Search

Using the Visual World Paradigm, the current study aimed to explore whether the mass/count distinction is determined by syntax in Mandarin Chinese, focusing on classified nouns in nominal phrases. By using dual-role classifiers, ontological count and mass nouns, and phrase structures with and without biased syntactic cues we found that the mass/count distinction is initially computed using phrase structure but can be overridden in cases where the syntax is incompatible with nouns' ontological meanings. The results indicate that in Mandarin Chinese, syntactic cues can be rapidly used to make predictions about upcoming information in real time processing.

The introduction to this book reviews detail the major claims put forth in RoN in 1970, and in particular, the claim that complex words, with deverbal nominals being the case at point, represent a formation that is neither predictable nor productive, and are hence lexically listed. This claim goes hand in hand, in RoN, with the claim that whatever similarities do hold between the deverbal nominal such as destruction and the verb destroy emerge from the existence of a category neutral listed form, DESTROY, which has a consistent subcategorization frame (an object in this case), which is realized identically in the syntax, in accordance with the X’-theory, and where the form DESTROY itself inherits its category from its categorial insertion context (N, V etc.). Since 1970, a rich body of studies has emerged which investigated the properties of lexical formations such as destruction and their relationship with the verb destroy, giving rise to multiple accounts of the emergence of complex words, as well as to the emergence of distinct argument structure combination in the context of nominalizations in particular, and word formation in general. Particularly influential was Grimshaw’s (1990) work, which introduced a typologically sound distinction between nominalizations with event structure (Complex Event Nominals, or Argument Structure Nominals) and nominals which lack event structure, and which may be result nominals or referential nominals or Simple Event Nominals, i.e. nouns which denote an event, but which do not have an event structure in the verbal sense (e.g. trip). More recently there has been the questioning of the partition between word formation and syntactic constituent building altogether, starting with Marantz (1997), and continuing with influential work by many of the contributors to this volume. This volume brings together a sample of contemporary approaches to nominalization, based on the historical record, but also branching into new grounds, both in terms of their syntactic approaches, and in terms of the range of languages considered.

In her chapter ‘Nominalizing verbal passives: PROs and cons’, Borer argues that nominalization, and by extension many other morphological processes, must be syntactic. Borer focuses on so-called short argument structure nominals (SASNs), i.e. ASNs which are missing an overt logical (external) subject, and which do not obligatorily take a by-phrase. Borer provides evidence that SASNs embed a passive structure, with the latter showing most of the syntactic properties of clausal verbal passive, including the promotion of the internal argument. Nominalization is thus an operation which can combine a passivized verbal extended projection with a higher nominal head. Long ASNs, in turn, are nominalizations which bring together a nominalizer with an active Verbal Extended Projection, ExP[V], complete with all its arguments, including the external. ASNs (de-verbal/de-adjectival), according to Borer, therefore must contain a verbal/adjectival ExP, and the argument array in ASNs is that which is associated with the embedded ExP[V] and ExP[A] respectively, and not with the noun. This in turn means that the operation Nominalization, which brings together a verbal/adjectival stem with a nominalizing affix, must be allowed to apply to the output of syntactic operations which involve complex syntactic phrases, including passive and movement.

This article argues that important constraints on the properties of derived nominals can only be explained if complex words, and specifically derived nominals, are syntactically derived and if noncompositional Content, in essence conceptual meaning, is constrained by syntactic locality.

Nous nous interessons dans cet article aux adjectifs (supposes) construits comme des expressions nominales. A partir de donnees de l’anglais, du francais, de l’hebreu et de l’espagnol, nous montrerons que ceux-ci appartiennent a deux classes distinctes. Une classe restreinte est constituee de noms veritables homophones avec les adjectifs correspondants, mais dont le sens n’est, bien que relie, pas directement derive de celui de l’adjectif. La classe la plus ample et productive est constituee de veritables adjectifs epithetes modifiant un nom nul. La variation des interpretations que le N nul peut recevoir, ainsi que la variation a travers les langues, s’avereront dependre des conditions d’identification et de legitimation des noms nuls dans une langue et une structure donnees.

