¹ Womersley, D., ‘From polybianism to perfectibilism: the influence on Gibbon of “Le Chevalier de Chastellux”, British Journal for Eighteenth-Century Studies 13 (1990), 47–55CrossRefGoogle Scholar, at p. 54. Cited as ‘Womersley’.

² Memoir E, p. 324 n. 48, The Autobiographies of Edward Gibbon (1896), ed. Murray, JohnGoogle Scholar. (Since it prints the six drafts of the memoirs consecutively, this edition is still preferable, despite inferior critical apparatus, to Edward Gibbon, Memoirs of My Life (1966), ed. Bonnard, G. A.Google Scholar. Drafts of the memoirs cited in the text as ‘Mem.’). Cf. ‘Gibbon's Dark Ages’, JRS 73 (1983), 1–23Google Scholar, at p. 18. Hereafter cited as ‘Gibbon's Dark Ages’.

³ Womersley supposes that this point of view, as presented in the Memoirs, is an attempt ‘to smooth the jagged edges of life into art’ (p. 54). This seems a curious inversion, when what Gibbon offers is so obviously puzzling (or jagged), and the new solution is so convenient; see below (11).

⁴ ‘Gibbon's Dark Ages’, 20. These words were pre-ceded by others: ‘The memoirs rarely, if ever, lie…’. Their fallibility is in fact at the level of minutiae: compare Mem. E 308 and The Letters of Edward Gibbon (1956), ed. Norton, J. E., no. 316Google Scholar (hereafter cited as Letters) and n. 167 below.

⁵ loc. cit.

⁶ To Ann Firth, 22 Sept. 1793, pr. The Girlhood of Maria Holroyd (1896), ed. Adeane, J. H., 239Google Scholar. As examples: (1) from September 1778 (Letters, 432) until Gibbon's arrival in England with a completed MS. in August 1787, the History was always composed with an eye to future dates of completion, sometimes two or three years in advance; (2) he was perpetually calculating his life expectancy, to establish the time available to him, Mem. E, 347 and n. 72, Add. MSS 34882 ff. 49–50, Note of a conversation on immortality, between G[ibbon] and H[olroyd]; (3) he was both sensitive to the ideological point behind French revolutionary chronology, and skilled in its usage, Letters, 859, 875; (4) for his sentiments on those who were ‘regardless of futurity’ and the calculation of time, e.g. DF vii.216.

⁷ cf. Memoirs of My Life, op. cit. (n. 2), pl. 5. Craddock, P. B., Edward Gibbon, Luminous Historian 1772–1794 (1989), implies, p. 371 n. 24Google Scholar, that Gibbon indulged in ‘memorial reconstruction of dates’. Generally, this is not so. Not every date is documented, but when so many are, they narrow the range of error for those which are not, and inspire confidence in the scruple of the author; at Mem.B, 148 Gibbon engages in self-imposed memory tests when writing his Memoirs.

⁸ Parallels to the History abound, as one might expect. Thus in the final drafts of the Memoirs (E and F) Gibbon was unable to restrain the habit of writing with full references. Again, besides his own papers, he sought to enlarge his fund of knowledge by research: this is well known in the case of his genealogical researches; see inter alia the request to Sheffield for the use of correspondence, Letters, 791; Norton, J. E., A Bibliography of the Works of Edward Gibbon (1940), 182Google Scholar; Craddock, op. cit. (n. 7), 290–1. His preoccupation with chronology is well to the fore here.

⁹ The History of the Decline and Fall of the Roman Empire (1781), iii.636.

¹⁰ The History of the Decline and Fall of the Roman Empire (1909–1914), ed. Bury, J. B., iv.178Google Scholar; hereafter cited in the text under ‘DF’. However, references to volumes of the History assume the volume divisions of the first edition (1776–88). As Dr William Smith pointed out in his edition of 1854–5 (iv.407 n.a), the emendation might then be construed as a reference to Naples and Spain.

¹¹ Mem. E, 324 n. 48. The attribution to the Prince of Beauvau is William Smith's, loc. cit (n. 10), and seems at least plausible given Gibbon's contacts with the Princess, ‘a most superior woman’ (see Letters, 391, 387, 452, 498), and the Beauvaus’ central position in that small group who carried on Anglo-French cultural interchange either side of the Channel; for Gibbon and the Prince, The English Essays of Edward Gibbon (1972), ed. Craddock, P. B., 213–14Google Scholar; hereafter cited as EE.

¹² The force of this reference depends upon its precise dating amidst the French revolutionary flux; and we may note the groundlessness of the view that the notes to Memoir E were written after 2 March 1791, the date with which Gibbon signs the text, just because the notes come after the text (Bonnard, op. cit. (n. 2), xxvii–xxviii). This would be to assume, in a parallel case, that Gibbon did not finish his History on 27 June 1787 – the date with which he signs the text – although we can be sure he did, both from his Memoirs (E, 333) and from the convention of signing the date, which would be meaningless if it were not a terminus. The notes to draft E have, of course, been fair copied (Add. MSS 34874 ff. 97–102b), but it is clear from the cases of draft F and the continuous draft of the Antiquities of the House of Brunswick (cf. n. 104), that Gibbon habitually wrote his notes or their gist on small or folded paper in parallel with the text (cf. Add. MSS 34874 ff. 124–7, 34881 f. 249b) – after all, a most convenient way of proceeding. Textual arguments may also be adduced in the same direction: cf. Mem. E, 325 n. 66 and Gibbon to Sheffield 5 Feb. 1791, Letters, 771; also Mem. E, 319 n. 41.

¹³ Womersley, 53. I am grateful to Dr Womersley for his kindness in elucidating his argument at this point.

¹⁴ He did not intend publication in 1788 (Mem. A, 353), 1791 or January 1793 (Letters, 791, 826). Subsequently, according to Sheffield in 1796, Gibbon stated in conversation that he would publish (The Miscellaneous Works of Edward Gibbon (1796 ed.) i.1 n. *). But this might reflect a misunderstanding on Sheffield's part, apparent in January 1793 (cf. Sheffield to Gibbon, 23 Jan. 1793, pr. Private Letters of Edward Gibbon (1896) ed. Prothero, R. E., ii.366Google Scholar). Sheffield had a clear interest in justifying publication (cf. Miscellaneous Works (1796 ed) i.v), but even if we set aside the dubiety of his evidence, it is well beyond the term of the composition of draft E of the Memoirs in early 1791.

¹⁵ History of the Decline and Fall … (1789 ed.), iii.636; the 1787 edition, also iii.636, is unchanged from that of 1781 – (for Gibbon's disinclination to revise this, Letters, 638) – and supplies a terminus a quo. On the dating and numbering of editions, Norton, op. cit. (n. 8), 51–3.

¹⁶ Dates of stay given with equal accuracy in Letters, 649, 702, 704 or Mem. E, 334, 336, 340.

¹⁷ Presence in London was a virtual necessity for Gibbon in transacting this sort of revisory business with the printer, e.g. Letters, 663, 666, 668.

