© Immanuel Wallerstein 1997.
[You are free to download this paper or send it electronically to others. If you wish to translate it into another language, or to publish it in a printed medium or on another web site, you must obtain formal authorization from the author.]
(Paper delivered at 91st Annual Meeting of the American Sociological Association, New York, Aug. 16, 1996)
World-systems analysis as an explicit perspective within social science dates from the 1970's, although of course it reflects a point of view that has a long history and builds on much earlier work. It never put itself forward as a branch of sociology or of social science. It did not think of itself as the "sociology of the world," side by side with urban sociology or the sociology of small groups or political sociology. Rather it presented itself as a critique of many of the premises of existing social science, as a mode of what I have called "unthinking social science."
It is for this reason that I, for one, have always resisted using the term "world-systems theory," frequently used to describe what is being argued, especially by non-practitioners, and have insisted on calling our work "world-systems analysis." It is much too early to theorize in any serious way, and when we get to that point it is social science and not world-systems that we should be theorizing. I regard the work of the past 20 years and of some years to come as the work of clearing the underbrush, so that we may build a more useful framework for social science.
If world-systems analysis took shape in the 1970's, it was because conditions for its emergence were ripe within the world- system. Let us review what they were. The prime factor can be summarized as the world revolution of 1968 both the events themselves and the underlying conditions that gave rise to the events.
Let us remember the shape of U.S. and world social science of the 1950's and 1960's. The biggest change in world social science in the 25 years after 1945 had been the discovery of the contemporary reality of the Third World. This geopolitical discovery had the effect of undermining the nineteenth-century construction of social science which had created separate theories and disciplines for the study of Europe/North America on the one hand and for that of the rest of the world on the other hand. After 1945, social science became, was forced to become, geographically integrated, so to speak. Thus it became legitimate, but only then, for persons called sociologists or historians or political scientists to do research on and in Africa or Asia or Latin America.[1]
[1] See the discussion in I. Wallerstein et al., Open the Social Sciences (Stanford: Stanford Univ. Press, 1996).
This was the era of area studies, and area studies changed the social organization of social science, first in the United States and then in most other parts of the world.[2] In seeking to justify itself intellectually, area studies' advocates faced a fundamental epistemological dilemma. They wished to argue that the theories of social science applied to all areas of the world, and not merely to Europe/North America. Previously the theories of the nomothetic social sciences had been applied de facto only to what was thought of as the modern "civilized" world, and only Europe/North America was considered as belonging to such a world. In this sense, area studies proposed "universalizing universalism." At the same time, however, proponents of area studies wished to argue that this could not be done simply by applying the generalizations previously developed in Europe/North America to the Third World. Conditions in the Third World, said the area studies people, were quite different. After all, if they had not been different, why would we have needed area studies?
[2] See my "The Unintended Consequences of Cold War Area Studies," in The Cold War and the University: Toward an Intellectual History of the Postwar Years (New York: New Press, forthcoming 1997).
Arguing that conditions are the same and arguing simultaneously that they are different is not the easiest thing to do. However, area studies people came up with a clever, and plausible, solution to the apparent dilemma. They based their work on a view that had already been widespread in the social sciences, to wit, that there exist stages through which society goes (and therefore societies go), and that these stages represent evolutionary progress. Applied to the Third World, this theory was baptized "modernization theory," or developmentalism. Modernization theory argued quite simply the following: All societies go through a defined set of stages in a process ending in modernity. The operational definition of a society was a state, presently in existence as either a sovereign member of the interstate system or a colony destined one day to become a sovereign member. The names of these stages varied among the theorists, but the general idea remained the same. The point of the theorizing was to figure out how states moved from stage to stage, to enable us to indicate at what stage given states presently were, and to help all states arrive at modernity.
The epistemological advantages of the theory were great. All states were the same, insofar as they went through identical stages for identical reasons. But all states were also different, in that they presently were at different stages, and the timing of the movements of each from stage to stage was particular. The political advantages of the theory were great as well. The theory enabled all and sundry to engage in applying the theory to the practical situation by advising governments how best to act to speed up the process of moving upward along the stages. The theory also justified a considerably increased allocation of governmental funds (more or less everywhere) to social scientists, especially to those who claimed to be working on "development."
