Clearly Robertson’s loose, novelistic approach to the whole notion of baking was driving a certain kind of person absolutely crazy. And then all at once, I was buoyed by this thought: I am not that kind of person! This was the moment when I decided I was ready to jump in. It was time to start my starter.
Considering what it is (a living thing) and what it does (leaven and flavor a bread), the instructions for starting a sourdough culture could not be much simpler. Take some flour, preferably a fifty-fifty mixture of white and whole grain, and mix it by hand in a glass bowl with some warm water until you have something that feels like a smooth pancake batter. Cover the bowl with a cloth and leave it in a cool spot for two or three days. If by then nothing has happened, wait a few more days and check it again.
Simple maybe, but not foolproof: My first attempt at starting a starter didn’t start. After a week of inactivity, the batter separated into a layer of cement beneath a layer of perfectly clear water. It remained absolutely inert and odorless. I did some reading to figure out what was supposed to be happening but wasn’t. Wild yeasts and bacteria were supposed to find their way into the batter, take up residence, and eventually organize themselves into a more or less stable microbial community. Curiously, none of the authorities I consulted could say with certainty just where these yeasts and bacteria came from or how they got here, if and when they did. They might already be in the flour, or on my hands (which is why Robertson suggests mixing by hand), or in the air. Indeed, one of the many mysteries of sourdough culture is the origin of its resident microbes, some of which—like the all-important Lactobacillus sanfranciscensis—have never been found anywhere on earth except in a sourdough bread culture.* This suggests these “wild” microbes are actually in some sense domesticated—dependent upon us (and our love of bread) to create and maintain their highly specialized ecological niche. But either I had failed to create a niche to their liking or the bugs had failed to find it, because even after two weeks my starter was as lifeless as plaster.
I started a new culture, but this time after mixing it I gave the bowl an hour or two outside in the sun, hoping to snag some airborne microbes. I also gave it some vigorous stirs whenever I remembered to, in order to work some oxygen into the mixture. Within a week, my batter was showing tentative signs of life in what seemed very much like an instance of spontaneous generation: proposing the occasional bubble and giving off a faint, not-unpleasant scent reminiscent of rotten apples. But a couple of days later, the odor had taken an unpleasant turn, veering toward strong cheese or worn sock. Something bacterial was definitely afoot. So, following Robertson’s directions, I discarded 80 percent of the starter, more or less, and fed what remained a couple tablespoons of fresh flour and warm water. Within a day, the bowl was burbling contentedly. I had a starter! Whether it was lively enough to leaven a dough, I wasn’t yet sure, but it was definitely alive.
A couple weeks later, when my starter seemed to be settling into a predictable daily rhythm, rising in the hours after its morning feeding and then subsiding again overnight, I embarked on my first loaf of naturally leavened bread.
Step one is to turn a small amount of the starter into a “sponge” or “leaven”—basically, use it to inoculate a much bigger mass of sourdough culture, which in turn would inoculate and leaven the entire dough I would mix the next morning. Placing a glass bowl on my new (digital) kitchen scale, I zeroed it out and added two hundred grams of flour (the same fifty-fifty mix I used to feed my starter), then an equal amount of warm water. To this I introduced a heaping tablespoon of my starter, mixed it all together, covered the bowl with a dish towel, and went off to bed.
I faced a test in the morning, one that many of the participants in the chat rooms and discussion groups on line had struggled to pass. To wit: Would this so-called sponge take on enough air overnight so that, when dropped into a bowl of water, it would float? If instead it sank, that would indicate there wasn’t enough microbial activity to leaven a loaf of bread.
The question would be decided while I slept: There was nothing to do now but wait, while my culture either did or failed to do its fermentative thing. Already this felt like a radically different way of “cooking” than I had done up to now, but not because it was any more exacting or precise. To the contrary: I’d delegated my accustomed kitchen powers and responsibilities to this invisible cohort of unidentified microbes.
Up to now, most of the things I’d cooked and ingredients I’d cooked with had been dead, after all, and therefore more or less tractable. The raw materials responded in predictable ways to physical and chemical processes that I controlled; whatever did or didn’t happen to them could be explained in terms of either chemistry or physics. Obviously those laws play an important part in baking, too, but the most important processes unfolding in a naturally leavened bread are biological. Though the baker might be able to influence and even manage those processes, “control” would be far too strong a word for what he does. It’s a little like the difference between gardening and building. As with the plants or the soil in a garden, the gardener is working with living creatures that have their own interests and agency. He succeeds not by dictating to them, as a carpenter might to lumber, but by aligning his interests with theirs. To use a metaphor a little closer to Chad Robertson’s frame of reference, what the baker does is a little like the surfer’s relationship to the wave.
This lack of control has never sat well with our species, which probably explains why the modern history of bread baking can be told as a series of steps aimed at taking the unruliness, uncertainty, and comparative slowness of biology out of the process. Milling white flour was the first such step. Whole-grain flours, as I would soon learn, are much more complex and biologically active than white flour. That’s because white flour consists chiefly of dead starch, whereas the germ and the bran removed in milling it contain living cells. Whole grains teem with enzymes and volatile oils that make their flours more perishable and fermentation more difficult to manage.
