Skip to main content

Marek Majer

The University of Lodz, Department of Slavic Philology, Faculty Member

Followers

532

Following

404

Co-authors

5

Public Views

BIO:
Asst. Prof., Dept. of Slavic Philology (2017–now, University of Lodz) (http://slavica.uni.lodz.pl)
Ph.D. in Linguistics (2017, Harvard University)
M.A. in Slavic Languages (2010, University of Lodz)
M.A. in English (2009, University of Lodz)

RESEARCH INTERESTS:
– Slavic languages, linguistics and philology;
– Indo-European linguistics
– the linguistic neighbors of Slavic (especially the Baltic countries and the Balkans, including the non-Slavic languages of these regions, particularly Albanian)
– general historical linguistics
Address: University of Lodz, Faculty of Philology
Department of Slavic Philology
Pomorska 171/173
90-236 Lodz, Poland

Uniwersytet Łódzki, Wydział Filologiczny
Katedra Filologii Słowiańskiej
ul. Pomorska 171/173
90-236 Łódź

less

Maya D Vlahova-Angelova

Bulgarian Academy of Sciences

Nadka N Nikolova

Shumen University

Larry Labro Koroloff

Ivan G. Iliev / Ivan D. Dobrev

University of Plovdiv

Texas A&M University

Rafie F E D A Y E V A Demirova

Trakya University

Живко Войников

Таня Нейчева

University of Plovdiv

Bulgarian Academy of Sciences

Iliyana Krapova

Università Ca' Foscari Venezia

InterestsView All (39)

Uploads

Papers by Marek Majer

Common Slavic *-nǫ-, *-ny- , or *-nu-? A New Look at the History of the Slavic Nasal Suffix. The Diachronic Background and the Rise of the Common Slavic Variation

by Marek Majer and Rafał Szeptyński

Slavistična revija, 2024

In a previous study we showed that in some peripheral dialects of Common Slavic, the aorist/infin... more In a previous study we showed that in some peripheral dialects of Common Slavic, the aorist/infinitive suffix of Class II verbs was *-nu-. In the present study, we focus on the diachronic explanation of the variation *-nǫ- ~ *-ny- ~ *-nu-in this suffix. We propose that its rise was facilitated by CSl phonological developments in final syllables, at the juncture of stem markers and the PIE endings: 2sg *-neh₂-s (or possibly 2sg *-neh₂-s(-s) and 3sg *-neh₂-s-t) > CSl *-ny; (possibly) 1sg *-new-m > CSl *-nǫ; 3sg and word-internally *-new-(t) > CSl *-nu(-). The resulting allomorphy could be leveled in different directions, whereby the causes traditionally invoked to explain *-nǫ-and *-ny- (perseverative nasalization, analogy to other verbal stems) played a role too.

Common Slavic *-nǫ-, *-ny-, or *-nu-? A New Look at the History of the Slavic Nasal Suffix. Slovene Evidence for *-nu-

by Marek Majer and Rafał Szeptyński

Slavistična revija, 2024

The current consensus is that the aorist/infinitive suffix of Class II verbs in Common Slavic was... more The current consensus is that the aorist/infinitive suffix of Class II verbs in Common Slavic was either *-nǫ- (in most of the area) or *-ny- (part of the western periphery). In this study, we argue that this picture is still inadequate as it does not account for part of the dialectal Slovene evidence. A careful analysis shows that some peripheral varieties of Slovene must reflect a pre-form *-nu-, lexicalized relics of which are also found more widely in Slovene. The existence of CSl *-nu- has thus far only been conjectured based on internal reconstruction and Indo-European parallels. (An attempt at an integrated diachronic explanation of the rise of the CSl variants will be the task of a separate ensuing study.)

Polabian <Waspéisede> '(go) backwards' and the type of Common Slavic vъspętь

(forthc.)

ABSTRACT: The entry <Waspéisede> ‘hinter sich gehen’ (‘go backwards’) in Hennig’s Polabian lexico... more ABSTRACT: The entry <Waspéisede> ‘hinter sich gehen’ (‘go backwards’) in Hennig’s Polabian lexicon clearly contains the same root as /paizdă/ ‘buttocks, arse’ (*pizda), but the traditional interpretations as either a substantive /våspaizdĕ/ (*vъs-pizd-ьje) ‘back part, area near the buttocks’ or as a verbal form /våspaizdĕ/ (<< *vъs-pizd-i(tь)) ‘goes backwards’ run into serious problems. In this contribution, it is proposed that the entry should instead be read /våspaizd/, with a rare instance of a double silent <e> in the spelling, as in <Risekenóna> /rüzgnonă/ ‘swollen’ (*orzgъnanoje). The word can then be analyzed as an adverbial form amounting to pre-Polabian *vъs-pizd-ь ‘backwards’. It is identical in structure with Common Slavic *vъs-pęt-ь ‘id.’ and was probably its direct replacement by way of root update (perhaps under the influence of MLG tōrügge).

The Polabian word for 'bird of prey, accipitrid' and PIE *lēk-eh₂-(ye-)

(forthc.)

The Polabian word /lekănaićă/, denoting a bird of prey, has no convincing etymology. Attempts to ... more The Polabian word /lekănaićă/, denoting a bird of prey, has no convincing etymology. Attempts to link it to Common Slavic *kanja ‘bird of prey, accipitrid’ have proved far-fetched. In this study, it is argued that /lekănaićă/ can continue a form *lěkanica ‘flyer, diver’, which would be a derivative from Common Slavic *lěkati ‘fly/dive repeatedly’ – itself unattested, but exactly superimposable on Latvian lȩ̃kât lȩ̃kãju ‘jump, leap, move rapidly, dance, fly’ and Greek ληκάω ‘have intercourse’ (Attic comic poets) / ‘dance to a song’ (Hesychius), a well-known equation documenting the class of Proto-Indo-European iteratives/intensives with the suffix *-eh₂-(ye-) and lengthened ē-grade in the root. The analysis is further supported by the association of the root *lek- with ‘flight’ in Balto-Slavic (Lithuanian lė̃kti lẽkia ‘fly, soar, run’, CSl *letěti ‘fly’ < *lek-t-) and by word-formation parallels involving the suffix.

Etimologjia e fjalës kot ‘pa dobi’

Seminari Ndërkombëtar për Gjuhën, Letërsinë dhe Kulturën Shqiptare

Propozojmë një etimologji të re të fjalës shqipe kot (edhe nkot) ‘pa dobi’. Argumentojmë që trajt... more Propozojmë një etimologji të re të fjalës shqipe kot (edhe nkot) ‘pa dobi’. Argumentojmë që trajta origjinale ka qenë nkot (khs., në shqipen e vjetër, tek Bogdani: kurraj Virgjina rrij nkot; tek Budi: për të nkottë) dhe që ajo është vazhdimësi e një shprehjeje (post)protoindoevropiane *ne kweh2-tu-m ‘për asnjë përfitim’, ku *ne është pjesëza origjinale mohuese pie. dhe *kweh2-tu- ‘përfitim’ përfaqëson një emër abstrakt nga rrënja foljore pie. *kweh2- ‘posedoj’ – burimi i shq. kam. Arsyeja për mungesën e zëshmimit të nk- në ng- mund të ketë qenë kufiri origjinal ndërmjet fjalëve, veçanërisht nëse pjesëza mohuese e vjetëruar *ne u riinterpretua si parafjalë n(d)ë. (Si një alternativë, mund të supozonim që mohimi origjinal *ne është eliminuar përmes procesit të njohur si ‘Cikli i Jespersenit’ – khs. latinisht nōn passum ‘asnjë hap’ > frëngjisht ne pas ‘nuk’ >> frëngjisht bised. pas ‘nuk’. Në këtë rast, varianti nkot do të duhej të rezultonte nga njëfjalëzimi me parafjalën në, siç supozohet zakonisht.)

We present a new etymology of the Albanian word kot (also nkot) ‘in vain’. We suggest that the original form is nkot (cf., in Old Albanian, Bogdani: kurraj Virgjina rrij nkot; Budi: për të nkottë) and that it continues a (post-)Proto-Indo-European expression *ne kweh2- tu-m ‘for no gain’, where *ne is the original PIE negation marker and *kweh2-tu- ‘gain’ represents an abstract noun built to the PIE verbal root *kweh2- ‘possess’ – the base of Alb. kam ‘have’. The lack of voicing of nk- to ng- can be explained by the persistent word boundary, especially if the obsolete negation marker *ne was reinterpreted as the preposition n(d)ë. (Alternatively, it could be assumed that the original negation *ne was elided through the process known as Jespersen’s Cycle; cf. Latin nōn passum ‘not a step’ > French ne pas ‘not’ >> colloquial French pas ‘not’. In this case, the variant nkot would have to result from univerbation with the preposition në, as usually assumed.)

Two studies in Albanian etymology: vobektë 'poor', send 'thing'

Demiraj, B. (ed.), The Albanian Language Area and its Surroundings from Late Antiquity to the High Middle Ages [Albanische Forschungen, vol. 50], 2024

The article contains a detailed study of the etymology of two Albanian lexemes. For vobektë 'poor... more The article contains a detailed study of the etymology of two Albanian lexemes. For vobektë 'poor' (primitive forms: Tosk vabëk, Geg vobeg), the standard explanation as a loanword from Slavic *ubogъ is maintained. However, it is argued that the Albanian forms can only be accounted for properly if one recognizes that the borrowing belongs to the earliest layer of Slavicisms. Then, the assumption of the pre-9th century Proto-Slavic input form [ōbăg-] can yield the attested Albanian forms. For send 'thing', an entirely new account is offered. The etymon is specified as Proto-Indo-European *ku̯endʰ-os n. 'suffering' (= Greek πένθος 'unhappy event, sorrow'), with reference to the typologically common semantic developments 'suffer' > 'toil, work' and 'work, deed' > 'thing, object' (as observed e.g. in Macedonian and Bulgarian rabota).

Потеклото на глаголските прилози од типот -ник

Прилози. Македонска академија на науките и уметностите. Оделение за лингвистика и литературна наука 46.1 [= Посветени на акад. Зузана Тополињска по повод 90-годишнината од раѓањето], p. 251-276, 2021

The East South Slavic gerundive (= verbal adverb, transgressive, etc.; Mac. глаголски прилог, Bg.... more The East South Slavic gerundive (= verbal adverb, transgressive, etc.; Mac. глаголски прилог, Bg. деепричастие) displays great formal diversity in the modern vernacular dialects. Almost all variants can be derived from fossilized case forms of the old present active participle in -ǫť-/-ęť-, often followed by diverse ‘extensions’ (nominal suffixes, adverbial markers, particles, etc.) However, certain variants cannot be explained in this way. One such problematic type are forms with initial -n- (-nik, -nički, -nica, -na, etc.) mostly known from the southernmost dialects. It is usually assumed that this type is based on the past passive participle in -n- with influence from the personal suffix -(n)ik. Such an unexpected diachronic source for a gerundive seems problematic for several reasons. In this study, it is suggested that gerundives of the type -nik arose from the old nom.sg.m of the present active participle in -ę (a well-known source of gerundives across Slavic) followed by adverbial ‘extensions’ such as -ica, by means of unetymological nasal vowel resolution as in the type *telę *telęte ‘calf’ → *tel-en-ъkъ; thus also *plač-ę ‘crying’ → *plač-en-ica etc.

The origin of East South Slavic gerunds of the type -nik

forthc.

