72 2008 V V V V V, 2001 Ohori, 2001

文 法 化 としてのトコロ 関 係 節 の 成 立 主 要 部 内 在 型 関 係 節 との 比 較 からみえるもの 1. はじめに 1999b Ohori (2001) Hopper & Traugott, 2003, Narrog, 2006; Traugott & Dasher, 2002

72 2008 V V V V V, 2001 Ohori, 2001

73 2. トコロ 関 係 節 と 関 連 構 文 (1) (1) a. b. c. (2) (2) a. b. c. (2) (1) (1a-c) (3) a.

74 b. c. (1) (1) (4) (4) a. b. c. (1) (4) state of affair mismatch 1

75 (4) (5) (5) a. b. c. (6) 2 (6) a. * b. * c. * Shimoyama, 1999 (7)

76 (7) a. b. (8) (8) a. b. (9) a. b. * (9b) (10) (10) a. b. *

77 3. 主 要 部 内 在 型 関 係 節 の 構 造 とその 成 立 3. 1. 主 要 部 内 在 型 関 係 節 の 構 造 Kuroda (1974-77) Cole (1987) Culy (1990) Navajo Lakhota Quechua 3 (11)

78 cf., 1999a (11) a. NP b. AdvP e saturate 1999a 1999b Ohara (1996) 2002 2001 Kikuta (2002)

79 3. 2. 主 要 部 内 在 型 関 係 節 の 成 立 1955 2000, 1955; Kuroda, 1974-77 4 John saw the girl swimming in the river Felser (1999) Wrona (2008)

80 -aku Wrona bridging context, 2000; Wrona, 2008 14 (12) a. 1.20 b. 5.843 c. 15.3667 e. 20.4468 f. 10.1949 (12b) (12f)

81 (13) a. b. (14) a. b. c.

82 (15) (15) a. b. 1955 1955 2000 1955

83, 2000 4. トコロ 関 係 節 4. 1. トコロ 関 係 節 の 成 立 : 先 行 研 究 から

84 (16) 19. 4288 (17) (17) a. b. (16) (18a) (18b) (18) a. b.

85 (17) Ohori (2001) (19) 5 (19) a. b. Ohori (3) Come to my place Come to me Ohori Ohori

86 12 13 (20) Ohori (2001) 6 4. 2. Ohori (2001) 説 の 問 題 Ohori (2001) (21) 1024-28 11C

87 (21) a. b. Ohori (2001) 5. トコロ 関 係 節 の 成 立 過 程 の 提 案 5. 1. 中 世 までのトコロ 節 Ohori (2001)

88 (18a) 5. 2. 中 世 前 期 のトコロ 節 とトコロ 関 係 節 Ohori (2001)

89 14 16C 12 13C Ohori 13C 14C (22) Ohori (2001) (22) a.... b. (23)

90 (23) a. b. c. (23) (24)

91 (25) (25) (25) (26)......

92 (27) (27) Ohori (2001) (27) (28) a. AdvP proi proi b. NP proi i (28a) (30b) (28a) (28b) Ohori (2001) (29) a. b.

93 (30) 7 Horie & Sassa, 2000 4.2 Ohori (2001) Come to my place Come to me

94 (31) (31) a. b. (32) 8 (32) (32)...

95 (33) (33)... Ohori (2001)

96 5. 3. トコロ 関 係 節 の 主 語 目 的 語 非 対 称 性 Ohori (2001) Ohori

97 15 6. 結 語

98 (Hopper & Traugott, 2003; Traugott & Dasher, 2002) Ohori (2001) Heine, Claudi, and Hünnemeyer (1991) person > object > space > time > quality

99 Hopper & Traugott, 2003) 5.3

100 1 Kuroda, 1974-77;, 1998, 2001 Ohara (1996) 2008 2 (5) (i) a. * b. * c. * 3, 1999a 4 Wrona 2008 (i) a. 15.3752 b. 20.4360 c. 20.4429

101 5 (19b) 6... 7 8 1955 3.2 2000 Cole, P. (1987). The structure of internally-headed relative clauses. Natural Language and Linguistic Theory, 5, 277-302. Culy, C. D. (1990). The syntax and semantics of internally-headed relative clauses. Ph.D. dissertation. Stanford University. Felser, C. (1999). Verbal complement clauses: A minimalist study of direct perception constructions. Amsterdam: John Benjamins. Heine, B., Claudi, U., & Hünnemeyer, F. (1991). Grammaticalization. Chicago: University of Chicago Press. Hopper, P., & Traugott, E. C. (2003). Grammaticalization, 2 nd ed. Cambridge: Cambridge

