Originalism and judicial restraint, part VI
(I, II, III, IV, V) My initial comments in part I framing this view of judicial restraint probably made it tolerably clear, but it is worth spelling out, because so many critics of forms of originalism seem not to recognize it, that originalism is not a theory that promises certainty about constitutional law based on the historical materials. Jeffrey Toobin claims, for instance, that "conservatives ... have claimed that they can identify the original intent of the framers and use their eighteenth-century wisdom to resolve any modern controversy." David Strauss lists, as the first among three "fundamental problems with originalism" (p. 18):
On the most practical level, it is often impossible to uncover what the original understandings were: what people thought they were doing when they adopted the various provisions of the Constitution. Discovering how people in the past thought about their world is the task of historians, and there is no reason to think that lawyers and judges are going to be good at doing that kind of history—especially when they are dealing with controversial legal issues that arouse strong sentiments.
Strauss elaborates here (p. 140), explaining one of "two principal problems with originalism":
[T]he originalist project [is] a particularly difficult, challenging form of intellectual history and one that often will, to the honest originalist, turn up the answer “I don’t know,” or “there were various ideas and none clearly prevailed,” or “they were just confused back then.” That is one difficulty with originalism. Too often, it will be just too hard to figure out the answers to the relevant historical questions.
Andy Koppelman comments (p. 1919),
[O]ne of the central stated purposes of originalism, and perhaps its chief selling point in the popular press, is to produce unique and indisputable answers to legal questions in order to eliminate the possibility of judicial discretion.
Koppelman thinks, of course, that this purpose is unfulfilled.
These sorts of comments (and, of course, much of the originalist advertising to which they respond) confuse the temporal/ontological issue of originalism—what the Constitution is—with the epistemic issue of how much access we have to that entity.
The controversy over originalism is over the temporal extent of the truthmakers for constitutional claims. When we say that X is constitutional, or unconstitutional, what exactly makes such claims true or false? Originalists of various stripes claim that some portion of that truthmaker is temporally confined to the time of the Founding, while living constitutionalists of various stripes claim that some portion of that truthmaker is temporally extended across generations, or is located in today's world. My original-textually-expressed-sense view assesses constitutionality based on the sense expressed by constitutional text at the founding, but also based on the reference-yielding facts that are true today, making it a zombie Constitution—half-dead, half-alive.
Originalism and its denials are about temporality and ontology, but the principles of judicial restraint and its kin that I have set out here are a matter of constitutional epistemology. And epistemology and temporal ontology are different. There are lots of facts that don't change, but which we nonetheless don't know. For instance, the number of letters in a particular copy of the Constitution is not changing—the letters aren't going anywhere, and new ones aren't appearing—but I still don't know how many there are. Goldbach's conjecture is either true or false permanently, but we don't know which. The number of eyeglasses owned by Benjamin Franklin isn't changing, but we still don't know what it is.
Now, it is true that originalism has sometimes advertised itself as a refuge of certainty (or at least relative certainty) from the raging seas of the living constitution. But if that advertising is understood as a necessary connection, it is mistaken. The examples above show that epistemic virtues like knowability are not conceptually tied to our constitutional truthmaker's temporal extent or location. It is possible to be both unchanging and unknown, or even unknowable. Neither is a temporally-extended constitutional truthmaker necessarily unknowable, as long as we can know the principles governing its change over time. Tort law and contract law, for instance, are temporally extended across generations, but still knowable. Common-law constitutionalism simply claims that the Constitution is like these common-law subjects. All four boxes are possibilities: unchanging and known, unchanging and unknown, changing and unknown, and changing and known. Indeed, within the unknowable subset of the unknown, we have both the unknowable unchanging and the unknowable changing.
The epistemologically-rooted principles of judicial restraint I have been discussing in these posts govern what to do precisely when sufficient certainty about historical materials is not available. If it is impossible to uncover what the original understandings were (with sufficient certainty given the interests at stake), we are simply ignorant of what the Constitution means. We then need principles for dealing with ignorance, to be sure, but modifying our assessment of the temporal extent of the Constitution should not, I think, be one of them. As with Justice Jackson's Level 5 turn to policy considerations simply he found original materials were inscrutable, that response is like looking for keys under a lamp simply because the light is better there.
