(Re)Constructing Constructivist International RelationsResearch

Center for International StudiesUniversity of Southern California

October 2001

2On October 6, 2001 the Center for International Studies at the University of Southern

California held a workshop entitled (Re)Constructing Constructivist International Relations

Research. This workshop joins a series held by CIS on cutting edge research in international

relations. Participants were sent a statement composed by Hayward Alker and a copy of Nicholas

Onufs paper and asked to bring written comments based on these two documents.

I am grateful to Hayward Alker and Nicholas Onuf for their important role in

conceptualizing and providing the original stimulus. Thanks also to Melissa Ince (Program

Coordinator) who worked tirelessly in all aspects of organizing, including compiling this

working paper.

Ann TicknerDirectorCenter for International StudiesUniversity of Southern California

33

(Re)Constructing Constructivist International Relations ResearchCenter for International Studies, USC

Saturday October 6, 2001

Panel One: Hayward Alker, School of International Relations, USC (Moderator)Legal-Historical Rule-Based Approaches to Constructivist IR ResearchNicholas Onuf, Florida International UniversityWayne Sandholtz, UC Irvine

Panel Two: Robert English, School of International Relations, USC (Moderator) Social Historical Constructivism

Cecelia Lynch, UC IrvineDaniel Lynch, USC

Panel Three: J. Ann Tickner, Director Center for International Studies & Professor School ofInternational Relations, USC (Moderator)Social Historical Constructivism IIRaymond Duvall, University of MinnesotaColin Wight, University of Wales, Aberystwyth

Panel Four: Saori Katada, School of International Relations, USC(Moderator)Computationally Oriented ApproachesSteven Majeski, University of WashingtonHayward Alker, USC

Workshop Participants:Hayward Alker is the John A. McCone Professor of International Relations with the School of InternationalRelations at the University of Southern California. He has just completed a project on the development ofinformation resources for anticipating, preventing, managing violent inter-group conflict and interstate conflictsaround the world. He is now working on a project reviewing major debates about world order in the 20th century. Heis the author of Rediscoveries and Reformulations: Humanistic Methodologies for International Studies CambridgeUniversity Press and an edited volume entitled Journeys Through Conflict: Narratives and Lessons Rowman andLittlefield (2001) as well as numerous journal articles and book chapters.

Raymond Duvall is a Morse-Alumni Distinguished Teaching Professor of Political Science, the Associate Directorof the Interdisciplinary Center for the Study of Global Change of which the Mac Arthur Inter-Disciplinary Programon Peace and International Cooperation at the University of Minnesota is a part. He was a Visiting Professor at theGraduate Institute of International Studies in Geneva, Switzerland. He is the co-editor of Cultures of Insecurity:States, Communities, and the Production of Danger University of Minnesota Press (1999). He has publishednumerous articles in journals such as American Political Science Review, International Studies Quarterly andJournal of Conflict Resolution.

Cecelia Lynch is an Associate Professor of Political Science at the University of California, Irvine. Her experienceis in international relations theory, social movements, political philosophy, ethics, and international organization.She is the author of Beyond Appeasement: Interpreting Interwar Peace Movements in World Politics CornellUniversity Press (1999), and co-editor, with Michael Loriaux, of Law and Moral Action in World Politics Universityof Minnesota Press (2000). She is currently working on two books: one, co-authored with Audie Klotz, onconstructivist methods, and the other, for which she was awarded an SSRC-MacArthur Foundation Fellowship, onreligion in world politics.

4Daniel Lynch is an Assistant Professor with the School of International Relations at the University of SouthernCalifornia. Lynchs most recent work focuses on the international origins of democratization. He is contrasting theexperiences of Taiwan and Thailand with those of China and Burma. He is also researching Chinese concepts ofcomprehensive security and how they relate to identity formation. Publications include After the Propaganda State:Media, Politics, and Thought Work in Reformed China Stanford University Press (1999).

Stephen Majeski is Chair and Professor of Political Science at the University of Washington and has researchinterests in international conflict and foreign policy making. He has published numerous articles concerning theseareas in journals such as American Journal of Political Science, Journal of Conflict Resolution, International StudiesQuarterly, and Conflict Management and Peace Science. He is currently finishing a project on how high level U.S.officials make policy with respect to Vietnam from 1961-65. He is also doing experimental work assessing howgroups make choices in situations of conflict and cooperation and using evolutionary theory to simulate how agentsinteract in complex environments.

Nicholas Onuf is a Professor of International Relations at Florida International University, where he and VendulkaKubalkova of the University of Miami organized the Miami International Relations Group focusing on constructivistscholarship. Dr. Onuf is widely credited for introducing the term constructivism to the field in 1989 and he has acontinuing interest in the many conceptual issues that this way of thinking has brought to the fore. In recent years hehas also studied the conceptual underpinnings of modernity. While on Sabbatical at the Center for InternationalStudies at USC (2001-02), is working on a book with his brother Peter Onuf, and another historian, linking the crisisof the U.S. federal union in the early 19th century to the rise of the liberal world. He also wants to assemble materialshe has written in the last five years into a book on the conceptual foundations of constructivist social theory.

Wayne Sandholtz is an Associate Professor of Political Science at the University of California, Irvine. His currentresearch includes projects on the evolution of international norms, the comparative study of corruption, and thedevelopment of supranational governance in the European Union. His past work has focused on the politics ofEuropean integration, including work on integration theory, high-technology cooperation, telecommunications, andmonetary union. He has had articles published in International Studies Quarterly, International Organization, andWorld Politics, and co-edited European Integration and Supranational Governance Oxford University Press (1998).

Colin Wight is Lecturer in International Politics at the University of Wales, Aberystwyth. He gained a first classhonours degree from the University of Wales and completed his PhD on the Agent-Structure Problem inInternational Relations Theory in 1998. Before going to Aberystwyth he worked as a journalist. Among Dr.Wights research interests are International Relations theory, social theory, the philosophy of science and socialscience and political philosophy. His publications include Political Thought and German Reunification: The NewGerman Ideology? Macmillan, (1999), a co-edited piece, They Shoot Dead Horses Dont They? Locating Agencyin the Agent-Structure Problematique in the European Journal of International Relations, (1999) andIncommensurability and Cross Paradigm Communication in International Relations Theory: Whats the FrequencyKenneth? Millennium, Journal of International Studies (1996).

Workshop Moderators

Robert English is an Assistant Professor with the School of International Relations at the University of SouthernCalifornia. Professor English's courses cover Russia, the former USSR, and Eastern Europe, with a focus that rangesfrom general issues of regional relations to specific questions of ethnicity, identity, and nationalism. He is presentlyworking on a book-length study entitled Our Serbian Brethren: History, Myth, and the Politics of Russian NationalIdentity.

Saori Katada is an Assistant Professor with the School of International Relations at the University of SouthernCalifornia. She teaches courses on Japanese foreign policy, international political economy, development andPacific Rim issues. Publications include Banking on Stability: Japan and the Cross-Pacific Dynamics of theInternational Financial Crisis Management University of Michigan Press (2001).

55

J. Ann Tickner is a Professor with the School of International Relations at the University of Southern California,and the Director of the Center for International Studies. Professor Tickners major current research interest is infeminist perspectives on international relations theory with a particular focus on ways of reconceptualizing security.Her latest book Gendering World Politics: Issues and Approaches in the Post-Cold War Era was published byColumbia University Press (2001).

Contacting Participants:From the School of International Relations, USCHayward Alker: alker@usc.edu Daniel Lynch: dlynch@usc.eduUniversity of Southern CaliforniaSchool of International Relations3518 Trousdale ParkwayVKC 330, University Park CampusLos Angeles, CA 90089-0043

From the Department of Political Science UC, IrvineCecelia Lynch: clynch@uci.eduWayne Sandholtz: wsandhol@uci.eduUniversity of California, IrvineDepartment of Political Science3151 Social Science PlazaIrvine, CA 92697-5100

Raymond Duvall: RDUVALL@polisci.umn.eduUniversity of MinnesotaDepartment of Political Science1414 Social Sciences Building267 19th Avenue SouthMinneapolis, Minn. 55455

Stephen Majeski: majeski@u.washington.eduUniversity of WashingtonDepartment of Political ScienceBox 353530Seattle, WA 98195

Colin Wight: Colin.Wight@aber.ac.ukUniversity of Wales, AberystwythDepartment of International PoliticsCeredigion, Wales UK SY23 3DA

6The Strange Career of Constructivism in International Relations

NICHOLAS ONUF

___________________________________________________________________________

This paper is forthcoming as Worlds of Our Making: The Strange Career of Constructivism in InternationalRelations, in Donald J. Puchala, ed., Visions of International Relations (Columbia: University of South CarolinaPress, January 2002).

