Mathematical Linguistics Vol. 16

Vol. 16 No. l (June, 1987)	Vol. 16 No. 2 (Sep., 1987)	Vol. 16 No. 3 (Dec., 1987)	Vol. 16 No. 4 (Mar., 1988)
Vol. 16 No. 5 (June, 1988)	Vol. 16 No. 6 (Sep., 1988)	Vol. 16 No. 7 (Dec., 1988)	Vol. 16 No. 8 (Mar., 1989)

Vol. 16 No. 1 (Jun. 20, 1987)

Classification:
paper
Author:
MOMOUCHI Yosio
Title:
Constructing The Frame of Direct Speech in Japanese Narrative Text
Pages:
1-15
Descriptors:
direct speech; frame; procedure; coherence; narrative text; Japanese
Abstract:
Direct speech in Japanese narrative text is quoted by quatation marks and embedded in background sentences. The frame of direct speech, a version of sentence frame, consists of speaker, addressee, direct speech and predicate. Syntactic relations of direct speech to background sentences and some cues for integrating direct speech splitted into two parts are examined. A skelton procedure for constructing the frame of direct speech is shown. Context information for restoring ellipted elements of the frame is considered.
Classification:
paper
Authors:
FURUGORI Teiji and HIRANUMA Kazumi
Title:
Statistical Characteristics of English Sentences Written by The Japanese and Detecting and Correcting Spelling-Errors
Pages:
16-26
Descriptors:
text processing; English; spelling-errors; detection and correction; Japanese English-users; English teaching
Abstract:

This paper describes statistical characteristics of English sentences written by Japanese students. The emphasis is put mainly on the lexical features such as the vocabulary, word frequencies, and numbers of misspellings. The facts found here will be useful for devising a program that will detect and correct spelling errors made by Japanese English- users. The study would also be of value to discover an effective way of English teaching.

Vol. 16 No. 2 (Sep. 19, 1987)

Classification:
paper
Author:
TOKUDA Katsumi
Title:
Relation between Errors of Kanji Reading and Writing and Attribute Factors of Kanji in Low Vision Children
Pages:
51-69
Descriptors:
Kanji; attribute factors; low vision children; reading error; writing error
Abstract:

The purpose of this study was to clarify relations between errors of Kanji and attribute factors in low vision children The study samples include 60 students with visually impaired and 88 normal students in the sixth grade. The phonetical errors in Kanji writing were concerned with concreteness, numbers of the same reading and familiarity with the Kanji character Regularity, familiarity and concreteness of Kanji contributed to the errors of shape in low vision children.
Classification:
report
Authors:
TAKEUCHI Haruhiko and HIEDA Ichirou
Title:
A Research on Speech Sound Characteristics
Pages:
70-80
Descriptors:
speech sound characteristics; concept; synesthesia; metaphor
Classification:
note
Author :
OGINO Tsunao
Title:
An Evaluation Method of Ability of Kana-Kanji Conversion Systems --- Using VJE-beta as an Example ---
Pages:
81-87
Descriptors:
Kana-Kanji conversion; word segmentation; dictionary registration

Vol. 16 No. 3 (Dec. 19, 1987)

Classification:
paper
Author:
OGINO Takano
Title:
Classification System of The Meanings in Japanese
Pages:
95-112
Descriptors:
meaning classification system; selectional restriction; hyponymy; multiple inheritance; semantic component
Abstract:

In the field of natural language understanding by computers, it is often insisted that semantic restrictions are useful when syntactic structure cannot be decided correctly by morphological an syntactic relations of the sentence elements alone. From this point of view, the author, fucusing on sentence predicates, aims to construct a meaning classification system governed by the predicate. The main purpose of systematizing the classification is to describe the semantic restrictions of word groups effectively on the basis of semantic inheritance of concepts included in each semantic item.
Classification:
note
Author:
ITO Masamitsu
Title:
A Concordance Compiling System for Classical Japanese Using Personal Computer --- KOMOREBI---
Pages:
113-127
Descriptors:
text by Chinese characters and kana letters; KOOWAKA-BUKYOKU ``NIHONGI''; concordance (KWIC); BASIC; KOMREBI

Vol 16 No. 4 (Mar. 16, 1988)

Classification:
paper
Author:
MAEDA Takehiko
Title:
Conversation Analysis of Shopping Bahavior
Pages:
141-156
Descriptors:
conversation analysis; verbal behavior; shopping behavior; interaction; system of categories; Task area; Interpersonal relation area; Fixation area; Speaking-orientation area
Abstract:
This paper analyzes conversations recorded at various venues of shopping. A system of categories for use in conversation analysis is proposed. The system consists of four main areas: Task, Interpersonal relations, Fixation, and Speaking-orientation. Each area consists of two or three alternative categories. In the analysis, four categories (one category in each area) were assigned to each utterance (one thousand utterances in all). Thirty-one venues were divided into four groups according to certain characteristics. As a result of analysis, in conversations in venues of groups C and D, such as department stores, a cosmetic shop, and a real estate office, the interaction was found to be in stable equilibrium. Conversely, in conversations at venues of group B, such as street stalls, the interaction was found to be unstable.
Classification:
paper
Author:
FURUGORI Teiji
Title:
Effects of Top-Down Process in Reading Comprehension
Pages:
157-164
Descriptors:
comprehension; top-down process; text; experiment; natural language processing; language teaching
Abstract:

In this paper we describe two experiments on text comprehension and then discuss the importance of top-down process in language teaching and natural language processing.
Classification:
report
Author:
OGINO Tsunao
Title:
The Origins and Historical Period of Acceptance of Japanese Loan Words
Pages:
165-174
Descriptors:
loan word; historical period; origin; borrowing
Abstract:

A famous dictionary of loan words in Japanese edited by M. Umegaki includes more than 16,OOO entries with information of the origin and the historical period of acceptance of each entry. In this paper, the author tallied this information and clarified from the quantitative point of view how these loan words were brought into Japanese. The historical tendencies of Japanese loan words are as follows:
(1) recent Japanese have accepted more words than previous generations, and
(2) the origin of loan words changes with the times. The main origin in each era is: Sanskrit in the Nara, Heian and Kamakura eras, Portuguese in the Muromachi era, Dutch in the Edo era, English since the Meiji era until today. These facts reflect the Japanese history of cultural exchange.
Classification:
note
Authors:
TOKUDA Katsumi and SATO Yasumasa
Title:
Effects of Contributory Factors of Kanji on Estimation of Difficulty
Pages:
175-181
Descriptors:
Kanji; difficulty; estimation; contributory factors

Vol. 16 No. 5 (Jun. 18, 1988)

Classification:
paper
Authors:
IMAZU Hideyo and TANAKA Hozumi
Title:
Design and Implementation of Electronic Dictionary of Japanese
Pages:
189-204
Descriptors:
electronic dictionary; dictionary look-up; longest match; trie structure; structured dictionary
Abstract:

Electronic machine readable dictionaries are very easy and quick to access an entry through our terminal. However
(1) flexible dictionary look-up operations are not al lowed since an exact entry name should be specified, and
(2) it is not easy for human beings to read the contents of each entry displayed without indentation.
In order to solve the first problem, a don't care symbol is allowed to specify an entry name. It enables us to access an entry in lexicographically reverse order. Furthermore, as an entry name entered is matched, to a dictionary entry using the longest match strategy, the nearest entry, which has similar name entered, will be displayed even though an exact match will fail. Secondly, since a dictionary entry is displayed with indentations that are automatically generated, it is very easy to read.
Classification:
report
Author:
MIZUTANI Sizuo
Title:
Distributions of Semantic Types in Compound Words Written in Chinese Characters
Pages:
205-211
Descriptors:
word written by 3 Chinese characters; word written by 4 Chinese characters; compound word; derived word; meaning in word-formation; code for semantic pattern; distribution of semantic patterns

Vol 16 No. 6 (Sep. 14, 1988)

Classification:
paper
Authors:
TAKEUCHI Haruhiko and UTSUGI Akio
Title:
An Analysis of Pain Concepts by Network Model
Pages:
233-245
Descriptors:
network model; Pathfinder; multidimensional scaling; linguistic concepts; pain description; word relatedness
Abstract:

Pathfinder is a scaling technique for constructing network representations (Schvaneveldt et al., 1985). Pain descriptions were analyzed by Pathfinder and multidimensional scaling (MDS). It is shown that Pathfinder is useful for compensating the local distortion which appears in the configuration constructed by MDS. Pathfinder was applied to the developmental difference data in linguistic intuition of word relatedness in a sentence. The relation between the number of the links in the Pathfinder solution and language development was discussed.
Classification:
report
Author:
MIZUTANI Sizuo
Title:
NBG-Based Representation of Japanese Grammar
Pages:
246-257
Descriptors:
present-day Japanese; syntax; NBG; class logic; generative grammar; recognition grammar
Abstract:

To specify the contemporary Japanese grammar this paper proposes a method of formal representation without any new descriptive device or special modelling for linguistics. The proposal is based on the well- known language of the axiomatic set-theorY NBG. Taking such a system, most problems of grammar, say parsing, are reduced to the proof-problems in this formal system. From the same rule prescribed in such way, we can deduce rules both for generation and for recognition, by logical operation of a routine work. Our system itself is free from any specified metaphysical disposition on the human language. So, the tentative grammar in this paper is written from the traditional point of view which has been maintained since the Edo period.
Classification:
paper
Authors:
IGARASHI Jitsuko and YAMADA Sae
Title:
A Logical Analysis of Negative Polarity
Pages:
258-270
Descriptors:
negation; negative polarity; quantifiers; aspects
Abstract:

It is noticed that the negative polarity phenomena and the generation of polar limits are interrelated. In order to make clear this relation we have analyzed some samples of negative polarity phenomena and also examined how the polar limits are produced. The model we used is a very simplified one, of which quantifiers are limited to non-negative integers. We have classified these quantifiers into several classes according to the accuracy of each denoted quantum, i.e., the accuracy of the quantum denoted by each quantifier is different one from another. In our formalization, we have used particular expressions with limit indications, to distinguish negative polarity expressions from ordinary negative expressions.
Classification:
report
Author:
MIYAZIMA Tatuo
Title:
Kanji Content in Akutagawa Prize Novels
Pages:
271-277
Descriptors:
Akutagawa prize novels; percentage of Kanji; word groups according to the origin

Vol. 16 No. 7 (Dec. 15, 1988)

Classification:
paper
Author:
MIZUTANI Sizuo
Title:
A Grammar for Exclamatory Sentences Called as Kantai in Classical Japanese
Pages:
281-312
Descriptors:
kantai; zyuttai; classical Japanese; discrimination between bun (sentence) and ku; conditional clause; kakabimusubi; ellipsis; insertion; non-Subj Pred structure; NBG-representation
Abstract:

It was YAMADA Yosio's proposal (1908) for Japanese sentences to be classified into two types of kantai and zyuttai: i.e. part of the so-called exclamatory sentences builds a special class named kantai. Verifying his assertion by analysis of some five hundred examples mainly until the thirteen century, I extracted three syntactic characteristics of kantai-sentences as mentioned below, on which I prescribe their rules of formation by a NBG class-logical representation. A kantai-sentence is formed by a kantai-ku or by an interjection (cluster). A kantai-ku is, so to speak, a noun phrase governed by a special postposition (cluster) whose grammatical function is sentence-making. The characteristics are:
(1) that any kantai-sentence does not base on the subject predicate structure, as YAMADA has already pointed out;
(2) that the syntactic device kakarimusubi which makes generally ku to be bun (sentence) is not applied to any kantai-sentence, whereas It functions in all zyuttai-sentences according to the traditional opinion since MOTOWORI Norinaga;
(3) that no ellipsis occurs In the kernel of any kantai-sentence therefore, of any kantai-ku. These ones have been essentially preserved to the contemporary Japanese. Collection of examples in classics is appended to the end of this paper.
Classification:
report
Author:
SANADA Haruko
Title:
An Automatic Conversion of Japanese Sentence Style: from Da-Style to Desu/Masu-Style
Pages:
313-326
Descriptors:
style conversion; style; common style; polite style; Da-style; Desu/Masu-styIe; conjugation of verb
Abstract:

The automatic generation of Japanese sentences is becoming very important as a subject of man-machine-interface. This paper is about an automatic conversion system used to convert Da-style to Desu/Masu-style. The following are the characters of the system.
(1) It is available on the personal computers; therefore, its dictionary files are very small.
(2) It converts only the end of the objective description.
(3) It converts verbs using a probability logic according to circumstances.
The automatic conversion system has two problems. One is about the subject of conversion, and the other is about the styIe of sentences in aim. There are two move targets for the former problem besides the conversion of the end of sentences, which is reported in this paper. These two are as follows:
(1) The conversion of predicates with limitation.
(2) The conversion of words to unify the word level used in a sentence.
As for the latter problem, it is necessary to consider the conversion of varied sentences to unified sentences, especially to Dearu-style.

Vol. 16 No. 8 (Mar. 16, 1991)

Classification:
paper
Author:
MIZUTANI Sizuo
Title:
A Method for Semantic Description of Exclamatory Sentences Called as Kantai
Pages:
339-352
Descriptors:
semantic pattern of kantai-ku; kantai; zyuttai; formal system for semantic description; syntacticosemantics; NBG-representation; program by Syusin
Abstract:

A method for semantic description of exclamatory sentences is Proposed on the basis of a Japanese grammar by NBG-representation (cf. Keiryo Kokugogaku. 16[7]). kantai-ku's, which can be sentences by themselves, are classified into four groups by means of their syntactic structures. Each meaning of four patterns is also described formally by NBG-representation. Based on such results, semantic characteristics of kantai-bun are compared with the case of zyuttai-bun. It is accented in this paper to be able to give logical descriptions of meanings even to exclamatory sentences.
Classification:
report
Author:
HAYATSU Emiko
Title:
Semantic Distribution of Paired and Unpaired Transitive Verbs
Pages:
353-364
Descriptors:
transitive verb; intransitive verb; transitive and intransitive pair; paired transitive verb; unpaired transitive verb; meaning of a transitive verb; semantic distributions of transitive verbs
Classification:
report
Author:
MIYAZIMA Tatuo
Title:
Frequency Distribution of Words in Classical Works
Pages:
365-377
Descriptors:
words in classical works; frequency distribution

Back to Homepage