Children, it is well known, go through a developmental stage in which they omit functional material, a fact that is often attributed to a missing of deficient functional structure in the early grammar. We argue that the systematic omission of functional material, on the contrary, argues for the presence of functional structure, as in the absence of such structure what is expected is not a systematic omission of functional material but rather its random (over) use. Random use of functional material is attested in agrammatic speech in which we suggest it may indeed stem from absent or deficient functional structure. On the other hand, the early grammar is characterized by full, albeit phonologically unrealized, functional structures. Such phonologically unrealized functional structures, we suggest, are interpreted in the early grammar through Discourse-linking using principles that are available through Universal Grammar and that are otherwise attested in natural language.

This chapter focuses on the categorial properties of roots and proposes that, just like event structure, these emerge in the context of particular functional structure and as a consequence of it. For Borer, functors, whether segments of extended projections or derivational categorizers, are viewed as elements that partition the categorial space. Borer explicitly argues against linking the emergence of a category to zero-realized n, v, and a.

This volume brings together chapters that discuss the nature of roots, the core elements of word meaning. It looks not only at their syntax, but also their phonology, semantics, and morpho-phonological role (or lack thereof), insofar as these do turn out to bear on their interaction with syntax. To what degree these roots carry syntactic and/or semantic information has been the subject of much research. Some of the key questions which this book looks to address are: Do roots have any meaning at all in isolation, or is all meaning associated with constituents larger than roots? If the root has no syntactic category, what are its properties? Do roots have phonology in isolation?

Recent accounts of process nominals postulate a VP within the nominalized structure. A verb becomes a nominal by a head raising operation to a nominal affix. This view contrasts with analyses of process nominals as (pure) nominals with partial verbal properties, originally due to Chomsky (1970). Contributing to this debate, we will argue that direct evidence indicates that English process nominals contain a VP. Our evidence comes from the distribution of adverbs on the one hand, and from the presence of the VP anaphor do so in process nominals on the other. We show that a portion of the verbal extended projection specifically excluding IP or CP is present in process nominals. An array of word order facts about process nominals falls into place when we further assume that the verb is raised from VP over the subject, the object, and adverbs, adjoining to a nominal affix. Our analysis moreover adds to the evidence for functional structure above VP and supports particular claims about the syntax-morphology interface.

This chapter provides a detailed analysis of the structure of ‘long’ AS-nominals (cases in which both subject and object are fully articulated and no by-phrase is used), and further puts forth a distinction between the nominalizing affixes -ing and -ation (the latter including its ‘kin’, i.e. -ance/ence, -ment,; -al, and possibly -age) based on aspectual properties. Importantly, the properties of nominal -ing are shown to part company with those of -ing in gerunds, casting serious doubt on any analysis which seeks to give the two unified characteristics. While nominal -ing and progressive -ing do share some important properties, a unification of the two, similarly, is not a task that can be easily or trivially accomplished without deep revisions of our understanding of notions such as category or functional structure.

This chapter introduces Part I of this book, which is devoted to the structure of derived nominals. It begins with a critical review of analyses of derived nominals offered in the history of generative linguistics, with a particular focus on the typology of argument structure nominals (AS-nominals) and referring nominals (R-nominals) put forth in Grimshaw (1990). Accepting her important descriptive results, questions are nonetheless raised concerning her specific explanation for them. The preliminary conclusion reached is that the presence of an attested verb or an adjective within AS-nominals is pivotal (contra Grimshaw), and that it cannot be maintained that it is the nominalizing suffix that determines the presence of event structure. The next section reviews the specific structure associated with event structure in the book as a whole, based on analyses developed previously in volume II of Structuring Sense. The final part outlines the structure of the book.

