Facts – empirical findings – are useless without appropriate accompanying theory. Employing the metaphor in the target article's Poincaré quotation, without theory the house is just a shapeless pile of stones. However, when theory is unsupported by empirical findings, there are no stones with which to form the house. Both good theory and accurate facts are necessary to understand behavior. Taste perception's empirical findings are complex, and therefore the theories explaining those findings must also be complex.
The target article critiques the traditional position that there are four basic tastes largely based on the argument that empirical support for this position is lacking. For example, the article shows that the four basic tastes concept predated empirical research, and speculates that one (nonempirical) reason for there being four is the limits of human cognition – that we can only easily remember less than ten members of any category. However, it is not clear how that argument explains the existence of precisely four basic tastes. Amoore and his colleagues, close to half a century ago, postulated the commonly accepted seven basic odors.
Another argument made against the four basic tastes is that it is not possible to take substances representative of the four basic tastes, combine those substances, and be able to identify the original substances in the mixture. However, when two basic colors of paint, yellow and blue, are combined the result is green paint, whose origins would be undecipherable to anyone not trained in color mixture. Yet this does not mean that yellow and blue are not basic colors.
A more fruitful approach to the four basic tastes, one that is followed by many scientists, is to think of them as taste concepts that describe or account for other tastes (Logue Reference Logue2004). Such an approach was used in the target article's experiment when the subjects stated what percentages of various comparison tastes were accounted for by different standard tastes. On average, the four basic tastes accounted for 84% of the comparison tastes, a greater percentage than for any other set of basic tastes or taste words. No statistical tests were reported, so it is difficult to confirm or disconfirm the target article's statement that “other stimuli do about as well as the basic … stimuli.” Regardless, it appears that it was possible for the participants to account for the huge majority of several comparison tastes using the four basic tastes.
Examining participants' ability to use the four basic tastes to describe or account for other tastes involves using the four basic tastes theory as a heuristic. Any heuristic's value depends on the degree to which it helps us to understand data. Does the four basic tastes theory help us to generate new, productive experiments? Can it help guide neuroscientists in looking for specific brain structures? In fact, the four basic tastes theory does seem to have assisted neuroscience research. Scientists have found evidence of receptors and brain structures dedicated to at least two of the basic tastes (sweet and bitter; see, e.g., Sugita & Shiba 2005). Without the four basic tastes theory, scientists might have taken longer to find these neuronal pathways.
Just because the four basic tastes were postulated before there was empirical research to support them does not necessarily mean that they are inconsistent with subsequent empirical research. However, the target article also argues that the current empirical evidence does not support the four basic tastes theory. The difficulty with this argument is not that it is inaccurate – current empirical evidence is indeed not consistent with there being just four basic tastes. The difficulty with this argument is that it is not new. Bartoshuk's excellent (1988) chapter clearly demonstrates that through the centuries many basic tastes have been postulated and investigated in addition to sweet, sour, salty, and bitter. As just a few examples, for years scientists have believed that there is a fifth basic taste, umami (Smith & Margolskee Reference Smith and Margolskee2001); for decades scientists have been investigating the taste of water (Bartoshuk Reference Bartoshuk, Atkinson, Herrnstein, Lindzey and Luce1988); and recently, the taste of polysaccharides has been added to the list (Sclafani Reference Sclafani2004).
That there are basic tastes – four or some other number – is an essential component of the labeled-line theory of taste coding described in the literature and in the target article. Consistent with its critique of the basic tastes, the target article repeatedly raises concerns with the labeled-line theory, stating that evidence does not support it. However, there is significant evidence, some of which is cited in the target article, supporting the labeled-line theory. For example, the target article cites Zhang et al. (Reference Zhang, Hoon, Chandrashekar, Mueller, Cook, Wu, Zuker and Ryba2003) as showing “data … suggestive of the receptor specificity claimed for sweet and bitter.” We also know that individual neurons in the chorda tympani nerve and in the brain tend to respond more to one of the basic tastes than to the others (Pfaffmann et al. Reference Pfaffmann, Frank and Norgren1979; Scott et al. Reference Scott, Giza, Yan and Murphy1998; Scott & Plata-Salaman Reference Scott, Plata-Salaman, Getchell, Bartoshuk, Doty and Snow1991).
Nevertheless, it is also the case that there are data that do not support the labeled-line theory. Most taste cells in mammals will respond to a wide variety of chemical substances (Scott & Plata-Salaman Reference Scott, Plata-Salaman, Getchell, Bartoshuk, Doty and Snow1991). Such data tend to support what Pfaffman called the across-fiber pattern theory in 1941, in which tastes are coded by patterns of neuronal activation, rather than by labeled lines (Bartoshuk Reference Bartoshuk, Atkinson, Herrnstein, Lindzey and Luce1988).
Taste coding is not simple. Many, probably most, taste researchers believe that there are more than the four basic tastes of sweet, sour, salt, and bitter. In addition, it is not the case that there is no evidence for the labeled-line theory. It is also not the case that there is no evidence for the across-fiber pattern theory. The facts are that how our bodies code tastes provides evidence to support both theories, to a degree. Most scientists are sophisticated enough to perceive, appreciate, and work with this complexity, and they should be credited with evidencing this sophistication. In the meantime, consideration of the four (or more) basic tastes theory, the labeled-line theory, and the across-fiber pattern theory, is likely to continue to prove useful in generating empirical findings that will help advance our understanding of taste coding.
