Rethinking Logic of Inference and Explanation in the Field of International Relations
| Date | 01 October 2012 |
| Author | Yong-Soo Eun |
| Published date | 01 October 2012 |
| DOI | 10.1111/j.1467-9256.2012.01437.x |
| Subject Matter | Research Article |
bs_bs_banner
P O L I T I C S : 2 0 1 2 V O L 3 2 ( 3 ) , 1 6 2 – 1 7 4
doi: 10.1111/j.1467-9256.2012.01437.x
Research Article
Rethinking Logic of Inference and
Explanation in the Field of
International Relationsponl_1437162..174
Yong-Soo Eun
University of Incheon
This article provides an investigation of the conventional logic of inference that underlies the search
for a causal explanation of social and political behaviour, including state behaviour, in world
politics. It proposes an alternative reasoning strategy which can complement the traditional (induc-
tive and deductive) logic of inference. In addition, the article enters into a discussion of how the
proposed alternative can make theoretical contributions to obtaining a better understanding of the
complex reality of world politics.
Keywords: explanation; international relations; methodology; state behaviour; theory
Introduction
When Giovanni Sartori wrote his seminal piece on methodology in the study of
politics, his ‘underlying complaint’ was that ‘political scientists eminently lack ... a
training in logic – indeed in elementary logic’ (Sartori, 1970, p. 1033). In his words,
‘most of the literature introduced by the title “Methods” (in the social, behavioral
or political sciences) actually deals with survey techniques and social statistics, and
has little if anything to share with the crucial concern of “methodology”’.
Although research methods and techniques, such as applied statistics, have become
more sophisticated since the 1970s, there is still a relative lack of effort in advancing
‘methodology’, logical structure and procedures of scientific inquiry. Most of the
contemporary literature dealing with research methods in politics is – with excep-
tions1 – not fully engaged in discussions about the core aspects of methodology:
logics of inference and approaches to theory development. Instead, just like tradi-
tional work on political science research methods, the newer literature concentrates
more on research ‘techniques’ than on principles and rationales – or what Sartori
called ‘logos’ – which underlie these ‘techniques’.
By presenting a (potentially) useful alternative to conventional approaches to
inference and theory development, this article attempts to advance methodology in
the study of international relations. The article begins by critically examining two
distinctive logics of inference, namely inductive logic and deductive logic, which
underlie our search for an answer to questions concerning the causation of social
© 2012 The Author. Politics © 2012 Political Studies Association
R E T H I N K I N G L O G I C O F I N F E R E N C E I N I R
163
and political behaviour, including the state’s behaviour in world politics. In the
following section, an alternative approach to reasoning, which can complement the
conventional logics of inference, is proposed. Finally, how this alternative can make
theoretical contributions to obtaining a better understanding of the complex reality
of world politics is discussed.
Modes of inference and explanation: rejecting the
prioritisation of either inductive or deductive logic
First of all, I argue that any analyst should eschew the so-called ‘narrow’ inductive
logic of inference, associated with an empiricist epistemology. It holds that on the
basis of direct observation and ‘sense experience’ we can produce objective findings
and develop generalisations about causal relationships between social phenomena
(Sander, 2002, pp. 48–52). In particular, empiricism – which has strongly influenced
behaviouralists’ methodological and/or epistemological positions and techniques –
claims that theoretical understanding can be obtained through a process of inquiry
that begins with ‘theory-free observation of all the facts up to now’ (Sander, 2002,
p. 52). In David Hume’s words, ‘all our ideas are nothing but copies of our [sense]
impressions ... the form and content of scientific law wholly derive from sense
experience’ (Hume, 1748, quoted in Losee, 2001, pp. 93–94). It is suggested that
‘scientific’ knowledge emerges only with the collection of observable data; and that
the collections of sufficient observable data, it is presumed, will lead to the identifi-
cation of patterns that will in turn allow the formulation of laws (see Smith, Booth
and Zalewski, 1996). Thus the ‘narrow’ inductive approach to social and political
analysis, based on such logical empiricism, overly emphasises observable empirical
data and correspondingly downgrades a priori theoretical reasoning. For strict
empiricists and behaviouralists direct observation can serve as an independent tool
for developing both objective and general knowledge.
