How Relativity Contradicts Presentism

Simon Saunders
Faculty of Philosophy, 10 Merton St., Oxford

1 Introduction
But this picture of a ”block universe”, composed of a timeless
web of ”world-lines”in a four-dimensional space, however strongly
suggested by the theory of relativity, is a piece of gratuitous meta-
physics. Since the concept of change, of something happening, is
an inseparable component of the common-sense concept of time
and a necessary component of the scientist’s view of reality, it is
quite out of the question that theoretical physics should require
us to hold the Eleatic view that nothing happens in ”the objective
world”. Here, as so often in the philosophy of science, a useful lim-
itation in the form of representation is mistaken for a de…ciency
of the universe (Black, 1962).

The theory of relativity has excited more philosophical commentary, and ex-
erted more in‡uence in mainstream philosophy, than any scienti…c theory,
with the possible exception of Newton’s theory of gravity. But it is a re-
markable fact that its in‡uence on metaphysics proper has been somewhat
marginal. That is probably a testimony to the anti-metaphysical attitude
that characterized so much philosophy in the last century, certainly in the
Anglo-American tradition, and certainly among more scienti…cally-minded
philosophers. Although the hey-day of logical empiricism is long-since past,
philosophers of physics have continued to remain cool to metaphysics. Since
they remain the ones best suited to explain the implications of relativity the-
ory for the philosophy of time, if they …nd no interesting links between these
disciplines, metaphysicians are unlikely to look for them.
I make this observation (and I promise to say no more in this vein) because
relativity theory, and speci…cally the special theory of relativity, does I believe
have a simple and direct bearing on a perennial question in the philosophy of
time. It would, I believe, settle this question, were special relativity the whole
of the story. I shall say something about the broader perspective of quantum
theory and general relativity at the end.

1
 What traditional question does special relativity decide on? It is whether
reality - what exists - is a four-dimensional web of worldlines (a “block uni-
verse”), or something less. There are competing versions of what this “some-
thing less” might be, but they are variations on what I shall call presentism,
the view that only the present is real. The argument is simple, and I will state
it with the minimum of technicality. Although, I say, it has been largely ig-
nored, versions of it have been stated before; it was stated in brief by Gödel,1
and at great length by Putnam.2 But this latter version of the argument has
been roundly condemned by Stein. This dispute between Putnam and Stein
is in fact well-known in the philosophy of physics literature, but insofar as
there is a consensus on it, it is that Stein was in the right.3 I will come on to
this dispute in due course.

2 Presentism
“Presentism”, as I shall understand it, is the thesis that the present is all that
exists. But this needs some unpacking. It is intended to be something more
than a platitude, and surely the present is all that exists now, as the future is
all that will exist, and the past all that did exist. Who will argue with these
claims?
One way to get clearer on the presentist thesis is to say that it is meant
tenselessly; that the copula in the sentence “the present is all that exists” is
not itself tensed. But that is a doubtful maneuver. Very often the presentist
will go on to deny that there is any meaningful, irreducibly tenseless use of
“is”, at least when it comes to the physical world; and that passe Mellor
and others, it is equally possible to give tensed truth conditions for tenseless
sentences as tenseless ones for tensed. Exercises in the philosophy of language
do not seem to be settling anything.
But nebulous though it is, there is surely something about the presentist’s
position which is perfectly clear: it is intended to be a thesis about what to
count as real. It is a realist thesis. It is a claim about temporal reality which
is supposed to hold independent of our state of knowledge or beliefs. We and
our works are not what the thesis is about; the presentist is making a claim
1
For Gödel the point was obvious; he immediately went on to consider the situation in
Einstein’s theory of gravity, speci…cally in the light of his rotating universe solutions to the
…eld equations of general relativity. The latter argument has been carefully analysed by a
number of authors: see Stein (1994), Earman (1996), and Savitt (1997).
2
Putnam (1967); versions of it were also stated by Rietdijk (1966) and Maxwell (1986),
and these Stein has also criticised; but their interests were a little oblique to our topic,
and their handling of it more muddled. Rietdijk’s treatment, in particular, was just careful
enough, and just muddled enough, to be conclusively refuted (see Torretti 1983, pp. 250-51l;
see also Landsberg 1970 p. 1146-47). I shall con…ne myself almost entirely to Putnam’s
argument, and to Stein’s response to it (Stein 1978, 1991).
3
Among those who have endorsed Stein’s objections to Putnam’s argument, see Clifton
and Hogarth (1995), Dickson (1998 p.165-73), and Shimony (1993), with quali…cations).
So far as I know only Callender (1998) has expressed any real reservations; but see also
Saunders (1996, 1998).

