From species to supertrees: Popperian corroboration and some current controversies in systematics

Abstract

Popper's ideas of falsification and corroboration have been used over a long period to justify cladistic parsimony to the exclusion of other methods for phylogenetic inference. This has been based on a restrictive interpretation that takes observed data as Popperian evidence, and sees Popperian background knowledge as restricted to 'descent with modification'. An alternative, 'inclusive', interpretation, developed over the past decade, does not focus on corroboration as an automatic by-product of supposed non-falsification, but instead focuses on the useful fact that the evidence (e) for an hypothesis provides Popperian corroboration only when it is improbable in the absence of that hypothesis. This perspective highlights key errors in cladistic arguments, including a critical failure to ever assess corroboration through examination of improbability of evidence and a failure to view background knowledge as playing a role in casting doubts about supposed positive evidence. The pitfalls of Popperian justification of phylogenetic methods are not limited to cladistics. Justifications of likelihood methods unfortunately have repeated these key errors of cladistic philosophy, including the failure to satisfy the Popperian corroboration requirement that tests of hypotheses be 'genuine', as indicated by assessment of improbability of evidence given only background knowledge. Criticisms of the inclusive perspective are recognised as incorrect, based on the realisation that evidence e can relate to the hypothesis, that background knowledge can reflect chance, that corroboration does permit tail probabilities, and that corroboration in the inclusive framework is linked to tests as attempted refutations and so remains compatible with falsification. On the positive side, the inclusive approach opens the door to growth of knowledge through assessment of evidence, with no philosophical restriction on the forms of evidence in phylogenetic inference. Whatever the evidence put forward, corroboration can be claimed if attempts to explain away that evidence fail, so providing an indication of the improbability of the evidence without the hypothesis. Further, the repeated use of a form of evidence may reveal its pitfalls—the kinds of occasions when it does not provide corroboration—so providing lessons that guide the future search for effective evidence in different contexts. This theme of resisting the a priori justification of one form of evidence, and focusing sceptically on many different kinds of evidence, is critical in the two other systematics contexts. First, a unified species concept only makes sense if there is a way to sort through the pluralism that remains at the level of evidence for species hypotheses. Second, for building the tree of life (TOL) we must combine various pieces of evidence from separate studies in a meaningful way to evaluate TOL hypotheses. In both cases, Popperian corroboration provides the necessary framework for judging how our evidence bears on our hypotheses.

Popper and phylogenetics, a misguided rendezvous

“…Some proponents of the likelihood method suggested that the degree of corroboration of a tree hypothesis is primarily indicated by the improbability of data as fit-as-evidence, rather than by how well the hypothesis stood against attempts to falsify it (Faith 1992(Faith , 1999(Faith , 2004(Faith , 2006Faith and Cranston 1992;Faith and Trueman 2001). In other words, not character-state distributions but goodness-of-fit serves as the evidence in a phylogenetic Popperian test, and instead of emphasising falsification, degree of corroboration is interpreted to depend on the improbability of evidence in the absence of the hypothesis, that is p(e,b) -the improbability of goodness-of-fit as evidence would significantly contribute to the corroboration of tree hypotheses (Faith 2004).…”

Parsimony and explanatory power

“…That directness proved illusory, for Faith never addressed Farris et al.’s (2001, p. 239f; quoted above) comments on likelihood and the obvious point probability P ( e ). Nor did he provide any explanation of how he arrived at his own position (Faith, 2004, p. 4f): 10…”

“…Faith’s (2004) argument from the Neptune example was even more obscure, as he cited Popper (1983) without providing page numbers. Fortunately, Popper (1983) mentioned Neptune in only three passages, beginning with (Popper, 1983, p. 237):…”

“…I called attention to Faith’s (2004) gratuitous conclusion of tail probabilities in a paper that I presented at the 2004 meeting of the Societé Française Systematique, in a session at which Faith was also present. Perhaps for that reason, Faith (2006) made no attempt to show by any quotation, or even gratuitous paraphrase, that Popper used tail or cumulative probabilities.…”

“…The situation was no better for the FCT assumption that “evidence” meant fit, Faith’s (2004) defense of which was also based on the Neptune example (Faith, 2004, p. 4):…”