Professional Documents
Culture Documents
Quine (1992), Pursuit of Truth
Quine (1992), Pursuit of Truth
92-5606
CIP
Preface to the Revised Edition
take off of John Austin 's takeoff of J ane Austen. The four
lectures appeared as a little b ook in Italian, La scienza e i dati
viii P REF A C E T O F IRS T E D I T I ON
W.V .Q.
CONTENTS
I . EV I D EN C E I
II . REF ER EN C E 23
9. Bodies 2)
10. V.lues of variables 25
11. Utility of reification 29
12. Indifference ofontology )1
I). Ontology defosed ))
I I I . ME A N I N G 37
'4. Thefold linguist's entering wedge )7
15. Stimulation again 40
16. To each his own 42
17. Translation resumed 44
18. Indeterminacy of translation 47
19· Syntax 49
x C O NTE NTS
I V. I N TE N S I O N 61
V. TRUTH 77
32. Vehicles of tru th 77
33. Truth as disquotation 79
34. Paradox 82
35. Tarski's construction 84
36. Paradox skirted 86
37. Interlocked hierarchies 88
38. Excluded middle 90
39. Truth versus warranted belief 93
40. Truth in mathematics 94
41. Equivalent theories 95
42. Irresoluble rivalry 98
43. Two indeterminacies 101
References 105
Credits 109
Index 111
PURSUIT OF TRUTH
O'W'�l.lI Tel cfHxI.1I6�1IQ.
PLATO
EVIDE NCE
2. Observation sentences
J. Theory-laden?
My definition of observation sentence is of my devising,
but the term is not. Philosophers have long treated in their
several ways of what they called observation terms or ob
servation sentences. But it has now become fashionable to
question the notions, and to claim that the purportedly ob
servable is theory laden in varying degrees. It is pointed out
-
any of them, and more recondite reports are obse rvatio nal
enough for some. I agree that the practical notion of obser
vation is thus relative to one or another limited community,
rather than to the whole speech community. An observa
tion sentence for a community is an occasion sentence on
which members of the community can agree outright on
witnessing the occasion.
For philosophical purposes we can probe deeper how,
ever, and reach a single standard for the whole speech com
munity. Observable in this sense is whatever would be at
tested to on the spot by any witness in command of the
language and his five senses. If scientists were perversely to
persist in demanding further evidence beyond what sufficed
for agreement, their observables would reduce for the most
part to those of the whole speech community. Just a few,
EV I D EN C E 7
such as the indescribable smell of some uncommon gas,
would resist reduction.
But w hat has all this to do with a sentence's being theory
la den or theory-free? My definition distinguishes obser
vation sentences from others, whether relative to special
communities or to the general one, without reference to
theory-freedom. There is a sense, as we shall now see, in
which they are all theory-laden, even the most primitive
ones, and there is a sense in which none are, even the most
p rofessional ones.
Think first of primitive ones, the entering wedge in lan
guage learning. They are associated as wholes to appropri
ate ranges of stimulation, by conditioning. Component
words are there merely as component syllables, theory
free. But these words recur in theoretical contexts in the
fullness of time. It is precisely this s haring of words, by
observation sentences and theoretical sentences, that pro
vides logical connections between the two kinds of sen
tences and makes observation relevant to scientific theory.
Retrospectively those once innocent observation sentences
are theory-laden indeed. An observation sentence contain
ing no word more technical than 'water' will join forces
with theoretical sentences containing terms as technical as
'H20' . Seen holop hrastically, as conditioned to stimulatory
situations, the sentence is theory-free; seen analytically,
word by word, it is theory-laden. Insofar as observation
sentences bear on science at all, affording evidence and tests,
there has to be this retrospective theory-lading along with
the pristine holophrastic freedom from theory. To impugn
their observationality thus retrospectively is to commit
what Firth (p. 100) called the fallacy of conceptual retrojec
tion.
8 P U R SU IT OF TRU TH
More sophisticated observation sentences, inc lud ing
those of specia li zed scientific communities , are sim ilarly
tw o-faced, even tho ugh learned by composition rather than
direct conditio ning. What qu alifies them as obse rvation
sentences is still their holophras tic associ ation with fIXed
ranges of sensory stimul ation, however that association be
acq uired. Holophrastical ly they function sti ll as theory
free, like C. I. Lewis's "expressive" sentences (p. 1 79),
tho ugh when taken retrospectively word by word the self
same senten ces are theory- laden , like his "ob jective" ones.
When epistemology rounded the linguistic tum, talk of
observable objects ga ve way to tal k of observation terms. It
was a g ood mo ve, but not good enough. Observation sen
ten ces w ere distinguished from theoret ical ones only
derivati vely, as containing observation terms to the exclu
sion of theory-laden or theoretical terms. Consequently
Rei chenbach and others felt a need for "bridg e p rincip les "
to re late the two kin ds of sentences. No bridge is wanted,
we now see, and bridging is the w rong figure. Start in g with
sentences as we have done rather than with terms, we see no
bar to a sharing of voc ab ulary by the two kinds of sentences ;
and it is the shared vocabulary that links them.
Starting with sentences has conferred the further boon of
freeing the definition of observation sentence from any de
pendence on the dist inction between the theory -free and the
theo ry-laden. Yet a third adv antage of this move is th at we
can then study the acquisition and use of observ ation sen
tences without p rejudging what objects, if any , the compo
nent word s are meant to refer to. We thus a re freed to
spec ulate on the nat ure of reification and its utility for
scientific theory -a topic for Chap ter II. Taking terms as
starting point wou ld h ave me ant finessing reification and
EVI D ENC E 9
con ceding objective reference out of hand, without consid
e ring what it is for or what goes into it.
4. Observation categoricals
The support of a theory by observation stands forth most
expl icitly in experiment, so let us look i nto that. The sci en
t ist has a backlog of accepted theory, and is considering a
hypothesis for possible in corporation into it. The theo ry
tell s him that if the hypothesis under consideration is t rue,
then, whenev er a certain obser vable situation is set up, a
cert ain effect should be observed. So he set s up the situation
in qu estion. If the p redi ct ed effect fail s to appear, he aban
don s hi s hypothesis. If the effect doe s appear, hi s hypothesis
may b e t rue and so can be tentatively added to hi s ba cklog of
theory.
Th us suppose a team of fi eld mineralogist s have tu rned
up an unfam iliar crystalline mineral of a distinc tively p in k
ish cast. They speak of it provisionally as litholite, for want
of a better name. One of them conjectures its ch emical
co mposition. This is the hypothesis, of whi ch I shall spare
m yself the detail s. From his ba cklog of ch emi cal lore he
reason s t hat if this chemical hypoth esis is true , then any
pie ce oflitholite should em it h ydrogen sulfide when he ated
above 180°C. Th es e last p rovi sion s a re the observables ; for
o ur min eralogist and hi s colleagues know litholite when
they see it and hydrogen sul fide when they smell it, and
t hey can read a the rmometer.
