Language Research

Abstract Clausal negation and post-syntactic ellipsis have been extensively investigated within the Generative framework, and Persian linguists have offered detailed analyses of both phenomena. The present study, however, examines a specific type of negation that has thus far remained largely unaddressed in the syntactic literature on Persian. Rather than employing the verbal negative prefix, this construction makes use of an independent negative particle to reverse the truth value of a proposition. This syntactic configuration, called discontinuous negation, is realized by the bipartite marker na … na (‘neither … nor’) and licenses the ellipsis of parallel constituents and heads in coordinate structures. Focusing on this construction and drawing a principled distinction between sentential negation and constituent negation, I first demonstrate that the head of Negation Phrase (NegP) in Persian, merged above TP, hosts the verbal negative prefix. I then argue, following Hedde Zeijlstra (2015, 2022), that the negative particle na is phrasal in nature: it constitutes an independent syntactic element realized in the specifier of NegP and is capable of negating a clause either on its own or in combination with its counterpart. Building on these assumptions, I analyze coordinate structures with a shared subject that occur in discontinuous negation contexts and exhibit deletion of part of the first or second conjunct. I identify four distinct syntactic patterns—distinct from canonical gapping constructions—in which the verb or its dependents are deleted from either conjunct. I show that in cases where the verb lacks the negative prefix, the derivation involves either TP-ellipsis or VP-ellipsis. By contrast, in constructions where the verb bears negation in the presence of the discontinuous na … na marker, the relevant phenomenon is best analyzed as constituent negation rather than clausal negation. Moreover, drawing on both empirical evidence and theoretical considerations, I argue that in the constructions under discussion, an Across-the-Board (ATB) movement mechanism forces identical subjects to raise to a single structural position. This movement leaves behind a gap in the structure that, at the level of surface representation, gives the appearance of identity-based ellipsis

Abstract In the non-lexicalist approach to grammar, all morphemes, whether lexical elements or function heads, are pieces that merge within the domain of syntax to form larger constituents. Consequently, what is referred to as an affix is essentially a morpho-syntactic category that occupies the head of a functional projection and selects another category as its complement (Embick 2015). The present study, adopting such an approach, analyzes the structure of the imperfective aspect in Persian and its morpho-syntactic interaction with the derivation of the perfective aspect. The imperfective aspect, one of the two grammatical aspects in Persian, is embodied in structures traditionally called progressive and continuous verbs. Like the perfective aspect, this verbal construction is used in both present and past tenses, though in contemporary Persian, the imperfective, unlike its perfective counterpart which appears in indicative, subjunctive, and sometimes imperative moods, solely indicates the indicative mood. Throughout this research, we argue that a lexicalist analysis and a feature-based theory, especially where the interaction of the prefix "mi-" with causative and perfective suffixes (“-ɑn” and “de/-te”, respectively) is concerned, fail to account for the certain inflectional forms of the verbs. Framed within non-lexicalist approach and more in the spirit of Distributed Morphology (DM) we explore the construction of the imperfective aspect in Persian and examine the interaction of its prefix with other verbal inflectional elements. In this process, we argue that in the hierarchy of functional projections of sentence in Persian, the perfective projection merges above the imperfective projection. In this structure, the head of PerfP which has an overt auxiliary verb dominates the head of ProgP which hosts a covert grammatical verb in Persian. We will see that this speculation is reinforced by empirical evidence of cliticization of the perfect auxiliary (“bɑš” in Persian and “have” in English) in the present perfect continous sentences.

