When a food or beverage is placed into the mouth, flavor is experienced as a multitude of sensations, including taste, smell, touch, temperature, sight, sound, and sometimes pain/irritation (Delwiche Reference Delwiche2004). Taste and smell particularly have many demonstrated psychophysical interactions (Small & Prescott Reference Small and Prescott2005). Nevertheless, all so-called pure tastants used in many experiments including that of Erickson have been assumed to stimulate primarily taste buds and thus results on taste perception were ascribed almost exclusively to taste sensations. Some of these tastants, however, are also olfactory stimuli delivered to olfactory receptors by retronasal routes (Mojet et al. Reference Mojet, Koster and Prinz2005). Additionally, there is increasing consensus that flavor perception is predominantly based on olfactory sensitivity (Mojet et al. Reference Mojet, Koster and Prinz2005). For example, many patients confuse loss of smell as a gustatory problem (Hadley et al. Reference Hadley, Orlandi and Fong2004). Taste and smell interactions can happen at several levels of information processing before and at the cortical level. For a definitive example, when a subject is presented with a subthreshold concentration of an odor compound (i.e., benzaldehyde – a cherry/almond aroma) in conjunction with a subthreshold concentration of a taste compound (i.e., sodium saccharin – a sweet taste), subjects are able to detect the combination (Small & Prescott Reference Small and Prescott2005). Erickson's neglect of the close and important relationship of taste and smell causes his results to be questionable both on methodological and theoretical grounds. Considering the close relationship of taste and smell, it is unclear why if both olfactory receptors and taste buds can perceive a large diversity of odor and taste qualities, respectively (Axel Reference Axel2005), only taste has a limited number of basic elements (Hadley et al. Reference Hadley, Orlandi and Fong2004).

In olfactory research many efforts have been made in the last centuries to classify odors, though none of them have gained the wide acceptance as the model of four basic tastes. After the first odor classification by Linnaeus in 1756, researchers have proposed a number of basic odors comparable to that of basic tastes, ranging from 4 to 9. The existence of a limited number of odors called basic or primary or reference (“standard”) odors, however, has been consistently questioned (Dravnieks et al. Reference Dravnieks, Bock, Powers, Tibbetts and Ford1978; Zarzo & Stanton Reference Zarzo and Stanton2006).

In everyday life, tastes come principally from foods and beverages in discrete time periods, whereas odors are ambient. The relative inaccessibility of the olfactory epithelium compared to the tongue, as well as the higher technical effort needed to present odors compared to tastants, limited early work on olfaction. An experiment on odor perception analogous to that of Hanig (Reference Hanig1901) on taste perception, whose misinterpretation led to the unproven but commonly seen tongue map of anatomical specialization for the different qualities of taste, simply did not occur.

In our opinion, Erickson's call for control experiments about the concept of basic tastes should be extended to presentation of single as well as combinations of tastants and odors. In such studies, three types of interactions between odor molecules in mixtures (i.e., synergy, suppression, and hypoadditivity), which originate from integration mechanisms taking place at the single olfactory sensory neuron, should be considered (Duchamp-Viret et al. Reference Duchamp-Viret, Duchamp and Chaput2003). Different concentrations of each molecule should be examined, because the ligand repertoire of odor receptors broadens at higher concentrations (Duchamp-Viret et al. Reference Duchamp-Viret, Duchamp and Chaput2000). A further observation of interest is that increases in odor and taste intensity ratings are stronger for harmonious taste/odor pairs or taste/odor pairs that are typically encountered together (Delwiche Reference Delwiche2004). Tastant as well as odor grouping would both likely become more complex, particularly in view of these interactions.