 358– 64, 366, 369–71, 476–77 pseudo-cohortative, 381, 386–96, 399–414, 416– cohortative, 422–23, 454 22, 424, 426, 429, 432– conversive, 442, 445, 447 33, 435–38, 473 imperative, 171–73, 175– qaṭal (suffix conjugation), 78, 189, 195, 199, 203, 40, 101, 106, 109, 115– 215, 222, 236, 240, 280, 17, 119, 123, 127–30, 308, 313, 422, 425, 441, 132, 145–46, 149–50, 454 186, 189–91, 193, 195, infinitive absolute, 31, 200, 198–202, 205, 215, 224, 204, 238–39, 280, 425, 236, 238, 268–69, 279– 476 80, 284, 294, 299, 306, infinitive construct, 8, 18– 313, 317, 319, 323–48, 20, 57–58, 65, 93, 95–96, 350–56, 358, 361–72, 98–99, 143, 155, 171, 377, 381, 442–43, 445, 186–87, 189–91, 195, 450–51, 459, 466, 468– 197, 199, 203, 206, 211, 70, 472, 479 216, 224, 236–37, 240– 41, 260, 262, 267–69, 275, 280, 284, 313, 319, waw-yiqṭol, 373–74, 376, 382–84, 424, 437 wayyiqṭol, 11, 176, 237–38, 325–27, 329, 335, 455, 241, 331, 332, 335–36, 465, 469, 471–72, 475, 373–87, 389, 390–91, 477 393–95, 397–99, 402–4, jussive, 423–24, 454 406–10, 412–17, 419–32, non-conversive, 442–43 434–35, 437, 442, 445, passive participle, 4, 194, 447, 450, 452, 466–73, 205, 222, 224, 237, 253, 478 255–58, 262, 265, 268, (way)yiqṭol, 171, 244–45 274–75, 280, 283, 285– weqaṭal, 64, 132–33, 224, 88, 304–5, 334, 347, 477 441–42, 444–50, 452–56, 458–60 Index we-qaṭálti, 443 we-yiqṭol, 64, 373–74, 376, 379, 381–84, 437, 441– 54, 456, 458–60, 467–72 537 verbal stem (binyan), 40, 183, 241, 265, 270, 289, 474 C-stem (hifʿil), 59–60, 64– 66, 69, 106, 144, 203, X-yiqṭol, 454 209–51, 254–55, 257, yaqṭel, 210, 238–40 259, 261–62, 265, 268– yiqṭol (prefix conjugation), 69, 275, 286, 291, 298, 19, 40, 171–78, 189–93, 365, 386, 391, 393, 395, 195, 199–201, 210, 213, 397–400, 404, 407–8, 215, 220, 222, 224, 236– 427, 430, 432, 437–39, 39, 241–43, 259, 262, 466–70, 472–73 266, 268–69, 271, 273– C-stem passive (hofʿal), 185, 74, 276, 279–80, 283–84, 200, 215, 234, 237–38, 299, 306, 313, 319–36, 255, 262, 334 338–39, 341, 343–47, D-stem (piʿʿel), 77, 197, 373–76, 378–79, 381, 223–24, 226, 253–77, 383, 385–86, 398–99, 279–88, 291, 301, 304, 409, 424–26, 441–43, 308–11, 333, 334, 370– 449, 452, 458, 466, 468, 71, 476–77 470, 472, 475–76, 479 verbal form (Aramaic) infinitive construct, 280 verbal noun, 281, 365 verbal stem (Aramaic) D-stem (paʿʿel), 255, 261, 266, 281, 284 Dt-stem (etpaʿʿal), 189, 193– 94, 290, 292–93 G-stem (peʿal), 228, 261, 266 D-stem B (piʿel B), 190, 266, 270–74, 280 D-stem B passive (puʿal B), 266 D-stem passive (puʿʿal), 77, 185, 200, 254–59, 261– 66, 274–76, 278, 280–83, 285–86, 289, 291, 296, 304–5, 308–9, 333–34 Dt-stem (hitpaʿʿel), 192–93, 202, 232, 254, 261–62, 538 The Historical Depth of the Tiberian Reading Tradition 267, 278, 285, 289–94, 296, 298–301, 304–9, 296–99, 302–18 312–18, 334–36, 350–51, 354, 360, 370–71, 465, G-stem (qal), 18–20, 31, 55– 468–69, 474–75 66, 77, 183–85, 187–200, 203–4, 206, 209–51, N(t)-stem (nifʿal B), 192–94, 253–77, 279–93, 296, 201–2, 266, 299–301, 298–301, 304–5, 308, 308, 316 331–36, 350, 354, 362– Nt-stem (nitpaʿʿal), 192, 63, 370–71, 386, 391, 197, 201, 278, 289, 303, 393, 395, 397–99, 404, 305, 308, 316 427, 430, 432–33, 437– N-stem passive (nufʿal), 186, 206–7 49, 451, 453, 460, 465, 468, 473–77 G-stem (qal) internal passive, 77, 79, 183–87, verbal system, 163, 166, 445– 46 vernacular, 22, 29, 65, 98, 194, 196, 198–200, 203, 104, 121, 134–36, 138, 140, 204–8, 215, 224, 232, 329, 430 238, 255–58, 262–63, versions, ancient, 6–7, 62, 64, 269, 283, 304–5, 333–36, 69, 71–72, 76, 78, 356, 465, 468 360–61, 367 G-stem (qal) stative, 210, 212–13 G-stem B (qal B), 266, 268– 69, 271, 273–74 vocalisation (niqqud), 1–2, 4–5, 7–8, 11, 13–17, 20, 24–26, 30, 38, 40–41, 46, 52, 56, 60–61, 66, 76, 83, 85–87, Gt-stem, 69 93, 101, 105, 109, 122–23, N-stem (nifʿal), 55–59, 61– 133, 135, 137, 141, 162, 66, 77, 79, 183–208, 224, 164, 171, 183–84, 