Front Page Titles (by Subject) Appendix II: THE PASSAGE ON ATONEMENT, AND A MANUSCRIPT FRAGMENT ON JUSTICE - Glasgow Edition of the Works and Correspondence Vol. 1 The Theory of Moral Sentiments
Return to Title Page for Glasgow Edition of the Works and Correspondence Vol. 1 The Theory of Moral Sentiments
The Online Library of Liberty
A project of Liberty Fund, Inc.
Search this Title:
Also in the Library:
Appendix II: THE PASSAGE ON ATONEMENT, AND A MANUSCRIPT FRAGMENT ON JUSTICE - Adam Smith, Glasgow Edition of the Works and Correspondence Vol. 1 The Theory of Moral Sentiments 
The Theory of Moral Sentiments, ed. D.D. Raphael and A.L. Macfie, vol. I of the Glasgow Edition of the Works and Correspondence of Adam Smith (Indianapolis: Liberty Fund, 1982).
About Liberty Fund:
Liberty Fund, Inc. is a private, educational foundation established to encourage the study of the ideal of a society of free and responsible individuals.
The Glasgow Edition of the Works and Correspondence of Adam Smith and the associated volumes are published in hardcover by Oxford University Press. The six titles of the Glasgow Edition, but not the associated volumes, are being published in softcover by Liberty Fund. The online edition is published by Liberty Fund under license from Oxford University Press.
©Oxford University Press 1976. All rights reserved. No part of this material may be stored transmitted retransmitted lent or reproduced in any form or medium without the permission of Oxford University Press.
Fair use statement:
This material is put online to further the educational goals of Liberty Fund, Inc. Unless otherwise stated in the Copyright Information section above, this material may be used freely for educational and academic purposes. It may not be used in any way for profit.
THE PASSAGE ON ATONEMENT, AND A MANUSCRIPT FRAGMENT ON JUSTICE
TMS II.ii.3 criticizes the view that the idea of justice arises solely from utility. In editions 1–5, the chapter ends with a paragraph of orthodox theological doctrine on retributive justice. The paragraph is unusual for Smith, both in its concluding firm endorsement of Christian revelation, and in the ‘high–flying’ rhetoric of an earlier pious phrase (‘neither can he see any reason why the divine indignation should not be let loose without restraint, upon so vile an insect, as he is sensible that he himself must appear to be’—slightly toned down for edition 3). In edition 6, the paragraph was removed and replaced by a single dry sentence: ‘In every religion, and in every superstition that the world has ever beheld, accordingly, there has been a Tartarus as well as an Elysium; a place provided for the punishment of the wicked, as well as one for the reward of the just.’
This important change, made in 1788–9, would naturally lead one to think that Smith had become more sceptical about orthodox religion; or perhaps that he felt less inclination or obligation to express pious sentiments once he had quitted a Professorship of Moral Philosophy. (It is clear from the Advertisement to edition 6 that some of the revisions then made had been contemplated long before.) There has in fact been a curious controversy about possible reasons for Smith’s withdrawal of the paragraph.
William Magee, Archbishop of Dublin, published in 1801 a volume of Discourses on the Scriptural Doctrines of Atonement and Sacrifice. In edition 2, 1809, he added a number of Illustrations and Explanatory Dissertations, and in one of these (No. XXII) he quoted part of Adam Smith’s paragraph on divine justice, taking great satisfaction in the thought that the orthodox view of the Christian doctrine of Atonement was endorsed by a distinguished philosopher, ‘and he too the familiar friend of David Hume’. Elsewhere in edition 2 of his book (Dissertation No. LXIX), Magee attacked Hume along with Bolingbroke and expressed his astonishment that ‘such a man as Adam Smith’ could describe Hume, after the latter’s death, as having come as near as possible ‘to the idea of a perfectly wise and virtuous man’. The emphasis of capital letters is of course Magee’s own addition to the words that Smith had used of Hume.
What happened next is pleasantly recounted by Rae, Life, 428–9. (Rae’s first quotation from Magee is not entirely accurate.) Magee had ended his illustration from TMS by saying that the views which Smith had supported ‘as the natural suggestions of reason’ were nevertheless ‘the scoff of sciolists and witlings’. ‘The sciolists and witlings’, writes Rae, ‘were not slow in returning the scoff, and pointing out that while Smith was, no doubt, as an intellectual authority all that the Archbishop claimed for him, his authority really ran against the Archbishop’s view and not in favour of it, inasmuch as he had withdrawn the passage relied on from the last edition of his work.’ Magee tried to extricate himself from his discomfiture by adding a footnote in edition 3 of his own book, 1812, attributing Smith’s withdrawal to ‘the infection of David Hume’s society . . . one proof more . . . of the danger, even to the most enlightened, from a familiar contact with infidelity’; and then Magee referred again to Smith’s obituary praise of Hume which had shocked so many of the conventionally religious. Rae himself joins in ‘returning the scoff’ with the comment that Smith’s ‘intercourse with Hume was at its closest when he first published the passage in 1759, whereas Hume was fourteen years in his grave when the passage was omitted’. But Magee was under a misapprehension. He thought (and, as we shall see, he was not the only one to think) that the passage had been withdrawn long before edition 6. In the added footnote in which he refers to the influence of Hume, Magee writes: ‘The fact is, that in the later editions of the Theory of Moral Sentiments, no one sentence appears of the extract which has been cited above, and which I had derived from the first edition, the only one that I possessed.’
Rae goes on to say that ‘there is no reason to believe that Smith’s opinion about the atonement was anywise different in 1790 from what it was in 1759, or for doubting his own explanation of the omission, which he is said to have given to certain Edinburgh friends, that he thought the passage unnecessary and misplaced’. The report of this explanation is in vol. ii, 40, of Memoirs of the Life and Works of Sir John Sinclair (1837) by his son, the Rev. John Sinclair. What Archdeacon Sinclair actually says of Smith is this:
In the second edition of his Theory of Moral Sentiments, he omitted, on the suggestion, as is supposed, of his sceptical friend, a splendid passage, referred to by Archbishop Magee as among the ablest illustrations of the doctrine of Atonement. In the suspicion thus excited, my father did not participate. He was anxious to think favourably of a venerated friend. Smith himself justified the omission alluded to, not on the ground that the doctrine of Atonement was unfounded, but that the paragraph was unnecessary and misplaced.
When Archdeacon Sinclair states that the passage was omitted in edition 2, he is obviously just following what he took to be the natural sense of Magee’s footnote and has not himself looked at the relevant editions of TMS. Sinclair goes on to evoke a moving image of Smith on his knees at his mother’s deathbed, praying to the Redeemer whom he was alleged to have rejected. This is supposed to be evidence either that the withdrawal of the passage on the Redeemer did not imply any loss of faith or that the faith was later regained. In fact Smith’s mother died (in 1784) several years before the withdrawal in edition 6. It is not clear from Archdeacon Sinclair’s account whether Smith’s reasons for the omission of the paragraph were given directly to Sir John Sinclair himself or, as Rae apparently infers, to others in Edinburgh. Presumably Archdeacon Sinclair was told the story by his father. If Smith did give the explanation to anyone, it must have been within the short space of time that intervened between the publication of edition 6 and his death. That would have happened only if this particular revision (a minor one, compared with others) had been noted and had excited remark very soon after publication. Oddly enough, Rae himself, despite the reference to Sinclair on p. 429 of the Life, says on p. 428 that ‘the suppression of the passage about the atonement escaped notice for twenty years’ until Magee quoted it.