¹⁸ At the time I suppose Gibbon to have emended the text of the ‘General Observations’, he was drafting the dedication of Volumes IV–VI of the History: ‘LORD NORTH will permit me to express the feelings of friendship in the language of truth: but even truth and friendship should be silent, if he still dispensed the favours of the crown’ (1788 Preface, DF i.xlvi). This makes Gibbon's sentiments on the flattery of authority – even of the English kind – abundantly clear; as he commented in 1790–1 (also when writing draft E of the Memoirs), ‘I disdained to sink the Scholar in the politician’ EE, 341; cf. Letters, 771; Essai sur l'Étude de la Littérature (1758–61), ch. XI n*. For his general view of the worth of kings, DF ii.454–5, iv.493 and n. 75, EE 340.

¹⁹ Mem. E, 347, cf. Letters, no. 702, DF i.xlvi (on ‘the Public’), vii.45 n. 4.

²⁰ DF iv.174 (‘General Observations’); Gibbon's stereotype of Arcadius and Honorius is, of course, expanded in the History, iii.195–418 passim.

²¹ For the distinction in Gibbon's mind between Louis XVI and Louis XV, see also Letters, 752.

²² Gibbon was acutely interested in the French monarchy and the basis of political consent underlying it. His criterion of stability was that natural to an Englishman and parliamentarian – the ability of the monarchy to raise taxes – as is evident in the memoranda from his 1777 Paris visit, which also display some typically English awareness of the unpopularity and regressive nature of French finance, EE 213–24 passim. These data were used in the History, where the French people are adjudged ‘industrious, wealthy, and affectionate’, DF ii.208, cf. iv.501. Gibbon's interest in the subject and his ultimately favourable assessment are explained and confirmed by his investment in the French funds in 1784 which, despite the expensive American war, he adjudged ‘at least as solid as our own’, 24 Jan. 1784 to Sheffield, Letters, 612, cf. 609. For the maintenance of this distorted, but typical and explicable view, Letters, 730, 752, 803.

²³ Besides the Beauvaus (n. 11), Gibbon corresponded with Leclerc de Septchênes, a secretary to Louis XVI; but this related to the latter's translation of the History – which Gibbon dismissed as feeble (Mem. E, 339 n. 63) – not the lure of monarchy (Letters, 364). Craddock, op. cit. (n. 7), 88 states it as ‘certain’ that Gibbon met the king, but gives no evidence. In fact Gibbon's principal source, and influence, on these matters was Necker, (Letters, 501, 623, 626Google Scholar, Mem. E, 331 n. 49); what warped his perspective was social and administrative elitism, not monarchical deference. Less tangibly but yet probably, the monarchy's patronage of learning might predispose him to assume not only its legitimacy but its beneficence – relevant here are (1) his preoccupation with the Royal Academy of Inscriptions and Belles-Lettres in the years c. 1757–69 (whilst its Mémoires are much cited in the History, despite the falling-off implied at Mem. B, 164), and (2) the profit he derived from using the Royal Library (Mem. E, 314, cf. DF i.295 n. 171, iii.84, n. 24; B, 201–2).

²⁴ Womersley, 54.

²⁵ We should note that there is a second, hitherto unnoticed reference to the 'Observations' in the Memoirs, which again places them in 1772 and is entirely free of any complicating allusion to the French monarchy. After describing the studies preliminary to the History of 1771–2, Gibbon states: ‘As I believed, and as I still believe, that the propagation of the gospel and the triumph of Christianity are inseparably connected with the decline of the Roman Monarchy, I weighed the causes and effects of the Revolution…’ (Mem. C, 285). This would seem a fair description of the ‘General Observations’ – a text assessing what it calls ‘this awful revolution’, and one which makes a central insistence on the connection of Christianity with the fall of Rome (DF iv. 175).

²⁶ Changes in the footnotes are listed in Womersley, 53–4, with the exception of Mémoires sur les Chinois, par les missionaires de Pékin (Paris, 1776), cited DF iv.177 n. 6, cf. iii.85 n. 25; Keynes, G., The Library of Edward Gibbon (1980 ed.), 196Google Scholar.

²⁷ The argument here involves the important but intricate question of Gibbon's practice in revision and redrafting, and I can only summarize my findings: (1) apart from Volume I of the History (and the publicised emendation to the ‘Observations’), Gibbon never in fact changed the text of any work after publication; (2) nor did he ever revise a manuscript he regarded as finished (a statement which, admittedly, requires elaboration); (3) thus most revision in detail and all redrafting in extenso took place at the time of composition and before publication; (4) anyway, when Gibbon revised – or at least contemplated revision, as in the case of the History in the 1780s – this meant principally stylistic revision or the addition of new materials and references (in his own terminology ‘improvements’) designed to buttress existing positions; (5) there was but little ‘correction’ in matters of fact, and no fundamental ‘change’ of intellectual position. In short, the proposal that Gibbon went through the ‘General Observations’ twice after finishing them, making substantial corrections so as to reflect changes in viewpoint, has no parallels in his practice, and we must decline to entertain it; cf. Womersley, 54.

²⁸ Essai, chs. XLVII, LXV; Mem. C, 270; cf. Le Journal de Gibbon à Lausanne 1763–4 (1945), ed. Bonnard, G. A., 55–6Google Scholar: ‘[Nardini] est enfin arrive au … Forum Romanum, où l'on ne peut faire un pas sans renconter les monumens de la religion, de la grandeur et de la politique des Romains’ (25 Sept. 1763). This view may be traced a long way back in embryo: ‘Common Place Book’ 1755, EE 17–19; ‘Remarques Critiques sur les Dignités Sacerdotales de Jules César', 1757 pr. The Miscellaneous Works of Edward Gibbon (1814), ed. Lord Sheffield (hereafter cited as MW), v.61–5.

²⁹ cf. Voltaire, , Essai sur les Moeurs (Paris, 1963), ed. Pomeau, R., i.180–6, 277–310Google Scholar; for Gibbon and Voltaire: Mem. F, 79; Mem. B, 148–9; EE 17–21. The connection between Christianity and the fall of Rome was drawn by many authors to whom Gibbon had access, the root being Macchiavelli (‘Gibbon's Dark Ages’, 19 n. 120); cf. Hume, Natural History of Religion, § X, and Bayle, Miscellaneous Observations on the Comet, cxli. Of course, behind Macchiavelli stands Augustine, but there is no evidence that Gibbon had read him before 1771–2, cf. DF i.8 n. 25.

³⁰ The argument here is pursued further in the Appendix, below. On dating, Craddock, op. cit. (n. 7), 8–14, generally concurs, but thinks the ‘General Observations’ may have been composed in 1773 rather than 1772. This supposes a complex and unexplained series of events – that Gibbon began his History, put it aside to write the ‘Observations’, and then returned to it – to which we may fairly apply Ockham's razor; the evidence cited (p. 8) is nihil ad rem. On the Essai: none of the long prose notes to chs 1–26, 34–55 are in the original 1758 text, these were thus added subsequently. For a majority no MS. survives, but in some important cases a MS draft or other evidence enables us to date the time of addition and so confirm the point: (1) Add. MS 34,880 f. 158b, note on Lucretius (cf. ch. XLVII); (2) ff. 155b, 184, on Augustus' donative (ch. xx); (3) Gibbon's Journal to January 28th 1763 (1929), ed. Low, D. M., 11 Feb. 1759Google Scholar, note on Newtonian chronology (ch. XXXVIII). The edition in MW incorporates a second stage of post-publication additions to the notes for chs XVII, XX, XLI, LXIII taken from an interleaved copy of the text given to Sheffield by Gibbon, MW (1796 ed.), i.ix; the autograph is untraced, but there are copies of these in Add. MSS 34882 ff. 54–9, and the authenticity of the interleaved material is attested in another context by Mem.C, 254–5, cf. MW (1815 ed.), iv.1–2. However, consistently with what else we know of Gibbon's indolence in post-publication revision, he never incorporated these changes in his lifetime, and indeed refused to sanction a new edition of the Essai in 1776, Mem. B, 171; cf. n. 27.

