The Fresh Loaf

News & Information for Amateur Bakers and Artisan Bread Enthusiasts

Sourdough baguettes: My version of Gosselin's "baguettes tradition"

  • Pin It
dmsnyder's picture

Sourdough baguettes: My version of Gosselin's "baguettes tradition"

I have made baguettes following many different formulas. Some of the most interesting have been various versions of “pain à l'anciènne,” including those of Reinhart in The Bread Baker's Apprentice and of Leader in Local Breads. Sometime back in 2008, I found an e-mail that Peter Reinhart had sent to a bread bakers' Usenet mailing list in 2003 which described the formula for pain à l'anciènne as he got it directly from Philippe Gosselin. The version that ended up in BBA was simplified somewhat by Reinhart, adding all the ingredients before the mixing, omitting the double hydration and delayed addition of the salt.

When I first made baguettes from Gosselin's original method, they were the best-tasting ones I had ever made. I finally got to taste Gosselin's baguettes tradition (from the rue Caumartin shop) last year. To my taste, they had a bit of a tang suggesting they might have been made with levain, so I modified the formula to use a liquid levain and found I preferred the result to that leavened with commercial yeast. In fact, I preferred what I had baked to Gosselin's own.

This is the version I used for today's bake:



Baker's %

Organic AP Flour

400 g


Ice Water

275 g



8.75 g


Liquid Levain

200 g


Instant yeast (optional)

¼ tsp



883.75 g


Notes: Accounting for the flour and water in the levain, the total flour is 500 g and the total water is 375 g, making the actual dough hydration 75%. The actual salt percentage is 1.75%.

For today's bake, I made 3/4 of the dough amount in the table above.

I mixed the levain the night before starting on these and retarded it in the fridge overnight. 


  1. The night before baking, mix the flour and levain with 225 g of ice water and immediately refrigerate.

  2. The next morning, add the salt and 50 g of ice water to the dough and mix thoroughly. (I did this by hand by squishing the dough between my fingers until the water was fully incorporated.)

  3. Place the dough in a lightly greased bowl with a tight cover.

  4. Ferment at room temperature until the dough has about doubled in volume. (3 hours for me) Do stretch and folds in the bowl every 30 minutes for the first two hours.

  5. An hour before baking, pre-heat the oven to 500ºF, with baking stone and steaming apparatus in place.

  6. Divide the dough into 4 more or less equal pieces and stretch each into a 12-14 inch long “baguette.”

  7. Score and bake immediately at 460ºF, with steam for 10 minutes, and for about 20 minutes total.

  8. Cool on a rack before eating.

Notes: In Step 7., I specify shaping the loaves by simply stretching the dough pieces into a rough baguette shape. This is a very slack dough and a challenge to handle as one might a lower-hydration baguette dough. If you are very comfortable handling slack dough, have a firm grasp of the “iron hand in a velvet glove” principal and are feeling up to the challenge, you can shape the pieces as you would shape a baguette ordinarily. That is, in fact, what I did for this bake.

You will also note that I scored these baguettes with a single, longitudinal slash. I find the results more satisfactory than the traditional 5 or 7 cuts when scoring a very sticky dough like this. However, the difference is merely cosmetic.

These baguettes had a chewy crust, except for the ears, which were crunchy. I think they could have baked 5 minutes longer, or I could have left them in the turned off oven for another 5-7 minutes to dry the crust. The crumb was nice and open. The flavor was sweet, complex and moderately tangy. I attribute this to a combination of factors – retarding the levain overnight and fermenting the dough, after the final mixing, at 85 dF.

This baguette is still a favorite.


Submitted to YeastSpotting


PiPs's picture

They look delicious ... 


lvbread's picture


Great job! Wish I can do that.

Happy baking!

dmsnyder's picture

They were very tasty 2 hours after baking. Four hours after baking, I tasted them again, and they tasted more sour. This morning, they were good but some of the flavor complixity was diminished. 


fancy4baking's picture

Very delicious looking baguettes. Well usual :D

Juergen Krauss's picture
Juergen Krauss



dmsnyder's picture


oferhalevi's picture

Have I understood the instructions properly? It looks like there is no final proof?

dmsnyder's picture

You understood correctly. There is no proofing after shaping. The loaves go right in the oven.

