On the difference between mono-, holo-, and paraphyletic groups: a consistent distinction of process and pattern

Abstract

About 50 years ago, the German entomologist Willi Hennig presented a new approach in biological systematics that he called a phylogenetic systematics. The main difference between his approach and traditional Linnean systematics was that he distinguished two new kinds of groups that he called mono- and paraphyletic groups, and whereof he considered only monophyletic groups to be natural groups. However, almost immediately after publication of his approach in English, some biological systematists commented that his monophyletic groups rather ought to be called holophyletic groups. The comment sparked a heated debate about the definition of the concept ‘monophyletic groups’, but the debate never reached consensus. In this paper, I claim that the controversy does not concern the definition of the concept monophyletic groups per se, but instead conceptualization of phylogenies (i.e. dichotomously branching processes) in a general sense. I discuss the relation between mono-, holo- and paraphyletic groups, and conclude that Hennig's conceptualization of phylogenies is both inconsistent and empirically wrong, whereas Linné's instead is consistent and correct.

INTRODUCTION

When Hennig (1950, 1966) presented his empirical analysis of phylogenetic relationships between biological species, he also distinguished two kinds of groups that he called mono- and paraphyletic groups, and then claimed that ‘only monophyletic groups are natural groups’. The distinction and subsequent claim founded a philosophy that became known as cladistics, and this philosophy today pervades biological systematics.

However, almost immediately after Hennig (1966) published his writings in English (in a book), Ashlock (1971, 1972, 1974, 1979) commented that Hennig's ‘monophyletic’ groups rather ought to be called ‘holophyletic’ groups because they not only include the monophyletic group, but also its ancestor. Ashlock's comment sparked a heated response from a group of biological systematists defending Hennig's terminology by discussing the priority of the definition of the concept ‘monophyletic groups’ (Nelson, 1973; Farris, 1974, 1990; but see also Mayr, 1974, 1978 and references therein). Ashlock and Hennig's defenders never agreed on the term or its extension, thereby leaving the controversy latent within biological systematics, sporadically emerging in similar arguments (Mayr & Bock, 2002 and references therein).

Here, I argue that the controversy does not concern the definition of the concept ‘monophyletic groups’per se, as Hennig's defenders turned it into, but instead the definition of the term ‘monophyletic’ in terms of the definition of the term ‘group’, and thus ultimately in terms of the term ‘thing’. I argue that the controversy concerns conceptualization of dichotomously branching processes on a generic level (i.e. of a consistent distinction of pattern and process on a processual level), and that Hennig's conceptualization is both inconsistent and empirically wrong. Processes of this kind can we jointly call phylogenies, and things that participate in such processes are, for example, mitochondria, cells and biological species (according to Darwin's model of the origin of species). Understanding of my argumentation is, however, greatly enhanced by thinking about phylogenies in terms of physically cohesive things, such as mitochondria or cells, instead of biological species, to prevent confusion of single things with several things (i.e. of single things with kinds of things).

ASHLOCK's POINT

THE CONFUSION

Hennig distinguished two kinds of ‘groups’ in phylogenies that he called mono- and paraphyletic groups. Now, Ashlock did not question Hennig's distinction of these kinds of groups per se, but only pointed at the fact that also the kind of group Hennig called paraphyletic are monophyletic. Also they have a single origin and hence are ‘mono’-phyletic. At that time, Hennig's defenders comprehended the point as a matter of definition of the concept ‘monophyletic groups’, but today they instead meet the point by the argument that paraphyletic groups do not include all the descendants of a common ancestor, but only some of them. However, this is actually Ashlock's point: the fact that Hennig only refers to the monophyletic groups that include an ancestor and all its descendants (not to those that include an ancestor and some of its descendants) means that this kind of group rather ought to be called a holophyletic group because it excludes the monophyletic groups that are paraphyletic.

Ashlock only tried to explain (which Hennig's defenders obviously also understand today) that there are two kinds of monophyletic groups: holophyletic (including an ancestor and all its descendants) and paraphyletic (including an ancestor and some of its descendants), and that Hennig confused monophyletic groups with holophyletic groups by calling holophyletic groups monophyletic groups. Hennig used mono = holo despite that mono = holo and para; He used mono = all despite that mono = all and some. He used ‘mono’ as if it is exclusive (i.e. excludes ‘some’) despite that it is not. He equalized one concept with one of its kinds by calling the kind by the name of the concept (like calling ‘yellow’‘colour’ and excluding ‘red’ from ‘colour’). He equalized a generic concept with one of its specifics by calling the specific by the name of the generic. The remaining specific, ‘para’, did he dismiss as not being ‘natural’. What this ‘dismissal’ actually means is difficult to understand, but it does of course not turn paraphyletic groups nonmonophyletic. Hennig can of course only confuse the concept holophyletic with the concept monophyletic; he cannot make paraphyletic groups nonmonophyletic. Paraphyletic groups will remain monophyletic independently of whether Hennig confuses the concepts holo- and monophyletic or not.

If Hennig had realized (acknowledged?) that he only referred to those monophyletic groups that include all descendants, he would have had to agree on Ashlock's correction that this kind of group rather ought to be called holophyletic groups. It would have forced him to admit that the distinction he called ‘mono- and paraphyletic groups’ actually splits monophyletic groups into holo- and paraphyletic groups, and that he thus actually claims that only holophyletic groups are natural groups (i.e. that he actually only recognizes holophyletic groups). However, this restriction is inconsistent because holophyletic groups include monophyletic groups, which, in turn, include paraphyletic groups. It is impossible to exclude paraphyletic groups from holophyletic groups because holophyletic groups include paraphyletic groups. In short, mono-, holo- and paraphyletic groups are actually two kinds of one kind (of group), and do thereby include each other independently of which kind we start with. Such complex of kinds is an enigma for a person who believes that kinds (i.e. concepts) exist, but is instead conceptualization itself for a person who does not. This particular such complex is simply monophyletic groups, which I will return to later.

The distinction ‘holo- and paraphyletic groups’ is moreover wrong in an empirical sense. Its problem is that it rests on a definition of single things as single line segments, because this definition equalizes time and space and does thereby contradict the empirically confirmed relativity of time. Time equalizing space does effectively exclude any possibility for time to be relative (i.e. to vary with speed in space). This incompatibility between the distinction and empirical evidence may appear irrelevant, but it does none the less falsify the distinction empirically. The falsification actually means that such things (i.e. kinds of groups) do not exist, nor can possibly exist per definition, because they are self-contradictions per definition. They are incompatible with single things; if they would exist, single things would not exist, and vice versa.

MONO-, HOLO-, AND PARAPHYLETIC GROUPS: HOW TO TELL THEM APART AND HOW TO JOIN THEM

THE DIFFERENCE AND RELATION BETWEEN MONO-, HOLO-, AND PARAPHYLETIC GROUPS

What, then, is the difference and relation between these kinds of groups? The answer is that holo- and paraphyletic groups are two kinds of processual parts of phylogenies (i.e. holo- and paraphyletic processes), whereas monophyletic groups are the existing parts of these processes. Process is real in light of things, and things are real in light of process, and holo- and paraphyletic processes are thus real in light of monophyletic groups of things, whereas monophyletic groups of things are real in light of holo- and paraphyletic processes. Holo- and paraphyletic ‘groups’ are non-existing processual parts, whereas monophyletic groups are the existing parts of these processes. This is why holo- and paraphyletic ‘groups’ are kinds of monophyletic groups. Monophyletic groups are ‘natural’ groups (i.e. existing, historically cohesive groups) that can change (i.e. evolve) and, similar to single things, they are thereby patterns in time, whereas para- and holophyletic ‘groups’ are such groups over time before and after change. All processes contain parts that are before, respectively, after change, and existing parts (presence) that are both before and after change (i.e. the things that changes). These parts are actually the parts that we use to conceptualize things, and they form complexes of the kind mentioned above. We conceptualize single things as one part that is before change and one part that is after change, and one existing part that is both before and after change (i.e. the part that change). Aristotle explained conceptualization by calling the things that change ‘genera’, and the parts before and after change ‘species’. Using this terminology, we can call monophyletic groups genera, and holo- and paraphyletic processes species. Monophyletic groups are the existing parts of holo- and paraphyletic processes. The fact that neither holo- nor paraphyletic processes exist may appear counterintuitive to a person who has looked too much on illustrations of phylogenies, but it is easily verified empirically by the fact that it is impossible to see a group consisting of an ancestor and its descendant(s) in reality. An ancestor in a dichotomously branching process does of course disappear in the moment its descendant(s) appears. However, the whole group has historical cohesion over the dichotomous branching, and is thus real and existing as a monophyletic group over the branching in this sense. Monophyletic groups are the concrete expressions (i.e. the existing parts) of holo- and paraphyletic processes in each moment of time. These definitions of the concepts and the relation between them is not a matter of definitions of the concepts per se, as Hennig's defenders tried to turn it into, but instead of understanding the concepts in terms of our conceptualization of things (i.e. in terms of process and pattern). Only this understanding equalizes single things and monophyletic groups of things consistently, which is a requirement for a consistent conceptualization of phylogenies. Neither pattern nor process can by themselves accomplish this equalization.