³¹ ‘Gibbon's Dark Ages’, 20.

³² The title (and substance) of , Womersley'sThe Transformation of the Decline and Fall of the Roman Empire (1988)Google Scholar exemplify this most fully; but his later article (n. 1, passim) pursues the idea in terms more easily measured by the historian.

³³ i.e. issues highlighted by Womersley, in seeking to illustrate the influence on Gibbon of the Marquis of Chastellux (op. cit. (n. 1), passim); but since Gibbon neither cited nor corresponded with Chastellux, I do not feel myself competent to pursue the speculation. The Transformation of the Decline and Fall also argues that Gibbon abandoned a belief in uniform principles of human nature in writing the history (e.g. pp. 4–6). In fact his adherence to this view is so comprehensive that space precludes adequate demonstration of the fact; but see the only direct evidence Womersley quotes on the subject (p. 210), against his thesis; and Robertson's, J. review, Notes and Queries 37 (1990), 477–8CrossRefGoogle Scholar, and nn. 111, 127 below.

³⁴ Given that the ‘Observations’ open with the most sustained hymn to Polybius anywhere in Gibbon's oeuvre (DF iv.172–3), it is one of Womersley's most marked paradoxes that they symbolize his departure from Polybianism (op. cit. (n. 1)). This stems from a prior misreading of one of Gibbon's sentences on Polybius and an ignorance of his pages thereon in Transformation of the Decline and Fall, 188. Such an error, and no isolated one, seems to the historian a consequence of a literary critical methodology, which places a premium on linguistic virtuosity in the analysis of minute portions of text.

³⁵ Chs 1–3 centre on the following themes from the ‘General Observations’: (i) the preservation of the Empire by (more or less) abstaining from the expansionist adventurism which was seen to have destabilized the Republic (DF i.1–10); (ii) the military, religious and political institutions identified by Polybius and their moral basis (DF i.10–20, 31–46, 65–79); (iii) the idea – far too little noticed – that Rome's basic transgression was against the idea of nationhood, at least in Europe. The unification of Italy, had the career of conquest stopped there, would have been acceptable (DF i.20–5, 36–47 (esp. 38), 61–4, 86–90).

³⁶ DF iii, 376–7; cf. v.523, Letters, 752 (15 Dec. 1789) for later opinions to the same effect.

³⁷ Failure to observe the elementary truth that Gibbon was a moralist rather than a political economist or sociologist has, however, led to some confusions: (1) underestimation of the importance of (vicious) luxury, e.g. Pocock, ‘Gibbon's Decline and Fall’, 148; Porter, Roy, Edward Gibbon: Making History (1988), 147Google Scholar, cf. D F iii. 196; (2) ignorance of Gibbon's moralism and moral consistency: e.g. Burrow, J. W., Gibbon (1985), 88Google Scholar (on Mandeville), cf. DF vii.309 and n. 104.

³⁸ e.g. Essai ch. LXXXIII; DF i.62.

³⁹ DF i.1; for a more intricate (and early) working of the same idea, i.58–61.

⁴⁰ DF iii. 311; cf. iii. 196–7, v.458, etc.

⁴¹ Further ‘Polybian’ consequences are apparent: the nobles’ distaste for military service and their inability to sustain a properly balanced constitution (DF iii.310).

⁴² e.g. DF ii.272 n. 34; iii. 140–1; vii. 157–8.

⁴³ And here Gibbon is too generous to his later self. As the foregoing discussion (pp. 133–4) of just one note in the Memoirs shows, the criticism attached to the Essai – ‘brevis esse laboro, obscurus fio’ (Mem. B, 173) – applies throughout Gibbon's life (though it is but the obverse of a laudable desire for concision). His abjuration of the style of Montesquieu (loc. cit.) is frequently noted, but we may as well note with Robertson a continued loyalty to Tacitus and the obliquity (or polyvalence) of meaning this entailed, MW ii.249, no. CXLIX.

⁴⁴ Mem. B, 173 referring to Essai chs XLIV–LV; for the central importance of these paragraphs to the History, below pp. 141 f., 146f. However, the specific links between the two are greatly enhanced by cognisance of two long passages cancelled by Gibbon in the original MS, Add. MSS 34880 ff. 151–2, 153–5, passages of such importance that I hope to publish them in due course (cf. also nn. 51, 105). Though Gibbon refused to allow the Essai to be reprinted in 1776, by c. 1790 he came to think this might represent excessive pride in his later work (Mem. B, loc. cit.); already in 1784 we find him ordering from Cadell six copies of the Essai alongside six of the History, to give as presents in Lausanne, Add. MSS 34886 f. 199. As is well-known (Mem. B, 137, 152), Gibbon dated his own intellectual maturity from the period of his first stay at Lausanne (1753–8), which again implicitly focuses attention on the Essai, the culmination of his studies from January 1756 on, cf. Add. MSS 34880 ff. 86–159b, Mem. B, 206.

⁴⁵ cf. below, pp. 144–6.

⁴⁶ Mem. E, 333; with typical candour, Gibbon marks signs of haste in the notes to the final two volumes, e.g. DF v.401, n. 168, 460, n. 88; vi.34 n. 81; vii.139 n. 2, 167 n. 6, 330 n. 63; cf. v.377 and n. 125 – an obvious slip. Another striking instance of self-mortification is the famous lament, that he had not given ‘the history’ of the age of the Antonines: this marginale in one of Gibbon's copies of the History [B.L. C.60 m.1] was not, pace Craddock, simply a ‘marginal comment’ (EE 338), but a new note designed for publication on page 1 of the text – in as prominent a position as it well could be.

⁴⁷ Letters, 638, 768.

⁴⁸ cf. below pp. 155–6.

⁴⁹ Mem. D, 412; cf. DF ii.74–5 (c. 15);seen. 54 below.

⁵⁰ DF i.208–9, 407–10, 433–4, 456–8; ii.276–8; iii.31–3, 199–205, 271–3, 279–83, 291–3, 304–48; iv.2–8, 21–2, 35–6. 48–9, 65–6, 201–5, 233–5. 285–6, 331–47, 427–38, 445–6, 470–542 passim; v.32–41, 273–331 passim; vii.138–40, 218–338. This list of figures makes an elementary point about the significance of the preliminary reconnaissance. It is, however, especially inadequate for Volumes I and VI: since we cannot know how Gibbon meant to distinguish in his original scheme between the history of ‘Roman’ institutions, such as the senate and the papacy, and the wider history necessarily involved therein, I have excluded references to these (cf. DF 1.407, ‘the form and the seat of government were intimately blended together’); but the passages which remain leave no doubt as to their internal coherence, in theme and detail.