After mentioning Sinclair’s report, Rae then writes:
As if taking an odd revenge for its suppression, the original manuscript of this particular passage seems to have reappeared from between the leaves of a volume of Aristotle in the year 1831, when all the rest of the MS. of the book and of Smith’s other works had long gone to destruction.
At the end of this sentence we are referred to a footnote, which simply says ‘Add. MSS., 32,574’, and so suggests that the manuscript which came to light in 1831 is now in the British Library (the British Museum). In fact this is not so. Additional Manuscript 32,574 in the British Library is Volume XVI of the Notebooks of the Rev. John Mitford (1781–1859), and it is the source of Rae’s information. The first entry in this volume is signed ‘J. Mitford’ and is dated ‘1855. Sept. 26.’ On leaf number 64, there is the following note:
on Adam Smith’s Moral Sentiments. ‘I am sorry to find sd Bp. Bathurst, that his Splendid Passage on the necessity of a Redeemer, was omitted in the Second Edition.’ The omission probably owing to his Acquaintance with Hume.
Bp Bathurst and Chalmers. +
‘Second Edition’ and ‘Splendid Passage’ show that Bathurst was simply relying on Sinclair’s book. The sign + at the end of Mitford’s original entry was presumably added later, together with this note on the facing verso of leaf 63:
A. Smith’s injunctions to his Executors to destroy all his loose Mss. were strictly followed, but that Passage so long[?] preserved, reappeared from between the folds of a Volume of Aristotle in 1831.
Discovered by Revd W. B. Cunningham of Preston Pans into whose hands Dr Smith’s Library had passed
The Rev. W. B. Cunningham of Prestonpans was the husband of one of the two daughters of David Douglas, Lord Reston, the cousin to whom Adam Smith bequeathed his books and other possessions. On the death of Lord Reston, Adam Smith’s library was divided between the two daughters. (See Bonar, Catalogue 2, intro. xvi–xvii.)
W. R. Scott, ASSP, 57–9, describes one of four manuscripts in the possession of the Glasgow University Library. This particular manuscript consists of one folio sheet of four pages, and the four documents together contain fifteen pages of writing. Scott says:
The date of these four documents is of great importance. The many avocations of Adam Smith during the first eight years he was at Glasgow [i.e. 1751–9] make it highly improbable, if not impossible, that they could have been written then, and thus they may be assigned to the Edinburgh period [i.e. 1748–51].
The one manuscript which is relevant to the present discussion is taken by Scott to be ‘introductory to a group of lectures’ on jurisprudence, delivered in Edinburgh and corresponding to the Glasgow lectures on jurisprudence, a Report of which was published by Edwin Cannan in 1896. Scott describes the manuscript as follows:
There had been a very brief account of moral obligation, and the surviving manuscript begins with the statement that ‘duty, for its own sake and without any further view, is the natural and proper object of love and reward, and vice of hatred and punishment’. Here follow the sentences on the Atonement, which appeared in the first five editions of the Theory of Moral Sentiments with small alterations. At this point in the Theory a chapter ends, and in the next a different aspect of the subject is begun. Here [i.e. in the manuscript] the discussion continues with material rewards and punishments. The sentry found asleep at his post is discussed, then the argument passes on to the institution of the civil magistrate. The authority of custom or statute law is traced back to the natural principles of justice, and the study of the rules which express it constitutes Natural Jurisprudence or the Theory of the General Principles of Law. Adam Smith indicates that he will give a particular discourse upon that subject. The concluding part of this paper discusses the relation between Justice and Benevolence and between the latter and resentment and punishment. No doubt the lectures went on (as indicated) to discuss how far these principles find expression in existing legal systems.
In footnote 2 to p. 58, Scott writes:
This manuscript may be that which was found in a volume of Aristotle in 1831 (Rae, Life, p. 261 [error for p. 429]) and described as that of a part of the Theory of Moral Sentiments. The beginning of it very closely resembles the corresponding part of the Theory. The remainder is quite different.
In footnote 5 to p. 320 of his book, Scott refers again to what he calls ‘the celebrated sentences on the Atonement’, and writes:
There is no reason to doubt Adam Smith’s own statement that they were withdrawn in the sixth edition of the Theory (1790) because they were misplaced. Rae mentions (Life, p. 429) that the manuscript, containing the sentences, was found in a volume of Aristotle in the year 1831. Whether Rae intended it or not, this has been taken to mean that the fragment discovered was a part of the Theory. It was not, being the part of the Edinburgh Lectures which is described in Part I, Chapter V [i.e. pp. 57–9]. This MS. begins with the conclusion of a discussion which is that ‘duty for its own sake’ is ‘the natural and proper object of love and reward’. As first written this lecture went on to examine in some detail the principles of legal sanctions. At an early revision the sentences on the Atonement were inserted. This obviously was not a happy arrangement. In the Theory the passage was expanded and made to close Part II, Section ii, Chapter iii. Then follows Section iii, which discusses the influence of fortune upon the sentiments of mankind, which is far from being an ideal collocation.
Scott did not print the text of the manuscript described in these quotations. It is given below, but first some comment needs to be made on several points in Scott’s account.
(1) Scott’s ground for assigning this and the other three short manuscripts (amounting in all to fifteen pages) to the period of the Edinburgh lectures is extraordinarily flimsy. Since Smith had the time, between 1751 and 1759, not only to write TMS but to compose courses of lectures which extended beyond the subject–matter of that book to natural theology, jurisprudence, and economics, why should it be supposed that his ‘many avocations’ made it ‘highly improbable, if not impossible,’ for him to compose these pieces which are all concerned with subjects that were included in his lectures as Professor of Moral Philosophy?
(2) Scott’s assertion that the initial words of the manuscript had been preceded by ‘a very brief account of moral obligation’ (how did he know that it was ‘very brief’?) is a figment of his imagination, produced by a misreading of the first word of the manuscript, which is ‘Deity’, but which Scott took to be ‘Duty’. In any case, Smith would never have said that duty is ‘the natural object of love and reward, and vice of hatred and punishment’. Obviously virtue is what corresponds to vice in this connection. The manuscript begins with words that do not make a complete sentence: ‘Deity, as it does to us, for its own sake and without any further view the natural and proper object of Love and Reward and Vice of hatred and punishment.’ Scott must have supposed that the word ‘is’ had been inadvertently omitted before ‘the natural and proper object’, and he conveniently ignored the phrase ‘as it does to us’. Clearly the word ‘Deity’ will have been preceded by some such words as ‘Virtue appears to the’. In editions 1–2 of TMS, the paragraph that was withdrawn from edition 6 contains the following sentence (in editions 3–5, the first words are revised to ‘Our untaught, natural sentiments, all’):
All our natural sentiments prompt us to believe, that as perfect virtue is supposed necessarily to appear to the Deity, as it does to us, for its own sake, and without any further view, the natural and proper object of love and reward, so must vice, of hatred and punishment.
