Over the holiday, a friend asked about the “is crema bad for coffee?” question that we recently posted. We provided an answer in the form of one of our Christmas dinner dishes. This then lead to a debate about the methodology used by today’s coffee tinkerers who are on a discovery mission to improve coffee.

Skimming the crema: a deconstructionist's act?As we wrote previously, some coffee professionals are questioning all assumptions — including whether crema is any good for espresso. The logic behind this last example being that crema, on its own, tastes rather bitter. And while there’s agreement that a healthy crema is a sign of a properly extracted espresso, some would even say the resulting espresso tastes better if you skim the crema off the cup.

How does our Christmas dinner fit into this? My wife decided to make a dish from The French Laundry Cookbook. Like many accomplished chefs, we’ve long considered Thomas Keller to be a great coffee pretender. However, earning three Michelin stars simultaneously at two different restaurants suggests that he knows a thing or two about food.

The French Laundry's crab salad: a metaphor for coffee flavor complexity?The dish in question is a Dungeness crab salad with a recipe that calls for daikon radish. Why did Mr. Keller opt to add the radish to a naturally sweet crab salad? On its own, the radish has a bitter flavor. But taste his recipe in its aggregate form, and the radish adds both a complexity to the flavor as well a texture the dish would otherwise lack.

To our own subjective taste buds, skimming the crema off an espresso is akin to Mr. Keller throwing out the daikon radish. An espresso without crema loses some of its flavor complexity, even if those components tasted on their own aren’t terribly appetizing. And as far as texture is concerned, espresso is a rare emulsion of gas, liquid, and solids all coming together with aromatics originating from each. Skim off the crema, and you’re merely left with the liquid minority.

In pursuit of independent variables

This crema-or-no-crema question typifies the sort of experimentation going on with coffee these days; the approaches we hear about are very binary or one-dimensional. But there’s good reason for that. Simple, baby-steps science encourages us to explore constituent influences in as much isolation as possible so that we may draw direct conclusions of cause-and-effect.

Thus we get primitive experiments such as crema vs. no crema, put portafilter handles in the freezer vs. leave them warm in the espresso machine group head, etc. And we get one-dimensional manipulations such as pressure profiling. Experimentalists are incented to fix all independent variables save for the one they’re monkeying with — to see how it contributes to the overall result. Tweak multiple things at once, and you cannot be sure what gets credit for the difference.

Illustration of the superposition principle of wavesThe problem with this deconstructionist approach is that it leads to a lazy, mistaken belief in superposition — i.e., the idea that all variables are entirely independent and that the whole is the mere sum of all the parts. Nature rarely works this way. Take something as multifaceted, subjective, and broad as coffee flavor. Optimizing your coffee growing, roasting, extraction, etc., for one characteristic — say, sweetness — often has major side-effects on the other characteristics — say, body. We’re certain that our past criticism of deconstructionist approaches must sound like some snobbish critique of post-modern art, but this is precisely what we mean by all that.

Today’s disproportionate appeal of single origin coffees and, more to the point, single origin espresso shots reflects the value currently placed in this simplistic, deconstructed approach. It is well-suited to the early developmental stages of today’s coffee consumer palates. Many single origin coffees exhibit simplistic one-note or two-note flavor profiles, offering basic building blocks for consumers’ budding flavor vocabularies.

Of course, not everyone prefers orchestral music to chamber music. But complex systems are so named for a reason. Saying “crab salad: sweet, good” and “radish: bitter, bad” reflects only the most primitive flavor profiling. Appreciating complex systems, where the sum may differ from the contributions of its constituent parts in isolation, means the difference between recognizing Thomas Keller for deliberate genius rather than for a random mistake.