The limitations of the theory were easy to discern as well. Modernization theory purported to be based on the systematic comparison of independent cases, and this presumed a dubious and totally unproven premise, that each state operated autonomously and was substantially unaffected by factors external to its borders. The theory further presumed a general law of social development (the so-called stages), a process furthermore that was presumed to be progressive, both of which arguments were also undemonstrated. And the theory therefore predicted that those states currently at earlier stages of development could, would, and should arrive at an endpoint in which they were essentially clones of whatever was considered by the theorist the model of the most "advanced" state or states.
Politically, the implications were clear. If a state at a so-called lower stage wanted to resemble a state at a so-called advanced stage in terms of prosperity and internal political profile, it had best copy the pattern of the advanced state, and implicitly therefore had best follow the advice of that state. In a world defined by the rhetoric of the cold war, this meant that states were adjured by some to follow the model of the U.S. and by others to follow the model of the U.S.S.R. Non-alignment was disqualified by objective scientific analysis.
Of course, these political implications were the object of ferocious refusal by the revolutionaries of 1968. It was an easy jump for them (and others) to deny the epistemological premises. This created the atmosphere in which there was receptivity for the kind of protest that world-systems analysis represented. It is important to remember this original intention of world-systems analysis, the protest against modernization theory, if we are to understand the directions in which it has moved since. I see four major thrusts to the work we have done collectively. None of these thrusts has been exclusively the work of persons involved in world-systems analysis per se. But in each case, those involved in world-systems analysis have played an important role in pursuing and defining the thrust.
1) The first thrust was globality. It followed from the famous concern with the unit of analysis, said to be a world- system rather than a society/state. To be sure, modernization theory had been international, in that it insisted on comparing systematically all states. But it had never been global, since it posited no emergent characteristics of a world-system, indeed never spoke of a world-system at all. World-systems analysis insisted on seeing all parts of the world-system as parts of a "world," the parts being impossible to understand or analyze separately. The characteristics of any given state at T(2) were said to be not the result of some "primordial" characteristic at T(1), but rather the outcome of processes of the system, the world-system. This is the meaning of Gunder Frank's famous formula, the "development of underdevelopment."
(2) The second thrust was historicity, and it followed from the first. If the processes were systemic, then the history the entire history of the system (as opposed to the history of subunits, taken separately and comparatively) was the crucial element in understanding the present state of the system. To be sure, for this purpose one had to make a decision on the temporal boundaries of the systemic processes, and in practice this has been the subject of contentious debate. Nonetheless, the overall thrust was to push analysis away from exclusively contemporary data, or even from data covering only the nineteenth and twentieth centuries, in the direction of Braudel's longue dur‚e. (3) The third thrust was "unidisciplinarity," and it followed from the second. If there were historically-emergent and historically-evolving processes in the world-system, what would lead us to assume that these processes could be separated into distinguishable and segregated streams with particular (even opposed) logics? The burden of proof was surely on those who argued the distinctiveness of the economic, political, and sociocultural arenas. World-systems analysis preferred to insist on seeing "totalities."
(4) The fourth thrust was therefore "holism." This thrust was historico-epistemological, and it followed from all the previous ones. The arguments of world-systems analysis led its advocates to be dubious of, even opposed to, the boundary lines within the social sciences, as they had been historically constructed in the period 1850-1945. These boundaries did not seem to hold water, and thus there was talk of restructuring knowledge. Indeed, holism leads to rethinking as well the historically-constructed and now consecrated great divide between the sciences and the humanities, and perhaps unthinking it as well.
It is important to distinguish these four thrusts from currents that used seemingly similar terminology but were in no sense intended as protests against the dominant modes of social science.
Globalism was not "globalization." As used by most persons in the last ten years, "globalization" refers to some assertedly new, chronologically recent, process in which states are said to be no longer primary units of decision-making, but are now, only now, finding themselves located in a structure in which something called the "world market," a somewhat mystical and surely reified entity, dictates the rules.
Historicity was not "social science history." As used by most persons in the last 25 years, "social science history" refers to the need for persons dealing with past data (so-called historians) to use that data to test social science generalizations derived from the analysis of contemporary data. Social science history is in many ways anti-historical process, and relegates empirical work (especially about the past) to the position of hierarchical subordination to so-called theoretical work. Social science history is compatible with globalization but not with globality.
