The Fresh Loaf

News & Information for Amateur Bakers and Artisan Bread Enthusiasts

Effects of fermentation and soaking question

venkitac's picture

Effects of fermentation and soaking question

I don't know whether this question is geeky, but anyway: most lean dough recipes I've seen call for a poolish/biga fermented for a few hours (or retarded overnight) and then mixed in with a significant quantity of flour (let's say half poolish and half fresh flour or some such) to create dough, which is then again fermented for a small number of hours and then baked. I've a few questions about the effects of these:

- What if I make the entire dough, yeast and all, ferment for a couple of hours (or not at all), then retard overnight, de-chill and bake, versus create poolish, mix retarded poolish with fresh flour to create dough and bake? (The "Gosselin Method" in BBA is like the former). Is there a difference in flavor between these methods?

- Instead of using a preferment, what if we just use the entire flour quantity called for as a soaker, and then mix the soaked dough (unsalted and unyeasted dough) the next day with yeast and salt and then ferment for 3 hours and bake? I believe this is the original  Gosselin method. What would be the flavor difference for this method vs the above two?

- I've read books where they say "too much preferment sours the bread". From my minimal amount of baking, this doesn't seem true if you retard the preferment quickly. In fact, the Gosselin method in BBA is basically all preferment. Does that sound right, or is my taste too undeveloped?:)

Essentially, the first two questions boils down to this: we can soak all the flour overnight, or some large percentage of it. We can add yeast to all the dough overnight, or some large percentage of it. If we're retarding, we can retard immediately after mixing the dough(if we do the BBA Gosselin)/poolish(most other things) or we can retard a couple of hours later. These 3 variables create 8 different combinations, and I'm wondering is there a marked difference between the various combinations. I suppose a better question would be about the effects of all the variables involved. Bread geeks, help!

I'm not a consistent enough bread baker to try all this and say that "the difference in taste is because of this method vs that method", nor do I have strictly controlled conditions, hence I thought I would ask.

pmccool's picture

If I can say it without sounding like a smart-aleck, how about doing a side-by-side of all three techniques (preferment vs. cold retarded ferment vs. soaker) and let us know what you find out?

I'm not really sure what your results might be, but I'd be interested in learning what happens.


venkitac's picture

I actually want to do exactly that but I'm not very confident of my own baking or tasting skills (as I mentioned at the tail end of my looong post) to do a repeatable experiment to isolate what variable caused what result. I was hoping that someone actually knows all the chemistry to give a scientific explanation. Someone like Debra Wink?:)

If all else fails, I'll try all the combinations and post, but the results might vary because of small variations in whatever I do, which is my fear.  Let's see.

proth5's picture


and an invitation to discussion.

First, if you are not at least a somewhat consistent bread baker, just let me respectfully suggest that you are tying yourself up in knots with your questions.  There are many methods of producing bread and they will all have different results in your hands and under the conditions that you bake.   It might be best to pick a method and see how that works for you and then expand your repertoire.   There is no "scientific" explanation on all this as we are dealing with taste and handling qualities that are somewhat subjective.  This isn't formulaic.  Large scale bakeries will go to great lengths and run many analytic tests to adjust to conditions and flours, but we, as home bakers, must learn to respond to the demands of the living organism with which we work.   Anyway to attempt to answer your question. 

I'm going to assume (since you used "yeast" in your questions) that you are asking about breads with commercial yeast. Some of my answers would be different if we are talking about wild yeasts.

In general, the longer the fermentation time, the more "taste."  In theory, breads made with almost no yeast and fermented for very long times would be the "most flavorful."  However, there are tradeoffs that must be made in terms of time and space constraints (and how the gluten in the flour reacts to a very, very long fermentation) and a number of methods have been devised to coax taste from flour.

The poolish or biga (I prefer the term pre ferment) method takes a percentage of the total flour in the formula at a certain hydration, incorporates a small amount of yeast and allows it to completely ferment normally (not in a cool environment - usually at the ambiguous "room temperature") until it is ripe.  Ripeness is judged by the appearance/aroma of the preferment but is usually done "overnight."  So here is the first set of conditions - the amount of the flour that is in the preferment, the hydration of the preferment, the amount of yeast in the preferment, the ambient temperature and the time.  Pre ferments are generally not "retarded" - that is allowed to ripen in a very cool condition - although some people do it.  It is more typical to control the rate of fermentation in the pre ferment by varying the amount of yeast.

Now the question is how much flour to preferment.  I have seen formulas from very low percentages to very high.  When a very high percentage of the flour is pre fermented, the sugars/starches in the flour can be totally used by the yeast in the preferment and then some diastatic malt may need to be added so that the full amount of yeast in the full amount of flour can ferment properly. If too much flour is pre fermented and malt is not added, the resulting bread will not rise well. Perhaps that is what your author meant about "souring" the bread (or the author may have been referring to natural leaven breads - another topic.)

Once again, the longer the fermentation time, the more "flavor" in the pre ferment.

In general (although I know not always - so don't y'all pounce on me...) a bread that has a pre ferment then goes through a normal fermentation process.  It is not given retarded fermentations after mixing the full dough.  In theory this is because the benefits of long, cool fermentation have been obtained through the pre-ferment.  Or that's my story.

