It has often been said that we taste with our eyes. How a wine appears in the glass matters a great deal, because ‘taste’ itself is a multimodal perceptive event involving a number of senses, including vision alongside touch, taste and smell. Even the information we have about a wine influences the actual perception of the wine: brain-scanning studies have shown that experienced sommeliers process the taste of wine in different ways to non-experienced wine tasters.
So the appearance of wine matters. It’s also an important cue in blind tasting, and the Master Sommeliers have a visual assessment grid that they use to help teach blind tasting. The visual points in this grid are as follows:
I thought it would be good to explore a little more what wine science can tell us about the visual cues that wine gives us. While the grid above is useful, is there anything we can glean from wine scientists that would help us refine it or add to it? There are also some aspects about wine’s appearance that are taught widely, but which may be without scientific basis. It’s always good to question what we are taught.
Let’s begin with color, and white wines. Some people think that the green tinge to the colour in some white wines is attributable to the presence of chlorophyll in slightly under-ripe grapes. I asked some wine scientists whether this might be possible. ‘I don’t think there is any chlorophyll in wine,’ says Liz Waters, R&D Program Manager at the Australian GWRDC. ‘This is reported by Zoecklein (1999), where he states that Amerine (1972) says “White musts can contain traces of chlorophyll, carotene, and xathophyll”,’ says Geoff Cowey, researcher at the Australian Wine Research Institute. ‘He doesn’t state if they are derived from the grapes, but I suspect it is more likely from leaf and stem contact during cold soak/pressing, which contain chlorophylls. The stability of chlorophylls at wine pH is also questionable.’
So what is it that creates colour in white wines? This is widely used in visual assessment. ‘White wine colour is due to polyphenolic-derived pigments,’ says Liz Waters. ‘Not all of them are known or identified, but xanthyllium pigments certainly play a role.’ Geoff Cowey expands: ‘as whites don’t contain coloured anthocyanins in their skins, and white generally have little to no skin contact, I suspect the green/yellow color of white wines are due mostly to hydroxycinnamic acids and phenolics in pulp, which can sometimes have a yellow color, which tend toward brown colours as the wine develops and these compounds become oxidised. There are some proanthocyanins in white grapes that could also contribute towards color.’ It seems as if this is a subject ripe for more research. ‘There doesn’t appear to be much work on it in recent times, but historically the majority of researchers mention flavonoid content, as affected by limited oxidation,’ says Cowey. ‘It is an interesting question: it leads me to think of white grape skin colour albeit with minimal skin contact during processing. The color of ripe chardonnay grape skin colours (yellow/gold) produces yellow/gold coloured must and wine, versus Sauvignon Blanc (bright green skin and must). But then Riesling grapes can come in very yellow but produce very green coloured juice and wine.’
For red wines, color indicates a number of things. First off, it tells us a bit about wine pH. More acidic wines with lower pH tend more to the bright cherry red end of the red spectrum, while those with higher pH are more purple/blue. You can see this effect when you add a bit of tap water (pH around 7) to a bit of red wine in the bottom of a glass (pH 3.2-3.8): the colour goes from red to blue. Second, the colour will vary with the level of initial extraction, the presence or absence of oak during ageing, and the level of sulfur dioxide additions (this can have a bleaching effect, resulting in loss of colour). Third, and perhaps most importantly, the color tells you about the oxygen history of the wine: how much oxygen it has encountered during the winemaking process and ageing. Young wines are typically vividly, brightly coloured with fresh red and purple colours dominating. As they see more oxygen the colour shifts more in the orange/brick red spectrum, tending towards brown with excessive oxygen exposure. Compare a vintage Port at 10 years old with a 10 year old Tawny for an extreme example of how oxygen exposure changes the colour of a wine. Bricking of an older wine is first seen at the rim, because this is where the colour is less intense.
It should be pointed out that the grape variety can also influence wine colour. A good example here is Pinot Noir: it lacks acylated anthocyanins, which are a form of anthocyanin that is more stable, and it generally produces lower levels of anthocyanins, resulting in wines that are generally lighter in colour. The use of oak in ageing wines also makes a difference, because barrels can both introduce phenolics to the wine, and also allow for more oxygen exposure during the winemaking process. Thus oak-aged whites are typically deeper in colour than those aged in tank.
