Postscript 1985

Robert Keohane's proposal to include my article published in Millennium in the summer of 1981 in this collection of readings is a challenge to

-239-

define my position in relation to the other texts selected. These other texts are all part of a single debate stimulated by recent works by Waltz and Gilpin. My article stems from a different--and very largely idiosyncratic--intellectual process. It does, however, touch upon themes that emerged in this debate, making of me a Monsieur Jourdain, writing prose without having been aware of the fact.

I have deliberately refrained from revising my text and have made only some strictly stylistic and editorial changes to the 1981 version so as to adapt it to the present volume. Once placed before the public, a text is entitled to respect for its own integrity. It has a life of its own, rich or poor. The author too is entitled to assume a certain independence of the text. My own views (as I hope those of most authors) have evolved since 1981. Accordingly, I prefer to try to make the link with the other readings through this postscript.

In the range of their different arguments, I find myself in agreement and in disagreement with aspects of each of the other authors' texts. I am, however, left with the general impression that this is a specifically American debate even though it is couched in terms of international or world systems. Stanley Hoffmann ( 1977) put it that international relations is an American social science. This is not (on my part any more than on Hoffmann's) to suggest that American thought is cast in a single mold. (I protest in advance my innocence of Robert Gilpin's strictures against lumping together authors whose views differ in important respects.) What is common, it seems to me, is (1) the perspective of the United States as the preponderant of the two major powers in the system and consequently the sharing of a certain measure of responsibility for U.S. policy, and (2) the organization of argument around certain obligatory themes of debates, notably those of power versus morality and of science versus tradition. The first of these is, to employ Waltz's language, a systemic conditioning of American thought. The second derives more from an explicitly American cultural process. One aspect of this process was the intellectual conversion of U.S. policymakers to the use of the accumulated physical power of the United States for the performance of a world system-creating and system-maintaining role. Important influences in this conversion were European-formed thinkers like Reinhold Niebuhr and Hans Morgenthau who introduced a more pessimistic and

-240-

power-oriented view of mankind into an American milieu conditioned by eghteenth-century optimism and nineteenth-century belief in progress. Another aspect was the need to legitimate this newfound realism in "scientific" terms. The second aspect can be read as the revenge of eighteenth-century natural-law thinking for the loss of innocence implicit in the first. Richard Ashley has well recounted the socializing process through which successive cohorts of American (and by assimilation Canadian) graduate students have been brought into this stream of thinking.

At this point, following Gilpin's example, an autobiographical reference is in order: The reader should know that this author did not experience the abovementioned process of intellectual formation. His introduction to international political processes came through practice as an "empathetic neutral" ( Cox and Jacobson 1977) in his role of international official in one of the less salient spheres of policy. His only formal academic training was in the study of history. Accordingly, he never shared a sense of responsibility for nor aspired to influence U.S. policy or that of any other country, though he has been well aware that his destiny, like that of the rest of mankind, is profoundly shaped by what he cannot influence. These circumstances have inclined him toward an initial acceptance of the realist position. The political world is at the outset a given world. Men make history, as Marx wrote, but not in conditions of their own choosing. To have any influence over events, or at the very least to forestall the worst eventualities, it is necessary to begin with an understanding of the conditions not chosen by oneself in which action is possible.

The intellectual influences that contributed to the formation of this idiosyncratic view share with realism a common source in Machiavelli. They diverge in having followed a historicist current, through Giambattista Vico to Georges Sorel and, above all, Antonio Gramsci. These thinkers were not concerned primarily with international relations; they addressed the problem of knowledge about society and social tramformations. Historians provided the more specific light on international structures--to saw extent the twentieth-century British Marxist historians, and more particularly Fernand Braudel and the French Annales school. Intellectual points of contact with influences upon other contributors to this volume include E. H. Carr (especially with Gilpin), Friedrich Meinecke, Ludwig Dehio, and Karl Polanyi (especially with Ruggie). So

-241-

much for autobiography: the point is that the itinerary to the Millennium article did not pass through neorealism; it contemplates neorealism from the destination reached.

