the rubric of ‘political science’, this latter term will also be brieﬂy clariﬁed. Third, thediscussion will shift to an examination of the behavioural movement, which containsthe most optimistic formulation of empirical political theory. This will also entaila cursory discussion of the idea of positivism. Fourth, the decline of the empiricalapproach, or, at least, the decline of its imperial ambitions, will be considered inthe light of critical responses and the development of ‘post-behaviouralism’. In thiscontext, there will be a succinct discussion of the after shocks of empirical theory onpolitical theory. The main after shock is rational choice theory.The shape of empirical theory in the 1950s was premised largely on a rejection ofboth institutional state theor y, historical and traditional normative theory—exceptwhere they could be shown to contain an empirically-veriﬁable content. Institutionalstate theory was seen to be hidebound by its formal attachment to institutions andthe historical comparative method. The task was to consider informal behaviour.The state also was seen, by many empirical theorists, as too vague and imprecisea concept. Further, the bulk of classical political theory was considered a body ofhighly questionable unveriﬁable assumptions. The only viable substance to classicaltheory was a very limited range of testable hypotheses. The history of this body ofquestionable assumptions was therefore considered as innocuous antiquarianism.At this point, as mentioned, there was a strong suggestion that political science waspolitical theory, in the sense that all the traditional senses of the term ‘political theory’had been vacated. This perspective on political theory, particularly in America, had astrong g rip until the late 1960s, when it came under criticism. However, one shouldnot imagine by any means that the issues were resolved. They merely faded fromdiscussion and could well arise again.Second, g iven the close correlation between empirical theory and political science, itis important to get some purchase on the development of the idea of ‘political science’itself. There are three uses of the term ‘political science’, which were all prevalentduring the late nineteenth century. The ﬁrst, and original use dates back to lateeighteenth century thinkers, such as Montesquieu, Condorcet, Adam Smith, AdamFerguson, and David Hume, where it was usually understood as the ‘science of thelegislator’. The Scottish Enlightenment thinkers were particularly signiﬁcant here. Infact, other areas, like political economy, were frequently viewed as a subset of politicalscience. Adam Smith, for example, in his Wealth of Nations, described ‘politicaleconomy’ as a ‘branch of the science of a statesman or legislator’ (Smith 1979: 428).Therewas thereforelittle or no demarcation of what might now be regarded as separatedisciplines. Smith’s Wealth of Nations blends political economy, moral philosophy,political theory, and history as part of a uniﬁed enterprise. The term ‘political science’was picked up by North American commentators, from the vocabular y of the ScottishEnlightenment, and used in debates over the new Constitution and Republic.34Political science was also linked to a more general demand for ‘social science’. Onemajor intellectual input into this process was the Enlightenment itself. It is problem-atic to gener alize about the Enlightenment, given its very differing manifestationsacross Europe and North America (see Haakonssen 1995; Schmidt 2000). Minimally,though, many Enlightenment thinkers were making an effort to grasp human affairs
We Have a Firm Foundation 53through the open use of reason, in order to perceive identiﬁable and veriﬁable causalpatterns. There was, in other words, a greater appetite for empirical facts concern-ing nature, human nature, and society. Theorists were often inspired by success ofNewtonian physics, and the new ‘experimental philosophy’, in searching for thesepatterns. There were, for these diverse writers, therefore parallels between the sci-ence of nature and the science of politics. For Hume, for example, ‘It is universallyacknowledged that there is a great uniformity among the actions of men, in all nationsand ages, and that human nature remains the same, in its principles and operations.The same motives always produced the same actions: The same events follow fromthe same causes’ (Hume 1975: 83). Inconstancy of human action was ‘no more thanwhat happens in the operation of the body, nor can we conclude anything from theone irregularity, which will not follow equally from the other’ (Hume 1981, Book II,Part III, Section 1: 403–4). Thus, theorists, such as Hume, Turgot, and Montesquieu,believed in the possibility of causal social laws. Political science was also viewed as an‘applied science’, which could spawn social projects for social and political improve-ment. It could potentially show how to increase the happiness of state populations.Thus, as many theorists of the period urged, every government concerned to max-imize the pleasure and minimize the pain of its citizens, should take serious note ofpolitical science. This early conception of political science was though still inclus-ive of—what we would now regard as—separate disciplines. Sound moral preceptswere regarded as both morally obligatory and empirically correct, that is, for humannature to achieve its political ends. Political science was consequently regarded as asubtle blending of moral and empirical generalizations. Only political economy camenearer to what we might now regard as ‘empirical science’, namely, creating empiricalgeneralizations, which did not have to be necessarily linked with moral precepts.35The second view of political science reﬂects the development of the idea of polit-ical studies in the late nineteenth century. This use of political science traded on aperception of the classical Greek view, where political science was, quite literally, the‘science of the polis’. Political science was therefore a basic synonym for both classicalpolitical theory and institutional theory. There was, though, a growing awareness ofthe signiﬁcance of p olitical science as a more uniquely empirical approach, but it wasstill regarded with scepticism. Ernest Barker (the ﬁrst professor of political sciencein Britain) noted, in his inaugural lecture, that ‘I am not altogether happy about theterm “science”. It has been vindicated so largely, and almost exclusively, for exact andexperimental study of natural phenomena …I shall use it, as Aristotle …to signifya method or form of inquiry by the name of Political Theory’ (Barker in King (ed.)1978: 18). In this sense, theorizing about politics meant the systematic linkage ofideas about politics. Barker, and many others, considered that this is what Plato’s andAristotle’s work on politics had been concerned with. Such a science blended empir-ical and more abstr a ct normative considerations. This use of political science alsocharacterized the Staatslehre tradition up to the 1920s, in Europe and North America.Political science therefore meant systematic institutional political theory. However,Staatslehre itself also began to be regarded as suspect during this later period. Giventhat it tended to unify legal, political, historical, and philosophical ideas, it also
