The traditional attempts to identify grammatical functions of intonation have recently been supplemented by attempts to identify holistic lexical functions. One such, proposed by Liberman & Sag 1974 and by Sag & Liberman 1975 is a ‘contradiction contour’ having the general shape . with the sentence accent normally at or near the lowest pitch. This contour, as an exemplar of the theory of intonational wholes, is examined in the light of an alternative theory of abstract meanings attached to rises vs. falls. ‘Contradiction’ is seen as a contextual inference from the more abstract meanings.

Consistent error patterns in English past-tense forms are reported for three age groups: preschoolers, 8–10-year-olds, and adults. It is argued that, although irregular forms are rote-learned, speakers make generalizations about such forms. Such a generalization is defined as a schema which describes general phonological properties of a morphological class, and is used in organizing and accessing the lexicon. Schemas for the English past tense develop and change with age, yielding implications for both acquisitional and diachronic theory.

Comparative analysis of spoken and written versions of a narrative demonstrates (1) that features which have been identified as characterizing oral discourse are also found in written discourse, and (2) that the written short story combines syntactic complexity expected in writing with features which create involvement expected in speaking. Quintessentially literary devices (repetition of sounds and words, syntactic structures, and rhythm) are shared by written literary language and ordinary spontaneous conversation because both are typified by subjective knowing and by focus on interpersonal involvement. In contrast, expository prose and content-focused oral genres, such as lectures and instructions, may be typified by objective knowing and by focus on content.

This paper argues that sentences in the have a V frame are not a jungle of idiosyncrasies, but exhibit orderly and systematic behavior, governed by strict semantic rules. These rules can be stated in precise formulae with full predictive power. Ten subtypes are singled out: each has a slightly different semantic formula, but all have a common core, which is a semantic invariant of the have a V frame. These semantic formulae, which account for differences both in acceptability (have a drink and *have an eat) and in meaning (drink and have a drink), are stated not in terms of ad-hoc features or labels, but in an independently justified semantic metalanguage based on natural language.

Alternatives are discussed for a number of assumptions fundamental to established linguistic theories. These alternatives lead to a coherent view of linguistic structure that treats grammar as a symbolic phenomenon, and emphasizes the importance of analys-ability to grammatical structure. A descriptive framework called space grammar is outlined, in particular its approach to semantic structure and grammatical valence relations. The concepts are applied first to nominal structure and then to the English passive construction. It is claimed that be, by, and the participial morpheme are all meaningful, and that passives are not derived from an active clausal source.

It has been observed that many languages exhibit a semantic, syntactic, or morphological correlation between the direct objects of transitive verbs and the surface subjects of certain inactive intransitives. The Unaccusative Hypothesis (UH) proposes, within the framework of Relational Grammar, that final subjects of this type are initial direct objects. This paper shows that several morphological and syntactic processes in Georgian refer to just these nominals, thus supporting the UH. It further shows that, although the semantic relations of nominals to verbs is not irrelevant, those morphological and syntactic processes which refer to initial direct objects cannot be stated simply on the basis of semantics.

English questions lacking an auxiliary (e.g. You happy now?) fail to occur in certain tenses. The first part of this paper suggests that the phenomenon derives from the interaction of the Recoverability Condition (Chomsky 1965) with markedness conventions for tense structures. The second part shows that reduced yes–no questions differ from reduced WH-questions as regards their sensitivity to person, number, morphological redundancy, and contraction. It is argued that reduced yes–no questions and reduced WH-questions do not constitute a unified phenomenon, and that the asymmetries in the two classes reflect properties of distinct components in the grammar.

