Showing papers in "Natural Language and Linguistic Theory in 2011"

Multiple agree with clitics: person complementarity vs. omnivorous number

Abstract: This paper capitalizes on the difference between person complementarity (e.g. PCC effects) and omnivorous number (e.g. the fact that a single plural marker can be used to cross-reference more than one plural argument) by proposing that the same syntactic mechanism of Multiple Agree is responsible for both. The widely divergent surface difference results from the fact that person features are fully binary, whereas number features are syntactically privative. Additionally, arguments drawn from a variety of verbal cross-referencing morphemes implicating phi-interactions between subject and object support the claim that these elements are clitics, necessitating a principled morphosyntactic difference between clitics and other DPs undergoing object shift, and revisitation of the clitic-affix distinction.

The gradual emergence of phonological form in a new language

Abstract: The division of linguistic structure into a meaningless (phonological) level and a meaningful level of morphemes and words is considered a basic design feature of human language. Although established sign languages, like spoken languages, have been shown to be characterized by this bifurcation, no information has been available about the way in which such structure arises. We report here on a newly emerging sign language, Al-Sayyid Bedouin Sign Language, which functions as a full language but in which a phonological level of structure has not yet emerged. Early indications of formal regularities provide clues to the way in which phonological structure may develop over time.

Hindi pseudo-incorporation

Abstract: This paper argues that Hindi incorporation is, in fact, pseudo-incorporation, involving noun phrases rather than nouns. Furthermore, it shows that there is no requirement that the incorporated nominal surface as a morphological or even a syntactic unit with the verb. Such loosely aligned nominals can nevertheless be identified as incorporation on the basis of semantic intuitions having to do with number interpretation, anaphora, and certain properties typically associated with lexical processes. Contrary to standard assumptions, it is argued that the target of pseudo-incorporation is specified for number. Singular incorporated nominals in Hindi are shown to be semantically singular, with number neutrality arising as a consequence of interaction with aspectual operators. Taking aspectual information into account is also shown to have interesting implications for current approaches to the semantics of incorporation, one in which the incorporated nominal introduces a discourse referent, and one in which it functions as a predicate modifier. A closer look at the effect of aspect on anaphora, for example, does not unequivocally support the predicate modification view of pseudo-incorporation. The paper draws on data from Hungarian, and to some extent Danish, to explore the cross-linguistic applicability of the claims made on the basis of Hindi. Most notably, a distinction between Hungarian verbs with respect to incorporation of bare singulars provides striking confirmation of the claim that number morphology is semantically visible in pseudo-incorporation. It also addresses restrictions on the productivity of pseudo-incorporation in light of the proposed analysis of pseudo-incorporation.

Left peripheral focus: mismatches between syntax and information structure

Abstract: In Czech, German, and many other languages, part of the semantic focus of the utterance can be moved to the left periphery of the clause. The main generalization is that only the leftmost accented part of the semantic focus can be moved. We propose that movement to the left periphery is generally triggered by an unspecific edge feature of C (Chomsky 2008) and its restrictions can be attributed to requirements of cyclic linearization, modifying the theory of cyclic linearization developed by Fox and Pesetsky (2005). The crucial assumption is that structural accent is a direct consequence of being linearized at merge, thus it is indirectly relevant for (locality restrictions on) movement. The absence of structural accent correlates with givenness. Given elements may later receive (topic or contrastive) accents, which accounts for fronting in multiple focus/contrastive topic constructions. Without any additional assumptions, the model can account for movement of pragmatically unmarked elements to the left periphery (‘formal fronting’, Frey 2005). Crucially, the analysis makes no reference at all to concepts of information structure in the syntax, in line with the claim of Chomsky (2008) that UG specifies no direct link between syntax and information structure.

