The World of Fine Wine

Charles Spence, The Color of Wine (Part 1)

What role does vision—and in particular, color—play in wine appreciation? Does a wine taste any better or worse if you drink without seeing it, as in a “dine-in-the-dark” restaurant, say, or when it is served in an opaque glass? And does altering the color of the wine change its taste? While it might come as little surprise to find that inexperienced wine drinkers can be fooled simply by manipulating the color of a wine, experts surely wouldn’t be. Or would they? Research now shows that color plays a far more important role in our evaluation and appreciation of wine than certain experts would have us believe.

Over the past decade or so, a small but growing band of psychologists and cognitive neuroscientists has started to become interested in studying the contributions of each of the senses to the multisensory perception of flavor.

In terms of wine, aroma (involving both orthonasal and retronasal smell) and taste clearly play a central role. Similarly, the oral-somatosensory feel of the bitter-tasting tannins on the back of the tongue and the spritz of carbon dioxide in certain red wines are also very important. Serving a wine at the correct temperature is, of course, crucial, too. But what about the sound that a sparkling wine or Champagne makes in the glass, or the golden-yellow color of your favorite Sauternes? Research from my laboratory in Oxford has shown that people’s perception of the carbonation in a fizzy drink depends, at least in part, on the sound it makes in the cup. One might also wonder whether Champagne would taste as good if you couldn’t hear the bubbles popping? Barry Smith, in his 2007 edited volume on wine appreciation Questions of Taste: The Philosophy of Wine, has even suggested that fine wines appear to make a distinctive gentle glugging sound when poured from the bottle, though I have yet to see any empirical evidence in support of this particular claim!

But what about the color of a wine? Just how important a role does it play in people’s evaluation and appreciation? Well, the first glance is obviously sufficient to distinguish between “red,” “white,” and “rosé” wines. Beyond that, however, it has been argued that the qualitative and quantitative aspects of color (such as brightness, clarity, limpidity versus dullness, lacklusterness) also provide clues as to the age and/or maturity of certain wines, as well as to the cultivar (think only of the salmon hue in a Pinot Gris wine). Indeed, one can often tell the age of a white wine from the slightly golden oxidation. Meanwhile, for reds, coloration can provide clues as to how extracted the wine is—that is, how much leeching of the color and tannins there has been from the skins.

That said, a wine’s general color and appearance only contributes up to five of a maximum of 100 points according to Robert Parker’s famous Wine Advocate Rating System (see http://erobertparker.com/info/legend.asp; accessed March 15, 2010). In fact, Parker states that most wines nowadays get a score of either four or five for their visual appearance. In other words, color and other visual appearance cues meaningfully contribute only 1 percent of the variance to a wine’s total score. But are the visual attributes of a wine really of so little importance? Recent laboratory research suggests that color may actually play a far more important role in the perception of wine than Parker, at least, seems willing to acknowledge.

Straight or curved?

In the 1980s and ’90s, sensory scientists demonstrated that people’s judgment of a wine’s aroma and/or flavor would often change when they evaluated it in the absence of any visual cues as to its color, such as when tasting from an opaque glass rather than a clear one. Using this technique, color was shown to affect trained assessors’ evaluation of the aroma of both Bordeaux and Port wines. Research from this period also demonstrated that, for certain wines, aroma and flavor acceptance could be predicted on the basis of a wine’s color. So, for example, the Czech consumers in a study by Pokorn and colleagues preferred Moravian white wines with a prevailing yellow hue and disliked those having a slight greenish coloration.

These early studies tended to rely on presenting wines with their natural colors, or else obscuring the color of the wine, and measuring any changes that resulted in people’s ratings. By contrast, experimental psychologists like to place their subjects (or participants, as we call them nowadays) under conditions of sensory conflict, where one of the subject’s senses tells them one thing, while another says something different. So, for example, in one classic early experiment, people were given a straight rod to run their hands up and down while wearing distorting glasses that made the rod look curved. Would they report that they were holding a straight rod (thus showing that touch dominates) or one that was curved (thus illustrating visual dominance instead)? Seeing how the brain resolves such situations of intersensory conflict has helped scientists understand just how important each of the senses is when it comes to making a judgment about the objects and events in the world around us. In JJ Gibson’s classic (1943) experiment, the subjects reported that the rod was curved, thus showing complete visual dominance over their perception of an object’s shape.

