The Evolution of Modern Metaphysics: Making Sense of Things (71 page)

The material mode of speech is in Carnap’s view widespread. To take another of his own examples, if I say, ‘This books treats of Africa,’ I have adopted the material mode. Had I said, ‘This book contains the word “Africa”,’ I would have adopted the formal mode.
⁴⁵
The latter would have been less misleading because ‘it is not a quality of Africa to be treated of in that book…. It is only a quality of the word “Africa” to be contained in the book’ (
Philosophy
, p. 65). However, there is no suggestion in this case that I am engaged in metaphysics. That suggestion arises only when I am practising the kind of logical analysis that is characteristic of philosophy, that is when I am clarifying the workings of some linguistic framework. If I
then
use the material mode, I give the impression, not only that I am making claims about the entities that form the subject matter of the framework, but that I am concerned with what it is for them to
be
those entities. An especially striking variation on this theme is the case in which I present reasons for or against adopting a particular linguistic framework, that is for answering a particular external question in a certain way, and, by using the material mode, give the impression that I am addressing an internal question. Thus I might say, ‘The indispensability of mathematics to physics is a good reason for believing that real numbers exist.’
⁴⁶
The less misleading formal mode counterpart of this is: ‘The indispensability of mathematics to physics is a good reason for adopting a linguistic framework that allows us to use the expression “real number” as we do.’
⁴⁷
In a nutshell, then, while Carnap is at once the staunchest advocate of philosophy and the fiercest critic of metaphysics, he recognizes that, because of the material mode of speech,
the one can all too easily appear as the other.
⁴⁸

(b) The Implications for Metaphysics on My Conception of Metaphysics

What are the implications of Carnap’s views for metaphysics when metaphysics is conceived as the most general attempt to make sense of
things?

As in Wittgenstein’s case (
Ch. 9
, §§2 and 8, and
Ch. 10
, §3) this question sub-divides according to what is meant by making sense of things. Its answers, too, are in line with those that it receives in Wittgenstein’s case. If making sense of things is understood as arriving at truths about the world, then we can say that, by Carnap’s lights, metaphysics is the most general of the natural sciences, and quite distinct from philosophy. If making sense of things is understood as providing for clarity in our thinking, which in Carnap’s terms means clarifying linguistic frameworks that we use, or might use, then we can say that by Carnap’s lights metaphysics is philosophy of the most general kind – where the generality in question is the generality of the concepts involved in the frameworks being clarified. On this second account the only truths that it is the prerogative of metaphysicians to assert are analytic truths determined by the rules of some framework. Metaphysicians are not in the business of discovering truths about the world. They are in the
business of supplying suitably lucid frameworks from which to select the best for couching truths about the world that have already been discovered, or that may yet be discovered.

One interesting feature of this second account is that it allows for clear Carnapian answers to the three questions that
I posed in §6 of the Introduction:

• (the Transcendence Question) there is no scope for metaphysicians to make sense of what is transcendent, if what is transcendent means what transcends sense experience, for there is no such sense to be made

• (the Creativity Question) metaphysics is an inherently creative enterprise

and, most significantly perhaps, redressing what I take to be the principal failing of the later Wittgenstein (see
Ch. 10
, §6, and §3(e)),

• (the Novelty Question) metaphysicians can, and sometimes should, make sense of things in ways that are radically new.
⁴⁹

There is also an important account that is intermediate between the two just considered. On this intermediate account, just as on the second, metaphysicians are in the business of supplying suitably lucid frameworks from which to select the best for couching truths about the world.
But they are also in the business of making the selection
. This is something they must do in the light of our various purposes and in the light of whatever truths about the world have already been discovered. In this respect the intermediate account is like the first: it gives metaphysicians some responsibility for telling the eventual story about what the world is like, albeit responsibility that they can discharge without dirtying their hands. The full significance of this intermediate account will become apparent at the very end of the chapter.

