The Old Persian construction type tya manā krtam ‘what I have done’, in which a genitive form of a nominal (manā), used to denote an agent, occurs construed with a participial form (krtam), has exact counterparts in Avestan and in Indic. The Old Persian construction is generally considered a possessive construction. This view is shown to be unacceptable; the Old Persian data require that manā krtam be classed as a passive construction, which accords with the facts of Avestan and Indic. It is also shown that the construction Ir. mana krtam, Ind. mama krtam arose through contact between nominal and verbal syntactic patterns as a variant of an older Indo-Iranian construction in which agent was denoted by an instrumental nominal form. The developments discussed have close parallels in Latin.

There are currently two prominent competing approaches to the study of the social significance of variation in language. One approach is the observation of the frequency of variants in actual speech and their correlation with social factors. The other approach, scalogram analysis, instead of recording the frequency of variants, systematically shows what the use of one variant implies for the use of another. It is argued here that the observation of frequency leads to generalizations not revealed by the use of scalogram analysis alone. Finally, some suggestions are made concerning the application of both models to problems of the teaching of reading and of socially standard English.

Sentence (as opposed to clause) structure involves a combination of predications into larger units much as in the statement calculus of formal logic, but requires a richer apparatus. An over-all taxonomic scheme—relevant to sentence structures in many parts of the world—is presented. After showing how structurally distinct parts of the sentence are relevant to the structure of paragraphs and discourses, a detailed classification of English sentence types is presented. The parameters of the resulting over-all system are suggestive of universal parameters of relation and contrast among sentences.

The American Heritage dictionary, in calling on a ‘Usage panel’ to give opinions on matters of disputed usage, has developed a technique close to that of jury testing, with juries made up of naïve expert speakers, as used by linguists. The Heritage points the way to satisfying public demand for adherence to standards of correctness, so-called. The technique, however, needs to be more rigorously and consistently applied. Jury testing can also be used on a same-different basis to determine whether sequences of sounds which are identical are instances of the same or different words. The sequence suit, in ‘He brought a suit’, is one form in a context of legal action, but another in a context of clothing.

The semantic structure of paired English adjectives is investigated, and the results are extended—or an extension of the result is suggested—to the vast lexicon of derived English adjectives. It is found that English adjectives are not semantic primitives, but rather are semantically based upon or derived from nouns or verbs. Categories of noun-based vs. verb-based adjectives are set up, and then contrasted with the categories of stative vs. active adjectives. It is claimed that the facts of semantic structure of English adjectives support the presence of lexical or pre-lexical transformations which map the deep semantic structure onto the surface lexical items called adjectives.

The declension of vocalic stems in Sanskrit as taught in Pāṇini’s grammar is presented here in such a way as to highlight those sections of the grammar that function in the formation of the declensional paradigm. Conclusions are drawn about the operation of a hypothetical grammar in which rules are ordered at random and in which the only significant principle of ordering is the context of the constituents in the base form.

A method is proposed for quantitatively comparing word-meaning lists in related languages, taking into account synonymous lexical representations. This is applied to the Indo-European languages to estimate rates of change of lexical representation for more than a thousand meanings documented by Buck. Comparisons between the distribution of rates for the lexicostatistic test list and other meanings, and between the distributions calculated from Indo-European data and those from Malayo-Polynesian, suggest generalizations of lexicostatistic theory.

The concept head (of a phrase) is frequently invoked in transformational theory, but it has not been formalized and is perhaps not formalizable within the framework of a phrase-structure categorial component. However, if dependency rules generate the structured strings of categories, a head or governor is formally specified for each phrase, thus supplying additional information needed to determine the applicability of some transformations. Moreover, the problem of pruning inappropriately labeled or vacuous non-terminal nodes, which arises when transformations are applied to structures generated by phrase-structure rules, does not arise when they are applied to dependency structures. Although dependency rules are not well-adapted to generating structure-sensitive sets of strings, there is evidence that the structure-sensitive properties of language are more appropriately accounted for by transformational rules, leaving the catégorial component both context-free and structure-free.

The use of the Russian indefinites in -to and -nibud’ is analysed from the viewpoint of generative semantics. It is argued that the choice between the two suffixes is determined by the underlying structure of the sentence, where the ordering of quantifiers is of primary importance. The relative positions of quantifiers is also shown to be relevant for the semantic difference between the English pronouns any and every.

This paper argues that Chomsky’s version of the history of linguistics is fundamentally false. With evidence drawn from recent reviews and some additional information, it argues that the term ‘Cartesian’ has no historical justification in regard to the linguistic theory embodied in the 1660 Port-Royal Grammar. The paper then examines the history of linguistics from 1660 to the Romantics, a period in which Chomsky finds much scattered evidence for the persistent dominance of the Cartesian tradition. Contrary to Chomsky’s opinion, however, Locke and not Descartes was the dominating force. Du Marsais was strongly anti-Cartesian and outspokenly pro-Lockean. The basic and most influential work in 18th century linguistic theory was Condillac’s Essai (1746), which is entirely Lockean; this work revived linguistic theory at the mid-century, including universal grammar. There is no conflict between universal grammar and the entirely Locke-derived theoretical work on the origin of language. Chomsky’s version of history is the product of serious deficiencies in knowledge and research, and is an obstacle to the creation of a true and significant history of linguistics.

Evidence is presented that the superficial Subject-Verb-Object word order of English arises by a transformation from an underlying constituent order in which clauses begin with verbs—or more correctly, predicates: the representation in question here is a semantic one in which noun, verb, adjective, conjunction, etc. are undifferentiated.