In fact, over the past 50 years or so, researchers have repeatedly shown that it is vision that normally dominates over the other senses across a range of different experimental situations.

So, what happens when you introduce the intersensory conflict situation to the study of wine by, for example, artificially coloring a white wine red? Does vision dominate once again? And does the answer depend on whether or not you are testing wine experts? Scientists only got around to conducting this particular version of the conflict experiment at the start of the 21st century.7 Consistent with the results of the research cited above, vision was shown to have a much greater role in the evaluation of wine aroma than Robert Parker’s scoring system might lead one to expect. The participants in Morrot et al’s rightly famous study were 54 students enrolled on an enology course at the University of Bordeaux. In the first experimental session, they were given two glasses of wine (both 1996 AOC Bordeaux)—one white, a Semillon/Sauvignon Blanc blend, the other red, from Cabernet Sauvignon and Merlot. They were asked to make a list of aroma descriptors for the two wines. A list of descriptors was also provided to them. For each of the descriptors that the subjects chose to use, they had to indicate which of the two wines presented that characteristic more intensely. The students used one set of terms (such as honey, lemon, litchi, and straw) to describe the odor characteristics of the white wine and another set of terms (for example, chicory, coal, prune, chocolate, and tobacco) to describe the red wine.

In the second part of the experiment, conducted a week later, the students were given two further glasses of wine, one white, the other red, and an alphabetical list of the odor descriptors they had chosen during the first session a week earlier. They then had to say, for each odor descriptor, which of the two wines presented that characteristic more intensely. This time, however, what looked like the red wine was in fact the same white wine they had been given originally, but colored to make it look indistinguishable from the red. The crucial, and rather surprising, result was that the budding enologists now chose the red-wine aroma descriptors for the inappropriately colored white wine. These results have been taken by many to demonstrate, at least in enology students, vision’s dominance over judgments of a wine’s aroma. When it comes to wine, then, it seems that people smell what they see! That, at least, is how the popular press typically describe such results.

But, one might ask, are experts less likely to be fooled than novices? My own experience would suggest otherwise. María José San Román, the Spanish chef at Monastrell restaurant (www.monastrell.com), and I were able to fool one of the foremost wine tasters in Spain using exactly the same “trick” at the massive food-and-wine conference Lo Mejor de la Gastronomía, held in Alicante, Spain, in 2009. María José managed to get one of her friends in the wine industry to deliver half a case of what looked like red wine in suitably authentic-looking red-wine bottles. When these bottles were uncorked in front of the audience at our session at the conference, we asked the expert to evaluate the bouquet. It must be said that he took some time before coming to his decision. In the end, though, his struggle seemed to be about which particular red-berry fruit flavor it was that he could detect in the wine—was it raspberry or strawberry? Given the above, albeit anecdotal, example, one might wonder whether Robert Parker himself, or other equally respected tasters, would also be fooled by an inappropriately colored wine? Well, the great man is famously reticent about taking part in any kind of competitive wine tasting. Researchers in New Zealand, however, have now repeated the experiment with 29 wine experts (including professional wine tasters and winemakers).

Their results make for interesting reading. It turns out that wine experts are susceptible to some degree of color-induced biasing of their olfactory judgments. For example, the experts’ descriptions of the aroma of a young barrique-fermented Chardonnay when it was colored red were more accurate when it was served in an opaque glass than when served in a clear glass. This is despite the fact that they had been explicitly told to rate each wine irrespective of its color. Thus, it appears that experts can—to a certain extent, at least—be fooled by a wine that has been colored inappropriately. However, the experts in the study by Parr et al were able to discern the white-wine samples that had been masked with color to simulate an aged white wine (by giving it a golden color through the addition of turmeric and caramel). Would non-expert drinkers respond any differently?

