The last quoted passage introduces the general point mentioned at the beginning of 15.1: if Hume makes what looks like a dogmatic ontological claim that causation is nothing but regular succession,19so that there is in fact nothing like Causation in the objects, one will best understand his claim if one adds the qualification ‘so far as we have any notion of it’, or ‘so far as it is anything to us’, or some such phrase. This qualification may for example be applied to the passage which immediately follows the ‘global subjectivism about necessity’ passage, which may at first seem like conclusive support for the standard view.20Having equated power and energy and efficacy with necessity (157/1.3.14.4), and having just argued in high subjectivist fashion that all necessity lies only in the mind, Hume continues as follows:
(15) The efficacy or energy [or necessity or power] of causes is neither plac’d in the causes themselves, nor in the deity, nor in the concurrence of these two principles; but belongs entirely to the soul. . . .’Tis here that the real power of causes is plac’d, along with their connexion and necessity. (166/1.3.14.23) And of course this must be so, if power (so far as we have any notion of it) is the same as necessity, and all necessity whatever is only in the mind. Thus Hume’s global subjectivism about necessity—and also, no doubt, a certain pleasure in creating a
18 And what wefind, in fact, is that matter and motion are constantly conjoined with thought just as billiard ball movements of type A are constantly conjoined with billiard ball movements of type B.
19 More fully: causation is nothing but regular succession (1st definition) which has a certain effect on human minds (specified in the 2nd definition).
20 It may also be taken as support for the Causation-only-in-the-mind view noted and put aside in nn. 7 and 16.
dramatic effect—leads him to make what looks like the outright, dogmatic, meta- physical claim that there is quite definitely no sort of Causation in objects.21
It is, however, central to Hume’s overall position that the passage may be expanded as follows:
So far as the efficacy or energy or necessity or power of causes is something of which we have any positively or descriptively contentful (E-intelligible) idea at all, and can positively- contentfully talk about at all, it is neither placed in the causes themselves, nor in the deity, nor in the concurrence of these two principles; but belongs entirely to the soul. . . . It is here that the real power of causes is placed—so far as we have any positively contentful (impression- grounded, E-intelligible) notion of power, that is; it is here that the connexion and necessity of causes is placed—so far as we have any notion of necessity, that is.
Two pages later, after a further dramatic passage, he makes this point explicit, in a remark already quoted. He says that he is
(16) indeed, ready to allow, that there may be several qualities both in material and immaterial objects, with which we are utterly unacquainted; (168/ 1.3.14.27)
and he adds that we may if we wish‘call these power or efficacy’. But when we use ‘power’ and ‘efficacy’ in this way, ‘meaning these unknown qualities . . . with which we are utterly unacquainted’, ‘’twill be of little consequence to the world’; and we must be quite clear what we are doing. Above all, we must not suppose, as so many misguided philosophers (believers in‘intelligible causes’) have done, that the terms ‘power’ and ‘efficacy’ ‘signify something, of which we have a clear idea’.
That’s all; there is no irony here. In the very next paragraph he refers three times to this utterly unknown power, in passages already quoted: the‘uniting principle among our internal perceptions’ and the ‘uniting principle . . . among external objects’ are utterly (E-)unintelligible to us, for‘experience . . . never gives us any insight into the internal structure or operating principle of objects’. That is, it never gives us any insight into the nature of Causation or power:
(17) we have [EP] no other notion of cause and effect, but that of certain objects, which have been always conjoin’d together. . . . We cannot penetrate into [REF] the reason of the conjunction.22
21
Elsewhere his inclination to sound as if he is making such a claim is (as noted in 15.1) perhaps to be explained principally by his inclination to think of‘the objects’ as nothing but immediate mental objects of experience, perception-constituted objects or content-constituted objects for which a regularity theory of causation is in a sense correct (6.3 and Appendix A).
