Journal of Philosophy, Inc. The Journal of Philosophy

Journal of Philosophy, Inc.
It Ain't Necessarily So
Author(s): Hilary Putnam
Source: The Journal of Philosophy, Vol. 59, No. 22, American Philosophical Association
Eastern Division: Symposium Papers to be Presented at the Fifty-Ninth Annual Meeting,
New York City, December 27-29, 1962 (Oct. 25, 1962), pp. 658-671
Published by: Journal of Philosophy, Inc.
Stable URL: https://www.jstor.org/stable/2023596
Accessed: 03-02-2020 19:58 UTC
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms
Journal of Philosophy, Inc. is collaborating with JSTOR to digitize, preserve and extend
access to The Journal of Philosophy
This content downloaded from 62.204.213.86 on Mon, 03 Feb 2020 19:58:28 UTC
All use subject to https://about.jstor.org/terms
658 THE JOURNAL OF PHILOSOPHY
what in fact we would say. Rather we now s'ay somethingy about

the situation. But where we cannot at present pass a judgment
one way or the other there is nothing else we can appeal to as
showing what is correct. There is no reason, a priori, why our
present usage should legislate for all hypothetical cases. Given
present circumstances, the correct thing to say is that all whales
are mammals. But whether this is, as we intend it, a necessary
truth or contingent is indeterminate. It is indeterminate because
the decision as to which it is would depend upon our beingr able
to say now what we should say about certain hypothetical cases.
And evidently we are not prepared to do that.
If this is so, Lewis's idea that what criteria are attached to a
given term as part of its meaning must be "fixed" in advance
of experience seems to be false. And the corollary, that it would
always be clear upon investigation whether or not the criteria at-
taching to one term are or are not included in those attachinlg
to another is likewise false.
This is not to say that the concept of lnecessity is useless here.
It might be thought of as an ideal rigidity in our judgments about
what to say concerninig hypothetical cases. I have dealt here only
with the first of the two examples used in the first part of the
paper. Whether a siinilar indeterminacy can be found concern-
ing, say, the statement that all cats are animals I have not dealt
with. If it can, then a more sweepilng reinterpretation of the
notion of necessity might be called for.
KEITH S. DONNELJLAN
CORNELL UNIVERSITY
IT AIN'T NECESSARILY SO *
T HIIE two statements that Donellan considered il hlis paper t are

both more or less analytic in character. By that I mean
that they are the sort of statemenit that miiost people would con-
sider to be true by definition, if they considered them to be neces-
sary truths at all. Olne might quarrel about whether 'all whales
are mammals' is a necessary truth at all. But if one considers it
to be a necessary truth, then one would consider it to be true
by definition. And, similarly, most people wouLld say that 'all cats
* To be presented in a symiposium on "Necessary Truth'' at the fifty!-

ninth annual meeting of the American Phlilosoplhical Associatiol, Eastern
Division, December 27, 1962.
t First paper of this symposium, this JOURNAL, 59: 647.
IT AIN'T NECESSARILY SO 659
are animials' is true by definition, notwithstanding the fact that

they would be hard put to answer the question, "true by what
definition ? "
I like what Donellan had to say about these statements, and I
liked especially the remark that occurs toward the end of his paper,
that there are situations in which we are confronted by a question
about how to talk, but in which it is not possible to describe one of
the available decisions as deciding to retain our old way of talking
(or "not to change the meaning") and the other as deciding to
adopt a "'new " way of talking (or to "'change the meaning ").
In this paper I want to concentrate mostly on statements that
look necessary, but that are not analytic; on "synthetic necessary
truths," so to speak. This is not to say that there are not serious
problems connected with analyticity. On the contrary, there cer-
tainly are. The general area of necessary truths might be broken
up, at least for the sake of preliminary exploration (and we phi-
losophers are still in the stage of preliminary exploration even after
thousands of years) into three main subareas: the area of analytic
truths, the area of logical and mathematical truths, and the area
of "synthetic a priori" truths. I don't mean to beg any questions
by this division. Thus, separating logical and mathematical truths
from analytic truths is not meant to prejudge the question whether
they are or are not ultimately anialytic.
I. ANALYTIC TRUTHS
The "analyticity" of 'all cats are animals' or, what is closely

