I’ve been intrigued by the question of whether there are principles that explain why certain flavors work well together while others don’t. Is it just a matter of historical accident that Italians are devoted to garlic, basil and tomatoes and the Chinese like soy sauce, ginger, and rice wine? Or can these canonical patterns be explained by more fundamental principles? Skeptical arguments opposed to the idea that culinary preparation is a genuine fine art often base their argument on the absence of such principles. But a new study casts doubt on that skepticism.

One popular hypothesis among chefs and food scientists is that ingredients that share flavor compounds are more likely to taste well together. Chocolate and blue cheese are alleged to share at least 73 flavor compounds suggesting that they ought to pair well together.

This new research published in Scientific Reports suggests that this hypothesis is true only in North American and Western European Cuisines. By contrast, East Asian and Southern European cuisines tend to use flavors that have distinctly different flavor compounds. In short, North American and Western European recipes depend on synergy with similar flavors enhancing each other, and Asian and Southern European cuisines depend on contrast.

The study analyses thousands of recipes by looking at the flavor molecules that explain the taste of each ingredient, allowing them to measure the degree of taste similarity between any two ingredients, and resulting in a “flavor network” which divides ingredients into flavor categories and shows their relative degrees of similarity.

The study further finds that the shared compound effect in North American/Western European cuisine is primarily the result of the use of milk, butter, cocoa, vanilla, cream, and egg. In the Asian/Southern European cuisines, the use of beef, ginger, pork, cayenne, chicken, and onion were responsible for the contrasting flavors. When these ingredients are statistically removed from consideration the effect of synergy and contrast largely disappears.

This study is a treasure-trove of information; it will take me some time to absorb it. But it strikes me that this is a landmark investigation. If there are indeed principles of culinary organization—analogous to what we fine in music theory for instance—this data should provide a foundation for their discovery.