THREE QUESTIONS ABOUT TREATISE I Iv 2

THREE QUESTIONS ABOUT TREATISE I iv 2

Abstract

This paper attempts to answer three puzzling questions about Treatise I iv 2, “Of scepticism with regard to the senses.” First, why does Hume think that the existence of objects “distinct” from the mind implies the “continu’d existence” of objects; i.e. that objects that enjoy a distinct existence could not be momentary objects that happen to exist only when perceived? Second, why does Hume think it obvious that the objects of the senses have a “broken and interrupted” existence, when it seems that this is true of acts of perceiving or perceptual episodes but not of the objects of sense perception? Third, why does Hume think that “coherence” alone, as opposed to “constancy,” is “too weak to support so vast an edifice, as is [belief] in the continu’d existence of all external bodies”?

THREE QUESTIONS ABOUT TREATISE I iv 2

Treatise I iv 2, “Of scepticism with regard to the senses,” is a notoriously difficult section of Hume’s masterpiece. In this paper, I examine three puzzling questions that arise from this section:

(1) Why does Hume think that the “continu’d existence” of the objects of the senses and their “existence distinct from the mind” stand or fall together?

(2) Why does Hume claim that the objects of the senses obviously have an intermittent or discontinuous existence?

(3) Why does Hume think that the “coherence” of our impressions (as opposed to their “constancy”) goes only a little way toward explaining the belief in body?

1. “Continu’d” and “distinct” existence

Hume says that our belief in the existence of bodies can be broken down into two beliefs, namely that

(A) the objects of the senses continue to exist while they are not being perceived

and

(B) the objects of the senses exist distinct from (independently of) being perceived.

Hume immediately adds that (i) if (A) is true, then (B) is true, and (ii) that if (B) is true, then (A) is true. As he puts it:

These two questions concerning the continu’d and distinct existence of body are intimately joined together. For if the objects of our senses continue to exist, when they are not perceiv’d, their existence is of course independent of and distinct from perception; and vice-versa, if their existence be independent of the perception and distinct from it, they must continue to exist, even tho’ they be not perceiv’d. (T 188)[1]

But while we may readily agree that (i) is correct, it may seem that (ii) is quite obviously wrong: why should the “distinct” existence of objects imply their continued existence? Why couldn’t there be momentary objects, which happen to exist only when being perceived, but nevertheless do not depend on being perceived for their existence?

Baldly stated, the answer I propose is that if objects of the senses existed only when being perceived, then there would be a constant conjunction between their existence and their being perceived. So, these objects would be causally dependent on being perceived, in Hume’s sense of causal dependence, and they would accordingly fail to enjoy a “distinct” existence. So given Hume’s views about causality, (B) does imply (A), just as he says.

But this bald statement is open to the following objection. If objects of the senses existed only when being perceived, this would make it the case that if an object of the senses exists at time t, then it is perceived at t; it would make it the case that an object’s existing is a sufficient condition for its being perceived. But this does not show that the object’s being perceived causes it to exist. Of course if an object’s existence were sufficient of its being perceived, then an object’s being perceived would be a necessary condition of its existing, but that does not mean that its being perceived would be a cause of its existing: a match’s igniting may be a necessary condition of its being struck, but its igniting is not a cause of its being struck.

The trouble is that on Hume’s definition of causality, as it is often construed, perceiving a momentary object cannot be said to cause its existence, anymore than a match’s igniting causes it to be struck. To see this, consider a schematic formulation of the version of the first definition of ‘cause’ that Hume gives in the Enquiry:

E1 causes E2 = df

(1) E1 is followed by E2

(2) All events similar to E1 are followed by events similar to E2[2]

In order to allow for cases where cause and effect are simultaneous, I shall replace “followed by” with “is accompanied by”, and for the sake of simplicity,” I shall replace the notion of similarity with that of sameness in kind (as Hume himself occasionally does).[3] This yields the following definition:

D1: E1 causes E2 = df

(1) E1 is accompanied by E2

(2) all events of kind K1 are accompanied by events of kind K2

(3) E1 is of kind K1 and E2 is of kind K2

It is plain that on D1, a cause is a sufficient condition for its effect. So, letting E1 stand for a momentary object’s existing and E2 stand for its being perceived, the definition provides no basis for saying that its being perceived causes it to exist, or that its existence is causally dependent on its being perceived.

