[ Pobierz całość w formacie PDF ]

biological devices, namely, these motor routines, to perform
their proper function. That passing flies historically caused
this system to issue the famous neural signals tells us not so
much what that system s proper function is, as how it has
served its proper function. And so in general for selected
devices that signal in response to features of the environ-
ment. Their proper function is switching on and steering
 consumer devices, often in the same organism. But
devices that can do this job flexibly enough for evolution to
like them, and simply enough for evolution to be able to
devise them, are ones that do it in a special way by issuing
signals that correspond, at least often enough, to values of
the relevant outside features.
Now, can these reflections lay the disjunction problem to
rest? It all depends, some philosophers would hold, on what
the proper function is of the motor routines themselves that
launch the frog s gulps. Historically these routines caused
host frogs to ingest a huge group of small dark moving
objects, many but not all of which were flies or other nutri-
tious (to frogs) insects. Then is the proper function of these
Why Austerity in Ontology Does Not Work 175
motor routines to effect ingestion of small dark moving
objects? In that case, the system that produces the famous
neural impulse must have enabled these routines to serve
their proper functions by matching its outputs to small dark
moving objects be they flies or BBs or bits of leaf. The
content of the neural impulse remains disjunctive.
The crucial question here, I have argued (Elder 1998b), is
what causally explains the proliferation among Rana pipiens
ancestors of just these gulping routines what causally
explains the fact that genes coding for these routines won
out over their alleles. Even if we are ignorant of some
details, we can be sure that in this causal explanation, the
darkness of the bits ingested plays no part. Neither does their
movingness (-prior-to-ingestion). What figures in the causal
explanation, we can be sure, is just the nutritiousness (to
frogs) of the bits ingested. So the proper function of the
motor routines is effecting ingestion of nutritious bits. And
the famous neural signal enabled the routines to do this just
to the extent that it corresponded to the presence of nutri-
tious bits. Therefore the content of that signal is  nutritious
bit here now.
The disjunction problem can then really be solved, in the
case of its most famous victim. Viewing the devices that
produce and respond to the famous neural signal as products
of natural selection permits us to see that what that signal is
supposed to correspond to its content is not just every-
thing in the environment that does in fact trigger its pro-
duction. It permits us to understand how the frog can make
wrong identifications of the small bits that cross its visual
field.
Can these insights be extended? Can we devise an exter-
nalist picture of the content of our thoughts, thus preserving
clarity on how we can be fallible about the nature of the
176 Chapter 8
kinds and stuffs we encounter, while invoking selectional
history to explain our fallibility in the judgements we make
as to when particular kinds and stuffs are present before us?
The answer is Yes, provided we are careful to admit just how
far the extension must reach. Ruth Millikan has extended the
teleosemantic account to cover the content of our thoughts
(Millikan 1984, 2000). The extension does have to reach far
too far for me to trace it in a chapter (or even in several chap-
ters). The reason why the extension must reach far is that
Rana pipiens s famous neural signal is designed to steer only
a single behavior, to steer it always, and to steer it immedi-
ately. It is as much a command  gulp thataway!  as an
indicative report  nutritious bit over there! Our thoughts
about our surroundings can be exclusively indicative reports.
They can be stored for later use, and when used can be
enlisted to attune actions that serve any of a range of desires.
But the ways our indicative thoughts guide and shape our
behaviors, as effectively as evolution has required them to
do or better, as effectively as evolution has required of the
cognitive and sensory programs that produce them has
much in common with the ways the neural signaler in Rana
pipiens won the favor of natural selection. Our indicative
thoughts have shaped behaviors effectively, as often as they
have been required to, by corresponding under definite
mapping rules to a variety of ambient circumstances. They
have not always shaped behaviors effectively, and do not do
so now: the outside circumstances which the rules require
the thoughts to map may be absent; we are fallible, both in
our thoughts about what nature s kinds are like, and in our
thoughts about when they are present before us. But natural
selection has never insisted on perfection.
Why Austerity in Ontology Does Not Work 177
8.3 Collective Cognitive Tracking
I now have sketched the reasons for thinking that if minds
are placed on the ground floor in a projective ontological
scheme if we allow, as austere ontologists must, that the
world really appears to minds to contain many familiar
medium-sized objects, but say that the medium-sized
objects are all mere appearances, projected by the minds onto
a world that is far more austere then those minds must
drag with them enough of biological reality to populate a
whole history of natural selection. They must likewise, and
for the same reason, drag with them the causation by which
they are naturally selected. I now will argue that if the world
is objectively to appear as rich as scientifically informed
common sense supposes it to be, the minds on the ground
floor must be minds in a community. They must drag with
them enough physical reality to secure their mutual dis-
tinctness and enable them to communicate linguistically
with one another, and enough social reality to undergird the
existence of a shared language. [ Pobierz całość w formacie PDF ]

  • zanotowane.pl
  • doc.pisz.pl
  • pdf.pisz.pl
  • dirtyboys.xlx.pl