How do we translate what we taste in the glass to what we write on the page? Jamie Goode delves deeply into this most complex of processes
Recently I re-read a notebook of mine from 1993, when I first tried writing tasting notes. These stuttering notes were laughable: I could barely muster a sentence. While readers may debate the relative merits of my current notes, one thing is clear: the fact that I can now fill the better part of a page with a description of almost any half-decent wine indicates the development of my descriptive language.
We probably all go through this sort of wine vocabulary acquisition process in our journeys from novices to professional drinkers, but how does it happen?
The language of wine is an important but rarely addressed subject. In the trade, there's a lot of talk about wine, whether it's merchants describing their wares, critics advising consumers, or WSET students swotting for their practical exams.
What outsiders find surprising is the richness of the language employed to describe the sometimes imprecise sensory experience of wine.
In particular, philosophers find it fascinating that we routinely share our inner sensory worlds – our private experiences – with each other in verbal form.
Indeed, to my knowledge there are currently three books on wine and philosophy at various stages of preparation. Two are multi-author works: the first is being compiled by Professor Barry Smith of Birkbeck College, University of London (who has authored an excellent entry on wine and philosophy in the forthcoming third edition of the Oxford Companion to Wine) and the second by Dr Fritz Allhoff of Western Michigan University.
The third is a single-author book on aesthetics by Norwegian philosopher Professor Ole Martin Skilleås of the University of Bergen.
This article draws together some different threads.
It compares two contrasting professional uses of wine language, sensory analysis and criticism, and looks at two academic studies that have tried to analyse the way language is used in tasting notes.
Finally, it looks at the nature of the tasting itself: the brain's perception of wine.
Along the way, it will address some core questions:
• In writing tasting notes we are attempting to capture in words our perception of the wine: what is the most appropriate strategy for doing this?
• How much does language itself shape our perception of wine?
• Does our ability to perceive certain aspects of wine depend to an extent on the available terms we have to describe those aspects?
• Are we more able to focus our attention on those particular aspects of wine for which we have appropriate descriptors?
• How much of the language we use to describe wine is 'realistic' and how much is a learned, self-referential code?
Sensory analysis
UC Davis researcher Dr Ann Noble is one of the leading practitioners of wine sensory analysis. Most readers will be familiar with her wine aroma wheel: a circular graphic containing a set of wine descriptors arranged by family.
'The concept of the aroma wheel was initiated in response to several things,' she recalls. 'In analytical sensory evaluation we objectively and quantitatively describe flavour using descriptive analysis. For this, terms are selected which describe the differences among the wines. Aroma standards are provided, judges are trained, and then the intensity of the terms is rated. These descriptive words must be specific, precise words which are not subjective.'
From her work in flavour chemistry, Noble was aware of flavour wheels that scientists in the UK had developed for beer and whisky, and decided it would be useful to do the same thing for wine.
'I started with terms we had compiled from several descriptive analysis studies, and then I included words on existing lists of descriptive words for wine,' she says.
'Then a group of winemakers and sensory people met to select terms which we thought were necessary'.
Noble then sent the list out to the wine industry and collected responses (few people replied). The wheel was finally published in 1984, with subsequent modifications in 1987 and 1990.
The terms given in the wheel present a useful set of wine words; in conjunction with aroma standards, this sort of tool can help people build a descriptive vocabulary that has some basis in reality and which is consistent across individuals.
Dr Leigh Francis from the Australian Wine Research Institute is another expert on sensory analysis. 'We run sensory studies very regularly with groups of experienced but not necessarily always “expert” judges.
We invariably start with a development of a sensory language relevant to the sample set being studied that is meaningful to the panel of assessors,' Francis explains.
'The first session involves individuals generating descriptors for each of the sensory aspects (appearance, aroma) of the samples presented. Experienced judges can easily list terms to describe sensations but there is wide variation in the terms used. We follow a method where the panel leader is not teaching the judges but simply collates the terms and encourages discussion about the words generated. The terms are listed on a white board and for one wine there might be as many as 60 terms generated just for aroma.
'In subsequent discussion people are usually able to agree quickly that different words generated, such as 'bubblegum', 'artificial fruit', 'banana lolly', 'confectionery' and 'estery' might represent the same sensation and can happily agree on one or two terms to represent a large group of words generated by the panel (such as estery/confectionery).'
How much variation do individuals show in their vocabulary for wine? 'The process always emphasises that there is large 'natural variation' in terms used by individuals: terms that are meaningful to them based on their previous experiences and memories, using everyday language,' says Francis.
