1. 1 (1) a. b. c. d. 1a b c d cf. Langacker 1987, 1990, 2000, Goldberg 1995 5 3 1) 2) 3) 2. 1997 3 i A. B. C. A (cf. Goldberg 1995, 2003 B (cf. Langacker 1991) A)B C 1
19942 (2) a. b. c. 1994:150-165 1994 2.a 2.b 2.c ii AC B 1 : 2 : 12 17 1/0 3. 8 iii 8 1000 8 8000 203 2
.1 FS16 iv FS211 { / / / / / } { / / / / / } FS 1 { / / / / } { / / / / / } { / // / / } { / / / / / / } X. _ X.. _ X.. _ X.. _ X.. FS 2 _ X.. X.. _ X.. _ X.. _ X.. _ X...1 1 FS1 FS2 FS2 1 11 FS2 10cf. 3
Romesburg 1989, Kohonen 19953 (3) a. b. 3 3a 0 1 (3) a 0 1 1 0 0 0 0 0 1 0 0 1 0 0 0 0 0 b 1 0 1 0 0 0 1 0 0 0 0 0 0 1 0 0 0.2 2 2233 1/0 1 SPSS Ver10.1 Win K-means 4. 4.1. 5 203 17 2 2 5 K=5 3 K=5 K=5/ 12/ 2.236.2 1 2 3 4 5 0.021 0 0.131 0.118 0.172 0.936 0.020 0 0.176 0.724 0.596 0 0.016 0 0.414 0 1 0.574 0.647 0 0.021 0.122 0.443 0.353 0.069 0.277 0.429 0.443 0.471 0.276 0 0.492 0.148 0 0.483 _ 0.660 0.286 0.672 0 0.310 _ 0.106 0 0 1.000 0 _ 0.021 0.041 0.016 0 0 _ 0.021 0 0 0 0 4
_ 1.000 0.102 1 0.824 0 0.170 0 0.115 0.118 0 _ 0 0.306 0 0 0.414 _ 0.021 0.224 0 0 0.517 _ 0 0.143 0 0.176 0 _ 0 0.204 0 0 0.069 47 49 61 17 29.3 4.2. 4.1 5 1 2 3 4 5 4.2.1 1 (4) a. b. c. : XYV : X motion Y LOC 4.2.2 2 (5) a. b. c. : XYV 1 2 4.2.3 3 (6) a. b. c. : XYV : X change Y STATE 5
4.2.4 4 (7) a. b. c. d. : XZY V : 4 4.2.5 5 (8) a. b. c. d. : (X)YZV : 4.2.6 FS1 FS1 FS2 construction v 4.3. 6
1 2 3 4 5 1 1.940 1.640 1.676 1.386 2 1.940 1.294 1.440 1.572 3 1.640 1.294 1.258 1.697 4 1.676 1.440 1.258 1.873 5 1.386 1.572 1.697 1.873.4 4 1 2 1.940 3 1.640 1 5 1.386 2 1 1.940 4 4.1 3 3 5 1 2 3 4 3 1 5 2 3 4 5 4.5 5 3 1 5 2 3 4 5. 4 810 7
(9) a.... b. (10) a... b. (11) a. b. 911a b 9 9a 8b 10a 4 10b 5 1011 12 (12) a. 19 b. c. 12 12a 12b12c 912 3 1 0.936 0.59510.659 2 10.428 0.491 3 0.5730.672 1 4 5 1 2 3 4 5 FS1 FS2.5 5 1 5 8
2 3 4 3 1 5 6. Croft, William1991Syntactic Categories and Grammatical Relations. Chicago: The University of Chicago Press. Goldberg, Adele. E.1995Constructions: A Construction Grammar Approach to Argument Structure. Chicago: The University of Chicago Press. 1997,. Kohonen, T1995Self-Organizing Maps, Berlin: Springer Series in Information Sciences, Vol. 30. Langacker, Ronald W.1987Foundations of Cognitive Grammar, Vol. 1, Theoretical Prerequisites, Stan- ford, Calif.: Stanford University Press. Langacker, Ronald W.1990Concept, Image, and Symbol. The Cognitive Basis of Grammar, Cognitive Linguistics Research, 1, Berlin-New York: Mouton de Gruyter. Langacker, Ronald W.2000 A Dynamic Usage-Based Model. Michael Barlow and Suzanne Kemmer (eds.) Usage Based Models of Language. Stanford, Calif. : CSLI Publications, pp.1-63. Michaelis, Laura A.2003 Word Meaning, Sentence Meaning and Constructional Meaning. In H. Cuyckens, R. Dirven and J. Taylor, (eds.), Cognitive Perspectives on Lexical Semantics. Amsterdam: Mouton de Gruyter. pp.163-210. 1991. 1993. PP.1-37 Romesburg, C.1989Cluster Analysis for Researchers. Florida: RobertE. Krieger Publishing Company. 9
Taylor, R. John.2003Cognitive Grammar. Oxford: Oxford University Press. 1982 1,. 1983 PP.467-547. 2000,. 2001 No.8 pp. 1-20. 2003.: No.3. CD-ROM 100 NTT CD-ROM. i 2004 ii LSA; Latent Semantic Analysis)LSALandauer & Dumais1997 http://lsa.colorado.edu/ iii CD-ROM 100 / /// /// iv FS1 CD-ROM v Langacker 1987, 1990 2000Michaelis 20032001, 2004 10
abstract Quantitive analysis of Japanese Particle NI A New Approach to Descriptive Grammar LEE, Che-ho (Graduate School of Kyoto University) The purpose of this paper is to examine the usage of Japanese particle NI by restrictions of the context. In previous studies, Ni is generally analyzed as a postpositional particle and acquires various meanings such as the goal of a motion, the result of the change, and the marker of an adverbial phrase. However, the relation between them is not so clear. Two questions I try to answer in the paper are as the following: (1) how many usage of NI we find based on form-meaning pairs, (2) what kind of relation exists between each example of NI usage. In order to provide a solution, I would like to introduce a new research methodology. Firstly, in order to observe the usage of NI I collect NI sentences as samples and create a corpus. This is the very first level to put this methodology into practice. Secondly, I analyze my data by conducting a quantification of the usage, based on a multi-variable analysis and also based on a theoretical/descriptive position of Construction Grammar. To give a more specific account, the following procedure is performed, a) all of the samples are collected from a corpus, b) all samples are identified as 1 and 0 by Distributed Representations, c) the samples are statistically analyzed using cluster analysis. As a concluding remark, the analysis result suggests the following facts. 1) Three kinds of asymmetric restrictions are observed while clustering. They are the meaning of the coinciding noun, the syntactic relation and both coinciding noun and syntactic relation. All these three restrictions affect the clustering. 2) The usage of the postpositional particle NI is understood by restrictions coming from the context.