Last month I discussed a report of a new study that showed correlations between wine and emotions. At the time I didn’t have access to the complete study, and was relying on a helpful synopsis by Becca Yeomans. I have since been able to view the original study, which is not yet available on line. The study is fascinating because it shows that we respond emotionally to wines which potentially opens a whole new arena for wine appreciation. Yet there are some very odd results that are hard to explain as I noted in my earlier post. Now that I’ve had a chance to read the study, I wanted to revisit the conclusions to be drawn from it, although it turns out that the puzzling results remain puzzling.
The study, conducted by the School of Agricultural, Food and Biosystems Engineering at the Universidad Politécnica de Madrid in Spain, consisted of a sensory evaluation of several wines by a trained panel in order to identify the organoleptic properties of the wine and a consumer evaluation of the same wines accompanied by an emotion response analysis. The wines evaluated were a Verdejo, Chardonnay, a rosé made from Garnacha, two Rioja Reserva level wines, one from 2012, the other 2013, and a Ribero del Duero Reserva from 2012.
The consumer panel consisted of 208 students of various ages, and roughly equal cohorts of men and women, whose only qualification was that they consumed wine at least once a month. After a warm-up flight, they were given wines in a random order to be judged according to how much they liked the wine and then were given a list of 26 emotion words from which to choose the emotions elicited by the wine.
Here is a summary of the results
- 22 of the 26 emotion terms showed a significant correlation with the wines.
- “good”, “happy”, “joyful”, “mild”, and “pleasant” were associated with fruity and floral aromas
- aromas/flavors of vanilla, clove, and licorice were associated with “aggressive” and “guilty”.
- Astringency, the drying sensations caused by tannin, was associated with “aggressive”.
- Subjects grouped wines differently based on sensory analysis vs. the emotion analysis.
- Men and older adults scored the emotion words higher than women or younger adults but women were more discriminating regarded the emotions elicited.
- Young adults scored the two white wines and the rosé higher for the emotions “good”, “happy”, and “pleasant”; the 2013 Rioja Tempranillo scored lowest on these emotions for this cohort.
- Older adults identified “mild” as the distinguish characteristic between the white wines, rose and the 2013 Tempranillo.
- Negative terms such as “guilty” or “worried” were more prominent for younger adults.
The gender and age differences are interesting but I have no idea how to explain the results so I will ignore them.
I have no training in reading statistical analyses so I’m not sure I have a sophisticated understanding of the paper. But here is what I find puzzling.
If I’m reading the charts correctly, liking a wine was positively correlated with the positive emotions. But the negative emotion terms such as “aggressive” or “guilty” were associated with aromas of vanilla, clove, and licorice, and with astringency. These are all descriptors associated with big, red wines such as at least some Tempranillo. The positive emotions were associated with white wines.
But surely we can’t conclude from this that tannic, red wines with vanilla and clove notes are unpopular because they evoke negative emotions. I looked up data for wine sales in Spain and they show a decided preference for red wine especially Tempranillo which is the signature grape of Spain.
The consumers were asked to assess their degree of liking a wine before identifying the emotions so these results can’t be easily explained by the emotion biasing the judgment about the wine’s likeability. The only explanation I can come up with is that consumers outside the laboratory don’t make the connection between likeability and emotion. Only in the laboratory when the test subjects know they’re expected to make the connection is the association salient thus leading the subjects to have an experience unlike that of ordinary consumers.
But there is one more hypothesis that is worth mentioning. The authors of the study state,“If a wine conjures up negative emotions for an individual, that person is probably not going to buy that bottle of wine in the future.” Whatever the relationship is between wine and emotion, I doubt that wines “conjure up” emotions. As I have argued elsewhere the association between wine and emotion is typically metaphorical. We associate astringent wines with aggressive emotions but they don’t make us feel angry or afraid. It’s an association, not a direct causal relationship between the wine and an emotion. “Conjuring up” may be the wrong way to describe this relationship. Similarly, a piece of music may express anger or fear, but we don’t feel angry or afraid when listening, and it’s that expression that we find enjoyable despite the fact the emotion may be classified as “negative”.
If in conducting the test, the experimenters were encouraging the study participants to think of wines as causing emotions rather than metaphorically expressing them, their experience of the wines might be quite atypical. A wine that causes one to feel aggressive or guilty might indeed be unpleasant; a wine that expresses aggression or guilt might be interesting.
I have no idea how much plausibility to assign either of these explanations.
On the surface, the study seems well done and the results are certainly interesting. But the fact that the results of the study seem to contradict what we know about wine preferences means that much more work will have to be done to tease out the factors that explain this relationship between emotion and wine.