Rasnitsyn, 2008. Theoretical Foundations of Evolutionary Biology

1.1. THE PROCESS OF EVOLUTION 1.1.1. SYNTHETIC THEORY OF EVOLUTION

1.1.2. EPIGENETIC THEORY OF EVOLUTION 1.1.2.1. Basic Principles 1.1.2.2. Adaptive Compromise 1.1.2.3. Problems

A.P. Rassinitsyn. Theoretical Foundations of Evolutionary Biology // V.V. Zherikhin, A.G. Ponomarenko, A.P. Rassinitsyn. Introduction to Paleoentomology. M.: KMK. 2008. 371 p. 1.2. METHODOLOGY OF PHYLOGENETICS, TAXONOMY, AND NOMENCLATURE 1.2.1. PHYLOGENETICS 1.2.1.1. Group Analysis 1.2.1.2. Trait Analysis 1.2.1.2.1. Difference Analysis 1.2.1.2.2. Similarity Analysis 1.2.1.3. Computer Cladistics

1.2.2. TAXONOMY 1.2.2.1. Cladistics 1.2.2.2. Phenetics 1.2.2.3. Phylistics

1.2. METHODOLOGY OF PHYLOGENETICS, TAXONOMY, AND NOMENCLATURE What was discussed above — “how things (e.g., evolution) actually happen” — is called ontology in philosophy. Let us now turn to methodology — to how phylogenetic relationships, the system of organisms, and the system of their names should be studied. In the course of ordinary research a scientist pays little attention to the methodological foundations of his work. However, when familiar standards and methods begin to fail, or simply cease to satisfy us, there is a temptation to stage something like a scientific revolution (Kuhn, 1970): to discard familiar methods and standards together with the methodology on which they rest, and to replace them with something else. Revolutions — even scientific ones — rarely pass without cost, in the sense that they discard approaches that are still quite serviceable, albeit in a somewhat different context, and results that are significant but require reinterpretation. It is therefore preferable to re-examine the methodological foundations of scientific activity in the hope of identifying and correcting weaknesses, thus preserving both the approach itself and the bulk of the results obtained through it. The current revolution in the theory of systematics (taxonomy) and the closely related field of phylogenetics has been under way for half a century and is generated, in my view, by the drive toward formalisation, automation, and standardisation of the scientific process — toward enhancing its objectivity and the reproducibility of its results. And correspondingly toward reducing the role of the individual researcher, especially his or her intuition. This revolution is linked to the computerisation of our lives (not only of science), and it began, naturally, in the West. Computerisation demands formalised, mathematised methods, free from intuition or at least appearing so. In systematics, phenetics was the first favourite, then cladistics came to dominate minds entirely, helped by the long-standing works of Willi Hennig finally being discovered by Americans and translated into English. Cladistics largely still corresponds to the Western, and above all American, worldview, although questions and scepticism are already arising even there. But even here cladism is beginning to grow from a fashion into a mode of thought, and yet a comparative analysis of cladistic and non-cladistic methodologies encounters less resistance here. Before proceeding directly to phylogenetics, taxonomy, and nomenclature, I should like to discuss some broader methodological problems and, in particular, the methodology of cognition itself. Let us begin with the goals of cognition. Two main approaches have long been distinguished — the deductive and the inductive — to which a third was added comparatively recently (less than a century ago): the hypothetico-deductive method, associated with the name of Karl Popper. Speaking very roughly, in the deductive approach a scientist starts from a system of postulates that he accepts as true and derives consequences from them by strict logical inference, which are thereby accepted as equally true. In the inductive approach a scientist draws conclusions not from postulates but from observations (“facts”), and being confident of the truth of those facts he likewise considers his conclusions to be true, i.e., to describe the actual state of affairs (what is “really” the case). This distinction does not, however, seem to me fundamental, since postulates are not plucked from thin air but are formulated — even if from general considerations — in such a way that the results of their application are compatible with experience and with reality. That is, deduction is also oriented toward observation and consequently rests to some degree on induction. On the other hand, in induction the results of observations are generalised as conclusions about the nature of things, which in further reasoning are then used as postulates. It is precisely this similarity that makes it possible to employ deductive and inductive methods as a unified approach whose goal is the apprehension of truth. In contrast, for the hypothetico-deductive method all conclusions drawn indifferently from observations, from postulates, or from any other generalisations remain hypotheses — that is, inferences requiring obligatory testing. A hypothesis cannot in principle be verified, i.e., proved true once and for all. As we shall see, it cannot be definitively falsified either; one can only reach a decision about its adequacy or inadequacy with respect to available data, i.e., about the possibility or inadvisability of relying on it in further cognitive or other activity. Thus, the hypothetico-deductive method, now recognised as best corresponding to the real process of cognition, sets as its goal not the apprehension of truth but the approach to it by means of effective cognitive methods. Each step of cognition comprises, in my view, six sequential stages (Table 1). Drawing on prior experience and knowledge, we set a problem and plan the collection of material. We then make observations and/or conduct an experiment and analyse the results, attempting to find familiar elements in an unfamiliar (not yet understood) visible pattern. Drawing on general experience and the results of previous acquaintance with similar objects, we identify among the familiar elements those more important for the purposes of our work — that is, those that may help us reveal important features of the still-incomprehensible pattern. In turn, we consider important those features that we suppose to exert strong influence on, or otherwise be closely correlated with, the structure of our pattern, and thereby allow us to understand it and predict its behaviour under different conditions and circumstances. For instance, when a systematist discovers a completely new group of insects and begins to classify them (to construct the internal system of the group), knowing of the existence of and possible differences between sexes and ontogenetic stages, he will exclude characters that distinguish males from females, adults from larvae, etc., from among the taxonomically important characters, however profound these differences may be in themselves. In a similar manner we usually prefer structural characters over colour characters. Table 1. Steps of cognition

1 problem statement and data collection planning

2 observation and experiment

3 searching for analogies: familiar elements in a new picture for us and identifying the more important ones – those that can reflect the deep structure of the system

4 creating hypotheses about the regularities and mechanisms underlying the observed picture

5 attempting to falsify these hypotheses: deriving consequences from them and comparing them with the results of observations and experiments, both previous and new, conducted specifically to test the hypotheses

6 evaluating the result using presumptions: if falsifying evidence is found, it is necessary to decide whether it is sufficiently reliable and convincing to abandon the hypothesis