Why are humans the only creatures to share food with strangers? Karen Gold ponders over a degustation of ancient dishes with an archaeologist.

To the squeamish and the vegetarian: look away now. For this is a story of bloody carcasses, rotting vegetation, decayed stomachs and fossilised faeces. And a dinner. And what it means to be human. Martin Jones's local butchers looked at him very strangely when asked to supply air-dried horse (bresaola) for a historical feast. Nevertheless, they arranged to fly it in from Italy. They drew the line only at his further requests for gazelle (a protected species) and then antelope - so Jones sourced that from South African meat suppliers via the web instead.

Given that the book Feast: Why Humans Share Food by the Pitt-Rivers professor of archaeological science at Cambridge University lists sow's udders and sea-scorpion among known Roman delicacies, my situation as his dinner guest was worrying. What would be on the menu? How far would prehistoric veracity take his male inclination to share the horsemeat kill with me? In the event, my worst concern was that he really had used a wild bird's egg to mix the mustard-seed cake from a 11,000-year-old recipe. (He didn't: it was the supermarket kind.) But the potential foreignness of content and manners in the meal (see the menu to the right) brought home how strange the act of sharing food is, how full of common, and commonly inexpressed, social, nutritional and sensory assumptions.

This is the point at the heart of his book. Primates share food only with close kin or for social/sexual/genetic self-interest; generally animal feeding is characterised by fields full of cows, all grazing alone. Yet we humans extend our biological need for food to the social activity of sharing it with strangers, placing it between us, making eye contact, baring our teeth and talking - all potentially threatening activities.

É«ºÐÖ±²¥

Jones has been interested in this nexus of biological and social humanity for 30 years, ever since as a natural sciences graduate he realised that his archaeology hobby might be a way of earning a living. Almost immediately, he says, he became intrigued by the archaeology of food.

Influenced by late 1960s anthropology, which redrew the picture of primitive societies as sophisticated entities, archaeology in the 1970s moved in the same direction. Finds that had gathered dust on shelves for decades suddenly yielded new information: pots told the story of cooking, tools and toothmarks of food retrieval and sharing. Tantalisingly, the social interactions of prehistory began to come into view.

É«ºÐÖ±²¥

And then came the science. As a bio-archaeologist, Jones was well placed to make the most of colleagues' new chemical and molecular analyses of microscopic food residues, human bones, teeth and hair, stomach contents and coprolites. This physiological evidence of ancient diet, combined with new perspectives on site contents and layout - cooking debris, hearth positioning - effects a new way of telling the story of human evolution. As the human brain enlarged, Jones explains, the stomach and teeth grew smaller, hence the need to develop forms of external digestion - chopping, rotting, fermentation - and ultimately - cooking. Tracing that story over huge timescales - "I don't tend to talk about Stone Age, Bronze Age, neolithic. I don't consciously avoid them. It's just that the kinds of questions I like involve the whole history of the human race" - Jones also picks out signs of key change. He notes the development of weaving, evidenced in Galilee 23,000 years ago, enabling the making of fishing nets; the shift from scattered to central hearths; symptoms of widespread malnutrition, seen in fossilised human teeth and hair, following the narrowing of diet and exposure to climatic variation as hunter-gatherers settled into agricultural society.

Many pieces of the jigsaw are missing - there is no evidence for food-mixing or cooking before pots to discover it in - yet the human urge to tell these stories is unavoidable both within the book's subject and in its shape. At the start of every chapter, Jones creates a fictionalised account of his excavated meals, painting a picture of social organisation and interaction around the food, homing in on detail - growth of food storage areas, better survival of the elderly on softened, cooked meat, bones worn from kneeling and kneading. "I wanted to get into the timescale of people eating a meal. But the fiction also forced me to ask other questions. If you go back to Boxgrove (the archaeological excavation site near Chichester), I spent ages trying to work out how the different bits of meat got to different places, how pregnant women got any meat at all. It's only when you build up a scene that you ask those questions, even if you have to fill in some gaps," Jones says.

He asked colleagues to read these mini-fictions before the book went to press. Some were the original excavators of the sites where the meals took place. He feared loss of professional credibility; instead, he says, he got enthusiasm, even when his readers did not necessarily agree with his interpretative detail. Archaeologists want to tell stories, he argues. "One colleague said: 'In future, we must do this in our site reports.'"

A feast is a story in itself, he adds. Its preparation involves narrative (add this, stir that, warm the oven first). Its consumption (unlike a meal, which can be eaten alone) is nothing without social ritual and conversation. We have feasts not only for calories but for cultural experience. The origin of the word "company" - " com panem " - literally means "with bread". The man carves (or used to) the roast. We eat the meat before the sweet. All these are stories.

As is the fact that both Jones and I, having dismissed the lentil cake on which whole civilisations have survived as virtually inedible, then gave up in the face of the fish.

É«ºÐÖ±²¥

Understanding the storytelling mode is a thread through the development of human cognition, Jones explains. "If you look at the flints Homo heidelbergensis were using at Boxgrove, they followed a format, almost as if they were made from a template. Now I'm working on a Moravian site about 25,000-30,000 years old, where people sat around campfires, looking at each other face to face and sharing food. They did other narrative things: decorating their bodies, weaving. Things you tell a story about. There's a whole series of vocalisations that primates make, but this idea of narrative, that you can tell someone else how to make a meal or where to find food, that makes us human."

Feast: Why Humans Share Food by Martin Jones is published by Open University Press on March 29, ?20.00.

A TASTE OF OUR FOREFAÉ«ºÐÖ±²¥RS' FOOD

The menu Martin Jones created for our dinner spanned half a million years of history. Each course was based on excavation in a specific place.

É«ºÐÖ±²¥

We started with horse, just like the huge one butchered by a group of flint-using Homo heidelbergensis and carried away, presumably for preserving, from the Boxgrove site (in Chichester) about 500,000 years ago.

Then we moved to 9000BC and the first-known working kitchen, in Jerf el-Ahmar in Syria, in which archaeologists found remains of cooked meat, nuts, seeds, cracked barley, mustard-seed cakes and signs of fermentation, implying wine's presence.

Wine was certainly on the menu in Roman Britain. Residues of fruit in the Colchester meal from about AD45 show how extensively Rome's far-flung colonies fed its ruling class. There is evidence of bread too now, though no clarity about the invention of leavening. Finally the fish - sprats, served whole - from a monastic meal in Picardie, around AD1372.

We ate our meal off leaves initially - even that probably a concession to modern manners - and with fingers, drinking from horn cups (Victorian, but evidence exists for ancient ones). During the first course, we also ate in silence, guarding our food, looking sideways and down, without eye contact: an eerie, strangely non-human experience.

É«ºÐÖ±²¥

But perhaps we got away lightly.

Meals mentioned in the book that might have appeared on the table but did not include tortoise, hedgehog, toad and lizard from Neanderthal Spain, and knotgrass from agriculturalised Denmark 2,000 years ago.