A follow-up experiment conducted in 23 “social drinkers” showed that they were so bad at reliably identifying the aromas present in the wine that it was difficult to discern any pattern in the data when an inappropriate wine coloring was added. Indeed, more generally, it turns out that people find it extremely difficult to identify aromas in the absence of any other non-olfactory cue; typically, they can correctly name only about a third of even common everyday odors.It has been argued that it may be precisely because olfaction, by itself, provides so little useful information concerning an object’s identity that we come to rely on other cues, such as a product’s color, label, or taste when trying to judge its flavor. And while experts, trained assessors, and novices can all distinguish red and white wines on the basis of their aroma, all three groups fail when it comes to discriminating rosé wines.What’s more, the wine experts perform no better than the novices on this particular sorting task.

Visual flavor

In well-designed, successful products, color normally provides useful information with regards to the likely flavor and taste. Hence, it is logical for us, under normal everyday circumstances, to expect to be presented with the taste, aroma, and flavor profile that may be indicated by the color and any other visual-appearance cues (such as clarity or opacity). In a very real sense, then, we taste with our eyes. (This is what is sometimes referred to as “visual flavor.”) Similarly, it has been argued that there is, in nature, a predictable transition from unripe (and sour) green fruits through to ripe (and sweet) reddish fruits. It is for this very reason, or so it has been argued, that coloring a solution green leads to enhanced sensitivity to sourness, while at the same time reducing our sensitivity to sweetness. By contrast, coloring a fruit-flavored solution red can increase perceived sweetness by anything up to 12 percent.

It may have been this very lack of expected sweetness that the Czech consumers, in the study noted earlier, were objecting to when they were served the green-tinted Moravian white wines. The expectation that many people have is that yellow wines result from the use of fully ripened grapes—hence, a yellow wine should be less acidic, sweeter, and possibly also show more pronounced fruity and floral notes than a green wine. Indeed, in order to keep a white wine tasting fresh (rather than dull and flabby), one needs the right balance between sweet (ripe fruit) and sour (acidity), and clues as to this balance may often be present in the color of a white wine. On this particular occasion, however, the consumers’ visual flavor expectations turned out to be wrong, for the green wines were actually no more acidic, sour, or otherwise than the yellower wines, whose color they preferred.

But what about red wines? Enologists often say that red wines tend to have a deeper and richer color in good vintages. Tromp and van Wyk have gone further, arguing that color is perhaps the single most important factor in determining the quality of a red wine, assuming that it doesn’t have any noticeable defects and the aroma and flavor don’t differ too much from the expectations raised in the taster by its color. Given that the density of coloration in a red wine depends primarily on its anthocyanin content, coloration can undoubtedly provide an objective guide to the properties of young wines of the same varietal and from the same region. Given the correlation between anthocyanin content and ripeness, a lack of extractable pigment in the grape material will, generally, mean a wine isn’t going to be high quality. There are, though, undoubtedly exceptions to this rule, such as lightly colored and transparent Burgundies that can look weedy and thin and yet actually pack a lot of power and weight. There is also the danger of overextraction. If wine critics see a black, inky,purplish wine, they may well think that the producer has overdone it. Sensory scientists have demonstrated that color density and pigment parameters provided an objective guide to the quality of certain red wines, such as young Beaujolais. Of course, as a red wine ages, the contribution of non-anthocyanic polymeric pigmentation to coloration will increase, while the contribution of anthocyanins decreases as the tannins soften.

When it comes to wines, then, an expert can often use their color and clarity to make a fairly accurate prediction about the type of wine, its age, its likely bouquet, and perhaps also its taste profile. For the expert, all this information can be ascertained even before the aroma and/or flavor has been experienced—hence, cutting down the cognitive load when tasting unfamiliar wines or tasting “blind.” As Michael Broadbent MW puts it in his book Wine Tasting (1982; p.47): “The enormous advantage color has over smell and taste is that it can be so much more easily described, recorded or matched, and conveyed from one person to another.” The meaning of such subtle differences in shading etc may, of course, be lost on the social drinker. And it is for this very reason that expert wine tasters may be more susceptible to being fooled by inappropriate wine coloring—because their expectations are that much stronger than those of the non-expert.In a very real sense, we perceive, at least in part, what we expect we are going to smell and taste. It turns out that the discrepancy between the color of a wine (or any other product, for that matter) and its aroma/flavor has to be pretty dramatic before people will notice it and start discounting the evidence before their eyes.

