The Hidden Science of Bread: How Dough Becomes Crumb

Take a loaf out of the oven and cut it open. The holes you are looking at were not made by the rise. They were established during mixing, when air was folded into the dough, and everything that happened afterwards - hours of fermentation, the violence of the first minutes in the oven - only inflated what was already there. That single fact explains more about bread than any recipe: why handling matters more than time, why a dough that is knocked back never recovers the same crumb, why two bakers with identical ingredients produce different loaves.

Elliot Marsh follows a loaf from grain to stale, and stops at each stage to ask what is actually happening.

The gluten network comes first: two proteins with opposite personalities, one that flows and one that resists, joined by bonds that break and reform under work - which is why dough is elastic rather than merely sticky, and why there is such a thing as mixing too much. Then fermentation, and the discovery that the most repeated claim about sourdough does not survive measurement: when researchers sequenced 500 starters from four continents, they found almost no geographic signal at all. The median starter held a single species of yeast. Seventy were found across the whole set, distributed by the microbes' own interactions rather than by the air of any particular kitchen.

Then the oven, where the loaf does most of its growing after the yeast is dead; the crust, where the reaction that makes bread smell like bread also makes a regulated contaminant; and staling, which is not drying out at all but starch quietly recrystallising - which is why a loaf goes stale faster in the refrigerator than on the counter, and why warming it back up partly undoes the damage.

There is also the industrial story: the 1961 process that replaced hours of fermentation with a measured eleven watt-hours of mechanical work per kilogram, delivered in five minutes, so that a country could bake acceptable bread from wheat that was not strong enough. It came to make most of Britain's bread. What it bought and what it cost are both set out here, along with what the evidence does not support.

This is a book for people who would rather know the mechanism than be handed a recipe. The numbers in it come from the published record - the starter survey in eLife, the gluten and staling literature, the regulation that defines enriched bread down to milligrams per pound, the agency guidance on acrylamide - and several popular figures are deliberately absent, because chasing them back to a primary measurement turned up nothing solid. Those omissions are marked where they occur rather than quietly filled in.

Twenty-four diagrams, including a three-stage picture of the crumb that shows the gas cells growing without ever increasing in number.