The t est of a hypothesis thu s hin ges on a lo gical rel ation
of i mpli cation. On one side, the theoretical, we h ave the
backl og of accepted theory plus the h ypothesis. This com
bination does the implying. On the other side, the observa-
10 P U R SU I T OF T R U TH
tional, we have an implied generality that the experimenter
can directly test, directly challenge-in this case by heating
some of the pink stuff and sniffing.
A generality that is compounded of observables in this
way-'Whenever this, that'-is what I call an observation
categorical. It is compounded of observation sentences. The
'Whenever' is not intended to reify times and quantify over
them. What is intended is an irreducible generality prior
to any objective reference. It is a generality to the effect
that the circumstances described in the one observation sen
tence are invariably accompanied by those described in the
other.2
Though compounded of two occasion sentences, the ob
servation categorical is itself a standing sentence, and hence
fair game for implication by scientific theory. It thus solves
the problem of linking theory logically to observation, as
well as epitomizing the experimental situation.
That situation is where a hypothesis is being tested by
an experiment. An opposite situation is equally familiar: a
chance observation may prompt us to conjecture a new
observation categorical, and we may invent a theoretical
hypothesis to explain it. For example, we might notice wil
lows leaning over a stream. This suggests the observation
categorical:
6. Holism
Let us recall that the hypothesis regarding the chemical
composition of litholite did not im p ly its observation
categorical single-handed. It implied it with the help of a
backlog of accepted scientific theory. In order to deduce an
observation categorical from a given hy pothesis, we may
have to enlist the aid of other theoretical sentences and of
many common-sense platitudes that go without saying,
and perhaps the aid even of arithmetic and other parts of
mathem atics.
In that situation, the falsity of the observation categorical
does not conclusively refute the hy pothesis. What it refutes
is the conjunction of sentences that was needed to imply the
observation categorical. In order to retract that conjunction
14 P URS U IT OF TRU T H
we do not ha ve to retract the hypothesis in question; we
could retract some other sentence of the conj unction in
stead. This is the important insight called holism. Pierre
Duhem made muc h of it early in this century , but not too
m uch.
The scientist thinks of his experiment as a test spe cifically
of his new hypothesis , but o nly because t his was the sen
tence he was wonde ring about and is prepared to re ject.
Moreo ver, there are also the situations where he has no
preconcei ved hypothesis, but just happens upon an anoma
lo us phenomenon . It is a case of his happening upon a
co unter-instance of an observation categorical which, ac
cording to his c urrent theory as a whole, ought to have been
true. So he loo ks to his theory with a c ritical eye.
Over-Iogici zing, we may pict ure the accommodation of
a failed observation catego rical as follows. We ha ve before
us some set S of pu rported truths that was fo und jointly to
imply the false categori cal. Implication may be taken here
simply as deducibility by the logic of t ruth functio ns,
quantification, and identity. (We can always pro vide for
more substan tial conseq uences by incorporating appropri
ate premisses explicitly into S.) Now some one or more of
the sentences in S are going to have to be rescinded. We ex
empt some membe rs of S from t his thre at on determining
that the fateful implication still holds witho ut their help .
Any p urely logical truth is thus exempted, since it adds
nothing to what S wo uld logically imply anyway ; and sun
d ry irrele vant sentences in S will be exempted as well . Of
the remaining mem bers of S, we rescind one that seems
most suspect, or least c rucial to our o verall theory. We heed
a maxim of mi nimum m utilation. If the remaining mem
bers of S s till conspire to imply the fa lse categorical, we try
EVI DENCE 15
7. Empirical content
Stimu lus meanings ha ve fu zzy boundaries, as witness ag ain
the black swan and the a lbino raven. If we i magine sha rp
dema rcation, howev er, we can bu ild up to a decepti ve ly
p recise but withal inst ructi ve definition of empirical con
tent.
C all an observation categorical a na lytic for a given
speake r if, as in 'Robins a re birds ', the affirmati ve stimulus
meaning fo r him of the one compon ent is inc luded in t hat
of the other. Otherwise synthetic. Call a sentence o r set of
sentences testable if it imp lies some synthetic observation
categorica ls . Call two observa tion categoricals synonymous
EVIDE N C E 17
REFERENCE
9. Bodies
There were advanta ges. we saw (§3) . in sta rtin g with obser
vation sentences rather than terms. One advan ta ge was that
the nature and utili t y of reification could be deferred for
consideration un ti l an e piste molog i cal se tti ng had b een
sketched in. We are now at that stag e .
while the free categ orical does not. However, the scenes
first associated with 'Raven' will show a raven at the salient
focus, and those first associated with 'Black' will show
P U R SU I T O F TR U TH
black at the salient focus. Insofar, the free categorical al
ready meets the focal demand. The difference between the
free and the focal in other cases, and between conjunction
and predication (§ 2.) , can gradually dawn on the child in its
own time.
By virtue of its narrowed focus, however, the focal ob
servation categorical-unlike the free one-has decidedly
the air of general discourse about bodies : willows in the one
example, ravens in the other. This is where I see bodies
materializing, ontologically speaking: as ideal nodes at the
foci of intersecting observation sentences. Here, I suggest,
is the root of reification.
For the very young child, who has not got beyond obser
vation sentences, the recurrent presentation of a body is
much on a par with similarities of stimulation that clearly
do not prompt reification. Recurrent confrontation of a ball
is on a par at first with mere recurrent exposure to sunshine
or cool air: the question whether it is the same old ball or
one like it makes no more sense than whether it is the same
old sunbeam, the same old breeze. Experience is in its
feature-placing stage, in Strawson's phrase. Individuation
comes only later.
True, an infant is observed to expect a steadily moving
object to reappear after it passes behind a screen; but this all
happens within the specious present, and reflects rather the
expectation of continuity of a present feature than the reifi
cation of an intermittently absent object. Again a dog's rec
ognition of a recurrent individual is beside the point; the
dog is responding to a distinctive odor or other trait, un
available in the case of qualitatively indistinguishable balls.
To us the question whether we are seeing the same old
ball or just a similar one is meaningful even in cases where it
REFERENCE 25
1J . Ontology de}Used
We found that two ontologies, if explicitly correlated one to
one, are empirically on a par; there is no empirical ground
for choosing the one rather than the other. What is empiri
cally significant in an ontology is just its contribution of
neutral nodes to the structure of the theory. We could rein
terpret 'Tabitha' as designating no longer the cat, but the
whole cosmos minus the cat; or, again, as designating the
eat's singleton, or unit class. Reinterpreting the rest of our
terms for bodies in corresponding fashion, we come out
with an ontology interchangeable with our familiar one . As
wholes they are empirically indistinguishable. Bodies still
continue, under each interpretation, to be distinct from
their cosmic complements and fro m their singletons; they
are distinguished in a relativistic way, by their roles relative
34 PURSUIT OF TRUTH
to one another and to the rest of the ontology. Hence my
watchword ontological relativity. But see further §20.