Reduced and unreduced phrasal comparatives

Abstract: Degree heads combine with individual (John is taller than [Mary]) as well as clausal arguments (John is taller than [Mary is]). Does the degree head have the same meaning in these two argument structures? Two kinds of answers have been proposed in the literature: I. there is a single meaning where the 2-place degree head combines with a degree predicate, with a reduction operation that derives the DP argument from a degree predicate denoting clausal argument, and II. there are distinct meanings for each argument structure, one combining with an individual denoting DP (3-place degree head) and the other with a degree predicate denoting clause (2-place degree head). We show that languages vary in which of these answers they choose: English goes for option I and Hindi-Urdu and Japanese for versions of option II. Our account of this variation assumes that the crosslinguistic distribution of 2-place and 3-place degree heads is not in itself subject to crosslinguistic parametrization; they are just syntactic projections of the basic meaning of comparison. We advance a specific proposal which derives the differences between the languages from the morphosyntactic properties of ‘than’ and a preference for minimal structure.

Inchoativity, change of state, and telicity: evidence from Spanish reflexive psychological verbs

Abstract: In this paper, we argue that by making a careful distinction between the notions of inchoativity and telicity, we can gain new insight into how changes of state can be expressed in natural language. Our argument is based on an analysis of Spanish reflexive psychological verbs (SRPVs) such as aburrirse ‘to be/become bored’ and enfadarse ‘to become angry.’ We present diagnostics that clearly support the claim that while these verbs are inchoative, they are not telic, nor do they denote changes of state. Additional tests indicate that these verbs are not dynamic, either; however, we show that they lack dynamicity for different reasons: aburrirse verbs, because they are stative; enfadarse verbs, because they denote truly punctual eventualities. We present a formal semantic analysis using the event ontology proposed in Pinon (1997). This analysis allows us to capture very naturally the similarities and differences between the two subclasses of SRPVs, to characterize their inchoativity, and to distinguish it from telicity. In addition, it supports a view of the Vendlerian aspectual classes on which the achievement class describes truly punctual eventualities and excludes certain predicates commonly (if not universally) assumed to belong to this class, such as the so-called degree achievement verbs.

Causatives across components

Abstract: The paper divides verbal alternations which, broadly speaking, involve the cognitive notion of Causation into (i) causativization, (ii) decausativization. The latter operation is identical across languages and applies universally in the lexicon. The former is argued to be lexical in some languages; in others, causative verbs are built in the syntax by means of a Cause predicate and an embedded one. The study presents novel empirical evidence, identifies and derives the various clusters of properties associated with the alternations, and formulates the precise mechanisms underlying them. The empirical array is drawn mainly from Hungarian and Japanese. The findings have direct implications for alternative conceptions of the division of labor between the syntax and the lexicon.

Mixed agreement, the person feature, and the index/concord distinction

Abstract: In mixed agreement, different agreement targets show different values for the same controller This paper offers an explanation for the existence of mixed agreement that accounts for the following Polite Plural Generalization: universally, any second person ‘polite plural’ pronouns (eg French vous), used honorifically for a single addressee, control syntactic (plural) agreement on all person targets, while non-person-agreeing targets such as predicate adjectives vary across languages, between syntactic and semantic number agreement Following Wechsler and Zlatic (The Many Faces of Agreement, CSLI Publications, 2003), person features exist only as features of referential indices (Index phi features), never as grammatical head features of the sort that are involved in adjective-noun concord (Concord phi features) Mixed agreement arises if the ‘polite plural’ or other pronominal controller is underspecified for Concord phi features But a pronoun has a referential Index, which is necessarily marked with phi features, so any Index agreement targets will appear in the second person plural form A diachronic explanation is offered for this bifurcation of agreement targets into Index and Concord targets: the former descend from incorporated pronouns while the latter have other sources

The subjective mode of comparison: Metalinguistic comparatives in Greek and Korean

