This article offers a detailed analysis of the English suffix -ee (employee, escapee, refugee, etc.) based on fifteen hundred naturally occurring tokens of some five hundred word types. The data suggest that formation of nouns in -ee is moderately but genuinely productive, and that analyses based on the syntactic argument structure of the stem verb are unsatisfactory. Instead, formation of -ee nouns systematically adheres to three essentially semantic constraints: first, the referent of an -ee noun must be sentient; second, the denotation of an -ee noun must be episodically linked (as defined below) to the denotation of its stem; and third, a use of an -ee noun entails a relative lack of volitional control on the part of its referent. I argue that these semantic constraints taken together amount to a special-purpose thematic role that actively constrains productive use of derivational morphology.

Quatrains in English folk verse are governed by laws that regulate the patterns of truncation (nonfilling of metrical positions) at the ends of lines. Each truncation pattern (we claim 26) is adhered to consistently through multiple stanzas and defines a verse type. Our descriptive goal is to account for why these and only these truncation patterns exist. Our crucial hypothesis is that the function of truncated lines is to render SALIENT certain layers in the natural constituency of the quatrain: the line, the couplet, or the quatrain as a whole. All three cannot be rendered salient at once, so the saliency constraints conflict. Each saliency constraint also conflicts with metrical constraints, which require that metrical positions be filled with appropriate syllables and stresses. The twenty-six well-formed quatrain types each represent a particular prioritization of the conflicting constraints.

We formalize this in optimality theory (Prince & Smolensky 1993): the inventory of types is derived as the factorial typology of our constraint set; namely, the set of outputs of all grammars obtained by freely ranking the violable constraints. We also account for differing text frequencies in our data corpus by assigning each constraint a range of possible strengths, and from this develop an optimality-theoretic account of gradient well-formedness judgments.

This article argues that identificational focus, which expresses exhaustive identification and occupies the specifier of a functional projection, must be distinguished in language description from information focus, which conveys new information and involves no syntactic reordering. The properties of the two types of focus are established on the basis of Hungarian and English material. It is argued that the cleft constituent is the realization of identificational focus in English. Only-phrases are analyzed as identificational foci carrying an evaluative presupposition. The feature specification of identificational focus is shown to be subject to parametric variation: the focus operators of various languages are specified for the positive value of either or both of the features [+exhaustive] and [+contrastive].

The historical diffusion of lexical and grammatical features from one pidgin to another has been well documented for the Pacific region, particularly by Baker (1993) who argued that items were spread individually in the early nineteenth century via an ad hoc foreigner talk register. Noting the profound similarities between Hawai‘i Creole English (HCE) and the Caribbean English Creoles (CECs) which led Bickerton (1981) to propose the language bioprogram hypothesis, Goodman (1985) suggested a stronger model of diffusion, one which involved the transmission of a structurally complex pidgin or creole from the Caribbean to Hawai‘i. Holm (1986) and Dillard (1995) have endorsed Goodman’s hypothesis. This study, drawing on a wealth of pidgin/creole data spanning the previous two centuries, finds little support for Goodman’s proposal. Textual evidence shows that the nineteenth-century pidgin of Hawai‘i lacked not only the structure of later HCE but also displayed far stronger links with neighboring Pacific pidgin Englishes than the CECs. Furthermore, the creole TMA system and for-complementation patterns are revealed to have developed late and primarily (though not entirely) within the population of native-born speakers, as predicted by the bioprogram. However, while the pace of creolization was fairly rapid in Hawai‘i, HCE did not form entirely within a single generation.

A grammatical affix undergoing phonetic erosion is sometimes abruptly replaced by a conveniently available lexical stem with which it shares one or more phonological segments. The new affix has the phonological shape of the old independent stem, but acquires the basic grammatical function of the old affix, though it may also bring in a portion of the stem's own morphological and semantic idiosyncrasies. Because the old affixal form is eliminated, the historical process can easily be misdiagnosed as reflecting the gradual compression of an original syntactic construction which includes the relevant independent stem. Recognition of the system-renewing element in this particular type of stem-to-affix grammaticalization leads to awareness of the pivotal role of the inherited system in grammaticalizations, and to a general critique of historical models which see grammaticalization as a straightforward syntax-to-morphology compression.