The importance of the distinction between term and ob
servation sentence shone forth in §§ 3 and 9 , and it does so
again here. 'There's a rabbit' remains keyed to the sensory
stimulations by which we learned it, even if we reinterpret
the term 'rabbit' as denoting cosmic complements or sin
gletons of rabbits. The term does continue to conjure up
visions appropriate to the observation sentence through
which the term was learned, and so be it; but there is no
empirical bar to the reinterpretation. The original sensory
associations were indispensable genetically in generating
the nodes by which we structure our theory of the world.
But all that matters by way of evidence for the theory is
the stimulatory basis of the observation sentences plus the
structure that the neutral nodes serve to implement. The
stimulation remains as rabbity as ever, but the corre
sponding node or obj ect goes neutral and is up for grabs.
Bodies were our primordial reifications, rooted in innate
perceptual similarities. It would be gratuitous to swap them
for proxies; the point was just that one could. But our on
tological preconceptions have a less tenacious grip on the
deliberate refinements of sophisticated science. Physicists
did first picture elementary particles and light waves in anal
ogy to familiar things, but they have gone on to sap the
analogies. The particles are less and less like bodies, and
the waves seem more like pulsations of energy in the void.
When we get to the positing of numbers and other abstract
objects, I have conj ectured in Roots oj ReJerence that we are
indebted to some fruitful confusions along the way. Lan
guage and science are rooted in what good scientific lan-
REFERENCE 35
guage eschews. In Wittgenstein's figure, we climb the lad
der and kick it away.
Some findings known as the Bose-Einstein and Fermi
Einstein statistics suggest how we might be led actually to
repudiate even the more traditional elementary particles as
values of variables, rather than retaining them and just ac
quiescing provisionally in their mysterious ways . Those
results seem to show that there is no difference even in
principle between saying of two elementary particles of a
given kind that they are in the respective places Q and b and
that they are oppositely placed, in b and Q. It would seem
then not merely that elementary particles are unlike bodies;
it would seem that there are no such denizens of space-time
at all, and that we should speak of places Q and b merely as
being in certain states, indeed the same state, rather than as
being occupied by two things.
Perhaps physicists will accommodate this quandary in
another way. But I p rize the example as illustrating the kind
of consideration that could prompt one to repudiate some
hypothetical objects. The consideration is not based on
positivistic misgivings over theoretical entities . It is based
on tensions internal to theory .
Theories can take yet more drastic turns: such not merely
as to threaten a c herished ontology of elementary particles,
but to threaten the very sense of the ontological question,
the question what there is . What I have been taking as the
standard idiom for existential purposes, namely quantifica
tion, can serve as standard only when embedded in the
standard form of regimented language that we have been
picturing : one whose further apparatus consists only of
truth functions and predicates . I f there is any deviation in
P U RSU I T OF TRUTH
ME ANIN G
J5 . Stimulation a gain
It would seem then that this matching of observation sen
tences hinges on s amenes s of stimulation of both parties,
the linguist and the informant . But an event of stimulation,
as I use the term (§ I ) , is the activation of some subset of the
subject's sensory receptors . Since the linguist and his infor
mant share no receptors, how can they be said to s ha re a
stimulation? We mi g h t say rather that they unde rg o similar
stimulation, but that would assume still an approximate
homology of nerve endings from one individual to another.
Surely such anatomi cal minutiae ought not to matter here.
I was expressing t h i s discomfort as early as 1 965 . 1 By
1 98 1 it prompted me to readjust my de fi n i tio n of observa
tion sentence. In my original definition I had appe aled to
sameness of s ti m ulus meaning between speakers . 2 but in
1 98 I I defined it rather for the sin gl e speaker, by the follow
ing condition:
If querying the s enten ce elicits assent from the given speaker
on one occasion. it will el icit assent likewise on any other
occasion when the same total set of receptors is triggered ;
and similarly for dissent. J
Then I accounted a
sentence observational for a whole com
munity when it was observational for each member. In this
I E . g . in a lecture " Propositional Obj ects, " published in Onto/i1gictJ/
Relll t ivity aM Othtr Essays .
1 7. Translation resumed
Our linguist then goes on tentatively identifying and trans
lating observation sentences. Some of them are perhaps
compounded of others of them. in ways hinting of our
MEANING 45
logical particles 'and', 'or', 'but', 'not'. By collating the
situations that command the natives' assent to the com
pounds with the situations that command assent to the
components, and similarly for dissent, the linguist gets a
plausible line on such connectives.
Unlike observation sentences, most utterances resist cor
relation with concurrent stimulations. Taking the initia
tive, the linguist may volunteer and query such a sentence
for assent or dissent in various situations, but no correlation
with concurrent stimulation is forthcoming. What next?
He can keep a record of these unconstrued sentences and
dissect them. Some of the segments will have occurred also
in the already construed observation sentences. He will
treat them as words, and try pairing them off with English
expressions in ways suggested by those observation sen
tences. Such are what I have called analytical hypotheses.
There is guesswork here, and more extravagrant guess
work to follow. The linguist will turn to the unconstrued,
nonobservational sentences in which these same words oc
curred, and he will project conjectural interpretations of
some of those sentences on the strength of these sporadic
fragments. He will accumulate a tentative Jungle vocabu
lary, with English translations. and a tentative apparatus of
grammatical constructions. Recursion then sets in, deter
mining tentative translations of a potential infinity of sen
tences. Our linguist keeps testing his system for its efficacy
in dealing with natives, and he goes on tinkering with it and
guessing again. The routine of query and assent that had
been his standby in construing observation sentences con
tinues to be invaluable at these higher and more conjectural
levels.
Clearly the task is formidable and the freedom for conjec-
PURSUIT OF TRUTH
ture is enormous. Linguists can usually avoid radical trans
lation by finding someone who can interpret the language,
however haltingly, into a somewhat familiar one. But it is
only radical translation that exposes the poverty of ultimate
data for the identification of meanings.
Let us consider, then, what constraints our radical trans
lator can bring to bear to help guide his conjectures. Con
tinuity is helpful: successive utterances may be expected to
have some bearing on one another. When several such have
been tentatively interpreted, moreover, their interrelation
itself may suggest the translation of a linking word that will
be helpful in spotting similar connections elsewhere.