2
about reality, not about what we know or say about it.
The claim can also be put in negative terms, as the view which is opposed
to the tenseless view of time (according to which all events exist on a par,
regardless of whether we consider them as past, present, or future). On the
tenseless view, talk of events as past, present or future is really talk about
ourselves, of the relation of events to how we are momentarily arranged. The
word “now” is like the word “here”; mention of “future” and “past” is like
pointing this way and that way in space.
That is all that is needed to bring out the con‡ict with special relativity.
If presentism is a thesis about ontology, and says that existence consists of a
three-dimensional spatial reality; if, in elaboration of this thesis, it opposes
the tenseless view of time, and denies that talk of events as past, present or
future is elliptical talk about the relation of events to our momentary selves:
then it contradicts special relativity. It contradicts it in the sense that it
implies that special relativity is badly de…cient as a fundamental theory of
the world.
Of course special relativity is an empirical theory. One might take the
view that presentism is concerned with a level or reality which is beyond the
reach of experimental methods.4 But I do not grant that physics is so limited
in scope, or that metaphysics can …nd anything deeper. But I will not argue
for either of these claims here.

3 How Relativity Contradicts Presentism

The di¢ culty posed by special relativity is extremely simple. According to
presentism, all that is physically real is the present - a system of physical
events all of which are simultaneous with each other. No other events are
real. Precisely what this system of events may be, now, as I snap my …ngers,
may not be known to me; but there is a fact of the matter as to what it is,
and it is a universal fact which embraces us all. It is an intersubjective reality
- now, as a snap my …ngers - and it is a reality which contains us only as an
incidental part. But even if one knew all that there is to know, consistent
with special relativity, one would not be able to say what this system of events
might be. According to presentism, therefore, special relativity is radically
de…cient as a description of reality. It is blind to the sequencing of what is
physically real.
There is no such problem in the Newtonian case. There, knowing all there
is to know, the set of events simultaneous with this event - as I snap my
…ngers - is unambiguously de…ned. It is all and only those events absolutely
simultaneous with it. In that theory there is postulated a relation, absolute
simultaneity, which partitions events into disjoint classes, namely instants of
time. It does so democratically: no one event of each class is singled out in
the de…nition of the partition. To be precise, this relationship of absolute
4
For a metaphysician who hovers uncomfortably between this view, and the view that
anyway relativity does not contradict presentism, see Smith (1993 p.2-4.).

3
simultaneity is re‡exive, symmetric and transitive, so instants of time are
equivalence classes of events. This relation, moreover, plays a crucial role in
the subsequent de…nition of the dynamical laws of motion (and of suitable
initial data for those laws); it could hardly play a more fundamental role in
Newton’s theory. And we ourselves, and our momentary and spatial arrange-
ments, are manifestly incidental to its de…nition; we are incidental to what
each reality consists in. Newtonian theory, gratifyingly for the presentist, is
attentive to what is physically real. Not so special relativity.
The argument is so simple that it speaks for itself. No technical result
is needed: it is of the essence of the theory of special relativity that absolute
simultaneity as such does not exist. Everyone knows that there is nothing
else to replace it - there is no other non-trivial symmetric and transitive
relation intrinsic to Minkowski space. Of course, making reference to the
matter content of spacetime as well, there may well be methods for de…ning a
partitioning of spacetime into spaces (for de…ning global instants, as required
by presentism), but none of them are likely to claim any fundamental status.
It is unlikely that any can be taken seriously, if we are concerned with the
de…nition of the totality of what is physically real. Only given a matter
distribution of exceptional symmetry - for example, a stream of particles all
moving inertially, with zero relative velocities - would a slicing of spacetime
into spaces at di¤erent times (a foliation of spacetime) be obviously privileged.
The presentist will literally need a river for there to be time, according to his
metaphysics.5
One need only consider the realistic candidates to see the di¢ culty. Given
any one inertial (straight) worldline, one can de…ne a natural slicing of Minkowski
space into spaces at di¤erent times, namely into the set of (parallel) timeslices
orthogonal to it. But how can the whole of reality - what is physically real
- depend on a single worldline? What is this thing which has this special
privilege - or who is it - that has this extraordinary status? In fact, from
a physically realistic point of view, there are no objects which always move
inertially. Nor, in an in…nite universe, does anyone know how to construct
such a world-line: one cannot de…ne the centre of mass of the universe as a
whole if it is in…nite; and one cannot make do with any part of it without
privileging that part. Why that part, and not some other? Even in the …nite
case the option of the centre of mass is not very attractive, involving as it does
messy and arbitrary conventions (there is no unique de…nition of the centre of
mass frame in the …nite case, again in contrast to the situation in Newtonian
theory). The presentist cannot be neutral on this score; to suppose it can
be settled by convention is precisely to take the view that what is real - the
5
For a more general criterion, less dependent on symmetry, suppose that hypersurfaces
are de…ned as everywhere orthogonal to the integral curves of the four-velocity …eld of the
‡uid. Then the ‡uid had better be irrotational (lacking “twist”), if the surfaces are not to
intersect. (For a simple geometric illustration, think of a twisted rope; it cannot be cut so
as to cut each strand of it orthogonally.) Gödel was led to his rotating universe solutions
to the Einstein …eld equations by considerations of just this sort: see Malement (1994).

