THE NOTION OF POWER OR WHAT IS LEFT OUT IN ROBERT GILPIN’S ‘TRICHOTOMOUS CATEGORISATION’: THEORETICAL AND PHILOSOPHICAL POSITIONS IN INTERNATIONAL POLITICAL ECONOMY AND THE GROUND FOR THEORETICAL MERGER

The article aims at contributing to philosophical understanding of issues in international political economy (IPE). The philosophical formations of different approaches to conducting research in IPE demonstrate that all its divisions are conditional, with permeable rather than impenetrable boundaries. To simplify introduction, some IPE textbooks follow Robert Gilpin’s “trichotomous categorisation”, or theoretical division of the field into realism, liberalism and Marxism/ structuralism, which fails to clearly define cultural and historical roots, ignores alternative philosophical approaches to IPE, and denies the possibility of theoretical mergers. With the main focus on the notion of power, this article aims at exploring the possible ways to overcome this trichotomous theoretical division, making the attempt to find ground for theoretical merger. The paper concludes that the theoretical boundaries can be transcended by merging different concepts (a “horizontal” merger), or by approaching the same phenomena in different epistemological ways (a “vertical” merger).


Introduction: three traditions in IPE?
T he subject of International Political Economy (IPE) revolves around the question of how politics and economics interact to synthesise knowledge in the field. The study of IPE is not only about what we study, it is also about how we approach our subject, which assumptions we make and what we are seeking to provide: empirical evidence with the possibility to make predictions or an understanding within a cultural, social or historical context . The philosophical formations of different approaches to conducting research in IPE demonstrate that all its divisions are conditional, with permeable rather than impenetrable boundaries. To simplify introduction, some IPE textbooks follow Robert Gilpin's "trichotomous categorisation", or theoretical division of the field into realism, liberalism and Marxism/ structuralism , where realism is sometimes defined as "economic nationalism" [O'Brien, Williams, 2010:9] or called a "paradigm of the politics of international economy relations", or "the statecentred PIER version of IPE", as it falls out of the political economy criteria of the subject [Strange, 1991:40].
This essay has the following structure: the first part consists of the methodological usefulness of the "trichotomous categorisation", followed by the specification of how the concepts of power and hegemony are formed up within realist, liberal and Marxist frameworks, as an example of conceptual difference of these traditions; the second part of the essay focuses on the theoretical and philosophical positions "left out" by the categorisation, and finally, the concluding part comprises the possible ways to overcome the trichotomous theoretical division, making the attempt to find ground for theoretical merger.

What the division of the IPE thought into Realists, Liberals and Marxists/ Structuralists provides us with?
The three perspectives, realist, liberal and Marxist/structuralist (Critical realism), have a foundational ontology that accepts the existence of a reality "out there", independent from our actions or thoughts [Burhill, Linklater, 2009]. The philosophical positions of Realism and Liberalism underpin the positivist epistemological reference to the possibility of science, providing causal explanation and empirically tested knowledge . Critical realist accounts stand on the edge between positivism and interpretivism: while accepting the possibility of scientific research in IPE, not all notions in the real world are claimed to be observable and testable. Ontologically, Gilpin's "trichotomous categorisation" is not only a simplification of a foundational positivist theoretical position, but also a shortcut of human thought in the field of IPE; as it leaves behind alternative anti-foundational interpretivist accounts with their ability to uncover reality from value-laden perspectives . If this division fails to uncover the full spectrum of possible inquiry, why do some basic textbooks follow it? By doing so they structure "scientific" methodology with analytical tools, provided by foundational ontology.
Foundationalism opens up the most important scientific methods of "inferential justification", such as deductivism, "strict inductivism", and "Liberal inductivism" [Poston, 2010]. These tools are widely used in analytical research and in theory formation: "foundationalists have two main projects: providing a suitable theory of noninferential justification and providing an adequate theory of proper inference". The analysis of the philosophical debate over the possibility of science is outside of the scope of this essay, however it is important to acknowledge that all textbooks begin with a reference to the ontological and epistemological positions of different theoretical approaches. Realists, liberalists and Marxists use the regress arguments to build up their epistemological positions but they do it differently from each other, setting up alternative starting points. That is how "trichotomous" division helps us examine the same political or economic phenomena from distinct perspectives, and concentrate on different concepts, combining research methods from various sci-ences: economics, history, psychology and sociology.