Unidisciplinarity was not "multidisciplinarity." Multidisciplinarity accepted the legitimacy of the boundaries of the social sciences, but asked the various practitioners to read and use each other's findings, in an additive fashion. It was the belief that more cooks often improve the broth. It resisted the study of totalities on the grounds that it is hard to specify the data in ways amenable to testable propositions, and therefore encouraged vague and non-diprovable argumentation.
And finally, holism was not a rehash of "general education." General education had accepted the basic premises of the modern divisioning of knowledge into three superdomains: the natural sciences, the humanities, and (in-between the so-called "two cultures") the social sciences. General education was the case for making all scholars (and indeed all educated persons) sensitive to the premises underlying each of the separate domains. Holism asks whether the superdomains are in fact different kinds of knowledge, or ought to be thought of in this way. This debate is directly relevant to the crucial question of the relation of the quest for the true and the quest for the good.
If I have emphasized not only what the thrusts of world- systems analysis have been but also what they have not been, it is because we are running the danger of success. It is because of the strength, and not the weakness, of our efforts that our terminology is in the process of being appropriated for other, indeed opposite, purposes. This can cause serious confusion in the general scholarly public, and even worse, may lead to confusion on our own part, thus undermining our ability to pursue the tasks we have set ourselves.
I have in my title used the phrase, "rise and future demise of world-systems analysis." So far, I have talked only about the rise. Wherein do I see a demise? The demise of a movement, and world-systems analysis has been essentially a movement within contemporary social science, derives from its contradictions and from the eventual exhaustion of its utility. We are not there yet, but we are clearly moving in the direction of such a demise, or if you will permit my prejudices, a bifurcation. What are the contradictions of world-systems analysis?
1) The first is that world-systems analysis is precisely not a theory or a mode of theorizing, but a perspective and a critique of other perspectives. It is a very powerful critique, and I personally believe the critique is devastating for a large number of the premises on which much of social science presently operates. Critiques are destructive; they intend to be. They tear down, but they do not by themselves build up. I called this earlier the process of clearing the underbrush. Once one has cleared the underbrush, however, one only has a clearing; not a new construction but only the possibility of building one.
Old theories never die, but they usually don't just fade away either. They first hide, then mutate. Thus, the work of critique of the old theories may seem never-ending. The risk is that we shall become so enamored of this task that we may lose ourselves in it and refuse the necessary risk of moving on ourselves. To the extent that we shall fail to do this, we shall become redundant and irrelevant. At which point the mutants come back, stronger than ever. The attempt in the 1990's to relegitimize modernization theory is an instance of this, albeit thus far one that has been rather weak. If I might continue the medical metaphor, the problem today of world-systems analysis is analogous to the problem of overused antibiotics. The solution is to move forward from medical therapy to preventive medicine.
2) There is a second problem with critiques, especially critiques that are past the moment of initial shock and vigor. Critiques are not that difficult to pseudo-coopt. I have tried already to indicate the ways in which our terminology, or something close to it, is being used for purposes other than we had in mind, which then can have the effect of corrupting what we ourselves do. So then this becomes a question of "physician, heal thyself." But I am making more than a general admonition always to be self-critical. I am suggesting that there is a tendency to forget our own original critical stance, as we hail those who seem to be emulating us, and that this tendency poses considerable risks both to the critical task and to the putative task of reconstruction.[3] At the end of the road, we risk finding ourselves in the situation of so many intellectual movements, a name that has become a shell.
[3] I have argued the nature of such risks in my article "Hold the Tiller Firm: On Method and the Unit of Analysis," in S.K. Sanderson, ed., Civilizations and World Systems: Studying World-Historical Change (Walnut Creek, CA: Altamira, 1995), 239-247.
(3) We have shifted over the years from criticizing the ways in which we analyze the contemporary situation in peripheral zones of the world-economy to criticizing the ways in which the history of the modern world has been written, to criticizing the theories that are supposed to explain the modern world-system, to criticizing the methodologies used in the historical social sciences, to criticizing the ways in which knowledge institutions have been constructed. We have been following the paths of our critiques and of answering those who have in turn been critical of our work. It is as though we have been going through doors to find other doors behind them, in a constant regress. Perhaps the problem is deeper than we have imagined.
