This paper tackles the syntactic properties of the plural marker men in Mandarin Chinese. We argue that men is a modifier (non-inflectional plural marker in the sense of Wiltschko 2008) adjoined to the n head, where men is interpreted as an additive plural marker. We propose that men will move in the two contexts: First, when attaching to common nouns, men agrees with the D head, giving rise to definite construals, and men undergoes an n-to-D movement if D hosts a pronoun. Second, when attaching to proper names, men moves along with the proper name in a pied-piping manner. In both cases, men ends up adjoining D, where men is interpreted associatively. Additionally, we discuss cases in which proper names and pronouns co-occur and the role of men in the derivation, using an Agree and labeling approach.
The elicitation of Information Focus (IF) remains a central challenge in linguistic research, particularly in crosslinguistic and multilingual settings. The standard method, the Question-Answer (QA) test, while widely employed, suffers from well-documented limitations. Specifically, in QA-based elicitation, fragment answers often obscure the syntactic and prosodic realization of IF (Merchant 2004; Krifka 2006), and even full responses may yield interpretive ambiguities, encoding contrastive or corrective readings rather than purely informational ones, thus reducing the reliability of the collected data. To address these methodological issues, Cruschina & Mayol (2022, 2024) propose the Question with a Delayed Answer (QDA) test, an adaptation of the Discourse Completion Task (DCT) used in prosodic research on Romance languages. By inserting intervening material between the wh-question and the response, the QDA encourages the spontaneous production of full, congruent sentences, thereby yielding syntactically transparent data. Building on this methodological innovation, the present paper investigates how IF interacts with argument structure in two pro-drop Romance languages, Italian and Spanish. Focusing on transitive and unaccusative verbs, we show that QDA-elicited data reveal systematic crosslinguistic differences: while Italian exhibits stronger constraints on postverbal subjects under IF, Spanish allows more flexibility. These findings have implications not only for the empirical description of Romance focus strategies but also for theoretical models of the syntax-information structure interface.
This study investigates the early parsing of subject–verb agreement in German-acquiring infants, focusing on whether morphosyntactic dependencies are parsed in real time when φ-feature morphology is the only available disambiguating cue. German offers a stringent test case due to the syncretic nature of the third-person pronoun sie (‘she’ / ‘they’) despite the language’s obligatory subject expression. Using a preferential-looking paradigm adapted from Gavarró & Keidel (2024), we tested 19 monolingual German-learning infants (17.5–22 months) on sentences containing either lexically specified Determiner phrases or ambiguous pronominal subjects. Eye-tracking results revealed that infants reliably identified referents based on subject–verb agreement morphology in both full-DP and pronominal conditions, although effects were more robust with overt subjects. These findings challenge claims of a production–comprehension asymmetry and support models that posit early syntactic competence, including the computation of agreement, not contingent on input frequency or semantic salience, but reflecting universal, structure-sensitive operations.
This introduction provides an overview of current research on partitivity. The contributions in this volume collectively address several persistent challenges in the analysis of partitivity, such as (i) the semantic underspecification of partitive markers, which are not always transparently associated with the part-whole relation; (ii) the connection between partitive markers and indefiniteness and negation; (iii) the diachronic development of partitive markers and their relation to other categories, such as determiners and prepositions; and (iv) the syntactic structure of binominal constructions. Overall, we hope that these contributions advance our understanding of the many empirical and theoretical problems that partitivity poses for linguistic theory.
This introduction provides an overview of current research on partitivity. The contributions in this volume collectively address several persistent challenges in the analysis of partitivity, such as (i) the semantic underspecification of partitive markers, which are not always transparently associated with the part–whole relation; (ii) the connection between partitive markers and indefiniteness and negation; (iii) the diachronic development of partitive markers and their relation to other categories, such as determiners and prepositions; and (iv) the syntactic structure of binominal constructions. Overall, we hope that these contributions advance our understanding of the many empirical and theoretical problems that partitivity poses for linguistic theory.
