There are different types of law review articles.  But over the years I have had a handful that I call my “mad scientist” papers. These papers spring from ideas that honestly seem a bit preposterous when they come into my mind but have enough heft that I play around with them to see if they write.* 

For example, one of my first articles argued that Erie Railroad Co. v. Tompkins should be understood as a nondelegation case; after all, if Congress gave the President carte blanche power to create rules governing contract or tort disputes between citizens of different States, surely that would be too much delegated legislative power, right?  Why should federal courts be different? That idea seemed off-the-wall when I first thought it, but also interesting enough to write about. 

Later, I remember thinking about Seminole Rock deference, and then out of the blue I wondered whether Chenery II was a “substitute” doctrine such that, if the Court were to overrule Seminole Rock, agencies could substitute away from ambiguous rules in favor of adjudication, i.e., no rules at all. After batting the idea around, I wrote it up.   A few years later, I wondered – again, out of the blue – whether ossification sometimes is a good thing for regulators because it allows them to make credible commitments, thus inducing long-term investment. After laughing at first (how could ossification be good for agencies?), I wrote that paper, too, because after thinking about it and talking it through with smart people, I didn’t see any obvious flaw in the theory. Several years later still, I wondered whether Chevron deference may be problematic not (just) because it enabled agency to act too aggressively with respect to big problems, but also because it created a collective-action problem with Congress that discouraged both Congress and agencies from solving small problems. So, I wrote article, too.

These are “mad scientist” ideas — ideas that may be right but, if not, at least hopefully are wrong in interesting ways.

In that spirit, I’m pleased to announce my latest “mad scientist” idea: Aggregating Original Meaning, forthcoming in the Columbia Law Review. This is an idea that has been gnawing at me for years and I’m pleased to finally see it written down.

Here’s the abstract:

Public-meaning originalism—the theory that the meaning of legal texts, including the U.S. Constitution, is fixed when authoritatively adopted and binding until authoritatively changed—is ascendant. Although scholars debate this theory, a majority of U.S. Supreme Court Justices describe themselves as originalists and advocates increasingly advance arguments sounding in public-meaning originalism. And even jurists who are not originalists often recognize history’s relevance.

The Court, however, has overlooked a conceptual puzzle—and so has everyone else: What should originalists do when deciding a case with multiple steps, each requiring its own originalist analysis? For example, courts routinely decide what a constitutional provision means and whether it has been incorporated against the States. Courts also determine whether a right has been violated and what the remedy should be. And courts often decide whether the Constitution creates a power and how to apply it to “edge” cases. Because finding original-public meaning is difficult, however, judges cannot be certain they have resolved each step correctly. Nonetheless, after resolving one step, they proceed to the next step as if resolution of the first was certain. Such inattention to possible error runs headlong into the mathematical concept of aggregation, which requires using the product rule or conditional probability when assessing the likelihood that multiple claims are true.

This Article introduces aggregation into the originalist literature and demonstrates how the premises of public-meaning originalism support accounting for aggregate uncertainty. Yet although aggregation may help judges reach decisions more consistent with original meaning, its disruptiveness also heightens the significance of many other concepts and debates, such as what it means to “prove” the law, the strength of stare decisis, the scope of construction, the validity of liquidation and tradition, the legitimacy of compensating adjustments to offset non-originalist precedent, and the value of corpus linguistics. Aggregation, however, should be more than just disruptive. Rather, it may add at least some additional stability to the law by shoring up familiar features of constitutional doctrines that originalists thus far have struggled to defend, including the presumption of constitutionality, clear-statement rules, and hybrid rights.

Constitutional theory is not my usual field, but the paper has a lot of applications for admin law. 

Here is the full draft. The article won’t be published until later this year (or early next year), so lots of time to incorporate comments. I’m excited about this idea — even if it is a bit mad.

* Most don’t.  But I often reflect on this observation from Frank Easterbrook (that I first read via Will Baude): “A free mind is apt to err – most mutations in thought, as well as in genes, are neutral or harmful – but .… [h]igh risk probably is an essential ingredient of high gain.”