The most frequently used methods are "quantitative" (economic) methods, such as statistics, based on the empirical evidence that can be calculated. Another dominant method in IPE is derived from psychology and is based on the explanation of behaviour, built upon the concept of rationality. Both realism and liberalism rest upon behavioralism in their attempt to provide value-free, objective analysis of IPE issues, by making sophisticated categorisations and empirical testing of behaviour . The Rational Choice approach, including game theory with different scenarios, such as the Prisoners' Dilemma, and the Chicken and Stag Hunt among others, aims at generalising knowledge to make predictions . The ability of game theories to make predictions is widely contested as the separation of facts from values withdraws individuals from the context in which they operate .
Rationality in all its occurrences, is instrumental or "bounded" when it doesn't "consider the full range of possible choices" [Hindmoor, 2006:16], but is one of the most important concepts in the study of IPE. For instance, at the state level of analysis, the protectionist policies advanced by some countries, could be viewed as a reflection of the Prisoner's Dilemma, where the actors are trying to avoid the risk of "free riding" [O'Brien, Williams, 2010:33]. Moreover, at the institutional level of inquiry, as in Keohane's theory of international institutions, rationality proves to be a working concept, by showing that the strength of institutions depends on the extent to which their member-states "share social values and have similar political systems" [Keohane, 2002:34], thus suggesting that states are acting rationally when they tent to cooperate with the similar states to eliminate uncertainty and reduce "the costs of making and enforcing agreement" [Keohane, 2002:30].
Proving that theories can have fluid boundaries, the theory of international institutions "share a great deal" with a "sophisticated version of realism", as it inte-grates into analysis what has been left out by Realism: institutions, transnational corporations, domestic politics and the role of three types of beliefs: "principles beliefs", and "causal beliefs" and worldviews in world politics (Keohane, 2002:6). Institutionalism is considered by some to be a methodology in the study of IPE, which stresses the role of formal and informal institutions in shaping individual behaviour. In fact, methods like rational choice and institutionalism can contain theoretical assumptions and be referred to as "theories rather than methods", but the most persuasive argument in this account is that they should be better viewed as methods which can be used by different theories [Burhill, Linklater, 2009:37]. For instance, analytical Marxists fortify historical materialism by grounding logical and linguistic analysis on rational choice methodology [Cohen, 2001].
Critical theorists with their "middle ground" epistemology [Savigny, Marsden, 2011:28] provide us with alternative theoretical tools for the study of IPE, such as Structuralism, with the dominant role of structures in shaping behaviour, or historical materialism, with its explanation of historical development through the lenses of "economic development within society". The emphasis of analysis on production forces, property relations and class struggle represents "an important counter-weight" to realist and liberal views on power relations, enabling the elaboration of "dialectical from of IPE analysis appropriate to the conditions of the late twentieth century" [Gill, 1991:51]. What are the main conceptual divergences between realists, liberalists and Marxists/ structuralists?

Power and hegemony from realist, liberal and Marxist/structuralist perspectives
"The aim in IPE is to look for the sources of power and the consequences of the exercise of power" [Strange, 1991:48]. According to the theoretical division, the concept of power is located differently in the realist, liberal and Marxists/structuralists' traditions. For Critical realists, "power is posi-tioned structurally", systematically and is not always directly observable, but could be conceptualised through its consequences [Savigny, Marsden, 2011:79]. In contrast, Liberals and Realists, with their positivist ontology, suggest that power is observable, but locate it differently.
For realists, power is attributed to states, the key selfish and rational actors in the anarchic political arena, where conflict is an inherent element and self-defence is the mode of survival. Realists acknowledge that states should try to acquire "relative strength" rather than absolute superiority [Waltz, 2010:168]. Most neo-realists agree that military polarity no longer fully describes the existing world order, as globalisation imposes new threats on states and their strength depends on the "distribution of capabilities", diplomatic skills or "balancing", economic, geographical and cooperative positioning within international society [Waltz, 2010:126]. Globalisation makes diplomatic tasks more complex: pressuring states not only to ally with existing or rising superpowers but also balance against their further empowerment, as "states ally to balance against threat rather than power alone" [Walt, 1990:5].