(c) Carnap on Alternative Conceptions of Metaphysics

I have distinguished between where Carnap’s views leave metaphysics on his own conception and where they leave it on mine. But I do not mean to suggest that Carnap is oblivious to the possibility of alternative conceptions to his own; nor indeed that he is oblivious to the possibility of alternative conceptions to his own that may lead to a sympathetic reassessment of the very material that, on his own conception, he is forced to decry. Here we do well to recall the point that I emphasized in §1: for a statement to lack what logical positivists call ‘literal’ meaning, in other words for it to fail to express a
truth or a falsehood, is by no means for it to lack all kinds of meaning. Some extremely significant possibilities are left open: witness the fact that logical positivists deny literal meaning to any statement of commendation or condemnation, whether ethical, aesthetic, or of any other kind; to any statement that serves as an optative, or a supplication, or a curse; to any statement of joy, wonder, or horror at the beauty and ravages of creation; and to countless other statements besides.
⁵⁰
What if one of the thinkers under attack from Carnap insists that his metaphysics is not an attempt to discover and state transcendent truths about the essence of things; that it is not an attempt to discover and state
truths
at all; that it is an attempt, rather, to convey some of these other kinds of meaning, using statements that belong to some of these other classes; and that it is therefore best viewed, not as a scientific exercise, but as an artistic exercise?
⁵¹

Carnap has plenty to say in response to such a suggestion. (See e.g.
Philosophy
, Lecture I, §5, and ‘Elimination’, §7.) But the gist of his response is gratifyingly straightforward and uncompromising. The metaphysical material that he has in his sights may well, he concedes, be an attempt to convey some of these other kinds of meaning; but if it is, then it is still subject to two serious criticisms. First, it is a
dissembling
attempt: it presents itself as what it is not, a system of logically interrelated truth-evaluable statements. Second, and much more damning, it is a
very poor
attempt: it is put to shame by other, more overtly artistic achievements. ‘Metaphysicians,’ Carnap writes with withering sarcasm, ‘are musicians without musical ability’ (‘Elimination’, p. 80).

6. Tu Quoque?

A common objection to logical positivism is that it cannot survive its own critique. Like Hume, Kant, and the early Wittgenstein, logical positivists seem to be under threat when they apply their own principles to their own espousal of them. Consider, in particular, the cardinal principle that a
statement has literal meaning if and only if either it or its negation expresses either an empirical truth or an analytic truth. Let us follow A.J. Ayer in calling this ‘the verification principle’ (Ayer (
1992
), p. 149).
⁵²
What is the status of the verification principle itself? Does it express an empirical truth? Or does it express an analytic truth? Or does it not express a truth at all? All three options appear unattractive: the first because it seems plainly inadequate to the force of the principle, in particular by casting it as a mere contingency; the second because it suggests that there is nothing more to the principle than how certain words are used; and the third because it means that, unless the principle expresses a falsehood, it lacks any literal meaning, which, despite the many other kinds of meaning that logical positivists acknowledge, seems intolerable.

I shall argue in this section that there is indeed an embarrassment for logical positivists here – or rather, hereabouts. For it is not quite where the common objection locates it. And it is of an altogether subtler kind than the common objection would lead us to suppose.

The first thing to appreciate is that all three options can actually be made to look attractive. All three have at some time been espoused by philosophers of a broadly positivist persuasion. If logical positivists do experience any embarrassment here, then, initially at least, it is liable to be felt as an embarrassment of riches.

Let us call the three options, respectively, the ‘empirical’ option, the ‘analytic’ option, and the ‘no-truth’ option. And let us first consider the empirical option. That is perhaps, at first blush, the least appealing of the three. But it is noteworthy that, if we allow ourselves the anachronism of saying where Hume stood on this issue, then this is the option he espoused. In Hume’s view, it was a matter of fact, confirmed by experience, that all our ideas are composed of simple ideas copied from previous impressions. Had that not been the case, then we could (for instance) have had, just as Descartes thought we did have, an innate idea of God, surpassing all our impressions in such a way as to furnish us with the knowledge that God exists, a truth that would have qualified neither as a relation of ideas nor as a matter of fact in Hume’s terms.