Numerous explanations have been put forth for the uses and derivations of tenses. This paper attempts to show that none of these explanations is satisfactory, since they cannot account for many ways in which tenses are used in English and other languages. While this paper does not try to present an adequate theory, it gives evidence that such a theory, at the very least, must be able to incorporate in its underlying structures concepts such as the interrelationship of the actual time of occurrence of events with the involvement of the speaker, hearer, or other participants in these events.

A small set of English verbs may take the inflections -ed or -t for preterit and past participle. The hypothesis that forms in divided usage are not randomly distributed was tested on groups of British and American subjects by presenting these forms in ‘forced-choice’ selection tests. Despite the well-known difference between American and British usage in preferring one or the other of the inflections, the -t forms were predominantly selected in participial use and for preterits in contexts suggestive of non-durative (‘effective’) aspect.

An explicit set of phrase structure and transformational rules is presented to account for the correct distribution of grammatical sentences which incorporate the coördinate conjunctions and, or, and nor, the quantifiers each, all, both, either, and neither, and the adverbs simultaneously, independently, en masse, mutually etc. Evidence is given to show that a Coordinate Conjunction Transformation is required in a grammar, and that it should be a substitution transformation, not a reduction transformation. A set of selection features based on properties of the quantifiers each, all, and both are discussed and shown to play a significant role in the grammar of coordinate conjoined structures. Of two hypotheses for describing such structures, one, the Phrase Structure Rule Hypothesis, is shown to extend naturally to the grammar of respectively sentences and reciprocal constructions, including constructions which incorporate the element each other. An alternative hypothesis, the Conjunction Reduction Hypothesis, is shown to have significant failings.

Almost all known varieties of vernacular Arabic are characterized by a functional and formal split of the old dual morpheme -ayn. It serves as a dual marker when appended to some nouns, as a plural marker with others; the two functions entail morphophonemic, sometimes also phonetic and syntactic, differences. Some of the features studied can be traced back to the ninth century; others are later developments.

Pre-Chomskyan linguistics recognized that the phonological systems of natural languages tend to form symmetrical patterns; this insight could be incorporated into generative phonology via a phonological base component, but is not adequately expressed by MS rules. A phonological base would also solve some difficulties in the theory of systematic phonetics. Current lack of interest in phonological patterning has resulted from the increased attention paid to morphophonemics; but the deep vs. surface distinction in phonology can be demonstrated on the basis of patterning even in languages lacking morpho-phonemic alternation. That distinction, it is argued, may be of only historical validity, whether or not alternations occur in the language.

This article deals with birhemic (two-term) sentences that contain no verb in their deep structure. If the predicate is an adjective or an indefinite noun, a copula is generated only if the tense is non-present and the mood non-indicative. If the predicate is a definite noun, a copula is generated in all cases, but is deleted later for the present indicative. Advancing the hypothesis that in birhemic predications the Subject-Predicate relation is never freely reversible, a hierarchy of definiteness is established; if two noun phrases are coupled, whichever occupies a higher position on this scale will be the subject.

This paper presents evidence for abstract phonology—that is, a phonology that recognizes underlying segments which are not realized phonetically. Two underlying vowel segments /ᴐ/ and /ε/ are posited for Nupe (a Kwa language of Central Nigeria), both of which are realized as [a], therefore merging (‘absolutely neutralizing’) with /a/. Arguments based on pattern congruity, naturalness, and a ‘weak’ form of alternation are brought forth in support of this solution. Central, however, is the demonstration of the psychological reality of these two segments. In addition, various alternate solutions are shown to be inadequate. Finally, implications of the abstract Nupe solution are discussed.

A theory has been proposed (Lieberman 1967) that accounts for some aspects of intonation in terms of two phonologic features, the breath-group and prominence. Acoustic and physiologic correlates of these features were derived by experimental procedures that made use of subglottal air pressure and flow measurements as well as acoustic analysis. Perceptual data indicated that listeners ‘decoded’ certain intonational signals by means of ‘motor theory perception’ structured in terms of the ‘archetypal’, i.e., primary, articulatory correlates of these features. In the present study this theory was tested by recording the electrical activity of the crico-thyroid muscle of the larynx for a set of 480 short statements and yes-no questions that sometimes had non-terminal [+prominent] syllables. Independently derived data of Fromkin & Ohala 1968 also were examined, and were found to be consistent with the theory proposed by Lieberman except that [+prominent] syllables in unmarked breath-groups had cricothyroid activity. In yes-no questions where the crico-thyroid was active at the end of the marked breath-group, non-terminal [+prominent] syllables had no crico-thyroid activity. The archetypal articulatory correlate of the marked breath-group is an increase in laryngeal tension; [+prominence] involves an increase in subglottal air pressure as well as increases in vowel duration and generally heightened muscular activity. Implementation rules relate the phonologic features to their archetypal and secondary articulatory correlates.

During the past five years it has become clear that phrase structure rules and transformations provide a grossly inadequate characterization of the notion ‘rule of grammar.’ The problem is this: phrase structure rules and transformations are local; they define well-formedness conditions on individual phrase-markers and on pairs of successive phrase-markers. However, certain rules of grammar are global in nature; they extend over entire derivations, or parts of derivations, and cannot be stated in full generality (if at all) by local operations. I have proposed that rules of grammar be considered as well-formedness conditions on derivations (or ‘derivational constraints’). In the most general case, rules of grammar will be global in nature. Phrase structure rules and transformations turn out to be special cases of derivational constraints. From the point of view of linguistic description, the theory of derivational constraints is as much an innovation over transformational grammar as transformational grammar was over phrase structure grammar. In this paper a few of the phenomena that require the postulation of global derivational constraints will be considered. Some of these are purely syntactic in nature; others involve the interaction of syntax and phonology. The cases that involve the interaction of syntax and semantic representations are not considered here, nor is the formal characterization of such constraints included.