The East South Slavic gerundive (= verbal adverb, transgressive, etc.; Mac. глаголски прилог, Bg.... more The East South Slavic gerundive (= verbal adverb, transgressive, etc.; Mac. глаголски прилог, Bg. деепричастие) displays great formal diversity in the modern vernacular dialects. Almost all variants can be derived from fossilized case forms of the old present active participle in -ǫť-/-ęť-, often followed by diverse ‘extensions’ (nominal suffixes, adverbial markers, particles, etc.) However, certain variants cannot be explained in this way. One such problematic type are forms with initial -n- (-nik, -nički, -nica, -na, etc.) mostly known from the southernmost dialects. It is usually assumed that this type is based on the past passive participle in -n- with influence from the personal suffix -(n)ik. Such an unexpected diachronic source for a gerundive seems problematic for several reasons. In this study, it is suggested that gerundives of the type -nik arose from the old nom.sg.m of the present active participle in -ę (a well-known source of gerundives across Slavic) followed by adverbial ‘extensions’ such as -ica, by means of unetymological nasal vowel resolution as in the type *telę *telęte ‘calf’ → *tel-en-ъkъ; thus also *plač-ę ‘crying’ → *plač-en-ica etc.

A new possible Slavic source for Romanian a privi 'to look'

forthc.

for their help in obtaining certain less easily accessible books and articles. The responsibility... more for their help in obtaining certain less easily accessible books and articles. The responsibility for all views is mine. 2 There are no grounds to consider the arbitrary Daco-Thracian etymology proposed by VINEREANU (2008: 677), where the pre-form is reconstructed as *pri-vili-, with subsequent palatalization and loss of *l, and associated with the PIE root *wel-'see ' (LIV₂: 675). This judgment applies to the methodological approach of the work in general, where Daco-Thracian etymologies are sought not only for words that may genuinely reach back to some kind of remote Balkan ancestry, but also -or perhaps first of all -for glaring Slavicisms. Cf., just to stay within the range of several pages from privi, the interpretation of Rom răstoacă 'backwater' as derived from PIE *reyǵ-'stretch, bind' rather than borrowed from SSl *rastoka or Rom praf 'dust' from PIE *pel-es-rather than from SSl *praxъ. 3 The other factor that needs to be reckoned with here is, of course, contamination (cf. LESCHBER 2010, 2012 for state-of-theart overviews and numerous examples). 4 It may also be noted that the verb displays many other meanings than the prototypical ones specified at the outset of the article (e.g. 'follow', 'believe', or even 'live'; cf. DLR VIII/5: 1521-1523); these, however, may be obtained far more easily from various shades of 'look, watch, regard, consider' than the other way round. Thus, it may be assumed that they represent Romanian-internal developments. Under certain etymologies it could be argued that some of these meanings preserve the original state, but this does not appear likely.

The Caland System as an explanatory device in Balto-Slavic etymology. A historical perspective

In: Jouravel, A. & Mathys, A. (eds.), Wort- und Formenvielfalt. Festschrift für Christoph Koch zum 80. Geburtstag, p. 251–272, 2021, 2021

The prehistory of the Proto-Slavic abstract suffix *-ostь and its Baltic congeners

Scando-Slavica (forthc.)

1 The problem PSl. 1 *-ostь is the parent form of a suffix immediately recognizable to anyone fam... more 1 The problem PSl. 1 *-ostь is the parent form of a suffix immediately recognizable to anyone familiar with a modern Sl. language. Reflexes of *-ostь productively derive abstract nouns from adjectives across the family (Ru. -ost , Ukr. -ist , Pol. -ość, Cz. -ost, Sln. -ost, Bulg. -ost etc.); 2 some examples are provided in Table 1. Table 1. Sample derivatives in *-ostь in the modern Sl. languages adjective abstract noun PSl. projection Ru. smelyj 'bold' → smelost ́ 'boldness' *sъmělъ → *sъmělostь -i Ukr. hidnyj 'dignified' → hidnist ́ 'dignity' *godьnъ → *godьnostь -i Pol. stary 'old' → starość 'old age' *starъ → *starostь -i Cz. hloupý 'stupid' → hloupost 'stupidity' *glupъ → *glupostь -i Sln. mlad 'young' → mladost 'youth' *moldъ → *moldostь -i Bulg. tvărd 'hard' → tvărdost 'hardness' *tvьrdъ → *tvьrdostь -i The suffix is, of course, well known also from early written corpora: OCS mǫdrъ 'wise' → mǫdrostь 'wisdom', ORu. radъ 'glad' → radostь 'gladness', etc. Although few of the forms presented in Table could actually be posited for PSl., there can be no doubt * A rudimentary version of the present theory was included in my PhD thesis (The Caland System in the North: Archaism and Innovation in Property-Concept/State Morphology in Balto-Slavic, Harvard University, 2017), written under the supervision of Jay H. Jasanoff. I am also grateful to the participants of those events at which the paper was presented in a more expanded form (14 th Annual Meeting of the Slavic Linguistics Society, Potsdam, 11-13 September 2019, as well as a guest talk at Vilnius University on

With all due respect, on Slavic abstracts in -y: The history of Proto-Slavic cěty ‘respect’ and some comparative notes on its congeners (ljuby ‘love’, cěly ‘healing, greeting’)

by Marek Majer and Rafał Szeptyński

Journal of Slavic Linguistics, 2021

The scarcely attested Proto-Slavic *cěty *-ъve ‘respect’ appears to be a mostly overlooked member... more The scarcely attested Proto-Slavic *cěty *-ъve ‘respect’ appears to be a mostly overlooked member of the small class of abstracts in *-y *-ъve; no precise accounts of the noun’s origin have been proposed so far. Two complementary approaches are put forth in the article: 1) inheritance from a PIE animate s-stem *keyt-ōs >> *koyt-ōs (paralleling a recent analysis of *ljuby ‘love’ < PIE *lewbh-ōs as well as its presumed secondary association with a verb in *-i-ti), or 2) inner-Slavic origin based on the formally similar *ljuby ‘love’ and *cěly ‘a healing’. The study also offers novel analyses—based on hitherto unexploited philological and lexicographical data—concerning various related issues (e.g. the status of PSl. nominal *cětъ, verbal *cětiti, and personal names in *Cěto/i-; the adposition *cětja; the semantic and pragmatic developments in *cěly ‘greeting, kiss’; the secondary rise of masculine *cělovъ/*cělyvъ ‘kiss’) with the purpose of integrating the entirety of the material concerning the root *cět- and the abstract type in *-y *-ъve into coherent pictures.

Slavic *ljuby and the heterogeneity of the inflectional class in *-y

Indogermanische Forschungen

The inherited Proto-Slavic feminine nouns in nom.sg *-y, oblique *-ъv-are conventionally thought ... more The inherited Proto-Slavic feminine nouns in nom.sg *-y, oblique *-ъv-are conventionally thought to go back to one source only, namely Proto-Indo-European stems in *-uh-. However, a handful of previous studies have already demonstrated the possibility that other inherited items could join this paradigm by phonologically developing a nom.sg in *-y from other sources (particularly *-ōs). Accordingly, in the present article, it is hypothesized that the well-known PSl. noun *ljuby *ljubъve F 'love' may go back to an amphikinetic s-stem *lewbʰ-ōs. Such a reconstruction makes it possible to explain a number of otherwise problematic features of the complex of *ljuby, such as: 1) the enigmatic OCS acc.sg ljuby (for expected ljubъvь) in the collocations (prě)ljuby (sъ)tvoriti or (prě)ljuby dějati 'commit adultery'; 2) the unexpected presence of *-sin the Balt. cognate, Lith. liaupsė̃ F 'praise, adoration'; 3) the occurrence of PSl. *-y as an abstract marker (a function in which PIE amphikinetic s-stems in *-ōs are securely, even if marginally, found, particularly in the domain of emotions and mental states-cf. Gr. αἰδώς αἰδόος F 'reverence, awe', ἔρως *ἔροος M 'love, desire', Ved. bhiyás-, Av. biiah-M 'fear'-while no such function can convincingly be reconstructed for a formation in *-u-h2).

The Indo-European prehistory of the Albanian word for 'sister'

proceedings of: Ljubilej – Ljubljäum – Ljubilee – Ljubljanniversaire. IG / SIES / SÉIE Arbeitstagung (100 years of comparative linguistics at the University of Ljubljana), Ljubljana, 4–7 June 2019

In the present article, which builds on an almost overlooked suggestion by Eric Hamp, it is argue... more In the present article, which builds on an almost overlooked suggestion by Eric Hamp, it is argued that Albanian motër 'sister' is not a direct reflex of Proto-Indo-European *meh₂ter- 'mother' (with an unexpected shift of meaning)-as assumed almost universally in the literature-but that it represents an early derivative of this word, probably *māt(e)rā < *meh₂t-(e)r-eh₂ 'of the [same] mother' or 'having the [same] mother'. Such an item may have been easily lexicalized as a replacement of the inherited word for 'sister'; cf. Greek αδελφ(ε)ός 'brother' < *sm ̥-gʷelbʰ-es-o-'of the same womb' or Spanish hermano 'brother' < Latin germānus 'of [one's own] kin, real, full'. Thus, despite widespread claims to the contrary, Albanian 'sister' can probably be explained via fairly trivial developments and does not document any extraordinary facts about (Pre-)Albanian kinship terminology or social structure.

Parahistoria indoevropiane e fjalës shqipe për 'motrën'

Seminari Ndërkombëtar për Gjuhën, Letërsinë dhe Kulturën Shqiptare, 2019

Shq. motër (në disa dialekte dhe në gjuhën e vjetër edhe motrë, motërë) konsiderohet përgjithësis... more Shq. motër (në disa dialekte dhe në gjuhën e vjetër edhe motrë, motërë) konsiderohet përgjithësisht si refleks i drejtpërdrejtë i fjalës indoevropiane për ‘nënën’ – *meh₂ter- (khs. skt. mā́tṛ-, gr. mḗtēr, lat. māter, sll. mati etj.). Dallimi kuptimor shpjegohet ose si rezultat i një ndryshimi semantik të gjerë – ndoshta me disa hapa të ndërmjetëm (p.sh. ‘nënë’ > ‘çdo femër e afërt’ > ‘motër’) – ose si pasojë e ndryshimeve në strukturën e përgjithshme të terminologjisë së farefisnisë për shkak të zhvillimeve jashtëgjuhësore (p.sh. është supozuar se motra e madhe përmbushte rolin e nënës). Këto mendime shprehen në të gjithë fjalorët etimologjikë të shqipes (Meyer 1891: 287–288; Çabej 1976: 371; Huld 1984: 96; Demiraj 1997: 279; Orel 1998: 275; Topalli 2018: 1009) dhe të indoevropianishtes (p.sh. NIL: 457–460), si dhe në punimet që trajtojnë historinë e leksikut të shqipes nga perspektivat indoevropiane (p.sh. Barić 1955: 57–58, Tagliavini 1965: 118; Matzinger 2018: 1789).

Megjithatë, një ndryshim aq rrënjësor semantik nuk duket shumë i besueshëm; gjithashtu, nuk ka arsye për të supozuar një ristrukturim të përgjithshëm të terminologjisë së farefisnisë (meqenëse emërtimet e tjera të trashëguara të farefisnisë shfaqen pikërisht me kuptimet e tyre të pritura, p.sh. vjehërr < pie. *sweḱuro- [khs. skt. śváśura-, gr. hekurós, lat. socer, sl. svekrъ] ose nip < pie. *nepot- [khs. skt. nápāt-, gr. anepsiós, lat. nepōs, sl. netьjь]). Në fakt, autoritetet etimologjike më të kujdesshme (p.sh. Demiraj 1997) pranojnë që zhvillimi semantik i fjalës motër mbetet një mister i plotë.

