Sex perception, the ability to discriminate accurately the sex of an observed other, is a central prerequisite for all interpersonal interactions. Of particular interest here are the cues that modulate sex perceptions from multisensory input. Work using various types of stimuli suggests the existence of “sex tuned” neurons (Jordan et al. 2006; 2005; Troje et al. 2006
at least some of which are multi-sensory [see also Eagleman 2001; Shimojo & Shams 2001; Kovacs et al. 2004; van der Zwan et al. 2009)]. Of those studies two are particularly interesting in the present context. Using combinations of olfactory and visual cues (Kovacs et al. 2004) showed that both male and female olfactory cues make sexually ambiguous faces appear more often to be, respectively, male or female. Similarly, (van der Zwan et al. 2009) combined unambiguous auditory sex-cues with ambiguous visual sex-cues to show perceptual integration: The sound of female footsteps made sexually ambiguous point-light walkers (Johansson 1973) appear more often to be female (van der Zwan et al. 2009).
Two similarities between those two studies immediately are apparent. Both paired visually ambiguous sex cues with a cue from a second modality that was both sexually unambiguous and consciously perceived. For example, (Kovacs et al. 2004) presented volatile sex hormone-like steroids, androstadienone and estra-tetraen-ol, mixed into a scented paste, caused observers to resolve sexually ambiguous faces into specific sex categories. While the classification of these olfactory stimuli as human pheromones is still only supposition, androstadienone did shift observer’s perceptions of sexually ambiguous faces such that they more often were judged to be male. Similarly, estra-tetraen-ol caused observers more often to perceive sexually ambiguous faces as female. Those observations were interpreted as showing that observers could use an unambiguous olfactory sex cue to resolve ambiguity in a visual cue to give rise to unambiguous sex perceptions (Kovacs et al. 2004).
In much the same way, van der Zwan et al. (2009) used auditory representations to shift observer’s vision-based perceptions of sexually ambiguous point-light walkers (Johansson 1973). van der Zwan et al. (2009) developed an auditory walking sequence (a series of foot-falls) that observers reliably rated as sounding female. They went on to show, using an aftereffects paradigm, that when that auditory walking sequence was paired with a sexually ambiguous visual walker, observers would subsequently report that walker to be male. Those data too were interpreted as evidence that observers confronted with visually ambiguous sex information could use an unambiguous and consciously perceived cue from another modality to resolve perceived sex.
While there is precedent for auditory cues affecting visual perceptions (2009) the capacity for neural processes to use olfactory cues to resolve visual ambiguities is less well understood. To that end, a discussion of the level at which olfactory/visual interactive processing might occur can provide some insights. For example, the capacity for combined stimuli to induce aftereffects has been taken as evidence of true perceptual integration (Ernst & Bülthoff 2004;Ernst 2006). Similarly, the capacity for sub-threshold stimuli mutually to influence resulting perceptions can be interpreted as evidence that the processes by which that integration occurs are perceptual rather than, say, cognitive.
With that in mind, the experiment reported here was designed to further explore the nature of olfactory-visual interactions. A number of studies have shown that olfactory cues, even when they are not consciously perceived, can affect physiological processing and behaviours (2007; Lundstrom et al. 2003; Lundstrom & Olsson 2005). To our knowledge, what has not previously been shown is whether sub-threshold olfactory cues can be used to mediate sex perceptions and specifically, visual sex perceptions in the same way. Thus, the aim of this experiment was to determine whether a sub-threshold olfactory sex cue could affect perceptions in the same way as such cues have been shown to affect mood and arousal. Specifically, this experiment tested the hypothesis that observers would more often judge visually ambiguous walkers to be male when performing the task in the presence of male sweat, compared to when the olfactory cue was absent. Further, we predicted they would do so even when not able to report the presence of the sweat odorant.