The Evolution of Modern Metaphysics: Making Sense of Things (68 page)

41
For a much more modest role that nonsense may play in fostering such understanding, see again §464, cited in n. 23.

42
I should note in advance that the underlying exegesis is by no means uncontroversial. For an approach to Wittgenstein that does much to forestall the objection, see the various essays in Baker (
2004a
), esp. those in Parts IA and IIA. For a rejoinder, see Hacker (
2007
).

43
Insofar as he would be prepared to venture any explicit statement
about
transcendental idealism, then no doubt he would say, just as his former self would have said, or at least just as the ingenuous version of his former self would have said, that it is a piece of metaphysical nonsense. Not that I mean to suggest that it is no longer even on his radar. Much of §§90ff., for example, where he is engaging with his former self, is an attempt to counteract what once made it, for him, such enticing nonsense. Some of what he says in the course of this discussion may itself sound transcendentally idealistic: see e.g. §§103, 104, and 114. In fact, however, he is merely emphasizing the contingency of our sense-making. To hear what he says in these sections as
already
transcendentally idealistic – to hear it as already making limitations out of essential features – is to commit the very error that he is trying to guard against: see e.g. §108.

44
They are advanced most famously in Williams (
2006k
). They are considered sympathetically in Lear (
1984
), (
1986
), Garver (
1994
), and Forster (
2004
). They are explored in Anscombe (
1981
) and Sacks (
2000
),
Ch. 6
. They are attacked in Bolton (
1982
), Malcolm (
1982
), and, less directly, McDowell (
1993
). What follows draws on Moore (
1997a
),
Ch. 6
, §3, and
Ch. 7
, §4, where I also discuss them.

45
Cf.
Tractatus
, 3.263.

46
§122.

47
Cf. §108.

48
This excludes (for instance) lions: see p. 223.

49
Cf.
Zettel
, §§313 and 314.

50
Cf. §207 and Davidson (
1984a
). Cf. also Lear (
1984
).

51
Here I am appealing to the Division Principle which features in the Limit Argument (see
Ch. 5
, §8), with sense-making understood in a suitably thin way.

52
Or at any rate such, for current purposes, we may treat as our rules. For discussion of how much this glosses over – itself interestingly relevant to what is to come – see Fausto-Sterling (
2000
).

53
See
Ch. 5
, Appendix, for the definition of empirical idealism.

54
Cf. Wittgenstein’s impatience at §§380 and 381 with the question, ‘How do I know that this colour is red?’, asked with inappropriate philosophical intent.

55
I do not claim, however, that he never himself feels the force of the temptation first to raise such pseudo-questions and then to follow where they lead. See e.g. pp. 226–227, where he shows some discomfort with simply saying, ‘Even though everybody believed that twice two was five it would still be four.’ His discomfort seems to me unwarranted. The situation is analogous to that which I identified parenthetically in
Ch. 1
, §3, where Descartes, over-cautiously in my view, refused to rule out the possibility of God’s making one plus two unequal to three. I would further contend, and have tried to argue elsewhere, that the temptation to raise such pseudo-questions, and eventually to embrace some version of transcendental idealism, is in part a temptation to try to express inexpressible understanding of ours, such as our understanding of what being an aunt consists in: see Moore (
1997a
), Chs 7–9. If I am right, this reinforces the connections between the early work and the later work. (On the idea that there can be such a thing as understanding what being an aunt consists in, even though there is no such thing as saying what being an aunt consists in, cf. §78.)

56
See again nn. 8 and 9.

57
There are deep connections between this problem and a problem that I signalled for Descartes in
Chapter 1
, §3: that we need to be able to tell introspectively when we have a clear and distinct perception, and that having a clear and distinct perception is to be understood normatively.

58
Cf. the hesitancy voiced in both that section of
Zettel
and the next.

59
Cf. Frege’s view, outlined in
Ch. 8
, §6.

60
Cf. §§54, 123, and 133, and
Culture
, pp. 86–87.

61
In fact, just before this remark about mathematics, he says that philosophy, by which he means good philosophy, ‘leaves
everything
as it is’ (emphasis added). Everything? Well, everything untainted by bad philosophy. And here, of course, the threat of circularity is again manifest.

62
See Moore (
2001a
),
Ch. 4
, §§1 and 2.

63
Wittgenstein himself has reservations about this: see esp.
Remarks
, Pt V. We shall return to this issue in
Ch. 14
, §§2 and 3.

64
See Moore (
2001a
),
Ch. 8
, §3, and
Ch. 10
, §§3 and 4. (A ‘real’ number is a number that can be expressed using an infinite decimal expansion, a ‘natural’ number a non-negative whole number.)

65
Cf. the distinction that he draws between the ‘calculus’ and the ‘prose’ in mathematical discourse, in
Vienna Circle
, p. 149.

