. 18
( 62 .)


Metaphysics, Mind, and Science: Two Views 103

realization itself is a context-sensitive notion. More poignantly, the claim
at the core of the standard view of realization “ that realizers are meta-
physically suf¬cient for the properties or states that they realize “ drives
one to this view. This presents those adopting the standard notion of re-
alization with a dilemma: Either give up or soften this claim of suf¬ciency
(but at the expense of a range of further physicalist claims), or admit that
realization, and so the metaphysics of the mental, is ineliminably context
sensitive. Either way, some widely held physicalist views need to be revised
or rejected.
In the next section, I offer a more rounded characterization of the
standard view of realization that brings out more explicitly two theses at
the heart of that view. This will make my chief objection to the standard
view easy to state and set the scene for an exploration of some context-
sensitive alternatives to it.

4 the standard view (i): realizers as
metaphysically suf¬cient
A widespread view amongst physicalists in the philosophy of mind, what-
ever their other differences, is that realizers satisfy what I shall call the
metaphysical suf¬ciency thesis:

Metaphysical Suf¬ciency Thesis: Realizers are metaphysically suf¬cient for the
properties or states they realize.

I want to say something about why this thesis is implicit in standard
conceptions of realization, particularly those used in the philosophy of
One reason is historical. As materialists came to be in¬‚uenced by the
way in which the computer metaphor suggested that mental states were
multiply realized in physical states, rather than strictly identical to those
states, the claim that physical states were metaphysically necessary and
suf¬cient for particular mental states, appropriate when considering an
identity theory, was weakened to one of suf¬ciency only.
A second reason is that many statements of what it means for mental
states to be realized by physical states presuppose or imply this claim. For
example, it is common to think of realization as a relation of determina-
tion (of mental states by physical states), and the suf¬ciency thesis is at
least a necessary condition for such determination. Also, in explaining
the one-many relationship between mental and physical states allowed
by the notion of multiple realization, it is common to point out not only
Individualism and Externalism in Cognitive Sciences

that this is not to be confused with the claim that there is a many-one re-
lationship between mental and physical states, but that such a possibility
would call physicalism itself into question. This possibility, that of emer-
gent realization, that is, of a physical realizer for a given mental property
that could realize some other mental property were the world different
in various ways, is precisely what is ruled out by the suf¬ciency thesis,
because such realizations would not in and of themselves determine the
properties they realize.
A third reason is that the suf¬ciency thesis is needed to make sense of
many of the positions that physicalists have adopted and the arguments
they have offered in support of them. An intuition at the core of physi-
calism is that all the relevant physical facts ¬x all the nonphysical facts,
and the notions of supervenience and realization have both been used
to articulate this intuition further. Supervenience, in all its varieties, is
itself a relation of determination, and if one thinks of realization as a
correlative notion, then it too must be determinative. (Alternatively, if
one holds that the physical realization of a given property is typically a
subset of the subvenient base properties, realizations are, at most, partial
determinants of the properties they realize, a view I return to discuss in
Chapter 6.) And as already bruited above, the suf¬ciency thesis not only
seems necessary for reductively identifying mental and physical states in
views such as Kim™s, but it also generates the recent wave of what Jerry
Fodor calls epiphobia among nonreductionists “ epiphobia being the
fear that one is becoming an epiphenomenalist.

5 the standard view (ii): realizers as
physically constitutive
Kim™s own reductionism about the mind is also guided by a second thesis,
one at least implicitly shared by many others, including Richard Boyd,
David Lewis, and Sydney Shoemaker. I shall call this thesis the physical
constitutivity thesis:

Physical Constitutivity Thesis: Realizers of states and properties are exhaustively
physically constituted by the intrinsic, physical states of the individual whose states
or properties they are.

