3.5 Contemporary Approaches to Wine Evaluation

Numerical ratings from wine critics have overshadowed scientific attempts
to systematise the evaluation of wine, such as Ann Noble's famous 'Flavour Wheel', pictured above.

The Language of Wine.

One of the first attempts to describe wine in English was by a Scottish Doctor, Alexander Henderson (1780-1863), as recorded in his book "The History of Ancient and Modern Wines" 1824. Henderson relates a familiar problem  - how to find words to convey flavours? "In most wines, it is true, certain tastes predominate; and these can be easily discriminated, especially when the wine is new" he writes."Port, for example may be called bitter, rough or astringent; Rhine wines sharp and acidulous; Malmsey [Madeira] bitter-sweet, &c.: but what terms will convey an adequate notion of that peculiar ethereal flavour which distinguishes each of these liquors when duly mellowed by age...? To tell us that it is penetrant, volatile, transient, and so forth, is nothing to the purpose...the only satisfactory and intelligible way in which the description can be given, as has been already observed, is by comparison with some other known sensation of taste, respecting which all men are agreed...In like manner, those properties of wine which we recognize by the organ of smell...are equally difficult of investigation and definition. In general, they bear some analogy to the tastes; but sometimes they are essentially different."(1)

Little has changed since Henderson's time, except perhaps our ability to state the problem more elegantly."We tasters feel to some extent betrayed by language", wrote the contemporary French oenologist, Emile Peynaud, concluding, "It is impossible to describe a wine without simplifying and distorting its image."(2)

A tendency towards simplification is understandable, given how complex wine can be, and frustrated wine writers with a fetish for being trendy have often tried to 'deconstruct' it. Their results sometimes remind one of the very worst Dadaist poetry, whereby a barrage of random words are melded together to create an absurd kind of 'anti-tasting note' :

"...Monolithic, fearless, even rude...Pencil shavings and patchouli compete with Mister Kleen and those socks there piled up in the corner. There's sweetness later on, much flowers and making up, and somewhere mid-swallow there's a perceptible tong sound which rings on for several minutes. Subtle overtones of toothpaste, orange juice, coffee and bacon".

Unfortunately, wine professionals and Master's of Wine have being just as culpable. One visiting English M.W., confronted with an Australian Shiraz, described earthy aromas as"...fresh droppings from a Suffolk lamb on a summer's day in a Sussex lane".No less imaginative is English wine writer, Michael Broadbent, noticing"...the smell of schoolgirls' uniforms" (3) in  a wine. And not to be out done, the late Auberon Waugh, in his wine column for Britain's Tatler, detected the aroma of"...a dead chrysanthemum on the grave of a still-born West Indian baby" (no, he wasn't fired, but he and his editor, Tina Brown, were brought before the Press Council to answer charges of insensitivity)".(4)  

In a seminal study of the subject from 1976, "Wines: Their Sensory Evaluation", Maynard A. Amerine and Edward B. Roessler, sought to relegate the obfuscations of wine critics to a "Romantic" tradition.  As professors at the University of California at Davis, an important center for the study of wine, they wanted to introduce some scientific foundation into the business of tasting. This study built on earlier work, when together in 1959, the pair had developed "The University of California at Davis 20 Point Scale System Organoleptic Evaluation Scoring Guide For Wine" as a method of rating the large number of experimental wines that were being produced at the university. Designed when making defect free wine was of primary concern, the systems intent was to expose flaws in wine rather than identify virtues, and although it is still used, its criteria is now widely considered obsolete: Wines with no faults could generally receive scores in between 17 and 20 points. Nevertheless, the system, (outlined below) is historically significant as one of the first scientific efforts to broach the language of wine, encouraging critics towards a less passive approach.