The translator will depend early and late on psychologi
cal conjectures as to what the native is likely to believe. This
policy already governed his translations of observation sen
tences. It will continue to operate beyond the observational
level, deterring him from translating a native assertion into
too glaring a falsehood. He will favor translations that as
cribe beliefs to the native that stand to reason or are conso
nant with the native's observed way of life. But he will not
cultivate these values at the cost of unduly complicating the
structure to be ascribed to the native's grammar and se
mantics, for this again would be bad psychology; the lan
guage must have been simple enough for acquisition by the
natives, whose minds, failing evidence to the contrary, are
presumed to be pretty much like our own. Practical psy
chology is what sustains our radical translator all along the
way, and the method of his psychology is empathy: he
imagines himself in the native's situation as best he can.
Our radical translator would put his developing manual
of translation continually to use, and go on revising it in the
light of his successes and failures of communication. The
MEANING 47
successes consist-to repeat-in successful negotiation and
smooth conversation. Reactions of astonishment or bewil
derment on a native's part, or seemingly irrelevant re
sponses, tend to suggest that the manual has gone wrong.
We readily imagine the translator's ups and downs . Per
haps he has tentatively translated two native sentences into
English ones that are akin to each other in some semantic
way, and he finds this same kinship reflected in a native's
use of the two native sentences. This encourages him in his
pair of tentative translations. So he goes on blithely suppos
ing that he is communicating, only to be caught up short.
This may persuade him that his pair of translations was
wrong after all. He wonders how far back, in the smooth
flowing antecedent conversation, he got off the beam .
1 8. Indeterminacy of translation
Considerations of the sort we have been surveying are all
that the radical translator has to go on. This is not because
the meanings of sentences are elusive or inscrutable; it is
because there is nothing to them, beyond what these fum
bling procedures can come up with. Nor is there hope even
of codifying these procedures and then defining what counts
as translation by citing the procedures; for the procedures
involve weighing incommensurable values. How much
grotesqueness may we allow to the native's beliefs, for in
stance, in order to avoid how much grotesqueness in his
grammar or semantics?
These reflections leave us little reason to expect that two
radical translators, working independently on Jungle.
would come out with interchangeable manuals. Their
manuals might be indistinguishable in terms of any native
PURSUIT OF TRUTH
behavior that t he y give reason to expect, and yet each
manual m ig h t prescribe some translations that the other
translator would rej ect Such is t he thesis of indete r min acy
.
of translation.
A manual of Jungle-to-English translation constitutes a
recursive, or inductive, definition of a translation relation
toget her with a claim that it correlates sentences comp a tibl y
with the behavior of all concerned. The thesis of indetermi
nacy of translation is that these claims on the part of two
manuals m i g h t both be true and yet the two translation
relations might not be usable in alternation, from s entence
to sentence, without issuing in incoherent sequences . Or, to
put it another way, the English sentences prescribed as
translation o f a given Jun gle sentence by two rival manuals
m igh t not be interchangeable in English contexts.
The use of one or the other manual might i n deed cause
differences in speech afterward, as remarked by Robert
Kirk in connection with the idioms of propositional at
titude; but the two would do equal j ustice to the s tatu s quo.
I have directed my indeter m ina cy thesis on a radically
exotic lan guage for the sake of pla us ib il i t y but in prin ciple
,
19. Syntax
Readers have supposed that I extended my indeterminacy
thesis to syntax. This puzzled me until I became aware,
recently, of a subtle cause of the misconception. In Word and
Object (pp. 5 5 , 68- 7 2) I claimed that our distinctive ap
paratus of reification and reference is subject to indetermi
nacy of translation. The apparatus includes pronouns, = " •
20 . Indeterminacy of reference
The difference between taking a sentence holophrastically
as a seamless whole and taking it analyticall y term by term
proved crucial in earlier matters (§§3 , 9 , 1 3 ) . It is crucial also
to translation. Taken analytically, the indeterminacy of
translation is trivial and indisputable. It was factually illus
trated in Ontological Relativity (pp. 3 5-36) by the Japanese
classifiers, and more abstractly above by proxy functions
(§ 1 3). It is the unsurprising reflection that divergent inter
pretations of the words in a sentence can so offset one an
other as to sustain an identical translation of the sentence as a
whole. It is what I have called inscrutability of reference;
'indeterminacy of reference' would have been better. The
serious and controversial thesis of indeterminacy of transla
tion is not that; it is rather the holophrastic thesis, which is
stronger. It declares for divergences that remain unrecon
ciled even at the level of the whole sentence, and are com
pensated for only by divergences in the translations of other
whole sentences.
Unlike indeterminacy of reference, which is so readily
illustrated by mutually compensatory adjustments within
the limits of a single sentence, the full or holophrastic inde
terminacy of translation draws too broadly on a language to
admit of factual illustration. Radical translation is a rare
achievement, and it is not going to be undertaken success-
MEANING 51
fully twice for the same language. But see Levy for a plausi
ble artificial example, based on measurement in deviant
geometries. Also there is Massey's sweeping example based
on the duality of affirmation to negation, conjunction to
alternation, and universal quantification to existential. His
rival translations, the homophonic and its dual, conflict on
every sentence. A weakness of this construction is that the
dual manual depends on viewing the natives' volunteered
sentences as denied rather than affirmed-a gratuitous re
versal of the translator's conventional orientation. Still, in
view of Levy's construction if not Massey's, one can
scarcely question the holophrastic indeterminacy thesis.
A thick and imposing periodical on the philosophy of
language is published twice a year in the Canary Islands
under the title Gavagai. A book by David Premack, on his
language experiments with chimpanzees, came out lately
under the title Gavagai. Hubert Dreyfus has California van
ity plates on his Volkswagen Rabbit that spell 'GA V AGAr.
The word has become the logo of my thesis of indetermi
nacy of translation, and now it is making its way in a wider
world. Ironically, indeterminacy of translation in the
strong sense was not what I coined the word to illustrate. It
did not illustrate that, for 'Gavagai' is an observation sen
tence, firmly translatable holophrastically as ' (Lo, a) rab
bit'. But this translation is insufficient to fix the reference of
'gavagai' as a term; that was the point ofthe example. It is an
extreme example of the indetenninacy of reference, the
contained term being the whole of the sentence. No room
is left here for compensatory adjustments, and none are
needed.
Kindly readers have sought a technical distinction be
tween my phrases 'inscrutability of reference' and 'ontolog
ical relativity' that was never clear in my own mind. But I
S2 PURSUIT OF TRUTH
can now say what o n tolo g ica l r el ati v i t y
is relative to, m ore
s uc cin ctl y than I did i n the lectures, paper, and book of that
title. It is relative to a manual of t ran s la tion . To say that
' gavagai ' denotes rabbits is to opt for a ma n u a l of translation
in which 'ga v ag ai ' is translated as ' rabbit' , instead of opting
for any of the alternative m anu als .
