In November 2006, food writer Mark Bittman published a recipe in The New York Times that promised to shake up the world of home baking. The recipe, developed by Jim Lahey of the Sullivan Street Bakery in Manhattan, did the seemingly impossible: It allowed the average home cook to bake a loaf of bread that looked like it had been produced in a professional bakery. The recipe, which instantly won legions of followers, was exceedingly simple: Mix a few cups of flour, a tiny amount of yeast, and a little salt together in a bowl; stir in some water until the ingredients just come together; and leave the dough to rise. After 12 to 18 hours, the dough is turned a couple of times, shaped, risen, and baked in a Dutch oven. An hour later, out comes the most beautiful-looking loaf most people have ever baked at home—and all with no kneading required.
At first, it seemed unlikely that there was anything to improve upon here. The no-knead recipe was remarkably easy and worlds better than other no-fuss breads. But there were some complaints amid all the praise. I decided to give the existing recipe to five inexperienced bakers in order to see what (if any) issues arose. I noticed a problem even before we sliced into the first loaf. While all were beautifully browned and crisp, the loaves varied wildly in size and shape, ranging from rounded mounds to flat, irregular blobs. Casting first impressions aside, I cut into each one and tasted a bite. Though the crusts were extraordinary—better than any I’d ever produced—the flavor of the crumb fell flat in every sample. It simply did not capture the complex yeasty, tangy flavor of a true artisanal loaf. I wondered if I could make this bread more consistent and better-tasting.
I decided to tackle the problem of shape first. Thanks to the ingenious use of a Dutch oven, the bread always acquired a dark, crisp crust, but the loaves took on a disconcertingly broad range of forms. After observing testers make the recipe a few times, I realized the problem: The wetness of the dough was making it too delicate to handle. Though it was well risen before baking, it was deflating on its way into the pot. In addition, because of its high moisture content, the dough was spreading out over the bottom of the pot before it could firm up properly. I analyzed the no-knead recipe and found that its dough is 85 percent hydrated— meaning that for every 10 ounces of flour, there are 8.5 ounces of water. Most rustic breads, on the other hand, max out at around 80 percent hydration, and standard sandwich breads are never more than 75 percent hydrated. So what would happen if I reduced the water?
To find out, I made a batch of dough in which I cut the hydration to 70 percent. Sure enough, this dough was much easier to handle and emerged from the oven well risen and perfectly shaped. But unfortunately, the texture was ruined. Instead of an open, airy crumb structure, it was dense and chewy, with rubbery pockets of unleavened flour. So more moisture led to an open but squat loaf, and less moisture led to a high but dense loaf. Was there a way to reconcile these two extremes?
How Water Affects Bread Texture
50 percent hydration: This loaf had the smallest ratio of water to flour, resulting in a weak gluten network. As a result, the loaf was small and dense with a tight crumb.
68 percent hydration: With a typical hydration level of 68 percent, this loaf rose and expanded well, possessing modest-sized holes.
80 percent hydration: This loaf had the greatest ratio of water to flour, causing the gluten network to be weak and diluted. Therefore, the loaf was flat and wide with large air pockets.
Many bread recipes call for a rest period after adding water to the flour but before kneading. This rest is called “autolysis” (although most bakers use the French term autolyse). In most recipes, autolysis is just 20 to 30 minutes, but the no-knead bread calls for something completely out of the ordinary: a 12-hour rest. Was there something in the mechanics of such a lengthy autolysis that could help me solve the textural problem? The most common explanation for the autolysis process is simply that it allows time for the flour to hydrate and the mixed dough to rest, making the dough easier to manipulate later on. But the word “autolysis” technically refers to the destruction of cells or proteins through enzymatic action. I decided to have a closer look at what really happens to the dough during the process.
The ultimate goal of making bread dough is to create gluten, a strong network of cross-linked proteins that traps air bubbles and stretches as the dough bakes, creating the bubbly, chewy crumb structure that is the signature of any good loaf. In order to form these cross-links, the proteins in the flour need to be aligned next to each other. Imagine the proteins as bundled-up balls of yarn that you are trying to tie together into one longer piece, which you’ll then sew together into a wider sheet. In their balled-up state, it’s not possible to tie them together; first you have to untangle and straighten them out. This straightening out and aligning is usually accomplished by kneading.