In the Viticulture and Enology department of the same university, another effort to introduce methodological rigor to wine tasting lexicon was being instigated by sensory scientist and flavour chemist, Ann C. Noble. The result was the development of the now famous "Wine Aroma Wheel". Created in 1990, the wheel  consists of three circles of increasingly specific descriptive words, all based on real-world smells: the most general descriptors such as fruity, vegetative, nutty, caramelized, woody, earthy, chemical, pungent,  floral, and spicy are located in the inner tier.  In the middle tier are 29 more specific categories including citrus, berry, tree fruit, tropical fruit and dried fruit. The outer circle offers 94 still more precise descriptors, allowing for such subtleties as strawberries vs strawberry jam.  Noble maintains a disdain for words like "harmonious,"  "fragrant," and "dense," which she says have no exact meaning in relation to wine.  Not surprisingly, she believes that,"winemakers favor fuzzy terms for a reason...The labels are for marketing an image". (5) (The Australian Wine Research Institute has since created a similar wheel based upon Red Wine Mouthfeel.)(6)

While the pioneering work at U.C. Davis helped to shift the language of wine reviews away from sheer poetic fancy ("this wine reminds me of my first kiss") and vague anthropomorphic similes('masculine' or 'feminine', 'noble' or 'common'), it was not until more recent research into the matching of smell and taste sensations with their assigned chemical compounds, that my own concept of creating a wine rating methodology began to crystalise. As I charted these correspondences (see Appendices: Assigned Chemical Compounds) the rationale of wine and food descriptors based on vocabulary supported by science began to unfold, begging the question:"What is it that we are really smelling and tasting?"

To describe an Australian Shiraz in terms of "liquorice", "plum" and "black pepper" still bemuses many wine lovers. Some interpret it to mean that these substances have actually been infused into the wine, in much the same way as one might infuse botanicals in neutral spirit to create Gin. That we can recognise a 'liquorice' characteristic in a wine at all is only because wine actually contains chemical compounds which correspond to those found in fruits, vegetables and other elements of our environment. If we wanted to be technically correct, we could describe the aroma and flavour of an Australian Shiraz as having notes of benzylbutanoate, phellandrelul dipenlene, followed by a touch of 2-methyl-3-buten-2-ol - that is plum, black pepper and aniseed.  As absurd as this might sound, it illustrates the reality that "...some of the most commonly observed fragrances in wines-toast, butter, vanilla, citrus, apples, cherries, pears, honey, herbs-are there because of volatile organic compounds that were either in the grapes themselves or that seeped into the finished juice."(7)

What to the layman has long appeared to be a rather fanciful use of descriptors, has in fact its basis in science, supporting the approach that most respected wine critics today have in the main adhered to, despite ongoing mockery from a cynical public: That is, to simply describe the specific aromas and tastes as they are experienced with as much brevity and precision as possible. And to their merit, it must be added "Given that most flavour descriptors have been established  [by wine tasters]in ignorance of their molecular grounds, it is astonishing what competent analysts wine tasters have turned out to be."(8)

Sensory Chemistry
Some examples of the chemical compounds common to both food and wine are listed below,
as identified by Dr Leigh Francis, sensory research manager at the Australian Wine Research Institute.
 Aroma / Flavour  Chemical Compound  Wine commonly associated with...
Tropical fruit, especially passionfruit thiols, sulphur containing compounds Sauvignon Blanc
 Cut grass / vegetal  hexanols  Sauvignon Blanc
 Capsicum / Asparagus  isobutyl-methoxypyrazine Cabernet Sauvignon,
particularly unripe examples
 Black pepper  rotundone  cool climate Shiraz
 Raspberry  beta ionone  Cabernet Franc & Pinot Noir
 Blackcurrant  Dimethyl sulfide  Cabernet Sauvignon
 Butter / butterscotch  Diaecetyl
 Chardonnay, especially those which go through malolactic / barrel fermention.
 Lychee / confectionary (Turkish delight)  cis rose oxide
(a monoterpene)
 Citrus fruit / flowers  linalool (a monoterpene)  Riesling
 Vanilla / coconut  oak derived / oak lactone  Oak matured red & white wines
 Kerosene  TDN trimethyl dihydronaphtlalene  found in diesel fuel and experienced
in some old Rieslings

Tasting by Numbers

The linguistic difficulties associated with describing a wine in 'three dimensions' in part explains why a growing number of people - from producers and retailers to wine writers and consumers - are now relying on numerical rating systems to describe wines in conjunction with tasting notes.

