Abstract: This paper introduces a novel unit of syntactic analysis, the catena (Latin for chain; plural catenae). The catena is defined in a dependency-based grammar as a word or a combination of words that is continuous with respect to dominance. According to this definition, any dependency tree or any subtree (complete or partial) of a dependency tree qualifies as a catena. The paper demonstrates that idioms are stored as catenae and that the elided material of ellipsis mechanisms (e.g., answer fragments, gapping, stripping, VP ellipsis, pseudogapping, sluicing, and comparative deletion) is a catena. Constituents are always catenae, but many catenae are not constituents. Based on the flexibility and utility of the catena concept, the claim is put forth and defended that the catena is the fundamental unit of syntax, not the constituent.

Abstract: . Bantu languages display a number of constructions that raise serious questions about the universality of the theory of abstract Case as currently realized in the Minimalist Program. Chomsky (2000) claims that positing uninterpretable features (like Case features) is not purely stipulation but that uninterpretable features are justified by their visible effects in the syntax. This paper presents evidence that the syntactic effects predicted by Case theory are not borne out in Bantu languages, which in fact display effects that are only predicted to be possible in the absence of uninterpretable Case features. Evidence includes constructions where Case-checking should require a DP to cease being active, but it does not do so (e.g., compound tenses and raising constructions), as well as instances where Case-checking could not have occurred, but the constructions are nonetheless acceptable (e.g., possible-constructions and locative inversion). It is claimed that uninterpretable Case features are not present in Bantu languages and that it is instead gender features that make a goal active for Agree (following Carstens 2011).

Abstract: In this article, we propose that three types of focus constructions in Japanese-clefts, in-situ focus, and sluicing/stripping-share the same underlying structure and are derived by syntactic "metamorphosis" from one structure to another. After revealing similarities between cleft constructions and in-situ focus constructions, we specifically propose that the latter underlies the former, which is derived by a focus movement followed by a heavy remnant movement. It is shown that various properties of cleft constructions follow from the syntax of in-situ focus constructions and movement operations. Then, it is further argued that cleft constructions underlie sluicing/stripping under certain conditions. The article also touches on the so-called clausemate condition found in multiple cleft sentences. A new set of data will be presented that poses a problem for syntactic explanations of this effect. (Less)

Abstract: Based on previously unnoted data concerning wh-the-hell attitude-bearing questions in Mandarin Chinese, I argue that integrating the pragmatic notion ''point of view'' into the syntactic computation enables us to capture certain properties of wh-the-hell questions in Mandarin Chinese in a systematic and illuminating way. The two central phenomena analyzed are (1) causal zenme how come is not compatible with daodi the-hell in Chinese, and (2) in Chinese wh-the-hell questions, when daodi the-hell takes matrix scope while staying overtly in the embedded clause, the person feature of the matrix subject cannot be third person. I argue that explaining these phenomena depends on recognizing the logophoricity of the negative attitudes carried by daodi and causal zenme. The valuation relation that I postulate between the Point-of-View operator and an unvalued Point-of-View feature of daodi and causal zenme not only explains these data but also formally captures certain properties of the Chinese logophor ziji (cf. Huang & Liu 2001).

Abstract: . The aim of this study is to show that fragment answers in Turkish are elliptical structures, where CP is targeted for ellipsis after a fragment answer moves to FP above CP. Evidence for the elliptical nature of fragment answers comes from binding, Case matching, scope, and postposition pied-piping, among others. An analysis is given of N-words as fragment answers and how (semantic) parallelism is satisfied in those cases. These elliptical structures are island-insensitive in Turkish. Fragment answers are island-insensitive because they move to a position that selects the elision site CP; thus, all the copies of a fragment answer that lack the ✓-feature due to crossing an island are elided in the elision site (Merchant 2004). This study also argues that fragment answers in English cannot be argued to be island-sensitive based on the data in Merchant 2004.

