Sunday, May 1, 2016

On Hume’s Copy Principle

One of the most powerful tools in Hume’s epistemic toolkit throughout the Enquiry is the copy principle (CP). Roughly, the idea behind the CP is that all mental content can be divided into two categories: impressions, which are perceptions received immediately through sensation (either from sense organs or internal emotional states), and ideas, which we form based on our impressions. Importantly, impressions and ideas are not two different kinds of mental content, but rather “all our ideas, or more feeble perceptions, are copies of our impressions, or more lively ones;” ideas and impressions are both essentially perceptions, and are only “distinguished by their differing degrees of force and vivacity.” Our idea of the color brown, for example, is nothing but an image of the exact same color that we received through the senses, and we recognize the second image as an idea instead of an impression because the color isn’t as lively or as forceful in our minds as the initial impression was (rather than being thrust upon us, for example, we bring it into our reflection). Hume maintains that this principle holds not only for our idea of the color brown, furthermore, but for every possible idea; no idea can contain any content that was not first copied from an immediate impression on the senses. Rather than demonstrably prove this claim, Hume challenges the opposition to come up with a counterexample, and moves on to assert that proper use of the CP would put an end to many (mainly metaphysical) disputes that center around the disagreement of philosophical terms. The idea is that, whenever we are suspicious that a philosophical term is meaningless, we should ask, “from what impression is that supposed idea derived? And if it is impossible to assign any, this will serve to confirm our suspicion.” Equipped with this principle, Hume is able to easily dissolve age old disputes about concepts like substance and self, maintaining that these ideas must be meaningless and unintelligible, since there are no impressions that they can correspond to. It is unclear, however, that all of our ideas are copied exactly from impressions in the ways spelled out by the CP. Rather, it seems that some of our ideas are more conceptual than perceptual, and so it seems that Hume may be unsupported in his claim that the CP applies to all ideas, and unjustified, therefore, in his use of the principle to dissolve these various debates.

If we take a step back and examine how Hume uses and states the CP, it seems plausible that we might be suspicious of the meanings of the philosophical terms ‘idea’ and ‘impression.’ Following Hume’s advice, then, we ask “from what what impression are these ideas copied from? What impression gave rise to your idea of ‘impression’ or ‘idea’?” It does not seem like Hume has an obvious answer to these questions. His idea of ‘impression’ is stated in conceptual terms, as a class of “more lively perceptions,” and ‘idea’ is similarly defined as “the less lively perceptions of which we are conscious when we reflect” on our sensations. Those do not seem like perceptual definitions, and Hume would be hard pressed to describe a specific impression that can explain the relative ‘liveliness’ of the two terms, as well as the fact that they are otherwise indistinguishable. If we continue to follow Hume’s advice, then, it seems that our suspicions have been confirmed; Hume cannot assign any particular impression to his ideas of ‘impression’ and ‘idea,’ and so these terms must not actually carry any meaning. Taking another step back, it seems that Hume is attempting to give a general account of human psychology, while simultaneously claiming that we can have no abstract ideas that were not originally copied from particular impressions, which seems somewhat contradictory. In order to account for the whole of human experience, in other words, it seems that Hume wants claim that ideas (in general) are always copied from particular impressions (in general), but this removes his justification for claiming anything about a general idea or impression. If all ideas are copied or compounded from simple impressions, then Hume can have no general conception of ‘idea’ or ‘impression’ separate from any particular impression (or composition of impressions), and so the CP seems to be more of a description of the particular ideas that Hume has had so far, rather than a law of human psychology. Ultimately, then, it seems that Hume is on shaky ground when he uses the CP to dispel various ideas as meaningless, since the principle itself disallows him from applying it universally to a conception of ideas in general.

Hume, however, would be reasonable to argue that this objection misses his point. Although his principles are certainly supposed to amount to a general account of human psychology, he never claimed to have any general notion of ‘idea’ or ‘impression.’ All that Hume claims is that “when we analyze our thoughts and ideas, however compounded or sublime, we always find that they resolve themselves into such simple ideas as were copied from a preceding feeling or sentiment.” In other words, Hume might as well have written that the idea of brown is copied from the impression of brown, and the idea of sadness is copied from the impression of sadness, etc. He uses the word ‘idea’ not to refer to some general notion, but rather to refer to any particular idea that one might be able to come up with, and chooses to sum up this list with the assertion that “every idea which we examine is copied from a similar impression.” Of course, the absence of a general notion of ‘idea’ opens Hume up to the objection that the CP may only apply to the particular ideas that he has experienced — perhaps some of our ideas are missing from his list —  but he remains confident that no one will ever be able to come up with an idea not derived directly from their impressions. Like our knowledge of causal necessity, Hume’s knowledge of the CP seems to be a matter of fact, based on reasoning from experience; but like our knowledge of causal necessity, the fact that Hume’s reasoning comes from experience does not diminish the certainty of the reasoning. In other words, even though we might discover that the CP is based on reasoning from experience, and so not entirely supported by demonstrative arguments “of the understanding, there is no danger that these reasonings, on which almost all knowledge depends, will ever be affected by such a discovery.” So, just as no one will ever be able to hit a cue ball against an 8-ball only to have them both fly upward at the speed of light, no one will ever be able to produce an idea not derived from impressions in accordance with the CP; it does not matter that we cannot demonstrate the universal necessity of either fact in the understanding, as long as they are continually confirmed by the vast uniformity of experience. Hume would likely concede, therefore, that he has not provided any universal account of ideas or their universal relation to impressions, but would maintain that this does not disallow him from using the CP to dispel various definitions as meaningless, since the vast uniformity of experience has confirmed it as a matter of fact.

