Saturday, August 29, 2020

Always Bee Cooking #13: On Recipe Writing, Or, A Review of Two King Arthur Flour Recipes for Sourdough English Muffins

When I began writing this, the goal was to review the Sourdough section of The King Arthur Flour 200th Anniversary Cookbook. It’s about 30 pages long, or 5% of the whole book. I’ve cooked almost everything from it multiple times; breads, obviously, but also sticky buns and cakes. Breakfasts and desserts. It’s an impressive book that I have had almost nothing but excellent results from.

Which is frustrating, because I think the recipes are badly written. I knew that going into writing this review; what I didn’t realize is how precise I would need to be to articulate my problems with the recipe writing.

So that review is now this: a comparative review of two recipes for sourdough English muffins. One from the KAF Cookbook, and one from The goal, here, is closer to performing a close reading of a text than it is product analysis. I hope that, in pointing out issues with these recipes, I can help readers focus in on their own needs for recipe writing and to better imagine the results of recipes they read in the future.


Let me get ahead of myself.

The Sourdough section of the KAF Cookbook is positioned near the end of its 600 pages, followed only by crafts (like paste and papier-mâché  you can make from flour) and appendices. I note this because I suspect some of the problems I’m about to lay out are either addressed earlier (in the portions of the book I haven’t engaged with), or at least assumed based on layout.

That is: the sourdough baker is presumed to have the knowledge and experience of the bread baker, the dessert baker, the muffins & pastas & pancakes-maker. It is the final step in learning to bake, in other words.

I find this assumption suspect, given my own personal circumstances. Sourdough is not the simplest thing in the world, but its difficulty is no more reliant on expert technique than baking with Active Dry yeast is. Now let me stop getting ahead of myself.

My favorite example of the shortcomings of the KAF Cookbook come from comparing the recipe for English muffins in the book itself with the recipe for English muffins available on the website. Let’s go step by step. First, the introduction:

From the Cookbook:

The best English muffins are made with sourdough and their characteristic “holes” are created by adding baking soda just before they are cooked on a griddle.

From the

Who doesn't love English muffins? Homemade sourdough muffins seem even more scrumptious, and some of the taste-testers here had to admit that these crusty, chewy, tangy gems were some of the best they'd ever eaten.

I would honestly give the edge to the cookbook here. It’s a better sell that includes a guiding hand, letting you know that there is an important technical aspect to watch out for. “Some” testers calling these “the best they’d ever eaten” is pretty useless, honestly.

Onto ingredients.

From the Cookbook

  • 1 cup sourdough starter
  • 1½ cups milk
  • 5½ to 6 cups King Arthur Unbleached All-Purpose Flour
  • 1 tablespoon sugar
  • 1 tablespoon salt
  • 1 teaspoon baking soda
  • Cornmeal to sprinkle on baking sheet


  • 2 tablespoons (25g) sugar
  • 2 cups (454g) warm water (110°F-115°F)
  • 1 tablespoon active dry yeast or instant yeast
  • 1 cup (227g) sourdough starter, ripe (fed) or discard; ripe will give you a more vigorous rise
  • 7 cups (843g) King Arthur Unbleached All-Purpose Flour
  • ½ cup (56g) Baker's Special Dry Milk or nonfat dry milk
  • 4 tablespoons (57g) butter, at room temperature
  • 1 tablespoon salt
  • ¼ teaspoon sour salt (citric acid), optional; for enhanced sour flavor
  • semolina or cornmeal, for coating

These are different English muffins, structurally speaking; which is why this isn’t a comparative review to find the “better” version.

The most obvious difference is that the website does entirely away with the baking soda that the Cookbook version prioritizes. There are no chemical leaveners at all. The characteristic “nooks and crannies” of these muffins are exclusively produced by the CO2 production by the dry/instant yeast and whatever is still active in the sourdough starter.

The website version also has more clear sources of flavor: sugar is doubled, salt trebled, butter and dry milk and citric acid added to add texture and intensity of already-existing flavors. All of that added moisture and they’ve only added a cup extra of flour, with no change in the amount of sourdough starter (except that its role is better explained).

As far as ingredients go, the KAF Cookbook is clearly superior if you’re impulsively starting some English Muffins; fewer ingredients, more focus on pantry staples. And I can say from experience that they are delicious. The website, on the other hand, is almost certainly going to leave you with more tender, flavorful muffins. As long as you have dry milk on hand (I, personally, never have). It’s also more thorough about what those ingredients are for and what slight variations are okay.

Onto the directions.