197–99, 230–32, 237, 253–54, 201–3, 205, 207–8, 237, 256, 261–63, 265–67, 240–43, 245–47, 270, 274, 272, 275, 280, 283, 285, 279, 298, 306, 314, 317, 287, 289–91, 293–94, 330, 332, 350–51, 355, Index 366–67, 369, 372–74, 377, 539 u-vowel, 185, 333 387, 413, 427–28, 430, 432, vowel point, 45 434, 438, 453, 456, 458–59, vowel shift, 34 464, 471, 476, 478 Vulgate, 7, 56, 57, 59–65, 68, vocalism, 273, 317, 459 70–72, 78, 146, 242, 311, vowel 323, 356, 358–59, 361, 456 a-vowel, 37, 46, 94, 96, Wisdom literature, 324, 338 103, 274, 373, 389 writer, 91, 329, 342, 365, 421 anceps, 103–4, 134 Writings, 161, 185, 217, 241– e-vowel, 96, 151, 362 43, 390, 392, 401, 403, 436, final, 390–91 439, 474 i-vowel, 146, 151 Yavne Yam, 33 linking, 146 Yavne Yam (Meṣad long, 30, 103 o-vowel, 32 short, 84, 185 Ḥashavyahu), 33, 127, 129 Zephaniah, 254 Cambridge Semitic Languages and Cultures General Editor Geoffrey Khan OPEN ACCESS Cambridge Semitic Languages and Cultures About the series This series is published by Open Book Publishers in collaboration with the Faculty of Asian and Middle Eastern Studies of the University of Cambridge. The aim of the series is to publish in open-access form monographs in the field of Semitic languages and the cultures associated with speakers of Semitic languages. It is hoped that this will help disseminate research in this field to academic researchers around the world and also open up this research to the communities whose languages and cultures the volumes concern. This series includes philological and linguistic studies of Semitic languages, editions of Semitic texts, and studies of Semitic cultures. Titles in the series will cover all periods, traditions and methodological approaches to the field. The editorial board comprises Geoffrey Khan, Aaron Hornkohl, and Esther-Miriam Wagner. This is the first Open Access book series in the field; it combines the high peer-review and editorial standards with the fair Open Access model offered by OBP. Open Access (that is, making texts free to read and reuse) helps spread research results and other educational materials to everyone everywhere, not just to those who can afford it or have access to well-endowed university libraries. Copyrights stay where they belong, with the authors. Authors are encouraged to secure funding to offset the publication costs and thereby sustain the publishing model, but if no institutional funding is available, authors are not charged for publication. Any grant secured covers the actual costs of publishing and is not taken as profit. In short: we support publishing that respects the authors and serves the public interest. You can find more information about this serie at: http://www.openbookpublishers.com/section/107/1 Other titles in the series Studies in the Masoretic Tradition of the Hebrew Bible Daniel J. Crowther, Aaron D. Hornkohl and Geoffrey Khan https://doi.org/10.11647/OBP.0330 Diachronic Variation in the Omani Arabic Vernacular of the Al-ʿAwābī District From Carl Reinhardt (1894) to the Present Day Roberta Morano https://doi.org/10.11647/OBP.0298 Sefer ha-Pardes by Jedaiah ha-Penini A Critical Edition with English Translation David Torollo https://doi.org/10.11647/OBP.0299 Neo-Aramaic and Kurdish Folklore from Northern Iraq A Comparative Anthology with a Sample of Glossed Texts, Volume 1 Geoffrey Khan, Masoud Mohammadirad, Dorota Molin & Paul M. Noorlander https://doi.org/10.11647/OBP.0306 Neo-Aramaic and Kurdish Folklore from Northern Iraq A Comparative Anthology with a Sample of Glossed Texts, Volume 2 Geoffrey Khan, Masoud Mohammadirad, Dorota Molin & Paul M. Noorlander https://doi.org/10.11647/OBP.0307 The Neo-Aramaic Oral Heritage of the Jews of Zakho Oz Aloni https://doi.org/10.11647/OBP.0272 Points of Contact The Shared Intellectual History of Vocalisation in Syriac, Arabic, and Hebrew Nick Posegay  Winner of the