Abstract This research aims to examine the grammaticalization process of the lexical verb raːħ ‘went’ in the Urban Jordanian dialect, and explore semantic differences and sociolinguistic factors (gender and age) that influence the use of raːħ variants. We used a forced-choice task to identify the dominant variants across demographics and then examined their semantic association with certainty and possibility. An acceptability judgment task was also employed, focusing on the influence of markers of certainty and probability on Urban Jordanians’ choices. It is shown that ħ - (the shortest variant of ra:ħ ) is dominant in certainty contexts, whereas raħ (the second shortest variant) is prevalent in probability contexts. Additionally, men and older speakers tend to favor the less reduced forms, particularly raːħ and raħ , while raːjiħ appears relatively more often among younger speakers. In contrast, women and younger speakers show a stronger preference for ħ- , the most grammaticalized variant, suggesting that women and younger speakers may be leading the ongoing language change. Overall, the variation is systematic, reflecting both semantic differentiation and socially conditioned change.
Abstract This paper draws on acceptability judgments and corpus data to investigate the alternation between infinitival (INF) and finite da -present complements (DPC) in Serbo-Croatian. We first provide experimental evidence that predictors suggested in the literature on the basis of limited empirical evidence do indeed shape complement choice: DPCs are more frequent with larger complements and root modality, while INFs are preferred with smaller complements and epistemic modality. Second, our corpus analysis shows that the previously reported effects of subject concreteness and animacy are better accounted for by subject theta role: DPC is favored by agentive subjects, INF by theme subjects. At the same time, however, we observe a notable mismatch between usage and acceptability: while speakers across the region rate both forms as acceptable to varying degrees, actual usage is sharply polarized, with DPCs prevalent in Serbian and nearly absent in Croatian. This pattern confirms the well-known finding that speakers’ intuitions preserve multiple grammatical options, while production data reveal how competing variants are resolved under communicative and normative pressure. But the discrepancy might also suggest that social meanings such as prestige, stigma, and identity, mediate production beyond grammar alone. We propose that INF vs. DPC constitutes a syntactic variable that carries social meaning, challenging claims in the sociolinguistic literature that syntax is socially inert, and we outline a research programme to identify and quantify such social meanings.
This research aims to examine the grammaticalization process of the lexical verb ra:(h) over bar 'went' in the Urban Jordanian dialect, and explore semantic differences and sociolinguistic factors (gender and age) that influence the use of ra:(h) over bar variants. We used a forced-choice task to identify the dominant variants across demographics and then examined their semantic association with certainty and possibility. An acceptability judgment task was also employed, focusing on the influence of markers of certainty and probability on Urban Jordanians' choices. It is shown that ra:(h) over bar- (the shortest variant of ra:(h) over bar) is dominant in certainty contexts, whereas rah (the second shortest variant) is prevalent in probability contexts. Additionally, men and older speakers tend to favor the less reduced forms, particularly ra (h) over bar and rah, while raji (h) over bar appears relatively more often among younger speakers. In contrast, women and younger speakers show a stronger preference for (h) over bar-, the most grammaticalized variant, suggesting that women and younger speakers may be leading the ongoing language change. Overall, the variation is systematic, reflecting both semantic differentiation and socially conditioned change.