And does the inde t erm ina cy or relativity extend also
somehow to the home language? In " On tolo g ica l Relativ
ity" I said it did, for the home language can be translated
into itself by permutations that depart materially from the
mere identity transformation, as proxy functions bear o ut .
But if we choose as our manual of translation the identity
transformation, thus taking the home language at face
value, the relativity is resolved . Reference is then ex pli ca ted
in disquotational paradigms anal o g ou s to Tarski's truth
paradigm (§3 3); thus 'rabbit' denotes rabbits, w ha tever they
are, and 'Boston' designates Boston.
21 . Whither meanings?
If we could contrive an acceptable relation of s ameness of
meaning, it would be a sho rt step to an acceptable definition
of me anings . For, as mo re than one philosopher has noted,
we could define the meaning of an expression as the class of
all expressions like it in m eaning . Conversely, if we had the
meanings to begin with, they and identity would provide
sameness of meaning, there being no entity without iden
tity. 8 In short, meanings and sameness of meaning present
one and the s a m e pro ble m.
8 This platitude has latdy been obscured b y a confusion over the axiom
of extensionality . which individuates sets . or classes. and has been sus
pended by some set theorists in an exploratory spirit. Might we not
likewise rec o gniz e meanings without identity ? No. Dropping extension-
MEANING 53
Translation does enjoy reasonable determinacy up
through observation categoricals and into the logical con
nectives. Thus one could make a stab at the interlinguistic
equating of empirical content (§7) . even in radical transla
tion. But empirical content pertains only to testable sen
tences and sets of sentences. We are still left with no general
concept of the meanings of sentences of less than critical
semantic mass.
It is not a conclusion that one readily j umps to or rests
with. One is tempted to suppose that we might define
meanings for sentences of less than critical mass, and even
for terms, by substitutivity. If we can interchange two ex
pressions without disturbing the empirical content of any
testable context. are they not alike in meaning? Well, the
plan collapses between languages. Interchanging expres
sions would tum the context into nonsense if the expres
sions belong to different languages. So the plan offers no
relief from the indeterminacy of translation.
22 . Domestic meaning
Lowering our sights. then, and giving up on "proposi
tions" as language-transcendent sentence meanings, we
might still look to the substitutivity expedient for a strictly
domestic, interlinguistically inapplicable notion of same
ness of meaning. Sentences are cognitively equivalent, we
might say, if putting one for the other does not affect the
empirical content of any set of sentences. This sounds right
in principle. For the most part it resists decisive application,
allty does not exempt sets from identity either. It only tables the question
of sufficient conditions for their identity . The notation 'x = y' stays on,
with sets as values of the variables. There is still no entity-no set,
nothing- without identity.
54 PURSUIT OF TRUTH
however, because of the rather visionary status of empirical
content (§7) .
Another approach looks to belief: to the speaker's assent
or dissent when the sentences are queried. This would not
do for standing sentences; it would equate all his beliefs.
But it works for occasion sentences, where we can check
each pair for concomitance over varying occasions. Two
occasion sentences may be accounted cognitively equiva
lent for a given speaker ifhe is disposed on every occasion to
assent to both or dissent from both or abstain. Derivatively,
then, they may be accounted cognitively equivalent for the
community if cognitively equivalent for each member.
When in particular the sentences are observation sentences,
we are back to stimulus synonymy (§ I 6) .
Cognitive equivalence s o defined then extends immedi
ately to terms, or predicates. They are cognitively equiva
lent, or we might now say cognitively synonymous, if their
predications-'It's an F', 'It's a G'-are cognitively equiva
lent. In view of our definition of cognitive equivalence of
occasion sentences, this boils down to saying that terms are
cognitively synonymous for a speaker if he believes them
to be coextensive, that is, true of the same things; and syn
onymous for the community if synonymous for each
member.
Some slight progress can then be made toward cognitive
equivalence of standing sentences. Certainly they should be
rated cognitively equivalent if one can be got from the other
by supplanting a component term by a cognitive synonym.
But this does not cover all the pairs of standing sentences
that we would want to regard as cognitively equivalent.
There is a third approach in a,udyticity . Once we have
analyticity, cognitive equivalence is forthcoming; for two
MEANING 55
23 . Lexicography
To question the old notion of meanings of words and sen
tences is not to repudiate semantics. Much good work has
been done regarding the manner, circumstances, and devel
opment of the use of words. Lexicography is its conspicu
ous manifestation. But I would not seek a scientific rehabili
tation of something like the old notion of separate and
distinct meanings; that notion is better seen as a stumbling
block cleared away. In later years indeed it has been more of
a stumbling block for philosophers than for scientific lin
guists, who, understandably, have simply found it not
technically useful.
Dictionaries are reputedly occupied with explaining the
meanings of words, and the work is neither myth-bound
nor capricious. How does it proceed? I hold that it is not
directed at cognitive equivalence of sentences, nor at syn
onymy of terms, and that it presupposes no notion of mean
ing at all. Let us consider then what the business of dic
tionaries really is.
Sometimes the dictionary explains a word by supplying
another expression that can replace it salva veritate at least in
positions uncontaminated by quotation or idioms of prop
ositional attitude. Sometimes, instead, a selection of in
formation is set down regarding the object or objects to
which the word refers. There is no pretense here of a dis
tinction between essential and accidental traits. It is a matter
MEANING 57
purely of pedagogy: the lexicographer wants to improve his
reader's chances of successful communication as best he can
in a small compass. Often, moreover, a dictionary entry
neither paraphrases the word nor describes its objects, but
describes, rather, the use of the word in sentences. This is
usually the way with grammatical particles, and it is often
the way also with terms. It is bound to be the way with a
term that neither refers to concrete objects nor admits of a
separable, self-contained paraphrase.
Behind this seeming disorder there is a unifying princi
ple. The goal may be seen always as the sentence. The
lexicographer is out to help his reader profit by the sen
tences that he sees or hears, and to help him react to them in
expected ways, and to help him emit sentences usefully.
But sentences are unlimited in their variety, so the lexicog
rapher organizes his teaching of sentences word by word,
teaching how to use each word in making sentences. One
way of teaching this, which is convenient when available, is
by citing a substitute expression; for the lexicographer thus
exploits the reader's presumed knowledge of how to use
that substitute expression in making sentences. And the
other sorts of dictionary entry likewise aim, in their differ
ent ways, at the same end: teaching the use of sentences.
When from semantics as pursued by philosophers we
move to lexicography, we shift our focus from likeness of
meaning to knowledge of meaning, so to speak; from syn
onymy of expressions, anyway, to the understanding of ex
pressions. The lexicographer's job is to inculcate under
standing of expressions, that is, to teach how to use them.
