Thoughts on the science, theory and philosophy of politics, Part II

In the first part of this series I distinguished between political and apolitical domains of social inquiry, separated by whether or not power is taken into consideration. I then discussed the contrast between political science and apolitical science, followed by the relationship between political science and political theory. I indicated that the task of political theory is to formulate concepts and hypothetical relationships that political science can test empirically. In this second part I will discuss the difference between political and apolitical theory. Following the introduction to the previous essay, “apolitical theory” can be read as “social theory which disregards power”.

In the book A Measure for Measures Ray Pawson stipulates five rules for empirically verifiable social theory (Pawson 1989 p.324-325). They are

  1. sociological theory takes the form of comparison of the probabilities of certain types of action in certain social groupings
  2. empirical hypotheses must pay attention to regularities, mechanisms and contexts
  3. empirical testing is most powerful in those disciplines employing formal networks of co-ordinated explanation
  4. empirical evidence is adjudicatory rather than verificatory
  5. data construction is irretrievably social

These rules indicate the rationale of meaningful social science. Although exact measurements are impossible, relative comparisons can be performed. Good theoretical hypotheses explain why one state of affairs is more probable than another.

Political theory ↔ Apolitical theory

Comparing political theory with apolitical theory, no differences come to my mind which would separate the two as far as rules 1-2 and 4-5 are concerned. But there is a very specific difference which is relevant to rule 3. It stems from the fact that political actions take place in a social environment stipulated by laws. In my previous essay I defined political action as action influenced by power, but it may equally well be defined as action influenced by law. The legal definition is in fact more precise since it excludes acts of coercion between individual citizens, acts that certainly aren’t political even though power, in the form of bodily strength or weaponry, may be exercised and obeyed. But the state exercises power through its laws. Citizens’ behaviour is guided by this power insofar as they conform to the law. It’s therefore clear that the line between political and apolitical phenomena is identical to the line between phenomena that are influenced by law and ones that are not.

The formal network of concepts which Pawson mentions in rule 3 is described at greater length earlier in the book:

“[In a formalist theory of meaning,] [t]he meanings of terms are established not by closed definition, but by their place in a postulate system. Thus what is specified is only the structure of relations a concept may enter, and not the substantive character of any particular property. (…) Such a conception is perhaps most intelligible by way of its contrast with the operationalist theory of meaning (…), [where] a concept has a clearly specified meaning, and is understood in exactly the same sense by all those who use it; all of which supposedly stems from the use of agreed-upon and replicable operational definitions.” (Pawson 1989 p.229-230)

The formal theory of meaning ties together with Pawson’s comparison between the methodology of physical science and that of social science. Formal theorizing is the only viable language in physical science. Although he recognizes that formal theory in social science won’t reach the mathematical precision of physical science, he nevertheless sees some degree of formalization of conceptual relationships as a prerequisite for meaningful empirical tests.

Social concepts specified in law have one thing in common with concepts in natural science. They are abstractions that draw their meaning from their relation to other abstractions in a theoretical system. Force, mass and acceleration depend on each other for their meaning. They can only be linked to empirical observations together. The parallel ends precisely there, since social laws are normative while natural laws are descriptive. But the important point is that concepts in man-made laws are not free-standing. They do not correspond to everyday language and they cannot be understood separately when they are used.

Legal concepts are constantly reinterpreted in courts in the light of similar precedents and general jurisprudential principles. This reinterpretation covers the law as a whole. No concept is reinterpreted on its own. As an example of legal interpretation, let’s look at the agreement on trade in bananas between the European Union and the United States of America (EUR-LEX 22010A0609(01)). Here’s a more or less randomly selected paragraph:

3. The EU further undertakes:

(a) to apply an MFN tariff-only regime for the importation of bananas and therefore not to apply measures affecting the importation of bananas into its territory in the form of quotas, tariff rate quotas, or import licensing regimes for bananas supplied from any source (other than automatic licensing regimes solely for market monitoring purposes) [2]; and

(b) not to apply any measure that discriminates between suppliers of banana distribution services based on the ownership or control of the service supplier or the origin of the bananas distributed.

At first sight it seems indisputable that the meaning of “bananas” in this agreement corresponds precisely to the meaning of the word in everyday language. But this is not so, because a legal banana is an abstraction which can only be understood in its legal context.

A legal interpretation of “bananas” might first of all distinguish between seeds, plants and fruits, or between different kinds of bananas, but these distinctions are of course common in everyday use also. More importantly, a legal interpretation identifies bananas as a product whose trade has previously been subject to certain earlier agreements, whose trade regulations will have direct implications on the rules governing the trade of certain other products (other fruits or agricultural products, perhaps other products that fall within the “MFN tariff-only regime”), or as a product whose trade regulations demand specific consideration of certain general principles of justice (e.g. equal treatment of suppliers, or fair trade with developing countries).  So the meaning of a legal “banana” is embedded in the meanings of other legal concepts which surround it. This connected network of legal concepts has a history of precedents and interconnections. It is also grounded in general jurisprudential principles.

Returning from bananas to political theory, what follows from the fact that political action is influenced by law and that legal concepts are interconnected in a formal system? What is the implication, in Pawson’s third rule, for the empirical testing of hypotheses in political theory, in contrast with hypothesis testing in apolitical theory? I suggest that an empirical test of a political hypothesis, framed in the legal language relating to that hypothesis, will better meet Pawson’s rule 3 than any test of an apolitical hypothesis for two reasons.

The first reason relates to Pawson’s rule 2: regularities, mechanisms and contexts. The context of social action in politics is set by the legal framework. Some apolitical mechanisms may also be formulated in contexts that appear to be well-defined, but the contrast to political mechanisms is evident in what the context excludes. Legal concepts have been specifically developed to exclude certain interpretations, and thereby certain actions. An interpretation that goes against the legal one, or actions that go against the law, are illegal and can therefore be disregarded (unless illegality itself is the subject of research, but that is an apolitical research topic). Even the most self-contained apolitical context imaginable will lack the formal boundaries that a legal system lays down.

The second reason relates to Pawson’s rule 5 and to his outline of the interview process which constitutes the empirical test (chapter 10 and p.325 in Pawson 1989). Not only is the context of social action clear to the researcher when it is specified in law, but it is also clear to persons being interviewed. The mutual knowledge which the researcher’s questions presuppose is better established when the institutional practices that govern social behavior are set in law than when they are set by informal convention. In other words, when people state the reasons for their behaviour, their statement can be more clearly interpreted when it is made with legal concepts than when it is made in an informal language. The empirical results can consequently be aggregated and compared with more certainty.

The fact that laws form a normative system, not a descriptive one, is actually of little consequence for political theory. Researchers can assume that legal norms guide the social phenomenon when they frame their theoretical hypotheses, or they can assume that they do not. They can test empirically which alternative is true. But either way, this is irrelevant for the methodology of hypothesis testing. What matters when theoretical hypotheses are put the test is formalization. Descriptive networks of formal concepts are hard to come by in social theory, but a normative network of formal concepts is better than no network at all. With the help of this network empirical questions can be narrowed – some social mechanisms can be excluded by definition, interviews can be focused the right way, and the empirical test will therefore be stronger. By making use of legal concepts, political theory can become more amenable to empirical testing than apolitical theory.

This idea would no doubt require a lot more elaboration and exemplification to be convincing, but that will have to wait for another time. In the last essay of this series I will try to demarcate political theory from political philosophy and see what characteristics of political philosophy separate it from apolitical philosophy.


Pawson, Ray, A Measure for Measures – a manifesto for empirical sociology, Routledge 1989.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: