Ontological

Relativity and
Other Essays

by W. V. Quine

Columbia University Press

New York and London
1969
68 Ontological Relativity
more inclusive theory. This similarity should perhaps not sur-
prise us, since both ontology and satisfaction are matters of 3
reference. In their elusiveness, at any rate-in their emptiness
now and again except relative to a broader background-both
truth and ontology may in a suddenly rather clear and even
tolerant sense be said to belong to transcendental metaphys- Epistemology
ics. 1s
Naturalized
Note added in proof. Besides such ontological reduction as
is provided by proxy functions (d. pp. 55-60), there is that
which consists simply in dropping objects whose absence will
not falsify any troths expressible in the notation. ColIlmonly
this sort of deflation can be managed by proxy functions, but Epistemology is concerned with
R. E. Grandy has shown me that sometimes it cannot. Let us the foundations of science. Conceived thus broadly, episte-
by all means recognize it then as a further kind of reduction. mology includes the study of the foundations of mathematics
In the background language we must, of course, be able to say as one of its departments. Spedalists at the turn of the century
what class of objects is dropped, just as in other cases we had thought that their efforts in this particular department were
to be able to specify the proxy function. This requirement achieving notable success: mathematics seemed to reduce alto-
seems sufficient still to stem any resurgence of Pythagoreanism gether to logic. In a more recent perspective this reduction is
on the strength of the Liiwenheim-Skolem theorem. seen to be better describable as a reduction to logic and set
18 In developing these thoughts I have been helped by discussions theory. This correction is a disappointment epistemologically,
with Saul Kripke, Thomas Nagel, and especially Burton Dreben. since the firmness and obviousness that we associate with logic
cannot be claimed for set theory. But still the success achieved
in the foundations of mathematics remains exemplary by com-
parative standards, and we can illuminate the rest of episte-
mology somewhat by drawing parallels to this department.
Studies in the foundations of mathematics divide symmetri-
cally into two sorts, conceptual and doctrinal. The conceptual
studies are concerned with meaning, the doctrinal with truth.
The conceptual studies are concerned with clarifying concepts.
by defining them, some in terms of others. The doctrinal-
studies are concerned with establishing laws by proving them,
 70 Epistemology Naturalized Epistemology Naturalized 71
some on the basis of others. Ideally the obscurer concepts Still there remains a helpful thought, regarding epistemol-
would be defined in terms of the clearer ones so as to maximize ogy generally, in that duality of structure which was especially
clarity, and the less obvious laws would be proved from the conspicuous in the foundations of mathematics. I refer to the
more obvious ones so as to maximize certainty. Ideally the bifurcation into a theory of concepts, or meaning, and a theory
del)nitions would generate all the concepts from clear and dis- of doctrine, or truth; for this applies to the epistemology of
tinct ideas, and the proofs would generate all the theorems natural knowledge no less than to the foundations of mathe-
from self-evident truths. -"
matics. The parallel is as follows. Just as mathematics is to be
The two ideals are linked. For, if you define all the concepts reduced to logic, or logic and set theory, so natural knowledge
by use of some favored subset of them, you thereby show how is to be based somehow on sense experience. This means ex-
to translate all theorems into these favored terms. The clearer !, plaining the notion of body in sensory terms; here is the eon-
these terms are, the likelier it is that the truths couched in /jeptual side. And it means justifying our knowledge of truths
them will be obviously true, or derivable from obvious. truths. of nature in sensory terms; here is the doctrinal side of the
If in particular the concepts of mathematics were all reducible bifurcation. ~
to the clear terms of logic, then all the truths of mathematics Hume pondered the epistemology of natural knowledge on
would go over into truths of logic; and surely the truths of both sides of the bifurcation, the conceptual and the doctrinal.
logic are all obvious or at least potentially obvious, Le., de- His handling of the conceptual side of the problem, the ex-
rivable from obvious truths by individually obvious steps. planation of body in sensory terms, was bold and simple: he
This particular outcome is in fact denied us, however, since
mathematics reduces only to set theory and not to logic proper.',
Such reduction still enhances clarity, but only because of the
interrelations that emerge and not because the end terms of
the analysis are clearer than others. As for the end truths, the
'tidentified bodies outright with the sense impressions. If com-
mon sense distinguishes between the material apple and our
ense im,pressions of it on the ground that the apple is one and
enduring while the impressions are many and Heeting, then,
Hume held, so much the worse for common sense; the notion
axioms of set theory, these have less obviousness and certainty of its being the same apple on one occasion and another is a v'
to recommend them than do most of the mathematical theo- vulgar confusion.
rems that we would d,erive from them. Moreover, we know Nearly a century after Hume's Treatise, the same view of
from Godel's work that no consistent axiom system can cover bodies was espoused by the early American philosopher Alex-
,mathematics even when we renounce self-evidence. Reduction ander Bryan Johnson.' "The word iron names an associated
in the foundations of mathematics remains mathematically and sight and feel," Johnson wrote.
philosophically fascinating, but it does not do what the epis- What then of the doctrinal side, the justification of our
temologist would like of it: it does not reveal the ground of knowledge of truths about nature? Here, Hume despaired. By
mathematical Imowledge, it does not show how mathematical 1 A. B. Johnson, A Treatise on Language (New York, 1836; Berkeley,
certainty is possible. 1947).