To change the world, we have to begin with an understanding of the world as it is, which means the structures of reality that surround us. "Understanding" is the key word here. The issues in the confrontation of approaches are linked to different modes of knowledge: positivism and historicism. Since these two terms have been used in contradictory ways in different texts included in this book, I reiterate my own usage here.

By "positivism" I mean the effort to conceive social science on the model of physics (or more particularly, physics as it was known in the eighteenth and nineteenth centuries before it had assimilated the principles of relativity and uncertainty). This involves positing a separation of subject and object. The data of politics are externally perceived events brought about by the interaction of actors in a field. The field itself, being an arrangement of actors, has certain properties of its own which can be called "systemic:" The concept of "cause" is applicable within such a framework of forces. Powerful actors are "causes" of change in the behavior of less powerful ones, and the structure of the system "causes" certain forms of behavior on the part of actors.

I use "historicism" to mean a quite different approach to knowledge about society which was well defined by Giambattista Vico ( 1774/ 1970) and has continued as a distinctive tradition to the present. In this approach, human institutions are made by people--not by the individual gestures of "actors" but by collective responses to a collectively perceived problematic that produce certain practices. Institutions and practices are therefore to be understood through the changing mental processes of their makers. There is, in this perspective, an identity of subject and object. The objective realities that this approach encompasses--the state, social classes, the conflict groups that Robert Gilpin (following Ralf Dahrendorf) refers to and their practices--are constituted by intersubjective ideas. As Gilpin says, none of these realities exist in the same way that individuals exist, but individuals act as though these other realities exist, and by so acting they reproduce them. Social and political institutions are thus seen as collective responses to the physical material context (natural nature) in which human aggregates find themselves. They in turn form part of the social material framework (artificial nature or

-242-

the network of social relations) in which historical action takes place. Historicism thus understood is the same as historical materialism. The method of historical materialism--or, in Robert Keohane's term, its research program--is to find the connections between the mental schema through which people conceive action and the material world which constrains both what people can do and how they can think about doing it.

The two approaches--positivist and historicist--yield quite different versions of the task of science. There can be no dispute about Kenneth Waltz's adherence to the positivist approach and he lays out clearly the tasks of a positivist science: to find laws (which are regularities in human activity stateable in the form of "if A, then B"); and to develop theories which explain why observable laws hold within specific spheres of activity. Laws and theories advance knowledge beyond what would otherwise be "mere description," i.e., the cataloguing of externally observed events. ³²

Insofar as this approach aspires to a general science of society, it cannot discriminate between times and places. All human activity is its province (though this activity is arbitrarily divided among a priori categories of activity of which international relations is one), all of it treated as raw material for the finding of laws and the development of theories. I believe this to be the root of the major defect in Waltz's approach pointed to by his critics (see especially Keohane and Ruggie): the inability of his theory to account for or to explain structural transformation. A general (read: universally applicable) science of society can allow for variations in technologies and in the relative capabilities of actors, but not in either the basic nature of the actors (power-seeking) or in their mode of (power-balancing). The universality of these basic attributes of the social system comes to be perceived as standing outside of and prior to history. History becomes but a mine of data illustrating the permutations and combinations that are possible within an essentially unchanging human story. Despite his wide historical learning, Waltz's work is fundamentally ahistorical. The elegance he achieves in the clarity of his theoretical statement comes at the price of an unconvincing mode of historical understanding.

The historicist approach to social science does not envisage any general or universally valid laws which can be explained by the development of appropriate generally applicable theories. For historicism, both human

-243-

nature and the structures of human interaction change, if only very slowly. History is the process of their changing. One cannot therefore speak of "laws" in any generally valid sense transcending historical eras, nor of structures as outside of or prior to history. ³³ Regularities in human activities may indeed be observed within particular eras, and thus the positivist approach can be fruitful within defined historical limits, though not with the universal pretensions it aspires to. The research program of historicism is to reveal the historical structures characteristic of particular eras within which such regularities prevail. Even more important, this research program is to explain transformations from one structure to another. If elegance is what Robert Keohane writes of as "spare, logically tight" theory (p. 197), then the historicist approach does not lead to elegance. It may, however, lead to better appraisal of historically specific conjunctures. One person's elegance is another's oversimplification.

