Let the philosopher pride himself on his precision as much as he will... I nevertheless dare defy him to make a single step in his progress without the aid of resemblance. Throw but one glance upon the metaphysical aspect of the sciences, even the least abstract of them, and then tell me whether the general inductions that are derived from particular facts, or rather the kinds themselves, the species and all abstract notions, can be formed otherwise than by means of resemblance. [20]
At the border of knowledge, similitude is that barely sketched form, that rudimentary relation which knowledge must overlay to its full extent, but which continues, indefinitely, to reside below knowledge in the manner of a mute and ineffaceable necessity.--Michel Foucault The Order of Things (1970) {1994, Vintage Books Edition} Translated by Alan Sheridan* (original Les Mots et les Choses, 1966), pp. 67-68
The position attributed to Hume by Foucault in the first paragraph above recalls Hume's claim (in the first Enquiry) that "it seems evident, that, if all the scenes of nature were continually shifted in such a manner, that no two events bore any resemblance to each other, but every object was entirely new, without any similitude to whatever had been seen before, we should never, in that case, have attained the least idea of necessity, or of a connexion among these objects." (EHU 5.8) Even so the view being attributed to Hume is also a bit awkward as an interpretation of Hume's accounts of natural and philosophical relations (a distinction articulated in the Treatise), and the role of resemblance and identity among them. I want to highlight three problems.
First, in his own voice, Foucault distinguishes natural and philosophical relations as follows: philosophical relations presuppose reflection. Now, in Hume reflection is a mechanism of the mind that turns ideas into new impressions. That is, "impressions of reflexion are only antecedent to their correspondent ideas; but posterior to those of sensation, and deriv'd from them." (Treatise 1.1.2.1.) The paradigmatic examples of impressions of reflexion are "desire and aversion, hope and fear." (Treatise 1.1.2.1.) That is to say, the passions and emotions. I mention this explicitly because on Foucault's interpretation, the elements or relata or building blocks of philosophical relations are impressions of reflection or the ideas derived from them. (And it would be surprising if the passions played a role in them.)
For, Hume does not explicitly claim that reflection is presupposed in any philosophical relation including identity (which is the one singled out by Foucault). After all he claims, "we extend it to mean any particular subject of comparison, without a connecting principle." (Treatise 1.1.5.1) So, it is strictly speaking false to suggest that in Hume philosophical relations require the operation of reflection.
But one may well think this is not the end of the matter. After all, Hume lists seven sources of comparison that can be "considered the sources of all philosophical relations. (Treatise 1.1.5.1, emphasis added.) These turn out to be viz. "resemblance, identity, relations of time and place, proportion in quantity or number, degrees in any quality, contrariety, and causation." (Treatise 1.3.1.1,) If reflexion is crucial to these (especially identity) than the spirit of Foucault's account can be saved. But none of them presuppose reflexion. In addition, Hume divides this list "into two classes; [first] into such as depend entirely on the ideas, which we compare together, and [second] such as may be chang'd without any change in the ideas." Crucially, again, reflexion is irrelevant to this. So, Foucault misrepresents Hume here on philosophical relations.*
To be sure, I don't mean to deny Foucault's claim that "no equality or relation of order can be established between two things unless their resemblance has at least occasioned their comparison." This is Hume's position, too:
The first [philosophical relation] is resemblance: And this is a relation, without which no philosophical relation can exist; since no objects will admit of comparison, but what have some degree of resemblance. (Treatise 1.1.5.3; emphasis in original)
But while it is true that for Hume identity is only a philosophical relation this is not true for resemblance in Hume. For, second, Hume recognizes two kinds or sites of resemblance, one is a philosophical relation and when it functions in a philosophical relation, although resemblance "be necessary to all philosophical relation, it does not follow, that it always produces a connexion or association of ideas." (Treatise 1.1.5.3) Only natural relations generate an association of ideas. And for Hume there are three such natural relations of which resemblance is one. (Treatise 1.1.4.1.) Which he does treat, as Foucault suggests, as a "gentle force." (Treatise 1.1.4.1) [As an aside, Foucault's 'calm' force is almost certainly the effect of translation either from Hume's English to Foucault's French Hume or from Foucault's text to my translation, etc. For this reason, I am treating resemblance and similitude as synonyms here even though Hume does not really use 'similitude' in the Treatise.]
