Immanuel Wallerstein on "the appropriate unit of analysis"

Louis Proyect lnp3 at
Mon Oct 23 12:46:33 MDT 2000

>From Immanuel Wallerstein's "The Rise and Future Demise of the World
Capitalist System: Concepts for Comparative Analysis" (1972)

Nothing illustrates the distortions of ahistorical models of social change
better than the dilemmas to which the concept of stages gives rise. If we
are to deal with social transformations over long historical time
(Braudel’s "the long term"), and if we are to give an explanation of both
continuity and transformation, then we must logically divide the long term
into segments in order to observe the structural changes from time A to
time B. These segments are, however, not discrete but continuous in
reality; ergo they are "stages" in the "development" of a social structure,
a development which we determine however not a priori but a posteriori.
That is, we cannot predict the future concretely, but we can predict the past.

The crucial issue when comparing "stages" is to determine the units of
which the "stages" are synchronic portraits (or "ideal types," if you
will). And the fundamental error of ahistorical social science (including
ahistorical versions of Marxism) is to reify parts of the totality into
such units and then to compare these reified structures.

For example, we may take modes of disposition of agricultural production,
and term them subsistence-cropping and cash-cropping. We may then see these
as entities which are "stages" of a development. We may talk about
decisions of groups of peasants to shift from one to the other. We may
describe other partial entities, such as states, as having within them two
separate "economies," each based on a different mode of disposition of
agricultural production. If we take each of these successive steps, all of
which are false steps, we will end up with the misleading concept of the
"dual economy" as have many liberal economists dealing with the so-called
underdeveloped countries of the world. Still worse, We may reify a
misreading of British history into a set of universal "stages" as Rostow does.

Marxist scholars have often fallen into exactly the same trap. If we take
modes of payment of agricultural labor and contrast a "feudal" mode wherein
the laborer is permitted to retain for subsistence a part of his
agricultural production with a "capitalist" mode wherein the same laborer
turns over the totality of his production to the landowner, receiving part
of it back in the form of wages, we may then see these two modes as
"stages" of a development. We may talk of the interests of "feudal"
landowners in preventing the conversion of their mode of payment to a
system of wages. We may then explain the fact that in the twentieth century
a partial entity, say a state in Latin America, has not yet industrialized
as the consequence of its being dominated by such landlords. If we take
each of these successive steps, all of which are false steps, we will end
up with the misleading concept of a "state dominated by feudal elements,"
as though such a thing could possibly exist in a capitalist world-economy.
But, as Andre Gunder Frank has clearly spelled out, such a myth dominated
for a long time "traditional Marxist" thought in Latin America.

Not only does the misidentification of the entities to be compare lead us
into false concepts, but it creates a non-problem: can stages skipped? This
question is only logically meaningful if we have" that "co-exist" within a
single empirical framework. If within a capital world-economy, we define
one state as feudal, a second as capitalist, an a third as socialist, then
and only then can we pose the question: can country "skip" from the feudal
stage to the socialist stage of national development without "passing
through capitalism"?

But if there is no such thing as "national development" (if by that" mean a
natural history), and if the proper entity of comparison is world-system,
then the problem of stage-skipping is nonsense. If a can be skipped, it
isn’t a stage. And we know this a posteriori.

If we are to talk of stages, then—and we should talk of stages—it must be
stages of social systems, that is, of totalities. And the only totalities
that exist or have historically existed are mini-systems and world systems,
and in the nineteenth and twentieth centuries there has been only one
world-system in existence, the capitalist world-economy.

We take the defining characteristic of a social system to be the existence
within it of a division of labor, such that the various sectors or areas
are dependent upon economic exchange with others for the smooth and
continuous provisioning of the needs of the area. Such economic exchanges
can clearly exist without a common political structure and even more
obviously without sharing the same culture.

