The constructive theory of causal inference, by which Hume connects his sceptical analysis of the causal relation with his final discovery of the impression of necessity in the felt determination of certain habits or customs in imagination, shows the limitations of such criticism as would dispose of Humes conception of experience as atomistic merely. It will be recalled that Hume begins the Treatise of Human Nature with an analysis of the perceptions of mind into impressions and ideas; and that, in the subsequent sections of Part I, he discloses the remaining elements of perception.
Therefore, it would be incorrect to identify perception with any one of its elements, or with all of them taken respectively in isolation. Only mere fancies or perfect ideas occur divorced from all associations. Normally, in the experience of mature persons, there occurs, at the least, a lively idea associated with a present impression; which is, by definition, the general nature of belief. These beliefs vary in elaborateness and force between the extremes of proof and mere chance; but only at the extreme of mere chance, or gratuitous fancy, do isolated impressions or ideas exist.
Ordinarily, the terms of Humes analysis of perception occur in the synthesis which he articulates in his theory of belief. Normal experience, then, will consist of perceptions, themselves the syntheses in habit which are beliefs. The substantial identity of things present here and now may be compared in direct perception. But only on the assumption that the causes of a things existence remain unaltered may the continued existence of a thing beyond perception be inferred.
Again, although times and places as such admit of comparison without inference, still any constancy or variation in such relations may be inferred to exist only as a result of causation. That relation, therefore, is the principle of all inferences about matters of facts. Nothing exists which may not be considered as either a cause or an effect; though it is plain there is no one quality, which universally belongs to all beings, and gives them title to that denomination (Hume, 185). Since, therefore, the origin of the idea of cause and effect is to be found in no quality of our perceptions, it must be derived from some relation between them.
Hume at once finds two such relations: causes and effects are contiguous in space and time, and the cause is always prior in time to the effect. Dr. Broad (120-2) points out that Humes proof of the temporal priority of causes is formally vicious. Hume himself seems to have had some doubts about its validity, for he writes: If this argument appear satisfactory, tis well. If not, I beg the reader to allow me the same liberty, which I have used in the preceding case, of supposing it such. For he shall find that the affair is of no great importance (225). But contiguity and succession do not afford a complete idea of causation.
A thing at once contiguous and prior to another still might not be considered its cause. There is a necessary connection to be taken into consideration, and that relation is of much greater importance than any of the other two above mentioned (Hume 211). Necessary connection is then the defining characteristic of the causal relation. The impression from which this idea is derived is therefore the one we are looking for. Yet the only relations between impressions Hume has found so far are those of contiguity and succession, which I have already regarded as imperfect and unsatisfactory (216).
And he proceeds to divide his problem into two questions: why we believe that every event must have some cause or other; and why we believe that the same cause must necessarily produce the same effect (Hume 223-6). Hume thus distinguishes the law of causality from the law of causation, and takes it that together they are what is meant by a necessary connection among events. Though a general maxim in philosophy, that every even must have a cause is not a matter of knowledge. This Hume demonstrates first on the grounds of his own view of the extent of knowledge.
The law of causality may be identified neither with resemblance, degrees of quality, contrariety, nor proportions in quantity and number. The law is therefore not known to be true. Hume thinks that anyone who would controvert this conclusion will be obliged to exhibit a relation at once identical with causality and known by direct inspection, which it will then be time enough to examine (224). He proceeds next to urge that the law in question is to be demonstrated by apagogic reasoning on no theory of knowledge, and therefore is neither intuitively nor demonstrably certain (228).
That every event must have some cause or other means that the ideas of cause and effect are necessarily connected. Was this the case, it would be impossible that those ideas should be separable. Yet, since they are distinct, the ideas of cause and effect are separable; and the denial of their necessary connection involves no contradiction. Here Hume relies on the principle of his atomism. Yet he need not have done so; for the contradictory of the law of causality being not self-contradictory, that law is not demonstrable by apagogic reasoning.
For since the relation of cause and effect is the principle of all inference about matters of fact, no inference to a probability can be independent of that relation. Hume takes his analysis thus far to have shown that our only notion of cause and effect is of certain objects constantly conjoined. We cannot penetrate into the reason of the conjunction. We only observe the thing itself, and always find that from the constant conjunction the objects acquire an union in the imagination (128). Our notion of cause and effect, as so far disclosed, is no more than a philosophical relation.
Thus though causation be a philosophical relation, as implying contiguity, succession, and constant conjunction, yet it is only so far as it is a natural relation, and produces an union among our ideas, that we are able to reason upon it, or draw any inference from it(Hume 131). And causation is more than a philosophical relation just so far as it is association. Concerning the nature of the transition from impression to idea in causal inference is thus that the transition is the work of associations or habits in imagination, not of reason.
So understood, the inference from impression to idea Hume declares to be one part of the definition of an opinion or belief; that it is an idea related to or associated with a present impression (Hume 137). Hume insists that the idea of necessary connection derives from the felt force of the natural relation of cause and effect. The ideas of cause and effect being separable, there can be no contradiction in denying their necessary connection. Here again, however, Humes conclusion is valid independently of the assumption on which he himself makes it out. For the contradictory of the law in question is conceivable.
And in going on to show the uniformity of nature to be indemonstrable, Hume points out on the one hand that we can at least conceive a change in the course of nature, which sufficiently proves that such a change is not absolutely impossible; and, on the other, that the uniformity in question being the presupposition of probable reasoning, any attempt at its demonstration by induction could only beg the question. Hume is giving a definition of cause and effect, so in conclusion he is describing the observed or felt nature of that relation. Those impressions may have causes Hume does not deny.
