Keep in mind that there is a distinction between legal behavior and legal conduct. The former, while intentional, eschews the agent’s reasons for action, while the latter necessarily incorporates those reasons. For an empirical study to be objective, according to the above definition, it must eliminate or severely restrict contestable assumptions in order to reach the designated level of contestable assumptions. Specifically, it must exclude the agent’s practical reasoning, including her political and constitutional reasons. The intentionality of the agent’s behavior will be preserved, but not her motivational and justificatory reasons. By making this move, however, we are no longer studying legal conduct. Consequently, we leave unexplained what relevance the empirical results have to the question in constitutional law which prompted the inquiry in the first place.
Consider the question of whether judges are political, a question recently addressed by Cass Sunstein et al. In answering this question, one might define "political" as “liberal or conservative.” We may then define "liberal" and "conservative" in terms of the ideology of the nominating president. Our inquiry is objective only if depicting the nominating president as “liberal” or “conservative” is objective. These depictions might fail the objectivity test because they include too high a level of contestable assumptions. Arguably no inquiry in any area of human knowledge is entirely free of contestable assumptions. This suggests the following rule regulating definitions applicable to legal empirical inquiry. The more objective the definition, the less relevant the answer is to the question which prompted the inquiry in the first place, namely, in this case, whether judges are political. Objectivity and contestability appear to be polar opposites. Hence, empirical inquiry and the language and reasons of constitutional law are incompatible.
To understand why this is so, let’s turn to the following question: Why do we want answers to the question of whether judges are political. The primary response, I think, is because questions of this kind are relevant to our understanding of the courts as well as to our own views, as attorneys, about adjudication. Even more importantly, answers to these questions are relevant to our political and constitutional convictions as citizens. We might want to know, for example, whether judicial supremacy comports with republican democracy in order to embrace or reject judicial supremacy.
We might also want to determine, for example, whether judicial practice can support the distinction between law and politics? If it can, the empirical data might help us educate attorneys and potential judges on how to apply the distinction in practice if such is desirable. However, if the study itself depends on too high a level of contestable assumptions about politics in general or about the meaning of particular political perspectives, then the results will be contestable. Hence, we cannot seriously aspire to educate others on how to apply the distinction in practice without innumerable qualifications such as “According to ideology A, X is the way to apply the distinction. According to ideology B, Y (not-X) is the way to apply the distinction. This just means that we are unable to generate objective results. Interestingly, and as an aside, if the empirical research cannot support the distinction between law and politics, we might want to inquire why judicial politics is superior to legislative politics.
In short, the questions driving doctrinal constitutional controversies, as well as the questions involved in theoretical controversies in constitutional law, occur in a different conceptual space than the conceptual space of empirical inquiry. To co-opt the words expressing the concepts in the question “Are judges political?" and give them operational definitions which exclude or severely restrict contestable assumptions is to pluck them from one conceptual space and put them in another. By doing so, the terms which have arisen within the conceptual space of constitutional law, theory, and adjudication necessarily lose their meaning. This problem cannot be answered, I submit, by appealing to the idea of complementarity. My position is not the simple rejection of one possible approach to understanding constitutional law. Rather, it is that the two particular approaches in this case—the language and reasons of constitutional law and the language and evidence of objective empirical inquiry—are incompatible approaches. When we conflate these terms neither constitutional law nor empirical science benefits. When we keep them distinct, the results are objective empirical knowledge about legal behavior which is irrelevant to the language and reasons of constitutional law, or non-objective results which are relevant to constitutional law, but contestable.
A third position might be possible. If we abandon the quest for objectivity and acknowledge that significant empirical inquiry in constitutional law will always include contestable assumptions, we must concede that using different contestable assumptions might produce different results. Accepting the inevitability of contestable assumptions in empirical inquiry will preclude developing a syncretic conception of social reality. However, we might achieve nevertheless interesting, though contestable, results about legal conduct.