A number of classical problems in the phonology of French are considered, especially as they relate to the vowel ‘shwa’ (e-muet). From the rule of Closed Syllable Adjustment (which relates shwa and [ε]), it is concluded that French phonology must be described in terms of representations making explicit provision for syllabic, as well as segmental, structure. The formal representation of syllable structure is discussed, and a concrete proposal is made. On this basis, it is suggested that the vowel shwa should be regarded as a phonological syllabic nucleus unfilled by any phonetic content in underlying structure—analogous to ‘empty nodes’ in syntax. It is shown that this representation provides a natural treatment of the facts of vowel alternation and elision that have been problematic for previous accounts of shwa. A particular advantage of this analysis is the resulting description of ‘h-aspiré' words: these can be characterized as vowel-initial words that are exceptions to a single general schema of ‘enchaînement'. Implications for other areas of French phonology, including the treatment of nasal vowels, are also suggested.

The pragmatic structuring of a Guaraní sentence is realized primarily by constituent structure. The syntactic elements of a sentence are partitioned into pragmatic components which, when related syntagmatically by scope and pragmatic function, typically form a right-branching structure. In the unmarked case, the pragmatic nucleus has unmarked Topic plus Core. An expression becomes marked when (a final segment of) the nucleus is given a binary pragmatic structure, highlighted by intonation and other features, with the marked expression as the first constituent. The remaining constituent may in turn be pragmatically structured. Marked pragmatic structuring reflects specific pragmatic conditions. This analysis accounts for most occurrences of marked wordorder, the placement of the intonation center and other intonation peaks, certain free subjects (both nouns and pronouns), various particles, and other boundary phenomena.

An analysis of English Expletive Infixation (as in fan-fuckin-tastic) in terms of a metrical theory of prosody is presented. It is shown that the major environment for Expletive Infixation—immediately before a stressed syllable—follows from independently motivated characteristics of this theory. Further support for this metrical theory is adduced from infixation in words with dactylic stress alternation and with internal stress-neutral junctures, and from the subordination of stress in forms after infixation.

The Western Muskogean languages Chickasaw and Choctaw have a subject/oblique opposition in nominal case-marking and syntactic relations, despite their ‘active’ system of pronominal agreement. In both languages, case assignment may reflect the operation of productive relation-changing rules. These facts call into question the classification of Choctaw by R. Van Valin and W. Foley (e.g. 1980) as ‘role-dominated’. Restrictions on NP density (the number of nominal arguments per clause) may account for some of the syntactic differences between the more familiar ‘reference-dominated’ languages and languages of the Western Muskogean type.

The distinction between contactive and non-contactive causation shows up repeatedly in languages (Masica 1976, Chap. 3; Cole 1976), and is claimed by Xolodovič 1969 to be ‘fundamental’ in their description. Arguing against earlier descriptions, this paper provides an alternate analysis of this important contrast.

This paper presents a description of an interestingly rich subset of English in terms of Generalized Phrase-Structure Grammar, and compares this description with some other recent treatments. The domain covered includes the auxiliary-verb system together with a number of subsystems that crucially interact with the analysis of auxiliaries: Do so Ellipsis, VP Fronting, VP Negation, VP ‘Deletion’, Sentential Adverb Placement, and ‘Subject-Auxiliary Inversion’. The theoretical framework is strongly equivalent to one using only context-free phrase-structure rules. Featurally complex category symbols, finite rule schemata, and metarules are employed to capture generalizations; but these devices do not alter the expressive power of the system, which is considerably more restrictive than systems of generative grammar in use elsewhere. Nonetheless, the description fares well when compared with rivals; e.g., a fairly large number of difficulties found in ‘Affix-hopping’ analyses, and not found in the present treatment, are noted. The present description is also compared briefly with that of Pullum & Wilson 1977, and found superior. Finally, the analysis of Akmajian, Steele & Wasow 1979 is criticized in considerable detail, and shown to compare unfavorably with the one given here.

The work of Jakobson—and later, of Greenberg—has developed the notion of MORPHOLOGICAL MARKEDNESS. In this view, certain morphological categories (e.g. noun singulars) are considered unmarked or more basic in relation to others (e.g. noun plurals). It is shown here that there are some principled exceptions to the general markedness conventions. One example of such LOCAL MARKEDNESS, as this type of markedness reversal will be called, is that nouns whose referents naturally occur in groups or pairs show many of the effects of being unmarked in the plural. The consequences of such local markedness in terms of language acquisition, paradigm regularization, lexical borrowing, and the formation of double morphology are explored in some detail.

