review 5

SmithPositivismandBeyond.pdf

Home >Political Science homework help >review 5

38 S TE VE S M ITH I Positivism and Beyond

Positivism and Beyond

STEVE SMITH

The author takes up the essential critique of positivism and empiricism-the impossibility of ''pure unvarnished perception" or "objective observation." He then examines alternatives to positivism: scientific realism, looking for "struc tures and things of the world that make science possible"; critical theory, drawn from the Frankfurt School- "that there can be no such thing as true empirical statements" apart from "knowledge-constitutive interest in control and predic tion"; feminist epistemology that identifies the dominance of male perspective to the virtual exclusion of feminine understanding; and postmodernism that observes how knowledge is "dependent on underlying power structures" and focuses on "the fundamental problem" of "how to match the subjective self, or knower, with an 'objective' or external world." The author concludes that "epis temological concerns are absolutely salient for contemporary international theory," but that these problems do "not mean accepting a less rigorous episte mological warrant" than positivism has offered.

For the last forty years the academic discipline of International Relations has been dominated by positivism. Positivism has involved a commit ment to a unified view of science, and the adop tion of methodologies of the natural sciences to explain the social world. The so-called "great debates" in the discipline's history, between ide alism and realism, traditionalism and behav iouralism, or between transnationalism and state-centrism, have not involved questions of epistemology. The discipline has tended to accept implicitly a rather simple and, crucially, an uncontested set of positivist assumptions which have fundamentally stifled debate over both what the world is like and how we might explain it. This is not true of those who worked either in the so-called "English school" or at the interface between international relations and political theory, because these writers never

bought into the positivist assumptions that dominated the discipline. But it has been the dominance of positivism that has accounted for both the character, and more importantly, the content of the central debates in international theory.

Viewed in this light, even the inter-para digm debate of the 1980s looks very narrow, because all three paradigms, (realism, pluralism and globalism/structuralism) were working under positivist assumptions. This helps explain just why they could be seen as three versions of one world, rather than three genuine alternative views of international relations. Similarly, the current "debate" between neo-realism and neo liberalism becomes much clearer when it is realised that both approaches are firmly posi tivist. My central claim here will be that posi tivism's importance has been not so much that it

Reprinted by permission from Steve Smith, Ken Booth, and Marysia Zalewski (eds.), International Theory: Positivism and Beyond (Cambridge, England: Cambridge University Press, 1996).

has given inten its empiricist e1 could be studie kinds of thing, ... The one approaches to i mon is a rejecti loosely descril here is meant ti ical Theorists feminist theori can be little de proaches reprt tional or main that this tradit been dominatf

It is becau temporary in writes of the "post-positivis indeed he add tional and post that it is the ti in internation 235-9), follov. tween idealisn 1940s, and be approaches to 1960s. In esser porary « critica sciously portn this makes it r what positivis involved in sp, tional theory.

The stake, much is clear theorists have approaches th, positivist (and ories"). One pc: been that of R< dential Addre Association in ate the rival r·

39 S T E V E S M I T H / Positivism and Beyond

has given international theory a method but that its empiricist epistemology has determined what could be studied because it has determined what kinds of things exist in international relations . . . . The one feature that the new "critical" approaches to international theory have in com mon is a rejection of the assumptions of what is loosely described as positivism. By "critical" here is meant the work of post-modernists, Crit ical Theorists (in the Frankfurt School sense), feminist theorists and post-structuralists. There can be little doubt both that these various ap proaches represent a massive attack on tradi tional or mainstream international theory, and that this traditional or mainstream theory has been dominated by positivist assumptions ....

It is because of these two features of con temporary international theory that Lapid writes of the current period constituting the "post-positivist era" of international theory; indeed he adds that the debate hetween tradi tional and post-positivist theory is so important that it is the third "discipline-defining debate" in international relations' history (1989, pp. 235-9), following the two earlier debates be tween idealism and realism in the 1930s and 1940s, and between traditional and scientific approaches to studying the discipline in the 1960s. In essence, then, because much contem porary "critical" international theory self-con sciously portrays itself as heing post-positivist, this makes it rather important to be clear as to what positivism means and what might be involved in speaking of a post-positivistinterna tional theory.

The stakes are high in such a debate. This much is clear from the way that mainstream theorists have responded to the rise of the approaches that Lapid groups together as post positivist (and which I above called "critical the ories"). One particularly important response has been that of Robert Keohane, who, in his Presi dential Address to the International Studies Association in I 988, spoke of the need to evalu ate the rival research paradigms of rationalist

(i.e. traditional neo-realism and neo-liberalism) and reflective (i.e. what I termed "critical") approaches in terms of their "testable theories", without which "they will remain on the margins of the field ... [since] ... it will be impossible to evaluate their research program" (1989, pp. 173-4). As is noted below, this form ofresponse reveals the dominance of positivism, since Keo hane issnes the challenge on grounds that are themselves positivist. Thus, positivism is pre cisely what is at issue in what Lapid calls the third debate because of its role in underpinning theory and, ultimately, serving as the criterion for judging between theory. Crucially, Keo hane' s central move is to propose that judge ment between rationalist and reflective theories takes place on criteria that not only favour ratio nalism, but, more importantly, are exactly the criteria that reflective accounts are attacking. But note that a failure to come up to the (posi tivist) mark will result in reflective work being confined to the margins. Important conse quences follow, then, from whether or not theo ries are positivist, and these consequences are not confined to questions as to what counts as knowledge but also involve the standing of the ories and theorists within academia. All of this makes it very important to be clear as to how positivism operates in international theory, and to show how it is seen not merely as one explicit alternative among many but rather as the implicit "gold standard" against which all ap proaches are evaluated.

But the stakes are also high because of the links between theory and practice. International theory underpins and informs international practice, even if there is a lengthy lag between the high-point of theories and their gradual absorp tion into political debate. Once established as common sense, theories become incredibly pow erful since they delineate not simply what can be known but also what it is sensible to talk about or suggest. Those who swim outside these safe waters risk more than simply the judgement that their theories are wrong; their entire ethical or

40 S T E V E S M I T H / Positivism and Beyond

moral stance may be ridiculed or seen as danger ous just because their theoretical assumptions are deemed unrealistic. Defining common sense is therefore the ultimate act of political power. In this sense what is at stake in debates about epis temology is very significant for political practice. Theories do not simply explain or predict, they tell us what possibilities exist for human action and intervention; they define not merely our ex planatory possibilities but also our ethical and practical horizons. In this Kantian light episte mology matters, and the stakes are far more con siderable than at first sight seem to be the case.

Having pointed out just how much is at stake in any discussion of epistemology in inter national theory, it is necessary, before examin ing the present debate about whether we can characterise the current situation in interna tional theory as one of "positivism and beyond", or simply "beyond positivism", to look at the history of positivism in the social sciences and then to examine various alternative epistemo logical positions. This will provide the intellec tual context for the debates in international theory, as well as serve as an introduction to the language and contexts of these wider philosophy of social science debates.

The History of Positivism

Positivism has a long history in the social sci ences. There are three main chronological vari ants of positivism in the history of the social sciences, with the third of these being the most relevant for international relations. What unites them is a strong commitment to a specific way of gaining knowledge about the world, as will be discussed in more detail below.

The first variant is that developed by Auguste Comte in the early nineteenth century (indeed it was Comte who coined the word "positivism", as well as "sociology"). Comte's aim was to develop a science of society, based on the methods of the natural sciences, namely observation. Its aim was to reveal the "evolu-

tionary causal laws" that explained observable phenomena. For Comte, positive science was a distinct third stage in the development of knowledge, which progressed first from theo logical to metaphysical knowledge and then to positivist knowledge. He saw the sciences as hierarchically arranged, with mathematics at the base and sociology at the top, and thought that each of the sciences passed through the three stages of knowledge. Crucially, he therefore thought that all sciences (including the sciences dealing with society) would eventually be uni fied methodologically. This view was enor mously important in the development of the social sciences throughout the nineteenth cen tury, fundamentally influencing writers as diverse as Marx and Engels, and Durkheim. It is an assumption that still dominates the discipline of International Relations insofar as scholars search for the same kinds of laws and regulari ties in the international world as they assume characterise the natural world.

