[The next several parts of this tutorial will draw very heavily on the essay I published in The Philosophical Review in 1960: "Hume's Theory of Mental Activity." The entire essay has been posted on box.net and is accessible via the link top of the home page of this blog.]
The explanation Hume offers of our belief in causal inference, and also the later, much more complicated, explanation he offers of our belief in the continued and independent existence of physical objects, rests on a theory of the functioning of the human mind that he nowhere spells out clearly and explicitly. Hume fails to present his theory in explicit form, I believe, because it is in conflict with his original intention, which was to develop a theory of human nature on the model of Newton's theory of physical nature. I have already noted that Hume's account of the association of ideas is misleading, because it is not ideas that attract one another by a "gentle force" and thus become associated together, but the mind that associates the ideas one to the other. The mind is active in the process of cognition and belief. Hume does not really want to acknowledge this fact. Indeed, in Part IV, in the midst of his extraordinary and very important account of personal identity, he explicitly rejects the notion that he is offering an account of the actions of the mind. Here is the passage, from Section vi, "Of Personal Identity," in Book I, Part IV:
"The mind is a kind of theatre, where several perceptions successively make their appearance; pass, re-pass, glide away, and mingle in a infinite variety of postures and situations.... The comparison of the theatre must not mislead us. They are the successive perceptions only, that constitute the mind; nor have we the most distant notion of the place, where these scenes are represented, or of the materials, of which it is compos'd."
But this simply will not do. Without directly confronting the theory of the mind's activity on which Hume is relying, we can make no sense of any of Hume's positive accounts of our beliefs. This was obvious to me long ago, in 1956, when I w s writing my doctoral dissertation, and therefore I undertook to reconstruct from the text a systematic account of the operations of the mind as Hume presents it to us, for all that he forswears it once he has done so.
The key to Hume's "theory of mental activity," as I called it, is the notions of propensities and dispositions or habits. A disposition or habit is a settled tendency of the mind to call up a certain idea when it is presented with an impressions of a particular character. The imagination, if subjected repeatedly to the gentle force of association, develops certain habits or customs. It comes to anticipate the conjunction of perceptions which past experience has exhibited. Much as Pavlov's dog would begin to salivate at the sound of a bell, so the mind, for example, generates the idea of an "effect" if presented with the "cause." [I so hope you are all familiar with the Russian behavioral psychologist Pavlov's famous experiments with dogs. If not, Google it.]
A "propensity" can be described as the disposition to develop a disposition, or as a "second-level" disposition. When we attempt to explain the behavior of Pavlov's dog, for example, it is not enough to state that Pavlov rang the bell every time the food was offered. That is, of course, a necessary part of the explanation, but we must add that the dog was capable of being conditioned. If it were not, no amount of bell-ringing could produce the conditioned reflex of salivation. [Long experience leads me to believe that one cannot, for example, condition cats.] Let us use the term "disposition" to describe the fact that an entity is prone to act or react in certain ways under certain conditions. Let us use the term "propensity" to describe the fact that an entity is prone to develop certain dispositions under certain conditions. Then we can say that the dog's salivation upon the ringing of the bell is a manifestation of a disposition (in this case a conditioned reflex) and the disposition in turn is a manifestation of the dog's propensity (to form dispositions of this sort).
There are several simple facts about stimuli, dispositions, and propensities which it will prove helpful to keep in mind when considering Hume's theory. First, the stimuli are, of course,
sensory in nature. They are the instigators or "proximate causes" of the conditioning process, as well as the triggers of the disposition already formed. Second, the sensory stimuli are the individuating conditions of the dispositions being formed. The nature of the stimulus determines the precise character, or content, of the disposition. For example, by choosing a different stimulus, we can make the dog salivate at the blowing of a whistle or a clap of the hands rather than at the ringing of a bell. Third, dispositions are distinguished from propensities by their logical type. As mentioned above, propensities are second-level dispositions, or dispositions to form dispositions. Therefore, dispositions depend upon the conditioning stimuli and follow after them, while propensities precede both stimuli and dispositions. Propensities are thus necessary conditions for the development of dispositions.