(3) In the footnote to his p. 58, Scott says that the manuscript ‘may be’ that which was found in a volume of Aristotle in 1831, though in the footnote to p. 320 he takes for granted that it was. Scott evidently did not look up Rae’s reference to the British Library manuscript, but there is in fact positive proof that the Glasgow manuscript is the one referred to in Mitford’s second note. Mitford says the manuscript was ‘Discovered by Revd W. B. Cunningham of Preston Pans’. In the margin of the first page of the Glasgow manuscript, there is written, in a later hand:
Free Church Manse Prestonpans.
(4) In the footnote to his p. 320, Scott says there is no reason to doubt Smith’s own statement that the paragraph of editions 1–5 was withdrawn because it was ‘misplaced’. This gives only half of the reason as reported by Sinclair and Rae, namely that Smith thought the passage ‘unnecessary and misplaced’. The addition of ‘unnecessary’ makes a difference.
(5) Judging from what Scott says in the note to his p. 320, he appears to think that the manuscript ‘as first written’ did not contain anything about the Atonement. ‘At an early revision the sentences on the Atonement were inserted.’ In fact, the manuscript ‘as first written’ had this as its second sentence:
The Justice of the Deity we think cannot surely be satisfied with [error for ‘without’] demanding some attonement, some expiation for the Offences of Mankind, and Revelation teaches us that this attonement has not only been demanded but has been paid for, at least, the more valuable part of Mankind.
Later, two sentences, corresponding to further words in the paragraph of editions 1–5 of TMS, were inserted before the sentence just quoted, but they are about our consciousness of human imperfection in the sight of God rather than about the idea of atonement.
(6) The remainder of the manuscript, that is to say, by far the greater part of it, does not correspond to anything in the so–called paragraph on ‘atonement’; but many of the later words of the manuscript correspond to other passages in the printed texts of TMS, and on the second page there occur the very words ‘the Theory of moral Sentiments’. Consequently some caution is needed before accepting Scott’s hasty conclusion that the manuscript was not a part of TMS but belonged to the Edinburgh lectures.
Having cleared Scott’s preconceptions from our path, it will be best to describe the manuscript anew, and to give its full text, before discussing further its relation to the printed editions of TMS.
The manuscript (Glasgow University Library, MS. Gen. 1035/227) was originally a single folio sheet (the two halves of which have now come apart) of four pages. The watermarks, briefly mentioned by Scott in a footnote ending on his p. 266, are similar to two of those which Scott describes on his p. 322, one of them being illustrated in Plate XV which faces that page. One half of the sheet has as its watermark ‘G.R.’ within a circular emblem and surmounted by a crown, with the word ‘Durham’ beneath. The other half–sheet has the watermark of Britannia within a picket fence, and the motto ‘Pro Patria’, as illustrated in Scott’s Plate XV, but in reverse; that is to say, Britannia is sitting on the left and facing right, with the motto at her right, while in Scott’s plate she is sitting on the right and facing left, with the motto at her left. The writing on the manuscript covers the whole of the first three pages, and three–quarters of the fourth page, indicating that it then came to an end. Catchwords at the foot of the first three pages make quite clear the order in which they were written. Each of the pages has a margin at the left, and the top half of the margin on the first page contains a lengthy insertion, preceded by a figure 2 or a sign resembling it, which Adam Smith was evidently accustomed to use for this purpose, as can be seen from two of the other manuscripts which were reproduced in facsimile by Scott (ASSP, 381, 383, 385). The place, in the original writing, at which the addition is to be inserted, is likewise indicated by a figure 2. The bottom half of the margin of the first page contains the later entry made by Mr. Cunningham. The margins on the other three sheets are left blank.
In the text of the manuscript, the words originally written, and most of the revisions and insertions, are in the hand of an amanuensis. Two or three of the revisions, however, are in the hand of Adam Smith himself; e.g. the word ‘men’ substituted for ‘Mankind’ in the first paragraph; and the figure 2 at the end of the first sentence, though not the corresponding figure 2 that precedes the inserted passage in the margin. Some of the remaining revisions are written above the relevant line. Others, however, are written on the same line as, and immediately after, cancelled words, showing that Smith made some changes as he dictated the piece. This feature of the document is one piece of evidence for the conclusion that it was written before the manuscript actually used for edition 1 of TMS.
In the text that follows, square brackets enclose words or letters that are struck out or over–written in the manuscript, while angle brackets enclose words or letters that constitute revisions of cancelled material or later insertions. It will be recalled that the manuscript begins in the middle of a sentence.
Deity, as it does to us, <for its own sake and without any further view> the natural and proper object of Love and Reward and Vice of hatred and punishment.<2> <2. Nay vice we are apt to fear should appear before the holiness of God more worthy of punishment than the imperfection of human Virtue can ever be of Reward. Man when about to appear before a Being of such perfect Sanctity can feel but little Confidence in his own merit[.]<;> [But the divine Justice etc.] and when he remembers the numberless blemishes and imperfections in his own Conduct must dread punishment rather than hope for Reward. The divine Justice etc.> The <divine> Justice [of the Deity] we think cannot surely be satisfied with<out> demanding some attonement, some expiation for the Offences of [Mankind]<men>, and Revelation teaches us that this attonement has not only been demanded but has been paid for, at least, the more valuable part of Mankind.
[There are indeed] Upon some occasions indeed we punish meerly from a View to the general interest of Society which [cann] we imagine cannot be otherwise supported. The punishments, for Example, which military discipline prescribes are all inflicted from this motive, and a Centinel who falls asleep upon his Watch [is] suffers death by the Laws of War because such carelessness might endanger the whole Army. In our hearts we cannot blame this necessary Severity. Nothing can be more just, than that one man [can]<should> be sacrificed to the security of thousands. But do we regard th[e]<is> punishment in the same light in which we look upon that of an ungrateful murderer or parricide[;]<?> Does our heart naturally applaud1 the same Ardor with which it goes along with the other? We look upon the one as an unfortunate Victime who indeed must be devoted to the interest of Numbers but whom in our hearts we would be glad to save, and we are only sorry that the Interests of [others] many should oppose it. If the other should escape from punishment it would excite our highest indignation and we [w]<sh>ould call upon God to avenge <in another world> that Crime which the injustice of Mankind had neglected to chastise upon Earth.
