26 January 2011

Color me surprised

[mosaic]So here’s a fun thing. Excerpted from Twitter, but with the graphics, etc. removed. It’s a dialogue between Evan Dawson, a journalist and spare-time (where does he find it?) wine writer from New York, and the auto-estimable James Suckling. Let’s count the evasions and logical fallacies, shall we?

Evan Dawson
If sunlight is best way to view color, why judge color indoors like you do?

James Suckling
Isn't that sort of a dumb question? Tasting outdoors doesn't work1. You know that.

Evan Dawson
Right. But 15% of the wine's score is color, and you admit you judge the color in sub-optimal conditions.

James Suckling
What do you do2? Do you have anything better to do today3...or is it a slow news day?

Evan Dawson
Just honestly curious. If a wine's color is 15% of its score, why judge in conditions that don't let you see it optimally?

James Suckling
May be you use your daylight flashlight when you visit cellars to taste4? Can [I] get one?

Evan Dawson
Ha! That would be great. But perhaps another reminder that assessing color for points is questionable.

Evan Dawson
And yes, I confess that I think it's strange to put so much weight on a wine's color. But your mileage may vary!

James Suckling
But just to be polite and answer your question. I have been tasting for 29 years5. I know how to judge color6.

Evan Dawson
A wine can go from 92 to 89 pretty easily all because you judged the score in artificial/lesser lighting. That matters!

James Suckling
Giving points for color works for me7, UC Davis8 and lots of people9.

So…by my count, that’s four evasions and five logical fallacies (though to be fair, some of the latter are reiterations of the same fallacy). I’d suggest that this is some sort of record, but then I remember that I’ve read/heard/seen political commentary more than once over my lifetime…

Since Mr. Suckling won’t actually answer Mr. Dawson’s (excellent) question about color – his defense of his self-alleged inerrancy, by the way, goes against science in the field – I’m free to opine.

Identifying color is fine, especially if it is any way unusual (opaque pinot noir, young wine not made from nebbiolo but with significant bricking, orange wine) but it is, to me, the least important category of descriptor. Why? Because it is so rarely useful in the note's afterlife. I often mock the fruit-salad tasting notes that writers (including me) tend to fall back on, because I doubt anyone has ever gone into a store and asked for a wine that tastes of “slightly bitter Rainier cherry skins and crisp, lemony acidity,” but even if that’s not true, I’m sure no one has gone in with an electromagnetic frequency range between which they wish to restrict their purchases. “No, sorry, that aglianico is just a bit too magenta for me.” Please.

But scoring color? Especially, as Dawson points out, in variable and sub-optimal conditions? Ludicrous. Of course, conditional variability can be a reason to suspect all components of wine scoring, but I’ve a pledge to myself that – the anti-scoring rant being well-worn territory – I won’t repeat what so many others have said on the issue, and yet here it’s especially damning. Unless the light source is being frequency-controlled across all wines in a peer group, it is impossible for wines tasted in different lighting to be scored for color in any reliable fashion. Especially when the color component forms as significant a portion of a total score as it does in Suckling’s methodology.

Why didn’t Suckling answer Dawson’s question, except with complaint, evasion, and logical fallacies? Because, obviously, he can’t. No one can.

(Yes, yes, I linked the word "science" to a Wikipedia article. I'll do penance in the afterlife. It was the best gateway to the actual science I could find in fifteen seconds of Google-fu.)


Evan Dawson said...

"Unless the light source is being frequency-controlled across all wines in a peer group, it is impossible for wines tasted in different lighting to be scored for color in any reliable fashion."

Absolutely. I have doubts that Mr. Suckling actually scores with a color component, having seen his videos scoring wines in a matter of seconds in wine stores. But he makes the claim, and of course others base score components on color.

Obviously wine is subjective on so many levels, but color seems the most silly and fleeting. I confess I never even considered the most damning part of it - that color changes so much as to be useless in a tasting note. I simply don't understand what "good" color is. Good Brunello is dark? I'd prefer Biondi Santi. And what of wines like Pinot blended with the Leon Millots and Corot Noirs of the world? To say nothing of Mega Purple, on which I am 101 points, in any lighting.

thor iverson said...

I have my color preferences as well, especially as predictors of style (which I employ along with label modernity, importer name, and all the other unopened-bottle indicators that any sensible buyer uses to cull in the absence of better information).

But it's not really necessary to get into preferences to identify the fallacies with which Suckling is working. Color can have significance, as I just claimed. But unless a taster...or, I guess, a viewer...is Geordi La Forge, they're only seeing an interpretation of the color, not the color itself. That's danger #1.

Danger #2 is, as you indicate, the light source, and there are two sources of error. First, the influence of the light source's frequencies on the wine's frequencies. Second, the difference between those influences across different light sources.

People who live with and for color -- graphic designers and such -- tear their hair out at Suckling's argument. I've seen it happen.

Keith Levenberg said...

Can I be Socratic for a minute?

Has Suckling explained *why* he scores for color? Does he consider the color an intrinsically important element of wine appreciation (such that blind people can't enjoy wine as much as sighted people) or does he consider it a proxy for something else? If the latter, what is it a proxy for, and why can't that component be judged directly rather than through a proxy? E.g., if you are scoring for ripeness and concentration, why not just give points for those qualities rather than giving points for a color that tends to connote ripeness and concentration?

Additionally, does he score color for "correctness" (i.e., a color appropriate to the age and grape variety of the wine), or against an objective ideal (e.g., the darker the better)? If the former, how can one do that if the wines are tasted blind? If the latter, does he account for the fact that different grape varieties have fundamentally different pigmenting material? And if it's OK to penalize a red for being lighter than another red, why isn't it OK to penalize a white for not being red? Because white and red grapes make fundamentally different types of wine? Aren't ploussard and grenache fundamentally different types of wine, too?

thor iverson said...

Good luck getting answers, Keith.

Some of the earlier queries he "answers," in a sense, by categorizing wine quality via color in the Twitter stream that led to Evan's questions.

But the rest...

Hank said...

I find the color of the bottle to be very important to my enjoyment of wine. I usually allocate up to 10 points to that, on my 110 point wine scoring system.

Thomas said...

I love it when wine rating is under the microscope and the answers to its questioners are along the lines of: Don't bother me with that; trust my many years of swami-ism.

Thomas said...

Hit the trigger too soon.

I also wanted to say that I've always questioned the 3-point weight given to color on the 20-point scale.

Thomas said...

This is crazy.

Forgive me--been off coffee for some time...

I've always questioned the 3 points given to "Appearance," color being one component of it. But still.

thor iverson said...

Hank: I prefer to judge on bottle weight and opacity. Bonus points if it doesn't fit in or on top of any racking in my cellar and doesn't stack well with other bottles.

Thomas: there's still time for you to learn how to use the internet. ;-)

Thomas said...

Am I on the Internet?

Wow! Who woodathunk?

Just ordered four 13 oz bags of coffee...

thor iverson said...

That wasn't coffee.

Thomas said...

Not only that, but I was wrong about Appearance--it counts for 2 out of 20.

Maybe I've lost a few brain cells after all--to ethanol?

King Krak, I Drink the Wine said...

Thomas, I think with ageing wine critics, the color must count more and more. The Emperor must give it 15 pts by now.