Some caveats

Having said that, it is important to highlight two potential caveats with regards to the interpretation of the intersensory-conflict experiments on wine that have been conducted to date. The first point to note is that the participants in these experiments were only ever asked to rate wine aroma. While Morrot et al’s participants were allowed to taste the wines before making their odor judgments, those taking part in Parr et al’s study could only sniff them. This minor methodological detail may, I think, turn out to be crucial, because food scientists have shown that inappropriately coloring a solution can have a qualitatively different effect on our perception of beverages that are merely sniffed, as compared to those that are actually tasted. Coloring a drink will likely result in people judging its aroma as more intense when they merely sniff it (via the orthonasal route) but as less intense when they actually taste it (via the retronasal route).

What explains this most unexpected of results? We simply don’t know yet. But one intriguing suggestion is that it may be much more important not to be fooled by inappropriate coloring when a foodstuff is actually in our mouths (that is, on the point of being swallowed) and where the potential for poisoning is that much greater than, say, when merely evaluating (for example, sniffing) a substance that is external to our bodies (such as prior to tasting). Whatever the correct explanation for this result turns out to be, the key point remains that while artificially coloring a white wine red may have one effect on our perception of its aroma when sniffed (as in Parr et al’s study), it may have a completely different effect on people’s perception when they actually taste it (as in Morrot et al’s study).

The second caveat concerns a person’s beliefs about the color of a drink and its relevance to aroma, flavor, etc—what is technically known as “the unity assumption.” In research conducted here in Oxford, we have demonstrated that if you tell a person that the color of a beverage might actually be misleading with regards to its aroma, people can sometimes (but not always) ignore the information provided by their eyes. Williams and colleagues (1984) reported a similar result when trained (but non-expert) assessors, who were evaluating the aroma of 1976 Bordeaux wines, were told to ignore the (in this case normally occurring) colors of the wines before them. Note, here, that the participants were given no reason to believe that the color of the wines they were evaluating was anything other than informative in either Morrot et al’s or Parr et al’s conflict studies. Hence, it is at present an open question as to whether people (be they experts or not) would still be “fooled” in the intersensory-conflict situation were they explicitly told that the wines under evaluation might have been colored inappropriately.

A final point to note here concerns the generalizability of the effects of inappropriately coloring wines. Just how specific are the results to the particular wines that happened to have been used in the studies reported by Morrot, Parr, and their colleagues? Systematic research has yet to be conducted on this question, but it would seem probable that the addition of red coloring ought to have far less of an effect on those white wines expressing aroma notes that are sufficiently intense to make the wine distinctive (as is, for example, often the case with a young Marlborough Sauvignon Blanc or a St Bris—a Sauvignon Blanc from Burgundy instead of the Loire). Furthermore, given that the correlation between a wine’s color and its aroma, flavor, and textural properties is likely stronger for red wines than for whites, artificially coloring a red wine white should probably also have less of an effect on a taster’s aroma judgments than coloring a white wine red.

Conclusions

So, to return to the question with which we started, just how much does a wine’s color really matter to wine evaluation and appreciation? Well, quite a lot. Though, as we have seen, the answer is complicated and depends
on a number of factors, including the strength of one’s expectations that a particular visual attribute constitutes a good predictor of a specific aroma/flavor attribute. Wine expertise also plays a role, as does whether the taster believes that the color is meaningfully related—that is, correlated—to the aroma/flavor. What appears clear is that a wine’s color, clarity, and general appearance sets up an expectation in the mind of the taster about the quality and likely flavor profile, bouquet, and taste attributes of the wine. Despite the fact that olfactory cues can often provide sufficient information to discriminate various odor qualities/characteristics of wines, beers, etc, as soon as we see the color of the drink before us, we typically cannot help but be influenced in our judgments by the expectations that have been set up by our eyes. The expectations that help reduce the cognitive load associated with evaluating a new wine end up shaping the perceived attributes of the wine when a taster actually comes to taste it. That said, color does not seem to affect (or bias) all aspects of flavor perception equally, with the perception of aroma being more susceptible to biasing by inappropriate coloration than the perception of flavor.