I have shortened the text and made a few changes for presentation at the Workshop on (Re)ConstructingConstructivist IR Research, Center of International Studies, University of Southern California, October 6, 2001___________________________________________________________________________

The Many Worlds of International Relations

The world that we live in is a place, both physical and social. We are physical beings capable of living in,and acting on, the world only as social beings. Agency is a social condition. The world that we make for ourselvesconsists of social relations that make sense, and use, of our physical circumstances.

For any of us as agents, the world is, as I said, the whole of our experience. As Aristotle might have said, itis sufficient in itself. Were it not sufficient for our needs, we would make it so or we would cease to be. As agents,we may participate in making our world what it is, and have it make sense to us, without recognizing ourselves asagents, somehow apart from that world and acting on it.

Conversely, we are quite capable of detaching ourselves from our world and observing it, our places in itand the consequences of our acts upon it. Once we stand back and become observers, the world ceases to be theseamless place that we take for granted. We see social relations to which we have little or no connection. We makethem into worlds in which we do not see ourselves as agents.

While each of us lives in a whole world that is uniquely our own, we share worlds by speaking about themto each other. Worlds that we speak of can only be partial and highly selective representations of what we see. Wemake these representations more or less compatible by telling each other not just what we (want them to) see butwhat we want them to do, and why they should do it. Speaking about worlds is always normative. By speaking, wenarrow down the number of worlds that we are collectively able to identify.

To the extent that some number of observers commonly represent some set of social relations as a world(whether they belong to it or not), then they have made a world for themselves collectively, but not necessarily forthe agents whose world (they say) it is. Moreover, they have made it normative (whether consciously or not). Theyhold themselves to this representation of the world, and they would have other observers adopt it. Indeed theywould have those agents whose relations (they say) make up this world stand back as observers, accept the world asrepresented to them and act accordingly.

By becoming agents in any world, observers make boundaries even harder to discern. Such is the case witha world for which the familiar name is international relations. The world of international relations has manyobservers. They include ordinary people who are quite sure that they have little effect on whatever they see. Theyalso include a relatively small number of us who make the study of International Relations a vocation. We knowfull well that what we see, and say about, the world of international relations has an occasionally significant andcumulatively substantial effect on it.

For ordinary people, what they observe are eventswars, world leaders holding summits, the rise and fallof stock markets, extradition hearings. Behind these events are social relations largely taken for granted. Whilesome few of these relations are highly conventionalized and always in view, others are shrouded in secrecy. Most

77

are invisible to ordinary observers because they are so far removed from daily experience. This world has nodiscernible boundaries.

Even for scholarly observers, the world of international relations does not have obvious boundaries. If welook deep into its historywhich is to say, deep into the history of the Western worldwe see a remarkably smallworld of set practices. The main agents in this world act on behalf of political societies that we only gradually cameto call states. Centuries ago these agents were rulers whose realms were indistinguishable from their selves andwhose relations were simultaneously highly stylized and intensely personal. As political societies dissociated fromroyal personae and agents came to represent states as abstractions, relations among agentsheads of state, theirministers, diplomatsremained much as they had always been.

In this still small world, agents are preoccupied with the many signs of standing, both personally and withrespect to their states. They attach great significance to ceremony. Most of their ceremonies involve thereciprocities of respect and the acknowledgment of relative position. Agents remember slights and they are quick totake offense. They are always ready to commit societal resources on a massive scale for vindication, and they areprone to believe in organized violence, not just as a last resort, but as an edifying spectacle. Rules matter, for theseagents live by an unwritten code of honor, some version of which is to be found in every small world that is left toitself.

The world that I just sketchedfor lack of a better name, the world of standing and statecraftsurvivesmore or less intact despite great changes in the world of events because of its relation to yet another worldtheworld of states. Paradoxically, this world is a large world spatially but much smaller socially. Membership in it isrestricted to states, which are large worlds in their own right but few in number. On the one hand, states are theproduct of long histories of arduous social construction. On the other hand, they exist only in formal relation to eachother, and the ways in which they conduct their relations are also formally limited.

Consider how the defining properties of statehood impose formal limits. To be states, other states mustrecognize them as such. Sovereign in principle, their rights and duties make them equal and apart. Furthermore,states are subject to general, formal rules, conventionally known as international law. These rules classify relationsof states in categories as confining as they are familiar (starting with peace and war), and they give rise to a largenumber of voluntarily incurred obligations (treaties) and institutions (international organizations), all in the samegeneral form.

States formally deal with each other through agents for whom the conditions of agency are also formallydetermined. So too are the venues for the conduct of their relations. Because the formalities of statehood soseverely limit the number of state agents and the exercise of agency, the grandly formal world of states supports asmall world of state agents whose relations are at once stylized and intensemuch as they were before states tookform. Neither world could continue to exist without the other.

State agents take the formalities of their relations exceedingly seriously. They are always careful to justifytheir conduct by claiming to act on behalf, and in the interest, of their states. By doing so, they make thepreoccupations of their small world weighty and impersonal. They have access to resources not otherwise availablein any world. Their importance is unequaled, in their own eyes and in the eyes of many others. Every deedconfirms them as indispensable, and the world of states as overwhelmingly, immediately realperilous perhaps butirreplaceable.

The world of states has a remarkable capacity, through its agents and through its effect on observers, toreproduce itself in a form that has changed very little over the last two centuries. Form is the key, once we grant thisworld its formality and thus its limits. Bounded out are all sorts of social relations that have direct and lastingeffects on a global or near-global scale. Conversely, a world that would encompass most such relations would haveto relegate states to the background. A variety of other institutions would come to the fore, and the number ofagents whose world it is would increase dramatically.

8For centuries, institutionalized practices of the Western world have had global effectsand not justincidental effects, as many scholars find it convenient to think. Nevertheless, designing global institutions andinsulating them from the pattern of relations characteristic of the mutually constituted worlds of states and stateagents only began in earnest with the rise of industrial capitalism. With industrial capitalism, there also arose aninterest in issues of technical facilitation and responsiveness to human needs. In what amounted to a designprinciple, state agents approached these issues one at a time, each time creating a specialized organization tocomplement rapid bureaucratic growth and differentiation within their states.

In these developments, observers saw a world in which states and their claims on agents mattered far lessthan the efficient performance of technical tasks. These observers supposed that technical successes would causethe new world to grow as state agents surrendered ever more responsibility to specialized organizations. The worldof states would shrink, and the small world of statecraft and standing would become increasingly irrelevant.Eventually, a single, benign and encompassing world of technical services would supercede all those worlds thatseemed to impede human progress.

While the passing years proved this functionalist prophecy to be wildly mistaken, the growth of technicalorganizations with global missions proceeded unabated. So too did technical task expansion within states. Thelong-term result is a vast, intricately organized world of services that depends on the formal relations of the world ofstates even as formers many successes indirectly support the latter. Meanwhile, the prodigious scale andcomplexity of the world of services defy most observers efforts to find its boundaries and put the whole of it inclear view. Indeed, much of the current rhetoric of globalization suggests a still inchoate sense of what that worldhas finally become (also see Onuf, 1998a:263-73; Onuf, 2001).

Observers of international relations who see beyond the world of events impose boundaries and emphasizesome sorts of social relations over others in order to make sense of what they see. By consensus, the world ofinternational relations is a world of worldsby definition, a world of states. Considerations of physical and socialscale yield three plausible versions of this world. As opposed to the world of events, which matters to a great manypeople in a variety of ways, the world of standing and statecraft is a world whose few agents matter a great deal toeach other. The world of states also matters to the agents of states because it perpetuates their own lived-in world.The world of services would matter to a great many people if they knew very much about it.

As an observer, I could no doubt propose any number of possible worlds of international relations byintroducing additional considerations to the limit of my imaginative and linguistic resources. Nor will any otherobserver see the four worlds enumerated here quite as I do. As observers, we never stand in the same place, and wenever see matters from the same angle. As agents, we can only experience our world from the inside.

Lived-in worlds have an immediacy and normative density that no observer can fully appreciate.Observers see that world from a distance and render it in drastically simplified terms, but they also live in otherworlds. For those who make observing international relations a vocation, one of these worlds consists of observerslike themselves. That lived-in world (which is my world) constitutes yet another world of international relations. Itsagents collectively make the world of states into an important and distinctive subject of study commonly calledInternational Relations. In doing so, they respond to normative demands that are too close, too immediate, for themto recognize. The world of scholarship exerts a normative pull over the world of statespulling it away from boththe world of standing and statecraft and the world of services, all the while pushing and pulling with the world ofevents.

Unsettled Times

The worlds that we make for ourselves as observers look the way that they do because of the ways that wego about the business of seeing. Gaze fixed on some world or another, scholars figure out ways of seeing it better.We could just as well sayand often do saythat we draw maps and blueprints, devise tools and procedures, buildmodels and frameworks, all of which make that world what we see it to be. Different worlds call for, and dependon, different ways of world-making.

