In Romanian, the definite article cannot be overtly realized if the maximal nominal projection co... more In Romanian, the definite article cannot be overtly realized if the maximal nominal projection contains only the lexical N and occurs in the complement position of (most) accusative-taking prepositions. Arguing that a definite D is present in the underlying syntactic representation, I describe article drop as a PF-phenomenon. I follow Dobrovie-Sorin's (2007) idea that article drop is conditioned by a complex head formation operation that assigns a X 0-status to DPs of the form [D+def N 0 ], but I show that her proposal of extending complex head formation to P is contradicted by the behavior of article drop in coordination. Therefore, I propose a different explanation for the limitation of article drop to the complement of P: based on the fact that article drop also occurs with prepositional case markers (which are K heads rather than Ps), I propose that article drop only occurs when D lacks Case. Underlying this analysis is a novel theory of case marking in Romanian: I propose that inflectional marking involves null Ks that trigger spreading of a Case feature to their complement (following Norris 2014, 2018), whereas prepositional marking involves an overt K and no feature spreading. Prepositions that trigger article drop are assimilated to K heads Isogloss 2024, 10(2)/3 Ion Giurgea 2 in that they take a DP, rather than a KP, complement, playing K's role of closing-off the nominal extended projection.
The goal of this paper is to reconcile the definite marking with the indefinite-like semantics of... more The goal of this paper is to reconcile the definite marking with the indefinite-like semantics of those superlatives that take a relative/comparative reading. Following Szabolcsi (1986) and Heim (1999), we will assume that the difference between absolute and relative readings of superlatives is represented at the syntactic level of semantic representation, LF (Logical Form). We will however depart from Heim’s hypothesis that what raises at LF is the superlative operator itself (EST). We will instead assume a quantificational-determiner analysis of EST, which involves two raising operations at LF: EST-raising to Spec,DP and raising of the whole superlative DP (Determiner Phrase) to a scope position in the sentence. We will examine the relative readings of quality superlatives. The generalizations and the proposed analysis are not assumed to extend to quantity superlatives.
I argue that a certain intonational pattern found in Romanian polar questions, which has been qua... more I argue that a certain intonational pattern found in Romanian polar questions, which has been qualified as neutral in previous studies, actually marks verum focus. This pattern consists of a L* nuclear accent on the verb and a HL% boundary tone, with the H part associated to the last stressed syllable, and the L part truncated when this syllable is the very last in the sentence. This pattern also characterizes early constituent focus (including focus on the verb, which is realized in the same way as verum focus). The neutral intonation is different, consisting of a L* nuclear accent on the last stressed constituent followed by a H% tone. This pattern is identical to the one found with narrow focus on the last constituent, as expected given the general prosodic properties of the language.
We describe the syntax and morphology of the genitival construction of Romanian, and we insist on... more We describe the syntax and morphology of the genitival construction of Romanian, and we insist on the syntactic analysis of some peculiar properties of these constructions: the alternation between genitives introduced by the agreeing particle al and those without al , the status of this agreeing particle, the similarity in distribution between genitives and agreeing possessors, and the alternation between agreeing and genitival forms in the paradigm of pronominal possessors. We argue that the absence of al after the suffixal definite article is the result of a PF-deletion rule. We conclude that al is probably ambiguous between an agreeing case-marker and a [pro-N+D] complex, and we show that this peculiar distribution results from the history of al (which goes back to a definite article). We argue that agreeing possessors – so-called “possessive adjectives” – are pronouns rather than adjectives, which is particularly clear in Romanian given their genitival distribution. The alternation between agreeing and genitive-marked pronominal possessors is shown to obey a morphological constraint, which disallows pronominal roots in combination with more than one set of inflectional phi-features. This explains why agreeing forms are only found where the inherent phi-features of the pronoun are marked on the root, whereas 3rd person pronouns, which mark the inherent phi-features on the inflectional morpheme, have genitive forms.