Interestingly, cooler vintages tend to produce deeper coloured red wines. In warm vintages, the grapes develop faster and reach ripeness while the anthocyanin production lags slightly behind. Also, stable colour is due to combinations between anthocyanins and polyphenolic compounds that take place during fermentation, and in the warmer vintages the polyphenolic composition will be different (the tannins will be ‘ripe’ and likely less reactive). This is aside from the extraction process involved, which of course can affect wine colour also. It’s worth mentioning here the phenomenon of co-pigmentation, where small amounts of white grapes (typically Viognier, at 1–6%) are added to a red wine fermentation. The colourless phenolic compounds from the white skins bind with anthocyanins and form stable colour complexes, and thus—paradoxically—the red wine with some white grapes added is actually darker in colour than without.
Is brightness or clarity in a wine just attributable to filtration/long settling? ‘Filtration and settling will increase brightness and clarity, and I can’t think of any other reason for wines to be clear, apart from particles being removed or settling out,’ says Liz Waters. ‘With greater filtration and/or settling, you achieve greater clarification as you remove suspended non-soluble material in the wine, and also decrease the turbidity,’ agrees Geoff Cowey. ‘Thus the clarity of the wine will improve, hence the term used in the industry of “cellar bright”. Note that wines can be really low in turbidity, and sterile filtered, and will have great clarity yet still have more of a dull appearance than bright, but this more a factor of the development of the wine.’
So what about tears? How are they created and what can they tell us about wine? Basically, they are formed by the evaporation of alcohol, which leads to what are known technically as Marangoni stresses, leading to climbing, thin films of wine. These films reach a point where they are no longer stable, and so the wine falls back down in a tear-like pattern. It’s hard to explain, but as the alcohol evaporates at a slightly higher rate from the wine in contact with the glass, this draws in the surrounding wine because of a surface tension gradient (the surface tension is less in the higher alcohol area) which then forms the film climbing the side of the glass. Thus the size of the tears is telling you about the concentration of alcohol in the wine, but also the temperature of the glass. If the glass is warmer, the alcohol will evaporate faster in the wine near the edge of the glass, and the tears will be bigger. Whether or not the tears are highly stained simply reflects the density of colour in the wine. ‘Both high extraction and cooler vintages would enhance colour in wine and thus in the tears, but there would be other reasons also,’ says Liz Waters. ‘This question is really about what leads to high colour in wine.’ In conclusion, tears are not that useful in the visual assessment of wine.
It needs to be pointed out here that there is a huge difference between viscosity and the sort of surface tension effects that create tears. Tears don’t have anything to do with the viscosity of the wine (and are sometimes wrongly attributed to the presence of glycerol in the wine), unless, of course, viscosity is in some way correlated with alcoholic strength, in which case this is correlation and not causation.
Finally, one question that has always interested me concerned the size of bubbles and their behaviour in Champagne. I’d always been taught that the finer the bubbles and the more persistent, the better the wine. But are the bubbles a property of the sparkling wine or Champagne, and do they really tell us about quality? I asked sparkling wine expert Dr Tony Jordan, previously head of Domaine Chandon, and now working for his own consultancy Oenotech.
‘Bubble size (which increases as a bubble rises in a glass of Champagne) and shape is affected by the composition of the wine, the nucleation sites present and the concentration of CO2 in the wine. Carbonated water has totally different bubble behaviour which shows that the wine’s composition (alcohol, surfactants, etc.) has an obvious influence, but it is a complex story.’ ‘Some people like to say that the finer the bubble the better the wine or the longer yeast age the wine has had. There may be some truth in the longer ( 10 years +?) cork aged Champagnes/ sparklings (see below) but it is not clear in 2 to 7 years tirage aged wines with moderate cork age and it has little to do with quality (however that is judged). If it was the case that longer yeast aged Prestige Cuvée Champagnes and sparklings had finer bubbles then, for instance, Taittinger would always pour a glass of Comtes de Champagne and show it had a finer bubble than shorter yeast age (NV or Vintage) products. This is usually not observed. It may be finer than some but then not finer than a commercial bubbly made with short yeast age!’ ‘Bubble formation requires the presence of nucleation sites with minute air pockets trapped inside them. The pockets of gas arise because of incomplete wetting of the nucleation sites when the wine is poured into the glass. These are now believed to be hollow cellulose fibres from paper and cloth rather than faults in the glass itself. CO2 diffuses into the pockets of gas until they grow big enough to lift off and interestingly when they do a small pocket of gas is left so another bubble forms from the same site and so on, hence the familiar stream of bubbles from one site. (The nucleation sites can be stuck to the glass or floating in the wine.)’
