Is Wine Tasting Perceptual Illusion?

[KMP]

Take 54 undergraduates from the Faculty of Oenology of the University of Bordeaux give them a white wine W (AOC Bordeaux 1996) containing semillion and sauvignon grapes and a red wine R (AOC Bordeaux 1996) containing cabernet sauvignon, also give them a list of odor descriptors and ask them to pick which descriptor is most intensely presented in each wine. Allow them to provide their own odor descriptors as well. One week later invite them back and give them the list they personally determined that described the wines (i.e. their individual list of chosen descriptors for the two wines) and ask them to apply those descriptors to two glasses of wine, wine W from the previous week and wine WR which is wine W colored red with purified grape anthocyanins (previously determined not to provide any odor to wine W).

What do you get? Well the number of terms used to describe wine W over the two tasting sessions is very similar. But in the second tasting for wine WR the subjects use many more odor descriptors of red wine than for white wine (about 3 to 1), even though the only difference to wine W is an odorless red color! What is also interesting is that in choosing odor descriptors in the first test they picked words represented by red or dark objects to describe wine R and yellow or clear objects to describe wine W.

OK, no big deal these are kids, albeit the future of the European wine industry. We all know what we are tasting, donÂ’t we? Our tasting descriptors arenÂ’t influenced by color, are they? Well the stimulus for this little test was an analysis done of the words used to describe red and white wine from 4 different sources; 3000 tasting comments by Jaques Dupont in La letter de Gault & Milan, a random selection of 3000 tasting comments from 32,000 in Guide Hachette du vin (a guidebook by a small group of professional tasters), 7000 comments from the Wine Advocate (Yes, Mr P himself), and 2000 tasting comments from a French winemaker. The most common characteristic was the use of odor descriptors for red or dark colors for red wine and yellow or clear descriptors for white wine.

(Morrot G, Brochet F, Dubourdieu D, The Color of Odors, Brain and Language 79 (2001) 309-320.)

Now, of course, its quite possible that the ODORS in red wine are best described by terms describing red or dark colors. After all who would argue with Mr PÂ’s nose when it comes to wine and especially from Bordeaux? Well there are these 54 students at the Faculty of Oenology of the University of BordeauxÂ…Â…Â…Â…!!

[GraemeG]

But I don't see what all this proves? I can understand - perhaps - tests done where attempts are made to separate visual input from sensory analysis. But to attempt to deliberately confuse the taster? What's the point? Is wine ever served like this? Do you open a bottle of white wine for your guests, add red food colouring, and bring it to the table?

It may even be defensible to colour both wine to colourless clarity - then tell the tasters that one is red and one white, and let them go from there. Perhaps something could be learned there about true preception of odours.

But if you're trying to measure aromatic perception, but then change some of the factors (colour) that tasters use as 'ballpark' markers - and not just change or make neutral, but deliberately mislead with them, what's the point? It doesn't achieve anything! What was the actual aim of the test?

cheers,
Graeme

[KMP]

But I don't see what all this proves? ..... What was the actual aim of the test?

cheers,
Graeme

Part of the stimulus for studies like this is that humans can discriminate odors well but have a problem providing verbal descriptions. By looking at how the 4 different wine tasting sources described odor they were able to show that this came down to associating descriptors of darker colors with red wine, and lighter color descriptors with white wine. That was supported with the first tasting test of the students. The students used odor descriptors associated with darker color for red wine, etc. So how do you test if the color of the wine is influencing the odor description? You don’t remove the color, you change the color. So the second tasting directly tests if the color red influences odor descriptors. Their conclusion is that “the tasters discounted the olfactory information”.

Personally I think its an interesting way to serve up a wine – much akin to putting a poor wine into an empty bottle from an outstanding vintage of Grange etc.

The point is that the color does influence how we describe a wine. The point to ponder, at least for me, is how much does the wineÂ’s color end up coloring how I describe the wine?

Mike

[AlanK]

I think the poor students were just confused.

If you give me a red coloured wine and asked me to pick the odour, I cannot help but would try to work out what grape variety it is...

Would be interesting if they

1) Blindfold the students, and asked them to describe say 5 red and 5 white wine.
2) Remove blindfold, and asked them to describe the same wines.
3) Compare results.

Cheers,
Alan

[KMP]

I think the poor students were just confused.

Alan, I'm not sure that confused is the best description, but if it is then you might well argue that they were very easily confused. They had two wines in front of them, both with the same odors but of different color. So what confused them, the color or the odor? The data appear to show that color was what influenced them. If it was simple confusion then among 54 individuals you might expect a more random use of descriptors. Its hard to extract exact numbers from the study but in the second test they used about 25 dark color descriptors and about 140 light color descriptors to describe the white wine. For the red colored white wine they used about 130 dark descriptors and about 30 light color descriptors - pretty much a complete reversal!

If you give me a red coloured wine and asked me to pick the odour, I cannot help but would try to work out what grape variety it is...

Certainly, something many of us try to do. But that's not what the students were being tested on. They were given a list of descriptors and asked to say which ones fit with which of the two wines.

Would be interesting if they

1) Blindfold the students, and asked them to describe say 5 red and 5 white wine.
2) Remove blindfold, and asked them to describe the same wines.
3) Compare results.

Cheers,
Alan

This seems like an interesting test to do at home, especially among a crowd of close friends. Let's see a show of hands for those brave enough!