Abstract: . The Multiple Spell-Out model of Chomsky (2000, 2001, 2004) and Uriagereka (1999) provides a straightforward characterization of the syntactic environments for contraction. It is well known that wanna contraction is possible in the subject-control case (Who do you wanna marry?) but blocked in the DP+infinitival case (*Who do you wanna marry you?). This asymmetry receives a straightforward account if the CP-phase boundary demarcates the domain for contraction. More complex cases of contraction discussed by Postal & Pullum (1978, 1982) fall into place if left-branching structures form a self-contained Spelled-Out domain as in Uriagereka’s model. The analysis can also derive a number of well-known generalizations regarding auxiliary reduction such as (a) the invisibility of a Case-marked trace for reduction and (b) the inhibitory effect of the gap on reduction. Generalization (a) is a natural consequence of the TP analysis for that-less finite complements whereas (b) is the result of the tight interaction between the syntactic system and the fine-grained prosodic classification of function words developed by Selkirk (1995). Particular details of the analysis support various principles of syntax such as the Economy of Projection and the notion of CP phase. This paper, therefore, achieves the minimalist desideratum of explaining linguistic phenomena solely in terms of interface conditions and computational efficiency.

Abstract: . Case marking can interact with semantic interpretation in various ways, including differential case marking phenomena. This paper explores the interaction of partitive case in Finnish with the property of divisibility. It is argued that the appearance of partitive case can be related to divisibility. Most occurrences of partitive case are explained by a uniform treatment, which assumes that the divisibility of a phase affects the morphological realization of the [Case] feature minimally contained in the phase. This account extends to partitive case licensing by events; the case marked nominal itself; clausal negation; and prepositions as well. The discussion highlights the fact that morphological case is not necessarily determined by the licensing head. Instead, it can be determined within a local domain (a phase), the output of a cyclic Spell-out system. The proposal also bears on issues such as the phasehood of DPs and the structure of restructuring constructions.

Abstract: . In the Scandinavian languages, a [+negative] phrase must be licensed by Spec-head agreement in overt syntax, necessitating leftward movement of negative objects, Negative Shift (NegS). Although string-vacuous NegS is possible in all Scandinavian varieties, there is a considerable amount of crosslinguistic variation as to non-string-vacuous NegS. In particular, the varieties differ with regard to which constituents may be crossed by NegS and whether crossing depends on the position of the main verb. Under the cyclic linearization approach (Fox & Pesetsky 2003, 2005a), non-string-vacuous movement must proceed via intermediate positions. The variation as to the distribution of negative objects in the Scandinavian languages can thus be accounted for by differences in the availability of these intermediate positions, which is taken here to be determined by a mechanism of feature transmission. The relevance of intermediate positions is corroborated as neither the intervening constituents nor the object’s base position nor its target position (to the left/right of the matrix main verb) can capture the observed variation by themselves.

Abstract: . This paper examines how Japanese-speaking children interpret implicit variables and the anaphor zibun‘self’ when both are concerned with referential and quantificational subjects and antecedents in conjoined stripping structures with a case marker. In previous studies on the availability of sloppy readings in child Japanese, it has been unclear whether the attested sloppy reading really stems from LF copying of linguistic antecedents. Those studies employed null object and soo-su‘do so’ constructions; however, there is debate about whether these constructions are instances of surface or deep anaphora. To show whether Japanese-speaking children have unambiguous accessibility to bound-variable or sloppy interpretations in the grammar of Japanese, I did three experiments using a stripping construction with a case marker, which is considered to involve LF copying of linguistic antecedents (Fukaya & Hoji 1999). It has been shown that Japanese-speaking children are able to associate implicit variables and zibun as a bound variable with referential and quantificational antecedents and that sloppy readings are available in child Japanese. This study provides evidence that, in both child and adult grammars, the Japanese stripping construction with a case marker involves LF copying of an antecedent IP and is therefore an instance of surface anaphora.