The purpose of this paper is to examine psycholinguistic work on attraction with a view to enriching our knowledge of the grammar of agreement. Following Franck et al. (2006), I assume that the different theories of agreement should relate to the way speakers err when they implement agreement operations. As an aberrant computation of the mind, attraction is interesting due to its frequency: in English experiments 13% of complex NPs (i.e. NPs which consist of two or more constituent NPs) establish incorrect agreement with the verb (as in *the key to the cabinets are in the kitchen; Eberhard, Cooper Cutting & Bock 2005). This is what makes it a magnet for both linguistic and psycholinguistic research. Here I examine the main findings and models in the psycholinguistic literature, and how they relate to existing theories of agreement in grammar. It will be argued that agreement cannot be properly understood unless models incorporate an adequate measurement of the size of the morphological component of every language studied, as agreement operations are continuously sensitive to this. The general idea, which I extend from Berg (1998) and Acuña-Fariña (2009) is that a strong morphosyntactic component blocks (rather than facilitates) semantic interference, and that languages opportunistically use more or less semantics in establishing agreement ties depending not only on morphological richness but also on the direction of encoding.

There have been claims in the literature that the variability of compound stress assignment in English can be explained with reference to the informativeness of the constituents (e.g. Bolinger 1972, Ladd 1984). Until now, however, large-scale empirical evidence for this idea has been lacking. This paper addresses this deficit by investigating a large number of compounds taken from the British National Corpus. It is the first study of compound stress variability in English to show that measures of informativeness (the morphological family sizes of the constituents and the constituents' degree of semantic specificity) are indeed highly predictive of prominence placement. Using these variables as predictors, in conjunction with other factors believed to be relevant (see Plag et al. 2008), we build a probabilistic model that can successfully assign prominence to a given construction. Our finding, that the more informative constituent of a compound tends to be most prominent, fits with the general propensity of speakers to accentuate important information, and can therefore be interpreted as evidence for an accentual theory of compound stress.

The paper offers a theoretical characterization of the middle Voice as distinct from the passive Voice, and addresses the cross-linguistic morphological variation in realizing these two non-active Voices in different classes of languages, represented by Hebrew, Greek and English. The two non-active Voices are the morphological realization of two distinct syntactic Voice heads generating middle and passive clauses respectively. The former are cross-linguistically interpreted as (i) anticausative, (ii) reflexive (and reciprocal), (iii) dispositional middle, and (iv) medio-passive, which is distinct from passive. This variation in the interpretation of the middle Voice reflects different properties of the root rather than the application of four different lexical rules postulated by lexicalist theories.

Most of the literature devoted to the study of deverbal nominalizations concentrates on the complex event reading (e.g. La concentración de partículas tiene lugar a temperatura ambiente ‘The concentration of particles takes place at room temperature’) and the object reading (e.g. El paciente tenía concentraciones de calcio en el hombro ‘The patient had calcium concentrations in the shoulder’), while nominalizations denoting states (e.g. La concentración de Sherlock Holmes duró cinco horas ‘Sherlock Holmes’ concentration lasted five hours') have remained, in general, understudied. In this paper we present their empirical properties and argue that, despite the empirical differences, state nominalizations and event nominalizations can receive a unified account. We show that in Spanish, Catalan, French, English and German the question of whether a deverbal nominalization denotes a state or an event, or is ambiguous between both readings depends on independent properties of the verbal base, allowing us to propose a unified account of both classes of nominalizations: the productive nominalizers in these languages can only denote the aspectual notions contained in the base's Aktionsart. We further argue that other languages, like Slovenian, have productive nominalizers that can operate over the external aspect of the predicate; in these cases, the nominalization can denote aspectual notions not contained in the base's Aktionsart.

Much recent work in phonology concentrates on the role of sonority in the phenomenon of vowel reduction, capitalizing on two facts: that reduction involves raising and/or shortening and that higher vowels and schwa are normally interpreted as having low sonority. This paper presents a different approach to vowel reduction in Standard Russian. It is proposed that the apparent sonority-driven effects in Russian are epiphenomenal. In particular, reduction to schwa is outside the domain of phonological computation in Russian, being an artifact of reduced duration. Other types of neutralization arising in vowel reduction are potentially amenable to a sonority-based analysis, but I argue that current approaches to sonority-driven reduction suffer from representational shortcomings. When these shortcomings are rectified, however, sonority is unnecessary as an explicit factor in vowel reduction: standard markedness mechanisms suffice to explain the data.