'Someone who has chewed bubblegum might be inclined to use 'bubblegum' or 'candy' as a term, while someone who has worked in a lab who knows the chemical class of esters might use this as a descriptor. Both sense the same aroma character but describe it differently. However, they can agree on one word that both are happy with, especially if a physical reference standard is available to help in this process (that is, some types of bubblegum or actual chemical compounds).'
What is the source of this variation? 'One source is physiology: some people will be very sensitive to some aroma compounds but not others and this will obviously affect perception and thus descriptors.
A person who is sensitive to the buttery compound diacetyl will describe a wine as “buttery” or “butterscotch”, while another person who is relatively insensitive to this compound, but sensitive to others, will describe the wine differently,' says Noble.
She adds: 'People experience things differently, but barring a specific deficit such as anosmia (inability to smell) or sensitivity to PROP [propylthiouracil, a bitter-tasting compound for which people fall genetically into three different sensitivity groups], people can perform sensory tests very similarly, sensitively and reproducibly.'
Here we come to an interesting, potentially important, question. Does having labels for sensory perceptions, such as the terms in Noble's wine aroma wheel, shape those perceptions in any way? In other words, does our language shape our sensory experience?
'All information affects what we perceive,' says Noble. 'Label a bottle with the name of a cheap wine and another bottle of the same wine with the name of a prestigious winery and people will use negative terms for the cheap wine and positive terms for the second wine.' This is called the error of expectation.
'There is a whole field of psychology studying this, and marketing people know lots about it as well,' adds Noble. 'One person may describe the aroma of wine as “floral, mango and pineapple”, whereas another sees it as “fruity” but may put different specific descriptors on the aroma notes. [This is] because of their different interpretations of each note (which we standardise by training with physical references) and because very low-intensity notes can be described equally validly by two similar terms since, at that low level, there is probably no way to pinpoint the note as one or other. You cannot ignore the power of suggestion here either, since perhaps the note could be perceived as pineapple when you go back looking for it.'
'I'm not sure if I can answer this question,' admits Leigh Francis, suggesting that it borders on philosophy of perception. 'The words used to describe sensations can be more precise or less precise and, depending on sensory experience and repeated practice of describing sensations with feedback from others, then the sensory experience can broaden or rather become more specific.'
'By being exposed to wines with a fruity type of aroma and initially describing them as fruity but, with further experience denoting them as apricot or citrus fruit character rather than the broader category, we're able to describe our perceptions better. That's part of the process of descriptive analysis with a panel: descriptors are generated and a term that might not have occurred to a someone can be accepted as an appropriate descriptor. That person will then use that term when they hadn't previously.'
Francis gives exposure to off-flavours, taints or faults as an example of this. 'Tasting a wine with oxidised flavour might result in someone deciding they don't like that wine but not being able to say what the flavour reminds them of. But after going through a process of generating terms such as “bruised apple” or “Sherry-like” or “wet cardboard”, the person can learn to describe this character and recognise it in subsequent wines.
'If we weren't thinking so much about aromas and flavours, and if we weren't concentrating hard on what we're perceiving and putting the experience into words, then we'd either be very broad about the sensations (“this apple tastes good”), somewhat broad (“this apple is quite crisp, sweet and acidic with a good apple flavour”), or detailed (“this apple has a fresh, estery flavour with a pear-like character, persistent sourness and some dryness”).
It's probably a combination of wide sensory experience, practice in using language to describe perceptions, and desire to concentrate on individual sensations that result in different sensory experience, rather than words limiting or shaping the experience.
'In research studies we'd always give the assessors the option of rating the intensity of an “other” term in addition to an agreed list. A term is really just a label and what precise term is used is not so important: lists of terms such as the wine aroma wheel or mouthfeel wheel are useful but shouldn't be considered as prescriptive. It can be misleading to, in any way, force people to use particular terms someone else has generated.'
Are some means of wine description more legitimate than others? 'This totally depends on the purpose of the tasting,' says Noble.
'If it is an analytical tasting to compare the sensory properties of wines from grapes aged in different oak barrels, then only by properly training the judges or randomising the samples is it valid. You don't ask which wine the tasters like better; if you just want an informal description then anything goes. But if a person perceived as knowledgeable says “this is a masculine heroic, full-bodied wine with stunning structure”, then no one else really knows what the hell they mean except that it must be an intensely flavoured wine with a lot of tannin.'
Noble adds: 'If you are trying to find out which wines people like, you must use typical consumers who drink wine of this type.'