In choosing between the two approaches, much depends upon one's idea of what theory is for. I have suggested two broad purposes corresponding to the two approaches: a problem-solving purpose, i.e., tacitly assuming the permanency of existing structures, which is served by the positivist approach; and a critical purpose envisaging the possibilities of structural transformation which is served by the historicist approach. The usefulness of all theory, whether problem-solving or critical, is in its applicability to particular situations. But whereas problem-solving theory assimilates particular situations to general rules, providing a kind of programmed method for dealing with them, critical theory seeks out the developmental potential within the particular.

Developmental potential signifies a possible change of structure. It can be grasped by understanding the contradictions and sources of conflict within existing structures; and this task may be aided by an understanding of how structural transformations have come about in the past. ³⁴ Thus the determination of breaking points between successive structures--those points at which transformations take place--becomes a major problem of method. John Ruggie raised this issue in pointing to the structural disjuncture between the medieval and modern world system, and to the inability of Waltz's structural realism to even consider let alone explain this transformation. The case is extremely important, since it contrasts two worlds constituted by quite distinct intersubjec-

-244-

tivities. The entities as well as the modes of relations among them are of different orders.

This case of transformation can be contrasted to the frequent invocations of Thucydides in neorealist literature in support of the contention that a balance-of-power system is the universal condition. What these invocations do establish is that there have been other periods in history where structures analogous to the balance of power of the modern states system have appeared. They do not consider that there have likewise been otherwise-constituted historical structures of which the medieval order of European Christendom was one. The instinct of structural realism may be to reduce the medieval order to its power model; but if so that would be to reject an opportunity for scientific exploration.

Ruggie suspects--and I share his suspicions--that the transformation from the medieval to the modern order cannot be understood solely in terms of a general international-systems theory (indeed, one could point out that the very term "international," derived from modern practice, is inapposite to the medieval world) but probably has also to be explained in terms of changing state structures and changing modes of production. This joins the substantive point of my argument: I have tried to sketch out a research program that would examine the linkage between changes in production, in forms of state and in world orders.

The relevancy of such a research program is strictly practical. It flows from the question whether the present age is one of those historical breaking points between world-order structures, whether the present world situation contains the development potential of a different world order. If this were to be the case, what then would be the range of future structural possibilities? What social and political forces would have to be mobilized in order to bring about one or another of feasible outcomes? The practical use of political theory should be to help answer such questions. That they are present in the minds of the contributors to this volume is clear--for instance in Keohane's primary concern to discover the means of bringing about peaceful change, and Gilpin's with the problems of change under conditions of declining hegemony. Neither of these authors sees clearly how structural realism can be a guide to the answers. My suggestion is that the approach of historical structures would be more apposite.

For Fernand Braudel ( 1958), a historical structure is the longue durée,

-245-

the enduring practices evolved by people for dealing with the recurrent necessities of social and political life and which come by them to be regarded as fixed attributes of human nature and social intercourse. But, particularly with regard to the world system, how long is the longue durée? Ruggie pointed to the breaking point between medieval and modern world orders, but have there been other breaking points since then? What is the proper periodization of world orders? I am inclined to answer that yes, there have been further breaking points, and to suggest a succession of mercantilist, liberal (pax britannica), neoimperialist, and neoliberal (pax americana) orders. At the same time, I would not want to give the impression that this was in some manner the uncovering of an ontological substratum of world history, that these successive world orders were real entities fixed in order of time within some immutable world-historic plan. This periodizing is an intellectual construct pertinent to the present and useful for the purpose of understanding how changes in economic and political practices and in the relations of social groups contribute to the genesis of new world orders. The approach is not reductionist in the sense of making one single factor or set of factors the explanation of all changes. It is grounded in the notion of reciprocal relationships among basic forces shaping social and political practice. ³⁵

Ruggie made another point in suggesting that Waltz's exclusive stress on power capabilities precludes consideration of other significant factors differentiating international systems, in particular the presence or absence of hegemony. Indeed, in neorealist discourse the term "hegemony" is reduced to the single dimension of dominance, i.e., a physical capabilities relationship among states. The Gramscian meaning of hegemony which I have used (see also Cox 1983), and which is important in distinguishing the pax britannica and pax americana from the other world orders of the sequence suggested above, joins an ideological and intersubjective element to the brute power relationship. In a hegemonic order, the dominant power makes certain concessions or compromises to secure the acquiescence of lesser powers to an order that can be expressed in terms of a general interest. It is important, in appraising a hegemonic order, to know both (a) that it functions mainly by consent in accordance with universalist principles, and (b) that it rests upon a certain structure of power and serves to maintain that structure. The consensual element

-246-

distinguishes hegemonic from nonhegemonic world orders. It also tends to mystify the power relations upon which the order ultimately rests.