The reply to this objection is that in the Treatise Hume implies, and in the Enquiry he virtually says, that a cause is not only a sufficient condition but also that, in the circumstances at hand, it is a necessary condition for its effect. In the Treatise he says not only that “the same cause always produce the same effect,” but also that “the same effect never arises but from the same cause … [so that] where several different objects produce the same effect, it must be by means of the same quality which we discover to be common amongst them.” (T 174-175) Leaving to one side the questionable point that alternative possible causes of a given effect must always have some common feature, this implies that a given effect can occur, in the circumstances at hand (viz. in he absence of alternative possible causes of the same effect), only if its actual cause occurs. In the Enquiry, Hume famously remarks that one “object” causes another when “if the first object had not been, the second never had existed.” (E 76) Allowing, as we must, for the possibility of alternative causes, this again means that in the circumstances at hand, the cause is a necessary condition of its effect. This point can be incorporated into D1 by expanding clause (2) as follows:

(2A) all events of kind K1 are accompanied by events of kind K2 and all events of kind K2 are, in the circumstances at hand, accompanied by events of kind K1.

Thus, for example, to say that striking a match causes it to ignite is to say that all such strikings are accompanied by ignition, and that in the circumstances at hand (there is no superheated air present, no other lit match is brought into contract with the one that ignites, and so forth), all such ignitions are accompanied by strikings.

Now for purely logical reasons we can say that

(a) All cases of an object’s being perceived are accompanied by cases of its existing.

But for momentary objects, we can also say that

(b) All cases of an object’s existing are, in the circumstances at hand (i.e. given the oddly perishable nature of such objects), accompanied by cases of its being perceived.

But this fit the model for saying that momentary objects are caused to exist by being perceived, in which case Hume is justified in saying that an object with a “distinct” or “independent” existence cannot be a momentary object--that (B) implies (A).

It might be objected that since (a) is true for purely logical reasons, the relation involved cannot be a causal one. But this objection is not decisive, for consider a parallel case:

(a1) All cases of murder are accompanied by cases of a person’s dying.

(b1) All cases of a person’s dying are, in the circumstances at hand, accompanied by cases of murder.

Suppose that, on the model of our expanded D1, we add that

(d) Oswald’s shooting Kennedy was a case of murder and Kennedy’s dying was a case of a person’s dying.

The fact that (a1) is logically true does not prevent its being the case that (a1), (b1), (c) and (d) jointly imply that Oswald’s shooting Kennedy caused Kennedy’s dying. The contingent nature of (b1) seems sufficient to preserve the causal connection between Oswald’s action and Kennedy’s dying. Likewise, the fact that (a) is logically true does not go against saying that perceiving a momentary object causes it to exist, because the contingent nature of (b) seems to preserve the causal connection between an object’s being perceived and its existing.

2. The intermittent existence of objects of the senses

Throughout Treatise I iv 2, Hume asserts that the objects of the senses have an intermittent or interrupted existence. Occasionally, he supports this view with the following Modus Tollens argument (T 210-211, T 214):

(1) If the objects of the senses continue to exist while not being perceived, then they exist independently of being perceived.

(2) The objects of the senses do not exist independently of being perceived.

______

... Therefore, the objects of the senses do not continue to exist while not being perceived.

To support this argument’s second premise, Hume cites the “experiments, which convince us, that our perceptions are not possest of any independent existence,” such as pressing one eye with a finger and seeing everything double. (T 210)

Hume, however, has a much more basic reason for holding that the objects of sense have a discontinuous existence. This is that our perceptions are obviously discontinuous, “broken,” or “interrupted.” Consider for example the perceptions you have when looking at the furniture in your study. Those perceptions are plainly discontinuous: they cease to exist each time you glance away from the furniture or shut your eyes, and they cease to exist for the much longer periods of time when you leave your study, or while you are sound asleep. This simple and obvious point, Hume thinks, incontrovertibly proves the falsity of the belief in the “continu’d” existence of the objects of the senses.

To grasp Hume’s thinking, it is crucial to understand that when he says that our perceptions are interrupted or discontinuous, he is not saying merely that our perceivings, or acts of perception, or perceptual episodes, are discontinuous, which is of course true. Rather, he is also claiming that the objects of our perceptions--the things that we perceive--are discontinuous. For Hume does not distinguish between perceptual episodes and the objects perceived in them; his use of the term “perceptions” to stand indifferently for both perceptual episodes and their objects is not just linguistic carelessness or sloppiness, but embodies a genuine non-recognition of the distinction between them. The only form in which Hume recognizes a distinction between perception and its objects is that of the philosophical theory of the “double existence of perception and objects” advocated by Descartes, Locke and others--a theory that Hume rejects. Except within the context of this theory, Hume sees no place for a distinction between perception and its objects.

Hume’s rejection of this distinction calls for an explanation, if only because the distinction is so elementary. Even ordinary, unsophisticated common sense would distinguish between perceiving something, or the perception of something, and what is perceived, e.g. between seeing an apple, and the apple which is seen; or touching a cat, and the cat which is touched; or smelling a rose, and the rose which is smelled, etc. Why then does Hume reject this elementary distinction?