 72 Epistemology Naturalized Epistemology Naturalized 73
his identification of hodies with impressions he did succeed in of the epistemology of natural knowledge. The other is resort~
construing some singular statements about bodies as indubita- to the resources of set theory as auxiliary concepts. The epiS-
ble truths, yes; as truths about impressions, directly known. temologist who is willing to eke out his austere ontology of
But general statements, also Singular statements about the fu- s~ense impressions with these set-theoretic auxiliaries is sud-
ture, gained no increment of certainty by being construed as fdenly rich: he has not just his impressions to play with, but
about impressions. t§~ets of them~ and sets of sets, and so on up. Constructions in
On the doctrinal side, I do not see that we are farther along the foundations of mathematics have shown that such set-
today than where Hume left us. The Humean predicament is theoretic aids are a powerful addition; after all, the entire glos-
the human predicament. But on the conceptual side there has sary of concepts of classical mathematics is constructible from
been progress. There the crucial step forward was made al- them. Thus equipped, our epistemologist may not need either
ready before Alexander Bryan Johnson's day, although John- to identify bodies with impressions or to settle for contextual
son di~ not emulate ,it. It was made by Bentham in his- theory definition; he may hope to find in some subtle construction of
of fictions/Bentham s step was the recognition of contextual sets upon sets of sense impressions a category of objects enjoy-
defini~on, or what he called paraphrasi.':JHe recognized that to in,!: just the formula properties that he wants for bodies.
explam a term we do not need to specify an object for it to IThe two resorts are very unequal in epistemological status.
" refer to, nor even specify a synonymous word or 'phrase; we Contextual definition is unassailable. Sentences that have been\
• need only show, by whatever means, how to translate all the given meaning as wholes are undeniably meaningful, and the
• whole sentences in which the term is to be used. Hume's and use they make of their component terms is therefore meaning-
Johnson'S desperate measure of identifying bodies with im- ful, regardless of whether any translations are offered for those
pressions ceased to. be the only conceivable way of making terms in isolatio~Surely Hume and A. B. Johnson would have
sense of talk of bodIeS, even granted that impressions were the used contextual definition with pleasure if they had thought of
only reality. One could undertake to explain talk of bodies in it. Recourse to sets, on the other hand, is a drastic ontological,
terms of talk of impressions by translating one's whole sen- ..!Q9_~e, a retreat from the austere ontology of impresSions.
te~ces abont bodies into whole sentences about impressions, There are philosophers who would rather settle for bodies out-
wlth~ut. equating the bodies themselves to anything at all. right than accept all these sets, which amount, after all, to the
• /ThIS Idea of contextual definition, or recognition of the sen- whole abstract ontology of mathematics.
v tence as .the primary vehicle of meaning, was indispensable to This issue has not always been clear, however, owing to de-
the ensum~ ~e:elopments in the foundations of mathematic~ ceptive hints of continuity between elementary logic and set
It was expliCIt m Frege, and it attained its full flower in Rus- theory. This is why mathematics was once believed to reduce
sell's doctrine of singalar descriptions as incomplete symbols. to logic, that is, to an innocent and unquestionable logic, and
Contextual definition was one of two resorts that could be to inherit these qualities. And this is probably why Russell was
expected to have a liberating effect upon the conceptual side content to resort to sets as well as to contextual definition when
 74 Epistemology Naturalized Epistemology Naturalized 75
,in Our Knowledge of the External World and elsewhere he trine must fall short of certainty. The other reason was that
addressed himself to the epistemology of natural knowledge, such constructions would deepen our understanding of our dis-
on its conceptual side. course about the world, even apart from questions of evidence;
I To account for the external world as a logical construct of it would make all cognitive discourse as clear as observation
I
~~nse data-such, in Russell's terms, was the program-:-rr was-
Carnap, in his Der logische Aufbau der Welt of 1928, who
(_,terms and logiC and, I must regretfully add, set theory.