The no-assertion-without-knowledge, no-knowledge-without-sufficient-evidence-given-the-stakes, and no-suppression-of-relevant-evidence principles can, I think, help judges from confusing their own values with those of the Constitution, but these principles do not just fall out of originalism by definition. Depending on what the history turns up, these principles might or might not fit with the meaning expressed by "judicial power" at the Founding; I suspect they do, but that depends on historical investigation. It is logically consistent with originalism, though, to have an unchanging Constitution that explicitly authorized judges to use a very low standard of proof regarding very-difficult-to-discern constitutional provisions. Such a Constitution would be originalist, but would, as Strauss and others fear, invite judges to see the Constitution as a mirror of their own values. Given sufficient inscrutability of other guides to constitutional meaning, the-Framers-were-sensible-people-so-they'd-probably-agree-with-me grounds might consistently tip the scales in favor of judicial policy preferences. However, that is not the only sort of unchanging Constitution we might have; if my principles are at least consistent with originalism, we have to resolve the contingent question of whether they fit with our actual Constitution's grant of "judicial power."
Originalism is a theory of constitutional interpretation, but my principles of judicial restraint are not interpretive principles as such—ways of attaching meaning to the Constitution. They are, however, (proposed as) part of constitutional adjudication: the ways that courts and perhaps others resolve tangible disputes about the Constitution. That is to say, they are part of "constitutional construction" as the term is defined by people like Larry Solum, Randy Barnett, and Keith Whittington.
Getting a handle on the epistemic principles relevant to constitutional adjudication can help solve a persistent worry, Solum and Barnett's central example of where constitutional construction is needed: the problem of vagueness. While vagueness has attracted a lot of attention from legal theorists, it has attracted an enormous amount more from philosophers. Timothy Williamson—he of forementioned knowledge-as-the-norm-of-assertion fame, though this is other, earlier work—has popularized the epistemic view of vagueness. The idea is that there is some particular number of hairs that will cause Harry to be bald (or bald-to-degree-1), but one fewer hair would render him not bald (or not bald-to-degree-1). There is such a number; we just don't know what it is.
We can apply this account of vagueness (though it has not caught on nearly as well as Williamson's views about the relationship of knowledge and assertion) to vagueness in constitutional interpretation. Here's an instance of vagueness in the law as I see it. I have an anti-outlier view of fundamental rights under the Privileges or Immunities Clause. Exactly how many states must give a particular privilege to their citizens for that privilege to count as a "privilege or immunity of citizens of the United States," which all states must respect? Well, it's vague. 28 out of 50 would clearly not be enough; 48 out of 50 would clearly be enough. In between I have varying levels of certainty. And depending on the stakes, my requisite level of certainly might vary. The line is uncertain, and as we approach it, we lose bits of confidence. 48 states? Definitely yes. 47 states? Definitely yes. 46 states? Definitely yes. 45? Well, pretty definitely yes. At some point, I know not quite where, my confidence level will slip below the confidence level I need for judicial review. Lack of knowledge about precisely when we lack knowledge perhaps seems unsettling, but it also matches Williamson's view of higher-order vagueness. (Put in epistemic modal logic terms, it amounts to the denial of the "4" axiom, , with the box representing knowledge.)
Now, on Williamson's account of vagueness, paired with the activismometer, we can only legitimately assert that something is a privilege of citizens of the United States—and thus, only strike down statutes on that basis—if we are far enough away from the boundary line that we meet the stakes-sensitive standard for knowledge. Relatively activist courts, who view judicial review as a relatively low-stakes proposition, will be willing to go closer to the line. Resolving issues of vagueness thus involves (a) deciding how much clarity is required for judicial review—i.e., where on our activismometer we should be on the Level-2-to-Level-4 spectrum—and (b) reserving judicial review for cases far enough away from the blurry boundary line to achieve that level of clarity. This isn't a terribly exciting account of how to respond to vagueness, but it seems fairly sensible to me, and I have reasonably high hopes that it matches the meaning of "judicial power."
That's all for now on this subject, though I continue to work on grounding all this philosophy historically and may blog about what I find. Perhaps more after the conference this weekend, or when this is all assembled into an article. This brief tour of judicial restraint is, of course, incomplete. Vagueness is not the only critical issue of constitutional construction (i.e., issue in constitutional adjudication not resolved by interpretation), of course, and several other issues in construction—the "constitutional adverse possession" issue of when to consider precedent entrenched, and the when-to-decide-constitutional-issues-at-all questions involved with political questions, standing and the like—are critical to a full theory of judicial restraint in constitutional cases. More on those when I have something to say.