Pra, në këtë studim shq. motër analizohet jo si refleks, por si derivat mjaft i vjetër i fjalës indoevropiane për ‘nënën’. Domethënë, kemi të bëjmë ose me një mbiemër pronor *meh₂t-(e)r-o- ‘i cili ka të njëjtën nënë’, ose me një mbiemër relacional *meh₂t-(e)r-o- ‘nga e njëjta nënë’, ‘i së njëjtës nënë’. Trajta femërore e një mbiemri të tillë (*meh₂t-(e)r-eh₂- > *māterā) mund të ketë qenë leksikalizuar si ‘gruaja që ka të njëjtën nënë’ / ‘gruaja që vjen nga e njëjta nënë’ dhe të ketë zëvendësuar fjalën e trashëguar për ‘motrën’ (pie. *swesor-). Zëvendësime të tilla leksikore janë shumë të përhapura në gjuhët indoevropiane, p.sh. gr. adelph(e)ós ‘vëlla’ < *sm̥-gʷelbʰ-es-o- ‘që vjen nga e njëjta mitër’, span. hermano ‘vëlla’ < lat. germānus ‘nga [i njëjti] fis/farë’ ← germen ‘farë’. Siç dihet mirë, edhe shq. vëlla mendohet të ketë origjinë të këtij lloji: *swe-lowdʰ-eh2- ‘[nga] fisi i vet’.

Me sa duket, një ide e ngjashme në lidhje me fjalën motër është paraqitur në disa vepra të Hamp-it (1986: 110; 1996: 140), por me disa dallime dhe pa një analizë të hollësishme të sfondit indoevropian. Duke ditur që vërejtjet e Hamp-it rreth kësaj fjale fatkeqësisht nuk janë përfshirë në asnjë fjalor etimologjik dhe mungojnë nga debati i përgjithshëm, punimi i pranishëm është një përpjekje për të ofruar një analizë të plotë dhe moderne.

[English] [NB: English version of the paper published separately]

Alb. motër (dialectal and Old Albanian also motrë, motërë) is universally considered to be a direct reflex of the Proto-Indo-European word for ‘mother’ – *meh₂ter- (Sanskr. mā́tr̥-, Gr. mḗtēr, Lat. māter, Slav. *mati *matere etc.). The difference in meaning is explained either as far-reaching semantic change – possibly including intermediate steps (e.g., ‘mother’ > ‘any female relative’ > ‘sister’) – or as the product of changes in kinship terminology caused by extra-linguistic developments (e.g., it has been assumed that the older sister fulfilled the role of the mother). However, such a radical semantic change does not appear likely; similarly, there is no reason to assume any general restructuring of kinship terminology, given that other inherited kinship terms surface precisely with their expected meaning (vjehërr ‘father-in-law’ < PIE *sweḱuro-, nip ‘grandson/nephew’ < PIE *nepot-). Thus, in the present study – which draws on a little-known idea by Eric Hamp – Alb. motër is analysed not as a reflex but as an early derivative of the IE word for ‘mother’. Namely, we are dealing either with a possessive adjective *meh₂t-(e)r-o- ‘having the [same] mother’ or with a relational adjective *meh₂t-(e)r-o- ‘of the [same] mother’. The feminine form of such an adjective (*meh₂t-(e)r-eh₂ > *māterā) may have been lexicalized as ‘woman having the same mother’ / ‘woman of the same mother’ and replaced the inherited word for ‘sister’ (PIE *swesor-). Such lexical replacements are widespread; cf. Greek adelph(e)ós ‘brother’ < *sm̥-gʷelbʰ-es-o- ‘of the same womb’ or Spanish hermano ‘brother’ < Lat. germānus ‘real, *from the [same] seed’.

Slavic 'i-stem adjectives' and their alleged inflection loss: The derivational prehistory and synchronic status of a category

proceedings of: Dérivation nominale et innovations dans les langues indo-européennes anciennes (Université de Rouen, 11–12 Oct 2018)

Pozostałości praindoeuropejskiego „Systemu Calanda” w języku prasłowiańskim – wybrane przykłady [Remnants of the Proto-Indo-European “Caland System” in Proto-Slavic – selected examples]

In: Jakubowicz, M. et al. (eds.), Etymologica slavica. Studia etymologiczne poświęcone prof. Franciszkowi Sławskiemu z okazji setnej rocznicy urodzin. Warszawa: Instytut Slawistyki PAN, p. 81–99, 2019

The ‘Caland system’ (CS) is a set of morphological rules governing the behaviour of many ‘prototy... more The ‘Caland system’ (CS) is a set of morphological rules governing the behaviour of many ‘prototypically adjectival’ (or ‘property-concept’) roots and their adjectival, nominal and verbal derivatives in PIE and several archaic IE languages (mostly Vedic, Greek and Latin). It consists in the productive use of a set of formants (e.g. adjectives in *-u- and *-ro-, comparatives in *-yos-, abstracts in *-es-), added directly to the root instead of a pre-existing adjectival stem. The present article reviews a few synchronically irregular Proto-Slavic forms, arguing that the observed morphological anomalies (e.g. the relationship between the positive and the comparative in adjectives like *deb-elъ ‘thick’, compar. *deb’-ьš-, or between the adj. *orvь-nъ ‘even’ and its old abstract *orvo *orvese), can be explained as (partially modified) CS inheritance, even if the roots themselves are lexical innovations.

The Slavic source of Albanian venit(em) 'wane', mahnit(em) 'astound' and certain other verbs

Contributions to the 22nd Annual Scientific Conference of the Association of Slavists (Polyslav), 2019

The shape of Albanian verbs borrowed from Slavic is normally determined by the Sl. infinitive (Sl... more The shape of Albanian verbs borrowed from Slavic is normally determined by the Sl. infinitive (Sl. *-ati → Alb. -at-; Sl. *-iti → Alb. -it-). Surprisingly, however, the element -it- also occurs in certain Alb. verbs whose Sl. source does not have an *-iti infinitive. Based on methods of external and internal reconstruction, it is argued that some such verbs (venit(et) ‘wane’, mahnit(et) ‘astound’) may have been borrowed from a Slavic dialect where the respective source verbs had an infinitive in *-nyti (*vę(d)nyti ‘wane’, *maxnyti ‘swing’; cf. e.g. Slovene veniti, mahniti) – rather than *-nǫti, as in most Slavic languages (*vę(d)nǫti, *maxnǫti ‘swing’; cf. e.g. Serbian venuti, mahnuti). Since Sl. *y normally yields i in loanwords into Alb., a Sl. infinitive in *-nyti may be expected to produce an Alb. verbal stem in -(n)it-. Subsequently, the type could have spread (through proportional analogy) to some other verbs borrowed from Slavic (e.g. pjetit ‘braid’, bodit ‘stab’). Since no Sl. varieties of the *-nyti type are currently found in the zones neighboring Alb., the theory might have interesting consequences for Balkan linguistic geography, perhaps suggesting a different past location of certain Sl. dialects (*-nyti type ranging further south) and/or of Albanian (further north).

Rindërtimi i morfologjisë së huazimeve sllave-shqiptare: rasti i shq. venit(em) dhe mahnit(em) [Reconstructing the morphology of Slavic-Albanian loanwords: The case of venit(em) and mahnit(em)]

Seminari Ndërkombëtar për Gjuhën, Letërsinë dhe Kulturën Shqiptare, 2018

Ndërsa disa prej huazimeve sllave në gjuhën shqipe janë relativisht të reja (pra burimi i tyre mu... more Ndërsa disa prej huazimeve sllave në gjuhën shqipe janë relativisht të reja (pra burimi i tyre mund të identifikohet lehtësisht me dialektet e njohura të serbishtes, të maqedonishtes etj.), disa të tjera janë mjaft të vjetra; duket se ato janë huazuar nga disa dialekte sllave që nuk janë të dëshmuara në mënyrë të drejtpërdrejtë.

Shumë prej huazimeve sllave janë folje. Ato ndahen në dy grupe kryesore: foljet që i përkasin tipit *‐ati në sllavishte i përgjigjen tipit ‐at(em) në shqipe, p.sh. shq. dumat ‘peshoj, mendoj’ ← sll. *dumati (khs. serb. dȕmati, rus. dúmat', pol. dumać), kurse foljet që i përkasin tipit *‐iti në sllavishte i përgjigjen tipit ‐it(em) në shqipe, p.sh. shq. vozit ← sll. *voziti ‘lëviz, çoj’ (khs. serb. vòziti, rus. vozít', pol. wozić).

Mirëpo, është më pak e qartë se përse tipi ‐it(em) në shqipe përmban gjithashtu disa huazime sllave që nuk kanë as paskajoren *‐iti as temën e kohës së tashme *‐i‐ në sllavishte. Dy shembuj të rëndësishëm janë folje me temën e kohës së tashme *‐ne‐ (klasa II në klasifikimin e foljeve sllave të A. LESKIEN‐it): shq. venit(em) ← sll. *vę(d)ne‐ (serb. vènuti vȅnē, rus. vjánut' vjánet, pol. więdnąć więdnie) dhe shq. mahnit(em) ← sll. *maxne‐ ‘tund, lëkund’ (serb. máhnuti mȃhnē, rus. maxnút' maxnët, pol. machnąć machnie). Siç mund të shihet, foljet e këtij tipi zakonisht kanë paskajoren *‐ǫti
në sllavishte (p.sh. *vę(d)nǫti, *maxnǫti), që bëhet rregullisht serb. ‐uti (p.sh. serb.vènuti, máhnuti). Nuk është zbuluar asnjë arsye e qartë pse sll. *vę(d)ne‐ (serb. vènuti) dhe *maxne‐ (serb. máhnuti) janë bashkuar me tipin ‐it(em) në shqipe – në vend se të ndertojnë një tip hipotetik të ri **‐ut(em), i cili do të mbështej mirë nga leksiku i trashëguar (khs. p.sh. ngut(em)).

Pra, këtu argumentohet se shqipja ka huazuar të dyja foljet nga një dialekt sllav i tipit *‐nyti, domethënë një dialekt në të cilin paskajorja ka pasur formën *vę(d)nyti, *maxnyti etj. dhe jo *vę(d)nǫti, *maxnǫti etj. Për këtë arsye, format shqipe janë venitem, mahnitem (zanores *y të elementeve sllave i përgjigjet në shqipe rregullisht zanorja i, khs. shq. pihatem ‘këputem, lodhem shumë’ ← sll. *pyxati). Ky konkluzion është shumë interesant, aq më tepër se në kohën tonë në zonën e kontakteve gjuhësore sllave-shqiptare nuk më njihet asnjë dialekt sllav i tipit *‐nyti (dialektet e këtij tipi më të afërta njihen nga Kroacia veriore). Kjo do të thotë se hipoteza jonë – edhe pse ajo duhet të mbetet e pasigurt për shkak të mungesës së të dhënave të mëtejshme – mund të krahasohet me gjeografinë dialektore dhe me historinë e migrimeve sllave në Ballkan (shpërndarja e dialekteve të tipit *‐nǫti dhe *‐nyti është një faktor i rëndësishëm, khs. ANDERSEN 1999).

[English] [NB: English version of the paper published separately]

While some of the Slavic loanwords in Albanian are recent borrowings of items familiar from the neighboring Slavic dialects, certain others are evidently of considerable age and probably borrowed from Slavic varieties that are not directly attested.

