Cause and Effect

On the flyleaf of Messrs Beauchamp and Rosenberg’s book about Hume’s theory of causation, Professor Donald Davidson says of it: ‘This is certainly the best available discussion of Hume and causality. It is much more than that, however: it is the best book-length treatment of causality.’ Professor Davidson is perhaps a little biased by the fact that the authors’ views on the nature of causality coincide so very closely with his own. I should not myself rank their book quite so highly, among those that have appeared in recent years, as J.L. Mackie’s The Cement of The Universe, to which indeed they pay respectful tribute. One of the merits of their book, which has, however, the defect of making it rather stodgy reading, is that with the exception of my own little book on Hume, in the OUP ‘Past Masters’ series – which, since they pay attention to my other writings, I take to have appeared only after their manuscript was completed – there is practically no modern contribution either to the philosophy of Hume or to the topic of causality that they fail to acknowledge and often to discuss, sometimes in greater detail than its interest seems to warrant. I was momentarily puzzled by a series of references to a philosopher called ‘Gertrude Anscombe’ until I remembered that Professor Elizabeth Anscombe’s initials were G.E.M.

The main theses which Beauchamp and Rosenberg ably defend are that historians of philosophy have been at fault in supposing that Hume’s conclusions about the nature of causality were primarily sceptical, that Hume was not so much concerned with eliciting what people ordinarily meant when they spoke in causal terms as with expounding a view of causality that would find room for it in nature, that the idea of necessary connection, which was a genuine feature of the concept of causality, did not represent any natural relation, but was projected by our minds through associations developing out of the experience of past regularities, that Hume’s two definitions of a cause as ‘An object precedent and contiguous to another, and where all the objects resembling the former are placed in like relations of precedence and contiguity to those objects that resemble the latter’, and as ‘An object precedent and contiguous to another, and so united with it that the idea of the one determines the mind to form the idea of the other, and the impression of the one to form a more lively idea of the other’, can be blended into a unified theory, that Hume rightly treats particular causal statements as entailing causal laws, and that causal relations hold between events.

I subscribe to most of these theses, though not to all of them. I think that the authors are right in treating Hume’s definitions of cause as complementary rather than inconsistent. They are right also, in my view, in construing Hume’s ‘sceptical’ analysis of causality as a demonstration that causality is not a logical relation, not, in Hume’s terminology, a relation of ideas, rather than an attempt to prove that all causal judgments are unwarranted. That his intention was not to discredit causal reasoning, which he equated with all non-deductive reasoning concerning matters of fact, is shown by his subtitling his Treatise of Human Nature as An Attempt to Introduce the Experimental Method of Reasoning into Moral Subjects, no less than by his being at pains to devise rules for judging of causes and effects, and by the confident use which he makes of causal arguments in his extensive writings on morals, economics, politics and history. Nevertheless I think that Beauchamp and Rosenberg do less than justice to the extent to which Hume himself considered the first book of the Treatise, at any rate, to be a work of scepticism. The principle ‘that instances of which we have had no experience, must resemble those of which we have had experience, and that the course of nature continues always uniformly the same’ is one that Hume believed to be at the root of all inductive reasoning, and he thought it important to prove, not only that this principle is not demonstrable, but that it cannot without circularity be shown to be even probable. It may be that later authors have been able to make out a case for applying canons of rationality to inductive procedures, but it was rather Hume’s way to contrast rationality with what he regarded as the legitimate force of custom. One gets a very different picture of Hume if one concentrates on the section of the Treatise entitled ‘Of Scepticism with regard to the Senses’, but for all their thoroughness, the present authors have chosen entirely to separate Hume’s treatment of causality from any consideration of his theory of the external world.

I think it likely that Beauchamp and Rosenberg are right in saying that Hume required of singular causal statements that they have the backing of universal generalisations, but wrong in supporting him on this point. My own view is that for a singular causal statement to be acceptable it need not have the backing of anything stronger than a general statement of tendency. This applies particularly in the field of human conduct. What happens is that we have a limited set of hypotheses concerning the ways in which some state of affairs is frequently brought about. Whenever we find that one of these is satisfied, and the others not, we say that its antecedent is the cause of the state of affairs in question, even though we do not believe the conjunction to be invariable.

I agree with Mackie, as against Beauchamp and Rosenberg, that if one insists on treating causality as a relation, rather than pursue what seems to me the more satisfactory course of subordinating its analysis to that of causal explanation, then its terms are better taken to be facts than events. One reason for this is the greater difficulty of individuating events. Our authors accept Davidson’s suggestion that events are to be identified by their causes and effects, but apart from the difficulty of delimiting the causes and effects in question, and the unwarranted assumption, on Davidson’s principles, of complete determinism, I cannot see how this avoids being circular.

One point on which I side with Beauchamp and Rosenberg, against Mackie as well as Von Wright, is in holding that no case has been made for distinguishing causal from merely temporal priority. I suspect, however, that they have not been able to explain, any more than I can, why the insertion of the concept of temporal priority into that of causality is not, as Russell once said it was, a purely arbitrary stipulation.

