More on Razian authority

Brian Leiter prefers that commenters refrain from pseudonymity, so I'll just continue this conversation here. I understand what Gardner and Green are saying, but if it's an accurate interpretation of Raz, it seems to be practically a reductio. To recap, my argument is as follows:

Imagine a legal system, SP, where the criteria of legal validity are: 1, only directives issued by official O are valid; 2, all those directives issued by O which do not conflict with moral considerations A and B are valid. Suppose O issues a directive, "Do X, and do not consider dependent reasons C, D, ... , J." This is clearly a soft positivist system. The question is whether such a system is possible. Gardner and Green claim that on a Razian account, it is not; such a system is incompatible with the conceptual truth that law claims authority.

I claim that, on the contrary, SP's law can easily claim authority. Obviously this claim may or may not be warranted, but, as Raz has been at pains to stress all along, this is not a problem; what Razians need to assert is the conceptual impossibility of SP. And not only is it possible, I believe that the legal system in SP is likely to have a stronger claim than one without such soft-positivist criteria.

Gardner claims that in my example "law-applying officials have no authority regarding the application of A and B. It is irrelevant that they have authority regarding lots of other things (such as C, D ... J)." But this seems to imply that to be an authority--to validly issue directives that can be taken as authoritative--one's directives must be taken as authoritative in all circumstances. Or as Green puts it, "if one assumes, with Raz, that it is necessary to the law that it claim authority, the law cannot include requirements, like A and B, concerning which the lawmaker plays no authoritative function."

To see why I feel this claim must be false, consider hard positivist "sources" criteria. The lawmaker, obviously, plays no authoritative function in deciding whether or not these criteria obtain; they either do or they do not. A bill was either passed in the proper way by the Queen in Parliament or it was not. Subjects must decide for themselves whether or not the directive issues from a valid Source before granting it authoritative status, and the officials might in fact be wrong; the Source-based criteria might fail in particular instances. Some directive might not in fact be encompassed by the authorizing legislation passed by the Queen in Parliament, despite the official's belief that it was (obviously there are complications, as the criteria may in a particular legal system have to do with the official's belief concerning ultra vires concerns rather than the fact of the matter, etc., but this can be put aside; we can imagine hard positivist criteria without such worries). In such a case, would we say, with Green, that "a[n] [executive official] issuing what he believes is a valid [legal] directive cannot consistently *claim* to be acting as an authority when issuing it, insofar as these requirements [ie, the Source-based criteria] are concerned"? No; the official will presumably believe that the criteria are satisfied, and hence be perfectly consistent in her claim (even if wrong). She is not acting as an authority on the question of whether or not the criteria are satisfied; she is acting under the authority conferred on her by the presumed satisfaction of those criteria.

This is exactly parallel to my official O in system SP consistently claiming that her directives are valid, even when she is wrong about A & B. If she is acting in good faith, she must obviously believe herself to be right about A & B; this is enough for her claim to be consistent. Just as the law can be wrong about its claim to authority--whether or not a directive ought in fact exclude and replace the dependent reasons underlying it--so too can officials be wrong about whether they have issued directives that pass the tests of validity within their systems. Are Raz, Green, and Gardner really denying this?

It might seem so. I'm having a bit of a hard time understanding Gardner's claim that "on the hard positivist view the directive must be interpreted so as to make the (legal) force of unexcluded reasons A and B turn on the law-applier's determination of their force." Is the claim is that whether or not the legal criteria of validity--even Source-based criteria--are satisfied turns on the official determination of this fact? Such an assertion seems truly odd, and does not flow at all from the self-image of the law as authoritative. It would seem to imply that officials can never be mistaken about applying the criteria of validity. But of course they can, just as different officials can disagree about whether or not they are satisfied. One official--an executive, say--might think he has valid authority; another--perhaps a judge--might disagree; perhaps a second set of judges will side with the first official. At each stage the officials will be trying to determine whether in fact the criteria of validity--whether Sources-based or otherwise--were satisfied (again, perhaps the judges will be looking at the different question of whether the official was justified in his possibly unauthorized action, but if so they are asking the slightly different question of whether the citizen had a legal duty to obey an invalid law in those circumstances, or a right to recover from the official's act, etc.).

Contra Gardner, soft posivitism does not efface the authority of a legal system's officials; what it does is narrow the scope of this authority. Hard positivist criteria of validity also narrow this scope. A legislature that can only pass bills with a 2/3 supermajority has "less" authority than one operating under simple majority. Same with one that can only legislate on Tuesdays. But this is no more incoherent or problematic than the idea that a doctor might have epistemic authority with respect to kidney diseases but not kidney diseases and brain tumors.