So-called respectively readings have posed serious challenges for theories of syntax and semantics. Sentences like George and Martha respectively denounced and were denounced by the governor (McCawley 1998) show that although the conjoined verbal expressions share the same syntactic subject, they do not predicate that subject in the same way; George (not Martha) denounced the governor, and Martha (but not George) was denounced by the governor. Postal (1998: 160–163) and Gawron & Kehler (2004: 193–194) show that this phenomenon poses problems for contemporary theories of grammar and argue that it is particularly acute for theories where subcategorization and predication are linked via unification. As these authors note, the problem is even more severe in respectively readings involving filler–gap constructions. In this paper I argue that the severity of these problems has been overstated and that the data do not entail any special dissociation between predication, subcategorization, or extraction. In this paper I propose an account which is fully compatible with unification-based theories of grammar. Gawron & Kehler (2004) propose an account of respectively phenomena which covers a remarkably wide range of cases. That approach relies on a Respf operator, which it stipulated to be optionally overt. However, I argue that this analysis is problematic because there are significant semantic differences between respectively readings with and without an overt realization of ‘respectively’. Rather, the data suggest that respectively readings may be special cases of more general phenomena which happen to create interpretations that are compatible with the semantics of the adverb respectively. This explains why respectively readings can arise without the adverb, and does not require us to posit a disconnect between predication and subcategorization. In fact, a sentence with a respectively reading will not differ in syntactic or semantic structure from sentences without such a reading.

Japanese has two peculiar types of relative clause (RC), No-RCs and De-RCs. In these types of relative clause, what looks like a (pivot) head noun appears at the left edge of the clause and is accompanied by no and de, respectively. This sharply contrasts with regular prenominal relative clauses in Japanese, which conform to the head-final word order pattern. The aim of this article is to investigate the syntax and semantics of these two types of relative clause in detail and reveal differences between them. Specifically, I will propose that (i) no in No-RCs is an appositive genitive particle licensed by a silent linker head, and (ii) de in De-RCs is a continuative/participial form of the copula da. Drawing a parallel with NP-no NP constructions and building on an idea from S.-Y. Kuroda's dissertation, it will be argued that No-RCs are derived by DP-internal inversion mediated by the linker. On the other hand, De-RCs will be shown to be relatives conjoined with the copula de. It will be further suggested that the fact that Korean and Mandarin Chinese lack equivalents of De-RCs is due to the absence of the appositive genitive particle and hence of DP-internal inversion.

The internal syntactic structures of restrictive and non-restrictive relative clauses are largely identical. This paper argues that despite a uniform internal structure, the external distributions – specifically, linearization with respect to the head – of non-restrictive relative clauses are subject to several conditions. In particular, sentential non-restrictive relative clauses with which and what can appear to the left of their heads in limited (and distinct) syntactic contexts. These lexical and syntactic constraints are represented within the framework of Sign-based Construction Grammar. In light of these observations, the paper revisits claims about the internal structure of parenthetical as-clauses. Prior claims that as cannot be a relativizer are shown to be unfounded, and new data are presented in favor of treating them as relative clauses with the external distribution of sentential adverbials. This is possible given the ability to state separately specifications of construction-internal syntactic structure and construction-external linear-order constraints.

This paper provides a unified syntactic account of the distribution of English have in causative constructions (e.g. John had Mary read a book) and experiencer constructions (e.g. John had the student walk out of his classroom). It is argued that have is realized in the context of an applicative head (Appl) and an event-introducer v, regardless of the type of v. Have is spelled out in the causative when Appl merges under vCAUSE, and in the experiencer construction when Appl merges under vBE. This proposal is extended to have in possessive constructions (e.g. John has a hat/a brother): have is realized in the context of vBE and Appl. The proposed account provides empirical evidence for expanding the distribution of Appl: (i) a causative can take ApplP as a complement, which was absent in Pylkkänen's (2008) typological classification, and (ii) Appl can merge above Voice, contrary to Pylkkänen's analysis in which Appl is argued to always merge below VoiceP, never above. Moreover, the proposed account supports the theoretical claim that argument structure is licensed by functional syntactic structure; in particular, it shows that the relevant functional heads are not aspectual heads, but Appl and v.

