Looking at our peasant household, what we generally have are large families on small farms. The households in these farms were not generally nuclear households, but extended ones. Pre-Han Chinese documents assume a household to include three generations: two elderly parents, their son, his wife, and their four children (eight individuals total). Ptolemaic and Roman census data reveal a bewildering array of composite families, including multi-generational homes, but also households composed of multiple nuclear families of siblings (so a man, his wife, his brother and then brother’s wife and their children, for instance), and so on. Normal family units tended to be around eight individuals, but with wide variation (for comparison, the average household size in the United States for a family is 3.14).
At the same time that households were large (by modern standards), the farms they tilled were, by modern standards, very small. The normal size of a Roman household small farm is generally estimated between 5 and 8 iugera (a Roman measurement of land, roughly 3 to 5 acres); in pre-Han Northern China (where wheat and millet, not rice, were the staple crops), the figure was “one hundred mu (4.764 acres)” – essentially the same. In Languedoc, a study of Saint-Thibery in 1460 showed 118 households (out of 189) on farms of less than 20 setérée (12 acres or so; the setérée appears to be an inexact unit of measurement); 96 of them were on less than 10 setérée (about 6 acres). So while there is a lot of variation, by and large it seems like the largest cluster of household farms tend to be around 3 to 8 acres or so; 5 acre farms are a good “average” small farm.
This coincidence of normal farm size and family size is not an accident, but essentially represents multi-generational family units occupying the smallest possible farms which could support them. The pressures that produce this result are not hard to grasp: families with multiple children and a farm large enough to split between them might do so, while families without enough land to split are likely to cluster around the farm they have. Pre-modern societies typically have only limited opportunities for wage labor (which are often lower status and worse in conditions than peasant farming!), so if the extended family unit can cluster on a single farm too small to split up, it will (with exception for the occasional adventurous type who sets off for high-risk occupations like soldier or bandit).
Now to be clear that doesn’t mean the farm sizes are uniform, because they aren’t. There is tremendous variation and obviously the difference between a 10 acre small farm and a 5 acre small farm is half of the farm. Moreover, in most of the communities you will have significant gaps between the poor peasants (whose farms are often very small, even by these measures), the average peasant farmer, and “rich peasants” who might have a somewhat (but often not massively so) larger farm and access to more farming capital (particularly draft animals). […] Nevertheless, what I want to stress is that these fairly small – 3-8 acres of so – farms with an extended family unit on it make up the vast majority of farming households and most of the rural population, even if they do not control most of the land (for instance in that Languedoc village, more than half of the land was held by households with more than 20 setérée a piece, so a handful of those “rich peasants” with larger accumulations effectively dominated the village’s landholding […]).
This is our workforce and we’re going to spend this entire essay talking about them. Why? Because these folks – these farmers – make up the majority of the population of basically all agrarian societies in the pre-modern period. And when I say “the majority” I mean the vast majority, on the order of 80-90% in many cases.
Bret Devereaux, “Collections: Bread, How Did They Make It? Part I: Farmers!”, A collection of Unmitigated Pedantry, 2020-07-24.