Abstract: In this paper, we present a striking parallel between Greek and Korean in the formation and interpretation of metalinguistic comparatives. The initial observation is that both languages show an empirical contrast between ordinary and metalinguistic comparatives realized in two places: (a) in the form of a designated metalinguistic comparative MORE, and (b) in the form of THAN employed. We propose (building on earlier ideas in Giannakidou and Stavrou 2009; Giannakidou and Yoon 2009) that the metalinguistic comparative is subjective and attitudinal, i.e. it introduces the point of view of an individual towards a sentence—and argue that the individual expresses invariably an attitude of preference: she prefers one sentence (the sentence itself, or the proposition it expresses) in a given context over another. The preference may come out as completely negative in certain cases, and this is manifested as yet another MORE lexicalization in Korean (charari), which selects nuni-THAN, which itself carries a negative expressive index (in the sense of Potts 2007b), we will claim. Expressive negativity is not equivalent to negation in syntax, as nuni alone cannot license NPIs that need negation. If our analysis is correct, it has one important implication that goes beyond just the metalinguistic comparatives in the individual languages we are considering. It allows the generalization that metalinguistic functions in language are indeed part of the grammar. In particular, they are reflexes of grammaticalization of perspective and subjective mode, on a par with predicates of personal taste discussed by Lasersohn (2005, 2008, 2009), mood choice, and similar phenomena. In comparatives, subjective mode is manifested as an attitude of preference, with possible addition of expressive meaning.

Biases in Harmonic Grammar: the road to restrictive learning *

Abstract: In the Optimality-Theoretic learnability and acquisition literature it has been proposed that certain classes of constraints must be biased toward particular rankings (e.g., MarkednessIO-Faithfulness; Specific IO-FaithfulnessGeneral IO-Faithfulness). While sometimes difficult to implement efficiently or comprehen- sively, these biases are necessary to explain how learners acquire the most restrictive grammar consistent with positive evidence from the target language, and how innova- tive patterns emerge during the course of child phonological development. This paper demonstrates that altering the mode of constraint interaction from strict ranking as in Optimality Theory to additive weighting as in Harmonic Grammar (HG) reduces the number of classes of constraints that must be distinguished by such biases. Using weighted constraints and a version of the Gradual Learning Algorithm (GLA), the only distinction needed is between Output-based constraints, which must be biased toward high weights, and Input-Output-based constraints, which must be biased to- ward the lowest weights possible. We implement this distinction within the HG-GLA model by assigning different initial weights and plasticity values to the two classes of constraints. This implementation suffices to ensure that restrictive grammars are

Locality and globality in phonological variation

Abstract: This paper explores the consequences of Harmonic Serialism (HS), a derivational variant of OT, for local and global phonological variation. Variation is local when each locus within a single form may vary independently, as in French schwa deletion. Variation is global when all loci within a single form must covary, as in labial (de)voicing in Warao. Within the framework of Optimality Theory (OT) with parallel evaluation, only global variation is predicted to exist. In this paper, I show how implementing a multiple-rankings theory of phonological variation within HS accounts for both local and global variation.

When agreement is for number and gender but not person

Abstract: In many languages, adjectives agree with a noun phrase in number and gender, but not in person. In others, ditransitive verbs can agree with their theme argument in number and gender, but not in person (the Person Case Constraint). However, a unified account of these two similar patterns has rarely been attempted. In this article, I review how a single syntactic principle from Baker (2008)—the Structural Condition on Person Agreement (SCOPA)—can explain both phenomena, in contrast to other existing proposals. I then go on to show how the SCOPA also accounts for five other environments in which verbs agree in number and gender (if relevant) but not in person. Special attention is given to two entirely new cases: subject raising constructions in Sakha and agreement with direct objects in Ostyak. Along the way, I also discuss the consequences of partial agreement for case assignment, using this to explain why non-SCOPA-compliant configurations sometimes result in legitimate partial agreement with a first or second person pronoun, and sometimes result in a structure being ineffable.