Sensory analysis is in some ways the 'purest' form of wine description: it aims to identify the different wine characteristics in an analytical way, focusing on what is actually in the glass. But it is also highly reductionistic. It attempts to break down wine into its component parts.
In this sense, it is useful for its intended purposes, but it could be argued that any description of wine that is based on isolating and identifying components is of relatively little use in actually describing what a wine is 'like' to drink.
If I were to describe a randomly chosen acquaintance of mine in terms of an assemblage of reductionist measures, such as his or her chemical composition, dimensions, or even their intellectual attainments, you wouldn't have much of an idea of what they are actually like as a person to relate to. Just as with people, the properties of wines that most interest drinkers emerge from the combination of their elements into a whole, which cannot be predicted from descriptions of the components, although these may indeed be useful clues. This is where we turn to the second main use of language in wine description: that employed by the critic.
The language of critics
Let's begin this section by looking at a few examples of the use of language by some well-known wine critics.
'Dark fruits and oak spices can be discerned in the aromatics of the 1998 Clos de Vougeot Le Grand Maupertui. Medium- to full-bodied, dense and concentrated, it is a fat, fruit-dominated wine. Blackberries, cherries and loads of spices cover up this wine's structured character… Smoked bacon, juniper berries, candle wax and cherries can be found in the nose of the medium to dark ruby-coloured 1999 Clos de Vougeot Le Grand Maupertui. This full-bodied, chewy-textured, highly-extracted wine is crammed with toast, blackberry syrup, Asian spices and bacon. It is intense, plush and delineated and has outstanding depth.'
Pierre Rovani, talking about the wines of Anne Gros in Parker's Wine Buyer's Guide, sixth edition.
'More reserved on the nose than most of the Ambroise wines, but also with exciting glimpses of what is hidden under the cloak. A slightly roasted, nutty scent, but not in the least dried out or too warm. On the palate there is even a coolness to the fruit, the elegance and freshness of which is even better preserved than in the Clos Vougeot. But there is at once more breadth, depth, richness and sustain. Harmonious all the way through the long finish. The sheer scale is slightly daunting, but the phenolic phalanx is marching in step, with its spears pointing in the right way, protecting rather than threatening the fruit. Regal wine, at the beginning of a long reign.'
Neil Beckett's tasting note on Bertrand Amboise Corton-Le Rognet Grand Cru 2003 in issue 5 of The World of Fine Wine
'A total whore of a wine (I mean that in a good sense) is the 2002 Grenache Old Vine. Made from 69-year-old vines planted in sandy soils, aged in a combination of new and old French and American wood, and cropped at 1.5 tons of fruit per acre, the vintage's cool drought conditions have resulted in a magnificent wine that represents the nectar of Grenache. Dense ruby/purple-colored with spectacular concentration in addition to sweet blackberry, kirsch, raspberry, pepper, and flower characteristics, this magnificently concentrated, full-bodied 2002 takes Barossa Grenache to new heights.'
Robert Parker commenting on a wine from new Barossa superstar Troy Kalleske
'Clear fresh cherry red. Nose tight but sweet and ripe. Lean and nutty, then more almond-like, growing in potency and finishing long.'
Hugh Johnson on Domaine Amiot-Servelle's 2000 Chambolle-Musigny Les Amoureuses in issue 3 of The World of Fine Wine
'A dark glowing colour followed by a thick, warm burr of blackberry fruits, like the embrace of a tropical night. The palate is dominated by plunging exotic blackberry richness, thick and almost pippy, its exuberance and sheer presence a testament to the solar wealth of 2003. This ample fruit is held impressively in balance as the palate unfolds, and nothing obtrudes or disconcerts; it all tapers away to a lively, breath-freshening finish. Magnificent, enduring wine.'
Andrew Jefford on Ligier-Belair's 2003 La Romanée in issue 5 of The World of Fine Wine
Just a few examples, but enough to give an impression of the rather different strategies that critics use to communicate their perceptions of wine in verbal form. One thing is immediately apparent: there's a big difference between the descriptive strategies used in the related professional activities of wine sensory analysis and wine criticism.
During a recent meeting with Elin McCoy, Robert Parker's biographer, we discussed Parker's use of language, and its power in influencing both consumers and the trade. 'Parker's language has become a “frame” for wine,' asserts McCoy. 'Language is a frame through which we judge something; we reframe something by using different language,' she explains. Her view is that the way people talk about wine influences how they then approach it. 'Parker has changed the paradigm for how people look at a wine, and they are influenced by that language.'