related, the redundancy of the expression 'a cat which is an
animal' seems to depend on the fact that the word 'animal' is the
name of a semantic category 1 and the word 'cat' is a member of
that category.
In an earlier paper 2 I called words that have an analytic defi-
nition "one-criterion words." Many words that are not one-
criterion words fall into semantic categories-in fact, all nouns
fall into semantic categories. 'House', for example, falls into the
semantic category material object, 'red' falls into the semantic
category color, and so on. Thus, for any noun one can find an
analytic or quasi-analytic truth of the sort 'a cat is an animal',
'a house is a material object', 'red is a color', and so forth. But it
hardly follows that all nouns are one-criterion words-in fact,
there are only a few hundred one-criterion words in English.
1 The notion of a "semantic category" is taken from J. A. Fodor and
Jerrold Katz, "The Structure of a Semantic Theory," to appear in Language.
2 "The Analytic and the Synthetic," in Herbert Feigl and Grover Max,
wvell, eds., Minnesota Studies in the P1iilosophy of Science, vol. III.
It is important to distinguish "analytic" truths of the sort 'all

cats are animals' from analytic truths of the sort 'all bachelors are
unmarried', in part because the former tend to be less necessary
than the latter. It might not be the case that all cats are ani-
mals; they might be automata!
There are, in fact, several possibilities. If some cats are ani-
mals in every sense of the word, while others are automata, then
there is no problem. I think we would all agree that these others
were neither animals nor cats but only fake cats-very realistic
and very clever fakes to be sure, but fakes nonetheless. Suppose,
however, that all cats on earth are automata. In that case the
situation is more complex. We should ask the question, "Were
there ever living cats?" If, let us say, up to fifty years ago there
were living cats and the Martians killed all of them and replaced
them all overnight with robots that look exactly like cats and
can't be told from cats by present-day biologists (although, let us
say, biologists will be able to detect the fake in fifty years more),
then I think we should again say that the animals we all call cats
are not in fact cats, and also not in fact animals, but robots. It is
clear how we should talk in this case: "there were cats up to fifty
years ago; there aren't any any longer. Because of a very ex-
ceptional combination of circumstances we did not discover this
fact until this time."
Suppose, however, that there never have been cats, i.e., genuine
non-fake cats. Suppose evolution has produced many things that
come close to the cat but that it never actually produced the cat,
and that the cat as we know it is and always was an artifact.
Every movement of a cat, every twitch of a muscle, every meow,
every flicker of an eyelid is thought out by a man in a control
center on Mars and is then executed by the cat's body as the
result of signals that emanate not from the cat's "brain" but
from a highly miniaturized radio receiver located, let us say, in
the cat's pineal gland. It seems to me that in this last case, once
we discovered the fake, we should continue to call these robots
that we have mistaken for animals and that we have employed as
house pets "cats," but not "animals."
This is the sort of problem that Donellan discussed in connec-
tion with the "all whales are mammals" case. Once we find out
that cats were created from the beginning by Martians, that they
are not self-directed, that they are automata, and so on, then it is
clear that we have a problem of how to speak. What is not clear
is which of the available decisions should be described as the de'
cision to keep the meaning of either word ('eat' or 'animal') un-
changed, and which decision should be described as the decision
to change the meaning. I agree with Donellan that this question