Around the same time that the advent of roller mills made white flour widely available in the 1880s, the introduction of commercial yeast gave bakers an even more decisive gain in control. Now, instead of having to rely on an unruly community of unidentified fungi and bacteria to leaven bread, as had been the case for thousands of years, they could enlist a single species of yeast to do the job on command. Called Saccharomyces cerevisiae, this species had been (as its name suggests) found in beer, selected over countless generations, and optimized for the role of putting gas in dough. Commercial bread yeast is a purified monoculture of S. cerevisiae, raised on a diet of molasses, then washed, dried, and powdered. Like any monoculture, it does one thing predictably and well: Feed it enough sugars and it will promptly cough up large quantities of carbon dioxide.
Though commercial yeast is alive, its behavior is linear, mechanical, and predictable, a simple matter of inputs and outputs—which is no doubt why it so quickly caught on. S. cerevisiae can be counted on to perform the same way everywhere and give the same results, making it supremely well suited to industrial production. Yeast could now be treated simply as another ingredient rather than as a locally variable community of organisms in need of special care and feeding. In fact, as microbes go, S. cerevisiae is notable for not playing well with others, especially bacteria. Compared with wild yeasts, commercial yeast cannot survive very long in the acidic environment created by lactobacilli.
While scientists have known about yeast since Louis Pasteur first identified it in 1857, the intricate microbial world within a wild sourdough culture like mine was a complete mystery until fairly recently—and remains at least a partial mystery even today. In 1970, a team of USDA scientist
s based in Albany, California, collected samples of sourdough starter from five San Francisco bakeries and conducted a kind of microbial census. Why San Francisco? Because the city was famous for its sourdough bread. The scientists were hoping to identify the local microbes responsible for the bread’s distinctive qualities. Their landmark 1971 paper, “Microorganisms of the San Francisco Sour Dough French Bread Process,” helped to spur a revival in naturally leavened breads and almost single-handedly established the (albeit still minor) field of sourdough microbiology.*
The USDA team discovered that unlike what happens in the straightforward fermentation performed by S. cerevisiae, no single yeast species was responsible for what takes place in a sourdough culture. Rather, the process depended on a complex, semisymbiotic association between a yeast (Candida milleri†) and a previously unknown bacterium. Assuming—wrongly, as it turned out—that the bacterium they had identified was unique to San Francisco’s famed sourdoughs, they named it Lactobacillus sanfranciscensis. It has since been found in bakeries all over the world. Oh well.
Though not exactly dependent on each other, the yeast and the bacteria are ideally suited to living together. Each microbe consumes a different type of sugar, so they don’t compete for food. And when the yeasts die, their proteins break down into amino acids that the lactobacilli need to grow.
At the same time, the lactobacilli produce organic acids that shape the environment in ways agreeable to C. milleri (which is acid-tolerant), but disagreeable to other yeasts and bacteria. L. sanfranciscensus also produces an antibiotic compound that prevents competing microbes from gaining a toehold in the culture, but which doesn’t trouble C. milleri in the least. Thus the sourdough culture defends itself from colonization by outsiders. This biochemical defense is a boon to us as well, since it extends the shelf life of the bread.
Perhaps the USDA team’s most important contribution was to demonstrate that a sourdough culture functions as a kind of ecosystem, with the various species performing distinct roles that lend stability to the culture over time. Once established, the system exhibits more cooperation than competition, so that no one organism ever dominates. Subsequent research in other parts of the world has greatly expanded the list of species found in sourdough cultures—at least twenty types of yeast and fifty different bacteria—but most of them seem to fall into similar niches, organize themselves into similar relationships, and perform similar functions. Same play, different actors. Presumably these yeasts and bacteria coevolved with one another, which might explain why many of them have been found nowhere except in sourdough cultures, their “natural habitat.” Which in turn suggests these microbes probably coevolved with us: Their culture depends upon our culture of bread making, and (until recently) vice versa.
In the microuniverse of a sourdough culture, the baker performs in the role of god, or at least of natural selection. It may well be that the requisite microbes are everywhere, but by shaping their environment—the food and feeding schedule, the ambient temperature, the amount of water—the baker, wittingly or unwittingly, selects which microbes will thrive and which will fail. Frequent feedings and warm temperatures tend to favor the yeasts, for example, creating an airier, milder loaf, whereas skipping meals and refrigerating the culture favors the bacteria, leading to a more acidic environment, and a more strongly flavored bread.
“Baking well really comes down to managing fermentation,” according to Robertson. The flavor and quality of a naturally leavened bread depends to a great extent on how skillfully the baker governs this invisible microbial world. And if the baker fails to care for his culture? It may take awhile, but once the sun of his attention goes dark, the culture eventually dies.
