There are a number of theories concerning taste factors. Some identify only four factors, i.e., sweet, sour, bitter, and salty. Others assume more factors like freshness of acidulous spring, pungent taste of alcohol, astringency (
shibumi) like tannin, or delicious taste (
umami) of glutamic acid, etc. The purpose of the present study is to examine to what extent these taste factors are independent.
1) Kato and others (cited in reference 2) measured absolute threshold and discrimination capacities of 10 veteran tasters for alcohol, succrose, succinic acid, cafeine, and salt. We have calculated the intercorrelations among these scores. The correlation matrix was factored by the centroid method. It is clear from Table 3 that three factors are enough to explain these five tastes. Fig. 1 indicates that alcohol has a very peculiar component, and the other four, though they are quite different from each other, have a common factor. We call it tentatively a “general tastability” factor.
2) By the same method as in (1), threshold and discrimination capacities of 10 tasters for hydrochloric acid, succinic acid, and lactic acid were analyzed. The result shown in Fig. 2 indicates that lactic acid has a somewhat distinct component besides its sourness.
3) By the Torgerson-Indow's multi-dimensional scaling, similarities among eleven substances, whose concentrations were 2
3 times above their threshold values, were analyzed. Those substances were sweet (succrose, glucose, cyclohexyl sulfate), bitter (cafein, quinine sulfate), sour (tartaric acid, citric acid, succinic acid), salty (sodium chloride),
umami (mono sodium glutaminate), and
shibumi (alum pottasium). -Exp. III-
The results may well be represented by Henning's taste tetrahedron. (see Fig. 3), but the characteristics of
shibumi and
umami were not very clear, and vectors of these substances lie out-side of Henning's tetrahedron. Succinic acid at this concentration level is a completely sour substance, and similarity with glutaminate was not noted.
4) By the same method as in (3), we examined the change of similarity pattern shown by pyramidal framework of four fundamental tastes and the remaining two, when their concentrations were increased to threshold×2
4. The results of both Exp. IV and Exp. V were similar to that of Exp. III (see Fig. 5 and 6).
5) Throughout Exp. III-V, the reliability of numerical judgments concerning taste similarity, by naïve subjects, ranged around 0.6-0.8 (see Table 4 and 8). These values were higher than those of odor similarity judgments reported in reference 7.
抄録全体を表示