International Relations. A self-Study Guide to Theory

constitute each other; each entity explains the other. This type of explanation 
is often called “constitutive”. This idea of “causality”, if taken seriously, is 
very different from the well-known Humean causation. This observation 
leads us to the next criterion for learning about theories of IR: the type or 
model of explanation and the notion of causality.
Before we discuss this criterion, let us sum up the previous one. The onto-
logical as well as the epistemological and methodological assumptions about 
the interrelation of agency and structure will be an important third criterion 
applied to our structured learning process about theories of IR. Each unit will 
ask if the explanation offered by the theory is structuralist or agency-centered 
or, alternatively, if the theory offers an explanation of a different type.
At the end you will know about explanation in terms of structural effects, 
explanation in terms of individual action, and explanation in terms of mutual 
constitution of agency and structure. The next question is about a more spe-
cific topic: the nature of explanation itself. What counts as an explanation of 
international politics? What is valid scientific knowledge about international 
relations? 

---

115 
3.
Epistemology and methodology 
Ontological questions are intimately linked to epistemological problems of 
knowledge and of how we come to gain that knowledge. In most theories, 
these problems basically revolve around explanation and/or understanding, 
what type of knowledge counts as valid knowledge, and how to gain valid 
knowledge. In this book, we will approach such epistemological and method-
ological assumptions by discussing different types of explanations offered by 
different theoretical approaches to IR. How can we best explain international 
politics? What is the nature of the explanation? What reasons can a theory of 
IR give so that an explanation is taken and accepted as “legitimate”, valid 
knowledge about international politics?
These are examples of guiding questions that help outline each theory’s 
specific answer to the core question/problem (Criterion 1) in the respective 
theoretical approaches. Three aspects will be particularly relevant. First, what 
is the interrelation of agency and structure found at the heart of the explana-
tion? This is primarily an ontological problem and has been introduced in the 
previous section through our discussion of the ontological core and the com-
bined epistemological and methodological implications. For systematic rea-
sons, I am mentioning it here again. Second, what is the “nature” or type of 
the explanation and the notion of causality therein? The third aspect will be 
the approach the theory takes to what is known in the social sciences as the 
level of analysis problem, which is in fact an analytical concept rather than a 
philosophy of science criterion. It will be used here in an indirect manner be-
cause it is a well-known “problem” in IR and most theories of IR refer to it. 
This also holds true for theories that reject the concept because of their dif-
ferent understanding of science: they refer to the level of analysis problem to 
clarify their position on science. Learning how different theories of IR ap-
proach the level of analysis problem will offer important insights into their 
general understanding of theory and science. 
3.1.
Type of explanation and causality 
The “type” of explanation is important with regard to a central epistemological 
criterion: how can one come to know something and what counts as valid 
knowledge? It requires taking a closer look at explanation itself – that is, how 
the interrelation between the ontological entities is perceived epistemological-
ly. In fact, when we ask about explanation, we ask about knowledge. What 
counts as valid knowledge of international politics in the discipline of IR? 

---

116 
Usually “knowledge” takes the form of a “causal relation” as it has been de-
scribed for the positivist account of science in Unit 2: law-like regularities that 
can be expressed in terms of if-then temporal sequences and that make predic-
tion possible. Take, for example, the theory of inter-democratic peace; based 
on empirical observation proved by statistics, the findings of the Correlates-of-
War project at the University if Michigan have been formulated into a quasi-
law that “democracies do not fight each other”. This finding can be trans-
formed into the following if-then statement: if a state is a democratic state, 
then the risk of war and conflict with other democratic states will be dimin-
ished. The knowledge has been used to legitimize a politics of democratization 
by the West in such areas as human rights and development policy, based on 
the conviction that democratization will contribute to peace and stability.
However, as you have learned in Units 2 and 3, neither “explanation” nor 
“causality” are fixed scientific terms, but instead depend on the underlying 
understanding of science. A scientific explanation does not necessarily take 
the form of a law-like generalization, nor does “causality” necessarily have to 
be perceived as a sequential relationship of cause and effect with the cause 
temporally preceding the effect. For example, social constructivist theory as 
formulated by Wendt points to a distinction between constitutive and explan-