20 January 2015

Joe Friday Was Nowhere to Be Found

The Supreme Court has begun issuing opinions for the October 2014 Term of Court. An opinion in a patent case — that, even more than normal, is actually a question of civil procedure — may turn out in the long run to be a critical commentary on judicial competence with reach far beyond patent law.

What is a "fact," under Fed. R. Civ. Proc. 52(a)(6) (which ultimately draws its authority from the Seventh Amendment, although no jury finding is actually at issue in this particular matter) that must be reviewed only for "clear error" by an appellate court? The majority opinion,1 by Justice Breyer, says that a judge's decision to credit one expert's testimony concerning what biochemists mean when they refer to large molecules as having a range of molecular weights over another expert's testimony on the same issue is a "question of fact" that must be reviewed for "clear error." Justice Thomas dissents that because the context in which the judge made his finding is the purely legal question of claim construction for a patent, the judge's decision was a "question of law" that is reviewed entirely anew ("de novo") by an appellate court. Leaving aside that both sides are wrong from the perspective of a chemist2 and deserve to be condemned to a semester of washing glassware in the lab (where they might actually learn something about the techniques by which molecular weights of complex biological compounds are actually determined), Teva is yet another in a line of cases in which judges and lawyers define what is "fact," and therefore not within the judicial competence to question, and what is "law."

Somebody does have to answer this question, because there really is a distinction to be made. The problem is that the wrong people, with the wrong backgrounds, are making the determinations of what may be reviewed only for "clear error"... and what constitutes "clear error." One need not go so far as eppur si muove; tobacco litigation is much nearer. Ultimately, though, this isn't even an error of fact, or an error of lab technique; it is an error of writing that should have been challenged by every lawyer in the chain, by the Patent Office's examiner, and by every judge. It is not indefinite; it is unclear, but only with 20-20 hindsight by a party seeking to evade (and not to learn or to advance knowledge or practice itself) — and that is a different kind of error that is badly managed by the law. In Teva, the fundamental error was in which kind of "experts" were called upon in the first place: The "correct" experts would have been the managing editors at high-end journals of chemistry and biochemistry, because the art in which one must be skilled is not that of lab technique, but of lab writeups. That assumes, of course, that an "expert" is necessary in the first place... when the legal doctrine that patent claims are limited to what the applicant actually demonstrated should have led elsewhere.

Teva has disturbing implications for non-patent statements of fact, too. I've ranted repeatedly against overinsulation of "legislative facts" from review, and sometimes those intersect. Sometimes a legislature makes a factual statement concerning abortion, or global warming, or disease transmission, in the factual preamble embedded in a statute; sometimes it's buried elsewhere in the legislative history that Justice Scalia is so adament has no meaning. It's even more apparent in administrative proceedings (such as determining the allowable concentration of a pollutant, and the way that is to be measured). There's no place for Joe Friday in Teva: He says he just wants the facts, and this is not an instance of the facts being unclear or contested — it's an instance of deciding what might be a fact in the first place.

But there's a meta-leval implication of Teva that is at least equally important. It is quite possibly right that judges are not competent to determine scientific facts. Or, as Justice Holmes noted in Bleistein, originality or creativity or damned near anything else related to the factual question of copyright infringement. What that says is that we're choosing the wrong people to be judges more than it says that these questions cannot, in fact, be resolved. That's not a good thing. If the judicial system were more willing to admit that the finality of its pronouncements does not, in fact, provide much more guarantee of truth (or factual accuracy) than any other appearance of words in print does, we'd be better off... and not questioning exactly how many generations of imbeciles are enough, for some value of "generation" and "imbecile."

Ultimately, though, there's yet a further meta-level problem with Teva. Contrary to his catchphrase, Joe Friday wasn't looking for facts. He was looking for story: A particular kind of story supported by the individual factual circumstances (whatever, metaphysically, they might be). That is also, ultimately, what is at issue in Teva: What are the boundaries of the story being told by this patent application? In this particular instance, legal doctrine says "authorial intent is the touchstone of interpretation"... and then proceded to not introduce the primary evidence of authorial intent as the critical elements in making its determination. All in all, this is a bad job by all of the lawyers involved, not excluding the judges. That it may have turned on a question of "what evidence is admissible?" just makes the absence of scientists from the decisionmaking process that much more apparent — and inexcusable.


  1. Teva Pharmaceuticals USA, Inc. v. Sandoz, Inc., No. 2013–854 (PDF) (20 Jan 2015).
  2. A.B. Washington University (St. Louis). As I've remarked before, sometimes the unjustified absence of scientists and engineers in the judicial system — and the legal profession as a whole — comes back to kick all of us in the gonads. This is one of those times.

    This is not a difficult issue: It was the lawyers' fault, and it's not factual at all. The proper practice would have been to specify the method explicitly, and that would have been required at the better journals in the chemical and life sciences had this patent application been an article submitted for publication. Further, even if the method had not been specified in the article, it would have been disclosed at one point or another in the review process, in lab notebooks, or other documentation — as opposed to lawyers' glosses on it. That the statement of the record includes no reference to any of this indicates that this is not a fact issue, but a typical lawyers' (all of them, including the judge) misunderstanding of what either a "fact" is or how science works.