The human mind has a small number of innate propensities, or "dispositions to form dispositions." When the mind is presented with perceptions conjoined in certain ways, its propensities are activated and it develops dispositions. These dispositions determine the mind to reproduce in imagination certain impressions when it experiences certain others. [The mind, of course, reproduces them as ideas, less vivid than the impressions of which they are copies.] The mind, in another of Hume's phrases, forms a "habit of association." The factors in cognition which Hume labeled impressions of reflection -- such as the impression of necessary connection, as we shall see in a moment -- are really dispositions, and the ideas of necessary connection, substance, and so forth, are not copies of impressions but ideas of mental dispositions. The innate propensities constitute the basic "machinery" of the mind. They are the necessary and universal conditions of all our ideas of causes and objects. In Hume's words:
"In order to justify myself, I must distinguish in the imagination betwixt the principles which are permanent, irresistible, and universal; such as the customary transition from causes to effects, and from effects to causes: And the principles, which are changeable, weak, and irregular; such as those I have just now taken notice of. The former are the foundation of all our thoughts and actions, so that upon their removal human nature must immediately perish and go to ruin."
With these few notions before us, we can now make a good deal of sense of Hume's explanation of our tendency to form causal inferences and to believe in their results. When we examine specific causal reasoning, Hume tells us, we find four components: an impression, an inference, an idea, and a belief. All causal reasoning begins with an impression of sensation which is present before the mind and acts as the anchor of the entire process. The second component is the inference by which the mind passes from the present impression to an idea. This idea, related to the impression, is the third component of the reasoning process. Fourth, there is the belief which we then repose in the idea.
The impression needs no explanation. To account for the inference to the idea, Hume introduces the factor of "constant conjunction." Upon examining causally related objects, and reflecting back over past experience, we discover that whenever we label one object cause of another, we can recall pairs of contiguous and successive objects which resemble the present impression
and idea. As the objects appear again and again in similar relations to one another, they become "associated" in the imagination. By a natural mental process, the habit is inculcated of conceiving of the one when the other is perceived. When the impression of the cause, let us say, is present to the mind, then the disposition created by past conjunction determines the mind to
reproduce the idea of the effect. The essence of the causal inference lies in this transition. There remains only the belief, which I shall come to presently.
Hume returns in Section 14 of Part iii to the first question, What is the nature of the idea of necessary connection? His analysis of particular causal inferences has produced three conclusions which, he believes, permit him now to answer the question. The conclusions are:
"[T]hat the simple view of any two objects or actions, however related, can never give us any idea of power, or of a connexion betwixt them: that this idea arises from the repetition of their union: that the repetition neither discovers nor causes any thing in the objects, but has an influence only on the mind, by that customary transition it produces."
But every idea, says Hume, is a copy of some preceding impression. As the idea of necessary connection is not derived from any quality or relation of objects, it cannot be a copy of an impression of sensation, so it must be an impression of reflection, that being the only other kind, according to Hume. The only available impression of reflection is the "feeling" attached to the disposition of the mind to pass from an object to the idea associated with it. Therefore the idea of necessary connection must be the idea of this mental transition.
This analysis of causal inference depends on three points, each of which requires some comment. The points are: (i) the nature of the "transition" from impression to related idea; (2) the nature
of the "impression of necessary connexion" on which we base our idea of causal influence; and (3) the nature of the mechanism of belief, whereby we assent to the inference and take the associated idea as objectively representative. Let us consider them in turn. The "transition" of which Hume speaks is not to be understood as a passing of the mind from one present perception to another present perception. What happens is that the mind, perceiving the first, "forms an idea of its usual attendant" (p. 93). In other words, the mind reproduces in imagination a copy of the impression which has lately been associated with the present impression. This process of recall is the manifestation of a mental disposition or "habit" which is inculcated by experience. The mind is confronted with a succession of resembling impressions, and thereupon develops a disposition to conceive the one when it perceives the other. As I have already noted, the proneness of the mind to develop such a disposition is itself a disposition; it is what I called a "propensity." Hume now sets to one side the aspects of causal reasoning which are independent of the mind's inferences. The contiguity and succession of objects is indeed separate from our
thought, as is the constant association of like objects. That is simply part of which is given to us in experience, and requires no further explanation. These relations are "independent of, and antecedent to the operations of the understanding," as Hume puts it. But the necessity of
their connection is a property of our inference upon them and derives from the mind's "habit." In the terminology which I have adopted, the associated impressions act as stimuli to activate an
innate propensity; the result is a mental disposition to imagine a related idea when presented with an impression.
