In a recently published review article, Portuguese microbiologist Manuel Malfeito-Ferreira argues that wine tasting is essentially synthetic or holistic in contrast to the dominant view in the wine industry that views odor perception as analytic.
The analytic model assumes aroma perception proceeds via the detection of key molecules. For instance the lychee aroma in a Gerwurztraminer is caused by terpenes. By contrast, a synthetic, or what he calls the configural, model views aroma perception as the perception of relations between molecules which lose their individual identity when in a mixture. That lychee aroma is the product of a complex interaction among several molecules on the synthetic model. The result of employing the mistaken analytic model of tasting, according to Malfeito-Ferreira, is that the current regime of wine tasting de-emphasizes the aesthetic properties of fine wine when assessing wine quality because aesthetic properties are synthetic, not analytic.
I discussed the first two premises in his argument in a former post. The first relies on the evidence coming from neurophysiology that signals from the olfactory bulb are directly processed by the orbital frontal cortex that regulates cognition, emotion, memory, and behavior. In other words, the olfactory image is shaped by “top-down” cognitive functions. What we perceive is not a simple matter of detecting molecules.
The second premise is the result of studies that show that 95% of trained individuals can recognize only three aromatic molecules in a single mixture. Thus, the long list of aromas wine tasters discover in a wine cannot then be because each are correlated with the detection of distinct molecules.
In this post, I want to point to two key explanatory premises in his argument. He writes:
According to this concept, the aroma of wine is characterised by a few molecules (about 30) that act as a buffer (the “aroma base”) because differences in their concentrations are not reflected in different aromatic perceptions. These molecules are the same in white or red wines and only the so-called “impact molecules”, or families of molecules, may break the aromatic “buffer”. Examples of impact volatiles are terpenes in varieties like Muscat or off-flavours. In the absence of impact molecules, wines may be described as odourless and different varietals are then indistinguishable (Campo et al., 2008) and have a light, sweet, pungent, alcoholic, and somewhat fruity aroma (Ferreira, 2010).
In other words, the basic aroma of wine is a synthetic perception that does not rely on the detection of individual aromatic compounds, since differences in concentration of these compounds do not affect the overall “generic” wine flavor. Only certain impact molecules, such as terpenes, will make a difference in aromatic perception.
However, even the perception of “impact molecules” is synthetic rather than analytic.
The simple binary mixture of ethyl isobutyrate (strawberry-like odour) and ethyl maltol (caramel-like odour), for example, is perceived as more typical of a pineapple odour, like allyl hexanoate (pineapple-like odour), than the odours of the individual components (Tromelin et al., 2020).
The mere presence of ethyl isobutyrate (strawberry-like odor) does not entail a taster will smell strawberry, and the mere presence of allyl hexanoate (pineapple-like odor) does not entail a taster will smell pineapple. The presence of individual aroma compounds is not a good predictor of what we smell. Thus, the primary way aroma scientists conceptualize aroma perception is misguided.
The impact of aromatic molecules on odour perception is frequently measured by OAV’s [odor activity values] obtained by the ratio between the odourant concentration in a mixture and its sensory detection threshold….In complex samples, the broad utilisation of OAV’s to predict the flavour does not seem reasonable, because several odours act as suppressors or enhancers of aromatic notes, even at sub-threshold concentrations (Ferreira et al., 2016).
Aroma is best described as “single aroma vectors” composed of several molecules—a synthetic or holistic perception—rather than a simple detection of a single compound. The perception of lychee in a Gewurztraminer is not just the detection of terpenes but the synthesis of many compounds including terpenes and a variety of tastes, textures, memories, beliefs, judgments and emotions that leave a distinctive impression on the taster. The influence of texture and mouthfeel on aroma perception is especially interesting:
Moreover, cross-modal interactions occur between aroma, taste and texture perception, changing the perception of equal odourant concentrations (Baldovini & Chaintreau, 2020). This is particularly relevant in the aroma of fine wines not dominated by single aromatic molecules or aroma vectors (McKay & Buica, 2020) and strongly dependent on the astringency sub-qualities that make sensory prediction from chemical composition rather difficult.
The very same combination of aromatic molecules can produce a different impression due to differences in taste and texture.
As you can see, the science of aroma and taste perception is moving away from a simple stimulus-detection-identification model of how aroma perception works. This has some potential implications for the value of wine descriptions which I will discuss in the next post in this series