has no clear sense. My own feeling is that to say that cats turned
out not to be animals is to keep the meaning of both words un-
changed. Someone else may feel that the correct thing is to say
is, "It's turned out that there aren't and never were any cats."
Someone else may feel that the correct thing to say is, "It's
turned out that some animals are robots." Today it doesn't seem
to make much difference what we say; while in the context of a
developed linguistic theory it may make a difference whether we
say that talking in one of these ways is changing the meaning and
talking in another of these ways is keeping the meaning un-
changed. But that is hardly relevant here and now; when
linguistic theory becomes that developed, then 'meaning' will itself
have become a technical term, and presumably our question now is
not which decision is changing the meaning in some future tech-
nical sense of 'meaning', but what we can say in our present
language.
II. SYNTHETIC NECESSARY STATEMENTS
Later I shall make a few remarks about the truths of logic and
mathematics. For the moment, let me turn to statements of quite
a different kind, for example, the statement that if one did X and
then Y, X must have been done at an earlier time than Y, or the
statement that space has three dimensions, or the statement that
the relation earlier than is transitive. All of these statements are
still classified as necessary by philosophers today. Those who feel
that the first statement is "conceptually necessary" reject time
travel as a "conceptual impossibility."
Once again I will beg your pardon for engaging in philosophical
science fiction. I want to imagine that something has happened
(which is in fact a possibility) namely, that modern physics has
definitely come to the conclusion that space is Riemannian. Now,
with this assumption in force, let us discuss the status of the
statement that one cannot reach the place from which one came
by traveling away from it in a straight line and continuing to move
in a constant sense. This is a geometrical statement. I want to
understand it, however, not as a statement of pure geometry, not
as a statement about space "in the abstract," but as a statement
about physical space, about the space in which we live and move
and have our being. It may be claimed that in that space, in the
space of actual physical experience, the notion of a "straight line"
has no application, because straight lines are supposed to have no
thickness at all and not the slightest variation in curvature, and
we ca.nnot identify by any physical means paths in space with
these properties (although we can approximate them as closely as

