Phonological evidence is presented in support of the hypothesis that all restrictions on the relative order of application of grammatical rules are determined by universal rather than language-specific principles. For a systematically representative set of synchronic and diachronic facts, previously accounted for by means of non-uniyersal extrinsic-ordering constraints, it is shown that there are alternative explanations of equal or greater generality in which the relative order of application of rules is either entirely unrestricted, or else fully predictable from the forms of the rules by a universal principle of proper inclusion precedence.

Evidence has been advanced recently in favor of incorporating a generalized version of derivational history (or global rules) into phonological theory. The innovation of this theoretical device constitutes a significant departure from the standard theory of phonology as articulated by Chomsky & Halle 1968, and vastly increases the range of possible formal grammatical relationships. However, any theory that incorporates a generalized version of derivational history characterizes a wider range of relationships than need be ascribed to natural language. All empirically defensible cases of derivational history are characteristically restricted to deletion phenomena. The restricted nature of this evidence permits—indeed, demands—a highly constrained version of derivational history. Thus the NULL SEGMENT HYPOTHESIS is here proposed as a possible and appropriate constraint. This hypothesis requires the selection of one derivational marker, i.e. the ‘null segment’, which is naturally assigned to deletion output strings and is carried through the derivation to be recovered in a similarly constrained fashion at a subsequent point in the derivation. The null segment hypothesis and its related constraints thus narrowly define the set of all and only those processes in the necessary empirically defensible derivational history relationship. The role of this hypothesis also has implications for representing certain information in lexical items.

Inadequacies in the Sound pattern of English feature system, in the realm of unattained natural classes and unaccommodated contrasts, are shown; and alternative solutions to these problems are considered. One involves added features; the other involves the notational innovation of the ‘complex symbol’.

'Normal stress'—a notion frequently encountered in phonology and, especially, syntax—has never been adequately defined. Linguists have apparently made a tacit assumption that the stress in citations elicited from an informant is the same as the stress used by a speaker making a minimum of special assumptions; but this is shown to be false. It is argued that ‘normal stress’, a notion inherited from structuralist linguistics, was required by assumptions inconsistent with those of the generative framework; and that this notion, even if it can be defined so as to be consistent with generative assumptions, is not a particularly useful one.

The notion of ‘antonymous adjective’, as used in recent and not-so-recent literature, is discussed here; and the criteria for antonymy suggested by different researchers are found to yield largely the same results. It is suggested that only noun-based adjectives can be ‘antonymous’, and that a useful delimitation of the term 'antonymous adjective can be achieved in terms of the inherentness of the underlying nouns: all scalar nouns taken to be inherently possessed by some entity may serve as the base for a set of an-tonymous adjectives. This conclusion is supported by a comparison with adjectives derived from so-called inalienable nouns: the two types of adjective are found to share many characteristics. Finally, some implications for grammatical descriptions are discussed.

It is demonstrated here that the Dravidian family of languages in South Asia is cognate with Elamite, an ancient language of West Asia. This demonstration is based on 57 Elamite words (mostly verb stems) paired with corresponding Dravidian terms. The correspondences are, on the whole, straightforward and interlocking. A beginning is made in reconstructing the phonology of Proto-Elamo-Dravidian.

The nature of the verbal sign, as of the sign in any semiotic system, must be specified through the relationship between the signans and signatum, as well as through the relations of the given sign to the other signs of the system, both at the level of the signata and at that of the corresponding signantia. In defining a sign of a semiotic system, then, we must consider not only the ‘vertical’ relationship between the two components of a sign taken in isolation, but also the twofold ‘horizontal’ relations existing between the respective components of the interrelated signs. The Saussurean thesis of the arbitrariness of the sign is partial and incomplete in that it specifies only the ‘vertical’ relations, disregarding the nature of the ‘horizontal’ ones. On the other hand, the opposite views on the motivated, iconic nature of the bond between the signans and signatum, as maintained by the adversaries of the Saussurean thesis, involve exclusively the sphere of the ‘horizontal’ relations and cannot refer to the ‘vertical’ relationship, which is characterized in principle by arbitrariness and conventionality. If the verbal sign is conceived of as a unity of the ‘vertical’ and ‘horizontal’ relations, the opposed propositions concerning the nature of the relations between the signans and signatum present themselves not as contradictory, but as complementary to each other (in Niels Bohr's sense of the term), specifying with necessary completeness the essence of the verbal sign.

The notion of consciousness is seen as being important to linguistics as well as psychology, and specifically to the linguistic distinction between GIVEN and NEW information. Given information is suggested to be that which the speaker assumes to be already present in the addressee's consciousness at the time of an utterance, and an example is discussed at length. Some attention is then given to contrastiveness, a separate phenomenon which is apt to be confused with new information. There follows a brief sketch of some related work in linguistics, together with a few remarks on the status of consciousness within the mainstream of modern psychology, as contrasted with the status of the notion in earlier times. Attention is then turned to the question of why a speaker treats certain information as given, and various grounds for doing so are illustrated. The converse question of why a speaker may stop treating something as given is also discussed. Finally, it is noted that, in certain definable instances, speakers exhibit a mild degree of egocentrism in their assumptions regarding the addressee's consciousness. This point is illustrated with reference to quotations and shallow memory.

Serious difficulties with existing continuity theories of the relationship between human language and systems of communication in other animals have caused many linguists to reject the possibility of investigating this relationship. Some of the difficulties seem to be inherent in the assumption that continuity theories must be theories of comparative intellect. These difficulties can be avoided, however, by theories which place their major emphasis on the adaptive functions of communication systems in their ecological context. Recent work on the ontogeny of bird song provides a basis for such a theory, within which a new evaluation may be made of the significance of recent experiments on language abilities in chimpanzees.

Reports of developing phonologies agree that the fricatives are generally acquired later than stops, and the hierarchy of distinctive features places [+ strident] at the foot of the list on the basis of linguistic evidence. Neurophysiological evidence seems to indicate that the differentiation of [+ strident] consonants is a function of degree of myelination of the auditory nerve and the cortical bodies with which it is connected. Evidence from pathological cases supports the hypothesis that differentiation of high frequencies is impaired in cases of demyelinating disease. It is concluded that the fricative consonants are acquired last because they are auditorily discriminated last by the developing child.