The article by Robert P. Erickson raises a most relevant issue concerning certain neglect about research on taste when compared with other sensory systems. As Erickson proposes, this neglect may have been due in part to previous misconceptions about the simplicity of taste detection and perception. Its primitive phylogenetic and ontogenetic origin may have contributed to the current state of affairs, since primitive is often misinterpreted as simple. Or, conversely, the complex nature of the processes involved in taste sensation, perception, and cognition may have hindered a more advanced knowledge in the taste field. Although the scenario appears to be simpler at the periphery, since distinct receptors and cells have been identified for some specific tastes – a finding common across species from drosophila to mammals – central taste processing involves complex reciprocal interactions among different brain levels. There are data supporting distinct subsets of neurons maximally responsive to the quality and also to the hedonics of specific tastes, originating at the brain-stem level (Yamamoto Reference Yamamoto2003). Moreover, at higher brain levels, taste perception engages combination, comparison, and interactions not only across taste qualities, but also from other sensory modalities, motivational state, and memory. A critical feature adding complexity to taste sensation is the dynamic nature of the gustatory system, as it is modified by learning, even at the lower brainstem level (Yamamoto & Yasoshima Reference Yamamoto, Yasoshima and Bermúdez-Rattoni2007). Moreover, in spite of a general misconception about the low cognitive level of taste memory, it shows complex learning phenomena dependent on the hippocampal system (Gallo et al. Reference Gallo, Ballesteros, Molero and Moron1999), and it shares molecular mechanisms with other complex types of memories (Shema et al. Reference Shema, Sacktor and Dudai2007).

Considering this complex scenario, it does not seem probable that a unique theory, either labeled-line or population coding, will be solely enough for explaining taste sensation and perception. The article by Erickson offers a much-needed opportunity for debate on the topic. Among the various issues the target article raises, that of the singularity of mixtures merits attention. It is stated that mixtures are mostly perceived as being singular. This assertion seems to be mainly based on psychophysical studies in humans. However, there is a bulk of evidence from behavioral studies in animals that may offer additional insights. The evidence shows that, depending on the behavioral procedure and the requirements imposed by the learning task, compounds may be treated either as a single configural cue, or as a combination of individual taste cues. In any of the cases, the mixture cannot be reduced to the sum of the elements. However, several phenomena that have been demonstrated in rodents by using conditioned taste preference and aversion tasks, support the idea that the segregation of individual taste cues in mixtures may take place as long as the task would require such discrimination.

First, in conditioned blocking, previous training with an individual member of a compound modulates learning about another member of the compound. Thus, a previously learned aversion or preference to an individual taste A retards the acquisition of a new learned response to a different taste B if the later is presented in a compound AB (Balleine et al. Reference Balleine, Espinet and González2005; Gallo et al. Reference Gallo, Ballesteros, Molero and Moron1999). The blocking phenomenon would not appear if the compound had been sensed as singular and the individual basic components had not been recognized by the animals.

A second example of phenomena showing the possibility of taste segregation in mixtures is perceptual learning. On the one hand, perceptual learning takes place when complex stimuli are required to be discriminated. The effect consists in facilitating the learning discrimination by previous non-reinforced exposures. On the other hand, previous non-reinforced exposure may retard learning if simple stimuli are used – an opposite well-known effect termed latent inhibition. Both perceptual learning and latent inhibition have been demonstrated with taste cues and taste compounds. The fact that previous exposures to the mixtures do not induce latent inhibition, but rather increased learning discrimination, does not support the idea that they are perceived as singular cues. Although different explanations have been proposed for explaining taste perceptual learning, the evidence does not support the idea that configural cues are generated by the presentation of mixtures, nor the latent inhibition of the common element as the sole explanation. Rather, a reduction in generalization between the taste mixtures by previous exposure, that it comes from interactions between the unique elements, has been proposed (Mackintosh et al. Reference Mackintosh, Kaye and Bennett1991).

Finally, the basic conditioned taste preference tasks are clear evidence showing that taste mixtures are not always sensed as individual cues. Learned taste preferences for an individual taste cue, such as quinine or saline, are developed by mixing each with sucrose. The effect cannot be explained by a sensory preconditioning effect in which the taste becomes associated with the positive hedonic features of sucrose, but by associations with the reinforcing caloric properties. Whatever the explanation, if the mixture would have been sensed as individual and distinct of the components, no learned preference to the individual components should be evident.

Overall, the animal behavioral data support the fact that individual taste cues can be perceived as distinct within taste mixtures. Further insight into this issue could be gained by research on preverbal human newborns and children. The fact that they can discriminate the basic tastes is supported by their distinctive facial response patterns. Recording behavioral reactions to taste mixtures could provide useful information.

A complete understanding of taste processing requires a multidisciplinary approach that integrates results from molecular biology, genetics, electrophysiology, neuroimaging, behavior, and psychophysics, both from evolutionary and developmental perspectives. Efforts aimed to increase research funding and give more attention to taste training at school are desirable because chemosensory processes are profoundly influencing our everyday life.