ISHPSSB 2001 || Quinnipiac University, July 18-22, 2001

Function, Teleology & Explanation

What counts as a biological explanation of function, especially with regard to processes that exhibit an element of projection, or so-called teleology? Two major themes have emerged, analytical- mechanical and etiological- selectionist. This session examines these traditions and their application, offering fresh perspectives and alternatives.

Organized by: Andrew Aavatsmark

Scott Thomson, Virginia Tech
"Teleology in Biological Explanation: Mayr's Teleonomy and Wright's Etiology"
Explanation in biology is made interesting by the need to encompass those biological phenomena which have traditionally been referred to as teleological. One important issue is how to speak of these things without using 'teleological language.' This has commonly been done by substituting words like 'goal-directedness,' 'function,' and 'teleonomy.' Two important and influential contributions in this field are Wright's 'consequence etiology,' and Mayr's division of teleology into teleonomy, teleomatic phenomena, adapted systems, and cosmic teleology. Although Wright's ideas have been more popular, especially in analyses of explanation, I argue that Mayr's philosophy ultimately offers a framework for a broader range of biological explanation by allowing for a 'second level' of causal explanation.

Andrew Aavatsmark, University of Connecticut
"Functional-Causal Explanation"
Discussions of explanation in biology tend to focus more on how function attribution is explanatory than on explanations where functionally individuated causes are cited. Clearly biologists do explain various phenomena as effects of items with functions--circulation is explained by the heart, for example. This is a functional-causal explanation. But, on a "selected effect" account, where functions depend on history, not structure or dispositions, functionally different items can be causally identical, and functionally identical items can be causally different. Thus functions are causally irrelevant. This paper develops an analysis of the explanatory relevance of functions. Functional kinds are explanatory because we can and do project generalizations about them--namely that a member of such a kind is likely to be biologically normal, and capable of producing its normal effects, including its proper functions. Certain malfunctions and accidental effects can also be explained on the same grounds. The projectibility of such generalizations is not infallible, and depends not on any intrinsic constitution of members of a functional kind, but on their common ancestry, and the action of similar external forces at work on them. This sort of generalization is typically only possible over members of the same species or genus.

Peter Schwartz, Harvard Medical School
"Functions in Molecular Biology"
Some of the most influential theories of function in biology claim that there are actually two concepts of function at work (c.f. Millikan 1989; Neander 1991; Godfrey-Smith 1994). According to these theories, which I will group together as the "bifurcation account," a trait-type X has the proper function F only if X has been favored by natural selection for doing F. But X can have the causal role function F even if X was never favored by selection in this way. All that is required is that X's doing F plays a role in an "analytical explanation" (Cummins 1975) of how the trait combines with others to produce the organism's complex capacities. Proper functions play a role in categorizing traits (e.g. as a "heart") and in determining which states of the organism count as dysfunctioning, while causal role functions do not carry these implications.
Some philosophers discussing the bifurcation account have suggestion that areas such as molecular biology or genetics only employ the concept of causal role function, leaving proper function to fields such as evolutionary theory and animal behavior (c.f. Kitcher 1993, Amundson and Lauder 1994, and Buller 1998). This assumption has initial plausibility. Research in molecular biology, for instance, uncovers exactly the sort of complex analytical explanations in which causal role functions appear, and such research rarely makes explicit claims about natural selection, or does so only in passing.
But this assignment of different concepts of function to different subfields of biology was not suggested by the proponents of the bifurcation account and has not been tested by examining the relevant science. Philosophers of biology are often guilty of making blanket assertions about the science without specifically examining the literature. In the function debate in particular, philosophers have rarely turned to actual biological studies to test their theories. And once we do turn our attention to the actual findings of molecular biology, I argue in this paper, we see that both concepts of function have important roles to play. I specifically examine a set of papers investigating the functions of dopamine receptors in the brain. Many of the references to "function" in these papers appear to refer to causal role functions, but others better fit the characterization of proper function. I use these examples to explore the interlocking theoretical roles of these two types of function ascriptions in molecular biology. Advances in the function debate will require such careful application and evaluation of conflicting theories to actual scientific research. In addition, identifying the role of proper function in molecular biology is an essential part of developing an explication of the content and structure of this field.


Schedule || Program || Conference Main Page