A Theory of Wine-food Interaction

These idea were more recently explored again here.

Unraveling the nature of wine pairings is tricky stuff. I’m sort of out of the game these days because I do not work with a tasting menu anymore (or even a changing menu for that matter). When this blog started I had posted quite a few accounts of pairing wines with certain dishes. We did quite a lot of tasting menus back at Dante and I was fortunate to do a lot of eating and drinking of them.

At various times in the past I’ve mentioned some ideas that governed the mechanics of food and wine interaction such as sensory after images (they are harnessed for many magic tricks) and the change in contrast detection exploited by “black art” theater.

I’ve also discussed reward systems (read this Wired.com article first) that govern the construction of our reality when perceiving food. For example, sweetness can suppress the perception of bitterness when experienced at the same time. Bitterness is seen by the body as a negative and is therefore harmonically dissonant (a taste for bitterness can be acquired, of course and my theory is that the acquiring of something so attentional is related to anxiety). If you tasted Campari before it was sweetened you would probably spit it out, but after Campari is sweetened by the producer, the perception of bitterness changes markedly. Our body’s reward system recognizes that though bitter, Campari is also redeemingly sweet and therefore nutritious which is why it constructs reality in way that makes the bitterness less dissonant.

At times, food & beverage interactions seem like black art theater (watch the video linked above) where comparisons between flavor divisions result in changes in contrast detection among other divisions (harmonically or inharmonically) while at other times interactions seems like they are governed by nutritional preference or warnings.

It might be possible to classify reactive pairings in two ways: nutritional preference pairings and “black art” contrast pairings. It is useful to revisit Dorneburg and Pages’ amazing text on wine pairing What To Drink With What You Eat and consider these two divisions.

Why does a dessert wine always have to be sweeter than a dessert? When pairing dessert wines with sweet foods, we want nutritional preference to go to the dessert wine, therefore it needs to be slightly sweeter. If the wine is not as sweet, reality will be constructed to show preference for the food and the dessert wine will be presented by the mind as thinner and stripped away of its richness.

“With a simple apple dessert like apple pie, Sauternes is a soft and sweet accompaniment. But if you serve the same apple pie with caramel sauce, it makes the wine taste flat.” -Madeline Triffon  from What To Drink With What You Eat a.k.a. “WTDWWYE”

Madeline’s results may be because the caramel sauce is sweeter than the wine and our reward system favors it over the Sauternes.

“Having birthday cake or wedding cake with a brut Champagne toast is horrifying! If the dessert is sweeter than the wine, it makes the wine taste drier. My favorite all-purpose sweet wine is Moscato d’Asti.” -Madeline Triffon from WTDWWYE

Let us consider another scenario: port can come after bitter chocolate, but black espresso which is also bitter like the chocolate cannot come after port. Nutritional preference will dictate how reality is constructed and the second stimulus will either be flattered and harmonically enhanced or ridiculed by the mind. When black espresso is consumed after something sweet and more nutritious, the bitterness is dramatically emphasized in our construction of reality.

Our body warns us of all sorts of things with its construction of reality, but why? Some seem so innocent.

Why does increasing temperature lower the threshold of perception of alcohol making it more apparent in hot drinks?

Why does lowering temperature such as chilling a red wine, lower the threshold of perception of tannin? The same happens when intensely oaked whiskeys are chilled. What are we being warned about?

Pairings related to black art style changes in contrast detection might not work the same as pairings related to nutritional preference (I’m rethinking this because nutritional preference seems to always linger).

The black on black of black art contrast pairings is typically the greatest attentional feature common to both the food and the beverage. In the case of wine and food that feature is typically acidity while in the case of dark chocolate and espresso that feature is bitterness. Scanning through WTDWWYE, most all the highest regarded pairings are related to matching acidity. When the major attentional feature is matched, contrast detection between other features is augmented and they are “elevated” to quote commonly used pairing language. In best case scenarios, an aroma from the food is “brought back into focus” and seemingly superimposed over the wine. Typically foods that do this have a significant sensory after image which may prove related to nutritional value.

**** I will analyze these when i get around to it.

holy grail pairings to ponder from WTDWWYE:

almonds : manzanilla sherry

asparagus : sauvignon blanc

ribs : zinfandel the sauce on ribs often has sugar, but the illusion of sweetness in the aroma of the zinfandel might be enough to create nutritional preference in the wine.

biscotti : vin santo the wine is sweeter than the biscotti which tips nutrional preference in its favor.  the aromas are also not too disparate which might influence the reaction.

cassoulet : tannic red wine pairing tannin and fat could be less a pairing that happens in the mind due to nutritional preference and more of a pairing that actually happens on the tongue as is the classic explanation.  within the mind though,  the tannins could end and provide relief from the cloying sensory afterimage of the fat.  i’ll have to ponder this one next time i find a really tannic wine.

caviar : champagne

ceviche : sauvignon blanc

charcuterie : beaujolais

fresh goat cheese : sancerre this pairing features an acid/acid comparison which induces a change in contrast detection similar to black art.

muenster cheese : gewurztraminer (esp. low-acid fruity)

roquefort blue cheese : sauternes

stilton british cow’s milk blue : barley wine

chocolate : banyul’s or port

choucroute (sauerkraut) : alsatian reisling or german kabinett

clams : muscadet

corn : chardonnay, buttery oaky california

crab : riesling, esp. german kabinett or spatlese crab is often referred to in language as “sweet” and often dressed up in very nutritious butter, so a reisling like a sweet spatleses might be needed to induce a simple nutritional preference pairing.