Two sorts of analyses have been proposed for the constituent structure of the infinitival complements found in sentences like John tried to leave and John appeared to leave. One analysis maintains that such ‘bare’ infinitives are sentential, the superficially absent complementizer and subject being represented by lexically empty categories. The alternative holds that they are a type of verb phrase complement, represented as they appear superficially, lacking both complementizer and subject. The present paper argues in support of the former view, and more generally that all infinitives are clausal. A specific analysis is presented, with a range of arguments bearing on the alternatives considered.

In conversations involving more than two people, most utterances are intended to be understood not only by the people being addressed, but also by the others. These utterances cannot be accounted for in current theories of speech acts unless several basic changes are made. In our proposal, the speaker performs two types of illocutionary act with each utterance. One is the traditional kind, such as an assertion, promise, or apology; this is directed at the addressees. The other, called an informative, is directed at all the participants in the conversation—the addressees and third parties alike. It is intended to inform all of them jointly of the assertion, promise, or apology being directed at the addressees. We present evidence that every traditional illocutionary act is performed by means of an informative.

‘Ad-Verbs’ modify (i.e. ‘change') verbs. Adverbs internal to a VP are Ad-Verbs modifying the verb that heads that VP. VP-external adverbs may be interpreted as Ad-Verbs only with respect to some higher verb (e.g. an auxiliary or a ‘verb' in the translation into an interpreted logic). Basic Ad-Sentences also exist; but neither they nor Ad-Verbs are adequately translated as functional operators applying to independently evaluated arguments. Adverbs typically translate into expressions like variable-binding operators which introduce the ‘variables’ which they ‘bind’. Adverbs signal re-evaluation of the expressions on which they operate, helping to ‘build’ logical form.

Black English has, in addition to the motion verb come, a come which expresses only speaker indignation. To some extent it shares grammatical properties with auxiliary verbs. Typically, it occurs in utterances that can be taken as identical to ones occurring in non-Black dialects; and it is apparently for this reason that it has remained undetected. The existence of this come indicates that an even greater difference exists between Black and White speech than previously thought. It is also of interest that the semi-auxiliary come occurs even in acrolectal varieties of Black English. This reveals that the postcreole continuum in the United States differs from that in Guyana as described by Bickerton 1975, who states that forms which are identical to ones in the base language, but functionally distinct, are not part of the acrolect.

Two approaches to the problem of interpreting utterances are sketched: a structural account at the levels of sentence and speech act, and an interactional account of their functions and consequences as constitutive of the social order in which they occur. For the problem of utterance interpretation, linguistic analyses can specify potential meanings and functions, but cannot indicate actual interpretations to which conversationalists are oriented. Linguistic analysis alone renders an account that is propositionally ambiguous, functionally equivocal, and interactionally indeterminate. In order to account for how talk becomes determinate for conversationalists, a pragmatically-based interactional approach is offered. A segment of actual conversation that occurred during a first-grade lesson is analysed here, with focus on a description of the contexts in terms of which talk is understood. One utterance in particular, characterized as a pragmatic counterfactual, is discussed at length as a unique product of an interactional account of conversation.

This paper has been inspired by ideas contained in W. P. Lehmann's work on syntactic typology (1978). It discusses a set of factors held to be relevant for the ways subject, verb, and object may be linearized in various languages: these are sentence depth, the cognitive factor, and relative structural independence. Various linearization patterns, i.e. SVO, SOV, VSO, VOS, OVS, and OSV conform to these factors in different degrees. A hypothesis is tested concerning the statistical correlation between the distribution of various linearization patterns among the languages of the world and their degree of conformity to the set of above-mentioned factors.