Taking the model constructed in Chapters 6–10 as its starting point, Chapter 11 investigates the way in which the model could be applied to Semitic morphology, a system which in terms of its phonological realization properties is about as distinct from English as can be imagined. Striking, and non-trivial support, in turn, emerges within the Semitic verbal system (and specifically Hebrew) for the distinction between roots and functors and for the distinction between root selection and functor-contingent properties. Even more strikingly, the investigation of autosegmental morphological forms provides direct and inevitable evidence for pieces of the account as outlined in previous chapters, and specifically for the proposal for the domain of Content, for the claim that Content is read of phonologically realized strings, and that, most importantly, Content is matched with strings which contain underspecified phonological information. Finally, and taking the distinct diagnostics of AS-nominals and R-nominals as a blueprint for the investigation of words with internal ExP-segments vs. words which lack them, a model is put in place which allows, on the one hand, the characterization of the vast non-compositionality of the Semitic verbal system (on a par with R-nominals), with, on the other, a nonetheless surprising degree of predictability in well-defined subdomains (on a par with AS-nominals).

This chapter is devoted to defining the syntactic domain within which atomic Content is matched with relevant strings. Assuming an Encyclopaedia, a reservoir of atomic Content items (by assumption conceptual and thus constrained), an interface mapping device (an ‘en-searching’ device) is proposed such that it matches an atomic Content unit with a qualifying string, the latter being, fundamentally, a (partially) phonologically realized syntactic constituent which is delimited by the first merging segment of an Extended Projection. Within the empirical domain, the chapter is built around two puzzles concerning the contrastive properties of AS-nominals and R-nominals While the latter may be non-compositional (‘transmission’, ‘transformation’, ‘constitution’) and may be built of roots that are not otherwise independently attested as verbs (‘nation’, ‘petulance’), the former are always compositional (‘*the transformation of the structure by the linguist’ under a technical reading) and do not allow stems that are otherwise not attested as verbs (‘*the petulance of the passenger for three hours’). Delimiting the domain of Content by the merger of ExP-segments, in turn, solves both these puzzles in a straightforward way, as it forces the verbal constituent within AS-nominals to be Contentful as a result of the presence of argument-structure assigning ExP-segments. Evidence is further provided for the matching of Content on the basis of phonological rather than phonologically abstract representations. An appendix discusses in some detail the differences between non-compositional ‘words’ and phrasal idioms as motivation for the claim that the latter must be otherwise handled.

This book is volume III of a trilogy which explores the difference between words however defined and structures however constructed. The trilogy sets out to demonstrate that the explanation of linguistic competence should be shifted from lexical entries to syntactic structure, from the memorizing of listed information to the manipulation of grammatical rules. Its reformulation of how grammar and listed items interact has profound implications for linguistic, philosophical, and psychological theories about the human mind and language. The book departs from both constructional approaches to syntax and the long generative tradition that uses the word as the nucleus around which the syntax grows. It argues that the hierarchical, abstract structures of language are universal, not language specific, and that language variation emerges from the morphological and phonological properties of grammatical functors, or more specifically, inflection. This volume applies this approach to the construction of complex words. The book develops a new model of word formation, arguing that the basic building blocks of language are on the one hand rigid semantic and syntactic functions, and on the other hand, roots, which in themselves are but packets of phonological information, and are devoid of both meaning and grammatical properties of any kind. Within such a model, syntactic category, syntactic selection and argument structure are all mediated through syntactic structures projected from rigid functions, or alternatively, constructed through general combinatorial principles of syntax, such as Chomsky's Merge. The meaning of ‘words’, in turn, does not involve the existence of lexemes, but rather the matching of a well-defined and phonologically articulated syntactic domain with conceptual Content, itself outside the domain of language as such. In a departure from most current models of syntax but in line with many philosophical traditions, then, the Exo-Skeletal model partitions ‘meaning’ into formal functions, on the one hand, and Content, on the other hand. While the former are read off syntactico-semantic structures as is usually assumed, Content is crucially read off syntactico-phonological structures.

This chapter takes as its starting point the conclusions of Chapter 9, according to which the best way to capture restrictions on the formation of complex words is by constraining the derivation itself, a conclusion that strongly suggests that the derivation must be by phase. Comparing the domain of spellout as articulated in Chapter 9 and the domain of Content matching (as likewise previously articulated), a specific proposal is made on the constraining of derivations by phase. Key claims and conclusions are: that Content matching is always optional; that Content may be matched on the basis of partially realized roots, but fully realized functional vocabulary; and that the domain of spellout, in contrast with the domain of Content matching, may extend beyond ExP-segments. The chapter, finally, brings together the various components of the grammar as they emerge from the model as a whole, including narrow syntax, formal semantics, spellout, and Content matching, proposing a model of utterances as conjunctions, in particular, of formal-semantic representations and Contented phonologically realized strings.

