In 3.1 we outline the methodological approach adopted in the subsequent delimitation and semantic analysis of Japanese taste terms. Details of the collection of data are given in 3.2, and the representational conventions employed for Japanese in the study are summarized in 3.3.

Delimitation and semantic analysis

The theory of lexical meaning presented in Chapter 2 is structuralist in nature: the meaning of a lexical item in a given language is basically a function of the intralingual semantic relationships which it contracts with other items. From this it follows that the elucidation of lexical meaning must necessarily be approached from within the language concerned. Such a view is supported by the well-known fact that lexical systems encoding ‘corresponding’ domains in different languages commonly prove to be anisomorphic, both in the placement of the boundaries of the domain and in its subdivision.

Before undertaking this study, the investigator was familiar, to a greater or lesser degree, with a range of Japanese lexical items whose meaning seemed to relate in some way to ‘taste’ in the widest sense, such as UMAI ‘good-tasting’, AMAI ‘sweet’, SHIBUI ‘astringent’, KOOBASHII ‘fragrant’, ABURAKKOI ‘fatty’, MIZUPPOI ‘watery, insipid’, SAPPARI-SHITE IRU ‘clean-tasting’, KORIKORI-SHITE IRU ‘resilient’, SUJIPPOI ‘stringy’, etc. What was not known, and what the investigation was intended to discover, was how such terms were semantically organized within the Japanese vocabulary. Do all such terms belong together in a single lexical system, or not? How is this to be shown? How can we decide in a linguistically principled manner what counts as a taste term in Japanese?

In this study, we approach these problems by means of a methodology based upon question-and-answer procedures.