Methods in Metaphysics
Methods in Metaphysics
Methods in Metaphysics
identical with the earlier brain, then there is no impossibility in surviving the
destruction of one’s body even if one is identical with one’s body.
Either way, then, Richard Swinburne’s argument is unsound.
Example 2: Time Travel and Backwards Causation
A second example involves imagining that one is traveling in a time machine,
back into the past. Again, it might seem that one can form a series of vivid pictures
of what it would be like to travel back to the year 1900, and to have experiences
there. Does this then provide good support for the idea that time travel into the past
is logically possible?
If it does, it also provides good grounds for concluding that backward
causation – causation in which later events cause earlier ones – is logically possible.
But if backward causation is logically possible, then mustn’t causal loops be logically
possible. But causal loops may involve self-supporting causal loops, or, more
dramatically, self-undercutting causal loops. If such things are problematic, then the
question arises whether what one imagines when one thinks of oneself as imagining
what it would be like to travel back into the past really provides good grounds for
concluding that time travel is logically possible.
The crucial question to ask in such a case is this: “What exactly can one
vividly imagine in such a case?” Try imagining travelling back a day into the past.
As regards what one can vividly imagine – as contrasted with what one can
conceive – is it not the case that it does not differ from what one can vividly imagine
if one imagined oneself being transported instead to an alternative spatiotemporal
realm that is just like the way this world was yesterday? If this is right, then one’s
description of what one is imagining as imagining that one has travelled back into
the past goes beyond the content of the vivid image that one has formed. The vivid
image provides no support, then, for the claim that time travel back into the past is
logically possible.
Method 2: The appeal to what one cannot imagine.
Such appeals to what one cannot imagine, in cases where, if the state of
affairs in question did exist, it would be perceivable, are sometimes used to support
claims that that state of affairs is logically impossible.
Example: The Incompatibility of Different Color Properties
Can one imagine something that is both completely red and at least partly
green? It seems that one cannot, and many philosophers have taken that as grounds
for concluding that it is logically impossible for something to be both completely red
and at least partly green.
In thinking about this, it is worth asking a related question: Can one imagine
something that is reddish-green in color? Here, too, it seems that one cannot. But is
it logically impossible that something should be reddish-green? Given that
something can be reddish-blue, reddish-yellow, greenish-blue, and greenish yellow,
is it reasonable to think that it is logically impossible for something to be reddish-
green?
3
Notice that the sets that one typically thinks of do not have themselves as
members: the set of all horses does not, for example, contain itself as a member, for
all of its members are horses, and no set is a horse.
Can one think of any sets that do belong to themselves? Well, must not the
universal set, U, defined as the set that contains every set, be a set that contains itself
as a member?
Next, define the ‘Russell set’ – call it R – as the set of all and only sets that are
not members of themselves:
! ∈ ! = def. ‘~(! ∈ !).
Question: Is it the case that ! ∈ !? Does the Russell set belong to itself?
Well, either ! ∈ !, or ~(! ∈ !)
(1) Suppose then that ! ∈ !. It then follows immediately from the definition of the
Russell set R that ~(! ∈ !). So we have a contradiction.
(1) Suppose, on the other hand, that ~ ! ∈ ! , It then follows immediately from the
definition of the Russell set R that ! ∈ !. So once again we have a contradiction.
In either case, then, we have a contradiction. So if the Russell set R exists, if there is
a set that contains all and only those sets that are not members of themselves, a
contradiction is true.
Like most philosophers, Russell held that there are no true contradictions,
and so he concluded that there could not be a set of all and only sets that are not
members of themselves. (Graham Priest, an Australian philosopher, and advocate
of dialethic logic, maintains that the correct conclusion to draw is that there are some
true contradictions.)
Example 2: The non-existence of the universal set, U, defined as the set of all sets.
The following principle seems extremely plausible:
For any set S, and any property P, one can define the set T that consists of all and
only members of set S that have property P.
But if that principle is right, if the universal set, U, defined as the set containing
absolutely every set that exists, exists, then one can define the Russell set, R, since
6
that is just the subset of U that consist of the members of that set that have the
property of not belonging to themselves.
Accordingly, if universal set, U – the set of all sets – existed, so would the
Russell set, R – the set of all sets that are not members of themselves, and so we
would again have a contradiction.
Method 8: The proof of propositions using logic plus conceptual analysis.
Analytic truths can be defined as truths that are derivable from logical truths
in the narrow sense that is, from purely formal logical truths – by the substitution
of synonymous expressions.
Historical Note
Immanuel Kant defined analytical truths as subject-predicate propositions
whose predicate concept is contained in its subject – as, for example, in the
proposition that all bachelors are unmarried. This definition is, however, too
narrow, since it means that analytically true propositions are restricted to
propositions of a subject-predicate form, whereas analytically true propositions can
have other logical forms.
What are some examples of analytically true propositions that are not of
subject-predicate form? Here are some candidates:
(1) The proposition that 5 < 7
(2) The proposition that if x < y, and y < z, then x < z.
(3) The proposition that if x is a subset of y, and y is a subset of z, then x is a subset of
z.
(4) The proposition that if x is heavier than y, and y is heavier than z, then x is
heavier than z.