suffered from the increasing emphasis on the segmentation of disciplinary areas inthe early twentieth century. In general, therefore, despite Er nest Barker’s nostalgicappeal, this more inclusive notion of political science, qua Staatslehre—as closelylinked to classical political theor y—was fading fast.The third use of political science developed from the 1920s. It is here that weﬁnd the groundwork for both the apparent separation of political theory and polit-ical science and subsequent attempts at the reabsorption of political theory into theimperium of empirical theory. This third use also forms the backdrop in the 1950s tothe sense of spiritual crisis in political theory that pervaded the writings of Strauss,Arendt, and Voegelin.36This third conception was an open attempt, in tandem withother social sciences such as sociology and anthropology, to emulate the methodsand achievements of the natural sciences. It not only separated out normative andhistorical political theory from political science, but also led, in some cases, to theattempt to colonize the whole concept of political theory. For some, therefore, polit-ical theory became political science. This latter notion still pervades some Americanconceptions of political theory, particularly in its rational choice mode—often nowcalled ‘positive political theor y’.This third sense of political science became, for a time dur ing the twentieth cen-tury, the dominant use. During the late 1920s a loose sense of identity began todevelop in the social sciences in America.37This third sense developed in NorthAmerica in two stages.38The ﬁrst stage, from the 1920s up to 1940s, has been seenas a prelude to behaviouralism. Largely under the leadership of Charles Merriamin Chicago University, the politics-profession in North America began to turn itsattention away from institutional and historical study towards more empirical andquantitative techniques.39Large political science conferences were held in Chicagobetween 1923 and 1925 devoted to the new empirical ‘science of politics’, which,in the words of one commentator, converted ‘virtually every leader of the profes-sion to the behavioural persuasion’ (Jensen in Lipset (ed.) 1969: 5). Chicago, underMerriam, subsequently became a centre of this new scientiﬁc approach to polit-ics. Under Merriam’s academic leadership graduate students such as Leonard White,V. O. Key, Gabr iel Almond, Harold Lasswell, Herbert Simon, and David Truman,amongst many others, devoted their talents to this new empirical discipline. Thisearlier period was, on one level, reacting to the legalism, institutionalism, and com-parativism of the earlier phase. However, an interest also developed in a more strictapproach to informal behaviour, focused on public opinion surveys, voting patterns,and socialization processes. This still entailed a blend of empirical political sciencewith continuing concerns about the normative importance of democracy.The second stage focused on behavioural political science, which had a powerfulimpact in the 1950s and 1960s period. This had a far more immediate and longerterm effect in America than in Britain or Europe. Disciplines like politics, soci-ology, and anthropology, all became enthralled with the prospect of attaining greaterscientiﬁc empirical rigour.40For proponents of behaviouralism one should distin-guish behaviourism and behaviouralism. Both shared the belief that the approachof the natural sciences was most ﬁtting for the study of humans. However, for
We Have a Firm Foundation 55David Easton, for example, political science ‘has never been behaviour istic’ (Eastonin Farr and Seidelman (eds.) 1993: 294; see also Farr in Farr, Dryzek, and Leonard(eds.) 1995: 202). For Easton, behaviourism ‘refers to a theory in psychology abouthuman behaviour’, as embodied in the work of psychologists such as J. B. Watson andB. F. Skinner, the founder of operant conditioning. There is a form of physiologicalreductionism in behaviourism, which behaviouralists found uncongenial. Politics interms of attitudes, meanings, and beliefs could not be reduced in this manner. How-ever, political theory critics of behaviouralism, such as Dante Germino, were quiteclear that there was little to choose between the two empiricisms and the distinctionwas merely rhetorical (see Germino 1967: 193–5).David Easton, in a retrospective article, saw seven main themes within behavioural-ism: a concern with discoverable uniformities in political behaviour; to be able to testand verify empirical generalizations; to focus on techniques for acquiring and inter-preting empirical data (i.e. questionnaires, interviews, sampling, regression analysis,factor analysis, and rational modelling); the precise quantiﬁcation and measurementof empirical data; the analytical separation of values or evaluative concerns from fac-tual data41; the concern to systematize the relation between research and theory; and,ﬁnally, the aim to engage, as far as possible, in pure science, but with an eventual eyeto ‘utilize political knowledge in the solution of practical problems of society’ (see alsoDavid Easton in Monroe (ed.) 1997: 14). The central preoccupations thus becamethe recording and quantifying of political behaviour. Political systems with input andoutput functions replaced the study of states; the study of democracy became electoralbehaviour and public opinion quantiﬁcation and surveys; pressure or interest groupbehaviour replaced the study of societies.The behavioural movement of the 1950s coincided with other important develop-ments. There was, ﬁrst, the coincidence with the end-of-ideology movement, whichrepudiated both normative political theory and political ideology (in some casesthe two terms were regarded as synonymous). This involved some degree of self-satisfaction with the role and achievements of liberal democracy in practice. Ideologyand normative theory had thus both become redundant (see Vincent 1995, ch. 1).There was, in addition, a clear belief in the 1950s, amongst a generation that hadlived through the 1930s and 1940s, with the wars, Gulags, show trials, Nazism, Jewishpogroms, and Stalinism, that ideological or normative-based politics embodied dan-gerous delusions. Ideologies might serve a function in developing immature societies,yet in industrialized democratic societies they no longer served anything more than adecorative role. Consensus and convergence on basic aims had been achieved in lib-eral democracies. Most of the major parties in industrialized societies had achieved,in the welfare mixed economy structure, the majority of their reformist aims. Theleft had accepted the dangers of excessive state power and the right had accepted thenecessity of the welfare state and the rights of working people. As Seymour MartinLipset remarked, ‘This very triumph of the democratic social revolution of the Westends domestic politics for those intellectuals who must have ideologies or utopias tomotivate them to political action’ (Lipset 1969a: 406; see also Bell 1965). With basicagreement on political values achieved, politics became focused on more peripheral
pragmatic adjustment, GNP, prices, wages, the public-sector borrowing requirement.All else was gesture and froth. As Lipset commented ‘The democratic struggle willcontinue, but it will be a ﬁght without ideologies’ (Lipset 1969a: 408).The ‘end of ideology’ also coincided with the heroic age of sociology—a science freefrom all superstition and yet embodying commitments to freedom and liberal demo-cracy. In the social sciences of the 1950s, ideology was the foremost superstition, whichneeded unravelling. The development of empirical social science therefore demandeda value-free rigour, scepticism, empirical veriﬁcation, or falsiﬁcation, unsullied bythe emotional appeals of ideological or normative political theory. A positivistic sep-aration of facts and values lurked beneath all these judgements. In addition, the endof ideology coincided with the ‘death of political philosophy’ movement (which willbe discussed in Part Two), consensus politics in Britain, and ﬁnally with the moredisturbing phenomenon McCarthyite anti-communist purges in North America.Apart from some extreme adherents of behaviourism, positivistic political sciencedid not always demand the complete elimination of normative theory and ideology.There were those who would have liked to see this elimination, or, at least, trans-mutation into rigorous empirical political theory. However many political scientists,such as David Easton, Robert Lasswell, Robert Dahl, Karl Deutsch, and Heinz Eulau,had been trained initially as more traditional political theorists. They did not there-fore construe political theory as a total waste of time. The historical and normativevision could offer hypotheses for empirical testing. In this sense, the hard contrast,which occasionally appears between political theorists and political scientists can bemisleading.42For John Gunnell, the crucial factor deﬁning the stance of behavioural theory,was tied to the political theory writings of the 1920s and 1930s émigré generation,including ﬁgures such as Strauss, Arendt, Brecht, Adorno, and many others, whoadopted a deeply-critical stance to political science, associating it with individualisticliberalism, relativism, potential nihilism, and social crisis. In this critical context,political scientists, for