Is Wine-Tasting Junk Science?

http://www.npr.org/blogs/13.7/2014/08/08/338839598/is-wine-tasting-junk-science

by 1337biz • 12 years ago

53 46 12 years ago

npr.org

tormeh• 12 years ago

Spoiler: Yes.

swombat• 12 years ago

Article: can npr headlines contradict Betteridge's law of headlines? Can this very headline do it?

u124556• 12 years ago

The answer is no. Not because it's not junk, but because it's not science.

raverbashing• 12 years ago

Blah blah blah, drink your wine and enjoy it.

If you like it great.

This doesn't mean all aspects are purely taste, like: acidity, tanninity, type of grape, etc

jnazario• 12 years ago

having come to that realization myself, that's more or less what i tell people who are new to wine. but i added something: learn how to describe what you like (or don't like) so that you can find new wines to try and are likely to enjoy. figure out how to communicate, say, a certain flavor with the term "tannin" or "oak" and when you talk with a wine expert you can learn new varietals to try.

since i've done this (in the past ten years) my enjoyment of wine has gone up.

bane2• 12 years ago

Yeah exactly, all the varietals and vintage and flavor descriptions are basically just ways of narrowing the window to help you find something you might like.

For example, I tend to like wines with medium tannins and oak flavor, deep plum and tobacco notes. If there's spice flavors or not I'm also fine. I find I usually can find what I want in Chilean Cabs, and have less luck with other wines.

However, I've liked other kinds of wines too. Where I live the local Cabs are generally terrible, but the local Chambourcins, which have few of the qualities I generally look for, are good drinks too.

mankyd2• 12 years ago

The brothers' taste is [...] vindicated by their ability to name and communicate what they notice in the wine, and to enable us to come to notice it, too.

But that's the point of calling it "junk". What they notice (and thusly communicate) about the wine often simply isn't there. They certainly do "notice" things, but what they notice is often imaginary.

A good bottle of wine can certainly be enjoyed. It's a question of where they enjoyment stems from, and how much credence we put in the idea of some wines being significantly different and better than others.

JoeAltmaier3• 12 years ago

Sure some are better than others. Some batches of coca cola are better than others. The 'junk' comes in when the expert starts naming imaginary features and rating them.

mankyd2• 12 years ago

Yes. I should elaborate on what I mean by "how much credence". We have an entire industry built up not just around wine, but around _tasting_ wine - judging it, speaking about it, writing about it, etc. It seems that, while some wine will certainly be better than others, the degree of difference from one quality wine to the next, and the amount of significance we ascribe to it is where the debate should really lie.

_ak• 12 years ago

I'm not sure whether things are actually imaginary, or whether they might just have a lower perception threshold.

In beer tasting, we talk about off-flavors like green apple, butterscotch, cooked cabbage, or flavors like spicy/peppery, cloves, bananas, citrus, passionfruit, etc. None of these are imaginary, as each of these are linked to a specific chemical compound, such as acetaldehyde, diacetyl, DMS, various phenols and esters, and of course hop oils and various humulones, cohumulones, etc. from the hops.

What would help in wine tasting would be to link each aroma and flavor to specific chemical compounds, it would make detection more objective and make it possible to tell apart the very real aromas and flavors from the imaginary ones.

grimman• 12 years ago

I'll just throw this whisky tasting link in for contrast: https://www.youtube.com/watch?v=Boj33DVSILQ

adwf• 12 years ago

That's a brilliant video. Scottish people and pretentiousness don't tend to go together very well :)

arethuza• 12 years ago

That's not strictly true - it's practically mandatory here in Edinburgh at this time of year ;-)

dkfmn2• 12 years ago

For contrast here's another I've always enjoyed, a little more playful: http://youtu.be/BW1te_miu5I

vadman• 12 years ago

I like Paterson, but I always appreciated Charles MacLean's more relaxed delivery -- https://www.youtube.com/playlist?list=PLC6CEFB7B0C73A51D

atmosx• 12 years ago

This video is awesome. The guy to my eyes - I don't drink alcohol, let alone whiskey - is as 'expert' as one can get on the subject. I don't expect any of my whiskey-lover friends to know any of these things.

