The Dough Before the Dough

On Yeasted Preferments

I still need to sit down and outline for you all what I hope to cover in Breaducation in detail (especially in light of your responses to my query last week), but for now I’ll say that my goal for it is to have it be a comprehensive overview of the fundamentals of bread baking. Given that it is going to be just a single volume made by a one person with limited time to produce it, it cannot cover everything in the same level of detail. But I do want it to at least touch upon every aspect of the bread baking process, even if only to send you to the right place should you want to delve further into one area or another. (It is going to be as much a reference manual as a cookbook and will be chock full of citations and recommendations for further reading.)

---

One subject I intend to cover in great detail is the use of preferments. A preferment is the catch-all name for any portion of a dough that is fermented ahead of the final mixing of the dough. Not all breads include a preferment, of course; doughs without them—those in which all of the fermentation takes place after the final dough has been created—are known as “straight” or “direct” doughs. Preferments can be made using sourdough cultures or commercial yeast; because sourdough is a perpetual process, it makes no sense to talk about a “straight” sourdough—some portion of the dough is always fermented ahead of the final mix. Sourdough fermentations are a big topic, and one I’ve covered at length here already. I’ll definitely have more to say about sourdoughs here again, but for the sake of today’s post, I’m going to limit myself to yeasted preferments (also a big topic).

There are a few categories of yeasted preferments, but the differences between them are less important than their effect on a dough, so let’s start with the numerous reasons why a baker might choose to use a preferment.

Why use a preferment?

The first reason is flavor. Long fermentations of any kind encourage the formation of acids and other compounds that lend breads complexity and depth of flavor; adding a preferment extends the overall fermentation time (backwards, in this case). Yeasted preferments can add milky, buttery, and/or tangy flavors to a dough otherwise lacking one.

Then there’s leavening power. Preferments usually start with a tiny percentage of yeast which proliferate as the dough ferments; later on, when the preferment is added to the final dough, it has far more leavening potential than that initial pinch of yeast would have had by itself. Many formulas, however, do not rely on the preferment for leavening, and include another, larger hit of yeast during the final mix. But there are some formulas that get their leavening potential from the preferment alone. (In that case, they tend to use a relatively large percentage of preferment in the final dough.)

Next is structure. Just as doughs develop structure as they sit—through the creation of gluten networks, along with the production of acids, which strengthen them—preferments contain developed gluten, so adding them to a dough gives it structure it wouldn’t otherwise have. (Preferments are also “pre-developed,” in a sense.) Using a preferment reduces the amount of mixing or folding necessary in the final dough.

Then there is extensibility. Preferments also contain compounds that encourage extensibility, or the ability for the gluten in a dough to elongate without springing back (extensibility is the opposite of elasticity, the ability of a dough to resist elongation). I don’t know for sure yet which compounds are responsible for this effect, but I’m guessing that it is glutathione and L-cysteine, two related molecules produced by yeast that are known to reduce the strength of gluten bonds. There’s always a delicate balance to strike between elasticity and extensibility—structure and strength versus workability—but it can be useful oftentimes to encourage the latter over the former, especially in the case of products that demand heavy dough manipulation, like croissants or pretzels (two breads I’ve been working with a lot recently).

Then there is convenience. This is one of those things that applies more to bread baking in a professional context than at home. All of the benefits I’ve already mentioned can be achieved without a preferment by simply extending the fermentation of the dough forward in time rather than backwards—using small amounts of yeast and/or fermenting it at lower temperatures, each of which encourage a long, slow fermentation. At first glance it would seem like there’s little advantage to using a preferment, since both approaches require the same amount of time to achieve their goal.

Unless you are working in a bakery and making multiple products that all could benefit from the inclusion of a preferment, that is. In that case, instead of making the dough for each product and long fermenting each of them separately, you can make a single, large batch of a preferment, let it proof, and then divide it among the individual doughs. (Preferments are also simple and quick to create, so they can be whipped up easily at the end of a shift, ready to use when the next baker arrives to start making doughs.)

