Eighth Annual Conference
SE Functions in Physiology
SE functions supervene on the selection history of a character or trait. Physiological outcomes depend on the current physical state of organisms (and the environment in which they happen to be situated) and not on their history. So it would seem to follow that SE functions have no role to play in physiological explanations. But I here argue that this popular and commonsensical view neglects the role of functional norms. I also try to elucidate that role.
Homologues as units of evolvability: the role of structure and function
I present an account of homology that integrates phylogenetic and developmental approaches to homology. Homology is the morphological organization of an organism into different homologues, where a homologue is construed here as a unit of morphological evolvability, i.e., as a part of an organism that can exhibit heritable phenotypic variation independently of the organism’s other homologues. This evolvability of independent homologues manifests itself in phylogenetic patterns, as recognized by phylogenetic approaches to homology. A developmental account of homology attempts to explain how an organism’s developmental constitution results in different bodily units that can vary and evolve independently of each other.
My talk pays particular attention to the way in which homology can be said to explain and how homology relates to structure and function. Function in the sense of causal role function or activity function is germane to homology in two ways. First, function features in different individuals can be homologous. For instance, a gene expression pattern, a developmental process, or a behavioural pattern can be a homologue. Second, homology as the presence of distinct units of evolvability is accounted for by certain features of an organism’s development, and such developmental processes include activity and causal role functions. However, bodily units in different individuals can be homologous even if they differ substantially in their form and function. Rather than the internal structural and causal-functional properties of a homologue, of particular importance are developmental and functional dissociations between homologues, i.e., relations that account for why different homologues can vary independently of each other.
The Analogy of Evolution by Selection
Many attempts to produce a theory of evolution by selection general enough to apply to empirical domains other than biology have used a strategy that might be called ‘generalization by abstraction’. First, one generalizes the evolutionary function of a class or entity (e.g. replicators), then one uses the generalization to explore if populations in other empirical domains do evolve by a process of selection. This is supposed make clear the necessary functional relationships that allow an understanding of the abstract structure of evolutionary theory. I will focus on the processes of generalization and abstraction, showing their logical structure and their shortcomings in order to produce a domain-free theory of evolution by selection: both allow for domain-specific (biological) concepts because of their use of biological background theories. These shortcomings can be solved by the use of the analogy process as a heuristic in the construction of such an abstract theoretical structure by inverting the logical processes of generalization and abstraction. The analogy process promises us a principled methodology to formulate a domain-free and operational abstract structure of the evolutionary theory, thus allowing the scientist to empirically check its validity within any domains.
Richard A. Richards
Functional Analysis and Character Transformation
It has been commonplace for biological systematists to distinguish pattern and process. Patterns are the products of evolutionary processes and are historical – in the branching tree-like structure of evolution, and morphological – in the distribution of shared characters among organisms. Traditional Evolutionary Systematists have typically assumed that since these patterns are the products of evolutionary processes, we can use what we know about these processes – natural selection in particular – to reconstruct patterns. This assumption has grounded the use of functional analysis to classify shared characters as homologies (due to a single origin in a common ancestor), and to reconstruct character transformation sequences that can then be used to reconstruct phylogenies. Cladistic (or “phylogenetic”) systematists have challenged this use of functional analysis, arguing instead for a parsimony principle.
Some of the central issues in this debate about functional analysis are revealed most clearly in a symposium at the 1979 Annual Meeting of the American Society of Zoologists, where Evolutionary Systematist Walter Bock presents and defends a method of functional analysis against the criticisms of Cladist Joel Cracraft. First, I lay out Bock’s method, as he presents it – in terms of form, function, faculty and biological role. Second, I explain Cracraft’s criticism, and how it seems to be based on “Wright’s Rule” and the assumption of a taxic model of macroevolution. In this case we can see how an apparent dispute about evidence, inferential strategy and research program is grounded on conflicting empirical and theoretical assumptions.
Karl J. Niklas
The Evolution of Plant Body Plans: Darwin’s Dilemma
How body plans are simultaneously conserved within different plant and animal lineages and yet remain adaptively plastic perplexed Darwin until his death. Today, the developmental mechanisms underlying different animal body plans have been extensively studied and are reasonably well understood from a molecular genomic perspective. In contrast, much less is known about the developmental mechanisms responsible for the evolution and diversification of plant body plans. In this lecture, I will review the major plant body plans and define each in terms of how it achieves its organized growth and development. Using a biophysical perspective, I will also explore why the same body plans reappear in dramatically different plant lineages and habitats. A computer model will be presented that mimics the early evolution of the land plants and helps to explain the “Devonian plant explosion”, which rivals the great Cambrian explosion of animal evolution.