What does this do to the taste?  It adds the taste of long fermentation to some of the flour.  What does that do?  Well, pre fermenting a large percentage of the flour will add more "taste", but what I find is that I face tradeoffs in crumb structure with my hands and my altitude.  I tend to pre ferment a smaller amount of the flour than most folks and do a more thorough bulk ferment.  I have worked and tuned these factors until I have exactly the bread that I want (Note that I say "I want."  I am baking for me and a small group of folk who seem to like what I do.  Great tasting bread for me might not be what you want.  This is a subjective business.)  I make a formula, bake, taste, evaluate, take notes, and then tweak.  You will need to do the same. There is theory and there is practice.  This is a hands on craft.

The Gosselin Method as I understand it (and you seem to refer to it in many different ways) is an interesting variant.  It does a retarded partial fermentation with all of the flour and then adds additional water, salt and yeast.  This gives the benefit of a long, cool partial ferment on all of the flour. Note that I say "partial ferment "on all of the flour.  If a complete ferment was given to all of the flour, when additional yeast and water was added, there would be no food left for the yeast.  The bread would not require additional bulk fermentation time and would not rise well.  The cool temperature and small amount of yeast in this method is what drives the partial fermentation. More extensibility?  Probably.  More "taste"? Probably.  In your hands? You must try, bake, and taste. (So in terms of too much pre ferment souring the bread, the Gosselin Method does not actually do a complete pre ferment on all of the flour, so the bread should not be "soured.")

Mixing a complete dough (salt, yeast, and all) and doing a long cool bulk ferment gives the advantage of a long fermentation to all of the flour.  More "taste"? Most likely.  More extensibility.  Probably.  The presence of the salt will have an impact on how the gluten develops, but probably not too much.  This method has certainly been used with good success to give flavorful, very open crumbed baguettes.

And with that, we must consider the effect of the amount of yeast in the final dough - do we want maturity to occur slowly or quickly?  We can use a large amount of yeast and make a long fermentation by retarding the bulk ferment or we can use a very small amount of yeast and make a long ferment without retarding. Each of us has different constraints that will make different methods more desirable.  For example, I must produce bread from the mixing of the pre ferment to the cooled loaf in less than 20 hours.   My process has also been tuned for that.

(And no matter what, I will paraphrase what I have heard from no less a baking luminary than Solveig Tofte  - that getting an open crumb in baguettes is not a function so much of hydration, but of getting the fermentation correct.  In all these methods there is a factor of removing the bread from the particular type of fermentation involved at the correct time and degree of fermentation.)

When you talk of soaking all or some of the flour, you are talking about putting water to the flour in the absence of yeast.  This will allow the gluten to develop without fermentation.  This is often used with whole wheat flours - where it also softens the bran so that it does not slice the gluten strands.  It is used less often with white flours.  Soaking the flour is not a pre ferment.  It can be done at cool temperatures for exactly that reason - to prevent any wild yeasts from getting involved in fermentation.  So, soaking the flour is not really done with goal of adding "taste" - it is done with a goal of getting gluten to develop more thoroughly.  This may have more effect on bread volume than taste.

There are infinite variations and there is no "one true way."  Few of us on these pages have access to precisely controlled environments, but I will repeat my recommendation to try to bake the same bread a few times and only then evaluate and vary.  It is unfortunate that we do not have absolute consistency in terminology for bread baking, but sometimes we need to take a deep breath.  Learn to think in terms of the amount of flour that is pre fermented, the hydration of the pre ferment, the total hydration of the bread and the percentage of yeast in both the preferment and the final dough.  This is a very useful way to think about bread techniques and although some may have interesting variants, you can at least see those variants in the context of "normal" methods.

I hope this helps

AndyM's picture

Hi Venkitac-

I'll wade in - it sounds like you are using standard flours and not whole/crushed grains for all of the doughs in question.  If this is the case, then your idea of using all of the flour in a "soaker" sounds essentially like an extended autolyse ("autolyse" is Calvel's term for incorporating all the flour and water together without salt or leavening and letting it rest for 15 min to an hour, then adding in the remaining ingredients before a final mix).  There are two primary purposes of autolyse - first is to hydrate the flour.  It takes some time for the water in a dough to work its way in to where it bonds in the flour (which is in the developing protein (gluten) molecules and in the starch molecules (mostly the damaged starch)).  An autolyse period allows this process of hydration to take place before active mixing begins, allowing gluten formation (which requires water) to get a head start.  This means that less mechanical mixing is necessary to fully develop that gluten, and reducing mechanical mixing time is desirable to prevent over-oxidation of doughs and the overdevelopment of elasticity at the expence of extensibility.  The other main purpose of an autolyse period is to give enzymatic activity a head start before the introduction of leavening agents such as commercial yeast, preferments, or sourdoughs (each of which have yeast organisms of some sort as the primary leaveners).  All of these leavening agents act in the same basic way - by fermenting simple sugars.  The sugars are present in very small quantities in flour already, but a far more plentiful supply of simple sugars is created by enzymatic activity in the dough.  Flours contain a variety of enzymes that, when activated by water, start to "snip off" molecules of sugars from the ends of longer, more complex, molecules of starch (and to a lesser extent, proteins).  This is the primary food source for fermenting agents, and by using an autolyse period, these enzymes start snipping off sugar molecules before there is any yeast to ferment said sugar molecules.  Thus, when the yeast is added, it enters an environment that is already rich with food.