Although such a narrow perspective is most noticeably found in ethnography or
behavioural anthropology (see, e.g., Graves, 2002; Metcalf, 2005) rather than in
the field of international relations (IR), and many ‘Western’ IR scholars would be
reluctant to subscribe to that perspective, it nonetheless still has a profound effect
on ‘area specialists’, particularly when it comes to the lively debate on ‘area studies
versus disciplinary studies’ in political science (see Fukuyama, 2004; Johnson,
1997; Johnson and Ijiri, 2005; Katzenstein, 2001; Pye, 2001). Peter Katzenstein
(2005, p. xi) explains: ‘Those who construct abstract theories dream of a future in
which polities all over the world are alike; academics in the field of area studies
dream of a past when each locale in the world was largely isolated’. In this debate
Shiping Tang (Tang, Li and Acharya, 2009, p. 3), a renowned Chinese scholar of
Asia and international relations, comments:
‘Contrary to the increasingly deductive approach in the United States and the
grand theorizing approach in Europe that have traditionally paid less attention to
empirical facts, we believe that empirical facts provide the ultimate foundation
for theorizing about international politics’.
From this perspective, the scientific method for acquiring valid knowledge is char-
acterised by induction – the movement from empirical observation and experiment
© 2012 The Author. Politics © 2012 Political Studies Association
POLITICS: 2012 VOL 32(3)
164
Y O N G - S O O E U N
to scientific law. As such, theory is treated as just a ‘language for recording exhibited
regularities ... [and thus a priori] theoretical assumptions are not required’ (Hay,
2002, p. 29).
I believe, however, that this ‘narrow’ inductive approach to political inquiry – that is,
prioritising ‘empirical facts’ and thus starting from a blank sheet without explicit
theoretical assumptions, and filling it anew with a maze of area (empirical) data,
hoping that it leads to something interesting and general – should be seen as
problematic, because prioritisation of being empirical and an appeal to pure experi-
ence is too ambiguous and/or ambitious: ‘experience’ can refer to both ‘what is
presented to us’ and the actual act of ‘experiencing’ (Hollis, 2002, p. 71). Besides,
even if we define facts as the actual act of experiencing, I strongly doubt that it could
ever be possible for any analyst to observe all the relevant facts and evidence. What
is more, there are many key factors – such as perceptions or beliefs – which cannot
be observed directly, but may exert causal power on the actions of individuals and (by
extension) states. Nonetheless, from the point of view of the inductive logic favoured
by strict empiricists and ‘area specialists’, any non-observation-based statements
could be rejected as ‘meaningless’ (Sander, 2002, p. 46) or ‘mere fictions’ (Patomäki
and Wight, 2000, p. 217). Such an epistemological stance is problematic in terms of
its epistemic range and nuances. In this vein, I consider that any analyst ought to
make deductive inferences derived from specific observations of relevant factors
(arbitrarily) chosen on the basis of their own explanatory expectations and theoreti-
cal assumptions. In other words, the logic of the deductive approach should be
employed in the first place in world political analysis.
However, this by no means suggests that we follow a strict and formal form of
deductive logic, often related to ‘deductive-nomological’ modelling: that is, first, a
general law is postulated; second, antecedent conditions are specified; and third, the
explanation of the observed event is deduced from the first and the second steps
(for a good overview of this modelling, see Hollis and Smith, 1990, pp. 35–45). Such
a strict form of deductive logic is essentially a mirror image of the ‘narrow inductive’
approach discussed above. The reason that we should not adhere to it in the
analysis of world politics is, then, rather clear: it is also too ambitious and ambigu-
ous. Stated differently, commencing by deriving generalisable and predictive
hypotheses from simple assumptions about the world is problematic in the study of
world politics where complexity, intricacy and variety are pervasive. Nevertheless
many IR scholars privilege the broad applicability and predictive and parsimonious
capacity of the theory: they claim that a good theory must ‘generalize across cases,
events, incidents and time frames’ pursuing ‘parsimony’ (Walt, 2005, pp. 23–28; see
also Przeworski and Teune, 1982; Waltz, 1979). Gary King, Robert Keohane and
Sidney Verba’s Designing Social Inquiry – which ‘has strongly influenced the methods
of study of many contemporary liberal, realist and even constructivist theorists in
IR’ (Kurki, 2007, p. 361) – also makes clear that...
To continue reading
Request your trial