Greek polydefinites are cases of adjectival modification where the adjective features its own definite determiner. We propose an account of the phenomenon that treats it as an instance of close apposition. Like close appositives, polydefinites in Greek instantiate multiple definite determiners, display a freedom in word order, and involve a restrictive interpretation. We propose that close apposition in Greek forms a complex DP out of two DPs which are in a sisterhood relationship through identification of the Referential roles within the DPs. This operation, semantically tantamount to set intersection, is constrained to apply only when the resulting set is not co-extensive with either initial set. This ensures the restrictive interpretation of one DP over the other. The fact that in polydefinites, it is always the DP containing the adjective that obligatorily satisfies the constraint has to do with the presence of noun ellipsis within that DP: (noun) ellipsis is known to come with a disanaphora requirement. We show that noun ellipsis is also responsible for the distribution of adjectives and adjective interpretations, as well as those discourse effects of polydefinites that have been thought of as the result of a DP-internal Focus projection. Finally, we make a proposal for the encoding of definiteness in Greek, consonant both with the existence of polydefinites in the language and with the prerequisite for set intersection among DPs: the overtly realized Greek definite determiner does not itself contribute an iota operator but preserves the <e,t>denotation at the DP level. Our proposal thus deals not only with the multiple occurrence of definite determiners in a construction that picks out a single discourse referent, but also with the compositionality problem that such a situation gives rise to. In the final part we tie the cross-linguistic (un)availability of expletive determiners of the Greek type to the (un)availability of morphologically realized case.

There has been a consistent pattern of criticism of the reliability of acceptability judgment data in syntax for at least 50 years (e.g., Hill 1961), culminating in several high-profile criticisms within the past ten years (Edelman & Christiansen 2003, Ferreira 2005, Wasow & Arnold 2005, Gibson & Fedorenko 2010, in press). The fundamental claim of these critics is that traditional acceptability judgment collection methods, which tend to be relatively informal compared to methods from experimental psychology, lead to an intolerably high number of false positive results. In this paper we empirically assess this claim by formally testing all 469 (unique, US-English) data points from a popular syntax textbook (Adger 2003) using 440 naïve participants, two judgment tasks (magnitude estimation and yes–no), and three different types of statistical analyses (standard frequentist tests, linear mixed effects models, and Bayes factor analyses). The results suggest that the maximum discrepancy between traditional methods and formal experimental methods is 2%. This suggests that even under the (likely unwarranted) assumption that the discrepant results are all false positives that have found their way into the syntactic literature due to the shortcomings of traditional methods, the minimum replication rate of these 469 data points is 98%. We discuss the implications of these results for questions about the reliability of syntactic data, as well as the practical consequences of these results for the methodological options available to syntacticians.

The aim of this paper is to provide an analysis of the positive effect that modification has on the distribution of noun phrases in otherwise illicit environments. I focus on de nominals in French. By focusing on these nominals, whose distribution is altered by the addition of modifiers, the paper shows that modifiers can do much more than simply modify: they can change the syntactic and semantic status of a noun phrase. The licensing property of modifiers is an intriguing topic and has not been greatly discussed in the literature. I argue that modifiers can come to play the role of determiners in French as long as they are accompanied by a head de, which is the spell-out of a Cardinal head (see Lyons 1999). My proposal goes back to an old idea put forward by Damourette & Pichon (1911–1940) according to which, in modified contexts, de functions as one half of the article while the adjective functions as the other half. More generally, articles in French are seen as dual entities comprising of a specifier and a head. In the absence of the determiner les, an adjective can raise to the specifier of CardinalP. This is achieved via phrasal rather than head movement.

This work examines multiple-modal constructions in Mandarin Chinese. It is shown that modals in Mandarin Chinese are ordered in specific ways, and, while a number of factors contribute to the determination of the hierarchical order of modals, the finiteness of the complement clauses that the modals take plays a particularly important role. A generalization proposed in this paper is that (i) if a modal takes a finite clause as complement, then the modal itself can only occur in a finite context; and (ii) if a modal takes a nonfinite clause as complement, then it can occur in a nonfinite context too. This generalization is shown to account for some ordering phenomena of modals in Mandarin Chinese that would otherwise remain mysterious.

In Finnish, negation is expressed via an auxiliary, and no other verb may occur above this auxiliary in the structure. This gives rise to a problem with respect to the modals of obligation and necessity, which take scope over negation yet appear below it. It is tempting to account for this in terms of LF-movement, but evidence suggests that there are in fact two modal phrases in Finnish, one above negation and the other below it, the higher of which encodes necessity/obligation. Evidence for the higher phrase comes from the negative imperative. Although the PF part of a verb in a negative sentence cannot move to the head of this higher phrase, the head itself is in the right position to take scope over negation. Thus, rather than attributing the scope properties of the modals to LF-movement, it will instead be argued that the LF-interpretable part of a head is merged precisely where it takes scope, and that the relation between the LF- and PF-interpretable parts of the modal is one of checking at a distance. Head-movement will be regarded solely as a PF phenomenon. It will be seen that the scope relations of the modals and the imperative mood can be accounted for under this hypothesis. Thus, Finnish provides evidence for a view of syntax which identifies syntactic structure largely with the LF-interpretable part of a sentence, and sees head movement as fundamentally a PF phenomenon. There are two morphological moods in Finnish, which seem to provide counter-examples to this hypothesis, which will be left as a problem for future research.

