<<

. 16
( 62 .)



>>

Wilfrid Sellars™ classic attack on the Myth of the Given, that links between
Individualism and Externalism in Cognitive Sciences
90

the normativity of the mental and externalism about the mind have
been forged most thoroughly, particularly in the work of John McDowell,
Robert Brandom, and John Haugeland. What these views share, and what
allows them to carve a path from normativity to externalism, is the idea
that in order to have at least certain kinds of meaningful, content-laden,
intentional states one must be subject to rules, standards, and conven-
tions, where these are not metaphysically determined by intrinsic facts
about the bearer of those states. Recognizing the normative dimension
to the mental takes one beyond the individual and into the social.14


5 narrow and wide content
Thus far, I have barely paused to register the various twists and turns
to the debate over individualism in psychology. But there is one twist
that can™t be missed: the distinction between narrow and wide content.
One intuitive response to the initial Putnam-Burge arguments against
individualism has been to concede that while there is a sense in which
even doppelg¨ ngers can have mental states with different content, there is
a
an equally important sense in which they must have mental states with the
same content. That content, content shared by doppelg¨ ngers no matter
a
how different their environments, is narrow content.
We can use Putnam™s own example to illustrate what narrow content is,
and why we might insist on its existence and importance. While we might
distinguish between the meaning of Oscar™s term “water” and Oscar*™s
term “water” (which we designate with “twater” in part to highlight its
distinct extension), we could equally ascribe a common, shared mean-
ing to their terms “water,” one neutral between H2 O and XYZ. Precisely
because there is so little difference between Oscar™s term “water” and
Oscar*™s term “water,” it is plausible to view those terms, even if not
strictly identical, as sharing so extensive a common core of meaning that
we would be overlooking something important were we simply to treat
them as semantically independent terms. Putting this in terms of the psy-
chological states that Oscar and Oscar* are in, we can say that although
there is some difference between the psychological ascriptions we would
make for each “ those that employ a notion of wide content “ there is also
much intentional psychology that is shared between Oscar and Oscar*.
In fact, there is much shared in their intentional psychology even when
we are considering their thoughts about “water.” In order to express these
common, intentional psychological states, we need some notion of con-
tent that Oscar and Oscar* share. That is, we need some notion of narrow
Individualism: Philosophical Foundations 91

content. Whatever else it is, narrow content is the type of content that
physical twins must share, however different their environments.
In introducing narrow content in this way, I have presupposed that the
Twin Earth thought experiments have shown that Oscar and Oscar*™s psy-
chological states differ in their propositional content. More generally, I
am assuming that our ordinary, pretheoretical notion of propositional
content, of intentionality, is not narrow but wide. This is certainly the
received view both of what the Twin Earth arguments show and of inten-
tionality, but it will pay to spell out just why.
The intentional content of so-called folk psychological states, such as
belief and desire, is what is speci¬ed by the that-clause of an ascription
of propositional attitudes. For example, when we say that
Peter thinks that wolves are placental

the content of Peter™s thought is that wolves are placental. This is what his
thought is about, what it represents as being the case. Peter™s thought
is about wolves, that is, he has some internal representation that refers
to wolves. At least when we are considering folk psychology, there is no
intrinsic feature of the representation itself that makes the corresponding
thought one about wolves. In fact, note that we can understand what
the content of Peter™s thought is in this example although I have said
nothing at all about the nature of the representation in and of itself.
Even supposing that Peter™s thought is instantiated in Peter in virtue of
there being a token of the sentence “wolves are placental” inscribed in
Mentalese in a place in Peter™s brain that we can call his “thought box,”
this fact about Peter™s internal organization is not suf¬cient for Peter™s
thought to be about wolves being placental. Someone else with just that
inscription in her thought box could have a distinct thought were facts
outside of Peter other than they are. In particular, this could be so were
“mere” social facts, facts about our communal conventions for the use
of natural language terms, different. The claim that this is true not just
of particular in-the-head facts but of all of them considered together “
they do not suf¬ce to ¬x the contents of one™s thoughts “ is another way
to express the conclusions drawn from the Putnam and Burge thought
experiments.
The intentionality of folk psychological states seems intuitively to lie in
some sort of relation between what is inside the head and what is outside of
it. It is plausible to think that this relationship is, broadly construed, causal
in nature. It is wolves that Peter thinks about because wolves are what
causally impinge on Peter either through his own direct sense experience
Individualism and Externalism in Cognitive Sciences
92