99

Scholars have always raised questions about accepted ways of seeing. Beginning in the 1970s, theirnumbers increased, in the 1980s dramatically. Some critics came to the radical conclusion that we know not whatwe see, and delude ourselves into thinking that we do. A few others began to see worlds as never-endingconstruction projects involving even themselves as agents, and realized that they needed new and differenttoolstools for making worlds and not just for seeing them.

Let me put these developments in more conventional, less obviously metaphorical terms. The 1980s wereunsettled times in the social sciences. The positivist quest for reliable, cumulative knowledge about the world cameunder assault, along with the positivist assumption that, deep down, nature and society have the same nature.Critics held positivist science, whether applied to nature or society, to be a central feature of what they called theEnlightenment project or, indeed, modernity. Looking back, we can see that this assault did not come out of theblue. Ennui had beset positivist social science, and so had a measure of smugness.

Critics decried positivist science as an emblem of modernity, understood as the rationalizing, aggrandizingpath that the Western world has taken over a period of several centuries and especially since the Enlightenment.Their critique did not proceed over matters of methodover the ways that we might acquire reliable,commensurable knowledge of the world. Positivists were always prepared to discuss such matters and likely toprevail on their own termsterms that emphasize the measurable properties of things (positivities) and theirrelations. Instead the assault began over prior matters: how are we capable of knowing anything at all about theworld or able to convey what we think we know to anyone else? Suddenly the term epistemology displaced the termmethodology as the signal that controversial matters were under discussion. So did the neologism post-modern, andthe term modern itself took on an unaccustomed resonance.

Epistemological radicals doubted that we can know anything for sure. Language deceives us. No matterwhat how much, or well, we think we are communicating with each other, no language, natural or formal, is capableof representing the world as it is. Instead, what we think we know is the product of provisional agreement onlyrelatively fixed in time and place, and these radical critics mocked the vain Enlightenment quest for firmfoundations or an objective vantage point. Obviously I mean vain in both senses of the word.

Positivists saw this assault as willfully destructive, even nihilistic, and not at all the liberating, playful,celebratory occasion that its advocates took it to be. Not only were these unsettled times. No one held out any hopefor settlement, if only because post-modern critics interpreted any effort to settle matters as a symptom of theEnlightenments rationalist pathology. The world of scholarship, long ordered into worlds (disciplines, fields)ostensibly corresponding to the diverse worlds of ordinary experience, found itself, and quite a number of itsconstituent worlds, torn in two.

As a field of study, International Relations was hardly exempt from these unsettling developments. HereRichard Ashley warrants particular mention (and note that Ashley had been a student of Hayward Alkers). In aseries of highly visible essays spanning the 1980s, each more slashing than the one before, Ashley left his audiencegaspingwhether in dismay or delight (Ashley, 1981; 1984; 1987; 1988; 1989). Others followed suit. As thedecade turned, dissidents joined in celebration (Ashley and Walker, 1990).

Retrospectively, these developments seem far less surprising than they did at the time. Already in the1970s, positivists showed signs of fatigue brought on by the demands of normal science. Fatigue made it all themore difficult to cope with dwindling resources as the Vietnam War came to an end, and fatigue did nothing to dullthe disappointments due to the meager results that normal science had posted. By the end of the 1970s, realists hadbegun to reassert themselves, International Political Economy consolidated its position at the fields center, theoryreturned to prominence, formal analytic skills earned instant prestige. It soon became clear that publication ofKenneth Waltzs Theory of International Politics (1979) was a defining moment for the field.

By the end of the next decade, International Relations was deeply dividedmore deeply than Kal Holstihad imagined when he called it the dividing discipline (1985). One response was to domesticate these disturbingdevelopments. This was Robert Keohanes clear intention when he edited Neorealism and Its Critics (1986). No

10

less was it Yosef Lapids when he sought to bring these developments into focus as an intelligible debate (Lapid,1989:238)the third great debate to punctuate the fields brief history.

A second response was my own. I announced it in World of Our Making (1989), and I named itconstructivism. (I had first used this term in Onuf, 1987an article few scholars in my world had occasion to see.)The term betrays its origins in the 1980s and the influence of social theory at that time. Anthony Giddens (1984)and Jrgen Habermas (1984) were particular sources of inspiration. By building a bridge across the divide (Onuf,1989:55-61), I aimed to provide for myself and other scholars a way between positivist social science and assaultson modernity then rampant (pp. 36-43).

This third way holds that ontology is the key to escaping the impasse between positivist complacency overepistemological matters and the wholesale post-modern dismissal of methodical pursuits. Ontological discussionsrefer to the worldany worldas if we could take its existence for granted, but not its properties. Constructivism,as I presented it in 1989, grants ontological parity to things and their relations. Conceptually speaking, neitheragents as members of society, nor society as the totality of agents and their relations, come first.

On ontological grounds, constructivism challenges the positivist view that language serves only to representthe world as it is. Language also serves a constitutive function. By speaking, we make the world what it is.Nevertheless, in making the world, we do not just make it up. Constructivism takes the linguistic turn but only sofar (Onuf, 1989:78-94; Shapiro, 1981, was particularly helpful as my guide into the turn).

Nevertheless, constructivism as presented is not post-modern because it does not challenge theEnlightenment belief in the possibility of meaningful knowledge about the world we live in. Constructivism treatssuch knowledge as if it were independent of the language that we use to represent the world, but only provisionallyso. As for method, constructivism is eclectic. Ontological openness warrants methodological diversity. Theconstructivist is a bricoleurone who makes what is needed out of available materials (Lvi-Strauss, 1966:16-22inspired my use of this term). All such materials are necessarily social, the result of continuous bricolage: world-making is bricolage.

Three premises structure my systematic rendition of constructivism in World of Our Making.

a) Society is what it does. Any coherent set of social relations (including international relations) is also,and always, a process in which agents and their worlds constitute each other. Co-constitution accounts for pervasivechange and the appearance of constancy in social relations.

b) Speech and its derivatives (rules, policies) are the media of social construction. People become agentsby living in a world of language. They depend on language to express their wishes, to translate their wishes intogoals, and finally to act on their goals. Performative speech is the basis of, and template for, normative conduct.Social construction is always normative.

c) As media, rules transform available materials into resources, eventuating in asymmetric opportunitiesfor control and the asymmetric distribution of benefits. This is rule, and rule is to be found in everysocietyincluding international society.

Not coincidentally, 1989 is the same year that Friedrich Kratochwils important book on Rules, Norms andDecisions appeared. Though widely cited as constructivisms foundational texts, neither his book nor mine has hada great deal of substantive impact. Indeed neither has had the impact of two papers (Wendt, 1987, Dessler, 1989)that also deserve to considered foundational. In an account of what I hope to show has been a strange career forconstructivism, it is entirely relevant to ask why this should have been so.

Both Kratochwils book and mine are technically demanding as I have acknowledged in my own case;see Kublkov et al., 1998:20). As for whether they are needlessly so, let me point out first that both books takemost readers into unfamiliar worlds. Adjusting to the unfamiliar always takes work. Both of us engage in a greatdeal of conceptual clarification, thereby calling on readers to do their share, so to speak, by reading closely. World

1111

of Our Making adds to the cost of careful reading, though very little in my opinion, by asking readers to supply someof the relevant connections to the worlds of international relations. After all, these are familiar worlds for mostreaders, and connections will come readily to mind.

Close, often critical reading of diverse texts is the method that I used in the first place to propound myversion of constructivism. This method yielded a large number of assertions whose source readers will recognizeeven after I reformulated and assembled them systematically. Pedigrees matter, especially to scholars. For other,less demanding readers, I recently outlined constructivisms essential propositions with no reference whatsoever toanyone elses work (Onuf, 1998b).

Unlike World of Our Making, this unadorned but systematic exposition is, I believe, easy to read. Whetherits systematic arrangement discourages just the sort of bricolage that I engaged in myself, and that socialconstruction always involves, remains to be seen. Here I might note that Harold Lasswell and Abraham KaplansPower and Society (1950) provided me with an explicit model for all of my systematizing efforts. However much Icontinue to find this magnificent book suitable for selective appropriation, others seem not to, for it is long out ofprint and rarely cited.

There are (at least) two other reasons why World of Our Making failed to have a greater impact than it did.In the first place, its emphasis on the connection between rules and rule disturbs liberals. I define rule as a conditionin which some people use rules, which are never neutral in content, to control the conduct of others. The conditionof rule always confers advantages on those who exercise it. International relations has rules, as any liberal wouldagree. Few liberals would further agree with me that these rules can only be understood as resulting in a conditionof rule, whether or not we choose to call it this.