Given the above, is it fair to say that Parker and most other critics underestimate the importance of color to wine? It would certainly be easy to come away from the research reviewed above with that impression. That said, it is important to remember that for most wines, color normally does provide very useful information about a wine’s type, “typicity,” age, and likely gustatory and olfactory attributes. This state of affairs contrasts markedly with the situation in the psychologist’s laboratory, where the predictable association between color and aroma/flavor has been broken artificially in the name of science. So, to the extent that the correlation between a wine’s color and its flavor profile remains high in the real world, Parker’s scheme might not be so far from the truth after all. Any variance in the quality of the wine (be it visible or not) will easily be captured by the 15 marks that Parker awards for the aroma of the wine (not to mention its “cleanliness,” which is both a visual and an olfactory property—for example, one can smell the fermentation faults and see the lack of clarity of the wine in the glass) and the 20 marks awarded to its flavor and finish. Luminosity and limpidity also help one to recognize a clean and stable wine. Hence, while the results of conflict experiments—which are so popular among psychologists and cognitive neuroscientists (your author included)—are undoubtedly impressive, it is important to bear in mind that real life doesn’t much resemble the intersensory-conflict situation.
Might things change in the future? Many fast-moving consumer-goods companies have already cottoned on to the marketing opportunities offered by deliberately manipulating the mapping between a food’s color and its flavor. Think only of the recently introduced Fruity Smarties or Skittles Confused product lines, where the colors and flavors of the individual candies have been deliberately mixed up. Similarly, some of the chefs that I work with—including Heston Blumenthal at The Fat Duck in Bray (www.thefatduck.co.uk)—have also started to play with their diners’ expectations, through dishes such Blumenthal’s signature beetroot-and-orange-jelly dish. How long, you might wonder, will it be before some forward-thinking wine producer tries the same thing with wine? I don’t think you should be too worried on this score, given that confusing the senses appears to be a strategy that appeals primarily to the younger section of the population, who are hopefully not the target of the marketers of wine. That said, it is worth remembering, in closing, that winemakers have, to a certain degree at least, been deliberating manipulating the color of their wines for centuries.

Acknowledgments

I would like to thank Barry Smith for providing so many helpful comments on an earlier draft of this paper.