One possible answer lies in Hume’s acceptance of the view that we perceive only our own impressions.[4] For impressions have exactly the same temporal characteristics, including notably the same duration, as perceptions, taken as perceptual episodes or acts of perceiving. So the thesis that we perceive only impressions implies that the objects of perception last no longer than the perceptions themselves. But in that case, the distinction between perceptions and objects collapses, at least for an empiricist like Hume. For from Hume’s empiricist point of view, the only thing that could have differentiated between perceptions and their objects is a temporal difference between them, because any “act” of perception distinct from the object perceived would not be something of which we could have any impression: it would be, so to speak, diaphanous or “transparent,” so that Hume’s empiricism would banish it as a meaningless notion. Thus it may seem that Hume’s acceptance of the philosophical view that we perceive only our own impressions explains his otherwise puzzling denial of the elementary distinction between perceptions and objects, which in turn explains his view that the objects of the senses have an intermittent existence.

This explanation, however, puts the cart before the horse. For as I have said, Hume regards the discontinuity of our perceptions, together with the collapse of any distinction between perceptions and its objects, as an incontrovertible proof that the objects of perception are discontinuous. But of course our sense impressions are discontinuous; so an explanation of why Hume holds that objects of the senses are discontinuous that appeals to his view that we perceive only impressions renders his claim that perceptions are discontinuous and his rejection of the perception/object distinction completely otiose, as a proof of the discontinuity of the objects of the senses: one need not appeal to the discontinuity of perceptions and the lack of a distinction between perceptions and objects in order to establish that the objects of the senses are discontinuous, if one assumes from the start that those object are only impressions. Hence, in order for Hume’s claim that perceptions are discontinuous, together with his rejection of the perception/object distinction, to serve as a (non-otiose) argument for his claim that the objects of the senses are discontinuous, neither premise of that argument can rest on the view that we perceive only impressions. So we need a reason for the rejection of the perception/object distinction that does not rest on that view.

I think that such a reason can be given, by pressing the point that there is no impression of any introspectible “act of perceiving” that could distinguish it from its object. For what then does distinguish perceptions from objects? It still seems that the only thing that could do so would be their temporal characteristics: objects can, so to speak, outlast the perceptions of them, and so must be distinct from those perceptions. This point’s central importance is recognized by Kant in his Analogies of Experience, where he explicitly distinguishes between the time-relations of perceptual episodes and the time-relations of their objects, saying that while perceptual episodes are always successive, their objects may be co-existent; for example that while the perceptions of the front of a house may occur before perceptions of the back, the front and back of the house that the observer successively sees co-exist. Hume recognizes the point in his own way, since he says that one of the chief components in our belief in body is precisely that bodies continue to exist unperceived. But unlike Kant, Hume does not argue that assigning a different set of time-relations to objects of perception from those of the perceptions themselves is a condition of the very possibility of experience, and is therefore epistemologically warranted. Rather, Hume sees the time-relations of perceptual episodes--what Kant calls the order of apprehension and some of his commentators call the subjective time-order--as the only one that is necessary for experience: what Kant considers to be the objective time-order enters into Hume’s thought only as the fictional product of the imagination working in accordance with merely contingent principles. So, these time-relations cannot, for Hume, be appealed to as an epistemically warranted way of distinguishing between perceptions and objects: objects do not knowably have different time-relations than perceptions. Within the range of what can be known, there are only the time-relations of the perceptions themselves, and since there is accordingly no other set of time-relations to appeal to, and yet no way to distinguish between perceptions and objects other than their time-relations, the distinction between perceptions and objects collapses.

3. The role of “coherence”

Hume finds two features of our impressions that explain why we believe in the continuous existence to the objects of the senses despite their intermittent existence: constancy and coherence. He discusses separately how constancy and coherence each contribute to this belief. His general position is that coherence alone is “too weak to support so vast an edifice, as is [belief in] the continu’d existence of all external bodies” (T I iv 2, 198-199), and he places much more weight on his complicated “system” about the way constancy leads to the belief in object-continuity. (T 199-210) Why does Hume think that coherence alone goes so little way toward explaining this belief?

Here is one of Hume’s characterizations of coherence:

Bodies often change their positions and qualities, and after a little absence or interruption may become hardly knowable. But here ‘tis observable, that even in these changes they preserve a coherence, and have a regular dependence on each other; which is the foundation of a kind of reasoning from causation, and produces the opinion of their continu’d existence. When I return to my chamber after an hour’s absence, I find not my fire in the same situation, in which I left it: But then I am accustom’d in other instances to see a like alteration produc’d in a like time, whether I am present or absent, near or remote. This coherence, therefore, in their changes is one of the characteristics of external objects, as well as their constancy. (T I iv 2, 195)