Hegemonic power is essential to maintain the existing world order, where liberalisation is more likely to take place with the assistance of a superpower, as "openness is most likely to occur during periods when a hegemonic state is in its ascendancy" [Krasner, 1976:323]. The realist theory attributes the hegemonic power to economic, political and military dominance, and pays scrutinising attention to the "theory of hegemonic stability" and the modelling of the rise and fall of superpowers, such as the United States [Gilpin, 2001:93], suggesting that bipolarity, as in the Cold War, could contribute to stability .
The motive for cooperation is one of the key distinctions between realism and liberalism: for Realists the reason for cooperation lies in maximisation of power, for liberals in mutual benefit or "positive sum" gain [Burchil, 2009]. For liberals, power relations developed domestically and internationally are interconnected, whilst realists view do-mestic and international politics as separate due to their distinct natures, i.e. hierarchical and anarchical [Burchil, 2009]. Developing the philosophical worldview of Immanuel Kant, with the notion of "Perpetual Peace", liberals locate power in structures and institutions that provide incentive for cooperation among republican states, with the state receiving the function of upholding the rules of the game, including property and contract law [Kant, 1795]. The state is an important actor, but in an era of growing economic, strategic, environmental and ideational interdependence, it is no longer unitary in the analysis, rendering its role to the international institutions [Keohane and Nye, 1977;. "The notion of transnationalism" and the way it transforms power relations among states is a challenge to the Realist approach to power relations [Cohen, 2007:202].
The liberal concept of power relations and commitment to free trade, with globalisation seen as a positive process which could eventually move us closer to "perpetual peace" among states, lies in the centre of attack by critical theorists. Marxists blame capitalism, and neo-imperialism as its globalised form, for creating inequalities through uneven accumulation of surplus value [Peet, Hartwick, 2009]. For most critical accounts, power resides in the "attendant" power structures. Kant's notions of freedom and equality are castigated as the "false consciousness", the hegemonic power and ideological brainwashing by the ruling class to preserve existing power relations through mass consent [Morton, 2007]. The further development of the Marxist criticism of liberalism is found in neo-Marxist theories such as dependency theory, world system theory and regulation theory. The Gramscian concept of the hegemonic power of ideas is a further theoretical development of the Marxist notions of social control through physical domination and ideological hegemony [Morton, 2007].
Capitalist class ideology legalises and normalises the interests of the ruling classes. The possibility of challenging the existing status quo attributed to the counter-hegemonic struggle at the international level in an effort to generate a new power block, moves national inter-class power relations onto an international level [Morton, 2007]. International organisations are the catalysts of the ideological dominance of capitalism at the international level, as they reflect the interests of core countries. The term rationality ceases to have the positivist ground, it is considered to be an "interpretative structure of thought" which is attributed to a "specific social group", with the understanding that "what appears as irrational to one rationality can, however, be quite rational to another" [Cox, 1987:26].
As we can see, the "trichotomous" division is useful in structuring different approaches to power relations in IPE: the Realist understanding of power is narrow, "one-dimensional", and reduced to state power maximisation. For liberals, the behaviour and the role of political actors in decision-making is central, but is still focused on observable conflict, including, however, the second dimension, or facet of power, the "mobilization of bias", or the power to set political agenda . Critical theorists manage to reach the "third dimension" of power, addressing a latent conflict by "manipulation and authority" aimed at "misleading" individuals by "false" consensus [Lukes, 1974:23].

What is left out?
There are three positions on which we could disagree with this division: the absence of clear boundaries of the historical roots of the three theories, the ignorance of alternative philosophical approaches, and the denial of the possibility of theoretical mergers due to the strict categorising of their epistemological positions. It is important to closer examine these arguments. The first deficiency is that the specification of the historical roots is questionable, which makes the positioning of the contemporary theories also questionable, as under a "close textual scrutiny", they lose their historical "anchoring" and no longer seem to represent a "clear-cut" division into Realism, Liberalism and Marxism [Watson, 2011:48] There are vivid examples of misinterpreta-tions and false positioning of some texts and authors within a certain theoretical framework.