The point of the "partial exclusion" critique is that while the law needs to claim authority, it need only claim authority contingent on validity. Indeed, this contingency is precisely what separates a fully articulated legal system with criteria of legality applied by a cadre of officials from a simple personalist regime where Rex or Regina is an authority on everything. Just as an official can say "Do X" and claim authority because she believes that she is authorized to do so by a bill passed in the proper way (and she believes that moral considerations justify the subjects of her directive treating it as authoritative and hence excluding A, B, ... , J), so too can an official say "Do X" and claim authority because she believes she is authorized to do so by the existence of the bill plus the fact that the directive in fact complies with A & B (and moral considerations justify the subjects excluding C-J). In both cases the official can be wrong both about validity and about whether the authority is warranted, but so what?

Is the claim that inserting any moral considerations into the criteria of validity destroys the possible service value, because any sort of moral evaluation is categorically more difficult to apply than the Sources-based tests the hard positivists envision? This doesn't seem at all persuasive to me, but I won't bother trying to refute it. At any rate, it doesn't seem the natural interpretation of Raz; I had thought the point wasn't that moral considerations were distinctly harder to evaluate, but rather that authority would be undermined by trying to peek beneath and look to all its dependent reasons. Soft positivism doesn't do this. Even if A & B (let's say these are rule-of-law values) are among the dependent reasons that the authority will consider in making its judgment, O provides a service to her subjects by giving her directive in place of C-J. Is is not merely question-begging but mistaken to say that unless all dependent reasons are excluded, no useful service is provided by the exclusion of the rest.

Whether or not O in fact provides this service is an open, moral question, just as are all claims of authority. But I submit that O might actually have an easier time justifying this claim than she would if she were in a hard positivist system lacking the extra criteria of conformity with A & B. This flows from the general claim that it is easier to justify authorities of lesser rather than greater scope. Think about justifications for rule-of-law values: they often point to the ways in which the formal aspects of law (prospectivity, generality, due process) may increase conformity with other, more substantive, moral values. As a potentially vulnerable subject, I may think it reasonable to treat O as authoritative with respect to C-J only if I need not take her as authoritative with respect to rule-of-law considerations A and B. Her conformity with the latter allows me to believe that, in fact, I am better off letting her directive replace my own judgment concerning the rest.

I'm not saying this argument always goes through; I'm generally skeptical concerning legal authority. But soft positivism, by limiting (but not effacing) the law's authority, makes it easier to justify.

I'm really not sure what to add to this. If there's something I'm still missing about Gardner or Green's argument, I'd love to know what; I continued to be baffled by the acceptance of Raz's claim, and I'd be happy to be shown what I'm missing.

Grumpy before my time

Any article titled "An Economist Goes to a Bar And Solves the Mysteries of Dating" is pretty much guaranteed to annoy me. Now, again, the research here is rather interesting: they ran speed-dating experiments for two years and thus have a huge data set on 4-minute interactions between Columbia grad students, at the end of which each decides whether or not they'd be interested in a future date. Very cool. One might question how generalizable are conclusions derived from Columbia grad students (or from speed-dating), but still, neat.

My complaint is simply that the author is doing the world a disservice by implying ("all of this rational-choice stuff") that if you want to study this sort of thing, you should look to economics. This is social psychology pure and simple (and indeed, two of the coauthors were psychologists). Now, none of this would matter much ... except that the social ecology of knowledge-production really is sensitive to considerations like this. It's not good for Truth when economics is seen as the only "real" social science, and bright students go for econ PhDs regardless of the questions they intend to study or the methods best suited for studying them.

UPDATE: to clarify a bit--I don't have a problem with interdisciplinarity. Interdisciplinarity, done well, is great. What I have a problem with is when the general public comes to have a distorted view about what various disciplines are about. If you go to grad school for econ, you'll spend your first years with Mas-Colell, Whinston, and Green; it's not the most natural place for learning how to conduct field studies to gather data on social interactions.

Against Raz's SCA argument for hard positivism

I think Leiter and Greenberg are quite right with their "partial exclusion" critique of Raz's argument that soft positivism is incompatible with the law's necessary claim to authority. I'm not sure their example is quite what they are looking for, though. With a slight modification, it could be immunized against Michael Steven Green's objection:

An official lays down a legal directive: "In instances of this sort, do X; do not consider C ... J." Moreover, the criteria for legal validity in this system include, 1, that directives be issued by this particular official, and 2, that they be consistent with reasons A and B, which are presumably dependent (moral) reasons applicable to a very wide variety of situations.

This is clearly a directive that could be authoritative: we can easily imagine (expertise, democratic respect, whatever) the official's directive being a reason to do X and to exclude C-J from our deliberations about whether to do X. It has served us, despite requiring us to consult A and B to test its legal validity.