British and Irish Association of Jewish Studies (BIAJS) Annual Book Prize https://https://doi.org/10.11647/OBP.0271 A Handbook and Reader of Ottoman Arabic Esther-Miriam Wagner (ed.) https://doi.org/10.11647/OBP.0208 Diversity and Rabbinization Jewish Texts and Societies between 400 and 1000 CE Gavin McDowell, Ron Naiweld, Daniel Stökl Ben Ezra (eds) https://doi.org/10.11647/OBP.0219 New Perspectives in Biblical and Rabbinic Hebrew Aaron D. Hornkohl and Geoffrey Khan (eds) https://doi.org/10.11647/OBP.0250 The Marvels Found in the Great Cities and in the Seas and on the Islands A Representative of ‘Aǧā’ib Literature in Syriac Sergey Minov https://doi.org/10.11647/OBP.0237 Studies in the Grammar and Lexicon of Neo-Aramaic Geoffrey Khan and Paul M. Noorlander (eds) https://doi.org/10.11647/OBP.0209 Jewish-Muslim Intellectual History Entangled Textual Materials from the Firkovitch Collection, Saint Petersburg Camilla Adang, Bruno Chiesa, Omar Hamdan, Wilferd Madelung, Sabine Schmidtke and Jan Thiele (eds) https://doi.org/10.11647/OBP.0214 Studies in Semitic Vocalisation and Reading Traditions Aaron Hornkohl and Geoffrey Khan (eds) https://doi.org/10.11647/OBP.0207 Studies in Rabbinic Hebrew Shai Heijmans (ed.) https://doi.org/10.11647/OBP.0164 The Tiberian Pronunciation Tradition of Biblical Hebrew Volume 1 Geoffrey Khan  Winner of the 2021 Frank Moore Cross Book Award for best book related to the history and/or religion of the ancient Near East and Eastern Mediterranean https://doi.org/10.11647/OBP.0163 The Tiberian Pronunciation Tradition of Biblical Hebrew Volume 2 Geoffrey Khan  Winner of the 2021 Frank Moore Cross Book Award for best book related to the history and/or religion of the ancient Near East and Eastern Mediterranean https://doi.org/10.11647/OBP.0194 The Historical Depth of the Tiberian Reading Tradition of Biblical Hebrew Aaron D. Hornkohl This volume explores an underappreciated feature of the standard Tiberian Masoretic tradition of Biblical Hebrew, namely its composite nature. Focusing on cases of dissonance between the tradition’s written (consonantal) and reading (vocalic) components, the study shows that the Tiberian spelling and pronunciation traditions, though related, interdependent, and largely in harmony, at numerous points reflect distinct oral realisations of the biblical text. Where the extant vocalisation differs from the apparently pre-exilic pronunciation presupposed by the written tradition, the former often exhibits conspicuous affinity with post-exilic linguistic conventions as seen in representative Second Temple material, such as the core Late Biblical Hebrew books, the Dead Sea Scrolls, Ben Sira, rabbinic literature, the Samaritan Pentateuch, and contemporary Aramaic and Syriac material. On the one hand, such instances of written-reading disharmony clearly entail a degree of anachronism in the vocalisation of Classical Biblical Hebrew compositions. On the other, since many of the innovative and secondary features in the Tiberian vocalisation tradition are typical of sources from the Second Temple Period and, in some cases, are documented as minority alternatives in even earlier material, the Masoretic reading tradition is justifiably characterised as a linguistic artefact of profound historical depth. As with all Open Book publications, this entire book is available to read for free on the publisher’s website. Printed and digital editions, together with supplementary digital material, can also be found at www.openbookpublishers.com Cover image: T-S AS 8.129. A leaf from a Cairo Geniza biblical codex containing Gen. 30.17–20 and showcasing Moshe Moḥe’s non-standard Tiberian pointing of the standard Tiberian pronunciation of Issachar (see within, ch. 4), courtesy of the Syndics of Cambridge University Library. Cover design: Jeevanjot Kaur Nagpal. e book ebook and OA editions also available