Does Oxygen Have a Function, or Where Should the Regress of Functional Ascriptions Stop in Biology?

Abstract

Biologists apply the notion of function to almost every type of structure and process that enters into descriptions of biological phenomena. They can also generate alarmingly long regresses: x can be the function of y, which is the function of z … and so on. But the functional regress must stop somewhere. This chapter investigates whether the philosophical theories restrict the regress of functional attributions by asking if they legitimate making such attributions to structures at elementary levels of organization (atoms and elementary molecules) and to structures at higher ones (organisms and species). First, I propose a classification of the current theories of functions into three categories, rather than the usual two: Larry Wright’s “etilogical theory” is definetly something different from the “selective etiological theories” that have been developed after him. Then I examine whether these theories can admit or not the ascription of functions to very low or very hig levels of organization. At the most elementary levels, functional ascriptions are unacceptable for the selective etiological theory of functions, because atoms or elementary molecules are not units of selection; they are less problematic for the systemic theory of functions, and also for Wright’s original “etiological theory,” provided that the composition and behavior of the parts constituting the system involved are precisely stated. At the level of organisms and species, functional ascriptions are possible within both the selective etiological and the systemic theory, but this will heavily depend on the theoretical framework involved in both cases. These limit cases show that the selective conceptions of functions are less tolerant than the systemic ones. They also suggest, as already noted by William Wimsatt, that functions are more convincingly ascribed to processes than to structural entities.