foie gras : sauternes this could be a simple nutritional preference pairing where the reward for sweetness trumps the reward for fat.  the reward for the sauterns could only be slightly greater than that of the fatty foie gras making the experiences not feel particularly distant.

olives : sherry, esp. fino, manzanilla or amontillado the most prominent attentional feature of both could either be acidity if the olives are fermented or olfactory umami created by the esterifcation of fatty acids.  upon fermenting olives take on an aroma like sausages and sherry is known for its “rancio” character.

oysters : chablis the most prominent attentional feature of both can be acidity if minionette or lemon juice are added to the oysters resulting of a change in contrast detection in the aroma of the wine.  the chablis may also get associated with the lingering salinity (after image?) and therefore enhanced by our reward system.

salt? p. 164

Does any of this conform to anybody else’s experiences?



Instant Aging: Vacuum Reduction Yields Barrel “Bouillion” Cubes

[As I’ve learned since through performing maturation experiments and reading a lot of journal articles, there are lots of flaws with my theory of what barrels contribute but you can still use the idea to create a very accurate barrel essence to add to other finished spirits]

After dissecting many spirits, I’ve come to realize that the vast majority of sensory attributes that a barrel contributes to a spirit are not volatile at the boiling points of ethanol and water [this fails to account for the significance of the angel’s share]. What this means is that we can remove the volatile constituents and end up with dried barrel essence. We can then introduce a solvent such as an un-aged fruit brandy or aromatic bitters and synthesize the character of barrel aging by having the perfect ratios of non-volatile compounds.

Vacuum reduction is a great tool for reducing a barrel aged spirit such as bourbon down to a dried powder because it reduces the boiling point of water enough that the flavor compounds will be as un-effected by heat as possible during evaporation (no process volatiles). My vacuum reduction rig is a Comeau vacuum aspirator (acquired for $75!) attached to a vacuum flask ($15) heated by a hot plate (a stove on low with a double boiler substitutes well). A double boiler always needs to be used because when you run out of water, the solids will scorch instantly. If your solids clump when you try and reconstitute them, they likely got scorched.

The cost of this process is essentially the cost in bourbon of the volume you want to artificially age. For example one ounce of bourbon is sacrificed for every one ounce of peach brandy you want to treat. An ounce of bourbon from a handle of Evan Williams costs about fifty cents.

For the proof of concept, 100ml of Old Granddad was reduced to powder and then the barrel essence was reconstituted with Kuchan brand Indian Blood Peach brandy. The results are very impressive with the un-aged spirit tasting very much aged. I never really enjoyed the un-aged peach brandy previously, because it seemed to resemble bubble gum, but the barrel essence seems to add attributes that push the ordinary into the extraordinary.

And this is all legal…! People are willing to spend tons of money reducing Campari to powder to rim a glass. This is significantly cheaper (after you spent $100 on some used lab equipment) and probably much more interesting on a sensory level.

[I’ve used this technique quite a bit, liters at a time, and have settled upon using an excalibur food dehydrator over a vacuum reduction rig. I’ve also digested quite a few brilliant journal articles on aging. This technique is a great predictive tool for distilleries doing new product development but you cannot be naive about the results and need to know how they differ from the real deal. A new make white dog, for example, would not be instantly aged and rather would require months to calm down and come to equilibrium. No angel’s share would be accounted for either. All in all, the results are better than any other fake aging technique and can be done affordably and accurately in volumes as small as 2 oz.]

DIY Barrel Proof Overholt

After dissecting so many spirits I got the idea that I could create a barrel proof rendering of a whiskey by using vacuum distillation paired with a second distillation. The non-volatile part would be separated from the volatile part then the volatile part concentrated in a typical copper reflux still and the two segments rejoined. The change would be a reduction of water and therefore an increase in proof and a concentration of aroma (there are some finer points where fidelity is lost).

In the process, I discovered that my simple glass laboratory vacuum still sucks and is not good at collecting the solvent (I think I know how to tweak it). Time is money and Overholt is cheap so I ended up merging two bottles into one. One bottle was vacuum reduced (using a simple aspirator-to-vacuum flask rig) as far as my patience would take it (from 750 mL to 180 mL, but next time I’ll go all the way). The second was was simply re-distilled at high reflux to 80% alc. The volatile part is not sensitive to heat, so using a conventional normal-temp-atmospheric-still does little to impact the aroma (this is true enough, but not really).

The sensory properties of each half are really interesting. A lot of what a barrel contributes to a spirit is not volatile so re-distilling an aged whiskey reveals something very close to the white dog that went into the barrel. The non-volatile parts were concentrated at about 60C and it was amazing to see how much aroma was there. The non-volatile aromas really seem to define Overholt. The vacuum reduced segment became really turbid and I was afraid I spoiled the color permanently, but after marrying the color went right back to the same beautiful barrel hue.

Unfortunately, patience got the best of me and because I only reduced the non-volatile half to 180 mL, I only ended up with a 55% alcohol finished product, but it still made a lovely Manhattan.

I was really impressed with the success of this technique and am going to pursue it further. Next up is Fernet 151!