Medieval historians and the readers of medievalist fiction understand the centrality of bread in the diets of most people, not only during the medieval period but throughout a long stretch of human history. From the birth of grain farming through to the present day for most people, breads, or at least foods with grains as a staple ingredient, remain a primary source of calories.

The sheer amount of labor devoted to trying to meet our insatiable need for flour is awesome to behold when you examine the process. Only a farming family that mills its own grain and bakes its own bread probably can fully understand the weight of this labor.

But we can approach some portion of an understanding through reading and research.

Setting aside (for now) the many processes that resulted in your possession of bushels of grain, you face the next task of turning the grain into flour. Milling grain was a labor-intensive process, until craftworkers and engineers figured out how to use cams (shaped pieces of metal that move other pieces of metal). On Google, look at the images or videos for pear-shaped cams, for example, and understand that these pieces of metal were literally life-savers. But they also took humanity quite some time to figure out.

You need tools to convert unmilled grain into milled grain, or flour. According to one source (Gimpel), if two people labor over a rotary hand-mill for ten hours, they can mill seventy kilograms (~154.34 pounds) of grain into flour. (See Google for images of the old stone hand-mills.)

Next, let’s consider the conversion ratios and daily requirements for one soldier. According to Donald W. Engels, you lose some mass at each step of the conversion. So 3.6 pounds of unmilled grains processes into 3.3 pounds of milled grains, and 3.3 pounds of milled grain bakes into three pounds of bread. An active soldier needs three pounds of bread as a minimum daily ration.

So by my calculation, one hour of work by two people with a hand-mill provides about two days of bread for each person.

Your 154 pounds of grain, or ten hours of work by two people, results in about 138 pounds of bread. You can provide bread for two soldiers for a little over three weeks.

But a water wheel, cam shafts, and larger mill stones make mass production more feasible. A mill with one seven-foot water-powered stone might be able to process 150 kilograms of grain per hour (Gimpel 8). The largest Roman milling complex could process approximately 2,400 to 3,200 kilograms of grain during a ten-hour period (Gimpel 9).

Since mills also could speed the processes of metal forging, fulling, and paper-making, they spread across every medieval water system with continuous annual flow. Gimpel writes that in England, for example, the Domesday Book recorded 5,624 mills operating on 9,250 manors, and about a third of the manors had one or more mills in operation (12).

The key takeaways for my writing, or for realistic medievalist fiction, are the amount of labor necessary to make bread and the quantity of bread necessary to feed a soldier or an army.

Even with normal practices like having soldiers carry three days of rations in their packs or having a wagon or mule train carrying about seven days of rations, an army on the march generally needed daily down time to make bread as well as constant resupply.

Given the amount of weight involved and the availability of an adequate local supply of grains, a commander’s options often narrowed to a choice between quick in-and-out cavalry raids or an army’s slow march, ideally along a river or a coastline.

See Jean Gimpel’s The Medieval Machine: The Industrial Revolution of the Middle Ages or Cathedral, Forge, and Waterwheel: Technology and Invention in the Middle Ages by Frances & Joseph Gies for further reading.

[An aside: I finished a substantive revision of Book Three yesterday. Now, I’ll ask my first and second readers for their feedback, and then I’ll do more revision and polishing editing. So far, I’m definitely on pace to meet my goal of publishing book three before year’s end. I originally set a December 2025 date, but I may be able to move the publication date forward into October or November 2025. We’ll see.]