Gunnell, ‘eventually felt constrained to make a choice’ (Gunnell1993b: 220). In the end, this was not so much a debate about method, as about theculture of liberalism and democracy. Gunnell thus notes that by ‘the early 1960s, theconﬂict was not simply one between individuals such as Easton and Strauss. It hadbeen passed to a new generation of scholars who had been trained in the new waysof political theory, denied by the émigrés and by the founders of the behaviouralmovement, and who had already begun to lose sight of the roots of the conﬂictbetween the paradigms into which they had been initiated’ (Gunnell 1993b: 250).43The intellectual background to behavioural political science lay in the popularity ofwhat might loosely be termed positivism in the twentieth centur y. One of the leadingphilosophers of Viennese positivism, Carnap, was teaching in Chicago during the1950s. A new generation of political scientists became familiar with this philosophicalposition. Positivism gelled with the idea of a genuine ‘empirical political theory’.Positivism was essentially though a broader programme tied up with a more generalconception of science. Theories in the natural sciences were viewed as uniﬁed systemsof explanation, incorporating laws, which were ‘controllable by factual evidence’
We Have a Firm Foundation 57(Nagel 1961: 4). The basic contention was that scientiﬁc theories could grasp anobjective reality through a neutral observation language. Reality was deﬁnitely notstructured or constituted by natural science theory. Theor ies tell us, in a moderatelydetached way, about a reality.44Explanations in the natural sciences, utilizing a neutralobservation language, could thus be deﬁned as systematically related propositionsabout an external reality, propositions which may, in certain contexts, be describedas laws supported by empirical evidence. The general framework within which thiskind of theory functions is usually called positiv ism. Political science, from the 1950sparticularly, stressed this approach.The concept positivism is, however, complex. It denotes two broad ideas. First, itindicates those who accept the designation positivist, such as Auguste Comte or theViennese logical positivism movement—although the latter are occasionally cited asneo-positivist. Comte’s legacy—especially via positivist sociology—formed a back-ground set of beliefs, which resonate with later positivist sympathizers in the twentiethcentury. Comte’s idea that positive science (or philosophy) would triumph ultimatelyover metaphysics and religion (both the latter being viewed as prior, more prim-itive, stages of human development); his insistence on a clear boundary betweenempirically-tested facts and imaginary theoretical constructions; his strong belief inprogress through science; and, his assertion of the linkage between moral and materialprogress (i.e. the knowledge that science provides would allow all manner of techno-logical control in both the natural and the social and political ﬁelds), all impacted onearly twentieth century positivist theory.The second sense of positivism, which reasserts many of the Comteian ideals,embodies a more general adherence to certain epistemological theses, for example: theunity of the sciences; the belief that the only valid standard of knowledge we have lieseither in the empirical sciences or logic and mathematics; the assumption of the realityof sense impressions; the conception of a scientiﬁc theorist as a dispassionate observerwho never asserts anything which has not been empirically proved; an intense dislikeand mistrust of metaphysical thought; adherence to a notion of philosophy as analysis,and its being parasitic upon science; the acceptance of the clear distinction betweenfact and value; more speciﬁcally, the belief that the natural and social sciences share acertain common methodology; also the belief in a growing body of empirically-testedpositive knowledge.There have been two broad manifestations of this latter positivist tendency thiscentury. The ﬁrst relates to the neo-Kantian distinction between theoretical andpractical reason, a distinction that is supposed to make room for autonomy andmoral judgement. Increasingly neo-Kantianism, in the twentieth century, becamesceptical of the moral autonomy that Kant had postulated. Values became increasinglysuspect. Facts though were certain. This distinction became a crucial plank in theneo-Kantianism behind Max Weber’s sociology work and his distinctions betweenvalue free social science and moral discourse. Weber was no simple-minded positivist.Moral and religious values were of importance to individuals, but he still adhered tothe idea that there was a clear heterogeneity between facts and values and that sciencehad no answers to the question how we ought to live. Under Nietzsche’s tutelage,
Weber raised the question: could there be any rational foundation for our basicvalues? For Weber the fact that he could not answer this question was a matter ofanxiety. The other positivist manifestation—which we are most familiar with—iswhat might be termed the Anglo-Saxon ‘liberal social science perspective’, whichadopts the positivist position, often on consequentialist grounds. There is somethingmore Comtian and utilitarian, than neo-Kantian, in this latter approach. However,it still contains all the expected positivist components. The separation between factsand values, particularly, is foundational. David Easton’s contemporaneous commenthere is quite typically positivist, ‘The factual aspect of a proposition refers to a part ofreality ; hence it can be tested by reference to the facts. In this way we check its truth.The moral aspect of a proposition, however, expresses only the emotional responseof an individual …Although we can say that the aspect of a proposition referringto a fact can be true of false, it is meaningless to characterize the value aspect of aproposition in this way’ (Easton 1953: 221).In summary, the concept of political theory aimed at by behaviouralists was seento be value free and objective. The overt aim was to emulate the natural sciences,namely, to collect empirical data, discover correlations, draw up generalizations,and formulate testable theories, which allowed prediction. As one exponent, GeorgeHomans, put it, ‘As we have come to accept …the standards of natural sciencefor testing the truth of propositions, so we should take more seriously [in the socialsciences] the standards of natural science in explanation. In that we have been laggard’(Homans 1967: 28). It is no surprise in this context that political behaviour couldtake on the alluring shape of the natural world—embodying empirical facts, whichcould be described and studied.The general conception of the theorist here was that of a neutral observer who care-fully describes and explains the objective world. The function of the theorist was notto interpret the world, but rather to explain it through rigorously-tested categories. Ingeneral, empirical theory resisted any historical, normative, metaphysical, or ethicalpresence. Values were seen in the context of emotive responses. Facts were regarded aspreconstituted givens—that is, prior to theory and representation. Empirical theoriesobserve, explain, generalize, and establish causal relations. Theories, in effect, orderthe empirical facts in a comprehensible manner. The substance of such empirical the-ories was often initially drawn from behavioural psychology, neo-classical economics,systems theory, mathematical modelling, and the like. Such theories explain politicalbehaviour outside the fr amework of political ideas, ideologies, or institutional frame-works. This tendency became the more dominant method of the discipline up untilthe late 1960s, although, as stressed, it has always had a much stronger following inNorth American political studies.Yet, it is also important to emphasize here that, in the understanding of empiricaltheory, everything that was of importance in normative classical and historical notionsof political theory, namely, a clear perception of the reality of politics, an understand-ing and explanation of its processes and a unambiguous set of prescriptions for howsociety should be organized, were all present in the aspirations of empirical theory.Social change and reform were an integral part of the vision of empirical theory.
