1. It is now a commonplace that one of the functions of intonation in English is to divide a discourse into ‘units of information’ and to structure these units by singling out within each one a focal part which contains what the speaker is presenting to the listener as ‘new’, in the sense of ‘newsworthy’ (Halliday, 1967; Quirk et at., 1972; Bolinger, 1958, 1972). The remainder of the information unit (if there is a remainder - some units are all focus) is said to contain ‘given’ information. 2 According to some versions of the theory, many information units – in fact the majority of those with ‘unmarked tonicity’ (Halliday), or with the focus in the ‘neutral position’ (Quirk et at.) – are ambiguous in structure, and the choice between two or more structural possibilities is made in accordance with the context. The principle is illustrated in the following examples, in which the A-sentences provide the context, the nuclear syllable is marked with an accent, and the focal unit is in capitals (Halliday's term for the focal unit is the ‘domain of focus’; Quirk et al. speak of the ‘new element’ or the ‘unit marked as new’)

In a recent paper in this journal, Hawkins (1980), it was observed that there was considerable, but quite principled, variation both in the types and in the quantities of word order co-occurrences found in Greenberg's (1966) data (cf. his Appendix I (AI) and Appendix II (AII), reproduced here as Tables 1 and 2 respectively). We formulated two types of universal statements to account for this data: implicational universals, which define permitted versus non-permitted co-occurrences of word orders (matching the attested versus non-attested co-occurrences respectively), and which are of the form ‘if a language has some property (or properties) P, then it will also have some property (or properties) Q’; and a ‘universal of language distribution’, entitled the principle of ‘Cross-Category Harmony’ (CCH), which predicts the different quantities of languages having the permitted co-occurences.2 The present paper is intended as a sequel to the last. We now relate our universals to the cross-category, X-bar, generalizations of generative grammar and to issues of markedness, suggesting that our universals and X-bar theory can be of mutual benefit to one another.

Recent work by Lightfoot (1976a, 1977) and Malsch (1976) has led to conflicting hypotheses about the underlying order of old English. Lightfoot hypothesizes that OE is an SOV language, while Malsch hypothesizes that it is a VO language. Although Malsch's argument is far more rigorous than Lightfoot's, both arguments are unconvincing because both suffer from a conspicuous lack of attention to the data available in surviving OE texts:Lightfoot asserts the nonexistence of constructions which do in fact exist (1977: 211), while Malsch cites only constructions drawn or elaborated from secondary sources (39, note 1). Furthermore, the two arguments are not directly opposed to each other. Lightfoot argues for a verb-final underlying order and an NP Postposing rule, while Malsch argues for a verb-initial underlying order and a Verb Postposing rule. Neither argues against the postposing rule of the other. At the same time, their only common assumptions seem to be that underlying representations of sentences have linear ordering and that surface auxiliaries in OE are derived from higher verbs.

Greenberg's pioneering work on word order universals (Greenberg, 1966) in which he proposed a total of universals concerning the ‘order of meaningful elements’ as well as a typology based on this order distinguishing 24 language types has truly had a great influence on linguistic research in the past few years. Not only has this important contribution to our understanding of language universals and typology precipitated such widespread interest in this particular field of linguistics which has culminated in a number of language specific monographs (see for instance Heine, 1976; Lehmann, 1974; Friedrich, 1975); symposia (see Li, 1975, 1976, 1977) and even larger projects (for example the Language Universals Project of Stanford University, and the Cologne project on language universals, see Greenberg, 1978; Seiler, 1973, 1978), but it also has influenced linguistics in general to such an extent that many consider syntactic description and research to be incomplete without a thorough discussion of word order, in some cases to the virtual exclusion of other topics.