Adjectival passives in Hebrew: evidence for parallelism between the adjectival and verbal systems

Abstract: This article focuses on Hebrew adjectival passives, showing that, as was claimed for other languages, the class of adjectival passives in Hebrew is not homogenous, but rather consists of two sub-classes. Former attempts to capture the non-homogenous nature of the class of adjectival passives in different languages relied mainly on the existence versus absence of an event in their interpretation. In contrast, I argue that the criterion distinguishing the two sub-classes of adjectival passives in Hebrew is the presence versus absence of an implicit Agent or Cause argument. Thus, the split parallels a very well-known split in the verbal system—that between passive and unaccusative verbs. Once this parallelism between the adjectival and the verbal systems is recognized, it is possible to claim that the same valence-changing processes (namely, saturation and decausativization) are operative in both systems. This assumption can predict the syntactic and semantic behavior of the two sub-classes of adjectives, as well as their composition, without resorting to operations unique to adjectival passive formation.

Relative embeddings: a Circassian puzzle for the syntax/semantics interface

Abstract: This paper documents and analyzes the pattern used in the Northwest Caucasian language Adyghe (Circassian) to express what the following five different constructions convey in other languages: headed and headless relative clauses, embedded declaratives, embedded polar interrogatives, and embedded constituent interrogatives. We argue that Adyghe encodes the meanings of all these embedded structures by means of the same syntactic construction, a relative clause. This pervasive use of relative clauses is possible due to mechanisms that are independently attested not just in Adyghe but also in more familiar languages like English. These mechanisms include concealed questions, polarity operators, and nominals such as fact and question that can connect propositional attitude verbs or interrogative verbs with embedded clauses. We suggest that this extensive use of relative clauses in Adyghe is triggered by the absence of non-relative complementizer. We further show that this use is facilitated by their morphological visibility: a relativizer realized as a prefix on the verb, verbal affixation, a rich system of applicative heads hosting indirect arguments, and the availability of a case marker suffixed to headless relatives. We conclude by discussing the implications of the Adyghe system for the general design of embedding and subordination in natural language.

NP-ellipsis with adjectival remnants: a micro-comparative perspective

Abstract: Two strategies of NP-ellipsis have been identified in the literature: (a) the elision strategy, and (b) the pronominalization strategy. The former has been said to be dependent on the presence of inflectional morphology (i.e., agreement) on the adjectival remnant. The latter strategy is used when the adjectival remnant does not carry any inflectional morphology. The aim of this article is to show that there are languages, among which Dutch, where morphological agreement appears to be the licensing factor, but where one-insertion (i.e., the pronominalization strategy) is the actual strategy. We arrive at this conclusion via an in-depth and systematic micro-comparative investigation of NPE in a number of closely related languages and dialects, more specifically: Afrikaans, Frisian, (standard) Dutch and dialectal variants of Dutch. English will be included in our analysis as well, since it is a core example of the pronominalization (i.e., one insertion) strategy. At a more theoretical level, it will be shown on the basis of close inspection of our micro-variation data that the pro-nouns involved in the pronominalization strategy have a composite structure. It will be shown that this decompositional analysis of pro-nouns brings together (i.e., unifies) the elision strategy and the pronominalization strategy. Another outcome of our study will be that languages/dialects may have available more than one NPE strategy.

Predication vs. aboutness in copy raising

Abstract: Copy raising sentences (Charlie looks like his prospects are bright) are ambiguous between a thematic and a nonthematic reading for the subject, corresponding to whether or not it is the perceptual source. On the basis of Hebrew and English data, this paper motivates a novel generalization: a pronominal copy in the complement is necessary if and only if the matrix subject is not thematic. This follows if (i) a nonthematic DP must be licensed by predication, (ii) the clausal complement is turned into a predicate by merging with a null operator, and (iii) the pronominal copy is the variable required by the operator. Contra previous analyses, I argue that the complement in copy raising may be propositional, forming an “aboutness” relation with the subject. When it is predicative, however, a null operator is necessary, since CPs are not natural predicates. The dichotomy between propositional and predicative CPs cuts across the gap/copy distinction, and is manifested in other constructions, also discussed (hanging topic vs. left dislocation, rationale vs. purpose clauses, and proleptic object constructions).