Now, you'll note that standard autolyse periods range from 15 min to one hour.  This is intentional.  The hydration benefit is complete by this time, so further autolyse should have no appreciable effect on hydration.  Enzymatic activity, however, continues as long as the enzymes are active and as long as there are appropriate molecules for those enzymes to attack.  Therefore, the enzymes will keep snipping off sugars, and while this sounds good, there can be too much of a good thing.  Two potential negatives: first, enzymes attack starch and protein molecules - if these molecules are snipped up too much, then two bad things can happen - the water that has bonded in the starch molecules will be shed, leading to a glossy, sticky dough that is hard to handle, and as the protein molecules are attacked, the very gluten structure you desire to develop in your dough is being broken down. Thus you could end up with a slack, glossy dough that is substantially reduced in strength from an autolyse that goes too far.  Second, when you do add the fermenting agent, it might be entering an environment that has too much simple sugar - yeast can be overwhelmed in an over-sugared environment, leading to slowed fermentation, which works directly against one of the purposes of autolyse.

As with most things in making bread, the question here is not what things are good and what things are bad, but finding the appropriate balance between each of the biochemical factors that are essential to the final product.  Lots of effort is expended trying to figure out the "right" level of enzymatic activity to balance with the "right" amount of leavening agent and the "right" amount of acidity.  However, these things are not really independent of each other - if you change the acidity of a dough, you will change the enzymatic activity, as well as the fermentation rate of the yeast.  Similarly, temperature is a variable that has substantial effects on fermentation, enzymatic activity, and even the hydration capacity of flour (colder flours hold more water).

Regarding mixing the yeast in before an overnight rest, if you leave out the salt, then you will essentially be creating a giant ball of pre-ferment.  Pre-ferments, however, have some potential limitations - particularly with regard to gluten strength.  They are vulnerable to the same kinds of protein degradation as extended-autolyse doughs are.  This is one reason why most prefermented doughs use only a percentage of flour in the preferment -this percentage is more fully fermented, but the fresher flour added into the fully mixed dough has its gluten at full strength.  Again, it is an attempt to balance two competing characteristics -the flavor of the more complete fermentation of a pre-ferment with the strength of flour that has not been degraded by extended enzymatic activity.  One other way to help maintain gluten strength is adding the salt - this strengthens the gluten structure while inhibiting the enzymatic activity and fermentation rate.  Therefore, in the scenario you present,  I would recommend including the salt before an overnight rise, thus essentially making the entire dough, even if it is not mixed/kneaded.  In this case, you are essentially using one of the very popular "no-knead" methods of making dough.  These have strong scientific underpinnings, as do formally mixed/kneaded doughs, and I have found them to work extremely well for a variety of breads.

Now, on to temperature. Refrigerating doughs accomplishes two things - it slows down fermentation rate/enzymatic activity, and it shifts the balance of biological activity away from yeast fermentation and toward bacterial activity.  A primary bacterial contributor to bread taste is lactobacillus, which produces, as a result of its activity, lactic acid.  As can be expected from the fact that this is an acid, when lactobacillus organisms are relatively more acitive, a dough will become more acidic, and therefore will likely have a more pronounced sour flavor in the final product.  This is desirable in "San Francisco" type sourdoughs, where more sourness is coveted.  In most old-world style breads, though, the goal is to produce a more balanced flavor profile.  So putting a dough straight into the refrigerator after mixing should lead to lower enzymatic activity (fewer sugars), reduced fermentation rate (longer rise time), and increased levels of sour tastes in the final product (also recall that temperature affects other dough characteristics, including hydration, meaning that doughs that will be refrigerated can be made wetter than room-temperature doughs).  Leaving the dough out at room temperature before putting it in to the fridge should allow the enzymes and the yeast to get going and create some of those other flavor characteristics before the lactobacillus take priority.  How long to leave a dough out would depend on the percentage of leavening in the dough and the ambient room temperature, and would probably require some experimentation.

If you're still reading down here, congratulations!  You're a bigger bread geek than I am!  I hope that these explanations might help you develop some expectation about the effects of some of these variables - but as you can see, the variables are not fully independent of each other, so that when you alter one variable, the others will respond in certain ways.  The trick, then, is to find the balance that creates the effect you are looking for.  I would encourage you to post updates of your experiments, along with photos if possible, so that we can all learn from your experiences.

proth5's picture

You refer to a retarded ferment as encouraging the action of lactobaccili (and even that is not 100% true under all circumstances, but close enough).  I totally get that in the context of naturally leavened breads.

But what about doughs with only commercial yeast?  My thinking would be that these have not yet had a chance to undergo any significant development and that the "souring" that you talk about should not occur.  Yes, the bread benefits from the long fermentation, but that should be about it.  Any thoughts?

And, of course, if we use a poolish (or a liquid levain) as the pre ferment we must autolyse in the presence of leavening agents (and a soaker is just not practical)- so yet another variation...

I was going to add to my post that I have worked with or talked to a number of folk who I would consider very qualified professional bakers and although they can explain they why's and wherefores, when asked about the choices in their formulas, they always end with "It works for me."  This is not an equation where we plug variables and good bread comes out the end.  Even in a large scale bakery (where mixing and machining qualities as well as consistency take precendence over taste) with all their testing it comes down to a highly experienced baker saying "yes, this is the right mix."

