Kevin Thievon
What were the causal mechanisms that led to the outbreak of the Iran-Iraq war in 1980? A realist lens would emphasise the power maximisation that the Iraqi president, Saddam Hussein, sought to become the regional hegemon towards the nascent threat represented by the new Islamic Republic of Iran; a liberal scholar would first point out the institutional failures – e.g. the little influence of the United Nations’ resolution 479 –, or the economic motivations of Saddam, such as getting full sovereignty over the strategic waterway of Shatt al-Arab; and a constructivist one would put forward the role of the idea according to which the majoritarian Shi’i population in Iraq could be seduced by the 1979’s Islamic Revolution, and then turned against the Sunni regime of Saddam.[1],[2]
This example is useful to outline the extreme difficulty to bring out a law that would explain an International Relations (IR) phenomenon – here, the invasion of Iran by Iraq. Were there objective patterns that could have helped to predict this conflict? The essay’s title is a worthwhile question given the great stake of drawing out laws from reality. A law could be defined as the ‘mechanistic processes that bring about standardised outcomes.’[3] And, precisely, a ‘nomothetic enterprise’ aims at exploring those ‘processes’[4] – nomos, in ancient Greek, signifies laws. Nomothetic knowledge would thus be constituted of verified large-scale social patterns that compose the reality of international politics, this so-called reality being a complex blend of universal laws.
The essay firstly explores the extent to which political science – that encompasses IR – is ontologically more likely to produce nomothetic knowledge than a close discipline: history. This comparison helps to argue that IR seemingly requires nomothetic outcomes to be policy relevant.[5] Secondly, it appears that some conditions are needed for specific outcomes to become “nomothetic.” Here, the essay starts to underline the difficulties of producing law-like statements in political science, as hinted at by the Iran-Iraq war above. The third section goes deeper in highlighting the profound subjectivity of knowledge to which IR, as a discipline, seems to lead. An analysis of the dichotomy positivism/interpretivism enlightens this subjectivity as a feature that restrains the advent of nomothetic knowledge. Finally, the essay argues that absolute nomothetic knowledge is impossible in IR; only a conditional one is achievable, depending on the definition that one gives to nomothetic.
The raison d’être of the IR discipline seems to require nomothetic outcomes
A deepening of the notion of nomothetic knowledge is necessary before moving to the comparison with history. A nomothetic approach proceeds from ‘precise measurement, prediction and […] investigations of large groups’ that allow ‘generalisation’ about large social patterns.[6] This implies a “scientific” methodology composed of quantitative methods such as statistical analysis and large-scale observations – which is the opposite of an idiographic approach that focuses on a narrower and single subject.[7] Some scholars have argued that within social sciences, while historians would be working on single events, ‘political scientists generalise about the relationships between variables and construct law-like statements about social behaviour.’[8] Indeed, Jack S. Levy qualified them as ‘nomothetically oriented social scientists.’[9] One of his argument refers to academic graduate programs. For students in political science, they would be more focused on methodology than their counterparts in historical programs.[10] The topic proposed for this essay goes in favour of this argument.
This idea of different approaches within social sciences is supported by the professor Bueno de Mesquita who contends that the political scientist ‘is more likely to emphasise general explanations of social phenomena, while the historian is more likely to emphasise particularistic, unique features of individual episodes of social phenomena.’[11] The reason of this difference could be found in the nature of the research outcomes: political scientists, and particularly IR scholars, would be supposed to be policy-relevant whereas historians would be bound to ‘understand an event as unique rather than typical.’[12] This would be due to the fact that history does not aim at proposing policies. Conversely, the ultimate goal of IR is to be both scientifically rigorous and socially and politically valuable.[13] That is why Suganami declares that ‘the nomothetic orientation in IR is rooted in the […] ethos of seeking knowledge useful for the management of social affairs.’[14] Ontologically, the IR discipline’s raison d’être seems then to be the production of nomothetic knowledge – namely, law-like outcomes based on scientific methodology – ultimately meant for decision-makers.
Nevertheless, in 2018, a study about the methodology used in IR stipulates that 85% of the researchers interviewed in the world proceed through qualitative approaches – 60% use them as their primary methodology.[15] Then, if the majority of IR scholars disregard quantitative approaches as their primary methodology, and given that a generalisation – a law-like – is apparently drawn from scientific methodology as seen above, how can nomothetic knowledge be actually produced in IR?
Conditions of possibility: the epistémè
It seems that nomothetic knowledge necessitates conditions of possibility. The approach of Foucault is interesting here since he correlates the condition of possibility of knowledge with history. More precisely, he asserts that knowledge, to be ‘possible to say true or false,’ needs to be contextualised through an ‘apparatus’ that he called epistémè.[16]Simply put, to produce a nomothetic discourse on a situation, it is required to incorporate the “way of thinking” – that is, the historical context, or the ‘historical a priori’[17]– of the community within which this discourse is produced. So, to Foucault, a pure nomothetic discourse on reality – included scientific ones – would be impossible given all the cultural-political-historical determinations of the law-like designer. Although this concept of epistémè might face shortcomings to illustrate this argumentation, the idea it proposes still clarifies the conditional feature of nomothetic knowledge. Going back to Levy, who argues that ‘generalisations of political scientists are limited to a domain defined by the analytical assumptions of the theory,’[18] the idea of an apparatus that would be necessary to contextualise a generalisation is therefore reinforced.