22
93/1.3.6.15. The device of adding some variety of the phrase‘insofar as we have any idea (or notion or conception) of it’ applies to all other passages that seem to support the standard view, e.g. 165/ 1.3.14.20,169/1.3.14.29, 171/1.3.14.33, 173/1.3.15.5. They come particularly thick and fast in the text immediately before and after Hume’s explicit statement of global subjectivism about necessity.
As remarked, Hume states his dramatic conclusion ten times at the climax of 1.3.14 (165–7/1.3.14.20–25): the necessity we take ourselves to be acquainted with in objects is not in fact in objects; it’s nothing but a determination of the mind, an impression of reflexion, a feeling we have when we experience or think about certain things; ‘the necessary connexion betwixt causes and effects’ with which we take ourselves to be acquainted‘is the transition [of our thought] . . . from one to the other . . . arising from the accustomed union’ (165/1.3.14.21). And for present purposes we can grant that his argument for the conclusion that we neither do nor can have any experience of necessity or causal power in the objects is entirely successful.23We can grant that it’s defensible even in its modern form, which makes no use of Hume’s premiss that true experience of Causation would have to confer the capacity for a priori certain inference about causal matters (it simply stresses the idea that all we can ever come upon, e.g. in scientific investigation, are regularities of succession). For the point that matters here is that it has no positive ontological implication. Hume isn’t saying that there’s definitely no such thing as Causation, only that we have no experience of it of a sort that could give us any insight into its nature (contrary to the rationalist philosophers he was arguing against).
Immediately after quotation (15) he goes on to say that this is a violently para- doxical claim. And so it is, both for ordinary people who are convinced that they have direct experience of causal force, and for philosophers who are convinced that they have some genuine understanding of the nature of causal power in the world, and that ‘the universe [is] in principle intellectually transparent’.24 But the violent paradox lies in the epistemological claim that we have no idea of Causation at all, and that the necessary connection we suppose ourselves to discern in objects lies only in the mind. It really is quite extraordinary that it should upon reflection turn out that we have absolutely no idea of the ultimate nature of Causation, given that we are ourselves genuinely Causally efficacious agents, and live in a world governed by Causation, by some inflexible ‘operating principle’, a world in which ‘everything is surely governed by steady, inviolable laws’ (Philo, Dialogues 125/174). It is, however, a complete mistake to think that the ‘violent paradox’ consists in the claim that there is definitely no such thing as Causation in reality—no reason at all why reality, whatever its ultimate unknown nature, is regular in the way that it is. Poor old Hume—lumbered with such a claim.25
There’s a another decisive passage in 1.4.2, which again requires care. Hume writes that
23 For doubts, see Anscombe op. cit. and 23 below. For an important qualification to the claim that this is Hume’s view, see 19.2.
24
Craig 1987: 38.
25 This reading is further confirmed by the comparable passage in the Enquiry, discussed on pp. 189–90 below.
(18) the only conclusion we can draw from the existence of one thing to that of another, is by means of [1] the relation of cause and effect, which shews, that there is [2] a connexion betwixt them, and that the existence of one is dependent on that of the other. [3] The idea of this relation is deriv’d from past experience [of constant conjunction] . . . (212/1.4.2.47).
[3] is the idea of the relation of cause and effect which we derive from experience of constant conjunction. [1] is the relation of cause and effect itself, as defined in the first definition of cause. It’s the relation of cause and effect insofar as it’s something of which we have some positively contentful conception, something which we can detect in things, something which can thereby ‘shew’ us something, something which we can make use of in drawing conclusions.26What is of principal importance is the distinction Hume makes between [1] and [2]. For [1]—the relation of cause and effect in the sense of something which we can detect, and which can thereby‘shew’ us something—is explicitly distinguished from [2], i.e. the actual Causal connection, the relation of real existential dependence between objects. If [1] is detected as holding between two (types of ) objects A and B, Hume says, that shows that something other than [1], i.e. [2], is present. So [1] and [2] can’t possibly be identified. Hume’s claim, to restate it, is that when the conditions laid down in the regularity-theory definition of cause look to be fulfilled by A and B, we can conclude from that that there is a real causal (Causal) connection between them, real existential dependence (whose intrin- sic nature we cannot however know). Observed constant conjunction between A and B shows us that, or is evidence that, there is real causal connection between A and B; conjunction is evidence for connection. Clearly, then, constant conjunction and real causal connection aren’t the same thing, and Hume is not a ‘Humean’.
