One of the most important discoveries of the last thirty years is the extent to which the pattern of anaphoric interpretations is determined by the geometry of syntactic structure. As our understanding of these phenomena has steadily grown, the theory of syntax has often been driven by discoveries in this domain, and it is no accident that Chomsky's Binding Theory was a centerpiece of the principles and parameters approach of the 1980s. However, what remained accidental in Chomsky's theory, and in (...) most of the theories that have followed it, is the apparently complementary distribution of forms that support anaphora for a given antecedent. This book argues not only that the complementary distribution in question is robust empirically, but that its existence is derived by a competitive theory of anaphora. It is demonstrated in detail that the competitive theory provides a far better explanation of anti-locality, anti-subject orientation and the range of apparently exceptional distributions that have been long been problematic for other approaches, such as Chomsky's Binding Theory and the influential predication-based theory of Reinhart and Reuland. (shrink)

It seems that when the term "coreference" is used, whether in linguistics or in philosophy, there is often presumed to be a consensus about what it is, or at least about what it is in the context where the term is introduced. I don't think the term deserves to have much use at all, insofar as it disguises more interesting linguistic and pragmatic relations between nominal forms in natural language. My preoccupation with these relations issues in part from some of (...) the central concerns and distinctions introduced in Evans (1980), an essay that has had wide and, in some cases, a very deep influence on how a variety of reference issues have been addressed in modern linguistics. As a linguist, my interest is in the way in which natural language shapes what we know and from that perspective, I want to understand how natural language sets boundary conditions on how linguistic forms can be used to achieve readings that pick out the same entity more than once. (shrink)

Broadly construed, anaphors are forms that must be anteceded in a discourse, and more narrowly, as syntacticians tend to use the term, anaphors are forms that must be anteceded within a bounded, syntactically defined domain. In this short note, I focus on the difference between these two notions of anaphor and some problems with approaches to anaphora that try to collapse them by linking all anaphors to their antecedents by syntactic operations. The latter approach permits syntactic operations to exceed the (...) bounds of sentence grammar and must still introduce familiar descriptive constraints to distinguish between anaphora in the general sense, on the one hand, and syntactic anaphors, on the other. The notion of syntactic anaphor most familiar to syntacticians is the class of forms that are susceptible to Principle A (1a) of Chomsky's (1981) Binding Theory.. (shrink)

This essay explores the place of coconstrual relations, such as antecedent-anaphor relations, in a theory of grammar informed by minimalist architecture. It has been argued that the logical space created by minimalist theorizing should favor an account of coconstrual derived from the tree-building operations of narrow syntax (Agree, feature theory, Merge and its subcase, Remerge), dispensing with rules or conditions that evaluate constructed trees. On such an account, it is argued, the explanatory power of narrow syntax is enhanced and the (...) role of the interpretive component can be circumscribed. However, if coconstrual cannot be reduced to the derivational relations of narrow syntax, then we must be prepared to reevaluate the role of syntax-sensitive interpretive rules, balancing the need for such rules against any complication of narrow syntax mechanisms just to account for coconstrual. It will be argued that dependent identity relations, the form of coconstrual that is sensitive to syntactic configurations, must be interpreted from the output of narrow syntax and are not expressed within narrow syntax at all. This result unburdens narrow syntax of a class of relations that bring theoretical and empirical complications, while providing a more elegant account of coconstrual in a broader conception of the interpretive interface. (shrink)

In recent years, certain analytic proposals have been appealed to that are incompatible with fundamental principles of structure−building that appear attractive. One such principle is Extension (Chomsky,1995), which ensures that what counts as the top of the tree at a given point in a derivation restricts the class of possible operations that can apply at that point. Another principle I will show to be desirable is Insularity, which bans on interarboreal movement.

Extension embodies an intuition that appears worth preserving, namely, the notion that at any point in a derivation, what counts as the top of the tree at that moment restricts the class of possible operations that can apply. This principle, if a version of it is correct, propels an aggressively derivational view of syntax. However, Chomsky's proposal is too restrictive to permit a wide range of syntactic analyses that appear to be worth preserving, all of which violate Extension by merging (...) constituents near the top of the tree. After some 'fixes' of Extension are rejected, including interarboreal movement, I then propose to revise Extension as Exxtension. The essential idea of Exxtension is to loosen, yet make precise and restrictive, the definition of what counts as (near) the top of the tree. The new definition, to be introduced after some structural assumptions have been established, has interesting consequences for the role of thematic selection and modification once it has been enhanced by the Selection Principle, also to be introduced below. Then all of the analyses discussed that are worth preserving will be shown to be viable within the Exxtension approach, except one, for which a different sort of analysis will be sketched. (shrink)

Ken Safir, Rutgers University Following a line of thought initiated by Kuno (1972), it has been suggested that the coconstrual of first person pronouns is a model for the coconstrual of a logophoric pronoun with its antecedent. This particular proposal has been extended to the forms of logophoricity that have been observed in some African languages (e.g., Ewe, as remarked in passing by Clements, 1975, and Amharic, as proposed by Schlenker, 2000).

