Changing the Mission of Theories of Teleology 31
explaining certain puzzling problems, and they need not be seen as competing mutually
exclusive options but rather as alternative parts of a larger picture. (Almost) nobody writes
papers defi ning when “S knows that p” anymore. Even papers declaring the nature of jus-
tifi cation have waned—in favor of examinations of the various aspects to justifi cation, be
they biological, ecological, psychological, social, political, and so forth. No one factor is
the magic single answer about justifi cation—they each have interesting things to tell us,
and which factor is the deciding one depends on the specifi c circumstances of a particular
instance.
This is the kind of result I envision, and would recommend, for modern theories of
teleological functions. To some extent we see these same kind of battle lines now being
drawn between present-looking systematic (Cummins-style) functions, backward-looking
views, both distant-past historical/evolutionary functions (Millikan 1984, 1986, 1989a,
1989b, 1989c, 1990, 1993, 2002; Neander 1991a, 1991b; Papineau 1987; Griffi ths 1993;
Allen and Bekoff 1995) and recent-past/goal contribution/design views (Boorse 1976,
1977, 2002; Godfrey-Smith 1994; Kitcher 1993), and forward-looking propensity func-
tions (Bigelow and Pargetter 1987). What I would recommend is a truce, and a realization
that all of these notions have their virtues, and each has the potential to explain things we
would like to explain. Let us stop the functions debate from imitating the kinds of chaos
and splintering that epistemologists have wisely moved away from.
2.5 DO Focus on Explanation: Pragmatic Teleo-Pluralism
I have now described four DON’Ts—things I think we should avoid in our theories of
teleofunctions. The upshot of the last one is that we should stop pursuing conceptual
analysis of some single unifi ed concept of “function” that will cover every case, defuse
every counterexample, and explain everything. There is no one intuitive notion of “func-
tion” (or “teleofunction”) for which we will fi nd the magic set of necessary and suffi cient
conditions.^8 There are many different kinds of questions to be answered, and different
phenomena to be explained, and various accounts of function can explain some and not
others (or perhaps different aspects of the same phenomenon). The urge to fi nd one account
that explains everything is bound to lead to complex, convoluted, gerrymandered theories
that may get some things right, but rarely sound like basic principles of nature.^9 So what
should we do? We should acknowledge the different advantages of the various views of
function, and use them all as they help explain things. For lack of a better term, I call this
attitude Pragmatic Teleo-Pluralism.
I am not alone in urging this cessation of hostilities. The maven of teleology, Millikan,
herself tried to avoid it in LTOBC by explicitly stating that her notion of “proper function”
was a technical term, not an analysis of what we all mean by “function.” More recently
(1989c, 2002) she recognized what she termed an “ambiguity” in the notion of function,