O N T O L O G Y AND IDEOLOGY 11
If moral experience is discussed in terms of the uniqueness of the impera-
five, instead of the quality, a similar analysis holds. Either the psycho-
analytic or behavioristic accounts o f the development of the superego or
conscience illustrate genetic explanations of moral experience in the impera-
tive mood. They provide empirical explanations which, if true, account
for the facts and do not lead to any undesirable metaphysic about the
nature of standards and our conformance to them. An analysis of the kind
here presented is thus "non-naturalistic" in the narrower meaning that
term has come to have in ethical discussion--insistence upon the inde-
finability of the moral predicates--yet naturalistic in the broader sense
of not admitting extranatural realms of being or ways of knowing. Clearly,
it is not necessary to deny some fairly obvious facts of human experience
in order to deny obiective reality to values. 6
NOTES
That I do not use "good" to mean "intrinsically good" and why I do not will be-
come clear in the sequel. But "good" is a one-term predicate, not a relation.
This point has been made by Gustav Bergmann. See his "Undefined Descriptive
Predicates," Philosophy and Phenomenotogical Research, vol. 8 (1947), especially
pp. 81-82. Also "Logical Positivism" in A History of Philosophical Systems, V. Ferm,
ed. (New York: Philosophical Library, 1950).
~The Philosophy of G. E. Moore (Evanston, It1.: Northwestern Universib~ Press,
1942), p. 589.
4 Compare the gestalt psychologists' insistence that such properties as well as mean-
ing and value are "in the world." But their psychological insight is frequently vitiated
by an accompanying idealistic metaphysic about meaning.
~W. K. Frankena, "The Naturalistic Fallacy," Mind, voL 48 (1939).
I have profited in my thinking on these things from frequent discussion with Pro-
lessor W. S. Sellars. See also his "Language, Rules and Behavior," in John Dewey:
Philosopher of Science and Freedom, S. Hook, ed. (New York: Dial Press, 1950).
Ontology' and Ideology
by W . V. Q U I N E
HARVARD UNIVERSITY
ON SEVERAL occasions 1 I have urged, in substance, that
I. The ontology to which an (interpreted) theory is committed com-
prises all and only the objects over which the bound variables of the
theory have to be construed as ranging in order that the statements
affirmed in the theory be true.
Bergmann 2 has lately proposed a different ontological standard, which, he
12 PHILOSOPHICAL STUDIES
feels, "represents a more adequate analysis of the traditional ontological
meaning of 'exist'."
Bergmann adopts, to begin with, the fiction of the ideal Language of a
possible world. He appreciates that this is a speculative sort of fiction, but
believes, perhaps rightly, that it can be helpful provisionally in clarifying
a variety of philosophical matters. Appealing then to this fiction, Bergmann
interprets my position (for comparison with his own, to be explained
hereafter) as follows: for me, he said,
II. "Properties of the first [logical] type exist in a world if and only if
quantification over the predicate variables of this type occurs in its ideal
language."
Quite properly for the purposes of his note, Bergmann is here limiting
his consideration of my doctrine to the case of predicate variables of first
type. But I have a minor quarrel with statement II, having to do with
the question of class versus property, which I should mention before moving
to more central matters. Of course unbindable predicate variables, viewed
simply as schematic letters, carry no ontological commitment; however,
once predicate variables are bound, the question arises whether they are
bound as variables referring to classes as values or as variables referring to
properties as values. If this question is not answered in an explicit interpre-
tation of the notation, we in practice look to implicit cues in the laws of
substitutivity of equivalents; thus we may take the values as classes if the
system obeys extensional substitutivity laws, but otherwise we must take
them as properties.
A more important point is that my own statement (I), unlike state-
ment II, explicates only the ontological commitments of a theory and not
the ontological truth about a world. B~gmann extends I to the latter
purpose, in II, by means of his own fiction of the ideal language of a
world. This extension is interesting and, granted his fiction, it is certainly
the step I should want to make; I merely mention that it is his.
Now let us proceed to the alternative ontological standard which Berg-
mann proposes. His expression of it, limited again to the case of predicate
variables of the first type, is as follows:
III. "Properties of the ~rst type exist in a world if in speaking about an
ideal language of this world, I find it to contain descriptive constants
that are substitution instances for its predicate variables of the first type."