⁵¹ On the Essai, compare chs LVI–LXXVII, DF i.31–6, ii.20–3, cf. n. 105 below; on Du Gouvernement Féodal, ‘Gibbon's Dark Ages’, 4, 10; on the Critical Observations: compare EE 146, DF i.33; EE 139, DF ii.17; EE 146–8, DF ii.20–3; EE 146, DF ii.357 n. 18; EE 155, DF ii.461 n. 15; EE 136–7, DF ii.465 and n. 25. Although Gibbon esteemed the Critical Observations, a published work, he did not cite it – presumably because he was already suffering remorse about its anonymity, Mem. C, 283; EE 310.

⁵² Gibbon's proposal for a supplementary volume of 17 Nov. 1790 (Letters, 768), comprised: ‘I. a series of fragments, disquisitions, digressions &c more or less connected with the principal subject. 2. Several tables of geography, chronology, coins, weights and measures &c; … 3. A critical review of all the authors whom I have used and quoted.’ (1) and (2) point to the Nachlass from the 1771 reconnaissance, as well as to several of Gibbon's ‘early’ or pre-1772 manuscript works, later published by his explicit sanction in the Miscellaneous Works (1788 Will, pr. Prothero, op. cit (n. 14), i.vi). In fact, the seriousness of Gibbon's intentions as to Volume VII may be questioned: the plan for a review of authors (3) was pre-empted by his occasional and latterly (Volumes V and VI) systematic practice of commenting on them in the History itself, and it was renounced in the Preface of 1788 (i.xlv–xlvi, cf. MW (1796 ed.), i.686 n. *); whilst even at the point of maximum financial interest, he was chilled by the fatigue of ‘these obscure labours’ (Letters, 768), for he would have had much new work to execute notwithstanding. But the principle of utilizing early material is clearly established. On Sur les Triomphes des Romains: EE 338; cf. ‘Gibbon's Dark Ages’, 7.

⁵³ On the dating of this text, ‘Gibbon's Dark Ages’, 14. We know that: (i) Gibbon made what was in effect a précis of the Digression in the annotations to his copy of the fourth edition of the History [B.L. C.135 h.3], (dateable in fact to 1781, cf. Add. MS 34882 ff. 176–9); (2) there are two fair copies of the text. Presumably that with Gibbon's own marginal corrections was made at the time of composition (Add. MSS 34,880 ff. 264–72b), and the second one later (Add. MSS 34,881 ff. 242–8b). The inference must be that having epitomized the Digression as a note in 1781, it then occurred to him to reproduce the essay as a whole. This parallels his expanding conception of to how to supplement the History.

⁵⁴ Gibbon wrote in Memoir D that ‘in this supplement I may perhaps introduce a Critical dissertation on the miraculous darkness of the Passion’ (p. 412). The memoir was not, as Bonnard notes, any sort of supplement, but it was written at precisely the same time that Gibbon was considering the supplementary volume (Edward Gibbon: Memoirs of My Life, xxvi and n. 4, cf. Letters, 768), and it seems almost certain that Gibbon was here thinking of Volume VII, which he always referred to as the ‘Supplement to the History…’, EE 342.

⁵⁵ On the repetitive nature of Gibbon's procedures before 1772, ‘the refurbishing or writing up of themes in Roman history which he had worked out previously’, ‘Gibbon's Dark Ages’, 14. Cf. the degree of repetition in the 1783 musings entitled ‘Notes’, to the first three volumes of the History: compare EE 319–21, DF ii.496–8; EE 321–2, DF ii.67; EE 324, DF iii.436–9; EE 325, DF ii.479–80; EE 327–9, DF iii.110 n.84; but also iv-535–7.

⁵⁶ cf. Jordan, D. P., Gibbon and His Roman Empire (1971)Google Scholar; but a too narrow focus on the perceived symbiosis between the History and the Memoirs is characteristic of modern criticism, cf. Roy Porter, Gibbon (1988), Conclusion, Carnochan, W. B., Gibbon's Solitude (1987), e.g. chs 1, 3Google Scholar – the best of recent historical and literary treatments of Gibbon.

⁵⁷ The view that Gibbon saw the Essai as ‘a minor work’ (locc. cit. (n. 56), here Jordan, p. 10) is emphatically contradicted by the prominence it is accorded in the Memoirs (Mem. B, 167–74; (C, 250–7); and by Gibbon's explicit descriptions of it as ‘a more elaborate composition’ than his other writings at Lausanne in the 1750s (Mem. C, 250), and as his ‘first work’ or ‘performance’ (Mem.B, 167–8), i.e. his first published work, and the first intended as such from the outset (cf. Letters, 21 and Appendix II 1.397–8), albeit with subsequent hesitations on this score. (We may, with Gibbon, discount the humorous episode of the ‘Age of Sesostris’, Mem. F, 79–81.)

⁵⁸ Chs 1–2. Ferguson's History of the Progress and Termination of the Roman Republic (1783) is a variant: narrative proper starts in Bk. I ch. 5 with the second Punic War, but the first four chapters, while containing much general matter, also endeavour to take in the ‘legendary’ period in snatches; hence the comment ‘I have … endeavoured to give, even to the first part of my labours, the form of narration’ (i.5) – which is its own testimony to contemporary expectation in the matter. Mitford, W., History of Greece vol. 1Google Scholar (1784), chs 1–2 is similar, working in a parallel context.

⁵⁹ His procedure here is a vast expansion of that followed in the History of Scotland (1759), Book I; another example of the elephantine introduction derives from Voltaire's insertion of the 1765 Philosophie de l'Histoire as a proem to subsequent editions of the Essai sur les Moeurs.

⁶⁰ A variant was the brief narrative sequel – Robertson's History of Scotland stopped in 1603, but its ‘Conclusion’ took in the years to 1707 in a few pages (Book VIII); likewise Ferguson's History of the Roman Republic runs from Tiberius to Nerva in Book VI ch. 7.

⁶¹ e.g. Hume's History of England (1754–61) in all its parts; Voltaire, , Le Siècle de Louis XIVGoogle Scholar (1751). Gibbon's Volumes V and VI might be regarded as falling under this rubric, but see below v.

⁶² History of Charles V, Book XII; Gibbon's Volume III is comparable, below p. 150.

⁶³ DF v.180; cf. Hume's History of England (1754–61), i.1. Hegel's recognition that ‘original’ history was no longer the prime category of historical writing, being subordinate to reflective or philosophical history, was an eloquent reflection of an achieved intellectual revolution, ‘Die Arten der Geschichtsschreibung’ (1822/8), in Die Vernunft in der Geschichte (1955), ed. Hoffmeister, J.Google Scholar. Of course, the great works of Enlightenment historiography have their roots in the antiquarianism of the previous two centuries; but whilst the latter had many uses, it did not have the centrality that history enjoyed progressively from the mid-eighteenth century on; cf. Essai ch. 11.