Asymmetries between person and number in syntax: a commentary on Baker’s SCOPA

Abstract: This paper is a commentary on Baker’s “When Agreement is for Number and Gender but not Person”. In many contexts, the behavior of person agreement departs from that of number and/or gender agreement; the central hypothesis advanced by Baker—the Structural Condition on Person Agreement (or SCOPA)—is an attempt to derive these departures from a single, structural condition on the application of person agreement.

Japanese loanword devoicing revisited: A rating study

Abstract: In Japanese loanword phonology, geminates optionally devoice when there is another voiced obstruent within the same stem, i.e., geminates may optionally devoice when they violate OCP(voice). This devoicing of OCP-violating geminates has received much attention in the recent phonological literature. However, the debates centering around this phenomenon have relied primarily on intuition-based data, and no systematic judgment experiments have been performed. This paper fills that gap. The experiment reported in this paper shows that Japanese speakers do find devoicing of geminates natural when there is another voiced stop within the same word, i.e., when the geminates violate OCP(voice). The experiment moreover finds other interesting aspects of devoicing: (i) the naturalness of devoicing of OCP-violating geminates correlates positively with the lexical frequencies of the words in question, (ii) the naturalness of devoicing of OCP-violating geminates is not significantly affected by place of articulation, (iii) speakers find (context-free) devoicing of geminates more natural than devoicing of OCP-violating singletons, and (iv) speakers find the devoicing of OCP-violating singletons more natural in word-medial position than in word-initial position.

The communicative function of English verb number

Abstract: This paper offers an account of English verb number solely in terms of communicative function. In this account, verb number is a full-fledged expressive device comparable to the number system for English nouns; all instances of verb number are accounted for in terms of meaning, with the caveat that meaning pertains to conceptualization, not reference. Two points emerge from this line of analysis. First, verb number is not controlled by subject number; each is a separate communicative choice governed by discourse-based principles of semantic coherence. Second, even though the subject usually precedes the verb, speakers can nevertheless think ahead and choose subject number so as to accommodate an anticipated choice of verb number just as well as the reverse, and appear to do both on different occasions. Thus, both choices are best understood in terms of symmetrical relations of textual cohesion, usually with each other but sometimes with other features of the context.

Predication and edge effects

Abstract: This paper investigates the properties of syntactic edges, with special attention being paid to two central issues in cyclic syntax: the domain and the nature of cyclicity. This paper argues for the premise that predication domains form a Spell-out domain, and that Spell-out results in order preservation of the predication domain. It is shown that elements externally merged at the edge of a predication domain observe a special ordering restriction, the Edge Generalization. The ordering restriction is explained by the interaction of two premises of cyclic syntax, coupled with a theory of probe-goal Search. Empirical evidence for the proposal comes from various sub-extraction phenomena out of edges of predication domains in Korean and Japanese. In particular, the interactions between floating numeral quantifier constructions and (primary and secondary) predication constructions are closely examined. The current proposal poses some interesting challenges to the proposition-based-phase system: it argues against the claim that only strong phases undergo Spell-out, and that edges are spelled out separately from the complement. The proposal also has some implications for the structure and typology of resultative and depictive predicates, and for the finer-grained structure of VP including aspectual adverbs and small clause complements.

Ranking and necessity: the Fusional Reduction Algorithm

Abstract: Understanding a linguistic theory within OT requires an exact characterization of the ranking conditions necessitated by data. These conditions determine the formal shape of the grammar while providing the crucial link between the data and its interpretation. We introduce an algorithm (‘Fusional Reduction’, FRed) which calculates the necessary and sufficient ranking conditions inherent in any collection of candidates and presents them in a maximally concise and informative way. The algorithm, stemming from the original proposal of Brasoveanu 2003, is set within the fusional ERC theory of Prince 2002a. In this context, the Most Informative Basis and the Skeletal Basis emerge as the two important types of reduced representations of ranking structure. We examine their properties and show how FRed produces them from data. Fine-grained FRed is compared with broad-stroke RCD (Tesar and Smolensky 1993, Tesar 1995 et seq.), and RCD is reinterpreted and embraced within FRed as a simplified, information-losing sub-case. Finally, FRed is compared with other related algorithms in structure, worst-case complexity, and relevance to the analytical enterprise. This paper revises Brasoveanu and Prince 2005, 2007; Prince and Brasoveanu 2010 gives a more formal perspective, with proof of the theorems.