The main goal of this paper is to provide a unified analysis of de nominals in French and noncanonical genitive nominals in Russian. Unlike previous literature that relies on partitivity, quantification and case assignment, we argue that a unified formal account can be substantiated by guaranteeing a uniform syntactic analysis of weak indefinite expressions not based on quantification, a semantic dependency of de nominals and non-canonical genitives to non-veridicality, and different linearization operations at Spell-out that guarantee concatenation of words or subwords (French de*n, Russian n⊕gen).
In this contribution, I present new evidence against any attempt to unify pseudo-partitive/quantificational and qualitative binominals (e.g., un grupo/montón de estudiantes ‘a group/lot of students’, una mierda de departamento ‘a shit of an apartment’). The observation is that pseudo-partitive/ quantificational nouns, unlike expressive nouns, license NP-ellipsis of the de-phrase (e.g., A: Vinieron estudiantes? ‘Did students come?’, B: Un montón (de estudiantes) ‘a lot (of students)’). Assuming that the codas of pseudo-partitive and quantificational nouns are NPs realized as primeval genitives (Pesetsky 2013), NP-ellipsis of the usual type is correctly predicted under wellknown licensing and identity conditions. In turn, the de-phrase in qualitative binominals is the specifier of an Equative Phrase, which relates the de-NP to an empty nominal ranging over properties (Saab 2022a). As I show, NP-ellipsis is banned in such environments for principled reasons. Thus, the ellipsis pattern is accounted for if the relevant binominals have a different structural source for genitive marking.
In this review article we discuss the main arguments in support of the hypothesis that in Contemporary Catalan and Italian nominal ne is neither a prepositional clitic, nor a partitive clitic nor a quantitative clitic. While the item from which Romance ne originates (Latin inde) pronominalized PPs with locative meanings, similarly to what happens with de, following a grammaticalization cline, ne comes to pronominalize different layers of nominal structure, and in contemporary Romance also expresses non-specific indefiniteness. We maintain that, syntactically, whereas partitivity is characterized as being a birelational category that necessarily connects a subset to a maximal superset, the indefinite use of ne is not birelational in the same sense. We support the hypothesis that nominal ne is a property-type anaphora that resumes a non-specific weak indefinite nominal expression. We review a number of syntactic and semantic arguments in support of the non-partitivity of nominal ne in Catalan and Italian, from which we conclude that it is a pro-NP, unmarked for gender and number, whose antecedent cannot have a unique or maximal reference.
Grounded in the principle of Antihomophony (Kayne 2019a), several grammatical/functional morphemes, previously considered to have different properties according to their distribution, are examined. A more abstract and fine-grained analysis of the contexts in which these morphemes appear confirms that they are the same element in all constructions. After reviewing elements such as there, will, Italian feminine marker -a or Spanish plural marker -s, the so-called partitive preposition of is argued to be the same element as the one in other constructions. Therefore, Antihomophony is a powerful principle that simplifies the functional lexicon.
Previous work shows that there are prosodic differences between agreeing and disagreeing responses to positive and negative polar questions (Goodhue & Wagner 2018; González-Fuentes et al. 2015, among others). In this study, we focus on the prosodic properties of responses to negative statements in Spanish, a research question that has received little attention and which has not been explored in any detail. That is, we study what characterizes prosodically the following types of responses to negative statements: A: Los profesores no corrigieron los ejercicios. ‘The teachers didn’t correct the exercises.’; B1: {Sí, No}, los profesores no corrigieron los ejercicios. ‘{Yes, No}, the teachers didn’t correct the exercises.’ [agreeing response]; B2: {Sí, No}, los profesores sí corrigieron los ejercicios. ‘{Yes, No}, the teachers did correct the exercises.’ [disagreeing response]. There are two noteworthy findings. On the one hand, there is a Fall-Rise contour of the response particle (RespP, sí / no), which is observed almost exclusively in disagreeing responses. On the other hand, within the sentence following the response particle, disagreeing responses display four characteristics: (a) a longer duration of the Polarity Particle (PolP); (b) a bigger fall in F0 between the PolP and the verb; (c) a slightly higher F0 of the PolP with respect to the end of the subject; and (d) a longer duration at the end of the subject (with the RespP No). Overall, the PolP in disagreeing responses displays a substantial degree of prosodic prominence, compared to responses in agreement. Also, our results indicate that the RespP does not have an effect on the variables we investigated.