But untangling and stretching out short pieces of yarn is much easier than untangling entire balls. This is where autolysis comes in. As the dough autolyzes, enzymes naturally present in wheat act like scissors, cutting the balled-up proteins into smaller segments that are easier to straighten during kneading. This is why dough that has undergone autolysis requires much less kneading than freshly made dough. And here’s where the hydration level comes in: The more water there is, the more efficiently the cut-and-link process takes place.
So this was the explanation for how the no-knead bread recipe published in The New York Times worked. With 85 percent hydration and a 12-hour rest, the dough was so wet and had autolyzed for so long that the enzymes had broken the proteins down into extremely small pieces. These pieces were so small that, even without kneading, they could stretch out and cross-link during fermentation and the brief turning step. At 70 percent hydration, there simply was not enough water in my dough for the enzymes to act as efficiently as they had in the original recipe. As a result, many of the proteins in my finished bread were still in a semiballed-up state, giving my bread the overly chewy texture.
What if the secret to making a better no-knead bread was actually adding in some kneading? I knew that even at a relatively dry 70 percent hydration, the proteins in my dough had already been broken down significantly by the long 12- to 18-hour autolysis. All they probably needed was a little kneading to untangle and create an airy, light crumb. I decided to make the leap.
I took the dough that I had resting from the day before and turned it out onto my board. Trusting that my understanding of autolysis was correct, I gave the dough the bare minimum of kneads—adding just 15 extra seconds to the no-knead recipe—and continued exactly as I had before. The dough emerged from the oven as beautifully browned and perfectly shaped as any I’d made so far. After letting it cool, I cut into it to reveal an ideal crumb structure: large pockets of air and stretched sheets of gluten. Not only that, I found that since such a small amount of kneading could develop gluten in such a forceful manner, I could actually reduce the minimum time of the rest period from 12 hours to 8 hours. That 15 seconds of kneading had reaped huge benefits.
No Substitute for Flavor
Now that I had bread with a great shape and texture, I turned my attention to the loaf’s lackluster taste. To get a better sense of what specific flavors I was missing, I bought a loaf of bread from a bakery that makes dough the old-fashioned way—with a fermented starter. Because a starter contains a much more varied assortment of yeasts than the ones found in a packet, it yields more complex flavor. Tasting the bakery bread side by side with the no-knead bread confirmed this. But creating a starter is a multiday process. How could I get the flavors that a starter produces without actually having to use one? Could I introduce a little tanginess another way?
Scanning the labels in our dry storage area, I saw that the majority of our bottled vinegars are 5 percent solutions of acetic acid—the same acid produced by bacteria during dough fermentation. Since other vinegars would introduce undesirable flavors to the bread, I experimented with different amounts of distilled white vinegar before settling on a single tablespoon.
My bread now had tang, but it lacked complexity. What I needed was a concentrated shot of yeasty flavor. As I racked my brain, I realized that beyond bread, there is another commonly available substance that relies on yeast for flavor: beer. Would its flavors compare to those produced in dough fermentation?
How Beer Boosts Bread's Flavor
During a starter’s fermentation, yeast produces alcohol, carbon dioxide, and sulfur compounds, all of which contribute to good bread’s unique flavor. These three elements are present together in another location—a bottle of beer. But why choose lager over other types of beer? It’s all about the fermentation. Most nonlager beers undergo a process called “top fermentation,” whereby yeast oats on top of the wort (grain mashed in hot water), which is exposed to oxygen and kept warm. Oxygen and warmth persuade yeast to produce spicy, astringent flavor compounds called phenols and fruity, oral compounds called esters that are desirable in beer but not in bread. Lagers, on the other hand, undergo “bottom fermentation,” where the yeast is kept submerged in the low-oxygen environment at the bottom of the wort at colder temperatures, which causes the yeast to produce fewer phenols and esters, so that the breadier yeast and sulfur flavors come forward.
For the most part, no. I started my testing with dark ales, thinking their rich taste would lead to better flavor. The resulting bread had a strange spicy, fruity aftertaste and smelled like beer. Then I tried a light American-style lager. This time, the loaf came out with a distinct “bready” versus “beery” aroma that could fool anyone who had not seen the lager go into the dough. Why is it that the lighter beer produced the better taste? It turns out that the yeast in lagers is treated in a way that closely resembles the way yeast acts in dough, resulting in the production of similar flavor compounds.
Through the simplest of tweaks—less hydration, the addition of vinegar and beer, and a few seconds of kneading—I had a loaf of bread that both looked and tasted incredible.