Without conducting a detailed survey of all the available judging systems in Australia and abroad, there are several assumptions implicit in any numerical rating scale for wine. The first is that a wine's qualities are quantifiable in the first place.  In theory, this is achieved by reference to a scale, whereby a taster arrives at a figure by quantifying the intensity of a given sensation (i.e. taste, smell etc.)  Scales are calibrated against "neutral" sensations, the most apt being "water", therefore, a red wine with a brilliant purple / crimson colour is judged a more intense sensation than that of a dull brown colour and so rated accordingly. Likewise,  an aroma that resembles a fine perfume, is judged a more intense sensation than a liquid with an almost indiscernible primary note, and a palate that exhibits complexity and mouth filling flavours, is judged more intense than a wine with a dilute flavour profile, and so on. While intensity of sensation is one of the physiological realities which the brain works with,  'intensity' alone is not only a reductive approach but a completely undemocratic way to determine a wine's quality. While Shiraz wines from any new world country might consistently rate 90+ points, varietals like Gamay (i.e.-Beaujolais),  that produce naturally lighter wines simply cannot. The critics themselves have acknowledged this alleged 'glass ceiling' : "When asked which varietals might get scores above 95, Tanzer and Matthews [prominent U.S. critics]listed the same ones: Cabernet, Pinot Noir, Syrah, Chardonnay and Riesling. Matthews added that the best Nebbiolos and Sangioveses from Italy might score that high. Both said Beaujolais could not. And while Sauvignon Blanc is widely popular, Matthews could think of only one Sauvignon Blanc-based wine worthy of 95 points: Chateau Haut-Brion Blanc, which sells for more than $200 a bottle".(9)

It is not surprising then, that wine writers like Hugh Johnson, have argued that the quality of a wine is altogether too subjective to be assigned a numerical rating with such a high degree of implied precision as "89 points". And naturally, as we have noted in previous chapters, sensation and perception are self-evidently subjective. But the question remains, to what extent?  It is not unreasonable to suggest that, as with all sensory experience, a common culture can enable a degree of calibration of perceptual representations to occur, as actually seems to be the case amongst professional wine critics.

Accepting that a wine's qualities are measurable (albeit imprecisely), the second implicit assumption is that its scale anticipates an ideal wine - (i.e. a 100 point score). But what determines the criteria for such a wine?
At the most basic level the criteria must be dictated by human physiology. (See the first section of this study). There is no sense of 'indulgence' in bitter tastes or 'moldy' aromas, rather we have a predilection towards those sensations which enhance life, and we tend to crave them. Preferences exist, but their is an underlying commonality to be found among them. Indeed, what remains is the remarkable degree of consensus that exists amongst wine critics internationally:"...there is near-universal accord about what attributes a top wine should have-appealing aromatics, ripe fruit, good structure, a sense of harmony in the mouth and a long finish."Indeed,"...the critics tend to agree about wines a lot more than they disagree..."(10)

Above: Robert Parker, regarded as the world's most influential wine critic.

The most controversial, and the most influential numerical rating system currently in use is that of American Wine Critic, Robert Parker. His "100-point" system was first devised along with his friend, Victor Morgenroth, in order to counter what Parker believed to be confusing or inflated ratings of other wine writers. The system was introduced to Parker's now famous publication "The Wine Advocate"  in 1978.  At the time, most reviewers did not use a scale, preferring prose that could 'politely mask' a wine's shortcomings, if need be. As imitation is the sincerest form of flattery, the major U.S. wine magazine, Wine Spectator, later adopted Parker's model in 1985 and other reviewers have since followed suit. These scales actually rank wines from 50 to 100 points (not 1-100), on colour and appearance, aroma and bouquet, flavour and finish, and overall quality level or potential. (Australia's James Halliday is an exception, working on a 75-100 point scale).