It was sad for epistemologists, Hume and others, to have to
came nearest to executing it. acquiesce in the impossibility of strictly deriving the science of
This was the conceptual side of epistemology; what of the the external world from sensory evidence. Two cardinal tenets
doctrinal? There the Humean predicament remained unal- of empiricism remained unassailable, however, and so remain
I teredo Carnap's constructions, if carried successfully to comple- to this day. One is that whatever evidence there is for science. J
/"
if '.,
tion, would have enabled us to translate all sentences about is sensory evidence. The other, to which I shall recur, is that all !
the world into terms of sense data, or observation, pl..,s logic inculcation of meanings of words must rest ultimately on'
and set theory. But themere£act that a sentence is couched in ·s,:;;;so;:y-';;;idence. Hence the continuing attractiveness of the
terms of observation, logic, and set theory does not mean that idea of a logischef Attfbau in which the sensory content of dis-
it can be proved from observation sentences by logic and set course would stand forth explicitly.
i theory. The most modest of generalizations about observable If Carnap had successfully carried such a construction
J [ traits will cover more cases than its ntterer can have had occa- through, how could he have told whether it was the right one?
• §ion actually to observe. The hopelessness of grounding natural The question would have had no point. He was seeking what
science upon immediate experience in a firmly logical way was he called a rational reconstwction. Any consh'uction of physi-
acknowledged. The Cartesian quest for certainty had been the calistic discourse in terms of sense experience, logic, and set
_ remote motivation of epistemology, both on its conceptual and theory would have been seen as satisfactory if it made the
its doctrinal side; but that quest was seen as a lost cause. To physicalistic discourse come out right. If there is one way there
endow the truths of nature with the full authority of imme- are many, but any would be a great achievement.
diate experience was as forlorn a hope as hoping to endow the But why all this creative reconstruction, all this make-
truths of mathematics with the potential obviousness of ele- believe? The stimulation of his sensory receptors is all the evi-
mentary logic. dence anybody has had to go on, ultimately, in arriving at his
VVhat then could have motivated Carnap's heroic efforts on picture of the world. VVhy not just see how this consh'uction
the conceptual side of epistemology, when hope of certainty on really proceeds? VVhy not settle for psychology? Such a sur-
the doctrinal side was abandoned? There .were two good rea- render of the epistemological burden to psychology is a move
~ sons still. One was that such constructions could be expected to a. t was disallowed in earlier times as circular reasoning. If the U
,,/ I:licit an.d clarify the sensory evidence for science, even if the epistemologisfs goal is validation of the grounds of empirical
t~nferentJaI steps between sensory evidence and scientific doc-
ffiscience, he defeats his purpose by nsing psychology or other
 76 Epistemology Naturalized Epistemology Naturalized 77
em~irical. science in the validation. However, such scruples vised to suit. This plan, however illuminating, does not afTer
agamst CIrcularity have little point once we have stopped any key to translating the sentences of science into terms of
dreaming of deducing science from observations. If we are out observation, logiC, and set theory.
simply to understand the link between observation and sci- We must despair of any such reduction. Carnap had de-'
ence, we are wen advised to use any available information in- spaired of it by 1936, when, in "Testability and meaning," 2 he
introduced so-called reduction forms of a type weaker than
cluding that prOVided by the very science whose link With' ob-
r definition. Definitions had shown always how to translate sen-
servation we are seeking to understand.
But there remains a different reason, unconnected with fears ..,/' tences into equivalent sentences. Contextual definition of
term showed how to translate sentences containing the term
a.
of circularity, for still favoring creative reconstruction. We
• should like to be able to translate science into logic and ob- Unto equivalent sentences lacking the term. Reduction forms of
[\ servatio~ terms. and set theory. This would be a great epis- Carnap's liberalized kind, on the other hand, do not in general
temologICal achIevement, for it would show all the rest of the give equivalences; they give i)11plications. They explain a new
1 f .
J concepts a science to be theoretically superfluous. It would term, if only partially, by specifying some sentences which are
legitimize them-to whatever degree the concepts of set the- implied by sentences containing the term, and other sentences
~ry, lOgic, and observation are themselves legitimate-by show- which imply sentences containing the term.
mg that everything done with the one apparatus could in prin- It is tempting to suppose that the countenancing of reduc-
ciple be done with the other. If psychology itself could deliver tion forms in this liberal sense is just one further step of liber~
a truly translational reduction of this kind, we should welcome alization comparable to the earlier one, taken by Bentham, of
it; b~t certainly it cannot, for certainly we did not grow up countenancing contextual definition. The former and sterner
learnmg definitions of physicalistic language in terms of a prior kind of rational reconstruction might have heen represented as
language of set theory, logic, and observation. Here, then, a fictitious history in which we imagined our ancestors intro-
would be good reason for persisting in a rational reconstruc- ducing the terms of phYSicalistic discourse on a phenomenalis-
tion: we want to establish the essential innocence of phYSical . tic and set-theoretic basis by a succession of contextual defini-
concepts, by showing them to be theoretically dispensable. tions. The new and more liberal kind of rational reconstruction
The fact is, though, that the construction which Carnap out- is a fictitious history in which we imagine our ancestors intro-
lined in Der logi,che Aufbalt der Welt does not give transla- ducing those terms by a succession rather of reduction forms of
tional reduction either. It would not even if the outline were the weaker sort.