Really awesome video, thanks.

sudhirj• 12 years ago

Wine tasting as layment know it and as the connoisseurs know it is a little different. Or as a sommelier knows it, rather.

I like wine generally, but most of what I know of it comes from the documentary Somm, (on Netflix) - a sommelier's job is to tell a story - and to do this s/he must very knowledgeable and articulate about wine, its history, and its present. The history is basically a memory test, tasting wine is articulating what the drinker might experience in words, and having a palate honed enough to determine exactly what the wine is and where it's from.

All three, except the articulation, are verifiable quantitative tests. Even the during articulation the sommelier cannot stray too far from what is conventionally accepted.

I'm not harping too much on the sommelier service test, because the article is about wine tasting, but other sides apply. People train themselves to learn about wines in detail, and to differentiate wines right down to the highest level of classification and regionality just by just smell and taste alone.

pappyo• 12 years ago

It's actually quite impressive what a master Sommelier can do. S/he can taste a wine, tell you the varietal, tell you the region it was made and the year it was produced. The flowery language behind describing the wines is for the person the Sommelier is explaining the wine to.

I highly recommend the documentary Somm [0]. It's not an easy thing to become a master wine taster.

[0] https://www.youtube.com/watch?v=O4zeyuk8hL8

judk2• 12 years ago

Have there been rigorous studies to test these claims?

okabat• 12 years ago

I walked away from Somm convinced that wine tasting is a load of crap. The organization that puts on the "Master Sommelier" exam doesn't share results or a scoring rubric, which is a huge red flag that the test isn't merit based. It screams of manufactured exclusivity.

The best trained wine tasters can use deductive reasoning to narrow down the wine's identity, but a significant portion seems to be visual. Arguably, they are only effective at this because wine taxonomy is defined by visual characteristics. In a double blind taste test, I highly doubt it's possible to precisely identify the wine.

programmer_dude• 12 years ago

What these examples remind us of is that perceiving isn't opposed to or independent of understanding and thought; it is, rather, an exercise of understanding and thought.

I think this is what people should take away from this article.

gabemart2• 12 years ago

I have wanted to perform a rigorous double-blind wine-tasting experiment for a long time, but have always lacked the resources.

The true test of the validity of wine-tasting isn't whether experts can tell "good" wine from "bad" wine, but whether experts can reliably describe the same wine with the same language with an acceptably small margin of error.

My ideal experiment would give blindfolded wine experts a number of wines to taste. They would rate each wine using some appropriate standardized scale that would measure a number of relevant qualities of the wine. The experts would be given multiple samples of each wine, randomized appropriately. The experiment would measure how closely each expert mirrored their own rating of each wine on subsequent tastings.

This is basically a slightly more complex version of the famous (and hypothetical) "Lady tasting tea" experiment [1].

I like this setup because the ratings of each expert are compared only to their own previous ratings. I think this minimizes ambiguity and external sources of error.

[1] https://en.wikipedia.org/wiki/Lady_tasting_tea

JoeAltmaier3• 12 years ago

Doesn't have to cost much. Just choose some supposed wildly different wines, and tear up a sheet to make blindfolds.

My father-in-law Thomas is from India. He likes whiskey including Indian brands; his colleague was a Scotch whisky snob. At a party he challenged the snob to name all the whiskeys on the bar.

Thomas went into another room with the bottles in a bag; came out carrying a tray with half a dozen shot glasses. The snob tasted each one, make erudite remarks about their qualities and selected a brand for each one. Thomas opened the bag and showed that only the Indian bottle had the seal cracked.

Needless to say Thomas doesn't get invited over there any more.

gabemart2• 12 years ago

Doesn't have to cost much. Just choose some supposed wildly different wines, and tear up a sheet to make blindfolds.

The hard part is finding qualified experts who are willing to participate. The experiment would be most interesting and uncontroversial if it was done with experts whose credentials were unassailable.

jghn• 12 years ago

This is the real problem. The extreme majority of the "snobs" as you put it really don't have the pallet or skill to really be able to discern what they think they can, it's mostly brand loyalty, marketing, etc.