He can be wholly successful in teaching the use of sentences
without considering in what sense they might be said to be
58 P U R S U I T OF T R U T H
equivalent. Nothing, apparently, could be more remote
than meanings from the lexicographer's concern. Why
should they be less remote from ours?
So we might try looking to the understanding of expres
sions, rather than to synonymy, as the operationally basic
notion of semantics. What sense can we make of it? In prac
tice we credit someone with understanding a sentence if we
are not surprised by the circumstances of his uttering it or
by his reaction to hearing it-provided further that his reac
tion is not one of visible bewilderment. We suspect that he
does not understand it if the event is drastically at variance
with those conditions. Still no boundary is evident, no gen
eral criterion for deciding whether he actually misunder
stands the sentence or merely holds some unusual theory
regarding its subject matter.
We can be more confident in imputing misunderstanding
of a word than of a sentence, for we can then observe some
one's use of numerous sentences, or his response to them,
aU of which contain the word. We can control our experi
ment, choosing and querying sentences ourselves. We may
find that he responds otherwise than would generally be
expected when the sentences contain the word in question,
and that he responds in the more usual ways to many sen
tences that lack that word but are much the same in other
respects.
In this matter of understanding language there is thus a
subtle interplay between word and sentence. In one way the
sentence is fundamental: understanding a word consists in
knowing how to use it in sentences and how to react to such
sentences. Yet if we would test someone's understanding of
a sentence, we do best to focus on a word, ringing changes
on its sentential contexts . Once we have thus satisfied our-
MEANING 59
INTENSION
25 . Perception extended
Observationality varies with the group of speakers con
cerned, and is also, within the group, somewhat a matter of
degree (§2) . Consequently the construction 'perceives that
'
p continues to ftourish when the content clause is not obser
vational, or not very. We even hear 'Tom perceives that the
train is late'.
Consider how one would get on to using that sentence.
People have ways of showing that they perceive that the
train is late, and these ways run to type. One way is by
saying that the train is late. Also people pace impatiently,
they look at the clock, they look along the track. Along
with acquiring such habits ourselves, we have learned to
observe similar manifestations on the part of others. We are
ready to see our own ways replicated in another person.
This readiness was what enabled us to teach observation
sentences to other persons, and to learn when to affirm ' x
'
perceives that p in observational cases; and the ability ex
tends beyond observation sentences to sentences like 'The
train is late' .
The evidence is not assembled deliberately. One em
pathizes, projecting oneself into Tom's situation and Tom's
behavior pattern, and finds thereby that the sentence 'The
train is late' is what comes naturally. Such is the somewhat
haphazard basis for saying that Tom perceives that the train
is late. The basis becomes more conclusive if the observed
behavior on Tom's part includes a statement of his own that
the train is late.
PURSUIT OF TRUTH
An occasion sentence of the form 'x perceives that p ' can
be true even when the content clause is a standing sentence,
such as 'Randy is a dog', rather than an occasion sentence.
What is then required, however, is not only that the perci
pient be prepared on that occasion to assent to that clause,
but also that he be just then becoming aware that it is true.
Ascriptions of perceptions call increasingly for background
knowledge and conjecture on the ascriptor's part as we
move away from observation sentences.
29 . Anomalous monism
We reflected in §24 that a neurological rendering of ' Tom
perceives that it i s raining' , appl icabl e to all su ch occa si on s
merely on Tom's part, would already be pretty formidable
even if Tom's neural make-up were known in detail. We
reflec ted further that a neuro lo gical renderin g of ' perceiv es
that it is raining', applicable to all comers, would be out of
the question.
Yet each percepti on is a sin gle occurrence in a particul ar
brain, and is fully specifiabl e in neuro l o gi cal terms once the
details are known. We cannot say the same for a belief,
which can be publicly shared. but we can say somewhat the
same for the instance of the belief in a si n gle believer. The
period during which I go on believing that the earth rotates
is distin guished from my earlier stages by at least some
verbal dispositions, which must reside in some distinctive
quirks in my nervous system.
Per ceptions are neural reali t ies . and so are the individual
instances of beliefs and other propositional attitudes insofar
as these do not fade out into irreality altogether (§27) . Phys
30 . Modalities
The modalities of necessity and possibility are not overtly
mentalistic, but still they are intensional, in the sense of
resisting substitutivity of i de�tity Here again we have the
.
31 . A mentalistic heritage
Appreciation of one another's perceptions is fundamental,
we saw (§24) , to the handing down of language. The men
talistic strain is thus archaic. We see it in animism, the
primitive ascription of minds to bodies on an excessive
scale. Perhaps there was a vestige of animism in Aristotle's
theory of the natural motion of substances: earth down-
INTENSION 75
TRUTH
32 . Vehicles of truth
What are true or false, it will be wi del y agreed, are proposi
tions . But it would not be so widely agreed were it not for
ambiguity of ' prop o si t i on ' . Some unders t an d the word a s
referring to sentences m eeting certain specifications.
Others understand it as referring rather to the meanings of
such sentences. What looked like wide agreement thus re
solves into two schools of thought: fo r the first school the
vehicles of truth and falsity are the sentences , and for the
second t hey are the meanings of the sentences.
A weakness o f this second position is the tenuousness of
the notion o f sentence meani ng s . The tenuousness reaches
the breaking p oin t if one is pursuaded of my thesis of the
indeterminacy of translation (§§ I 8, 2 1 ) . Even apart from that
thesis, it seems perverse to bypass the v isi b le or audible
sentences and to center upon sentence meanings as truth
vehicles; for it is only by recurring to the sentence that we
can say which sen tence meaning we
have in mind.
There was indeed a motive for p res sing to the sentence
meanings . Many sentences in the same or different lan
guages are deemed to be alike in meanin g . and distinctions
among them are indifferent to truth; so one narrowed the
field by ascribing truth rather to the meanings. This motive
PURSUIT O F TRUTH
would be excellent if the notion of sentence meani n g were
not so elusive. But as matters stand we fare better by treat
ing direct l y of sentences. These we can get our teeth into.
There was also a second motive, eq ual and opposite to the
first, for pressing on to the sentence m eani n g s ; n a m el y, that
one and the same sentence can be true o n some o ccasion s
and false on others. Thus 'The Pope wi ll visit Boston' was
true but turned false after his last vis i t . ' I ha ve a headache' is
true or false depending on who says it and when. Ambi
g u ity or vagueness of terms, also, can cause the truth value
of a sentence to depend in part on the sp eaker ' s in ten tio n .
Propositions, thought of as sentence m ea nin g s , were th e
meanings exclusively of sentences of a fi rm e r sort, not sub
ject to such vacillations; w ha t we may call eternal sen ten ces . t
My obvious response, then, i s that those eternal sentences
themselves can serve as the t ru t h vehicles. Just think of T ,
'yo u ' , 'he', 'she', 'here', and ' t here ' as supplanted by names
and addresses or other i d enti fyin g particulars as needed .