Firstly, many prominent critical realists, like Robert Gilpin, Stephen Gill and David Law, attribute Friedrich List's nationalism to be a predecessor of modern-day realism, whereas there are "clear elements of antirealism" in List's position, deriving from his long-run national interests dominance over "short-run economic bargaining power" [Watson, 2011:34]. Secondly, the historical roots of liberalism within IPE are also "poorly specified", with Smith being placed within the liberal tradition, while his work contains "clear anti-liberal arguments": while considered to be a supporter of laissez-faire, in reality his analysis is much more complicated and aims at finding out weak points in the free market philosophy, especially in the point when he draws his critiques over the dehumanising effects of the capitalist production system. In this respect Smith's approach could be referred to as critical rather than Liberal, as though taking a different path, he arrives at the same conclusion as Marx: Smith's division of labour or Marx's reproduction of the capitalist system involves a high "psychological costs" paid by individuals routinized in the factory work, which, however, for Smiths (contrary to Marx) could be compensated for by "statefinanced education" [Watson, 2011:43-44].
The second deficiency of the trichotomous categorisation is that it no longer represents the "new IPE", as there is so much left out: the way that interests are constructed, the way "policies are subject to the processes of international diffusion", and the transformation of world politics [Keohane, 2009:34]. In fact, those who try to fit IPE into this narrow division, or those who sink in theoretical conceptualisation suffer from "myopia when it comes to the world around them" [Strange, 1994:209]. The "new IPE" theorists face the problem of unequivocally attributing themselves to any of the strands offered by the methodological division. For instance, Susan Strange is opposed to being called a theoretician at all, Robert Keohane acknowledges that the Rationalist framework he developed in "Ideas and For-eign Policy" (1993) was "closer to what has come to be called constructivism than to the materialistic rationalism of the OEP (open economy politics) orthodoxy" [Keohane, 2009:38].
Even the prominent realist Robert Gilpin, though denying the basic tenets of constructivism, which are that (1) shared ideas rather than material forces determine human structures and (2) the product of these shared ideas are the interests and identities of human beings [Gilpin, 2001:19], suggests that constructivism should be taken into account according to its "increasingly influential position", and that any theory should "seek to integrate both ideas and material forces" [Gilpin, 2001:20]. The interpretivist accounts, though acknowledged by many authors, are still neglected by some IPE textbooks. Further diversion from the orthodox theoretical tradition could be found in Poststructuralism, a philosophical stream that goes beyond positivism and Structuralism, denying the precedence of the structure over agency, focusing on the deconstruction of the traditional discourses in order to reveal what is omitted and covered in subtext [Foucault, 1979;. This interpretivist account, by denying a single objective truth, opens up the possibility to discover alternative "truths", locating power in any substance and contesting the orthodox understanding of reality.
The last deficiency covers the positioning of theories as mutually exclusive, with little possibility of merging. Various theories focus on different aspects of IPE, or investigate the same phenomena from distinct perspectives. Therefore they could be used as tools for problematising the issues of IPE, combining both foundational and interpretivist accounts, linking qualitative and quantitative methods to provide deeper understanding. Thus a theoretical merger, or "methodological triangulation", has a potential for conceptual innovation [Tashakkori, Teddlie, 1998]. The necessity of transcendence beyond the boundaries of orthodox methodologies is dictated by the structural changes and complexities of contemporary IPE issues. Susan Strange first outlined the problem of logical inconsistency between contemporary narrow-state-minded realism and self-exclusive political and economic liberalism [Strange, 1994:216].