This is all a bit gnarly, so let me recapitulate in a few words. We observe repeated cases in which similar pairs of sense impressions occur right next to one another in space and time. Fire-heat, fire-heat, fire-heat. The mind has an innate propensity to develop a habit of associating such constantly conjoined pairs, so that when I observe one of them but not the other, I conjure up an idea of the missing one. [I see a fire, but do not feel heat. This triggers the habit of association, and I call up the idea of heat, thinking to myself, "I bet that fire is hot."]
The second element of Hume's analysis is the "impression of necessary connexion" from which the idea of causal influence is copied. We are here presented with a common problem of textual
interpretation. What Hume says is not the same as what Hume says he says. The natural development of his argument leads him in one direction, but the prior strictures of his system require a different and conflicting move. Hume talks as if the impression of necessary connection, like an impression of love or envy, were a directly observable mental content arising
from the workings of the imagination. But the description which he actually gives of this impression is peculiar, to say the least.
The "internal impression," he claims, is the "propensity, which custom produces, to pass from an object to the idea of its usual attendant." Again, the "customary transition is the same with the power and necessity." Hume is not saying here that the impression arises from the transition or is conjoined with the transition or is dependent upon the transition. He is saying that the impression is the transition. Now this is plainly an error in classification. It is what in the 1950's and 60's we used to call a "category mistake," following Gilbert Ryle's usage of that phrase in his very influential book The Concept of Mind. [Full disclosure: I wrote my undergraduate honors thesis on that book in 1953, and out of it managed to get my very first professional publication, a little note in MIND, a tasty bit with the enticing title "Professor Ryle's Discussion of Agitations."]
"Customary transitions" and "propensities" are mental operations or powers, not contents of consciousness. If the idea of necessary connection is a copy of the transition from an impression to its usual attendant, then it is a copy of a mental activity. It is in fact the idea of the mind's disposition to reproduce related perceptions in imagination. In these passages we can observe Hume shifting toward explanation in terms of mental activities, while still tied to the language of mental contents. Later on we shall see him move even further in this direction when he deals with the problem of external objects.
Finally, there is the problem of belief. Hume rejects the scholastic view that to believe in the existence of an object is to conjoin an idea of existence to the idea of the object. He casts about, therefore, for some quality common to all the ideas and impressions in which we repose belief, and he hits upon "the manner of our conceiving them." The belief in an idea is nothing but an increased force and vivacity by which our conception is heightened and enlivened." This theory, as Hume himself recognized, is open to serious objections. One objection is the fact that there are many common beliefs which cannot be explained as "enlivened ideas." When we form an inference upon the evidence of historical traces, and reason from a present impression of manuscripts to the past existence of a historical personage, the resultant belief cannot be attributed to the force of the impressions of the text before us. We could not, for example,
strengthen the belief by illuminating the page more brightly or recasting the print more sharply. Hume appears to take account of this objection in several passages, the first of which contains a
turn of phrase suggesting a possible revision:
"For suppose I form at present an idea, of which I have forgot the correspondent impression, I am able to conclude from this idea, that such an impression did once exist; and as this conclusion is attended with belief, it may be ask'd, from whence are the qualities of force and vivacity deriv'd, which constitute this belief? And to this I answer very readily, from the present idea. For as this idea is not here considered as the representation of any absent object, but as a real perception in the mind, of which we are intimately conscious, it must be able to bestow on whatever is related to it the same quality, call it firmness, or solidity, or force, or vivacity with which the mind reflects upon it, and is assur'd of its present existence."
The key phrase is the last clause but one of the final sentence. The present idea, serving as the "impression" from which the inference proceeds, must possess that firmness or solidity or force
or vivacity with which the mind reflects upon it. The force is not a quality of the perception, but rather a quality of the way in which the mind conceives it. Hence any idea or impression, however weak, may serve as the starting point for a causal inference, if only the mind reflects upon it with the solidity of belief. Now in the passage I have quoted, Hume uses this locution only to explain our inference from a present idea to the past impression of which it is a copy. But in the Appendix to the Treatise, he extends the same description to all belief. There he states that the "feeling" of belief is a "firmer conception, or a faster hold, that we take of the object." Belief is not a hold that the object takes on us, but that we take on the object. Here again we see a shift in emphasis from the characteristics of perception to some activity of the mind.