This chapter provides a detailed analysis for the structure of ‘short’ AS-nominals — nominals in which the logical subject is either missing altogether or is expressed with a by-phrase. It is suggested that such cases emerge from the application of a passive within the AS-nominal, effectively suppressing or demoting the subject and promoting the object, even when it occurs post-nominally in English (e.g. ‘the destruction of the city’). Cases where the logical object occurs pre-nominally (e.g. ‘the city's destruction’) are in turn analyzed as involving an additional raising to a pre-nominal specifier. Evidence is provided for the absence of a PRO subject in such short nominals, as well as for the occurrence of short AS-nominals exactly in contexts in which passive is otherwise licit.

This chapter returns to the general model, and offers a perspective on phrase structure, constituent structure, linearization and functors. While Chomsky's Bare Phrase Structure is adopted, the assumption that external merge involves lexically listed items (complete with category) is rejected, and is replaced with the claim that external merge involves roots, otherwise unspecified for category, and functors. The nature of C-functors is discussed in detail, and they are argued to be defined them as fundamentally syntactic functions which divide the categorial space, defining a categorial space as their domain, and, obligatorily transitive, defining their complement set as a distinct categorial space (their Categorial Complement Set). The notation CX[Y] is thus proposed, for a C-functor which projects, defines, domain X and which defines its complement domain as Y. The role of S-functors, fundamentally semantic functions, in turn is defined relative to the semantic value which they assign to otherwise empty sets, which are subsequently projected as segments of Extended Projections (ExP-segments). Extended Projections, in turn, are defined relative to their shared Categorial Complement Set. A somewhat speculative discussion concerning adjunction and linearization ends this chapter.

This chapter provides an introduction to the book as a whole. It begins with a detailed critical review of historical approaches to Words, focusing, in particular, on the rationale for listedness and lexical word formation, and attempting to show that from both a phonological and syntactic perspective, that move is conceptually and theoretically problematic. This is followed by a preliminary presentation of the specifics of the system to be used throughout the book, eventually to be justified in detail in Chapters 6–10. Specifically, roots are introduced as a phonological index, and two distinct functors are introduced—a categorial functor (C-functor), responsible for the division of the categorial space, and a semantic functor (S-functor) responsible for valuing otherwise null terminals to project — once valued and thus categorized — as segments of Extended Projections. Extended Projections are likewise defined as a universally fixed set of segments of Extended Projections dominating a categorial core (C-core). The chapter ends with an outline and discussion of the general organization of the book.

This chapter is devoted to articulating a model of categorial determination based fundamentally on the conceptualization of categorial selection as a partition of the categorial space, and as establishing equivalence classes. In essence, functors define not only the category which they, themselves project, but also define a complement categorial domain, which comes to be associated with their complements. If such complements are otherwise category-less (e.g. roots) they thus come to be equivalent to a category (e.g. V-equivalent, N-equivalent and so on). If the complement is already categorial (e.g. itself headed by a functor), the existence of a complement categorial domain amounts, effectively, to a checking or a selection mechanism ruling out, e.g. the merger of a V-selecting functor such as -ation with a derived adjective such as ‘available’. Crucially, the model of categorization put forth is committed to the categorization of form in ‘the form’ as N or of form within ‘formation’ as V without the presence of additional structure, i.e., in both these cases ‘form’ is crucially a terminal and mono-morphemic. As a consequence, the account is committed to the absence of zero-affixes marking ‘form’ as N or V respectively. Much of the chapter, consequently, is devoted to arguing against the existence of zero-instantiated C-functors in English. Final comments concern the status of multi-function functors such as -ing and the status of adjectives.