Notes

1. See C Spence, C Levitan, MU Shankar, & M Zampini, “Does Food Color Influence Taste and Flavor Perception in Humans?” Chemosensory Perception 3 (2010), pp.68–84; RJ Stevenson, The Psychology of Flavor (Oxford University Press, Oxford; 2009) for recent reviews.
2. Humans have two relatively distinct senses of smell: one (older) system associated with the inhalation of external odors, known as orthonasal olfaction; the other (newer system involving the posterior nares) associated with the detection of the olfactory stimuli emanating from the food we eat, since odors are periodically forced out of the nasal cavity when we chew or swallow food
or taste wine, known as retronasal olfaction. We humans are purportedly the only creatures to possess the latter system. The distinction is important because the subjective attributes/quality of specific food and wine odors can vary quite substantially as a function of whether they are presented orthonasally or retronasally. See A Gilbert, What the Nose Knows: The Science of Scent in Everyday Life (Crown Publishers, New York; 2008).
3. See M Zampini & C Spence, “Modifying the Multisensory Perception of a Carbonated Beverage Using Auditory Cues,” Food Quality and Preference 16 (2005), pp.632–41; BC Smith (editor), Questions of Taste: The Philosophy of Wine (Oxford University Press, Oxford; 2007), p.53.
4. Note here that many other wine experts appear to place a far higher value on the visual appearance and color of wine than does Parker himself. So, for example, Emile Peynaud writes: “The first sense to come into action is sight. It informs us about the wine’s appearance. Its color (intensity, shade)—what we call the robe. We say a wine is ruby, garnet, purplish, or brickish. The first indication, often instructive, has a great influence on those that follow. The taster will be tempted to judge a wine more severely if it is cloudy or of abnormal color. The color intensity of a red wine is an indication of it’s ‘body’ and its volume; the hue, of its age. The color of a white wine shows its state of oxidation. The eye, therefore, plays an important part in tasting and is a preparation for it. To taste without seeing considerably increases the difficulty: It is not always easy to pick out among a series of wines tasted without looking which are the dry whites, the rosés, and the reds when these have little tannin. This blind-tasting exercise is very instructive.” Knowing and Making Wine (John Wiley & Sons, London; 1984), p.4. See also M Broadbent, Wine Tasting (Mitchell Beazley, London; 1982) for another writer who values the visual attributes of wine more than Parker seemingly does.
5. For example, see J Pokorn , M Filipu°, & F Pudil, “Prediction of Odour and Flavour Acceptancies of White Wines on the Basis of Their Colour,” Nahrung 42 (1998), pp.412–15; F Sauvageot & M Chapon, “La Couleur d’un Vin (Blanc ou Rouge) Peut-elle Etre Identifée sans l’Aide de l’Oeil?”, Les Cahiers de l’ENSBANA 4 (1983), pp.107–15; AA Williams, SP Langron, & AC Noble, “Influence of Appearance on the Assessment of Aroma in Bordeaux Wines by Trained Assessors,” Journal of the Institute of Brewing 90 (1984), pp.250–53; AA Williams, SP Langron, CF Timberlake, & J Bakker, “Effect of Color on the Assessment of Ports,” International Journal of Food Science & Technology 19 (1984), pp.659–71. Pokorn et al tested 10 Muscat, 11 Rulander, and 11 Traminer wines from the southern Moravia crop of 1994–96. These wines tend to be relatively sour, with the grapes typically being gathered before becoming fully ripe due to the relatively cold and humid climate.
6. Interestingly, when the subjects were told to close their eyes, the stick suddenly felt straight again, thus showing that the sense of touch was itself providing accurate information about the shape of the rod; JJ Gibson, “Adaptation, After-Effect, and Contrast in the Perception of Curved Lines,” Journal of Experimental Psychology 16 (1943), pp.1–31. For a recent review of the literature on the conflict situation, see C Spence & A Gallace, “Making Sense of Touch,” in H Chatterjee (editor), Touch in Museums: Policy and Practice in Object Handling (Berg Publications, Oxford; 2008), pp.21–40.
7. It turns out that JJ Gibson, who conducted the curved-stick experiment, and his wife Eleanor Gibson were also interested in the question of wine expertise, though never, as far as I am aware, conducted such a conflict experiment with wine. See JJ Gibson & EJ Gibson, “Perceptual Learning: Differentiation or Enrichment,” Psychological Review 62 (1955), pp.32–41. Instead, the experiment was finally carried by G Morrot, F Brochet, & D Dubourdieu, “The Color of Odors,” Brain and Language 79 (2001), pp.309–20.
8. Note that people generally tend to use yellow, orange, and clear sources of odor to describe white wines (think of lemon, honey, butter, and hazelnut, etc), while red-wine odors are typically described in terms of red and black fruits/or dark objects (blackcurrant, juniper, licorice, cherry, etc). See J Ballester, H Abdi, J Langlois, D Peyron, & D Valentin, “The Odor of Colors: Can Wine Experts and Novices Distinguish the Odors of White, Red, and Rosé Wines?” Chemosensory Perception 2 (2009), pp.203–13. F Brochet & D Dubourdieu, “Wine Descriptive Language Supports Cognitive Specificity of Chemical Senses,” Brain & Language 77 (2001), pp.187–96.