filled in. 'fobe crucial point comes where Carnap is explaining This, however, is a wrong comparison. The fact is rather that
how to aSSIgn sense qualities to pOSitions in phYSical space and the former and sterner kind of rational reconstruction, where
time. These aSSigmnents are to be made in such a way as to definition reigned, embodied no fictitious history at all. It was
fulfill, as well as possible, certain desiderata which he states nothing more nor less than a set of directions-or would have
and with growth of experience the assignments are to be re: 'Philosophy of Science 3 (1936),419--471; 4 (1937), 1--40.
 78 Epistemology Naturalized Epistemology Naturalized 79
been, if successful-for accomplishing everything in terms of rmeanings of typical statements about the external world are.
phenomena and set theory that we now accomplish in terms of l!naccessible and ineffable.
bodies. It would have been a true reduction by translation, a . How is this inaccessibility to be explained? Simply on the
legitimation by elimination. Definil'e est eliminal'e. Rational ground that the experiential implicatio~s of ~ typical ~tatement
reconstruction by Carnap's later and looser reduction forms about bodies are too complex for fimte aXlOmahzatlOn, how-
does none of this.
To relax the demand for definition, and settIe for a kind of
ever lengthy? No; I have a different exp~anation. It ~s t~at :liii"\
I
typical statement about bodies has no fund. of expenentlal Im-
reduction that does not eliminate, is to renounce the last re- plications it can call its own. A substantlal ~as.' of th~ory,
maining advantage that we supposed rational reconstruction to taken together, will commonly have experientlallmphcatlOns;
have over straight psychology; namely, the advantage of trans- this is how ~e make verifiable predictions. We may not be able
I" lational reduction. If all we hope for is a reconstruction that to explain why we ;':;:-ive 'attheorie;;-;;;bich make successful
, links science to experience in explicit ways short of translation, predictions, but we do arrive at such theories. .,...!I
then it would seem more sensible to settle for psychology. Bet- Sometimes also an experience implied by a theory falls to
ter to discover how science is in fact developed and learned come off; and then, ideally, we declare the theory false. But
Lthan to fabricate a fictitious structure to a similar effect. the failure falsifies only a block of theory as a whole, a con-
iThe empiricist made one major concession when he de- junction of many statements. The failure shows that one. or
spaired of deuucing the truths of nature from sensory evi- more of those statements is false, but it does not show whIch.
dence. In despairing now even of translating those truths into
terms of observation and logico-mathematical auxiliaries, he
rthe predicted experiences, true and false, are not implied by
.anyone of the component statements of the theory rather than
makes another major concessi'?r0 For suppose we hold, with ianother. The component statements simply do not hav~ em-
• the old empiricist Peirce, that the very meaning of a statement
consists in the difference its truth would make to possible ex-
Ipirical meanings, by Peirce's s~andard; but a sufficwntly mclu-
lEve portion of theory does.! If we can asp"e t.o a ~OIt of l
perience. Might we not formulate, in a chapter-length sentence logischer Aufbau del' Welt at all, it must be to one m whlCh the
in observational language, all the difference that the truth of a texts slated for translation into observational and logico-mathe-
given statement might make to experience, and might we not matical terms are mostly broad theories taken as wholes, rather
veii
then take all this as the translation? Even if the difference that than just terms or short sentenc~ The translation of a :heoq ~I
the trutI, of the statement would make to experience ramifies would be a ponderous axiomatization of all the expenentlal
indefinitely, we might still hope to embrace it all in the logical difference that the truth of the theory would make. It would
implications of our chapter-length formulation, just as we can be a queer translation, for it would translate the whole but
II axiomatize an infinity of theorems. In giving up hope of such \
i ~one of the parts. We might better speak i~ such a case no: of
translation, then, the empiricist is conceding that the empiric::!.1 Lftranslation but simply of observational eVldence for theones;

(If!);> v'

~.