This paper draws on acceptability judgments and corpus data to investigate the alternation between infinitival (INF) and finite da-present complements (DPC) in Serbo-Croatian. We first provide experimental evidence that predictors suggested in the literature on the basis of limited empirical evidence do indeed shape complement choice: DPCs are more frequent with larger complements and root modality, while INFs are preferred with smaller complements and epistemic modality. Second, our corpus analysis shows that the previously reported effects of subject concreteness and animacy are better accounted for by subject theta role: DPC is favored by agentive subjects, INF by theme subjects. At the same time, however, we observe a notable mismatch between usage and acceptability: while speakers across the region rate both forms as acceptable to varying degrees, actual usage is sharply polarized, with DPCs prevalent in Serbian and nearly absent in Croatian. This pattern confirms the well-known finding that speakers' intuitions preserve multiple grammatical options, while production data reveal how competing variants are resolved under communicative and normative pressure. But the discrepancy might also suggest that social meanings such as prestige, stigma, and identity, mediate production beyond grammar alone. We propose that INF vs. DPC constitutes a syntactic variable that carries social meaning, challenging claims in the sociolinguistic literature that syntax is socially inert, and we outline a research programme to identify and quantify such social meanings
This paper investigates how information-structural categories like focus can be linearized inside the DP. Starting from the claim that information-structural marking possibilities in the DP-structure often mirror marking possibilities in the CP-structure within the same language, we show that while this assumption holds for various European languages, it breaks down when Mabia languages are taken into account. Focusing on Likpakpaanl, we provide an example for a language, where, due to particular restrictions on information-structural marking, notions like focus cannot be marked at all inside the DP and require a clausal marking strategy.
The orders of noun (N) and adjective modified by an adpositional phrase (A PP) in the world's languages have been explained by the Head Final Filter (HFF) (Williams 1982) and the Final-over-Final Constraint (FOFC) (Biberauer et al. 2014, Sheehan 2017). Alexeyenko & Zeijlstra (2021) argue that the HFF can be violated in languages with rich agreement morphology, and propose the Phrase Continuity Requirement (PCR). However, there are some languages that have rich morphology but obey the HFF. As an alternative approach to the orders of N, A and PP, we generalize the HFF and the FOFC into a prosodic constraint, the Minimal Prosodic Boundaries constraint (MPB), which requires that a constituent be externalized with minimal boundaries between its sub-constituents. The variation in possible orders is correctly predicted by an additional rule, the Mapping of Left Bracket Sequence (MLBS), and by language-specific stress patterns. We further argue that our prosodic account can also explain the counterexamples to the HFF and the FOFC.
The literature on second position elements (a.k.a. second position clitics) is characterized by a longstanding discussion as to whether these elements are base-generated in second position or whether they are base-generated in first position and only arrive in second position as the result of a subsequent transformation. In this paper, I will provide two arguments for the transformational approach based on a typological study of second position coordinators (Weisser 2024), i.e. coordinators that appear inside of one of their conjuncts rather than in the semantically transparent peripheral position (e.g. in between the two conjuncts). The first argument will be based on the observation that, although the coordinators usually occupy a second position within their coordinands, they also show positional alternations in some syntactic contexts. These alternations, I argue, can be derived in both transformational and base-generation models. However, we find that the kinds of empirically attested alternations are systematically restricted in a way that falls out of a transformational approach but that cannot be captured in a base-generation approach. In other words, the base-generation approach massively overgenerates predicting much more alternations than we actually find. The second argument is based on opaque interactions between the second position placement and other morphosyntactic operations.
Two competing views regarding the syntax-semantics of imperatives can be distinguished. A minimal view assumes that imperatives are properties in semantics, and their syntax is rather bare, while the modal approach maintains that imperatives are propositions that contain a (covert) performative modal in syntax. This paper supports the modal approach from data patterns in two Chinese languages. It is observed that the performative modal particle can be overtly pronounced in imperatives in Chinese. I also examine the Imperative-and-Declarative (IaD) sentences, which have been employed to counter the modal approach. Confirming earlier generalizations, Chinese provides empirical evidence for the claim that there are two distinct types of IaDs. The possibility of an overt performative modal appearing in one but not the other allows for a clear distinction between genuine imperative antecedents and their imperative-like counterparts. Such distinction provides further empirical support for the modal analysis of imperatives.
Clitic placement with European Portuguese uninflected verbs is problematic for prosodic analyses. While 'proclisis triggers' give rise to obligatory proclisis with inflected verbs, they often lead to optional proclisis/enclisis with uninflected verbs (Raposo & Uriagereka 2008), except for aspectual/quantificational adverbials with head-like behaviour (Martins 2013a). These patterns provide support for a syntactic 'Anti-V2 requirement', whereby clitics usually raise to adjoin to T and either: (a) an XP raises to/through spec FinP and the verb stays in Asp (resulting in proclisis); or (b) the verb raises through T to Fin (resulting in enclisis); but not both (building on Fern & aacute;ndez-Rubiera 2009, 2010). The special behaviour of uninflected infinitives can be explained by the possibility of clitics attaching to a lower verbal host and then moving with that host to T resulting in enclisis (Raposo and Uriagereka 2008). Aspectual/quantificational heads block verb movement to T, ruling out enclisis via 'low fusion'.
