A Phenomenological Model For Generating the Tasting Description of Japanese Sake

Introduction

The description process (typically by sommeliers) of tastes and odors is widely thought to involve the following assumptions:

• 1.

  The wine has some flavors,

• 2.

  The taster detects the flavors via olfactory receptors,

• 3.

  The taster selects one of the flavors,

• 4.

  The taster identifies the exact word for the flavor.

This can be called “a binomial model of the sense and the expressions,” which leads to a myth: There would be a “right perception” and a “right expression.” However, in actual situations, our perceptual representations of taste or smell are often vague, and a set of verbal expressions is not given. In this paper, we adopt a phenomenological model that claims that the taste of the sake does not exist in the sake itself: It appears or arises between sake and the subject (taster) as a personal image. In other words, the object (in this case, the taste of sake in the real world) cannot exist independently of the subject (the taster). In this paper, the author tries to deal with not the object itself, but the “experienced representation of the object” from a first-person point of view, which highlights “my tasting experiences” as “I taste something.” This paper seeks to develop a multi-faceted model for generating tasting representations.

Background

The studies of the narrative are expanding toward various neighboring domains. shows the expanding structure of the narratology as “A comprehensive framework for research in narrative generation.” In the framework, Ogata classifies the research into four phases; the Narrative generation systems, Interdisciplinary narratology, Narratological approaches to various fields, and Related studies in the broadest sense.

This study is an attempt and an illustration of the analyzation of the multi-faceted narratological description of the sensory information. More specifically, the aim of this study is to show the relationships between the two modes of description – the verbal description and the pictorial description – of the sense of taste. This study would be classified into the “Narratological approaches to various fields” in the Ogata’s framework.

Among the inter-disciplinal domain of narratology, one of the most living areas would be the cognitive narratology. A series of cognitive scientific studies by Kanai et al. (Kanai & Kodama, 2010; Kanai & Ogata, 2004a, 2004b) has tried to reveal the relationships among story, non-story, or discourse, especially in films.

The author's study can be lined up in these cognitive narratological studies. Yet, the study presented in this chapter deals with rather lower – sensory level– cognition.

A natural language is generally an audiovisual-dominant symbol system. It categorizes the world into words: Cats are differentiated from dogs, which, in turn, are differentiated from wolves. Cats, dogs, and wolves are grouped together as a category of animals, which is different from fruit, which includes apples and oranges. Thus, a language operates according to the principle of differentiation. The principle works strongly or weakly in different domains because each domain has its intrinsic discriminable power. The discriminability of the sensory perception of taste is considerably weak, and thus, the number of words that name tastes directly is limited to not more than 10 words in the case of Japanese; these are the words for the five basic tastes, astringency, and dryness, plus a few words for complex tastes like koku (or ‘rich taste’).

It can be said that producing the representations of taste has two difficulties: the lack of discriminability or differentiation and the lack of expressions for the domain of taste. However, the door to the verbal expression of the taste is not completely shut.