This paper argues against the LF-Copy VP-ellipsis model in Oku (1998) and for a PF-deletion approach. Simplex (SE) anaphor data from Thai is used to show that long-distance person blocking patterns are incompatible with elided content being copied in at the interface, post-agreement. This analysis adopts Reuland’s (2011) φ-feature agreement approach to SE-anaphor valuation, with Giblin’s (2015) modifications for long-distance anaphors (LDAs). It is shown that positions where Reuland’s account predicts valuation via φ-feature agreement do not licence Vehicle Change (VC), the phenomenon where interpretations of elided elements would require different anaphora in overt counterparts. In LDA blocking contexts it can be seen that SE-anaphors must be valued independently of each other. This interaction in syntax between overt and elided material is only possible under a PF deletion analysis. This has consequences for analyses of ellipsis and reconstruction, as well as the more general division of labour between interfaces.
This article argues that A'-dependencies can be accounted for in terms of Form Copy (Chomsky et al. 2023). To make this argument, we draw on evidence from the distribution and interpretation of resumptive pronouns from Kallulli (2009, 2012), and we also show how the account could be extended to other phenomena, notably Left Dislocation in English. Specifically, we make the following four claims. Firstly, A'-dependencies are reducible to Form Copy. Second, we follow Kallulli’s (2009, 2012) proposal that resumption involves a pronoun in a concealed relative clause, which we model here in terms of Form Copy. Third, we show how this analysis can be extended to English Left Dislocation, bringing this construction into line with Kallulli’s analysis of resumption. Fourth, an important corollary of our approach is that Form Copy allows partial PF-copies; it relates syntactic objects, which do not have to be fully identical in formal features, still less in terms of segmental phonology. This is true not only for A'-dependencies but also for A-dependencies. The theoretical contribution of this article is thus our extension of the scope of Form Copy to A'-dependencies.
In this paper I propose that Spanish passive imperatives with the passive reflexive pronoun SE have all the semantic and formal properties of imperative sentences in spite of the fact that their subject cannot be coreferent with the addressee, since it is generally a non-human object which is interpreted as a patient. I propose that the special and intriguing properties of passive imperatives come from the fact that they perform a deferred directive speech act. An analysis in terms of Speech Act Phrase is provided. The passive imperative sentence merges as an open expression, since the Addressee argument does not receive a referential index, but it is a free variable whose value will be fixed later on. Passive morphology allows a formal subject to merge whose reference is not bound to the addressee. The analysis provides an explanation to the fact that passive imperatives typically occur in written texts, advices, and instructions, i.e., the type of texts that allow differed speech acts.
This work describes and compares a relatively new construction in their Italian and English variants, namely not-NOM-that constructions and not-ACC-ing constructions respectively: Non io che perdo le chiavi di casa per la seconda volta quest’anno / Not me losing my housekeys twice this year. Despite the presence of a negative marker, these sentences retain an affirmative meaning, expressing irony, sarcasm, or self-deprecation (rephrased as ‘It is ironic and surprising that I lost my housekeys twice this year’). They are predominantly used by younger generations on social media and in informal speech. We compare the English construction with its Italian counterpart, focusing on their syntax. More specifically, we provide an in-depth structural analysis of English not-ACC-ing constructions claiming that they contain a focalized Acc-ing construction (in the sense of Lowe 2019) and Expletive Negation (following Greco 2020). We also claim that the main structural and semantic properties of Italian not-NOM-that constructions also derive from: (i) the presence of a high negation merged in the CP layer, (ii) the focalization of a whole chunk of structure.
