Baking Roman Bread: Reproducing the Herculaneum Loaf

The Herculaneum loaf - British Museum image

A friend recently drew my attention to a video at the British Museum site dating from their blockbuster Pompeii and Herculaneum exhibition in 2013, in which chef Giorgio Locatelli tries to reconstruct the famous carbonized loaf featured in that same exhibition.

Giorgio Locatelli - image from www.britishmuseum.org

Locatelli is a great cook of course, and produced a nice loaf of bread, but I was unconvinced by his effort. From the video itself, a few problems arise. The most obvious I think is the implausible use of string to recreate the obvious two-tier structure of this and similar loaves.
It seems to me much more likely that this was built from two pieces of dough, like a number of other breads in different times and cultures. The nearly-obsolete English cottage loaf is an example some readers may remember, but there is also an interesting parallel with the way eucharistic bread is made for the liturgy of the Greek Orthodox Church, where just this happens--two pieces are stacked and the top marked. It would be bold to claim a direct connection with ancient bread-making, but I don't think it can be ruled out either.

Prosphora - blogs.ancientfaith.com

It might be worth mentioning that in some online discussions I have seen a suggestion that a mould was used to produce the shape of the top of the ancient loaf, but I don't find that compelling either. I think it is more likely that the dough was fairly dense and retained a shape given by scoring the top.

The BM video depicts a reasonably plausible approach to making the dough, using a biga to leaven the bread. When you dig into their website however they have a recipe that loses the plot entirely, with added gluten as well as yeast. That doesn't seem to be Locatelli's own recipe from the video, however.

So this seemed to invite some further exploration, and possible improvement.

I've just made one attempt so far, and only with moderate success. I built a leaven from my sourdough starter. Roman bread did not use brewer's yeast; if it had any leavening, it was something more like sourdough or "old dough" methods. I then made a dough from spelt and bread flour. My baker's instincts trumped my reconstructionist ones here, because my dough was really too hydrated for the purpose. I made two discs, but allowed a final proof (maybe not required?) during which the top just flowed over the bottom.

I slashed or scored it into the eight pieces required, then baked it in a cooler oven than I might otherwise have done, partly imagining that Roman bakers might not always have had access to 400 degrees Celsius. The result was actually rather like Locatelli's (see below).

I've already noted one mistake, in effect, the high hydration. I think that the dough from Herculaneum was probably quite a bit dryer and denser, and while likely the result was perhaps fairly stodgy by modern standards this may have been the desired ancient outcome.

It's worth remembering that these breads were not intended for sandwiches or toast. The scoring indicates the loaf was designed to be broken into portions and eaten, by itself or with other foods, as a substantial part of a meal.

The loaf from Herculaneum was found still in the oven, so it didn't actually finish baking, and its shape might not have been what Celer who prepared it (his name is on the stamp) intended, but on the other hand it is not very different from other examples and depictions.

While I have not found as much documentation about the carbonized loaf as I would like yet, I have come across the suggestion that emmer, far, was prevalent in Campania at this time. So the wheat was probably less elastic and inclined to spring than bread wheat, triticum. While I used some spelt, next time I think it's all spelt or similar. It may also have been made with a coarser kind of meal, closer to modern semolina, which might be worth considering for another attempt at getting something closer to the bread from Herculaneum out of a modern oven.