Instead, liberals imagine that rules need do no more than establish the conditions of agency.Anarchyrule by no-oneis the presumptive result. If anarchy confers advantages on particular agents, it does notdo so with the consistency that the concept of rule suggests. Most students of international relations, includingrealists, hold this view, which makes them liberals by my reckoning. It also makes them wrong. Insofar as thesesame students think that international relations are endemically anarchic, and distinctively so, then they are doublywrong.

Second, World of Our Making starts with the claim that International Relations is a field lacking adistinctive subject. Rather more implicitly, Kratochwil made the same claim. Any such claim is deeply threateningto those who have made their careers on the premise of distinctiveness. This quest for distinctivenessan elusivequest for a general theory of international relations (Ferguson and Mansbach, 1988)is indeed one way to writethe fields history. As a field, International Relations is no less a world because it lacks a distinctive subject tovalidate its claim to autonomy. This world exists because we, who are its agents, say it does, and because observersby and large defer to what we say about ourselves.

Changing Worlds

Long centered in the United States, the field of International Relations has experienced a number ofchanges necessarily related to changes in the world of events, not to mention other worlds of scholarship. I havealready commented on the declining vitality of positivist social science. By the 1980s, rationalist theory, notpositivist science, dominated the field. This tendency toward formalization culminated in the twinning ofneorealism and neoliberalism (Baldwin, 1993). If the twins stood astride the field, around the margins there hadassembled a ragtag crowd of dissidents.

As the Cold War ended, the twins had almost nothing to say about the scale of change (see Kegley andRaymond, 1994; Lebow and Risse-Kappen, 1995, for useful discussions). National identity emerged as a large issueof interest. However loosely conceived, social construction offered an explanation for the formation of identities. Itis in this context that Alexander Wendt (1992) popularized the term constructivism. Suggesting that the basicsociological issue of identity- and interest-formation gives constructivism its identity, he deliberately

12

dissociated himself from recent epistemological debates (p. 393). With dissidents sent to the sidelines, manyyoung scholars saw the way open for contextually enriched, normatively aware scholarship.

Though described as constructivist, much of this work is reminiscent of the liberal institutionalistscholarship that had fallen out of favor decades before. Wendts turn away from philosophical issues turned out tobe a mixed blessing. As the field gained from an expanded agenda, constructivism lost what should have been itsmost distinctive features. Any concern with language was left to the dissidents; as a derivative of performativespeech, rules ceased to matter; without rules as media of social construction, the co-constitution of agents andstructures became an airy abstraction, needlessly subject to just the sort of debate that Wendt seems to have hadenough of. Ironically, Wendt holds some of the responsibility for making the agent-structure debate so confusedand pointless that most scholars (I among them) will have nothing to do with the subject (see Gould, 1998, for thewhole, ugly story).

In due course well-established scholars began to identify themselves as constructivists. Peter Katzensteinbecame a leading figure as editor of The Culture of National Security (1996). Soon thereafter John Ruggie(1997:11-3) announced that he had been a constructivist all along. Although Ruggie is something of an exception,most of the late-joiners had no interest in the social-theoretical and philosophical backdrop of constructivism as Ihad presented it in World of Our Making. In keeping with Wendts sociological inclinations (1992:393), the newconstructivists drew on the new institutionalism in sociology for inspiration (see Finnemore, 1996a for an overview).Routinely positivist and conceptually anemic (see for example Scott, Meyer and associates, 1994:1-112), thisliterature served mostly to reinforce a deplorable tendency to talk about culture in the vaguest terms imaginable.

Constructivism today is still construed as a third way, or middle ground, between positivism and post-modern epistemological radicalism (Adler, 1997:321-3; Checkel, 1998:327). In practice, however, many scholarsincreasingly treat it as the main alternative to the neo-twins, which, after all, have very little separating them(Ruggie, 1997:9-11 aptly called them both neo-utilitarian). Post-modern scholars must be surprised to findthemselves thrown together with constructivists (see Jepperson, Wendt and Katzenstein, 1996:46, Ruggie, 1997:35,for examples of this practice). If indeed the third debate that Lapid announced in 1989 can now begin (Adler,1997:348), it is only because so many scholars believe that there are only two ways to proceed. If they are notrational choice theorists, then they must be constructivists. Either way, empirical research will suffice to settledifferences. Indicatively, two recent surveys bear the exact same section title: A Constructivist Research Agenda(Adler, 1997:341-7, Hopf, 1998:186-92).

The end of the cold war induced another significant change in the scholarly world of international relations.As Marxists declined in influence, many European scholars joined this world, but not without reservations. Theyresented the longstanding dominance of scholars in the United States, whose predilection for a narrow constructionof international relations as a world of states they did not share. Constructivism offered an attractive alternative.While taking off in a number of directions (see for example Fierke and Jrgensen, 2001), European constructivistsare generally more sensitive than their counterparts in the United States to the links between language, rules andrule. In this respect, Maja Zehfusss discussion of constructivism (2001) is far better than other surveys and reviewessays recently published in English (Adler, 1997; Checkel, 1998; Hopf, 1998).

As the field globalizes, constructivism is everywhere, perhaps in danger of becoming all things to allscholars, finally suffering the fate of all fads. Symptomatically, talk of constructivism has seeped across the alwaysleaky boundary between International Relations and International Law as a field of study (Slaughter, Tulumello andWood, 1998). No less symptomatic are the flurry of essays sizing up constructivism (cited in the precedingparagraph), not to mention acknowledgment in general surveys (Dougherty and Pfaltzgraff, 1996:162-3; Knutsen,1997:279-82) and journals of opinion (Walt, 1998:38, 40-1). Perhaps the greatest measure of constructivismsiconic value is the juxtaposition of rationalist approaches and constructivist approaches to organize themonumental special issue of International Organization sizing up its fifty years of publication (Katzenstein,Keohane and Krasner, 1998). Yet another major event was the appearance, several months after I was done withthis essay, of Wendts long-awaited, ambitious, demanding and undeniably important book (Wendt 1999). Myengagement below with some of Wendts recent claims does not take this book into accountthe terms ofengagement would not have changed very much had I done so.

1313

There are signs that constructivism can survive all this attention without slipping into blandness. Feministand post-colonial studies have managed to do so. Furthermore, there are significant opportunities for reinforcementamong the survivors, all of which share a concern for rules in relation to language on the one hand and rule on theother. Indicatively, Elisabeth Prgl (2000) draws on feminist theory, and Lily Ling (forthcoming) on post-colonialtheory to do just this.

Constructivisms Enduring Strengths

In the decade since I introduced the term constructivism, it has come to be used in ways that I had notforeseen and, more to the point, in ways that vitiate what I meant to convey by the term. Notwithstandingconstructivisms strange career, I see no need to revise the premises that guided me initially, nor even to change verymany of the systematically related features of the text that appeared over a decade ago.

If I had been a better salesman, I would have spent the years since extolling constructivism as I continue tounderstand it. Belatedly, let me offer five reasons why any scholar should seriously consider working within thisframework. These are the enduring strengths of constructivism in its strong form. They derive from, and furthersubstantiate, the basic premises of constructivism, as I presented them above.

First, the strong version of constructivism fosters a sensitivity to language as doing. It fosters a dispositionto use metaphors of work, as against drama (cf. Harr, 1993:148-85, on Social Action as Drama and Work[chapter title]). I do not want to suggest that the metaphorical language of play does not have its uses. Actually, Iuse this sort of language all the timeany bricoleur would. Nevertheless, there are two reasons why I think that weshould prefer the metaphorical language of work.

Realists prefer the language of play, most emphatically in reference to war. The deployment of theatricaltropes and the terminological conventions of game theory help constitute a world that realists claim merely toobserve, but further claim is distinctiveindeed uniquely set apart from the mundane world of ordinary people andthe way they carry on with their lives (also see Fierke, 1998:31-43). Just as theater creates a world of elementalsimplicity to which an audience responds with heightened awareness, so too does the realist rhetoric of struggle,violence and fear. I have already indicated why I think that the quest for distinctiveness is a mistake. I would addhere that success in this respect has the effect of making the world of international relations distinctively terrifyingbut, paradoxically, a further step removed from the immediate world of experience that we routinely call reality.

I might also note that the post-modern penchant for play, spectacle and narcissismfor theatricalposturingbetrays a deep connection to realism. More precisely, both betray a common heritage in the Romanticimpulse of the 19th c. It is just this impulse that neo-realism would banish through the disciplining effects of formallanguage nevertheless dependent of the metaphors of work. The term structure, which is so conspicuously central toWaltzs refurbishment of realism (1979), illustrates the point (here again see Fierke, 1998:47-50).

