So I realize things around here have been slow, but what with the Girlfriend becoming the Wife (Mrs. Drinksnob?), the ensuing festivities, and us both getting pneumonia, I’ve been a little preoccupied.  I’ve been lured out of my lair temporarily (coughing and sniffling) by an appalling mis-use of food science by Ms. Coco Krumme in Slate, a web publication whose point I’ve never quite understood.  Normally wine is not really within what I consider my purview, but food science – especially flavor chemistry – is, and so are rhetorical fallacies like naked appeals to authority.

The article starts pleasingly enough by retreading the old “wine snobs are classists/morons/coating-their-tongues-with-dollars” argument.  Those wine snobs!  Always talking with silly words about that wine!  Good thing we have journalists to prove that, oh my, yes, Bourdieu was right!  Since it’s by-now well known that extrinsic factors do in fact affect the sensory perception of wine (that is, if we know a wine is considered “better” it will actually taste better), this is not news.  But it’s always good for a slow week as a nice bit of non-inflammatory class warfare.

Where Ms. Krumme takes a really wrong turn is in turning to food science to justify her point.

In defense of critics, one might argue that these correlations exist because expensive wines actually taste like focused cassis, where cheap wines are just juicy. But, along with Quandt and Taber, food scientists think that’s doubtful. Recently, researchers have been smashing apart wines in mass spectrometers, looking for odorants like ethyl 2-methylbutanoate, which smells like apples. While they are finding minor differences between varietals, the similarities are more striking. Merlots contain slightly more earthy compounds than cabernets, but the two are otherwise indistinguishable. It’s impossible, furthermore, to pick apart differentiating flavors of specific spices or flavors of earth in any wine. Granted, the human nose is more agile than a mass spectrometer, which only detects the mass and structure of molecules. It’s unlikely, however, that experts have such precise senses that they can identify minute variations of tastes and odors that a sophisticated machine cannot observe at all.

“[F]ood scientists think [it’s] doubtful” that wines might have hugely complex flavors?  Let’s see her citation.  Oh.  One article (linked in the blockquote).  Which, if actually read, proves to contain only quantitation without any flavor impact study.  I’m not criticizing the paper; the paper appears to have been carefully written and based on a totally reasonable experimental design, and I have no reason to doubt its conclusions.  But its conclusions are that the compounds which were identifiable by GC-MS/GC-O in Merlot and Cabernet Sauvignon are similar (but not identical).  This is very, very different from concluding that “it’s impossible, furthermore, to pick apart differentiating flavors of specific spices or flavors of earth in any wine.”  The researchers in fact attempt to do so by assigning odorants to specific flavor categories.  The search for the peppery flavor in Shiraz, as Harold McGee documented years ago, is very real.

Now, I’m sick and tired, and my EndNote library is no longer full of wine research papers (its full of cheese research papers!), so I’m just going to do this the sloppy way.  Here is one way in which one could learn about whether or not wines are instrumentally differentiable by bottle or by varietal.  Here is some of (the probably more relevant) sensory science research on the subject.  You don’t just pick one paper and, I’m going to assume, read only the abstract, because you think it supports your already chosen position.

Which leads me to the second, possibly worse problem with this article.  This is argument by appeal to authority, in which Ms. Krumme shores up her argument not through good citation, but by appealing to capital-S Science.  She throws in ethyl 2-methylbutanoate because it sounds science-y, but it’s the kind of mean bluff middle-school kids play when they talk about sex: “of course I know what ‘blow job’ means – I was seeing if you knew!”  Esters like ethyl 2-methylbutanoate, of course, are abundant in most alcoholic beverages, but are unlikely to be considered particularly important because of their undistinguished fruity flavors and high aroma thresholds.  But she doesn’t know that; it just sounds good and science-y.

This is made even more clear by the end of the paragraph, in which Ms. Krumme demonstrates her complete misunderstanding of the sensitivity differences between mass spectrometers and human noses.  Human noses may or may not be more “agile”, but they are certainly far more sensitive than current mass spectrometers, a fact that anyone attempting to instrumentally quantify, for example, geosmin will attest sorrowfully to.  Nevertheless, these machines are “sophisticated” because of Science, and so are the ultimate arbiter of the point Ms. Krumme wishes to convey.  Good thing for us we’ve got the machines scheduled to work out that whole flavor thing by Tuesday, so that I can go find a more interesting line of work.