The violation of Justice is what Mankind will never submit to from their Equals. It provokes the Resentment of the injured and incites them to take vengeance upon the Offender. They feel that Mankind applaud and go along with <t>h[i]<e>m when they punish him,2 and they imagine that they become contemptible when they do not. That civil Society may not be a Scene of Bloodshed <confusion> and disorder every man revenging himself at his own hand whenever he fancies himself injured, the Magistrates in all Governments that have acquired considerable Authority employs the power of the commonwealth to enforce the practice of Justice, and to give Satisfaction to the injured either by punishing the offender or by obliging him to compensate the wrong that has been done. The magistrate promises to hear all complaints of injustice, to enquire diligently into the circumstances alledged upon both Sides, and to give that redress which to any impartial person shall appear to be just and equitable. Hence the origin of both civil and criminal Jurisdiction. The Rules by which the magistrate[s] in [all] <each> countries actually regu[l]lates all his discisions of this kind [which]<whether> established upon express Statute, upon acc[o––?]<iden>tal custom or upon their own evident equity constitute the civil and criminal Jurisprudence of that Country. The Rules by which it is most suitable to the natural principles of Justice, or to the Analogy of those Sentiments upon which our Sense of it is founded that such descisions should be regulated, const[–?]<i>tute what is called Natural Jurisprudence, or the Theory of the general principles of Law. they make a very important part of the Theory of moral Sentiments. I shall not at present, however, stop to analyse them, as I intend hereafter to give a particular discourse upon that Subject.
When our benevolence to each particular person is exactly proportioned to the importance of those circumstances which point them out to our favourable regard, we are, by a metaphor, said to do them Justice[:]<;> and we are said to do them injustice when it is otherwise. When we chuse [–?]<r>ather, for exemple, to do a good Office to a new acquaintance than to an Old friend we are said to do Injustice to the latter. This, however, is a different Species of Injustice from that which we have been treating of above. It does not consist in doing hurt, but in not doing good according to the most perfect propriety. In the Schools it has been distinguished by the name of d[e?]<i>stributive Justice, as the former, which can alone properly be called Justice, has been denominated commutat[––?]<iv>e Justice. In the observation of distributive Justice consists the proper exercise of all the social and beneficent Virtues. It cannot be extorted by force. The violation of it does no positive harm, and therefor, exposes to no punishment. The Rules which determine the external actions which it prescribes, are loose and unaccurate and fall short of that exact pre[s]<c>ision, which, as I shall show hereafter, is peculiar to the Rules of what is properly called Justice. The Rules of punishment have been by most Writers referred to distributive Justice as well as the Rules of Beneficence, and they seem to have imagined that improper vengea[––?]<nc>e was an impropriety of the same kind with improper Benevolence. There is indeed a certain degree of looseness and inaccuracy [of] in what may be called the natural principles of punishment. What is the extent of the Right which is violated, and wherein consists its v[e?]<i>olation, can in almost all cases be determined with exact precision. But what degree of Resentment or punishment is due for this violation cannot easily be fixed exactly by general Rules which have any great foundation in nature; but varies with every variety of Circumstances: And so far the principles and rules of punishments resemble those of Beneficence. But they differ from them in another Circumstance which is much more essential, and which those Writers have not perhaps, sufficiently attended to. Improper punishment, punishment which is either not due at all or which exceeds the demerit of the Crime, is an injury to the Criminal, may and ought to be opposed by force, and if inflicted, exposes the person who inflicts it to punishment in his turn. But meer improper Beneficence cannot be opposed by force and exposes the person who exercises it to no punishment.
Compare the following set of extracts from TMS. References are given to the arrangement of chapters and paragraphs in the present (and so in the sixth) edition, but since the manuscript is earlier than any of the printed texts, the actual words and punctuation of the quotations are taken (except for the sixth extract) from edition 1.
All our natural sentiments prompt us to believe, that as perfect virtue is supposed necessarily to appear to the Deity, as it does to us, for its own sake, and without any further view, the natural and proper object of love and reward, so must vice, of hatred and punishment. . . . If we consult our natural sentiments, we are apt to fear, lest before the holiness of God, vice should appear to be more worthy of punishment than the weakness and imperfection of human virtue can ever seem to be of reward. Man, when about to appear before a being of infinite perfection, can feel but little confidence in his own merit, . . . he can easily conceive, how the numberless violations of duty, of which he has been guilty, should render him the proper object of aversion and punishment; . . . Some other intercession, some other sacrifice, some other atonement, he imagines, must be made for him, beyond what he himself is capable of making, before the purity of the divine justice can be reconciled to his manifold offences. The doctrines of revelation . . . show us . . . that the most powerful intercession has been made, and that the most dreadful atonement has been paid for our manifold transgressions and iniquities.
(II.ii.3, final paragraph—the one suppressed in edition 6)
Upon some occasions, indeed, we both punish and approve of punishment, merely from a view to the general interest of society, which, we imagine, cannot otherwise be secured. Of this kind are all the punishments inflicted for breaches of what is called either civil police, or military discipline. . . . A centinel, for example, who falls asleep upon his watch, suffers death by the laws of war, because such carelessness might endanger the whole army. This severity may, upon many occasions, appear necessary, and, for that reason, just and proper. When the preservation of an individual is inconsistent with the safety of a multitude, nothing can be more just than that the many should be preferred to the one. Yet this punishment, how necessary soever, always appears to be excessively severe. The natural atrocity of the crime seems to be so little, and the punishment so great, that it is with great difficulty that our heart can reconcile itself to it. . . . A man of humanity . . . must make an effort . . . before he can . . . go along with it . . . It is not, however, in this manner, that he looks upon the just punishment of an ungrateful murderer or parricide. His heart, in this case, applauds with ardour, and even with transport, the just retaliation which seems due to such detestable crimes, . . . He looks upon the centinel as an unfortunate victim, who, indeed, must, and ought to be, devoted to the safety of numbers, but whom still, in his heart, he would be glad to save; and he is only sorry, that the interest of the many should oppose it. But if the murderer should escape from punishment, it would excite his highest indignation, and he would call upon God to avenge, in another world, that crime which the injustice of mankind had neglected to chastise upon earth.
There is, however, another virtue, . . . of which the violation exposes to resentment, and consequently to punishment. This virtue is justice: the violation of justice is injury: . . . It is, therefore, the proper object of resentment, and of punishment, . . . As mankind go along with, and approve of, the violence employed to avenge the hurt which is done by injustice, so they much more go along with, and approve of, that which is employed to prevent and beat off the injury, . . .
Among equals each individual is naturally . . . regarded as having a right both to defend himself from injuries, and to exact a certain degree of punishment . . .
As the violation of justice is what men will never submit to from one another, the publick magistrate is under a necessity of employing the power of the commonwealth to enforce the practice of this virtue. Without this precaution, civil society would become a scene of bloodshed and disorder, every man revenging himself at his own hand whenever he fancied he was injured. To prevent the confusion which would attend upon every man’s doing justice to himself, the magistrate, in all governments that have acquired any considerable authority, undertakes to do justice to all, and promises to hear and to redress every complaint of injury. In all well–governed states too not only judges are appointed for determining the controversies of individuals, but rules are prescribed for regulating the decisions of those judges; and these rules are, in general, intended to coincide with those of natural justice. . . . In no country do the decisions of positive law coincide exactly in every case with the rules which the natural sense of justice would dictate.
