East-west diffusion

Why did crops and inventions spread further along a continent's width than along its length?

Put two maps side by side. The wheat-and-barley package domesticated in the Fertile Crescent turns up, over a few thousand years, in Greece, in Egypt, in the Indus valley, eventually in Ireland — a span of thousands of kilometres, mostly east and west. Maize, domesticated in southern Mexico, took far longer to become a staple in eastern North America, and the llama never made the journey north at all.

The distances are not obviously different. So the interesting question is not how far these things travelled but what they had to survive on the way — and whether the direction of travel changes that.

Start with what a crop actually is. Not a technique, but a population of organisms tuned by generations of selection to a particular regime: a length of growing season, a temperature range, a rainfall pattern, and — less obviously — a day length, since many plants use the changing photoperiod as the cue to flower.

Now move that population. Travel a thousand kilometres east and almost none of those variables change much: same latitude, same day length through the year, broadly similar seasonality. Travel a thousand kilometres north or south and every one of them shifts at once. The plant may germinate and grow and simply never flower, because the signal it is waiting for never arrives.

What follows for the two kinds of journey? Along a latitude, a crop can move as itself — carried, planted, and productive on arrival. Across latitudes, it must be re-bred: generations of selection for a shorter season or a different photoperiod response before it is worth growing. That is not impossible, only slow. Maize did eventually reach from Chile to Canada, and the varieties at the ends of that range differ substantially from their Mexican ancestors, which is exactly what the argument predicts.

This is Jared Diamond's continental axis argument. Eurasia's long dimension runs east-west; the Americas and Africa run north-south, and both are also pinched by barriers where the axis is narrowest — the Sahara, and the tropical lowlands and narrow isthmus between Mesoamerica and the Andes. His conclusion is that Eurasia's founder crops and livestock could accumulate into a single package shared across a continent, while the Americas' two great agricultural centres largely developed separately.

Does the argument extend to inventions, as the question assumes? Here I would be more cautious than the popular version is. A wheel does not care about photoperiod. What a technique needs is not a matching climate but people to carry it, and the same corridor that suits crops also suits caravans and conquering armies, so contact and crops travel together. That makes the axis effect on technology real but indirect — it works through the traffic, not through the technology — and correspondingly weaker as an explanation.

And there are honest problems with the thesis itself. It has been criticised as environmental determinism — James Blaut's objection was that it explains European dominance by geography while quietly reproducing the conclusion it set out to explain. The east-west corridor of Eurasia is itself crossed by the Himalayas, the Central Asian deserts and the steppe, so latitude alone hardly makes it open. The Americas' diffusion story is messier than a simple axis account suggests, with real long-distance movement of crops and considerable local domestication rather than blockage. And the argument is difficult to test: the number of continents is small, the cases are not independent, and it is easy to select the examples that fit.

The claim worth defending is narrower than the slogan. It is not that crops cannot cross latitudes; plainly they do. It is that crossing latitudes imposes an adaptation cost that crossing longitudes does not, and that a cost paid repeatedly over thousands of years shows up as a difference in the rate at which agricultural packages assemble and spread.

Two refinements connect the pieces. First, direction is not the only variable in play: barriers — desert, mountain, isthmus, sea — matter at least as much, and the strongest cases against north-south diffusion tend to be places where a barrier and a latitude change coincide. Second, the mechanism is strongest for domesticates, weaker for techniques, and weakest of all as an account of who ended up ruling whom, which is where most of the argument's critics are aiming.

The temptation to resist is treating the axis as a master key. It is a real constraint on one class of things, doing real work over long timescales, embedded in a case that remains genuinely contested among historians and geographers.

• Why so few large mammals were domesticable, and how that interacts with the axis argument.
• How photoperiod sensitivity was bred out of modern crops, and what that did to their range.

Every term the collection defines is gathered in the glossary.