Immanuel Wallerstein and the system that makes the parts

Ask why Britain grew rich and Ghana did not, and you will usually be handed a comparison. Two countries, two sets of institutions, two endowments of coal and schooling. Line them up, hold what you can constant, and watch which variable moves. That is how development was taught for most of the twentieth century, and it is not stupid.

Immanuel Wallerstein's objection was that it smuggles in its own conclusion. To compare Britain and Ghana as two instances of one kind of thing — the national economy, arriving at modernity early or late — you must first assume they are separable. Wallerstein thought they had one history between them. The cotton that fed Lancashire and the terms on which it was grown elsewhere were a single arrangement, described from opposite ends.

He was not saying that Britain robbed Ghana, though he thought a great deal of robbery had occurred. He was saying the unit of analysis had been chosen badly: if wealth and poverty are produced by the same process, the sentence "Ghana failed to develop" has no clear subject, because no Ghana-sized process existed that could have succeeded. The replacement is duller and harder: what does a position in a world division of labour require of whoever occupies it, and what must change, and where, for the occupant to move?

He found the world-system by way of Africa, and the route matters#

Wallerstein was born in New York on 28 September 1930 and educated at Columbia: BA in 1951, MA in 1954, doctorate in 1959, with Korean War service in between. The dissertation was on Ghana and the Ivory Coast, and became The Road to Independence in 1964. Africa: The Politics of Independence had appeared in 1961, Africa: The Politics of Unity in 1967; he chaired the African Studies Association in 1973. Decolonisation had handed him a puzzle the comparative method could not hold: a colony became a state, acquired a flag, a seat at the United Nations, a central bank, and its economic position did not move. If sovereignty were the variable that mattered, something ought to have shifted. Mostly nothing did.

He taught at Columbia from 1958 to 1971, at McGill until 1976, and then at Binghamton, as Distinguished Professor of Sociology, until 1999. There in 1976 he founded the Fernand Braudel Center for the Study of Economies, Historical Systems, and Civilizations, running it until 2005 and editing its journal, Review. He was president of the International Sociological Association from 1994 to 1998, and from 2000 a senior research scholar at Yale. He died at Branford, Connecticut, on 31 August 2019, at 88.

One Columbia detail is hard to unsee. He was on the faculty during the occupation of 1968 and published University in Turmoil: The Politics of Change the following year; the world revolution he would later treat as a structural event had, for him, happened in his own building.

Volume I dated the beginning of capitalism to the wrong century, deliberately#

The Modern World-System, vol. I: Capitalist Agriculture and the Origins of the European World-Economy in the Sixteenth Century appeared with Academic Press in 1974. Volume II, on mercantilism and consolidation from 1600 to 1750, followed in 1980; volume III, on the second era of great expansion from the 1730s to the 1840s, in 1989; volume IV, Centrist Liberalism Triumphant, 1789-1914, in 2011, by then with California. Thirty-seven years separate the first from the last. He was eighty-one when IV appeared, and the series stops there.

The provocation is chronological. Capitalism, on the standard account, begins with the factory, in England, around 1780. Wallerstein put its origins in the long sixteenth century, in an agrarian Europe of grain, sugar and silver, and located them in a world-economy rather than a country: one economic division of labour spanning many political units with no roof over it. That feature carries the definition. Empires had a single authority and a single economy; this had one economy and many authorities, so capital could play states off against each other and no power could tax the system into stillness. The debt to Fernand Braudel is plain, and Wallerstein put the man's name over the door of his research centre.

The semi-periphery is the load-bearing wall, not the waiting room#

The three-zone vocabulary is the part of Wallerstein that travelled, and the part most often garbled. Core, semi-periphery and periphery are not a league table of countries, and they are not geography. They are positions in a division of labour, defined by the production concentrated in them: core-like processes are relatively monopolised and so high-profit, peripheral ones competitive and so low-profit, and exchange between them moves surplus steadily one way. Countries are places where mixtures of these processes sit; a country's zone is a summary of its mix, which is why movement between zones is possible and does happen.