The wisdom of every state or commonwealth endeavours . . . to restrain those who are subject to its authority, from hurting or disturbing the happiness of one another. The rules which it establishes for this purpose, constitute the civil and criminal law of each particular state or country. The principles upon which those rules either are, or ought to be founded, are the subject of a particular science, of all sciences by far the most important, but hitherto, perhaps, the least cultivated, that of natural jurisprudence; concerning which it belongs not to our present subject to enter into any detail.
(VI.ii.intro.2. This passage was first added in edition 6.)
It might have been expected that the reasonings of lawyers . . . should have led them to aim at establishing a system of what might properly be called natural jurisprudence, or a theory of the general principles which ought to run through and be the foundation of the laws of all nations. . . . I shall in another discourse endeavour to give an account of the general principles of law and government, . . . I shall not, therefore, at present enter into any further detail concerning the history of jurisprudence.
In one sense we are said to do justice to our neighbour when we abstain from doing him any positive harm, and do not directly hurt him, . . . This is that justice which I have treated of above, the observance of which may be extorted by force, and the violation of which exposes to punishment. In another sense we are said not to do justice to our neighbour unless we conceive for him all that love, respect and esteem, which his character, his situation, and his connection with ourselves, render suitable and proper for us to feel, and unless we act accordingly. It is in this sense that we are said to do injustice to a man of merit who is connected with us, tho’ we abstain from hurting him in every respect, if we do not exert ourselves to serve him . . . The first sense of the word coincides with what Aristotle and the Schoolmen call commutative justice, and with what Grotius calls the justitia expletrix, which consists in abstaining from what is anothers, and in doing voluntarily whatever we can with propriety be forced to do. The second sense of the word coincides with what some have called distributive justice [Added footnote: ‘The distributive justice of Aristotle is somewhat different . . .’], and with the justitia attributrix of Grotius, which consists in proper beneficence, in the becoming use of what is our own, and in the applying it to those purposes either of charity or generosity, to which it is most suitable in our situation that it should be applied. In this sense justice comprehends all the social virtues. There is yet another sense in which the word justice is sometimes taken, . . . Thus we are said to do injustice to a poem or a picture, when we do not admire them enough, . . . In the same manner we are said to do injustice to ourselves when we appear not to give sufficient attention to any particular object of self–interest. In this last sense, what is called justice means the same thing with exact and perfect propriety of conduct and behaviour, . . .
The decision of this question . . . will depend . . . secondly, upon the precision and exactness, or the looseness and inaccuracy of the general rules themselves.
Secondly, I say, it will depend partly upon the precision and exactness, or the looseness and inaccuracy of the general rules themselves, . . .
The general rules of almost all the virtues . . . are in many respects loose and inaccurate, . . .
There is, however, one virtue of which the general rules determine with the greatest exactness every external action which it requires. This virtue is justice.
Beneficence is always free, it cannot be extorted by force, the meer want of it exposes to no punishment: . . .
The printed editions of TMS do not contain several sentences found towards the end of the manuscript, concerning the difference between improper vengeance or punishment and improper benevolence. It may be thought that this is because Adam Smith wanted to reserve the topic for his projected book on jurisprudence, but a more likely explanation is that he had changed his view by the time he came to publish TMS. Two points require notice.
(1) In the manuscript, Smith says, of the natural principles of punishment, that the extent and character of the violation of a right can be determined with precision, but not the degree of resentment or punishment due, since this latter varies with circumstances. Now there is evidence in the manuscript that, at the time when Smith dictated it, he had not yet thought out his theory of the impartial spectator (a theory which underwent considerable development between the publication of editions 1 and 6 of TMS, as can be seen in the elaboration of the account of conscience in Part III, first for edition 2 and then again for edition 6). When discussing the function of the magistrate, the manuscript says that he ‘promises to hear all complaints of injustice, . . . and to give that redress which to any impartial person shall appear to be just and equitable’. TMS reproduces this simply as ‘promises to hear and to redress every complaint of injury’. If Smith had included at this time the reference to ‘any impartial person’, he would certainly have written of the impartial ‘spectator’ instead. Once he had formulated his theory of the impartial spectator, he of course took the view that the proper degree of resentment or punishment was that which had the sympathy of the impartial spectator, as we can see from II.ii.2 of TMS. This chapter relates to resentment. That Smith would hold the same view of punishment is obvious enough, but can be confirmed from the two extant Reports of his lectures on jurisprudence. LJ(A), a full Report of lectures delivered in 1762–3, contains the following sentences at ii.89–90.
Now in all cases the measure of the punishment to be inflicted on the delinquent is the concurrence of the impartial spectator with the resentment of the injured. If the injury is so great as that the spectator can go along with the injured person in revenging himself by the death of the offender, this is the proper punishment, and what is to be exacted by the offended person or the magistrate in his place who acts in the character of an impartial spectator. . . . In all cases a punishment appears equitable in the eyes of the [unconcerned spectator] <rest of mankind> when it is such that the spectator would concur with the offended person in exacting.
In LJ(B), a summarized version of lectures delivered in 1763–4, the corresponding passage is at 181 (Cannan ed., 136): ‘Injury naturaly excites the resentment of the spectator, and the punishment of the offender is reasonable as far as the indifferent spectator can go along with it. This is the natural measure of punishment.’
Consequently Smith would no longer accept the view that the rules of punishment resemble those of beneficence in being imprecise. That is why the relevant sentences of the manuscript are not reproduced in TMS, either at II.ii.1, where justice and beneficence are compared, or at III.6.8–10, where Smith distinguishes the precision of the rules of justice from the looseness of the rules of other virtues.
(2) Having noted an apparent similarity between improper punishment and improper benevolence, the manuscript goes on to contrast them, in that improper punishment may and ought to be opposed by force and renders the inflicter of it liable to punishment in his turn as having done an injury, while improper beneficence cannot be opposed by force and exposes to no punishment. The conclusion about ‘improper’ (i.e. want of proper) beneficence is reproduced in TMS at II.ii.1.3, but not that about improper punishment. It is indeed surprising that nowhere in TMS does Smith repeat the statement in the manuscript that ‘Improper punishment, punishment which is either not due at all or which exceeds the demerit of the Crime, is an injury to the Criminal’. This is not only a sound expression of what Smith would call our ‘natural moral sentiments’; it is a point which one would expect Smith, as an upholder of the retributive or desert theory of punishment, to include in his criticism of the utilitarian account of justice. Why does he not do so?
The fact is that Smith found himself in a cleft stick on this issue and had not thought out his position consistently. In the manuscript he says that the sentinel is punished for reasons of utility (‘meerly from a View to the general interest of Society’), but he then writes: ‘In our hearts we cannot blame this necessary Severity. Nothing can be more just [our italics], than that one man should be sacrificed to the security of thousands.’ In the printed text of TMS, the sentences just quoted are modified and elaborated.