The widespread pattern of “noun-less” variants of DPs (e.g. two/these [nP/NumPØ] vs. two/these ca... more The widespread pattern of “noun-less” variants of DPs (e.g. two/these [nP/NumPØ] vs. two/these cars) might lead to the expectation that the definite determiner should be able to head DPs with a null complement. The best candidate for representing the structure [Ddef [nP/NumPØ]] are 3rd person personal pronouns (Elbourne, 2005; a.o), an analysis supported by evidence for NP-ellipsis (or N-anaphora) in pronouns. However, equating pronouns with THE+[nP/NumPØ] faces a number of issues: the different behavior of pronouns and other definite descriptions with respect to binding; the differences in syntactic and semantic features between pronouns and DPs headed by THE (e.g. gender); the one-to-many structure-form relationship between the single structure THE+[nP/NumPØ] and different series of pronominal forms (strong and weak forms); and, a gap in the attested combinations of THE and [NØ]. Our account addresses these issues. We propose that the D used in pronouns, which we label Dpron, has the semantics of THE, but THE and Dpron differ in their formal features. We posit that a null NP complement needs to be licensed by features on the determiner the NP merges with, and that Dpron bears such features. Thus, we capture the fact that typical THE does not surface in cases where there is no overt element in the complement of D (i.e. what we call total emptiness) via the absence of these features on THE. Additionally, to further explore the differences between strong and weak forms in Romanian, we present the results of a corpus study on these forms.
It has long been recognized that sentences with passive se obey a Person constraint:<br> th... more It has long been recognized that sentences with passive se obey a Person constraint:<br> the subject cannot be 1st or 2nd person. I discuss a further constraint on the subject,<br> manifest in Romanian: not only 1st or 2nd person pronouns, but all those DPs that<br> must be marked by the prepositional object marker accompanied by clitic-doubling<br> when functioning as direct objects are excluded from being subjects of se-passives.<br> Following Richards (2008), I propose that these DPs, which are high on the Person/<br> Animacy scale, have a Person feature (manifested by clitic-doubling when they are<br> case-licensed by v*), whereas those that can occur as subjects of se-passives lack the<br> Person feature completely. The ban on +Person internal arguments in se-passives is<br> due to the intervention of the Person feature associated with the external argument.<br> I argue that the element saturating the external argument...
I discuss various theories of sentence topics against the background of attested examples of topi... more I discuss various theories of sentence topics against the background of attested examples of topicalization in Romanian. I argue that non-referential topical phrases constitute a problem for the concept of aboutness topic proposed by Reinhart (1981) and can better be accounted for in a theory where topicalized constituents are indicators of the discourse topic (von Fintel 1994). I examine various types of topicalized constituents, classified according to the relation they entertain with the previous sentences and the discourse topic. Furthermore, I discuss another function of sentence topics, which can be characterized independently of discourse relations: that of anchoring the new information in the common ground. I suggest a way in which this function can be subsumed under the general function of indicating discourse-topic dynamics.
This paper investigates non-finite verbal constituents that qualify as modal nonfinite relatives ... more This paper investigates non-finite verbal constituents that qualify as modal nonfinite relatives in Romanian, French, and Italian. We show that these constructions are reduced relatives based on the relativization of the object. The analysis extends to reduced relatives in ...
I discuss the semantics of a sub-type of focus fronting in Romanian, which is neither associated ... more I discuss the semantics of a sub-type of focus fronting in Romanian, which is neither associated to focal particles nor mirative or exclamative. I show that this type is not "contrastive" in the sense of involving a closed set of contextually identifiable alternatives, as has been claimed in previous studies, and does not necessarily have a corrective import. The conditions this type is subject to are: (i) the fronted constituent is a true focus (in the sense of an element introducing alternatives in the interpretation, see Rooth 1992, Krifka 2008) and not just new information; (ii) an exhaustivity presupposition or implicature (described as rejection of all other alternatives) and an existential presupposition (one alternative is true) are introduced, in a way similar to English it-clefts, as discussed by Büring and Križ (2013). I further present some differences with respect to English clefts. Finally, I discuss the status of the negation that precedes a fronted focus, a...