we desire). However, this is not relevant in the present context.
Approximate straight lines will do. In fact, it is possible to state
one of the differences between a Riemannian and a Euclidean world
without utsing the notion of "straightness" at all. Suppose that
by a "place" we agree to mean, for the nonce, a cubical spatial
region whose volume is one thousand cubic feet. Then, in a
Riemannian world, there are only a finite niumber of disjoint
"places." There is no path, curved or straight, on which one
can "go sight-seeing" in a Riemannian world and hope to see
more than a certain number N of disjoint "places." Thus, the
statement that one can see as many distinct disjoint places as one
likes if one travels far enough along a suitable path is one which
is true in any Euclidean world, and not in any Riemannian world.
Now, I think it is intuitively clear that the two propositions
just mentioned: that one cannot return to one 's starting point
by traveling on a straight line unless one reverses the sense of
one's motion at some point, and that one can visit an arbitrary
number of distinct and disjoint "places" by continuing far enough
on a suitable path-that is, that there is no finite upper bound on
the total number of "places"-had the status of necessary truthls
before the nineteenth century.
Let me say, of a statement that enjoys the status with respect
to a body of knowledge that these statements enjoyed before the
nineteenth century and that the other statements alluded to enjoy
today, that it is "necessary relative to the appropriate body of
knowledge." This notion of necessity relative to a body of knowl-
edge is a technical notion being introduced here for special pur-
poses and to which we give special properties. In particular,
when we say that a statement is necessary relative to a body of
knowledge, we imply that it is included in that body of knowledge
and that it enjoys a special role in that body of knowledge. For
example, one is not expected to give much of a reason for that kind
of statement. But we do not imply that the statement is neces-
sarily true, although, of course, it is thought to be true by someone
whose knowledge that body of knowledge is.
We are now confronted with this problem: a statement that
was necessary relative to a body of knowledge later came to be
declared false in science. What can we say about this?
Many philosophers have found it tempting to say that niothino
significant happened, that we simply changed the meanliilg of words.
This is to assimilate the case to Donellan's first case, the whales-
mammals case. Just as 'whale' may, perhaps, have two meanings
-one for laymen and oiie for scientists-so 'straight line' may
have two meanings, one for laymen and one for scienitists. Paths
that are straight in the layman's sense may not be straight in the
scientist's sense, and paths that are straight in the scientist's sense
may not be straight in the layman's sense. Onie caanInot come back
to one's starting point by proceeding indefinitely on a path that
is straight in the old Euclidean sense; when the scientist says that
one can come back to one 's starting point by continuing long
enough on a straight line, there is no paradox and no contradiction
with common sense. What he means by a straight line may be
a curved line in the layman's sense, and we all know you can come
back to your starting point by continiuilng' long enough on a closed
curve.
This account is not tenable, however. To see that it isn't, let
us shift to the second statement. Here we are in immediate diffi-
culties because there seems to be no difference, even today, between
the layman's senise of 'path' and the scientist's sense of 'path'.
Anything that the layman could trace out if we gave him but
world enough and time, the scientists would accept as a "path,"
and anything that the scientist would trace out as a "path," the
layman would accept as a "path." (Here I do not count micro-
scopic "paths" as paths.) Similarly with 'place'. To put it an-
other way: if Euclidean geometry is only apparently false owing
to a change in the meaning of words, then if we keep the meanings
of the words unchanged, if we use the words in the old way,
Euclidean geometry must still be true. In that case, in addition
to the N "places" to which one can get by following the various
paths in our Riemannian space, there must be infinitely many ad-
ditional "places" to which one can get by following other paths
that somehow the scientist has ov-erlooked. Where are these
places? Where are these other paths? Tn fact, they don't exist.
If someone believes that they exist, then he must invent special
physical laws to explain why, try as we may, we never sncceed in
seeing one of these other places or in sticking to one of these other
paths. If someone did accept such laws and insisted on holding
on to Euclidean geometry in the face of all present scientific
experience, it is clear that he would not have simply "made a
decision to keep the meanings of words unehanged'"; he would
have adopted a metaphysical theory.
The statement that there are onily finitely miany disjoint
"places" to get to, travel as yoit )may expre.,ses a downirio-ht "con1-
ceptual impossibility" within the framework of Euclidean geom-
etry. And one cannot say that all that has happened is that we
have changed the meaning of the word 'path', because in that
case one would be committecl to the metaphysical hypothesis that,
in addition to the "paths" that are still so ealled, there exist others
which are somehow physically inaccessible and additional "places"
which are somehow physically inaccessible and which, together with
what the physicists presently recognize as places and paths, fill out
a Euclidean space.
Insofar as the terms 'place', 'path', and 'straioght line' have any
application at all in physical space, they still have the application
they always had; something that was literally inconceivable has
turned out to be true.
Incidentally, although modern physics does not yet say that
space is Riemannian, it does say that our space has variable
curvature. This means that if two light rays stay a constant
distance apart for a long time and then come closer together after
passing the sun, we do not say that these two light rays are follow-
ing curved paths through space, but we say rather that they follow
straight paths and that two straight paths may have a constant
distance from each other for a long time and then later have a
decreasing distance from each other. Once again, if anyone wishes
to say, "Well, those paths aren't straight in the old sense of
'straight'," then I invite him to tell me vhich paths in the space
near the sun are "really straight." And I guarantee that, first,
no matter which paths he chooses as the straight ones, I will be
able to embarrass him acutely. I will be able to show, for example,
not only that light rays refuse to travel along the paths he claims
to be really straight, but that they are not the shortest paths by
any method of measurement he may elect; one cannot even travel
along those paths in a rocket ship without accelerations, decelera-
tions, twists, turns, etc. In short, the paths he claims are "really
straight" will look crooked, act crooked, and feel crooked. More-
over, if anyone does say that certain nongeodesics are the really
straight paths in the space near the sun, then his decision will have
to be a quite arbitrary one; and the theory that more or less
arbitrarily selected curved paths near the sun are "really straight"
(because they obey the laws of Euclidean geometry and the
geodesics do not) would again be a metaphysical theory, and the
decision to accept it would certainly not be a mere decision to "keep
the meaning of words unchanged."
III. THE CLUSTER CHARACTER OF GEOMETRIC NOTIONS
Distance cannot be "operationally defined" as distance accord-

ing to an aluminum ruler, nor as distance according to a woodepi
ruler, nor as distance according to an iron ruler, nor as distance
according to an optical measuring instrument, nor in terms of any
other one operational criterion. The criteria we have for distance