That middle zone is where the theory does its work, and the part readers skip. In the theoretical reprise closing volume I, Wallerstein is blunt:

The semiperiphery is a necessary structural element in a world-economy... These middle areas (like middle groups in an empire) partially deflect the political pressures which groups primarily located in peripheral areas might otherwise direct against core-states.

The argument is political rather than economic. A two-tier system is a standing invitation to unified opposition: everyone at the bottom has the same enemy. Three tiers give you a middle stratum with something to lose and something to hope for, whose interests align with neither end, and which therefore breaks up the coalition that would otherwise form. Stability, on this account, comes from the geometry of stratification rather than from consent or force. It is the claim his sharpest critic went straight for.

The disciplines are nineteenth-century furniture, and we are still sitting on them#

Wallerstein spent his last decades arguing about departments. Economics to the economists, politics to the political scientists, the past to the historians: he held that the carve-up told you nothing about how social reality is jointed and a great deal about the nineteenth-century liberal state, which needed market, state and civil society to be separate spheres and built universities that taught them separately. If the divisions are products of the system, they are not neutral instruments for studying it.

The institutional form of that argument was the Gulbenkian Commission on the Restructuring of the Social Sciences, which he chaired: ten members, six social scientists and two natural scientists and two from the humanities, meeting in 1994 and 1995 on Calouste Gulbenkian Foundation money. Its report, Open the Social Sciences (Stanford, 1996), runs to about a hundred pages and has been translated into seventeen languages, startling for a committee document on curriculum design. World-Systems Analysis: An Introduction (Duke, 2004) is shorter still at 128 pages, half of it about where the modern disciplines came from rather than about trade.

That makes him a theorist of knowledge as much as of the economy, nearer to Foucault on the production of the normal and to the Cambridge school's history of political languages than his reputation suggests. It also leaves a problem he never closed: if the categories of social science are artefacts of the system, so is world-systems analysis. He conceded that more readily than his opponents allow, and never answered it. Anyone watching the current round of department closures in Britain will notice that his case against disciplinary borders is now being made, with rather different motives, by finance officers.

1968 was a verdict on the movements that had already won#

The best thing he wrote about politics concerns the movements he sympathised with. The antisystemic movements of the nineteenth and twentieth centuries — communist parties, social democracy, national liberation — differed in almost everything except strategy. All three worked in two steps: take state power first, transform the world afterwards.

By the middle of the twentieth century, step one was substantially complete. Communist parties governed from Berlin to Beijing, social democrats held office across western Europe, national liberation movements had inherited the colonial administrations. Step two did not arrive. What arrived instead were governments with budgets to balance and foreign creditors to satisfy, inside a world division of labour their victory had not touched.

That, in his reading, is what 1968 was about, and why the revolt was aimed at least as much at the Old Left as at capital. He put the case in The Agonies of Liberalism: What Hope Progress?, in New Left Review I/204 in the spring of 1994, treating the year as the hinge between an era in which liberal progress could still be believed and one in which it could not. His account of the mechanics is deflationary to the point of rudeness: expanded higher education produced graduates in numbers, the promised jobs and status did not follow at the same rate, and the universities were where the trouble started. The most Wallersteinian thing about the argument is that it treats the movements' success as the explanation of their failure.

The objections were serious, and at least two of them landed#

Robert Brenner's The Origins of Capitalist Development: A Critique of Neo-Smithian Marxism, in New Left Review I/104 for July-August 1977, remains the most damaging thing written about him. The charge is that Wallerstein, like Andre Gunder Frank, locates the dynamic of capitalism in trade rather than in the relations of production, which makes him a Smithian in Marxist dress. "Trade cannot determine the transformation of class relations of production," Brenner wrote. What makes a system capitalist is a class structure in which producers have been separated from their means of subsistence and all are compelled to compete; markets alone do no such thing. His counter-example is hard to shrug off. Poland's grain estates in the sixteenth century were deeply integrated into the European market, exactly as Wallerstein described, and the result was not wage labour but a tightening of serfdom.

The Wallersteinian reply is that Brenner's definition restores the national container, turning capitalism into an English event that later diffused, which is the framing under dispute. Both halves have force, which is why the argument still runs fifty years on, and why anyone arriving from Marx on the commodity will find the two sides answering different questions.