4
breaking down of spacetime into spaces at di¤erent times - is not of funda-
mental import, but a matter of convenience, a matter of convention. That is
precisely what the tenseless theory says. This is not an option available to
the presentist.
In the general case, and in an in…nite universe, in practise one seizes on a
segment of a particular worldline or worldtube of some body (or bodies slowly
moving with respect to each other). At any point of it a spacelike hyperplane
can be constructed to which it is orthogonal. If approximately inertial, a
family of hyperplanes can be constructed which are non-intersecting, at least
locally. This is the technique used in positional astronomy: Ephemeris Time
is precisely such a system for partitioning events in the history of the solar
system into spaces at di¤erent times (more precisely, it is a based on the
relative con…gurations of the Earth-Moon-Sun system). By its means, if there
ever does come into being a community of astronauts in space, moving about
the solar system, it will still be possible to agree on what events in whose
lives get to take place at the same time. The criteria will be public and
intersubjective. No particular person will be singled out in counting what
is real (of what belongs to which moment in time). But the community as
a whole is singled out. The Earth-Moon-Sun system acquires a very special
status. As the basis for the criterion of what is real, it is parochial. It will
hardly do for metaphysics. It is as embarrassing as Newton’s hypothesis as
to what is really at absolute rest (the centre of mass of the Solar System).
What is so special about the Earth, Moon and Sun?
The presentist has little option but to hold out for some as yet unknown
criterion for determining what is physically real - for splitting Minkowski
spacetime into spaces at di¤erent times. Special relativity, the presentist must
conclude, is radically incomplete. But the alternative view is that progress in
physics has counted against presentism. Physical theories were once compat-
ible with it, but then they were not.

4 The Dispute Between Putnam and Stein

This argument is I believe unassailable, but it is similar to Putnam’s, and
Putnam’s has been roundly condemned. On what grounds?
First Putnam’s argument. He considers what he calls “the view of the
man on the street”. It is the same as presentism: it is the view that “all (and
only) things which exist now are real”. Putnam now assumes:

1. I-now am real. (Of course, this assumption changes each time

I announce that I am making it, since “I-now”refers to a di¤erent
instantaneous “me”.)
II. At least one other observer is real, and it is possible for this
other observer to be in motion relative to me. (Putnam 1967 p.
240.)

5
He also assumes what he calls “the principle that There Are No Privileged
Observers”:

III. If is the case that all and only the things that stand in a
certain relation R to me-now are real, and you-now are real, then
it is also the case that all and only the things that stand in the
relation R to you-now are real. (ibid p.240).

Putnam gives no argument for this principle, and in his subsequent use of it
only transitivity is explicitly mentioned; but by the letter of III it is clear that
R must also be symmetric. And so it should be, if of all the events which are
real, no one of them is to be privileged. If not, then if we start from a …duciary
event x (there is at least one real event, by I), and de…ne the others that are
real as those which stand in relation R to x, then the set fy : Rxyg that we
end up with cannot also be de…ned by starting with another element of this
set. In other words, each set will have to be speci…ed by the relationship R
and a particular element of it (not any element of it); so for each set one
element of it would have to be privileged. Likewise if R is symmetric but not
transitive. Since R is surely re‡exive, “no privilege” exactly forces R to be
an equivalence relation. When the elements in question are events in the lives
of observers - person-stages - the principle is “no privileged person-stages”.
(“No privileged observers” has a better ring.)
Finally, Putnam requires:

IV. R is de…nable in special relativity.

The similarity of Putnam’s argument to the one I have given should be

perfectly clear. Indeed, given a partitioning of Minkowski space M , one can
always de…ne an equivalence relation in its terms. Let the timeslices of M
- the partitions - be labelled by a parameter t, denote fMt g: Then R is the
co-membership relation

R = f< x; y >; 9t such that x 2 Mt and y 2 Mt g: (1)

It is obviously an equivalence relation, and, given fMt g, it is obviously de…n-
able in special relativity. The question that remains is how the partitioning
of Minkowski space was arrived at. If by a relationship on M - essentially re-
versing the procedure just sketched - then we are back to Putnam’s approach.
But whatever the method, a principle analogous to III will apply. I would add
that not only can it not privilege any particular person, but it had better not
privilege any particular community, either. A metaphysics which is explicitly
community bound is not worthy of the name.
Putnam considered only partitions de…ned by a relationship R, not the
more general case. A quick result follows on the narrow reading of IV, that
R must be de…ned in terms of the geometry of M . Then, trivially, there are
only the two equivalence relations, f< x; y >; x; y 2 M g - all events are real

6
- and f< x; y >; x = yg - only the …duciary event x is real. If these are the
only possibilities, it is not hard to see which we should choose.
But Putnam did allow that R may be de…ned by reference to the matter
distribution. The result still follows, failing any special symmetries, but whilst
intuitively plausible it is harder to prove. Putnam did not prove it.6 He
only considered the most obvious candidates for it, in particular Einstein
synchrony, denote Ein; this obviously fails, because for points on an arbitrary
collection of timelike lines it is neither symmetric nor transitive. He also
considered the worldline-independent relation of past causal connectibility,
denote Con. Let x y i¤ x is in or on the past light cone of y; then Con
is the relation f< x; y >; x yg: Although re‡exive and transitive, it is not
symmetric. This too is of no use to the presentist.
Putnam did not put it in quite these terms, however. In this latter part
of his paper, where he introduced Con, he spoke rather of truth-values of
statements, not of the reality of events. He considered Con in this context,
as the suggestion that only statements about events in the lower half of my
light-cone have a truth value. Of this he remarked:

This last move, however, ‡agrantly violates the idea that there
are no Privileged Observers. Why should a statement’s having or
not having a truth value depend upon the relation of the events
referred to in the statement to just one special human being, me?
(ibid p.246)

The point is not entirely self-evident. A statement may fail to have a truth
value because it may fail to refer to anything, and whether or not a statement
refers (and what it refers to) may well depend on whether it has a relation
to a particular human being - so much is true of any statement containing an
explicit or implicit indexical. And Putnam did not make it clear that one can
hardly insist on his requirement III, of “No Privileged Observers”, as it was
originally stated, in this new context, for no-one will demand symmetry in
this case. By shifting to the question of what statements have truth-values,
it is surely intended that we include statements referring to past events as
well as to present ones. Putting it in non-linguistic terms, it is not the thesis
of presentism, but rather the the thesis that only the present and the past
is real. Call it “possibilism”7 If de…ned by a relation, the possibilist will
obviously not want it to be symmetric. If an earlier event is real in relation
to a later one, that should not imply that the later one is real in relation to
the earlier. Given possibilism, we should not expect to obtain a democracy
of timelike-separated observers.
6
It is an immediate corollary of the result of Clifton and Hogarth (1996), namely that
in the absence of symmetries, there is no worldline-dependent relation which is so much
as transitive. Perversely, they took this to lend support to Stein’s argument, rather than
Putnam’s (for it generalizes Stein’s proof that Con is the only intrinsically-de…nable non-
trivial transitive relation on M ; see below).
7
In line with Savitt’s terminology (Savitt 1998).

7
 But possibilists will still insist on the remaining requirement built into
III, that if you are real to me, then what is real to you is real to me as well.
Transitivity is necessary here as before. And now one might be lulled into
thinking that transitivity is su¢ cient for possibilism: that possibilism di¤ers
from presentism only in that it demands a weaker version of III.
The waters are now seriously muddied, for of course transitivity is not
enough for possibilism. The constraint is operating, here as before, that what
is present - meaning, according to possibilism, what is the boundary of all
that is physically real - is an intersubjective and non-parochial a¤air. But
this latter constraint can no longer be imposed by formal conditions on R, if
R relates all and only the events which the possibilist considers as real.
Were transitivity all that is required by the possibilist, the relation Con
would be just the ticket. This is exactly Stein’s response to Putnam: it
is enough to de…ne a relation “already happened”, and it is enough if this
relation is transitive and de…nable in terms of the intrinsic geometry of M
(re‡exivity as before can be stipulated). Con does the job. Stein made this
clear in his …rst paper on the subject (Stein 1968, p.5); in his second he went
on to prove that Con is essentially the only such transitive relation on M
(Stein 1991).
As for Putnam’s error, Stein located it in the passage just cited:

The answer is that “having or not having a truth value”, in this

question, must be understood classically to mean “at a given time”
.... but “at a given time”is not a relativistically invariant notion,
and the question of de…niteness of truth value, to make sense at
all for Einstein-Minkowski space-time, has to be interpreted as
meaning “de…niteness at a given space-time point (or event) - to
be vivid, “de…niteness for me now”. The “Privileged Observer”
(or, rather, privileged event) is - in e¤ect- named in the ques-
tion, and therefore has every right to be considered germane to
the answer. Putnam’s objection has an exact analogue, whose in-
appropriateness is plain, in the pre-relativistic case; namely, the
question “why should a statement’s having or not having a truth
value depend upon the relation of the events referred to in the
statement of just one special time, now ?” (Stein 1968 p.15).

According to Stein, Putnam presupposes notions that are simply not available
in special relativity. He has failed to take note of the changed situation in that
context, that “de…niteness to the present”has to be replaced by ”de…niteness
at a given space-time point”. Clearly, on making this replacement, one cannot
rule out reference to a particular point, no more than in the non-relativistic
case can one rule out reference to a particular instant.
All well and good, but clearly this changed situation is simply no longer
hospitable to presentism. In the pre-relativistic case “an instant” is a public
reality, on which all who were included in it could intersubjectively agree. It
o¤ered room enough for an account of the whole of reality (it was a plausible

8
reality, at least for some). But nothing like this can be said of a spacetime
point. If this is really all that special relativity provides, short of the whole
of Minkowski spacetime, we have no option but to opt for the latter. If it is
true that “the now” can only be a spacetime point - as Stein seems to imply
- then presentism is obviously untenable. A single point in spacetime cannot
be all that is physically real.
Stein is in fact perfectly indi¤erent to presentism.8 He makes this clear in
the paragraph that follows:

...in Einstein-Minkowski space-time an event’s present is consti-

tuted by itself alone. In this theory, therefore, the present tense
can never be applied correctly to “foreign” objects. This is at
bottom a consequence (and a fairly obvious one) of our adopt-
ing relativistically invariant language - since, as we know, there
is no relativistically invariant notion of simultaneity. The appear-
ance of paradox only con…rms that the space-time of Einstein and
Minkowski is quite di¤erent from pre-relativistic space-time. (ibid
p.15)

The tenseless point of view is so natural that it is not even worthy of comment:
of course the present of an event is constituted by itself alone; what else is one
to think in special relativity! Stein is not concerned with the metaphysical
thesis of presentism. He is impatient with talk of the view of the man in the
street - he …nds it curious that special relativity should be held hostage to
that - because for Stein, it is a fairly obvious consequence of relativity theory
that an event’s present is constituted by itself alone (the second of the two
trivial equivalence relations). Obviously the presentist’s position is untenable,
given special relativity. Precisely so; the only question is why Stein disputed
Putnam’s conclusions.
There is one more respect in which the shift of topic, to possibilism and
de…niteness of truth value, may have led to confusion. What is the status
of future contingencies? Rietdijk advanced an argument similar in certain
respects to Putnam’s, but used it to conclude that special relativity implies
determinism. Almost twenty years later, and without reference to either Put-
nam or Rietdijk, so too did Maxwell. It was this that prompted Stein’s second
paper on the subject, in which he proved that Con was essentially unique. It
was titled “On Relativity Theory and the Openness of the Future”. Clifton
and Hogarth’s generalization of this result, to include worldline-dependent
relations (in the absence of symmetries), was entitled “The De…nability of
Objective Becoming in Minkowski Spacetime”. Black shifted without com-
ment from the view that reality is a 4-dimensional whole, to the view that
change is unreal; Shimony, approvingly citing Stein’s response to Putnam, was
8
Compare Torretti (1983 p.250): ‘Each event is (tenselessly) real and determinate, in
this absolute sense, as its own worldpoint. No tensed, frame-indepenent statement can add
to it or detract from its reality...’ I do not believe that the presentist, or anyone even loosely
associated with presentism, can agree with this statementl.