This paper describes and analyses the pattern of word stress found in the standard dialect of Koshur (Kashmiri) spoken in Srinagar. The significance of Koshur for studies of stress lies in that taken together, its pattern of stress assignment and a pervasive pattern of syncope conspire to produce a four-way syllable weight distinction that has sometimes been expressed as the scale CVːC>CVː>CVC>CV. The interesting feature of this type of scale is that closed syllables, CVːC and CVC are preferred as stress peaks over open syllables with vowels of the same length. Other researchers have noted that in languages with this scale, or the abbreviated ternary version CVː>CVC>CV, CVC syllables behave ambiguously with respect to stress. They seem to be heavy in relation to CV when CVː syllables are absent. In a stress clash context however, CVC defers to CVː. ‘Mora-only’ accounts of other languages with this scale have interpreted the ambiguous behaviour of CVC as evidence that CVC syllables are bimoraic where their behaviour seems to group them with CVː but monomoraic elsewhere (e.g. Rosenthall & van der Hulst 1999, Morén 2000). To account for the CVːC>CVː effect, mora-only accounts have been forced to assume that CVːC are trimoraic. We show that a mora-only analysis does not offer a satisfying account of the Koshur facts, and we argue instead that the origin of the CVC>CV and CVːC>CVː effects is the presence of a coda that branches from the final mora of a syllable, making the closed syllables more harmonic as prosodic heads. Under this view, branchingness emerges as another dimension of the mora, along with moraic quantity and the quality of segments linked to moras, which contributes to syllable prominence.

In an article in this journal, Bao (2005) proposes a constraint on functional transfer that he claims accounts for features of colloquial Singapore English (and other language contact varieties) better than the congruence constraint proposed by Siegel (1999) and subsequently developed in later works (e.g. Siegel 2003, 2008a). More specifically, Bao argues that the requirement of surface syntactic similarity for transfer is too strong. His analysis uses Mandarin to exemplify the Chinese substrate languages that were the source of transfer, following the view that there is a universal Chinese grammar (Chao 1968: 13). However, the present article shows that Bao's claim is unjustified because the actual source of transfer was a variety of Chinese that differs significantly from Mandarin in the area of grammar he examined.

San Duanmu's Syllable Structure: The Limits of Variation raises a number of questions that are of general interest for phonological theory. Of special interest here are: the genesis and management of linearity in complex segments, the place of analogy (or paradigm uniformity) in grammar, the role of morphology in accounting for phonological patterns, the balance of static (distributional patterns) and dynamic (phonological processes) evidence for syllable structure, the role of stress in syllabification, and the import of corpus-based data for phonological analysis. In each case, Duanmu's proposals are evaluated according to their intrinsic consistency, the empirical record and the relevant body of literature. Alternative ways of handling the phenomena are offered, and these are fairly traditional in most cases. Duanmu's book is particularly relevant in the current constitution of the field where the see-saw movement between computation and representations seems to swing back in direction of the latter after having long been immobilised on the computational end. Standing clearly on the representational side, the theory exposed in the book aims to show that all surface strings may be reduced to a fixed and invariant syllable template, C(onsonant)V(owel)X. This enterprise is interesting especially in presence of another representationally-oriented theory, CVCV (Lowenstamm 1996, Scheer 2004), which also aims at reducing surface variation to an invariant syllabic skeleton, made of a monotonic sequence of CV units. However, the CVX and the CVCV templates are quite distinct, and the strategies that are used in order to accommodate the surface string are opposite (shrinking in the former case, expanding in the latter).

In a 1982 paper in the journal Glossa, Pullum outlined a set of arguments for treating English infinitival to as a defective auxiliary verb. Twenty years later, in his entry on infinitival constructions in The Cambridge Grammar of the English Language (CGEL, 2002), Huddleston argues that several distributional facts about auxiliaries fit poorly with this hypothesis. He proposes, on the basis of significant structural parallels, that to is a subordinator (complementizer). I show that Huddleston's arguments constitute a flawed analysis in CGEL's otherwise superb coverage of English descriptive grammar, and that the facts run strongly counter to his claims, often falling out independently from generalizations about auxiliaries that Huddleston overlooks. Several of these points were anticipated in Pullum's paper, but recent research on an idiosyncratic auxiliary-specific pattern of English nonrestrictive relative clause formation provides a powerful new argument in support of the auxiliary claim. In this respect, as in all others, the assignment of to to the class of auxiliaries provides the simplest and broadest account of its syntactic behavior.