This week I’m continuing to ponder the question of whether, in wine criticism, preferences can be separated from evaluation. Philosopher David Sackris argues that preferences are inherently bound up with evaluation so that wine criticism reflects only a critic’s preferences not an objective evaluation of a wine. I tend to agree with Sackris up to a point but it’s important to separate the good arguments from the bad and I find myself resisting many of his inferences.
Several premises in his argument involve an interpretation of empirical studies of how wine critics evaluate wines.
The first is a study of wine descriptors used in tasting notes carried out by Frederic Brochet, an oenologist at the University of Bordeaux, in 2001. Brochet found that, in tasting notes from mostly professional critics, there is a strong association between certain descriptors and wine quality. For example, Brochet found that non-evaluative terms such as ‘‘earth’’, ‘‘cedar’’, ‘‘tannins’’, ‘‘herbs’’, and ‘‘jammy’’ are associated with non-preferred wines.
This is supposed to demonstrate that critics’ preferences figure in their evaluations. But I don’t quite get the inference. Why does the fact that critics have a distinctive vocabulary for higher vs lower quality wines show us anything about whether preferences can be separated from evaluation?These are wine tasting notes that are supposed to provide an evaluation. Why wouldn’t they include descriptors that indicate their evaluative conclusion?
Perhaps the reasoning is that, if the evaluation is not colored by the critic’s preferences, the descriptive analysis should come first and the evaluation would then be derived from the description. Perhaps, but this example fails to demonstrate the conclusion Sackris seeks. “Earth”, cedar, tannins, herbs, and jammy are generic descriptors that don’t indicate anything very distinctive. Why would these not be appropriate descriptors for a simple wine that lacks distinction?
Brochet also found that descriptions are prototypical based on the type of wine being tasted—critics use specific vocabularies for specific varietals or regions. This is certainly true but how does this show anything about a critic’s preferences distorting her judgment? I suppose the inference is that if we are using a prototype to identify aromas and textures, we can’t be sure we’re tasting what is in the wine rather than what we expect we should taste in the wine.
If I know I am going to be drinking a Sauvignon Blanc, especially if I am around people who are ‘in the know’, then I better be smelling cat’s pee on a gooseberry bush. This is what Parr et al. refer to as perceptual bias: knowledge about wine styles and what they should taste like interfere with the actual tasting process.
Well, only if you want to make a fool of yourself. There are many Sauvignon Blanc that don’t smell like cat pee. Warm climate Sauvignon Blanc usually doesn’t. No doubt expectations can influence what we taste. That is why when tasting professionally it’s crucial to go back and critically re-assess your notes before drawing conclusions. Do all critics do this? Probably not. But Sackris is not arguing that critics too often jump to conclusions. He is arguing that it is impossible to evaluate a wine without allowing our preferences to distort our evaluation.
I don’t see how either of these examples make that case.
He then uses the famous disagreements between Robert Parker and Jancis Robinson on the relative virtues of big, alcoholic wines to show that whether one likes a wine cannot be separated from one’s evaluation.
Jancis Robinson, as well as many winemakers, regret the influence Robert Parker had over wine styles in the 1990’s and 2000’s, which they claim was guided by his preference for mammoth, alcoholic wines. Many winemakers and critics like Robinson preferred wines with more restraint, but the winemakers made wines in Parker’s preferred style to earn higher scores. Sackris argues that If pleasure and personal preference can be separated from aesthetic appreciation, Robinson and these disgruntled wine producers should be able to aesthetically appreciate this style as well as Parker. Thus, Robinson should give evaluations close to Parker’s even though she doesn’t prefer these wines. This of course is not what happened, especially in the often discussed case of the 2003 Chateau Pavie.
This is a telling point. It’s obvious that personal preference drives many of the evaluations that critics such as Parker and Robinson make. However, I think what this shows is that critics and winemakers are advocates for their preferences and their evaluations reflect that. They are selling their point of view. It seems to me this is what we want from a critic—we want to know what she thinks about a wine. That is what satisfies her subscribers, readers, or advertisers.
It doesn’t follow that should they be given the task of accurately describing the wine, they couldn’t do so. In fact the success that Master of Wine candidates have in their blind tasting exams is evidence that tasters can, when it matters, separate preferences from descriptions at least well enough to pass the exam.
Sackris is arguing that whether one likes a wine or not is intrinsically connected to one’s evaluation—they cannot be separated. He is not making the weaker claim that critics often allow preferences to influence their judgement. They surely do, but he needs the stronger claim for his thesis and I don’t think these arguments quite do the work.