Food around the world

For most of human history people fed themselves by hunting animals, fishing and gathering wild plants. A hunter-gatherer food system depends on detailed knowledge of the local environment: which plants are edible in each season, where animals move, and how to prepare food. Processing was not confined to farmers: Revedin and colleagues’ 2010 study identified wild-plant starch on grinding tools from three European sites dating to about 30,000 years ago, evidence of plant processing and possible flour production.

Mobility, diet, storage and population size differed between places and periods. Seasonal movement could help people reach different resources, but it should not be treated as a universal rule. For example, archaeological evidence from the early Natufian period in the Levant indicates occupation of substantial buildings for much of the year while people used wild foods.

Seasonal shortages remained a possible challenge; assess the available resources and storage practices in the particular community rather than assuming a single hunter-gatherer diet or way of life.

Some of the earliest agricultural food systems emerged roughly 10 000 to 12 000 years ago; domestication occurred independently in several regions at different times. Consider three interacting sets of factors rather than a single inevitable cause. First, local climate and water conditions affected food resources. For example, increased rainfall and climatic stabilisation after the Younger Dryas accompanied more settled living in the Jordan Valley; this regional sequence is not a universal explanation for farming everywhere.

Second, some regions contained wild species suited to domestication: plants that people could cultivate for food (wheat, barley and lentils in south-west Asia; rice and millet in China; maize, beans and squash in Mesoamerica; potatoes in the Andes). Selection affected traits such as seed retention in cereals; not all crops were grown for edible seeds. Animal traits that could favour management included social organisation, lower reactivity to humans and the ability to reproduce under human care.

Sheep, goats, cattle and pigs became livestock, but they did not all share the same feeding habits: pigs are omnivores. Third, knowledge and management practices mattered: repeated tending, harvesting and propagation changed relationships with food species. Cultivation could begin among mobile as well as more settled communities. Domestication involved changes across generations, not a single invention or simply taming one animal.

Early farming had clear potential advantages for communities. Grain could be stored, so a surplus could carry a village through winter or a poor season and feed people who were not themselves producing food, such as potters, builders or priests. Agriculture contributed to the growth of settlements and populations, but storage and more settled living did not begin only after domestication.

Kuijt and Finlayson’s 2009 study of Dhraʹ in Jordan describes purpose-built granaries about 11,300–11,175 calibrated years before present, before domesticated cereals; the evidence concerns cultivation of wild plants during a gradual transition. Farming also created new challenges in feeding populations. Heavy dependence on a few staples could make a failed harvest dangerous if other foods, stores or exchange could not make up the loss. Health comparisons need regional evidence.

Stock and colleagues’ 2023 skeletal study found that decreases in body size preceded farming in some regions, while stature and body mass remained relatively stable through the transition in the Levant and China. A shorter skeleton alone therefore does not demonstrate that farming caused poor nutrition. Dense settlements could favour disease transmission; cultivation also required labour, and repeated cropping could deplete soil nutrients.

These risks should not be turned into a universal claim that every farmer was less healthy or worked longer than every forager. Both systems have attributes and challenges, so a strong answer compares them rather than treating one as simply superior.