The second variant is that of logical posi tivism ( or as it is sometimes called logical empiri cism), which emerged in the 1920s in what was known as the Vienna Circle. This variant domi nated English-speaking philosophy into the late 1960s, and is the starkest variant of the three sum marised here. The central shared proposition of the members of this circle was that science was the only true form of knowledge and that there was nothing that could be known outside of what conld be known scientifically. Hence, statements were only cognitively meaningful if they could be falsified or verified by experience. Moral and aes thetic statements were seen therefore as cogni tively meaningless since they could not be in principle verified or falsified by experience. Such statements are merely expressions of preferences or feelings and emotions, but they are not knowl edge. Thus, for instance, logical positivists rejected Comte's notion of causal laws explaining observable phenomena as metaphysical and therefore unscientific. In international relations, such a view would mean that it was simply not

possible to structure o "objective))I

The thi most influe, fifty years. I but moved, for what cm ist view ( cor edge shoul, physics. Ch, four main£ first, logidsn mation of s the canons< verificationL that are eith, (synthetic) , scientific; tl: tion, the vit between ob servations t finally, the I that establisl of discoveri ship betwee,

This vi, social scien 1950s and 1 ideas of th, Carnap, Na fledgling so, important " dally 1966 , extremely i, valved in eXJ event is exp! era! law. Ust tive argume:: postulated, ( fied, and (3 event deduc known as th and Hempe applied to 1

41 ST EVE S M I TH / Positivism and Beyond

possible to speak of unobservables such as the structure of the international system or the "objective)) laws of human nature.

The third variant is the one that has been most influential in the social sciences in the last fifty years. It emerged out of logical positivism, but moved away from its extremely stark criteria for what counts as knowledge and its reduction ist view (contra Comte) that all cognitive knowl edge should be based on the principles of physics. Christopher Lloyd has summarised its four main features as follows (1993, pp. 72-3): first, logicism,the view that the objective confir mation of scientific theory should conform to the canons of deductive logic; second, empirical verificationism,the idea that only statements that are either empirically verifiable or falsifiable (synthetic) or true by.definition (analytic) are scientific; third, theory and obsetvation distinc tion, the view that there is a strict separation between observations and theories, with ob servations being seen as theoretically neutral; finally, the Humean theory of causation, the idea that establishing a causal relationship is a matter of discovering the invariant temporal relation ship between observed events.

This view was extremely important in the social sciences, where the orthodoxy of the 1950s and 1960s was one of trying to apply the ideas of the main proponents of this view, Carnap, Nagel, Hempel and Popper, to the fledgling social science disciplines. Particularly important was the work of Carl Hempel ( espe cially 1966 and 1974) because he developed an extremely influential account of what is in volved in explaining an event. He argued that an event is explained by "covering" it under a gen eral law. Usually this takes the form of a deduc tive argument whereby (I) a general law is postulated, (2) antecedent conditions are speci fied, and (3) the explanation of the observed event deduced from (I) and (2). This model is known as the "deductive-nomological" model, and Hempel argued famously that it could be applied to the social sciences and to history

(1974). He also put forward an alternative, the "inductive-statistical" model, whereby statistical or probabilistic laws are established inductively and are used to show how a specific event is highly likely given the established law (1966, p. ll).

This third variant underpins much of the literature of international relations since the 1950s. Its features seem to me to have become somewhat detached from their philosophical roots as they have taken hold in international relations. In my judgement, positivism in inter national theory has had four underlying and very often deeply implicit assumptions, which deal with many of the points raised in Lloyd's four-point summary noted above but in an alto gether less philosophically conscious and explicit way. The first is a belief in the unity of science (including the social sciences). This was especially influential in international relations, and many would argue continues to be so. Accordingly, the slow pace of development of international theory could be explained by com paring it to the development of physics, which took centuries to create powerful theories. In short, the same methodologies and epistemolo gies apply to all realms of enquiry. In philosoph ical language this is known as naturalism, of which there are strong and weak versions; the strong view is that there is no fundamental dif ference between the social and the '(natural)) worlds; the weaker version is that despite differ ences between the two realms the methods of the natural sciences can still be used for the analysis of the social world. In international relations an example of the strong version might be the view that the international system is essentially the same as the systems of the natural world; the weaker version is illustrated by the claim that scientific methods can be used to understand the beliefs of decision-makers even though this does not mean that these beliefs fol low some (strong) laws of behaviour.

The second influential assumption is the view that there is a distinction between facts and

42 S TE V E S M I TH / Positivism and Beyond

values, and, moreover, that "facts" are theory neutral. This fitted particularly well into the rush towards behaviouralist quantification in the 1960s, and was a position very much to the fore in the USA during the debates over US involvement in Vietnam. In philosophical terms this is an objectivist position, one that sees pbjective knowledge of the world as possible despite the fact that observations may be sub jective. Thirdly, there has been a powerful belief in the existence of regularities in the social as well as the natural world. This, of course, licenses both the "deductive-nomological" and the "inductive-statistical" forms of covering law explanation. Again, in international relations terms, this kind of assumption lies at the heart of debates about polarity and stability, or about long cycles in world history. Finally, there has been a tremendous reliance on the belief that it is empirical validation or falsification that is the hallmark of "real" enquiry, a position we have already seen explicitly taken by Keohane; in philosophical language this is the adoption of an empiricist epistemology.

The impact of positivism on international theory has been surprisingly unreflective, with the canon of positivism significantly shaping the discipline since the 1950s. But there has been lit tle in the way of a discussion of what positivism actually means .... [T]he terms "positivism" and "empiricism" are conflated and confused in international relations . ... But it is certainly evi dent that in international relations positivism has tended to involve a commitment to a nat ural science methodology, fashioned on an early twentieth-century view of physics; that is to say a physics before the epistemologically revolu tionary development of quantum mechanics in the 1920s, which fundamentally altered the pre vailing view of the physical world as one which could be accurately observed. Accordingly, pos itivism in international relations, as in all the social sciences, has essentially been a method ological commitment) tied to an empiricist epistemology: together these result in a very re-

stricted range of permissible ontological claims. Thus whilst the terms ((positivism}) and "empiri cism" are used interchangeably, in both the phi losophy of social science and the philosophy of natural science, I think that it is absolutely nec essary to maintain a conceptual distinction between the two. However, this is not easy to do because whereas for the logical positivists ( or logical empiricists) positivism operated as an epistemological warrant about what kinds of knowledge claims might be made, in the third variant of Hempel, Popper et al. summarised above, positivism was much more than just a commitment to an empiricist epistemology. If it is difficult to differentiate between the use of the terms in the philosophy of natural and social science) in international relations and the other social sciences the two terms are virtually syn onymous.

An answer to the question "what does posi tivism mean in international relations?" can now be given. Positivism is a methodological view that combines naturalism (in either its strong ( ontological and methodological) or its weak (methodological) sense, and a belief in regulari ties. It is licensed by a strict empiricist epistemol ogy itself committed to an objectivism about the relationship between theory and evidence. Whilst the usage of terms such as "epistemol ogy'', "methodology)) and "ontology" may be various and inconsistent it is important that we separate "epistemology" conceptually from both "ontology" and "methodology", and then sepa rate positivism from empiricism. Thus, as to the latter, I do not accept the view that empiricism= positivism = epistemology + methodology; rather positivism is a methodological position reliant on an empiricist epistemology which grounds our knowledge of the world in justifica tion by ( ultimately brute) experience and there by licensing methodology and ontology in so far as they are empirically warranted.

