The Supposed Science of Food Pairings, the Culinary Genome, and the Creative Process

Whenever we talk about pairing, I remember a short conversation I had years ago with a pickle maker.

It happened on way to a picnic in Cambridge with friends, heading down to the river near Harvard Square. We’d already grabbed supplies at Darwin’s, sandwiches and a big package of Oreos, which we were munching when we stumbled on a small farmers market.

It was mostly fruit stalls, but I’ve always been a salty guy, so I went straight for the Moon Brine pickle stand for some lunch suppliments.

I hesitated between the hots and the dills, and asked the proprietor which I should try first. The guy took a look at the open bag of cookies in my hand and said, “I’m not sure what really goes best with Oreos.”

“Hey man,” I replied, “Everything goes with everything.” and bought a jar of the hot pickles.

Nowadays, working for culture I spend a lot of time thinking about this stuff, and I’m not the only one; chefs, recipe developers and food manufacturers spend a lot of time agonizing over pairings and ingredient choices. Over at Chemical & Engineering News, writer Carmen Drahl has a fascinating explanation of the current scientific research into these combinations, and in particular, so-called flavor pairing theory.

Originally proposed by Fat Duck owner Heston Blumenthal and flavor chemist Francois Benzi, the theory states that harmonious flavor combinations occur when you combine ingredients that share certain odorants, molecules responsible for the distinct aromas and flavors of different foods and beverages.

One combo [chemist & blogger Martin Lersch] put forward was strawberry and coriander, after he deduced from the literature that the two share at least one key odorant, (Z)-3-hexenal. Bloggers who tried it confirmed the combination was tasty.

You can find shared odorants in a surprising selection of foods: famously, white chocolate and caviar have the odorant trimethylamine in common, while oysters, kiwi fruit, white wine and blue cheese all have methyl hexanoate as one component of their smell and taste. Single odorants aren’t the whole story, of course: most food’s flavor (and texture) comes from a combination of many odorants along with non-aromatic chemicals like salt and starch. Jasmine, pork liver and, uh, feces all share the a common odorant (indole), but they’re obviously a bit different on the plate.

Flavor pairing theory has been around now for several years now, as a part of the larger molecular gastronomy movement, and as Drahl notes in her article, it’s become accepted enough that there are several companies now selling proprietary tools for finding foods and beverages with common odorants. But don’t fire your sommelier yet: a simple relationship between chemistry and good taste combinations remains elusive. Just last year, a study by researchers at Northwestern indicated that preferred flavor combinations are strongly determined by cuisine, not common odorants.

While Western European and American cuisines had many dishes whose ingredients shared key odorants, Southern European and Asian cuisines favored combinations which shared very few: they combined ingredients that were highly dissimmilar. This indicates that, even if flavor pairing theory can correctly identify new pleasing combinations, the results are culturally-determined. Human tastes may be highly mutable, meaning there are lots of food and beverage combos out there that won’t “work” according to pairing theory, but will be tasty nonetheless.

Furthermore, Drahl notes in her article that some preliminary studies are now showing that novelty rather that acutal deliciousness might account for the popularity of some of these flavor-pairing combinations. In fact, the complexity of human chemical perception has led some critics to dismiss the theory outright, including plain-spoken tomato-flavor researcher Harry J. Klee, (speaking to Drahl):

“That whole flavor-pairing crap is just a gimmick by a chef who is practicing biology without a license.”

But between scientists wrangling over slippery sensory data, and businesses looking to cash in by selling patened databases, I have my own, uh, theory about pairing theory:

I suspect pairing theory isn’t especially good at giving us “correct” pairs. As Drahl notes in her article, new pairs discovered by the theory typically require skilled preparation to make them work together: just dropping an oyster into a glass of white wine doesn’t cut it. Not to mention all those great cuisines (Italian, Chinese, Thai, Greek, etc.) whose ingredients tend not to share common odorants.

Instead, I believe flavor pairing theory is a culinary Oulipianism, a constraint which unlocks creative possibilities. Like the Oulipo writers, who invented new poetic forms by, say, forbidding the use of the letter “e” or requiring each line of a poem to be a single letter longer than the previous line, the molecular gastronomist using flavor pairing is restricting the set of all possible choices in order to produce something previously unimagined.

It’s not that banana “naturally” pairs with cloves: banana happens to rhyme with cloves, and blue cheese rhymes with white wine, and Parmasan cheese with rum. The chef or sommelier can use that to help shape their choices, but they can always change the scheme and create equally-delicious combinations that run against theory: eggs and onions, hazelnut and chevre, blueberry and mozzarella.

As an aside, the Futurists were actually doing this decades ago with their recipes, serving fish with jelly on an engine block. They were more concerned with combining the metaphorical meanings of their food than their flavor, or edibility, but their goal was to shock your senses and create new possiblities.

How does this affect us, the cheese-lovers? Well, pairing theory accounts for some common wisdom: if you check the map in the sidebar, you’ll see that beer and white wine share many common odorants with cheese, and pairing with them is often easier than working with tannic reds or other beverages like coffee or tea. And there are some surprising revealitions, like the closeness of rum and blue cheese, or (specifically) roqefort and coconut, that seem worthy of a tasting. But there certainly are matches to be found on the far side of the flavor map, too, including classic match-ups like mozzarella and tomato and peaches and cream.

The downside of aesthetic theories of any kind is that they can provide cover to explain decisions that folks would otherwise have to account for with their own taste. It’s easier to say “the theory says x” than to say, “I like x”. This is why there’s so much money at stake: the snack developers over at Procter & Gamble want to cover their butts when they try Cinnamon Bacon Pringles out on the public, and the corporate chef wants someone to blame when their favorite honey-cilantro-mango smoothies aren’t a hit. In the same way, plenty of folks want to have a guarantee that the new Syrah they bought will really match their favorite Manchego.

I hate to burst anyone’s bubble, but if you want certainty, you’re in the wrong place. Of course, at culture we always do our best, as will just about any professional or organization, to ensure a great pairing, but I think the essential message, one that’s frequently repeated but seldom emphasized, is to trust yourself. Trying new things is an adventure, and success can not be guaranteed, whether it’s by an algorithm or a panel experts. You are the expert on what you like. Everything goes with everything.

Above: “making pickles” via Lori_NY

Will Fertman

Will Fertman is a writer and food business entrepreneur living in Berkeley, CA. A former staff writer and web manager for culture, Will wrote the Ruminations column for the magazine and spent lots of time wrangling social media. Today Will is the Director of Content and co-founder at the monger, a data-driven platform for the specialty food industry, supplying accountable information and transactions for producers, distributors, retailers, foodservice and consumers.