Think of tenses as dropped; we can use dates, the predicate
'earlier than', and the like as needed. Think of ambiguities
and vaguenesses as resolved by paraphrase-not abso
lut ely , but enough to immobilize the truth value of the
parti cul ar sentence. The truth values need not be known,
b u t the y must be stable.
The attitude is the one that is familiar in the teaching of.
l og ic . When we take illustrative sentences from everyday
l In my logic books O f 1 940, 1 94 1 , and 1 9 50, and revised editions down
the years, my word for them was ' state m en t '; but I became chary of it
because of its cu stomary use rather for an act . ' Eternal sentence', along
w ith 'standing sentence ' (§4) , d a t e s from Word and ObjtCl. ' St an di n g
sentence' is more inclusive. 'The Timts h as come' is a standing sentence,
for it can com m and assent all day independently of interim stimulation ;
but it is not eternal .
TRUTH 79
JJ . Truth as disquotation
Such being what admit of truth, then, w he rein does their
t rut h consist? They qualify as true, one is told, by corre
sponding to reality. But correspondence word by word will
not do; it in vites the idle cluttering of reality with a bizarre
host of fancied obj ects, just for the sake of correspondence.
80 PURSUIT OF TRUTH
A neater plan is to posit facts, as correspondents of true
sentences as wholes; but this still is a put-up job. Objects in
abundance, concrete and abstract, are indeed needed for an
account of the world; but facts contribute nothing beyond
their specious support of a correspondence theory .
Yet there i s some underlying validity t o the correspon
dence theory of truth, as Tarski has taught us. Instead of
saying that
'Snow is white' is true if and only if it is a fact
that snow is white
we can sim ply delete 'it is a fact that' as vacuous, and there
with facts themselves:
' Snow is white' is true if and only if snow is
white.
To ascribe truth to the sentence is to ascribe whiteness to
snow; such is the correspondence, in this example. Ascrip
tion of truth just cancels the quotation marks. Truth is dis
quotation.
So the truth predicate is superfluous when ascribed to a
given sentence; you could j ust utter the sentence. But it is
needed for sentences that are not given. Thus we may want
to say that everything someone said on some occasion was
true, or that all consequences of true theories are true. Such
contexts, when analyzed logically, exhibit the truth predi
cate in application not to a quotation but to a pronoun, or
bound variable.
The truth predicate proves invaluable when we want to
generalize along a dimension that cannot be swept out by a
general term. The easy sort of generalization is illustrated
by generalization on the term 'Socrates' in ' Socrates is mor-
TRUTH 81
tal'; the sentence generalizes to 'All men are mortal'. The
general term ' man' has served to sweep out the desired
dimension of generality. The harder sort of generalization is
illustrated by generalization on the clause 'time flies' in 'If
time flies then time flies' . We want to say that this com
pound continues true when the clause is supplanted by any
other; and we can do no better than to say just that in so
many words, including the word 'true'. We say "All sen
tences of the form 'If p then p ' are true. " We could not
generalize as in ' All men are mortal' , 1?ecause 'time flies' is
not, like 'Socrates', a name of one of a range of objects
(men) over which to generalize. We cleared this obstacle by
semantic ascent: by ascending to a level where there were
indeed objects over which to generalize, namely lingui�tic
objects, sentences.
Semantic ascent serves also outside of logic. When Ein
stein propounded relativity, disrupting our basic concep
tions of distance and time, it was hard to assess it without
leaning on our basic conceptions and thus begging the ques
tion. But by semantic ascent one could compare the new
and old theories as symbolic structures, and so appreciate
that the new theory organized the pertinent data more sim
ply than the old. Simplicity of symbolic structures can be
appreciated independently of those basic conceptions.
As already hinted by the correspondence theory, the
truth predicate is an intermediary between words and the
world. What is true is the sentence, but its truth consists in
the world's being as the sentence says. Hence the use of the
truth predicate in accommodating semantic ascent.
The disquotational account of truth does not define the
truth predicate-not in the strict sense of 'definition'; for
definition in the strict sense tells how to eliminate the
PURSUIT OF TRUTH
defined expression from every desired context in favor of
previously established notation. But in a looser sense the
disquotational account does define truth. It tells us what it is
for any sentence to be true, and it tells us this in terms just as
clear to us as the sentence in question itself. We understand
what it is for the sentence 'Snow is white' to be true as
clearly as we understand what it is for snow to be white.
Evidently one who puzzles over the adj ective 'true' should
puzzle rather over the sentences to which he ascribes it.
'True' is transparent.
For eternal sentences the disquotational account of truth
is neat, we see, and simple. It is readily extended, more
over, to the workaday world of individual utterances; thus
an utterance of ' I have a headache' is true if and only if the
utterer has a headache while uttering it.
34 . Paradox
It seems paradoxical that the truth predicate, for all its trans
parency, should prove useful to the point of indispensabil
ity. In the matter of paradox, moreover, this is scarcely the
beginning. Truth is notoriously enmeshed in paradox, to
the point of out-and-out antinomy.
An ancient form of the antinomy of truth is the Paradox
of the Liar: 'I am lying', or 'This sentence is not true'. A
looser and fancier version was the paradox of Epimenides
the Cretan, who said that all Cretans were liars . The under
lying antinomy can be purified for logical purposes to read
thus:
40 . Truth in mathematics
What now of those parts of mathematics that share no em
pirical meaning, because of never getting a pplied in natural
science? What of the hi g her reaches of set th eo ry? We see
them as mean in gful because they are couched in the same
grammar and vocabulary that generate the applied parts of
mathematics . We are just sparing ourselves the unnatural
gerrymandering of grammar that would be needed to ex
clude them. On our two-valued approach they then qualify
as true or false, albeit inscrutably.
They are not wholly inscrutable. The main axioms of set
theory are generalities operative already in the applicable
TRUTH 95
part of the domain. Further sentences such as the con
tinuum hypothesis and the axiom of choice, which are inde
pendent of those axioms, can still be submitted to the con
siderations of simplicity, economy, and naturalness that
contribute to the molding of scientific theories generall y .
Such considerations support COders axiom of construcribil
ity, 'V = L'. 5 It inactivates the more gratuitous flights of
higher set theory, and incidentally it implies the axiom of
choice and the continuum hypothesis. More sweeping
economies have been envisioned by Hermann Weyl, Paul
Lorenzen, Errett Bishop, and currently Hao Wang and Sol
omon Feferman, who would establish that all the mathe
matical needs of science can be supplied on the meager basis
of what has come to be known as predicative set theory. 6
Such gains are of a piece with the simplifications and econo
mies that are hailed as progress within natural science itself.
