That quote comes from the conclusion of Judge Cecilia Altonaga’s Order granting defendant’s motion to exclude plaintiffs’ experts in In Re Denture Cream Products Liability Litigation. If you’re interested in the problem of causation in toxic tort cases this little gem would make a great primer.

Of particular interest are the following: (a) the court’s dissection of the experts’ arrangement of bits of unrelated data into what to the casual observer might appear to be a sound deductive argument; (b) the court’s discussion of causal inference via induction and its implicit requirement that experts assess ( weigh) and then disclose the strength of their belief in each premise of the argument; and, (c) the court’s recognition of the problem of selection bias when paid experts are allowed to opine as to causation using a so-called differential diagnosis or differential etiology approach – they tend, presumably subconsciously, to populate the line-up of suspect causes with the defendant’s product and three or four easily ruled-out alternatives.

We especially liked the following about the opinion of plaintiffs’ experts: "This theory is not ridiculous, but neither is it necessarily true; it is ripe for testing." That’s what we said about Milward. All the plaintiff had was a plausible yet untested (because, said her expert, it was untestable) theory. Traditionally, that hadn’t been nearly enough.

There’s a clear split between the circuits with most requiring evidence that a theory’s prediction has been confirmed at least once before it can go to a jury while a couple of others require only plausibility and an expert willing to say that it’s so based on his or her scientific judgment. So it’s the empiricists vs. the epistemological anarchists; Popper v. Feyerabend. If the Supremes decide to have a look it’ll be a very big deal.