Sluicing has been classically taken as a construction that involves embedded wh-questions. However, in the last two decades some studies have shown that it may also occur in some kinds of relative clauses, namely in Free relatives. In the current article, we present novel data that permit to distinguish Relative Sluicing from Standard Sluicing in European Portuguese. This paper is a first attempt to characterise Relative Sluicing in European Portuguese, comparing this language with Brazilian Portuguese, but also with Hungarian and Gungbe, and to evaluate the capacity of the current syntactic proposals to account for Relative Sluicing in this language. The data show that Relative Sluicing involves Free relatives which occur in a nominal context, as claimed by previous analyses, that the fronted constituent is a wh-phrase, not a head, and suggest that a focus-based approach is an adequate way to deal with Standard and Relative Sluicing in European Portuguese.
In Spanish, feminine singular nouns beginning with a stressed /a/, such as arma ‘weapon’ and asa ‘handle’, do not take the standard article forms la and una, but el and un. This paper examines the non-standard uses of el and un observed in morphologically complex words, such as prefixed nouns (el microalga) and diminutives (un alita). The hypothesis is that what makes these variants possible is the adjunct status of prefixes and diminutive suffixes. It is claimed that adjuncts do not obligatorily trigger the reassignment of standard forms la and una, like a head would, because they can be late-merged.
The present study reviews manifestations of vocatives in a wide range of typologically related and unrelated languages and identifies three major patterns: particles, suffixes and non-concatenative forms such as intonation contours. It is shown that regardless of their shape vocatives systematically encode not-at-issue meaning, particularly the speaker’s evaluation of the type of social relationship with the addressee and the perceived physical distance. It is demonstrated that optional vocative markers are most efficiently treated as modifiers of speech acts, whereas mandatory vocative morphemes function as nominal modifiers that mark the noun as incompatible with argument positions and only appropriate for the speech act type address.
Truncation of proper names (e.g. English Robert > Rob) has been claimed to be a marginal or extragrammatical word-formation process due to the presumed unpredictability and irregularity of its outputs (Dressler & Merlini-Barbaresi 1994 et seq.). In this study, the hypothesis is pursued that the coexistence of multiple truncation patterns in a single language is in some cases due to language change: older patterns coexist with more recent ones, thus creating the false impression of random variability. This hypothesis is tested in an experiment where participants are asked to choose the most appropriate short name for a person of a certain age group in a dialogue. The experiment was conducted over the crowdsourcing platform AlpiLinK (https://alpilink.it/) for speakers of Romance and Germanic varieties of Northern Italy. The results of the experiment show that the truncation patterns tested for Italian and German names differ significantly in the age that participants associate with them. A second finding of this study is that diachronically close truncation patterns differ minimally in the grammars that correspond to each pattern in a formal model of truncation implemented as a typological system in Optimality Theory (Alber & Arndt-Lappe 2022). Minimal grammatical change defined inside the typological system has been hypothesized to be a characteristic feature of language change (Alber & Meneguzzo 2016; Alber & Kokkelmans 2022; Apostolopoulou 2022), a hypothesis confirmed by the results of the experiment in this study. Name truncation therefore is characterized by structured, rather than random variability: truncation patterns can be ordered by age and conform to an expected path of (minimal) diachronic change.
We examine the derivation of intransitive verbs from adjectives in Hungarian. Two phonologically unrelated suffixes are added to adjectival stems in almost complementary distribution: -ul~ül and -od~ed~öd-. We show that this suffixation abides with various phonological and morphological constraints in addition to being lexically conservative: the stem of the intransitive verb must occur also as the stem of its transitive counterpart.
This paper examines the morphology and phonology of hypocoristic formation in Gujarati, an Indo-Aryan language, within the framework of Optimality Theory and applying the Theory of Prosodic Morphology (McCarthy & Prince 1986). The intention of the speaker using a hypocoristic to address someone can range from affection to annoyance to even contempt or anger (Desai 1994). Consequently, there are different types of hypocoristic forms found in the language. Gujarati hypocoristics are formed through the truncation of a personal name, attaching a suffix to a personal name, or by attaching a suffix to a truncated personal name. These can be classified according to the suffix attached and the extent of truncation involved. The data shows that the hypocoristic formation process satisfies the Prosodic Morphology Hypothesis (McCarthy & Prince 1990) as the base is always a prosodic unit: a heavy syllable (CVC) or a prosodic word, thereby meeting the Template Satisfaction Condition.