This severity may, upon many occasions, appear necessary, and, for that reason, just and proper. When the preservation of an individual is inconsistent with the safety of a multitude, nothing can be more just than that the many should be preferred to the one. Yet this punishment, how necessary soever, always appears to be excessively severe. The natural atrocity of the crime seems to be so little, and the punishment so great, that it is with great difficulty that our heart can reconcile itself to it. Though such carelessness appears very blameable, yet the thought of this crime does not naturally excite any such resentment, as would prompt us to take such dreadful revenge.
Smith is still prepared to say it is ‘just’ (as well as ‘proper’) to inflict, for utilitarian reasons, a punishment whose severity exceeds the ‘natural atrocity’ of the crime. But in these circumstances he could not say elsewhere, as the manuscript does, that ‘punishment . . . which exceeds the demerit of the Crime, is an injury to the Criminal’, for ‘injury’ means a breach of justice.
The extant Reports of the lectures on jurisprudence quote Smith as again using the example of the sentinel. In LJ(B) 182 (Cannan ed., 136), he is still prepared to call the punishment ‘just’: ‘if a centinel be put to death for leaving3 his post, tho’ the punishment be just and the injury that might have ensued be very great, yet mankind can never enter into this punishment as if he had been a thief or a robber.’ But in LJ(A) ii.92, Smith expresses himself more cautiously:
In the same manner the military laws punish a centinell who falls asleep upon guard with death. This is intirely founded on the consideration of the publick good; and tho we may perhaps [our italics] approve of the sacrificing one person for the safety of a few, yet such a punishment when it is inflicted affects us in a very different manner from that of a cruel murtherer or other attrocious criminall.
Apart from the few sentences comparing improper punishment with improper benevolence, the whole of the substance of our manuscript fragment, often with the self–same words, is included in different parts of TMS. It is interesting to observe that, even when writing new material for edition 6, Smith was prepared to repeat some of the thought of the manuscript fragment, and even to introduce a brief phrase (but perhaps only by chance) that had occurred in the manuscript and that he had not previously used in TMS.
There can be no doubt that the manuscript is earlier than edition 1 of the book. As we have already observed, some of the corrections in the manuscript were made at the first dictation of the material, and it is the revised words, together with insertions made subsequently, that find a place in edition 1 of TMS. Often, too, the version in the printed text expands and improves upon the thought of the manuscript. The discussion of the sentinel is one, but not the only, clear instance of such improvement. Then again there is the evidence already cited that in the manuscript the theory of the impartial spectator has not yet been explicitly formulated.
Having established that the manuscript preceded the one submitted to the printer for edition 1 of the TMS, we can now consider for what purpose it was written. There are two possibilities, (1) that it was part of an early draft of the book, and (2) that it was part of a lecture. The first hypothesis receives some support from one piece of evidence, namely the occurrence in the manuscript of the very words ‘the Theory of moral Sentiments’. As against that, however, there are two considerations which point, one of them strongly, to the alternative hypothesis of a lecture.
First, in the manuscript Smith states his intention ‘hereafter to give a particular discourse’ on natural jurisprudence or the theory of the general principles of law. In the last paragraph of TMS he says he will endeavour ‘in another discourse’ to give an account of the general principles of law and government, and of the history of jurisprudence. Here ‘another discourse’ means, of course, another book; but that cannot be the meaning of the words in the manuscript, ‘I intend hereafter to give a particular discourse’. To give a particular discourse can only mean to deliver a lecture, or possibly a series of lectures.
The second consideration is less compelling, though worth mentioning. The manuscript stops before the end of a page, as do the two manuscripts of similar length reproduced in facsimile by Scott in ASSP, 379–85. Scott mentions (58) that these three manuscripts differ from the larger ‘early draft of part of The Wealth of Nations’ (the text of which he prints on pp. 322–53) in that a new chapter, in the latter work, follows on from the previous one, on the same page, if there is room, instead of beginning on a fresh page. The first printed versions of TMS and WN often begin a new chapter on the same page as the end of the previous chapter, if there is adequate space left, and one can infer that this was Smith’s practice in the manuscripts for those books, since the printers of the first editions appear to have followed their copy closely in other respects. We can therefore agree with Scott that the blank space at the end of three of the short manuscripts, including the one discussed here, is a reason for regarding them as the final portions of lectures.
Whatever may be said, however, of the two other fragments, which deal with economics,4 there is no reason to assign our particular manuscript to Smith’s Edinburgh lectures, which, so far as we know, did not deal with ethics. Scott attributed this lecture to the Edinburgh period because he took it to be an introduction to lectures on law and because he thought Smith must have been too busy, during his first years in Glasgow, to write about law and economics. Since it is now perfectly clear that this particular manuscript covers the subject–matter of parts of TMS, the obvious conclusion is that it comes from one of the lectures which he gave as Professor of Moral Philosophy at Glasgow.
The lecture was, of course, on justice, and the single sentence about the Christian doctrine of the Atonement is relatively incidental. Smith’s retention of this sheet cannot possibly have been due to any desire he might have had to re–arrange the position in TMS of the paragraph on divine justice. If he did have that purpose, he would have kept the whole paragraph, not just a sheet which begins in the middle of it; and in any event he would have used the fuller version of the paragraph that was printed in editions 1–5, and not the shorter version that he had originally written for his lecture course. It is perhaps idle to speculate on the reasons why this particular sheet happened to be placed in one of Smith’s books and so preserved, but if conjecture may be allowed, one can hazard a different suggestion.
We have already noted that certain sentences at the end of the manuscript were not used in TMS, partly because Smith had altered his view on one point and partly because he would have seen a difficulty in his position on another. Now in these sentences of the manuscript, he is discussing what he takes to be the scholastic view of distributive justice, as distinguished from commutative justice, which, he has said, can alone properly be called justice. The distinction between different senses of the term ‘justice’ is described in TMS at VII.ii.1.10. We are there told that one sense ‘coincides with what Aristotle and the Schoolmen call commutative justice’, and that a second sense ‘coincides with what some have called distributive justice’ [our italics]. At this point Smith inserts a footnote to explain that ‘The distributive justice of Aristotle is somewhat different’. The note goes on to give a succinct explanation of Aristotle’s view and ends with a reference to the Nicomachean Ethics. Now Mitford’s information was that Mr. Cunningham found the manuscript in a volume of Aristotle, and it is not too fanciful in the circumstances to suppose that the volume was, or included, Aristotle’s Ethics. In preparing his earlier thoughts for publication, Smith would have checked many of his statements, and in this instance he would have found, by reference to Aristotle, that some qualification was needed to the bare statement in the lecture that ‘in the Schools’ the name of distributive justice was used for the proper allocation of beneficence. Following up this line of thought, one can even suggest an identification of the particular volume in which the manuscript was found. Adam Smith had a copy of the Works of Aristotle in Greek and Latin, edited by du Val and published in four folio volumes at Paris in 1629. (Bonar, Catalogue 2, 10, incorrectly gives the date as 1729.) Apart from this, he had separate editions of the Rhetoric and of the Poetics but no separate edition of the Ethics. Volume III of the du Val edition includes Aristotle’s ethical writings, and it is quite likely that this volume was the one in which the manuscript was found. Smith’s copy of the du Val Aristotle was certainly among the books that were bequeathed to Mrs. Cunningham, and it is now in the Library of the Queen’s University of Belfast. It would be pleasant to be able to report that Volume III shows some line of discoloration as the result of having secreted a folded sheet of paper for some seventy years, either at the relevant part of the Nicomachean Ethics or in the endpapers, but the Sub–Librarian at the Queen’s University tells us that there is no such trace.