This chapter provides detailed empirical evidence for the presence of a verbal or an adjectival Extended Projection within Argument Structure Nominals (AS-nominals), which is in turn responsible for the emergence of event interpretation within such nominals. Arguments are based primarily on Hebrew and on English, with some additional evidence coming from French. Data arguing against the existence, specifically, of verbal structure within derived nominals, based primarily on Chomsky (1970) for English and on Siloni (1996) for Hebrew is reviewed in detail, and while some of it is found lacking, some facts are provided with an alternative explanation. Direct evidence for verbal structure within AS-nominals is based, in Hebrew, on the distribution of object markers, by-phrases, and adverbs, as well as on constituent structure distinctions between AS-nominals and other nominals, derived or otherwise. For English, further evidence is based on ellipsis, on the limited distribution of adverbs, and on the distinct properties of de-verbal nominalizations and de-adjectival nominalizations, the latter valid in French as well. Finally, evidence is presented for adverbial raising to adjectives in English derived nominals, leading the way to an account of the limitations on the distribution of adverbs in English derived nominals.

This chapter provides a summary of the major results presented in the book, as well as an overview of their significance, particularly in light of past and future research agendas. A summary is presented of the syntactic, semantic, and phonological properties of the main building blocks proposed in this book: roots, C-functors, S-functors, and Extended Projections. In view of the conclusions reached regarding the role of phonological indices and phonological realization in the syntactic derivation, this chapter urges a re-evaluation of the role of some aspects of phonology in the syntactic derivation. It specifically questions the claim that is sometimes put forward that phonology as a whole cannot be part of narrow syntax.

This chapter provides evidence for roots as phonological units devoid of Content. Borrowing a chapter from Aronoff, (1976), evidence for the pure phonological nature of roots is based on English bound roots as well as from Semitic consonantal roots. Additional evidence comes from phonological selection by roots. The claim that roots select arguments, as is claimed at times, is directly challenged. Additional evidence for the absence of Content for roots comes from a careful investigation of coercion, and specifically, from the presence of coercion for Contentful, fully categorized items, as e.g. in ‘there is much dog on the floor’ vs. its absence within non-compositional words (e.g. ‘flaky’). More concretely, it is argued that if Content is assigned to roots, the contrast is inexplicable. If, however, Content is assigned to constituents larger than the root, and specifically, e.g. to [Ndog] in the context of ‘the’, and to ‘flaky’ as a whole, rather than to ‘flake’, an account is possible, insofar as whatever Content is assigned to the constituent ‘dog’ is in conceptual conflict with the semantics of ‘much’. The inevitable conclusion, however, is that the root ‘π√flake’ in itself cannot have Content or a similar conflict would emerge. The chapter includes a general discussion on phonological relatedness and puts forth the strong claim that accounting for phonological relatedness is the primary task of any morphological system, and that any morphosyntactic system that fails to capture such relatedness it not, in actuality, accounting for the relevant subject matter.

This paper puts forth a hypothesis on the nature of the transition from VSO word order to SVO word order and, more specifically, on the status of grammars which are in a state of transition. I argue that among such grammars one expects to find one in which [SPEC,IP] is never a so-called A-position and in which syntactic verb movement to I is optional and that Modern Hebrew is an example of such a language. The first part of the paper is dedicated to evidence in favor of the optional nature of verb movement, while the second part explores the properties of [SPEC,IP] and the consequences which its non-A-status has for quantification, movement and word order.

The purpose of this paper is to argue for a maturational schedule in language development involving the gradual reformulation of UG-determined biunique relations. We investigate an early stage (about 2;0) in child Italian which shows object agreement with a participle for lexical objects, a phenomenon not attested in standard adult Italian. We provide an analysis for participle agreement in adult Italian, based on Spec-Head agreement. We then argue that the child analyzes participle phrases with direct objects as APs, rather than as transitive VPs, resulting in an extension of Spec-Head agreement to the relations between the participle and its object. This error on the part of the child, we argue, derives from the existence in the child's grammar of the maturationally-determined Unique External Argument Proto Principle (UEAPP), requiring every predicative element to have its own unique subject. Independent evidence for UEAPP is provided by the confirmed prediction that children in the object agreement stage do not use unergative verbs in the passato prossimo. Additional evidence from early Polish further confirms our analysis, and we further show that the null subject stage in child language follows from UEAPP. We conclude by outlining the maturational schedule we propose.