AndyM's picture

Thanks for your comments above, as well as for these questions.  I completely agree that what works for one person might not work the same way for another person, and that there is no real substitute for experience.  That being said, we can develop some expectations of the potential effects of certain kinds of changes made to a dough.


First, I would suggest that the difference between naturally leavened doughs and those made with commercial yeast can be viewed as a continuum.  Commercial yeast is a particular variety of yeast that has been standardly bred because of its properties of rapid fermentation and relative stability.  This is important for commercial bakers, who value efficiency and consistency, and home bakers, who often keep yeast around for a long time before using it.  Sourdough cultures are very complex systems that have achieved a relative level of stability over time (this typically takes about 1-2 years, in my experience).  They usually contain several different varieties of yeasts along with many different bactierial strains.  Once these organisms reach their balance, a starter can be quite predictable in its behavior, and in the taste profile of the bread it produces - but no one starter is exactly the same as any other starter.


With commercial yeast, most standard processes of fermentation move so quickly that the activity of the yeast dominates any other type of activity.  Even though there are lactobacillus (and other bacterial) organisms present in the flour and ambient air, these organisms take some time to get to work and to reproduce enough to have an appreciable effect on the final taste.  A sourdough culture gives the bacteria a head start by already having a stable community of organisms ready to go.  Now, if you use a very small amount of commercial yeast, together with an extended fermentation time, you can generate enough bacterial activity in the dough that it adds that extra "tang".  This is essentially what a poolish or biga does - it allows for enough time to develop a bacterial side to the activity of the dough, while also allowing the yeast to create the alcohols and CO2 that are the products of their fermentation.  Theoretically, with a small enough amount of yeast, and a long enough fermentation time, you could perhaps achieve the same effects as a sourdough, although the wild-card in this scenario is the enzymatic activity, which will continue and ultimately could degrade the starch/protein so much that you would lose certain desirable characteristics of strength and water-holding ability.  This leads to one other comment on enzymes - my understanding is that in "complete fermentation", the issue is not that the yeast runs out of food - the enzymes that create yeast food continue to be active until they run out of starches, which would take much longer than the cycle of fermentation of yeast.  Instead, what happens is that the yeast cells, which cannot move independently, start to encase themselves in alcohol and CO2, which are, in a very real sense, their own waste products.  Once they create a complete bubble of alcohol and CO2, they can no longer reach their food source, and they die.  Thus it is not the lack of food, but the yeast's inability to access the food, that defines "complete fermentation".  This is also why "punching down" a dough can have a rejeuvenating effect - the punchdown punctures the waste-bubbles around the yeast cells, and moves the yeast cells into different locations where they can once again access their food source.


Regarding poolish/pre-ferments, one could prepare a preferment with a part of the flour, then autolyse the rest of the flour at the time of mixing (i.e. the autolyse would include unfermented flour and water), then after autolyse add the preferment and any further yeast content that is desired.  Something similar is described in Peter Reinhart's books, I believe, and may be one of the variants of the Gosselin method.  In a sense, I suppse that the pre-ferment itself may be considered to be subject to an "autolyse", but this use of the term would not be in keeping with Calvel's intent, which was that flour is hydrated in the absence of any leavening agent.


I completely agree with your comment that bread baking is not an "equation" that can be crunched through - it is far too complex a system.  For any change you make in one element of that system, the rest of the system will be affected by that change.  This makes experimenting very difficult, because to know what change led to your effect, you would want to keep everything except your one altered variable the same.  But if many variables change because you altered one variable that is of interest to you, then you have lost what might be thought of as "experimental control".  It becomes very difficult to figure out if the change you know you made is what caused the outcome, or if the outcome really resulted from one of the unknown changes that occurred indirectly as a result of your intended change.  In the absence of a scientifically controlled environment, then, personal experimentation, together with some sense of the kinds of results that can be expected from a given change (and a whole lot of flexibility), is probably the best bet for most people.

dmsnyder's picture

That's the sound of a quantum jump in the level of discourse.

Okay, then. If there is minimal bacterial action in a cold retarded, yeasted dough, what would be the expected difference between a dough with a long, slow bulk fermentation (with a very small amount of yeast) at room temperature and a dough that is cold retarded for a longer time? 

It seems to me that either there would be no perceptible difference or the difference would have to be due to bacterial action. (There is another set of activities due to chemical reactions controlled by enzymes already present in the flour, but I don't know if these influence flavor.)


proth5's picture

I'll make no secret that I miss mulling over minutae with bwraith. 

I mull over the two in my post below.

Good thing I'm just a raggedy home baker/miller or I'd get intense...

AndyM's picture

Hi David-

Nice question!  Theoretically again (I haven't done this test), I think there would be a difference, and I think you're right in thinking that the difference would be due to bacterial activity.  I mentioned the lactobacillus bacteria earlier; this organism leads to a component of lactic acid in the final product.  Another type of bacteria that is relevant is acetobacillus, which produces acetic acid.  Acetic acid gives a much more sharp taste and creates a more acidic environment.  Temperature affects these two types of bacteria differently, with lactic acid production relatively preferred at higher temperatures, and acetic acid production relatively preferred at lower temperatures.  So, with low yeast percentages fermenting at room temperature, I would expect that the bacterial activity that would be present would favor lactic acid, which tastes more mild and full in the mouth.  A dough that is cold-fermented for a similar amount of time (which would have to have a higher yeast content, because more yeast would be required to ferment the same dough in the same time at a lower temperature) would likely favor acetic acid, which has a more astringent, sharper taste.  However, because the bacterial activity in yeasted doughs is relatively low (compared to that in sourdough breads), the final amount of either of these acids will be fairly low, and the taste differences should be somewhat subtle (though noticeable).