As to the separation of epistemology, methodology and ontology, the three are indeed fundamentally interrelated. Methodology (why

STEV

use that method?) temology (answer: inates between ''ti range of what we "false"); whereas 01 world like and wha epistemological wa itself license a met should not be ta because I have com, via an analysis of er the epistemologica empiricism, althoug has underpinned m, national relations f< do maintain that ep it determines of wh moreover, it is not · undermine its imp fashionably the cas, philosophers and (1 ontology is prior to very tricky ground b that I see neither on prior to the other, I them as mutually an Thus, just as epistem, mining what can be ontology affects wha cally. In th.is light, pr as has certainly beer philosophy of know mology) and in pos prioritises ontology), such a move sets up , the two and implies t the other ....

Empiricism,as no the only grounds for jt rest ultimately on obs work of philosophers John Locke, its centre must be based on a ph that is to say the not about phenomena whi

r ' f ' S TE VE S M ITH / Positivismand Beyond ' ' I

ims. piri- phi- ,yof nee- tion odo (or

; an s of hlrd ised st a If it "the ,cial :her yn-

OSI-

LOW

1ew >ng eak 1ri- ,ol- the LCe.

ol- be we ,th }3-

the ,~ gy; on .ch :a- re- far

lY, ed hy

I f

f ! !

I f I

use that method?) needs the warrant of an epis temology ( answer: because this method discrim - inates between ('true" and '(false" within the range of what we could know to be "true" or "false"); whereas ontological claims (what is the world like and what is its furniture?) without an epistemological warrant is dogma and will not itself license a methodology .... My argument should not be taken as meaning that just because I have come to this point in this chapter via an analysis of empiricism this means I think the epistemological warrant is or should be empiricism, although this is the assumption that has underpinned much of the literature of inter national relations for the last forty years. But I do maintain that epistemology matters because it determines of what we can have knowledge; moreover, it is not possible to wish it away, or undermine its importance, by arguing, as is fashionably the case amongst post-modernist philosophers and (philosophical) realists, that ontology is prior to epistemology. All of this is very tricky ground but my main claim is simply that I see neither ontology nor epistemology as prior to the other, but instead see the two of them as mutually and inextricably interrelated. Thus, just as epistemology is important in deter mining what can be accepted ontologically, so ontology affects what we accept epistemologi cally. In this light, prioritising one or the other, as has certainly been the case in work on the philosophy of knowledge (prioritising episte mology) and in post-modernist work (which prioritises ontology), misses the point because such a move sets up a false distinction between the two and implies that one is separable from the other ....

Empiricism,as noted above, is the view that the only grounds for justified belief are those that rest ultimately on observation. Based upon the work of philosophers such as David Hume and John Locke, its central premise is that science must be based on a phenomenalist nominalism, that is to say the notion that only statements about phenomena which can be directly experi-

enced can count as knowledge and that any state ments that do not refer to independent atomised objects cannot be granted the status of justified knowledge (Kolakowski, 1972, pp. 11-17). Em piricists believe that science can rest on a bedrock of such.pure observation, and from this bedrock can be established, by induction, the entire sci entific structure. Put simply, basic beliefs, war ranted by direct perception, provide the basis for induction so that we can move away from a very narrow foundation for knowledge to much wider inductive generalisations. Of course, it is much more complicated than this .... The key point, however, is the simple one that if one grants that such a bedrock, however narrow, exists, then one can see how there might be an empirical foundation to knowledge.

But there are serious, and in my view insu perable, limitations to an empiricist epistemol ogy. I will simply outline three fundamental ones. The first is that the epistemological war rant offered by empiricism is very narrow indeed, if in the end it has to be based on direct observation. Such a warrant rules out any con sideration of(unobservable) things like social or international structures, or even social facts ( to use Durkheim's phrase which refers to those shared social concepts and understandings such as crime, which he believed should be treated "as things"). Thus, many philosophers point out that a strict empiricism actually allows us to know very little about only a very restricted amount of "reality". The second limitation is that, strictly speaking, empiricism does not allow us to talk about causes since these are unobservable. The best we can do, following Hume, is to talk about ('constant conjunctions", and therefore eschew notions of necessity. Causation is thus reduced to mere correlation, and our enquiry is there fore limited to that of prediction and cannot involve explanation. Hempel's covering law model, for example, can tell us what we might expect to happen, but not why it happens. Moreover, the logical structure of the covering law model allows us to make correct predictions

S T E V E S M I T H / Positivism and Beyond 44

from false premises. Overall, then, the kind of knowledge we can gain from an empiricism that refuses discussion of unobservables such as causes is very limited indeed.

But it is the final problem that really under mines empiricism most fundamentally: put sim ply it is the objective that the kind of pure unvarnished perception talked ofby empiricists is simply impossible. There can be no "objective" observation, nor any "brute experience". Obser vation and perception are always affected by prior theoretical and conceptual commitments. Empiricism, in other words, underestimates the amount of theory involved in perception or observation. To describe what we experience we have to use concepts, and these are not dictated by what we observe; they are either a priori in the mind, or they are the result of a prior theoretical language. The problems for empiricism arising out of its underestimation of the role of theory have been most clearly expressed in Quine's famous essay "Two Dogmas of Empiricism" (1961), in which he disputed both aspects of the empiricist view of the relationship between the ory and facts. First, he noted that there was no easy distinction between analytic and synthetic statements, and that even analytical statements (that is those deemed true by convention) are susceptible to revision by experience. For empiri cists this matters because they treat theory (or, in Hume's phrase the "relations of ideas") as "merely"a series of tautologies, or to use Martin Hollis's phrase (1994, p. 52) a "filing system"; no truths about the world can come from this filing system, such truths can only come from observa tion (or again to use Hume's phrase, from "mat ters of fact"). Quine's first objection, then, punctures the neat distinction between analytic and synthetic statements that is central to the empiricist view of the role of theories.

His second objection is equally fundamen tal. Whereas empiricism rests on pure observa tion, Quine argues that this is simply not possible, since theory is involved in all empirical observation. Even the most simple acts of"pure"

observation involve a web of belief which is both far removed and far more complex than the indi vidual act of observation. Our senses cannot give us access to "the truth" since there is no way of describing experience independently of its inter pretation. There are, therefore, no brute facts, no facts without interpretation, and interpretation always involves theory. Thus, whereas the em piricist view of knowledge clearly implies that if experience differs from our prior beliefs then we should change our beliefs (since these can never reveal any truths about the world), Quine shows that the web of belief in which individual acts of observation occur can give us reasons to think that it is our interpretation of observation that is mistaken. Any individual observation, therefore, can be revised or redefined. Theories define what the "facts" are, and it is always an alternative to reject the "facts" and thereby save the theory. Facts are always theory dependent, and not inde pendent as empiricism maintains. To quote Quine "it is misleading to speak of the empirical content of an individual statement. . . . Any st~tement can be held true come what may .... Conversely . .. no statement is immune to revi sion" (1961, p. 43).

These are absolutely fundamental objec tions to the empiricism that has dominated international theory, and, despite protests to the contrary by many contemporary empiricists, I think they apply to much of the work currently being undertaken in international relations ....

Contemporary Epistemological Debates

If, as I have claimed, positivism and its episte mological foundation, empiricism, are seriously flawed, if rationalism (empiricism's long-stand ing rival) is currently out of favour, and if prag matism seems to run into a series of objections from those who want to retain notions of a foundational truth, then maybe there are other positions that might be of interest to those inter national theorists attempting to locate their work in an epistemology other than empiricism.

Broadly, in the philoso ularly promi tional theory. hermeneutics furt School SE differentiate i "critical them post-positivis, point episten epistemology. approaches m entific realism with rationali standpoint a1 links to pragrr,

The centr. it makes sense rience; that is ing the structt make science I cist conceptior tic) is entirely theoretical cor are to be treat "facts". As a n piricist, with e objects of scie1 and account fc structures and distinguishes l the empirical: t mechanisms rn events, and tht ence. The prob looked at the ti ing the other t about what is how we knm Bhaskar puts i epistemic fallac ing the ontolo order, priority i claims to beir Bhaskar rejects

45 S TE V E S M IT H / Positivism and Beyond

h ,-

,f r ,o ,n

if ,e er vs of ik is ·e, at to y. e

te :al 1y

ri-

ed he , I tly

te sly .d ,g ,ns fa 1er er eir m.