The hegemonic concept has analytical applicability at the national as well as the international level (indeed, Gramsci developed it for application at the national level). I would differ from Gilpin when he (and Stephen Krasner 1978a, in line with him) suggests that it is possible to distinguish a national interest from the writer of particular interests, if they mean that such a general will exists as some form of objective reality. I can accept their proposition if national interest is understood in a hegemonic sense, i.e., as the way in which the dominant groups in the state have been able--through concessions to the claims of subordinate groups-to evolve a broadly accepted mode of thinking about general or national interests. Unfortunately, Gilpin (and Krasner) end their inquiry with the identification of national interests. When the concept of hegemony is introduced, it becomes necessary to ask what is the form of power that underlies the state and produces this particular understanding of national interest, this particular raison d'état--in Gramscian terms, what is the configuration of the historic bloc?

Finally, there is the troublesome question of the ideological nature of thought--troublesome insofar as the imputation of ideology may appear to be insulting to the positivist who draws a line between his science and another's ideology. I should make it clear that I do not draw such a line; I accept that my own thought is grounded in a particular perspective; and I mean no offense in pointing to what appears to be a similar grounding in other people's thought. Science, for me, is a matter of rigor in the development of concepts and in the appraisal of evidence. There is an inevitable ideological element in science which lies in the choice of subject and the purposes to which analysis is put. The troublesome part comes when some scientific enterprise claims to transcend history and to propound some universally valid form of knowledge. Positivism, by its pretensions to escape from history, runs the greater risk of falling into the trap of unconscious ideology.

There are two opposed concepts of history, each of which is intellectually grounded in the separation of subject and object. One is a methodological separation wherein events are conceived as an infinite series of objectified data. This approach seeks universal laws of behavior. Struc-

-247-

tural realism, as noted, is one of its manifestations. The other sees the subjectivity of historical action as determined by an objectified historical process. It seeks to discover the "laws of motion" of history. Both of these concepts of history lend themselves readily to ideology: the one becoming an ideology reifying the status quo; the other an ideology underpinning revolution by revealing the certainty of a particular future. Both remove the element of uncertainty inherent in the historicist expectation of dialectical development arising out of the contradictions of existing forces--a conception in which, as argued above, subject and object are united.

Neorealism, both in its Waltzian structuralist form and in its gametheoretic interactionist form, appear ideologically to be a science at the service of big-power management of the international system. There is an unmistakably Panglossian quality to a theory published in the late 1970s which concludes that a bipolar system is the best of all possible worlds. The historical moment has left its indelible mark upon this purportedly universalist science.

To the American social science of international relations, Marxism is the great "other," the ideology supportive of the rival superpower. It is also that most readily associated with the alternative mode of separation of subject and object. In the works of this American social science, Marxism is politely recognized but usually reduced to a few simple propositions which do not impinge upon its own discourse. If there is any dialogue between the American science of international relations and Marxism, it is a dialogue de sourds. Gilpin was justified in protesting the richness and diversity of realist thought, but it is at least as justifiable to point to the diversity of Marxist thought. It cuts across all the epistemological distinctions discussed above. There is a structuralist Marxism which, as Richard Ashley has indicated, has analogies to structural realism, not in the use to which theory is put but in its conception of the nature of knowledge. There is a determinist tradition (perhaps less evident at present) which purports to reveal the laws of motion of history. And there is a historicist Marxism that rejects the notion of objective laws of history and focuses upon class struggle as the heuristic model for the understanding of structural change. It is obviously in the last of these Marxist currents that this writer feels most comfortable. Were it not for the contradictory diversity of Marxist thought, he would be glad to

-248-

acknowledge himself (in a parody of Reaganite rhetoric) as your friendly neighborhood Marxist-Leninst subversive. But as things stand in the complex world of Marxism, he prefers to be identified simply as a historical materialist.