She recalled being on a panel at a Pinot Noir conference in New Zealand with Michael Broadbent, Pierre Rovani, Bob Campbell MW and Michel Bettane, discussing how best to evaluate Pinot Noir.
People used to associating the terms 'subtle', 'finesse', elegant' and 'sappy' with Pinot Noir (in a positive way) will be looking for different things when they approach a wine to those who have grown up on descriptors such as 'rich', 'concentrated', 'dark', 'sweet' and 'powerful'. 'People aren't always aware that the frame is in their heads,' says McCoy. 'I think it is very important: it is being aware of the place that we are judging wines from.'
This takes us back to the question: is the language we use somehow influencing the nature of our perception? It could be that as we learn more about wine and progress from being novices to tasters with a broad vocabulary for wine, the sorts of words we carry in our minds will lead us to look for those aspects or qualities in the wines we taste.
They may also help us to recognise aspects of wine that we wouldn't have spotted without a tasting term to act as a peg on which to hang the sensation.
This brings us to the studies of cognitive psychologist Frédéric Brochet. He has studied the practice of wine tasting as carried out by professionals.
'Tasting is representing,' says Brochet, 'and when the brain carries out a “knowledge” or “understanding” task, it manipulates representations.' In this context a 'representation' is a conscious experience constructed by the mind on the basis of a physical experience, in this case the taste, smell, sight and mouthfeel of a wine.
Brochet uses three methodologies in his work, one of which is textual analysis: looking at the sorts of words that tasters use to verbalise their representations in a statistically rigorous way. Brochet used five data sets, consisting of tasting notes from Guide Hachette, Robert Parker, Jacques Dupont, Brochet himself and notes on eight wines from 44 professionals, collected at Vinexpo. Employing textual analysis software called ALCESTE, Brochet studied the way the different tasters used words to describe their tasting experiences.
He summarises his six key results as follows.
1 The authors' descriptive representations are based on the types of wines and not on the different parts of the tasting.
2 The representations are 'prototypical': that is, specific vocabularies are used to describe types of wines, and each vocabulary represents a type of wine. To put it another way, when you come to taste a wine, the first thing you do is (unconsciously) decide what sort of wine it is, which leads you to describe it in a certain way, using the terms you possess for this type of wine.
3 The range of words used is different for each author.
4 Tasters possess a specific vocabulary for preferred and non-preferred wines. No taster seems to be able to put aside their preferences when their representations are described. Brochet adds that this result is well known from the fragrance world.
5 Colour is a major factor in organising the classes of descriptive term used by the tasters, and has a major influence on the sorts of descriptor used.
6 Cultural information is present in the sensorial descriptions.
There's another question we need to address here. Which linguistic tools are appropriate or permissible in descriptions of wine? Some argue that we should go no further than simile (it 'tastes like'); others see metaphor as a crucial tool in our endeavour to share what we are experiencing.
Should we be attempting to describe a wine as plainly and accurately as possible, breaking it down in a reductionistic manner into its constituent flavours and aromas (as in sensory analysis), or do we use more figurative and creative language to build a more holistic description?
A fascinating academic project is underway at the Department of Modern Languages at the University of Castilla-La Mancha in Spain, involving Drs Ernesto Suarez-Toste, Rosario Caballero and Raquel Segovia, called 'Translating the Senses: figurative language in wine discourse'.
The initial stage of the project involves collecting a data set consisting of 12,000 tasting notes from a range of publications (Wine Advocate, Wine Spectator, Wine Enthusiast, Wine News, Decanter and wineanorak.com). The text is cut, pasted and cleared of all extra information. The types of metaphor used are tagged, then a concordance tracks each instance of any metaphor type of interest.
'Wine folk use metaphor all the time,' says Suarez-Toste. 'Aroma wheels are OK for identifying aromas, but the structure and mouthfeel almost always demand the use of figurative language. It has nothing to do with waxing poetical about something sublime; we use metaphor even for the most average grape juice around,' he states.
'For one thing we personify wine most of the time. Not simply by saying it has a nose instead of a smell. It has character, it's endowed with human virtues and vices. It can be generous, sexy, voluptuous, whimsical, shy, demure, bold or aggressive. We almost cannot conceive wine without personifying it.'
We reach for metaphors because of the impoverished language we have for describing tastes and smells. 'Because there is no single lexicon with the expressive potential to cover all the range of sensorial impressions, the intellectualisation of sensorial experience is inextricably linked to the figurative uses of language,' explains Suarez-Toste.
'There is no problem with this as far as such areas of human life as poetry are concerned, but the inherent subjectivity of sensorial experience represents innumerable difficulties when technical discourse is under scrutiny.'
