When someone I respect as much as Peter Reinhart claims to have discovered a technique that “has the potential to change the… bread landscape in America,” it’s worth taking notice. In The Bread Baker’s Apprentice Reinhart writes:

[Delayed fermentation] has the potential to change the entire bread landscape in America. I’ve begun teaching it to my students…. Within the next few years I fully expect to see variations of this method appearing in both artisan bakeshops and at the industry level. It is the next frontier in breads. When we deconstruct the process, it takes us beyond fermentation, actually beneath fermentation, down to the level of enzymes. It is the enzyme that serves as the catalytic converter, freeing up the sugars that are bound up in the complex starches of flour. The delayed-fermentation technique, revealed to me by Gosselin, and intuited by many others without knowing why, is all about how enzymes affect fermentation and release flavor. (24)

Delayed fermentation, also called retarding, is the method of making bread by mixing the dough and refrigerating it overnight or longer. The dough is then baked following the usual method, as if it had been  made fresh. Most of the recipes in The Bread Baker’s Apprentice take advantage of this technique in some form. By the time he published Whole Grain Breads: New Techniques, Extraordinary Flavor, he was even more convinced of the virtues of delayed fermentation, particularly for fussy whole-grain breads:

The delayed fermentation method used in this book capitalizes on enzyme activity while also incorporating the parallel drama of the living microorganisms: yeast and bacteria. It uses a new approach to draw out the full flavor potential of the grain, delaying much of the fermentation until after the enzymes have done their work. (Conventionally, fermentation begins right away, so enzyme activity happens concurrently with fermentation.) This new approach addresses the whole grain baking puzzle by combining several distinct entities to create a dough that performs well and yields a loaf of bread with wonderful flavor. (46)

I have learned more about bread from Peter Reinhart than any other author in print; the methods he presents in these two books are the basis for the way I make bread. Many times have I made bread over two or three days, following his delayed fermentation method. I’ve never had cause to complain about the resulting bread.

But as I have learned to rely less directly on recipes in my bread baking and start to improvise and develop my own techniques, I have begun to wonder if delayed fermentation is worth the trouble: it means you have to plan to make bread over a day in advance and that you have to somehow clear enough space in the refrigerator to accommodate a covered mixing bowl of dough (or, even worse, a half sheet). Was the payoff—a supposed marked improvement in flavor—worth this effort?

I decided to do a test. I made a basic bread dough (4 oz starter, 2 tsp sea salt, 1 tsp yeast, 7 3/4 oz water, 10 1/8 oz AP Flour, 2 1/8 oz WW flour) using my handy-dandy bread calculator at 68% hydration; my standard level of wetness: not overly dry but manageable. I kneaded the dough for ten minutes. I made my first batch of the dough at 5 PM on Saturday. I started an identical batch around noon on Sunday: at 12:15, the second batch was ready to go through its first rise and I took the refrigerated batch from the night before out of the refrigerator so it could begin its first rise. At this point, the doughs looked almost identical, although up-close the delayed-fermentation dough had a smoother surface, possibly from its night of rest.

After three hours of rising time, neither dough had quite doubled but there were signs of yeast activity. The air pockets visible through the bottom of my glass bowls were more evenly-distributed than in the delayed fermentation dough, which had several large air pockets. By this time the doughs’ temperatures had more or less evened out; the non-delayed dough was a few degrees warmer. Shaping the doughs into boules and transferring them to proofing baskets I could feel no significant difference.

After 90 minutes of proofing in which both doughs seemed to grow evenly in their respective baskets, I was ready to bake. I had my oven set to 475°F and baking stone in place, with a sheet pan ready to serve as a steam distributor. As I turned the doughs out onto my peel, I noticed that the delayed-fermentation dough held more tightly to the shape of the proofing basket; the fresh dough was more apt to spread out. In order to tell the breads apart after baking, I labeled them with an F (for fermented — both doughs are actually fermented and very little extra fermentation happens in the refrigerator since yeast are nearly dormant at this temperature; as of baking time, I did not realize all this) and a U (for unfermented).

Into the oven for 30 minutes, they emerged nicely brown and crusty.

In the baked breads the differences between the two doughs are more apparent: the delayed fermentation bread was slightly taller and less spread out – it very nicely matched the oval shape of the basket. The fresh dough had spread out a lot more and was consequently a bit squatter. Cutting the loaves open revealed starker contrasts: while the delayed-fermentation dough had small holes from trapped gases, the fresh dough bread had a more even texture.

Cosmetic differences aside, what about the taste? Expecting great things, Martha and I eagerly sat down with our two slices of bread. I took a bite from the delayed fermentation dough first. It was good! Next, I tried the fresh loaf. It was also good! The flavor differences, if they existed, must have been subtle, because to my coarse palate these breads tasted the same. Martha, who is much more taste-sensitive than I, agreed. There was also very little textural difference: I thought I could detect more crunchiness in the fresh crust and more chewiness in the delayed fermentation crust. The delayed-fermentation bread also felt slightly denser.

Consulting additional references after baking, I saw that the differences I did detect — the tighter rise and larger gas pockets of the delayed-fermentation dough — were scientifically predictable results. As Shirly O’Corriher writes in Cookwise:

Reduces ovenspring and loaf volume Chilling for a period as long as overnight reduces ovenspring (oven rise) and reduces loaf volume. Alcohol that was made by the yeast from the beginning changes to a gas in the hot oven and is a contributor to ovenspring. During a lengthy cold period, alcohol simply evaporates from the surface of the dough and some goes to vinegar. So, there is a reduced amount of alcohol to contribute to ovenspring…

Opens texture slightly Chilling dough retards the action of yeast and influences texture, though much more subtly than breadmaking method. The air-bubble nuclei worked into the dough during mixing and kneading are enlarged by gases from the yeast. Initially these tiny bubbles contain air, which is mostly nitrogen (78 percent) and oxygen (21 percent). Immediately the yeast uses up the oxygen so you have mostly nitrogen in these tiny bubbles. The gases from yeast are essentially carbon dioxide, and they enlarge the same bubbles. The longer the dough rises, the more carbon dioxide there is in the enlarged bubbles.