NOTES

1.	Among critics of the world systems approach, note especially Skocpol ( 1977 and 1979) and Brenner ( 1977).
2.	I use the term "world order" in preference to "interstate system," as it is relevant to all historical periods (and not only those in which states have been the component entities) and in preference to "world system" as it is more indicative of a structure having only a certain duration in time and avoiding the equilibrium connotations of "system." "World" designates the relevant totality, geographically limited by the range of probable interactions (some past "worlds" being limited to the Mediterranean, to Europe, to China, etc.). "Order" is used in the sense of the way things usually happen (not the absence of turbulence); thus disorder is included in the concept of order. An interstate system is one historical form of world order. The term is used in the plural to indicate that particular patterns of power relationships which have endured in time can be contrasted in terms of their principal characteristics as distinctive world orders.
3.	E. P. Thompson ( 1978:231-242) argues that historical concepts must often "display extreme elasticity and allow for great irregularity."
4.	Kenneth Waltz ( 1980) asked the question "will the future be like the past?," which he answered affirmatively--not only was the same pattern of relationships likely to prevail but it would be for the good of all that this should be so. It should be noted that the future contemplated by Waltz was the next decade or so.
5.	A recent example of this argument is Stephen Krasner ( 1978). The normative intent of the new realism is most apparent as a polemic response to liberal moralism. This was also the case for E. H. Carr ( 1946), who offered a "scientific" mode of thinking about international relations in opposition to the "utopianism" of the supporters of the League of Nations in Britain. Dean Acheson and George Kennan, in laying the foundations of U.S. Cold War policy, acknowledged their debt to Reinhold Niebuhr, whose revival of a pessimistic Augustinian view of human nature challenged the optimistic Lockean view native to American culture. Krasner's chosen target is "Lockean liberalism," which he sees as having undermined the rational defense of U.S. national interests.
6.	See, for instance, R. G. Collingwood ( 1942) distinction between dialectical and eristical reasoning. Collingwood takes dialectic back to its Greek origins and spares us the assertions of theological Marxism concerning "Diamat."
7.	As in Krasner ( 1978b) and Katzenstein ( 1978). The United States is represented by these authors as a state which is weak in relation to the strength of civil society (or more particularly of interests in civil society), whereas other states--

-249-

	e.g., Japan or France--are stronger in relation to their societies. Civil society is thus seen in the U.S. case as limiting the effectiveness of the state.
8.	Gramsci saw ideas, politics, and economics as reciprocally related, convertible into each other and bound together in a blocco storico. "Historical materialism" he wrote, "is in a certain sense a reform and development of Hegelianism. It is philosophy freed from unilateral ideological elements, the full consciousness of the contradictions of philosophy" ( 1975:471, my rough translation).
9.	The notion of a framework for action recalls what Machiavelli ( 1531/ 1970:105-106) called necessità, a sense that the conditions of existence require action to create or sustain a form of social order. Necessità engenders both the possibility of a new order and all the risks inherent in changing the existing order. "Few men ever welcome new laws setting up a new order in the state unless necessity makes it clear to them that there is a need for such laws; and since such a necessity cannot arise without danger, the state may easily be ruined before the new order has been brought to completion."
10.	In this regard, Stanley Hoffmann ( 1977) has written: "Born and raised in America, the discipline of international relations is, so to speak, too close to the fire. It needs triple distance: it should move away from the cotemporary world toward the past; from the perspective of a superpower (and a highly conservative one), toward that of the weak and the revolutionary--away from the impossible quest for stability; from the glide into policy science, back to the steep ascent toward the peaks which the questions raised by traditional political philosophy represent (p. 59).
11.	Taylor ( 1965) points out that expectations with regard to negotiating behavior are culturally differentiated in the present world. Garrett Mattingly ( 1955) studied the origin of the ideas outlined in this paragraph which are implicit in the modern state system.
12.	Collective images are not aggregations of fragmented opinions of individuals such as are compiled through surveys; they are coherent mental types expressive of the world views of specific groups such as may be reconstructed through the work of historians and sociologists--.e.g., Max Weber's reconstructions of forms of religious consciousness.
13.	Gramsci's principal application of the concept of hegemony was to the relations among social classes e.g., in explaining the inability of the Italian industrial bourgeoisie to establish its hegemony after the unification of Italy and in examining the prospects of the Italian industrial workers establishing their class hegemony over peasantry and petty bourgeoisie so as to create a new blocc storico (historic bloc)--a term which in Gramsci's work corresponds roughly to the notion of historic structure in this essay. The term "hegemony" in Gramsci's work is linked to debates in the international Communist movement concerning revolutionary strategy and in this connection its application is specifically to classes. The form of the concept, however, draws upon his reading of Machiavelli and is not restricted to class relations; it has a broader potential applicability. Gramsci's adjustment of Machiavellian ideas to the realities of the world he knew was an exercise in dialectic in the sense defined above. It is an appropriate continuation of his method to perceive the applicability of the concept to world order structures as suggested here. For Gramsci, as for Machiavelli, the general question involved in hegemony is

