This paper starts by defining a typical case of grammaticalization (or grammaticization) as a process whereby independent lexical items become dependent forms and acquire grammatical functions, and provides a set of criteria for characterizing this process. They are: abstractness of meaning, paradigmaticity, obligatoriness of marking, morphological boundedness, and interaction with the rest of grammar. Further, extended cases of grammaticalization are discussed, namely the rise of polyfunctionality and the emergence of grammatical constructions. It is shown that, especially in the latter, discourse function plays a crucial role in the conventionalization of a syntagm into a grammatical construction. In the last section, some considerations are given to the motivation behind grammaticalization. Two major forces are metaphor (i.e. cross-domain mapping) and metonymy (i.e. domaininternal profile shift). Finally, general constraints on semantic change are discussed.

View full abstract

This study presents a derivational account of the semantics of gozaru, arrived at by classifying tokens of gozaru found in Feiqe Monogatari (the romanized Amakusa version of the Tale of the Heike) based on their grammatical functions, and comparing them against those occurring in the original version of the Feiqe Monogatari. In spatial existence sentences, where the meaning of gozaru remains closest to its origin, nearly all tokens of gozaru occur in honorific expressions, whereas its use in polite expressions predominates in other cases. The fact that polite expressions, unlike honorifics, lack selectional restrictions as to their subjects, thus allowing for a wider range of use, is the reason suggested for this predominance. Moreover, the process of grammaticalization, and the shifts undergone by existence verbs in their lexical categorization, served among other factors to influence the development of gozaru's meaning towards a polite one.

View full abstract

This paper aims to describe the process of grammaticalization in Japanese potential expressions with reference to the situation in other languages. The following points will be illustrated: (a) As with other languages in the world, completives and spontaneous expressions are a source of potential expressions in Japanese. The multiplicity of forms of the latter origin is a characteristic of Japanese (§3). (b) Within potential expressions, the development of ABILITY>POSSIBILITY is observed cross-linguistically. In Japanese, however, the direction of change is the reverse: POSSIBILITY>ABILITY>MODALIZED ABILITY (§4). (c) Japanese potential expressions further come to express directives and epistemic possibility in some dialects, but the former is normally limited to prohibitive expressions and the latter type of change constitutes only marginal cases (§5). In sum, forms which originally denoted meanings other than ability and possibility have repeatedly come into use as potential expressions in Japanese, and have further changed their meanings within the realm of potential expressions. However they seldom change into other expressions with the exception of prohibitive. Japanese potential expressions normally give way to other forms which are grammaticalized in a similar fashion, and cease to be used as potential expressions.

View full abstract

In Old Japanese, there are two types of noun clauses for the complement sentences: juntai-type noun clause, in which the complement sentences are formed with rentai (attributive) verbal forms as in [Tomo no enpou yori otozuretaru] wo yorokobu 'someone rejoices at the visiting of a friend from a distant place', and koto-type as in [kaze no fuku koto] yamaneba 'since wind blowing does not end'. When a juntai-type noun clause is used as a subject, its predicate is restricted to a stative one. Examining cases of juntai-type clauses in Old Japanese, the paper clarifies that a juntai-type noun clause is used only for the object of one's feeling, emotion, or judgement. On the other hand, there is no such restriction in the use of koto-type noun clauses. The paper demonstrates that although the function of koto-type noun clause has hardly changed from Old Japanese to Modern Japanese, the no-type noun clause in Modern Japanese has inherited the function of the old juntai-type noun clause. The form no originally was a pronoun indicating 'substance', but it has gone through the process of grammaticalization and extended to a form for expressing a 'matter.' The paper argues that it is this development of no that consequently led it to displace juntai-type noun clauses, resulting in their disappearance.

View full abstract

The aim of this paper is to discuss some aspects of grammaticalization in Modern Japanese, especially focusing on the continuum between content words and function words. After reviewing and examining the definition of grammaticalization, the paper points out the significance of synchronic study on grammaticalization. The paper claims that the significance of synchronic study on grammaticalization lies mainly in the following points: (1) it will make possible the demonstration of the continuum between content words and function words, and (2) the semantic analysis of a grammaticalized function word may be facilitated by the comparison of its corresponding content word. Based on this point, the paper gives as exhaustive a review as possible on grammaticalization phenomena in Modern Japanese, suggesting the direction of future research.

View full abstract

This paper examines the process of grammaticalization of the directional case particle -sa, the objective particle -koto/-toko and the past tense marker -teatta used in the Tohoku dialects. The original meaning of the particle -sa is directional, that is, it indicates the direction toward which some directional movement or action proceeds (ex. Tokyo sa iku 'I will go to Tokyo'). However, on the Japan Sea of the Tohoku district, this form has the added meaning of locative (ex. Tsukue no ue sa hon aru 'There is a book on the desk'). In the past, the objective particle -koto/-toko was used only with animate nouns (ex. Taro Jiro toko nagutta 'Taro hit Jiro'), but now, it has extended its grammatical function with inanimate nouns as well (ex. Taro boru toko ketobashita 'Taro kicked a ball') in this same region. The past tense marker -teatta has the continuative aspect in the dialects of the western part of Aomori and the northern part of Akita (ex. nondeatta 'was drinking'), but this aspectual meaning is new. Originally, it signified the completive aspect (ex. iteatta 'someone was there', nondeatta 'drank'). Thus, it was found that on the Japan Sea side of the Tohoku district, these grammatical forms have grammaticalized their functions more than the other regions.

