But all the appetites which take their origin from a certain state of the body, seem to suggest the means of their own gratification; and even long before experience, some anticipation or preconception of the pleasure which attends that gratification. In the appetite for sex, which frequently, I am disposed to believe almost always, comes a long time before the age of puberty, this is perfectly and distinctly evident. The appetite for food suggests to the new-born infant the operation of sucking, the only means by which it can possibly gratify that appetite. It is continually sucking. It sucks whatever is presented to its mouth. It sucks even when there is nothing presented to its mouth, and some anticipation or preconception of the pleasure which it is to enjoy in sucking, seems to make it delight in putting its mouth in the shape and configuration by which it alone can enjoy that pleasure. There are other appetites in which the most unexperienced imagination produces a similar effect upon the organs which Nature has provided for their gratification.--Adam Smith (“Of the External Senses,” 79, p. 165)
Smith clearly commits himself to the existence of what (in my book-manuscript-in-progress) I label "proto-passions." In fact, the two examples he offers (appetite for sex and appetite for food) are two of the original passions (not unlike the natural sentiment of resentment).
Here, Smith is adamant that such proto-passions are innate (“long before experience…the new-born infant…the most unexperienced imagination.”) So, while it is, of course, possible that some proto-passions are themselves a consequence of habitual experience (this is implied, perhaps, by a passage that I have discussed here), Smith appears to think that a group (“there are other appetites”) of the proto-passions are innate (in his terminology: they are provided by Nature, not experience). Of course, that proto-passions are innate is compatible with the further claim that they require environmental cues or triggering objects to be activated. While below I describe some pre-conceptions that according to Smith do require such triggering objects, Smith clearly thinks some of those innate proto-passions are self-activating: the infant “sucks even when there is nothing presented to its mouth.” Even though Smith was a life-long, childless bachelor, he clearly showed an active interest in child-development (and I think this makes him so insightful on the mutual emotional regulation that Helen calls attention to).
Now, Smith’s position is not the minimal claim that prior to experience these proto-passions exist as kind of very vague drives; rather the proto-passion already have considerable content about the world built into them. This content is not just that the anticipation of the motivational significant pleasure that is built into the proto-passion, but also some of the means by which to gratify the passion. So, the infant may lack the idea or representation of a mother’s breast, but it certainly is more than blank slate about the structure of its world. As Smith puts it (in treating the self-motion of a “new-born animal”): “the very desire of motion supposes some notion or preconception of externality; and the desire to move towards the side of the agreeable, or from that of the disagreeable sensation, supposes at least some vague notion of some external thing or place which is the cause of those respective sensations.” ("External Senses," 85, EPS 167)
To be clear: the original passions that do have associated proto-passions are triggered by states of one’s body (they take “their origin from a certain state of the body” – see also The Theory of Moral Sentiments, 1.2.1) and not by (what I call) the natural triggering object that they are about (say, mother’s breast). Even so, the content of the natural proto-passion is not the state of one’s body, but much more akin to the content one would expect in the (normal) original passion. Smith’s treatment of smell, which simultaneously draws on and illuminates his account of the proto-passion that leads to infant-sucking behavior, nicely illustrates these issues:
The Smell not only excites the appetite, but directs to the object which can alone gratify that appetite. But by suggesting the direction towards that object, the Smell must necessarily suggest some notion of distance and externality, which are necessarily involved in the idea of direction; in the idea of the line of motion by which the distance can best be overcome, and the mouth brought into contact with the unknown substance which is the object of the appetite. That the Smell should alone suggest any preconception of the shape or magnitude of the external body to which it directs, seems not very probable. The sensation of Smell seems to have no sort of affinity or correspondence with shape or magnitude; and whatever preconception the infant may have of these, (and it may very probably have some such preconception,) is likely to be suggested, not so much directly by the Smell, and indirectly by the appetite excited by that Smell; as by the principle which teaches the child to mould its mouth into the conformation and action of sucking, even before it reaches the the object to which alone that conformation and action can be usefully applied.
The Smell, however, as it suggests the direction by which the external body must be approached, must suggest at least some vague idea or preconception of the existence of that body; of the thing to which it directs, though not perhaps of the precise shape and magnitude of that thing. The infant, too, feeling its mouth attracted and drawn as it were towards that external body, must conceive the Smell which thus draws and attracts it, as something belonging to or proceeding from that body, or what is afterwards denominated and obscurely understood to be as a sort of quality or attribute of that body. ("External Senses" 80-81, EPS 165-66).
According to Smith our senses have some structure of the world built into their normal functioning from the start. When we smell we automatically presuppose not just “some distance and externality, which are necessarily involved in the idea of direction,” but even “some vague idea or preconception of the existence of that body.” So, our proto-passions help deliver to us a structured external world. (In larger context of "External Senses," Smith is addressing issues generated by both Humean external-world skepticism as well as Berkeley's response to Molyneux’s Question. [Brian Glenney has written perceptively on this.] The worldly structure presupposed by our proto-passions can sometimes be relatively precise (as in the “shape or magnitude” of the aimed-for breast when the infant is sucking).
Of course, in the example above, smell is triggered by an object other than a state of one’s own body. But that does not undercut Smith’s claim here. We do not smell bodies that are not present. But the way and what we smell when triggered to do so draws on innate (proto-passion) mechanisms that have significant features of the world’s structure built into them
My argument, thus, goes against the standard reading of treating Smith as an “empiricist” (see, for example, Raphael (2007) 26, 49–50). Raphael treats the attribution of the Smith=empiricist Meme as so uncontroversial that he never bothers to offer evidence for this claim. Presumably, Raphael would cite the broad similarity between Hume’s and Smith’s philosophy—a view very common in scholarship and popular perception. Yet, Raphael fails to note that Smith is committed to the existence of the very un-Humean notion of “preconceptions.” (There are, in fact, Stoic and Epicurean precedents for attributing such structures to the native mind.) Moreover, I have not found a single place where Smith lets Hume’s so-called “copy-principle” do epistemological work for him nor does Smith ever commit to what we call “meaning empiricism.”
Of course, the absence of stated agreement is not very good evidence for Smith’s distance from Hume’s empiricism. Moreover, even if one were to accept that Smith does not share Hume’s epistemological empiricism, one can still claim that Smith is a kind of moral empiricist in the way that Hume can be taken to be one (this is, in fact, what Raphael’s claim amounts to). Even so, Smith is often a (friendly) critic of Hume whose moral psychology deviates significantly from Humean principles. An especially striking example of this is Smith’s treatment of moral reason as an active principle (see also the treatment of parental sympathy as an active principle)—a most un-Humean thought, but about that some other time more.