9. A control study confirmed that the red coloring itself (grape anthocyanins) imparted no aroma/flavor to the wine.
10. WV Parr, KG White, & D Heatherbell, “The Nose Knows: Influence of Color on Perception of Wine Aroma,” Journal of Wine Research 14 (2003), pp.79–101. It is worth noting here that while Morrot et al’s (2001) experiment employed a single trial design, Parr et al’s participants rated the aromas of 16 wines. This difference may be important because visual dominance (or capture, as it is sometimes called) is more likely to be observed in a single trial experiment than in a longer experiment where the repeated exposure to incongruent odor-color pairings may set up a context in which the taster starts to distrust the evidence before their eyes.
11. For reviews of the evidence, see DA Zellner, AM Bartoli, &
R Eckard, “Influence of Color on Odor Identification and Liking Ratings,” American Journal of Psychology 104 (1991), pp.547–61, or MU Shankar, C Levitan, & C Spence, “Grape Expectations: The Role of Cognitive Influences in Color-Flavor Interactions,” Consciousness & Cognition 19 (2010), pp.380–90.
12. The participants in this study (Ballester et al, 2009) were presented with 18 different wines (six each of red, white, and rosé) in dark glasses and had to try to categorize them as red, white, or rosé. All the wines were considered as premium or popular premium brands (including Côtes du Rhône and Coteaux du Languedoc) and came from different French vineyards from the 2005 vintage. See also Sauvageot and Chapon (1983).
13. JA Maga, “Influence of Color on Taste Thresholds,” Chemical Senses and Flavor 1 (1974), pp.115–19; for a review of the literature on color-flavor interactions, see Spence et al (2010).
14. See MG Jackson, CF Timberlake, P Bridle, & L Vallis, “Red Wine Quality: Correlations between Color, Aroma and Flavor and Other Parameters of Young Beaujolais,” Journal of the Science of Food and Agriculture 29 (1978), pp.717–27; TC Somers & ME Evans, “Wine Quality: Correlation with Color Density and Anthocyanin Equilibria in a Group of Young Red Wines,” Journal of the Science of Food and Agriculture 25 (1974), pp.1369–79. CF Timberlake & P Bridle, “Colour in Beverages” in AA Williams & RK Atkin (editors), Sensory Quality in Foods and Beverages: Definition, Measurement and Control (Ellis Horwood Ltd, Chichester; 1983), pp.140–54. A Tromp & CJ van Wyk, “The Influence of Color on the Assessment of Red Wine Quality,” Proceedings of the South African Society for Enology and Viticulture (1977), pp.107–08. Interestingly, Tromp and van Wyk showed that tasters could guess wine quality from its color but not wine color from its aroma and flavor.
15. For early evidence supporting this claim, see RM Pangborn, HW Berg, & B Hansen, “The Influence of Color on Discrimination of Sweetness in Dry Table Wine,” American Journal of Psychology 76 (1963), pp.492–95. The experts and non-experts in their study had to try to assess the sweetness of white wines that had been colored yellow, brown, pink, red, or purple to simulate Sauternes, Sherry, rosé, claret, and Burgundy, respectively. The experts rated the rosé-colored wine as being sweeter than the untainted white wine. By contrast, the sweetness ratings of the non-experts were not influenced by the color of the wine.
16. For reviews, see HNJ Schifferstein, “Effects of Product Beliefs on Product Perception and Liking,” in L Frewer, E Risvik, & H Schifferstein (editors), Food, People and Society: A European Perspective of Consumers’ Food Choices (Springer Verlag, Berlin; 2001), pp.73–96, or Shankar et al (2010).
17. BJ Koza, A Cilmi, M Dolese, & DA Zellner, “Color Enhances Orthonasal Olfactory Intensity and Reduces Retronasal Olfactory Intensity,” Chemical Senses 30 (2005), pp.643–49. That said, it is important to remember that in wines, at least, the retronasal route delivers so much more to the wine-drinking experience than the orthonasal route.
18. It is rather more difficult to color a white wine red; typically, a bleaching agent such as sulfur dioxide is used first to remove any red coloration before the addition of any golden coloration (using odorless food dyes). See CF Timberlake, “Colours in Beverages,” Food Flavourings, Ingredients, Processing, Packaging 4:7 (1982), pp.14–15 and 17. It is interesting to note here that Timberlake demonstrated that bleaching a red wine using sulfur dioxide until it became clear had no observable effect on the perception of the aroma of a Bordeaux rouge. It would be nice to see this controversial result replicated.
19. M Lelièvre, S Chollet, H Abdi, & D Valentin, “Beer-Trained and Untrained Assessors Rely More on Vision than on Taste when They Categorize Beers,” Chemosensory Perception 2 (2009), pp.143–53. See also Williams et al (1984).
20. See Spence et al (2010).

The World of Fine Wine

Archive