In this article, I first point out that Generative Grammar addresses the problem of “explaining” the diversity of human languages, by, somewhat ironically, proposing the concept of UG (Universal Grammar), a common cognitive capacity of homo sapiens. Theoretical explanations require a minimum set of theoretical postulates, and from them, try to derive (in the best cases, deductively) as many empirical phenomena as possible. I argue that the two earliest attempts at the comparative syntax of English and Japanese – Fukui (1986, et seqq.) and Kuroda (1988) – actually tried to propose theoretical explanations in this sense, rather than simply claiming, as widely – and mistakenly – assumed in the literature, that certain features (φ-features) are absent in Japanese. These authors tried to show that there are certain important “clustering effects” that ought to be derived from a single parameter, if the theory of UG is properly modified. I argue that the proper status of these two approaches, vis- à-vis a huge amount of comparative work that followed them in the generative framework, can be appropriately established under this interpretation. I also suggest that recent attempts at deriving clustering effects in various ways can in fact be shown to be rooted in earlier approaches in the 1980s, including the two proposals just mentioned. Exploration into the nature of “clustering effects/patterns” observed in the work of Fukui/Kuroda are likely to unveil the very nature of the “parameter” concept in general linguistic theory.

This paper draws on the issue of wa and ga, which I assert to be an appropriate topic to review the development of the study of Japanese within the framework of Generative Grammar (more specifically, the Standard Theory). After reviewing two notable works on case attachment in the grammar of Japanese, namely Kuroda (1965a) and Kuno (1973a), I propose a set of noncyclic case assignment rules that crucially employ the notion of [+/– transitive] in sentence grammar. The proposal precludes the surface manifestation of sequences of redundant case particles, which were inevitable in the earlier works. I then proceed to observe the interpretative characteristics of wa and ga, for which I provide an account in terms of notions such as new/old information and a “spotlighting” function in discourse. Through the course of discussion, the interactions between sentence grammar and discourse grammar are clarified, signaling the importance of both in accounting for characteristic properties of Japanese.

Drawing from notions starting from Locke’s semiotics and extending through Peirce’s theory of signs, the author finds applications in the interpretation of various artistic works (with particular focus on Japanese works, including haiku) to show how an understanding of aspects of the nature of signs makes possible insights not available through linguistic analysis alone. The vagueness in one poet’s use of the abstract noun mono ‘thing’ is shown to force the reader to contemplate properties ascribed thereto at the phenomenological level Peirce calls “firstness”. Sound textures of words, visual components of Chinese characters, etc. are shown to produce powerful meaning effects beyond direct linguistic representation. Examples of one-to-many mappings from written characters to words, intertextuality, violations of selectional restrictions, and plays on words, etc. are examined to show how aspects of signs function to evoke layers of meaning easily ignored in linguistic analyses that concentrate only on the “thirdness” of everyday language.

This paper is an enlarged version of the author’s lecture delivered at the 78th general meeting of the Linguistic Society of Japan. Two subjects are discussed: tonogenesis and ergative constructions in the TB languages. First, the sub-grouping of TB languages is mentioned briefly. With a number of examples the correspondence of cognate words in toneme dialects and tonemeless dialects within both the Chin and Tibetan groups is shown. The author expresses his opinion concerning tonal development in the Choni dialect of Tibetan.

After considering the formation of toneme patterns in the disyllabic words of Lhasa Tibetan, the author mentions the interesting phenomena of disyllabic words reduced to monosyllables evident in the Choni dialect in China, Dzongkha in Bhutan and Lhomi in Nepal.

In the second part he explains the ergative construction of Tibetan and discusses the opposition of ergative and non-ergative constructions. It is known that Tibetan originally had no passive construction as such, but in fact, when a passive meaning is called for, it is formed by the topicalization or focusization of the object of the sentence, as is clear in examples given which contrast with modern Chinese.

Finally, the author points out what were probably ergative forms of Rawang of Northern Burma, Chiang in Sichuan and Moso in Yunnan; and he assumes the development of the ergative construction in Moso.

In the history of linguistics, several revolutionary developments in methodology have occurred, including comparative grammar versus philology, synchronic linguistics versus historical linguistics, phonology versus phonetics, generative transformational grammar versus structural linguistics, etc. It is a peculiar characteristic of linguistics, however, that the new does not completely invalidate the old, and in the course of time the old and the new turn out to complement each other.