A mini-system is an entity that has within it a complete division of labor,
and a single cultural framework. Such systems are found only in very simple
agricultural or hunting and gathering societies. Such mini-systems no
longer exist in the world. Furthermore, there were fewer in the past than
is often asserted, since any such system that became tied to an empire by
the payment of tribute as "protection costs" ceased by that fact to be a
"system," no longer having a self-contained division of labor. For such an
area, the payment of tribute marked a shift, in Polanyi’s language, from
being a reciprocal economy to participating in a larger redistributive

Leaving aside the now defunct mini-systems, the only kind of social system
is a world-system, which we define quite simply as a unit with a single
division of labor and multiple cultural systems. It follows logically that
there can, however, be two varieties of such world-systems, one with a
common political system and one without. We shall designate these
respectively as world-empires and world-economies.

It turns out empirically that world-economies have historically been
unstable structures leading either towards disintegration or conquest by
one group and hence transformation into a world-empire. Examples of such
world-empires emerging from world-economies are the so-called great
civilizations of pre-modern times, such as China, Egypt, Rome (each at
appropriate periods of its history). On the other hand, the so-called
nineteenth-century empires, such as Great Britain or France, were not
world-empires at all, but nation-states with colonial appendages operating
within the framework of a world-economy.

World-empires were basically redistributive in economic form. No doubt they
bred clusters of merchants who engaged in economic exchange (primarily
long-distance trade), but such clusters, however large, Were a minor part
of the total economy and not fundamentally determinative of its fate. Such
long-distance trade tended to be, as Polanyi argues, "administered trade"
and not market trade, utilizing "ports of trade"

It was only with the emergence of the modern world-economy in
sixteenth-century Europe that we saw the full development and economic
predominance of market trade. This was the system called capitalism.
Capitalism and a world-economy (that is, a single division labor but
multiple polities and cultures) are obverse sides of the same coin. One
does not cause the other. We are merely defining the same in. divisible
phenomenon by different characteristics.

How and why it came about that this particular European world economy of
the sixteenth century did not become transformed into are distributive
world-empire but developed definitively as a capitalist world-economy I
have explained elsewhere. The genesis of this world historical
turning-point is marginal to the issues under discussion in paper, which is
rather what conceptual apparatus one brings to bear the analysis of
developments within the framework of precisely such capitalist world-economy.

Let us therefore turn to the capitalist world-economy. We shall to deal
with two pseudo-problems, created by the trap of not analyzing totalities:
the so-called persistence of feudal forms, and the so-called creation of
socialist systems. In doing this, we shall offer an alternative model with
which to engage in comparative analysis, one rooted in the historically
specific totality which is the world capitalist economy. We hope to
demonstrate thereby that to be historically specific is not to fail to be
analytically universal. On the contrary, the only road to nomothetic
propositions is through the historically concrete, just as in cosmology the
only road to a theory of the laws governing the universe is through the
concrete analysis of the historical evolution of this same universe.

On the "feudalism" debate, we take as a starting-point Frank’s concept of
"the development of underdevelopment," that is, the view that the economic
structures of contemporary underdeveloped countries is not the form which a
"traditional" society takes upon contact with "developed" societies, not an
earlier stage in the "transition" to industrialization. It is, rather, the
result of being involved in this world-economy as a peripheral, raw
material producing area, or as Frank puts it for Chile, "underdevelopment .
. . is the necessary product of four centuries of capitalism itself."

This formulation runs counter to a large body of writing concerning the
underdeveloped countries that was produced in the period 1950—70, a
literature which sought the factors that explained "development" within
non-systems such as "states" or "cultures" and, once having presumably
discovered these factors, urged their reproduction in underdeveloped areas
as the road to salvation.

Frank’s theory also runs counter, as we have already noted, to the received
orthodox version of Marxism that had long dominated Marxist parties and
intellectual circles, for example in Latin America. This older "Marxist
view of Latin America as a set of feudal societies in a more or less
pre-bourgeois stage of development has fallen before the critiques of Frank
and many others as well as before the political reality symbolized by the
Cuban revolution and all its many consequences. Recent analysis in Latin
America has centered instead around the concept of "dependence."

