The language Jarawara (Arauan, spoken in Brazil) exhibits a puzzling set of passive-like properties in its “O-Construction” (Dixon 2000, 2004). We argue that O-Constructions have a type of passive voice in some person combinations but not in others, and that they are unified in that they always have topic agreement on C with the internal argument. We relate this approach to recent research on Algonquian inverse systems (especially Oxford 2023a,b, 2024) which have also been argued to involve a passive-like voice-based alternation for specific person combinations. Our analysis captures facts about case, word order, divergences between C and T agreement, and the distribution of the passive-like prefix hi- (among other properties). Our findings provide support for the approach to person restrictions embodied in Oxford’s work and also demonstrate how topic agreement and the A system can interact. More generally, this work shows how a nuanced approach to passive constructions, and a willingness to separate agreement from voice, can lead to a cross-linguistically grounded analysis of what seems prima facie like an “unusual” construction.
This article addresses two issues that emerge from a close look at extraction out of Russian indicative čto-clauses – the first being that these clauses show unexpected weak-island behavior, and the second the generally problematic question of how it can be possible for weak islands to allow “marginal” extraction at all (in argument cases), a grammaticality status never traditionally explained in pre-minimalist literature and theoretically impossible on core minimalist assumptions. An approach is proposed for weak islands under Minimalism that eliminates the non-minimalist principles that were claimed to account for their behavior (especially Subjacency and the Empty Category Principle [ECP]) and also allows for an understanding of why Russian indicative čto-clauses show the partial opacity observed.
This article examines the judgment and interpretation of Mandarin relative clauses (RCs) by heritage speakers (HSs) and second language learners (L2ers) of Mandarin, using an acceptability judgment task (AJT) and a picture-based truth-value judgment task (TVJT), in order to examine whether there is dominant language transfer in this domain, and whether HSs and/or L2ers exhibit a subject-extracted RC (SRC) advantage. The HSs and L2ers were both English-dominant and matched on Mandarin proficiency. The AJT tested whether participants knew that Mandarin RCs are head-final, unlike head-initial English RCs. The TVJT tested whether participants correctly interpreted Mandarin RCs with two animate nouns. In the AJT, both HSs and L2ers rated head-final RCs significantly above head-initial RCs, overcoming English transfer of RC headedness with increased proficiency. In the TVJT, HSs performed similarly to L2ers; neither group showed a clear SRC advantage, contrary to the predictions of the Noun Phrase Accessibility Hierarchy (NPAH) (Keenan & Comrie 1977). Compared to HSs/L2ers of Korean (O’Grady et al. 2001), HSs/L2ers of Mandarin in the present study acquire RCs more successfully, possibly due to high Mandarin proficiency, task format, and/or the lack of case markers in Mandarin RCs. While there was a slight HS advantage over L2ers in the AJT, there was no HS advantage in the TVJT, suggesting that Mandarin word order is not particularly difficult for either group, in contrast to case marking in Korean.
Legate (2021) deconstructs passive into three characteristic properties – agent demotion, theme promotion, and morphological marking – and shows that these properties vary independently across languages. She concludes that this range of variation supports an approach, such as Minimalism, in which universal grammar includes no information specific to voice. This brief note takes a similar approach to antipassive, a clause type whose typology has been investigated by Polinsky (2017). I first deconstruct antipassive into two characteristic properties – demotion of the internal argument, which comes in several subvarieties, and voice marking – and then suggest that these properties vary independently across languages. The data are drawn primarily from Austronesian languages.
Ch’ol is a Mayan language with verb-initial order and preverbal topic and focus positions (Vázquez Álvarez 2011; Clemens & Coon 2018). This paper presents the results of a systematic investigation of Ch’ol word order across various focus environments, including (i) broad focus, (ii) subject focus, (iii) object focus, (iv) contrastive subject focus, and (v) contrastive object focus. We analyze semi-spontaneous responses to questions designed to elicit these focus types from 31 Ch’ol speakers. Both verb-initial and subject-initial clauses are present across five focus conditions, revealing a more nuanced relationship between information structure and word order than previously reported. We also find that while contrastive focus is predominantly marked via fronting of the focused constituent, more variation is found with information focus, and fronting is found to be nonobligatory in every focus condition.
This chapter discusses the inner structure of superficially different manner-adverbial patterns in Dutch, with a special focus on the surface pattern A+P, as in hardop (loud-up; ‘aloud’). In the spirit of the generative linguistic quest for cross-constructional symmetry, it is proposed that these adverbial patterns are all manifestations of one and the same abstract, underlying syntactic configuration, namely the Extended Adpositional Phrase (XPP). After this in-depth, single-language study of manner adverbials, the chapter continues with a more global, cross-linguistic perspective on the inner structure of manner adverbials, starting from the hypothesis that, at a more abstract level, the adverbial patterns attested cross-linguistically all have an adpositional design. Finally, the chapter briefly discusses the relationship between inner structure and outer behavior (i.e., distribution) of manner adverbial expressions. Specifically, the question is addressed to what extent “being prepositional” or “being postpositional” matters for the distributional behavior of manner adverbials. More in general, this chapter aims to provide another illustration of the fruitful interaction between in-depth investigation of individual languages, and the comparative-linguistic study of a larger sample of languages.
This paper provides a comprehensive cross-linguistic overview of overcounting, whereby a target numeral is expressed by counting toward the next-higher multiple of the base. I identify three major morphological patterns in overcounting numerals: P-connector, V-connector and no overt connector. I then zoom in on the structure of overcounting numerals in Ch’ol (Mayan). I argue that these numerals are construed with a covert, latively interpreted P, whose silence is due to P-drop.
In the Mandan (Western Siouan) complex verb, the makeup of the prefixal field features two loci for φ-feature marking, separated by what the Siouanist literature calls “preverbs” (PV). The pre-PV φ-slot is for marking first person plural (1PL); the post-PV φ-slot is for markers of speech-act participants. The question central to this paper is what explains the positioning relative to the preverbs of the 1PL marker and the other φ-morphology of Mandan. The Mandan 1PL prefix, which has a morphologically transparent dual inclusive reading, is syntactically represented in the form of a comitative phrasal structure involving the asyndetic coordination of a plural pronoun and a combination of the first person singular and second person pronouns: ‘we, viz., I with you’. While 1PL is a morphosyntactic complex occupying SpecTP (preceding preverbs), the first and second person markers are agreement inflections (following preverbs), linked to pro’s in A-positions.