Balkinization: Critical Reflections on Leiter's "Naturalistic Jurisprudence"

Critical Reflections on Leiter's "Naturalistic Jurisprudence"

Brian Tamanaha

Chicago law professor Brian Leiter is widely known for his various blogs, but he is also a serious legal philosopher who has staked out a unique position within the field. For more than a decade, he has advocated “Naturalizing Jurisprudence” (The essays developing his view are collected here.) Leiter explains that by “naturalism” he means

"a methodological doctrine about how we should approach philosophical inquiry. On this view, philosophy proceeds (as the most familiar metaphor has it) ‘in tandem with the sciences,’ that is, as the abstract and reflective branch of the empirical sciences as they limn the causal structure of the world.”

Leiter agues that legal philosophers ought to adopt this naturalistic approach to philosophical questions about law, which promises to be more fruitful than current conceptual or normative inquiries.

Jurists who engage in socio-legal theory (or “sociological jurisprudence”), as I do, would seem to be natural allies for Leiter’s approach. But there appears to be a difference between these endeavors. Broadly speaking, sociological jurisprudence assumes a thoroughly social view of law, and draws from social science to inform theorizing about law. It is less a discrete field of study than an umbrella that encompasses theoretical work on law with a strongly empirical orientation.

Leiter’s arguments, while sharing some of the same assumptions, are pitched in a different direction: his arguments are directed at other legal philosophers, in effect telling them they are engaged in misdirected abstract inquiries that amount to little more than dressing up their intuitions in fancy philosophical clothes. I’ll let legal philosophers fight that battle (although my view is that there is value in all kinds of theory).

Instead, I thought it might be informative to examine Leiter’s arguments from the standpoint of someone already committed to sociological jurisprudence. A few critical observations about the potential pitfalls of Leiter’s naturalized jurisprudence follow after the break.

In a recent paper, Leiter offers three examples of how one might “naturalize jurisprudential questions.” I will take up only the first example here, as that will suffice to articulate my concerns about his approach.

The following is a full paragraph that sets out his core argument on the first example, although I take the liberty of splitting the paragraph and injecting a clarifying comment. Leiter writes:

[A] naturalized jurisprudence might ask what must law be if current social-scientific theory of adjudication (namely, the Attitudinal Model) is to be true and explanatory? For the Attitudinal Model to be true and explanatory, there has to be, among other things, a clear demarcation between the ideological attitudes of judges (which are causally effective in determining the decisions) and the valid sources of law which are central to the Legal Model’s competing explanation of judicial decisions. Thus, implicit in the Attitudinal Model is quite plainly a concept of law as exhausted by authoritative texts (precedents, statutes, constitutions) which are the raw material of the competing Legal Model, and which exclude the ideological attitudes central to the Attitudinal Model.

[Up until this point, Leiter’s argument has two crucial steps: 1) asking what “must law be” if the model of judging is “true and explanatory”?; and 2) asserting that the Attitudinal Model implicitly contains a concept of law (the Legal Model). Leiter goes directly from this to make a series of claims about the concept of law within legal philosophy…]

The concept of law, in turn, that vindicates this assumption is none other than Raz’s “hard positivist” notion of a rule of recognition whose criteria of legality are exclusively ones of pedigree: a rule (or canon of interpretation) is part of the law by virtue of having a source in legislative enactments, prior court decisions or constitutional provisions. That is the view of law required by the Legal Model, and it is the view of law required to vindicate the Attitudinal Model as providing the best explanation of judicial decision. Raz’s Hard Positivism, in short, captures what law must be if the Attitudinal Model is true and explanatory. To be sure, this defense of Hard Positivism is very different from Raz’s own account—to which we’ll return in a moment—but for the naturalist it suffices that the Hard Positivist concept of law figures in the best explanatory account of legal phenomena.

The reasoning in this paragraph (split in two parts by me) strikes me as problematic in crucial ways that illustrate the risks of Leiter’s enterprise. Leiter brings together philosophical arguments and social science arguments in ways that misshapes both, making strange moves and coming to a strange conclusion. My critical examination will focus on just three steps (numbered).