We Have a Firm Foundation 59Science was viewed as a social inst rument. Thus, normative and historical theory inthis context, were literally superﬂuous.Consequently, David Easton and a number of North American political scientistswere clear that political theory had, in future, to be much more empirically r igorousin order to even survive academically. Easton, in his famous article, ‘The Decline ofPolitical Theory’, saw the majority of classical normative political theorists as simplyacademic parasites, feeding on past ideas and retailing antiquarian useless infor mationabout past values. Herbert Simon, at the same time, bewailed that ‘there will beno progress in political philosophy if we continue to think and write in the loose,literary, metaphysical st yle …The standard of rigour that is tolerated in politicaltheory would not receive a passing gr ade in an elementary course in logic’ (Simon1952: 494–6). Political theory needed to mutate into empirical political theory. This isthe complete reversal of Ernest Barker’s lament, in his 1928 inaugural lecture, wherepolitical science becomes normative and institutional theory. In Easton’s vision, apuriﬁed normative and historical political theory becomes empirical political theory.As William C. Mitchell signalled optimistically in 1969, political theory in future ‘willbecome increasingly logical, deductive, and mathematical. In terms of its content wewill make increasing use of economic theory, game theory, decision theory, welfareeconomics, and public ﬁnance’ (Mitchell in Lipset (ed.) 1968: 129).45Oddly, Mitchell’s comment is not too distant from the conclusions of BrianBarry’s 1990s essay, ‘The Strange Death of Political Philosophy’, where he identi-ﬁes, anachronistically, the hopeful lines of future political theory as studies of votingbehaviour, game theory, welfare economics, and value analysis (Barry 1991). How-ever, in Barry’s case, this is more of a general alliance with economic analysis. Theoddity is that this latter judgement is written by a normatively inclined political theor-ist who worked through part of the earlier behavioural phase. In Barry’s case, though,it is more of a reaction to the impoverished nature of Oxford analytical politicaltheory, in the 1960s period, and the woeful shor tcomings (as he perceived it) of thehistory of political theory as an approach. However, Barry’s odd assessment of futuredevelopments in theory is neither the kind of suggestion that gets the pulse racing,nor does it actually represent what really took place in the last two decades of thetwentieth century.Since the 1970s, and the so-called ‘post-behavioural revolution’, there has beenmore circumspection about the ‘scientiﬁc’ position. Most empirical theorists in thisperiod became more hesitant. In fact, Easton, the doyenne of the earlier behaviouralpersuasion, recategorized himself as ‘post-behavioural’ (see Easton 1953, 2nd edition1971, Epilogue, Part A). For Easton, the reasons for this post-behavioural develop-ment lie within the criticisms of the counter culture movements of the late 1960s,the utter inability of the behavioural movement to deal with the complex normativeissues arising out of the Vietnam war and the detailed civil rights debates, all of whichgripped the minds of most students studying politics.46Behavioural political sciencehad no way of addressing the deep social, moral, and legal debates concerning gender,war, race, rights, and social justice that dominated the late 1960s and 1970s moral andpolitical arguments. Political science seemed to be completely mute on such issues.
Contrary to the basic premises of behavioural theory, ideology, and normative theorydid seem to be more effective in addressing such issues.However, it was also argued by a number of political theorists, during the 1970s,that empiricism was both a challengeable epistemological and ontological thesis.In fact, the epistemology revealed the character of the ontology. Empirical polit-ical theory was a clear example of a deeply-embedded ontology. It revealed not somuch any foundational truths about politics, as certain embedded and unchallengedways of understanding our ‘political being’. Thus, empirical political theory hadto be considered as just another epistemology. It was a philosophically-contestableepistemology, amongst other epistemologies. The basic foundational distinctionsbetween, for example, explanation and interpretation, or facts and values, madewithin the epistemology of empirical theory, were not therefore categorically true.They were, conversely, philosophically-challengeable assumptions. In this context,empirical political theory began to lose its privileged and hegemonic status.The above point was further underscored by critical de velopments in the philo-sophy of science. Reﬂection on the methods of natural science did not cease with theclaims of hypothetico-deductive methods or logical positivism. The collective phe-nomenon of ‘post-empiricist science’, developed in writers such as Thomas Kuhn,Michael Polanyi, Peter Winch, Paul Feyerabend, and Mary Hesse, which g rew overthe 1970s and 1980s, raised a new series of detailed questions about the way in whichwe view natural science explanation and by default all empirical theories. Western sci-ence, as envisaged within this post-empiricist programme, was not the high point ofcivilization and human knowledge, conversely, it was an epistemolog ical moment.47As such, we do, in fact, have a great deal to learn from careful and sensitive examina-tion of different cultures and distinct knowledge structures. We also need to pay morecareful attention to self-reﬂexive critique within our own systems of knowledge, thatis to say, purportedly objective empirical data is not, in reality, so easily detach-able from theoretical models. Interpretations can have a constitutive effect. Theoriescan be seen, ironically, as the facts of natural science. This post-empiricist view ofscience throws considerable doubt on the projects of veriﬁcation, covering law the-ory and hypothetico-deductive methods—all pervasive in behavioural and empiricistinvestigations.Although the post-empiricist programme did not deny the separate role of naturalscience language, a number of points were made which linked, fortuitously, withideas in both interpretive and normative theorizing. First, theory is neither aboutreality, nor an adjustment to reality, rather it has some role to play in constitutingreality. Assumptions implicit in certain theories confer meanings and shape the world.There are no br ute facts, which are not permeated with interpretative assumptions.There is thus no unmediated or uninterpreted realit y. Valid knowledge is not theputative representation of something external.48In consequence, it is more difﬁcultto speak of the clear truth or falsity of beliefs or their measurement against someexternal empirical standard. Theories can be more or less persuasive or fruitful inthe way in which they constitute realities. Truth or falsity would be premised onalternative ideological or theoretical schemes. Such schemes would also be subject tohistorical change.
