Tuesday, June 08, 2004

Basic structure and secondary rules

I've been agonizing for a long time on the exceedingly open ended nature of Rawls's basic structure of society.

The principles that are presented in A Theory of Justice (TJ) serve primarily as guides to order social institutions, and not directly as principles of individual conduct. What is more, they are not to determine the internal decisions and procedures of each institution, but only the public rules that govern the interaction among them. Rawls calls these public rules the basic structure of society: "the way in which the major social institutions distribute fundamental rights and duties to determine the distribution of advantages from social cooperation." (TJ, p. 6)

The idea of the basic structure was introduced in TJ to limit the scope of the theory to the most important institutions, those that significantly impact individuals' life chances.

The problem is that institutions—in Rawls no less than in the sociological literature from which he draws—are defined quite loosely as "social practices". From as far back as "Two Concepts of Rules" (1955) Rawls had been using this definition. But how do you address a social practice in a theory that counts only constitutions, laws, regulations, and judicial decisions in its political toolbox

Here I fond H.L.A. Hart's distinction between primary and secondary rules helpful. Social practices, when they are not institutionalized in a modern legal corpus, consist of primary rules but no secondary rules. They can therefore not be directly reformed. What can be reformed are the laws that apply to them.

Attentive readers the blogerature will notice that this post was sparked by Lawrence Solum's recent entry in his Legal Theory Lexicon on primary and secondary rules. Read the whole thing, and by that I mean the entire Lexicon, not just the latest post.

This makes more acceptable the distinction between institutions to which the principles of justice apply directly (such as the tax code) and those to which it applies only indirectly (such as the family). Social practices like the family are part of the basic structure insofar as they are a kind of social fact, but they are not directly subject to the basic structure because they have no secondary rules to define how the rules affecting the institutions are constituted, validated, and changed.

One may object that the family does have such rules, that it is legally constituted as Marthan Nussbaum has convincingly argued. And this is true, but only in part. We can imagine that if all (public) laws pertaining to the family were abolished many things about the institution would change— tax benefits, default rules governing inheritance, presumptions of paternity—but may others would stay the same, whether because of private norms (such as religious codes) or tradition and cultural practice.

But this squares quite nicely with Rawls's own position on the family:

[P]olitical principles do not apply directly to [the family's] internal life but they do impose essential constraints on the family as an institution and guarantee the basic rights and liberties and fair opportunities of all its members. (Justice as Fairness: A Restatement, p. 164)


If I am correct, the distinction between primary and secondary rules helps to narrow the scope of the theory of justice even further, and to make its implementation more predictable and effective. This is not to say that it clarifies all of Rawls's ambiguities. There are outstanding questions of what counts as an institution that is part of the basic structure, and whether it is legitimate to try to make some of these institutions more or less relevant in the interest of justice.

It is also not to say that the solution will satisfy everyone. That we should abstain, as a matter of principle, from tinkering with many social practices which we (rightly) consider unjust, and limit the action of the state to the passage of laws that only indirectly compensate for such inequities will strike many as resignation at best and, at worst, hypocrisy. To them, I can only say that we see different roles for political philosophy in the world.

Indeed, I am becoming increasingly convinced that students of Rawls (at however many degrees of separation they stand from the professor) will come to be divided into two camps—radical and moderate Rawlsians, corresponding roughly to those who emphasize Rawls's democratic or liberal tendencies, respectively. If that turns out to be the case, I stand firm in the moderate liberal side of the divide.