Statement III supposes the ideal Language to be of a form which con-
rains primitive constants substitutable for predicate variables. Now I cer-
tainly favor limiting our considerations conveniently to certain species of
language forms--viz., those involving quantification of variables, rather
than alternative expedients such as combinators. But I dislikc assuming
ONTOLOGY AND IDEOLOGY 13
that there are primitive constants of a sort substitutable for bindable vari-
ables; for, as I have argued elsewhere, 8 such constants are always eliminable,
and there are broad advantages in supposing them (in theory) eliminated.
When they are eliminated, we are left still with primitive predicates; but
not in the status of class names or property names, substitutable for bind-
able class variables or property variables. The bound variables themselves
become the sole channel of reference to classes, to properties, or to any-
thing at all.
W e might try recasting statement III in the following form, which is
no longer subject to the above criticism.
IV. Properties of the first type exist in a world if in speaking about an
ideal language of this world I find it to contain (a) quantification over
predicate variables of the first type, and (b) descriptive constant predi-
cates which take individual variables as arguments.
Against IV, however, there is the objection that almost any descriptive
language of usual form, whether nominalistic or realistic in its presupposi-
tions by ordinary lights, may be expected to contain some descriptive
constant predicates which take individual variables as arguments; hence
the clause (b) of IV adds little or nothing to II.
Anyway it is not evident why there should be a connection between
constant predicates used and entities presupposed. Surely the mere occur-
rence of a predicate in the formulation of a theory is not sufficient in
order that the theory presuppose a corresponding universal entity--a cor-
responding class or property. Nor is it necessary; for we are familiar with
theories which imply that there are indenumerably many classes or prop-
erties (even of the first type), though the available predicates are ne-
cessarily denumerable.
Indeed, the naturalness and appropriateness of my unmodified state-
ment (I) seems scarcely open to question as long as 'ontology' is taken
to mean literally 'doctrine of what there is,' and as long as the theory
whose ontological commitments are being examined is expressed in quanti-
fictional form subject to the customary interpretation. For the universal
and existential quantifiers mean simply 'every entity (of appropriate type)
is such that' and 'some entity (of appropriate type) is such that.' The
theory presupposes all and any of those entities whose nonoccurrence
within the ranges of the variables of quantification would render Farts of
the theory false.
There is doubtless more to metaphysics than ontology in the above sense;
and some of this additional matter is perhaps thought of also as ontology
in some sense. But it accords with etymology, and also with some portion
of the philosophical tradition, to limit 'ontology' to just that part of meta-
physics which asks what there is.
14 PHILOSOPHICAL STUDIES
Given a theory, one philosophically interesting aspect of it into which
we can inquire is its ontology: what entities are the variables of quantifica-
tion to range over if the theory is to hold true? Another no less important
aspect into which we can inquire is its ideology (this seems the inevitable
word, despite unwanted connotations) : what ideas can be expressed in it?
Now the spirit of Bergmann's proposal (III), as I see it, is this: in an
effort not to omit important issues of ideology he would warp ontology
around to include them. It is clearer, I think, to recognize in ontology and
ideology two distinct domains of inquiry.
The ontology of a theory stands in no simple correspondence to its
ideology. The ontology' of the usual theory of real numbers is indenumer-
able, but the ideology--the range of severally expressible ideas--is de-
numerable. The ideology of the theory of real numbers embraces individual
ideas of just denumerably many of the indenumerably many real numbers.
On the other hand the ideology also embraces many such ideas as sum,
root, rationality, algebraicity, and the like, which need not have any onto-
logical correlates in the range of the variables of quantification of the theory.
Two theories can have the same ontology and different ideologies. Two
theories of real numbers, for example, may agree ontologically in that each
calls for all and only the real numbers as values of its variables, b u t they
may still differ ideologically in that one theory is expressed in a language
into which the sentence
i. the real number x is a whole number
can be translated, while the other theory is not. Note the importance of
this particular example; Tarski 4 has proved the completeness of a certain
elementary theory T of real numbers, and we know from G6ders proof of
the incompletabitity of the theory of whole numbers that Tarski's achieve-
ment would have been impossible if i were translatable into T.