But also: Preferments retain their positive attributes for awhile, particularly when kept refrigerated. This is the sort of convenience that is thus useful for pros and home bakers alike: You can mix a preferment on one day, set it into the fridge, and then use it whenever you are ready to mix the final dough the next day. Within reason, of course—preferments degrade over time, and most are used within 24 hours of formation, even when refrigerated. (You can also freeze a preferment, which will allow you to hold it even longer. Which attributes are retained well in the deep freeze and which are lost is something I need to do some experimenting around; that said, it’s safe to say that the flavor and acid-related structural benefits of a preferment are likely pretty stable to freezing.)

Finally, there’s economy. Again, this is more applicable in pro settings than at home, but some preferments are simply leftover or otherwise unusable doughs that get “recycled” into a new dough, avoiding waste. (Sourdough “discard” recipes are a common form of dough recycling in a home-baking context.) One baker I know freezes the inevitable scraps from his croissant doughs to use as a preferment in his brioches later on, for example.

Kinds of Preferments

While there are a handful of differently named preferments, there are really just two distinct kinds of preferments: those that are composed of a simple mixture of flour, water, and yeast, and those that are closer or identical in composition to the final dough (meaning they also include enrichments like butter, eggs, or milk, and/or salt). The former category includes both poolish and biga; the latter refers mainly to something known as pâte fermentée, or “old dough.” (A sponge sometimes refers to a type of preferment that is halfway between these styles.)

While many bakers use poolish or biga preferentially, the differences between these two preferments are more semantic than practical. A poolish is a usually a high-hydration mixture made up of equal parts water and flour, closer to a batter than a dough. (The word poolish is sometimes said to be a derivation of “Polish,” but there’s no evidence that the technique actually originated in Poland.) Biga—the Italian term for a preferment—often refers to a stiffer preferment closer in texture to a dough than a poolish (something along the lines of 60% hydration), but there are plenty of high-hydration biga formulas out there too, so it is a distinction without a difference. And there’s little practical difference between the two either once you account for the difference in hydration. A lower-hydration preferment might be slower to ferment than a higher-hydration one, but—provided they are both fermented to the same degree—the effect each has on the final dough is essentially the same. (I haven’t actually tested this yet, but I’ve also never seen anyone argue otherwise either.)

One impractical difference between a stiff preferment and a loose one is that the latter is much harder to incorporate into a dough, particularly when doing so by hand; a poolish is usually quite easy to stir into the liquid portion of the dough just before mixing.

A sponge is either a non-technical term for any yeasted preferment (the sort of thing you’d find in old fashioned bread recipes for home bakers), or the name for a preferment that includes some of the other components of the final dough, like butter, milk, or eggs; it usually does not include salt, though it might. I’ve yet to see any practical reason to use this style of sponge rather than a simple preferment, though I could see instances where it would be hard to leave enrichments out, especially when they account for a high percentage of water in the dough (like a brioche where most or all of the liquid comes from the eggs and not water).

Pâte fermentée is unique among preferments in that it contains all of the components of the final dough, including salt, and has the same hydration. That’s because it is usually “made” by simply taking a portion of “yesterday’s” dough and adding it to the next one (hence its other name, “old dough.”) This can be a way to recycle excess or otherwise unusable doughs (scraps, for example, or slightly overproofed doughs), though it can also be done deliberately, in order to gain all the other benefits of a preferment. In the latter case, each batch of dough makes both the breads it is destined for and a seed for the next batch. This allows a baker making the same products day-in, day-out to forgo making a separate preferment.

That said, you can also make a pâte fermentée from scratch with the sole purpose of using it in a dough later on. (It is usually not necessary to use the full amount of yeast nor to develop the gluten to the same extent as you’d do in the final dough.) I’m not sure yet why you’d choose to do this rather than opt for a simple preferment, but I have seen formulas that do so. It’s possible that the lower hydration and the presence of salt could make a pâte fermentée more stable and long-keeping than a poolish.