On another note, the enzymatic activity does play a role in all this - the enzymes are what prepare the food source for the yeast and create residual sugars that play important roles in baking.  Ultimately, they have a huge influence on flavor, contributing a lot of the sweetness that balances the sour and bitter profiles of other activity in the dough.

Mini Oven's picture
Mini Oven

I do hope you stick around as you do seem to know your way around.  ---:)  Mini

proth5's picture

Yes, the enzymatic activity is important.  Most commercial flours have malt added to correct falling number to an acceptable range though, so we tend to put it out of mind. Boosting enzymatic activity may also be a driving factor in adding malt to high % preferment breads.

Whole wheat, however, tends to have a very high falling number and normally we don't correct it.  Yet in my hands, whole what tends to ferment more quickly than white. This takes us down the path of the much higher ash content supporting the fermentation process.

But as the results of the experiment below tell us - soaking is for the gluten - long, slow fermentation for the taste.

Now to sort through the many permutations of each... 

arzajac's picture

From my experience at home-brewing beer, there is a big difference in the flavors produced by yeast at high versus low temperature.

Traditionally, beer is made using one of two types of yeast - ale yeast or lager yeast.  Ale yeast work best at higher temperatures and finish fermenting in a matter of days.  Lager yeasts work best at low temperatures and typically take several weeks to months to complete.  Lagers have a smoother flavor than ales. 

It is possible, and often desirable, to use ale yeast at lager temperatures and vice versa.  Steam Ale is a popular beer that is brewed at higher temperatures but using lager yeast.  At high temperature, yeast produces various organic compounds which can impart complex flavors to the final product.  This is what gives "Anchor Steam Beer" a unique flavor.  Some beers have fruity aromas and flavour, others have a caramel aftertaste, and some beers have a very clean taste.  All of these beers can be brewed with the same ingredients, just at different temperatures. 

So, for the same amount of yeast activity (for example, if you measured the amount of CO2 produced), yeast produces different by-products depending on temperature, among other things, of course.  But, I don't think bacterial activity is as significant at changing the flavor or bread in non-sourdough breads as yeast by-products.

What about enzyme activity?  Beer wort is made from keeping the malted barley at an optimal temperature for the amylase to convert the starches to sugar.  The sugars are sparged (strained) into a pot and boiled for a long period of time - long enough to denature any remaining enzymes in the wort.  So the development of flavor in beer is not reliant on enzymatic activity such as in bread.


proth5's picture

I tend to look at all this from a very practical standpoint. so, here goes.

While understanding the theoretical of using  a small enough amount of commercial yeast and a long enough retarded fermentation to allow bacterial action, in essence what we are doing as home bakers is using a moderate amount of commercial yeast and a short enough cold bulk fermentation time (usually no more than 8-12 hours) that I'm just not sure this theoretical occurance has time to occur.  Add to that the fact that high hydration "sourdoughs" kept at relatively cool temperatures tend to encourage yeast growth (as I have been told and sort of expereienced) and we're into the territory of "maybe it happens" in my tiny mind.  Or am I completely off base here?

I'd agree that a yeasted pre ferment develops a flavor, but here we are talking about fermenting at "room temperature" (in general) for 8-15 hours and I'm not yet ready to sign on that it develops the significant lactic/acetic qualities that we would get with a levain.  I would hope that they would be different - because this allows us different taste profiles.  I've taken a liquid levain pre ferment and put it into a commercially yeasted dough.  This is a very different flavor than a commercial yeast preferment in a commercial yeast dough (and I did not find it to be my "best bread ever"...)  I'm open to convincing, but my baking to date does not do it.

I'll go down with the ship over adding a liquid levain or a 100% hydration preferment to the autolyse.  I've talked about this at great length with folks who ought to know what Calvel meant.  I'm not budging on this.  You can't autolyse the full amount of flour with the remaining liquid if you do a high hydration pre ferment.  You cannot.  The pre ferment must be added and it's no crime.  If you think it is, then you must make a firm pre ferment to use autolyse.  The Gosselin method (and I hope this thread doesn't get found on "the Google" when people really want to know about those folks with 8 kids) has a small amount of yeast in the full amount of flour for a long, cold ferment.  I'm not sure that is a real autolyse - but that's where knowing the baseline helps us think a method through.  The method gives a long, cool, but not too thorough ferment at a low hydration to all of the flour.  This flour will have pretty nice gluten development, but since the yeast is starting to drown in its own waste but hasn't completely used all the available food it is a good candidate to get some more water, some more yeast, some knocking about and it will work well.  It should have a nice flavor because of all that flour that was fermented so long and will have good elasticity for the same reason.

