There is a lie running through your cookbooks. No, it’s not that you can substitute crackers for apples in your pie and no one will know the difference (though, come on, let’s be decent to each other, folks: Knock that off.) The lie goes much deeper than all that, and is the source of what I call the Cookbook Paradox.

The Cookbook Paradox is simply this: Most cookbooks do not actually teach you how to cook.

What they teach you to do instead is to memorize, and perhaps even replicate, recipes. But cooking isn’t just replication. It’s taking the ingredients you have on hand, or the ingredients that appeal to your tastes, and turning them into the food you want. It is, in other words, a skill—and one that most cookbooks will not help you develop.

The problem lies in the format: Memorizing recipes is just not a good way to learn to cook. So what’s a better way? In her new book, How To Bake Pi, mathematician/baker Eugenia Cheng offers a novel, mathematical approach to cooking—one that presents a potential solution to the Cookbook Paradox.


How To Bake Pi is more than a mathematically-minded cookbook. It is just as much a book about mathematical theory and how we learn it. The premise at the heart of the book is that the problem that stops a cookbook from teaching us how to cook is the same problem that makes math classes so bad at actually teaching us to do math.

“If math is hard it might be because nobody told them what it was for,” Cheng says, of those who went through math class alternately bored or frustrated. Math, she argues, is not for crafting strained little word problems to calculate the speed and distance traveled by cars. (Cheng correctly points out that any reasonable person would solve such a problem by looking at their car’s dash.) Instead, she says, we should be teaching math to address harder questions, to give us a way to get answers beyond what we can get from a calculator or a Google search—and one way to do that, is through our kitchens.

This collision of math and the kitchen is not, by itself, a particularly new one. Recipe conversions and measurement adjustment problems are so common as to be almost math class cliches. The mathematical territory that Cheng suggests navigating through cooking, though, is a largely unexplored one. A custard recipe doubles as a lesson in the principles of logic. A story about making-up a plum cake recipe becomes an introduction to the process of crafting mathematical generalizations. Instructions for making a Baked Alaska also serve as an explanation of the structural elements of category theory (Cheng’s own mathematical speciality.)


These are recipes, in short, that are less concerned with teaching you how to make a specific kind of chocolate cake and more concerned with getting you to understand the basic principles behind how that cake is put together. This approach perhaps explains why the recipes in the book are often structured not so much as instructions to be followed as they are equations that can be applied to the food “variables” that suit your circumstances at any given moment.

That doesn’t, however, mean you can simply careen around the kitchen tossing whatever substitutions pop into your mind into the mixing bowl and expect to have something resembling the promised cake at the end. The baking formulas can be remarkably rigid when they need to be. “(Egg yolks + sugar) + milk ≠ egg yolks + (sugar + milk),” Cheng admonishes cooks who think they can run a shortcut on her custard formula, a point she further illustrates a page later with a series of branch diagrams. Once you have learned the formulas for the recipes, though—and taken care to note precisely where the locked parameters fall—they offer a relaxed, uncluttered approach to baking that is genuinely enjoyable.

There are recipes in the book for which Cheng’s stripped-down, mathematical approach works better than others. The Baked Alaska recipe was frustratingly light on specifics when the bowl of raw egg whites I’d spent 20 minutes whipping simply refused to meringue. Cheng’s basic cake recipe, on the other hand, easily accommodated my improvisations (I was out of cocoa, but did have plenty of raspberries that I could toss into the batter instead). And that barebones custard equation above turned out to be particularly dreamy in practice.


In fact, my biggest disappointment with the book was that these recipes/formulas were mostly limited to the top of each chapter. I could easily see a more exhaustive mathematical cookbook, one that branched out further than desserts, being a welcome addition to math classes and kitchens alike.

Top image: art by Tara Jacoby, Bottom Image: My raspberry/dark chocolate spin on Cheng’s conference cake recipe