In the next chapter we shall see Quine trying to distil from logical positivism the basic empiricism in which he takes its true worth to lie. And although this will involve his challenging the very terms of the current dispute, it is noteworthy, again, that he will count the empiricism with which he is left as itself an empirical truth. (See §3 of the next chapter.)

It is the other two options, however, that have found greatest favour with logical positivists. On the analytic option, the verification principle expresses a truth in virtue of how those who are party to this dispute use the relevant words. The worry about this is that it makes the principle trivial. But does it? In any troublesome sense? After all, logical positivists cast the whole of pure mathematics as analytic. There is room, given our psychological limitations, for a kind of substantiveness in such matters – as even Hume knew. The analytic option is certainly defensible. It is in fact Carnap’s view (see
Philosophy
, Lecture I, §7, and Lectures II and III passim).

But Carnap also expressly considers the no-truth option. Or rather, he considers one version of the no-truth option, which is what he (wrongly) thinks we find in Wittgenstein’s
Tractatus
.
⁵³
This is to regard the verification principle as being of a piece with the metaphysical nonsense that it is supposed to help us banish, serving its purpose in some curiously indirect, self-abnegating way. Carnap sees no hope for such a view (
Logical Syntax
, pp. 283–284, and
Philosophy
, pp. 37–38).

He has not however considered the most compelling form of the no-truth option. The most compelling form of the no-truth option is something much less exotic: to regard the principle as a prescriptive definition of the expression ‘literal meaning’. (This is not so different from Carnap’s own view, of course, because such a definition can be regarded as part of a linguistic framework that straightway allows for the expression of analytic truths in accord with it.) Those who espouse this option will point to the importance of prescriptive definitions in other sciences, as a way of allaying any concern that such a thing cannot be of any scientific interest. This is the option that Carl Hempel espouses (Hempel (
1959
), p. 125). It is also the option that A.J. Ayer espouses. Admittedly, in
Language, Truth and Logic
Ayer is not clear about this. Though he calls the principle a ‘definition’, he also insists that ‘it is not supposed to be arbitrary’ (Ayer (
1971
), pp. 20–21). Later, however, in response to a direct challenge to say what status he takes the principle to have, he says that he was never inclined to regard the principle either as empirical
or
as analytic, then continues:

Happily not everything that the verification principle failed to license was cast by me on the pyre of metaphysics. In my treatment of ethics, I made provision for prescriptive statements…. Accordingly, in …
Language, Truth and Logic
, I treated the verification principle as a prescriptive definition. (Ayer (
1992
), p. 149)

We shall return shortly to some discomfort that Ayer nevertheless feels with this. But notice first that the choice between the no-truth option and the analytic option can be cast in terms of the Limit Argument which appeared
in
Chapter 5
, §8. That argument purported to reveal an incoherence in the project of drawing a limit (in the sense of a limitation) to what we can make sense of – which is, in effect, the logical positivists’ project.
⁵⁴
We have already seen that, on a suitably thick conception of sense-making, the Limit Argument can be resisted. Such a conception brings both its premises, the Limit-Drawing Principle and the Division Principle, into question. Now the logical positivists have, in effect, a thick conception of sense-making, whereby making sense of something involves producing statements about it which have literal meaning and which either express empirically knowable truths about that thing or express analytic truths that make provision for such knowledge. In these terms, those who espouse the no-truth option could say that they have drawn a limit to what we can make sense of without making sense of that limit,
contra
the Limit-Drawing Principle; while those who espouse the analytic option could say that they have made sense of a limit to what we can make sense of without making sense of anything on its ‘far’ side,
contra
the Division Principle.