We argue against the prevailing view that metathesis is somehow less natural phonetically than other processes and distinguished by a relatively greater phonological motivation. We survey cases of consonant-vowel metathesis—both synchronic processes and diachronic changes—with the goal of understanding how metathesis sound changes arise. We identify two types of CV metathesis, with distinct synchronic properties and distinct historical origins, and we argue that the two types do have natural, phonetic bases and fundamental commonalities.

Normally, the negation of a sentence S serves to reverse the assertion that S would have made, and does not affect other types of information S would have conveyed, such as presuppositions, implicatures, and so on. Occasionally, however, it seems as if negation is directed precisely at presuppositions, implicatures, or even at purely formal aspects of a sentence (such as intonation, pronunciation, and so on). The following are cases in point: Mary didn’t visit the pizzeria in the Vatican, because there is no pizzeria in the Vatican./I’m not tipsy: I’m drunk./He didn’t call the police, he called the POLICE.

I call such sentences DENIALS, and I argue against the Unitarian approach advocated by Horn and van der Sandt, among others, according to which denial is a homogeneous phenomenon that calls for a unified analysis. According to the alternative theory proposed here, there are several mechanisms of denial, but each of these is needed for independent reasons, and therefore no ad hoc mechanisms are necessary.

We discuss the Labovian view of the speech community against the backdrop of data from research on variation in minority languages. While members of the same speech community normally share a set of norms for social and stylistic constraints on variation and normally share a common grammar, a number of researchers have noted that some speech communities include subgroups of speakers that are unlike the rest of the community in that they observe different rules or constraints on variable usage. We provide an overview of the main types of discontinuities in variable usage which have been attested both in majority and minority languages and discuss twelve cases of discontinuity which have been documented in the speech of Franco-Ontarian adolescents residing in minority Francophone communities. We also attempt to account for the existence of these discontinuities and consider their implications for the concept of the speech community.

In some languages there are dependencies between grammatical systems, e.g. there may be fewer tense choices in negative than in positive polarity. We examine the direction of dependencies between eight types of grammatical systems, and establish a dependency hierarchy. Polarity is at the top of the hierarchy—the choices available in another system may depend on polarity, but the possibility of positive/negative specification never depends on any of the other systems considered here. Next come systems associated with the predicate (or perhaps with the clause as a whole): tense, aspect, and evidentiality. Next come systems associated with predicate arguments—person, reference classification (covering gender/noun class, classifiers, and human/nonhuman or animate/inanimate); then number. And finally case, which marks the function of a predicate argument. The rationale for this hierarchy is considered. An appendix adds systems of definiteness to the discussion.

This article presents a semantic analysis of the English progressive as an intensional operator within the framework of modal semantics proposed by Kratzer. This treatment allows a combination of the central idea of Dowty's influential analysis, that the progressive's meaning has a major modal component, with the insights of other scholars (Parsons, Vlach, and Landman). My claim is that using a more sophisticated background theory of modality allows natural solutions for the problems raised for the modal account.

Current work in syntax reexamines basic properties of movement. Under the minimalist assumptions of Chomsky (1995), movement is prohibited unless forced by grammatical considerations. From a set of comparable derivations, the one involving the least amount of moved material should therefore block other derivations. Within this framework, any cases of optional movement are problematic. We addressed this issue with experiments on stranding and pied-piping in relative clauses in 115 English learners, aged 3;5 to 11;11, and an adult control group. All subjects participated in an elicited production experiment and a grammaticality judgment experiment. Our findings suggest that pied-piping is possible in young children’s grammars only when stranding is ruled out, as predicted by minimalism. We claim the children’s responses represent the ‘natural’ grammar while the adults’ responses reflect a prescriptive artifact. We also found a discrepancy in all subject groups between production and judgments of the genitive pied-piping construction. We account for this finding with Kayne’s (1994) analysis of relative clauses.

