Abstract
How pleasant or unpleasant we find an odor depends on our past experience. Similarly, the perceptual properties of an odor and how similar it seems to other odors are also subject to change. The ways in which liking for an odor and its perceptual properties can be altered have been examined in two main kinds of experiment. One involves what we refer to as odor–odor learning. Participants are asked to sniff some mixtures, each containing two odors and later to rate the individual component odors in various ways. Thus, two mixtures can be represented as AX and BY, where A and B are unfamiliar odors that participants find difficult to identify, whereas X and Y are more familiar and more easily identified. Sniffing the mixtures, AX and BY, a number of times changes the way in which the odors are subsequently rated. Thus, A is perceived as being more X-like and B as more Y-like. For example, after sniffing a mixture of champignol (mushroom-smelling) and citral (lemon-smelling), participants rate champignol as more lemonlike than participants who had sniffed a mixture of champignol and cherry. Furthermore, when asked to rate the similarity of pairs of odors, A is rated as more similar to X than it is to Y and B more similar to Y than to X. Thus, odors can acquire new properties simply as a result of being experienced in a mixture with another odor (e.g. Stevenson et al., 2003).