We Have a Firm Foundation 61For many exponents of empirical political theory, there are deep problems withsuch a view. For example, how could one gain any reliable or testable empirical datafrom such elusive ideas? Further, it is not possible to quantify interpretations. Iftheory is constitutive in this way, then the whole empirical project looks suspect.The debate between empirical theory and political theory has not been resolved atall in political science. There have been some modiﬁcations within political science.There is now an awareness of the bewildering variety of approaches occasioned bythe post-behavioural phase. Thus, some more recent political scientists have triedto accommodate themselves to what is called ‘methodological pluralism’. For others,though, this variety generates dismay and anxiety. Felix Oppenheim suggested thatthis post-empiricist perspective does inevitably lead to the rejection of the olderforms of positivism and behaviouralism. But, he contends that political scientistsshould now avoid both the Scylla of old-fashioned behaviouralism and the Charybdisof simple-minded relativism. He also notes that ‘to reject …behaviouralism is not toabandon empiricism’ (Oppenheim 1981: 194). For Oppenheim, constructing goodexplicative deﬁnitions and explanations in political science still has loose parallelswith good natural science, in demanding accuracy and simplicity. Ye t, he admits thatthis would not, in political science, produce fully ﬂedged empirical covering laws, inthe older sense of positivism.One after shock of empiricism, which might be said to be now carrying the torchof empirical political theory to the present day, is rational choice theory (see Eastonin Farr and Seidelman (eds.) 1993: 302 ff.).49The origins of rational choice lie withinthe discipline of neo-classical economics as well as offshoots of utilitarianism.50Interms of the actual serious development of rational choice, it appeared preciselyat the point of the decline of behavioural theory during the 1950s and 1960s—although, initially, it was a very marginal and rather occult specialism, out on alimb as it were from mainstream economics. Despite its economic base, it is stillregarded as somewhat quaint by mainst ream economists. The seminal books, whichconstitute the cornerstones of the perspective, are Kenneth Arrow’s Social Choice andIndividual Values (1951), Anthony Downs An Economic Theory of Democ racy (1957),and Mancur Olson’s The Logic of Collective Action (1965). Another signiﬁcant text,which also had an important impact was James Buchanan and Gordon Tullock’swork The Calculus of Consent (1962), which drew the analogy between voters andmarket-based consumers. However, rational choice did not make overtly optimisticclaims for itself until the 1980s. Yet, in the last two decades of the twentieth centuryit became, in North America, the fastest growing element of political studies, andhas even been blessed with a distinctive title (to indicate its special status), namely,‘positive political theory’—which presumably makes the rest of political theory looka triﬂe negative.51It would now be true to say that in North America it has takenover the empirical mantle from institutional theory, behaviouralism, and pluralism.It has also moved conﬁdently into related disciplines, such as International Relations.As one synoptic study concludes, ‘scarcely an area of political science has remaineduntouched by its inﬂuence’ (Green and Schapiro 1994: 2). Some keen exponents ofrational choice consequently see this as a great triumph for the perspective (WilliamRiker 1990: 177–8).
Although there are a number of variants of rational choice theory (includingMarxist rational choice), the gist of the perspective is ‘the application of the ana-lytic method and techniques of modern economics to the study of political processes’(Brennan 1997: 89; see also Roemer 1986; Carver and Thomas (eds.) 1995). A basicdeﬁnition of rational choice is therefore ‘the economic study of non-market decision-making, or simply the application of economics to political science. The subjectmatter of public choice is the same as that of political science: the theory of thestate, voting rules, voting behaviour, party politics, the bureaucracy, and so on. Themethodology of public choice is that of economics, however. The basic behaviouralpostulate of public choice, as for economics, is that man is an egoistic, rational utilitymaximizer’ (Mueller 1989: 1–2). In other words, it is concerned with governmentor politics viewed through the market. As always happens in such theories, it hasdivided fairly quickly between various schools, for example, the Virginia School ofPublic Choice (associated with Buchanan and Tullock) and the Chicago school (asso-ciated with George Stigler and Olson), sometimes referred to as the ‘private interestregulation’ school.The key assumptions of rational choice are fairly rigorous and parsimonious. Theyare: ﬁrst, the centrality of individuals for all forms of social explanation, includ-ing groups. Thus, rational choice is methodologically individualist. Second, eachindividual is assumed to be rational. Third, rationality denotes agents choosing the‘option which they believe best fulﬁls their purposes’ (Brennan 1997: 98); or, as Rikerputs it, ‘that, within certain limits of available information …actors choose so as tomaximize their satisfaction’ (Riker 1990: 173). This notion of rationality is whollyinstrumental and says nothing about the contents of options or preferences.52A clearanalysis of basic incentives will go a long way towards explaining human behaviour.Fourth, the individual is self-interested. This does not entail either complete egoismor the impossibility of collective action. Far from it, for rational choice exponents,it provides a more logically satisfying way of explaining public choice and collectiveaction.53Fifth, the actual process of rational choice is a form of decontextualizedutility maximization. Each agent is trying (rationally) to maximize their utilities andminimize their losses. Faced with a number of options, the agent will pick one whichbest serves or maximizes efﬁciently her objectives. Essentially we are looking at theinteguments of, what is often referred to as, homo economicus. Rational choice alsoassumes that there will be a consistency in choices and options; preferences will beranked according to their utility for us. This constitutes a basic equilibrium. Thus,from any collection of preferences, the agent is able to calculate a choice from whichshe can expect the greatest utility payoff. This particular line of reasoning has ledmany rational choice theories into ‘game theory’ and various forms of mathemat-ical modelling. Finally, all rational choice analysis shows ‘a predilection for formaldeductive method, deriv ing ‘interesting’ (i.e. non-obvious, often counter intuitive)propositions via sometimes long and complex chains of logical reasoning from aminimal set of plausible axioms’ (Brennan 1997: 96). These assumptions of rationalchoice are considered to be universal, empir ical (in the sense that they form the basisfor testable research programmes), and scientiﬁcally orientated. Indeed, even some
We Have a Firm Foundation 63of its more rigorous critics still applaud its empirical and scientiﬁc aspirations as itsmost valuable asset (see Green and Schapiro 1994: 10).Rational choice in many ways ideally ﬁlls the more optimistic self-perception ofempirical political theory. It embodies a purportedly rigorously empirically testableand scientiﬁcally-based research programme and yet, at the same time, it can fulﬁl allthe requirements (for its proponents) of a normative political theory, once one hasaccepted the above foundational assumptions.The problem on the empirical front is that it is far from clear that it has had muchempirical success. The central contention of a recent synoptic study of rational choicetherefore notes, ‘curiously …the stature of rational choice scholarship does not reston a readily identiﬁable set of empirical successes’. The authors comment that mostcritics do not in fact focus on the empirical or operationalized aspect of the doctrine.They note that this aspect of rational choice work (which they examine exhaustively)is generally ‘marred by unscientiﬁcally chosen samples, poorly conducted tests, andtendentious interpretations of results. As a consequence, despite its enormous andgrowing prestige in the discipline, rational choice theory has yet to deliver on itspromise to advance the empirical study of politics’ (Green and Schapiro 1994: 5and 7). Part of the problem here, for the authors, is that this empirical weakness isrooted in the desire to establish a universal empirical theory of politics, which hasresulted in rational choice being ‘method