Abstract The phenomenon of the head- and constituent ellipsis is one of the most discussed issues that grammarians have given a different name to each of its types. The present paper deals with a specific type of ellipsis within the clause, which is referred to, in the tradition of generative grammar, as the Right Node Raising, and it is distinct from the deletion of the second coordinate clause, namely gapping. In the aforementioned ellipsis, the verb alone, or together with its dependents, is removed from the first coordinate clause. While examining such structures, the present article analyzes the ellipsis from the coordinate clauses which contain a complement or adjunct small clause, and therefore, due to the simultaneous deletion of the primary predicate (from the main clause) and the secondary predicate (from the subordinate clause), they represent more syntactic complexity of its own. In the analysis, which is based on the assumption of two processes, namely the “verb movement out of the vP” and the “object shift”, the discussed structures are divided into two types: coordinates with a common subject and ones with a non-common subject; and thus, a different explanation is provided for each kind. In coordinates that have a common subject and host a small clause in their argument structure, in fact, two vPs, and not two clauses, are coordinated, and the process known as Across-the-Board movement, which is applied to the verb and the subject of the clause, leaves some gaps in the first coordinate clause in place. In contrary, the ellipsis phenomenon is applied to coordinates with non-common subjects. Since this phenomenon occurs at PF level, it is expected that its occurrence is related to the phase domain. In this research, based on the Phase Theory, as well as the empirical evidence of the Persian, we argue that in the latter structures, ellipsis occurs at the higher phase stage, and the elements that have already reached the edge of the phase are pronounced at the phonetic level, following the Phase Impenetrability Condition.

Abstract It is widely accepted by traditional grammarians and linguists that Persian verbal roots fall into two forms depending on their tense feature: present stem and past stem. In contrast to this traditional view, the present study considers an alternative notion of allomorphy, indicating that a few verbs use a basic form, called “morph”, in all morphosyntactic contexts while some other roots allow for having two or more allomorphs. Within the framework of Distributed Morphology, especially on the basis of underspecification condition, I will review the empirical evidence adduced in favor of the hypothesis that the tense feature has no role in producing and selecting the so-called present and past stems. In this approach, the systematic distribution of verbal allomorphs like /suz/ and /sux/ (derived from “ÖSUZ” = burn) is constrained by some other formal features, such as adjacency to the morphophoneme {D}, absorbing the causative suffix (“-ān” or its free variation “-un”) or appearing in the boundary of a word. Finally, to avoid terminological and empirical confusion, I will suggest that we should avoid the use of the term “stem” altogether and speak of root, morph and allomorphs only. It should be added that the different variations in phonetic shape of lexical roots are divided into two classes: the first one follows from the phonological rules and the second one is affected by the morphophonological (or readjustment) rules. Phonological changes are predictable and produce automatic alternations, but unpredictable morphophonological changes lead to non-automatic alternations. Following the general trend of DM, I restrict the term “allomorphy” to those cases in which the variation in phonetic shape of a morpheme does not follow from the automatic phonological rules of the language. This analysis raises the theoretical question of whether language-specific readjustment rules (i.e. the alteration of root or affix phonology) are needed for morphological theory. Rejecting readjustment rules, I show that allomorphs of verbal roots are independently stored in Vocabulary.

Abstract Secondary predicates are among the syntactic structures that have been neglected by the traditional Persian grammarians, but in the course of the last four decades of studies in generative grammar, almost all kinds of this structure have been explored. Secondary predicates are of two major kinds: the so-called depictive predicates and resultatives. A depictive predicate attributes a property to its subject which holds of that subject all the time that the event denoted by the matrix verb is going on. Resultatives attribute a property to their subject which holds of that subject at the point that the event denoted by the matrix verb ends or culminates. In the present study, following Rothstein (2006), I show that the Persian depictive and resultative predicates can be distinguished from each other just by semantic properties; they have no syntactic differences. Moreover, I will discuss that depictives can be subject- or object-oriented, while resultatives can only be object-oriented. Both subject- and object-oriented secondary predicates are adjunct small clause, the former adjoins to the vP and the latter to the VP (or more correctly, to the main PredP) and the argument modified by the second predicate, sideward moves from the adjunct clause to the main clause leaving a PRO in its base position.