However, recently Ernesto Laclau has made an attack on Frank which, while
accepting the critique of dualist doctrines, refuses to accept the
categorization of Latin American states as capitalist. Instead Laclau
asserts that "the world capitalist system . . . includes, at the level of
its definition, various modes of production." He accuses Frank of confusing
the two concepts of the "capitalist mode of production" and "participation
in a world capitalist economic system."

Of course, if it’s a matter of definition, then there can be no argument.
But then the polemic is scarcely useful since it is reduced to a question
of semantics. Furthermore, Laclau insists that the definition is not his
but that of Marx, which is more debatable. Rosa Luxemburg put her finger on
a key element in Marx’s ambiguity or inconsistency in this particular
debate, the ambiguity which enables both Frank and Laclau to trace their
thoughts to Marx:

"Admittedly, Marx dealt in detail with the process of appropriating
non-capitalist means of production [N.B., Luxemburg is referring to primary
products produced in peripheral areas under conditions of coerced labor—
I.W.] as well as with the transformation of the peasants into a capitalist
proletariat. Chapter XXIV of Capital, Vol. 1, is devoted to describing the
origin of the English proletariat, of the capitalistic agricultural tenant
class and of industrial capital, with particular emphasis on the looting of
colonial countries by European capital. Yet we must bear in mind that all
this is treated solely with a view to so-called primitive accumulation. For
Marx, these processes are incidental, illustrating merely the genesis of
capital, its first appearance in the world; they are, as it were, travails
by which the capitalist mode of production emerges from a feudal society.
As soon as he comes to analyze the capitalist process of production and
circulation, he reaffirms the universal and exclusive domination of
capitalist production [N.B., that is, production based on wage labor—I.W.]."

There is, after all, a substantive issue in this debate. It is in fact the
same substantive issue that underlay the debate between Maurice Dobb and
Paul Sweezy in the early 1950s about the "transition from feudalism to
capitalism" that occurred in early modern Europe. The substantive issue, in
my view, concerns the appropriate unit of analysis for the purpose of
comparison. Basically, although neither Sweezy nor Frank is quite explicit
on this point, and though Dobb and Laclau can both point to texts of Marx
that seem clearly to indicate that they more faithfully follow Marx’s
argument, I believe both Sweezy and Frank better follow the spirit of Marx
if not his letter and that, leaving Marx quite out of the picture, they
bring us nearer to an understanding of what actually happened and is
happening than their opponents.

What is the picture, both analytical and historical, that Laclau
constructs? The heart of the problem revolves around the existence of free
labor as the defining characteristic of a capitalist mode of production:

"The fundamental economic relationship of capitalism is constituted by the
FREE [Wallerstein's emphasis] labourer’s sale of his labour-power, whose
necessary precondition is the loss by the direct producer of ownership of
the means of production.  now confront Frank’s affirmation that the
socio-economic complexes of Latin America have been capitalist since the
Conquest Period . . . with the currently available empirical evidence, we
must conclude that the ‘capitalist’ thesis is indefensible. In regions with
dense indigenous populations— Mexico, Peru, Bolivia, or Guatemala—the
direct producers were not despoiled of their ownership of the means of
production, while extra-economic coercion to maximize various systems of
labour service . . . was progressively intensified. In the plantations of
the West Indies, the economy was based on a mode of production constituted
by slave labour, while in the mining areas there developed disguised forms
of slavery and other types of forced labour which bore not the slightest
resemblance to the formation of a capitalist proletariat."

There in a nutshell it is. Western Europe, at least England from the late
seventeenth century on, had primarily landless, wage-earning laborers. In
Latin America, then and to some extent still now, laborers were not
proletarians, but slaves or "serfs." If proletariat, then capitalism. Of
course. To be sure. But is England, or Mexico, or the West Indies a unit of
analysis? Does each have a separate "mode of production"? Or is the unit
(for the sixteenth—eighteenth centuries) the European world-economy,
including England and Mexico, in which case what was the "mode of
production" of this world-economy?

Louis Proyect
Marxism mailing list:

More information about the Marxism mailing list