There often are people who can discern these things but they're much fewer and further between than people realize. Of course, those in the first category who agree with my statements will never have their mind changed, they'll just assume that they're in the second category. The likelihood of that being true has a lot of zeros after the decimal.

The problem with articles like these is that many "experts" fall into the first category, not the second. This is not without merit though as often these are the same "experts" that people trust to make buying decisions.

applecore• 12 years ago

Sorry, but that story is pretty dumb. If an expert is tricked into drinking the same shitty Indian whisky six times, they're going to be thrown off.

timtadh• 12 years ago

My wife tried this on me with yogurt. I like yogurt a lot and like to try every type of yogurt available in the store. She thought I just like buying the most expensive stuff. So one night she setup a blind taste test for me. I was able to distinguish and name the different yogurts (there were only a couple in the test).

But, I do think that Thomas' test was a bit unfair. He arranged as if it was a test of different whiskeys but then gave the same one every time. That isn't a double blind trial and it physiologically manipulates the participant. It makes a good story but I don't think it proves anything. Furthermore, you really need a larger sample size to say whether or not "snobs" can distinguish the real thing from the fake. If you are testing just one "snob" it would help to run multiple randomized trials.

mikeryan• 12 years ago

What you are describing is a sommelier exam.

There's an interesting documentary on some guys taking their master sommelier exam on Netflix called "Somm" who show an impressive array of these skills.

shalmanese• 12 years ago

One cheap test I like to do when with friends is to order a flight of something, have each person taste it in turns, then have one person shuffle the glasses around while the others aren't watching, recording how they were shuffled (if you want everyone to participate, just have a second person do a second shuffle).

It's then everybody's job to taste again and see if they can put the glasses back in order. So far, nobody I've done this with has performed better than chance with vodka tastings and people barely perform better than chance with red wine tastings, even when they're 3 different varietals.

The test tends to come out after at least a few drinks though so most people's palates are at least somewhat compromised.

WorldWideWayne• 12 years ago

Was anyone calling Wine-Tasting a science to begin with?

misiti3780• 12 years ago

Is it just me or does the art example goes against the entire concept of "thin slicing" in Malcolm Gladwell's book Blink. He gave an example of the exact opposite - a bunch of art experts thinking a statue is authentic, then one guy walking in, looking at the piece for 5 seconds, and saying it is fake (which it was).

diego_moita• 12 years ago

There are a lot of experiments showing that wine experts are very unreliable: by serving them the same wine in 2 different bottles (cheap and expensive), by adding food colorant to white wines, etc. Daniel Kahneman describes some of them in "Thinking Fast and Slow".

The same thing has been done with other fields dealing with vague and loosely-defined concepts: stock-picking, sports commentary, etc.

blakeja• 12 years ago

So is this article saying that trying to quantify what makes a wine "good" or "bad" is a matter of perspective and lacks any kind scientific method? I mean, do we need a scientific method if it is something we collectively can pretty much agree upon? They can do all the studies they want and criticize this all to the finest minutiae, but in my own personal experience I can say with certainty that wine recommended me to by trusted experts is generally better than just blindly picking a bottle. (Well not quite blind, but with a really limited knowledge of regions and vintages). At the end of the day for all us casual wine drinkers, I would think that is all that really matters. It may be interesting to see some statistics on casual vs serious wine drinkers, could help clarify a lot of this.

I would also say getting to know an expert to the point where they know your own individual tastes really helps. Perhaps it is this aspect that people are getting hung up, everyone will have their own tastes, enter the expert who can categorize this for you, etc.

glenra• 12 years ago

No, it's saying that people who claim to be able to identify which wines are "better" than others (and how they differ) can't reliably do that.

Wine recommended to you by trusted experts is not better than just blindly picking a bottle. Or at least, that is not something you can "say with certainty" if you haven't tested it. (And if you do test it, the results are likely to surprise you.) An awful lot of the experience of wine turns out to be the result of your expectations of it.

aruggirello• 12 years ago

While everything about wine-tasting is debatable, it must be said that sommeliers / wine tasters spend a lot of time training themselves to discern the scents and flavours of the different wines.

In fact there are several wine tasting training sets for sale, with tens of different wine aromas.