 80 Epistemology Naturalized Epistemology Naturalized 81
and we may, following Peirce, still fairly call this the empirical single sentences but only as larger blocks of theory, then the
meaning of the theories. indeterminacy of h'anslation of theoretical sentences is the
These considerations raise a philosophical question even f~'atural conclusion. And most sentences, apart from observa- It
about ordinary unphilosophical translation, such as from Eng- ~on sentences, are theoretical. This conclusion, conversely~~.,
lish into Arunta or Chinese. For, if the English sentences of a once it is embraced, seals the fate of any general notion of
theory have their meaning only together as a body, then we propositional meaning or, for that matter, state of affairs.
can justify their translation into Arunta only together as a Should the unwelcomeness of the conclusion persuade us to
body. There will be no justification for pairing off the com- abandon the verification theory of meaning? Certainly not.
ponent English sentences with component Arunta sentences, The sort of meaning that is basic to translation, and to the
except as these correlations make the translation of the theory learning of one's own language, is necessarily empirical mean-
" as a whole come out right. Any translations of the English sen- ing and nothing more. A child learns his first words and sen-
ences into Arunta sentences will be as correct as any Gther, so tences by hearing and using them in the presence of appropri-
I.o....n g as the net empirica~ implica~io.ns of the theory as a whole ate stimuli. These must be external stimuli, for they must act
lru
,. a~e preserved III translatIOn. But It IS to be expected that many
dIfferent ways of translating the component sentences, essen~
both on the child and on the speaker from whom he is learn-
ing.' Language is socially inculcated and controlled; the in- 1
tially different individually, would deliver the same empirical culcation and control turn strictly on the keying of sentences to
implications for the theory as a whole; deviations in the trans- shared stimulation, Internal factors may vary ad libitum with-
lation of one component sentence could be compensated for in out prejudice to communication as long as the keying of lan-
the translation of another component sentence. Insofar, there guage to external stimuli is undisturbed, Surely one has no
can be no ground for saying which of two glaringly unlike choice but to be an empiricist so far as one's theory of linguistic
translations of individual sentences is right. 3 meaning is concerned.
For an uncritical mentalist, no such indeterminacy threatens. What I have said of infant learning applies equally to the
Every term and every sentence is a label attached to an idea linguist's learning of a new language in the field. If the linguist
simple or complex, which is stored in the mind, When on th~ does not lean on related languages for which there are previ-
other hand we take a verification theory of meaning seriously, ously accepted translation practices, then obViously he has no
the indeterminacy would appear to be inescapable, The data but the concomitances of native utterance and observable
Vienna Circle espoused a verification theory of meaning but stimulus situation.'JNo wonder there is indeterminacy of trans- J
e
t Ifdid not take it seriously enough. If we recognize with Peirce
I hat the me~ning of a s,entence turns purely on what would
/...£ount as eVIdence for Its truth, and if we recognize with
lation-for of course only a small fraction of our utterances re-
port concurrent external stimulati0;;:Franted, the linguist will •
end up with unequivocal translations of everything; but only
Duhem that theoretical sentences have their evidence not as by making many arbitrary chOices-arbitrary even though un-
3 See above, p. 2 if. 4 See above, p. 28.
 82 Epistemology Naturalized Epistemology Naturalized 83
conscious-along the way. Arbitrary? By this I mean that experimentally controlled input-certain patterns of ..!.;:!>1dia-
different choices could still have made everything come out tion in assorted frequencies, for instance-and in the fullness
right that is susceptible in principle to any kind of check. "--;;£time tile subject delivers as output a description of the three-
. Let me Iil}.k up, in a different order, some of the pOints I dimensional external world and its history. The relation be-
• : have made')The crucial consideration behind my argument for tween the meager input and the torrential output is a relation
the indeterminacy of translation was that a statement about that we are prompted to study for somewhat the same reasons
the world does not always or usually have a separable fund of that always prompted epistemology; namely, in order to see
empirical consequences that it can call its own. That consider- how evidence relates to theory, and in what ways one's theory
ation served also to account for the impossibility of an episte- of nature transcends any available evidence.
mological reduction of the sort where every sentence is Such a study could still include, even, something like the old
equated to a sentence in observational and logico-mathemati- rational reconstruction, to whatever degree such reconstruction
cal ter~ And the impossibility of that sort of epistemological is practicable; for imaginative constructions can afford hints of
reduction disSipated the last advantage that rational recon- actual psychological processes, in much the way that mechani-
struelion seemed to have over psychology. cal simulations can. But a conspicuous difference between old
Philosophers have rightly despaired of translating everything epistemology mid the epistemological enterprise in this new
into observational and logico-mathematical terms. They have psychological setting is that we can now make free use of em-
despaired of this even when they have not recognized, as the pirical psychology.