Several prominent English reviewers refrain from the 100 point system, preferring instead a system out of 20 points (three points are allocated for colour and condition, seven points for nose, and ten points for palate), as in the cases of Jancis Robinson and Clive Coates.  

In preference for one or another system, it could be argued that a score out of three or seven is more difficult to assign than a score out of ten,**  and that small scales allow for only a very narrow spread of criteria. This means limited scope for detail in accounting for a wine's total profile. While these are valid criticisms, they ignore the common fault of all numerical rating systems currently in use: Whether the criteria are spread over 20, 30 or even 100 points is immaterial, if little or no detailed account as to how a score is arrived at is ever given in the first place. A typical breakdown of a 100 point scale from one of America's most prestigious wine review magazines illustrates this:

    * 95-100 = Classic: a great wine
    * 90-94 = Outstanding: a wine of superior character and style
    * 85-89 = Very good: a wine with special qualities
    * 80-84 = Good: a solid, well-made wine
    * 75-79 = Mediocre: a drinkable wine that may have minor flaws
    * 50-74 = Not recommended

Parker is no more illuminating: for him, an 80-89 point wine is"a barely above average to very good wine displaying various degrees of finesse and flavour."(11)Yet this is so vague an explanation that it could easily be describing most of the wines in the world. Exactly what criteria make for a 95 point wine as against a 65 point wine? Even more contentiously, why does a given wine score 89 points and not 90? Was it the colour, the aroma or the palate that failed to impress, and in what respect?  

When such questions arise, they usually remain unanswered, and any perception of 'scientific rigor' suddenly disappears. We are left to conclude that the rating is arrived at 'intuitively': The taster is experienced and exposed to many different wine styles, varying in their method of cultivation, vinification, maturation, age and origin, so over time, assigning a mark to a wine becomes 'second nature'. Publications like Wine Spectator would no doubt defend this approach, maintaining "... great efforts to ensure consistency from its critics. Wines are tasted blind, but each tasting includes a non-blind, previously scored wine to set the scale. Also, the magazine gathers its critics once a year to taste wines together and calibrate their individual scales".(12)

While intuitively established ratings do not necessarily lack value, the are generally considered without any concern for whether the results would be replicated the next day and are largely impervious to academic, professional or consumer scrutiny with regard to how the numbers where arrived at. Tim Atkin, taking this very issue to task in his article 'Tasting by Numbers' "...challenged [a professional wine critic] to blind taste 30 wines he had already scored in his book to see if he could replicate the numbers.[Atkin] issued a similar challenge to the well known American taster Steve Tanzer at a wine fair in Washington a few years ago. Neither took me up on the offer"he says. He concluded that the reason was that"They[the critics]know as well as you and I do that scores, like wines, vary from day to day and taster to taster. Today's 87 is tomorrow's 91."(13)  David Shaw of the Los Angeles Times tried to test Robert Parker's consistency, which Parker touts as his greatest asset, by asking him to do a tasting of six wines twice over consecutive days."Tellingly, Parker refused, saying, "I've got nothing to gain and everything to lose". (14)

The 100 point scale is now so firmly entrenched in the most influential wine markets around the world that it is demanded by consumers everywhere wanting to reduce their risk of purchase. Quite simply, whatever an individual's wine knowledge, they understand immediately that a 95 point rating is a good thing and a 75 not so good "almost viscerally".The wider impact that the popularization of these systems has had on international wine markets, wine prices and styles is seen as twofold:
A positive contribution of point systems has been to jettison the elitist wine jargon of times past,  and destroy the 'feudal system' which continually rewarded only those vineyards on the most prestigious sites - both factors that could too easily veil quality from the consumer. In this respect, Parker as pioneer and consummate consumer advocate, "...earns his legitimacy at least in part by being an unimpeachable champion of the proletariat drinker, a poor pleb who knows himself too often beaten into submission by the grand tradition and price of wine, but who is also keenly aware of his own limitations in that vast world".Consequently, "Parker's wine credentials turn out to be no less moral than oenological".(15)Small and large vineyards around the world now occupy a meritocracy where equal opportunities for high ratings exist, so encouraging indifferent winemakers to innovate and improve overall quality.