The ideology of a theory is a question of what the symbols mean; the
ontology" of a theory is a question of what the assertions say or imply that
there is. The ontology of a theory may indeed be considered to be implicit
in its ideology; for the question of the range of the variables of quantifica-
tion may be viewed as a question of the full meaning of the quantifiers.
As a subdivision of ideology there is the question of what ideas are
fundamental or primitive for a theory, and what ones derivative. Berg-
mann suggests how this distinction might be drawn with the help of an
antecedent distinction between logical and extra-logical truth. Various
points in Bergmann's paper, this among them, are best viewed as con-
tributions not to ontological but to ideological inquiry.
In the foregoing paragraphs I have contrasted the ontology of a theory
with the ideology of a theory. But the contrast carries over also into abso-
lute terms; in absolute ontology we ask what there really is, and in abso-
ONTOLOGY AND IDEOLOGY 15
lute ideology we ask what ideas can legitimately be had, or what primitive
ideas are given to us as a basis for thinking. Bergmann has suggested how
the fiction of ideal languages can help us make sense of these issues.
I have described the ideology of a theory vaguely as asking what ideas
are expressible in the language of the theory. Urgent questions of detail
then arise over how to construe 'idea.' Perhaps, for what is important in
ideological investigations, the notion of ideas as some sort of mental en-
tities can be circumvented. Much that belongs to ideology can be handled
in terms merely of the translatability of notations from one language into
another; witness tile mathematical work on definability by Tarski and
others. 5 A typical theorem of ideology in this vein is the above observa-
tion that i is not translatable into T.
Both ontology and ideology in their relativized aspects--the ontology
of a theory, the ideology of a theory--belong to what is commonly called
semantics. But, as I have urged elsewhere,6 a fundamental cleavage needs
to be observed between two parts of so-called semantics: the theory of
reference and the theory of meaning. The theory of reference treats of
naming, denotation, extension, coextensiveness, values of variables, truth;
the theory of meaning treats of synonymy, analyticity, syntheticity, en-
tailment, intension. Now the question of the ontology of a theory is a
question purely of the theory of reference. The question of the ideology
of a theory, on the other hand, obviously tends to fall within the theory
of meaning; and, insofar, it is heir to the miserable conditions, the virtual
lack of scientific conceptualization, which characterize the theory of mean-
ing5 But a partial analogue of the theory of meaning is contained within
the theory of reference itself; here extension takes the place of intension,
coextensiveness of predicates takes the place of synonymy of predicates, and
truth takes the place of analyticity. Much in the way of ideological study
can be usefully pursued thus within the theory of reference. Such in par-
ticular is the status of the mathematical studies of definability mentioned
above.
NOTES
1E.g., in "Designation and Existence," Journal of Philosophy, 36:701-9 (1939);
"Notes on Existenceand Necessity,"ibid., 40:113-27 (1943).
2Gustav Bergmann, "A Note on Ontology," Philosophical Studies, 1:89-92.
Mathematical Logic (Cambridge, Mass.: Harvard University Press, 1947), sec. 27;
Methods o[ Logic (New York: Holt, 1950), sec. 37.
'Alfred Tarski, A Decision Method /or Elementary Algebra and Geometry (Santa
Monica, 1948).
5Atfred Tarski, "Sur tes ensembles dtfinissabtes de nombres rtels," Fundamenta
Mathematicae, 17:210-39 (1931); "Einige methodologischeUntersuchungenfiber die
Definierbarkeitder Begriffe,"Erkenntuis, 4: 80-100 ( 1935-36); Julia Robinson, "Defina-
bility and Decision Problems in Arithmetic," Journal of Symbolic Logic, 14:98-114
(1949); W. V. Quine, "A Minimum Primitive for Number Theov¢," ibid., at press.
"Semantics and Abstract Objects." To be published among the papers presented at
the meeting of April 29, 1950, of the Institute for the Unity of Science.
See my "Two Dogmasof Empiricism,"forthcomingin Philosophical Review, vol. 60.