<p>The geographical varieties of Romanian spoken in Romania, the Republic of Moldova, and a... more <p>The geographical varieties of Romanian spoken in Romania, the Republic of Moldova, and adjacent regions are largely mutually intelligible. More important are the differences between these varieties (known as "Dacoromanian") and the South-Danubian varieties of Aromanian, Meglenoromanian, and Istroromanian, which have been separated from (Daco-)Romanian for a very long time, but qualify as dialects of Romanian from a historical and comparative Romance perspective.</p> <p>Standard Romanian is based on the southern dialect of Dacoromanian, in particular the variety of Muntenia, but also includes features taken from other dialects (e.g., the 3<sc>pl</sc> imperfect -<italic>au</italic>, the absence of "iotacism" in verb forms—<italic>văd</italic> instead of the etymological <italic>vă(d)z</italic> 'see.1<sc>sg' &lt; L</sc>at. *<sc>uidi̯o</sc> <italic>&lt;</italic> <sc>uideō,</sc> with the regular sound change <italic>-di̯-&gt;-dz-&gt;-z-</italic>). A unified standard language was established around the middle of the 19th century.</p> <p>Some of the differences between the high and the colloquial register of standard Romanian are due to innovations characterizing southern varieties: the demonstrative system (high register <italic>acest(a), acel(a)</italic> versus colloquial <italic>ăsta, ăla</italic>), the future (high register <italic>voi</italic> [inflected] + infinitive versus colloquial <italic>o</italic> [uninflected] + subjunctive), the use of the infinitive (more restricted in the colloquial register than in the high register), and the presumptive mood (mostly colloquial, representing a modal epistemic specialization of a future form <italic>oi</italic> + infinitive, which is itself an innovation with respect to <italic>voi</italic> + infinitive).</p> <p>Some of the features by which substandard varieties differ from the standard language represent innovations: the replacement of the inflectional dative and genitive by prepositional constructions, the change of the relative pronoun <italic>care</italic> into a complementizer, and the loss of the number contrast in the 3rd person of verbs (the latter representing a recent development, mostly found in the southern varieties, but also in parts of Crişana and Transylvania). The loss of agreement with the possessee on the genitival article <italic>al</italic> is an innovation that first appeared in the northern dialect and subsequently gained ground across substandard varieties.</p> <p>Northern varieties, especially in peripheral areas (Crişana, Maramureş, northern Moldova), preserve a number of archaic features that disappeared from the standard language, for example, the productivity of verb-clitic word orders (with both auxiliary and pronominal clitics), the use of <italic>al</italic>-Genitive-N word orders, the conditional periphrases <italic>vream</italic> + infinitive and <italic>reaş</italic> + infinitive (the latter in Banat), and, as a widespread phenomenon, the 3<sc>sg=3pl</sc> homonymy in the perfect auxiliary (in the form <italic>o</italic> &lt; <italic>au</italic>). Compared to the colloquial standard language, northern varieties preserve the infinitive better. An innovative feature characteristic of northern varieties is the use of periphrastic forms for the imperfect and pluperfect.</p> <p>As conservative features found in some nonstandard southern varieties, we may cite the use of the synthetic perfect (which in the standard language is restricted to the written register) and the stress on the oblique determiner/pronominal endings (<italic>ăstúia</italic> vs. <italic>ắstuia</italic>).</p>
The Romance pro-N clitic (Fr., Cat.) en / (It.) ne is problematic for most theories of cliticizat... more The Romance pro-N clitic (Fr., Cat.) en / (It.) ne is problematic for most theories of cliticization, because some phenomena support a movement analysis, while others seem to speak against movement. Endorsing the view that the pro-N clitic is the result of movement, this paper argues that the problems of this analysis can be solved by assuming that this clitic does not represent an N or NP, but rather a functional head intermediate between D/Q and N (probably Num), whose complement is empty after the DP-phase is completed. Consequently, I propose that the theory of cliticization should be modified in order to allow "dynamic minimality": only minimal elements may cliticize, but non-minimal elements may become minimal in the course of the derivation, if their complement has a null spell-out.