Broadly speaking there are five alternatives in the philosophy of knowledge that look partic ularly promising for post-positivist interna tional theory. These are (a) scientific realism, (b) hermeneutics, (c) Critical Theory (in its Frank furt School sense, hence the capital C and T to differentiate it from the more generic notion of "critical theories" which is used to refer to all post-positivist approaches), (d) feminist stand point epistemology, and (e) post-modernist epistemology. Some of these are related to the approaches mentioned above: notably, both sci entific realism and Critical Theory share much with rationalism, and hermeneutics, feminist standpoint and post-modernism have close links to pragmatism.

The central claim of scientific realism is that it makes sense to talk of a world outside of expe rience; that is to say, it is interested in uncover ing the structures and things of the world that make science possible. Accordingly, the empiri cist conception of the role of theories (as heuris tic) is entirely wrong in that the existence of theoretical concepts such as electrons or classes are to be treated in the same way as so-called ((facts". As a result, its epistemology is nonem piricist, with epistemology being the transitive objects of science which we create to represent and account for intransitive objects such as the structures and furniture of the world. Bhaskar distinguishes between the real, the actual and the empirical: the first refers to what entities and mechanisms make up the world, the second to events, and the third to that which we experi ence. The problem with empiricism is that it has looked at the third of these as a way of explain ing the other two so that it reduces questions about what is (ontology) to questions about how we know what is (epistemology). As Bhaskar puts it: "It is important to avoid the epistemic fallacy ... [which] consists in confus ing the ontological order with the epistemic order, priority in being with priority in deciding claims to being" (1978, p. 250). Similarly, Bhaskar rejects rationalism since it too reduces

ontology to epistemology by its reliance on the role of theoretically necessary conceptual truths to make sense of the world. In contrast, realist science is an attempt to describe and explain the structures and processes of the world that exist independently of our perception of them. Indeed, for science to be possible, the world must be made up of real structures and pro cesses. Bhaskar fundamentally disputes the pri macy given to epistemology by both rationalism and empiricism, since they reduce ontology to epistemology. For Bhaskar, ontology is prinlary, but is nothing like as flexible as the pragmatists imply; in this sense scientific realism disputes the primacy of epistemology over ontology in rationalism and empiricism, yet sees pragma tism as mistaken in its claim ( or more accurately its implication) that what is true is only what is "good in the way of belief'.

Hermeneuticsdevelops out of the work of Dilthey, Husserl, Weber, Heidegger, Wittgen stein and Gadamer. Its central claim is antinatu ralist in that it does not see the social world as in any sense amenable to the kind of treatment that empiricism, and especially positivism, assumes. Developing out of textual analysis, hermeneu tics, as developed by Dilthey in the nineteenth century, starts from the premise that the analy sis of nature and the analysis of the mind are very different enterprises. For Dilthey, each required a very different form of analysis, contra positivism, and these forms of analysis are what we now call explaining and understanding. Hermeneutics can at first sigllt be seen as a con cern with how to understand a text or an actor, and the work of Collingwood (1946) and Skin ner (see especially, Tully, 1988), are prinlarily concerned with this essentially methodological hermeneutics. However, the much more radical work of particularly Heidegger (1962), and Gadamer (1994) raises ontological questions about the nature of being: what does it mean for us to interpret and understand the world? Cru cially, hermeneutics reverses the argument of traditional epistemology and instead of a being

S T E V E S M I T H / Positivism and Beyond 46

interpreting a world sees a being formed by tacit know-how which is prior to the interpretation of facts, events, or data. Individuals are caught up in a hermeneutic circle whereby we can only understand the world by our being caught up in a web of significance. Hermeneutics, in short, has ontological significance, which means that the traditional concerns of epistemology are inappropriate for understanding and making sense of our beliefs, since they posit the inter pretive or observing subject as in some way prior to questions about the nature of being. In its place, Gadamer stresses the importance of the embeddedness of all analysis in language and history; individuals analyse and act within what Gadamer refers to as an "horizon", by which he means their beliefs, preconceptions and situat edness and which both enables and constrains them. Crucially, for Gadamer, this embedded ness means that notions of truth and reason are themselves historically constituted, so that the kinds of claims about objective knowledge that have dominated epistemological discussions between rationalism and empiricism are funda mentally mistaken. What Gadamer proposes is an ontology of knowledge, reason and truth which shows how they are embedded in history rather than being above it. Epistemology, in its traditional sense discussed above, can therefore never be something prior, neutral, foundational or decisive, but instead has to be seen as sec ondary to ontology.

Critical Theory has a more recent history, emerging out of the work of the Frankfurt School in the inter-war years. Its most influen tial thinker has been Jurgen Habermas. The main implication for epistemology has derived from his work on a broader conception of rea son than the instrumental view which domi nates Western science and his development of a non-positivist jnethodology for the social sci ences. In his book Knowledge and Human Inter ests (1987, first published in 1968), Habermas puts forward the view that there are three types of knowledge: empirical-analytical (the natural

sciences), historical-hermeneutic ( concerned with meaning and understanding), and critical sciences ( concerned with emancipation). Haber mas claims that each of these types of knowledge has its own set of "cognitive interests", respec tively, those of a technical interest in control and prediction, a practical interest in understanding, and an emancipatory interest in enhancing free dom. The epistemological implication of this transcendental claim is that there can be no such thing as true empirical statements, for example in the realm of the natural sciences, independent of the knowledge-constitutive interest in control and prediction. Since the late I 960s, Habermas has moved away from this rather restricted notion of knowledge-constitutive interests towards the development of what he terms a theory of communicative action {1984; 1987), in which he is concerned with developing an epistemology based on the notion of universal pragmatics or discourse ethics, whereby he sees knowledge emerging out of a consensus theory of truth. Central to this is his idea of an "ideal speech situation", which he sees as implicit in the act of communication, and as rationally entailing ethical and normative commitments. The "ideal speech situation" is based upon the notion that acts of communication necessarily presuppose four things: that statements are comprehensible, true, right and sincere. It is not that Habermas thinks that the ideal speech situ ation is something that is commonly found in communicative actions, only that it is presumed by the very acts themselves. Habermas believes that we could in principle reach a consensus on the validity of each of these four claims, and that this consensus would be achievable if we envis aged a situation in which power and distortion were removed from communication so that the "force of the better argument prevails" (Outh waite, 1994, p. 40). Thus, his epistemological position is one which seeks to avoid the simple objectivism of positivism whilst at the same time stopping short of embracing the kind of rela tivism implicit in traditional hermeneutics.

STE

He proposes ti proceed as do tl must see action tor involved; b, not mean that t cise these persp, cal grounds. Cn the existence of ments between work as a dire, enlightenment F both a major s modernists and those who want to emancipation

Feminist we and only one var ... Sandra Hardi strands in fem feminist empiri< feminist post- me mately rests on tl ogy discussed ab, do with the work discussed in the of feminist star expressed in the , who set out to ". logical conseque: lives differ strw (1983, p. 284). Western, and wh knowledge has I partial and whi women. Develop master-slave rela argue that this 1 epistemic advan1 better knowledg, than can the men point knowledge emancipatory pot ated by feminist , feminist standpoi enormous since i

S T EVE S M I TH I Positivism and Beyond

ed cal :::r- .ge !C

od 1g, '.e ns ch •le nt ol as ,d ts a ), n al ,s

y ,l n y ;,

y e ,t

He proposes that the social sciences cannot proceed as do the natural sciences, and instead must see action from the perspective of the ac tor involved; bnt he maintains that this does not mean that the social sciences cannot criti cise these perspectives on theoretical and ethi cal grounds. Crucially, Habermas' s emphasis on the existence of foundations for making judge ments between knowledge claims places his work as a direct descendant of the Kantian enlightenment project, a position that has been both a major source of criticism from post modernists and yet a great source of strength to those who want to link foundational knowledge to emancipation.