The language of work and the language of play do have one feature in common. Both convey a sense thathuman activity is intentional. The language of play emphasizes the role of language (notice the metaphor) inrepresenting the worldin mimicking it. In a complementary way, the language of work, of doing, underestimatesthe importance of speaking as a tool (again, notice the metaphor) for getting on with the tasks that we set forourselves. Once we grant that speech is the most powerful tool available to us, then metaphors of work tell us howpervasive social construction is, even in the absence of immediate intention and conscious design.

Let me turn to a second strength of constructivism in its strong form. It suggests due attention to process,and thus to social construction as an element in any social process. In turn, social construction points to the workthat rules do. By presenting agents with choices, rules affect conduct. Conversely, the pattern of choices affectrules, strengthening those that agents choose to follow and weakening those that they do not. In aggregate, rulesperform regulative and constitutive functions because each and every rule simultaneously produces both kinds ofeffects. This is so whether rules are informalso informal that many scholars call them norms and conventionsor

14

highly formalso obviously set apart from other rules that, by convention, we call them law and divide them byfunction.

In the context of international relations, constructivists soon discover that informal rules are ubiquitous butthat their properties and effects are infuriatingly difficult to pin down. The constructivist emphasis on rules alsovalidates scholarly attention to the formal rules making up international law and international organizations (cf.Adler, 1997:335, but note Martha Finnemores unconsidered complaint, 1996b:139, that constructivists giveinternational law insufficient attention). Whether regulative or constitutive in intent, these rules affect agents in avariety of ways and through them a variety of rules, formal and informal. As Harvey Starr has reminded me, it ishard to study international law and not become an intuitive constructivist. Furthermore, the constructivist emphasison rules leads to a consideration of the ways that agents justify their choices and thus to the place of ethics ininternational relations (Onuf, 1998c).

Constructivism in its strong form has a third strength that few self-styled constructivists will recognize assuch. Constructivism undermines the liberal tendency to ignore rule (asymmetries of control and privilege) as aconstant feature of social relations. If, as Michael Smith suggested to me, constructivism is little more than hard-headed liberalism, then few liberals manage to be hard-headed in the systematic way that constructivismencourages them to be. While I have more to say about the limitations of liberalism below, here I want to make mypoint about constructivism more affirmatively.

Constructivism fosters appreciation of rules that turn raw materials into resources and make rule possible.In the first instance, rules constitute resources by making material conditions generally intelligible. Linking materialconditions to human needs and goals is an integral part of this process. Rules also constitute resources by definingthe terms of agency. They tell us which agents, under what circumstances, have access to materials that other ruleshave assigned uses to.

Among agents whose goals and circumstances inevitably differ, differential access confers advantagesunevenly. For that matter, so does the analytically prior process by which rules make material conditionsintelligible, though perhaps more subtly. It is important not to be misled by the fact that rules always give agentschoices, along with some indication of the benefit that they may expect to gain from following the rules and the costthat should expect to incur from not doing so. Whatever choices agents make, rules cannot distribute consequencesneutrally because of their content, and not because of the way that they work.

Finally, rules constitute resources by giving some agents the opportunity to use the materials to which therules give them access to make and support still other rules that benefit them. For example, agents can use materialsat their disposal to influence the choices that other agents make with respect to following rules. Enforcement anddeterrence describe this process, which characterizes a familiar form of rule. If resources make rule possible, rulesmake resources what they are. They also make rule in some form an unavoidable feature of social relations.

As a fourth, somewhat related point, constructivism in its strong form holds that material conditions alwaysmatter, but they never matter all by themselves. On the one hand, constructivism abjures the vulgar materialism thatrealists are allegedly prone to. Wendt, for example (1995:73), charged neorealists with a desocialized view ofmaterial capabilities. The charge is unfair. In responding to Wendts many charges, John Mearsheimer (1995)simply ignored this one. Ever since Harold and Margaret Sprout (1965) vetted the relevant issues with exceptionalthoroughness, few realists construe capabilities in the brute, physical sense that Wendt supposed them to.

On the other hand, constructivism is not philosophically idealist. As for Wendts claim thatconstructivism has an idealist (or idea-ist) view of structure (Wendt, 1995: 73), I concur with this view onlybecause I believe that structures are nothing more than observers descriptions of the stable patterns that rules andpractices exhibit in any world (Onuf, 1998b:61-2). As soon as agents act on these descriptions (which may indeedstem from their own observations), structures enter the world, so to speakthey enter into the process of socialconstruction. The process itself has the effect of institutionalizing structures, including the terms of agency. Anyworld, and every institution, is always and necessarily dependent on socially mediated material conditions.

1515

Wendt (1995:73) has affirmed that material conditions are necessarily social, but not the converse. As aconsequence, capabilities interest him less than structures of shared knowledge (p. 73). Such a view ends upreducing the social world to mental states, and social construction to the diffusion of ideas. Identity becomes themain preoccupation of agents, as if they had nothing else to do. Norms are everywhere, but they seem to do verylittle besides granting agents their identity.

This sort of idealism is a perennial tendency for progressive liberals. It is also a particularly striking featureof the limp constructivism of recent fashion (Katzenstein, 1996, is chocked with examples.) No term illustrates thistendency more graphically than norm (also see Onuf, 1998c). Taken as more or less synonymous with sharedexpectations, norms exist outside of minds only incidentally. That norms cannot be shared without taking the formof linguistic statements brings them into the world of artifacts. As such, they are rules, if not directly linked tomaterial circumstances, then linked to other rules that are. Agents use rules in their social relations for a variety ofpurposes, most, if not all, of them with material implications.

In strong form, constructivisms last strength is its methodological openness. Consider Hayward Alkersmastery of diverse methods to advance his wide-ranging concerns, many of which center on the uses of language(Alker, 1996). Constructivism is perfectly consistent with rational choice theorizing, high positivist quantitativeanalysis, thick description and whatever else most of us actually do in the name of scholarship. Constructivismcannot be reconciled with claims that any particular set of scholarly activities are the only ones accuratelyrepresenting what happens in the world (any world). Nor can constructivism be reconciled with claims that facts (bywhich I mean the represented properties of worlds) are homogeneous, knowledge is cumulative, and worldsthemselves stand in some sort of Leibnizian or Comtean order. By contrast, bricolage makes do with heterogenousworking materials.

All such claims mark the divide between positivism and its critics. Stationed on one side or the other, mostself-styled constructivists fail to see any way to reach across the divide (cf. Hopf, 1998:181-5 on critical vs.conventional constructivists). The reason, I suspect, is the widely shared belief that ones position on the fact-value distinction compels a choice between incompatible methodological allegiances. Either one accepts theconventional positivist position that facts and values are always separable in principle and that being neutral onmatters of value is an important scholarly value. Or one takes the position that facts are always laden with valuesand that one can never been neutral on matters of value.

Constructivism acknowledges the fact-value distinction. From an observers point of view, values arefacts. As such, they are identifiable in what agents do (and saying is doing). Once identified, values as facts arecapable of being separated from other kinds of facts, at least in principle. Indeed, moral reflection depends on justsuch an operation.

Constructivism also holds that, in practice, values pervade social relations. For any agent, speaking is, as Isaid above, always normative. Speaking is inextricably related to the achievement of the speakers goals, and thus itis always laden with value. As soon as the observer (or perhaps I should say auditor) talks about, or otherwise actson, the facts, however ascertained, then that observer becomes an agent, and those facts take on value.

While observers and agents may occupy separate worlds (this is a positivist ideal), observers usuallyparticipate, as agents, in a world that they stand apart from only provisionally. As a consequence, any distinctionthat they make between facts and values (as facts of a different kind) is also provisional. It is no less useful for thefact of being provisional. As with people in general, constructivists have it both ways: they have no particulardifficulty dealing with values, as facts, when they stand aside to do so, and they know what happens to these factswhen they proceed to act upon them.

Worlds Apart?

Throughout this essay, I have argued that worlds have porous boundaries, that all of us belong to manyworlds and constantly traffic among them, and that a world of worlds is subject to constant social construction. It ispossible to overstate these claims. Some worlds baffle observers from other worlds. To the extent that they succeed

16

in making sense of what they see, what they say makes little or no sense to the agents whose world it is. In turn,observers acquire only the most limited agency for themselves.

On occasion, we hear talk of paradigms as if they were incommensurable worlds. If this were so, then mostother talk about paradigms misses the point: there is nothing that we as outsiders can say about worlds so utterlydifferent from our own. We might better say that paradigms resemble distant worlds that have little need for eachother. They coexist in mutual disregard, their agents caught up in concerns too different to attract each othersattention.

We might also ask if constructivism (which is to say, my construction of constructivism) is a paradigm inthis more limited but still quite general sense. My answer to this questiona negative answercalls for a fullerconsideration of the term paradigm. If constructivism is not a paradigm in any general sense of the term, then whatindeed is it? I conclude the essay with a partial, and perhaps not very satisfactory answer to this second question.