9
concerned to deny that change was illusory. Evidently there are two further
questions at stake in all this: one, whether indeterminism is consistent with
special relativity, and two, whether the tenseless view of time is committed
to the view that change, or becoming, is unreal. But one can answer these
questions either way and yet reject presentism;9 Putnam made no mention of
either of them.
The one clear respect in which Stein ‡atly denies a step in the argument
as I hanve given it is this:

...in e¤ect what he calls the principle of No Privileged Observers

just requires R to be an equivalence relations. But such a re-
quirement has in fact no connection with the privilegedness of
observers; and it is moreover extremely inappropriate to Einstein-
Minkowski space-time - in which (unlike pre-relativistic space-
time, with its temporal decomposition) there are no intrinsic geo-
metrical partitions into equivalence classes at all, besides the two
trivial ones...(ibid p.19)

Stein denies that to partition a set into classes by means of a relation, in

such a way that each class can be de…ned independent of the choice of any
particular element of that class, the relation must be an equivalence relation.
I say he is mistaken.10
How might Con be used to de…ne a partitioning of Minkowski space?
There is a near neighbour to it which does the job quite easily. Let x y i¤
x is on the surface of the past light-cone to y. De…ne Berk as the relation
f< x; y >; y xg: Berkeley’s criterion, indeed, was that to be real (to x) is
to be seen (by x, so to be on the past light-cone to x).11 Let us now partition
up M into disjoint sets in the obvious way, as a nesting of lightcones. We
do not obtain a foliation in this way, for the partitions are not spacelike
surfaces, but put that aside. The real di¢ culty is that each partition has a
distinguished point - the apex of each cone - and the partitioning as a whole
clearly distinguishes a unique timelike line - the locus of these distinguished
points. Such is the price for using a relation which is not an equivalence
relation. Now suppose this partitioning has the metaphysical signi…cance
9
The presentist, of course, may disagree; so too will “A-theorists”, who …nd compelling
McTaggart’s argument that the B-series is inadequate to the description of change. I have
more sympathy for Maxwell’s claim, that special relativity poses problems for indetermi-
nateness (see my 1996, 1998, for further discussion).
10
Compare Sklar, ‘...why one should think that such a doctrine of “No Privileged Ob-
servers” would lead one immediately to a¢ rm the transitivity of “reality for”, given that
one has already relativized such previously nonrelative doctrines as that of simultaneity, is
beyond me.’ (Sklar 1981 p.130). The answer is that Putnam’s aim was exactly to show that
once “simultaneity with”is relativized in a way which is non-symmetric, and non-transitive,
then “reality for” must be relativized simillarly: in which case it is unacceptable.
11
Of course Berkeley was not proposing that all that exists is what is visible to a unique
x - a unique and particular event in space and time; rather, he proposed that it is what is
visible to God.

10
accorded to it by the presentist: a given one of them is to de…ne the whole
of what is physially real, a mater on which all will agree. So what is this
worldline - the worldline of what or of whom - that is to have this extraordinary
signi…cance? It can hardly be one of the obvious candidates from the tenseless
point of view (one’s own worldline, or the worldline of the earth, or of the
centre of mass of the sun, earth and moon), wherein no ontological signi…cance
resides (for on the tenseless point of view the fundamental ontological reality
is the whole of Minkowski space, and everything in it). We have been over
this before
On the tenseless view, all events are real; the sign…cance of the partitioning
is quite di¤erent. If asked which relationship gives the “correct”partitioning,
the answer will depend on how the question is construed. Berk is a natural
candidate, but so is Ein - in each case referred to a particular timelike line.
If the question is construed as what to count as the past, or what to count
as the present, most of us will settle for a foliation of spacetime, on which we
can reach community-wide consent. It is Ein, and Ephemeris Time, that we
will choose.12

