Imagined Causes: Hume’s Conception of Objects by Stefanie Rocknak (review)

In Imagined Causes, Stefanie Rocknak tackles the vexing problem of interpreting Hume’s considered view about objects. She attributes to Hume a transcendental belief that she locates first in Treatise 1.3.2. Hume analyzes the idea of perfect (diachronic) identity into uninterruptedness and invariability. Impressions do not exhibit these features. Rocknak concludes that belief in perfectly identical objects is produced by the transcendental imagination. Hume critically examines other beliefs about objects. In Treatise 1.4.2, Rocknak distinguishes two vulgar (non-philosophical) beliefs (namely, the mere identification of resembling perceptions and the belief in the continued existence of perceptions) and the (false) philosophical belief, which rejects the vulgar beliefs by perceptual relativity arguments and distinguishes perceptions and objects. Ancient and modern varieties of philosophical belief are examined further in Treatise 1.4.3 and 1.4.4. Here I focus on Rocknak’s account of the transcendental belief.

Hume divides relations into two kinds, natural and philosophical. Identity is one of seven philosophical relations, which are comparisons of ideas. “Two objects, tho’ perfectly resembling each other, and even appearing in the same place at different times, may be numerically different” (T 1.3.1.1/69). To determine whether two objects are identical, we compare the ideas by causal reasoning, not abstract reasoning. For example, we might judge “that if we had kept our eye or hand constantly upon [an object], it wou’d have convey’d an invariable and uninterrupted perception” (T 1.3.2.2/74).

Because no perception exhibits continuity, Rocknak concludes, “Hume is employing a transcendental imagination” (84, cf. 49). But in describing causal conclusions about identity as “beyond the impressions of our senses” (T 1.3.2.2/74), it seems to me that Hume may simply mean that our conclusions are about unobserved, not unobservable, objects. For Hume, if I perceive a continuant, then what I perceive now also exists when I happen not to perceive it. So conceived, ‘continued existence’ requires the ideas of existence (that is, an idea of the object), particular times, and identity (cf. Treatise 1.2.6). Hume states that there is no non-empirical idea in the philosophical conception of objects as distinct from perceptions; the philosophical imagination “borrows all its ideas from some precedent impressions” (T 1.4.2.54/216).

Rocknak’s main evidence for the transcendental interpretation is T 1.3.2.2/73–74, where Hume claims we “always” imagine a “secret cause” connecting two observed objects (93). Rocknak holds that this secret cause is the perfectly identical object, connecting different appearances. It is always imagined, because it is in principle unobservable. I am not convinced that this passage supports Rocknak’s interpretation. For Hume, there is “always” a “secret cause” in uniform cases of contiguity or remoteness, not identity and difference. He aims to explain why different objects are always experienced as contiguous or remote from each other (cf. 95–96): the unobserved cause either unites or separates the objects. The “secret cause” need not be some unobservable cause; it may be some object that happens not to be observed (for example, an off-stage puppeteer who separates two puppets). After discussing contiguity, Hume writes, “The same reasoning extends to identity” (T 1.3.2.2/74). Clearly Hume means that causal reasoning is involved in judging the philosophical relation. But it is not clear that he requires unobservable “secret causes” to secure identity (cf. 96–103). Hume asks whether the present object and the resembling remembered object are identical. If there were a “secret cause” producing their resemblance, they would be different, not identical.

Furthermore, Hume explicitly states that the vulgar attribute identity to perceptions themselves (T 1.4.2.31/202; cf. 1.4.2.14/193). This conflicts with Rocknak’s interpretation that we always transcendentally imagine secret causes (142–51). Rocknak cleverly suggests that the problem is Hume’s, and...