I understand this thesis broadly such that stronger and weaker versions
of it could be articulated in terms of the notions of supervenience, type-
identity, or token-identity. In the philosophy of psychology, this thesis
might be thought to have its methodological counterpart in the popular
Metaphysics, Mind, and Science: Two Views 105

endorsement of the idea that homuncular functionalism and functional
analysis involve the decomposition of psychological capacities into their
constituent capacities, a claim we will have reason to consider more care-
fully later.5
Since physical realizations have been claimed to provide a metaphys-
ical and explanatory basis for the higher-level properties they realize,
it is not surprising that these links between functionalism, realization,
and constitution structure (or perhaps derive from) a broader physical-
ist metaphysics, one that accords microstructure a central role. As Kim
says, speaking in the ¬rst instance of our common sense conception of
chemical kinds, but clearly with a more general view in mind:
. . . many important properties of minerals, we think, are supervenient on, and
explainable in terms of, their microstructure, and chemical kinds constitute a mi-
crostructural taxonomy that is explanatorily rich and powerful. Microstructure
is important, in short, because macrophysical properties of substances are deter-
mined by microstructure. These ideas make up our ˜metaphysics™ of microdeter-
mination for properties of minerals and other substances, a background of partly
empirical and partly metaphysical assumptions that regulate our inductive and
explanatory practices.

As Kim says a little later, “[t]o have a physical realization is to be physically
grounded and explainable in terms of the processes at an underlying
level.” Such a view is also manifest in Kim™s one-time enthusiasm for the
prospects of understanding “mind-body supervenience as an instance of
mereological supervenience,” that is, the supervenience of wholes on
their parts.6

6 smallism, the standard view,
and the fragile sciences
In Chapter 1, I claimed that smallism, discrimination in favor of the small,
lurked in the background of contemporary individualism and nativism.
The constitutivity thesis is certainly smallist, but it might well be thought
that its in¬‚uence is quite limited. After all, recall that “realization” is a
term of art with its contemporary origin in a speci¬c literature “ that
of Turing machine functionalism in the philosophy of mind. However,
I want to suggest that an explicit appeal to or an implicit reliance on
that concept and thus on the suf¬ciency and constitutivity theses can
also be found in a range of positions in the fragile sciences, and thus
the reach of the standard view of realization is signi¬cantly broader than
one might think. Consider two such positions that are related to issues
Individualism and Externalism in Cognitive Sciences

¬rst discussed in Chapter 1 in exploring the relationship between various
forms of individualism and nativism.
One such position concerns the status of collectives in the social sci-
ences. It is common to hold that social-level entities, such as electorates,
institutions, practices and rituals, and their properties are, in some sense,
nothing over and above the individuals who are involved in them and
their properties. Although this relationship has seldom been expressed
in terms of the standard notion of realization, that concept would seem
ideal to capture the “nothing over and above” aspect to that relationship:
Individuals physically constitute social-level entities, and having the in-
dividuals and their particular properties is metaphysically suf¬cient to
have social-level entities and their corresponding properties. Consider
an electorate that is in the process of voting in a new government. That
electorate is physically realized at that time by a given number of individu-
als, and by a certain majority of those individuals voting for the opposition
party the electorate is thereby in the process of voting in a new govern-
ment. Nothing more than these mundane individual-level facts is needed
for this to be true. Nonreductionists and reductionists about collectives
disagree about what this realization relation implies about social ontol-
ogy. As we saw in Chapter 1, methodological individualists in the social
sciences hold that the relevant properties of individuals are psycholog-
ical and that this has methodological implications for how to do social
A quite distinct arena in which there is likewise an implicit reliance
on the conjunction of the constitutivity and suf¬ciency thesis is develop-
mental biology in which genes are held to be both physically constituted
and metaphysically determined by particular DNA sequences. The claim
about the physical constitution of genes in general “ they are strings of
DNA “ is one of the triumphs of twentieth-century biology. That partic-
ular sequences of DNA are held to be metaphysically suf¬cient for the
presence of a given gene underlies not only comparative molecular phy-
logenetic inferences that identify the same (or, as it is typically put, a
homologous) gene across organisms belonging to two different species
or other cladistic groups, but also the robustness of the appeal to DNA
sequences in talking of the “gene for” a given phenotypic trait. Thus, it
would make sense of key aspects of how genes are conceptualized in ge-
netics and developmental biology to say that they are realized in sequences
of DNA, even if that is not how biologists have in fact put the matter. In
fact, I shall argue later in this chapter that talk of realization has quite
general application within the biological sciences.
Metaphysics, Mind, and Science: Two Views 107