Analyses of word-meaning can usually be divided into those which assign a meaning to a word IN ISOLATION FROM a specific context of use, and those which regard the meaning of a word as being largely, if not entirely, DEPENDENT UPON a specific context of use. The ultimate expression of the latter ‘polysemantic’ approach is probably that of the later Wittgenstein who argued that ‘every difference in a word’s use is a consequence of and evidence for a difference in its meaning’ (Wertheimer, 1972: 49). One recent example of the former ‘monosemantic’ approach is that of Bolinger (1977) whose stated purpose is to ‘reaffirm the old principle that the natural condition of a language is to preserve one form for one meaning, and one meaning for one form’ (p. x); also, the proponents of lexicalism are likewise ‘monosemantic’ in inclination in that they stress the primacy of lexical units over the syntactic relations which exist between them. Both these approaches have their advocates, and it should be stated at the outset that neither is necessarily right or wrong; each can only be judged according to whether the phenomena it is used to interpret are thereby illuminated.

The Slavonic notion of ‘vid’ (= ‘aspect’) has recently received much attention from the viewpoint of general linguistics. Not surprisingly, perhaps, some of the major contributions to the study of aspect as a general metalinguistic category are offered by linguists who are anxious to treat aspect in terms so general that the category applies to as many specific problem areas in non-Slavonic languages as possible (such as the opposition between the so-called progressive and non-progressive forms in English and the opposition between the imparfait and the passé simple in literary French). However, in my view, a satisfactory assessment of aspect in non-Slavonic languages can only be arrived at if we obey the following two conditions: (1) the general metalinguistic category of aspect must be defined independently of problem areas in non-Slavonic languages; (2) the relationship between aspect and other categories traditionally associated with the verb (especially tense and Aktionsart) must be specified.

Existential sentences have usually been defined on the basis of their morpho-syntactic characteristics. In English, the term has been used to designate those sentences in which the unstressed, non-deictic there occurs. It has been further observed that most such sentences contain the verb be, an indefinite NP and a locative adverbial following there in that order. Despite this syntactic characterization, however, the term ‘existential sentence’ has been taken, erroneously, to refer to some semantic features of the sentence as well, and so it has been generally assumed that existential sentences always assert the existence of some entity.1

In Biblical Hebrew, the tendency for rhythmic stress was primarily manifest in stress retraction, which moved word-final stress to the preceding syllable so as to avoid ‘stress clash’ with an immediately following word-initial stress, subject to certain restrictions. It was discussed in detail by the traditional grammarians of Hebrew; recent generative accounts include Prince (1975) and McCarthy (1979). In Modern Hebrew, the tendency for rhythmic stress can be realized in a variety of ways, of which stress retraction is by no means the commonest.

One of the most characteristic features of German sentence structure is the so-called ‘frame’ (Rahmen, Klammer). According to the ‘frame’ principle, complex verbal expressions in main clauses are split up in such a way that the finite portion occurs in second position (in interrogative and imperative clauses in first position), while the remainder is relegated to the end, thus enclosing most of the clause in a kind of frame.

Recent transformational analyses of liaison in French have postulated in one form or another a simple general syntactic basis for this phenomenon (Schane, 1968; Selkirk, 1974; Klausenburger, 1978). This contrasts with the traditional approach in which all the syntactic contexts in which liaison is obligatory, impossible, or depends on stylistic factors are painstakingly listed (Fouché, 1959; Delattre, 1947, 1955, 1956). This may appear to be due to the fact that the traditional approach lacked an appropriate syntactic theory which would unify the apparently unmotivated series of contexts in which liaison capriciously wandered. Working within the framework of the Syntactic theory, Selkirk (1974) was able to discover a small set of simple rules to account for a wide array of previously recalcitrant data. This success has sometimes been taken as an argument for the superiority of the syntactic theory, cf. Jackendoff (1977).

In this paper various conclusions will be drawn from a comparison of productive and non-productive German verb morphology; more specifically, it examines some of the properties of the morphological component of a generative grammar of German.1

Two assumptions are made at the outset, which clearly distinguish the position adopted from that of the standard model of transformational phonology as exemplifed particularly in Chomsky & Halle (1968) (SPE). The first of these refers to the position of morphological rules in a grammar: it is that, for at least a large number of languages, the phenomena of morphology are not appropriately dealt with by syntactic, readjustment and phonological rules, in the manner of SPE, but that distinct morphological rules are required, rules which are, in particular, quite distinct from phonological rules.

