This paper is an attempt to formulate the derivational process of Japanese and English sentence adverbials based on universal rules and principles along the line of Greenbaum (1969), Ross (1973), etc., and to account for four semantic constraints imposed on sentence adverbials concerning the ‘scope’(‘focus’) of Q, NEG, IMP, and Sentential Pronominalization (abbreviated as SP) in terms of the concept of ‘functional stratification’(Langacker 1975 and Sawada 1975). The constraints are represented by (1):
(1) Sentence adverbials, in principle, are not included within the scope (focus) of Q, NEG, IMP, and SP.
Predicate adverbials, by contrast, are free from such constraints as in (1).
The explanation of (1) proposed in this paper is mainly based upon a recent linguistic and philosophical study of ‘performative sentences’ or ‘illocutionary force’ of an utterance (Austin 1962, Searle 1969), and also upon the concept of ‘chinjutsu’(‘modality’) in traditional Japanese linguistics (Yamada 1936, Tokieda 1941). I propose and argue the following principles:
(2) An utterance (or sentence) consists of three strata: a ‘performative stratum’(highest), an ‘attitudinal stratum’(middle), and a ‘propositional stratum’(lowest). The first and the second ones both indicate ‘illocutionary force’, and the last one, ‘propositional content.’
(3) The scope (focus) of Q, NEG, IMP, and SP does not stretch into an ‘attitudinal stratum.’
(4) Sentence adverbials belong to an ‘attitudinal stratum.’
Principles (2)-(4), which are of epistemological character and are deeply related with speech act theory, account for naturally and systematically the otherwise mysterious and inexplicable constraints in (1).
Finally, the above principles can be justified on independent grounds, and sentence adverbials prove to be a sub-set of a wider range of ‘subjective’ expressions, as is the case with ‘epistemic modals, ’ or ‘parentheticals, ’ which also belong to an ‘attitudinal stratum.’

View full abstract

The dialect dealt with in this paper is the one spoken in Rankin Inlet. N. W. T., which the author learned in 1977 from Mr. S. T. Mallon of the Government of N. W. T. and others, most of whom were native speakers of the dialect.

View full abstract

The kinds and number of word accent types of Japanese differ according to the dialect. Some dialects lack phonological distinction of word accent types. Systematization of accent types and distribution of dialectal accent areas have already been investigated by many scholars but quantitative studies concerning the accuracy of identification of accent types have rarely been tried before.
Using synthetic stimuli, the auther conducted some identification tests of word accent types on high school students in six cities, which covered four different dialectal accent areas. In addition, stability of dialectal accent types pronounced by many informants in Nagasaki was compared with that in Osaka, and the Nagasaki dialect was found more unstable in producing accent types than the Osaka dialect.
As the result of some perception experiments, individual differences were found both in the categorical boundary and in the identification accuracy. Dialectal difierences in boundaries were also found between certain accent types when one of the types was meaningless for certain dialectal subjects. The identification accuracy of subjects in Osaka, Tokyo, and Osaka had no significant difference, while it was significanly lower in Nagasaki where production of accent was unstable, and also in Fukui and Yonezawa whose dialects lack phonological distinction of accent types.
Accent perception was presumed to have some connection with accent production.

View full abstract

In the past, linguistic geography has been sharply differentiated from sociolinguistics. The former assumed homogeneity in a community and dealt with local variations in a language, whereas the latter assumed heterogeneity and dealt with social variations. The author's computerized Generalized Linguistic Atlas Printing System (GLAPS) makes it easy for linguistic geographers to produce a wide variety of linguistic atlases using computers. It thus opens up the possibility of advancing the field of linguistic geography by the easy incorporation, and analysis, of complex sociolingustic variables. The following investigation illustrates this.
A team of the Department of Linguistics, University of Tokyo, conducted an investigation of Tokunoshima Island, Kagoshima Prefecture, in July 1976. The team selected and interviewed 124 informants in order to examine distribution patterns of word-forms within the local dialect. Our findings have been reported in a book listed in the bibliography. The author, a member of the team, was assigned to analyse four items, including the word “palm ”.
Fig.5, an ordinary linguistic map of “palm ”, is a ‘GLAPS’ output. Its distribution pattern suggests how various words spread from Kametsu, the largest community of Tokunoshima Island. Other ‘GLAPS’ outputs show the data of fig. 5 in terms of sociolinguistic variables. For example, figs. 6-10 show the correspondence between word-forms and age. Fig. 11, a cross tabulation, reveals that people who often go to Kametsu tend to use words widely used there. Other variables are indicated in fig. 4, lines 3 through 6.
The inclusion of sociolinguistic variables in linguistic geography research thus allows us a more sophisticated understanding of dialect distribution patterns.

View full abstract