1. First, it is essential to note that the inquiry Leiter begins with is awkwardly framed: “what must law be if [the Attitudinal Model] is true and explanatory?” By framing it this way, which he does for philosophical reasons tied to naturalism, Leiter interjects theoretical points that have no place in the social scientific inquiry at issue.

The Attitudinal Model is a simplified model that political scientists rely upon to design and structure studies of judicial decision making. The basic idea behind the model is the assumption that judges decide cases oriented toward achieving their political preferences, with the law serving mainly as a constraint. If the results of a given study bear out the model to a high degree of confidence (reaching statistical significance and actual significance), then it “explains” the decisions.

What makes Leiter’s statement odd is that social scientists would never think to ask “what must law be [if the model is true and explanatory]?” For social scientists, the “explanatory” value of the model is confirmed (or not) by the results of study. Social scientists are not immediately concerned with the assumptions their model makes about law, but with whether the model produces a reliable account of judicial decision making. They are trying to isolate on and test the causal connection between the personal preferences of judges and their legal decisions.

These comments merely point out that social scientists and philosophers are engaged in different inquires, which is to be expected, but the following two points show how the failure to pay close attention to this leads to mistakes.

2. Several basic points must be made about the Attitudinal Model and the Legal Model. They are both simplified models for the purposes of research design. The idea behind this starkly posed contrast is to test whether judges decide cases in accordance with the law (legal model) or in accordance with their personal preferences (attitudinal model). Political scientists understand that judging is more complex than either model implies, but the contrast is useful nonetheless. (It should be mentioned that the Strategic Model is perhaps the hottest version within the field today, but Leiter does not mention it, so I will leave it out. It is worth noting in passing, however, that Leiter’s philosophical analysis relies heavily on the binary contrast between the legal and attitudinal models, whereas the strategic model involves a more nuanced integration of the two.)

Here is the crucial move Leiter makes when moving from the social science inquiry to the philosophical inquiry: he transforms the assumptions of their simplified model into what is “quite plainly a concept of law as exhausted by authoritative texts.” Following this reasoning, Leiter goes on to assert that the Legal Model is “vindicated” by Raz’s Hard Legal Positivism.

The vast chasm that separates these respective inquiries bears emphasis: Although researchers have constructed a simplified legal model to contrast with a simplified attitudinal model, social scientists, Leiter tells us, implicitly made far more profound claims: they implicitly relied upon a “concept of law,” and this concept of law is none other than Raz’s legal positivism (a controversial theory of law within legal philosophy). Leiter suggests that this complementary match lends gravitas to the model, as well as credibility to Raz’s hard legal positivism

Two related points will be made against his arguments. First, it is a supreme stretch to assert that a simplified model of judging makes any claim at all, implicitly or otherwise, about a “concept of law.” It can be said that these two alternative models make assertions about law and judging, but these are merely posited assumptions for the purposes of the study. To call this a “concept of law” gives it a standing and makes an additional set of claims that are not entailed within the notion of a “model.”

Leiter knows that the social scientists using these models do not explicitly claim to be producing a concept of law, but he seems to be arguing that legal philosophers can extract one anyway. What Leiter fails to recognize, however, is that there are different formulations of the Legal Model; indeed almost every study that tests the Legal Model posits a different set of assumptions about what the model requires. That’s why it is hazardous to assert that the Legal Model adopts any particular concept of law.

Even if one grants Leiter’s leap from a model to a concept of law for the sake of argument, secondly, it is wrong to assert that the Legal Model necessarily adopts Raz’s hard positivism rather than some other competing concept of law (like, for example, Soft or Inclusive legal positivism, a major rival of Raz’s position).

A deep lesson about the perils of his project lies in this objection: Leiter has drawn his erroneous conclusion because he confuses the social scientific question with the legal philosophy debate. They sound similar, but they are not.

Leiter’s mistake can be made plain in connection with his assertion that the Legal Model “excludes the ideological attitudes central to the Attitudinal Model.” Let’s assume that this is correct; the point remains that excluding the “ideological attitudes” of judges is very different from excluding “moral inquiries” in legal analysis. Leiter treats these in some sense as equivalent, or he appears to assume that the former entails the latter, but they are distinct. (I am reconstructing Leiter's argument here because he does not actually explain his reasoning on this point.)