David Hume’s Treatise of Human Nature, which is usually judged to be his principal contribution to philosophy, though he himself preferred his later and more polished Enquiries Concerning Human Understanding and Concerning the Principles of Morals, was first published in 1739. His friend Adam Smith’s most famous work An Inquiry into the Nature and Causes of the Wealth of Nations appeared in 1776, the year of Hume’s death. Hume had been born in 1711 and Smith in 1723. Smith, who lived until 1790, had published his The Theory of Moral Sentiments in 1759 and had in a charming letter been congratulated by Hume on its success. Smith in his turn was ready to acknowledge his admiration for Hume both as a philosopher and as a man. The obituary notice, in the form of a letter to William Strahan, which Smith allowed to be published, along with Hume’s few pages of autobiography ‘My Own Life’, in 1777, concluded with the words: ‘Upon the whole, I have always considered him, both in his life-time and since his death, as approaching as nearly to the idea of a perfectly wise and virtuous man, as perhaps the nature of human frailty will admit.’

In expressing himself so strongly, Smith was undoubtedly sincere, but he may also have been making amends for the disappointment he had caused Hume by refusing to make himself responsible for the publication of Hume’s sceptical Dialogues Concerning Natural Religion. If so, his success probably exceeded his expectations: So great was the offence given to the pious by his encomium that ‘a single, and as I thought, a very harmless sheet of paper, which I happened to write concerning the death of our late friend, Mr Hume, brought upon me ten times more abuse than the very violent attack I had made upon the whole commercial system of Great Britain.’

It has been suggested that Smith was indebted to Hume for some of his ideas about economics, but this question is not argued in Knud Haakonssen’s scholarly book. Dr Haakonssen is primarily interested in Smith’s theory of jurisprudence and the conception of justice with which it is allied, and it is here that he seeks and discovers the influence of Hume.

While Hume regarded the sense of justice as the source of our moral obligations, he also classified it as an artificial virtue. What he meant by this is that it does not arise from any natural motive. For the most part, men do not have any such passion as the love of mankind and their natural feelings of benevolence towards a limited number of persons would favour injustice, in that it would lead them to favour these persons at the expense of the community. Neither does the sense of justice proceed wholly from self-interest, though self-interest plays a leading part in its formation.

The position is that because of his physical weakness a man can survive and prosper only as a member of a group, and that the scarcity of goods which nature has provided brings different groups into competition for them. If this competition were unrestrained, no man could ever count on ‘the peaceable enjoyment of what he may acquire by his fortune and industry’. To avoid such a calamity, men have found it in their interest to subscribe to a set of conventions which establish rights to property and lay down the conditions in which its transfer from one man to another is legitimate. This is not a version of the theory of the social contract. The conventions are not promises, though Hume regards the performance of promises as one of the factors on which the peace and security of human societies depend. For promises themselves depend on a convention. The conventions are the means adopted by men in their own interest to cope with a grudging environment. If men had had the good fortune to be surrounded with such natural abundance that all their material appetities, however luxurious, could easily be satisfied, then, as Hume put it, ‘the cautious jealous virtue of justice would never have been dreamed of.’

But where does morality come in? In this case, through our sympathy with the victims of injustice, which works upon our moral feelings. Dr Haakonssen gives a very accurate summary of Hume’s moral theory. He represents Hume as holding that ‘moral virtues and vices are those qualities in a person which have a tendency to create such effects as by means of a sympathy, unbiased by the actual success of or personal relation to the person judged of, cause a pleasure or pain in the observer – a pleasure or pain which by association of impressions calls forth calm, as opposed to violent, versions of love or hatred ... and those two passions are what is properly called moral approval or disapproval.’

The question then arises what qualities evoke such moral evaluation. Hume, as correctly reported by Haakonssen, considers them to be reducible to ‘four broad, non-exclusive groups: those which are immediately agreeable to ourselves (that is, to the person with the quality), or to others; and those which are useful to ourselves, or to others’. Justice is included in the group of those that are useful to others. There is thus a utilitarian element in Hume’s moral philosophy, as there is in Adam Smith’s, but it is not the straightforward utilitarianism that was later to be propounded by Bentham and John Stuart Mill.

Haakonssen objects to Hume that he does not explain how our sympathy can extend as far as the operation of the sense of justice would require. Adam Smith surmounts this difficulty by having recourse to the fiction of an impartial spectator, whose observation of the injuries inflicted by injustice governs the moral sentiments of those who place themselves imaginatively in his position. This introduction of something approaching an absolute standard is, however, relativised by Smith to the social circumstances of the moral community whom the ideal observer represents, and his conception of the sense of justice as depending on sympathy, though in some ways more elaborate, is not fundamentally different from Hume’s. A distinctive feature of Smith’s theory is that he supposes nature to have endowed man, ‘not only with a desire of being approved of, but with a desire of being what ought to be approved of: or of being what he himself approves of in other men’. This allows Smith to diverge from Hume by finding a source of moral judgment in the estimation of propriety.

The larger part of Haakonssen’s book is devoted to an account of the way in which Smith matches his ideas of the development of law with his historical theory of the progress of society from the Age of Hunters through the Age of Shepherds and the Age of Agriculture to the Age of Commerce. This is all very clearly, if prosaically set out, and even readers who are not students of jurisprudence should find that the work affords them a valuable addition to their stock of knowledge.