This is positivist, because the fact that A and B rather than C and D are among our criteria for legal validity is a consequence of the social practice that we happen to have; it is soft positivist because, as John Gardner puts it in his comment, the A and B reasons do the validating independent of the official's assertion, should he make one, that A and B are in fact satisfied.

Ta-da! Soft positivism. Has this argument not been made in print before? The 1989 S. Calif. Law Review symposium on Raz had a few pieces that came close to this argument (I'm thinking of Stephen Perry's, I believe).

Words have meanings, part XXVI

In the midst of an otherwise-excellent article summarizing what appears to be quite nice research on the degree to which auto repair shops correctly diagnose problems and prescribe unnecessary services, David Leonhardt inserts a rather curious claim: that this research somehow exemplifies "freakonomics," against which there has been a (possibly justified) backlash, but does so in a good way. In doing so he links to an older NYTimes piece that seems just atrocious, one that runs together a number of totally different trends:

(1) economists studying non-market interactions (Beckerites, public choice folks, etc.)
(2) economists distancing themselves from political advocacy, seeing themselves as impartial scientists
(3) economists turning to statistics (or empirical work more broadly) as opposed to formal modeling
(4) economists turning to micro- as opposed to macro- level questions
(5) economists studying "the everyday"

Whatever categories you think Steve Levitt's work falls into, it's just insane to claim that these 5 things are somehow identical. Nor do they necessarily go together. Examples are easy enough to come by--David Friedman's "Hidden Order" is clearly 5, 1, and 4, but certainly not 2 or even 3.

Schneider's research is at the heart of economics: the study of market interactions. It's probably true to say that it's in some sense Not What Most Economists Do, but let's be a bit more precise about what we're talking about. He's studying "the everyday," but the everyday of the market, and while it's micro and empirical it doesn't at all sound like one of those "this instrumental variable regression will explain once and for all a complicated multicausal phenomenon that other social sciences, whose work I shall not read, let alone cite, have studied for decades" findings. Can we please reserve the term "Freakonomics" for the book and the blog? No? Sigh.

On gaps, contradictions, and originalism

Just a brief clarification concerning the last post, as Larry Solum has emphasized in a response to another interlocutor that "One of the points of the New Originalism is that the semantic content of the Constitution sometimes underdetermines its application. That's a feature of the theory, not a bug." While I discussed the problem of gaps and contradictions in my post, I wasn't entirely clear about distinguishing two separate levels of indeterminacy that give originalists trouble in different ways. So let me try to restate this argument, which was merely implicit before.

While Solum acknowledges that semantic content underdetermines its application, he misses the force of the full critique. On the one hand, if we are genuinely restricting ourselves to literal semantic meaning or something very close to it, "underdetermination" is a huge understatement: literal semantic meaning gets us nothing but the really obvious provisions about 2 senators per state, etc. (Keep in mind, though, that deference even to literal semantic meaning requires smuggling in some assumptions, with which I take issue; I think it's quite possible for our criteria of constitutional validity to contradict and override even this thinnest, least controversial sort of meaning. This can be safely put aside, however.) But, as I argued, literal semantic meaning is just not the stuff constitutions are made of: if it were, the idea that they can effectively constrain and structure governments would be absurd. So we have to import some degree of pragmatic content--whether it's Solum's "clause meaning" or something else--to get anything workable at all.

Once we do this, however, we run smack into the interpretive pluralism--the deep disagreement about the criteria of constitutional validity--that Leiter so rightly brings up. (As I noted, it's actually an issue even with LSM, but less obviously so.) Whether we should view this pluralism as generating contradictions (a plurality of inconsistent but valid "constitutional moves") or gaps (when you have a plurality of inconsistent criteria covering some area of consitutional import, then this means you lack any criteria at all) is an interesting philosophical question, but the bottom line is the same. Semantic originalism is not just a theory that leaves certain applications underdetermined; whether or not one may or must use it is itself underdetermined (at least, I would say, from a mainstream positivist standpoint) on any realistic appraisal of American constitutional practice. It is because our highest strata of officialdom are populated by originalists and structuralists and legal process types and Dworkinians and on and on that gaps/contradictions exist in the norms of constitutional validity. And that can't be wished away--or at least, not at once; if every Con Law professor converts to semantic originalism tomorrow, then perhaps in 40 years ...

In short: there are at least two levels of indeterminacy related to constitutional meaning. One is the sort Solum highlights: the rule may not determine the application; semantic meaning won't get us very far (though, as highlighted above, this bites deeper than he acknowledges). The second is at the level of rule-identification itself--and it is this, I insist, that is fatal to the semantic originalist's claim of conceptual necessity.