102 University Press. Narrog, H. (2006). A renaissance for historical semantics. English Linguistics, 23, 554-572. 1995 grammaticalization KLS 15: Proceedings of the Nineteenth Annual Meeting, 90-99. Horie, K. (1998). On the polyfunctionality of the Japanese particle NO: From the perspective of ontology and grammaticalization. Ohori, T. (Ed.). Studies in Japanese grammaticalization: Cognitive and discourse perspectives (pp. 169-192). Tokyo: Kuroshio. Horie, K. (2008). The grammaticalization of nominalizers in Japanese and Korean: A contrastive study. López-Couso, M.J., & Seoane, E. (Eds.), Rethinking Grammaticalization: New perspectives (pp. 169-187). Amsterdam: John Benjamins, Horie, K., & Sassa, Y. (2000). From place to space to discourse: A constrastive linguistic analysis of Japanese tokor[u] and Korean tey. M. Nakayama & C. J. Quinn, Jr. (Eds.), Japanese Korean linguistics, Vol. 9 (pp. 181-194). Stanford, CA: CSLI. 1955. Kikuta (2002). Integrating the complement and the adverbial analyses of Japanese internallyheaded relative clause. Proceedings of the eighth international conference of Head-Driven Phrase Structure Grammar. On-line. Available at: http://csli-publications.stanford.edu/ HPSG/2. 2008 V V 81-82: 115-165. 2000. Kuroda, S-Y (1974-77) Pivot independent relativization in Japanese. Reprinted in Kuroda, S-Y. (1992) Japanese syntax and semantics: Collected papers (pp.114-174). Dordrecht: Kluwer Academic. 1999a (pp. 27-103).. 1999b (pp. 105-162).. 2000 WEB http://nikkoku.jkn21.com/). 1998 47, 39-49,. 2001 No. 1 (pp. 229-255)..

103 Ohara, K.H. 1996. A constructional approach to Japanese internally-headed relativization. Ph.D. dissertation. U.C. Berkeley. 2002 II (pp. 11-35).. Ohori, T. (2001). Clause integration as grammaticalization: A case from Japanese tokorocomplements. Horie, K., & Sato, S. (Eds.). Cognitive-functional linguistics in an East Asian context (pp. 279-301). Tokyo: Kuroshio. Shimoyama, J. (1999). Internally-headed relative clauses in Japanese and E-Type anaphora, Journal of East Asian Linguistics, 8, 147-182. Traugott, E. C., & Dasher, R. B. Regularity in semantic change. Cambridge: Cambridge University Press. 1993 2: 76-87. 1995 3: 79-91. Wrona, J. (2008). The Old Japanese complement system. Kent: Global Oriental. 1986 1957-69 1989- (http://www.nijl.ac.jp/contents/d_library/index.html)

104 Synopsis The Development of Tokoro-Relative Clause: Analogy as an Indirect Factor Promoting Grammaticalization Chiharu Kikuta This paper argues that constructional analogy has played a crucial role in facilitating reanalysis leading to the development of tokoro-relative clauses. In modern Japanese, the tokoro-relative clause is often compared to the internally-headed relative clause (IHRC). This is firstly because they have basically the same surface structure, the only difference being the choice of complementizer, either no or tokoro, and secondly (and more importantly) because they both share an interesting syntax-semantics mismatch whereby an apparent clausal complement, which ought to denote an event or a state of affairs, in fact denotes an entity involved in the event. Nevertheless, the two constructions are not completely identical, and the tokoro-relative is known to show some peculiarities which are not likely to be attributed to the semantics of tokoro. Assuming that the differences result from the degree of grammaticalization to which the two constructions have achieved, this paper examines how IHRCs and tokoro-relative clauses have developed. Following the usage-based account of grammaticalization, this paper assumes that changes contributing to grammaticalization proceeds very gradually out of an ambiguous context where both the existing and the innovative interpretations are possible. It has been observed that IHRCs started to grow in the Heian Period (9C-12C), when the clausal complement of a predicate of direct perception, which denotes the perception of a state

105 of affairs can also be interpreted as that of a salient participant in the state. Tokoro-relatives, on the other hand, developed much more slowly. Even though ambiguous contexts were available in the Heian period, tokororelatives were fully established only in the Muromachi Period (14C-16C), according to Ohori (2001). This paper accepts Ohori s (2001) claim that the development of circumstantial clauses of tokoro-ni in the Kamakura Period (12C-14C) has been instrumental in giving rise to the tokoro-relative clause, but argues that the ambiguous interpretation based on the place-participant metonymy explains only half of the story. What is really crucial for the development of tokoro-relatives is that the circumstantial tokoro-ni clauses, which was on the increase in late MJ, are not exactly denoting the place of an event but are denoting events as background circumstances, sloppily juxtaposed. This kind of sloppy clause linkage is typical of Classical Japanese narratives, and the circumstantial adverbial clause which is often phonologically indistinguishable from (thus syntactically ambiguous with) IHRC is the paramount example. Therefore, when the circumstantial tokoro-clause started to have the sloppy clause linkage function in late MJ, the analogy with IHRC was prompted, inducing the metonymical reanalysis contingent on other syntactic conditions. Theoretically, this paper has argued for the possibility of constructional analogy as an indirect factor facilitating the metonymical reanalysis. Recent studies of grammaticalization in the usage-based approach, which attach prime importance to metonymy as the fundamental mechanism, focus on the speaker-hearer interaction and the context of use, and how a linguistic form can induce different meanings and implications. In other words, they emphasizes on how a single form can be associated with plural meanings

106 in actual speech contexts before the form undergoes reanalysis. However, they pay less attention to how other forms of a similar function influences on inducing the reanalysis. Admitting that analogy is a powerful mechanism unless it is duly constrained, this paper points out the importance of analogy based on constructional network for language change including grammaticalization. So long as language is a system, language users must always be aware of the paradigmatic relation of the linguistic form within the system, and this awareness ought to exert influence on the way language is changed.