⁶⁴ The Essai sur les Moeurs, for example, was a ramshackle assemblage of material frequently altered and re-sited – its enormous introduction (n. 59 above) being partly balanced at the end by a much slighter ‘Résumé’; the latter is indeed ‘philosophical’, being a series of general reflections on history, but it is not, like the ‘General Observations’, a causal or analytical review, ch. XCXVII (1963), ed. R. Pomeau. That Voltaire felt the tug of convention is also evident from the similar ‘Discours sur l'Histoire de Charles XII’, originally placed at the end of the first edition of the Histoire de Charles XII (1731–2), but then moved to the beginning, being further preceded by ‘Remarques sur l'Histoire’ (1742) in the 1756 edition; Oeuvres Historiques (1957), ed. Pomeau, R., 1660–1, 1670Google Scholar.

⁶⁵ Smith to Gibbon 10 Dec. 1788, in The Correspondence of Adam Smith (1977), ed. Mossner, E. C. and Ross, I. S., no. 283Google Scholar; cf. Strahan to Hume 12 Apr. 1776, pr. Oeuvres Historiques, p. 193 n. 1; Hume to Gibbon 18 Mar. 1776, Mem. E, 311 n. 30.

⁶⁶ On this collocation, below p. 141 and n. 68.

⁶⁷ Momigliano, A., ‘Gibbon's contribution to historical method’, repr. Studies in Historiography (1966), 40–55Google Scholar.

⁶⁸ Essai ch. XLIII; the association of ‘philosophy’ with causal analysis was of course universal, see Robertson to Gibbon, MW ii.417, no. CCXIX. Other attributes of the ‘philosopher’ may be mentioned: (a) detachment; (b) universality of standpoint; (c) [from (a)] association with Stoicism and its moral values; (d) hostility to metaphysics; (e) from all of these, the ability to speak a portion of the truth, cf. n. 111 below. ‘Remote history’ or ‘the distant view of history’ (DF iv.86, cf. Letters, 498) was seen as compatible with all these attributes. Porter's assertion, op. cit. (n. 37), 136, that from their title the ‘Observations’ could not supply causal analysis, is specious and no more. ‘General Observations’ supplied those ‘general pictures which compose the use and ornament of a remote history’ (DF v. 180), the large view being one taken by that central Enlightenment fiction, the ‘impartial’ or ‘philosophic spectator’ (e.g. DF ii.348; cf. iv. 176 ‘a philosopher may be permitted to enlarge his views’, v.258–9; Essai ch. XLVI; Raphael, D., ‘The Impartial Spectator’, Proc. Brit. Acad. 58 (1972), 335–54)Google Scholar. Thus, from their title, ‘General Observations’ were philosophic, here performing one of the central functions of philosophy, viz. causal analysis. This is of course evident from the text of the essay, and in the prominent equation of ‘general observatons’ and causal enquiry at DF vii.329. Naturally, on other occasions summary ‘observations’ are linked to other functions of the philosopher, notably judgement and assessment, where no causal problem exists: vii.73–5, cf. vi. 462–6.

⁶⁹ Table Talk, 15 Aug. 1833, to appear in the Princeton ed. of the Collected Works vol. 14 (forthcoming); cf. Morison, J. Cotter, Gibbon (1878, English Men of Letters), 130–1Google Scholar, who espouses a two-tier model similar to Gibbon's but cannot see it in the History – perhaps because the latter does not talk the language of ‘social evolution’ (or perfectibilism). But though both authors were demonstrably poor readers of Gibbon in detail, a substantial point remains: see p. 143.

⁷⁰ Essai ch. LV, emphasis added. This chapter, although written in 1761 (Add. MSS 34880 f. 175), summarizes one of the long cancels in the original 1758 MS. (Add. MSS 34880 ff. 151–2) which was to make way for his thoughts on the evolution of paganism. To suppose (in the final text) that these ideas might be put ‘entre les mains d'un Montesquieu’ was thus perhaps a blind – they had already been worked at some length by the hand of Gibbon.

⁷¹ DF i.84, my emphasis, cf. iv. 191, 471. This reworks Voltaire's mot in the ‘Résumé’ of the Essai sur les Moeurs – ‘toute cette histoire est un ramas de crimes, de folies, et de malheurs’ (ed. Pomeau, 804) – so as to highlight the evidential point. That Gibbon did not think reality and the evidential register were at one is evident from e.g. Essai ch. 1, DF i.293, iii.364, v.27 and n. 60. His passionate dislike of war is also relevant – even ‘in its fairest [contemporary] form … a perpetual violation of humanity and justice' (iii.429, cf. Letters, 32). Gibbon was opposed to every foreign war fought by England in his lifetime. (Commentators habitually overlook the scathing irony behind his comments about the Captain of the Hampshire grenadiers, which was really a ‘peaceful service’, Mem. B, 190.)

⁷² On Livy and Tacitus, Essai ch. LII; cf. Letters, 341 where they are equated with Hume and Robertson respectively. The Swiss History was an earlier, but hardly encouraging, excursion into narrative. As will be evident from n. 50, the History of the City would not have been primarily narrative, but structural, concentrating on buildings, spectacles, games, laws, population, etc.

⁷³ cf. IV below.

⁷⁴ As he had written in 1758: ‘L'Esprit Philosophique en trou vera sans difficulté les raisons [of Roman decline]’, Add. MSS 34880 f. 152, c.56.

⁷⁵ See for example the revealing terminology in Mem. C, 284: ‘I more seriously undertook (1768) to methodize the form, and to collect the substance of my Roman decay, of whose limits and extent I had yet a very inadequate notion’; cf. ‘Du Gouvernement Féodal’ (1768), MW ii.185. This shows how open-minded Gibbon was empirically, but that the underlying fact of Roman decay was never in doubt; which ambivalence is another central idea he sought to express in the Essai: ‘Qu'il [the historian] se gardât bien de chercher un systême; mais qu'il se gardât bien davantage de l'éviter’ (ch. 11). On the Swiss and Florentine histories: Mem. B, 196–7, a free version of Gibbon's Journal to January 28th, 1763 (1929), ed. Low, D. M., at 26 July 1762Google Scholar. Horace Walpole's pre-publication criticism of Gibbon's Volume II is comprehensible only in the light of his own a priori, against ‘so disgusting a subject as the Constantinopolitan history’: thus he begins by referring to matter treated in the first and third volumes (the last not in his possession) before turning to Volume II itself, Horace Walpole's Correspondence (1939–1983), ed. Lewis, W. S., to Mason, W. 27 Jan. 1781, xxix.97–9Google Scholar.

⁷⁶ Mem. C, 284; E, 308. Given the pre-eminence of profound causes in his conception of history, Gibbon was fascinated by sheer duration, and statements about the tongue durée abound, wholly divorced from any narrative consideration (e.g. DF i.64, 230, ii.168, 306, iv.180–1, 470, v.42; cf. Add. MSS 34880 f. 151, Essai MS. cancel c.52). Unlike Braudel, however, he could see that there were surprising cases of survival at the contingent level – as in the case of the empire, or of Islam, whose monotheistic rationality ran contrary to Gibbon's view of what a weak human nature required of a popular religion: ‘It is not the propagation but the permanency of [Mahomet's] religion which deserves our wonder’ (DF v.419). ‘Surprise’ or ‘wonder’ may have been what seduced Gibbon into the excessive length of Volumes II and III. Nevertheless, given a conceptual scheme where politics were seen as superstructural, this could lead to no new revelation, and thus in the subsequent case of Islam he was content merely to posit surprise, maintain reserve, and achieve brevity.