My (current) preference

I am a poolish guy, at least for the time being. Like I said, there’s little practical difference between a stiff preferment and a loose one, and there’s no good reason (that I know of yet) for a home baker to use pâte fermentée over a simple preferment, so I opt for the simplest and quickest to produce: equal parts water and flour, plus a pinch of yeast. Not only does it take less than a minute to weigh and mix, it’s also far easier to incorporate into the final dough later on.

How much preferment to use?

Before answering this question, it’s important to distinguish between preferment percentage and prefermented flour percentage, which are related but not identical. Preferment percentage refers to the weight of preferment, relative to the total amount of flour in the formula. (I’ve seen recipes that relate the preferment to the weight of flour in the final dough, which is not useful or logical.)

Prefermented flour percentage describes how much of the total flour in the dough goes into/comes from the preferment; in other words, it tells you how much of the flour in the overall formula is fermented ahead of time. Thus a final dough that has 1000g of flour in it and 400g poolish (made from 200g flour and 200g water), contains 20% prefermented flour (200g/1000g = 0.2) and 40% poolish (400g/1000g = 0.4).

The amount of prefermented flour in formulas can vary from 10 to as much as 40%, but most recipes fall on the lower-middle end of this range. (The sweet spot for many formulas is ~10% prefermented flour/~20% poolish.) You want to use enough preferment to gain the desired effects and make the effort and time to produce one worthwhile, but not so much that the bread is aggressively sour or weak in structure from an excess of acids. As I mentioned above, those formulas that use higher percentages of prefermented flour might do so to push the fermentation heavily, and in this case would not include any additional yeast.

How much yeast to use in a preferment?

Most preferments use a tiny amount of yeast relative to flour, something on the order of 0.02 to 0.3% (when using dry yeast). The amount varies relative to the length of time the preferment needs to hang out at room temperature—more for a quick, same-day fermentation, less for a long overnight rise.

But you can also use a hybrid approach: from zero to a couple of hours at room temperature, followed by an overnight stay in the fridge. The advantage of this technique is that the poolish is ready to use first thing in the morning, and stays that way for the entire day.

---

A basic poolish formula

Here’s my standard overnight poolish formula. This can be scaled up or down as needed (this makes 400g):

200g high-protein all-purpose or bread flour (100%)

200g room-temperature-ish water (100%)

0.5g (1/8 teaspoon) instant yeast (0.25%)

Combine the ingredients in a bowl or straight-sided container, cover, and let sit at room temperature until slightly bubbly, 2 to 4 hours, then transfer to the fridge for at least 12 hours. It should be bubbly, domed, and about doubled in volume at this point; if necessary, let it sit at room temperature until it is.

If you need to use the poolish on the day it is made, let sit out at room temperature until doubled in volume, 3 to 5 hours. Use immediately or refrigerate for up to 6 hours before using.

---

How do you know when a preferment is ready for use?

Much like a well-ripened sourdough levain, a ready-to-use preferment should be expanded in volume (to about twice its starting volume), bubbly, slightly domed on top, and aromatic, with milky and/or tangy aromas. It should pass the “float test." It should not smell sour, and it should ideally not have risen and then collapsed.

How long is a preferment good for?

This is a question I haven’t got a good answer for yet. Most recipes have you use the preferment within 24 hours of its formation, especially if refrigerated, but I don’t know how long you can stash one away before it’s no longer going to perform as it should. I’m pretty sure that how long a preferment “keeps” depends upon the benefits you are looking for from it—the flavor benefits are probably more stable than leavening ones, for example, and extensibility might actually increase with time, with the accumulation of beneficial compounds.

(Now that I’ve gotten this far into this piece, I realize that one of my “projects” for the book is—as I’ve done here—to lay out what I know while also sorting out what I’ve yet to understand. That will tell me what experiments remain to be done.)

—Andrew