Feminist work on epistemology is diverse, and only one variant of it will be dealt with here. ... Sandia Harding (1986) has noted three main strands in feminist work on epistemology: feminist empiricism, feminist standpoint and feminist post-modernism. The first of these ulti mately rests on the kind of empiricist epistemol ogy discussed above, while the third has much to do with the work on post-modern epistemology discussed in the next section. The central claim of feminist ·standpoint epistemology is well expressed in the work of Nancy Hartsock {1983 ), who set out to "explore some of the epistemo logical consequences of claiming that women's lives differ structurally from those of men" (1983, p. 284). For Hartsock, the male (and Western, and white) domination of science and knowledge has produced knowledge which is paitial and which excludes or marginalises women. Developing the Hegelian notion of the master-slave relationship, standpoint feminists argue that this marginality can be turned to epistemic advantage since women can have a better knowledge of male-dominated science than can the men involved; thus feminist stand point knowledge will have both explanatory and emancipatory potential for greater than that cre ated by feminist empiricists. The implication of feminist standpoint analysis for epistemology is enormous since it requires us to challenge the

traditional epistemological assumption that the identity of the knower was irrelevant in the process of knowing, an assumption found in both rationalism and empiricism. Indeed, not only is a feminist standpoint epistemology cor rosive of such a claim, but so is feminist empiri cism, since each is claiming that there can be no "master" narrative and that the situatedness of the mind is irrelevant. Moreover, feminist work on epistemology has made it abundantly clear that the knowing mind of traditional epistemol ogy is axiomatically a male mind. This has rad ical consequences) since it requires us to abandon the idea of a disinterested and de tached knowing subject, a move whicli is equally problematic for both rationalism and empiri cism; in its place feminists propose the idea that knowledge is a social activity. In this light, the knowledge produced by this process cannot but fail to be influenced by the social location of those who construct it. The fundamental ques tion for epistemology therefore becomes "whose knowledge?"

Post-modem work on epistemology is extra ordinarily diverse, and defies easy summary ... because its central tenet is one which seeks noth ing less than the overthrow of virtually all pre ceding positions on epistemology. There are extensive debates over what is post-modernism and how it differs from post-structuralism, let alone quite what a post-modern epistemology looks like. It is a genre of work that has been attacked and dismissed, usually by people who have not bothered to engage with the complex issues involved (and maybe without reading the texts either!) ....

The central implication of Foucault's work for epistemology comes from his concern with the historically specific conditions in which knowledge is generated. In his book The Order of Things (1970), he undertook "an archaeology of the human sciences" to show how the human sciences were not "natural" modes of enquiry but rather were made possible by an underlying structure of thought. His work on prisons

S T EVE S M I TH I Positivism and Beyond 48

(1977), on the rise of "modern" medicine (1975) and on madness (1967) was concerned not so much with the content of the knowledge generated in these areas but rather on the rela tionship between knowledge and practice-that is to say on the relationship between power and knowledge-a relationship so intertwined that he referred to it as power/knowledge; each is always involved in the operation of the other. His later work on genealogy ( 1986) sought to show how specific academic" discourses" emerged not as a neutral result of scholarly enquiry, but as the direct consequence of power relations. In short, power is implicated in all knowledge sys tems, such that notions like reason or truth are the products of specific historical circumstances. The concept of truth has to alter accordingly, since it can no longer refer to an underlying or foundational notion of truth, but rather to the idea of multiple truths. Epistemology is there fore decidedly not the centrepiece of philosoph ical enquiry, but is instead dependent on underlying power structures. This is a very rad ical move away from empiricism, rationalism or even pragmatism; it was also a source of consid erable dispute between Foucault and Habermas. Finally, Foucault saw the concept of truth not as an empirically valid concept but rather as a tool for resisting power. Accordingly, the central fea ture of epistemology, a concern with the criteria for determining the truth, is replaced by a much more practical notion of truth-as-tool.

Derrida's work can be seen as an attack on one central assumption of traditional work on epistemology, what he calls the "metaphysics of presence". His claim is anti-essentialist and anti foundationalist in that he refuses to see the knower as a given, and instead as merely one more construction of language and culture. Accordingly, he sees as flawed the central tenet of empiricist and rationalist epistemology, namely that the funda~ental problem of epistemology is how to match the subjective self, or knower, with an "objective" or external world. His work on deconstruction forces us to examine the "reason-

ableness of reason", since he locates it within a specific cultural and historical setting of thinking and writing. His meticulous examinations of specific pieces of writing (see as examples, 1976, 1978, 1982) show how arbitrary and particular are the logocentric structures which mark lan guage and thought; it is this which removes the foundations from rationalism and empiricism, since it completely subverts the idea of a prior presence (the metaphysics of presence). Instead, the knower is always caught up in a language and mode of thinking which, far from interpreting a world, instead constructs it ....

Positivism in International Theory

Having mapped out the terrain of the debate in the social sciences, this section will look at posi tivism in international theory, and the next will speculate about the possibilities open for devel oping a post-positivist international theory. The first question is whether international theory is currently dominated by positivism, or whether instead it has in some way moved beyond posi tivism.

At the outset I want to stress just how ill defined "positivism" is in international theory. Basically there are three common ways of using the term. The first treats positivism as essentially the same thing as empiricism, which is to say that each is seen as an epistemology.An episte mology deals with how it is that we might know something about the world. A seconduses posi tivism in a methodological way, by which is meant a set of rules for the actual practice of sci ence or study. Thirdly, positivism is often equated with behaviouralism,by which is meant a very restrictive reliance on quantitative data, and a disregard for what goes on inside actors' heads, as the basis for knowledge claims. Usu ally, it is the first of these uses that has charac terised international relations, but the overlap of usages has been especially marked, and it is quite common to be able to discern more than one of the above usages in any given enquiry.

Within p series of diffe Martin Hollis seen by philc piricism ( an (the view tha are one and which has bot] consequences: methodologic, cal assumptic Andrew Sayer itivism and err chooses not tc realist rejecti01 (1992, p. 7).

The prob] theorists have 1 and easy wayf philosophical, tantly, that th relations rese, rested implici Thus, while m, were aware of and indeed in cates of its st1 1976, which ir tetchy defence an unthinking Kuhnian norr debate or them sciousness. Thi when the inter cal, methodolo of positivism a1 unaware of th, over, there has sion as to wh might look like either that "wf positivism is nc it once was, anc comes the ina behaviouralisrr

49 S T E V E S M I T H / Positivism and Beyond

Within philosophy the term involves a series of different kinds of commitments. As Martin Hollis shows, ... positivism is usually seen by philosophers as entailing both em piricism (an epistemology) and naturalism (the view that the natural and social worlds are one and the same kind of thing, a view which has both ontological and methodological consequences); pos1tmsm thereby entails methodological, ontological and epistemologi cal assumptions and commitments. Indeed Andrew Sayer has remarked that the terms pos itivism and empiricism are so contested that he chooses not to use them in his (philosophical) realist rejection of these two associated positions (1992, p. 7).

The problem is not just that international theorists have tended to use the term in very free and easy ways, unaware of the depths of the philosophical waters involved, but, more impor tantly, that the vast majority of international relations research over the last 30 years has rested implicitly on positivist assumptions. Thus, while many of the leading behaviouralists were aware of their commitment to positivism and indeed in many cases were powerful advo cates of its strengths (see Hoole and Zinnes, 1976, which includes a particularly robust and tetchy defence by Singer), many others adopted an unthinking positivism, and worked within a Kuhnian normal science thereby foreclosing debate or theoretical and philosophical self-con sciousness. This has been especially problematic when the interrelationship of the epistemologi cal, methodological and ontological entailments of positivism are ignored, and when theorists are unaware of the consequences of these. More over, there has been little in the way of discus sion as to what an alternative to positivism might look like. It is often erroneously assumed either that "we are all positivists now" or that positivism is now much more sophisticated than it once was, and that this '<neo-positivism" over comes the inadequacies of the positivism-as behaviouralism that characterised international

theory in the 1950s and 1960s (such a position is well illustrated by Hermann, Kegley and Rose nau, 1987, pp. 18-22). In short, mainstream international theory has never really bothered to examine its positivist assumptions, nor what alternatives are available. Keohane's response to one such move to develop an alternative illus trates this poverty of imagination.