⁷⁷ See below pp. 143–4, 154–5.

⁷⁸ DF iv. 174. A similar abridgement is effected in the cancelled MS for the Essai of 1758, where Gibbon simply notes of the period between Claudius Gothicus and Augustulus, that ‘L'Esprit Philosophique … attendoit la prochaine dissolution.’ Add. MSS 34880 f. 152, c.56.

⁷⁹ Milman made the important point, perhaps difficult to appreciate today, that without any prior historiographical tradition ‘the whole period … seems to offer no more secure footing to an historical adventurer than the chaos of Milton – to be in a state of irreclaimable disorder’, so making Gibbon's firm insistence on Roman decay a necessity rather than a wearisome iteration, ‘Guizot's Edition of Gibbon’, Quarterly Review 50 (Oct. 1830), 273–307Google Scholar, at 287. This may be confirmed from widely varying contemporary perspectives: Walpole, H., Historic Doubts on … Richard III (1768), Preface, iv–viiGoogle Scholar; Robertson to Gibbon 12 May 1781, MW ii.249.

⁸⁰ Gibbon's most frequent usage in describing the moral history of the empire was neither ‘decline’ nor ‘fall’ but ‘decay’ – so expressing a state of corruption over the very long term most graphically (e.g. DF iv.173–4; Mem. C, 270, 284). Of course, ‘decline’ is virtually synonymous, but so, too, is ‘fall’ in one of its principal meanings (OED s.v., 5c, 16, also 1b); and whether we consider ‘the period of the fall of the [Western] empire’ (DF iii.73) – which covers one century or a quarto volume – or the soundlessness of its final ‘extinction’ (iv.56 and n. 134), there is justice in supposing this the closest to Gibbon's central meaning. The title ‘Decline and Fall’ is, then, a tautologous insistence on the moral theme so repeatedly invoked in the text. The other central term in this vocabulary, ‘ruin’, equally evokes the longue durée and gradual declivity, moral and physical, cf. p. 151 below.

⁸¹ e.g. DF ch. 26, iii.442–54 passim. The link between ‘situation’ and ‘manners’ or ‘character’ is made explicit at i.249 n. 73.

⁸² Coleridge by his terminology (‘fathom’, loc. cit. (n. 69)), and Porter, op. cit. (n. 37), 136, both suppose that if there had been an ‘ultimate’ cause, which they deny, it would have been hidden, ‘some grand arcanum’, cf. Womersley, op. cit. (n. 32), 211. In fact the resolution of this paradox lies in the idea that ‘profound’ causes of historical motion are ‘secret’ to contemporaries, but visible to the philosophic historian with the advantage of the ‘distant’ or ‘general’ view, as is repeatedly emphasized, DF ii.212, iii.196, 289–90; Essai ch. LV, cf. Letters, 609. (Interpretation of ‘secret’ as ‘occult’, ‘random’ or ‘vertiginous’ is a modern invention, Womersley, op. cit. (n. 32), 184.) This was not to patronize the past, but to remember that, in Gibbon's view, no age, however enlightened, could shake off its historically conditioning prejudices, Essai ch. XLVII.

⁸³ cf. IV below.

⁸⁴ This is not to question the general rectitude of Momigliano's seminal thesis – of Gibbon's linkage of philosophy to erudition or Belles-Lettres – though it is argued without direct reference to the History (op. cit. (n. 67); cf. Murray, O., ‘Momigliano e la cultura inglese’, Rivista Historica Italiana 100 (1988), 422–39, at 427–8Google Scholar): Gibbon indubitably saw himself as trying to illustrate this link in the Essai, and was indeed successful in doing so at a more detailed level in the History – p. 147 below.

⁸⁵ The 1776 Preface looks forward ‘most probably’ to one more volume to 476; as late as March 1779 Gibbon refers to bringing out ‘the second Volume’; September 1779 is the first sure indication that he will take up two more volumes; these went to press the following June (Letters, 445, 457, 477). (From his silence we may presume that Gibbon had by then completed his text, as with Volumes IV–VI, in contrast to the procedure he describes for Volume I, Letters, 315–16).

⁸⁶ Usesor variants in: DF ii.348; iii.377; iv.6, 173 n. 4, 224, 364 n. 2; v.46, 53, 392; vi.238; vii.146, 208. It is clear from the ‘General Observations’ that Gibbon was trying to improve on the fatalism of the Renaissance historians’ appeal to mere ‘Fortune’ (cf. iv.i26, vii.65), without, however, falling into the system-builder's trap of assuming that his two tiers necessarily moved in parallel, cf. Essai ch. LV (pr. p. 141 above). His failure was not, it should be stressed, necessarily conceptual, but one of literary construction.

⁸⁷ DF iii. 197, 243, 265, 304; later, even the ‘name’ of Rome becomes limited, near extinction and then finally abhorred, iii.379, 421–2, 507; iv.502.

⁸⁸ DF ii.168, my emphasis; on this see, however, pp. 145–6 below.

⁸⁹ DF iii. 196, iv.57 n. 137; cf. iii.318–19 and n. 51.

⁹⁰ One of the most striking is that, despite the evaporation of Roman spirit, discipline is revived more or less at will by every commander from Claudius Gothicus to Aetius. Gibbon is clear that discipline is in itself something superficial as compared to moral fibre (DF i.12, 250, ii.303), and is thus consistent in supposing this to happen. Still it is unsatisfactory: one suspects that ‘discipline’ is brought in to explain Roman victories post hoc propter hoc; and if superficial discipline is so powerful, what is the worth of the ‘profound’ analysis from manners and morals?

⁹¹ Mem.E, 334; notwithstanding his prejudices, Walpole's dissection of the sheer difficulty for the contemporary reader of Volumes II–III is devastatingly exact: to W. Mason 3 March 1781, Horace Walpole's Correspondence, xxix. 114–15.

⁹² Even before publication, Gibbon harboured seeds of doubt about Volumes II and III, which were to harden over the next decade. Though they retained the pictorial quality of philosophic instruction and commercial amusement he so valued – ‘un tableau interessant et instructif’ – he upheld only his continued diligence, and was thus ‘tout preparé à un jugement moins favorable de la part de mes lecteurs’ (Letters, 498). This is but one step away from the view of the Memoirs (E, 323–4).

⁹³ i.e. given Gibbon's much lesser knowledge of the sources after the sixth century, ‘Gibbon's Dark Ages’, 2–3, n. 167 below.