This, matters because, third, thus, in Hume's Treatise similitude/resemblance doesn't just reside mutely 'below knowledge' acting as a natural relation behind the scenes as it were constraining our mind, but it is also the content of knowledge (and so made explicit). So, here, too, Foucault is misrepresenting Hume's position.
Now, so far, I have skipped the material Foucault quotes and references with endnote [20] in the material I have quoted above. It's natural to read Foucault here as quoting Hume as evidence for his position. But Hume scholars may well wonder what the source of the quoted passage is. (It's not Hume after all.) In fact, the passage quoted by Foucault that starts with "Let the philosopher pride..." is not, first impressions notwithstanding by Hume, but rather is as endnote [20] indicates, a quote from Merian's (1767) Reflexions philosophiques sur la ressemblance, a book I had never looked at before.
Merian is a translator of Hume's first Enquiry and an important critic of Hume, which gave the world the term 'phenomenalism' in order to describe Hume (in 1793). See this (1997) paper by Laursen and Popkin for details (here). Sadly, their paper does not mention the Reflexions philosophiques. With help from Profs. Sophie Roux and Charles T. Wolfe, I located an online copy of the work, which is not very long. Merian does not seem to mention Hume or natural & philosophical relations in his Reflexions! So, Foucault's reading of Hume is not indebted to this work by Merian! (In fact, Merian merely seems here to be capturing Hume's insight of the first Enquiry 5.8 -- a work he had translated just a few years before -- that I quoted above.)
We can conclude then that here on Hume, Merian is not a source for Foucault, but that Merian is used by Foucault to articulate Hume's position (that Foucault recognizes in Merian's 1767 text). At this point one may well wonder if in the quoted passage above, Foucault is relying primary on the first Enquiry, but that would be curious here because he does directly refer to the Treatise (1.3.3 & 1.4.4--the latter a bit odd because it is Hume's account of the modern philosophy) in a endnote in a later chapter on the relationship between Hume's account of causation and its relationship to similarity.
To be sure I don't mean to suggest that you need to follow my interpretation of Hume's account of natural and philosophical relations. That's notoriously contested terrain. But rather that Foucault gets rather basic features of Hume's position wrong. Given that all the other references to Hume in The Order of Things are actually rather formulaic (and the kind of shorthand one may find in post-Kantian philosophy) one may well wonder how carefully Foucault read his Hume. (Oddly he never cites Merian again in the book.)
You may think that this is just scholarly book-keeping. That would be fine actually. However, the significance of this is that it undercuts Foucault's larger claim (which concludes the section) that "from the seventeenth century, resemblance was pushed out of the boundaries of knowledge, toward the humblest and basest of its frontiers." (71) This is, as we have seen, not true for Hume. I also would deny this is true of, say, of Newton's natural philosophy. If anything, one of the key issues in Newton's account of space and time is that Newton cane give a clear account of what Newton calls the 'species and magnitude' of the measure of space, but that he has a difficulty to do so for the measure of time. (See the postscript to my seventh chapter in my Newton's Metaphysics: Essays.) So the resemblance of spaces is much easier secured by Newton (and can do all kinds of work for him through geometric reasoning) than the putative resemblance of time. (This is no trivial matter if we look ahead to Einstein.)
I actually suspect that in the background early modern empiricism (here meaning Locke, Newton, and Hume) in all their variety are tacitly assimilated by Foucault to 20th century (Carnapian) empiricism where resemblance is indeed foundational (although problematized by Goodman and Quine). But that's just a hunch.
*It's possible, for example, that reflection gets assimilated to philosophical relations if one wrongly assumes that reflection is characteristic of philosophical activity.
Comments