Returning from these speculations to the fairly solid facts established earlier, there are some further inferences that may be drawn from the manuscript. To those who know the content of TMS, the title of the book seems a little strange, for the basic concepts of Smith’s distinctive theory are sympathy and the impartial spectator, not moral sentiments. In the manuscript, Smith uses the phrase ‘the Theory of moral Sentiments’ as parallel to ‘the Theory of the general principles of Law’, and from this we can see that the title of his book is not meant to describe his own individual contribution to ethics but is his name for the scope of the subject in general (just as more recent writers have entitled their books The Theory of Good and Evil or The Theory of Morals, meaning that they are writing essays in ethical theory, not that their own views alone can properly be called ‘the’ theory). Once we realize that Smith’s title is his name for the subject, it no longer seems strange. He was first taught ethics by Hutcheson, but Hume was the thinker who stimulated him to form a theory of his own. In Smith’s eyes Hume had demonstrated conclusively that moral judgement and action are not based on reason but on ‘sentiment’ or feeling. Hume had further suggested that the ‘peculiar sentiment of morals’ is mediated by sympathy; and Smith found this suggestion attractive in principle but over–simplified in its assumption that there was a single, ‘peculiar’, moral sentiment to be explained. He therefore elaborated a more complex account of sympathy, that would explain the distinction between several different forms of moral sentiment, the ‘sense of propriety’, of virtue, of merit, of duty. Hence he regards the task of ethical theory as that of giving an account of ‘moral sentiments’ in the plural.
Another inference from the manuscript that can be made with confidence is this. It has always been supposed, from hearsay and intrinsic probability, that Smith worked up TMS from his lectures on ethics. The manuscript provides definite proof that he did so, even to the extent of repeating many of the very words of his lectures in their written form. Some, slightly hazardous, internal evidence pointing in this direction can be found in the book itself, as is mentioned in section 1(a) of our Introduction (p. 4) and in editorial footnotes at II.i.1, IV.2.7 and 9, and VII.iii.1.2. But comparison of the manuscript with relevant parts of the printed texts puts the matter beyond any doubt.
This leads to yet another point. Rae (Life, 260–1) reports the opinion of J. R. McCulloch that Smith dictated WN to an amanuensis but wrote the manuscript of TMS in his own hand, and that this accounts for a difference in the style of the two works. Rae is sceptical, since there is no evidence that McCulloch had anything more to go on than his own impression that the style of WN is more diffuse than that of TMS, and Rae himself does not share that impression. We can be fairly certain, from our knowledge of Smith’s extensive use of amanuenses, that WN was indeed dictated. Now if Smith was using an amanuensis even for his lectures, it seems likely that he would have done so for the manuscript of TMS.5 Evidence from another quarter, however, makes this less certain. A comparison of details of antique spellings in edition 1 of WN with corresponding details in letters written in Smith’s own hand, shows clearly enough that he himself did not write the manuscript used by the printers. A similar scrutiny of details of spelling and contractions in edition 1 of TMS, on the other hand, shows little deviation from, and indeed a good deal of correspondence with, Smith’s practice in his letters. So it is possible that he did write the manuscript of his first book in his own hand. Nevertheless, we have already seen that much of the actual phraseology and construction of sentences repeats material in the lectures, which he had dictated to an amanuensis. If, as McCulloch believed, dictation produced a more diffuse style, the effect should be apparent in TMS too. One can of course account for differences of style in the two books, if differences there be, simply by the difference in Smith’s age at the respective dates of composition. Certainly the passages added in edition 6 of TMS tend to be more diffuse than the writing of the original book, but the simplest explanation of this is that in 1759 Smith was a man in his thirties, while in 1788–9 he was in his sixties.
So much for the manuscript fragment. A good deal can be learned from it about the composition of TMS, but nothing about the reasons why Smith withdrew the paragraph on divine justice. In returning to this question, we want to suggest that Archbishop Magee was not so silly as Rae supposed, but we must state clearly at the outset that our suggestion on this issue is to a certain extent speculative.
The paragraph withdrawn from edition 6 occurred at the end of a chapter that considers the extent to which the sense of justice depends on utility. Earlier in the chapter, Smith gives partial support to a utilitarian theory of justice, and speaks of ‘the account commonly given of our approbation of the punishment of injustice’ (II.ii.3.7). In LJ(A) ii.90, he again discusses utilitarian theories of ‘the originall measure of punishments’, which, he says, have been held by ‘Grotius and other writers’. But if one thinks of a utilitarian account of justice in general, Smith must surely have regarded Hume as the main contemporary proponent. His description of the utilitarian account in TMS II.ii.3.6 seems to refer particularly to Hume’s view that utility pleases through sympathy. Writing in the 1750s, Smith was bound to recall that Hume, in the Enquiry concerning the Principles of Morals (1751), had argued strongly for the view that justice, unlike benevolence, arises solely from utility. Consequently, even though Smith’s criticism of the utilitarian view of justice may have been aimed originally at a wider target, in the particular context of TMS II.ii.3 it must have seemed, and indeed have been intended, to be primarily directed against Hume, as are Smith’s criticisms of utilitarianism elsewhere in the book.
This conclusion is confirmed by a couple of minor revisions of the disputed paragraph that were introduced in edition 3. Editions 1 and 2 state that a utilitarian view of divine justice ‘is not the doctrine of nature, but of an artificial, though ingenious, refinement of philosophy’, and that ‘All our natural sentiments prompt us’ to take a non–utilitarian view. In edition 3, the first phrase becomes ‘is not the doctrine of untaught nature but of an artificial refinement of reason and philosophy’, and the second phrase becomes ‘Our untaught, natural sentiments, all prompt us’. The addition of ‘untaught’ in both sentences plainly takes account of Hume’s distinction between different senses of the term ‘natural’ (Treatise of Human Nature, III.i.2, III.ii.1; ed. Selby–Bigge, 474–5, 484).