View full abstract

This study presents a contrastive study of grammaticalization processes in Japanese and Korean from a typological perspective. Grammaticalization processes involving (i) nouns, (ii) aspectual constructions, and (iii) verbal inflections are presented to aid understanding of the similarities and differences between the two languages. Though superficially similar in many respects, intriguing differences are found between the two languages regarding the productivity and pathways involved in the specific grammaticalization processes (i)-(iii). There is ample scope for future contrastive studies that pay particular attention to dynamic/synchronic dimensions of grammaticalization-conventionalization of linguistic forms/constructions repeatedly occurring in discourse and the concomitant emergence of specific grammatical meanings-to explore the grammaticalization processes between the two languages.

View full abstract

Suffixation and auxiliation of morphologically independent words and syntactic structures has been recognized as one of the most salient features of grammaticalization from early on. However, mainly for theory-internal reasons, most contemporary studies of grammaticalization rather focus on semantic and pragmatic features of the process. With respect to Japanese this is rather unfortunate, because the language offers abundant material for the study of the morphological aspect of grammaticalization. In this paper, I will show that four general rules can be observed in the grammaticalization of morphemes in verb morphology: (1) suffixation of originally morphologically independent material, (2) fusion of morphemes, (3) phonological reduction, and (4) conversion of derivational suffixes into inflectional suffixes. Although these processes are very well known in grammaticalization theory, it is crucial to remember that they can only be pinned down concretely with the help of a rigorous methodology. I further point out that two processes, namely analogy and exaptation, are at work in exceptions to the above-mentioned rules.

View full abstract

Some postpositions in Japanese include in their composition non-finite ren'yo forms or -te forms of verbs. We can assume that these postpositions are derived from verbs. In the field of grammaticalization, omotte and niyotte, which express cause, have been viewed as examples of grammaticalization from the verbs motsu 'to hold' and yoru 'to approach', respectively. However, based on studies of kanbun kundoku, the traditional reading of Chinese text using Japanese, it can be seen that these postpositions have relations with forms found in kanbun text (text heavily influenced by kanbun kundoku). If they are loan translations from Chinese, a foreign language, borrowed into Japanese through kanbun kundoku, then it is inappropriate to view these postpositions as examples of grammaticalization. Through historical research and analysis distinguishing between kanbun text and wabun, i.e. strictly Japanese, text, we can prove that 1. the several meanings of omotte found in Middle and later Japanese can be seen as borrowings, or extensions from borrowings. On the other hand, mochi(te) marking 'instrument' that we see from the ancient times but was subsequently lost is a true example of grammaticalization from the verb motsu; 2. One of the meanings of the verb yoru, 'be attributed to', which has been argued to be the basis of grammaticalization of niyotte, can not be found in wabun text in Old and Middle Japanese, which invalidates the proposed account. We conclude that when taking Japanese as the target of grammaticalization research, we should strictly distinguish the borrowing situation, in consideration of the complex and important influence Chinese has had on Japanese.

View full abstract

This paper makes the following observations which are typologically meaningful. Although it has rarely been mentioned, the tense-aspect systems of Late Medieval Japanese and Modern Korean have some common features: In both languages, the standard past forms -ta/-ess-ta can also express the present situation, like the resultative, and the basic verb forms of both languages can also express the present situation, like the progressive. The reason that the two languages have these common features can be explained from the viewpoint of grammaticalization. The paper presents some evidence to show that the lexical meaning of existential verbs iru, aru/iss-ta in both languages has a great influence on the usage of the aspect forms -teiru, -tearu/-ko iss-ta, -e(a)iss-ta. That is, the degree of grammaticalization of -teiru, -tearu/-ko iss-ta, -e(a)iss-ta is relatively low. The paper also argues that in both languages, the standard past forms-ta/-ess-ta and the basic verb forms express the present situation in order to compensate for the insufficiency of the aspect forms in these languages. The paper proposes that both languages are at similar stages in the progress of the change from the aspect forms to the tense forms. In addition to this, the tense-aspect systems of both languages show a pattern of the grammaticalization in which the usage of aspect forms extends under the influence of the meaning of "existence."

View full abstract

An example of grammaticalization is found in the Yaeyama group of the Luchuan (Ryukyuan) language, comparing morphemes of verbal conjugation. Miyara verbal conjugations are divided into two major groups. Group I corresponds with the Japanese 5-dan type, and Group II corresponds with the 1-dan type. However, Group I has one additional conjugational type, the Active, which Group II lacks.[table]Yonaguni dialect has the verb i-(r-) which is equivalent to suru ('to do', intransitive) in Common Japanese.[table]It is concluded that the Miyara morpheme i-(r-) originated in the verb ^*i-, which means that acts or things have come into existence, uttered when the speaker has a look at these acts or things. It seems that in Miyara dialect, the grammaticalization progressed from Group II to Group I.

View full abstract

In this paper, I aim to assemble the various current results of research concerning the history of conditional expressions and reconsider them from the viewpoint of grammaticalization. With junsetsu (expected and reasonable) suppositional conditions, the development of the forms nara(ba) and tara(ba) are related to the shift from mizenkei+ba to izenkei (kateikei)+ba. With junsetsu kakutei (established) conditions, I confirmed the situation of the development of aida and hodoni. On gyakusetsu (concessive) kakutei conditions, keredomo (one of the modern conjunctive forms) is related to the process of transition from the auxiliary verb maji to majii or mai. Moreover, the development of the conjunctive form tokorode shows aspects of the transition from formal noun to conjunction.

View full abstract