In Romanian, the definite article cannot be overtly realized if the maximal nominal projection co... more In Romanian, the definite article cannot be overtly realized if the maximal nominal projection contains only the lexical N and occurs in the complement position of (most) accusative-taking prepositions. Arguing that a definite D is present in the underlying syntactic representation, I describe article drop as a PF-phenomenon. I follow Dobrovie-Sorin's (2007) idea that article drop is conditioned by a complex head formation operation that assigns a X 0-status to DPs of the form [D+def N 0 ], but I show that her proposal of extending complex head formation to P is contradicted by the behavior of article drop in coordination. Therefore, I propose a different explanation for the limitation of article drop to the complement of P: based on the fact that article drop also occurs with prepositional case markers (which are K heads rather than Ps), I propose that article drop only occurs when D lacks Case. Underlying this analysis is a novel theory of case marking in Romanian: I propose that inflectional marking involves null Ks that trigger spreading of a Case feature to their complement (following Norris 2014, 2018), whereas prepositional marking involves an overt K and no feature spreading. Prepositions that trigger article drop are assimilated to K heads Isogloss 2024, 10(2)/3 Ion Giurgea 2 in that they take a DP, rather than a KP, complement, playing K's role of closing-off the nominal extended projection.
The goal of this paper is to reconcile the definite marking with the indefinite-like semantics of... more The goal of this paper is to reconcile the definite marking with the indefinite-like semantics of those superlatives that take a relative/comparative reading. Following Szabolcsi (1986) and Heim (1999), we will assume that the difference between absolute and relative readings of superlatives is represented at the syntactic level of semantic representation, LF (Logical Form). We will however depart from Heim’s hypothesis that what raises at LF is the superlative operator itself (EST). We will instead assume a quantificational-determiner analysis of EST, which involves two raising operations at LF: EST-raising to Spec,DP and raising of the whole superlative DP (Determiner Phrase) to a scope position in the sentence. We will examine the relative readings of quality superlatives. The generalizations and the proposed analysis are not assumed to extend to quantity superlatives.
I argue that a certain intonational pattern found in Romanian polar questions, which has been qua... more I argue that a certain intonational pattern found in Romanian polar questions, which has been qualified as neutral in previous studies, actually marks verum focus. This pattern consists of a L* nuclear accent on the verb and a HL% boundary tone, with the H part associated to the last stressed syllable, and the L part truncated when this syllable is the very last in the sentence. This pattern also characterizes early constituent focus (including focus on the verb, which is realized in the same way as verum focus). The neutral intonation is different, consisting of a L* nuclear accent on the last stressed constituent followed by a H% tone. This pattern is identical to the one found with narrow focus on the last constituent, as expected given the general prosodic properties of the language.
We describe the syntax and morphology of the genitival construction of Romanian, and we insist on... more We describe the syntax and morphology of the genitival construction of Romanian, and we insist on the syntactic analysis of some peculiar properties of these constructions: the alternation between genitives introduced by the agreeing particle al and those without al , the status of this agreeing particle, the similarity in distribution between genitives and agreeing possessors, and the alternation between agreeing and genitival forms in the paradigm of pronominal possessors. We argue that the absence of al after the suffixal definite article is the result of a PF-deletion rule. We conclude that al is probably ambiguous between an agreeing case-marker and a [pro-N+D] complex, and we show that this peculiar distribution results from the history of al (which goes back to a definite article). We argue that agreeing possessors – so-called “possessive adjectives” – are pronouns rather than adjectives, which is particularly clear in Romanian given their genitival distribution. The alternation between agreeing and genitive-marked pronominal possessors is shown to obey a morphological constraint, which disallows pronominal roots in combination with more than one set of inflectional phi-features. This explains why agreeing forms are only found where the inherent phi-features of the pronoun are marked on the root, whereas 3rd person pronouns, which mark the inherent phi-features on the inflectional morpheme, have genitive forms.