define the notion collectively, not individually, and the connection
between any one criterion and the rest of the bundle may be
viewed as completely synthetic. For example, there is no contra-
diction involved in supposing that light sometimes travels in
curved lines. It is because of the cluster character of geometrical
concepts that the methods usually suggested by operationists for
demonstrating the falsity of Euclidean geometry by isolated experi-
ments would not have succeeded before the development of non-
Euclidean geometry. If someone had been able to construct a
very large light-ray triangle and had shown that the siim of the
angles exceeded 1800, even allowing for measuring errors, he
would not have shown the ancient Greek that Euclidean geometry
was false, but only that light did not travel in straight lines.
What accounted for the necessity of the principles of Euclidean
geometry relative to pre-nineteenth-century knowledge? An an-
swer would be difficult to give in detail, but I believe that the
general outlines of an answer are not hard to see. Spatial loca-
tions play an obviously fundamental role in all of our scientific
knowledge and in many of the operations of daily life. The use
of spatial locations requires, however, the acceptance of some
systematic body of geometrical theory. To abandon Euclidean
geometry before non-Euclidean geometry was invented would be
to "let our concepts crumble."
IV. TIME TRAVEL
I believe that an attempt to describe in ordinary language

what time travel would be like can easily lead to absurdities and
even downright contradictions. But if one has a mathematical
technique of representing all the phenomena subsumed under some
particular notion of "time travel," then it is easy to work out a
way of speaking, and even a way of thinking, corresponding to
the mathematical technique. A mathematical technique for repre-
senting at least one set of occurrences that might be meant by
the term 'time travel' already exists. This is the technique of
world lines and Minkowski space-time diagrams. Thus, suppose,
for example, that a time traveler-we'll call him Oscar Smith-
and his apparatus have world lines as shown in the diagram.
From the diagram we can at once read off what an observer sees
at various times. At to, fdr example, he sees Oscar Smith not yet
a time traveler. At time t1 he still sees Oscar Smith at place A,
but also he sees something else at place B. At place B he sees,
namely, an event of "creation"-not "particle-antiparticle crea-
KEY
t2
t 2 W 1>\ / Oscar Smith World Line
E|I Apprtus World Line
ti
to _4
A B
S PAC E
tion," but the creatioli of two iiiacro-objects which separate. One

of these macro-objects is easily described. It is simply an older
Oscar Smith, or an individual resembling in all possible ways an
older version of Oscar Smith, together with the apparatus of a time
machine. The world-line diagram shows that the older Oscar-
let's call him Oscar3-leaves his time machine. The other object
that was created in the same event is a somewhat peculiar object.
It is a system consisting of a third Oscar Smith, or, more precisely,
of a body like that of Oscar Smith, seated in a time machine. But
this system consisting of the Oscar Smith body and the time ma-
chine is a very exceptional physical system. If we talke a moving
picture of this physical system during its entire period of existence,
we will find that if that movie is played backward then the events
in it are all normal. In short, this is a system running backward
in time-entropy in the system is decreasing instead of increasing,
cigarette butts are growing into whole cigarettes, the Oscar Smith
body is emitting nloises that resemble speech sounds played back-
ward, and so forth. 'T'his system that is running(I backward in
time continues to exist until the time t2, when it merges with
Oscar Smith, and we see annihihation-not "particle-antiparticle
annihilation," but the annihilation of the real Oscar Smith and
the running-backward system. During a certain period of time,
there are three Oscar Smiths: Oscar Smith1, Oscar Smith3, and
the Oscar Smith who is living backward in time (Oscar Smiith., We
shall call him). We can even predict subjective phenomena from
the world-line diagram. We can say, for example, what sort of
memories Oscar Smith3 has at the moment of his "creation." He
has, namely, all the memories that someone would have if he had
all the experiences of Oscar1 up to the moment of his annihilation
and then all the experiences shown as occurring to the living-
backward Oscar, Oscar., on a movie film, provided the movie film
is reversed so that these experiences are shown as happening in
the normal order. I have no doubt whatsoever as to how any

