Biological Approaches to Mental Representation 563representations refer toCs if mechanisms were adapted for being caused byCsto
produceRs. We get the desired result if we instead suggest thatR-representations
refer toCs if mechanisms were adapted for being caused byCsquaCs (i.e., in
virtue of their beingCs) to produceRs. That is, the frog’sRs refer to small, dark,
moving things, on this proposal, if the frog’s perceptual mechanism was selected
for being causally sensitive to the property of something’s being small, dark and
moving.
(5) Swampman. In his original manifestation, Swampman is a doppelganger
(molecule for molecule duplicate) of Donald Davidson at a time ([Davidson, 1987],
though earlier versions of Swampcreatures exist, e.g. [Boorse, 1976]). Where
Swampmman differs from Davidson is in his history. Swampman comes about
by sheer accident as a result of a random collision or coalescing of sub atomic
particles and his resemblance to Davidson is sheer coincidence. According to
a teleosemantic theory, Swampman has no intentional content, since he has no
history of the appropriate sort: no evolutionary history of selection whatsoever,
ontogenetic or phylogenetic. The problem is that it is intuitive to think that
Swampman could think about things. Or it is at least intuitive to think that it
will seem to Swampman that he does. Does he not see the world about him? Does
he not think about it? Indeed, how could it even seem to him that he has such
mental states unless he at least hasthatintentional state?
One issue is how seriously we should take this type of imaginative counter-
example. Arguably, the objection fails for versions of teleosemantics that are
offered as attempts to describe the a posteriori necessary conditions for intentional
kinds, which are psychological kinds that are individuated by their contents (e.g.,
perceptions of red, and a child’s concept of an object). Teleosemantics treats
intentional kinds as analogous with biological kinds. Just as it is not a strong
argument against an historical definition of a species concept that it precludes
Swampman from being a member of Homo sapiens, it does not seem to be a
strong argument against an historical theory of mental content that it precludes
Swampman from having content, if these are similar kinds of kinds (see [Dretske,
1996]; [Neander, 1996]).
Notice that this is not the same as claiming that we need only concern ourselves
with actual intentional systems. That claim is unsatisfying if we want to know
what the metaphysical conditions for mental content are. A posteriori necessary
conditions hold in all possible worlds. On this view, “content’ (in “intentional
content”) is similar to “water” and “tiger”, insofar as all three refer to kinds that
could have hidden or unknown essences. On this view, were Swampman to exist,
he would not have intentional content, despite his having something superficially
similar to it. Just so, on this view, God could not make a tiger de novo, if tigers
have an essence of origin, and twin-water would not be water, if water is essentially
and necessarily H20.^8
(^8) I am referring to Putnam’s imaginative example, which involves a world (Twin-Earth) just
like ours except that in the place of water there is a superficially indistinguishable liquid, twin-
water, which is not H20. Twin-water is XYZ. Putnam’s intuition, following Saul Kripke, is that