Unfortunately, upon closer examination, this optimistic response turns out to be unsatisfying; we can grant that Hume was not operating with any general notions of ‘idea’ or ‘impression,’ and it still does not seem to follow that the vast uniformity of experience confirms the certainty of the CP. We can suppose that, rather than stating the CP in any general terms, Hume wrote out a list of all the ideas he can come up with, and the impressions that they are copied from. Somewhere along this list is a phrase that says, roughly, ‘the idea of the table is copied from the impression of the table, and is similar in every way, except in its diminished force and vivacity.’ So, it seems that Hume has an idea of an idea of the table (a second-order idea of the table), that describes the idea of the table (the first-order idea of the table) and its relation to the impression of the table. But, presumably, the impression and first-order idea of the table do not include considerations about their relations to each other. The impression of the table simply is brown, five feet long, forceful, and vivacious; the idea of the table is brown, five feet long, weak, and not very lively; the second-order idea of the table is not brown, or five feet long, nor is it merely a less forceful and vivacious version of the impression. Rather, the second-order idea of the table seems indescribable in perceptual terms; it is certainly about the brownness of the table in the idea versus the impression, but it is not itself brown. Even more, the second-order idea of the table is also about the first-order idea of the table’s conformity with the CP, and its role in Hume’s larger human psychology, which seem unaccounted for in the first-order idea and the impression. It seems, therefore, that Hume’s second-order idea of the table is not merely some perceptual image or feeling that is less forceful than the table he initially saw and felt, but rather must be a conceptual description of those perceptions and their relation to other pieces of his philosophical framework. So, since Hume himself seems to employ ideas that are conceptual in nature, and not directly traceable back to impressions, he cannot justifiably accuse the conceptual terms of other authors as being meaningless or unintelligible.


Importantly, this objection doesn’t necessarily disprove the CP, but rather shifts the burden of proof back onto Hume. As mentioned earlier, instead of attempting to deliver a demonstrative proof of the CP, Hume simply maintains that it is confirmed by the uniformity of experience, and leaves the burden on his opponents to propose an idea that violates the rule. If an opponent is able to produce and idea that they claim is not derived from an impression, then Hume concedes that, in order to defend the CP, “it will be incumbent on [him] … to produce the impression or lively perception which corresponds to [that idea].” I claim to have delivered such an idea (the second-order idea of the table), and so the burden is now on Hume to come up with an impression that the idea was copied from. Furthermore, I maintain that Hume will not be able to reasonably argue that the second-order idea of the table is copied solely from the impression or idea of the table (with less force and vivacity); the second-order idea contains relational considerations that the impression and idea do not, and the impression and idea are brown, five feet long, and qualitatively perceptual while the second-order idea is merely about these perceptual qualities. This does not, however, mean that Hume has no way of showing that the second-order idea of the table was derived in accordance to the CP. While it may not be copied directly from the impression or idea of the table, copying impressions is not all that our minds can do; our ideas are either direct copies of impressions, or are created through “compounding, transposing, augmenting, or demising the materials afforded to us by the senses and experience.” In other words, while I claim to have shown that Hume’s second-order idea of the table cannot have been copied directly from the impression or idea of the table, it remains possible that he can show the second-order idea to be some composition, or transposition, of various different impressions and ideas. Perhaps we transpose our impression of force and vivacity with our impression of two things being related, and compose this complex idea with both the impression and idea of the table — Hume may be able to construct a second-order idea of the table through a story with a similar flavor. I leave this as a possibility, but would like to further comment that it seems like quite a tall order. In general, second-order ideas seem fundamentally different from impressions and first-order ideas; it seems that while we may be able to adequately describe first-order ideas and impressions in fully perceptual terms (color, size, shape, etc), perceptual terms will never be able to give an account of second-order ideas, no matter how much they are transposed and composed. Second-order ideas seem to be about abstract, conceptual relations, that cannot really be pictured, heard, or felt, and Hume’s empiricism simply cannot account for this conceptual flavor. Ultimately, therefore, while the door is still open for Hume to show that second-order ideas can be built up through the composition and transposition of perceptual impressions, it seems that the essentially conceptual nature of second-order ideas will prevent this sort of explanation from being fully satisfying.