At this stage of the essay, and given that some political scientists seem to recognise generalisations as more conditional than universal,[19] the nature of these conditions of possibility should be further examined. To be robust and large-scale, generalisations ‘require a single, well-specified, and integrated theoretical structure and validation over an empirical domain that is carefully selected.’[20] In other words, a law-like must be drawn upon a theory empirically relevant and conscious of its limitations. Here comes a paradox: the wider a theory is, the closer nomothetic knowledge is getting – because it is generalised –, and the less robust and credible this knowledge becomes. But what if we correlate “sub-generalisations” – i.e. by domain of expertise – together? Could this render possible a wide theory without losing credibility? An answer to that may be found with the critical theorist Horkheimer. Indeed, he contends that in social sciences, ‘transdisciplinarity’ is a necessary condition to reach the most objective knowledge possible.[21]
In IR, this transdisciplinarity condition appears more and more as a leitmotif to avoid what could be called a “niche law-like.” Indeed, to generate the robust and contextualised generalisations mentioned above, IR scholars should depart from what Van Evera calls the ‘cult of the irrelevant.’[22] What he denounces here is the pitfall that some IR scholars encounter while being stuck in their niche-like research – that is, empirically irrelevant such as endless theoretical debates : ‘organizing the social sciences around disciplines rather than problems […] distracts social scientists from addressing the problems of the real world.’[23] This led to the trend in American universities to adopt a problem-oriented approach as evidenced by the multiplication of gender studies or postcolonial studies.[24] Natural sciences, deemed as more inclined to produce law-like statements, could ‘show the way’ to social sciences since they are already ‘organized around solving problems’ – e.g. Ecological Engineering is focusing on sustainable agriculture and permaforestry.[25] Finally, Horkheimer and Foucault are arguably among the emblematic figures of social sciences that, from different paradigms, have contested the possibility to grasp the reality of social facts in an absolute manner. Therefore, they both argue that, epistemologically, positivism is misleading – because a law-like is conditional.[26] This leads the essay to a core debate within the IR discipline itself: positivism against interpretivism. To facilitate the discussion, the essay will link with the latter both constructivist and subjectivist approaches.
The limits of positivism
Following the arguments above, pure nomothetic knowledge seems impossible to reach given the insatiable need of context, apparatus, and any other condition of possibility. To go further in this demonstration, it is indispensable to examine the positivist paradigm in IR. Indeed, it ‘still represents the dominant approach to political science in the United States.’[27] What seems important to understand is that, for positivists, it does not really matter what one believes about social facts because patterns of behaviour, with sufficient repeated instances, are enough to build objective laws.[28] For example, a positivist realist in the 1970s would have surely argued that the invasion of Iran by Iraq was inevitable given that all states are self-interested and seek to maximise their power when there is an opportunity. Nonetheless, what has been demonstrated so far is that an absolute generalisation – that is, an unconditional law – makes no sense in social sciences. The inherent risk of positivism in IR is therefore to substitute ‘statistical generalisations for laws.’[29] Even positivist scholars, who assert that general laws are achievable, talk from somewhere, from a paradigm, from an historical and social context – that they probably hardly ignore.
This is what makes interpretivism an attractive approach in this epistemic debate: ‘interpretivists deny the feasibility of objective theories of social behaviour.’[30] For them, unlike positivists, what counts is the meaning, not the law, it is to understand, not to explain.[31] Interestingly, understand does not imply a normative purpose, in contrast to explain. Indeed, ‘science, especially social science, is embedded in a social context and often serves to legitimate and uphold the power structure of which it is a part.’[32] Legitimate and uphold: this is the normative purpose. To illustrate that, Lebow is helpful when he argues that, in the United States, the current realist discourse’s role is to justify the American power.[33] As a consequence, if nomothetic knowledge means conditional knowledge which falls within a multidimensional context – in reference to ‘transdisciplinarity’ –, and which aims at understanding instead of explaining or justifying, then this definition becomes arguably relevant.
Thus, IR research seems to be compelled to a form of subjectivity because of a constantly evolving reality. Constructivists, who claim that what should be studied are social facts such as beliefs or identities, also argue that ‘there is no such thing as a balance of power, a social class or a tolerant society; […] positivists make a category error when they equate them with features of the world.’[34] Indeed, Lebow takes the example of the balance of power that became effective only with the performative discourse of leaders talking about it – and with the nascence of states. Henceforth, it is important to notice that social facts, constructed by humans, are evolutive – unlike physical realities such as molecules that were here before humans. In short, what is considered as a social fact by the constructivists is arguably something that has appeared with human life, and which varies along successive social constructs. That is why human phenomena are evolutive and, thereby, in contradiction with the idea of an irremovable fundamental law.
Conclusion
It has been initially argued that the raison d’être of IR is presumably to provide nomothetic knowledge, given the necessity for this discipline to be policy-relevant. Thus, the scientificity of IR research – somewhat synonym of “objective laws” – is claimed so based on its quantitative approaches. However, this essay tried to contest positivist scholars who pretend to produce general laws in IR, for the following reason: reality is entangled in a historical, political, and social context. Consequently, to become “nomothetic knowledge,” the essay attempted to demonstrate that the latter must be understood as conditional and evolutive. Thus, this relative nomothetic knowledge would not be oxymoric, but necessary to avoid conducting IR research as a demiurge. IR scholars should try more to understand reality the best they can than to pretend to explain it with laws once and for all. To that end, it might be worthwhile to study the distinction between understand and explain in more depth while confronting it to a specific IR phenomenon.
No comments:
Post a Comment