The widespread pattern of “noun-less” variants of DPs (e.g. two/these [nP/NumPØ] vs. two/these ca... more The widespread pattern of “noun-less” variants of DPs (e.g. two/these [nP/NumPØ] vs. two/these cars) might lead to the expectation that the definite determiner should be able to head DPs with a null complement. The best candidate for representing the structure [Ddef [nP/NumPØ]] are 3rd person personal pronouns (Elbourne, 2005; a.o), an analysis supported by evidence for NP-ellipsis (or N-anaphora) in pronouns. However, equating pronouns with THE+[nP/NumPØ] faces a number of issues: the different behavior of pronouns and other definite descriptions with respect to binding; the differences in syntactic and semantic features between pronouns and DPs headed by THE (e.g. gender); the one-to-many structure-form relationship between the single structure THE+[nP/NumPØ] and different series of pronominal forms (strong and weak forms); and, a gap in the attested combinations of THE and [NØ]. Our account addresses these issues. We propose that the D used in pronouns, which we label Dpron, has the semantics of THE, but THE and Dpron differ in their formal features. We posit that a null NP complement needs to be licensed by features on the determiner the NP merges with, and that Dpron bears such features. Thus, we capture the fact that typical THE does not surface in cases where there is no overt element in the complement of D (i.e. what we call total emptiness) via the absence of these features on THE. Additionally, to further explore the differences between strong and weak forms in Romanian, we present the results of a corpus study on these forms.
It has long been recognized that sentences with passive se obey a Person constraint:<br> th... more It has long been recognized that sentences with passive se obey a Person constraint:<br> the subject cannot be 1st or 2nd person. I discuss a further constraint on the subject,<br> manifest in Romanian: not only 1st or 2nd person pronouns, but all those DPs that<br> must be marked by the prepositional object marker accompanied by clitic-doubling<br> when functioning as direct objects are excluded from being subjects of se-passives.<br> Following Richards (2008), I propose that these DPs, which are high on the Person/<br> Animacy scale, have a Person feature (manifested by clitic-doubling when they are<br> case-licensed by v*), whereas those that can occur as subjects of se-passives lack the<br> Person feature completely. The ban on +Person internal arguments in se-passives is<br> due to the intervention of the Person feature associated with the external argument.<br> I argue that the element saturating the external argument...
I discuss various theories of sentence topics against the background of attested examples of topi... more I discuss various theories of sentence topics against the background of attested examples of topicalization in Romanian. I argue that non-referential topical phrases constitute a problem for the concept of aboutness topic proposed by Reinhart (1981) and can better be accounted for in a theory where topicalized constituents are indicators of the discourse topic (von Fintel 1994). I examine various types of topicalized constituents, classified according to the relation they entertain with the previous sentences and the discourse topic. Furthermore, I discuss another function of sentence topics, which can be characterized independently of discourse relations: that of anchoring the new information in the common ground. I suggest a way in which this function can be subsumed under the general function of indicating discourse-topic dynamics.
This paper investigates non-finite verbal constituents that qualify as modal nonfinite relatives ... more This paper investigates non-finite verbal constituents that qualify as modal nonfinite relatives in Romanian, French, and Italian. We show that these constructions are reduced relatives based on the relativization of the object. The analysis extends to reduced relatives in ...
I discuss the semantics of a sub-type of focus fronting in Romanian, which is neither associated ... more I discuss the semantics of a sub-type of focus fronting in Romanian, which is neither associated to focal particles nor mirative or exclamative. I show that this type is not "contrastive" in the sense of involving a closed set of contextually identifiable alternatives, as has been claimed in previous studies, and does not necessarily have a corrective import. The conditions this type is subject to are: (i) the fronted constituent is a true focus (in the sense of an element introducing alternatives in the interpretation, see Rooth 1992, Krifka 2008) and not just new information; (ii) an exhaustivity presupposition or implicature (described as rejection of all other alternatives) and an existential presupposition (one alternative is true) are introduced, in a way similar to English it-clefts, as discussed by Büring and Križ (2013). I further present some differences with respect to English clefts. Finally, I discuss the status of the negation that precedes a fronted focus, a...