While I think that it is the constitutivity thesis that is problematic in
talk of the realization of mental states, my general challenge is to the con-
junction of the suf¬ciency and constitutivity theses for at least a variety of
properties and states, including not only mental properties and states but
those from across the fragile sciences. Context can feature in an account
of realization in a number of ways, but feature it must, and I see no way of
adequately representing the role of context in such an account that does
not undermine either the suf¬ciency thesis or the constitutivity thesis. A
bald statement of my chief objection to the standard view of realization
is that the suf¬ciency and constitutivity theses are typically not true of
the same putative realizers. Often the realizations that are metaphysically
suf¬cient for the properties they realize are not exclusively physical con-
stituents of individuals with those properties. Conversely, sometimes the
physical constitution of an individual with a given property is not meta-
physically suf¬cient for that property to be present. Mental properties
are no exception here.
Physicalists who understand realization as a relation of metaphysical
determination, as most do, should embrace the idea that at least some
states and properties, including mental states and properties, have re-
alizers that extend beyond the individual instantiating them. States and
properties that have what I shall call a wide realization are prevalent in
both common-sense thinking and in the biological and social sciences.
Perhaps because there has been no general framework for such a view of
realization, this view has not been explicitly endorsed in the literature on
mental properties, although it is the view of realization that makes most
direct metaphysical sense of the widespread recognition that a range
of mental properties are not individualistic, and a view that externalists
should readily agree with. This advocacy of wide realizations represents
one way of developing a context-sensitive notion of realization.
There are initially less striking ways in which realization is context
sensitive, however, and I shall discuss two of them next.

7 context-sensitive realization
and metaphysical suf¬ciency
As a way of introducing the idea that realization is context sensitive, con-
sider the mental state of pain and the Ur-example of its realizer, C-¬ber
stimulation. As Sydney Shoemaker has pointed out, C-¬ber stimulation
is, at best, a partial realization of pain. What Shoemaker calls a core realiza-
tion of that mental state is the speci¬c part of the central nervous system
Individualism and Externalism in Cognitive Sciences

most readily identi¬ed as playing a crucial, causal role in producing or
sustaining the experience of pain. But when an individual is in pain other
parts of her central nervous system are also activated, and the activity of
these parts is crucial for C-¬ber stimulation to play the causal role that is,
according to functionalists, de¬nitive of pain.7
In general, the physical states that are partial realizations of a prop-
erty or state will be metaphysically context sensitive in that they will realize
that property or state only given their location in some broader physical
system. Considered just in themselves, they do not satisfy the suf¬ciency
thesis. Additionally, in the special case of the core realization of a prop-
erty, conceived of as the most salient part of some larger system in which
that property is instantiated, we have an epistemic dimension to the con-
text sensitivity of the realization. What we ¬nd of greatest causal salience
depends on our conceptual and perceptual abilities. It also depends on
the questions we ask, the background information we have, and, more
generally, our epistemic orientation.
The context sensitivity of partial and core realizations should be un-
controversial, but might be thought of as having little relevance here
because such realizations do not and have never been claimed to satisfy
the suf¬ciency thesis. Even if core realizations of a property are what
we most readily call to mind in thinking of the realization of that prop-
erty, there are more complete physical states of which core realizations
are a part that do satisfy the suf¬ciency thesis. Any interesting context-
sensitivity thesis about realization should apply to them, not simply to
core or other partial realizations. Following Shoemaker, we might de¬ne
a total realization of a property as just such a state of a system.
By talking of a given higher-level property, P, and the system, S, in
which P is realized, we can characterize the general distinction between
core and total realizations as follows:
(a) core realization of P: a state of the speci¬c part of S that is most
readily identi¬able as playing a crucial causal role in producing or
sustaining P
(b) total realization of P: a state of S, containing any given core real-
ization as a proper part, that is metaphysically suf¬cient for P
In particular cases, “S” is to be replaced by the appropriate system,
whether it be psychological, biological, economic, computational, chem-
ical, and so on, and their more determinate forms. In the case of pain,
the appropriate system is the nociceptive system, containing mechanical
and polymodal nociceptors in the skin (muscles and viscera), myelinated
Metaphysics, Mind, and Science: Two Views 109

and unmyelinated axons (the latter being the famed C-¬bers), spinal
neurons, parts of the brainstem and thalamus, and the somatosensory
area of the cerebral cortex. More generally, while P is a property of some
individual entity, such as an organism, S need not be identical to that
entity but, as in the example of pain, may form a part of it. Paradigms
of such systems are those in which bodily functions and their associated
properties are realized “ for example, the respiratory system, the digestive
system, the circulatory system “ that are a part of each creature with the
respective properties. Total realizations of P are exhaustively constituted
by a core realization of P plus what I will refer to as the noncore part of the
total realization.


. 18
( 62 .)