5 Prospects for Presentism

Putnam made mistakes in his argument, but that does not explain its recep-
tion. Dickson has diagnosed the fault with it - as we have seen Stein says the
same - as the sheer inappropriateness, in special relativity, of Putnam’s as-
sumptions; that “before special relativity can have anything to say about the
doctrines in question, they must be expressed in a language that is meaning-
ful in a relativistic context”(Dickson p.170). Stein’s sympathy with Carnap’s
philosophy is well-known (Carnap was committed to the doctrine of incom-
mensurability long before Kuhn): evidently it retains its appeal for Dickson
as well.
But we can meet this objection head-on. The requirement of intersubjec-
tivity is certainly relativistically meaningful. The inference from that to the
requirement that R be an equivalence relation was independent of spacetime
considerations altogether. To infer from that that there must exist a privi-
leged foliation to Minkowski spacetime is precisely to spell out the doctrine
in relativistically invariant terms. The fact that this doctrine is then ruled
out by special relativity was precisely Putnam’s point.
At the other extreme is the de‡ationary, commonsensical reading of pre-
sentism, along the lines sketched by Savitt and Moratio. According to them,
it is obvious that the past did exist, that the future will exist, and that only
the present is (presently) real. What, I wonder, is obvious to them? That
only my momentary self is (presently) real? That what is (presently) real is
what is related by Ein? Or is it Con, or is it Berk? They may of course reply
that it makes no di¤erence which, but that is not a de‡ationary reading of
anything.
12
For a recent defence of the virtues of Ein over Berk, see Sarkar and Stachel (1999).

11
 Somewhere in between is quietism, the view that special relativity can-
not adjudicate on the matter unless supplemented by some form or other of
veri…cationism (Sklar 1981). So it was with Einstein’s elimination of ether;
his approach to special relativity was explicitly veri…cationist, and only given
this could he conclude that the ether did not exist. Indeed, the presentist has
only to suppose that there is, in fact, a unique resting frame, albeit that no
measurement can tell us what it is: this is the foliation he has been after all
along. He can claim it serves an explanatory function in physics as well, citing
Lorentz, for whom it was a “matter of taste”, and citing Bell, for whom it was
the best way “to teach the subject.”13 But here I think Stein is exactly right
when he says (Stein 1991 p.155) that if it is veri…cationism that is needed,
to do away with an absolute “up” in the face of rotational symmetry, then it
is a form of it that is perfectly defensible, that we should all of us embrace:
the form of it which eliminates an absolute state of rest in the face of the
relativity principle. And, obviously, one can draw the same conclusion on the
basis of a realist view of Minkowski space (not of course available to Einstein
circa 1905).
I have maintained that presentism is a substantive position that places
clear demands on the theory of special relativity. They are demands which
I do not think can be met, consistent with that theory. This fact needs to
be clearly appreciated, if there is to be movement on this subject. Movement
there is, as soon as we consider the wider perspective of Einstein’s theory of
gravity, and dynamics proper.
Of course general relativity, just like the special theory, is committed to the
principle of arbitrariness of foliation. Nevertheless, for an important class of
spacetime models - hyperbolically complete spacetimes, for which the Cauchy
problem is soluble - there is a natural de…nition of a global foliation, which
has a number of desirable, dynamical properties. It is essentially unique; it is
what is actually used in numerical calculations in geometrodynamics; it also
has links to a number of open theoretical questions, particularly questions
concerning the nature of scale in the classical theory.
I give this example, called York time,14 after its discover James York,
not because I am convinced it is fundamental, in classical theory, but as
an example of the new avenues that are opened up as soon as one considers
gravitational dynamics proper.15 Certainly none of the arguments I have given
13
This, the so-called “Lorentz pedagogy”, has recently been defended by Brown and
Pooley (2001); but in their view it is not committed to the view that any one frame of
reference is truly the resting frame; they suppose that the forces which yield the contraction
and dillation e¤ects may be explanatory, even if there is no fact of the matter as to what
they really are.
14
For a simple exposition of its uses in Hamiltonian formulations of general relativity, see
Wald (1984).
15
And to repair my unhappy neglect of it in my (1996). The class of hyperbolically-
complete spacetimes is of course only a sector of the full theory, and there are already
problems with this in the quantum case, if black-hole evaporation is anything to go by; but
it is an important sector all the same, and black-hole evaporation equally causes problems
for unitarity.

12
here tell against it. And ultimately, of course, one must look to a quantum
theory of gravity, where the interpretation of time in canonical approaches
to quantization is anyway in dispute. If one throws into the equation the
foundational problems of quantum mechanics, and the evident di¢ culty, in
that context, of de…ning a Lorentz-covariant stochastic dynamics,16 it is clear
that here there is everything to play for. But we are not about to make
progress with any of these …elds if the metaphysics of presentism, in the most
simple case of classical special relativity, is still in dispute. In this most simple
case, I have argued, it can …nally be laid to rest.

