558 Karen Neander
to the subsumption strategy. We can easily imagine biologists express-
ing their analyses in a form analogous to the schematic diagrams of
electrical engineering, with special symbols for pumps, pipes, filters,
and so on. [Cummins, 1975, 760–1]This is all true but it omits the peculiar kind of idealization that is used in
biology in providing such functional analyses. When biologists provide such ex-
planations, they make sure to distinguish between normal and abnormal capacities,
or in other words they make sure to distinguish between what different parts of
a system contribute to the overall operation of the system as a whole when they
are functioning normally or properly and what they contribute when they are not
functioning normally or properly.
This point is not new. It was familiar to philosophers of biology before Cummins
wrote his piece, but it has often gotten lost since Cummins wrote his piece, and it
is important to the present discussion. To be clear, on its own it does not establish
the need for an SE-analysis. However, it is worth noting that SE-functions mesh
well with this idealization task. Organic systems are tremendously complicated
organized systems and they could not exist except for the fact that they are the
products of selection. Furthermore, they mostly only operate well — in terms
of survival and reproduction, and in terms of the vast array of capacities that
contribute to them — to the extent to which each of their intricately co-adapted
parts is performing the causal role for which it was selected. That’s because of how
selection works: basically (I simplify, but not in ways that need concern us here) it
tends to keep the parts that contribute to survival and reproduction and it tends
to eliminate the parts that hinder them. A further consideration is that there
are many, often billions of instances, of each type of system that the physiologist
seeks to describe: consider all the humans that have ever lived (96,100,000,000,
according to one estimate that I have read) let alone all the ants or fruit flies.
At least some of the immense variation among individuals can be set apart in
a theoretically significant fashion, when describing a type of system (a “species
design”, so to speak) by distinguishing between normal and abnormal variation.
Neurophysiology is no different. It also aims to provide an idealized functional
analysis of a tremendously complicated organized system. Only in this case, as
we currently understand it, it is a cognitive system designed for processing and
using information. Therefore, a normative notion of a function, of the kind used in
ordinary somatic physiology, is already deeply implicated in cognitive science. It
is already intimately involved in scientific talk of information-processing functions.
On the assumption that the normative notion of a function is the SE-notion, the
general teleosemantics enterprise is a parsimonious one. It makes good sense to
see how far we can go on the basis of the notions already in play.