From the Cookbook

Making the Sponge: In a ceramic bowl, mix together the starter, milk and about 3 cups of flour. Cover this with plastic wrap and leave it to work for anywhere from 2 to 24 hours. You might mix this up just before you go to bed so you can have fresh English muffins for breakfast the next morning.
Making the Dough: When the sponge has developed, mix the sugar, salt, baking soda and 2½ cups flour together in a separate bowl. Stir these into the sponge as thoroughly as you can and cover the resulting dough with plastic wrap and let it work for anywhere up to an hour. This allows the gluten in the flour you’ve just added to absorb some moisture and relax.
Kneading and Shaping: Flour your kneading board and hands well as this dough will be soft when you turn it out. Knead for only 2 to 3 minutes until the dough is smooth and no longer lumpy. With a floured rolling pin, roll it out, like a pie dough, from the center to the outside, until it is between ¼ and ½ inch thick.
Cut out circles between 3 and 4 inches in diameter (the muffins will shrink in diameter as they cook). A large tuna-sized can with both ends removed works well, or you can even throw tradition to the wind and cut squares.
Place the muffins on a cookie sheet that has been sprinkled with cornmeal and let them rest for at least 15 minutes.
Cooking: Place 4 or 5 circles on a lightly greased skillet on low, low heat with the cornmeal side down first. Cook slowly for 10 minutes, gently flip the muffins over and continue cooking for a further 10 minutes.
Serving: Cool your muffins, split with a fork to make the most of their wonderful open texture, toast and enjoy right away, or store the cooled muffins in a plastic bag to use at your leisure. English muffins also keep well in the freezer.


1. Combine all of the dough ingredients, except the cornmeal/semolina, in a large bowl.
2. Mix and knead — by hand, electric mixer, or bread machine — to form a smooth dough. The dough should be soft and elastic, but not particularly sticky; add additional flour if necessary.
3. Place the dough in a lightly greased bowl, cover, and set it aside to rise for about 1 1/2 hours, or until it's noticeably puffy. For most pronounced sour flavor, cover the bowl, and immediately place it in the refrigerator (without rising first). Let the dough chill for 24 hours; this will develop its flavor.
4. Gently deflate the dough, turn it out onto a lightly floured work surface, cover it, and let it sit for a few minutes, to relax the gluten. Divide the dough in half. Working with one piece at a time, roll ½" thick, and cut in 3" rounds. Re-roll and cut any remaining scraps. Repeat with the remaining half of dough.
5. Alternatively, divide the dough into 24 pieces (total). Shape each piece into a round ball, then flatten each ball into a 3" round. For a somewhat more even rise as the muffins cook, flatten each ball slightly larger than 3", and trim edges with a 3" cutter (or trim all around the edge with a pair of scissors). Muffins with cut (rather than flattened) sides will rise more evenly.
6. Place the rounds, evenly spaced, onto cornmeal- or semolina-sprinkled baking sheets (12 per sheet). Sprinkle them with additional cornmeal or semolina, cover with plastic wrap, and let them rise until light and puffy, about 45 to 60 minutes. If the dough has been refrigerated overnight, the rise time will be about 2 hours.
7. Carefully transfer the rounds (as many as a time that will fit without crowding) right-side up to a large electric griddle preheated to 350°F, or to an ungreased frying pan that has been preheated over medium-low heat.
8. Cook the muffins for about 10 to 12 minutes on each side, or until an instant-read thermometer inserted in the center of a muffin registers 190°F. The edges may feel a bit soft; that's OK.
9. Remove the muffins from the griddle, and cool on a rack. Store tightly wrapped at room temperature for 4 or 5 days; freeze for longer storage.

It’s here, I think, that the differences really show themselves. I’ll try to unpack how, because it’s my central issue with the cookbook itself: the sourdough recipes are really, really excellent. As long as you bring enough knowledge to fill in the gaps. When you’re trying to cook “by the book” they are incredibly frustrating.

Before that, though, I should acknowledge that the KAF Cookbook was initially published in 1990. There are reasons for the different approaches here; thirty years, for example. But also the book is a tome, and costs money (unlike the website), and has more explicit limits on space (page layouts) than the website does, and so has a different presumed audience. Those thirty years also saw massive shifts in the discourses around home cooking, which maybe one day I’ll properly be able to get into. I bring that up here to make it clear I’m not discounting these historical and material differences when getting into the comparison.

Let me get ahead of myself again. The first problem will come second. Here is the second problem: look at the cook times. The KAF Cookbook calls for 3-4 inch rounds flattened to between ¼ and ½ inch thick, while the website specifies 3-inch rounds at ½ inch thickness. So the cookbook version is a little wider and a little thinner, on average, with more space for variation.

The KAF Cookbook calls for “low, low heat” for 10 minutes per side. The website asks for 10-12 minutes per side on “medium-low heat.” Can I tell you something, from experience? Low, low heat for ten minutes per side is nowhere near long enough to cook through the English muffins you get from the cookbook. Keeping it low meant my first batch had to be cooked for something closer to thirty minutes per side. Which is annoying.

