On Probability and Systematics: Possibility, Probability, and Phylogenetic Inference

Abstract

In phylogenetic systematics, an ongoing debate has revolved around the appropriate choice of methodology for the construction of phylogenetic trees and inference of ancestral states. A recent paper by Mark Siddall and Arnold Kluge (Siddall and Kluge, 1997) advocates a privileged status for parsimony analysis, to the exclusion of other, statistically based, phylogenetic methods. Though hardly alone in championing this stance (see, for example, Kitching et al.'s 1998 textbook Cladistics), narrowly focusing on Siddall and Kluge's conceptual arguments justifying this position proves insightful. Rather than try to address every point made by Siddall and Kluge, I draw out two underlying general lines of argument that highlight assumptions that may lead to misplaced concerns and are in need of critical conceptual analysis. The two lines of argument that I identify are what I term Siddall and Kluge's (i) argument from falsificationism, and (ii) argument from probability. The first of these has been addressed elsewhere both by philosophers and biologists, and will merely be commented upon below. The argument from probability, though, is the primary focus of this article. I show that Siddall and Kluge's argument from probability is ambiguous, e.g., between metaphysical and epistemic possibility. Upon disambiguation, the argument from probability is either invalid, unsound, or simply misses the intended target. In working through this disambiguation, I precisely identify and clarify Siddall and Kluge's concerns, and show that statistical phylogenetic techniques ought not be considered problematic for the reasons cited by Siddall and Kluge. Broadly speaking, Siddall and Kluge have two main lines of argument implicit in their paper: (i) the argument from falsificationism; and (ii) the argument from probability. I will explore Siddall and Kluge's argument from probability in more detail in the following section. First, though, some brief comments on their argument from falsificationism. Siddall and Kluge's argument from falsificationism can be schematized as follows: F1 The desired scientific methodology is (some kind of) Popperian falsificationism. F2 Parsimony is a phylogenetic method consistent with Popperian falsificationism. F3 Statistical phylogenetic techniques are not consistent with Popperian falsificationism. F4 Parsimony is the only available phylogenetic technique consistent with Popperian falsificationism. F5 Therefore, parsimony is the only phylogenetic technique that conforms to the desired scientific methodology. Siddall and Kluge's argument from falsificationism is part of an ongoing debate in the systematics literature, e.g., the recent exchange in Systematic Biology between Kluge and DeQueiroz and Poe (de Queiroz and Poe, 2001; Kluge, 2001; de Queiroz and Poe, 2003). Statistical phylogeneticists have tended to argue against Siddall and Kluge in one of two ways. The first is to argue that statistical phylogenetic techniques do, in fact, conform with Siddall and Kluge's characterization of a falsificationist scientific methodology (i.e., to deny premise F3 and F4). The other strategy has been to argue that Siddall and Kluge are offering a mistaken interpretation of Popperian falsificationism, which does not qualify as a criterion by which statistical methods ought to be judged (i.e., to argue that Siddall and Kluge mischaracterize falsificationism in F1). There is a further question of how the falsificationism espoused by Siddall and Kluge resembles that which has been discussed in the philosophical literature (see Farris, 1983; Hull, 1983; Sober, 1988; and Hull, 1999, for earlier treatments of cladist characterizations of falsificationism) There do seem to be at least some important differences (e.g., there appears to be some incongruence over the treatment and classification of Fisherian statistics; Gillies, 1990; Urbach, 1991; Siddall and Kluge, 1997), though ultimately I do not think it much matters how closely cladistic accounts of falsificationism resemble philosophical accounts of falsificationsim. Philosophers have also evaluated Popperian falsificationism (though only a few have done so in the context of systematics (Hull, 1983, 1988, 1999; Sober, 1983, 2000). Most contemporary philosophers of science are critical of the idea that falsificationism is the only acceptable scientific methodology (Kuhn, 1970, 1996; Lakatos, 1970; Grunbaum, 1976; Kitcher, 1982; Giere, 1988, 1997; Salmon, 1998; Sober, 2000). Indeed, some philosophers have gone so far as to question whether falsificationism is even a very good scientific methodology (Howson and Urbach, 1993). The reasons for these objections are many and varied, and I will not rehash them here. Suffice to say that most philosophers of science would be mildly surprised that very few attempts have been made to deny premise F1 of Siddall and Kluge's argument from falsificationism. Though I think this strategy might prove fruitful, it is not within the aims of the present paper to explore this further. It should be noted that in denying the sole province of Popperian falsificationism one is not denying that testing hypotheses and (possibly) proving them false is an important component of scientific examinations. One must not confuse falsificationism with any act of falsifying hypotheses; to do so is to get caught in what could be dubbed the fallacy of persuasive terminology. Siddall and Kluge also argue for the privileged status of parsimony techniques on the basis of their interpretation of probability. What I term their argument from probability runs roughly as follows: P1 Phylogenetic trees are unique historical entities. P2 Probabilities cannot be assigned to unique historical entities. P3 Therefore, phylogenetic trees are not the kinds of things to which can be Statistical methods to phylogenetic Therefore, statistical methods that to phylogenetic trees are not to the of phylogenetic In what I of the and show that is either or Upon disambiguation, it is that the argument from probability is premise is I will ultimately an argument that Siddall and Kluge's concerns are argument both (i) precisely and Siddall and Kluge's and (ii) that parsimony and statistical phylogenetic techniques in a position to phylogenetic the of Siddall and Kluge's it is to their of and with their of historical as There in and A exchange between a and a how it is to the is the of on It simply is the that has or could elsewhere in the have a good idea of the and for on the of appropriate and from would not that can the of there the might have been I question to a kind of question is as to that the in this is to historical on this is a to the term the context of this I have argument for this an to the to whether elsewhere in the it that the context of the term in this question would to as kind of and as the as on even Siddall and Kluge's it that on has or could elsewhere in the on might have to a from a part of the and may to other it might to either of these out to be on is merely a part of in the There concerns this that point to in Siddall and Kluge's Siddall and Kluge are to on to be a unique historical to as to the of on and that of of things historical are of metaphysical possibility. should what is by and how does this to the scientific is what is and what is not not out by what by the of by the of and by the of possibility. the to be a over metaphysical possibility. is that which is not out by is in it is not not Most philosophers argue that it is that the could have been than the it is to that the the is is not the the to that for it to be that might not have been is to say that it is that could have been to that this is not on what do of the so it is not that Siddall and Kluge are of of metaphysical possibility. for them to a this appears metaphysical is a to in scientific to be with which there is a what is that should that as a what or hypotheses are consistent both with what do of that and also with other hypotheses to which are I will this to it from the metaphysical by Siddall and Kluge. not out by either what or the other scientific to which are is a scientific possibility. Philosophers would this as a of epistemic and within the scientific of scientific may not simply be as the by Siddall and as to scientific A question to is are Siddall and Kluge with matters of and The can be in what their What Siddall and Kluge is that a might have been the from (e.g., on to (e.g., the of things that are as and have the of in systematics (i.e., ought to be considered more 1997; Hull, 1976; Siddall and Kluge are in their have a to to phylogenetic Siddall and Kluge are not for the that a of which their metaphysical and scientific to between the historical and phylogenetic The term is can to either or phylogenetic other The term to either the historical of or to a of that of is is from the context in which it is and this is not the term as I it should be as to of the historical of also phylogenetic should be as a or of a of the historical of to of the is to an historical that is a part of the unique historical of to of a phylogenetic is to a the how Siddall and Kluge their following probability to to there has to be a of only one can be are of trees is by to the that there is only one historical though, that Siddall and Kluge's is a metaphysical possibility. this how most would a of the is not are not denying that there is a unique historical are simply that the of a of or hypotheses the are consistent with what of that epistemic of trees can be as a scientific not the metaphysical by Siddall and Kluge. to scientific the is over things a of hypotheses a Siddall and Kluge this one what the hypotheses are in for there are of which are of part of the premise for any phylogenetic of The is that the of the is not for (e.g., the of discussed in et the is in and in is not a metaphysical There are a of hypotheses or that the in question and are consistent with that The question for of is how to which of these is or with scientific there are many methods available to the is one scientific method Kluge, to systematics, this is so it is in parsimony techniques There are other scientific methods of statistical methods. can be assigned to with an the of probability. can also hypotheses though these are speaking, of of the of scientific Siddall and Kluge confuse by not between scientific and metaphysical in systematics The with the is that it to from this as trees are assigned a probability, and more than one can be the are not of that not phylogenetic trees are assigned the probability. by Siddall and Kluge, the phylogenetic with the probability might also be considered the is not to Siddall and Kluge's characterization of techniques in as I one could techniques this to this is and A for example, would be to techniques as a 1997; The main as is that Siddall and Kluge to between the historical and phylogenetic In the it is which of are on one from one to the other in the Siddall and Kluge are also assigned to or with assigned to or is an important which will be discussed in more detail below. The point of this is to highlight Siddall and Kluge's to between are of the historical and are to phylogenetic the is a unique historical phylogenetic trees are not historical or hypotheses of that premise P1 of Siddall and Kluge's argument from Phylogenetic trees are unique historical entities. speaking, P1 is It to the important I the following as a to a argument that will Siddall and Kluge's important The historical of is a unique historical which is in Phylogenetic trees are hypotheses the of the Siddall and Kluge to be with a interpretation of probability. In with whether of probability can be assigned to historical entities. Rather than historical I to is a historical entities. the question whether a probability can be assigned to a Siddall and Kluge are that to a probability to a fact, this is a for present can it there are other of probability available that do for an this is so it is to these of probability, and how a to be to the made between metaphysical and scientific so is there a between and of probability. of probability are that probability to be a of the that of of probability, on the other to be of of a of some or of the do not in the of A brief can draw out the of this for example, that I a that to be though the of that that I both an and a what the probability that the would upon The might with by probability, I can say of the is that the probability of that is not The probability of the is that can upon and that the is the probability cannot be The on the other might to the question as by probability, I have for that the is either or so the only of is that it is to be in either I can a probability to the that the will Upon and will be in of one is not to be what kind of interpretation of probability is there is of of probability and the at In the both to of the as the on which classification of probability Siddall and Kluge are most the of in for example, the is to in a and to get a the is has only a of a this to the of In the first to get a in at is the have are to the either will or will not get a and there is probability that can be assigned to that one on one alone is will prove can how a (i.e., ought to be considered with to in a of probability. will some of the over in the Siddall and Kluge of The that I will are the and of probability. The interpretation of probability as the of an in a of is very the is or or or is of of the interpretation of probability. Though these are important for the at can be and I will of these as though need only be very can be by the of an in an of and from this to a of probability are that probability is a in the of are of things in the Most a probability simply is the of an in the or of Siddall and Kluge point any is a which is part of the of in which the of either or does not A in does not a can a be from a the only that could be is or as Siddall and Kluge is It is for this that are not assigned to the probability of simply is the of that in the appropriate the probability of is on a what would it to a probability to the historical with the of the probability would be on some There are two here. as Siddall and Kluge point is that it is not what the might is the probability at the of these One for the probability is the probability of the historical a of this can not be it a whether are on a (i.e., of or a kind of (i.e., the the probability is more probability here. is the that Siddall and Kluge to be the of their are not this is by the of and the of the historical that would be of a from the that the of simply are not of this the are what the probability is of the historical on some The is that are not in a position to this question as it is one simply do not with the of the historical that is what is at to try to address the would have to the question as to be the probability of some phylogenetic some though, that are the probability of a phylogenetic on some far from to the of There are two things to here. The first is that simply are not to the probability of phylogenetic trees some the probability this to the scientific and more than phylogenetic do so on (i.e., of it the that the could be and as would phylogenetic trees on The to is that even to phylogenetic trees on are not assigned to historical or are assigned to which a kind of the of even Siddall and Kluge are much of their their or I that the appropriate on which a probability can be is is a by any probability assigned in the that the probability of is on a the is a of many of which may have a of are to as the that might be over the of as a of the or in a against or In fact, the of in these or of these has the of and might a the of in the The might argue that the are made by a very of and that it is not at for the of an in a of a to from the of that in the The though, is with two one the may be to qualify as a from which a probability can be The that a are to a the other is a What does a probability have to any of the the of an in any might from the of that in the or on what can that a probability has any to any and other are more as In fact, the of the interpretation of probability, and to address these it is a brief at these are of the of probability, and have to address these One of the first of these One of the primary to the to to the interpretation of the ultimately to the it that must be must be of the these out to be to It is this which to the status of within the interpretation of probability. the of probability, the is an Rather than identify probability with the of an probability as the of an to The interpretation probability as to a which is to be and it on the that a probability is a of some with the in is by of and in a that probability may be to be a of the the of this with to probability to a probability as probability of a this a to the probability of a an can say that the a a probability to the that it is an or in with the than to the that it is a of a In this a may have a probability even though it may only for probability is a of can Siddall and Kluge's between the and the the historical as a might the probability of the as a of the in which this are by as of to of the interpretation to the that are of these is to to the probability of a as a of the to be by a or statistical than by an upon how Siddall and Kluge are to I to that Siddall and Kluge in the strategy of their argument from probability. to by the of to there is what Siddall and Kluge should have is that the in methods ought to be as The would be whether are an appropriate in the of phylogenetic strategy would as as the of probability is consistent with the of probability, must also a and of how ought to be It may be that this is a to probability There are many of and some that any of a interpretation that the ought to be considered some kind of 2000). Though in systematics as may prove fruitful, as I not of any to in any phylogenetic of will have to be in of Siddall and Kluge's to an from to the either will or will not get a and there is probability that can be assigned to that one on one alone is is and the at whether a is or not upon the one has been It may be to most as a of a on a might be more this to a of probability 2000). is important for at least two First, it is that do, in fact, on the may be to do that an of how probability with probability is a of the of a has in a and these can be and can be by upon it is can be by to are with the One of is that a probability the probability for the a could a probability to the that will get a in a The would need to on as much as e.g., how at with on in the The of the and of to things in a this conforms with It is to a will get a or of is part of the of the of many The are how to which is and it it would be a much at least much to also conforms with of things The more a is with a the is to both which is and that to though, is not only of the also of the and the must identify the the must identify the and the identify the upon which to are much a will in a probability, or a or a probability will be to how much is available to The one is with a the more one will be in the assigned to of that The of this of of probability to premise P2 of Siddall and Kluge's argument from Probabilities cannot be assigned to unique historical entities. it P2 is not precisely to the of probability available to the get the of argument be to Siddall and Kluge, that most closely resembles the in their cannot be assigned to can be assigned to can be assigned to of The at P3 does not from the get the Therefore, the historical is not the kind of to which a probability can be to be consistent with what Siddall and Kluge are in of their of historical and their characterization of probability (see The in the is an and things ought not have assigned to as the by is a scientific need to a Phylogenetic as hypotheses of the of of the are of are more or by the of phylogenetic trees as a scientific as

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 17 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.