THE CONSEQUENCES OF HENNIG'S CONFUSION OF HOLO-AND MONOPHYLETIC GROUPS

Hennig's confusion of holo- and monophyletic groups is a simple confusion of the concepts time and space, and thereby also of the concepts process and pattern, before and after, ancestor and descendant, the abstract and the concrete, our comprehension of reality and reality, etc., and thereby in practice an exclusion (i.e. denial) of both change (i.e. evolution), the difference between reality and our illustration of it, and thus also of things. However, the confusion does not actually delete reality in front of our eyes, but only denies it as well as our traditional way to discuss it. Mono-, holo-, and paraphyletic groups themselves remain unconfused because they are just as real as the difference between reality and our comprehension of reality is, and thus just as real as our conceptualization of single things in terms of process and pattern is. Hennig's confusion only denies them.

HOW TO AVOID THE CONFUSION

We avoid Hennig's confusion by realizing that the external line segments in phylogenetic trees represent single things (i.e. single pieces of pattern), whereas the internal line segments represent single processual parts (i.e. single pieces of process) each consisting of two things in a row (i.e. two pieces of pattern): the first a descendant and the second an ancestor. This helps us keep holo- and monophyletic groups apart: the first including both the descendant and the ancestor of internal line segments by their ancestral line, and the second including only the latter, thereby keeping process and pattern apart, and thus also keeping the illustration and reality apart. These distinctions are extremely important for empirical science because confusing them thus contradicts empirical evidence, thereby leading onto the confusing road of belief. Consistent conceptualization of phylogenies requires realization that the internal line segments represent two pieces of pattern in a row.

CONCLUSION

A consistent dissection of the controversy between Ashlock and Hennig's defenders reveals that Ashlock (and Mayr) was consistent and right: Hennig did indeed confuse holophyletic groups with monophyletic groups, either unconsciously or on purpose. The consistent and empirically correct comprehension of these concepts is that mono-, holo-, and paraphyletic groups are inter-related by holo- and paraphyletic groups being processual parts of monophyletic groups. Monophyletic groups are existing, historically cohesive groups that can change, whereas holo- and paraphyletic processes are such groups before and after change. Monophyletic groups exist in time, whereas holo- and paraphyletic processes instead are distinguished over time. It means that these kinds of groups have to be kept apart, but are impossible to separate. They are defined with respect to the definition of things and to each other in a conceptualization of monophyletic groups in phylogenies. Monophyletic groups can we organize consistently using nested categories (i.e. categories of categories) in a system of the Linnean kind. The Linnean classification is thus a consistent conceptualization of phylogenies.

Hennig's confusion of mono- and holophyletic groups/processes, and ‘denial’ of paraphyletic processes is inconsistent and empirically wrong. It is in practice an inconsistent exclusion (denial?) of the distance between reality and our comprehension of reality, and thereby an equalization of space and time that is falsified by empirical evidence in the form of the relativity of time. The relativity of time instead supports our traditional notion that there is a distance between our comprehension of reality and reality itself, thereby also a notion that process and pattern are different and real, and thus also a notion that monophyletic groups and holo- and paraphyletic processes are real. The relativity of time thereby also supports empirical science's basic assumption, that is, that things exist. Hennig's confusion of mono- and holophyletic groups creates a notion that things do not exist, but it is only an illusion arising from its confusion of single things with several things in a row. The purpose of Hennig's confusion was probably to bridge the gap between our comprehension of reality and reality that makes discussions about phylogenies difficult, but, unfortunately, this gap is not bridged consistently and correct by conceptual confusion. A consistent and correct bridge is instead constructed by a consistent understanding of the concepts mono-, holo-, and paraphyletic groups.

The fundamental problem in conceptualization of phylogenies is that the distance between reality and our comprehension of reality means that a consistent conceptualization (distinguishing pattern and process) has to be ambiguous on the specific level (i.e. can distinguish several, equally correct, but incompatible systems of monophyletic groups), whereas an inconsistent conceptualization (confusing pattern and process) instead is self-contradictory on the generic level. It means that conceptualization of phylogenies is inherently ambiguous. However, whereas consistent conceptualization (i.e. specific ambiguity) is empirically correct (i.e. fitting facts), inconsistent conceptualization (i.e. generic self-contradiction) is not. It means that the Linnean conceptualization of phylogenies is consistent and correct, whereas Hennig's conceptualization instead is inconsistent and erroneous. The kind of group Hennig called ‘monophyletic groups’ is actually a confusion of two kinds of groups: holo- and monophyletic groups, whereof the first is a kind of the latter, and thus also a confusion of process and pattern. By contrast, the Linnean conceptualization of phylogenies keeps process and pattern apart, and is thus both consistent and correct.

REFERENCES