Hard legal positivism precludes moral inquiries, and Leiter appears to take that to be the key that links the Legal Model to hard positivism, but the Legal Model cares about political preferences. A range of theories of law that do not exclude moral inquiry nonetheless assert that judges should not be influenced by their personal preferences when engaging in such inquiry (the assumption of the legal model); these theories of law, accordingly, are also consistent with the Legal Model.

This point will be demonstrated in connection with Raz's main rival--inclusive legal positivism. Drastically simplified, here is the core difference between these theories: inclusive legal positivists assert that questions of law can turn on moral inquiries (when the rule of recognition provides for it), whereas hard positivists insist that what is valid law must not depend upon a judge answering moral questions (but is a matter of pedigree).

The Legal Model can easily accept the notion within Inclusive Legal Positivism that the law may require judges to make moral decisions (nor does this matter for the Attitudinal Model). For example, let’s say that open-ended legal questions (fairness, reasonableness, cruel and unusual) require the judge to decide based upon the community’s moral views of these issues. The Legal Model says nothing at all about this one way or the other; all it says is that judges should not be influenced by personal preferences when making legal decisions. If the law requires these types of open inquiries, there is no problem; fidelity to law by judges in their decisions can be tested by either the Attitudinal or Legal Models.

Here is an example: A given study might find that a liberal judge deciding cases with open-ended questions of this sort makes decisions that correlate to a high degree with conservative political outcomes; presumably the explanation for this pattern would be that those outcomes are consistent with the moral view of the community (as distinct from the judge's moral views) that the law incorporates. These decisions, which can turn on the judge findings answers to moral questions, would not be influenced by personal preferences (This example is not far fetched because studies show that on many issues liberal judges rule more often in a conservative than a liberal direction). In this example, the judge would decide in a fashion contrary to what the Attitudinal Model predicts, lending support for the Legal Model. But these results--and the models they test--have no bearing at all on any philosophical theory of law.

The bottom line is that the Legal Model does not necessarily endorse hard positivism—or indeed any other concept of law—because the assumptions of the Legal Model are consistent with a range of theories of law. A model or account of how judges judge (whether they are influenced by personal values) is different from a theory of law, and can align with the latter in various ways.

3. The strangest part of Leiter’s argument comes at the end: “…Hard Positivism, in short, captures what law must be if the Attitudinal Model is true and explanatory….for the naturalist it suffices that the Hard Positivist concept of law figures in the best explanatory account of legal phenomena.

I won’t repeat earlier objections to this line of reasoning but instead will point out how this conclusion twists matters into a logical pretzel.

For social scientists, because these two models are often set out as competing explanations, the confirmation of the attitudinal model—a demonstration that it is “true or explanatory”—often entails the refutation or diminishment of the legal model. But Leiter obscures this antipodal relationship, reasoning that the hard positivist concept of law (embodied in the Legal Model) must be correct “if the Attitudinal Model is true and explanatory.”

That is a weird way of describing matters. The vindication of the Attitudinal Model by social scientists brings down the competing Legal Model, whereas for Leiter the Attitudinal Model is true if the concept of law that the Legal Model is based upon (or so he asserts) is true. Leiter thus claims the concept of law purportedly vindicated by the Legal Model is “true and explanatory” even when the model does poorly in the social scientist’s actual study. The actual results of the study, in other words, don’t mean anything for Leiter’s philosophical point.

There is a consistent logic to Leiter’s reasoning, but it strains common sense, and has great potential for confusion.

Ultimately Leiter backs away from this argument, noting that the attitudinal model has relatively weak predictive power. So he does not go further with this particular analysis. It is offered by him as an example how how a naturalized jurisprudence might work. And that is the way it should be understood.

The problem, what I hope to have demonstrated in this lengthy post, is that it is dangerous to translate social scientific inquiries into inquires in legal philosophy. They have different interests and concerns that are not easy to translate.

Nothing in this post refutes Leiter’s philosophical argument for naturalizing jurisprudence (which I do not address), but legal philosophers who follow his lead must tread with caution.

Posted 12:10 PM by Brian Tamanaha [link]