driven’ rather than ‘problem driven’ (Greenand Schapiro 1994: 202–3; see also Schapiro 2002).The bulk of the criticism on the normative front focuses on a number of well-trodden paths. Rational choice is clearly premised on an unquestioned empiricistmetaphysics. However, its basic foundational elements neither seem very plausible,nor ultimately very reasonable to its critics. Basically—apart from the fact that they arenot really empirically veriﬁed—the above set of assumptions are regarded frequentlyas simply false and misleading. The assumption of the isolated or atomized individualis highly questionable and sociologically and historically contentious. It embodies anexcessively narrow and slightly weird perspective on human beings. Minimally, itsimply cannot account for the complexity and idiosyncrasies of human individuals,when they act morally or politically. To reduce all individual action and choice toinstrumental personal preference rankings, utility maximization and self-interestsdoes little or no justice to human nature or human action. The same point holds formore orthodox utilitarianism. It might give us some very partial insight into somecollective actions, but that it is about it. It also employs an excessively narrow anddeeply-arbitrary conception of human rationality.Apart from certain more random and idiosyncr atic offshoots in, for example,Marxist rational choice, critics see not so much a universal foundational empir icalpolitical theory, as a somewhat pessimistic ideological doctrine, driven by a parochialNorth American conception of neo-classical liberal market economics and utilitariancalculus (which even orthodox economists would feel uneasy with). Its importancereﬂectsmore on the power and inﬂuence of North America, rather than anytheoreticaldepth or long term intellectual signiﬁcance. Intrinsic to this model are a number ofdeeply-questionable foundational assumptions. It is essentially inclined ideologically
to be pessimistic about all government-led initiatives—simply because they are notgenerated through market choice. It is deeply cynical of all human motivations, seeingself-interest and personal utility maximization at the root of all morality and politics.Essentially its view of human beings is profoundly sterile. It has close connections witha range of public policies concerning the slimming down of government (rolling backto the state), the reduction of public expenditure, the movement from progressive toproportionate taxation regimes, the mar ket-based privatization of government, thewholesale introduction of competitive market processes into all areas of government,administration, and public service. It provides ideological succour for ideas suchas cost–beneﬁt analysis, private ﬁnance initiatives, value for money policies, cost-effectiveness measurement, market testing, int roducing competition in the deliveryof all public service, and the like, many of which have permeated public policydebate in Britain and North America (see Peter Self 1993, 2000). As the original1960s rational choice theorists wanted to model the democratic voter on the market-based consumer, so rational choice in the 1990s and 2000s has wanted to modelgovernment bureaucracies, health care, education, and the like, on the private ﬁrm.This movement in public policy is not only due to rational choice theory. However,rational choice is still complicit in a more general ideological shift.In terms of the general perception of political theory, there are some odd sociolo-gical parallels with behaviouralism. Like behavioural political theory, rational choicehas mainly been a North American phenomenon, taking up a powerful niche in thecontemporary politics academy. It also shares a basic positivist ﬁdeism. However,there is a key difference to behavioural theory. Behavioural theory, in the 1950s and1960s, faced a comparatively weak and demoralized profession of political theory.Apart from the European contingent of émigrés theorists, such as Arendt or Strauss,the bulk of theory (outside empirical theory) was constituted by forms of logicalpositivism and linguistic philosophy (to be examined in Part Two). These forms ofphilosophy were rooted (unwittingly in many cases) in an empiricist foundationalism,which gave immediate credence to empirical claims as genuine ﬁrst-order knowledge.Thus, behaviouralism was able, fairly easily, to roll over the opposition, for a time.However, rational choice—despite its success in the academy—developed during the1980s. This coincided with the so-called rediscovery of normative political theor y, theearly conﬁdent halcyon days of the methodological debates around Skinner’s work,Rawls justice-based argument, postpositivist and many other diverse critiques. In thissense, it encountered a wide-ranging diverse opposition from within other domainsof political theor y. This has considerably (and thankfully) limited its scope.In conclusion, the dominant aims of political science still remain tied to theinformal and empirical, rather than the formal, institutional, historical, or normative.Although some of the more extravagant mid-twentieth century claims of empiricalpolitical theory to ‘colonize’ completely the whole of political theory have now con-tracted, political science is still the far more dominant partner within North Americanand European political studies. Empirical political theory, despite the post-empiricistand post-positivist arguments, still remains committed to the measurable, quanti-ﬁable, and testable. However, the aspiration for empir ical theory (particularly in
We Have a Firm Foundation 65offshoots such as rational choice) to absorb political theory, in toto, is not absent, butrather dormant.IDEOLOGICAL POLITICAL THEORYIdeology is one of the most contested conceptions of political theory. The gist of thisperspective is that political theory is and always has been (unless obscured by historicalor abstruse philosophical theories) a deeply practical mode of thought, which isconnected directly with the sphere of political action. Ideology, in other words, is thetruth about political theory. In this perspective, when the political philosophers of thepast were writing and thinking about politics—Machiavelli, Hobbes, and Locke—theywere actually writing as ideolog ists. Thus, ideology, as a practical political engagementtrying to navigate the political realm, change perceptions, and construct public policy,is the reality of political theory. Thus, ideology draws our attention, minimally, toone important dimension of theory—the practical, engaged dimension—which canoccasionally and unexpectedly get ignored in the sheer welter of abstract theorizing.However, in claiming this kind of role for itself, ideology not only conﬂicts quitedirectly with some dominant perceptions of normative political theory, but alsowith dimensions of historical and empirical theories.The relation with normative theory is the most difﬁcult and sensitive. After a briefintroduction concerning the concept ideology, the debate over the relation betweenideology and political theory will be analysed in terms of, ﬁrst, attempts to fully integ-rate ideology and political theory, in other words, to make them indistinguishable;second, in terms of efforts to completely demarcate them. Both of these categories—integration and segregation—have positive and negative poles, which therefore givesrise to two further sub-categories for each response. Some of the arguments havealready have been touched upon in previous sections of Part One, thus the expositionswill be brief.The concept of ideology is a comparatively new political word dating from theearly 1800s, and not in any recognizable form until the 1840s, and again not in anypopularized form till the late nineteenth and early twentieth century (see Vincent1995, ch. 1). Its ﬁrst use, in the writings of Destutt de Tracy, focused on the Enlight-enment orientated idea of an ‘empirical science of ideas’. It had no immediate politicalconnotations. In Marx and Engels’ use, in the mid-nineteenth century, it took on adeﬁnite political and critical sense. However, it was not until the twentieth centurythat it really came into its own within popular political discussion. However, giventhat political theory was also, etymologically, a novel term, dating from the mid-to late-nineteenth centur y, neither concept can really claim great longevity, exceptrhetorically. In many ways, despite its commonplace use in academic and ordinaryspeech, it still remains the poor and often viliﬁed cousin of political theory.Thus, beginning ﬁrst with the negative integration thesis. One of the ﬁrst toimply that political theory and political ideology could be fully integrated was Marx.However, Marx, and the subsequent Marxist tradition, present a complex picture.
