A Supreme Court opinion yesterday exposed the dirty little secret of textualist interpretation: That it is at most a means to throw out the least-plausible meanings. It does not, and cannot, lead to indisputably-correct interpretation. Especially when there's money or "policy" involved.
First, a methodological — mathematical — premise. How might one solve this equation:
x = cos(a+b) / sin(a-2b)
for arbitrary values of a and b? Speed-demons who rush to do simple algebra as fast as they can (that is, lawyers) will either defer a boundary-check to later or not do it at all, because there are multiple conditions under which the equation has either an undefined result or no meaningful result. The most obvious one is easy to show:
if a = 2b then
x = cos(a + a/2) / sin(a-a)
x = cos(1.5a) / sin(0)
x = cos(1.5a) / 0
oops, undefined (divide-by-zero error)
And now the fun with words begins.
Yesterday's decision in HollyFrontier Cheyenne Refining, LLC v. Renewable Fuels Ass'n, No. [20]20–472 (PDF) depends upon a different kind of exception that leads to an undefined result. The simplest, and easiest, way to put the question in context is this: When an exemption has lapsed through passage of time, and a statute passed in 2005 (amended in 2007) allows an otherwise-eligible beneficiary of that exemption to apply for an "extension", does that lapse cut off the right to an extension? That is, does the lapse mean that it's no longer an "extension" that is being applied for, but something else? Thus followed the battle of the dictionaries and other direct evidence of the meaning of "extension" Friday morning in St. James's Library…1 Omitting repetitions and cross-references between the opinions:
| Justice Gorsuch (majority) |
Justice Barrett (dissent) |
|---|---|
"extension" not defined in statute (slip op. at 4)
not imposing uniformity across all law (8)
|
Most natural reading, not just any possible reading, is best baseline
|
My focus is on the evidence offered, not interpretive methods based on that evidence (see, e.g., Justice Barrett's dissent at 7–8 citing the… umm… neither-precisely-objective-and-disinterested-nor-semantically/semiotically/culturally-inclusive A[ntonin] Scalia & B[rian] Garner, Reading Law: The Interpretation of Legal Texts). And in this sense, we're back in St. James's Library with a potential divide-by-zero error at issue in both opinions.
But nobody did any boundary checking. The real question here is not what "extension" means to a government official or lawyer, because this statute isn't for government officials or lawyers. It is for, and directed to the conduct of, "small" operations in the oil industry. The probability that most of them even know about the Oxford English Dictionary, or the New Oxford American Dictionary, is fairly limited. (Hell, the probability that the Congresscritters and their staffs knew about them — let alone referred to them — when writing this legislation is also fairly limited.)
Most of all, HollyFrontier is about bad writing arising from failure to boundary check. Let's jump back up to that initial equation at the top and consider this: Is x a meaningful result for our purposes if it's a negative number (slightly less than half of the circle, and we are just not going to get into non-Euclidean geometry here)? If the Congresscritters, their staffs, and the industry lobbyists had had enough brains to flip a light switch, the specific circumstances would have been foreseeable, and nobody would have allowed a statute that required this level of gyration regarding a nontechnical term. It could have been solved by defining "extension" to require continuity (or not); or by requiring that the application be submitted while an exemption was in force (or not); or by adding the word "continued" in front of "extension" (or "not-necessarily continuous"). This dispute is a consequence of bad writing. And it resulted in more — reams more (since these briefs, etc. were submitted on paper) — bad writing.
This dispute was also resolved improperly, by treating "data" and "reasoning" as of equivalent weight and priority. But that's for another time, and falls definitely — oh, so definitely — on the side of the Moderns.
- Fortunately, because this is a relatively recent statute, we need not delve into corpus linguistics analysis and confront the racial, gender, religious, and class biases inherent in those sources, nor the preservation problems, nor the applicability of the kind of discourse in written sources to "ordinary meaning" in a largely nonliterate population. Neither the Ancients nor the Moderns are at issue; they're all Moderns. (Obviously not Postmoderns; we are talking about Congresscritters and judges and oil barons here.)
- Editions omitted for brevity; they are all recent enough that linguistic and colloquial drift are probably minimal. On this point, at least, Justice Gorsuch has by far the better of the argument (slip op. at 7) — at least for nontechnical terms. Were we arguing about the meaning of the term "influencer," however, even five years might prove too great a gap.
- disclosure The Oxford American Dictionary has long been my preferred source to ferret out both the edges of the ordinary meanings of words, in ordinary American English as used by native speakers of American English, and the scope and sources of exceptions. I ensured that more than one performance report (with preprinted blocks on the forms) and more than one set of other paperwork has "benefitted" from it over the years…
- Here, however, there is a problem. Anderson's Dictionary was published in 1889, and reprinted in 1996 (thus the "1996" quotation). It is impossible to determine whether the particular reference was updated (whether in the reprinted edition or during the course of legal research, because this particular citation by Justice Barrett is through/to a brief) without going back to the 1889 edition… and the Bluebook discourages that level of inquiry. Which says volumes, in and of itself, about battles among legal books… and the ethics lying deep behind legal research.