Current generative grammar predicts a larger set of transitive structures than the six types normally mentioned in the typological handbooks (SVO, SOV, VSO, etc). The main cause of this discrepancy is that, whereas generative grammar investigates the hierarchical positions of the verb and its arguments, most typological research is concerned with their relative surface order. In order to bring together these two lines of research, we have to translate the predictions of generative grammar into a more sophisticated typology in linear terms that can be taken as the point of departure for future typological research. This programmatic article is written in the hope that the generative grammar may help typologists to find certain so far unknown typological differences between languages, and that typologists, in turn, may help generative grammar by providing the relevant typological data that are needed to evaluate the competing theoretical proposals and to improve the most successful ones. Keywords: word order typology, phrase structure theory, verb movement, adverb placement.

In this paper, we introduce what we consider to be a new problem for syntactic analysis. We show that German allows for discontinuous coordination of DPs, where material that is not part of the coordinate structure may intervene between the DP-coordinates. We further argue that discontinuous coordination of DPs differs from standard DP-coordination in one important aspect: The coordinate structure as a whole is not accessible to material c-commanded by the first coordinate in overt syntax. This difference is a non-trivial problem for any analysis of discontinuous DP-coordination. Keywords: coordinate structures; discontinuous constituents; DP-coordination; German syntax; linear order

This paper proposes that case decomposes into a number of separate functional projections, which are ordered in a universal functional sequence. Cross-linguistic variation in position and complexity of case morphology is then accounted for in terms of differential noun phrase movement within the invariant sequence (Cinque 2005). I further investigate the possibility that variation in movement may be reduced to variation in the shape of the actual lexical entries. In order to implement this idea, the model of cyclic spell out by Starke (2009b) is adopted. In this model, each step of external merge is followed by lexical access. Consequently, evacuation movements may be triggered after each step of external merge in order for successful lexicalization to take place. Keywords: agglutination; case; case paradigm; flexion; lexicon; movement; nanosyntax; parameter; phrasal spell out; variation

‘Extraposition’ is a cover term for two distinct phenomena (cf. also Fox & Nissenbaum 1999; Kiss 2005). The first, which I assume to be derived by parallel construal (following Koster 2000), targets only RCs, blocks reconstruction of the extraposed constituent to the base position of the source, and as such has the effect of bleeding condition C. The second, which I claim is derived via scattered deletion, targets post-head complements/modifiers, triggers obligatory reconstruction of the source+complement/modifier to base-position and leaves extraposed complements open to subextraction, all else being equal. The scattered deletion of spec+head and complement is argued to be derived from Uriagereka's (1999) simplification of the LCA, and the copy theory of labelling (cf. Sheehan 2010). Keywords: linearisation; extraposition; multiple spell-out; LCA; labelling; reconstruction; relative clause; complement PP; scattered deletion

In this paper a novel approach is proposed for a particular type of cleft-like construction in Somali, the waxaa sentence, in which wax is a scope marker. As such, it operates a function choice in the alternative set overtly expressed in the embedded clause, which serves as a semantic restrictor. In this line of analysis, wax requires the realization of a bi-clausal, copular construction in which the relativized presupposition (i.e. the restrictor) is merged in subject position, while the focused phrase is the predicate. Indefinite wax is located in the position dedicated to focused constituents in Somali and incorporates the Focus Marker baa, allowing for the realization of an in situ Focus (an option otherwise excluded in this language) and obtaining a specific interpretation for the unfocused part of the sentence. Indeed, based on corpus analysis and interface considerations, evidence is provided that baa and waxaa constructions are by no means interchangeable narrow Focus constructions. Focusing by means of waxaa implies a context in which unfocused information is not properly ‘given’; rather it conveys a type of ‘weakly familiar’ information that contributes somehow to the informative part of the sentence. A discussion about the function and formal properties of internal Topics in Focus-prominent languages concludes the work. Keywords: cleft construction; discourse; (weak) familiarity; (narrow) Focus; presupposition; relative clause; scope marker; specificational (copular) sentence; internal Topic