It is likely that Hume would have discussed with Smith in conversation their radical difference of opinion on the place of utility in moral judgement, and the minor revision noted above may be due to Hume’s criticism. It is also possible that Hume may have teased Smith about his acceptance of conventional orthodoxy on theological matters. However that may be, edition 3 of TMS contains some further minor revisions of the paragraphs on divine justice, and of later remarks on the character of the clergyman, toning down the categorical affirmations of the original version. Editions 1 and 2 said that a non–retributive view of divine justice ‘can, by no means, be so easily admitted’; edition 3 alters this to ‘seems repugnant to some very natural feelings’. Editions 1 and 2 stated firmly that man can see no reason why he should not be the subject of divine indignation; edition 3 says ‘he thinks he can see no reason’. In editions 1 and 2, man ‘is sensible’ that he appears to God to be a vile insect, and ‘is conscious’ that he is undeserving of happiness; in edition 3, he only ‘imagines’ the first and ‘suspects’ the second. In editions 1 and 2, repentance, sorrow, humiliation, and contrition ‘are’ the sentiments which become him; in edition 3, they ‘seem’ so. A similar problematic note is struck by a revision in the preceding paragraph (II.ii.3.12) of ‘religion authorises’ to ‘religion, we suppose, authorises’. Then again, at V.2.5, the clergyman, who in editions 1 and 2 ‘is’ the messenger of serious tidings and ‘is’ continually occupied with the grand and solemn, becomes in edition 3 one who ‘seems to be’ the former and ‘is supposed to be’ the latter. It is of course possible that the more cautious statements of edition 3 represent Smith’s original views, the expression of which he felt would be injudicious as coming from a Professor of Moral Philosophy but which honesty obliged him to make clear after he had quitted his Chair. At any rate we ought to note that, in Smith’s revision of TMS, the withdrawal of the paragraph on divine justice in edition 6 was not the first suggestion that he might have moved away from orthodox theology.
Whether or not the changes made in edition 3 were influenced by discussion with Hume, there is no doubt that the criticism of utilitarianism which ends with the paragraph on divine justice was first and foremost a criticism of Hume. In the light of this, let us consider what Smith might have been ready to say in 1759 and reluctant to let stand after the death of Hume in 1776. (Edition 5 of TMS appeared in 1781, but at this juncture we can properly say that Smith’s ‘many avocations’ prevented him for a long time from making the radical revisions that he had contemplated.) It will be recalled that Magee had been shocked by Smith’s estimate of Hume’s character, and in this Magee was not alone. The phrase that Magee quoted had given great offence to the faithful at the time of its original publication. It came in the last sentence of Letter 178 addressed to William Strahan, dated 9 November 1776, soon after Hume’s death. The letter was written for publication along with Hume’s short autobiography. Smith knew very well that he was stirring up a hornet’s nest, but although he was temperamentally averse from public controversy on matters of religion, he deliberately ended his letter with the statement that Hume had come as near to perfect virtue as human frailty allowed. It was written with the deepest sincerity, on the death of Smith’s greatest friend, whom the world called an ‘atheist’. When Smith came to revise his book on ethics, he must surely have felt some revulsion from concluding a criticism of Hume with a paragraph whose language echoed the sermons of those ‘high–flying’ preachers who had been the bitterest detractors of Hume. It is no wonder, if Archdeacon Sinclair’s report is authentic, that Smith should have thought the paragraph ‘unnecessary and misplaced’. It was unnecessary because the preceding criticism of utilitarian theory stood firmly enough on its own ground of appeal to our ‘natural sentiments’. It was misplaced because it was (after 1776, at least) quite the wrong spirit in which to end a polemic directed as much against his dead friend as against anyone else. So instead of Christian doctrine about expiation and atonement, Smith made his own atonement by substituting a sentence so Humean in tone that it might almost be called a libation to Hume’s ghost: ‘In every religion, and in every superstition that the world has ever beheld, accordingly, there has been a Tartarus as well as an Elysium; a place provided for the punishment of the wicked, as well as one for the reward of the just.’
Rae concludes his account of the controversy by saying (Life, 429–30) that ‘Smith gives a fresh expression to his belief in a future state and an all–seeing Judge in one of the new passages he wrote’ for edition 6, showing ‘that he died as he lived, in the full faith of those doctrines of natural religion which he had publicly taught’. Certainly Smith never abandoned natural religion. The new passages (there are in fact two of them, at III.2.12 and III.2.33) about the all–seeing Judge seem at first sight to be very near in doctrine to the suppressed paragraph, and indeed one might wonder why, if Smith really wanted to retain the paragraph in another place, he did not insert it there. Yet a closer look at the ‘all–seeing Judge’ passages gives a different impression. Rae quotes from the first of them but does not mention that Smith reverts to the idea more fully towards the end of the same chapter. The new passages, like the suppressed paragraph, are about the doctrine of divine reward and punishment in an afterlife, but Smith does not now give unqualified support to the doctrine as preached by Christians. The notion of heavenly reward, says Smith, is the only comfort for unrecognized innocence and virtue, but it has too often been taught in a form that contradicts our moral sentiments by confining divine salvation to the religious. He quotes and derides an address of Massillon to the effect that soldiers cannot hope for the heaven which one day of penance and mortification in a monk’s cell can bring. ‘To compare, in this manner, the futile mortifications of a monastery, to the ennobling hardships and hazards of war . . . is surely contrary to all our moral sentiments’. A paragraph of support for theology is followed by two paragraphs of scorn for ‘monks and friars’ as contrasted with ‘heroes, . . . statesmen and lawgivers, . . . poets and philosophers . . . all the great protectors, instructors, and benefactors of mankind; all those to whom our natural sense of praise–worthiness forces us to ascribe the highest merit and most exalted virtue’. Smith then ends his chapter by quoting Voltaire’s satirical couplet on the Christian concept of hell:
If Smith had added to this honours list of the ancients a similar one for the moderns, he would have put Hume at the head of it to correspond to Plato (and Plato’s Socrates). Smith’s derision of ‘monks and friars’ and ‘the futile mortifications of a monastery’ has a familiar ring. It was Hume who wrote (Enquiry concerning the Principles of Morals, IX.i; ed. Selby–Bigge, §219) that ‘penance, mortification, . . . and the whole train of monkish virtues . . . are . . . everywhere rejected by men of sense’. And it was Adam Smith who deliberately imitated the last sentence of Plato’s Phaedo by ending his epitaph to Hume with the judgement that Hume had approached ‘as nearly to the idea of a perfectly wise and virtuous man, as perhaps the nature of human frailty will permit’.
 The word ‘with’ has presumably been omitted.
 In the final word of this clause, the amanuensis appears to have begun to write ‘th’ (i.e. ‘them’?) and then changed it to ‘h’.
 Perhaps Smith in fact said ‘sleeping at’ and was misheard by the student whose report is copied in LJ(B). The sentinel would have been more culpable if he had deliberately left his post.
 Ronald L. Meek and Andrew S. Skinner, ‘The Development of Adam Smith’s Ideas on the Division of Labour’, Economic Journal, lxxxiii (1973), 1094–1116 (especially 1104–6), give reasons for assigning these two fragments to an even later period, the 1760s.
 The draft revision of 1759, sent to Sir Gilbert Elliot, is in the hand of an amanuensis.