Its actually been done in various ways for many years. If my memory serves me correctly Wine and Spirit Buying Guide used to do a thing at wine shows (way back in the late '70s?) where they would blindfold you and then ask you to identify whether a wine was red or white. I think the first year this was done many people complained that the wines were too similar to tell apart! Interesting conclusion because its been shown that if wine is put into dark glasses, so the color cannot be seen, tasters "modify" their description of the wine.

Mike

[AlanK]

Mike,

I guess I have to organize something with my friends now for the sake of science!

Cheers,
Alan

[Guest]

Mike,

I guess I have to organize something with my friends now for the sake of science!

Cheers,
Alan

Alan:

Great, let us hear how it goes!

I have an old drinking buddy visiting from Oz next week so I'm sure there will be something similar being done on this side of the Pacific. (Always wanted an excuse to buy black wine glasses. )

Mike

[Guest]

far, far, far, far, too much emphasis on wine evaluation is placed on the verbalising of descriptors - so what?

this adds nothing to good critique. I refer the WSET approach, youthful/developed and intense or medium weight.
do we really care if something smells like passionfruits on a summer's day in autumn.
youthful and floral covers it.

[DaveL]

This reminds me of the introduction to Jancis Robson's "Wine Tasting Workbook".
There she suggests that when many wines are tasted blind (I mean unable to see the colour of the wine) most people find it very hard to accurately distinguish between red and white. Now this makes it very easy to believe that people could be so easily influenced by the colour of the wine they were drinking.

All seems the same, very interesting, but I find if I try to drink blindfolded I very quickly end up tipping wine on myself.

[GraemeG]

far, far, far, far, too much emphasis on wine evaluation is placed on the verbalising of descriptors - so what?

this adds nothing to good critique. I refer the WSET approach, youthful/developed and intense or medium weight.
do we really care if something smells like passionfruits on a summer's day in autumn.
youthful and floral covers it.

Agree entirely. Tell me about acid, tannin, attack, palate coverage, balance and interplay of components. Who wants a list of the local grocer's shelves?

At an MW seminar once I heard Anthony Corrigan talk about sensory evaluation. Very informative (although his comments were really aimed at blind tasting). Make an observation and draw a conclusion from it. "I detect this, threerfore I conclude that. Makes you think about what you're tasting. Highly recommended I reckon.

cheers,
Graeme

[TORB]

Agree entirely. Tell me about acid, tannin, attack, palate coverage, balance and interplay of components. Who wants a list of the local grocer's shelves?

GG,

Agree totally but a flavor profile does have its users. For example, if it has charry oak or heavy prune flavors I want to know about it. Likewise, Mr OakHater Walsh will want to know if there is dill on the nose.

[KMP]

Agree entirely. Tell me about acid, tannin, attack, palate coverage, balance and interplay of components. Who wants a list of the local grocer's shelves?

cheers,
Graeme

Now we are getting somewhere. Thanks, Graeme.

I think those of us who learnt (i.e. were taught) to taste 20-30 years ago agree with this. And you must have noticed that most of what you list are not odor descriptors. However the problem is that we cannot ignore odor as its a very vital clue to many aspects of a wine, including the grape variety. The problem is how much emphasis should we place on odor.

I posted this topic for a couple of different reasons:

1) Many comments on wines do not place sufficient emphasis on the structural components (acid, tannins, mouthfeel etc).

A randomly picked example (not from any forum) Dark violet/purple, the color shows great concentration and youthful primary hues. The nose is all about dark plums, black berries and currants with excellent lift and brightness to the aromatics. Medium-full bodied there is terrific fruit purity with delicious black fruit flavors wrapped inside ripe, sleek tannins. Give this 2-3 more months and this will really come together and should drink exceptional for a couple of years.

The wine sounds truly wonderful, but I can't even evaluate if all that stuff is in balance. It really is all about what the wine smells like! But...

2) The comments of expert wine evaluators seem to indicate a close association exists between the color of a wine and the color of the descriptors, and

3) The study with the students suggests that odor descriptors can be influenced by the color of the wine irrespective of its odor.

So, for me, there is a very serious question here. When a wine is described in glowing terms that almost all revolve around its smell how confident can I be that this is an accurate sensory evaluation? Has the taster waxed so lyrical because the color, and no doubt the odor, has caused a domino effect leading to an exaggeration of odor descriptors to the detriment of a more meaningful description of the structural components of the wine? Does the taster even understand the importance of structure?

DonÂ’t get me wrong here folks. IÂ’m as guilty as most of trying to find too much in a wine, especially a favorite and IÂ’m not pointing the finger at anyoneÂ’s tasting notes. IÂ’m just worried that wine evaluations today seem to be placing too much emphasis on how a wine smells and not enough on how its structured.

Mike

[KMP]

You've got to laugh about all this. Here is the Wine Spectator's evaluation of the wine described in my previous post.

Bright, jazzy, swaggering style shoots off sparks of mint and white pepper around a lavish core of blackberry fruit, finishing with fine tannins and juicy acidity. Drink now through 2006.

Now were the swaggering and the sparks evident before or after they'd finished tasting?

Mike

[Guest]

Simply put: Our perception is influenced by past experiences.

There are many examples of this in regard to visual stimuli- why should smell be any different