reason for this irreducibility, that the statements largely do not The old epistemology aspired to contain, in a sense, natural
have their private bundles of empirical consequences. And science; it would construct it somehow from sense data. Epis-
some philosophers have seen in this irreducibility the bank- temology in its new selting, conversely, is contained in natural
ruptcy of epistemology. Carnap and the other logical positiv- science, as a chapter of psychology. But the old containment
ists of the Vienna Circle had already pressed the term "meta- remains valid too, in its way. We are studying how the human v
physics" into pejorative use, as connoting meaninglessness; and subject of our study posits bodies and projects his physics from
the term "epistemology" was next. Wittgenstein and his fol- his data, and we appreciate that our position in the world is
lowers, mainly at Oxford, found a residual philosophical voca- just like his. Our very epistemological enterprise, therefore,
tion in therapy: in curing philosophers of the delusion that and the psychology wherein it is a component chapter, and the
there were epistemological problems. whole of natural science wherein psychology is a component
But I think that at this paint it may be more useful to say book-all this is our own construction or projection from stim-
rather that epistemology still goes on, though in a new setting ulations like those we were meting out to our epistemological
and a clarified status. Epistemology, or something like it, sim- subject. There is thus reciprocal containment, though contain- J
ply falls into place as a chapter of psychology and hence of ment in different senses: epistemology in natural science and
natural science. It studies a natural phenomenon, viz., a physi- natural science in epistemology.
cal human subject. This human subject is accorded a certain TIlis interplay is reminiscent again of the old threat of cirou-
 84 Epistemology Naturalized Epistemology Naturalized 85
larity, but it is all right now that we have stopped dreaming of what? Are Gestalten prior to sensory atoms because they are
deducing science from sense data. We are after an understand- noticed, or should we favor sensory atoms on some more subtle

·
ing of science as an institution or process in the world and we ground? Now that we are permitted to appeal to physical stim-
do not intend that understanding to be any better tha~ the sci- ulation, the problem dissolves; A is epistemologically prior to B
ence which is its object. This attitude is indeed one that Neu- A is causally nearer than B to the sensory receptors. Or, what
is in some ways better, just talk explicitly in terms of causal

G
rath was already urging in Vienna Circle days, with his para-
ble of the marmer who has to rebuild his boat while staying proximity to seusory receptors aud drop the talk of epistemo-
afloat in it.' .. gical priority.
. One ~ffect of seeing epistemology in a psychological setting Around 1932 there was dehate in the Vienna Circle ,over
IS .th~t It resolv:s a stubborn old enigma of epistemological what to count as observation sentences, or Pro'tok~llslltze,5
pnontr·IOur retInas are irradiated in two dimensions, yet we One positiou was that they had the form of reports of sense
see thmgs as three-dimensional without conscious inference impressions. Another was that they were statements of an ele-
VVhich is to count as observation-the unconscious tw~-dimen~ mentary sort about the external world, e.g., "A red cube is
s~onaI reception or the conscious three-dimensional apprehen- standing on the table." Auother, Neurath's, was that they had
sIOn? In the old epistemological context the conscious form had the form of reports of relations between percipients and exter-
priority, for we were out to justify our knowledge of the exter- ual things: "Otto now sees a red cube on the table." The worst
nal world by rational reconstruction, and that demands aware- of it was that there seemed to be no objective way of settling
ness. Awareness ceased to be demanded when we gave up the matter: no way of making real sense of the question.
trying to justify our Imowledge of the external world by ra- [Let us now try to view the matter unreservedly in the con-
tIonal reconstruction. What to count as observation now can be text of the external world. Vaguely speaking, what we want of
'\ sc:tled in terms of the stimulation of sensory receptors, let con- observation sentences is that they be the ones in closest causal v
SCIOusness fall where it may. proximity to tl1e sensory receptm::! But how is such proximity
The Gestalt psychologists' challenge to sensory atomism, to be gauged? The idea may be rephrased this way: observa-
v:
~
hICl: seemed so relevant to epistemology forty years ago, is . on sentences are sentences which, as we learn language, are

hkewIse deactivated. Regardless of whether sensory atoms or most strongly conditioned .1<'L"QEcur~_~E!:_yensory stimulation
Gestalten are what favor the forefront of our consciousness it rather than to stored collateral information:TnilSTet"u-Si;;;ag::-'
ine a sentenan:111e-ried f~?our-veraicl'as"to whether it is true or /)l~',:\T ,,\~, ""~'
is simply the stimulations of our sensory receptors that are b~st
-false; queried for our assent or dissent. Then the sentence is
I
i. .. :.'
C""",,,
looked upon as the input to our cognitive mechanism. Old
paradoxes about unconscious data and inference, old problems an observation seutence if our verdict depeuds only on the " _iv. c.,;;"
about chaius of inference that would have to he completed too L~eusory stimulation present at the time. )/i" ..