⁹⁴ A simple table makes clear the widely varying density of coverage in the different Volumes of the History. For all its crudity, this was clearly an exercise performed by Gibbon himself, cf. 1782 Preface, DF i.xli, v.180:

Volume I covers c. A.D. 180–324, or c. 144 years per volume of text

Volumes II–III cover c. 324–476 or 76 years per volume of text

Volume IV covers c. 476–628 or 152 years per volume of text

Volumes V–VI cover c. 628–1453 or 422 years per volume of text

At no point, even in Volumes II–III, is Gibbon writing an evenly proportioned narrative, as he occasionally reminds us: ‘The general design of this work will not permit us minutely to relate the actions of every emperor after he ascended the throne’ (DF i.313, cf. ii.457). However, as the table implies, and despite a unity of subject, Gibbon drew a clear distinction between Volume I and its two successors by always exempting it from criticisms of the latter (e.g. Letters, 677). Though piously noting Hume's objection to the too ‘concise and superficial narrative of the first reigns from Alexander to Commodus’, he makes it plain that this was an isolated viewpoint (Mem. E, 308). If we suppose Volume I to start with Augustus, the elision of the period between the Principate and c. 180 represents a further, radical decision about narrative structure. However, Gibbon's subsequent doubts about his effective omission of this period do not weaken the point, since he only sighed after giving its ‘history’, which, as he emphasizes, was not mere description, EE 338. (Craddock's supposition to the contrary, EE 588, n. 1, is confuted by her reference to OED.) Below p. 146 on Volume IV.

⁹⁵ Even in authors who suppose that the ‘true’ or mature Gibbon is to be found only in the final volumes of the History – Jordan, op. cit. (n. 56), Womersley, op. cit. (n. 32) – coverage of the text is in inverse proportion to their theses.

⁹⁶ DF v.182; cf. Letters, 638, where Gibbon states, even as he writes (December 1786), that he is allowing himself the ‘utmost latitude’ in the choice and treatment of subject matter. He continues to the end the bizarre process of minuting the decline and fall of the empire – its ‘last and fatal stroke’ (DF vi.523) is still five chapters and 100 years short of 1453 – but since he is really writing ‘the eventful story of the barbarians’, albeit linked to the Byzantine annals (v.183, cf. Letters, 677), the defect or curiosity is no longer central.

⁹⁷ DF v.183, cf. Letters, 518; Horace Walpole to Lady Ossory 8 Nov. 1789 (approving the procedure but deprecating the arrangement), Correspondence xxxiv.79. For the ‘consistent procedure’: vi.135–6 (barbarians after the Saracens), 305–6 (the Crusades); vii.11 (Genghis Khan), 225–6 (the Papacy) etc. In acting thus Gibbon might fairly claim to be following the example of his principal sources, DF iv.421 n.13 (Procopius). His control of overall length in Volumes IV–VI is as perfect as the handling of Volumes II–III was inflationary. In September 1783 – before he had devised the organizational plan of Volumes V and VI – he forecast three more volumes; shortly after its devising, he wondered if the radical economy it permitted might not bring him down to two volumes (October 1784); by December 1786 (halfway through the composition of Volume VI, Add. MSS 34882 f. 175) he forecast three volumes again, apologizing to his publisher that they might be ‘somewhat thinner, perhaps, than their predecessors’, but on publication day – aided doubtless by such fillers as the ‘Digression on the House of Courtenay’, DF vi.466–74, cf. Mem. B, 195–6 – he managed to give almost identical value for money, 1950 pages of quarto text and notes compared to 1955 in the first three volumes (1st edition). Given Gibbon's acute awareness of literary length and commercial value – ‘I can exactly compute, by the square foot, or the square page, all that remains to be done’ (January 1787) – this was almost certainly deliberate: Letters, 592, 623, 638, 642.

⁹⁸ Of course, it has long been recognized that the ‘General Observations’ represent Gibbon en philosophe, Giarizzo, G., Edward Gibbon e la cultura europea del settecento (1954), 231Google Scholar; but this has not been a favourable or sympathetic evaluation.

⁹⁹ As early as December 1777 he was aware of the lengthy coverage which the ‘age of Constantine’ would require (Mem. E, 315), but it was not until September 1779, when he must have been composing the chapters which make up the period after the death of Julian (or Volume III) – he went to press the following May (Letters, 467) – that he became aware that an extra volume would be necessary, cf. n. 85; i.e. he was diverted from plan by his later rather than earlier material. DF ii.138 (ch. 16) also looks forward to ‘the second volume of this history’ in terms implying the centrality of Constantine. This point of view would also explain the portrayal of the new system of government under Diocletian and Constantine (chs 13, 17) as pathological (cf. p. 144 and n. 88 above on ‘rapid decay’), when to all appearances it is part of an imperial revival.

¹⁰⁰ DF iv.226, Mem. E, 326; cf. Mem. F, 79 on ‘the Age of Sesostris’ with explicit acknowledgement to Voltaire. This non-linear method of computing time may also be discerned in the 1776 Preface, DF i.xxxix–xli. The Voltairean ‘Siècle’ is in turn indebted to Bossuet's ‘Époque’, another author who affected Gibbon significantly, Mem. F, 86. Volume IV could, of course, equally be described as the ‘Age of Procopius’, DF iv.224–6.

¹⁰¹ DF iv.174 (pr. above p. 143), cf. iv.179. The only significant reference to emperors apart from Constantine is to Arcadius and Honorius (iv.174, 177–8), but they are conspicuous by their absence, and could thus hardly be seen or foreseen as a problem relating to the length of the text.

¹⁰² Ch. XLIX, cf. DF i.238.

¹⁰³ ibid., ch. L.

¹⁰⁴ DF i. 1, my emphasis; cf. v. 183: ‘each circumstance of the eventful story of the barbarians…’ The term is frequently evoked in the detailed texture of the History, e.g. DF i.127, 238, 303; ii.457;iii.38 and n. 100, 324, 378 and n. 1, 457 n. 39; iv.62, 90, 104 n. 143,152 etc. The idea was, of course, common, e.g. Hardwicke to Gibbon 20 Sept. 1781, MW ii.254–5. History as denned by its leading circumstances is also evoked in the 1776 Preface (i.xxxix–xl) and the summaries Gibbon gives at the end of chs 38 and 71 (though here the notion of a priori interest intersects, iv.170–1, vii.338); cf. Mem. B, 193–4, Letters, 463, MW v.487–8. This idea accords with what we know of Gibbon's methods of literary composition, where he frequently starts with discrete fragments of text and works out from these. The Antiquities of the House of Brunswick is a major and complete example of the process. Here we have two MS states: (1) single or sometimes two-page folio fragments, Add. MSS 34,880 ff. 326–353b; which are then (2) smoothed into a continuous prose draft, Add. MSS 34,880 295–324 and 34,881 f. 149b. (Alas, the two states are garbled out of all recognition in EE 398–531.) Compare the so-called ‘Circumnavigation of Africa’, Add. MSS 34,880 ff. 355–66, which represents the first state only and must therefore be regarded as unfinished. Gibbon's statement about the MS. of the History (Mem. E, 334) does not necessarily stand against this point of view, though the apparently seamless drafts of the Memoirs as a whole do represent a quite different model.