Detractors of the 100 point system say that it limits the spectrum of wines that sell well, and that winemakers have become sycophantic servants, fashioning wines only to cater for the palates of high profile critics in order to achieve big scores and quick sales. This is epitomised by a consultancy service in the U.S., 'Enologix', which uses chemical analysis to"assist winemakers in...boosting average national critics' scores". (16)Inevitably, this is leading to a homogenization of wine, in which terroir or regional distinctiveness is becoming increasingly irrelevant. Many in the French Wine Industry already consider Parker to be the embodiment of globalization, one reason why he is reportedly now unwelcome in Burgundy.

An overview of current approaches to wine evaluation serves to illustrate the controversy surrounding efforts to quantify wine quality. Clearly, any serious methodology must take all the issues into consideration. Accepting 100 points as the International standard, we must also realise the potential complexity of wine, and make no apologies if our rating systems reflect this reality. The challenge remains to create a comprehensive and transparent way of both representing and 'rating' the qualities of a wine, that is reproducible and accountable, and based upon scientific vocabulary, yet one which also allows mere mortals to participate and analyse the positive and negative qualities of a wine with confidence. 
My response to this rather daunting challenge was to develop the 'Wine Spider' system, which is the subject of the next chapter.

The Australian Show Judging System

In the 1998 Royal Melbourne Wine Show Catalogue of Results, the "Notes on Wine Judging" are worth scrutinizing, because they reflect the basis on which some awards are given out. Exhibits in all Australian Wine Shows are judged on a comparative basis, i.e.. all the entries are judged at the same time.  The other method used in Varietal classes in Adelaide is the International system where the wines are judged and pointed individually. Exhibits are judged on a point score with a maximum of 20 points.  On the judging sheet this is divided into:

3 points for colour and condition; usually wines with poor colour are rejected for defect on both nose and palate.
7 points for nose.
10 points for palate: a wine with an excellent nose may fall down on the palate due to excessive acidity or tannin defects.  Conversely, some wines with only a fair nose may have excellent balance on the palate.

Medals are awarded for the following points:

18.5-20 - Gold Medal
17.0-18.4 - Silver Medal
15.5-16.9 - Bronze Medal

All exhibits are judged "blind".  Judges do not handle the bottles. In Melbourne the judges are arranged into four groups, each with three judges, with a Chairman of judges who arranges the group of judges, divides the classes to be judged between four groups, gives guidance and occasionally resolves disagreements. Each class is set up and poured by associates for  the wine judges. The judges assess the wines, point them, and then collaborate on the medals awarded.   The points listed in the results section represented the total of the three judges assessments. Judges are generally selected from those who work in or are connected with the wine industry.  Potential judges are recommended to the Royal Agricultural Society of Victoria and are given experience as associate judges.  New judges are selected from the best performers.

The Davis Scoring System

The Davis system is quite straightforward. It assigns a certain number of points to each of ten categories which are then totaled to obtain the overall rating score for a given wine.
17 - 20 Wines of outstanding characteristics having no defects
13 - 16 Standard wines with neither outstanding character or defect
9 - 12 Wines of commercial acceptability with noticeable defects
5 - 8 Wines below commercial acceptability
1 - 5 Completely spoiled wines

Appearance (2 points) - The wine is given 2 points if it is brilliant, with no dullness, murkiness, or particles of sediment. If it is clear, but not flashing with light reflections, it rates 1 point. If full or cloudy it gets 0.

Color (2 points) - Acceptable colors for white wines are varying shades of yellow, gold, straw. Flaws are any amber tones, indicating oxidation. A rosé can be a true pink, or by reason of its grape source, tinged with deeper red or orange. Overly violet tints, brown tints of amber or deep red are faults. The color of red wines depends greatly upon the grape variety. Pinot Noir may be light enough to verge on transparency. Cabernet or Zinfandel will be deep red. Newer or younger wines will often have blue-purple edges, as older wines will show bronze edges.