It is a matter of tightening and streamlining our global sys
tem of the world.
42 . Irresoluble rivalry
So we may limit our attention to Case 2. Let us limit it
further to global systems of the world, so that there is no
question of fitting the rival theories into a broader context.
So we are imagining a global system empirically equivalent
to our own and logically compatible with ours but hinging
on alien terms. It may seem that as staunch empiricists we
should reckon both theories as true. Still, this line is unat
tractive if the other theory is less simple and natural than
ours; and indeed there is no limit to how grotesquely cum
bersome a theory might be and still be empirically equiva
lent to an elegant one. We do better, in such a case, to take
advantage of the presence of irreducibly alien terms. We can
simply bar them from our language as meaningless. After
all, they are not adding to what our own theory can predict,
any more than 'phlogiston' or 'entelechy' does, or indeed
'fate' , 'grace', ' ni rvana ' , ' mana'. We thus consign all con
texts of the alien terms to the limbo of nonsentences.
We have here an encroachment of coherence considera
tions upon standards of truth. Simplicity and naturalness
are making the difference between truth and meaningless
ness.
We might still choose to enrich our original theory with
any novel findings of the other theory that do not use the
alien terms. It would be a matter of welcoming information
from a presumed dependable outside source, much as sup
plementary truths of number theory are got by excursions
TRUTH 99
43 . Two indeterminacies
There is an evident parallel between the empirical under
deter mina tion of global science and the indeterminacy of
translation. In both cases the total ity of p o s s ible evidence is
insufficient to cli n ch the system uniquely. But the indeter
minacy of translation is additional to the other. If we settle
upon one of the empirically equivalent systems of the
world, however arbitrarily, we still h a v e within it the inde
terminacy of translation.
Another distinctive point about the indeterminacy of
tran slation is that it cle arly has nothing to do with inaccessi
ble facts and human limitations. Dispositions to ob s e rvable
behavior are all there is for semantics to be rig h t or wrong
about (§ I 4) . In the case of sy s tems of the world, on the other
hand, one is prepared to belie ve that reality exceeds the
scope of the human app aratus in unspecifiable ways.
Let us now look more closely to parallels. On the one
hand we have the two incompatible but equally faithful
systems of translation; each propounds some translations
that the other rej e cts . On the other hand we have two in-
102 PURSUIT OF TRUTH
compatible but empirically equivalent systems of the
world. We noted in § 1 8 that we can reconcile the two sys
tems of translation by recognizing them as defining differ
ent relations, translation! and translation2' We noted in §4 I
that we can reconcile the two systems of the world by simi
larly splitting one or more theoretical terms.
What the indeterminacy of translation shows is that the
notion of propositions as sentence meanings is untenable.
What the empirical under-determination of global science
shows is that there are various defensible ways of conceiv
ing the world.
RE F E R E N C E S
CREDITS
INDEX
REFERENCES
pp· 43 - 5 5 ·
Parsons. Charles. Mathematics and Philosophy . Ithaca: Cornell Uni
versity Press, 1 98 3 .
1 06 REFERENCES
Poincare. Henri. Scimct and Hypothesis. New York. 1 905.
Popper. Sir Karl. The Logic oj Scientific Discover. New York: Basic
Books. 1 959.
Premack. David. eavagai. Cambridge: MIT Press. 1 986.
Pumam. Hilary . "Mathematics without Foundations . " Journal oj
P#ailosophy 64 ( 1 96 7) . pp . 5 -22 .
Quine. W. V. Mathemaliad Logic. New York. 1 940. Corrected ed.
(thanks to Wang) . Cambridge: Harvard University Press. 1 95 1 .
--- Elemmt4ry Logic. Bos to n. 1 94 1 . Rev. ed . • Harvard Univer
sity Press. 1 96 5 .
-- Methods oJLogic. New York . 1 950. 3 d ed 1 972. 4th, Harvard
. •
--- TItt Ways oj Paradox and Other Essays. New York, 1 966.
Enlarged ed Harvard U niv ers i ty Press, 1 976.
.•
Dummeti. Michael, 93
Carnap, Rudolf, 56
Cause, 7 5-76
ennis paribus, 1 8. 74 Ecumenical line, 99- 1 00
Checkpoints, 10-21, 3 8, 43-47 Einstein, Albert. 20, 3 5 . 8 1
Chisholm, Roderick M . , 65n, 7 1 n Empathy. 42-43, 46. 61-63 ,
Classes, 88-90 68-69
Coherence, 93, 98 Empirical content, 1 6- 1 8. S J -54,
Communication, 44 94-95
Concatenation, 69, 8 5 Empirical equivalence. 1 7, 95- 1 0 1
112 I N D EX
Empi rici sm , 1 9 Indeterminacy : of reference.
Epistemology, 1 -2, 8, 1 2, 1 9- 20, 50- 5 2 ; of trans la tion . 37.
42 47-5 1 ; 53. 77 . 1 0 1 - 1 02
Equivalence, 1 7. 54- 5 5 , 95- 1 0 1 Indexicals. 78
Essence. 5 6 . 74 Intensionality. 68, 71 -73
Eternal sentences, 78-79
Evidence. 1 -2 . 5. 1 2- 1 3
Excluded middle. 90-92 Kant. Immanuel. 1 8
Existence, 27-28. 92 Kirk, Robert, 48
E x periment . 9- 1 0. 1 4
Exten si onal ity . 5 2n. 72
Learning. 5- 8 , 2 3 - 24, 29-30, 3 9.
44 ; empath y in. 42. 6 1 -62
Leon elli , Michele, viii
Facts, 80
Fallibility, 2 1
Levy, Edwin, 5 1
Feferman, Solomon, 95 Lewis, Clarence I. . 8
Lexicography. 56- 59
Firth. Roderick, 7
Logic, 1 4, 27-2. 8 . 3 5 - 36• 45 .
FsUesdal. Dagfinn. viii. 4 1 . 99
78-79. 8 1 . 92.
Freedom, 91
Frege. Gottlob, 69
Lorenzen. Pau l, 95
LOwenheim-Skolem Theorem. 3 2
Future. 90-9 1
Massey. G. J. . 5 1
Gavdgai, 4�, 5 1 Mathematics. 1 5 . 56. 94-95
Geometry. 5 I . 96 Meurings. 3 7, 52-56. 77. 1 02.
Gibson. Robert F . •
viii. 99
Mind, 6 1 , 72, 74-75
God, 90-9 1
Minimum mutilation. 14. I S. S6
GOdel, Kurt. 3 2 . 95 Modal operato rs , 3 0 , 73
GreUing. Kurt. 86-87
Moore, G. E 5 5 . •
Ullian. Joseph S. . zo
Under-determination. 96- 1 02
Understanding. 5 8-59 Yosida. Natuhiko. 1 8