World of Our Making identifies three senses of the term paradigm. One is relevant to the books centralpremises, another to its systematic properties. Neither draws inspiration from Thomas Kuhns familiar use of theterm (Kuhn, 1970; see generally Onuf, 1989: 12-27), which I called puzzle paradigm and which, I believe, is the onein general use. If International Relations could indeed be said to have built a paradigm around a central theoreticalpuzzlefor example, how is anarchy tenable over any period of time?then indeed we might want to concede thedisciplinary claim of distinctiveness. The price of doing so, I might add, is a field of such modest proportions andlimited prospects that few among us would be content to work within its confines.

Borrowing from Sheldon Wolin (1980), I prefaced paradigm with the modifier operative to capture thesense of the term as it relates to constructivisms central premises. An operative paradigm is an ensemble of rulesand practices, which agents speak of in world-defining terms and respond to as normatively compelling. Liberalismand the liberal world are aspects of a single operative paradigm. Liberalism as a way of talking about socialrelations makes the liberal world a coherent whole. At the same time, the evident coherence of the liberal worldconfirms the normative force of liberalism.

Seen from the outside, the liberal world lacks the coherence that liberals take for granted. Liberals prefer tothink that rules in the form of rights minimize the need for rule. I would say instead that an ensemble of rights andduties constitutes an enduring form of rule, often styled the rule of law. Thanks to the rules, some agents benefitbeyond their duethey rule without the vestments of rule. Liberal societies exhibit this form of rule to a degreeunmatched in other societies. Yet they combine this form of rule with others in complex arrangements that hardlymake sense if liberalism is the only way of talking about social relations. For the observer, there is more to theliberal world than its operative paradigm can ever convey.

As I remarked earlier, liberals describe the world of international relations in the terms that they knowbestliberal termsbecause their liberal beliefs prevent them from seeing the way that rule works. The world ofinternational relations is a world of rules and rule, even if no-one ever claims to rule. No world is ruled by no-one:the very idea of anarchy is a contradiction in terms. Insofar as this world consists of states bound together by rightsand duties, it is a liberal society ruled to the benefit of some states over others, despite the apparent absence of rule.Yet the world of international relations is more complex than this. Insofar as it displays features that I identifiedwith functional institutions (the world of services) and status-conscious agents of state (the world of standing andstatecraft, other forms of rule coexist, as do other operative paradigms.

Agents rule the world of international relations as they rule in every world. They rule with rules that otheragents must take into account in their conduct. The paradigm that I see operating wherever there is rule I calledpolitical society. Within this paradigm is a place for the operative paradigm that liberals take for granted, not tomention places for other operative paradigms that liberals take little interest in. The world of international relationsis a political society in the first instance. Thanks to the operative paradigm that makes it so, it works pretty much asany political society does.

1717

Most worlds do have porous boundaries, forms of rule blend together, operative paradigms overlap. Asobservers make some things clearer, they must impose order on what they claim to know. If constructivism is to doits job, it has to be systematic in a way that worlds made by many and diverse agents can never be. Drawing onRobert Merton (1968:64-72)and Talcott Parsons (1978:352-3), I used the modifier codificatory to capture the senseof the term paradigm as it relates to the systematic features of World of Our Making. A codificatory paradigm is afully worked out system of categories, a world made whole. As a tool, it helps any observer bring order to messyworlds. Comtean positivism, with its ontological levels of irreducible causal relations (Onuf 1998a: ch. 8), offersanother such paradigmone that constructivists might see as inhibiting opportunistic bricolage.

It was surely a mistake to adopt the term codificatory paradigm, and not just because it soundspretentiouswhich it does. Dealing with three discrete senses of the term paradigm, two of them unfamiliar,undoubtedly taxed readers ability to keep all the relevant distinctions in mind. My systematic intentions might justas well have been served without invoking the term paradigm in the third sense, and possibly in any sense. This isnot to say that I have second thought about the writing a book that does function as a codificatory paradigm, if onlyfor myself and a few other like-minded souls.

Constructivism is not a paradigm in the general sense of the term, even if it draws attention to meshedworlds as operative paradigms. Instead, constructivism is a way of studying any world of social relations (Onuf,1998b:58). My labeling it this way is deliberate, but perhaps unduly vague. Better to call it a framework, as I didearlier in this essay (also see Kublkov et al., 1998:19, and note the subtitle of Lasswell and Kaplan, 1950: AFramework for Political Inquiry). Indeed constructivism offers an inclusive framework for the study of socialrelations. Notice that this claim suits my metaphorical inclinations, and it tells us broadly what constructivism is infunctional termswhat we can use it for.

As a framework, constructivism makes it possible for observers to propose any number of theories, orgeneral explanations, for what happens in any world, and it allows observers to fit theories together. Constructivismis not a theory itself (Onuf, 1998a:188-9), although I confess that I may have confused the issue recently by callingit a theoretical stance (Onuf, 1997:7). Few self-proclaimed constructivists seem to be confused; they do not seemto treat constructivism as a theory even if, as bricoleurs, they talk the language of theory. Wendt is an exception.

More precisely, Wendt has become an exception in his recent work. Constructivism is a structural theoryof international politics: intersubjective structures explain much of what happens in a world of states (Wendt,1994:385, 1996:48). Early on (1987), Wendt emphasized the co-constitution of agents and structures, not as anexplanation, but as a description of how any world works. By the mid-1990s, we find him having abandoned thisposition (expressly in Wendt, 1996:48-9), evidently on the mistaken belief that this description denies the possibilityof explaining anything in general terms. Because co-constitution is a comprehensive description, and constructivisma capacious framework, there is plenty of room for Wendts structural theory. There is room too for any number ofother theoriestheories that start with social relations and end up with agents, theories that run in the otherdirection, theories that start with rules as the media of social construction and run both ways.

Wendts commitment to building theory is commendable. If, by calling World of Our Making meta-theory (Wendt, 1991), he meant to criticize me for not developing a theory, structural or otherwise, then this sort ofcriticism misses the mark. Frameworks come first. I put together a large and sturdy framework from the diversetheoretical materials of many disciplines because I thought that International Relations needed one. Insofar as theterm meta-theory has become a epithet for self-indulgent epistemological posturing, I am even less inclined toaccept it. Let me shift metaphors. Constructivism finds a way between epistemology and methodology by takingthe ontological turna turn that opens up the road to theory.

References:

Adler, Emmanuel (1997) Seizing the Middle Ground: Constructivism in World Politics, European Journal ofInternational Relations 3:319-63.

18

Alker, Hayward R. (1996) Rediscoveries and Reformulations: Humanistic Methodologies for International Studies.Cambridge: Cambridge University Press.

Ashley, Richard K. (1981) Political Realism and Human Interest, International Studies Quarterly 25:204-36.

______ (1984) The Poverty of Neorealism, International Organization 38:225-86.

______ (1987) The Geopolitics of Geopolitical Space: Towards a Critical Theory of International Politics,Alternatives 12:403-34.

_______ (1988) Untying the Sovereign State: A Double Reading of the Anarchy Problematique, Millennium:Journal of International Studies 17:227-62.

_______ (1989) Living on Borderlines: Man, Poststructuralism, and War, pp. 259-321. InInternational/Intertextual Relations: Postmodern Readings of World Politics, edited by James Der Derianand Michael J. Shapiro. Lexington MA: Lexington Books.

Ashley, Richard K., and R. B. J. Walker (1990) Speaking the Language of Exile: Dissidence in InternationalRelations, special issue of International Studies Quarterly 34 (3).

Baldwin, David A., ed. (1993) Neorealism and Neoliberalism: The Contemporary Debate. New York: ColumbiaUniversity Press.

Checkel, Jeffrey T. (1998) The Constructivist Turn in International Relations Theory, World Politics 50:324-48.

Dessler, David (1989) Whats at Stake in the Agent-Structure Debate? International Organization 43:441-474.

Dougherty, James E., and Robert L. Pfaltzgraff, Jr. (1996) Contending Theories of International Relations: AComprehensive Survey, 4th ed. New York: Longman.

Ferguson, Yale H., and Richard W. Mansbach (1988) The Elusive Quest: Theory and International Politics.Columbia SC: University of South Carolina Press.

Fierke, K. M. (1998) Changing Games, Changing Strategies: Critical Investigations in Security. Manchester:Manchester University Press.

______ and Knud Erik Jrgensen, eds. (2001) Constructing International Relations: The Next Generation. ArmonkNY: M. E. Sharpe.

Finnemore, Martha (1996a) Norms, Culture, and World Politics: Insights from Sociologys Institutionalism,International Organization 50:325-48.