Primarily, political theory and ideology are reduced to the same category, althoughboth denote an illusion. The material conditions of economic life form the real basisto social existence. Cultural and political structures can only be understood via thesematerial conditions and the ensuing class struggles. Since the mater ial basis is primary,all ideas have to be explained via their connection to the material base. They cannot beexplained in themselves. They constitute the ideology of a society. Marx, in one of hissynoptic semi-autobiographical pieces of writing, the Preface to the Cr itique of PoliticalEconomy, called the above idea the ‘leading thread’ of his studies (Marx and Engels1968: 182). It is understandable, in this reading, that Engels, and others, should thushave referred to all ideology (including political theory) as the ‘false-consciousness’.Its chief delusion is its inability to see its own class basis. The history of ideology istherefore subsumable under a history of class interests. Political philosophers are quiteliterally professional ideologists or professional purveyors of illusions. The social andeconomic sciences therefore need to explain the eruptions of ideology and politicaltheory.Ideology and political theory thus become social objects to be explained within abroader empirical social theory. Much twentieth century sociology—both structur-alism and functionalism—continued to view political ideas as aspects of a broaderscience of society. Social science, in general, has often seen both political theories andideologies as social objects for study. In fact, for Durkheim and Talcott Parsons, soci-ology per se, contained a complete social epistemology, which provided clear answersto all the older philosophical problems of knowledge. Humans (and their cognitiveexistence) have no distinctive attributes outside of society. A science of society thusexplains political theory and ideology.We have already encountered the above basic argument within empirical politicaltheory. With the rise of empirical theory in the mid-twentieth century, the ‘illusory’dimension of normative political theory came to the fore. This view was encapsulatedin the perspective of behavioural political science. The general frame of the ‘end ofideology’ perspective also caught the same drift of argument. Social science, in effect,offered a science of society. The development of empirical theory demanded a valuefree rigour and clear veriﬁcation processes, unsullied by appeals to normative politicaltheory or ideology. As Edward Shils commented, ‘science is not and never has beenpart of an ideological culture. Indeed the spirit in which science works is alien toideology’ (Shils 1968: 74). The only salvation for political theory or ideology was tomutate into empirical political theory.This was the general view of the behavioural movement. Classical normative the-ory, the history of political theory and ideologies persisted with a use of theory ‘thatlingered from an earlier period in the discipline’s history’; and as James Farr noted‘In being empirical and explanatory, however, theory in behavioural research was tobe value-free and objective. There was, it was argued, a logical gulf between fact andvalue, between “is” and “ought”, which in no way could be spanned. Normative topicslike freedom, justice, or authority—the staples of a prescientiﬁc study of politics—were best understood in terms of one’s subjective emotions or expressive states. The ywere also laced with a “strong dose of metaphysical discourse” ’. Farr continues that,
We Have a Firm Foundation 67for behaviouralists, ‘endlessly reinterpreting the great books of dead men and tire-lessly disputing the meaning of the good life had nothing to do with science’ (seeFarr in Farr, Dryzek and Leonard (eds.) 1995: 204; also Heinz Eulau 1963: 8–10).Ideologies and political theories, in the older sense, could serve cohesive functions (associal objects to be studied) in developing societies, however, in large industrializeddemocratic societies they were largely redundant or decorative. Consensus on basicsocial and political aims had been agreed. All the rest was froth.54Moving to the second thesis of positive inte gration. In this context, the integrationof political theory and political ideology should not be a matter of concern. There are,again, however, different perspectives. Many adhere, unwittingly, to the integration ofthe terms, that is, where ideology becomes an unwitting synonym for political theory.Thus, one often encounters quite unselfconscious references to ‘liberal ideology’, andthe like, in discussions, which otherwise appear to be exclusively focused on thecategory political philosophy.Even more ironically, this unwitting use appears within the ‘second wave’ of historyof political theory writings (outlined earlier). This is, in fact, doubly ironic given theovertly close attention to language, and the avoidance of anachronism, characteristicof the second wave theories (see Leslie 1970). If we bear in mind that the term ‘ideo-logy’ is a neologism from the nineteenth century, carrying a baggage of uses, it is,to say the least, strange to ﬁnd Quentin Skinner, in a number of writings, referringto, for example, ‘History and Ideology in the English Revolution’ or ‘The Ideolo-gical Context of Hobbes’ Political Thought’ (see Skinner 1965, 1966). James Tully,explicating Skinner’s method, also reﬂects this usage. Thus, for Tully, the new methoddemands we place all texts in an ‘ideological context’. Tully continues, that an ideologyfor Skinner, ‘is a language of politics deﬁned by its conventions and employed by anumber of writers. Thus, scholasticism, humanism, Lutheranism, and Calvinism areideologies and both scholasticism and humanism comprise the general ideolog icalcontext of the Italian city-states during the Renaissance’ (Tully 1988: 9). Luther andCalvin thus become political ideologists!55Placing an idea or text in context—thesacred mantra of second wave theory—makes ‘political theory, …a part of p olitics,and the questions it treats are the effects of political action’. Tully continues, that ‘sincea p olitical ideology represents a political action …to change some of the conventionsof the ideology is to change the way in which some of that political action is represen-ted’ (Tully 1988: 10–11). Consequently, Tully describes Skinner’s whole substantivetwo-volume opus, The Foundations of Modern Political Thought (1976), as ‘a map ofthe great political ideologies of early modern Europe’ (Tully 1988: 12). Thus, politicaltheory and ideology become one. One searches in vain, in such contextualist writ-ings, for a glimmering of recognition that the concept of ideology itself, is a deeplytroubled, comparatively quite new, idea containing deep unresolved tensions.Another semi-conscious response on this same issue is provided by the communit-arian movement of the 1980s and 1990s (who will be examined in more detail in PartThree). One of the hallmarks of their arguments is the association of theory withsituated communal practices. Thus, ideas cannot be deﬁned independently of thehuman relationships, which constitute them. Communitarianism argues, therefore,