This article is on the relationship between case and agreement. A noun phrase is assigned the structural case that it bears through agreement with a functional head. Several recent works assume this thesis, referred to as the George and Kornfilt Thesis, as a basic premise to provide an account of structural case assignment. The central thesis of the article is that there is at least one more dependency that needs to be assumed in case phenomena, namely that, in some languages of the world, structural object case, or more particularly accusative case, is dependent on subject agreement—the Subject Agreement–Accusative Case Conjecture. The article proposes the Jump-start Hypothesis in order to explain this dependency. According to this hypothesis, in a finite construction, case assignment to each argument is activated by a single source of agreement. The argumentation relies on data from Turkish nominalizations and restructuring infinitives.

In this paper I propose that ellipsis is licensed by overt movement. Examining variation in VP-ellipsis across English dialects, I show that movement is crucially implicated in whether or not a given element can license ellipsis. I discuss well-known restrictions on VP-ellipsis and present new data that shows that a movement-based account of these restrictions is superior to previous ones. I show that the proposed account can be extended to other cases involving A′ movement with empirical benefits, and I conclude by sketching the technical implementation of the theory, arguing that ellipsis is a ‘repair’ operation that prevents a linearization failure following non-deletion of a lower copy. I suggest that types of movement that are unable to spell out lower copies (i.e. A-movement) do not license ellipsis, thus explaining ellipsis licensing in terms of general conditions on copy deletion. Keywords: ellipsis licensing; VP-ellipsis; sluicing; Copy Theory of Movement; A/A′-distinction

In this paper I explore the elements that make up the German distributive universal quantifier jeder, and the structural relationships among them. I argue that jeder consists of three overt morphemes je-d-er, which are heads in an extended adjectival projection (xAP). Their relative order is derived by movement [ xAP je d er t jeP]. Je corresponds to the adjectival stem, -d- is an adjectival article (which in turn is analyzed as a relative complementizer) and -er is an agreement head, AgrA. The xAP further contains movement traces/ copies of the nominal which jeder quantifies over. One of these copies is in the complement of je where, I claim, it supplies the restriction (distributive key or range). The components of the proposal are all motivated independently of jeder: (i) the morphology of jeder identifies it as adjectival, hence an analysis of it must incorporate an (independently motivated) adjectival syntax; (ii) a comparison with the distributive dual quantifier beid- ‘both’ further informs the syntactic analysis internal to the word jeder; and (iii) a comparison of je in jeder and in other je-words suggests that je takes an N(P) complement, a fact that confirms the expectations regarding the selectional properties of je raised by the preceding discussion. Finally, a comparison of jeder with counterparts of it in other languages, as well as with other complex determiners in German, will broaden the scope and corroborate important aspects of the present proposal.

We discuss the phenomenon of closest conjunct agreement with a special focus on head-final languages; we present data from two such languages, Hindi and Tsez, which allow agreement with the rightmost conjunct when the verb follows the conjoined phrase. This contrasts with head-initial languages, such as Arabic, where close conjunct agreement is with the leftmost conjunct in clauses with VS order. In addition, both languages exhibit certain flexibility of word order at root clause level; when the verb precedes the conjoined phrase, it can also agree with the leftmost conjunct. The empirical data from the two languages raise the following questions. First, is the typological difference between head-initial and head-final languages in the context of coordination due to a difference in the structure of coordination in these two groups? Second, to what extent is the syntactic configuration relevant to the computation of closest conjunct agreement? Third, what is the role of linear proximity in closest conjunct agreement? These questions have wider implications for the analysis of agreement and the relation between syntax and the morpho-phonological component.

Kashmiri has two structural positions for negation: a high position associated with focus and a low position associated with tense. In verb second environments, the difference between these two positions is neutralized on the surface and irrespective of the location where negation is generated, it appears as a suffix on the finite verb. But in non-verb second environments such as conditionals and correlatives the two negations can be teased apart. In these environments, the high negation appears as a suffix on material such as relative phrases and the conditional marker while the low negation appears as a suffix on the finite verb. The high vs. low distinction has semantic implications: in certain environments where the negation is arguably ‘expletive’, negation can only be high.