Structure and Nature

Author(s): W. V. Quine
Source: The Journal of Philosophy , Jan., 1992, Vol. 89, No. 1 (Jan., 1992), pp. 5-9
Published by: Journal of Philosophy, Inc.

Stable URL: https://www.jstor.org/stable/2026889

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms

is collaborating with JSTOR to digitize, preserve and extend access to The Journal of
Philosophy

This content downloaded from

130.225.157.199 on Fri, 13 Oct 2023 03:29:51 +00:00
All use subject to https://about.jstor.org/terms
 THE JOURNAL OF PHILOSOPHY
VOLUME LXXXIX, NO. 1, JANUARY 1992

+_ 0 -4-

STRUCTURE AND NATURE

W E are familiar with three adequate but incompatible ways

of modeling number theory in set theory, or the theory
of classes; and there are infinitely many more. We bandy
our numbers without caring which classes we are bandying from
among this wealth of alternatives. We are just content that we are
operating somewhere within the ontology of classes to which we
have committed ourselves anyway for other purposes. This situation,
long since remarked upon by Paul Benacerraf' and others, receives
formal expression in Ramsey sentences.2 Such a sentence merely
affirms that there are classes obeying the desired laws-arithmetical
laws in this case-and then proceeds to apply such unspecified
classes to whatever arithmetical business we had in mind.
That is the structuralist treatment of number. It is just a way of
eliminating an idle question-"What is number?"-and a gratui-
tous decision among indifferent alternatives. Ontological economy
is not its point; the same old classes are presupposed.
David Lewis3 has ventured the next step: a structuralist treatment
of classes themselves. Now just as structuralism in number theory
depended on the various ways of reducing numbers to classes, so
structuralism in class theory depends on reducing classes in turn
to-what? For Lewis, as for most of us, I should like to think, there
are only classes and concrete individuals. So his structuralism is in-
deed ontologically significant. It professes a reduction of classes to
individuals-hence out and out nominalism.
Arithmetical structuralism, as expressed in Ramsey sentences, de-
pended on there being classes in which the required arithmetical
structures could be defined and realized, though we were freed of
choosing among alternatives. Similarly for Lewis's structuralism of

' "Mathematical Truth," this JOURNAL, LXX, 19 (November 8, 1973): 661-79.

2 Frank Ramsey, The Foundations of Mathematics (New York: Routledge,
1931), pp. 212-36.
3 Parts of Classes (New York: Blackwell, 1991).

0022-362X/92/8901/5-9 (?) 1992 The Journal of Philosophy, Inc.

This content downloaded from

130.225.157.199 on Fri, 13 Oct 2023 03:29:51 +00:00
All use subject to https://about.jstor.org/terms
6 THE JOURNAL OF PHILOSOPHY

classes; it depends, for its truth, upon there being individuals in

which the desired classial structures can be defined and realized.
Thus he needs a theory of individuals.
What he uses is mereology, or the theory of part and whole, with
atoms. By ingenious constructions he carries the day, contingently
on there being enough atoms, and granted George Boolos's4 view of
second-order logic as just a classless logic of plurals. If there are not
enough atoms, then the higher reaches of set theory turn out mean-
ingful but false. How much of axiomatic set theory is true comes to
depend on the size of the concrete universe: on how many atoms
there are. If there are only finitely many, then so much the worse for
classical set theory even at its weakest. Truth in mathematics comes
to depend on the richness of nature. Lewis appreciates all this.
Uncertain as to the richness of nature, we can still pursue higher
set theory in a conjectural spirit, finding out how it would be if
nature were up to it. In any event, Lewis's construction renders the
whole of set theory meaningful from his nominalistic point of view,
however much of it is thereby rendered true and how much false.
And, ironically enough, the scientific utility of applied mathematics
is independent of all that.
Structuralism for classes, hence for all abstract objects, is undeni-
ably congenial. They are things that are known anyway only by their
structural role in cognitive discourse; never by ostension. By osten-
sion and extrapolation we do learn what individuals qualify as
members of the class of cats, but that class itself is just an abstract
entity to which each cat bears a cryptic epsilon relation. Your class
of cats and mine can still be different things, and your membership
relation and mine can be different, if it makes any sense to say so,
though we see eye to eye on every cat. Such is the appeal of struc-
turalism regarding abstract objects.
Much though I admire Lewis's reduction, however, it is not for
me. My own line is a yet more sweeping structuralism, applying to
concrete and abstract objects indiscriminately. I base it, paradoxical
as this may seem, on a naturalistic approach to epistemology. Natu-
ral science tells us that our ongoing cognitive access to the world
around us is limited to meager channels. There is the triggering of
our sensory receptors by the impact of molecules and light rays.
Also there is the difference in muscular effort sensed in walking up
or down hill. What more? Even the notion of a cat, let alone a class
or number, is a human artifact, rooted in innate predisposition and
cultural tradition. The very notion of an object at all, concrete or
abstract, is a human contribution, a feature of our inherited appara-
tus for organizing the amorphous welter of neural input.
4 "Nominalist Platonism," Philosophical Review, XCIV, 3 (July 1985): 327-44.