This paper investigates ex-situ focalization in Ikale, a Yoruba dialect of Nigeria, which appears to violate the Final-Over-Final Condition (FOFC) due to its clause-final morphological focus marking. Starting with a description of focus realization in the language, I propose that Ikale's ex-situ focus particle realizes a head-final FocP, which dominates a head-initial TP. I argue that such a FOFC-violating structure is best analyzed using the phase-based approach to FOFC's enforcement (a la Richards 2016, Erlewine 2017), rather than through extended projections. I further discuss two prominent alternative approaches to FOFC, (Biberauer et al.'s (2014) & Biberauer's (2017a) asymmetric approach and Zeijlstra's (2023) symmetric approach), and the potential challenges they face. I then propose that the current analysis may be extended to two other Benue-Congo languages (Igede and Nupe) and two other dialects of Yoruba (Ondo and Ok & imath;t & imath;pupa) with similar ex-situ clause-final morphological focus marking.
This paper proposes a theory of circumfix linearization in Distributed Morphology. The core of the proposal is that circumfixation arises whenever one of two exponents of the same morphosyntactic feature has a prosodic subcategorization frame. This will be exemplified with data from Indonesian ke- -an nominalizations where facts regarding stress placement, the domain of nominalization and other grammatical processes point towards a prosodic-cyclic analysis. The analysis is extended to mode ɩP-circumfixes in Meskwaki, circuminfixes in Tlahuitoltepec Mixe, and prosodically conditioned circumfixes in German past participles.
The agreement morpheme in the Turkish verbal domain surfaces in different paradigms depending on the preceding TAM marker. Kornfilt (1996) has proposed that this difference in spell-out signals a deeper syntactic difference, in that z-paradigm but not k-paradigm agreement morphemes are preceded by a silent copula. The present study is concerned with yet another, more recently documented paradigm attested in colloquial speech. Its key empirical finding is that these new forms are hybrids that share properties with both the k- and the z-paradigm. Its main theoretical claim is that this finding also affects our understanding of the older two sets of forms. Accordingly, the paper develops a novel allomorphy analysis of the three agreement paradigms. The allomorphy grammar proposed here and Kornfilt's copula grammar can coexist within a single speaker, and the former might have developed diachronically out of the latter.
In this paper, we examine the cross-linguistic syncretisms between markers of negation and 'expletive negation' (exn), i.e., a formal instance of negation lacking negative meaning, in fear-clauses. We argue that these syncretisms reflect structural proximity to the negative functional sequence (cf. De Clercq (2020) and Baunaz & Lander (2023)) and that exn realises a high, epistemic modal feature (cf. Makri (2013) and Tsiakmakis & Espinal (2022) for Modern Greek). We then explain how the syncretisms can be captured in a uniform way. For this, we adopt the Nanosyntactic model of lexicalisation (cf. De Clercq et al. 2025), and so-called stored 'complex left branches' (cf. De Clercq 2019). Ultimately, we show that, under the assumption that exn is modal, all patterns can be accounted for by varying the shapes of lexical items, while keeping everything else equal.
This paper addresses vocative truncation in Romanesco (Italo-Romance), which has been described as conditioned by a host of constraints at different structural levels. We crowdsourced the data collection with an online questionnaire devised to ask the three research questions whether the acceptability of truncation depends on (a) grammatical number, (b) the individual lexical items, and (c) speaker's age. Respondents' ratings on a 5-point Likert scale were analysed by means of Random Forests and Conditional Inference Trees. The results confirmed the significance of number and lexeme as conditioning factors.
