Four Levels of Originalism

In my constitutional law class we have been discussing at various points the concept of "originalism," and as it quite frequent with that word the discussion can get very easily confused. There are, after all, lots of different concepts that go by the name, or invoke the concept of "original" in some way. This post will propose a classification of types of originalism, with some discussion of how legitimate each is. The inspiration for my classification scheme comes from something from my undergrad con-law class with Steve Calabresi, where he presented five modes of interpretation on a scale from most to least binding: text, structure, originalist stuff, precedent, and policy. (I'm actually not sure which order the originalist stuff and precedent were in, but it will shortly become irrelevant.) The idea is that, in trying to answer any constitutional question, you start by looking at the text. If that solves your problem, great, but at this late hour at least it probably doesn't. To resolve ambiguity, you might then ask which solution would be most consistent with the overall structure of the Constitution. If that is of no help, you consider what the Court has said on the matter, or ones like it, before. At some point in the process you examine the history from around the enactment of the relevant constitutional provisions; that's the originalist part. And if none of those sources of actual authority are of any help, if it is truly a 50/50 proposition which side ought prevail in the trial based on the actual law of the Constitution, you are free to choose whichever side ought prevail on policy grounds. If and only if, of course.

So here's my theory: originalist inquiries aren't another level of that scale. Rather, each of the other four levels can be informed by originalist/historical inquiry, and as the level of authority declines with each step from text to structure to precedent to policy, so the level of legitimacy of originalism declines. Let's examine each variety in turn.



Original Text
This is by far the most valid form of originalism, and is what people mean when they suggest that everyone is an originalist now, left and right alike. Often when Scalia writes pieces defending originalism this is what he describes, though it's certainly not the only form or even the dominant form he practices. The concept at this level is that when we say the constitutional text is what binds us, we must mean specifically the text as it would have been read when it was written. Not what glosses or interpretations would have been put on it when it was written, just literally what the words would have meant. This is the "dictionary originalism," where you get out some dictionary from the time of the Founding or from Reconstruction and see how it defined a given word. And it is perfectly valid: the simple thought experiment in which over the next hundred or so years the English language evolves such that any English sentence written through the present day would still be perfectly intelligible, but would have a meaning completely removed from its current one. If future Americans read the Constitution in the language of their day it would be gibberish, or worse, it might happen to be a Constitution, just one whose meaning was completely different from ours. But it wouldn't be law; the Constitution as we know it would still be. Any smaller-scale linguistic shifts in meaning over the past 224 years similarly lack the power to alter the supreme law of the land, so in reading the document we must be wary lest such shifts improperly color our understanding.

Though this is by far the most valid kind of originalism, it is also the least useful. There just haven't been that many radical shifts in meaning over this period, and even once we correct with them we are left with ink on paper. Nothing about dictionary, pure-text originalism can help us deal with the ambiguities in the Constitution, it can only prevent us from accidentally applying something which is not the actual Constitution as written in its own language. So we need something else.

Original Structure
Structural analysis of the Constitution is very interesting. In one sense it is the art of drawing inferences from the constitutional text alone which are not obvious simply from reading the language of each clause. Nowhere in the Constitution does it say, for instance, that states are prohibited from levying taxes specifically on organs of the federal government. To maintain otherwise, however, would be inconsistent with the constitutional structures created by the text, for the reasons expounded by Chief Justice Marshall in McCulloch v. Maryland: the federal government is supreme over the states, though of limited power, and the power to tax is the power to destroy, so giving states this power would effectively negate federal supremacy. The idea is that there is a structure underlying the Constitution that tells us certain things about how that document must operate, even if it does not say so. But actually, though I said the trick is to draw inferences solely from the text, the historical context of the text can help us understand what structures it creates. I think the Reconstruction Amendments are the best example of this. There is good reason to think that the Fourteenth Amendment was occasioned by a sense of the questionable constitutional status of the Civil Rights Act of 1866. It would be just plain strange, therefore, if the Fourteenth Amendment were not sufficient to sustain that Act. If our reading of the text suggests otherwise, we are given reason to be skeptical of that reading. We might even go so far as to infer that some or all of that Act's requirements were baked into the Constitution through that Amendment, though only so far as is consistent with the text. Moreover, the structural influence solely from the text of Section One which says that, as that Section expresses limits against the state governments the states should be given little or no deference in Fourteenth Amendment cases is surely strengthened by the historical context, in which the Amendment was meant to break the back of the old Southern view of "state sovereignty" that had just ripped the country apart for the better part of a decade.