However, there have been notable excep tions to this silence about the role of positivism, but at first sight these seem to have been rela tively ineffective in altering the orientation of the mainstream discipline. But on reflection such a view may simply reflect a massive (and very common) underestimation of the hold that positivism has had over international relations. One seemingly important reaction came in the so-called "great debate" of the mid-1960s, when Hedley Bull attacked the "scientists" for their methodological assumptions ( see Knorr and Rosenau, 1969, for a collection of the main papers dealing with this "debate"); in place of "science" Bull proposed more traditional and historical analysis. Now the problem with the debate was that it ignored both epistemology and ontology; it was instead a very narrow dis pute about what methods were appropriate for the study of international relations. A second celebrated debate came in the late I 960s with the induction/deduction dispute between Oran Young and Bruce Russett (see Young, 1969 and Russett, 1969); this was essentially a debate about whether enquiry was led by observation or by theory. But again this, although it seemed more important at the time, was really only another debate about the appropriateness of dif ferent methods. A third, and much more sub stantive, response was the critique of Charles Reynolds (1973, 1992), which noted the weak nesses of empiricism and in its place proposed a form of Collingwoodian history whereby the

. task of the theorist was to understand events as perceived by the individuals involved in them. But this too was largely ignored in the main stream literature. The reason seems to me to be

S T E V E S M I T H / Positivism and Beyond 50

that the mainstream was simply too wedded to positivism, and crucially, wedded in an implicit rather than an explicit way. On an anecdotal level, I recall many discussions with leading spe cialists in my own research area at the time (for eign policy analysis) who denied outright that they were positivists, associating it with a crude form of behaviourism which had failed 'to pro duce the goods'. This was very evident in the mid-1970s when the comparative foreign policy movement lost its impetus (see Kegley, 1980, Rosenau, 1976 and Smith, 1986). The response was not to abandon positivism, since this seemed untouched by the problems facing this sub-field of the discipline, but rather to reject the excessive reliance on quantitative data char acteristic of behaviouralism, and also to ques tion the belief in an inductive route to general theory. But, of course, neither of these moves involved a rejection of positivism, only one component of it, and a rather extreme one at that.

More recently, a series of post-modernist writers (most notably Jim George I 988, pp. 67- 70 and 74-85; 1994, pp. 41-68) and Richard Ashley (1984) have written about the limita tions of positivism. But their work has been largely ignored by many traditionalists precisely because they write from theoretical positions that traditionalists do not accept as capable of providing "real" or "proper" knowledge. Ken neth Waltz's ( 1986) response to Ashley's (1984) critique, John Mearsheimer's quite amazingly confused comments on "critical theory" (1995), Kai Holsti's (1993) worries that such interven tions do not lead to "progress», Roy Jones's (1994) reaction to Rob Walker'swork, and Kai Holsti's (I 989) and Tom Biersteker's (1989) responses to Lapid's article on post-positivism discussed above, are very clear examples of the reaction to post-structural or post-modern attacks on positivism. Rare indeed is the attempt to engage with these criticisms .... Yet again the reaction is essentially that writers such as Ashley and George are not doing "proper" international

relations. Part of the reason for this quite consis tent rejection of attacks on positivism in interna tional relations is, as stated above, that many international relations specialists tend to think of it in a very narrow sense; in this light, the cri tiques mentioned above "do not apply to them". But there is more to it than that, and in my judgement the explanation lies in the epistemo logical and ontological features of positivism. Thus the methodological aspects of positivism can be rejected as unduly quantitative or be haviouralist, but doing so does not mean that positivism's epistemological basis, and there by its range of possible statements about what exists (its ontological realm), are also rejected. Positivism-as-methodology may be rejected but all too often positivism-as-epistemology contin ues to play the same role as before.

My basic argument, then, is that positivism is usually equated with a specific epistemological position, but almost always involved method ological commitments; together these result in a very limited range of possible ontological claims. It is precisely this linkage that lies at the heart ofKeohane's view of what is at stake in the rationalist/reflective debate; and it is this link age that explains just why his proposed test of the alternatives is so problematic. Keohane accepts as settled exactly the kind of epistemo logical and methodological questions that non positivists want to problematise, and therefore his proposed test results in only a small range of ontological statements being deemed accept able. Positivism is therefore used in a variety of ways, lacks a common definition, conflates and confuses very distinct philosophical concepts (so that positivism is sometimes used to refer to an epistemology and at others to refer to a methodology), is implicitly and explicitly pow erful, and through the 1980s has come to be increasingly criticised. All of which begs the question of whether we can move beyond posi tivism, and what might such theories look like; or, perhaps whether there can now only be post positivist theory?

Positivism a

At the outset no such thin1 post-positivi: this before , there were ti positivist inti atory and < foundational called traditi whelmingly , positivist th, post-positivi historical so, as foundatio1 feminist theo fication can _ account of tl tivist approa epistemologi

My clai1 international cause it is p< positivist wo use Keoharn counts are w epistemologi there is no 1 alternative. R the fact that i the realist co1 to be largely · temology, if epistemologi, of approache the five altern marised abo~ tivist approac traditional ir works within tional theory distinctly diff

ToillustI my recent di·

y ,f

S T E V E S M I T H I Positivism and Beyond

Positivism and Beyond

At the outset I want to claim that there is really no such thing as a post-positivist approach, only post-positivist approaches. I have written about this before (Smith, 1995), pointing out that there were two main debates concerning post positivist international theory, between explan atory and constitutive theory and between foundational and anti-foundational theory. I called traditional international theory as over whelmingly explanatory in character, and post positivist theory as constitutive, and among post-positivist approaches saw critical theory, historical sociology, and some feminist theory as foundational and post-modernism and some feminist theory as anti-foundational. This classi fication can now be improved on so as to take account of the fact that the various post-posi tivist approaciles operate within very different epistemological positions.

My claim is that whilst the vast bulk of international theory is indeed explanatory be cause it is positivist, and though a lot of post positivist work is constitutive ( or reflective to use Keohane's term), the post-positivist ac counts are working with distinctively different epistemologies. It is this which explains why there is no prospect of them constituting an alternative. Rather, historical sociology, despite the fact that it has been central in undermining the realist concept of the state, seems to me now to be largely working within an empiricist epis temology, if not an outright positivism. The epistemological positions of the other three sets of approaches seem to fit within at least three of the five alternative epistemological stances sum marised above. Thus, althougll the post-posi tivist approaches are united in an opposition to traditional international theory, one of them works within the same epistemology as tradi tional theory, and the others are operating in distinctly different epistemologies to eacil other.

To illustrate these differences I will broaden my recent division of approaches into founda-

tionalist and anti-foupdationalist. In its place I will use four criteria, those of objectivism, em piricism, naturalism and behaviouralism, to assess the precise commitments of the alterna tive epistemologies. Following Martin Hollis's definitions of these ... objectivism can be de fined as referring to the view that objective knowledge of the world is possible; naturalism as meaning that there is a single scientific method whim can analyse both the "natural" and the social worlds; empiricismas involving the claim that knowledge has finally to be justified by experience; and behaviouralismas meaning that we do not need to worry about what actors think they are doing to explain their behaviour ....