As far as the verbal material is concerned, many Slavic borrowings enter the Albanian verbal class in -at (mid./pass. -atem). These are, in principle, the Slavic verbs in *-ati (cf. Alb. dumat ‘think out’ ← Sl. *dumati: Bosnian/Croatian/Serbian dȕmati, Russian dúmat', Polish dumać). Another big class of Sl. loanwords are the verbs in -it (mid./pass. -item). These are also clear as long as they represent Sl. verbs in *-iti (cf. Alb. vozit ‘row, drive’ ← Sl. *voziti: Bosnian/Croatian/Serbian vòziti, Russian vozít', Polish wozić).

However, it is less obvious why the Albanian class in -it(em) also includes some Slavic loanwords that exhibit neither an infinitive in *-iti nor a present stem in *-i- in Slavic. Here belong, most notably, two Slavic verbs with a present stem in *-ne- (Leskien’s class II): Alb.venit ‘cause to wane’ (usually mid./pass. venitem ‘wane, wither’) ← Sl. *vę(d)ne- (B/C/S vènuti vȅnē, Ru. vjánut' vjánet, Pol. więdnąć więdnie) and Alb. mahnit ‘astound, stupify’ (mid./pass. mahnitem ‘be astounded’) ← Sl. *maxne- ‘swing, brandish’ (B/C/S máhnuti mȃhnē, Ru. maxnút' maxnët, Pol. machnąć machnie). As can be seen, these verbs generally display an infinitive in *-ǫti, cf. *vę(d)nǫti, *maxnǫti, which regularly becomes B/C/S -uti, cf. vènuti, máhnuti. No clear reason for Slavic *vę(d)nǫ and *maxnǫ ending up in the Albanian class in -it(em) – rather than a hypothetical class in **-ut(em), which would be well supported by native Albanian lexicon, cf. ngut(em) ‘hurry’ – has ever been identified.

Accordingly, it is argued here that Albanian might have borrowed the relevant verbs from a *-nyti dialect, i.e. one in which the infinitives had the form *vę(d)nyti and *maxnyti rather than*vę(d)nǫti and *maxnǫti (the reflex of South Slavic *y as i is regular in Albanian, cf. e.g. Alb.pihatem ‘be exhausted’ ← *pyxati). This is especially interesting in view of the fact that no *-nyti dialects are currently known from the geographic area neighboring Albanian, so that the hypothesis ‒ although it must remain tentative in view of the scanty material – can be compared with the dialectgeography and settlement history of the South Slavic area (including the distribution of the *-nǫti and *-nyti dialects as studied in ANDERSEN 1999).

Review of: Slavic Nominal Word Formation. Proto-Indo-European origins and historical development, R. Matasović, 2014

Journal of Slavic Linguistics, 2017

review of: Slavic Nominal Word Formation. Proto-Indo-European origins and historical development,... more

Common Slavic *-nǫ-, *-ny- , or *-nu-? A New Look at the History of the Slavic Nasal Suffix. The Diachronic Background and the Rise of the Common Slavic Variation

by Marek Majer and Rafał Szeptyński

Slavistična revija, 2024

In a previous study we showed that in some peripheral dialects of Common Slavic, the aorist/infin... more In a previous study we showed that in some peripheral dialects of Common Slavic, the aorist/infinitive suffix of Class II verbs was *-nu-. In the present study, we focus on the diachronic explanation of the variation *-nǫ- ~ *-ny- ~ *-nu-in this suffix. We propose that its rise was facilitated by CSl phonological developments in final syllables, at the juncture of stem markers and the PIE endings: 2sg *-neh₂-s (or possibly 2sg *-neh₂-s(-s) and 3sg *-neh₂-s-t) > CSl *-ny; (possibly) 1sg *-new-m > CSl *-nǫ; 3sg and word-internally *-new-(t) > CSl *-nu(-). The resulting allomorphy could be leveled in different directions, whereby the causes traditionally invoked to explain *-nǫ-and *-ny- (perseverative nasalization, analogy to other verbal stems) played a role too.

Common Slavic *-nǫ-, *-ny-, or *-nu-? A New Look at the History of the Slavic Nasal Suffix. Slovene Evidence for *-nu-

by Marek Majer and Rafał Szeptyński

Slavistična revija, 2024

The current consensus is that the aorist/infinitive suffix of Class II verbs in Common Slavic was... more The current consensus is that the aorist/infinitive suffix of Class II verbs in Common Slavic was either *-nǫ- (in most of the area) or *-ny- (part of the western periphery). In this study, we argue that this picture is still inadequate as it does not account for part of the dialectal Slovene evidence. A careful analysis shows that some peripheral varieties of Slovene must reflect a pre-form *-nu-, lexicalized relics of which are also found more widely in Slovene. The existence of CSl *-nu- has thus far only been conjectured based on internal reconstruction and Indo-European parallels. (An attempt at an integrated diachronic explanation of the rise of the CSl variants will be the task of a separate ensuing study.)

Polabian <Waspéisede> '(go) backwards' and the type of Common Slavic vъspętь

(forthc.)

ABSTRACT: The entry <Waspéisede> ‘hinter sich gehen’ (‘go backwards’) in Hennig’s Polabian lexico... more ABSTRACT: The entry <Waspéisede> ‘hinter sich gehen’ (‘go backwards’) in Hennig’s Polabian lexicon clearly contains the same root as /paizdă/ ‘buttocks, arse’ (*pizda), but the traditional interpretations as either a substantive /våspaizdĕ/ (*vъs-pizd-ьje) ‘back part, area near the buttocks’ or as a verbal form /våspaizdĕ/ (<< *vъs-pizd-i(tь)) ‘goes backwards’ run into serious problems. In this contribution, it is proposed that the entry should instead be read /våspaizd/, with a rare instance of a double silent <e> in the spelling, as in <Risekenóna> /rüzgnonă/ ‘swollen’ (*orzgъnanoje). The word can then be analyzed as an adverbial form amounting to pre-Polabian *vъs-pizd-ь ‘backwards’. It is identical in structure with Common Slavic *vъs-pęt-ь ‘id.’ and was probably its direct replacement by way of root update (perhaps under the influence of MLG tōrügge).

The Polabian word for 'bird of prey, accipitrid' and PIE *lēk-eh₂-(ye-)

(forthc.)

The Polabian word /lekănaićă/, denoting a bird of prey, has no convincing etymology. Attempts to ... more The Polabian word /lekănaićă/, denoting a bird of prey, has no convincing etymology. Attempts to link it to Common Slavic *kanja ‘bird of prey, accipitrid’ have proved far-fetched. In this study, it is argued that /lekănaićă/ can continue a form *lěkanica ‘flyer, diver’, which would be a derivative from Common Slavic *lěkati ‘fly/dive repeatedly’ – itself unattested, but exactly superimposable on Latvian lȩ̃kât lȩ̃kãju ‘jump, leap, move rapidly, dance, fly’ and Greek ληκάω ‘have intercourse’ (Attic comic poets) / ‘dance to a song’ (Hesychius), a well-known equation documenting the class of Proto-Indo-European iteratives/intensives with the suffix *-eh₂-(ye-) and lengthened ē-grade in the root. The analysis is further supported by the association of the root *lek- with ‘flight’ in Balto-Slavic (Lithuanian lė̃kti lẽkia ‘fly, soar, run’, CSl *letěti ‘fly’ < *lek-t-) and by word-formation parallels involving the suffix.

Etimologjia e fjalës kot ‘pa dobi’

Seminari Ndërkombëtar për Gjuhën, Letërsinë dhe Kulturën Shqiptare

Propozojmë një etimologji të re të fjalës shqipe kot (edhe nkot) ‘pa dobi’. Argumentojmë që trajt... more Propozojmë një etimologji të re të fjalës shqipe kot (edhe nkot) ‘pa dobi’. Argumentojmë që trajta origjinale ka qenë nkot (khs., në shqipen e vjetër, tek Bogdani: kurraj Virgjina rrij nkot; tek Budi: për të nkottë) dhe që ajo është vazhdimësi e një shprehjeje (post)protoindoevropiane *ne kweh2-tu-m ‘për asnjë përfitim’, ku *ne është pjesëza origjinale mohuese pie. dhe *kweh2-tu- ‘përfitim’ përfaqëson një emër abstrakt nga rrënja foljore pie. *kweh2- ‘posedoj’ – burimi i shq. kam. Arsyeja për mungesën e zëshmimit të nk- në ng- mund të ketë qenë kufiri origjinal ndërmjet fjalëve, veçanërisht nëse pjesëza mohuese e vjetëruar *ne u riinterpretua si parafjalë n(d)ë. (Si një alternativë, mund të supozonim që mohimi origjinal *ne është eliminuar përmes procesit të njohur si ‘Cikli i Jespersenit’ – khs. latinisht nōn passum ‘asnjë hap’ > frëngjisht ne pas ‘nuk’ >> frëngjisht bised. pas ‘nuk’. Në këtë rast, varianti nkot do të duhej të rezultonte nga njëfjalëzimi me parafjalën në, siç supozohet zakonisht.)

We present a new etymology of the Albanian word kot (also nkot) ‘in vain’. We suggest that the original form is nkot (cf., in Old Albanian, Bogdani: kurraj Virgjina rrij nkot; Budi: për të nkottë) and that it continues a (post-)Proto-Indo-European expression *ne kweh2- tu-m ‘for no gain’, where *ne is the original PIE negation marker and *kweh2-tu- ‘gain’ represents an abstract noun built to the PIE verbal root *kweh2- ‘possess’ – the base of Alb. kam ‘have’. The lack of voicing of nk- to ng- can be explained by the persistent word boundary, especially if the obsolete negation marker *ne was reinterpreted as the preposition n(d)ë. (Alternatively, it could be assumed that the original negation *ne was elided through the process known as Jespersen’s Cycle; cf. Latin nōn passum ‘not a step’ > French ne pas ‘not’ >> colloquial French pas ‘not’. In this case, the variant nkot would have to result from univerbation with the preposition në, as usually assumed.)

Two studies in Albanian etymology: vobektë 'poor', send 'thing'

Demiraj, B. (ed.), The Albanian Language Area and its Surroundings from Late Antiquity to the High Middle Ages [Albanische Forschungen, vol. 50], 2024

The article contains a detailed study of the etymology of two Albanian lexemes. For vobektë 'poor... more The article contains a detailed study of the etymology of two Albanian lexemes. For vobektë 'poor' (primitive forms: Tosk vabëk, Geg vobeg), the standard explanation as a loanword from Slavic *ubogъ is maintained. However, it is argued that the Albanian forms can only be accounted for properly if one recognizes that the borrowing belongs to the earliest layer of Slavicisms. Then, the assumption of the pre-9th century Proto-Slavic input form [ōbăg-] can yield the attested Albanian forms. For send 'thing', an entirely new account is offered. The etymon is specified as Proto-Indo-European *ku̯endʰ-os n. 'suffering' (= Greek πένθος 'unhappy event, sorrow'), with reference to the typologically common semantic developments 'suffer' > 'toil, work' and 'work, deed' > 'thing, object' (as observed e.g. in Macedonian and Bulgarian rabota).