As humans OTOH, our smell is inferior to, say, dogs.

acd• 12 years ago

I've been at wine tastings where the sommelier who is a sommelier teacher can pick from what exact bordeaux region a certain wine is from without looking at the label. Respects that expert skill different from computers but mastery. I think it is like in all fields there are self proclaimed experts and real experts and not all who claim to be an expert are one some fake it.

However its also been proven that people think that an expensive wine tastes better than a cheap one. So psychology and expectations play a big role.

I would advocate that we get mass spectrometers out there in smart phones so that we can scan the wine and tell its contents. Not only do we want to test the wine contents, we also want to scan our meat to check that it is what is says it is and we also want to scan vegetables for pesticide residue.

dbbolton2• 12 years ago

Mass spec isn't terribly useful on a mixture with a large number of unknown chemicals (like wine). The resulting spectrum would likely have overlapping peaks all over the place and wouldn't really tell you much about the composition.

dbbolton2• 12 years ago

Wine tasting itself isn't necessarily a futile activity.

The "junk science" part is when self-described "experts" assign arbitrary percentile scores based on the tasting.

reubenswartz• 12 years ago

A few weeks ago I was cooking dinner for a couple of couples at our house. I opened a fancy bottle of wine about an hour ahead of time, and I also had an $8 bottle left over from cooking breakfast the day before. (You'd think from all this that I'm some kind of chef, but it doesn't take much to throw a little wine on some mushrooms.)

I had a couple of sleeves for wine bottles, so I decided to put them over the bottles and do a blind taste test. Of the 5 other people, 3 preferred the cheaper wine, 1 was neutral, and 1 preferred the more expensive one.

You don't need wine sleeves-- we've used paper bags, or you can decant in a double-blind safe way.

No one said science was easy, but someone's got to do it. ;-) You might even save yourself some money.

nsomniact2• 12 years ago

In college we used to pour handles of McCormick vodka into left over Grey Goose bottles(same ones used over and over). I think you all can guess that no one ever knew the difference. :p

bane2• 12 years ago

There's not really a good correlation between price and "goodness" of wine. The real rule of wine is, if you like it, it's good wine. If you don't like how it tastes, it's not good wine. That's it.

The wine world is notoriously pretentious.

It's good to know locations, varietals and aging a little because it can help you narrow the window down to find a wine you'll probably like well enough vs ones you won't. I for example, tend not to like Malbecs. Why? I dunno, just something in the flavor. I also don't like Mountain Dew.

Wine can help accent or suppress flavors in your food as well. Just like drinking beer with pizza tastes good, you probably wouldn't drink coke with your morning cereal.

If you really want to get into wines, there seems to be a correlation between rating and flavor complexity. And there's all kinds of ways to taste (or miss) complexity in wines. Just like a good bbq sauce has elements of sweet, sour, spicy, etc., wine can be more or less complex in flavors. But not everybody likes complex flavors and that's fine. That's easier in fact since complexity is usually a result of more expensive wine making processes like aging so your local Two-Buck Chuck might make those people happy. But detecting the differences or preferences for flavor complexity is like detecting the differences between two different batches of your grandmother's home-made Spaghetti sauce. You can probably describe it (the first had more garlic, the second more basil), but absolute comparisons are really impossible and don't really matter all that much. It's like asking "which is better a sponge or the number 7?"

Something revealed to me during trips to various places where a glass of wine with lunch or dinner is normal, is that wine is just another drink. Some of the better meals I've had were had with a 2 euro 1/2 L carafe of the local watered down table wine. In these places, wine is just a drink, like juice, water or soda. People who drink wine with every meal really don't care what the wine is so long as it doesn't taste bad.

You can probably find more than enough wines anybody will like for under $10 a bottle.

It's just a drink.

nsomniact2• 12 years ago

I don't trust anyone who can "smell" winners or losers. Seems like it's going on in the startup world lately

twoddle• 12 years ago

For your average Joe, yet is probably is (it certainly is for myself), but those who are 'trained' in the field or have had experience will be able to taste subtle differences between a £10 bottle of wine from Bordeaux and a £10 bottle of wine from Cali.

Is Wine-Tasting Junk Science?

Related Stories