This project collected linguistic data for spatial relations across a typologically and genetically varied set of languages. In the linguistic analysis, we focus on the ways in which propositions may be functionally equivalent across the linguistic communities while nonetheless representing semantically quite distinctive frames of reference. Running nonlinguistic experiments on subjects from these language communities, we find that a population’s cognitive frame of reference correlates with the linguistic frame of reference within the same referential domain.

An increasingly influential current in creole studies considers there to have been no appreciable break in transmission or simplification in the birth of creole languages, instead treating creoles as gradual, relatively nondisruptive developments of their lexifiers amidst conditions of heavy language contact. Central to this superstratist view is the claim that creole is a sociohistorical, but not synchronic, term. This article outlines three features which in fact render creoles synchronically distinguishable from other languages, all three of them clear results of a break in transmission followed by a development period too brief for the traits to be undone as they have been in older languages. It is also shown that an expanded data set reveals flaws in the sociohistorical argumentation behind the superstratist framework.

Shakespeare’s verse is studded with alliteration and paronomasia. The more fundamental question of a patterned relationship between sound and meaning in his Sonnets has not been answered, partly because no method of uncovering such correspondences was available. However, once groups of sounds, specifically sonorant and obstruent sequences, are examined as the locus of the sound-meaning nexus, it emerges that Shakespeare consistently aligns these sequences with relational meanings defined by the dyad of freedom and constraint. The coherence between sound and sense is thus shown to be iconic.

This report examines the behavior of a morpheme um in Toba Batak (Malayo-Polynesian) which alternates as a prefix or an infix, and argues that the observed variation is conditioned by constraints on consonant clusters. Thus, Toba Batak shows that segmental factors may crucially influence the linear position of affixes, in contrast with a wealth of cases cited in the prosodic morphology literature in support of claims that the distributional properties of morphemes are often conditioned by prosodic structure. In addition, evidence from several sources on Toba Batak suggest that um in this language has been migrating from infixed to prefixed positions over time, and that taken as a whole, the stages involved in this change are linguistically coherent. This pattern of shift has interesting implications for typology and for theories of language evolution.

One of the aims of linguistic theory is to account for language variation. This article contributes to that objective by examining resumptive restrictive relative clauses crosslinguistically. The major claims are (1) the core-grammar distinction between conventional and resumptive restrictive relatives is due to the feature composition of the relative complementizer; (2) the prohibition against pied-piping that some languages adhere to correlates with the lack of lexical relative pronouns; (3) particular grammars need to accommodate language-specific properties such as preferences for which elements may, must, or cannot acquire a phonological matrix; and (4) resumptive pronouns which appear in relatives in the absence of an island are inserted at PF for other than interpretive reasons.

Jaeger, Lockwood, Kemmerer, Van Valin, Murphy, and Khalak 1996 (Language 72.3) reported an experimental study that provided reaction time and PET neuroimaging data said to support Pinker’s (1991) theory of inflectional morphology in which rule-governed forms and exceptions are processed by separate mechanisms. The results were also taken as evidence against connectionist accounts in which a single processing system generates both types of forms. We provide a critical analysis of the study that yields three main conclusions: First, Jaeger et al.’s data do not provide strong evidence that rule-governed forms and exceptions are processed in separate brain regions. Second, there are problems with the design of the study that contaminate critical comparisons between conditions. The results therefore afford alternative interpretations related to experiment-specific factors rather than the regular-irregular distinction. Third, the dissociations between rule-governed forms and exceptions observed in studies such as Jaeger et al.’s can be accommodated by the connectionist theory. We conclude by offering suggestions for future research that would overcome the major limitations of this study and provide more decisive evidence bearing on the issues.

Neidle, Kegl, Bahan, Aarons, & MacLaughlin 1997 argues that rightward WH-movement in ASL constitutes a counterexample to claims by Kayne (1994) that all phrasal projections exhibit specifier-head-complement order and that syntactic movement is leftward. Petronio & Lillo-Martin 1997, although not adopting Kayne's antisymmetry framework, offers a critique of our analysis and a proposal involving leftward WH-movement. Here, we argue that Petronio and Lillo-Martin's interpretations of the data are incorrect and that their analysis cannot account for the facts of the language. We therefore maintain our position that ASL WH-phrases move rightward in ASL, and that universal grammar must allow the option of rightward movement.