that political and moral goods cannot be determined by abstract reasoning. All human‘goods’ arise from particular historical communities. There is no concept which standsapart from a social context. Morality is neither invented nor discovered, but inter-preted as already existent (see, for example, Walzer 1987: 21). We ‘read off’ an existingtradition of discourse. The community becomes the locus of the good. Unwittingly,again, this appears to tie all political theory closely to political ideology, neither hasany distinguishing marks. They are simply different names for the same communaldiscourse.Not all in the historical domain, however, are as unaware in the ‘use’ of the conceptideology. In an open and explicit use of ideology to denote both political activ-ism and political theory, Richard Ashcraft comments that, ‘only an ideologicallygrounded approach with respect to current political problems can provide a br idgebetween the t raditions of political philosophy and the perception of what counts as“political” phenomena’ (Ashcraft 1975: 20). Political philosophers are, or should be,considered unequivocal ideologists. In response to the idea that philosophy is some-thing higher or more saintly than ideology, Ashcraft asks, ‘how is it even possiblefor …epistemological presuppositions to stand apart from the very conﬂict they pro-pose to “study” and are assumed to tr anscend’ (Ashcraft 1975: 26). Ashcraft, appearstobe directing his ﬁre at both historians of political theory and analytical philosophers,arguing that ‘some of the responsibility for the divorce of traditional political theoryfrom present concerns of political life rests squarely with those teachers of polit-ical theory who have encapsulated the meaning of politics within the frozen worldsof “analysis” or “history”’ (Ashcraft 1975: 19). As Ashcraft continues, for many polit-ical philosophers, the title ideologist is though ‘the original sin’ (see Ashcraft 1980:695). Ideology appears to relinquish all claims for universality. For Ashcraft, however,this universality is well worth losing. He suggests that most theorists in the past wereactually concerned about problems in society and were actually, what we now thinkof as, ideologists. To make them just philosophers is a modern self-indulgence.Further, it is worth noting that more contemporary neo-Marxism does not alwaystake a negative view of ideology (as contrasted to genuine science). It can also take animmensely positive v iew of the integration of political theory and ideology. Thus, themost noted twentieth-century Marxist, Antonio Gramsci, saw proletarian ideologyas an effective tool of political struggle against bourgeois ideology. The hegemony ofpolitical ideas was thus considered of immense importance. Ideas take on a partialautonomy from the material base. There can, in other words, be an authentic anduseful Marxist ideology, qua political theory. Aspects of this ‘partial autonomy’ vieware reﬂected in some later twentieth century Marxist writers such as Gramsci. It isalso the predominant view of twentieth century critical theory.In a more general, late twentieth century scenario, language overall—in both theoryand ideology—has not always been viewed as a transparent conveyor of meaning.Languages, even the sophisticated languages in political philosophy, cannot gain anyreal distance or neutrality from the subject of politics. Ideology and political theoryare both focused on language and language is focused on social action. Speaking cantherefore be considered a way of acting. The study of ideologies and political theories
We Have a Firm Foundation 69can therefore be seen as the study of the social world itself. The medium of languageitself is embedded within an historical and political inheritance. Thus, languagecannot stand back from social conﬂict, it is the medium of expression and experienceof such conﬂict. In other words, ideology and political theory neither reﬂect neutrallyon, nor simply represent the world, but rather partly constitute it. Ideology andpolitical theory are enmeshed in complex relations and struggles of power. To analysethis process is the self-appointed task of, for example, discourse analysis, forms ofstructuralist Marxism, psychoanalysis, semiotics, and much postmodern genealogy.All reject the ‘neutralist’ thesis concerning ideology and theory, stressing conversely,the constitutive and expressive role of language.This critique of the language of political theory and ideology has been especiallycharacteristic of Michel Foucault’s writings (which will be examined in Part Four).Foucault even suggested abandoning the concepts of ideology and political theoryaltogether. They would be replaced by painstaking genealogical explanation, whichexamines how certain discourses and regimes of truth (epistemes) come about. ForFoucault, all knowledge related to power and domination. As he stated, ‘what oneseeks then is not to know what is true or false, justiﬁed or not justiﬁed, real or illus-ory …One seeks to know what are the ties, what are the connections that can bemarked between mechanisms of coercion and elements of knowledge, what games ofdismissal and support are developed from the one to the others, what it is that enablessome element of knowledge to take up effects of power assigned in a similar systemto a t rue or probable or uncertain or false element, and what it is that enables someprocess of coercion to acquire the form and the justiﬁcation proper to a rational,calculated, technically efﬁcient, and so forth, element’ (Foucault in Schmidt (ed.)1996: 393). Knowledge always conforms to restraints and rules and power also needssomething approximating to knowledge.56Thus, for postmodern-inclined writers,neither political theory nor ideology represent any external objective reality. Ideo-logy and political philosophy, for Foucault, are both subjects for genealogy. We arealways encultered beings who express, contingently, our diverse communal narrativesthrough theory or ideology. There is no external reality, which we can represent.A related dimension to this attack on representation theory concerns the broadtradition of twentieth century purported nonfoundationalism. Although not directlyfocused on this integration thesis, there are a number of the arguments within thistradition (taken as broad category), which facilitate the conceptual linkage betweenpolitical theory and ideology. For example, for nonfoundationalists there are nogivens and no raw data in the world. The idea of an empirical given is a ‘myth’.Further, there is nothing external to our symbolic systems. We live and think inseveral worlds with distinct, often incommensurable systems of symbols. In addition,there is an abandonment of correspondence accounts and a focus on coherence.Statements therefore become true, not by referring to an external given world, butrather in terms of whether they cohere with distinct systems of symbols (see Goodmanand Elgin 1988: 8). In a similar vein, for Richard Rorty, poetic creativity must nowreplace representations of reality; irony and gaming are set over against knowledgeclaims. Rorty summarizes this drift of argument by completely identifying political