On Bagehot's Criticism of Adam Smith's The Theory of Moral Sentiments

Professional philosophers — and perhaps other academics — have a ritual of politeness and inquisitiveness in asking each other, ‘what are you working on?’ The implicature of this question is sufficiently broad that it is perfectly legitimate to answer it in terms of what one has recently completed. A few weeks ago, my friend, Tom Holden (Santa Barbara), answered my version of this question over dinner in Los Angeles. But I was confused by his answer because he talked about his paper on a nineteenth-century philosopher he called, Badjut. Eventually, I realized he was talking about Bagehot (here)!

As it happens, the relationship between spoken language and literary style figures into Bagehot’s treatment of Smith, in his (1876) “Adam Smith as a Person:”

A really great style, indeed, Adam Smith’s certainly is not. Lord Mansfield is said to have told Boswell that he did not feel, in reading either Hume or Adam Smith, that he was reading English at all; and it was very natural that it should be so. English was not the mother tongue of either. Adam Smith had, no doubt, spoken somewhat broad Scotch for the first fourteen or fifteen years of his life; probably he never spoke anything that could quite be called English till he went to Oxford. And nothing so much hampers the free use of the pen in any language as the incessant remembrance of a kindred but different one; you are never sure the idioms nature prompts are those of the tongue you would speak, or of the tongue you would reject. Hume and Adam Smith exemplify the difficulty in opposite ways. Hume is always idiomatic, but his idioms are constantly wrong; many of his best passages are, on that account, curiously grating and puzzling; you feel that they are very like what an Englishman would say, but yet that, after all, somehow or other, they are what he never would say;—there is a minute seasoning of imperceptible difference which distracts your attention, and which you are for ever stopping to analyse. Adam Smith’s habit was very different. His style is not colloquial in the least. He adheres to the heavy “book” English which he had found in the works of others, and was sure that he could repeat in his own. And in that sort of style he has eminent merit. No one ever has to read him twice to gather his meaning; no one can bring much valid objection to his way of expressing that meaning; there is even a sort of appropriateness, though often a clumsy sort, in his way of saying it. But the style has no intrinsic happiness; no one would read it for its own sake; the words do not cleave to the meaning, so that you cannot think of them without it, or of it without them. This is only given to those who write in the speech of their childhood, and only to the very few of those—the five or six in every generation—who have from nature the best grace, who think by inborn feeling in words at once charming and accurate. (pp. 24-25)

According to Bagehot, qua stylist of the English printed language, Adam Smith has “eminent merit," and thereby ranks quite above Hume, but as an author in a living, indigenous language Smith falls far short of the natural aristocracy — the five or six in every generation — of style that Bagehot presupposes. It does not occur to Bagehot that, perhaps, Hume’s linguistic choices are deliberate; that making you stop in your tracks in order to analyze is part of his point.

Anyway, Bagehot wrote two serious essays on Smith. In addition to “Adam Smith as a Person,” there is also “Adam Smith and our Modern Economy” first published, I think, in Economic Studies (1880) and both can be found in The Works and Life of Walter Bagehot, vol. 7 (Economic Studies and Essays, a work from 1915) in the Online Library of Liberty hosted by Liberty fund. In Economics Studies, Bagehot calls Smith the initiator of English political economy, but “But what he did was much like the rough view of the first traveller who discovers a country; he saw some great outlines well, but he mistook others and left out much.” (p. 107)

“Adam Smith and our Modern Economy” has not aged very well. It is a rather tedious critique of the doctrines of Wealth of Nations often from the perspective of then modern political economy. But more often than not Bagehot’s readings of Smith’s book are not especially careful. This work has had little impact on subsequent scholarship. Somewhat oddly, while his judgment on Wealth of Nations is not much higher — he called it “a very amusing book about old times” (p. 25) — in his more influential “Adam Smith as a Person,” Bagehot does rate it much more highly than The Theory of Moral Sentiments (hereafter TMS). which he treats of “inconsiderable philosophical value.” (pp. 13-14)

In his negative judgment on TMS, Bagehot treats himself as part of a general consensus and also a corrective to earlier generations:

The Theory of Moral Sentiments was, indeed, for many years, exceedingly praised. One sect of philosophers praised it, as it seems to me, because they were glad of a celebrated ally, and another because they were glad of a celebrated oppo nent: the first said, "See that so great an authority as Adam Smith concurs with us"; and the second replied, "But see how very weak his arguments are; if so able an arguer as Adam Smith can say so little for your doctrines, how destitute of argumentative grounds those doctrines must be". Several works in the history of philosophy have had a similar fate. But a mere student of philosophy who cares for no sect, and wants only to know the truth, will nowadays, I think, find little to interest him in this celebrated book. (p. 12)

In fact, what is notable about this judgment, and Bagehot will be explicit about this a few lines down is that the “friends” of TMS,

are the school of "moral sense" thinkers, because he is on their side, and believes in a special moral faculty, which he laboriously constructs from sympathy; his enemies are the Utilitarian school, who believe in no such special faculty, and who set themselves to show that his labour has been in vain, and that no such faculty has been so built up. (p. 12)