I'm also interested in the complete fermentation.  Yes, punching down a dough does re distribute the waste products and allow the yeast to grow again, but at some point, the food is exhausted.  We see this in over proofed dough - or dough that has had too long of a bulk ferment.  Doesn't matter what you do - the thing is not coming back.  I know that if I stir down a fully ripe 100% hydration pre ferment, I won't be seeing much more action from it.  Perhaps the difference in the hydration levels between the pre ferment and the final dough are the difference.  On a practical level, I have seen as much as 42% of the flour pre fermented in a formula.  If that 100% hydration pre ferment is mature, you have not much food left for the yeast in the rest of the flour.  Something must be added. Or so the baker says.  What am I missing here?

I'm not a scientist (although I are an engineer).  I tend to try to do "single factor" experiments (although, my baking has reached a point where I can consistently turn out the same bread with the same formula week after week) but I know if I were a scientists I would find them replete with flaws.  What I have found useful though is to set parameters: I need to get the baking done from pre ferment to cooled loaf in 20 elapsed hours.  I want to use natural leaven.  Then I establish my desired mixing technique - by hand.  By this very set of parameters I have eliminated certain techniques (reatrded bulk fermentation and pseudo Gosselin for example...).  Then the variables that I mentioned can be varied to design the bread.  As in any discipline if we leave things too open ended we can glide into the paralysis by analysis mode and neglect making excellent bread in the pursuit of "perfect bread" (and if "my teacher" heard me say that he/she would not believe it.  Perhaps my time at "the place" did me good, after all)  What I have focused on in my bread evaluations is to identify the probable source of the problem.  Insipid taste?  Probably a problem with fermentation.  If I was careful about my bulk fermentation, this is probably an issue with how I conducted my preferment.  If I know that it was ripe, I will try to vary the percent of flour that I pre ferment.  Just can't get the thing to hold its shape?  Well, if I've been careful at all of the other steps, perhaps I have pushed the hydration too high.  That's why I emphasize baking a single formula (even if it is sub optimal) until you have achieved reasonable consistency.  That way when you vary one factor you have the memory of how the thing has performed in the past and can judge if you are closer or further from your goal.  When that bread is consistent and if you want to "improve" it - then you can vary one factor at a time once again.  Bear in mind that because of the tender age that I started baking bread and my own "decrepitude" I've been messing with bread now for a half a century.  I won't say I'm an expert, but I've had a lot of dough pass through these hands and frankly that colors my thinking.

Jeez we can get going on this stuff.  :>)  That's why, while I won't cop to ignorance on a lot of these topics, I tend to endorse putting them to back of the mind and let the bread do the talking.  But it is fun to chew over some of the more theoretical...

dmsnyder's picture

I'm going to go stretch and fold some dough now.


foolishpoolish's picture

and correct me if I'm wrong but I believe the 'Gosselin method' is indeed a yeastless autolyse. Flour and water held overnight before yeast and salt are added. There's no yeast fermentation going on during that long resting period. It was Reinhart who modified the technique in BBA to add yeast before refrigeration. Perhaps you had it confused with Anis Bouabsa's baguette-making process?


proth5's picture

I really am not too familiar with the "Gosselin" technique, so I looked it up on these pages on dmsyner's blog and found that yeast was being added.  Since I don't own a copy of BBA any longer it was hard to find other reference material.

So, I figured David gave me the straight scoop.  I stand corrected and more informed.

venkitac's picture

Nit: I believe the BBA Gosselin technique adds yeast, but the original Gosselin technique used by Gosselin doesn't. Reinhart modified the original technique is my understanding - the "original one" mixes flour with ice cold water, soak for 12 hours in the fridge, then add salt+yeast and ferment and proof and bake. Correct me if I'm wrong...

dmsnyder's picture

FYI, the following link has the formula Gosselin gave Peter Reinhart. In the Usenet post where I found it, Peter says Gosselin gave slightly different instructions to other visitors to his bakery.à-l039ancienne-according-peter-reinhart-interpretted-dmsnyder-m

And here is the original message from Peter, in its entirety: 

From: "s.reinhart" <>

Subject: Pain a la ancienne--as originally shown to me

Date: Sat, 25 Jan 2003 08:48:36 -0500

Dear List Members,

Here is what Philippe Gosselin showed me, though others who have met him 

said that he did it differently for them. First, he mixes a dough of about 65% water to flour, with no yeast or salt, using very cold water. This is held overnight in the refrigerator, or what the French might call a long "autolyse."  The next day he remixes the dough, adding 1% fresh yeast, approx. 1.75% salt, and another 5% cold water. He suggested that with American flour it would be best to use 75% water, or maybe even more, depending on the flour. It makes a ciabatta-like dough, wetter than regular baguette dough (but not unlike the "Retrodor" baguette that is becoming popular in Paris these days--see the new Jeffrey Steingarten book, "It Must Have Been Something I Ate," for more on that excellent method). Gosselin mixes this until thorougly incorporated, about 4 to 6 minutes or so. This dough is allowed to ferment for 6 hours at room temperature, during which time it awakens and doubles in size (This is a big batch, so a small batch might awaken faster). The dough is then divided into baguette size (about 12 ounces), formed into a six inch torpedoes, rested for about ten minutes, in a bed of flour, then gently pulled, not rolled, to baguette length, placed on a baking cloth (couche) until enough are shaped to fill the oven. They are immediately taken to the oven, scored with a blade just as any baguette, and baked like other baguettes (about 460 F., equivelent). They nearly double in size in the oven, resulting in a crumb with holes somewhere between a regular baguette and a ciabatta.