Потеклото на глаголските прилози од типот -ник

Прилози. Македонска академија на науките и уметностите. Оделение за лингвистика и литературна наука 46.1 [= Посветени на акад. Зузана Тополињска по повод 90-годишнината од раѓањето], p. 251-276, 2021

The East South Slavic gerundive (= verbal adverb, transgressive, etc.; Mac. глаголски прилог, Bg.... more The East South Slavic gerundive (= verbal adverb, transgressive, etc.; Mac. глаголски прилог, Bg. деепричастие) displays great formal diversity in the modern vernacular dialects. Almost all variants can be derived from fossilized case forms of the old present active participle in -ǫť-/-ęť-, often followed by diverse ‘extensions’ (nominal suffixes, adverbial markers, particles, etc.) However, certain variants cannot be explained in this way. One such problematic type are forms with initial -n- (-nik, -nički, -nica, -na, etc.) mostly known from the southernmost dialects. It is usually assumed that this type is based on the past passive participle in -n- with influence from the personal suffix -(n)ik. Such an unexpected diachronic source for a gerundive seems problematic for several reasons. In this study, it is suggested that gerundives of the type -nik arose from the old nom.sg.m of the present active participle in -ę (a well-known source of gerundives across Slavic) followed by adverbial ‘extensions’ such as -ica, by means of unetymological nasal vowel resolution as in the type *telę *telęte ‘calf’ → *tel-en-ъkъ; thus also *plač-ę ‘crying’ → *plač-en-ica etc.

The origin of East South Slavic gerunds of the type -nik

forthc.

The East South Slavic gerundive (= verbal adverb, transgressive, etc.; Mac. глаголски прилог, Bg.... more The East South Slavic gerundive (= verbal adverb, transgressive, etc.; Mac. глаголски прилог, Bg. деепричастие) displays great formal diversity in the modern vernacular dialects. Almost all variants can be derived from fossilized case forms of the old present active participle in -ǫť-/-ęť-, often followed by diverse ‘extensions’ (nominal suffixes, adverbial markers, particles, etc.) However, certain variants cannot be explained in this way. One such problematic type are forms with initial -n- (-nik, -nički, -nica, -na, etc.) mostly known from the southernmost dialects. It is usually assumed that this type is based on the past passive participle in -n- with influence from the personal suffix -(n)ik. Such an unexpected diachronic source for a gerundive seems problematic for several reasons. In this study, it is suggested that gerundives of the type -nik arose from the old nom.sg.m of the present active participle in -ę (a well-known source of gerundives across Slavic) followed by adverbial ‘extensions’ such as -ica, by means of unetymological nasal vowel resolution as in the type *telę *telęte ‘calf’ → *tel-en-ъkъ; thus also *plač-ę ‘crying’ → *plač-en-ica etc.

A new possible Slavic source for Romanian a privi 'to look'

forthc.

for their help in obtaining certain less easily accessible books and articles. The responsibility... more for their help in obtaining certain less easily accessible books and articles. The responsibility for all views is mine. 2 There are no grounds to consider the arbitrary Daco-Thracian etymology proposed by VINEREANU (2008: 677), where the pre-form is reconstructed as *pri-vili-, with subsequent palatalization and loss of *l, and associated with the PIE root *wel-'see ' (LIV₂: 675). This judgment applies to the methodological approach of the work in general, where Daco-Thracian etymologies are sought not only for words that may genuinely reach back to some kind of remote Balkan ancestry, but also -or perhaps first of all -for glaring Slavicisms. Cf., just to stay within the range of several pages from privi, the interpretation of Rom răstoacă 'backwater' as derived from PIE *reyǵ-'stretch, bind' rather than borrowed from SSl *rastoka or Rom praf 'dust' from PIE *pel-es-rather than from SSl *praxъ. 3 The other factor that needs to be reckoned with here is, of course, contamination (cf. LESCHBER 2010, 2012 for state-of-theart overviews and numerous examples). 4 It may also be noted that the verb displays many other meanings than the prototypical ones specified at the outset of the article (e.g. 'follow', 'believe', or even 'live'; cf. DLR VIII/5: 1521-1523); these, however, may be obtained far more easily from various shades of 'look, watch, regard, consider' than the other way round. Thus, it may be assumed that they represent Romanian-internal developments. Under certain etymologies it could be argued that some of these meanings preserve the original state, but this does not appear likely.

The Caland System as an explanatory device in Balto-Slavic etymology. A historical perspective

In: Jouravel, A. & Mathys, A. (eds.), Wort- und Formenvielfalt. Festschrift für Christoph Koch zum 80. Geburtstag, p. 251–272, 2021, 2021

The prehistory of the Proto-Slavic abstract suffix *-ostь and its Baltic congeners

Scando-Slavica (forthc.)

1 The problem PSl. 1 *-ostь is the parent form of a suffix immediately recognizable to anyone fam... more 1 The problem PSl. 1 *-ostь is the parent form of a suffix immediately recognizable to anyone familiar with a modern Sl. language. Reflexes of *-ostь productively derive abstract nouns from adjectives across the family (Ru. -ost , Ukr. -ist , Pol. -ość, Cz. -ost, Sln. -ost, Bulg. -ost etc.); 2 some examples are provided in Table 1. Table 1. Sample derivatives in *-ostь in the modern Sl. languages adjective abstract noun PSl. projection Ru. smelyj 'bold' → smelost ́ 'boldness' *sъmělъ → *sъmělostь -i Ukr. hidnyj 'dignified' → hidnist ́ 'dignity' *godьnъ → *godьnostь -i Pol. stary 'old' → starość 'old age' *starъ → *starostь -i Cz. hloupý 'stupid' → hloupost 'stupidity' *glupъ → *glupostь -i Sln. mlad 'young' → mladost 'youth' *moldъ → *moldostь -i Bulg. tvărd 'hard' → tvărdost 'hardness' *tvьrdъ → *tvьrdostь -i The suffix is, of course, well known also from early written corpora: OCS mǫdrъ 'wise' → mǫdrostь 'wisdom', ORu. radъ 'glad' → radostь 'gladness', etc. Although few of the forms presented in Table could actually be posited for PSl., there can be no doubt * A rudimentary version of the present theory was included in my PhD thesis (The Caland System in the North: Archaism and Innovation in Property-Concept/State Morphology in Balto-Slavic, Harvard University, 2017), written under the supervision of Jay H. Jasanoff. I am also grateful to the participants of those events at which the paper was presented in a more expanded form (14 th Annual Meeting of the Slavic Linguistics Society, Potsdam, 11-13 September 2019, as well as a guest talk at Vilnius University on

With all due respect, on Slavic abstracts in -y: The history of Proto-Slavic cěty ‘respect’ and some comparative notes on its congeners (ljuby ‘love’, cěly ‘healing, greeting’)

by Marek Majer and Rafał Szeptyński

Journal of Slavic Linguistics, 2021

The scarcely attested Proto-Slavic *cěty *-ъve ‘respect’ appears to be a mostly overlooked member... more The scarcely attested Proto-Slavic *cěty *-ъve ‘respect’ appears to be a mostly overlooked member of the small class of abstracts in *-y *-ъve; no precise accounts of the noun’s origin have been proposed so far. Two complementary approaches are put forth in the article: 1) inheritance from a PIE animate s-stem *keyt-ōs >> *koyt-ōs (paralleling a recent analysis of *ljuby ‘love’ < PIE *lewbh-ōs as well as its presumed secondary association with a verb in *-i-ti), or 2) inner-Slavic origin based on the formally similar *ljuby ‘love’ and *cěly ‘a healing’. The study also offers novel analyses—based on hitherto unexploited philological and lexicographical data—concerning various related issues (e.g. the status of PSl. nominal *cětъ, verbal *cětiti, and personal names in *Cěto/i-; the adposition *cětja; the semantic and pragmatic developments in *cěly ‘greeting, kiss’; the secondary rise of masculine *cělovъ/*cělyvъ ‘kiss’) with the purpose of integrating the entirety of the material concerning the root *cět- and the abstract type in *-y *-ъve into coherent pictures.

Slavic *ljuby and the heterogeneity of the inflectional class in *-y

Indogermanische Forschungen

The inherited Proto-Slavic feminine nouns in nom.sg *-y, oblique *-ъv-are conventionally thought ... more The inherited Proto-Slavic feminine nouns in nom.sg *-y, oblique *-ъv-are conventionally thought to go back to one source only, namely Proto-Indo-European stems in *-uh-. However, a handful of previous studies have already demonstrated the possibility that other inherited items could join this paradigm by phonologically developing a nom.sg in *-y from other sources (particularly *-ōs). Accordingly, in the present article, it is hypothesized that the well-known PSl. noun *ljuby *ljubъve F 'love' may go back to an amphikinetic s-stem *lewbʰ-ōs. Such a reconstruction makes it possible to explain a number of otherwise problematic features of the complex of *ljuby, such as: 1) the enigmatic OCS acc.sg ljuby (for expected ljubъvь) in the collocations (prě)ljuby (sъ)tvoriti or (prě)ljuby dějati 'commit adultery'; 2) the unexpected presence of *-sin the Balt. cognate, Lith. liaupsė̃ F 'praise, adoration'; 3) the occurrence of PSl. *-y as an abstract marker (a function in which PIE amphikinetic s-stems in *-ōs are securely, even if marginally, found, particularly in the domain of emotions and mental states-cf. Gr. αἰδώς αἰδόος F 'reverence, awe', ἔρως *ἔροος M 'love, desire', Ved. bhiyás-, Av. biiah-M 'fear'-while no such function can convincingly be reconstructed for a formation in *-u-h2).

The Indo-European prehistory of the Albanian word for 'sister'

proceedings of: Ljubilej – Ljubljäum – Ljubilee – Ljubljanniversaire. IG / SIES / SÉIE Arbeitstagung (100 years of comparative linguistics at the University of Ljubljana), Ljubljana, 4–7 June 2019

In the present article, which builds on an almost overlooked suggestion by Eric Hamp, it is argue... more In the present article, which builds on an almost overlooked suggestion by Eric Hamp, it is argued that Albanian motër 'sister' is not a direct reflex of Proto-Indo-European *meh₂ter- 'mother' (with an unexpected shift of meaning)-as assumed almost universally in the literature-but that it represents an early derivative of this word, probably *māt(e)rā < *meh₂t-(e)r-eh₂ 'of the [same] mother' or 'having the [same] mother'. Such an item may have been easily lexicalized as a replacement of the inherited word for 'sister'; cf. Greek αδελφ(ε)ός 'brother' < *sm ̥-gʷelbʰ-es-o-'of the same womb' or Spanish hermano 'brother' < Latin germānus 'of [one's own] kin, real, full'. Thus, despite widespread claims to the contrary, Albanian 'sister' can probably be explained via fairly trivial developments and does not document any extraordinary facts about (Pre-)Albanian kinship terminology or social structure.

Parahistoria indoevropiane e fjalës shqipe për 'motrën'

Seminari Ndërkombëtar për Gjuhën, Letërsinë dhe Kulturën Shqiptare, 2019

Shq. motër (në disa dialekte dhe në gjuhën e vjetër edhe motrë, motërë) konsiderohet përgjithësis... more Shq. motër (në disa dialekte dhe në gjuhën e vjetër edhe motrë, motërë) konsiderohet përgjithësisht si refleks i drejtpërdrejtë i fjalës indoevropiane për ‘nënën’ – *meh₂ter- (khs. skt. mā́tṛ-, gr. mḗtēr, lat. māter, sll. mati etj.). Dallimi kuptimor shpjegohet ose si rezultat i një ndryshimi semantik të gjerë – ndoshta me disa hapa të ndërmjetëm (p.sh. ‘nënë’ > ‘çdo femër e afërt’ > ‘motër’) – ose si pasojë e ndryshimeve në strukturën e përgjithshme të terminologjisë së farefisnisë për shkak të zhvillimeve jashtëgjuhësore (p.sh. është supozuar se motra e madhe përmbushte rolin e nënës). Këto mendime shprehen në të gjithë fjalorët etimologjikë të shqipes (Meyer 1891: 287–288; Çabej 1976: 371; Huld 1984: 96; Demiraj 1997: 279; Orel 1998: 275; Topalli 2018: 1009) dhe të indoevropianishtes (p.sh. NIL: 457–460), si dhe në punimet që trajtojnë historinë e leksikut të shqipes nga perspektivat indoevropiane (p.sh. Barić 1955: 57–58, Tagliavini 1965: 118; Matzinger 2018: 1789).

