Académique Documents
Professionnel Documents
Culture Documents
By Nathaniel J. Kan
cannot exist; at least, no temporal relations exist. That it is illogical to say that one event
is simultaneous to, earlier than or later than another event, or that one event is in the
present, past or future. McTaggart offers no alternative theory of time; he only claims
time's unreality. In this paper I will attempt to show that there are significant logical holes
in his proof.
McTaggart begins by positing that there are two ways to view the ordering of
time. First, the series of times ordered running through the past up until the present,
through the present, and then continuing on through the future. This he terms the A-
series. The second series, which he calls the B-series, is the series that orders events by
"earlier than" and "later than," beginning (presumably) with the earliest time, and running
McTaggart goes on to state that there cannot be time without change. I believe by
this he means that if all the particles in the universe were to maintain the same positions,
there would be no time, as this seems to be the only way his proposition could be true
(we will return to this). Therefore, the B-series alone cannot constitute time, because
nothing in the B-series can ever change; any event will always stand in the same (B-
series) temporal relations to every other event, and the only way for events to change is
for them to change in temporal relations. Only one class of characteristics of events can
change: the temporal relations with regards to the present. These are the basic
Kan 2
I maintain that time is possible without change. The type of change that
particles were to maintain the same positions, we would still have time. Imagine a world
where by a quantum fluke all particles on one side of the universe were to maintain their
positions for one Planck's time, and then continue their normal course of motion. We
would not consider time to have "frozen" for this one instant in half the universe,
especially because normal motion continued for the other half of the universe. Now
imagine a world where by another fluke all particles in the universe maintain their
positions for one Planck's time, and then continue their motion. By the same logic would
we half to believe that time continued as normal for that one Planck's time, it would
simply be inconceivable to us that it did so, because there would be no way to tell.
constitute time. While it seems true that time would be inconceivable without any change
at all, here we (and he as well) are talking about change in the spatial distributions of
matter and energy, not in temporal relationships. Imagine a universe where nothing
besides temporal relations change (Classically all particles stay in the same position;
Quantum mechanically all particles continue to maintain the same waveforms which
never collapse). The only thing that would change in this world is which moment is the
present, and thus, every moment's relationship to the present. But time would also be
inconceivable in this world, and for the same reason McTaggart says time's existence
requires change. In this world also there would be no way of ever knowing that time was
progressing. Thus an A-series alone could not constitute time, at least not by McTaggart's
Kan 3
definition. By claiming first that time depends on the existence of change (in spatial
relationships of matter) and then saying that the only the A-series offers the possibility of
In this way McTaggart claims that an A-series is necessary for time. Past, present
and future are relations to something outside the time series, for if they were relations
within the time series they would necessarily be fixed, because the relation between two
moments never changes. But past, present and future are mutually exclusive relationships
to stand in. This seems explainable because it can be said that a moment "is present, will
be past, and has been future" (McTaggart 32). But then if some moment T is present,
there is no past time in which T will be past, but T is past in some future moment T'. But
at some still later moment T", T' will be a past moment. And thus T is a past moment at
a past moment. The same can be worked out for T' and T" and so on, ad infinitum.
What if T having the present relation and T' having the past relation happen not at
the same time, but in succession? Thus T is a present moment in the present, T' is a past
moment in the future, and therefore we need the past/present/future relations to relate the
McTaggart claims this leads to another infinite series, which does not allow the first
moment T to escape from contradiction. Thus, because there can be no A-series, time is
the requirement of an infinite series, but not one that is vicious. McTaggart writes:
The attribution of the characteristics past, present, and future to the terms
them successively. This means … that they have them in relation to terms
And, since this continues infinitely, the first set of terms never escapes
Thus moments stand in temporal relations to the 'now' entity that is outside the A-series
(McTaggart labels this X), and these temporal relations stand in temporal relations, and
these temporal relations stand in temporal relations, and so on. However, let us examine
the situation from before, with T and T'. T is in the 'present' relation with X, which
means we would say "T is happening now." What relation does this 'present' relation
have with X? 'Present', as if we ask the question, "When will it be true that it holds?" the
Likewise consider T'. In the B-series, T' is later than T. While T' stands in the
past relation to X, it does so in the future. And when does the relationship of "T' will be a
past moment in the future" hold? In the present and in the past, but not in the far future.
And because of the need to discriminate thus the need for another relation continues.
A parallel can be draw here to Bradley's Regress, but in this case instead of an
external relations between objects, we have temporal relations between moments in time.
Why does the presence of infinite layers of temporal relations necessarily mean those
temporal relations do not exist? While arguing against Bradley's theory we can claim that
the existence of an external relation is a "brute" fact and that there is no need of an
explanation for why it holds1. This does not seem to apply here, because the temporal
relations are necessary to define the succession of relations on the level above. One
argument that might apply, however, is that there does not seem to be any good reason
for why there shouldn't be infinite levels of temporal relations. While it might be said that
infinite levels of temporal relations are inconclusive, I would claim that infinite levels of
temporal relations do absolve the primary level (that level regarding moments) of
1
Van Inwagen, Peter. Metaphysics. Westview Press: 2002. Pg 36.
Kan 6
contradiction, as each level has an explanation of it's succession: the level beneath it.
One final important thing that McTaggart fails to point out is that his argument is
dependent on the truth of logical determinism, which he does not make a case for and
which does not seem to be clearly resolved yet (if it ever will be). Because he assumes
that the only type of change possible is that of temporal relations in the A-series, then
under his system the fact that an event will occur is always a fact, even before it's
occurrence. In order to decide if McTaggart's argument is true, the reader would have to
In conclusion, I do not propose that time is either real or unreal. However, the
initial assumption that there cannot be time without change in temporal relationships is
false; McTaggart is first claiming that there must be spatial relation change for time and
then claiming that spatial relation change is not change, only temporal relation change is.
Even if we accept these questionable basic assumptions, the conclusion that the infinite
series of temporal relations leads to a contradiction is lacking evidence. For these reasons