My version is designed to make it more user friendly for non-professionals who don't have the luxury of baking shifts, waiting for 6 hours, etc. I think the results are fairly comparable, though a home oven can only take small lengths, not full baguette, so I advise making the pieces about 6 ounces for mini-baguettes. Of course, I also use the dough for ciabatta, pizza, and focaccia, though Gosselin doesn't use it for those purposes.

If anyone tries the more difficult method described above, I'd love to hear how it turns out for you. Good luck!


     Best Regards,



venkitac's picture

Thanks for the *super detailed* explanations, all. I can't say I have digested all this, I'll probably have to sleep over this and re-read this a couple of days:) I'll chime in with just one thing: the  comment about soaking doing nothing to flavor seems spot on. I took Paul's suggestion and ran 2 experiments with whole wheat bread (70% whole, 30% white, actually):

- Cold soak & retard all the dough with no yeast or salt for 12 hours, add salt+yeast, ferment for 3 hrs, proof for 45 mins, bake.

- Mix all the dough, salt, yeast and all. Immediately retard for 12 hours, dechill, proof, bake.

The latter was much stronger in wheat flavor than the former. In fact, I have never had such an intense whole-wheat flavor in any bread I have eaten, almost too strong a whole-wheat flavor. The latter bread, in fact, would have to be eaten as such and not with olive oil or in a sandwich, because the wheat flavor would mask everything else. (The only other food with which I have ever experienced such a strong wheat flavor is an Indian steamed wheat dish - pulverize wheat, autolyse 15 mins, steam, eat, really strong whole-wheat flavor). OTOH, the gluten development in the former was remarkable, it was like an all-white-flour bread, I could do anything with the dough really. But the flavor was much milder, and surprisingly, I got less of a rise out of it. (From one of the above posts, I now think it makes sense, it was probably over-autolysed and the yeast didn't do it's job fully).

Thanks again. I'm gonna print out the entire thread and stick it on the fridge. I'm kind of a bread geek even if not a great baker, and I've a read a bunch of books on this. But the above thread replies are actually more informative than any book I've read! Viva fresh loaf and thanks, all.

foolishpoolish's picture

From my experience and what I have read about the process - the benefit from 'soaking' (other than gluten development) comes largely from a longer period of amylase, thus freeing up more sugars. This has an impact both on perceived sweetness in the final product and also on crust colour and flavour. The 'gosselin' method that is often cited applies specifically to baguettes (or Pain à l'Ancienne) where both these attributes are important. Wholegrain breads may not benefit so much from this method (the dough in your experiment was 70% WW) - I believe someone here on TFL has come to similar conclusions from their own experiments.

My lay understanding of the enzymatic processes would lead me to think there is no reason why one should expect an increased 'wheatiness' from extended 'soaking' - this is something your experiment appears to confirm. However in my experience, slow fermentation or pre-fermentation of whole grains does bring out flavour in a way that fast fermentation of a direct dough does not. Again your experiment seems to bear this out. This is one of the reasons why, for example, when making a mixed wheat/rye bread, often the bulk of the rye appears in the levain/'rye sour' or pre-ferment.

In conclusion, what I think I'm trying to say is: the two different processes lead to different applications. Neither should be discounted as a blanket rule in favour of the other - but rather employed appropriately depending on the type of bread being made.

Cheers, FP

AndyM's picture

proth- thanks again for the thought-provoking comments - I think you are completely correct that including a liquid preferment in an autolyse is a valid use of the term.  I've even used this exact process in I use only firm preferments these days and my memory of high hydration preferments is rusty. So thanks for re-awakening my thoughts on this.

I'm also interested in the questions you raised about the contributions of temperature, enzymatic activity, yeast fermentation, and possible bacterial action.  You are right that at a certain point, the yeast in a pre-ferment has fermented all it is going to ferment.  But there might be more than one reason why its activity might stop - true, the food source could be used up, but it is also possible that due to increasing acidity in the preferment, the pH of the environment might enter a range where the yeast is no longer able to function optimally.  Baker's yeast (Sac. Cereviciae) prefers an environment that is only slightly acidic - if the pH goes to low, it will not perform well.  Could this be what happens in a preferment that loses its steam?

Similarly, with an overproofed dough, the same question arises- it is true that there might be no more food left, but could it also be that the acidity of an overproofed dough causes the yeast to act increasingly poorly?  With doughs, one could also look at the action of proteases (the enzymes that attacks the protein structure (i.e. gluten) of the dough).  If the gluten breaks down enough, then the yeast could potentially still be fermenting sugars, but the dough might have lost so much stuctural integrity that it can't hold in the yeast by-products.  The result could look quite similar to dead yeast.

Now, i am not a chemist either, and I don't know the truth about these questions.  However, it is the case that even in "fully fermented" doughs, there continues to be substantial enzymatic activity through the final stages of proofing and baking - this is what produces residual sugars that contribute to the maillard reaction and to the final sweetness profile of the loaf.  If the enzymes continue to be active even in a fully fermented loaf, then it seems that they would have the potential to continue to produce food for further fermentation, should the yeast be available for that fermentation.  This idea makes me lean toward the thought that the enzymes (particularly amylases) have sugar-producing capacity that consistently exceeds the yeast's capacity to ferment that sugar.