reasonable scientist would describe these events, if they actually
transpired. He would interpret them in terms of the world-line
diagram; i.e., he would say: "There are not really three Oscar
Smiths; there is only one Oscar Smith. The Oscar Smith you call
Oscar Smith1 lives forward in time until the time t2; at t2 his
world line for some reason bends backward in time, and he lives
backward in time from t2 back to the time t1. At t1 he starts
living forward in time again and continues living forward in time
for the rest of his life."
I remember having a discussion concerning time travel with a
philosopher friend a number of years ago. I told him the story
I have just told you. My friend's view, in a nutshell, was that
time travel was a conceptual impossibility. The phenomena I de-
scribed can, of course, be imagined; but they are correctly de-
scribed in the way I first described them; i.e., Oscar Smith1 lives
until t2, at which time he collides with the strange system. When
the two systems merge, they are both annihilated. At the time t1
this strange physical system was created, as was also another
individual very much resembling Oscar Smith1, but with an en-
tirely fictitious set of memories including, incidentally, memories
of Oscar Smith1's existence up to the time t2.
Let us ask ourselves what makes us so sure that there is here a
consistently imaginable set of circumstances to be described. The
answer is that it is the mathematical representation, i.e., the world-
line diagram itself, that gives us this assurance. Similarly, in the
case of space with variable curvature, near the sun, the only thing
that makes us sure that there is a consistently imaginable set of
phenomena to be described is their represcintation in termns of the
mathematics of non-Euclidean geometry.
The present case also exhibits disanalogies to the geometric
case. In the geometric case we could not go on using language in
the old way-if to preserve Euclidean geometry is to go on using
language in the old way-without finding ourselves committed to
ghost places and ghost paths. In the present case, we can go on
using language in the old way-if my friend's way of describing
the situation is the one which corresponds to "using language
in the old way" -without having to countenance any ghost entities.
But there are a host of difficulties which make us doubt whether
to speak in this way is to go on using language without any change
of usage or meaning. First of all, consider the sort of system that
a physicist would describe as a "human being living backward in
time. " The same system would be described by my friend not as a
person at all, but as a human body going through a rather nauseat-
668 THIE JOURNAL OF PHILOSOPHY
ing succession of physical states. Thus, on my friend's account,

Oscar2 is not a person at all, and Oscar1 and Oscar3 are two quite
different persons. Oscar1 is a person who had a normal life
up to the time t2 when something very abnormal happened to him,
namely, he vanished, and Oscar3 is a person who had a normal life
from the time t1 on, but who came into existence in a very abnormal
way: he appeared out of thin air. Consider now the legal prob-
lems that might arise and whose resolution might depend on
whether we accepted the physicist's account or the account of my
friend. Suppose Oscar1 murders someone but is not apprehended
before the time at which he vanishes. Can we or can we not
punish Oscar3 for Oscar1 's crime? On the physicist's account,
Oscar3 is Oscar1, only grown older, and can hence be held re-
sponsible for all the actions of Oscar1. On my friend's account,
Oscar3 is only a person under the unfortunate delusion that he
is Oscar1 grown older, and should be treated with appropriate
kindness rather than punishment. And, of course, no one is re-
sponsible for Oscar2 's actions on this view, since they are not really
actions at all. And if Oscar1's wife lives with Oscar3 after t2,
she is guilty of unlawful cohabitation, while if she lives with Oscar3
prior to t2, the lady is guilty of adultery. In this kind of case,
to go into court and tell the story as my friend would tell it would
be to use language in a most extraordinary way.
This case differs importantly from Donellan 's cases in that,
although our problem can be described as a problem of how to
speak, it is not merely a problem of how to speak, since moral and
social questions depend on how we decide to speak.
V. CONCLUSIONS
In the last few years I have been amused and irritated by the
spate of articles proving that time travel is a "conceptual im-
possibility." All these articles make the same mistake. They
take it to be enough to show that, if we start talking about time
travel, things go wrong with ordinary language in countless places
and in countless ways. For example, it makes no sense to prevent
an occurrence that is in the past, yet a time traveler about to
visit the Ice Age in his time machine may well take an overcoat
to keep from freezing to death several million years ago. Exactly
similar objections could have been raised against the notion of
there being only finitely many "places" prior to the development
of Riemannian geometry. It is precisely the existence of the
world-line language that makes us sure that all these apparently
insurmountable difficulties about "preventing," "expecting," etc.,
can be met. For example, the proper way to reformulate the