References
Bell, J. (1987), ‘How to Teach Special Relativity’, in Speakable and Un-
speakable in Quantum Mechanics, Cambridge University Press: Cambridge.
Black, M. (1962), ‘Review of G.J. Whitrow’s “The Natural Philosophy of
Time”’, Scienti…c American, CCVI, p.181-2.
Brown, H., and O. Pooley (2000), ‘The Origin of the Spacetime Metric: Bell’s
“Lorentzian Pedagogy”and its Signi…cance in General Relativity’, in Physics
Meets Philosophy at the Planck Length, C. Callender and N. Huggett, eds.,
Cambridge University Press: Cambridge.
Callender, C. (1998), ‘Shedding Light on Time’, Philosophy of Science (Pro-
ceedings), 67, S587-S599.
Clifton, R. and M. Hogarth (1995), ‘The De…nability of Objective Becoming
in Minkowski Spacetime’, Synthese 103, 355-87.
M. Dickson (1998), ‘Digression: The Block-Universe Argument’, in Quantum
Chance and Non-Locality, Cambridge University Press: Cambridge..
Earman, J. (1996), Shrieks, Bangs,Whimpers and Grunts, Oxford University
Press: Oxford.
Gödel, K. (1949), ‘A Remark About the Relationship Between Relativity
Theory and Idealistic Philosophy’, in Albert Einstein, Philosopher-Scientist,
A. Schilpp, ed., Open Court, La Salle.
Landsberg, , P. T. (1970), ‘Time in Statistical Physics and Special Relativity’,
Studium Generale, 23, pp. 1108-59).
Malement, D. (1994), ‘Commentary’, in The Collected Papers of Kurt Godel,
Vol.3, S. Fe¤erman, J. Dawson, W. Goldfarb, C. Parsons, and R. Solovay,
eds., Oxford University Press: Oxford.
Maxwell, N. (1985), ‘Are Probabilism and Special Relativity Incompatible?’,
Philosophy of Science, 52, 23-43.
Putnam, H. (1967), ‘Time and Physical Geometry’, Journal of Philosophy,
64, 240-47, reprinted in Philosophical Papers, Vol. 1, Cambridge University
Press: Cambridge, 1975.
16
See my (1996) for illustrations of the di¢ culties, and the reasons why they do not arise in
the Galilean-covariant case. For the di¢ culty in generalizating the pilot-wave theory to the
relativisic case, see my (1999); the latter are less directly linked to the foliation-arbitrariness,
however.

13
Rietdijk, C. (1966), ‘A Rigorous Proof of Determinism Derived from the Spe-
cial Theory of Relativity’, Philosophy of Science, 33, 341-4.
Sarkar, S., and J. Stachel (1999), ‘Did Malament Prove the Non-Conventionality
of Simultaneity in the Special Theory of Relativity?’, Philosophy of Science,
66, 208 -20.
Saunders, S. (1996), ‘Time, Quantum Mechanics, and Tense’, Synthese, 107,
19-53.
Saunders, S. (1999), ‘The “Beables” of Relativistic Pilot-Wave Theory’, in
From Physics to Philosophy, J. Butter…eld, and C. Pagonis, eds., Cambridge
University Press: Cambridge.
Saunders, S. (1998), ‘Tense and Indeterminateness’, Philosophy of Science
(Proceedings), 67, S600-611.
Savitt, S. (1994), ‘The Replacement of Time’, Australasian Journal of Phi-
losophy, 72, 463-74.
Savitt, S. (2000), ‘There’s No Time Like the Present (in Minkowski Space-
time)’, Philosophy of Science (Proceedings), 67, S563-74.
Shimony, A. (1993), ‘The Transient Now’, in Search for a Naturalistic World
View, Vol.2, Cambridge University Press: Cambridge.
Sklar, L. (1981), ‘Time, Reality, and Relativity’, in Reduction, Time and
Reality, R. Healey, ed., Cambridge University Press: Cambridge.
Smith, Q. (1993), ‘Introduction’, in The New Theory of Time, J. Oaklander
and Q. Smith, (eds.), Yale University Press: Newhaven.
Stein, H. (1968), ‘On Einstein-Minkowski Space-Time’, Journal of Philoso-
phy, 65, 5-23.
Stein, H. (1991), ‘On Relativity Theory and the Openness of the Future’,
Philosophy of Science, 58, 147-67
Stein, H. (1994), ‘Commentary’, in The Collected Papers of Kurt Godel,
Vol.3., S. Fe¤erman, J. Dawson, W. Goldfarb, C. Parsons, and R. Solovay,
eds., Oxford University Press: Oxford.
Torretti, R. (1983), Relativity and Geometry, Pergamon Press: Oxford.
Wald, R. (1984), General Relativity, Chicago University Press: Chicago.

14