Smith's Theory of Moral Sentiments (TMS) tends to arouse sharply divergent reactions among the philosophers who pick it up. Kant is said to have considered it his favorite among Scottish moral sense theories, but others have dismissed it as devoid of systematic argument, or derivative, in its theoretical aspirations, of Hume. What explains these disparate reactions is one and the same feature of the book: that it consists largely of what Smith himself calls “illustrations” of the workings of the moral sentiments (TMS, “Advertisement”)—short vignettes, elegantly described, that attempt to show what frightens us about death, what we find interesting and what dull or distasteful about other people's love affairs, how moral luck factors into our assessment of various actions, or how and why we deceive ourselves. To some, this provides the detail and psychological acuity that they find lacking in most moral philosophy; to others, it seems something more properly taken up by novelists or empirical psychologists, not the business of a philosopher. Indeed, one prominent view of TMS is that it is a work in descriptive psychology or sociology, not a contribution to normative moral theory (Campbell 1971; Raphael 2007). It is hard to square this reading with the many normative judgments in TMS (see Hanley 2009, chapter 2 and Otteson 2002, chapter 6), and it misses the force of Smith's insistence that the proper way to make normative judgments is to consider the details of a phenomenon from an impartial perspective. To judge the workings of our moral faculties, then, we need to consider them, and their uses, in appropriate detail. Laying out in detail how they work can help us see how they can be corrupted, and therefore to avoid that corruption, at least to some extent (see TMS 61–6, 92–104). If this was Smith's goal—and it fits the text of TMS very well—then he was engaged not in the sociology or psychology but the phenomenology of morals, describing the workings of our modes of moral judgment as carefully as possible from within, and believing that the comprehensive view that results can itself help guide us in moral judgment. Moral phenomenology is normative moral theory, for him, and there is no more foundational theory—no set of general principles—of which we might avail ourselves. Justification for how we make moral judgments can only be found within the way we actually do make moral judgments; both moral justification and moral critique must be immanent to, not transcendent of, our moral practice (compare TMS 313–4).
A few implications of this approach. First, Smith is an anti-reductionist. He does not think morality can be reduced to a set of natural or divine laws, nor that it is simply a means for producing “the greatest happiness for the greatest number of people, ” in the phrase coined by his teacher, Frances Hutcheson. He indeed says explicitly, against the proto-utilitarianism of Hutcheson and Hume, that philosophers in his day have paid too much attention to the consequences of actions, and he wants to focus instead on their propriety: the relation they bear to the motive that inspires them (18–19). At the same time, he argues that the moral systems proposed by Samuel Clarke, William Wollaston, and Lord Shaftesbury overstress propriety, which is just one “essential ingredient” in virtuous action (294; see also 265 and 326). His own view attempts to take account of all the essential ingredients in virtue, and moral judgment, and to resist the temptation to reduce those ingredients to a single principle.
Second, and relatedly, Smith's way of approaching virtue often resembles Aristotle's—who has also sometimes been seen as too fond of the description of virtue, and who tried to acknowledge the many diverse elements of virtue, and the judgment of virtue, rather than to reduce them to a single principle. Smith says at the end of TMS that his system corresponds “pretty exactly” with Aristotle's (271). The attentive reader of TMS will have noticed this earlier: when he characterizes propriety as lying between the excess and defect of passion (27), for instance, or when he distinguishes the restraint of appetite out of self-interest from the virtue of temperance (28), or when he emphasizes habit (152, 324), or the superiority of friendships of virtue over friendships of pleasure (224–5).
Finally, Smith's phenomenological method is deeply interwoven with strong leanings toward particularism. He insists that general moral rules are “founded upon experience of what, in particular instances, our moral faculties, our natural sense of merit and propriety, approve, or disapprove of” (159; see also 160 and 320), and that our notions of right and wrong bottom out in these reactions to particular cases (320; see also 187 and Gill forthcoming). His account of virtue as depending on our attempts to adjust ourselves as closely as possible to the feelings of the particular others we encounter also suggests that what is virtue in one set of circumstances may not be so in somewhat different circumstances. These commitments entail that moral theorists will give us little moral guidance (Smith thinks moral theory should help guide moral practice: TMS 315) if they present just the general structure of right and wrong. A fine-grained phenomenology of how we carry out various kinds of moral judgment, and the errors or infelicities to which we are prone in that process, will be far more helpful.