Hausman on Certainty and Necessity in Hume

47. Hausman on Certainty and Necessity in Hume Professor Hausman in the course of a painstaking and often illuminating examination of my paper "Hume on Intuitive and Demonstrative Inference" fortunately has occasion to make some positive suggestions of his own regarding the best way to interpret Hume's philosophy. One of the most interesting and provocative of these suggestions is that we should discount Hume's claim to have found an impression of reflection which is at the same time one of necessity. Rather, we should regard the impression of reflection in question as one of certainty. Such a manner of proceeding would, moreover, in Hausman' s view have the advantage öf explaining why we all believe that causal connections are necessary when in fact they are not. For the same impression of certainty, as Hausman envisages it, will be generated by both the cognitive act of comparing two relata wherein genuine necessity is reflected and the determination of an idea by a preceding impression wherein causal inference consists and wherein genuine necessity plays no role. But how can Hausman be so sure that they will be the same? For he insists that the cognitive act of comparing two relata is radically different from the determination of an idea by a preceding impression. Why then should we expect that observation of the two should generate comparable feelings of certainty? And if they do not generate comparable feelings of certainty what is the purpose served by postulating them? Moreover, I use the term "postulating" advisedly: by Hausman' s own admission there is little or no textual evidence for their presence in Hume's theory of causal inference. Hausman' s purpose would be no worse served - at least incoherence would be avoided - by emphasizing what a cognitive act of comparing two relata and a determination of an idea by a preceding impression have in common as opposed to what distinguishes them. They have, moreover, more in 48. common than he seems to think. For, while the one requires reflection according to Hume and the other does not, they 2 both involve a determination of the mind. Nor is Hume's use of the concept of intuition in connection with the former incompatible with such a determination, as Hausman seems to think. On the contrary, it implies it. For intuition as Hume understands it seems to pertain to immediate inference. And such inferences consist as far as he is concerned in determinations of the mind to move from one 3 idea to a second one. Indeed, it is just the supposed infallibility of such determinations that allows him to distinguish them from the more haphazard principles of as4 sociation. A cognitive act of comparing two relata and a determination of an idea by a preceding impression are sufficiently alike, then, to explain why we all believe that causal connections are necessary when in fact they are not. This does not mean, of course, that necessity need reside in such a cognitive act of comparing. As a matter cf fact, as we shall see in a moment, this is not the view that Hume usually takes of the matter. But a cognitive act of comparing, as we shall also see in a moment, is close enough to the area in which necessity does reside to allow Hume to explain in the manner indicated how our mistaken belief about causal connections arises. Hausman' s comparable feelings of certainty are not required. Nor has Occam's razor finished cutting. More specifically , the idea of real necessity of which Hausman speaks could have, as far as I can see, no role to play in Hume's philosophy which is not already played by the cognitive act of comparing two relata. There is, moreover, no such idea anymore than there is any idea of causal necessity. The "idea of entities in a particular sort of relation" presumably like resemblance will not do. For resemblance is not necessity. And to say as Hausman does that "it is because we can read off a resemblance, say, that we pronounce the relationship necessary" is to state a problem, not to 49. ? ?. 6 solve it. What then is the solution? It seems...