Stefanie Rocknak’s research while affiliated with Hartwick College and other places

According to Husserl, Hume’s empirical method was deeply flawed—like all empiricists, Hume did not, and could not adequately justify his method, much less his findings (PRP 113–114, LI 115–117, 406). Instead, Hume gives us a “circular” and “irrational” “psychological explanation” of “mediate judgments of fact,” i.e. of inductive inferences (LI, 117). Yet Husserl was certain that he could justify both his own method and his own findings with an appeal to the phenomenological, pre-theoretical, pre-naturalistic “epoché” (I1 §§59–60). However, whether or not Husserl’s notion of an epoché is justified, or even viable, is not our focus here. Rather, our issue is with Hume, particularly: How could Hume have responded? In this paper, I show that in Book I of the Treatise, Hume did—however implicitly—appeal to a “pre-theoretical” notion of belief, which meets Husserl’s demands for a pre-theoretical grounding, i.e. a justificatory grounding. And so, his method is, at least in this respect, justified. But this belief is by no means “prenaturalistic.” Rather, it is a function of empirical data. In particular, it is a product of the “constant and coherent” impressions that seem to naturally obtain of experience. As a result, Hume’s method, which, in agreement with Husserl, we may characterize as psychology, does admit of a certain degree of regularity, but an empirical regularity. And according to Hume, contra Husserl’s complaints, this is the best that we can hope for; regularity will have to suffice, not certainty. However, it must be noted up front that this justification is implicit—Hume is not nearly as forthright as he could have been. Thus, making his position more explicit is the task of this paper. I am not presenting my own position, where I offer my own arguments. Instead, I present textual evidence, contextualized with ample explication. It is my hope that by doing so, Hume may more effectively make his case.

This article shows that in 1.4.2.15-24 of the Treatise of Human Nature, Hume presents his own position on objects, which is to be distinguished from both the vulgar and philosophical conception of objects. Here, Hume argues that objects that are effectively imagined to have a “perfect identity” are imagined due to the constancy and coherence of our perceptions (what we may call ‘level 1 constancy and coherence’). In particular, we imagine that objects cause such perceptions, via what I call ‘indirect causation.’ In virtue of imagining ideas of objects that have a perfect identity, our perceptions seem to be even more constant and coherent (what we may call ‘level 2 constancy and coherence’). Thus, in addition to seeing that Hume is presenting his own position on objects in this section of the Treatise, we see that he is working with a previously unrecognized kind of causation, i.e., indirect causation, and that he has two kinds of constancy and coherence in mind: level 1 and level 2.

Traditionally, scholars have argued that 1.4.2 may be split into two general sections: one that concerns the “vulgar” conception of objects, and another that concerns the “philosophical” conception of objects. I argue that there is a third position: Hume’s, which includes two more accounts of how we transcendentally conceive of perfect identity. We examine one of those accounts of perfect identity here (which constitutes Hume’s second account of how we transcendentally conceive of perfect identity) and the other in Chap. 7 (which constitutes Hume’s third account of how we transcendentally conceive of perfect identity). In Chaps. 8 and 9, I show why all three instances of how we transcendentally conceive of perfect identity are not to be confused with the vulgar position on objects, nor with the philosophical position.

In 1.3.9, Hume introduces two levels of reality. He does this to show why the relations of resemblance and/or contiguity cannot reflexively produce vivacious ideas in the manner that causation can (T 1.3.9.2; SBN 107). But the implications of Hume’s account of reality are far-reaching. In fact, if we don’t take his two systems of reality into account, we can’t understand his notion of an object, his many forms of belief, nor his notion of justification. Oddly though, Hume’s two systems of reality are largely overlooked in Hume scholarship, if not ignored altogether (with some exceptions, e.g. Kemp Smith 1941; Owen 1999; Loeb 2002).

In Part II of this book, we saw that Hume thinks that we always imagine ideas of objects that admit of perfect identity—by way of transcendental causation. However, while explaining the “philosopher’s” position in 1.4.2, Hume claims that we only imagine causes in reaction to the vulgar, where we do not employ transcendental causation. In this chapter, we examine the philosophical position in detail. In Sect. 1, I explain why vulgar perspective II falls apart—at the hands of the philosophers. In Sect. 2, I explain why the philosophers think that it is reasonable to think that mind-independent objects exist. In Sect. 3, I explain why Hume thought the philosophers were mistaken.

Proto-objects, I claim, are the necessary conceptual building blocks for an idea of an object that admits of perfect identity. Also, ideas of objects that admit of perfect identity, must, according to Hume, be imagined. In this chapter, we examine Hume’s somewhat implicit first account of perfect identity, given in 1.3.2. In the course of doing so, we begin to see how and why proto-objects enable us to imagine objects that admit of a perfect identity. However, the reader should note that this chapter merely serves as an introduction to Hume’s theory of imagined causes and perfect identity, while Chaps. 6, 7, and 8 provide us with a more fully-developed version.

Immediately after discussing what we identified in Chap. 6 as transcendental causation in 1.4.2, Hume introduces his four-part system. A detailed explanation of this system is lacking in the literature, although Kemp Smith (1941), does parse it into respective parts, and gives a brief explanation of it as a four part system (pp. 474–487). However, in the next few chapters I give a much more exhaustive account.

In this chapter, I make two major claims, which conflict with most, if not all readings of 1.4.2, First, the vulgar perspective should be divided into two sections, which correspond to, respectively, parts 2 and 3 of Hume’s system of identity. We may refer to these sections as vulgar perspective I, and vulgar perspective II. Second, neither vulgar perspective I nor II should be confused with the three accounts of perfect identity that were explicated in Part II of this book.

With the general concepts of impressions, ideas, cause and effect, belief and reality in mind, we may now turn to Hume’s notion of an object. In this chapter, we focus on a traditionally overlooked sense in which he uses the word ‘object’—“objects” that are either impressions or are ideas that “exactly represent” impressions. I show that such objects could not be what Hume has in mind by ideas of objects that admit of a “perfect identity.” This is the case because impressions and ideas that exactly represent impressions (simple or complex) do not represent both of the properties of uninterruptedness and invariability, while ideas of objects that we imagine to admit of perfect identity do. In fact, as we will see in the following chapters, impressions, and ideas that exactly represent impressions (simple or complex) must be understood as the necessary psychological building blocks for ideas of objects with a perfect identity. Accordingly, we may think of, and hereafter refer to, those impressions, and ideas that exactly represent impressions as “proto-objects.”