You might notice that the bottom of the cross sectional slice has a more dense crumb. This is where the seam was sealed, and that part of the loaf was de-gassed more than the top part. The gas distribution might have evened out with some proofing. I don't know. This would not have been an issue if I had simply stretched the dough pieces into shape.


dabrownman's picture

delicious looking.  fine crust and crumb as usual.  A nice World Bread Day example!  Makes one want to consider simplifying slashing even more, if you were handicapped like me, to go to Pierre Nury's no slash, stretch and flop Light Rustic Rye :=)

Thanks for sharing David

dmsnyder's picture

Reinhart in his BBA version calls for no scoring. That would probably work for this very extensible dough. Your choice.


baybakin's picture

Wow, these look lovely.  The smaller baguette has just about (in my opinion) perfect ears for this type of bread.

I have also played with the Leader version of this formula, and found it quite nice.  I actually use the double hydration technique with salt water for nearly every bread I make, placing it after the autolysis in my process.

I'll have to have you give me some tips if you're ever out to east bay.

dmsnyder's picture

I'm no expert on double hydration, but my understanding is that it is primarily useful in very high-hydration doughs.


baybakin's picture

Indeed, you can develop the gluten easier in a lower hydration dough, then adding more water (and salt in this case) once the gluten has had a chance to develop a bit.  My standard bread recipe comes in at about 78%  and I've found that adding more water along with the salt after the autolysis both cuts down on kneeding time/number of folds needed, as well as makes the salt easier to incorporate into the dough.

Floydm's picture

Beautiful, David.

dmsnyder's picture


FlourChild's picture

These look amazing!  Love the shiny, open crumb, and I'm imagining what these must have tasted like- a sourdough version of Gosselin's pain a la ancienne  :)  Appreciate your sharing the details of Gosselin's original method.

dmsnyder's picture


Song Of The Baker's picture
Song Of The Baker

Nicely done David!  Inspiring me to lean towards baguettes again.  I thought of you in the last few days after baking 4 loaves of SF Sourdough and having failed at a good score bloom all four times.  I seem to have better luck with slashing rye breads than sourdough.  The slash always turns out flat, with the edges instead of turning upwards and creating ears, stay level with the crust.  Anyway, back to YOUR drawing board post :)


alfanso's picture

Hi David,

I baked your formula for this ficelle this morning for the first time with a few tweaks.  My short journey to here was through  the Bouabsa baguettes:

to my first levain - the Fromartz formula:


To your baguettes made with liquid levain:

To DonD's baguettes a l'ancienne w/ dbl cold retardation:

And finally to your Baguette Tradition after Phillip Gosselin:

The tweaks that I mentioned above:

  • 200 French folds after incorporation of salt, yeast & 50g of ice water.
  • Divide the dough into two batches - I will not allow 4 across my oven deck - it's not wide enough to comfortably accommodate more than 3 without affecting the inside sidewalls of the breads.  
  • Half of the dough went back into fridge for additional 30 minutes of cooling, therefore giving me time to overlap schedule for 2 bakes, each with 2 ficelles.
  • 4 S&F in 2 hours, every 30 minutes.
  • Rest after final S&F - 5 minutes. 
  • Divide & pre-shape - 15 minutes.
  • Roll ficelles and prove - ~30-35 minutes.
  • Bake 10 minutes steam (1.5 cups very hot water on lava rocks), separate & rotate 180 front to back, then 10 minutes more after rotating.
  • Errored on the first batch (the traditional slash) - did not lower oven temp to 450 (grrr).
  • Applied your full length grigne slash to the final two (first time I tried that one - me like!).

My journey across these few breads over a relatively short time has been an educational one.  As I'd mentioned to you, I have a strong desire to push the caramelization factor - hence the consistently dark crusts.  

Once more, I can attribute a portion of my success so far to the inspiration that I've received from reading your posts at TFL, as well as abiding by your formulas.


dmsnyder's picture

They all look well-baked, and the crumb photo is super!

Although it doesn't seem traditional, I too prefer a bolder bake for baguettes, as I'm sure you can tell.


alfanso's picture

Hi David,

I only rotate through the levain Gosselin baguettes every so often, maybe my 3rd or 4th time yesterday.  I'm happy with the results, and I probably don't bake them more often because I tend toward bakes that require fewer hours on bake days than these do (I bake when my wife takes off for a few hours of her myriad activities on Tue., Thu. and Sat.).