66
I myself see nothing wrong with it. This is related to nn. 11 and 14 of the previous chapter. See also n. 4 above. And see Friedrich Waismann (
1959
), p. 359, where Waismann discusses the same example in a broadly Wittgensteinian framework, but with what seems to me an altogether cooler head than Wittgenstein.

67
For a fuller discussion of this example, see Moore (
2011
). For a superb discussion of the more general issues raised in both this section and the next, see Williams (
2006n
), §4.

68
For the record, there are also clear answers to the Transcendence Question and the Creativity Question, suitably construed. For there is a clear sense in which Wittgenstein will deny that metaphysicians can make sense of what is transcendent (§126), and a clear sense in which he will deny that their enterprise is one of discovery (§3).

69
See above, §1. And see in particular n. 6 for references to his pejorative uses of ‘philosophy’.

70
See in particular n. 88 of that section.

71
See in particular
Ch. 21
, §§6 and 7(c). See also Conclusion, §5.

Chapter 11 Carnap The Elimination of Metaphysics?

1. Logical Positivism

Here is a cartoon sketch. Hume was appalled by the metaphysical excesses of his predecessors. He opposed them with a radical empiricism. But Kant thought that Hume’s empiricism was too radical. On Hume’s account, earlier metaphysicians had not only professed to know what they could not know, they had professed to know what they could not even think. Kant believed that they were very often guilty as charged in the first of these respects, much less often in the second. He opposed them with something more subtle. But the subtleties of Kant’s view, combined with its own uneasy relation to itself, meant that many of his successors felt that they now had license to try to make sense of things in ways that Hume would have regarded as far more egregious than anything he had been trying to combat in the first place. And so it was that, in the twentieth century, within the analytic tradition, there was a neo-Humean backlash, a reversion to a radical empiricism that could be used to mount a full-scale semantic attack on these new excesses, reducing them to the status of literal meaninglessness. This was the movement known as logical positivism.
¹

That this movement should have arisen in analytic philosophy is hardly surprising. The Humean attention to sense that it demanded was very much of a piece with the attention to sense that had come to be one of the defining features, if not the defining feature, of analytic philosophy. By the same token logical positivists were able to make use of various analytic tools, in executing their Humean project, that had not been available to Hume himself, most notably the tools of the new formal logic that Frege had established.
In sum – and remember, this is a cartoon sketch – logical positivism was Humeanism made analytic.
²
,
³

As a species of Humeanism it involved a descendant of Hume’s distinction between relations of ideas and matters of fact. As an analytic movement it involved a version of Frege’s distinction between analytic truths and synthetic truths. Indeed, the one was the other. But how
could
the one be the other, without important questions being begged against Kant? Frege, when he introduced his distinction between analytic truths and synthetic truths, claimed that he was merely appropriating Kant’s distinction (
Ch. 8
, §2). And Kant, when he first introduced that same distinction, introduced it as one of a pair of distinctions, the other being the distinction between truths that are knowable
a priori
and truths that are not, his very point being that Hume’s single distinction had been unequal to the task of accounting for every kind of truth (
Ch. 5
, §3). ‘What,’ Kant would have asked – as for that matter would Frege, who likewise introduced his distinction alongside this second one – ‘about the synthetic
a priori
?’

It is unfair, however, to suggest that the logical positivists simply begged this question. They addressed it; and they gave it a non-Kantian answer. They strenuously denied that there
is
any such category.
⁴
This was precisely one of the hallmarks of their reversion to Hume. The single contrast between that which is analytically true, and therefore knowable
a priori
, and that which is knowable only by suitable appeal to sense experience, and therefore synthetically true, was as much as they thought was needed.

And it gave logical positivism its distinctive stamp. Logical positivists insisted that unless a given statement expressed a truth that could in principle be known in one or other of these two ways, or a falsehood whose negation could in principle be known in one or other of these two ways, it did not express a truth or a falsehood at all and lacked any literal meaning.
⁵
This so-called verification theory of meaning arguably entailed that each of the following statements lacked any literal meaning:

Pure being and pure nothing are the same.

The good is more identical than the beautiful.

Time passes at one second per second.

God moves in mysterious ways.

Bananas look purple when they are not being observed.

No human being is ever justified in killing another.

This miscellany of examples should both ring bells from previous chapters and, particularly in the case of the last few, indicate how destructive the view was capable of being. To be sure, it is important not to exaggerate the destructiveness. Logical positivists were denying that certain statements had a certain kind of meaning, what I have followed them in calling ‘literal’ meaning. That is, they were denying that these statements expressed truths or falsehoods. But this left open the possibility that some of them expressed something else, hence that they had meaning of some other kind. In particular it left open the possibility that some of them expressed feelings, prescriptions, or proscriptions – any of which the last statement in particular might be thought to express.
⁶
It left open the possibility that, on some reasonable conception of what it is to make sense of things, some of these statements could contribute to doing just that. Nonetheless, it is clear how logical positivists took their view to inflict the damage on earlier metaphysical excesses that they wanted it to inflict.

2. Carnap’s Version of Logical Positivism. Linguistic Frameworks