Changing the Mission of Theories of Teleology 29
social, economic, physiological, physical, and perhaps also genetic factors that have little
or nothing to do with intentions.^3
We see a similar kind of balancing on a more individual basis in psychology and psy-
choanalysis. The best psychological explanation of certain behaviors may include assign-
ing functions to behaviors (including the design and creation of artifacts) that are far
different from those the subject would describe, and assigning intentions vastly different
from what the subject would say his or her intentions were.
So when we seek to explain behavior and function of artifacts, we should of course give
the designers’ (and manufacturers’ and users’) intentions signifi cant weight. But to require
a theory of artifactual functions to be intentionalist would ignore important methodological
considerations in psychology, sociology, and anthropology. We must value the reports of
the people we’re studying, yet also be willing to acknowledge that the best explanation of
their behavior, and the functions of their artifacts, is sometimes not given by their inten-
tions but by other factors.
2.4 Lessons from Epistemology: DON’T Let Theories of Teleological Function
Spiral Out of Control into Defi nitional Oblivion
It seems to me that the development of the functions debate since Millikan’s LTOBC is
similar to the way theories of knowledge and justifi cation developed in the 1960s and 70s
after the famous “Gettier Problem” became their focal point in 1963. In a short space I
can give only the briefest overview of this theme in the development of epistemology, but
I hope I can show how analytic approaches to knowledge and justifi cation met and over-
came a serious roadblock in their development.
In his Theaetetus, Plato was the one who gave us the standard equation: Knowledge Is
Justifi ed True Belief. This formula stood fairly well for quite a long time, through the
disputes between rationalists and empiricists that raged during the Enlightenment. Then
in 1963 Edmund Gettier wrote a short paper that presented a case in which one allegedly
has a true belief that is justifi ed, yet intuitively does not seem like something we should
consider knowledge.^4 The trick is that part of the belief is true, and another part justifi ed,
and minor logical manipulation and conjunction yields a belief that is both true and justi-
fi ed, but not knowledge.^5 Gettier’s paper caused an uproar, and the game was afoot—fi nd
an additional factor to add to truth and justifi cation to yield knowledge and escape the
Gettier problem.
The dust cleared somewhat to show that what seemed to be needed was a way of des-
ignating justifi cation as “defeasible” or “indefeasible”—having true, indefeasibly justifi ed
beliefs equals having knowledge. Many competing defi nitions of defeasibility were pro-
posed, some bizarrely complex. One solution memorable for its length and extravagance
is Swain’s (1974)^6 :