This content downloaded from

130.225.157.199 on Fri, 13 Oct 2023 03:29:51 +00:00
All use subject to https://about.jstor.org/terms
 STRUCTURE AND NATURE 7

Language, the vehicle of science, is linked to our neural input by

neural mechanisms of association or conditioning. Early in our
learning of language we learn to associate certain expressions with
certain ranges of neural input. I call these expressions observation
sentences. Grammatically, some are indeed sentences, e.g., 'It's rain-
ing', and some are nouns or adjectives, e.g., 'Cat' or 'Milk' or
'White'. To begin with, they are just expressions associated ho-
lophrastically with ranges of neural input. They are as if to say, with
William James, "Hello, thingumbob again"-except that even 'thin-
gumbob' hints more of objectual reference than I could wish. The
observation sentence names nothing, to begin with, neither neural
input nor external object. But it is learned from adults who have
learned the ways of reference, so it is indeed an expression, such as
'It's raining' or 'Cat' or 'Milk' or 'White', that is destined for even-
tual integration into a system of objectual reference on the child's
own part as he matures. This growing system of terms and pur-
ported objects is our science of nature, our organization of the
welter of neural input.
Reification and implication are the key principles by which that
organizing proceeds. We master implication in the course of learn-
ing to use the logical particles, such as 'not', 'and', 'or', 'if', 'every',
'some'. For instance, our learning to use 'not' and 'or' consists in
part in learning to assent to 'q' whenever we have assented to 'not p'
and 'p or q'. Our learning of 'every' consists in part in learning to
assent to 'a is a G' once we have assented to 'a is an F and 'Every F
is a G'.
It is in implications of this last sort that reification makes its con-
tribution, in our having reified the object a and the other Fs and Gs.
The reification of bodies comes in stages in one's acquisition of
language, each successive stage being more clearly and emphatically
an affirmation of existence. The last stage is where the body is recog-
nized as identical over time, despite long absences and interim modi-
fications. Such reification presupposes an elaborate schematism of
space, time, and conjectural hidden careers or trajectories on the
part of causally interacting bodies. Such identifications across time
are a major factor in knitting implications across the growing fabric
of scientific hypotheses.
At more sophisticated stages in the development of language and
science, implications are enhanced by positing further objects, no
longer observable; thus subvisible particles, also numbers and other
classes.
The implications thus forged are what relate our evolving scien-
tific theory to our heterogeneous neural input and hence ultimately
to the external world that our scientific theory purports to describe.

This content downloaded from

130.225.157.199 on Fri, 13 Oct 2023 03:29:51 +00:00
All use subject to https://about.jstor.org/terms
8 THE JOURNAL OF PHILOSOPHY

