Trade diasporas

Why did long-distance trade so often pass through settled communities of resident foreigners?

The pattern repeats across places that never influenced one another: Armenian merchants from New Julfa working the Indian Ocean, Hanseatic factors lodged in their compounds at Bergen and Novgorod, German traders in the Venetian fondaco, Hausa kola networks across the savannah, Chinese communities in the port towns of Southeast Asia. In each case the trade ran through a settled colony of people who were not from the place they sat in. That is an odd shape. If you want to sell cloth in a distant city, why not simply sell it to somebody who lives there?

Because you cannot be there. That is the whole difficulty, and it is easy to underrate. When a cargo takes months to arrive, the price it will fetch is unknown at the moment you dispatch it, so somebody at the far end must hold it, judge the market, choose the moment and set the price. You are delegating discretion across a distance you cannot see over.

Consider what your agent can do with that discretion. Report a lower price than he obtained. Claim spoilage. Sell to a relative below market. Or, once, simply keep the lot. And consider your remedies. Sue him — in whose court, under whose law, and with what power to compel a man in another sovereign's city? Cross-jurisdiction enforcement is precisely what did not exist. Take a bond — but he must hold wealth where you can seize it. Rely on repeat dealings with you alone — but your trade with him is intermittent, and a single cargo may be worth more than a lifetime of commissions.

So the discipline cannot be bilateral. It has to be multilateral: news of a cheat must reach everyone who might ever employ him, and they must all refuse. That needs three things at once — a channel carrying credible news faster than he can rebuild elsewhere, a shared understanding of what counts as cheating, and members who value things the group controls beyond this one deal.

Now look at what a community of resident kinsmen and co-religionists supplies. Correspondence flows along it constantly, carrying prices and conduct in the same letters. It has its own norms and elders to apply them. And it holds marriage prospects, credit, worship and burial. The direction of that argument is the mechanism: the community was not built for trade. It already existed, and trade repurposed it, which is why the arrangement was cheap enough to appear independently in so many places.

That explains the two features that looked strange. Why resident rather than visiting? Because catching a market needs continuous presence — and continuous presence is what dissolves a man's ties to home, so the network had to keep an active hold on him through correspondence, marriage back into the group, and often family left behind. Why foreign rather than a local hire? Because a local hire stands outside the punishment network, and unreachability is the one thing you cannot tolerate.

Two corrections, one of them to my own account.

The first is that the host ruler wanted this shape too, so a story told from the merchants' side is half of it. A walled compound — fondaco, funduq, kontor — is taxable at one gate, searchable, and its occupants are a standing hostage for their home city's good behaviour. Rulers granted quarters and internal jurisdiction as a bargain, not a tolerance, which is why these communities so often have an address rather than being diffuse.

The second is that the reputation mechanism is genuinely contested. Greif's influential reading of the eleventh-century Maghribi traders as a coalition enforcing by collective boycott has been challenged directly — Edwards and Ogilvie argued from the same letters that these merchants also used courts and did trade with outsiders. The honest position is that private reputation and public law were complements more often than substitutes, and the strong claim that diasporas existed because law was absent outruns the evidence. Competing explanations sit alongside it: brokerage between languages, currencies and legal customs, which needs bilingual residents whatever the enforcement story; and exclusion, where host societies barred minorities from land or office and pushed them into trade rather than trade selecting them.

The load-bearing claim is narrower and testable. If the diaspora form is doing agency work, it should erode where states began supplying enforceable contract and cheap verifiable information — consular courts, the telegraph, marine insurance, graded commodities, the bill of lading — and persist longest in goods whose quality cannot be verified at a distance, such as gems, art and uncollateralised credit. The refuting observation would be sustained, large-scale trade in hard-to-verify goods between strangers with neither a shared community nor a reachable court. As for how large these communities were, the surviving counts are tax or lodging records of unknown coverage, so I will not give numbers.

• How the commenda and similar partnership contracts split profit and loss to align an agent's incentives inside the same networks.
• Why several of these communities converted from trade to finance as impersonal enforcement arrived, rather than dissolving.

Every term the collection defines is gathered in the glossary.