principle that it makes no sense to prevent an occurrence in the
past is to introduce the relativistic notion of proper time, i.e., time
along the world line, and to say that it makes no sense to prevent
an occurrence not in one's proper future. Also, even if an event
is in one 's proper future, but one already knows its outcome,
say, because it is in the objective past and has been recorded,
then it cannot be prevented (although one can try). For example
if reliable records show that an older self of you is going to freeze
to death two million years ago, then, try as you may, you will
not succeed in preventing this event. But this actually introduces
nothing new into human life; it is the analogue of the principle
that, if you know with certainty that something is going to happen
to you in the future, then, try as you may, you won't succeed in
forestalling it. It is just that there is a new way of knowing
with certainty that something is going to happen to you in your
proper future-namely, if your proper future happens to be also
your present past, then you may be able to know with certainty
what will happen to you by using records.
The principle of the transitivity of the relation earlier than
involves similar considerations. If the world line of the universe
as a whole happened to be a closed curve, then we should have to
abandon that principle altogether. As G6del has pointed out,3
there is no contradiction with General Relativity in the supposition
that the universe may have a closed world line.
The history of the causal principle is yet another case in point.
Before quantum mechanics, if we found that an event A was some-
times succeeded by an event B and sometimes by a different event
B', this was taken as conclusive proof that there were factors, say
C and C', differing in the two situations. Physicists abandoned
the principle that in such cases one should always postulate such
unknown factors only because a worked-out mathematical language
of an acausal character appeared on the scene, the mathematical
language of unitary transformation and projections that is used
in quantum mechanics. This is a mathematical way of represent-
ing phenomena, but it is not just a mathematical way of repre-
senting phenomena. Once again, it influences the way in which
ordinary men decide questions. When cases arise of the sort
Donellan foresaw in his paper, then ordinary men may choose one
way of speaking over another precisely because one way of speaking
links up with a scientific way of talking and thinking whereas the
8 Kurt Godel, " A Remark about the Relationship between Relativity

Theory and Idealistic Philosophy," in Paul Arthiir Sehilpp, ed., Albert Ein-
stein: Philosopher-Scientist.
670 TIIE JOURNAL OF PHILOSOPHY
other way of speaking does not link up with any coherent con-
ceptual system.
The causality case is analogous to the geometry case in that
the decision to preserve the older way of speaking-that is, to say,
whenever an event A appears to produce either of two different
events B and B', that there must be some hidden difference in the
attendant circumstances-involves postulating ghost entities. The
ghost entities in question are called "hidden variables" in the
literature of quantum mechanics.
I am inclined to think that the situation is not substantially
different in logic and mathematics. I believe that if I had the
time I could describe for you a case in which we would have a
choice between accepting a physical theory based upon a non-
standard logic, on the one hand, and retaininlg standard logic and
postulating hidden variables on the other. In this case, too, the
decision to retain the old logic is not merely the decision to keep
the meaning of certain words unchanged, for it has physical and
perhaps metaphysical consequences. In quantum mechanics, for
example, the customary interpretation says that an electron does
not have a definite position prior to a position measurement; the
position measurement causes the electron to take on suddenily the
property that we call its "position" (this is the so-called "quan-
tum jump"). Attempts to work out a theory of quantum jumps
and of measurement in quantum mechanics have beeni notorioulsy
unsuccessful to date. Recently it has been pointed out 4 that it is
entirely unnecessary to postulate the absence of sharp values prior
to measurement and the occurrence of quantum jumps, if we are
willing to regard quantum mechanics as a theory formalized
within a certain nonstandard logic, the modular logic proposed in
1935 by Birkhoff and von Neumanni, for precisely the purpose of
formalizing quantum mechanics.
There seems to be only one conclusion to come to, and I here-
with come to it. The distinction between statements niecessary
relative to a body of knowledge and statements contingent relative
to that body of knowledge is an important methodological dis-
tinction and should not be jettisoned. But the traditional philo-
sophical distinction between statements necessary in some eternal
sense and statements contingent in some eternal sense is not
workable. The rescuing move which consists in saying that if a
statement which appears to be necessary relative to a body of
4 This was pointed out by David Finkelstein in his lecture to the informal
subgroup on Measurement in Quantum Mechanics at the Boulder Symposilim
on Mathematical Physics, sponsored by the American Mathematical Society in
the summer of 1960.
APA ABSTRACTS 671
knowledge at one time is not necessary relative to the body