Ken Safir, Rutgers University ABSTRACT: It is argued that the indexicality of first person pronouns is arises from a restriction on the pronouns themselves, as opposed to any operator that binds them. The nature of this restriction is an asyntactic constant function that picks out individuals to the context of utterance (following Kaplan, 1989)). Constant function pronouns do not require an antecedent, neither an operator nor an argument, although this does not prevent them from participating in bound readings if an (...) appropriate antecedent is introduced. The notion that agents of contexts and agents of propositional attitudes are versions of the same operator-variable relation is thus rejected, along with certain less fine-grained versions of the nature of de se interpretation. Consequently, indexical pronouns such as first person ones contrast with logophoric pronouns, which are necessarily operator-bound by perspectival operators introduced by propositional attitude verbs. Scope-sensitive properties of operator-binding and the perspectival interpretations that are imposed on logophoricity distinguish the latter from constant function phenomena, which are sensitive neither to scope, as it is usually treated, nor perspectival shifts. Constant function phenomena are also detectable as restrictions on third person forms, and two such examples are lightly explored: the English generic pronoun one and the proximate/obviative distinction in languages like Fox. (shrink)

Utrecht, and at the École des Hautes Études in Paris in 2002. It is a partial description of a book ms. presently in preparation that will have the same title. The ms. will also explore the interpretation of 'generic person', 'impersonal agency' and 'arbitrary person'.

As a generative linguist and syntactician, I share the general outlook established since the 1950's that the goal of syntactic research into the nature of the language faculty is to understand the nature of syntax as it is determined by Universal Grammar. My particular research foci have been, and continue to be, on the contribution of syntactic form to semantic interpretation, on the one hand, and on the nature of the linguistic typology made available by the innate language faculty.

In this essay I will explore the syntactic expression of the notion ‘clause’ by focusing on some syntactic and semantic properties of bare infinitive (BI) complements to perception verbs in English. I shall argue briefly that perception BI complements must be clausal, and then turn in more detail to the issue of what sort of clause the BI complement must be. It will be established that the categorical nature of the perception BI complement as IP or VP is contingent on (...) selectional properties of the perception verb, properties which may be contextually influenced by modality and negation. More specifically, an interpretative distinction between ‘vision’ and ‘non-vision’see first suggested by Napoli (1988) will be shown, contrary to her proposal, to provide the key to the contrast in the selection of clausal types. The IP/VP distinction will be shown to correlate with the viability of interpretations for the BI clause that are ‘temporally independent’ of the matrix clause. Given the IP/VP distinction, it will be demonstrated further that the peculiarly limited distribution of the expletive formativethere in BI complements may be predicted from independently motivated syntactic principles that determine clausal structure. (shrink)

It is argued that most anaphors have semantic content and that the semantic content of a given anaphoric atom plays an active role in determining both its distribution and the interpretation of the sentences in which it is employed. It is first demonstrated that semantic distinctions between semantically relational anaphoric atoms predict differences between their distributions. It is then argued that all of the semantically relational anaphoric atoms respect Principle A, while semantically contentless anaphors often do not.

The primary goal of this paper is to distinguish binding from reflexivity in domains where they appear to overlap. In so doing I will argue that Principles A and B of the Binding Theory are symmetric in the domains to which they apply. This symmetry derives from a deeper unity that permits us to dispense with Principles A and B and replace them with interpretive principles that distinguish reflexivity and binding.

This essay is a contribution to the discussion, now going on for many years, concerning what sorts of identity relations should be represented in the syntax and semantics of formal grammar and what properties those relations should have. In what follows, I will use the neutral cover term coconstrual to refer identity relations of one sort or another between nominals when no particular syntactic or semantic analysis is presupposed (among which are dependent identity, covaluation and coreference). The central claim made (...) here is (I), with its consequence in (Ia) and the far-from-innocent background assumption in (II). Most of my argumentation is designed to establish (Ia) to argue that it should be understood as the natural consequence of (I) and with respect to strict readings in ellipsis contexts. Although I will also argue for (II), I rely on Safir (2004b) for a more general defense of it. (shrink)