Now the question of acids - lactic and acetic acid are two fairly well-known acids that contribute to the flavor profile of a finished product - they are more prevalent in sourdough breads, presumably because the sourdough culture has an active colony that includes lacto- and acetobacillus bacteria.  But yeated breads have these acids too, in lower amounts.  These acids must come from somewhere.  Here's where I get to the limit of my current knowledge, though - I have always thought that yeast did one and only one thing: convert simple sugars into CO2 and alcohol  That's all, nothing else.  If this is the case, then something else must be creating the complex of organic acids (lactic and acetic among them) that create the fullness of the final flavor profile.  My thought is that the alcohol that the yeast produces is then the food source for bacterial strains that are present in all doughs, though in low concentrations in non-sourdoughs.  Each bacterial strain converts the alcohol into whatever variety of organic acid is produced by that kind of bacteria.  Chemically, this is also a process of "fermentation", and I had always assumed that this was what was meant by "secondary fermentation" - the primary fermentation is the yeast acting on sugars, and the secondary fermentation is the bacteria acting on the product of the primary fermentation.

Trouble is, I don't really know if this is true.  It also seems possible that the alcohol could enter into a chemical reaction with something else in the dough that results in a variety of organic acids - this could be catalyzed by some non-biological entity, such as an enzyme, or could perhaps even be catalyzed by temperature, thus leading to the observed fact that breads leavened at different temperatures end up with different flavors - maybe different temperatures catalyze different chemical reactions with the yeast by-products?

Or, perhaps my very simplistic view of the activity of yeast is off-base - could yeast actually directly produce a greater variety of acidic by-products?

This is where we need the services of a resident bio-chemist - does anyone have one of those lying around the house who could shed some light on these processes?

And thanks again, to proth and to everyone who has contributed to this thread - I haven't thought this stuff through to this degree of detail in more than ten years, and it does feel good to stretch that mental muscle...

proth5's picture

to purchase a pH meter that has a wider range and more accuracy than my pH papers.  I make cheese and some types of cheeses require very precise pH's at certain steps in the process - so I figured I could treat myself.

I just think it would be interesting to try a few tests while doing bakes and pre ferments. 

My word is my bond on this.  Sometimes it takes me awhile to get things done, but that's how bonds are...

Just as an aside, what is often seen on an over fermented dough is lack of browning (uh, and I just answred my own question about "what's going wrong with my bread?" - I've been making some adjustments to my "old faithful" process to see their impact and I've not been completely thrilled with the results - doh!) which does indicate lack of sugars (or lack of available sugars based on amylase activity?) rather than low pH.

I'll probably retreat to the "Well, that's what works" in the long term, but I'd like to understand the theoretical underpinnings. 

davidg618's picture

It's true that there are bacteria that convert alcohol to acid. In the food industry the most common example would be vinegar (primarily acetic acid) is created by exposing alcohol to a specific acetic acid producing bacteria that "eats" ethanol. However, this is not the bacteria that is present in sourdoughs.

Some bacteria, just like yeast convert sugars into their respective waste products, i.e., CO2 and acids, CO2 and alcohol respectively. So some acid producing sugar eating bacteria may live side by side in some environments with some yeasts and compete for the available sugars. But eventually, due to different reproduction rates one species would dominate.

So how do sourdough starters continue to live stable lives for alledgely hundreds of years--if you believe the folklore? Therein lies the beauty of sourdough starters.

Sugar molecules aren't all the same. We've all seen, on food labels, words like fructose, glucose, sucrose, maltose, dextrose, and who-knows-what-ose. These are all different sugars, and their difference is evidenced at the molecular level. Rather then spend our time looking at the myriad differences, for this discussion it's important to know that some yeasts can't eat some sugars, and similarly some bacteria can't eat some sugars. If a particular yeast lives side-by-side with a particular bacteria, in an environment that contains sugars that each can eat, and each can't eat the other's preferred sugar; Eureka! we've got the Jack Spratt and his wife relationship. They can exist side-by-side, and lick the platter clean.

To get slightly more specific, from my Google derived reading--I'm not a biologist, neither micro nor macro, but I am a scientist, and I can read, and math is math--the most scientifically studied sourdough is--surprise, surprise--San Fransciso sourdough. The biologists have discovered the primary bacteria in SF sourdough is unique--appropriately, they included sanfrancisco in its appointed name, and it can, and does eat maltose. It's teamed with a variant of a relatively common yeast that can't eat maltose. Moreover, the SF bacteria produces other byproducts that prevent other organisms from living in the same environment--much like penicillin--that its yeast neighbor is immune to. There is a lot more going on too in a flour and water environment: e.g., temperature tolerance, pH tolerance, enzyme production, ionic bonding, and so forth; but this is the gist of their relationship: they can coexist, and so long as there is food for both, they coexist in a stable environment--for centuries, if you believe the folklore. So, simply put, we feed our starters, and keep them comfortable.

An aside: Subsequent reasearch has found the sanfrancisco bacteria in sourdough starters im many other places throughout the world, teamed with similar yeasts.

That is not to say there aren't other stable yeast, bacteria relationships in other starters, however, it is safe to say that to be stable their coexistance requires similar attributes.

David G