of knowledge at a later time, then it is not really the same state-
ment that is involved, that words have changed their meaning,
and that the old statement would still be a necessary truth if the
meanings of the words had been kept unchanged, is unsuccessful.
The rescuing move which consists in saying that such statements
were only mistaken to be necessary truths, that they were con-
tingent statements all along, and that their "necessity" was
"merely psychological" is just the other side of the same blunder.
For the difference between statements that can be overthrown
by merely conceiving of suitable experiments and statements that
can be overthrown only by conceiving of whole new theoretical
structures-sometimes structures, like Relativity and Quantum
Mechanics, that change our whole way of reasoning about nature
-is of logical and methodological significance, and not just of
psychological interest.
HILARY PUTNAM
MASSACHUSETTS INSTITUTE Or TECHNOLOGY
AMERICAN PHILOSOPHICAL ASSOCIATION

EASTERN DIVISION
Abstracts of papers to be read at the Fifty-ninth

Annual Meeting, December 27-29, 1962
ETHICS
THE DISTANCE BETWEEN HUME AND EMOTIVISM
BY JOHN SWEIGART
The place of Ilume, with respect to the status of moral state-

ments, has often been a point of controversy. In the broad tradi-
tion Hume has generally been considered a naturalist in ethics,
but with the rise of emotive theory Ayer and other emotivists, as
well as many who teach classes in ethics today, believe that Hume
well supports the noncognitivist position. This emotivist view
was founded, of course, by Carnap and Ayer, whose treatments
require recognition of a sharp dichotomy between the factual and
the moral.
However, it can be shown that for Hume matters of morality
are actually a certain kind of matters of fact. He held the basic
data of perceptions to be no less applicable to morality than to

Journal of Philosophy, Inc. The Journal of Philosophy

Uploaded by

Copyright:

Available Formats

Journal of Philosophy, Inc. The Journal of Philosophy

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Journal of Philosophy, Inc. The Journal of Philosophy

Uploaded by

Copyright:

Available Formats

Journal of Philosophy, Inc.

what in fact we would say. Rather we now s'ay somethingy about

T HIIE two statements that Donellan considered il hlis paper t are

* To be presented in a symiposium on "Necessary Truth'' at the fifty!-

are animials' is true by definition, notwithstanding the fact that

The "analyticity" of 'all cats are animals' or, what is closely

It is important to distinguish "analytic" truths of the sort 'all

to change the meaning. I agree with Donellan that this question

II. SYNTHETIC NECESSARY STATEMENTS

these properties (although we can approximate them as closely as

III. THE CLUSTER CHARACTER OF GEOMETRIC NOTIONS

Distance cannot be "operationally defined" as distance accord-

other one operational criterion. The criteria we have for distance

IV. TIME TRAVEL

I believe that an attempt to describe in ordinary language

t 2 W 1>\ / Oscar Smith World Line

E|I Apprtus World Line

tion," but the creatioli of two iiiacro-objects which separate. One

the normal order. I have no doubt whatsoever as to how any

ing succession of physical states. Thus, on my friend's account,

can be met. For example, the proper way to reformulate the

8 Kurt Godel, " A Remark about the Relationship between Relativity

knowledge at one time is not necessary relative to the body

AMERICAN PHILOSOPHICAL ASSOCIATION

Abstracts of papers to be read at the Fifty-ninth

THE DISTANCE BETWEEN HUME AND EMOTIVISM

The place of Ilume, with respect to the status of moral state-

You might also like