Dec 02 2013

Champagne Tasting

One of my primary goals for this blog is to reinforce, strongly and frequently, the notion of neuropsychological humility – the understanding that our perceptions and memories are deeply flawed and biased. There appears to be almost no limit to the extent to which people can deceive themselves into believing bizarre things.

Psychologists have documented these flaws and biases in numerous ways, and when confronted with demonstrations of such people tend to be amused, as if they were being entertained by a magic show, but do not necessarily apply the lessons to themselves and their own lives. This is one of the key differences, in my opinion, between skeptics (critical thinkers) and non-skeptics – a working knowledge of self-deception.

With that in mind, here is yet another study showing that extrinsic factors and expectation affect our sensory perceptions. I will say right off that this is a small study involving only 15 subjects, but given the nature of the results it is still useful.

The researchers tested 4 champagne experts, 6 intermediates, and 5 novices with a blinded test of 7 different champagnes and sparkling wines ranging in price from £18 to £400. In addition to being blinded to the price, they were blind-folded so they could not see the color of what they were tasting. The results were essentially all over the place, showing no clear correlation to price or expertise. The intermediates and experts rated the £40 the highest, while the novices rated the £75 the highest. The novices and intermediates rated the £400 bottle among the lowest while the experts rated it average.

That much is not surprising – price does not necessarily relate to objective quality. For those on a limited budget, it is nice to know that inexpensive sparkling wine fares just as well as pricey champagne. Essentially, find something cheap that you like.

The tasters were also asked to estimate the ratio of red to white grapes contributing to the sparkling wine/champagne. Conventional wisdom has it that this ratio primarily determines the  unique taste of each product. The study showed that even the experts were completely unable to infer this ratio from blindly tasting the sparkling wine or champagne. Rather, their judgements seemed to be influenced by dose and alcohol content.

This small study (which should be repeated with larger numbers) is in line with earlier studies showing essentially the same thing. For example, Morrot, Brochet, and Dubourdieu performed a study in which they colored white wine red, and then had 54 tasters describe the wine. They used red metaphors to describe the wine, as would typically be used to describe a red, rather than white, wine.

This demonstrates that what we see influences what we taste. There are many more examples of how one sense will influence another – essentially our brains integrate multiple sensory streams into one coherent narrative. What we experience is in no way objective, it is altered to make everything seem to fit together.

In another study researchers exposed wine tasters to positive and negative statements about the wine; they did this both before and after they tasted the wine, but before they reported their impression of the wine. The comments affected the tasters’ evaluations of the wine when they occurred before, but not after, tasting the wine. This suggests that the tasters were not just following what they were told about the wine despite their experience, but rather that the positive and negative statements actually affected their experience of the taste of the wine.

A recent analysis of wine contests over several years, including expert judges being given the same wine from the same bottle multiple times, shows general inconsistency in rating wine (poor intra-rater and inter-rater reliability).

In yet another study the price tag associated with wine affected the experience of “pleasantness” of the wine. There is a consistent pattern in the research – subjective experiences can be modulated by suggestion, expectation, and other sensory cues.

Of course this has profound implications for areas beyond wine tasting. Suggestion and expectation, for example, would also influence our subjective reporting of symptoms in a clinical study.


Wine-tasting may be particularly subjective, but there is plenty of other research to indicate that the basic phenomenon of subjective experience being highly modifiable by external and internal factors, is generalizable. This has implications for everyday life, in addition to research into any subjective phenomenon.

It is for this and other reasons that good scientific protocol always strives to minimize or eliminate any subjective variables from research.

26 responses so far