-250-

	the nature of power, and power is a centaur, part nun, pan beast, a combination of force and consent. See Machiavelli ( 1513/ 1977:149-150) and Gramsci ( 1971:169170).
14.	A recent discussion of the reciprocal character of these relations is in Gourevitch ( 1978).
15.	I have been engaged with Jeffrey Harrod in a study of production relations on a world scale which begins with an examination of distinctive patterns of power relations in the production process as separate historical structures and which then leads to a consideration of different form of state and global political economy. Bringing in these last two levels is necessary to an understanding of the existence of the different patterns of production relations and the hierarchy of relationships among them. One could equally well adopt forms of state or world orders at the point of departure and ultimately be required to bring the other levels in to explain the historical process.
16.	Keohane cites as others who have contributed to this theory Charles Kindleberger, Robert Gilpin, and Stephen Krasner. "Hegemony" is used by Keohane here in the limited sense of dominance by a state. This meaning is to be distinguished from its meaning in this article, which is derived from Gramsci--i.e., hegemony as a structure of dominance, leaving open the question of whether the dominant power is a state or a group of states or some combination of state and private power, which is sustained by broadly based consent through acceptance of an ideology and of institutions consistent with this structure. Thus a hegemonic structure of world order is one in which power takes a primarily consensual form, as distinguished from a nonhegemonic order in which there are manifestly rival powers and no power has been able to establish the legitimacy of its dominance. There can be dominance without hegemony; hegemony is one possible form dominance may take. Institutionalized hegemony, as used in this essay, corresponds to what Keohane calls a "strong international regime." His theory can be restated in our terms as: dominance by a powerful state is most conducive to the development of hegemony. In this present text, the term "hegemony" is reserved for a consensual order and "dominance" refers only to a preponderance of material power. Keohane's discussion of hegemony is developed in his later work ( 1984) but without affecting the distinction made here.
17.	Two classic studies relevant particularly to the interwar period are Karl Polanyi ( 1957b) and E. H. Carr ( 1946). Stephen Blank ( 1978) comments on postwar British economic policy; as does Stephen Krasner ( 1976). Also see R. F. Harrod ( 1951).
18.	The international implications of the New Deal are dealt with in several passages in Arthur M. Schlesinger Jr. ( 1960: vol. 2). Charles Maier ( 1978) discusses the relationship between the New Deal and the postwar ideology of world order. Richard Gardner ( 1956) shows the link between New Deal ideas and the institutions of world economy set up after World War II in the Bretton Woods negotiations.
19.	The basic point I am making here is suggested by a passage from Gramsci ( 1971:176-177; 1975:1562) which reads: "Do international relations precede or follow (logically) fundamental social relations? There can be no doubt but that they follow. Any organic innovation in the social structure, through its technical-military expressions, modifies organically absolute and relative relations in the international

