In the final section of his treatise, "Of Systems of Moral Philosophy", Smith takes a step back from his own theory of morality to assess other theories of what constitutes virtue, and of how we are able to know what virtue is. Lastly, he considers the nature of the practical rules of morality.
Smith first assesses those who locate virtue mainly in propriety -- in particular, Plato, Aristotle, and the sect of philosophers known as Stoics (271). Smith has made clear by now that he holds propriety in the highest regard, much as the ancient philosophers did. Plato believed in living life virtuously by meditating on what is just; Aristotle believed in acting so that the principles of propriety are actively expressed; the Stoics believed in the expression of propriety through the passive acceptance of fortune. (In the case of adversities that indicated that the whole of life had turned against an individual, the Stoics endorsed the option of voluntary death, or suicide.) Smith's issue is that these philosophies are too passively contemplative, and lack the means to explain beneficent or hurtful actions, which merit reward or punishment in ways that cannot be explained by propriety alone.
Smith then considers the philosopher Epicurus' theory that virtue consists of prudence (294). Epicurus argues that all pleasures and pains are bodily in nature, and that virtue entails acting to maximize the former and minimize the latter. Smith disagrees with this idea on the basis that there are too many pleasures and virtues to be explained solely by bodily pleasure and pain, appealing as such explanations may be to a philosopher.
He then considers systems that locate virtue in benevolence, such as those championed by the Eclectic philosophers, as well as those of Dr. Henry More, John Smith of Cambridge, Dr. Ralph Cudworth, and Dr. Hutcheson (300-301). Such systems are based on the idea that the will of God is benevolent, and that virtue consists of obeying the will of God -- thus, virtue lies in benevolence. Smith says the problem with this idea is that benevolence may be sufficient for a perfect God, but that imperfect humans require external motives to lead virtuous existences (or existences of any sort). Further, Smith says the only logic behind following the will of God is the promise of either eternal reward or eternal damnation, or perhaps the aspiration of an imperfect, limited human being to obey an all-powerful, all-knowing God. In the former case, this philosophy is just another way of saying that virtue consists of prudence; in the latter case, it is simply a way of saying that virtue consists of propriety.
Lastly, Smith considers what he calls the "licentious system" of Dr. Mandeville, which revolves around the idea that all seemingly moral actions are actually performed out of vain self-interest (306-308). Smith denies the validity of this theory by saying that, even if people act out of self-interest in acting morally, the moral motives of self-interest -- love of glory, love of virtue -- are not vain, but admirable. Smith sees Dr. Mandeville's theory as a disservice to moral philosophy, and as one which should never have gained much traction, because it contains no sense of truth or familiarity that resonates with the people who study it. (This, for Smith, is the only type of proof to which moral theory is open.)
Smith then considers theories of how we come to understand what is virtuous: Hobbes' theory that understanding of virtue comes from self-love, Dr. Cudworth's theory that it comes from the faculties of reason, and Dr. Hutcheson's theory that it comes from a peculiar perceptual faculty (314-322). Smith refutes the idea that our notion of the virtuous comes from self-love with the counter-example of sympathy, by which we put ourselves in imaginative positions that help us to feel what others are feeling. This basic societal action demonstrates that we do not, as Hobbes suggests, participate in society purely out of self-love. Smith then refutes Cudworth's theory by citing the fact that we have an immediate, instinctive sense of the moral content of actions prior to applying reason to them -- this is how we go about forming the general rules that constitute duty. Last of all, he refutes Hutcheson's theory by pointing out that no one until recent history had naturally noticed the sensory faculty that Hutcheson discusses. Because of this, it seems vastly unlikely that such a faculty exists. And even beyond this, approbation and disapprobation are extremely sensitive to particular instances and circumstances, and possess no unifying perceptual feeling in the way that a color, a sound, or an emotion like anger does.
At the very end of his book, Smith discusses the practical rules that may be set forth regarding morality. In keeping with the difference he has already set forth between the hard-and-fast nature of justice and the loose, interpretive nature of other virtues, he rejects casuistry as a futile exercise. For Smith, casuistry glazes over the nuances of circumstance which so influence propriety. He believes instead that the entirety of practical moral rules consists of jurisprudence, wherein the specific laws for enforcing justice are delineated.
Smith's philosophical prowess is displayed to powerful effect in his refutation of counterarguments. His reasoning regarding what constitutes virtue echoes his analysis of beauty as the average form of something: as Smith says, theories that argue that virtue is grounded in propriety, prudence, or benevolence are all appealing to us because such theories all reflect true parts of our moral nature. It therefore makes sense that virtue is a composite function of these three concepts; in fact, virtue may draw in even more concepts, which Smith has not enumerated.
The piety typical of Smith's time also limits his discourse here: for instance, when considering why one would follow the will of God, Smith presupposes that it is morally right to fear impiety. One may wonder whether Smith's theory, which in some places is so reliant on God, could hold its own against fully-developed theories which do not presuppose a deity.
Smith also emphasizes that the question of how we come to understand virtue bears no relevance to the actual theory of moral sentiments, though this question is fascinating in its own right. I would add that the matter of what constitutes virtue also has little bearing on the theory itself. Smith has taken pains to define various moral characteristics -- humanity, beneficence, justice, etc. -- and the matter of what constitutes virtue does not impact how any of virtue's constituent characteristics play into our overall moral sentiments.
Thus, while this section is certainly an interesting read, the reader ought to note that it is primarily a discussion of terminology (except for the last chapter, which examines jurisprudence). Smith's theory has been laid out in its entirety by the start of this section; he is seeking only to refine broader moral concepts to better fit the context of his theory.