Megjithatë, një ndryshim aq rrënjësor semantik nuk duket shumë i besueshëm; gjithashtu, nuk ka arsye për të supozuar një ristrukturim të përgjithshëm të terminologjisë së farefisnisë (meqenëse emërtimet e tjera të trashëguara të farefisnisë shfaqen pikërisht me kuptimet e tyre të pritura, p.sh. vjehërr < pie. *sweḱuro- [khs. skt. śváśura-, gr. hekurós, lat. socer, sl. svekrъ] ose nip < pie. *nepot- [khs. skt. nápāt-, gr. anepsiós, lat. nepōs, sl. netьjь]). Në fakt, autoritetet etimologjike më të kujdesshme (p.sh. Demiraj 1997) pranojnë që zhvillimi semantik i fjalës motër mbetet një mister i plotë.

Pra, në këtë studim shq. motër analizohet jo si refleks, por si derivat mjaft i vjetër i fjalës indoevropiane për ‘nënën’. Domethënë, kemi të bëjmë ose me një mbiemër pronor *meh₂t-(e)r-o- ‘i cili ka të njëjtën nënë’, ose me një mbiemër relacional *meh₂t-(e)r-o- ‘nga e njëjta nënë’, ‘i së njëjtës nënë’. Trajta femërore e një mbiemri të tillë (*meh₂t-(e)r-eh₂- > *māterā) mund të ketë qenë leksikalizuar si ‘gruaja që ka të njëjtën nënë’ / ‘gruaja që vjen nga e njëjta nënë’ dhe të ketë zëvendësuar fjalën e trashëguar për ‘motrën’ (pie. *swesor-). Zëvendësime të tilla leksikore janë shumë të përhapura në gjuhët indoevropiane, p.sh. gr. adelph(e)ós ‘vëlla’ < *sm̥-gʷelbʰ-es-o- ‘që vjen nga e njëjta mitër’, span. hermano ‘vëlla’ < lat. germānus ‘nga [i njëjti] fis/farë’ ← germen ‘farë’. Siç dihet mirë, edhe shq. vëlla mendohet të ketë origjinë të këtij lloji: *swe-lowdʰ-eh2- ‘[nga] fisi i vet’.

Me sa duket, një ide e ngjashme në lidhje me fjalën motër është paraqitur në disa vepra të Hamp-it (1986: 110; 1996: 140), por me disa dallime dhe pa një analizë të hollësishme të sfondit indoevropian. Duke ditur që vërejtjet e Hamp-it rreth kësaj fjale fatkeqësisht nuk janë përfshirë në asnjë fjalor etimologjik dhe mungojnë nga debati i përgjithshëm, punimi i pranishëm është një përpjekje për të ofruar një analizë të plotë dhe moderne.

[English] [NB: English version of the paper published separately]

Alb. motër (dialectal and Old Albanian also motrë, motërë) is universally considered to be a direct reflex of the Proto-Indo-European word for ‘mother’ – *meh₂ter- (Sanskr. mā́tr̥-, Gr. mḗtēr, Lat. māter, Slav. *mati *matere etc.). The difference in meaning is explained either as far-reaching semantic change – possibly including intermediate steps (e.g., ‘mother’ > ‘any female relative’ > ‘sister’) – or as the product of changes in kinship terminology caused by extra-linguistic developments (e.g., it has been assumed that the older sister fulfilled the role of the mother). However, such a radical semantic change does not appear likely; similarly, there is no reason to assume any general restructuring of kinship terminology, given that other inherited kinship terms surface precisely with their expected meaning (vjehërr ‘father-in-law’ < PIE *sweḱuro-, nip ‘grandson/nephew’ < PIE *nepot-). Thus, in the present study – which draws on a little-known idea by Eric Hamp – Alb. motër is analysed not as a reflex but as an early derivative of the IE word for ‘mother’. Namely, we are dealing either with a possessive adjective *meh₂t-(e)r-o- ‘having the [same] mother’ or with a relational adjective *meh₂t-(e)r-o- ‘of the [same] mother’. The feminine form of such an adjective (*meh₂t-(e)r-eh₂ > *māterā) may have been lexicalized as ‘woman having the same mother’ / ‘woman of the same mother’ and replaced the inherited word for ‘sister’ (PIE *swesor-). Such lexical replacements are widespread; cf. Greek adelph(e)ós ‘brother’ < *sm̥-gʷelbʰ-es-o- ‘of the same womb’ or Spanish hermano ‘brother’ < Lat. germānus ‘real, *from the [same] seed’.

Slavic 'i-stem adjectives' and their alleged inflection loss: The derivational prehistory and synchronic status of a category

proceedings of: Dérivation nominale et innovations dans les langues indo-européennes anciennes (Université de Rouen, 11–12 Oct 2018)

Pozostałości praindoeuropejskiego „Systemu Calanda” w języku prasłowiańskim – wybrane przykłady [Remnants of the Proto-Indo-European “Caland System” in Proto-Slavic – selected examples]

In: Jakubowicz, M. et al. (eds.), Etymologica slavica. Studia etymologiczne poświęcone prof. Franciszkowi Sławskiemu z okazji setnej rocznicy urodzin. Warszawa: Instytut Slawistyki PAN, p. 81–99, 2019

The ‘Caland system’ (CS) is a set of morphological rules governing the behaviour of many ‘prototy... more The ‘Caland system’ (CS) is a set of morphological rules governing the behaviour of many ‘prototypically adjectival’ (or ‘property-concept’) roots and their adjectival, nominal and verbal derivatives in PIE and several archaic IE languages (mostly Vedic, Greek and Latin). It consists in the productive use of a set of formants (e.g. adjectives in *-u- and *-ro-, comparatives in *-yos-, abstracts in *-es-), added directly to the root instead of a pre-existing adjectival stem. The present article reviews a few synchronically irregular Proto-Slavic forms, arguing that the observed morphological anomalies (e.g. the relationship between the positive and the comparative in adjectives like *deb-elъ ‘thick’, compar. *deb’-ьš-, or between the adj. *orvь-nъ ‘even’ and its old abstract *orvo *orvese), can be explained as (partially modified) CS inheritance, even if the roots themselves are lexical innovations.

The Slavic source of Albanian venit(em) 'wane', mahnit(em) 'astound' and certain other verbs

Contributions to the 22nd Annual Scientific Conference of the Association of Slavists (Polyslav), 2019

The shape of Albanian verbs borrowed from Slavic is normally determined by the Sl. infinitive (Sl... more The shape of Albanian verbs borrowed from Slavic is normally determined by the Sl. infinitive (Sl. *-ati → Alb. -at-; Sl. *-iti → Alb. -it-). Surprisingly, however, the element -it- also occurs in certain Alb. verbs whose Sl. source does not have an *-iti infinitive. Based on methods of external and internal reconstruction, it is argued that some such verbs (venit(et) ‘wane’, mahnit(et) ‘astound’) may have been borrowed from a Slavic dialect where the respective source verbs had an infinitive in *-nyti (*vę(d)nyti ‘wane’, *maxnyti ‘swing’; cf. e.g. Slovene veniti, mahniti) – rather than *-nǫti, as in most Slavic languages (*vę(d)nǫti, *maxnǫti ‘swing’; cf. e.g. Serbian venuti, mahnuti). Since Sl. *y normally yields i in loanwords into Alb., a Sl. infinitive in *-nyti may be expected to produce an Alb. verbal stem in -(n)it-. Subsequently, the type could have spread (through proportional analogy) to some other verbs borrowed from Slavic (e.g. pjetit ‘braid’, bodit ‘stab’). Since no Sl. varieties of the *-nyti type are currently found in the zones neighboring Alb., the theory might have interesting consequences for Balkan linguistic geography, perhaps suggesting a different past location of certain Sl. dialects (*-nyti type ranging further south) and/or of Albanian (further north).

Rindërtimi i morfologjisë së huazimeve sllave-shqiptare: rasti i shq. venit(em) dhe mahnit(em) [Reconstructing the morphology of Slavic-Albanian loanwords: The case of venit(em) and mahnit(em)]

Seminari Ndërkombëtar për Gjuhën, Letërsinë dhe Kulturën Shqiptare, 2018

Ndërsa disa prej huazimeve sllave në gjuhën shqipe janë relativisht të reja (pra burimi i tyre mu... more Ndërsa disa prej huazimeve sllave në gjuhën shqipe janë relativisht të reja (pra burimi i tyre mund të identifikohet lehtësisht me dialektet e njohura të serbishtes, të maqedonishtes etj.), disa të tjera janë mjaft të vjetra; duket se ato janë huazuar nga disa dialekte sllave që nuk janë të dëshmuara në mënyrë të drejtpërdrejtë.

Shumë prej huazimeve sllave janë folje. Ato ndahen në dy grupe kryesore: foljet që i përkasin tipit *‐ati në sllavishte i përgjigjen tipit ‐at(em) në shqipe, p.sh. shq. dumat ‘peshoj, mendoj’ ← sll. *dumati (khs. serb. dȕmati, rus. dúmat', pol. dumać), kurse foljet që i përkasin tipit *‐iti në sllavishte i përgjigjen tipit ‐it(em) në shqipe, p.sh. shq. vozit ← sll. *voziti ‘lëviz, çoj’ (khs. serb. vòziti, rus. vozít', pol. wozić).

Mirëpo, është më pak e qartë se përse tipi ‐it(em) në shqipe përmban gjithashtu disa huazime sllave që nuk kanë as paskajoren *‐iti as temën e kohës së tashme *‐i‐ në sllavishte. Dy shembuj të rëndësishëm janë folje me temën e kohës së tashme *‐ne‐ (klasa II në klasifikimin e foljeve sllave të A. LESKIEN‐it): shq. venit(em) ← sll. *vę(d)ne‐ (serb. vènuti vȅnē, rus. vjánut' vjánet, pol. więdnąć więdnie) dhe shq. mahnit(em) ← sll. *maxne‐ ‘tund, lëkund’ (serb. máhnuti mȃhnē, rus. maxnút' maxnët, pol. machnąć machnie). Siç mund të shihet, foljet e këtij tipi zakonisht kanë paskajoren *‐ǫti
në sllavishte (p.sh. *vę(d)nǫti, *maxnǫti), që bëhet rregullisht serb. ‐uti (p.sh. serb.vènuti, máhnuti). Nuk është zbuluar asnjë arsye e qartë pse sll. *vę(d)ne‐ (serb. vènuti) dhe *maxne‐ (serb. máhnuti) janë bashkuar me tipin ‐it(em) në shqipe – në vend se të ndertojnë një tip hipotetik të ri **‐ut(em), i cili do të mbështej mirë nga leksiku i trashëguar (khs. p.sh. ngut(em)).

