Crossmodal Correspondence between Wine Aroma and Visual Cues
Heatherly, M., M. Foley, J. P. Munafo, and C. R. Luckett.  2018.  Chemical Senses, Volume 43, Issue 4, 23 April 2018, Pages e1Ėe136.

Abstract:
Crossmodal correspondence, or the mindís inclination to associate sensory stimuli from at least two different sensory modalities, is an area of growing interest for those who study food perception. When eating or drinking, the background noise, the taste/aroma of the product and even the packaging itself can all play a role in our experience. Studies, using packaging, have demonstrated that changing the color of the bottle, or the label itself can influence the taste perception of the liquid inside (1-4). Generally, crossmodal research regarding aromas and colors and shapes have used simple aromas (i.e. a single odorant) and involve line scales anchored with abstract visual 2-D shapes with varying degrees of sharpness/roundness (5-6). The purpose of this study was to investigate whether odor-color-shape crossmodal correspondences exist in a complex odor matrix and in a more real-world setting. In both experiments, five odorants were used to create: buttery, citrus, floral, smoky, and vegetable forward wines. All odorants were diluted into 100ml Kendall Jackson Chardonnay to the respective serving concentrations. In Experiment 1, visual stimuli consisted of two sets of 3D printed shapes. A projective mapping task was performed, in which panelists were instructed to distribute the wine aromas and shapes onto a sensory space, based on similarities and differences. In Experiment 2, visual stimuli consisted of eight wine labels affixed to wine bottles. Two levels of angularity were presented on the label backgrounds: rounded and sharp, as well as four colors: red, brown, yellow, and green. Panelists were asked to rate how pleasant each wine aroma was, how well the aromas matched each of the wine labels, and how well they liked the labels. Experiment 1 confirmed the existence of odor-color-shape crossmodal correspondences in a complex aroma matrix. Experiment 2 confirmed the existence of odor-color crossmodal correspondences in a real life setting, but failed to identify the existence of odor-shape crossmodal correspondences with more subtle stimuli. Weak correlations existed between hedonic scores and groupings, as well as with matching scores. Further research should focus on determining the role hedonics play in crossmodal correspondences. Furthermore, the specific impacts of odor complexity on olfactory-visual crossmodal correspondence needs to be further investigated.