Behavioural Processes 61 (2003) 13–25 Preexposure to the stimulus elements, but not training to detect them, retards human odour-taste learning Richard J. Stevenson ∗ , Trevor I. Case Department of Psychology, Macquarie University, Macquarie, NSW 2109, Australia Received 22 April 2002; received in revised form 22 July 2002; accepted 8 August 2002 Abstract Odours are judged to smell sweeter following simultaneous oral pairings with the tastant sucrose and sourer after parings with the tastant citric acid. This effect may result from human participants perceiving and encoding a unitary odour-taste percept. This study examined two factors thought likely to disrupt such encoding; (a) preexposure to the mixture elements and (b) training to spot the elements of taste-odour mixtures. Half of the participants were trained to identify tastes and smells and half received no training. All participants were preexposed to two odours (A, B) and two tastes (X, Y), followed by pairings of these stimuli (AX, BY) and then by pairings between two non-preexposed odours and the same tastes (CX, DY). This process was then repeated on a second session. Odour-taste learning was retarded following preexposure, but was unaffected by training. These findings suggest; (1) that odour-taste mixtures may be cognitively impenetrable and (2) that preexposure leads to encoding of A and B, which are then resistant to interference when further pairings are presented (i.e. AX, BY). © 2002 Elsevier Science B.V. All rights reserved. Keywords: Odour; Taste; Flavour; Learning; Training; Preexposure 1. Introduction Although the human olfactory and taste systems are discrete (Pierce and Halpern, 1996), they evi- dence considerable interaction at a psychological level (Frank and Byram, 1988; Rozin, 1982). One probable consequence of this is odour-taste learning, in which an odour can acquire taste-like properties (Stevenson et al., 1995). In a typical experiment participants sniff two target odours, rating each on a variety of dimen- sions, most notably how sweet and sour they smell. The conditioning period then commences and partici- pants are asked to swill and spit a series of solutions, ∗ Corresponding author. Tel.: +61-2-9850-8098; fax: +61-2-9850-8062. E-mail address: rstevens@psy.mq.edu.au (R.J. Stevenson). some of which contain one target odour in sucrose and some the other target in citric acid. Participants later return to the laboratory and rate the same odours again, judging how sweet and sour they smell. A consistent finding, over several studies, has been that sucrose paired odours come to smell sweeter and less sour, whilst citric acid paired odours come to smell sourer and less sweet (Stevenson and Boakes, 2003a). One way in which odour-taste learning can be explained is by reference to configural encoding (Stevenson et al., 1998). According to this account, when a mixture is sampled by mouth participants simultaneously experience the odour and the taste, whilst paying little attention to either element. The consequence is a unitary perceptual experience which is encoded as such in memory. Later, when the partic- ipant smells the target odour alone, it automatically 0376-6357/02/$ – see front matter © 2002 Elsevier Science B.V. All rights reserved. PII:S0376-6357(02)00166-3