Can a function be ascribed to the immune system?
Can a function be ascribed to the immune system? At first glance, the immune ‘system’ is, as a system, a particularly obvious and convincing example of functional ascription: the immune system is ‘supposed’ to distinguish between self and nonself, that is, between what belongs to the organism, and what is foreign to it. This view has been defended, in particular, by Matthen and Levi (1984). Besides, in an analysis of how the immune system may dysfunction in a case like autoimmune diseases, Matthen and Levi interpret the immune system’s function in intentional terms. In my talk, I’ll try to show that:
- It is not possible to speak of the immune system in intentional terms.
- It is not true that the immune system distinguishes self from nonself.
- No satisfying ascription of a function, in an etiological sense, to the immune system, has been proposed.
- Because it seems difficult to ascribe a function to the immune system in an etiological sense, I suggest that it is more satisfactory to adopt a systemic conception (following Cummins 1975) of functions when the immune system is dealt with.
- It is possible to account for the immune system’s dysfunctions (e.g. autoimmune diseases) thanks to the systemic conception of functions.
I’ll try to show what the precise content of such a systemic conception of the immune system’s function would be.
Cummins, Robert (1975), Functional Analysis, The Journal of Philosophy 72, 741-764.
Matthen, Mohan et Levi, Edwin (1984), Teleology, Error and the Human Immune System, The Journal of Philosophy 81(7), 351-372.
V. Louise Roth
Iteration, Reiteration, Information, Homology
The notion of function in the psychiatric debate
In The normal and the pathological (1943), Georges Canguilhem criticized at length the theoretical assumptions underlying what he called “Broussais’ principle” — in reference to Joseph Victor Broussais, a French physician who was one of the first to claim that one could give a purely objectivist account of the pathological phenomena. According to Canguilhem, Broussais didn’t manage to fulfill the two following requirements for any objectivist account:
1) First, it has to rely on a clear definition of the normal state as a matter of fact.
2) Secondly, it has to give a clear construal of how to distinguish between the normal and the pathological states.
In the psychiatric debate, Christopher Boorse and Jerome Wakefield can be characterized as the two best proponents of “Broussais’ principle” nowadays. Aiming at defending the legitimacy of psychiatry as a genuine branch of medicine, they have both put forward a renewed objectivist account of mental disorder. In my talk, I will assume that their accounts convincingly highlighted that the concept of natural function is the key factual component for defining the normal state in a purely objective way. However, I will raise some doubts concerning the ability for each account to deal, empirically, with the demarcation problem (how to identify a dysfunction, and how to distinguish between normal and abnormal mental functioning). Boorse, in his late version of the Bio-Statistical Theory (BST), claims that the difference between the normal and the pathological state amounts to a difference of degree with respect to the statistical distribution of the efficiency of a certain part-function in a group composed of individuals of a specific age and sex. I will argue that, although this account is appealing from a theoretical point of view, its possible application as a tool for clarifying psychiatric issues appears unrealistic. Wakefield, as to him, put forward what he coined the Harmful Dysfunction Analysis (HDA) of mental disorders. His analysis relies on a “black box essentialist” account of scientific concepts which suggests that our current evolutionary understanding of mental functioning succeeds in clarifying the psychiatric categories. According to Wakefield, most of the diagnostic categories, even though they are hypothetical, can be strengthened by the circumstantial evidence that they are failures of naturally selected functions. I will demonstrate that, in the absence of any substantial scientific evidence, this proposal unfortunately turns out to be a merely rhetoric defense of psychiatric labels.
Pim (W.M.C.) Edelaar
Homology – an expression of selection or constraints?
Diverse homologies typically clearly represent the influence of natural selection and developmental constraints. They provide a vivid case of the classic form versus function debate. Before we can boil this down to the empirical issue that seems to most interest us, some conceptual issues must be clarified. We then present a study on crossbill beak morphology performed by Edelaar that tries to estimate the relative contribution of natural selection and developmental constraint on relatively recent diversification of crossbills. This provides one data point in the general debate along with a methodology that can be more generally applied.
Homology and Contemporary Study of Animal Behaviour
In this paper, I assess the relevance of the homology concept in contemporary investigations of animal behaviour. I undertake this analysis from the standpoint of behavioural ecology, a perspective that allows probing of both ontogenetic and phylogenetic elements of what is broadly understood as animal behaviours. I will defend the idea that, although the concept of homology has little relevance in ecology-oriented investigations of behaviour, it is still a heuristically important tool. After I illustrate this, I then proceed to infer the main properties of the concept.