of wolves or through his causal location in a linguistic community. This is
not to imply that being in such a causal relationship is suf¬cient for one to
have thoughts about wolves, but that these causal relationships are partial
metaphysical determiners of the intentionality of one™s thoughts. And
this shows that folk psychological content is wide, since individuals who
differ in this beyond-the-head relation may have thoughts with different
contents, no matter how similar they are in in-the-head respects.
Consider now the claim that there is, in addition, a notion of narrow
content to be found lurking in folk psychology. The intuition that the
above appeal to the twins Oscar and Oscar* drew on was that their inter-
nal identity does generate some sort of intentional identity or similarity,
even if there is also a respect in which they have thoughts with different
contents. One way to express this is to say that their “water” thoughts
share a narrow content, even though they differ in their wide content.
There are three motivations for the notion of narrow content, none of
which presupposes individualism about the mind, and each of which has
served as the basis for a distinct proposal about what narrow content is.
The ¬rst concedes that individual representations do not suf¬ce to ¬x
content but challenges the generalization of this point to cover all indi-
vidual representations considered together. That is, once one considers
not just the intrinsic features of a given mental token but its relations to
other mental tokens, the claim that what™s inside the head doesn™t ¬x any
notion of content loses its plausibility. This is because there can be, and
typically is, a rich causal structure to mental representations that begins
and ends in the head, a point that can be drawn from the functional-
ist view of mental states. For example, Oscar and Oscar* both have a
mental token, “water,” that is causally related to the same perceptual and
mental inputs, and the same mental and behavioral outputs, and this is
suf¬cient for those tokens to share some type of content. The proposal
that the narrow content of a mental state is its narrow conceptual role
is a development of this idea, an idea also developed independently as
procedural semantics in early work in arti¬cial intelligence.
The second motivation reaches back to the origins of the anti-
individualistic perspective on mental content in the philosophy of lan-
guage to recognize two aspects to the meaning of natural language terms
and the close relationship between them. Consider the term “I.” The
referent of a spoken token of the word “I” is the speaker, and that is
who is meant by that utterance; the referent of “you” is the person being
spoken to. But “I” also has a common meaning when different people
utter it, as does “you”: “I” refers to the speaker, and “you” refers to the
Individualism: Philosophical Foundations 93

hearer. Thus, when two people say to each other “I love you,” there is
clearly some sense in which they have said exactly the same thing to one
another, and we need some notion of content that captures this. What
these utterances and their constituents share is their narrow content, this
being a sort of rule “ such as that “I” refers to the speaker/writer “ that is
then contextualized to derive the referent of a particular token utterance
or inscription. This has generated the proposal that narrow content is a
function from contexts to truth conditions: Take the narrow content of a
mental state, add a beyond-the-head context, and one arrives at its (wide)
propositional content.
The third motivation is the idea that the beyond-the-head differences
between Oscar and Oscar* make absolutely no difference to how the
two see the world: Their worlds are phenomenologically identical. But
how the world appears to one is mediated not just by any old internal
machinery but by content-laden internal machinery, such as concepts
and ideas, thoughts and beliefs. If physical twins are phenomenologi-
cally identical, then they must be intentionally identical in some sense,
and the corresponding notion of content must be narrow, not wide.
This has given rise to the idea that narrow content is phenomenological
content.15
These various proposals provide a strategy for limiting the signi¬cance
of the Putnam-Burge arguments against individualism since they sug-
gest that there is no strict incompatibility between an individualistic and
an intentional psychology. Thus, one can concede the conclusions that
Putnam and Burge draw but rely on the presence of some notion of nar-
row content either to continue with content-laden psychology as it has
developed to date, or to look for alternative ways to connect computa-
tional and neuroscienti¬c approaches to the mind with our existing folk
psychology.


6 functionalism, physicalism, and individualism
For many philosophers of mind, individualism has been attractive be-
cause of a perceived connection between that view and physicalism and
functionalism in the philosophy of mind, both of which have been widely
accepted over the last thirty years. Physicalism or materialism has been
expressed in various ways, perhaps most commonly in terms of the no-
tion of supervenience that we have already met: All facts, properties,
processes, events, and things supervene on the physical facts, properties,
processes, events, and things, as they are posited in elementary physics.
Individualism and Externalism in Cognitive Sciences
94