Abstract The syntactic expression of sentential negation has given rise to a lot of research in most linguistic frameworks. During the last few decades of research in generative grammar, studies on negation focus on a range of topics, including the position of Negation Phrase. Indeed, not only do languages vary with respect to the form of negative elements, but the underlying position of these functional elements is also subject to cross-linguistic variation. Within the framework of Distributed Morphology (Halle & Marantz 1994), the present study investigates the position of NegP in Farsi. Contrary to the claim in the literature, it will be argued - both on theoretical and empirical grounds - that in this language NegP immediately dominates TP and verbal elements absorb the functional head of Negation Phrase via a post-syntactic process, called morphological merger. Moreover, the syntax of negation is connected to the phenomenon of negative polarity, i.e. items whose distribution is limited to a number of negative contexts. So, this article examines the conditions which license the occurrence of N-phrases (N-words) in indicative structures. Clearly, studying N-phrases provides more insight not only into the nature of such context-sensitive elements, but also into the syntactic nature and position of sentential negation itself.

Abstract Once DS is abandoned in the Minimalist framework and structure building is carried out by merge, multiple root syntactic objects in a single derivational step are not only allowed in principle, but must indeed be employed in some derivations in order for the extension condition to be satisfied. The crucial aspect of this theoretical model is that it permits instances of sideward movement, whereby a given constituent moves from a syntactic object K to an independent syntactic object L. Following Hornstein and Nunes (2014), the present study examines the sideward movement of the subject in the Persian past participle clauses assuming that these non-finite adverbial adjuncts hang below the vP of main clauses. As in these structures the subjects of the adjunct clauses are understood as being the same as the subjects of the matrix clauses, it will be argued that the head T in this language has a strong EPP feature that triggers movement of an overt or covert subject into its Specifier. Based on these assumptions, a DP has to move sideways from the subject position of the participle clause to merge with the main vP allowing the external theta-role of the matrix verb to be discharged. In the next step, the same DP moves upward to the Specifier of TP to satisfy the EPP feature of the main clause. Sideward movement also provides an escape hatch for the subject to have its case checked. Were this constituent not to move, the derivation would crash at LF as it could not get its uninterpretable case feature checked within the adjunct clause.

Abstract Whether or not there is a verbal passive construction in Persian has been ‎highly controversial. Traditional grammarians and most linguists have ‎argued that there is a structural passive construction in Persian, similar to ‎that observed in English. Within the framework of the Minimalist ‎Program, but more in the spirit of Emonds’s (2006) division of analytic ‎passives into verbal and adjectival, it will be argued that the Persian ‎so-called verbal passive construction is just an instance of adjectival ‎passive, or indeed, unaccusative CPr with a past participle serving as ‎its NV element. In line with what holds for verb movement in simplex ‎and complex predicates, it will be discussed in detail that Modern ‎Persian lacks verbal passive constructions and hence passive ‎projections, and in contrast to the grammatical verb budan ‘to be’ ‎which can be used either as a light verb in complex predicates or as a ‎perfect auxiliary in present/past perfect tenses, the grammatical verb ‎šodan ‘to become’, which is traditionally considered as a passive ‎auxiliary, is indeed a light verb hosted by an NV element. According ‎to this line of analysis, in Persian adjectival passives, the head [A-en] is ‎present in both LF and PF and the past participle is obligatorily ‎interpreted as an adjective.

Abstract Topicalization and focus movement have been discussed in literature with respect to a vriety of their properties. In Persian, syntactic objects within Tense Phrase, including phrasal arguments, may freely be displaced to a higher position and the reordering of these elements is mainly considered to be the result of discourse functional elements such as topic and focus. The present study is an effort to address the question of the formal motiviation of these movements, and their formalization within a feature-based system. Moreover, it tries to account for the surface position of subject in Persian, based on a VP shell analysis and the strength of Dfeature in this language. Thus one of the main contributions of this study is that the uninterpretable D-feature on the head of TP is strong in Persian and as a result, the structural subject has to move to the specifer position of its own clause.