(19) Constant conjunction . . . proves a dependence (4–5/1.1.1.8). It doesn’t constitute one.
(20) Constant conjunction [between two things] sufficiently proves, that the one . . . is the cause of the other (174/1.3.15.9).
Constant conjunction is evidence—good evidence, a sufficient proof—of causal connection. It’s not the same thing. If constant conjunction of A and B were the same thing as causal connection between A and B, it couldn’t do anything as weak as ‘sufficiently prove’ causal connection between them.
(21) We are never sensible of any connection between causes and effects, and . . . ’tis only by our experience of their constant conjunction, [that] we can arrive at any knowledge of this relation (247/1.4.5.30).
26 See 21 for the point that the two definitions of cause are only held to capture what causation is so far as it’s something of which we have some positively contentful conception, some more than merely ‘relative’ idea. Hume explicitly states that they fail to capture the nature of causation as it is in itself, in reality.
There’s more to causal connection than constant conjunction, but this is all the knowledge we can have of it. We are ignorant of the nature of things, as any sceptic knows. We fall short.
That’s one very clear implication of these passages. Another is that there’s some- thing importantly right about a point made by Rosenberg and Beauchamp discussed in 8. This is the point that Hume sometimes distinguishes—in a way that seems very odd to us—between [1] causal relations (‘the relation of cause and effect’), on the one hand, and [2] the operations of nature, or the actual connection between objects whereby‘the existence of one is dependent on that of the other’, on the other hand. He makes just such a distinction in passage (18).
The explanation of this fact has been provided. When Hume makes such a distinction, by [1], the‘relation of cause and effect’, he means the relation of cause and effect so far as it is anything to us, positively-contentfully speaking, and accord- ingly distinguishes it from [2], the actual connection between objects in virtue of which‘the existence of one is dependent on that of the other’.
The reason why it’s important to note his tendency to make such a distinction is that it makes the following way of putting things seem possible. ‘Yes, Hume does indeed hold a regularity theory of causation. He does indeed hold that“the relation of cause and effect” is (first definition) just a matter of regular succession between objects, of a sort (second definition) which affects the mind in a certain way. But he also, and fully compatibly with this, believes in Causation—in something in the objects which is not just this regularity, given which they are regular in the way they are. He just doesn’t call this “the relation of cause and effect”, for he uses this expression in the theory-of-ideas-sanctioned way. He calls it“the ultimate connex- ion” between them (“the uniting principle”, the “reason of the conjunction” of objects, etc.) He doesn’t call it “the relation of cause and effect” because by “the relation of cause and effect” he means the relation of cause and effect so far as it’s something of which we have a positively contentful conception.’
I think this is an important part of the truth about Hume’s text. It solves many difficulties. One reason why it can’t be the whole truth is that Hume later makes such claims as that experience and observation can give us no‘acquaintance with the nature of cause and effect’ (68/7.18). Here, by ‘cause and effect’, he means the actual unknown ‘ultimate connexion’—since experience and observation do give us acquaintance with regular succession, and also with the feeling of determination in the mind. This matter will be discussed in Part 3.
The complexity of Hume’s text produces many complications in commentary, but the basic point is simple. The ‘necessity-we-attribute’ is only in the mind. The relation of cause and effect so far as we have any positive conception of it is fully captured by the two definitions of cause. There is (of course) such a thing as Causation, the‘ultimate force and efficacy of nature’ (159/1.3.14.8). There is natural necessity, in that sense. But its nature is perfectly unknown to us insofar as it’s something more than regular succession.