-251-

	field too." Gramsci used the term "organic" to refer to relatively long-term and permanent changes, as opposed to "conjunctural."
20.	E. J. Hobsbawm ( 1977:15) writes: "The men who officially presided over the affairs of the victorious bourgeois order in its moment of triumph were a deeply reactionary country nobleman from Prussia, an imitation emperor in France and a succession of aristocratic landowners in Britain."
21.	Among analysts who concur in this are Karl Polanyi ( 1957b); Gunnar Myrdal ( 1960); and Geoffrey Barraclough ( 1968).
22.	George Lichtheim ( 1971) has proposed a periodization of imperialisms, and I have taken the term "liberal imperalism" from him.
23.	Max Beloff ( 1961) was perhaps the first to point to the mechanisms whereby participation in international organizations altered the internal policymaking practices of states. R. W. Cox and H. K. Jacobson et al. ( 1972) represented the political systems of international organizations as including segments of states. R. O. Keohane and J. S. Nye ( 1974) pointed to the processes whereby coalitions are formed among segments of the apparatuses of different states and the ways in which international institutions facilitate such coalitions. These various works, while they point to the existence of mechanisms for policy coordination among states and for penetration of external influences within states, do not discuss the implications of these mechanisms for the structure of power within states. It is this structural aspect I wish to designate by the term "internationalization of the state." Christian Palloix ( 1975:82) refers to "internationalisation de l'appareil de l'Etat national, de certains lieux de cet appareil d'Etat" by which he designates those segments of national states which serve as policy supports for the internationalization of production. He thus raises the question of structural changes in the state, though he does not enlarge upon the point. Keohane and Nye ( 1977) linked the transgovernmental mechanism to the concept of "interdependence." I find this concept tends to obscure the power relationships involved in structural changes in both state and world order and prefer not to use it for that reason. Gourevitch ( 1978) does retain the concept of interdependence while insisting that it be linked with power struggles among social forces within states.
24.	There is, of course, a whole literature implicit in the argument of this paragraph. Some sketchy references may be useful. Andrew Shonfield ( 1965) illustrated the development of corporative-type structures of the kind I associate with the welfare-nationalist state. The shift from industry-level corporatism to an enterprise-based corporatism led by the big public and private corporations has been noted in some industrial relations works, particularly those concerned with the emergence of a "new working class," e.g., Serge Mallet ( 1963), but the industrial relations literature has generally not linked what I have elsewhere called enterprise corporatism to the broader framework suggested here ( R. W. Cox 1977). Erhard Friedberg ( 1974:94-108) discusses the subordination of the old corporatism to the new. The shift in terminology from planning to industrial policy is related to the internationalizing of state and economy. Industrial policy has become a matter of interest to global economic policymakers (see William Diebold 1980, and John Pinder , Takashi Hosomi and William Diebold, for the Trilateral Commission, 1979). If planning evokes the specter of economic nationalism, industrial policy, as the Trilateral Commission study points out, can be looked upon with favor from a world