quickly-these no longer matter. Buta verdict cannot depend on present stimulation to the
I~ the o.ld .auti-psychologistic days the question of epistemo- exclusion of stored information. The very fact of our having
logICal pnonty was moot. What is epistemologically prior to 5 Carnap and Neurath in Erkenntnis 3 (1932),204-228.
 86 Epistemology Naturalized Epistemology Naturalized 87
learned the language evinces much storing of information, and speakers of the language give the same verdict when given the
of information without which we should be in no position to same concurrent stimulation. To put the point negatively, an
give verdicts on sentences however observational. Evidently observation sentence is one that is not sensitive to differences
then we must relax our definition of observation sentence to in past experience within the speech community.
read thus fa sentence is an observation sentence if all verdicts {This formulation accords perfectly with the traditional role
VV'''' on it depend on present sensory stimulation and on no stored of the observation sentence as the court of appeal of scientific
information beyond what goes into understanding the sen- theories. For by our dennition the observation sentences are
tence·1 the sentences on which all members of the community will
Thls formnlation raises another problem: how are we to dis- agree under uniform stimulation. And what is the criterion of
tinguish between information that goes into understanding a membership in the same community? Simply general fiuency
sentence and information that goes beyond? This is the prob- of dialogue. This criterion admits of degrees, and indeed we
lem of distinguishing between analytic truth, which issues may usefully take the community more narrowly for some
from the mere meanings of words, and synthetic truth, which studies than for others. What count as observation sentences
depends on more than meanings. Now I have long maintained for a commuuity of specialists would not always.so count for a
that this distinction is illusory. There is one step toward such a larger community. (
distinction, however, which does make sense: a sentence that is There is generally no subjectivity in the phrasing of observa-
true by mere meanings of words should be expected, at least if tion sentences, as we are now conceiving them; they will usu-
it is simple,_to be subscribed to by all Huent speakers in the ally be about bodies. Since the distinguishing trait of an obser-
• communityJ Perhaps the controversial notion of analyticity can vation sentence is intersubjective agreement under agreeing
be dispensed with, in our definition of observation sentence, in stimulation, a corporeal subject matter is likelier than not.
./
favor of this straightforward attribute of community-wide ac- IThe old tendency to associate observation sentences with a
ceptanc~ subjective sensory subject matter is rather an irony when we
This attribute is of course no explication of analyticity. The reflect that observation sentences are also meant to be the
community would agree that there have been black dogs, yet intersubjective tribunal of scientific hypothese.:J The old ten-
none who talk of analyticity would call this analytic. My rejec- dency was due to the drive to base science on something firmer
tion of the analyticity notion just means drawing no line be- and prior in the subject's experience; but we dropped that
tween what goes into the mere understanding of the sente';ces project.
of a language and what else the community sees eye-to-eye on. The dislodging of epistemology from its old status of first
I doubt that an objective distinction can be made between philosophy loosed a wave, we saw, of epistemological nihilism.
meaning and such collateral information as is community- This mood is reHected somewhat in the tendency of Pollmyi,
wide. Kuhn, and the late Russell Hanson to belittle the role of evi-
Turning back then to our task of defining observation sen- dence and to accentuate cultural relativism. Hanson ventured
tences, we get this: an observation sentence is one on which all even to discredit the idea of observation, arguing that so-called
 88 Epistemology Naturalized . Epistemology N ahlralized 89
observations vary from observer to observer with the amount meaning is fundan1ental too, since observation sentences are
of lmowledge that the observers bring with them. The veteran the ones we are in a position to learn to understand first, both I
physicist looks at some apparatus and sees an x-ray tube. The as children and as field linguists. For observation sentences are
neophyte, looking at the same place, observes rather "a glass precisely the ones that we can correlate with observable cir-
and metal instrument replete with wires, reRectors, screws, cumstances of the occasion of utterance or assent, indepen-
lamps, and pushbuttons." 6 One man's observation is another dently of variations in the past histories of individual infor-
man's closed book or Hight of fancy. The notion of observation mants. They afIord the only entry to a language.
as the impartial and objective source of evidence for science is ene observation sentence is the cornerstone of semantics.
bankrupt. Now my answer to the x-ray example was already F'~r it is, as we just saw, fundamental to the learning of mean-
hinted a little while back: what counts as an observation sen- ing. Also, it is where meaning is firmest. Sentences higher up in
tence varies with the width of community considered. But we theories have no empirical consequences they can call their
can also always get an absolute standard by talong in all own; they confront the tribunal of sensory evidence only io
speakers of the language, or most.' It is ironical that philos- more or less inclusive aggregat~~JThe observation sentence,
ophers, finding the old epistemology untenable as a whole, situated at the sensory periphery of the body scientific, is the
should react by repudiating a part which has only now moved minimal. verifiable aggregate; it has an empirical content all its
into clear foeus. own and wears it on its sleeve.