What about the good old tasting note?
'This relies heavily on a combination of terms articulating the remembrance of the taster's repository of aromas and flavours, connotations and, above all, figurative language which, although it may be perceived by the layman as deliberate obscurity, is a valuable tool that allows the (only partially satisfactory) communication of the experience of tasting wine. The vocabulary used points to various figurative phenomena, all of which are indispensable tools for articulating what is an intrinsically sensorial experience.'
So we have wine as a living creature, wine as a piece of cloth and wine as a building. It's easy to make fun of this sort of description, but metaphors are borne of necessity. While we'd like to have a more exact way of sharing our experience of wine in words, such precision doesn't exist, and those who restrict themselves merely to naming aromas and flavours miss out on some of the more important aspects of the character of wines that can't be described in this way, such as texture, structure, balance and elegance.
'Currently we're obsessed with structure and mouthfeel,' Suarez-Toste explains.
'These usually demand architecture and textile metaphors. One curiosity that our audiences enjoy is that a wine can be described in the same tasting note as “silky” and “velvety”.
'Of course for them, the terms are mutually exclusive. The idea is that both are different (but almost synonymous for the critic's purpose) realisations of a textile metaphor. The connotations are smooth and expensive, fresher in silk (more used for whites) and warmer in velvet (more frequent in reds), but essentially the same. And that's just the beginning. Lists of materials are boring when compared to words that unconsciously betray the textile metaphor: this wine is seamless, is bursting at the seams, the fruit is cloaked by tannin, wears its alcohol well, a core of tannin is wrapped in layers of fruit.'
The perceptual event: wine on the brain
Finally, let's look at the way the brain processes wine. In sensory analysis we use our senses of taste and smell as a measuring device, much like a scientific instrument which objectively assesses the liquid in the glass. Without dismissing the importance of this sensory work, I'd argue that, rather than acting as a passive measuring device, in tasting wine our chemical senses are active participants in the process.
Putting this another way, the experience of wine is not a property of the wine itself, but more the interaction between taster and wine – we bring something to the tasting experience that helps shapes the experience.
Most people in the wine trade have a rather simplistic view of 'tasting' that takes into account input from just taste, smell and touch receptors in the mouth and nasal cavity.
These receptors turn chemical and physical information into electrical signals that are processed by the brain. The simplistic view sees a linear pathway from the detection of elements of wine by the taste receptors in the tongue and the olfactory receptors in the nasal cavity to the mental representation of the wine's properties. This is wrong, and a much more nuanced and complicated view of the process is called for.
Rather than act as a straightforward measuring instrument, the brain 'models' the world around us. Our sensory systems are bombarded by a mass of information, which, if attended to uniformly, would swamp our perceptive and decision-making processes.
So the brain is able to extract from this sea of sensory data just those features that are most relevant to us at that moment. This is done by a procedure known as higher-order processing. We may think that our sensory system is revealing the world around us in an accurate and complete way, whereas what we actually experience is an edited version of reality based on the information that is most relevant to our survival and functioning.
For almost all purposes it does no harm to think of the world around as revealed to us as 'reality' – indeed, life would become quite complicated if we operated any other way. But for the purposes of this discussion, it's useful to realise that the version of reality we experience is an edited and partial one.
We are constantly bombarded with chemical stimuli and the brain filters this information so that only the important bits get through. It seems that much of the brain is dedicated to producing an edited view of reality, just as the staff in a newsroom work hard sifting through the journalists' output to produce a 15-minute news bulletin for broadcast.
The fact that we model the world around us rather than perceive an exact representation of it has relevance to wine appreciation.
Let's return to Robert Parker: if he rates a wine at 94/100, the tendency among wine lovers is to regard this score as a property of the wine ('this is a 94-point wine').
Yet it is more accurate to think of the score as a property of an interaction between Parker and the wine: what he is rating is his perceptual experience, which depends in part on him and only in part on the properties of the wine itself. Critics are reporting on their interaction with the wine by rating the conscious representation of that wine in their brain.
Where does this leave the language of wine? It suggests (and I'd not want to use a stronger term) that there's good basis for championing the use of figurative language in wine descriptions.
The nature of perception, and limits of reductionistic approaches to wine description mean it may be the most effective way of communicating what we experience. It also hints that we should exercise a degree of humility as we write down our impressions of the wines we experience.
So, yes, there is still room for poetry, magic and mystery in wine appreciation.
HAVE YOUR SAY
Click on the button to include your comments on this story
Want more stories delivered to your inbox?