This study discusses clitic placement in restructuring contexts in Italian and the Italo-Romance dialects spoken in Italy. The data come from AIS map 1086 and a judgement task experiment conducted on bilectal speakers of Italian and the dialect at six representative points in the Italian territory. The variation between clitic climbing and enclisis turns out to be much more complex than previous literature has suggested. The two clitic positions are available to all varieties, but at different rates and with different degrees of optionality. Optionality can be analysed as a function of language dominance in an intricate fashion.
This paper investigates the morphosyntax of numerals. Based on Russian and Dutch, we show that a division into three subcategories is necessary: (i) numeral ONE; (ii) Cardinals >= TWO; (iii) Indefinite Numerals such as MANY (cf., Barbiers 2007). We introduce a bi-partite structure, in which a low Classifier projection is connected to a high Classifier projection by an (abstract) preposition. Building on Borer (2005), Barbiers (2007), Kayne (2019), Corver (2021) we argue that ONE is an underspecified lower classifier, exhibiting meaningful parallels with the underspecified anchoring head at the clausal level (cf., Ritter & Wiltschko 2009, 2014). Cardinals are phrasal. Indefinite Numerals are subject to variation in whether they head ClassP-low or ClassP-high. Based on the morphosyntactic properties of the three categories we conclude that in some environments the PP and ClassP-high layers are not projected.
This study investigates the potential emergence of a fused lect in the Northern Kasaragod Variety of Malayalam (NKV-M), a speech variety shaped by prolonged contact with Malayalam, Kannada and Tulu. We examine verb-internal language mixing patterns using the exoskeletal framework with late insertion (originally proposed in Borer 2003), integrating sociolinguistic and formal approaches. Our findings reveal that the underlying syntactic skeleton of NKV-M is primarily derived from Malayalam. However, it incorporates agreement features into the T-projection, resulting in a fused T-Agr projection within the Tense Phrase. This fusion, which is absent in Standard Malayalam, reflects contact-induced retention of South Dravidian features. Morphological levelling among younger speakers and increasing conventionalisation of mixed forms indicate a trajectory towards a fused lect as defined by Auer (1999, 2014). This study demonstrates that NKV-M exhibits systematic, rule-governed variation rather than spontaneous code-mixing, and we argue that the exoskeletal model effectively captures this mixing pattern.
Fixed-VSO word order is understood to be a relatively recent innovation within the Mayan language family (England 1991), although it is a matter of recent debate whether this innovation is best understood as a prosodic (Clemens & Coon 2018) or a syntactic (Little 2020b) one. This paper adjudicates between these two proposals by closely examining the fixed-VSO Mayan language Mam. The data from Mam are consistent with Little's (2020b) syntactic approach by which objects raise overtly in the narrow syntax, where they hold certain structural and interpretational properties consistent with their high position.
This paper proposes a theory of circumfix linearization in Distributed Morphology. The core of the proposal is that circumfixation arises whenever one of two exponents of the same morphosyntactic feature has a prosodic subcategorization frame. This will be exemplified with data from Indonesian ke- -an nominalizations where facts regarding stress placement, the domain of nominalization and other grammatical processes point towards a prosodic-cyclic analysis. The analysis is extended to mode (sic)P-circumfixes in Meskwaki, circuminfixes in Tlahuitoltepec Mixe, and prosodically conditioned circumfixes in German past participles.
This paper discusses the distribution of subject clitics in nominal copular constructions in the variety of Este (PD) which is a partial pro-drop language where null subjects alternate with subject proclitics: when a postverbal subject NP is present, no subject clitic is allowed with lexical verbs. However, subject proclitics are allowed with postverbal subjects in inverse copular sentences where the postverbal NP agrees with both the copula and the subject proclitic. This paper explores the available discourse semantic interpretations of inverse copular sentences resulting by this inflectional pattern: while the preverbal predicative NP is a topicalized intensional element that instantiates a description, the postverbal NP represents new information. The preverbal predicative NP does not refer to an entity, but to a property representing the subset that includes the postverbal subject NP which a focal element introduced, just in this configuration, by a subject proclitic.