/clarification of the notion of observation sentence is a good The predicament of the indeterminacy of translation has
thing, for the notion is fundamental in two connections. These little bearing on observation sentences. The equating of an ob-
two correspond to the duality that I remarked npon early in servation sentence of our language to an observation sentence
v this lecture: the duality between concept and doctrine, be- of another language is mostly a matter of empirical generali-

j. tween knowing what a sentence means and knowing whether

it is tr~ The observation sentence is basic to both enterprises.
Its relation to doctrine, to our knowledge of what is true, is
zation; it is a matter of identity between the range of stimula-
tions that would prompt assent to the one sentence and the
range of stimulations that would prompt assent to the other.'
........-.• very much the traditional one: observation sentences are the It is no shock to the preconceptions of old Vienna to say that
! rerository of evidence for scientific hypotheses. Its relation to epistemology now becomes semantics. For epistemology re-
mains centered as always on evidence, and meaning remains
6 N. R. Hanson, "Observation and interpretation," in S. Morgenpesser,
ed., Philosophy of Science Today (New York: Basic Books, 1966). centered as always on verification; and evidence is verification.
7 This qualification allows for occasional deviants such as the insane What is likelier to shock preconceptions is that meaning, once
or the blind. Alternatively, such cases might be excluded by adjusting we get beyond observation sentences, ceases in general to have"
the level of fluency of dialogue whereby we define sameness of lan-
guage. (For prompting this. note and influencing the development of
any clear applicability to single sentences; also that epistemol-
this paper also in more substantial ways I am indebted to Burton
Dreben.) 8 CI. Quine, Word and Object, pp. 31-46, 68.
 90 Epistemology Naturalized
ogy merges with psychology, as well as. with lioguistics.
This rubbing out of boundaries could contribute to progress, 4
it seems to me, in philosophically interesting'inquiries of a sci-
entific nature. One possible area is perceptual norms. Consider,
to begin with, the linguistic phenomenon of phonemes. We Existence
form the habit, in hearing the myriad variations of spoken
sounds, of treating each as an approximation to one or another and
of a limited number of norms-around thirty altogether- Quantification
constituting so to speak a spoken alphabet. All speech in our
language can be treated in practice as sequences of just those
thirty elements, thus rectifyiog small deviations. Now outside
the realm of language also there is probably only a rather lim-
ited alphabet of perceptual norms altogether, toward which
The question whether there
we tend unconsciously to rectify all perceptions. These, if ex-
are numbers, or qualities, or classes, is a metaphysical ques-
perimentally identified, could be taken as epistemological
tion, such as the logical positivists have regarded as meani~g­
buildiog blocks, the working elements of experience. They
less. On the other hand the question whether there are rabbIts,
might prove in part to be culturally variable, as phonemes are,
or unicorns, is as meaningful as can be. A conspiCUOUS differ-
and in part universal.
ence is that bodies can be perceived. Still, this is not all that
Again there is the area that the psychologist Donald T.
matters; for we can evidently say also, meaningfully and with-
Campbell calls evolutionary epistemology. 9 In this area there
out metaphysics, that there are prime numbers between 10 and
is work by Hiiseyin Yilmaz, who shows how some structural
20.
traits of color perception could have been predicted from sur-
What typifies the metaphysical cases is rather, according to
vival value.1O An~ a more emphatically epistemological topic
an early doctrine of Carnap's,' the use of category words, or
that evolution helps to clarify is induction, now that we are
Allworter. It is meaningful to ask whether there are prime
allowing epistemology the resources of natural science."
numbers between 10 and 20, but meaningless to ask in general
j} D. T. Campbell, «Methodological suggestions from a comparative
whether there are numbers; and likewise it is meaningful to
psychology of knowledge processes," Inquiry 2 (1959), 152-182. ask whether there are rabbits, or unicorns, but meaningless to
10 Hiiseyin Yilmaz, "On color vision and a new approach to general ask in general whether there are bodies. .
perception," in E. E. Bernard and M. R. Kare, eds., Biological Proto- But this ruling is unsatisfactory in two ways. The first dIffi-
types and Synthetic Systems (New York: Plenum, 1962); "Perceptual
invariance and the psychophysical law," Perception and Psychophysics culty is that there is no evident standard of what to count ~s a
2 (1967), 533-538. category, or category word. Typically, in terms of formalIzed
11 See "Natural Kinds," Chapter 5 in this volume.
I-Carnap. Logical Syntax of Language, p. 292.