Theda Skocpol's critique, in the American Journal of Sociology of March 1977 (vol. 82, no. 5, pp. 1075-90), goes at the structure of the explanation rather than its content. Her first charge is functionalism: parts of the system are accounted for by their contribution to reproducing the whole, which tells you what a thing is for and passes it off as an account of how it came to be. The semi-periphery is the clearest case. To say that a middle zone exists because it deflects pressure from the core is to name a consequence and call it a cause. Her second charge is that the states in his model are thin, largely transmission belts for economic position, with military and administrative capacity given far less weight than her own work on revolutions gives them. Both charges stuck.

The third objection is the cheapest and the most persistent: that the framework explains everything and therefore predicts nothing. A country rises and it is mobility within the system; a country falls and it is hierarchy reproducing itself; calm is equilibrium, turbulence is structural crisis. There is something in this, though not everything. In July 2002, in Foreign Policy, Wallerstein published "The Eagle Has Crash Landed", arguing that the United States had been declining as a global power since the 1970s and that the response to the 11 September attacks would accelerate it. Bush's approval rating stood above 70 per cent. One may think the essay wrong. It was not unfalsifiable, and it was not safe.

What the framework is good for now, and where it turns into hope#

The test case is China. Marilyn Grell-Brisk, in Palgrave Communications in September 2017, clustered national income data and found China leaving in the mid-1990s the peripheral grouping it had shared with India, entering the semi-periphery around 2000, and sitting fully in the middle stratum by 2006. Read one way that vindicates the categories: they are what made the movement legible as something other than growth. Read another way it strains them, since Grell-Brisk also finds the three-layered distribution giving way to a multimodal one, with far more of the world's population in the middle. A semi-periphery holding a fifth of humanity is not obviously doing the small buffering job Wallerstein assigned it.

Elsewhere the fit is better. Supply chains are the framework's natural object: what the pandemic and the export-control years exposed was that some steps in a chain are monopolised while others are commodity-competitive, and the margin sits with the monopoly. Labour moves toward the zones that command it. The costs of a warming climate fall heaviest where the profits from carbon did not accumulate. And the present argument about American hegemony is conducted in something close to Wallerstein's vocabulary by people who have never read him.

His late claim was that the system has entered a structural crisis, and it is routinely misread as a collapse theory. He set it out in Monthly Review in March 2011, in an essay called "Structural Crisis in the World-System: Where Do We Go from Here?". The ordinary cyclical corrections have stopped restoring equilibrium; borrowing Ilya Prigogine's language, the system approaches a bifurcation, "a chaotic situation in which a stable equilibrium cannot be maintained". He dated the crisis from the 1970s and guessed it would run to about 2050. "The only certainty is that the existing system... cannot survive. What is impossible to know is what the successor system will be." Two camps contend over the outcome: a spirit of Davos, wanting a non-capitalist order that keeps hierarchy and polarisation intact, and a spirit of Porto Alegre, wanting one relatively democratic and egalitarian.

Half of that is a real analytical claim, and a sharp one: a system far from equilibrium behaves differently from a system near it, so the methods that work in ordinary times stop working. That is a serious limit on what social science can say.

The other half shades into consolation. "When the system is far from equilibrium, every little input has great effect" is the sentence his readers quote, and it does more work than it can carry. Undetermined means undetermined. It does not follow that your effort is likelier to be decisive, and the step from ignorance to optimism is taken without argument. Nor is the periodisation checkable: a crisis due to resolve around 2050 cannot be assessed by anyone reading the essay in 2026, which is a convenient property for a prediction to have.

Wallerstein published a numbered commentary twice a month for over twenty years. No. 500, dated 1 July 2019, was the last, and he died two months later. "I still think that," he wrote, "and therefore I think there is a 50-50 chance that we'll make it to transformatory change, but only 50-50." An even-money bet, delivered with the flatness of a finding, is an odd close to four decades of systemic analysis. It does at least mark the seam between the part of the work that is argument and the part that is a wager. What survives, whatever becomes of the boxes and arrows, is the habit he brought back from Accra. Before explaining anything, know what the thing is you are explaining, and be ready for it to be much larger than the map on the wall.