This paper critically examines Chomsky's position on meaning, considering its development over the years, its significance for other areas of grammatical theory and for related areas outside linguistics, and its methodological, empirical, and philosophical bases. Focus is on the points of disagreement between Chomsky's position and the author's, and it is argued that a close examination of the bases for Chomsky's position shows that it should be rejected.

Passive sentences often have different meanings and communicative intent from the corresponding active sentences. It is proposed that the ‘extra’ attitudinal and modal meanings of passive sentences are derived by Gricean conversational implicature. It is therefore possible to state the relation between active and passive sentences in syntactic terms without reference to meaning. Implicatures are more likely to be associated with passive sentences if the construction is ‘marked’, as measured by the restrictions placed on which NP's may be promoted to subject position, and what other roles the derived subject may have. Conversationally conveyed meanings are based on what appears as topic in subject position.

The present study re-examines the data of contemporary Russian conjugation in the framework of a semiotic concept of linguistic structure based on markedness as the informing principle. Emphasis is placed on demonstrating the coherence of expression and content that allows grammatical facts to subsist as such. This is achieved by analysing in detail how the relational values of the pertinent grammatical categories are represented diagrammatically in their phonological and morphophonemic expression. Setting explication of grammar as its goal, rather than description or predictability, the analysis articulates a radically different and explicitly hermeneutic perspective for linguistic inquiry.

Automatic alternations, those traditionally understood to be required by phonetic structure, have not been a part of transformational generative phonologies, where alternations are described by feature-changing P-rules. ‘Conspiracies’, the justification of two or more alternation P-rules by some requirement of phonetic structure, have also gone unexpressed. Phonetic structure is properly expressed by non-transformational, non-directional neutralization rules, of which alternation P-rules are partial repetitions. Automatic alternations can be considered alternations of completely with incompletely specified phonemes, resulting from a neutralization of features. If the features subject to neutralization are so marked in lexical entries, this effectively represents the automatic allomorphs, permitting the non-directional phonetic-structure rules to accomplish automatic alternations by selecting the correct alternants in environments of neutralization. (An illustration of this is an automatic-alternation conspiracy in Kambata, a language of Ethiopia—resulting from historical assimilations, vowel deletion, and metathesis.) The result is a non-transformational phonology consistent with the general tendency of leveling—one in which automatic alternations, conspiracies, and all non-controversial generalizations are expressed in a grammar which prohibits ordered rules and requires that all rules and representations be surface-true.

Not only the evidence of Gothic, but also that of North and West Germanic, speaks against the postulation of Proto-Germanic ‘umlaut’ allophones (§1). Furthermore, since even the results of the ‘earlier umlaut’ phenomena affecting */i e u/ in the various dialects point to the rise of allophonic variations in partly different environments, all the pre-literary vowel modifications produced by both vocalic and consonantal influences should be ascribed to the prehistory of the individual Germanic languages (§§2–3). The scarcity of forms with /e/ by a-umlaut of */i/ is explained through the assumption that the assimilation exerted by *[-a] on */i/ was widely resisted in order to avoid a merger of */i/ with */e/ (§2).

Some conservative Italian dialects possess simplified color systems, sometimes having fewer color terms than Classical Latin. Thus a hypothesis formulated by Berlin & Kay 1969 is confirmed: evolution of color systems is not always linear; regressions and subsequent re-elaboration may be observed. Italian dialects show some of the modalities of diachronic change of color systems.