Pra, këtu argumentohet se shqipja ka huazuar të dyja foljet nga një dialekt sllav i tipit *‐nyti, domethënë një dialekt në të cilin paskajorja ka pasur formën *vę(d)nyti, *maxnyti etj. dhe jo *vę(d)nǫti, *maxnǫti etj. Për këtë arsye, format shqipe janë venitem, mahnitem (zanores *y të elementeve sllave i përgjigjet në shqipe rregullisht zanorja i, khs. shq. pihatem ‘këputem, lodhem shumë’ ← sll. *pyxati). Ky konkluzion është shumë interesant, aq më tepër se në kohën tonë në zonën e kontakteve gjuhësore sllave-shqiptare nuk më njihet asnjë dialekt sllav i tipit *‐nyti (dialektet e këtij tipi më të afërta njihen nga Kroacia veriore). Kjo do të thotë se hipoteza jonë – edhe pse ajo duhet të mbetet e pasigurt për shkak të mungesës së të dhënave të mëtejshme – mund të krahasohet me gjeografinë dialektore dhe me historinë e migrimeve sllave në Ballkan (shpërndarja e dialekteve të tipit *‐nǫti dhe *‐nyti është një faktor i rëndësishëm, khs. ANDERSEN 1999).

[English] [NB: English version of the paper published separately]

While some of the Slavic loanwords in Albanian are recent borrowings of items familiar from the neighboring Slavic dialects, certain others are evidently of considerable age and probably borrowed from Slavic varieties that are not directly attested.

As far as the verbal material is concerned, many Slavic borrowings enter the Albanian verbal class in -at (mid./pass. -atem). These are, in principle, the Slavic verbs in *-ati (cf. Alb. dumat ‘think out’ ← Sl. *dumati: Bosnian/Croatian/Serbian dȕmati, Russian dúmat', Polish dumać). Another big class of Sl. loanwords are the verbs in -it (mid./pass. -item). These are also clear as long as they represent Sl. verbs in *-iti (cf. Alb. vozit ‘row, drive’ ← Sl. *voziti: Bosnian/Croatian/Serbian vòziti, Russian vozít', Polish wozić).

However, it is less obvious why the Albanian class in -it(em) also includes some Slavic loanwords that exhibit neither an infinitive in *-iti nor a present stem in *-i- in Slavic. Here belong, most notably, two Slavic verbs with a present stem in *-ne- (Leskien’s class II): Alb.venit ‘cause to wane’ (usually mid./pass. venitem ‘wane, wither’) ← Sl. *vę(d)ne- (B/C/S vènuti vȅnē, Ru. vjánut' vjánet, Pol. więdnąć więdnie) and Alb. mahnit ‘astound, stupify’ (mid./pass. mahnitem ‘be astounded’) ← Sl. *maxne- ‘swing, brandish’ (B/C/S máhnuti mȃhnē, Ru. maxnút' maxnët, Pol. machnąć machnie). As can be seen, these verbs generally display an infinitive in *-ǫti, cf. *vę(d)nǫti, *maxnǫti, which regularly becomes B/C/S -uti, cf. vènuti, máhnuti. No clear reason for Slavic *vę(d)nǫ and *maxnǫ ending up in the Albanian class in -it(em) – rather than a hypothetical class in **-ut(em), which would be well supported by native Albanian lexicon, cf. ngut(em) ‘hurry’ – has ever been identified.

Accordingly, it is argued here that Albanian might have borrowed the relevant verbs from a *-nyti dialect, i.e. one in which the infinitives had the form *vę(d)nyti and *maxnyti rather than*vę(d)nǫti and *maxnǫti (the reflex of South Slavic *y as i is regular in Albanian, cf. e.g. Alb.pihatem ‘be exhausted’ ← *pyxati). This is especially interesting in view of the fact that no *-nyti dialects are currently known from the geographic area neighboring Albanian, so that the hypothesis ‒ although it must remain tentative in view of the scanty material – can be compared with the dialectgeography and settlement history of the South Slavic area (including the distribution of the *-nǫti and *-nyti dialects as studied in ANDERSEN 1999).

Review of: Slavic Nominal Word Formation. Proto-Indo-European origins and historical development, R. Matasović, 2014

Journal of Slavic Linguistics, 2017

review of: Slavic Nominal Word Formation. Proto-Indo-European origins and historical development,... more

The Development of the Bulgarian Literary Language: From Incunabula to First Grammars, Late Fifteenth – Early Seventeenth Century, transl. M. Majer, K. Gucio. Lanham – Boulder – New York – London 2021 (Studies in Slavic, Baltic, and Eastern European Languages and Cultures, ISBN: 978-1-4985-8607-8)

by Ivan N. Petrov and Marek Majer

Ivan N. Petrov’s "The Development of the Bulgarian Literary Language: From Incunabula to First Gr... more Ivan N. Petrov’s "The Development of the Bulgarian Literary Language: From Incunabula to First Grammars, Late Fifteenth–Early Seventeenth Century" examines the history of the first printed Cyrillic books and their role in the development of the Bulgarian literary language. In the literary culture of the Southern Slavs, especially the Bulgarians, the period that began at the end of the fifteenth century and covered the sixteenth and seventeenth centuries is often seen as a foreshadowing of the pre-national era of modern times. In particular, the centuries-old manuscript tradition was gradually replaced by the Cyrillic printed book, which—after the incunabula of Krakow and Montenegro—was published in such centers as Târgoviște, Prague, Venice, Serbian monasteries, Vilnius, Moscow, Zabłudów, Lviv, Ostroh, and many others. Petrov shows how the study of old Slavic prints is closely linked to the processes that determined the emergence of modern literary languages in the Slavia Orthodoxa area, including the influence of the liturgical Church Slavonic language shared by the Orthodox Slavs, which was increasingly standardized and codified at that time. The perspective of a language historian brings new light to the complex and multidimensional issues of this important transitional period of Slavic history and culture.

Полабското ją 'e' (Polabian ją 'is')

Talk delivered at: Пролетен лингвистичен семинар, Институтът за български език „Проф. Любомир Андрейчин“ , 2024

The 3. sg. pres. of the copula in Polabian has the hitherto unexplained form ją, with a nasal vow... more The 3. sg. pres. of the copula in Polabian has the hitherto unexplained form ją, with a nasal vowel. The form cannot reflect Common Slavic *je(stь). We assume that the form is in fact an intrusion from the pronominal system and continues the old nom. sg. masc. of the relative pronoun *jь- (cf. OCS i[že] ‘which’). It is known that in Polabian, as in other West Slavic languages, the nom. sg. masc. of certain pronouns has been extended with a nasal element (cf. OCS tъ ‘this’, but Polish, Czech ten, Upper Sorbian tón etc. < *tъ-nъ). In Polabian, a nasal vowel develops in this context (cf. tǫ ‘this’ < *tъ-nъ). Similarly, the pronoun *jь could be extended to *jь-nъ (cf. Cz jen[ž] ‘which’), which is expected to yield precisely Plb ją. From the typological point of view, the use of pronominal forms in the function of the copula is well-attested cross-linguistically, including in Slavic (cf. Russian X èto Y, Polish X to Y ‘X is Y’).

Sound change versus analogy in some irregular Slavic comparatives

by Marek Majer and Rafał Szeptyński

Lautwandel und morphologische Analogie (Arbeitstagung der Indogermanischen Gesellschaft), Cologne, 12–14 September 2023, 2023

In most Slavic languages, comparatives of many adjectives are derived via a suffix -š-, added dir... more In most Slavic languages, comparatives of many adjectives are derived via a suffix -š-, added directly to the root and often accompanied by a yod-mutation of its final consonant (e.g. z > ž, s > š). This is the reflex of *-jĭš-, one of the allomorphs of the Common Slavic CPV suffix. However, several dialect areas of Slavic attest comparatives with an entirely unexpected alternation z, s > k, unparalleled in the respective standards. Cf., for the adjectives CSl *blizъkъ ‘close’, *nizъkъ ‘low’, and *vysokъ ‘high’:

Slk blíz-k-y : dial. CPV blik-š-í ‘close’
Ukr blyz'-k-yj : dial. CPV blyk-š-yj

Slk níz-k-y : dial. CPV nik-š-í ‘low’
Ukr nyz'-k-yj : dial. CPV nyk-š-yj

Slk vys-ok-ý : dial. CPV vyk-ší ‘high’
Ukr vys-ok-yj : dial. CPV vyk-š-yj
Sln, BCMS vis-ok-(i) : dial. CPV vik-š-i

Since, in these examples, the -š- of the CPV suffix would clash with an expected root-final mutated -š (or -ž devoiced to -š), it is likely that the rise of these peculiar forms is connected with the dispreference for the arising geminate -šš- and its simplification (which would obscure the morphological composition (although this state is tolerated in some systems: Cz [vɪs-ok-iː] ‘tall’ : [vɪʃ-iː], cf. etymological spelling <vys-ok-ý> : <vyš-š-í>.). However, the source of the innovative alternations remains unknown and has rarely been investigated. So far, the phenomenon has only been referred to vaguely as due to dissimilation or resuffixation (for Slk: Benedek 1983; for Ukr: AUM 1, Hryščenko 1978) – both difficult to substantiate. Furthermore, the facts of the respective languages have been described in isolation from each other.

We consider the material in its entirety, based on in-situ queries of dialectal data sourced from (mostly unpublished) materials of the Slavic Linguistic Atlas (OLA) as well as local, language-specific atlases. We offer an alternative, principled explanation. Note that root-final -š- arose via the
yod-mutation in the CPV also when the root of the adjective ended in the velars -x- (e.g. CSl *sux-ŭ ‘dry’ : CPV *suš-ĭ š-) and -g- (CSl *dorg-ŭ ‘dear’ : CPV *dorž-ĭ š-, later -žš- > -šš-). Such adjectives could analogically reinstate the velar from the POS form, replacing the inherited -šš- with analogical (-xš-, -gš- >) -kš-. As long as these inherited and analogical variants coexisted, it was possible for adjectives with inherited s ~ š or z ~ ž alternations to develop such variants too. Taking dialectal Slovak as an example:

‘dear’ CSl *dorg- CPV *dorž-ĭ š-
Slk dial drah- CPV draš-š- (inherited variant)
Slk dial drah- CPV drak-š- (variant analogical to POS, with [hš] > [kš])

‘high’ CSl *vys- CPV *vyš-ĭ š-
Slk dial vys- CPV vyš-š- (inherited variant)
Slk dial X X = vyk-š-

I.e., the synchronic competition of variant forms such as drak-š- beside inherited draš-š- may have led to the emergence of an analogical vyk-š- as a variant of inherited vyš-š-. This explanation can, mutatis mutandis, be applied to the other adjectives in Slk, as well as to the situations in Ukr, Sln and BCMS (cf. the existence of Ukr dial dorokšyj beside dorožšyj [-šš-] ‘dearer’ as a possible model for vykšyj beside inherited vyššyj; etc.).

The plausibility of this approach is supported by an outlier Polish dialectal CPV form vyr-š-y from vys-okj-i ‘high’. It may, in turn, owe its existence vis-à-vis inherited vyš-š-y to the situation in the adjective star-y ‘old’: here, the two competing variants were star-š-y and sta-š-y (the two variants may be considered competing reflexes of earlier *stař-š-y < post-CSl *star’-ĭ š- (<< CSl *star-ě-jьš-). Cf. standard Pol starszy, along with, e.g., Cz starší, Ukr staršyj), the synchronic coexistence of which may have spawned vyr-š-y as a variant of vyš-y < vyš-š-y. Note that all four areas – Pol, Slk, Ukr, Sln/BCMS – must have conducted their innovations independently.

The process is interesing inasmuch as it does not conform to the classical pattern of four-part analogy involving two pairs of forms differing in grammatical function, morphological class, etc. To consider the Slk situation schemed above, the POS stems drah- and vys-ok- remain without any tangent point and cannot be juxstaposed directly. Rather, it is the synchronic competition of two variant forms of the CPV of drah- that enabled the creation of a competitor for the inherited CPV of vys-ok-. Thus, the mechanism bears resemblance to the proposed type of diachronic change known as ‘product-oriented innovation’ (Bybee & Slobin 1982 etc.; previous applications to Slavic languages cf. Wandl 2020), although it can still be formalized as a proportional analogy. The scenario devised for the irregular Slavic comparatives invites considering the question whether (at least some) other processes previously explained as product-oriented innovations could not be accounted for in a similar way, especially if dialectal and substandard data are taken into account.

