Adam Smith (and Buffon) on Abstraction and the measure of metaphysical-ness (or complexity)

The man who first distinguished a particular object by the epithet of green, must have observed other objects that were not green, from which he meant to separate it by this appellation. The institution of this name, therefore, supposes comparison. It likewise supposes some degree of abstraction. The person who first invented this appellation must have distinguished the quality from the object to which it belonged, and must have conceived the object as capable of subsisting without the quality. The invention, therefore, even of the simplest nouns adjective, must have required more metaphysics than we are apt to be aware of. The different mental operations, of arrangement or classing, of comparison, and of abstraction, must all have been employed, before even the names of the different colours, the least metaphysical of all nouns adjective, could be instituted.--Adam Smith (1767) Dissertation on the Origin of Languages

The Dissertation on the Origin of Languages (hereafter Languages) was added to the third edition of The Theory of Moral Sentiments (TMS). In Languages Smith develops an account in which language and mind co-develop. The argument unfolds slowly in Smith’s hands, and the passage quoted above is excerpted from a much longer argument.

The very possibility of applying what Smith calls a “noun adjective,” presupposes having the capacity for contrastive reasoning and a mental capacity for classification and conceptualization of, as well as abstraction from, these experiences. On Smith’s view objects must be conceived to be bearers of properties before adjectives can be applied to features of these objects. Moreover, in the quote Smith clearly conceives of nouns adjective with different degrees of ‘metaphysical--ness’--presumably here meant in terms of abstraction from the appearances. Abstraction turns out to be the key variable when Smith turns to prepositions: “The invention of such a word, therefore, must have required a considerable degree of abstraction… Whatever were the difficulties, therefore, which embarrassed the first invention of nouns adjective, the same, and many more, must have embarrassed that of prepositions,” (Languages).

Abstraction is a mental power. For example, traditionally, geometric properties were understood as abstracted from sensible qualities. This conception of abstraction involves a kind of subtraction or stripping away from sensory input (where one focuses on only some elements and throws away the rest). So, for example, when the French natural historian, Buffon, describes the application of geometry to nature, he writes, “it is necessary that the phenomena we are concerned with explaining be susceptible to being considered in an abstract manner and that their nature be stripped of almost all physical qualities. For mathematics is inapplicable to the extent that subjects are not simple abstractions.” (“Initial Discourse” Histoire naturelle [recall];* Smith refers to Buffon in his Letter to Edinburgh Review). Sometimes also Smith uses “abstraction” in this sense (recall here for more discussion).

But that's not Smith’s way of understanding abstraction in Languages. There, for Smith, a thought is more abstract than another thought when it requires more discrete mental “operations,” or steps, to introduce during the history of human development. So, in addition to being a mental power, abstraction is itself a kind of measure for metaphysical-ness. A thought is very abstract if starting with sensations, as a kind of baseline, it requires many further mental operations to be thought. That can include the traditional kind of abstraction, of course (if the stripping of qualities is done in careful order), but also many other kinds. So, Smith’s general picture of mind is that it is akin to a kind of machine capable of different, countable and sequential operations. Smith borrows the idea of mental operations from Hume (see, especially, Treatise 1.3.8.15, where Hume also notes “’tis very difficult to talk of the operations of the mind with perfect propriety and exactness,”) but not the treatment of metaphysical-ness.” It's a neat idea and anticipates the development of measures of complexity.

Unfortunately, Smith does not explain some of the tacit commitments he relies on here. So, for example, he seems to assume a kind of natural minimal, unit for each mental operation. He also seems to presuppose a kind of mental-least-action-principle such that in the course of human development we naturally find -- or we can rationally reconstruct post-facto -- the least required mental operations in each step of abstraction. Finally, as Steven Horst pointed out to me, it is also unclear if for Smith the degree of abstraction is ultimately dependent on empirical facts about human cognitive architecture or the logical (ahh) complexity of the represented content.

*For an excellent paper on the significance of this kind of abstraction, see Domski.

Comments

[Sveinung S. Sivertsen]

Apologies for dredging this. But could it be said that Smith here is presenting a kind of conjectural history (in the sense of Dugal Stewart), with part of it being the conjectural history of adjectives and nouns, and this being abstraction by way of discrete mental steps?

Regards,
Sveinung S. Sivertsen

[Schliesser, Eric]

Apologies for slow reply--your comment ended up in the typepad spambox! Yes, the larger context is what Stewart describes as Smith's "conjectural history." I have written a paper (Articulating Practices as Reasons, Adam Smith Review, 2006) in which criticize Stewart's conception and have offered a slightly different characterization.