‘The bubbles at 'lift off' vary in size (10 to 20 micrometres) depending on the nucleation site. As the bubbles rise they continue to grow because CO2 continues to diffuse into them as they rise. Have a look at a rising stream of bubbles and you can easily see they are bigger as they approach the surface, they grow to up to 1mm in diameter if they have risen through about 10cm ( If the nucleation site is high in the glass then they haven't the time to grow as much).’
‘In the sense that an old cork aged (say 10+ years under cork) Champagne or sparkling may have lost pressure due to diffusion of CO2 through and past the cork, the bubbles can be expected to not grow as much as they rise because of the lower concentration of CO2 dissolved in the wine (bubble growth is directly proportional to dissolved CO2 concentration). That is, the bubbles at a given nucleation site is the same size no matter what the wine but by the time the bubbles have travelled to the surface there could be a difference , bigger for the full pressure bottles, lesser for lower pressure bottles.’
Jamie Goode is a London-based wine writer with a scientificbackground. After completing a PhD in plant biology he worked as ascience editor. Smitten by the wine bug, he began consumer-focusedwebsite www.wineanorak.com in 2000. The success of this site led tomore work, and in 2005 he landed a national newspaper column and abook deal. Now he devotes all his energies to wine, and his secondbook, Authentic Wine, was published by University of California Pressin September 2011, authored in conjunction with consultant winemakerSam Harrop.
"Send" was hit before I got a chance to say, "Please and thank you!!!!"
I know your very informative article was written forever ago, but I do have a question that none of the posts below have addressed. I just read your article today. I can totally understand the concept of the nucleation sites for gas/CO2 originating on the sides of the glass due to inconsistencies with the surface of the glass, obviously on a microscopic scale. What I do not understand is how the nucleation site can occur in the middle of the glass in the liquid and sometimes in several places in the middle of the glass! This would not seem to address the "nucleation site"- could you please explain how this would occur?
Great Article and very informative.
There are two aspects that I do miss in the “tearing “ discussion so far and I think it is not enough just focussing on alcohol.
1. The effect of residual sugar: think of a classic Riesling TBA – low alcohol and intense tearing.
2. The fact that pure alcohol doesn’t tear at all! Obviously the solution with other liquids plays an important role.
Thanks for the comments and the great questions. I don't know of any research about whether dry extract will have any effect on tearing. My guess would that it wouldn't be much of a factor. Jake, you are right that high alcohol wines would tend to be more extracted - they are usually made from riper grapes, and both the alcohol and ripeness of the skins would lead to more extraction. From quizzing the various scientists I contacted in researching this piece, I came away with the impression that there are quite a lot of unanswered questions.
William, its been a long time since high school chemistry, but color change is basically at the heart of measuring pH. pH and color change go hand-in-hand. What color the pH changes to depends on the type of pH indicator you are using. Remember doing chemistry projects with the pH strips? The measurement relied on the change of color.
Anthocyanins are a natural indicator, and tend to go red in solutions of lower pH and blue/violet in solutions of higher pH (they actually turn green/yellow once you go past neutral and start getting into the alkaline end of the scale). Wikipedia has some good stuff on this.
In regards to colour in reds, Is there a similar colour-changing effect when higher pH wines (purple/blue) are adjusted lower in the winery?
I have heard some people mention that dry extract in a wine may affect how quickly or slowly tears fall and how thick or thin they may be. Do you know of any research that has been done on the levels of dry extract in a wine and how levels of dry extract affect tearing? It seems that there would be a direct relationship between full bodied wines (wines higher in dry extract) and high alcohol wines stylistically.
My personal belief is that Jamie's conclusion -- that tearing is not useful in analyzing the visual appearance of wine, is slightly overstated. If you have a wine that is high in alcohol, it will show significantly different than a wine with more modest levels. However, his point is well taken since the temperature of the glass, and other factors, can contribute to differences in tearing. I think tearing is most useful when it is at its extremes, or shows extraction of color.
Eye opening! Thanks for the great research Dr. Goode. Would it be fair to assume that a Mosel Riesling @ 8% alcohol would "sheet" rather than "tear" because of a lack of alcohol and therefor less tension or do you conclude that tears tell us nothing at all about % of alcohol in the wine?
Great Read! Thanks, Jamie!