The Evolution of Modern Metaphysics: Making Sense of Things (27 page)

To be sure, if effort is to be expended on the creation of new concepts, then it needs a rationale. But there are all sorts of rationales that would be congenial to Hume. It is instructive in this connection to consider some stories that he himself tells about concept creation, such as the story that he tells about how we come by our concept of a promise. ‘In order … to distinguish [interested commerce from disinterested commerce],’ Hume writes, ‘there is a
certain form of words
invented for the former, by which we bind ourselves to the performance of any action. This form of words constitutes what we call a
promise
’ (
Treatise
, pp. 521–522, emphasis in original).
⁴⁶
The rationale in this case is thoroughly practical. But so too, on Hume’s conception, is the rationale for every operation of the human mind whereby we make sense of things.
⁴⁷
There is certainly nothing in the practicality of concept creation to prevent his acknowledging it as a significant component in any worthwhile attempt to make sense of things at the highest level of generality.

Hume’s view does admittedly preclude conceptual innovation of the most radical kind, the kind that is not restricted to working with old material. For the ultimate grounding for any conceptual innovation must, on Hume’s view, be the imaginative reconfiguration of extant simple ideas to form new complex ideas. Even so, if the view is correct, then mathematics shows how far such reconfiguration can take us.

In sum, then, it takes neither scepticism about Hume’s empiricism nor susceptibility to talk of the transcendent to wonder whether there are, in the libraries, untold volumes of metaphysics – including perhaps volumes in which concepts are created by using the various resonances of the word ‘substance’ – that should be snatched back from the flames.

Appendix: Scepticism About Human Reasoning

In the
Enquiry
Hume treats the distinction that he draws between relations of ideas and matters of fact as reasonably robust. But we should pause to note some earlier scepticism, which he voices in the
Treatise
, about how robust any such distinction can be, or at least about what form any such robustness can take (I.iv.1). Hume there reminds us that, whenever
something is established by a lengthy chain of pure deductive reasoning, as is the case for much of what we accept in mathematics, then our confidence in it must be based, in part, on confidence in various relevant matters of fact about the proper functioning of our associated faculties (p. 180; cf. p. 144). And, as we reflect on this, we cannot but wonder, not just whether our faculties have been functioning properly on this or that occasion, but whether there is any such thing as a proper functioning of our faculties; whether they are faculties for arriving at true beliefs at all. For Hume, this is itself a question about a matter of fact. It is a question about the relation between ‘our reason … considered as a kind of cause’ and ‘truth[, its] … natural effect’ (p. 180). It is also clearly a variation on Descartes’ Reflective Question (see
Ch. 1
, §3). And Hume, unlike Descartes, sees no hope for meeting such scepticism by redeploying the very faculties whose certification is in question.
⁴⁸
The more we use our resources to corroborate our resources, the greater, not the less, will be our concern about those resources. Instead, we must appeal to the natural impulses that guide us where our beliefs concerning any other matters of fact are concerned, impulses that lead us to accept that whose acceptance we could never justify by processes of pure deductive reasoning. ‘Whoever has taken the pains to refute the cavils of this total scepticism,’ Hume comments, ‘has … endeavour’d by arguments to establish a faculty, which nature has antecedently implanted in the mind’ (p. 183, emphasis removed).
⁴⁹

This is of interest not only in its own right, but also in relation to themes that will dominate subsequent chapters. Wittgenstein, in his later work, is acutely conscious of the ‘matters of fact’ that make it possible for us to have practices of mathematical calculation and reasoning.
⁵⁰
Yet he is also very keen to separate mathematical questions from questions concerning any such ‘matters of fact’. This is part of what he means when he denies, as he does in his earlier work too, that calculation is a kind of experiment.
⁵¹
More than once in his manuscripts, at a point where he denies this, he also adds, as a kind of item of marginalia, ‘The limits of empiricism’ (Wittgenstein (
1967a
), pp. 197 and 379).

We should beware, however, of seeing anything in Humean empiricism as a target. Hume still
has
his distinction between relations of ideas – in particular, relations of mathematical ideas – and matters of fact. He is as
keen as Wittgenstein to ward off any confusion of the former with the latter. And Hume, too – witness both the scepticism to which we have just seen him respond and his response – knows all about ‘the limits of empiricism’.
⁵²
Hume’s error, I believe, lies not in his failing to see where the limits of empiricism lie, but in his
failing to see how much of metaphysics lies within them.

1
Throughout this chapter I use the following abbreviations for Hume’s works: ‘Abstract’ for Hume (
1978b
);
Enquiry
for Hume (
1975a
); 2nd
Enquiry
for Hume (
1975b
);
Natural History of Religion
for Hume (
1976
); and
Treatise
for Hume (
1978a
). In giving non-page references to the
Treatise
I adopt the convention whereby ‘I.iv.2’ names Bk I, Pt IV, §II, and so forth.

2
I do not mean to suggest that Hume’s reflections were entirely without precedent. Locke (
1965
) and Berkeley (
1962a
) were important precursors.

3
For a good account, see Williams (
2006e
). In
Treatise
, p. 272, Hume wrote, ‘Generally speaking, the errors in religion are dangerous; those in philosophy only ridiculous.’