The contrast between positivism and empiricism with its deductive logic on one side, and the institutional and historical analysis which accepts interpretivist accounts on another, was intelligibly described by Benjamin Cohen in his "transatlantic divide" between American and British IPE. The pragmatic, quantitative and "increasingly standardised" American approach was juxtaposed with the less abstract British school, the integral part of which is intellectual openness, normative "ambitiousness" and critical inclination. In effect, the two schools represent the "two cultures divided by the common subject" [Cohen, 2007:200]. Most in academia acknowledge that the "mega-trends" of the "quantitative-Liberal" American school, such as the application of neoclassical economic analysis to politics, international relations and other social sciences, contains "model bias", such as the inadequacy of "self-adjusting systems", and "functionalist explanation", constructed with the aim of preserving "the virtues of American hegemony" [Wade, 2009:106-121]. Therefore, in the future the American school might undergo conceptual changes, a revolutionary transformation, bypassing the "transatlantic" boundaries of the methodological division.

Concluding the possibility of a theoretical merger
Conceivably, these theoretical boundaries could be transcended in two directions: horizontally and vertically. A horizontal merger brings cumulative effects or dialectical synergy on the conceptual basis: it presupposes the inclusion of a concept introduced by one theoretical tradition into the analysis leading onto another ontological ground. Susan Strange's eclectic approach, the two-level game theory or Keohane's institutionalism could serve as examples of such a merger. The vertical merger implies the possibility of transcending the boundaries set by various philosophical positions when analysing the same concepts or phenomena. The goal of this merger is to integrate different episte-mological frameworks outlined by separate ontologies in the same field of inquiry, with the aim of providing broader understanding. For instance, a vertical merger represents the possibility on the ground of the role of ideas in IPE: between Gramscian concepts of ideological hegemony and the constructivists' contingent nature of ideas, or on the basis of gendered identities: between liberal, critical and poststructural feminism.

"Horizontal" merger
An eclectic approach implies the possibility of transcending the orthodox realist and liberal approach by focusing on "two exogenous factors of technology and markets" as forces for structural change, other than power relations of states [Strange, 1991:39]. The three stages of an eclectic analytical framework, from the acceptance of the insights of other disciplines, through "conceptual synthesis" to the recognition of the role of the enterprise, brings to the core three "dimension[s] of diplomacy", and ends the "mutual isolation of the three standard paradigms of international relations, which so far had produced only a dialogue of the deaf" [Strange, 1991:34]. The best example of the rising influence of firms on the polices of states would be the decision-making process in petro-states, such as Kazakhstan, where "state-firm" and "firm-firm" diplomacy proceed in close correlation with "statestate" foreign affairs. Importantly, the forces for structural change are not equally empowered in every situation, and depend on the historical context, which makes these forces more or less dynamic relative to each other.
The force, or exogenous factor, which is more dynamic in a particular situation, will determine the direction of the major structural change. Strange's structuralism, with the power residing in the "key structures" of security, production, finance and knowledge, as well as in the "secondary structures" of transport, trade, energy and finance [Strange, 1988], transcends not only the realists' statism, liberal institutionalism, but also the Marxist' class-based or neo-Gramscian production-based Structural-ism. However, an eclectic approach takes its roots from critical theories, and epistemologically has features of critical realism, neoconstructivism and neo-Gramscianism, in the sense that it also "seeks to overcome the positivist distinction between the subject and the object of analysis" [Talani, 2004:56]. As Strange puts it, the "notion of enveloping structures" developed from Marxist, or in Robert Cox's case, Gramscian, perceptions of the importance in IPE of the relations of production as an important determinant of power relations" [Strange, 1991:34].
It is true that Strange's focus on the structure that production takes is partly rooted in the Cox's analysis of social forces in the formation of the state's policies, which is to "achieve dominance over coexisting modes" and determine "the whole complex structure of production" from which to "extract sufficient resources to continue to exercise its power" [Cox, 1987:399]. Strange, however, goes beyond the limits of Cox's conclusion that the power relation stems "only from the relation of production" [Talani, 2004:58] and that the crucial role in IPE is played solely by states. She introduces the notion of a "trilateral system of transnational diplomacy", where governments hold the same bargaining power as firms, especially transnational corporations, and that sometimes important decisions are taken between TNCs without governmental interference, therefore the nation-states are no longer the sole "managers of national economies" [Strange, 1994:212].