What’s especially notable about Bagehot’s characterization is that shortly thereafter, the greatest expositor of utilitarianism, and its chief exponent in the subsequent period, Sidgwick, argues that "The theories of Hume and Adam Smith taken together anticipate, to an important extent, the explanations of the origin of moral sentiments which have been more recently current in the utilitarian school.” (Outlines of the History of Ethics, fifth edition 1905, p. 218) In fact, Sidgwick treats Smith as a useful corrective to Hume, and he denies that Smith is a moral sense theorist at all: “Adam Smith, in his Theory of Moral Sentiments (1759), like Hume, regards sympathy as the ultimate element into which moral sentiments may be analysed, and holds that there is no ground for assuming a peculiar "moral sense": nor does he dispute the actuality or importance of that sympathetic pleasure in the happy effects of virtue, on which Hume laid stress.” (p. 213)¹

Bagehot’s main argument against TMS is that “unquestionably its arguments are very weak and attractive to refutation” (p. 12-13, emphasis in original) Rather than illustrating this case by case, Bagehot offers two (related) arguments against what he takes to be TMS’ main supposition, that is, “founding morals on sympathy.” (p. 13) First, according to Bagehot this involves a confounding of two familiar sentiments. For “We often sympathise where we cannot approve, and approve where we cannot sympathise.” (p. 13) And, Bagehot suggests, more damingly yet, that ‘party spirit’ often effaces the “distinction between the two.” (p. 13)

Given that Bagehot explicitly cites the 1759 edition, it’s possible he did not read later editions of TMS. And so would have missed that Smith is especially concerned with the deleterious effects of party spirit on moral judgment. In addition, Smith emphasized more strongly that being worthy of approval is more important than approval in the structure of his meta-ethics.

So, let’s turn to the second objection, which Bagehot tends to treat as continuous with the first. I quote:

Even the wisest party men more or less sympathise with the errors of their own side; they would be powerless if they did not do so; they would gain no in fluence if they were not of like passions with those near them. Adam Smith could not help being aware of this obvious objec tion; he was far too able a reasoner to elaborate a theory without foreseeing what would be said against it. But the way in which he tries to meet the objection only shows that the objection is invincible. He sets up a supplementary theory — a little epicycle — that the sympathy which is to test good morals must be the sympathy of an "impartial spectator". But, then, who is to watch the watchman? Who is to say when the spectator is impartial, and when he is not? If he sympathises with one side, the other will always say that he is partial. As a moralist, the supposed spectator must warmly approve good actions and warmly disapprove bad actions; as an impartial person, he must never do either the one or the other. He is a fiction of inconsistent halves; if he sympathises he is not impartial, and if he is impartial he does not sym pathise. The radical vice of the theory is shown by its requir ing this accessory invention of a being both hot and cold, because the essence of the theory is to identify the passion which loves with the sentiment which approves. (p. 13)

Now, Bagehot inclines to the idea that sympathy must express partiality. And there is a good deal of common sense here. But what he misses is that such sympathetic partiality is supposed to be the effect of impartial judgments about who is the victim and who is the aggressor in contexts of, say, moral conflict. That is, Smith recognizes that if one always stays neutral in each moral conflict or disagreement, one ends up being partial to the strong or to the moral aggressor. Taking sides can be the proper effect of impartial evaluation of the facts or the truth of the matter. (American newspapers are wholly oblivious to this—and, perhaps, Bagehot’s oblivion may be due to his own journalistic practice.) Either way, the impartial spectator is not an indifferent or neutral spectator (as Bagehot seems to imply); and so it is no mere epicycle, but integral to the structure of the moral theory.

But in the quoted passage, Bagehot has another criticism. That, in practice, one may not be in a good position to judge whether one’s own impartiality in one’s sympathy is really impartial. Even if one grants Smith that as a meta-ethical construction that the impartial spectator is the source of what one ought to feel, as a matter of moral epistemology and moral practicality, it is by no means obvious it can do the job for each of us properly. And Bagehot invites the thought that in order to secure the impartiality of the impartial spectator, we would need a meta-impartial spectator, and so on by way of infinite regress.

It is my sense, that this objection may be original to Bagehot. Now, in one sense the objection misses the details of Smith’s account that true impartiality in the impartial spectator is itself quite an intellectual and socially embedded achievement of the cultivation of judgment in an individual moral agent (and her educators). But Bagehot may grant this and still worry about the regress.

How one answers Bagehot’s worry is revealing. If you think that an infinite deity is or ought to be the backstop in Smith’s theory, you may well bite the bullet here (recall the debate between the flatteners and inflaters). If you think, perhaps under the influence of Hume in some of his more literary moments, that the unquestioned conviction of moral certainty is itself a problem in moral life, you may wave away the objection as a feature and not a bug; we need to learn to live with a certain amount of practical self-doubt. Or you may suspect that a certain amount of moral overconfidence, in which the righteous have an elevated self-worth (TMS 3.3.4) — and so risk getting things badly wrong —, is necessary for the proper working of the impartial spectator, perhaps shaped by our imitation of a proper moral exemplar.

1. Somewhat awkwardly, my reading of Smith is critical of both Bagehot and Sidgwick. But that’s for another time.