4
We might well expect, in view of the example of substance with which we began, that another case in point would be the doctrine of transubstantiation. In fact, in his
Enquiry
, Hume directed scepticism of the more modest kind at this doctrine, borrowing an argument due to John Tillotson (p. 109). We must however wonder, as so often in reading Hume, whether he was understating his case for various rhetorical and prudential reasons. (Incidentally, to say that religious thought does not concern matters of genuine belief, or does not make sense, is not
obviously
to indict it: cf. Wittgenstein (
1980a
), pp. 30ff. This is not to deny that it would have counted as a straightforward indictment for Hume.)

5
Deleuze decries this definition in Deleuze (
1991
), pp. 107ff. I think I am less at odds with him than I appear to be. For one thing, there is my qualification ‘for many purposes’. For another, I accept that his reservations are included in the difficulties that would have to be confronted when it came to examining what is involved in derivation. See
Ch. 21
, §2(d), for a further brief discussion of this issue.

6
Cf. Hume’s claim in
Treatise
, p. xviii, that no science or art ‘can go beyond experience, or establish any principles which are not founded on that authority.’

7
Cf. Pears (
1990
), p. 10.

8
I am using ‘knowledge’ in its customary sense, not the more restricted sense that Hume introduced in
Treatise
, p. 124.

9
For examples of this new interpretation, see Craig (
1987
),
Ch. 2
; and G. Strawson (
1992
). For debate, see the essays in Read and Richman (
2000
). For (what seem to me to be) correctives, see Pears (
1990
); Bennett (
2001
), §§273–275; Millican (
2002
), (
2007
), and (
2009
).

10
But nor can I resist a brief comment on the final paragraph of the chapter by Edward Craig cited in the previous note, in which he comments on what he describes as ‘an amazing paragraph’ from Moritz Schlick (
1938
). Schlick there argues that one consequence of the semantic empiricism that he and his fellow positivists claim to find in Hume, and that I likewise claim to find there, is that there is nothing we cannot know, at least in principle. Craig objects both to what he perceives as the absurdity of this consequence in its own right, ridiculing the ‘potential omniscience’ with which it credits us, and to the idea that there is anything of the sort in Hume, citing a passage from the
Enquiry
in which Hume makes clear that ‘our ignorance’ is not ‘a good reason for rejecting any thing’ (pp. 72–73). On the first point Craig confuses there being nothing we cannot know with our being able (simultaneously) to know everything. On the second point he confuses our being unable to know something with our (simply) not knowing it.

11
E.g. Strawson (
2000
), §2.

12
E.g. Millican (
2007
), §IV.

13
For an interesting defence of Hume against this kind of worry, see Everson (
1988
).

14
Cf. the famous empiricist formula, ‘
Nihil est in intellectu quod non prius fuerit in sensu
’ (‘Nothing is in the intellect that has not first been in the senses,’). For interesting material on the origin of this formula, see Cranefield (
1970
).

15
Hume himself, notoriously, cites a counterexample to it: the case of a man who has an idea of a particular shade of blue deriving from the impressions of other shades near it on the scale from darker to lighter. Hume says, ‘This instance is so singular, that it is scarcely worth our observing, and does not merit that for it alone we should alter our general maxim’ (
Enquiry
, p. 21).

16
See e.g.
Enquiry
, §IV, revealingly entitled ‘On “The” Idea of Necessary Connexion’ (double quotation marks added), esp. the first few pages.

17
For others, see
Treatise
, I.iv.3.

18
I am here ignoring the distinction between a term’s being used without any corresponding idea and its being used with a corresponding idea that (some of) its uses serve to misdescribe. (No doubt Hume would concede that Descartes has
something
in mind when he uses the term ‘God’.) In many cases this distinction is merely terminological.

19
For a helpful discussion, see Broackes (
2002
).

20
In what follows I shall prescind from the question, a variant of which will be prominent in
Ch. 12
, §8, of whether impressions, or items of sense experience, are entities of the right
sort
for beliefs to be derived from them;: see McDowell (
1996
), esp. Lecture II.

21
For a particularly helpful contribution to this debate, see Millican (
2002
).

22
See also
Treatise
, I.iv, passim, esp. pp. 225–226. There is also of course the issue of how the natural is defined; see further n. 24.

23
That this is Hume’s primary concern is nowadays relatively uncontroversial. It is a dominant theme of Stroud (
1977
): see esp. pp. 1–8. Cf. also Craig (
1987
), pp. 81ff.; and, for an interesting discussion, Biro (
1993
).

24
This suggestion is not meant to preclude artifice in the construction of the concepts in whose terms the beliefs are framed, a caveat that is particularly significant where beliefs about justice are concerned; see
Treatise
, III.ii, passim. The artificial is in any case opposed to the natural only on a narrow definition of the natural; see
Treatise
, p. 484, and
Enquiry
, pp. 307ff.

25
See
Enquiry
, pp. 54–55. Also relevant is that famous passage from the
Treatise
in which Hume, having noted how little help he is afforded by pure deductive reasoning in dispelling various doubts by which he is afflicted, comments, ‘Most fortunately it happens that … nature herself suffices to that purpose, and cures me of this philosophical melancholy and delirium…. I dine, I play a game of back-gammon, I converse, and am merry with my friends; and when after three or four hours’ amusement, I wou’d return to these speculations, they appear so cold, and strain’d, and ridiculous, that I cannot find it in my heart to enter into them any further’ (p. 269; cf.
Enquiry
, pp. 159–160). Not only are the processes of pure deductive reasoning unable to do the work done by the natural processes, then, but they are unable to subvert it.