‘Suffixless’ comparatives across Slavic: Typology, origins, implications

by Marek Majer and Rafał Szeptyński

18th Annual Meeting of the Slavic Linguistics Society, Bratislava, 24–26 August 2023, 2023

All Sl languages have a grammaticalized comparative degree. Its structure may be either 1) analyt... more All Sl languages have a grammaticalized comparative degree. Its structure may be either 1) analytic (Ukr bil’š važlyvyj ‘more important’; Mac poubav ‘more beautiful’) or synthetic, i.e. 2) additive-affixal (Pol tward-y ‘hard’ : tward-sz-y ‘harder’; Cz mil-ý ‘nice’ : mil-ejš-í), 3) substitutive-affixal (Slk šir-ok-ý ‘wide’ : šir-š-í; Sln glob-ok-(i) ‘deep’ : glob-lj-i) or 4) suffixless (BCMS mlad-(i) ‘young’ : mlađ-i; USorb wus-k-i ‘narrow’ : wuš-i).

It has been noticed in recent research (Majer & Szeptyński, 2021) that while types 1) and 2) are easily accomodated by the oft-quoted cross-linguistic generalizations/typologies of comparative forms (Jakobson, 1965; Dressler, 1986; Bobaljik, 2012 etc.), type 3) is unusual in that the comparative is not more complex – and sometimes less so – than the positive. The widespread attestation of type 3) in Slavic – as well as in some other branches of Indo-European, though primarily in their earlier stages – requires the adoption of a less strict typology than commonly assumed.

We observe that type 4) is just as problematic in this regard, if not more. However, surface suffixlessness may be interpreted in various ways; the morphological makeup of such forms is not always evident. A comprehensive and nuanced approach is required.

We attempt to fill two gaps in the existing scholarship, namely: i) to register all Sl languages and dialectal areas in which such suffixless comparatives are found, and ii) to arrange them by types and offer a basic analysis. For i), apart from standard publications, we also rely on extensive in-situ queries of dialectal data sourced from (mostly unpublished) materials of the Slavic Linguistic Atlas (OLA) as well as local, language-specific atlases. This guarantees an unprecedently high degree of coverage, at least regarding the most common adjectives.

For ii), our finding is that Slavic suffixless comparatives fall into three types, arranged here in a rising order of typological significance: a) phonetically conditioned, b) phonologically conditioned, and c) morphologized. Type a) refers to cases where the surface realization of the form may be explained by synchronically active phonetic rules: Cz [vɪs-ok-iː] ‘tall’ : [vɪʃ-iː] (cf. etymological spelling <vys-ok-ý> : <vyš-š-í>); Cz dial [mokr-iː] ‘wet’ : [mokr̝̊-iː] (/mokr̝- ʃiː/). The underlying structure may thus easily be substitutive or additive. Far rarer but more interesting is type b), apparently limited to cases where the root-final consonant in the cpv allomorph is -ž- and the -š- of the cpv suffix is ‘consumed’ by it: Cz dial draž-y (lit. draž-š-í) ‘more expensive’, Slk dial ňiž-í (lit. niž-š-í) ‘lower’, Cash dial vąž-ï (lit vąž-š-ï) ‘narrower’, in a way Ru dial mlaž-ij (lit. mlad-š-ij) ‘younger’. Progressive assimilation /žš/ > [ž] is never found in these systems outside such comparatives and therefore cannot be phonetic; still, at a deeper level, a suffixal analysis is possible. In type c) we include cases where such an analysis is clearly dispreferred (beside the well-known BCMS type mlađ-i: also Sln dial mlaj-i ‘younger’, Ukr dial šyr’-yj ‘wider’, molodž-yj ‘younger’). Finally, we inspect the diachronic rise of these types. We accept the standard view that type c) originates from the reanalysis and generalization of the CSl nom.sg allomorph (*molď-ьjь) at the cost of the obl allomorph (*molď-ьš-, generalized in most of Sl), but we observe that some degree of interaction with types a) and b) was likewise crucial here.

Splošnoslovansko *minǫti ali *minuti? Nov pogled na rekonstrukcijo pripone glagolov II. vrste

by Marek Majer and Rafał Szeptyński

Slovenski dialektološki posvet (SDP4), Inštitut za slovenski jezik Frana Ramovša, ZRC SAZU Ljubljana - 22.9.2022, 2022

It is generally acknowledged that the CSl class II (type OCS drъznetъ 3SG ‘dare’) continues PIE v... more It is generally acknowledged that the CSl class II (type OCS drъznetъ 3SG ‘dare’) continues PIE verbs in *-new-/-nu- and (more elusively) *-neh-/-nh̥-. However, the form of the suffix as found in the CSl infinitive/aorist stem is famously problematic. Traditionally, the suffix was reconstructed as CSl *-nǫ- (OCS drъznǫti INF ‘dare’); more recent research (early on Trubačev 1968: 374; at length Andersen 1999) often also operates with a by-form *-ny-, surviving in Sln, dialectal B/C/M/S, and possibly the westernmost parts of WSl. As universally admitted, neither variant can descend straightforwardly from the presumed source, PIE *-new-. There is no consensus on the origin of the rarer *-ny-, whereas *-nǫ- is usually explained as due to the optional nasalization (cf. *nuditi > *nǫditi ‘compel’) of the expected CSl *-n(j)u-, itself considered unattested except indirectly in the prevocalic allomorph *-nov- of the passive participle.

We draw attention to the hitherto unappreciated fact that traces of CSl *-nu- apparently survive directly. Evidence is necessarily limited, as large swaths of Sl conflate *u with *ǫ. Still, at least one language, in both its historical forms and modern dialects, attests genuine reflexes of *-nu- in a straightforward way, namely Slovene. The relevant data from 16th-c. Protestant prints as well as Upper Carniolan and Carinthian dialects, virtually limited to the single verb minuti ‘pass’, have already been discussed in the literature (Ramovš 1922: 5; 1936: 179). Besides clarifying the crucial issues raised by Ramovš that affect the diachronic interpretation of those facts, we point to the previously unnoticed regular, systemic reflexes of *-nu- preserved in the variety of the Carinthian dialect spoken in the Jaun Valley in Austria and the Resian dialect spoken in Italy.

Thus, by examining both early (Boduèn-de-Kurtenè 1875) and modern accounts (Steenwijk 1992), we challenge the widely-held view that the Resian data reflect *-nǫ- and thus support its archaic status. Additionally, we comment on instances of -nu- found in OCS writings (Vaillant 1948: 45).

Our identification of CSl *-nu- may simply be integrated into the traditional explanations, in fact corroborating them as a missing link – the direct non-prevocalic reflex of PIE *-new-. However, the coexistence of as many as 3 allomorphs (*-nǫ-, *-ny-, *-nu-) is highly unexpected, posing an even greater challenge for a credible reconstruction. While we agree that the primary source of *-nǫ- must lie in the nasalization of *-nu-, we speculate whether some elements of the complex situation might not also have arisen in the original (PIE imperfect →) CSl aorist paradigm due to the operation of Auslautgesetze involving various personal endings. Sequences like 1SG *-new-m, *-neh₂-m, 2SG *-new-s, *-neh₂-s and 3SG *-new-t, *-neh₂-t would have yielded a mix of * nu, *-ny (from *-neh₂-s), *-nǫ, and *-na (possibly also other variants, e.g. *-novь); subsequent generalizations would have led to the attested situation. Potential direct evidence for *-na- might in fact also be cited (chiefly from Croatian and Slovene), although we determine it unusable.

What could the Indo-Europeans compare with? The status of PIE comparative adjectives

by Marek Majer and Rafał Szeptyński

Proto-Indo-European Reconstruction: Problems, possibilities and new perspectives (Cambridge University, 1-2 October 2021), 2021

This paper deals with the morphological relations between positive and comparative adjectives rec... more This paper deals with the morphological relations between positive and comparative adjectives reconstructible for Core PIE. It is largely accepted—even if often tacitly—that at a certain early stage of PIE the only way in which the comparative could be derived from the positive was via suffix substitution; this is the so-called primary comparative of the type CPV *su̯eh₂d-i̯ os- ← POS *su̯eh₂d-u- ‘sweet’. On the other hand, in typological literature, it is often stated that comparative forms are morphologically marked with an additional affix in relation to positive adjectives, as in familiar structures of the type sweet-er ← sweet (cf. Jakobson 1965, Haspelmath 2008). Thus, as far as majority opinions are considered, the modern PIE reconstruction is at odds with a typological tendency. While there is no doubt that the latter claim in its radical versions (cf. Dressler 1986) is untenable, the question concerning the typological status of the PIE reconstruction remains open.
The starting point for our analysis is a restrained version of the typological generalisation, viz.: if a language has special comparative forms, marking with an additional affix (or periphrasis) is a productive means of their derivation from positive adjectives. This approach accounts for the peculiarities of the documented IE systems, which—having inherited the atypical primary relation—are far from adhering to it exclusively due to various innovations, cf. exx. (i) – (iii):

(i) emergence of the so-called secondary comparative in *-ter-o- in Indo-Iranian and Greek, which took over the role of the productive formation, cf. Gr. POS μωρός ‘silly’, CPV μωρότερος
(ii) simplification of the morphological structure of positive adjectives by merging thematic suffixes (*-o-, *-u-, *-i- etc.) with desinences; since the comparative remained a transparently suffixed formation, its substitutive character was thus replaced with an additive one, cf. PGmc. POS *þunn-uz ‘thin’ [R-E], CPV *þunn-iz-ō̄̄̄ [R-S-E] << PIE POS *tn̥h₂-u-s [R-S-E], CPV *tenh₂-yōs-∅ [R-S-E]
(iii) decomposition of some suffixes (*-ro-, *-no-, *-mo- etc.) and the absorption of their first segments by adjectival roots, parallel to (ii), cf. Lat. POS ruber ‘red’ (PIE *h₁rudʰ-ro-), CPV rubr-ior ‘redder’

By reviewing the typological tendency in question cross-linguistically—both within and outside the IE family—we aim at demonstrating that the reconstruction of the substitutive model as the only one for the formation of the PIE comparative lacks parallels and should therefore be avoided. Thus, we draw attention to three features which possibly distinguished the protolanguage from the early IE idioms. Firstly, the original prototypical positive of gradable adjectives was potentially a root formation, vestiges of which can be identified in Indo-Iranian, cf., e.g., Ved. jū́- ‘quick’ → jávīyas- (Rau 2009); note that here the structure of the comparative would have been purely additive and thus unremarkable typologically. Secondly, the comparative formant could retain its complex nature with the first element—of contested origin (cf. Rau 2014, Szeptyński 2018)—being of a substitutive character and the second—i.e. uncontroversially the nominal suffix *-s-— realising the additive model. Thirdly, it has been pointed out in the literature (e.g. Puhvel 1973, Sihler 1995) that the comparative may not have been fully grammaticalised in the protolanguage yet; it is important to stress that the comparative function of a genetically uniform formation could spread by diffusion. The latter view fits in with the recent hypothesis (Szeptyński 2018) that the comparative is an areal feature characteristic of Western and Central Eurasia as well as North Africa, propagated relatively late through language contact; PIE might, in fact, have been the source or early participant in this process.