He says the ultimate causes of sense-impressions are, in my opinion, perfectly inexplicable by human reason, (223) and he finds the alleged necessity that they have a cause to be not demonstrable. Nor are his arguments that impressions are prior to and productive of ideas advanced as a denial that impressions are thus productive. And the attraction of association is also assumed and its origins are regarded as inexplicable. Yet this means that impressions, ideas, and the attraction of association are found to exist in constant conjunction, not in necessary connection.
That the than the logical necessity of Malebranche, means that what has been called a necessary connection is in fact habitual; not that from this conclusion we may infer the non-existence of causes. For the fact that the rational necessity of causation is not to be demonstrated plainly does not imply that nothing in the nature of a cause can exist. If we do not know the laws of causality and causation to be true, neither do we know them to be false. Hence there is no reason, the contradictory of which would be inconceivable, why causes should be or should not be assumed.
The law of causation, being demonstrable by neither apagogic nor inductive reasoning, if demonstrable at all, will be on the ground that necessary connection in fact is disclosed within sense-perception. Since Humes failure to find that logical necessity obtains between the elements of sense-perceptions has been held to require his own analysis of experience, it may be well to consider briefly the fact that in other interests, and through a conception of experience not that of Hume, the same conclusion had been reached by three of the Cartesians.
Hume may well emphasize the conclusion that all of our beliefs that are justified by experimental enquiry and all of our accurately successful causal inferences will depend upon the operation in the understanding of those fundamental habits by which cases of constant conjunction are disclosed and inferred. The nature of the understanding thus is what constitutes the foundations of induction.
That the habits of which the understanding consists can in no case yield demonstrably certain conclusions, means that the foundations of induction are essentially illogical, to be neither demonstrated nor denied either by the reason of the Cartesians or by inductive theory itself. It is, finally, of the nature of the understanding that logic proper consists: the pretensions of our scholastic headpieces and logicians are simply to be set aside (312). The assumption that the elements of experience are intrinsically self-identical is thus requisite to the view of impressions as complete in themselves.
But the finding of elements by analysis is itself not the further explanation that these elements may be regarded as self contained because, like being simple, resemblance is not the name of a qualifying predicate. Whether or not the theory of philosophical relations be rejected along with the doctrine of impressions and ideas, the conclusion, as such, that apagogic reasoning is powerless in matters of existence, remains no less free of that theory and that doctrine, than is the conclusion itself of Humes failure to find necessary connections among matters of fact.
For, as Professor Kemp Smith has pointed out, it was Hume who first perceived the falsity of the Cartesian, rationalistic view of the causal relation (537). Malebranche could discover no necessary connection between events, yet he continued to conceive of the causal relation as being intelligible to the pure understanding, and, as a consequence of his theory of knowledge as the vision in God, failed to draw the conclusion that the law of causation is neither intuitively nor demonstrably certain.
For a real cause, Malebranche says, is a cause between which and its effect, the mind perceives a necessary connection (Rome 94). This conclusion drawn, Hume can attack the root of any assumption that the law of causation may be justified by experience. The attempted justification could only be inductive; and the law of causation is the presupposition of induction. Since causal inference is found to be neither rational nor merely sensory, if explicable at all, it will be so through an analysis, not of the fancies of the philosophers, but of the imagination that is the foundation of the senses and the memory.
It is thus found that probable inference consists of the habits of imagination, or beliefs, which are the perceptions that constitute the mind, and of which the more firmly established in the imagination are the understanding. To conclude the work we should say that Humes chief innovation in associations theory is his inclusion of cause and effect among the natural relations, or modes of association.
Yet even a moderately detailed examination of Humes theories of causal inference and belief in substance may suffice to indicate how groundless is the charge of total scepticism, while at the same time it discloses the character of unanalysed experience in Humes view. The relation between his critical analysis of causation and that of the Cartesians, as well as the logical nature of Humes arguments in that regard, make it plain that his analysis here is independent of his chief psychological dogma. Works Cited Baillie, James. Hume on Morality. London: Routledge, 2000.
Broad, Charlie Dunbar. Perception, Physics, and Reality; an Enquiry into the Information that Physical Science Can Supply About the Real. New York: Russell & Russell, 1972. Hall, Roland. Fifty Years of Hume Scholarship: A Bibliographical Guide. Edinburgh: Edinburgh University Press, 1978. Hendel, Charles William Jr. Studies in the Philosophy of David Hume. Princeton: Princeton University Press, 1925. Hume, David. A Treatise of Human Nature. New York: Penguin Classics, 1986. Kemp Smith, Norman. A Commentary to Kants Critique of Pure Reason. Houndmills: Palgrave Macmillan, 2003
Noonan, Harold W. Philosophy Guidebook to Hume on Knowledge. London: Routledge, 1999. Potkay, Adam. The Fate of Eloquence in the Age of Hume. Ithaca: Cornell University Press, 1994. Rome, Beatrice K. The Philosophy of Malebranche: A Study of His Integration of Faith, Reason, and Experimental Observation. Chicago: H. Regnery Co. , 1963.
Stewart, John B. The Moral and Political Philosophy of David Hume. New York: Columbia University Press, 1963. Strawson, Galen. The Secret Connexion: Causation, Realism, and David Hume. Oxford: Clarendon Press, 1989. Stroud, Barry. Hume. London: Routledge, 1977.