This fabric of interlocking sentences hooks on to the neural input at

the observation sentences. Some of these, we saw, were learned by
primitive conditioning to ranges of neural input. Many further ob-
servation sentences, more sophisticated ones, are scientific sen-
tences retroactively associated with observation through the inter-
vention of theory; thus the chemist sees at a glance that there was
copper in the solution, and perceives at a sniff that sulphur dioxide
escaped. Now the way implication links theory to neural input is that
a bundle of scientific hypotheses is found to imply that the fulfill-
ment of one of two given observation sentences should always en-
sure fulfillment of the other. This dictates an experiment: he brings
about fulfillment of the one and checks for the other. Failing the
other, the bundle of hypotheses is discredited; one or another of
them must be revoked.
Such, in schematic caricature, is surely the evidential relation of
scientific theory to the triggering of our neuroreceptors by the ex-
ternal world. I have ventured more detail elsewhere,5 and much
more wants venturing. Perhaps a fair bit of science might be found
on fuller analysis to be beyond the reach of evidence altogether. But
this sketch will suffice for what I want to say about structuralism.
The point I now want to make is one that over the years I have
repeatedly made in terms of what I call proxy functions. The point
is that if we transform the range of objects of our science in any
one-to-one fashion, by reinterpreting our terms and predicates as
applying to the new objects instead of the old ones, the entire evi-
dential support of our science will remain undisturbed. The reason
is twofold. First, implication hinges only on logical structure and is
independent of what the objects, the values of the variables, may be.
Second, the association of observation sentences with ranges of
neural input is holophrastic. It is independent of reifications, inde-
pendent of whatever objects the observation sentences or their parts
may be taken to refer to as terms.
The conclusion is that there can be no evidence for one ontology
as over against another, so long anyway as we can express a one-to-
one correlation between them. Save the structure and you save all.
Certainly we are dependent on a familiar ontology of middle-sized
bodies for the inception of reification, on the part both of the indi-
vidual and of the race; but once we have an ontology, we can change
it with impunity. For abstract objects this is unsurprising, and quite
in the spirit of Ramsey, Lewis, and Benacerraf. For familiar bodies it
is less intuitive. But we must bear in mind that an observation sen-

5 The Roots of Reference (Peru, IL: Open Court, 1974); and Pursuit of Truth
(Cambridge: Harvard, 1990).

This content downloaded from

130.225.157.199 on Fri, 13 Oct 2023 03:29:51 +00:00
All use subject to https://about.jstor.org/terms
 STRUCTURE AND NATURE 9

tence, say, 'That's a rabbit', retains all its original visual associations;
it is only the purported object that is gratuitously displaced.
This global ontological structuralism may seem abruptly at odds
with realism, let alone naturalism. It would seem even to undermine
the ground on which I rested it: my talk of impacts of light rays and
molecules on nerve endings. Are these rays, molecules, and nerve
endings themselves not disqualified now as mere figments of an
empty structure?
Naturalism itself is what saves the situation. Naturalism looks only
to natural science, however fallible, for an account of what there is
and what what there is does. Science ventures its tentative answers in
man-made concepts, perforce, couched in man-made language, but
we can ask no better. The very notion of object, or of one and many,
is indeed as parochially human as the parts of speech; to ask what
reality is really like, however, apart from human categories, is self-
stultifying. It is like asking how long the Nile really is, apart from
parochial matters of miles or meters. Positivists were right in brand-
ing such metaphysics as meaningless.
But early positivists were wrong if and when they concluded that
the world is not really composed of atoms or whatever. The world is
as natural science says it is, insofar as natural science is right; and
our judgment as to whether it is right, tentative always, is answerable
to the experimental testing of predictions.
We saw that reference could be wildly reinterpreted without vio-
lence to evidence. We see now that that is just part of a wider pic-
ture. Presumably yet more extravagant departures, resistant even to
sentence-by-sentence interpretation into our own science, could
conform equally well to all possible observations. If we were to en-
counter such a case and somehow recognize it as such, we might
even proceed to master it and then switch back and forth for richer
perspectives on reality. But naturalism would still counsel us that
reality is to be grasped only through a man-made conceptual
scheme, albeit any of various.
My global structuralism should not, therefore, be seen as a struc-
turalist ontology. To see it thus would be to rise above naturalism
and revert to the sin of transcendental metaphysics. My tentative
ontology continues to consist of quarks and their compounds, also
classes of such things, classes of such classes, and so on, pending
evidence to the contrary. My global structuralism is a naturalistic
thesis about the mundane human activity, within our world of
quarks, of devising theories of quarks and the like in the light of
physical impacts on our physical surfaces.
W. V. QUINE

Harvard University

This content downloaded from

130.225.157.199 on Fri, 13 Oct 2023 03:29:51 +00:00
All use subject to https://about.jstor.org/terms