But how does one show that these are analytic when analytic truths are
defined, not as Kant did, but as truths that are derivable from logical truths in the
narrow sense – that is, from purely formal logical truths – by the substitution of
synonymous expressions? Here is an illustration in the case of the fourth of the
above propositions.
1. Introduce the term “outbalances”, where x outbalances y is defined as follows:
When x and y are placed on opposite sides of a balance scale, the side with x on goes
down, and the side with y on it goes up.
2. So defined, there can be possible worlds where x outbalances y, and y outbalances
z, but z outbalances x.
3. Now define the concept of being heavier than as follows:
“A is heavier than B” = def. “A outbalances B, and it is true that (x)(y)(z)(If x
outbalances y, and y outbalances z, then x outbalances z)”.
4. Then one can use the following proposition, which is a purely formal logical
truth:
If (1) A outbalances B and (x)(y)(z)(If x outbalances y, and y outbalances z, then x
outbalances z), and (2) If B outbalances C and (x)(y)(z)(If x outbalances y, and y
7
certainly the view of some philosophers – most notably Bas van Fraassen, who has
argued at length that the principle of inference to the best explanation should be
rejected.
I think that van Fraassen is right in maintaining that the principle of inference
to the best explanation cannot be a basic principle of inductive logic. But that does
not mean that it cannot be a sound principle, for it may be possible to derive the
principle of inference to the best explanation from fundamental principles of
inductive logic, together with an analysis of the notion of explanation, and quite
possibly also an analysis of the concept of causation.
But how would the more basic inductive principles be formulated? The
answer is given by the next method that metaphysicians have used:
Method 11: The use of a system of logical probability to show that certain things
are likely to be the case, or that certain things are unlikely to be the case.
Illustrations:
(1) The mathematician Thomas Bayes (1702-1761), in his posthumously published
"An Essay towards Solving a Problem in the Doctrine of Chances” (1763), starting
from the assumption that all propensities are equally likely, showed how statistical
information about the behavior of an object could justify probabilistic conclusions
concerning that object’s underlying propensities.
(2) If metaphysically robust, governing laws of nature are logically impossible, then
it is very unlikely that the world contains any cosmic regularities.
(3) If metaphysically robust, governing laws of nature are, on the other hand,
logically possible, then it can be very likely that the world does contain cosmic
regularities.
Method 12: The use of inference to the best account of the truth conditions of
statements of a certain type.
Example 1: David Lewis's account of the truth conditions of statements about
logical possibilities.
Lewis’s idea of possible worlds is as follows. Possible worlds are concrete
entities. They are concrete spatiotemporal worlds of concrete objects, where
different possible worlds are completely unrelated to one another, worlds that stand
in no temporal, spatial, causal, or other external relation to one another.
Given this notion of possible worlds, David Lewis argues that one was
justified in postulating the existence of such things since they are needed to provide
truthmakers for statements about logical possibilities. Consider, for example, the
proposition that the existence of a talking donkey is logically possible. Lewis’s view
is that what makes that proposition true is that there is a concrete, spatiotemporal
world that is not our world, and that is not connected spatially, temporally, or
causally, or in any other way to our world, and which contains a talking donkey.
What is a truthmaker? The idea of a truthmaker for a given proposition is the
idea of a state of affairs that serves to make that proposition true, where as state of
affairs consists either of some entity’s having a property, or two or more entities’
standing in some relation, or some more complex combination of these things.
9
is a necessarily true proposition, there aren’t possible worlds W and V such that p is
true in W but false in V.
Example 2: The postulation of second-order relations between universals to serve
as truthmakers for statements of laws of nature.
Method 13: The appeal to direct acquaintance.
Examples
(1) The existence of emergent, sensuous properties.
In philosophy of mind, one of the major issues concerns the existence of what
are referred to as ‘qualia’ (Singular: ‘quale’ – pronounced either ‘kwalay’ or
‘kwalee’.) Qualia are qualitative properties involved in experiences, such as colors in
visual perception. Their existence is controversial, since if they exist, it is hard to see
how they can be reduced to the particles, forces, properties, and relations that entre
into theories in physics.
(2) The existence of a flow of time.
Some people hold that in experience one is aware of the ‘flow’ or ‘passage’ of
time, where such flow or passage is something more than its merely being the case
that events stand in relations of temporal priority.
(3) Phenomenological approaches to philosophy, and phenomenological states.
Philosophers who adopt a phenomenological approach typically claim that
one can be directly aware of properties of psychological states other than qualia of
the sensory variety. Thus it is often claimed, for example, that one can be aware of
cognitive qualia, such as what it is like to be a belief, or a thought, or a desire, or that
one can be aware of the intentionality of some mental states, of what it is for a
mental state to be about some other state of affairs.
Method 14: The appeal to non-intellectual “seemings”
Mike Huemer, in his book Skepticism and the Veil of Perception, advances the
following principle:
(PC) If it seems to S as if P, then S has at least prima facie justification for believing
that P. (99)
If Huemer’s (PC) principle is correct, then one can appeal to non-intellectual
seemings in support of a variety of important metaphysical propositions, including:
(a) There is a mind-independent physical world.
(b) God exists.
(c) Humans have libertarian free will.
Comment: Everything turns upon whether (PC) is sound. I believe that it is not, for
a variety of reasons. For one thing, I do not think that there is any concept of
seeming that has all of the properties that Huemer’s (PC) principle requires.