The morning after starting my sponge, I woke up eager to head down to the kitchen to see what, if anything, had happened overnight. When I’d mixed the stuff the night before, the heavy paste of flour and water filled a two-cup measuring bowl halfway to the top. Incredibly, it had doubled in volume overnight, and I could feel it had lightened considerably, achieving a consistency reminiscent of marshmallow. Through the glass I could see that the paste had become a gassy foam, shot through with millions of air bubbles. I felt certain it would float.
So into a larger bowl I measured out the quantity of warm water called for in the recipe (750 grams), and then, using a spatula, scooped out the sponge. It slid into the warm bath and then bobbed up to the surface of the water like a raft, buoyant. I was in business! Next I added 900 grams of white flour and 100 grams of whole-wheat flour. I mixed everything together by hand, squeezing the flour and water through my fingers to make sure there were no unhydrated lumps of flour—what bakers call “chestnuts.” The result was a dough wetter than anything I had ever worked with before. This promised to be a challenge.*
Before any salt is added, the dough gets to rest for twenty minutes or so. Called the “autolyse,” this period gives the flour a chance to fully hydrate, the gluten to begin to swell and get itself organized, the enzymes to begin cleaving complex starches into simpler sugars, and the fermentation of those sugars to commence. Salt acts as a check on all these processes, which in its absence would proceed too rapidly. The goal is a long, slow fermentation in order to build maximum flavor. As one nineteenth-century cookbook put it, salt serves as the bridle on the wild horse of fermentation.
After I mixed in the twenty grams of salt, the dough felt dull and sticky to the touch—a wet, heavy, lifeless clay. I covered the bowl with a towel and went back to work, setting my phone to alert me in forty-five minutes. “Bulk fermentation” was now under way—a period of between three and four hours during which the principal development and fermentation of the dough takes place.
A complex drama unfolds during the bulk fermentation, one that the baker cannot see but can infer by the evolving texture, smell, and taste of his dough. Within the dough, a spongiform structure is taking shape, a three-dimensional lacework of air. The structure is the result of two separate developments—one chemical in nature, the other biological—that in a dough made from wheat flour happen, fortunately for the panivore, to coincide and intersect just so.
The chemical development is the formation of gluten (the word means “glue” in Latin), an interesting if somewhat problematic substance that is found primarily in wheat, and to a much lesser extent in rye, another species of grass. To be precise, gluten as such is not found in wheat itself, but, rather, its two precursors are, the proteins gliadin and glutenin, which when moistened in water combine to form the mesh of proteins known as gluten. Unprepossessing on its own, each of these proteins contributes a different but equally important quality to a bread: extensibility on the part of gliadin, and elasticity on the part of glutenin. As in the fibers of a muscle, these qualities exist in a productive tension, the former allowing the dough to be stretched and shaped, while the latter impels it to bounce back to something close to its original form. In fact, the Chinese call gluten “the muscle of flour,” and all bakers speak in terms of a dough’s “strength” or “weakness,” qualities that correspond to the amount of gluten in it.
The pliable yet rubbery properties of gluten make it the ideal medium for trapping air, which happens to be the crucial by-product of the second, biological development under way in a wet mass of fermenting dough. While the gluten network is forming and gaining strength, the community of yeasts and bacteria introduced by the starter are dining on starches “damaged” during milling, when some of them are broken into sugars. Various enzymes (some of which are present in the flour, others produced by the bacteria and yeasts) go to work on the undamaged starches and proteins, breaking them down into simple sugars and amino acids to feed the microbes. Thus fed, the bacte
ria proliferate, producing lactic and acetic acids, which help to strengthen the gluten while contributing new flavors. And, most important of all, the yeasts are busy transforming each molecule of glucose they consume into two molecules of alcohol and two of carbon dioxide. The carbon dioxide gas, which is a by-product of alcohol production, would simply escape into the atmosphere if not for the rubbery matrix of gluten, which stretches like a balloon to contain it. Without the extensible and elastic gluten to trap the carbon dioxide, bread would never rise.
The properties of gluten have commended wheat to humanity since the Egyptians first recognized what it could do. Before that, wheat was just one edible grass among many, part of a crowded field that included millet, barley, oats, and rye and, later, corn and rice. Barley barely registers in our eating lives today, but before the invention of bread it was just as important a staple food in the West. It grows more quickly than wheat, and in more places, from the tropics to the Arctic Circle. Highly nutritious, it was the food of choice of the Roman gladiators, who were in fact called hordearii, the barley eaters. But though barley made nourishing porridges and flat breads (and beer, as I would discover), no amount of leavening could raise it off an oven floor.
Wheat’s own ancestors couldn’t rise, either. Einkorn, the earliest known form of wheat, has been cultivated in southeastern Turkey for nearly ten thousand years, but eaten mostly as a porridge or brewed as beer. It has too much gliadin and not enough glutenin to trap fermentation gases. The ancestry of bread wheat is tangled and still a subject of botanical debate, but it took thousands of years of accidental crosses and mutations before a civilization-altering curiosity showed up in a farmer’s field somewhere in the Fertile Crescent: a stalk of wheat with big fat seeds that just happened to contain the proteins gliadin and glutenin in just the right proportions. Gluten, and with it the possibility of leavened bread, had come into the world.*