Another example of the theoretical merger is the two-level approach, or "metaphor for domestic-international interactions" [Putnam, 1988:433], which cannot be fully attributed to any of the theories suggested by the division as it encapsulates the features of all three. The supposition that domestic politics determines international relations could find some resemblance in Marxist, or more precisely, Lenin's, critiques of capitalism, arguing that imperialism is the application of capitalism from the domestic level onto the international level, when the domestic resources of capitalist development are exhausted and the state needs external sources of capital and power ac-cumulation. Realist features can be seen in domestic and international "conflict behaviour", which proves that in some situations, before the formation of regional integration, the state remains the "central decisionmaker", when "national governments seek to maximise their own ability to satisfy domestic pressures" [Putnam, 1988:432-434]. The liberal position finds its reflection in the "synergistic linkage" of political "indifference curves" at Level I that affects Level II [Putnam, 1988:447].

"Vertical" merger
The vertical merger of theories of different ontologies is possible where they are not mutually exclusive, where they do not oppose, but complement each other. The possibility of such a merger could be developed from Wendt's supposition that constructivists, neo-realist and Marxists are all structuralists, as they all emphasise the impact of structures on identities and interests [Wendt, 1995;Reus-Smit, 2011]. The difference between neo-realism and neo-Marxism with its rational ontology, on one side, and constructivism with its anti-foundational background, on another, leads to disagreement upon the question of the structure's nature: "neo-Realists think it is made of material capabilities, whereas constructivists think it is also made of social relationships, where systems of ideas, beliefs and values have structural nature" [Wendt, 1995:73].
Another disagreement is over the formation of interests concerning social inter-action: whether they are exogenous to it, as rationalism suggests, or endogenous as the Constructivism proposes [Reus-Smit, 2011:216]. The constructivists understanding of society as a constitutive realm which produces knowledgeable social and political actors could be combined with the rationalist view of society as a "strategic domain", a place where actors rationally pursue their interests [Reus-Smit, 2011:216]. Interestingly, the theoretical merger between these contradictory philosophical positions is seen as possible by constructivists while being denied by realists. Some constructivists believe that "productive engagement" with rationalism could be pursued on the grounds of the dynamism of interests in IPE, from "interest formation" by institutionalised norms to "interest satisfaction" through strategic action :20, cited in Reus-Smit, 2011.
To connect the two approaches, Constructivists would have to undertake the task of explaining how actors come by their Diagram 1. An Example of Horizontal merger: An Eclectic Approach preferences while Rationalists explore how they exercise them. Therefore, constructivism is seen as a "complementary" rather than rival theoretical perspective to rationalism :20, cited in Reus-Smit, 2011. To analyse the relation between forces in society we could combine the neo-Gramscian concept of unobservable hegemony of "production and reproduction of knowledge" [Cox, 1987:39] with the analysis of institutions and norms which shape social identities to generate their hegemonic power. Attributing Gramsci to solely Marxism/structuralism would be a false interpretation, as his philosophy transcends the boundaries of theoretical division. We could go further and claim that neo-Gramscian accounts could be combined with the Poststructuralist model of IPE to present capitalist hegemony as the outcome of a political discourse rather than a structure [Gibson-Graham, 1996, cited in Savigny, Marsden, 2011.

Diagram 2. An Example of Vertical merger: Interest formation
The division of historical and contemporary thought in the field of IPE into realist, liberal and Marxist/structuralist traditions fails to clearly define cultural and historical roots, ignores alternative philosophical approaches to IPE, and denies the possibility of theoretical mergers. The "trichotomous" categorisation, or any broader theoretical one, including anti-foundational accounts, can be very useful when we study IPE, as it introduces different analytical tools, enabling us to structure knowledge. However, it offers little help when the aim is to con-duct research, when we need to make our own assumptions. Our ability to understand reality has been widely contested and lies at the centre of philosophical debate among a plethora of social scientists. The possibility to transcend theoretical boundaries by merging different concepts (a "horizontal" merger), or by approaching the same phenomena in different epistemological ways (a "vertical" merger), represents a valuable contribution to the development of, where possible, scientific knowledge, and the philosophical understanding of issues in IPE.