-252-

	economy perspective as a necessary aspect of policy harmonization: We have argued that industrial policies are needed to deal with structural problem in the modern economies. Thus, international action should not aim to dismantle these policies. The pressure should, rather, be toward positive and adaptive industrial policies, whether on the part of single countries or groups of countries combined. Far from being protectionist, industrial policy can help them to remove a cause of protectionism, by making the process of adjustment less painful" (p. 50). It may be objected that the argument and references presented here are more valid for Europe than for the United States, and that, indeed, the very concept of corporatism is alien to U.S. ideology. To this it can be replied that since the principal levers of the world economy are in the United States, the U.S. economy adjusts less than those of European countries and peripheral countries, and the insitutionalization of adjustment mechanisms is accordingly less developed. Structural analyses of the U.S. economy have, however, pointed to a distinction between a corporate internationally oriented sector and a medium and small business nationally oriented sector, and to the different segments of the state and different policy orientations associated with each. Cf. John Kenneth Galbraith ( 1974) and James O'Connor ( 1973). Historians point to the elements of corporatism in the New Deal, e.g., Schlesinger ( 1960).
25.	The Zaire case recalls the arrangements imposed by western powers on the Ottoman Empire and Egypt in the late nineteenth century, effectively attaching certain revenues for the service of foreign debt. See Herbert Feis ( 1961:332-342, 384-397).
26.	The evidence for the existence of a transnational managerial class lies in actual forms of organization, the elaboration of ideology, financial supports, and the behavior of individuals. Other structures stand as rival tendencies--e.g., national capital and its interests sustained by a whole other structure of loyalties, agencies, etc. Individuals or firms and state agencies may in some phases of their activity be caught up now in one, now in another, tendency. Thus the membership of the class may be continually shifting though the structure remains. It is sometimes argued that this is merely a case of U.S. capitalists giving themselves a hegemonic aura, an argument that by implications makes of imperialism a purely national phenomenon. There is no doubting the U.S. "origin of the values carried and propagated by this class, but neither is there any doubt that many non-U.S. citizens and agencies also participate in it nor that its world view is glow and distinguishable from the purely national capitalisms which exist alongside it. Through the transnational managerial class American culture, or a certain American business culture, has become globally hegemonic. Of course, should neomercantilist tendencies come to prevail in international economic relations, this transnational class structure would wither.
27.	Sam industries appear as ambiguously astride the two tendencies--e.g., the automobile industry. During a period of economic expansion, the aspect of this industry dominated in the United States, and the United Auto Workers union took the lead in creating world councils for the major international auto firms with a view to inaugurating mutlinational bargaining. As the industry was hit by recession, protectionism came to the fore.
28.	See Cox ( 1978). This tendency can be seen as the continuation of a longterm direction of production organization of which Taylorism was an early stage, in which control over the work process is progressively wrested from workers and

-253-

	separated out from the actual performance of tasks so as to be concentrated with management. See Harry Braverman ( 1974).
29.	Recent news from Brazil indicates restiveness on the part of Sao Paulo workers whose unions have been subjected to a state corporatist structure since the time of President Vargas.
30.	The World Bank promotes rural development and birth control. The concept of "self-reliance," once a slogan of anti-imperialism meaning "decoupling" from the imperial system, has been co-opted by the imperial system to mean self-help among populations becoming marginalized--a do-it-yourself welfare program.
31.	I have borrowed the term from Hartmut Elsenhans (n.d.)
32.	The term "description" as used in positivist discourse (often preceded by "mere") is meaningless in historicist discourse. Description, for the historicist, is inseparable from interpretation or understanding--i.e., the appraisal of a unique fact through the medium of an explanatory hypothesis. The task of theory is to develop such hypotheses and the concepts of limited historical applicability in which they are expressed--i.e., concepts like mercantilism, capitalism, fascism, etc. The difference between "description" (positivist) and "understanding" (historicist) is reflected in the words used to denote the object of study: datum (positivist) versus fact (historicist). The distinction is less self-evident in English than in Latin languages, where the corresponding words are past participles of the verbs "to give" and "to make." Positivism deals with externally perceived givens; historicism with events or institutions that are "made"--i.e., that have to be understood through the subjectivity of the makers as well as in terms of the objective consequences that flow from their existence.
33.	Nor can one speak of "cause" in historicist discourse, except in a most trivial sense. The "cause" of a murder is the contraction of the murderer's finger on a trigger which detonates a charge in a cartridge, sending a bullet into the vital parts of the victim. Explanation is the purpose of historicist inquiry. It is much more complex, requiring an assembling of individual motivations and social structures to be connected by explanatory hypotheses.
34.	This does not imply the presumption that the future will be like the past. But there can be (in the historicist approach) no complete separation between past and and future. The practical utility of knowledge about the past is in the development of explanatory hypotheses about change. Fernand Braudel ( 1958) employed the metaphor of a ship for such hypotheses. The hypothesis sails well in certain waters under a range of conditions; it remains becalmed or it founders in others. The task of theory is to explore the limits of validity of particular hypotheses and to devise new hypotheses to explain those cases in which they fail.
35.	Waltz writes of reductionism and reification in a curious way in saying that systems are reified by political scientists when they reduce them to their interacting parts (p. 61). In my reading of his work, Waltz comes close to the opposite of this position, reifying the international system by treating it not as an intellectual construct but as a "cause," and deriving the behavior of its parts, i.e., states, from the system itself; thus international relations is reduced to the workings of a reified system.

-254-