Langendoen and Bever (1973) contended that both expressions in (1) are ungrammatical in English, despite the acceptability of the first to native speakers of English.

(1) (a) a not unhappy person

(b) a not sad person

We came to this conclusion because we believed, first, that all grammars of natural languages should be subject to a constraint ‘M’ that no syntactic rule is able to make use of the morphological structure of lexemes (1973:402); and, second, that we had shown that no grammar that satisfies M can distinguish the grammaticality of(1a) from that of (1b). Our exact wording of M was as follows: ‘no syntactic transformational rule is permitted to make use of the internal morphological structure of lexical items’. Since the only syntactic rules of the standard theory, within which we were operating that are not transformations are base-categorial rules, which by definition make no use of the internal morphological structure of lexical items, the word ‘transformational’ in our formulation of M is unnecessary.

This paper raises the question whether transitive or intransitive semantics is the more basic. Recently, the transformational analysis of causatives (Lakoff, 1970; McCawley, 1968) has proposed that transitives are derived from intransitives by a causative transformation — thus implying that intransitive semantics is more basic. Because the transformational model has been so influential2 this claim has been carried over, often unquestioningly, into the grammar of many languages such as Hindi (Kachru, 1966, 1971; Kleiman, 1971). This paper shows that of significance in determining basicness is the question whether intransitive verbs of the form (O_) represent natural activity or agent-motivated activity; and using this criterion, we expect most transitives to be basic.

In this paper I shall present a truth conditional semantics for simple adjectival comparatives. The leading idea is that the semantics should hug closely to the surface syntax of the construction and should not depend on postulating degrees or extents as primitive entities. The remainder of this introductory section will raise some objections to existing accounts of the semantics of comparative constructions. In Section 2 it will be argued that an adequate analysis of compared adjectives must take into account the vagueness of positive adjectives, and that the latter can be captured by interpreting degree adjectives as partial functions. Section 3 contains an analysis of degree modifiers, and in Section 4 this analysis provides the basis for the interpretation of comparatives. Section 5 spells out some consequences of the theory, while the last section contains a summary and conclusion.

English rather than clauses display obvious surface similarities with comparative clause constructions. The parallelism between expressions where rather than is followed by clause remains, and such ordinary comparative expressions as better than, taller than, etc., followed by a clause or clause remains, suggests that the rather than expressions are just comparatives, derived in much the same way as the more familiar comparative forms. However, the derivation of rather than expressions as comparatives is not entirely straightforward. First, although rather has the form of a comparative adjective, rath+er (as required in comparative clause constructions), this is apparently a fossil form: in current English there is no adjective rath of which rather is the comparative. Thus, any treatment that derives comparatives from structures containing compared adjectives (e.g., -er much tall in Bresnan [1973]), would have to postulate an ajective rath, never realized in modern English, in order to give a unified treatment of rather than expressions as comparatives. Part of the purpose of this paper will be to argue for just such an account.

The industry of explaining away apparent cases of extrinsic rule-ordering in phonology has been in recession recently. Kenstowicz and Kisseberth, in a recent textbook (1979), make rather soothingly conservative noises about ordered-rule solutions. In this lull, during which (it is clear) ordered-rule solutions are still reckoned to be a possibility, Vago's claim (1977) to be able to substantiate such a position has gone unchallenged.

In 1980 Koutsoudas published a paper entitled ‘The Question of Rule Ordering: Some Common Fallacies’. In this paper he attacks various types of criticism of the theory of intrinsic rule ordering, and gives a set of procedures by which he allows his version of that theory (henceforth the KSN theory2) to be falsified. In the present article we utilize the procedures prescribed by Koutsoudas to demonstrate that this theory is not tenable.