Although the five views have been influen tial in developing post-positivist international theory) it is interesting to note that some have been mucil more influential than others. Scien tific realism has been used very little (see Wendt, 1987; Dessler, 1989), which is rather surprising given that it has enormous potential for those who wish to construct an account of the influ ence of unobservable structures (such as the international system or the society of states?) on behaviour. Similarly, hermeneutics, as distinct from work on perceptions, has been little used, despite its obvious implications for the study of decision-making (for one inlportant exception see Shapcott, 1994; see also the essays in Little and Smith, 1988). Feminist standpoint episte mology has been supported by Keohane (1991) as offering additional insig!,ts ( to those of the mainstream), and used by Tickner (1992) in her book on international relations, but, as Zalewski (1993) argues, has otherwise been largely ig nored in international theory. The other two epistemological perspectives, Critical Theory and post-modernism have been far more widely in corporated into international theory (for paradigmatic examples see, respectively, Cox, 1981, 1987; Hoffman, 1987; Linklater, 1990, 1992; and George, 1994; Ashley, 1987; Walker, 1993; Der Derian, 1987; Campbell, 1992). There is evidently still a large amount of epistemological

ST EVE S M IT H I Positivism and Beyond 52

space for the development of post-positivist approaches in international theory.

In conclusion, my task has been to show that positivism has had an enormous influence in international theory primarily because it has critically influenced what the discipline could talk about; in this sense it matters because its epistemological assumptions have had enor mous ontological consequences. International theory now has a number of post-positivist approaches, which are opening up space not merely for other ways of thinking about interna tional relations but also for other international realities. But there is no hope of a (single) post positivist approach because some very distinctly different and mutually exclusive epistemologi cal positions underlie post-positivist interna tional theories. My aim has been to say something about both the nature of positivism and its traditional alternatives, and about the kinds of epistemological positions open to those who want to move beyond positiviSt interna tional theory. This does not mean, however, that anything goes epistemologically, or that one can adopt a pick and mix approach to these alterna tive epistemological positions. Indeed, my main argument is that epistemological concerns are absolutely salient for contemporary interna tional theory and to claim that the move away from positivism does not mean accepting a less rigorous epistemological warrant for theory. In my judgement, the weaknesses of positivism are so fundamental that the positivist project cannot be resurrected. At the same time, positivism's dominance of the discipline has been, and con tinues to be, so great that it has come to be seen as almost common sense. But more important still has been positivism's role in determining, in the name of science, just what counts as the subject matter of international relations. Its epistemology has had enormous ontological effects, and these have affected not only the study but also the practice of international rela tions. In positivism's place, international theory needs to develop strong post-positivist theories

based on a variety of epistemologies because much more than epistemology is at stake.

References Ashley, Richard (!984), 'The Poverty of Neorealism',

International Organization, 38 (2), pp. 225-86. (1987), 'The Geopolitics of Geopolitical Space:

Toward a Critical Social Theory of International Politics', Alternatives, 12 ( 4), pp. 403-34.

Bhaskar, Roy (1978), A Realist Theory of Science (Brighton: Harvester).

(1979), The Possibility of Naturalism: A Philosophical Critique of the Contemporary Human Sciences (Brighton: Harvester).

(1989), Reclaiming Reality: A Critical Introduction to Contemporary Philosophy (London: Verso).

Biersteker, Thomas (1989), 'Critical Reflections on Post Positivism in International Relations', International Studies Quarterly, 33 (3), pp. 263-8.

Campbell, David (1992), Writing Security: United States Foreign Policy and the Politics of Identity (Manches ter: Manchester University Press).

Collingwood,R. G. (1946), The Idea of History (London: Oxford University Press).

Cox, Robert (1981), 'Social Forces, States and World Orders: Beyond International Relations Theory', Millennium: Journal of International Studies, 10 (2), pp. 126-55.

(1987), Production, Power and World Order: Social Forces in the Making of History (New York: Colum bia University Press).

Der Derian, Jaines (1987), On Diplomacy: A Genealogy of Western Estrangement (Oxford: Blackwell).

Derrida, Jacques (1976), Of Grammatology (translated with an introduction by Gayatri Spivak) (Baltimore MD: Johns Hopkins University Press).

(1978), Writing and Difference (translated with an introduction by Alan Bass) (London: Routledge).

(1982), Margins of Philosophy (translated with addi tional notes by Alan Bass) (Brighton: Harvester).

Dessler, David (1989), 'What's At Stake in the Agent Structure Debate?', International Organization, 43 (3), pp. 441-73.

Foucault, Michel (1967), Madness and Civilisation: A History of Insanity in the Age of Reason (first pub lished 1961) (London: Tavistock Publications).

(1970), The Order of Things: An Archaeology of the Human Sciences (first published 1966) (New York: Random House).

(1975), The Birth of the Clinic: An Archaeology of Med ical Perception (first published 1963) (New York: Vintage Books).

s '.

(!977), Discif (first publ Lane).

(1986), 'Neit2 now (ed.), Peregrine I

Gadamer, Hans revised edi1

George, Jim (1 S tions and Knowledge pline', in R lnternation tional Uni,

(1994), Disc, (Re)Introd, der, CO: LJ

George, Jim an of Dissent Critical Sc tions', Inte 269-93.

Habermas, Jilrgt ests (first p1

(1984 [1991]: Vol. 1: Ret (Cambridg

(1987 [1987]: Vol. 2: The bridge: Pol

Harding, Sandn nism (Milt

(1991), Who, Keynes: Oi;

Harding, SandH ogy(Miltor

Hartsock, Nanc oping the ( torical Mat, Hintilcka ( t spectives on andPhiloso de!), pp. 28

Heidegger, Mar Blackwell).

Hempel, Carl ( (Englewoo1

(1974). 'Reas Explanatim ophyofHis pp. 90--!05

Hermann, Char] (eds.)(198: Policy (Lon

S T E V E S M I T H / Positivism and Beyond

( 1977), Discipline and Punish: The Birth of the Prison (first published 1975) (Harmondsworth: Allen Lane).

(1986), 'Neitzsche, Genealogy, History', in Paul Rabi now (ed.), The Foucault Reader (Harmondsworth: Peregrine Books), pp. 76--100.

Gadamer, Hans-Georg (1994), Truth and Method, 2nd revised edition (New York: Continuum Books).

George, Jim (1988), 'The Study of International Rela tions and the Positivist/Empiricist Theory of Knowledge: Implications for the Australian Disci pline', in Richard Higgott (ed.), New Directions in International Relations (Canberra: Australian Na tional University), pp. 65-142.

(1994), Discourses of Global Politics: A Critical (Re)Introduction to International Relations (Boul der, CO: Lynne Reinner).

George, Jim and Campbell, David (1990), 'Patterns of Dissent and the Celebration of Difference: Critical Social Theory and International Rela tions', International Studies Quarterly, 34 (3), pp. 269-93.

Habermas, Jtirgen (1987), Knowledge and Human Inter ests(first published 1968) (Cambridge: Polity).

(1984 [1991)), The Theory of Communicative Action Vol. 1: Reason and the Rationalisation of Society (Cambridge: Polity Press).

(1987 [1987]), The Theory of Communicative A.ction Vol. 2: The Critique of Functionalist Reason (Cam bridge, Polity).

Harding, Sandra (1986), The Science Question in Femi nism (Milton Keynes: Open University Press).

(1991), Whose Science? ¾'hose Knowledge? (Milton Keynes: Open University Press).

Harding, Sandra (ed.) (1987), Feminism and Methodol ogy (Milton Keynes: Open University Press).

Hartsock, Nancy (1983), 'Feminist Standpoint: Devel oping the Ground for a Specifically Feminist His torical Materialism', in Sandra Harding and Merrill Hintikka (eds.), Discovering ReaHty: Feminist Per spectives on Epistemology, Metaphysics, Methodology and Philosophy of Science (Dordrecht, Holland: Rei del), pp. 283-310.

Heidegger, Martin (1962), Being And Time (Oxford: Blackwell).

Hempel, Carl (1966), Philosophy of Natural Science (Englewood Cliffs, NJ: Prentice-Hall).