______ (1996b) National Interests in International Society. Ithaca NY: Cornell University Press.

Giddens, Anthony (1984) The Constitution of Society: Outline of the Theory of Structuration. Berkeley and LosAngeles: University of California Press.

Gould, Harry D. (1998) What Is at Stake in the Agent-Structure Debate? pp. 79-98. In International Relations ina Constructed World, edited by Vendulka Kublkov, Nicholas Onuf and Paul Kowert. Armonk NY: M.E. Sharpe.

Habermas, Jrgen (1984) The Theory of Communicative Action,1, Reason and the Rationalization of Society,translated by Thomas McCarthy. Boston: Beacon Press.

1919

Harr, Rom (1993) Social Being, 2nd ed. Oxford: Blackwell.

Holsti, K. J. (1985) The Dividing Discipline: Hegemony and Diversity in International Theory. Boston: Allen &Unwin.

Hopf, Ted (1998) The Promise of Constructivism in International Relations Theory, International Security23:171-200.

Jepperson, Ronald L., Alexander Wendt and Peter J. Katzenstein (1996) Norms, Identity, and Culture in NationalSecurity, pp. 33-75. In The Culture of National Security: Norms and Identity in World Politics, edited byPeter J. Katzenstein. New York: Columbia University Press.

Katzenstein, Peter J., ed. (1996) The Culture of National Security: Norms and Identity in World Politics. New York:Columbia University Press.

Katzenstein, Peter J., Robert O. Keohane and Stephen D. Krasner (1998) International Organization at Fifty:Exploration and Contestation in the Study of World Politics, special issue of International Organization52 (4).

Kegley, Jr., Charles W., and Gregory Raymond (1994) A Multipolar Peace? Great-Power Politics in the Twenty-first Century. New York: St. Martins Press.

Keohane, Robert O., ed. (1986) Neorealism and Its Critics. New York: Columbia University Press.

Knutsen, Torbj_rn (1997) A History of International Relations Theory, 2nd ed. Manchester: Manchester UniversityPress.

Kratochwil, Friedrich (1989) Rules, Norms, and Decisions: On the Conditions of Practical and Legal Reasoning inInternational Relations and Domestic Affairs. Cambridge: Cambridge University Press.

Kublkov, Vendulka, Nicholas Onuf and Paul Kowert (1998) Constructing Constructivism. In International Relations in a Constructed World. Armonk NY: M. E. Sharpe.

Kuhn, Thomas S. (1970) The Structure of Scientific Revolutions, 2nd ed., enlarged. Chicago: University of ChicagoPress.

Lapid, Yosef (1989) The Third Debate: On the Prospects of International Theory in a Post-Positivist Era,International Studies Quarterly 33:235-54.

Lasswell, Harold D., and Abraham Kaplan (1950) Power and Society: A Framework for Political Inquiry. NewHaven CT: Yale University Press.

Lebow, Richard Ned, and Thomas Risse-Kappan, eds. (1995) International Relations Theory and the End of theCold War. New York: Columbia University Press.

Lvi-Strauss, Claude (1966) The Savage Mind. Chicago: University of Chicago Press.

Ling, L. H. M. (forthcoming) Conquest and Desire: Postcolonial Learning between Asia and the West. London:Macmillan.

Merton, Robert K. (1968) Social Theory and Social Structure, enlarged ed. New York: Free Press.

Mearsheimer, John J. (1995) A Realist Reply, International Security 20:82-93.

20

Onuf, Nicholas (1987) Rules in Moral Development, Human Development 30:257-67.

______ (1989) World of Our Making: Rules and Rule in Social Theory and International Relations. Columbia SC:University of South Carolina Press.

______ (1997) A Constructivist Manifesto, pp. 7-17. In Constituting International Political Economy, edited byKurt Burch and Robert A. Denemark. Boulder CO: Lynne Rienner.

______ (1998a) The Republican Legacy in International Thought. Cambridge: Cambridge University Press.

______ (1998b) Constructivism: A Users Manual, pp. 58-78. In International Relations in a Constructed World,edited by Vendulka Kublkov, Nicholas Onuf and Paul Kowert. Armonk NY: M. E. Sharpe.

______ (1998c) "Everyday Ethics in International Relations," Millennium: Journal of International Studies 27:669-93.

______(2001) Agency, Institutions and Structuration in Late Modern Civil Society, paper presented at Conferenceon Global Civil Society: Critical Engagements, Gregynog, Wales. Copy on request.

Parsons, Talcott C. (1978) Action Theory and the Human Condition. New York: Free Press.

Prgl, Elisabeth (2001) The Global Construction of Gender: Home-based Work in the Political Economy of the 20thCentury. New York: Columbia University Press.

Ruggie, John Gerard (1997) Constructing the World Polity: Essays on International Institutionalization. London:Routledge.

Scott, W. Richard, John W. Meyer and associates (1994) Institutional Environments and Organizations: StructuralComplexity and Individualism. Thousand Oaks CA: Sage.

Shapiro, Michael J. (1981) Language and Political Understanding: The Politics of Discursive Practices. NewHaven CT: Yale University Press.

Slaughter, Anne-Marie, Andre S. Tulumello and Stepan Wood (1998) International Law and InternationalRelations Theory: A New Generation of Interdisciplinary Scholarship, American Journal of InternationalLaw 92:367-97.

Sprout, Harold, and Margaret Sprout (1965) The Ecological Perspective on Human Affairs, with Special Referenceto International Politics. Princeton NJ: Princeton University Press.

Walt, Stephen M. (1998) International Relations: One World, Many Theories, Foreign Policy, No. 110:29-46.

Waltz, Kenneth N. (1979) Theory of International Politics. Reading MA: Addison-Wesley.

Wendt, Alexander (1987) The Agent-Structure Problem in International Relations Theory, InternationalOrganization 31:335-70.

______ (1991) Bridging the Theory-Meta-Theory Gap in International Relations, Review of International Studies17:383-92.

______ (1992) Anarchy is What States Make of It: The Social Construction of Power Politics, InternationalOrganization 46:383-425.

2121

______ (1994) Collective Identity Formation and the International State, American Political Science Review88:384-96.

______ (1995) Constructing International Politics, International Security 20:71-81.

______ (1996) Identity and Structural Change in International Politics, pp. 47-64. In The Return of Culture andIdentity in IR Theory, edited by Yosef Lapid and Friedrich Kratochwil. Boulder CO: Lynne Rienner.

Wolin, Sheldon (1980) Paradigms and Political Theories, pp. 160-91. In Paradigms and Revolutions, edited byGary Gutting. Notre Dame IN: University of Notre Dame Press.

Zehfuss, Maja (2001) Constructivisms in International Relations: Wendt, Onuf, and Kratochwil, pp. 54-75. InConstructing International Relations: The Next Generation, edited by Karin M. Fierke and Knud ErikJrgensen. Armonk NY: M. E. Sharpe.

22

Construction Sites

Wayne SandholtzUniversity of California, Irvine

Is there a constructivist research program and, if so, what are its chances for future progress? HaywardAlkers note prodded us to reflect on these and other questions. I have been hoping against hope that someonewould come up with convincing, positive answers, because I have not been able to. If a research program is acollective enterprise built around common questions, holding shared standards of progress and generatingcumulative knowledge, we do not have one, nor are we likely to any time soon. Does this conclusion mean thatconstructivism is a fad that will leave no enduring traces in the world of international relations scholarship?Paradoxically, perhaps, the answer is, No.

If constructivism has not established a genuine research program, it is not because the concerns, or thepractitioners, are fundamentally misguided. On the contrary: constructivism probes questions that will always be atthe heart of social inquiry, and its adherents have produced creative, insightful research that expands the pool ofhuman knowledge. The problem lies in the broader field of international relations, and it is (not surprisingly) one ofsocial norms and expectations. Our institutions reward theoretical precociousness. Every IR scholar must (or feelsthat s/he must) make a theoretical contribution. As a result, the dominant incentives are to expose the flaws inexisting theories and to propose something better. Collaborative research is difficult enough (the herding catsproblem), but building a research program involving more than a handful of scholars seems out of the question (seeHolsti 1985; Ferguson and Mansbach 1988).

So, we have many flavors of constructivism, with diverse questions, concepts, and methods. A growingnumber of researchers find persuasive the fundamental constructivist insight, that our worlds are at least as muchsocial as they are material. And clearly, scholars are deploying various constructivist ideas and tools to ask fruitfulquestions about the world and find illuminating answers. Perhaps, then, we should be more conscious of the virtuesof pragmatism: scholars use the tools that help them to solve the intellectual problems at hand. Constructivism willgenerate progress in this more diffuse, though pragmatic, sense because it zeroes in on what has always been thecore of social science, namely, the ways in which people act in a collective world of which they are both the creatureand the creator. Unlike other broad perspectives, constructivism recognizes that our capacity to navigate, and actupon, the physical and social worlds emerges out of collectively held understandings and rules. In this sense,constructivism addresses the questions that have been vital at least since Weber and Durkheim, if not Aristotle.

