||Date: 04/13/00 08:47:11
Subject: Metaphor does not constitute good philosophical practice.
|Reference to Sartre and poetry lead to
consideration of this very important question.
Can metaphor play a serious and useful role in philosophical reasoning?
Sartre and the more poetic of philosophers (or indeed philosophical of poets) would
consent that it does. However without a clear understanding of metaphor the debate will be
fruitless. I will present a brief analysis of metaphors and then draw some conclusions
from this analysis.
(I) An analysis of metaphors.
Use of a metaphor involves using some words whose literal meanings are not intended but,
sensitive to the context, have some other content. Usually there is no a simple way of
conveying the meaning directly, without recourse to metaphor. But note that use of the
metaphoric content of the same word can be different, for example
Men are dogs
could be used to mean a number of things, here are two possibilities:
(a) Men are predatory and are at a loss when not in a pack
(b) Men are lower forms of life deserving of lower forms of treatment.
Exactly what is meant is determined by the context of the utterance and even then it is
not entirely clear.
I shall confine my analysis here to metaphorical expressions which are denoting in a broad
sense: noun phrases (allowing abstract nouns) and the like.
Let w be a metaphorical expression and let S be the sentence in which the metaphor of w
occurs. Let S[x/w] be the sentence resulting in replacing w for x in S. Then S is
There is an x and a class of predicates C (on entities of the type of w and x) such that
C(w) and C(x) and 'S[x/w]' is true.
The point of the existential generalisation is to remove any unnecessary content contained
in w that renders S too strong (in most cases literally false). Applying this to the
example above we have, w = 'dogs':
There is an x and a class of predicates C such that C(dogs) and C(x) and 'All men are x'
The context determines the contents of the class C, C is there to capture 'what is meant
by the metaphor'. Notice that the predicates contained in C may have no corresponding
expressions of the language in question.
(II) Do metaphors make for good philosophical practice.
Notice that in the analysis above there is no systematic method of determining the
contents of the class of predicates C, we have to know from the context. It is plain to
see that for the most part it is impossible get be confident that we know exactly what the
speaker means. At least, no where near as confident as we would be if he had said what he
meant right directly. If the dogs metaphor means (a) above then simply stating (a) is far
more helpful than using the metaphor.
To make matters worse, in written philosophical texts the context is often unclear. How
can it be otherwise? The whole point of a book is that it inputs information to its reader
independently of where he is or what he is doing. Constraints on length, intelligibility
and interest force a book to underspecify the context of its words, there is no avoiding
The purpose of philosophical reasoning should be to lead people to some truth. In order to
do this the reasoning must be clear and intelligible. Metaphors, in general, are far from
clear and intelligible. If my analysis of metaphor is correct then the understanding of a
metaphor involves knowledge of which class of predicates C to apply in existentially
generalising out the metaphorical expression. But this is work that is done by the person
to whom the philosophy is subjected, not by the philosophy itself. The philosophy may as
well tell him to work out all the answers for himself, hardly an appropriate attitude.
Possibly I have been a little unfair. It is true that in some cases the members of C are
easily found or that there is no other method of conveying the intended content. But such
cases are rare, especially in the case of well constructed deep philosophical analysis. If
the concepts involved are far removed from natural discussion/thought then C will
certainly not be obvious. But if there is no other way of expressing what is meant then I
cannot see how the philosopher can have access to this mysterious true meaning of the
metaphor. Even it he has such access then the metaphor should be used as a teaching tool
rather than a platform for philosophy, for if C is not known then the metaphor is little
better than meaningless.
In general C is not known, so in general metaphors are little better than meaningless
(from the point of view of attaining clearly expressed ultimate truth).
Web Forum © by Bravenet Web Services
Philosophy Study Group Forum