After a long rise, the bubbles contain a very high percentage of carbon dioxide and a very low percentage of nitrogen. The longer the rise, or after multiple punch-downs, the higher the percentage of carbon dioxide. After a very long rise or two or three shorter rises, the bubbles contain essentially 100 percent carbon dioxide.

Carbon dioxide dissolves very well in cold water. So, if the dough is chilled, the carbon dioxide in a great number of bubbles will completely dissolve. The dough now has fewer bubbles for the gases from the yeast to enlarge. When the dough is warm again, the yeast puts much more gas into these fewer bubbles, creating bigger bubbles and a coarser bread. (29)

Neither of my breads had the open texture that I prefer, but of the two the delayed fermentation bread had larger air bubbles, giving it a slightly more rustic appearance.

Harold McGee in On Food and Cooking concurs about the effects of delayed fermentation, with a little less detail on the process:

In addition to giving the baker greater flexibility, retarding has useful effects on the dough. Long, slow fermentation allows both yeasts and bacteria more time to generate flavor compounds. Cold dough is stiffer than warm dough, so it’s easier to handle without causing a loss of leavening gas. And the cycle of cooling and rewarming redistributes the dough gases (from small bubbles into the water phase, then back out into larger bubbles), and encourages the development of a more open, irregular crumb structure. (539)

Both McGee and O’Corriher (just before the quoted section) agree with Reinhart that the enzyme activity engendered by delayed fermentation improve the flavor of bread. In this one test, I didn’t find that to be the case; at least not noticeably so. If I haven’t planned ahead, I know I can still making pretty flavorful bread without relying on a rest in the fridge. But that doesn’t mean I’m ready to give up on delayed fermentation. The benefits I did observe — the open texture (which can be further improved by alterations to my kneading technique) and the firmer, easier to manipulate dough — are a big step in the direction of my search for ideal loaves of bread. And as long as you’re capable of a little advanced planning, delayed fermentation can actually help with time management, since it allows you to mix the dough on one day and complete the baking on another. As long as there’s room in the fridge.

Since I don’t have to work on the weekend I avoid that nagging pressure to interact with the other humans by baking bread. The rhythm of bread baking is such that it is both leisurely and consuming.

The first thing I did in the morning was cover 10 grain hot cereal mix with boiling water as the starter for the multigrain bread that was featured in Cook’s Illustrated a year or two ago. This is my go-to basic bread to have around the house, great for toast and sandwiches. The crumb is not so dense as a lot of multigrain breads thanks to copious amounts of white flour and instant yeast makes for a predictable rising schedule. The flavor is earthy, as multigrained breads should be, and tinged with honey. This recipe is simple, not very time consuming and reliably produces a pretty pair of loaves.

With that out of the way, I went to the store to get more flour (3 lbs was not going to cut it). I had refreshed my sourdough starter the day before and left it out all morning to bubble and grow. I needed to make pitas for dinner and I also wanted to make a couple of european-style breads. I like a loose crumb so I was aiming for about 66% hydration, and I wanted to use a pound of my starter, mostly to facilitate storing the rest of it in the fridge. I was figuring I wanted about 3# of final dough for my european breads and at least 2# to be able to make 8 four ounce pitas. Math time!

A pound of starter was 8oz water and 8oz flour. I wanted to end up with about 80 oz of dough and I was going to have 2/3 as much water as flour (and I like round numbers that my scale can handle) so I planned to add 40 oz of flour. That gave me 48 oz of flour total, 66% of which is 31.68, which I rounded to a nice even 32 (which meant I would add 24 additional ounces).

With my formula figured out, I proceeded to mix and knead the dough, adding a good amount of salt. With wild yeast rising times are pretty unpredictable, and salt inhibits rising, so I had no idea when the dough was going to double in size. I set out to work on some other things.

Around 5 p.m. I got tired of waiting and decided to weigh and shape my dough. I recently discovered that weighing individual dough pieces so they are the same really helps with the uniformity of the final product (durh) so I got out my scale. 1.5# of dough went to a boule which I left to proof in a heavily oiled and floured glass bowl. Three .5# pieces got rolled into baguettes and placed on my couche to grow. The remaining 8 4 oz pieces (plus a small leftover piece) I shaped into small boules to rest for pitas.

Mostly because we were hungry I rolled out the pitas within an hour of shaping the dough even though they had not grown much at all. Apparently, that is not very important with pitas. Here they are in the oven:

That is what we like to call pocket-city. Below you’ll see them in a nice little stack (wrapped in a towel to keep them warm and chewy):

After dinner I decided it did not make much sense to wait up till all hours for the rest of my breads to proof in the meantime wasting all that heat in the stone in my oven, so the baguettes were transferred to a floured peel, scored and baked. For the record, they were underproofed, but I think they turned out fine.

Lastly, the boule. I apparently did not flour my bowl enough because this bread stuck to the sides and did not want to land on the peel. When it finally did, it was horribly deformed. It is pretty annoying to spend this much time on a piece of bread only to have some stupid mistake destroy everything you worked and fought for; much can be learned about life from bread baking, evidently. I made a feeble attempt at scoring this bread artfully to make the mistakes look intentional and then slid it onto the stone. I was surprised by a good amount of oven spring, but upon eating the bread the next day found dense areas that suggested underproofing. Great crumb for the most part though thanks to the relatively high hydration!

And so passed another Saturday.