This ontological formulation of physicalism is often accompanied by an
explanatory thesis, which states that physical explanations are, in some
sense, the ultimate explanations for any phenomenon whatsoever.
Individualism has been thought to be linked to physicalism since it im-
plies, via the supervenience formulation, that there is no psychological
difference without a corresponding difference in the intrinsic, physical
states of the individual. Those rejecting individualism have sometimes
been charged with endorsing a form of dualism about the mind, or mak-
ing a mystery of mental causation by ignoring the causal powers that
mental states have. Causation between mental states and events is ulti-
mately causation between an individual™s physical states and events, and
scienti¬c taxonomies of mental states must respect this feature of men-
tal causation. Connecting this up with the methodological formulations
that have had in¬‚uence in cognitive science itself, individualism has been
claimed to be a minimal constraint on arriving at psychological explana-
tions that locate the mind suitably in the physical world, a psychology that
taxonomizes its entities by their causal powers.
Individualists themselves disagree about what this implies about the
substantive nature of psychology. For example, Jerry Fodor thinks that
this minimal constraint provides a way of seeing how folk psychology in
particular and the notion of mental content more generally, have a re-
spectable place in the cognitive sciences. At one point Fodor held that this
in turn required cognitive science to use a notion of narrow content, al-
though in subsequent work he has suggested that ordinary propositional
content was adequate for the tasks of theory construction in the cogni-
tive sciences. Stephen Stich, by contrast, thinks that individualism implies
that cognitive science should jettison both folk psychology and the notion
of mental content altogether, arguing that content should be eliminated
from psychology, in part because of the con¬‚ict between the minimal
constraint of individualism and the wide nature of intentionality. Both
of Fodor™s positions suggest that individualistic approaches to psychology
will be recognizable descendants of current work in the cognitive sciences
that incorporates and builds on folk psychology. Stich™s view, by contrast,
implies that an individualistic cognitive science will be divorced from
folk-contaminated research traditions, being fashioned instead from the
cloth of computational intelligence or cognitive neuroscience.16
Functionalism is the view that psychological states and processes
should be individuated by their causal or functional roles, that is, by their
place within the overall causal economy of the organism. It has been com-
mon to suppose that these functional or causal roles are individualistic.
Individualism: Philosophical Foundations 95

Functionalism was introduced in the 1960s as a way of understanding the
relationship between mental and physical states that meshed with two
perspectives on minds then nascent but that have had a lasting effect on
how philosophers think about the mind.
The ¬rst was the rise of the computer metaphor and particularly the
analogy between the distinctions between software or program and hard-
ware, on the one hand, and mind and brain, on the other. Mental states
were not strictly identical to brain states but involved some sort of abstrac-
tion from those states, in much the way that computer programs were not
strictly identical to electronic states of computers but abstractions from
them. Functionalism cohered with the computer metaphor because it
offered a causal understanding of the mind that also seemed to charac-
terize the less mysterious relationship between programs and computers.
Functionalism also provided the conceptual underpinnings for the very
idea of arti¬cial intelligence.
The second was the related desire for an account of the mind that
made it perspicuous how creatures very much unlike us in many respects
could nonetheless share a mental life very much like our own. This desire
was typically expressed as the requirement that an account of the mind
must allow mental states to be multiply realized, that is, instantiated even
when the underlying physical realizers vary. The most attention-grabbing,
putative cases of multiple realization were those where the physical real-
izers for mental states were extremely different “ humans versus possible
silicon-based Martians. But there was also a concern for biological differ-
ences across species, as well as physical differences in a given individual
over time.
Functionalism seemed perfect for an account of the mind that meshed
with these two perspectives, since what mattered for functional identity
was not the nature of the physical stuff but, rather, the way in which that
stuff was structured or organized. Again, the computer metaphor was apt
here: The very same program could be instantiated on physically quite dif-
ferent machines. Crucial to program identity is functional organization;
likewise for minds.
Functionalism has been understood in various ways, but the two ways
most prevalent in cognitive science “ in terms of the notion of compu-
tation, and in terms of the idea of analytical decomposition “ both lend
themselves to an individualistic reading. Computational processes, con-
ceived as operating solely on the syntactic properties of mental states,
have been plausibly thought to be individualistic. And it is natural to view
analytical decomposition as beginning with a psychological capacity, such
Individualism and Externalism in Cognitive Sciences
96

as memory or depth perception, and seeking the intrinsic properties of
the organism that create and constitute that capacity.


7 the appeal to causal powers
Appeals to an entity™s causal powers are prominent in discussions of the
relationship between individualism and physicalism, and they are at the
core of an in¬‚uential and persistent argument for individualism ¬rst of-
fered by Jerry Fodor. Although criticisms of the argument seem to me

<<

. 16
( 62 .)



>>