<p>The geographical varieties of Romanian spoken in Romania, the Republic of Moldova, and a... more <p>The geographical varieties of Romanian spoken in Romania, the Republic of Moldova, and adjacent regions are largely mutually intelligible. More important are the differences between these varieties (known as "Dacoromanian") and the South-Danubian varieties of Aromanian, Meglenoromanian, and Istroromanian, which have been separated from (Daco-)Romanian for a very long time, but qualify as dialects of Romanian from a historical and comparative Romance perspective.</p> <p>Standard Romanian is based on the southern dialect of Dacoromanian, in particular the variety of Muntenia, but also includes features taken from other dialects (e.g., the 3<sc>pl</sc> imperfect -<italic>au</italic>, the absence of "iotacism" in verb forms—<italic>văd</italic> instead of the etymological <italic>vă(d)z</italic> 'see.1<sc>sg' &lt; L</sc>at. *<sc>uidi̯o</sc> <italic>&lt;</italic> <sc>uideō,</sc> with the regular sound change <italic>-di̯-&gt;-dz-&gt;-z-</italic>). A unified standard language was established around the middle of the 19th century.</p> <p>Some of the differences between the high and the colloquial register of standard Romanian are due to innovations characterizing southern varieties: the demonstrative system (high register <italic>acest(a), acel(a)</italic> versus colloquial <italic>ăsta, ăla</italic>), the future (high register <italic>voi</italic> [inflected] + infinitive versus colloquial <italic>o</italic> [uninflected] + subjunctive), the use of the infinitive (more restricted in the colloquial register than in the high register), and the presumptive mood (mostly colloquial, representing a modal epistemic specialization of a future form <italic>oi</italic> + infinitive, which is itself an innovation with respect to <italic>voi</italic> + infinitive).</p> <p>Some of the features by which substandard varieties differ from the standard language represent innovations: the replacement of the inflectional dative and genitive by prepositional constructions, the change of the relative pronoun <italic>care</italic> into a complementizer, and the loss of the number contrast in the 3rd person of verbs (the latter representing a recent development, mostly found in the southern varieties, but also in parts of Crişana and Transylvania). The loss of agreement with the possessee on the genitival article <italic>al</italic> is an innovation that first appeared in the northern dialect and subsequently gained ground across substandard varieties.</p> <p>Northern varieties, especially in peripheral areas (Crişana, Maramureş, northern Moldova), preserve a number of archaic features that disappeared from the standard language, for example, the productivity of verb-clitic word orders (with both auxiliary and pronominal clitics), the use of <italic>al</italic>-Genitive-N word orders, the conditional periphrases <italic>vream</italic> + infinitive and <italic>reaş</italic> + infinitive (the latter in Banat), and, as a widespread phenomenon, the 3<sc>sg=3pl</sc> homonymy in the perfect auxiliary (in the form <italic>o</italic> &lt; <italic>au</italic>). Compared to the colloquial standard language, northern varieties preserve the infinitive better. An innovative feature characteristic of northern varieties is the use of periphrastic forms for the imperfect and pluperfect.</p> <p>As conservative features found in some nonstandard southern varieties, we may cite the use of the synthetic perfect (which in the standard language is restricted to the written register) and the stress on the oblique determiner/pronominal endings (<italic>ăstúia</italic> vs. <italic>ắstuia</italic>).</p>
The Romance pro-N clitic (Fr., Cat.) en / (It.) ne is problematic for most theories of cliticizat... more The Romance pro-N clitic (Fr., Cat.) en / (It.) ne is problematic for most theories of cliticization, because some phenomena support a movement analysis, while others seem to speak against movement. Endorsing the view that the pro-N clitic is the result of movement, this paper argues that the problems of this analysis can be solved by assuming that this clitic does not represent an N or NP, but rather a functional head intermediate between D/Q and N (probably Num), whose complement is empty after the DP-phase is completed. Consequently, I propose that the theory of cliticization should be modified in order to allow "dynamic minimality": only minimal elements may cliticize, but non-minimal elements may become minimal in the course of the derivation, if their complement has a null spell-out.
Uploads
Papers by Ion Giurgea