In the first year of its existence, Leonard Bloomfield asked and answered the question, ‘Why a Linguistic Society?’ When it was twenty-one years old, he was satisfied that the Society had ‘come of age’, and he enumerated some of the benefits that had accrued from its activity. Now, in its fiftieth year, we can say that it has achieved a comfortable maturity; but it is also faced with new problems and responsibilities. What in 1945 was still a modest young society has shown extraordinary growth, beyond anything that could have been dreamed of in Bloomfceld's day. As we pause briefly to look back, are we justified in thinking that the Society has fulfilled the hopes of its founders ? Can we draw on the experiences of these years in charting our course into the future?

A special abstractness problem in generative grammar—that of intermediate derivational stages containing elements or sequences that are not well-formed as surface representations—is examined. Suspicious analyses employing such FALSE STEPS are cited from the phonological and syntactic literature. It is argued that false steps cannot be ruled out in general, because there is support for many false-step analyses, in both syntax and phonology; a Welsh morphophonemic case is treated in some detail.

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.

The notion of deep structure has served to crystallize a number of recent controversies in syntactic theory. This paper examines several arguments which have appeared in the literature, chiefly in Chomsky's ‘Some empirical issues in the theory of transformational grammar’, in support of the notion. It concludes that these arguments are not compelling, and that the issue of the existence of deep structure is not an empirical one at present.

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.

If the grammars which linguists write are to reflect psychologically real constructs, then purely structural evidence is not sufficient proof of them; some form of psychological evidence is required. The psycholinguistic test described here was designed to show whether or not a Hindi word such as [ghõ:sla:] ‘nest’, which appears only in this form on the surface, has an underlying form with an ‘abstract’ , i.e. ghõ:sla:. The results show that some speakers have the abstract underlying form, others do not.

Data are presented to show that a grammar of English should generate NP's with internal structures like [NP [Det the][N planets][NP Mars and Venus]], as well as NP's with structures like [NP [NP the two moons of Mars][NP Phobos and Deimos]]. The transformation required by the analysis works in conjunction with Chomsky's Subjacency Condition to account for the semantic and syntactic properties of surface-structure forms.

Historical linguistics has primarily used phonological evidence for genealogical classification. The evidence consists of ‘characteristic facts and details’, in Meillet's terms. Syntactic evidence can now be used. Subjectivity, i.e. the subjective dominance of the principal verb in a sentence, is cited as an illustration of a ‘characteristic fact’ in Proto-Indo-European and its dialects.

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’.

This paper offers a functional explanation for the existence and for the special properties of movement rules in natural languages. A survey of reasonably well-established movement transformations in English reveals certain asymmetries in the classes of such rules encountered empirically and in their formal properties. It is argued that the special formal properties of backing rules correlate with their function, which is different from the function of raising, lowering, and fronting rules. The latter three types can be given a uniform functional characterization in terms of the notion ‘objective content’, a notion that has figured prominently, though implicitly, in the history of generative grammar. This notion is discussed in preliminary terms, as is that of relative ‘prominence’, whose syntactic relevance has been clearly established. The hypothesis is advanced that raising, lowering, and fronting rules all serve the function of increasing the prominence of objective content in surface structure. This hypothesis accounts for the asymmetries noted in regard to the movement rules of English, and it also provides a motivated explanation for the fact that backing rules are upward bounded. Movement rules are shown to be only one facet of a broad conspiracy to ensure the surface prominence of objective content. This conspiracy is further explored in the context of information theory and generative semantics.

Double nominative and discontinuous possessive structures in Chinese are investigated. It is demonstrated that base rules as postulated in current syntactic theories cannot adequately handle these cases. It is argued that predicates can consist of an entire sentence; i.e., VP can be written as S. With minor modifications, this proposed rule will also accommodate problematic ‘topic-comment’ sentences.

Many ‘physical phonetic ‘ variables are continuous, so physically there is an infinitely large range of possible sounds. The generative phonologists, unlike Bloomfield, believe in a universal level of ’systematic phonetics’, where only a fixed finite number of distinct values are possible for each phonetic variable, and hence there is only a finite set of possible segments. To justify this hypothesis it would be necessary (a) to show that facts follow from it which are arbitrary on Bloomfield's theory, and (b) to explain away any counter-evidence. It is argued here that hardly any true predictions have been shown to follow from the hypothesis, and evidence is offered against it.

'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.

An Arabic theme–predicate structure may incorporate, as its predicate element, an embedded structure which itself consists of theme plus predicate. Authorities have not stated how far this process can be continued by further embeddings. Usage indicates, however, that a second embedding occurs only under certain limiting restrictions, and a third embedding not at all.

Intelligibility tests between dialects give results with three characteristics that need to be taken into account in their interpretation: asymmetry, sensitivity to changes in threshold, and incompleteness. If we interpret dialects as networks of communication arrived at by optimization, we make adequate use of these properties.

Grammarians working within the framework of semantically-based grammar have argued that syntactic representations are not arbitrary, but rather have their basis in semantics. In this paper it is further proposed that syntactic processes (rules and constraints) may have a semantic basis. Evidence is presented that the constraint against backward pronominalization with indefinite antecedents, which is found in a wide variety of languages, derives from the semantic properties of various classes of definite noun phrases.

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.

The basic formulation of a Telugu vowel syncope rule is discussed here, and certain peculiarities of its application to /CVC/ sequences containing a retroflex obstruent are pointed out. These peculiarities can be explained within the framework provided by the theory of natural assimilation in Schachter 1969. Situations arise where the syncope rule must choose the vowel to be deleted among several potentially deletable vowels. In certain cases of this sort, it is the output of a rule applying later than the syncope rule which provides a means of predicting which vowel is deleted. If we allow the syncope rule to look ahead in derivations at possible outputs, and impose on it a constraint derived partly from Schachter's assimilation theory, the phenomena of syncope can be explained.

Two kinds of variables have been used in generative phonology: Abbreviatory variables, which have only one expansion in the application of a given rule to a string; and Essential variables, which have many values simultaneously, e.g. the ‘star notation’ of Spe. Essential variables can be dispensed with entirely if rules are allowed to apply iteratively rather than simultaneously. Abbreviatory variables can be severely constrained by the Relevancy Condition proposed here, which allows only Irrelevant material to intervene between the Focus and Determinant of a rule. The result is a more highly constrained and more explanatory theory of phonology.

A generative treatment of the early stages of development of two children suggests that the main learning process in pronunciation consists of discovering and gaining control over the articulatory features required to make sounds heard. The lexical representation differs from the phonetic output at all stages because actual pronunciations are constrained by primitive or acquired articulatory processes that have the effect of imposing a phonotactic filter on the speech output. Rules and representations depict ‘competence’ and have ‘psychological reality’ by describing the articulatory analysis made by the learner, together with the constraints imposed by the filter. The phoneme concept required is similar to Sapir's: no invariance or bi-uniqueness conditions are imposed, but phonemes are much more concrete than those of current generative phonology. The abstractness of the latter results from a methodology that enthrones economy criteria requiring assumptions about human memory and learning which are almost certainly wrong. Alternatives are discussed.