This isn’t a problem because it’s wrong. Most recipes are wrong, because it is impossible to figure out variables in things like burner strength, pan availability, personal patience, and on and on. It’s a problem because of what it chooses to prioritize difference within.

The cookbook wants to be generous with the size and shape but not the cooking time. The website calls for specific dimensions with variation in the cooking time. The latter is generally more useful, in my experience, because it signals that even if you make things perfectly even there are variables at play that you should anticipate. The former simply says: do what you like, and here is when it will be done.

The first problem, then. It is one that plagues so many of KAF Cookbook’s recipes, and it shows up in the first line: “In a ceramic bowl, mix together the starter, milk and about 3 cups of flour.” Then you cover this, whatever it is. Is it a batter? A shaggy dough? A smooth dough? A fucking rock with loose flour everywhere? Who cares! Cover it!

If there is one thing I’ve found deeply annoying when reading recipes to try to learn baking, it’s the use of vague terms to refer to the consistency of a dough. What is a “shaggy” versus a “smooth” dough? How stiff is a “stiff peak?” What does “silky” feel like in any context?

The most frustrating thing about this is that the answer usually ends up being fairly obvious. But it’s only obvious in the doing, rather than the reading. You can (and will) overshoot it or undershoot it. But even though they use vague descriptors, a lot of the standard terminology of baking is actually incredibly useful, at least in terms of helping you identify when you’re ready to move to the next step.

The KAF Cookbook is lacking these descriptors in almost every important area. You have no opportunity to be annoyed by the vagueness of a “silky” knead, because you are expected to knead for the prescribed amount of time and pray.

A side note: the vague way dough consistency is described is both the funniest and most liberating thing, for me. I had long bought into the idea of baking as the “scientific” (which is to say math-oriented) side of cooking. It turns out to be the place with some of the most flowery (flour-y? yes.) language in the whole recipe world, even as they insist on precise percentages. C’est la vie. Or, more precisely: bake weird shit, fuck ‘em.

Compare that total lack of an indication of what you will be covering with plastic wrap to the website’s version: “Mix and knead — by hand, electric mixer, or bread machine — to form a smooth dough. The dough should be soft and elastic, but not particularly sticky; add additional flour if necessary.” What a wild difference.

For instance: the website’s version anticipates the question of tools. This is a problem with baking recipes; it’s not uncommon to find recipes that reveal much too late that they can only be done with a stand mixer unless you are prepared to hand-mix for an hour, at which point the timing of everything else will fall to pieces. This reflects the assumption that the KAF Cookbook makes that I discussed above, where using “advanced” ingredients assumes you have a certain grasp on technique.

Finding decent baking recipes that don’t assume that you know advanced technique just because they include a slightly involved (or “advanced”) ingredient often feels impossible. Ingredients and technique are different things. What the “by hand, electric mixer, or bread machine” does is to say: We are not going to tell you how long it takes to get where you need to go, because it doesn’t matter. We are going to tell you where you need to end up, because that does.

And where you need to go is a “smooth” dough. “Soft and elastic, but not particularly sticky.” That is going to feel different to every pair of hands connected to every of brain that tries this recipe. But it’s at least a guideline.

An important point: if you try this recipe and fuck it up, you have a potential point of failure here. If you try it again, you can remember: maybe I didn’t add additional flour when it was necessary, maybe it was stickier than intended (or less soft, or whatever). If you fuck up the KAF Cookbook version, there’s no reference. You simply did what was asked and it didn’t work; it therefore must be a bad recipe.

I would also like to return to the “ceramic bowl,” briefly. I could say that it’s an indication of the bizarre, broken priorities of the KAF Cookbook to insist on specific equipment rather than conveying useful information. I just did, in fact. More importantly: it’s not important. Whatever bowl you have is fine. Maybe some will serve you better than others, which is a journey worth taking. Prescribing the bowl rather than telling you what you should be striving for is terrible recipe writing.

This is the meat of the issue, and what I’m hoping most strongly to convey. Recipe writing is about producing good food in clear ways. It’s about choosing optimal ratios of ingredients and developing them with explicable techniques. It’s a product-oriented writing: the goal is to end up with something delicious and/or healthy and/or novel, but always edible.

An aspect of this that can get lost, though, is that recipe writing is also about teaching process as well as product. Whether the writer intends it or not. Because recipes are product-oriented, they will inevitably cut out explanations of why they suggest a certain method of cooking or include a specific ingredient. This lack of explanation can lead to an incurious relationship toward food. You simply plug in the recommended variables and receive something delicious at the end. Voila.

This is especially a problem when a recipe works out well. If you don’t need to know what stage to knead dough to, then you’ll never know you need to ask. And so when you try to introduce variations, you have no way of knowing which questions to ask. And if you don’t know which questions to ask, you have no way of knowing why things work or don’t.

A good recipe will end in good food. A great recipe will help you learn how to make good food.

Blog Archive