Perhaps the problem is the entire thought-system of the capitalist world-economy. This has been suggested, to be sure, by the so-called post-modernists. I am sympathetic to many of their critiques (most of which, however, we have been saying more clearly, and indeed earlier). However, I find them on the whole neither sufficiently "post"-modern nor sufficiently reconstructive. They will certainly not do our job for us.
To be a movement within social science had, and has, certain distinct advantages. It enables us to group forces, to clarify our critiques, and to sustain each other in a sometimes hostile environment. On the whole, I give us good marks for how we have conducted ourselves. On the one hand, we have allowed multiple views to co-exist, and thus avoided becoming a sect. On the other hand, we have not defined our program so loosely that it has lacked critical teeth, which is what would have happened if we had followed the recurrent suggestions that we rename ourselves (and therefore blend into) "the sociology of development," or "political economy," or "global sociology."
Nonetheless, being a movement has certain distinct disadvantages. I am often appalled by the two-line summaries of our perspective one can find in the books of others who have manifestly read virtually nothing of what we have written. I am equally appalled by the suavity with which our research findings are appropriated (and misappropriated) not only without credit but even more important without any integration of the underlying approach that gave rise to the research findings. This is in part inevitable, since movements tend to talk to themselves, and after a while this constrains radically their impact.
There is of course an alternative road we might follow that might overcome the limitations of being an intellectual movement. That road is that of moving into the very center of social science, not as a movement but as consensual premise. How might we do that?
The facetious answer would be that we should be writing, or some of us should be writing, general textbooks for first-year students of social science. The real answer is that persons involved in world-systems analysis should be addressing, and addressing urgently, some very fundamental questions, questions that in my view can only be satisfactorily addressed if one has unthought nineteenth-century social science and structures of knowledge and thoroughly absorbed the lessons of world-systems analysis.
Allow me to list some of these fundamental questions:
1) What is the nature of the distinctive arena of knowledge we may call social science, if there is one? How do we define its parameters and social role? In particular, in what ways, if any, is such a field to be distinguished from the humanities on the one side and the natural sciences on the other?
2) What is the relation, theoretically, between social science and social movements? between social science and power structures?
3) Are there multiple kinds of social systems (I would prefer the concept, historical systems), and, if so, what are the defining features that distinguish them?
4) Do such historical systems have a natural history or not? If so, can this history be called an evolutionary history?
5) How is TimeSpace socially constructed, and what differences does this make for the conceptualizations underlying social science activity?
6) What are the processes of transition from one historical system to another? What kinds of metaphors are plausible: self- organization, creativity, order out of chaos?
7) What is the theoretical relation between the quest for truth and the quest for a just society?
8) How can we conceive our existing historical system (world-system)? And what can we say about its rise, its structure, and its future demise, in the light of our answers to the other questions?
As you can see, the last is the question with which we started. A number of the other questions have been worrying various persons who consider themselves part of the network of scholars involved in world-systems analysis. Furthermore, of course, many other scholars, present and past, have worried about these questions, or at least some of them. The point however is to see that these questions are interrelated, and can really only be answered in relation to each other, that is, from a world- systems perspective.
The other point is that world-systems analysts are, on the whole, better trained than most social scientists today to address these questions as an interrelated set. When we do begin to address them in this way, we shall no longer be acting primarily as a movement within social science, but we shall be laying claim to formulating the central questions of the enterprise. Is this hybris? Not really. As world-systems analysts, we know that intellectual activities are not simply a matter of intelligence or will but of social timing, in terms of the world-system. It is because the historical system in which we live is in terminal crisis that there exists the chance of addressing these questions in ways that can make possible substantively rational social constructions. This was not a possibility available to nineteenth-century scholars, however insightful or masterly they were. It is because the legitimacy of the hierarchies that are fundamental to the capitalist world- economy hierarchies of class, of race, of gender are being fundamentally challenged, both politically and intellectually that it may be possible to construct, for the first time, a more inclusive and relatively more objective social science.
It is the times that make it possible, again for the first time, to stand on the shoulders of those nineteenth-century giants and see something beyond, provided we have the energy and the will. It is the times that permit us, without disgracing ourselves, to follow Danton's exhortation: "De l'audace, encore de l'audace, et toujours de l'audace." These are our times, and it is the moment when social scientists will demonstrate whether or not they will be capable of constructing a social science that will speak to the worldwide social transformation through which we shall be living.
(Go to top of list of papers)
(Go to Fernand Braudel Center Home Page)