Anyway, below are two pics of yesterday's bake.  The first is what has become my standard "cheesecake" shot.  As you can see, I am not at all disappointed with the results, rather the opposite - delighted.  As I don't get to these often, I forget how slack the dough is at shaping time, but as I go along, I tweak an instruction/activity or two and record them along with what to expect from the dough at that stage.  And shaping is a bit of a challenge, just as you had originally outlined.

The crumb shot has a snub-nose thanks to my wife lopping off the tip before I could cut it (an ongoing "issue").

Okay, so what is my "problem"?  It is with my inability to produce a more sour levain outcome.  My levain seems to be quite happy and healthy, and I refresh it every two-three weeks from the refrigerated stock.  I have been refreshing the levain based on Mariana's suggested 3 stage build, moving a 100% starter base progressively through 62%, 61% and finally to a 66% refreshed starter at the end of stage three.  And it is this refreshed starter that I use for creating a formula-ready liquid levain.

Functionally, this provides me with a healthy and happy starter, but not with any level of even vaguely pronounced tang.  So...I next took a portion of that starter as a base and used dabrownman's 3 stage build as a new base to make a liquid SJSD levain for the Gosselin dough.  A wonderful outcome, as I mentioned and as you can see above.  But (the big 'but' here), no improvement in tang.

I've been around the block long enough to understand that some things take more time, but at this point, I'm at a loss to understand why I can produce an active starter for totally acceptable output, but I'm currently incapable of controlling the level of tang/sour aspect of the final product.  It has such a mild flavor composition as to be just about negligible.  A tasty and quality bread, for sure, but not what I am aiming for in terms of tang.

I use a 50/50 WW/Rye flour mix for my refreshes.

Any clues to discern or suggestions for the next round?



BTW, we are headed back to Portland to spend the summer there.  So, unlikely that I'll bake during that time, but after an almost three year departure from that city which we love so much, we are quite looking forward to revisiting there, and may use this as an ongoing template for future summer excursions.

dmsnyder's picture

First, I have no experience with starter fed with that flour mix. However, in theory, it should produce a more sour levain than my mix does.

Let's see. To get more acid into a dough there are several approaches:

1. Use a higher percentage of pre fermented flour, and let the levain get very mature.

2. Paradoxically, use a lower percentage of pre fermented flour and use a long, cool bulk fermentation.

3. Use a firmer levain.

4. Retard the firm levain for couple days before mixing the dough.

5. Retard the formed loaves.

Now, these tactics can be used in various combination. However, you should consider that some - the first one, for example - may increase sourness at the expense of other desirable flavor elements.

Enjoy Stumptown! It doesn't look like we will get there this Summer.


adrianjm's picture


These look awesome, and just how I want to perfect my baguettes. I'm just starting out, and was hoping you could clarify a couple of points.

1. The first step asks to 'mix' the dough and water. For those without a mixer, does this mean knead? I found the hydration quote low compared to the sponge I make, so I found I had to knead the mix to fully incorporated. Is this correct? This leads me to my second question:

2. Why is it important to hold back some water till the next day?

Thanks again for providing the keys to what looks like a very tasty baguette!


dmsnyder's picture

1. These baguettes are made without any mechanical mixing. The first step is to mix the flour AND the levain with most of the water. I do this with a "Danish Dough Whisk," but you can use a wooden spoon, a spatula or your hands. 

2. The so-called "double hydration technique" is usually used in high hydration doughs. When hydration gets very high, the water molecules actually interfere with gluten network formation. They physically get between the gluten molecules and prevent the formation of the weak chemical bonds between folds of the long gluten molecules which make up the gluten network of CO2-trapping alveoli that give the dough its "structure." So, you start with a lower hydration dough, and you give the dough time to form these bonds, then you mix in the remaining water. This technique is most helpful with even wetter doughs, ciabatta, for example.

Note: This formula is my own adaptation of the Gosselin formula. In mine, a liquid levain contains a significant portion of the water in the total dough. If you leave out the levain in my formula, your dough will be much too dry.

I hope that helps.


adrianjm's picture

Thanks for the detailed info. Makes perfect sense now. I did find that I could not fully incorporate the flour in step 1 by hand mixing with a wooden spoon. Perhaps my levain was too stiff. I make it a bit less than 1:1.

I'll see how they turn out today!