(1974), 'Reasons and Covering Laws in Historical Explanation', in Patrick Gardiner (ed.), The Philos ophy of History (Oxford: Oxford University Press), pp. 90-105.

Hermann, Charles, Kegley, Charles and Rosenau, James (eds.) (1987), New Directions in the Study of Foreign Policy (London: Allen & Unwin).

Hoffman, Mark (1987), 'Critical Theory, and the Inter Paradigm Debate', Millennium: Journal of Interna tional Studies, 16 (2), pp. 231--49.

Hollis, Martin (1994), The Philosophy of Social Science: An Introduction (Cambridge: Cambridge Univer sity Press).

Hollis, Martin and Smith, Steve (1990), Explaining and Understanding International Relations (Oxford: Clarendon Press).

(1991), 'Beware of Gurus: Structure and Action in International Relations', Review of International Studies, 17 (4), pp. 394--410.

(1992), 'Structure and Action: Further Comment', Review of International Studies, 18 (2), pp. 187-8.

(1994), 'Two Stories About Structure and Agency', Review of International Studies, 20 (3 ), pp. 241-51.

(forthcoming, 1996), 'Why Epistemology Matters in International Theory', Review of International Stud ies,22 (!).

Holsti, Kai (1989), 'Mirror, Mirror on the Wall Which are the Fairest Theories of All?', International Stud ies Quarterly, 33 (3), pp. 255--61.

(1993), 'International Relations at the End of the Mil lennium', Review of International Studies, 19 (4), pp. 401-8.

Hoole, Francis and Zinnes, Dina (eds.) (1976), Quanti tative International Politics: An Appraisal (New York: Praeger).

Jones, Roy (1994), 'The Responsibility to Educate', Review of International Studies, 20 (3), pp. 299- 311.

Kegley, Charles (1980), The Comparative Study of For eign Polio/: Paradigm Lost? (Columbia, SC: Univer sity of South Carolina).

Keohane, Robert (1989), International Institutions and State Power: Essays in International Relations Theory (Boulder,CO: Westview).

(1991), 'International Relations Theory: Contribu tions of a Feminist Standpoint', in Rebecca Grant and Kathleen Newland (eds.), Gender and Interna tional Relations (Milton Keynes: Open University Press), pp. 41-50.

King, Gary, Keohane, Robert and Verba, Sidney ( 1994), Designing Social Enquiry: Scientific Inference in Qualitative Research (Princeton, NJ: Princeton University Press).

Knorr, Klaus and Rosenau, James (eds.) (1969), Con tending Approaches to International Politics (Prince ton, NJ: Princeton University Press).

Kolakowski, Leszek (1972), Positivist Philosophy (Har mondsworth: Penguin Books).

Kuhn, Thomas (1970), The Structure of Scientific Revolu tions, 2nd edition (Chicago: University of Chicago Press).

54 S T E V E S M I T H / Positivism and Beyond

Lapid, Yosef (1989), 'The Third Debate: On the Pros pects of International Theory in a Post-Positivist Era', International Studies Quarterly, 33 (3), pp. 23:>-54.

Linklater, Andrew ( 1990 ), Beyond Realism and Marxism: Critical Theory and International Relations (Bas ingstoke: Macmillan).

(1992), 'The Question of the Next Stage in Interna tional Relations Theory: A Critical-Theoretical Point of View', Millennium: Journal of International Studies,21 (!), pp. 77-98.

Lloyd, Christopher (1993), The Structures of History (Oxford: Blackwell).

Mearsheimer, John (1995), 'The False Promise of Inter national Institutions', International Security, 19 ( 3 ), pp. 5--49.

Outhwaite, William (1987), New Philosophies of Social Science: Realism, Hermeneutics and Critical Theory (Basingstoke: Macmillan).

(1994), Habermas: A Critical Introduction (Cam bridge: Polity).

Quine, W. V. 0. (1961), 'Two Dogmas of Empiri cism', in his From a Logical Point of View, 2nd edi tion (Cambridge, :MA: Harvard University Press), pp. 20-46.

Reynolds, Charles (1973), Theory and Explanation in International Politics (Oxford: Martin Robertson).

(1992), The World of States: An Introduction to Expla nation and Theory (Aldershot: Edward Elgar Pub lishing).

Rorty, Richard (1980), Philosophy and the Mirror of Nature (Oxford: Blackwell).

(1989), Contingency, Irony, and Solidarity (Cam bridge: Cambridge University Press).

(1991), Objectivity, Relativism and Truth: Philosophi cal Papers, VoL 1 (Cambridge: Cambridge Univer sity Press).

( 1993 ), 'Human Rights, Rationality, and Sentimental ity', in Stephen Shute and Susan Hurley (eds.), On Human Rights: The Oxford Amnesty Lectures 1993 (New York: Basic Books), pp. 111-34.

Rosenau, James (1976), 'Puzzlement in Foreign Policy', Jerusalem Journal of International Relations, 1 (1), pp. 1-10.

Russett, Bruce (1969), 'The Young Science of Interna tional Politics', World Politics, 22 (1), pp. 87-94.

Sayer, Andrew (1992), Method in Social Science: A Real ist Approach, 2nd edition (London: Routledge).

Smith, Steve (1986), 'Theories of Foreign Policy: An His torical Overview', Review of International Studies, 12 (!), pp. 13-29.

(1994a), 'Rearranging the Deckchairs on the Ship Called Modernity: Rosenberg, Epistemology and

Emancipation', Millennium: Journal of Interna tional Studies, 23 (2), pp. 395-405.

(1994b), 'Foreign Policy Theory and the New Europe', in Walter Carlsnaes and Steve Smith (eds.), European Foreign Policy: The EC and Chang ing Perspectives in Europe (London: Sage Publica tions), pp. 1-20.

(1995), 'The Self-Images of a Discipline: A Geneal ogy of International Relations Theory', in Ken Booth and Steve Smith (eds.), International Relations Theory Today (Cambridge: Polity), pp. 1-37.

Stanley, Liz and Wise, Sue (1993), BreakingOut Again: Feminist Ontology and Epistemology (London: Routledge).

Sylvester, Christine (1994), Feminist Theory and Inter national Relations in a Postmodern Era (Cambridge: Cambridge University Press).

Tickner, J. Ann ( 1992 ), Gender in International Relations: Feminist Perspectives on Achieving Global Security (New York: Columbia University Press).

Tully, James (ed.) (1988), Meaning and Context: Quen tin Skinner and his Critics (Cambridge: Polity Press).

Vasquez, John (1995), 'The Post-Positivist Debate: Reconstructing Scientific Enquiry and Interna tional Relations Theory After Enlightenment's Fall', in Ken Booth and Steven Smith (eds.), Inter national Relations Theory Today (Cambridge: Polity), pp. 217-40.

Walker, Rob (1993), Inside/Outside: International Rela tions as Political Theory (Cambridge: Cambridge University Press).

Waltz, Kenneth (1986), 'Reflections on Theory of Inter national Politics: A Response to My Critics', in Robert Keohane (ed.), Neorealism and its Critics (New York: Columbia University Press), pp. 322-45.

Wendt, Alexander (1987), 'The Agent-Structure Prob lem in International Relations Theory', Interna tional Organization, 41 (3), pp. 335-70.

(1992), 'Anarchy is What States Make oflt: A Social Construction of Power Politics', International Organization,46 (2), pp. 391-425.

Wiggershaus, Rolf (1994), The Frankfurt School (Cam bridge: Polity).

Young, Oran (1969), 'Professor Russett: Industrious Tailor to a Naked Emperor', World Politics, 21 (3), pp. 586-611.

Zalewski, Marysia (1993), 'Feminist Standpoint Theory Meets International Relations Theory: A Feminist Version of David and Goliath?', The Fletcher Forum of World Affairs, 17 (2), pp. 13-32.

CHAJ

MAJORACT

Int is t reaJ tha me,

ries

inte COf]

lnte Tre.: SOV{

orga intei

For state bure voice forei these: ship. effec polit

consi Reali tion, decisi