Reaching agreement as a core syntactic process Commentary of Bock & Middleton Reaching Agreement

Abstract: In their paper Reaching Agreement, Bock and Middleton (2011) review a vast array of psycholinguistic experiments on semantic influences in agreement which they argue provide critical empirical evidence to the longstanding debate about the role of meaning in syntax The authors propose to unify these findings within the Marking and Morphing model, the reference framework for many psycholinguistic studies of agreement production In this commentary, I discuss four concerns about the approach advocated by Bock and Middleton: (1) the pervasive confusion with respect to the definition of agreement, and its conceptual consequences on the debate about the role of meaning in syntax, (2) the infelicitous comparison between pronouns and verbs providing the empirical foundations of Marking and Morphing, (3) the existence of a set of experimental findings invalidating the assumption of the model with respect to the relation between feature transmission and morphology, (4) the lack of assumptions of Marking and Morphing with respect to the process of feature transmission, hence its inability to account for the structural effects on attraction In response to these concerns, I present an alternative model, Selection and Copy, and sketch a line of research that explores the workings of the Copy component I then address the criticisms raised by Bock and Middleton against this research and question the explanatory force of Marking and Morphing as a model of agreement defined as a core syntactic process

High tone in Moro: effects of prosodic categories and morphological domains

Abstract: This paper presents a description and analysis of the tonal system of Moro, a Kordofanian language of Sudan, showing that the distribution of H(igh) tone is sensitive to a number of morphological and prosodic factors. First, we demonstrate that the distribution of H on nouns is sensitive to the OCP, both within roots and with affixes. Nouns also exhibit lexical distinctions between forms that exhibit unbounded rightward spreading of H and those that show no spreading. We model this distinction using cophonologies. While the distribution of H on Moro verb stems bears some similarities to nouns, crucial differences emerge. Rightward H tone spreading is binary on verbs, and sensitive to the weight of the syllable in terms of both the presence of an onset and a coda. We model this effect as H tone spreading within a binary foot. Furthermore, unlike nouns, underlying representations play little role in the distribution of H on verb roots. H tone is predictably distributed within a morphological category, the derived stem (D-stem), similar to a constituent recognized in Bantu languages (e.g. Downing 2000). Finally, we analyze competition between H associated with the D-stem and H associated with affixes. This H tone competition is an OCP-driven effect occurring within the macrostem.

Phi-feature inflection and agreement: An introduction

Abstract: The linguistic literature abounds with discussions of phi-feature inflection. The formalist tradition has, in a variety of ways, approached this phenomenon predominantly in terms of an agreement relationship between two terms of a syntactic structure—for instance, the subject and the finite verb (typically showing agreement for person and number), or the object and a past participle (which, e.g. in the Romance languages, may agree for number and gender, but not for person). But not all these agreement relationships affect all phi-features equally, which raises the question of whether there is to be a unified approach to phi-features in general. And two terms that can entertain an agreement relationship for certain phi-features do not seem to engage in such agreement every time they might be expected to so do. On the surface, plural subjects can co-occur with singularly inflected finite verbs, and vice versa; and sometimes a subpart of the subject seems to control the selection of the inflection on the finite verb, in so-called ‘attraction’ cases. These kinds of phenomena give rise to an in-depth exploration of the nature and reality of agreement relationships, including the possibility of an analysis treating phi-featural properties as autonomous vis-a-vis one another, assigned to each term separately, not under agreement. This special issue brings together a collection of papers and commentaries reflecting on these matters in various ways. In this introduction, I set the stage for the discussion to follow.