Those who are feeling the force of the adage exemplum docet, exempla obscurant should skip to Part 3 (they may soon feel it again). Those who are still unconvinced may consider this: we should perhaps grant that Hume’s explicit discussion of causation in the Treatise is essentially contestable. It may be very hard to see—to feel—the force of the present account of Hume until one has seen that it’s unavoid- able, given (a) the nature of his strictly non-committal scepticism with regard to questions about what may or may not exist, (b) his connected remarks about what we may intelligibly suppose to exist, as opposed to what we may contentfully conceive of, (c) his natural pre-philosophical—but also fully philosophically endorsed—realism or at least basic realism with respect to external objects,27(d) his entirely unquestioned belief that there must be something about the nature of reality in virtue of which it is regular in the way that it is—i.e. Causation, and (e) his constant use of referring expressions referring to Causation. To these must be added (f) his careful rephrasal of his position in the Enquiry in response to misunderstanding of the Treatise.28For this rephrasal contains no support for the ‘standard’ view at all (as I’ll shortly try to show). It’s not seen by Hume as a significant revision of his earlier views, and he endorsed it as afinal statement of his ‘philosophical statements and principles’ in his plea to future readers—his ‘Advertisement’, his warning or notice—at the beginning of the Enquiry.
This said, I realize that it may not be thought possible to win the argument decisively on the ground of Hume’s explicit discussion of causation in the Treatise. For although there is a great deal of direct evidence for the present view (e.g. (e)), and massive—decisive—indirect evidence (e.g. (a), (b), (c), (f )), and although the appar- ent counter-evidence can be fully accounted for, it is arguable that the standard view is also compatible with the evidence of the explicit discussion of causation.
15.6 Conclusion
In sum: Hume doesn’t hold anything like the strong, realist (ontological), regularity theory of causation in the Treatise. He believes there is such a thing as Causation (natural necessity) in reality, but that we can’t know anything about its nature. This claim is of course, and as always, in some tension with his view that the notion of Causation or power in the objects is ‘unintelligible’ (E-unintelligible). But the appearance of tension is superficial, because to say that the notion of Causation is unintelligible is simply to say that we have (given his theory of ideas) no positive, descriptively contentful conception of the nature of Causation (apart from regular succession), and to say this is not to say that it doesn’t exist, or that we can know that
27 See again (e.g.) 187/1.4.2.1, 218/1.4.2.57, and also 222–4/1.4.3.9.–10. To say that Hume naturally believes in external objects is not to say that the belief somehow doesn’t really count for him, philosoph- ically speaking. It’s to say that he really does believe in the existence of external objects.
it doesn’t exist, or that we can’t achieve any reference to ‘it’ at all. It’s not to say that we can’t intelligibly suppose that it exists (6.5), or that we can’t possess a ‘relative’ idea of it. We do possess a‘relative’ idea of it: it is that in reality—the ‘ultimate force and efficacy of nature’ (159/1.3.14.8)—in virtue of which reality is regular in the way that it is.
On balance, it’s best to represent Hume as claiming only that causation so far as we know about it in the objects is nothing but regular succession, while assuming (with Newton and others) that there is something more to it in fact. He says that
(i) this multiplicity of resembling instances . . . constitutes the very essence of power or connexion (163/1.3.14.16);
that
(ii) necessity . . . is nothing but an internal impression of the mind (165/1.3.14.20); that this necessary
(iii) connexion, tie, or energy lies merely in ourselves (266/1.4.7.5); that it is
(iv) the constant conjunction of objects, along with the determination of the mind, which constitutes a physical necessity (171/1.3.14.33).
All these claims are remarks about the true and highly surprising origin (the impression-source) of our supposedly genuinely contentful idea of Causation or power or necessity. They all require the addition ‘so far as we have any notion of it, so far as it is (positively-contentfully) anything to us’. They’re all entirely com- patible with the view that there is power or Causation in objects. The point is the same: even though there is or may be such a thing, that’s not what our supposed idea of power or Causation is really an idea of, contrary to the convictions of those rationalist philosophers who believe in intelligible causes. Our idea of necessity simply can’t mean or reach out to Causation in the way we think, given its origins. It’s