Our convenors also asked that we reflect on the question of what is to be done, where constructivist workmight move so as to achieve progress. Given what I have said about fragmentation and herding cats, any attempt totell everybody else what we should be doing would be fruitless and pretentious. Instead, I will outline several areasof inquiry that I see as both fundamental to constructivism and liable to offer decent payoffs in the form of expandedunderstanding of our worlds. I start with the foundations laid out by Nick Onuf. I see the major task for rule-centered constructivism as converting highly abstract social theories into mid-range theories of internationalrelations that can guide systematic empirical research. The thoughts that follow are therefore not a work plan forconstructivists, but rather a sketch of the kind of work I expect to be doing over the next few years.

The foundations are, briefly stated:

1. Rules are central, because they link agents and social structures. Rules both establish agency, enablingand constraining action, and define institutions they are thus the means of mutual constitution.

2. Rules are in constant flux. Most of what actors do and say has an effect on the relevant rules, whether toconsolidate or erode them (Onuf 1994).

3. Rules create the condition of rule, which is another way of saying that rules and power are inextricablyintertwined.

Given these foundations, I explore three areas in which constructivism could build: the dynamics of rulechange, the connection between rules and rationality, and the relationship of rules to power.

2323

Dynamics of rule change

If rules are the basic building blocks of all social structures, including international relations, thenexplaining rule change is crucial to explaining social change. Indeed, change in international society means changein the rules that constitute actors and roles and govern their interactions. Shifts in basic rules, like sovereignty rules,imply changes in the nature of international society. IR scholars have recognized the importance of changing norms,and tracked the emergence of new norms, including norms that abolished the slave trade, promoted decolonization,sanctioned apartheid, justified humanitarian intervention, created weapons taboos, and so on. Others have describedwhat I would call the political mechanics of international norm change, focusing on transnational activist networks,norm cascades and spirals, and domestic adaptation to transnational norms (Keck and Sikkink 1998; Risse andSikkink 1999; Brysk 2000). What is still lacking is a coherent account of why rules evolve. Why do somenormative claims emerge and why do they prevail over alternatives? The transnational models of norm change arereally based on pluralist politics: governments respond to mobilized political pressures. Why, in the first place, dothe claims of norm entrepreneurs win over followers and allies?

One way to begin is to recognize that norm entrepreneurs and activist networks are participants in a broaderprocess of normative evolution. If rules are in constant flux, why is that, and how does the process work? I suggestthat there is an endogenous dynamic of rule change. That is, rule systems contain the seeds of change. Even if onecould, as a thought experiment, hold power and technology constant, rule change would be constant and ubiquitous.Rule change occurs in a process that can be depicted as an endlessly repeating cycle linking rules to actions,discourses (argumentation), and change.

The cycle begins with the constellation of existing rules, which provides the normative structure withinwhich actors decide what to do and evaluate the behavior of others. Because rules cannot cover every contingency,and because conflicts among rules are inevitable (Hart 1994), actions regularly trigger disputes. The arguments areabout which norm(s) apply, and what the norms require or permit. The dominant form of argumentation is byanalogy: actors assert analogies between the act in question and some set of prior cases. When the analogy ispersuasive, other actors will agree that the current dispute should fall under the same norms that covered the earlier(analogized) cases. But the argument does not end there, for it remains to be determined what the norms require inthe present instance. Again, players argue by analogy with similar cases, in order to establish how the rules shouldapply to the case in question (if there are mitigating factors, if the case qualifies as an exception, and so on).

The outcome of such discourses is always to change the norms under dispute. If everybody agrees that thenorms apply without qualification, then the norms have been strengthened and the scope of their applicationclarified. If the relevant actors agree that the disputed act qualifies as a justified exception to the norms, then thescope of their application has also been clarified (the proliferation of exceptions, of course, can weaken a rule, whichis also a norm change). If the participants in the discourse fail to reach consensus, then that also modifies the normsin question, leaving their status weakened or ambiguous. Disagreements over the meaning of the rules, and over thejustifiability of specific acts, can continue unresolved over long periods of time. By the same token, when broadagreement exists, governments can fairly quickly formalize a rule (against slavery, or piracy, for instance) bysigning and ratifying a convention.

The crucial point, however, is that the cycle of normative change has completed a turn. In a givennormative structure, actions trigger disputes. Argument ensues, grounded in analogies with previous experience.The outcomes of these discourses modify the rules, whether by making them stronger or weaker, clearer or moreambiguous. The cycle returns to its starting point, the normative structure, but the normative structure has changed.The altered norms establish the context for subsequent actions, disputes, and discourses.

A few insights derived from this conception should resonate with various strands of constructivist research.First, arguments (or discourses) are crucial, because they produce the varying degrees of consensus anddisagreement that modify the rules. We have a lot of work to do to understand the workings of analogicalargumentation, and the bases of persuasion. Second, we find that historical connectedness is inescapable. The rawmaterials of the normative entrepreneur, or bricoleur, are the ideas, experiences, and norm discourses of the past.

24

Analogical argument, the primary mode of persuasion where rules are concerned, is by definition referential tohistory.

Rules and rationality

The foregoing discussion has already raised the issue of rationality, in the assertion that the dominant modeof rationality in discourses about rules is analogical. Arguments arise at the point where (necessarily) general rulesmeet the particularity of experience. Though in principle some such disputes can be resolved by strict logicaldeduction, specific cases routinely and inevitably cross the boundary of what can be so determined: the rules areunclear, the proper classification of the facts ambiguous, and so on (MacCormick 1978). Argument by analogytakes over. The task of disputants is to persuade that the case at hand resembles (or does not) past instances inwhich the rules were applied in particular ways. All of us routinely engage such arguments of fit, in disputes thatarise in families, sports, workplace, neighborhoods, and churches not to mention the courts.

The committed rationalist will respond that all decision-making is based on a utility-based rationality, inwhich one calculates the consequences of alternatives and maximizes expected payoffs. People favor theapplication of the rules that will make them better off. Normative justifications (and thus the choice of analogies)are strictly self-serving window dressing. In order to demonstrate that the realm of normative reasoning and choiceis at least partially autonomous and thus amenable to a distinct analytical approach, let us accept as a starting pointthe fully selfish, rational maximizer. We accept, only for the sake of argument, that the words and actions of thismaximizer derive solely from calculations of advantage (subjectively defined). When confronted with a disputeover the application of rules, the maximizer will argue for the application of whichever rules justify her utility-driven acts.

At this point we note something curious. The utility maximizer, in order to avoid costs, is motivated to winthe dispute. In order to win, she must offer the most persuasive arguments and analogies. But the determination ofwhich arguments are likely to prevail has nothing to do with utility calculations, and everything to do with socialstandards of fit, relevance, and interpretation. At this moment, the maximizer has entered the world of normativediscourse and reasoning by analogy. Naturally she wants the greatest possible payoff, but her success depends onher skill in understanding the groups historically evolved standards of similarity and precedent, and offeringpersuasive analogies to past cases and decisions. In short, though driven ultimately by selfish, utility-basedobjectives, the maximizer is compelled to operate within a normative rationality, based on shared, historicallycontingent standards of precedent and fit. Achieving the greatest payoff means mounting the most convincinginterpretation of the rules and their past applications. And that effort is not subject to internal utility calculations,but rather to external norms and understandings.

People are rational, in both the utility sense and the normative sense. I see the two modes of rationality notas mutually exclusive but rather as complementary; indeed, utility maximization requires normative rationality, orthe ability to understand and manipulate arguments based on analogy and fit. I further assume that people constantlyand routinely reason about both utility and norms, and that both kinds of considerations affect their choices. Thoughutility may trump norms in some cases, in others norms will outweigh immediate utility. Actors may even developcomplex ways of balancing norms and utility. What we need is a better understanding of norm rationality, and theways in which it interacts with utility calculations.

Rules and power

International actors deploy both arguments and material resources to bring others to their view. At oneextreme, actors with sufficient power resources can impose their preferred solutions on other actors, though theywill simultaneously offer arguments designed to show that their choices are also normatively justified. Butunipolar moments are exceptionally rare, and never absolute. Britain did not dictate the terms of the 19th-centuryPax Britannica, it negotiated those rules with the continental European powers (McKeown 1986). Similarly, theUnited States has found that its status as sole superpower in the post-Cold War era by no means allows it to imposeits preferences on the rest of the world. Pluralism, not unipolar hegemony, seems to be the usual condition of

2525

international relations. Under pluralism, no single actor can impose a solution, hence normative arguments aboutwhat course of action is justified are crucial in establishing consensus among multiple intereste