[+wh] complementizers drive phonological phrasing in Fukuoka Japanese

Abstract: Evidence from the intonation of wh questions in the Fukuoka dialect of Japanese is used to explore a recent proposal by Richards that the syntax-prosody interface includes a requirement for wh elements and their associated complementizers to be contained in the same phonological phrase. An examination of multiple wh questions and nested wh questions determines that this phrasing requirement originates with the complementizer, not with the wh element.

Overwriting as optimization

Abstract: Recent optimality-theoretic analyses of fixed segmentism reduplication and root-and-pattern morphology invoke a technique of overwriting: Phonotactic constraints prevent the cooccurrence of an affix with the unmodified reduplicant or base word, and the resulting conflict is resolved by replacing part of the lexical material with the affix (Ussishkin 1999; Alderete et al. 1999). Nevins (2005) claims that this type of approach both overgenerates and undergenerates since it predicts unattested overwriting patterns, and cannot account for specific phenomena in Hebrew and Hindi. In this paper, we show that a substantial part of the overgeneration argument is empirically flawed, and argue that all remaining problems find a straightforward solution in the independently motivated parametrization of optimality-theoretic constraints and a general metacondition on possible constraint rankings.

Relatives with a Leftward Island in Early Modern English

Abstract: I describe a type of relative clause found in 16th–19th century English This construction, the Relative with a Leftward Island or RLI, is characterised by a cluster of unusual properties The relative pronoun is a definite, anaphoric pronoun, apparently semantically identical to that found in regular English appositive relatives, but the syntactic structure containing that pronoun is quite distinct from that of regular relative clauses RLIs are biclausal structures, syntactically independent of the antecedent of the relative pronoun, with the first clause left-adjoined to the second The relative pronoun occurs at the left edge of the subordinate, left-adjoined clause I provide a synchronic analysis of this construction, and a sketch of the diachrony of relative clauses around this time, a period in which many constructions emerged, spread to some extent, and then disappeared within a century or so, without ever becoming fully widespread The analysis offered here touches on many areas of syntactic theory, including island pied-piping, null subjects in non-pro-drop languages, resumption, the distribution of adjoined positions, and properties of movement and binding

And now for something completely different: Reid on English verb number

Abstract: Reid (2011) provides an innovative, thoroughly worked out, and rigorously argued alternative to “agreement” as an account of verb number marking in English, from a communicative and functionalist perspective. There are some points of contact between his paper and the other contributions to the present issue, but the differences far outweigh the similarities, so the contrasts will be the main focus of this commentary. The similarities include the general topic of “agreement” and a distinction between what counts as linguistic knowledge—arbitrary units of grammatical structure—and what counts as extralinguistic, such as information derived from context or knowledge of the world. Differences include the status of the notion of “agreement”, what counts as data, what counts as an explanation, the theoretical constructs posited and what counts as evidence in support of them, what is meant by “meaning”, and what kinds of predictions are made.

Remarks on omnivory and complementarity: a commentary on the paper by Andrew Nevins

Abstract: This commentary examines Nevins’ Multiple Agree (MA) approach to complex agreement phenomena, and in particular the two patterns that Nevins identifies as omnivorous number and person complementarity. Nevins analyzes both as outcomes of the MA mechanism and attributes the categorical split between person and number to ontological differences in the feature inventories: person features are binary and fully specified, while number features are unary and underspecified. I argue that the opposition between person and number is strained insofar as there exist contexts where person, too, patterns as though it were underspecified, giving rise to the omnivorous agreement pattern. I also show that the MA mechanics do not fully predict that number agreement should be omnivorous across the board. Auxiliary assumptions restricting possible probe structures are required.

Erratum to: Spotting, collecting and documenting negative polarity items

Abstract: Richter, Frank, Manfred Sailer, and Beata Trawinski. 2010. The collection of distributionally idiosyncratic items: An interface between data and theory. In Corpora, web and databases: Computer-based methods in modern phraseology and lexicography, eds. Stefaniya Ptashnyk, Erla Hallsteinsdottir, and Noah Bubenhofer. Vol. 25 of Phraseologie und Paromiologie, 247–262. Baltmannsweiler: Schneider Verlag Hohengehren.