Aroma and Bouquet (4 points). Aroma is the sensory impression arising from the mouth with the assistance of the sinus chimneys. Bouquet is fragrance detected by the nose. Either may be vinous, smelling like wine, but without any grape variety characteristics. The intensity may be light, medium, or high. Negative factors are odors that can be described as alcoholic, excessively woody, moldy or corked.

Volatile Acidity (2 points) - This is the term for Vinegary-ascensence. Does the wine smell of vinegar? If not, it rates 2 points. A slight vinegar smell will rate 1 point. If it smells of vinegar it rates 0.

Total Acidity (2 points) - Felt in the mouth, around the edges of the tongue, it is wine's refreshing zing. If low, the wine is flat, flabby or soapy. It can be too high with unpleasant sharpness.

Sweetness/sugar (1 point) - Sugar and total acidity go together. Overly set for the wine's type is a fault, as is overly dry.

Body (1 point) - This is the wine's viscous nature identified as mouth-feel, also the binges, or alcoholic strength.

Flavor (1 points) - The flavor should correspond with the bouquet and aroma, being clean, fruity, full or balanced. It should not be metallic, steamy, or alien in character.

Astringency (2 points) - Tannins give a wine astringency (or bitterness), and so does the wood in which wine is aged. Younger wines will be rougher than older wines. The ideal is mellow softness, velvet, roundness. A young wine is not discounted for a natural tannin.

General Quality (2 points) - The only category for subjective appraisal, adjusting the score on the basis of the wine's total performance.


Bibliography & Footnotes

1. The History of Ancient and Modern Wines; title page, pp.134-136.
London, 1824. Baldwin, Cradock & Joy. "This is probably the first book in the English language to give anything like an accurate account, based on some personal traveling, of the wines of Europe and also of Persia and the Cape of Good Hope". See http://www.winelit.slsa.sa.gov.au/topdozen.htm for more.
2. The Taste of Wine: The Art and Science of Wine Appreciation. By Emile Peynaud, Jacques Blouin Translated by Michael Schuster, M. Broadbent. Published 1996. John Wiley and Sons.
3. Cherries, Berries, Asphalt, and Jam. Why wine writers talk that way.
By Mike Steinberger Posted Friday, June 15, 2007, Slate Magazine.
4. Ibid.
5. Jabberwiney: A professor of viticulture puts a stopper on wine hooey. By Jon Cohen. Posted Thursday, Jan. 20, 2000 Slate Magazine.
6. Both Wheels can be purchased at their respective websites:
7. The World  of Fine Wine Magazine. Alex Hunt, in an article titled "The Foundations of Flavour," (2006)
8. Ibid.
9. Are ratings pointless? The highs -- and lows -- of the 100-point scale. W. Blake Gray, The San Fransciso Chronicle. Friday, June 15, 2007
10. Do You Taste What I Taste, The Physiology of a Wine Critic, Mike Steinberger, Slate Magazine June 2007
11. www.erobertparker.com
12. Are ratings pointless? The highs -- and lows -- of the 100-point scale. W. Blake Gray, The San Fransciso Chronicle. Friday, June 15, 2007
13.  Tasting by Numbers. By Tim Atkin MW, The World of Fine Wine Magazine, Issue 7, 2005.
14. The Great & Powerful Shnoz.  Does the Emperor of Wine wear any clothes? Mike Steinberger, June 17, 2002. Slate Magazine.
15. The Philosophy of Wine Tasting. Christopher Shields. The Times Literary Supplement, November 28, 2007. A review of "The Question of Taste. The Philosophy of Wine", by Barry C.Smith. Signal Book Ltd.
16. The Chemistry of 90+ Wine, by David Darlington.  August 7, 2005, The New York Times.

** The human being has two hands of five fingers, a total of ten units, and more often than not learns to count using them. It's the basis of our digital numerical system.