How valid are these inferences about structure drawn from historical context? Significantly, I think, but only when understood as establishing minimum requirements, not maximum ones. Continuing with the Fourteenth Amendment, the historical context makes absurd the idea that the Amendment does not protect African-Americans from discrimination, but our knowledge that protecting one specific racial minority was the overriding motivation for its passage does not remotely suggest that applying it beyond the field of racial discrimination is improper. The inference is not that a constitutional provision only does the thing which its context makes clear was its central purpose, just that it almost certainly does do that thing, perhaps as well as other things commanded by the text.

Original Precedent
This level and the next level are quite closely related, but they are somewhat distinct, I think. Or, at least, it is my hope that they can be distinguished and untangled in any particular case, though this level will doubtless often be contaminated with the next. The concept at this level is that, however strong the general interest in following precedent, interpretations of a constitutional provision from very shortly after its enactment deserve especially strong weight. Unlike at the previous level I think this one is most naturally applied to the original Constitution and the Bill of Rights, where many of the Supreme Court Justices who created the formative case-law had been in the First Congress or at the Philadelphia Convention or what have you. The idea is that the original interpreters of a piece of the Constitution will have been better acquainted with what it was supposed to do, and how it was understood to operate in practice when it was enacted. That's true enough as far as it goes, I suppose, but we are now coming to the point where the originalism gets less and less valid, partly because the kinds of authority involved get weaker and weaker.

Consider the Slaugherhouse Cases. They were decided in 1873, about as "original" as you can get in interpreting the 1868 Fourteenth Amendment. And they got it wrong. Obviously, blatantly wrong. They held that the Privileges or Immunities Clause, which to all the world reads like a robust guarantee of substantive rights against intrusion by state governments, and which there is particular historical/structural reason to think was understood to refer to the (also wrong) interpretation in Corfield v. Coryell of the Privileges and Immunities Clause of the original Constitution which gave just such a list of substantive rights, instead protected only those rights peculiar to national citizenship, such as holding federal office or traveling between states. Taking arguments from original precedent as particularly binding would suggest following this precedent, as the Court has always done (though it has found ways to read the proper P or I Clause into the rest of the Fourteenth Amendment). But the 1873 Supreme Court had no authority to change the Fourteenth Amendment, so if we have compelling evidence from text and structure that they misconstrued it we have ample reason to disregard their interpretation and use the correct one instead. This illustrates a key point: text is prior to structure, structure is prior to precedent (or it should be, at least, if you're not making arguments in a lower court), and precedent is prior to policy. That doesn't change depending on whether you've got an originalist aspect to your analysis. Original precedent does not override even the least originalist of structural arguments, if that argument is sufficiently strong, and historical inferences about structure cannot override text.

Original Policy
Similarly, original policy preferences cannot override anything. We have reached the level at which validity goes to zero. People in 1868 thought racially segregated schools were great, and didn't violate the Constitution's new equality mandates. They were also racists. Perhaps that's not a coincidence. Of course, not all of them liked segregation, and not all of them thought it was consistent with the Fourteenth Amendment. But lots of them were racist segregation-lovers who went ahead and segregated their schools having just passed the Fourteenth Amendment. So what? They were wrong. And even if they weren't wrong, they had no right to bake their own opinions into the Constitution except by actually writing them there, which they didn't do. About half the country apparently didn't think there was anything about anti-sedition laws inconsistent with the Free Speech Clause of the First Amendment. They were wrong.

As I said, it can be tough to distinguish this level from the previous one. But I think it can be done, often at least, and wherever we find that some historical understanding of what a new constitutional provision did was based not on what it said it did or on any particularly sound legal principles we should consider ourselves utterly free to disregard that understanding. If there is no other particularly compelling argument from text or structure or actually respectable precedent to guide us, we may even substitute our own modern preferences or moral intuitions for those of the framing generation. The alternative would be deeply undemocratic. In a way this relates to the limitation on arguments from original structure. What people understood to be the central purpose of some part of the Constitution we should be highly reluctant to say it did not do, but if people understood some part of the Constitution not to do something simply because they all didn't want to do it, despite the fact that it looked like it did it if you just read the words or examined the structures, that's their problem. As Scalia himself has been known to say, it is not what they intended but what they did that binds us.



So that's my typography of originalist argument styles. The next time you see someone making an originalist argument, ask which kind it is, and then ask whether that method really gets them where they are trying to go. If they are invoking the original meaning of the text, ask how much that really resolves the ambiguities of the written document. If they are drawing inferences as to structure from the original context, ask whether they are trying to establish the inclusion of some principle or the exclusion. If they are appealing to original interpretations in court precedent, ask whether that interpretation is in conflict with the text or structure. And if they are appealing solely to the original prejudices and preferences of the framers, just ask why. There won't be a good answer. Keep these distinctions in mind, and hopefully you'll be able to avoid the kinds of bait-and-switch that Scalia so often does, claiming to be doing one kind of analysis and then moving to a different level without telling you.