¹⁰⁵ i.e., apart from general historical-philosophical use, literary criticism and taste, and natural science. Gibbon's discussion of sources throughout his text, according to both style and substance, illustrates the first; as to science, the History (i.101–3) relies directly on the Essai (ch. XLI) with respect to the natural history of animal species; compare also ch. XLIII and IV.441–2 and n. 59, vii.319–20. Note, too, linked interests in climate (i.231–3, vii.219 and n. 2), demography (i.233–41 passim), natural disasters (iii.72, iv. 461–9, vii.318–20), and sub-Lamarckian ideas as to the effect of environment on human evolution (iii.74–80, 442–3; iv.400 and n. 84,411–12 and n. 103; v.335)–and the history of science generally (e.g. vi.5–6, 28–34, vii.85–6). Such inquiry was a necessary concomitant of history founded on the premiss that there were laws of human nature, though Gibbon carried from science to history (via his friends Buffon or Dr Watson vi. 10 n. 21, ct. Mem. E, 317 n. 35) rather than vice versa.

¹⁰⁶ The History converses with the reader at many levels: (1) generally, reading in this period was seen as ‘familiar converse’ or ‘intercourse’ (e.g. DF iv.269 n. 117; vii.128, 136. Mem. B, 141); and dialogue with ‘his friends’ (Mem. E, 346, cf. Add. MSS 34874 f. 95b) rather than egotism is the principal function of Gibbon's Fielding-like insertion of the author into the text, cf. Letters, 247, 251; p. 151 below; T. B. Macaulay, ‘Francis Bacon’ (1837), Critical and Historical Essays (1907, Everyman), ii.291–2. (2) English and ‘domestic’ accentuation in the text continually appeals to the home audience (starting with the promotion of Britain out of the circuit of Roman provinces in ch. 1, DF i.3f, and ending at vii.300 n. 86), but other national or regional audiences are identified too (e.g. i.240 n. 38, iv.526 n. 167, Lausanne and the Swiss, ii.304–5, iv.309 n. 32 Paris and the French, iii.42–7 the Scots); (3) finally, there are directly personal messages, to Mme. Necker (iii.409–12, cf. Letters, 498), Lord Loughborough (vi.333 n. 148), his Scots friends (11.509 n. 15, vi.465 n. 89), ?Deyverdun (vi.483 n. 23), Sir William Jones (iii.84 n. 20, iv.527 n. 173, cf. MW ii no. CLI).

¹⁰⁷ On eunuchs, DF ii.260–2, iii.380–2; on witchcraft et al., i.393–4, iii. 17–20; on Roman spectacles and buildings n.50 above, p. 151 below. Gibbon's concern with other issues still topical, such as paederasty, women and slaves, needs no labouring; his great omission is the history of ideas, pre-empted or evacuated by a focus on manners and morals.

¹⁰⁸ DF i.211; cf. v. 180 on ‘those general pictures which compose [?comprise] the use and ornament of a remote history.’

¹⁰⁹ Essai, ch. L.

¹¹⁰ ibid., ch. XLIX, pr. above p. 146.

¹¹¹ Porter, op. cit. (n. 37), 136, a representative modern view. Gibbon saw clearly that each person's perspective was different, relative to their role or situation in life, be it patriot, philosopher, saint, member of the menu peuple or whatever: ‘some animals are made to live in water, others on the earth, many in the air, and some, it is now believed, even in fire [the House of Commons]’ (Letters, 609). The love of knowledge was thus one among many (irrational) ‘passions’ or endowments, Mem. E, 344, C, 248, and one should not be snobbish or unrealistic about this. Few had the leisure to pursue such a career (DF i.238); historically, the philosopher's view was frequently a minority one (e.g. iv. 80); and though Gibbon set a high value on truth, he seconded John Vataces who ‘without deciding the precedency, … pronounced with truth that a prince and a philosopher are the two most eminent characters of human society’ (vi.477, cf. v.220), having himself oscillated between careers in politics and truth-telling, Letters, 571. Nonetheless, the philosopher dealt in (portions of) absolute truth, which the text clearly assumes to exist. This position entails one assumption: that the relative category of ‘the philosopher’ (and thus all the other types) is viewed as timeless rather than historical – and such is the case in fact. For example, Theodoric, Caliph Ali and Louis IX are all ‘heroes’ regardless of the apparently specific, classical connotations of the term (DF iv. 184, cf. 182; v.412; vi.374).

¹¹² ‘The truth of history’ is frequently an appeal to the limitations of such knowledge; all the same it is a species of truth, which leads, for example, to specific prescriptions – such as the advocacy of religious toleration – and which after all qualification has been made, is not vacuous, since history still ‘undertakes to record the transactions of the past for the instruction of future ages’ (DF ii.87, cf. i.197; ii.68, 306; iii. 119, 135; iv.471; vi.374, 511 etc.).

¹¹³ To W. Mason 3 March 1781, Correspondence xxix. 115; cf. n. 119 below for a parallel judgement, with a different valuation attached.

¹¹⁴ cf. Milman, op. cit. (n. 79), 288–9, 291–2. Of course, such pluralism invited the further inquiry and debate Gibbon hoped his History might inspire, p. 151 below.

¹¹⁵ Gibbon's handling of the concept of ‘manners’ is a great advance in rigour over that of the philosophes, but also exposes its primitive simplicity. In effect ‘manners’ was a catch-all of moral factors which lost analytical identity under the close examination he gave it: thus it was not a purely social (or sociological) concept, since it included Polybian military institutions, for example; even ‘the discipline and tactics of the Greeks and Romans form an interesting part of the national manners’ (DF iii.489, cf. iv.172–4); and it was quite consistent with an undifferentiated evocation of the Zeitgeist (e.g. ii.456, iii.260). Gibbon and the philosophes stand somewhere between the moralism of Christian tradition and the sociological development of ‘manners’ pioneered by Durkheim and Weber; and evaluation of ‘manners’ as a concept may be influenced by the view taken of any or all of these three.

¹¹⁶ This may be shown by a silence in the criticism of the History, or by the vitality of moralistic analyses of society in the England of the 1780s: Joanna Innes, ‘Politics and morals. The reformation of manners movement in later eighteenth century England’, in Hellmuth, E. (ed.), The Transformation of Political Culture (1990), 57–118Google Scholar, cf. DF iv.536, vi.274 etc.

¹¹⁷ See eg. p. 143; in Volumes IV–VI Gibbon applauded the ‘new Barbarians who … enjoy the advantage of speaking their own language and relating their own exploits’, Letters, 677.

¹¹⁸ DF ii.1, my emphasis. This is, however, a unique utterance; normally Gibbon does not make such direct juxtapositions, which might allow the need for some reconciliation of ideas. While adhering to this view, it may also be suggested that Gibbon saw barbarian virtues as a reflex of those of the Romans, i.e. adhering to the primacy of internal decay, but linking the internal and the external explanations thereof. Such is the burden of the cancelled MS of the Essai in 1758, Add. MSS 34880 f. 152, chs 55–56; due to the paucity of ‘barbarian’ sources, this argument is rarely made explicit in the History (DF i.271, iv.364), but it underlies Gibbon's concern with the power of the Roman name (n. 87 above).

¹¹⁹ 14 Feb. 1776, MW ii.154; cf. Letters, 638; EE 338 (on Hume, superlatives and positives); Mackintosh's table talk, 12 Oct. 1830, pr. Memoirs of … Sir James Mackintosh (1836), ii.476 etc.

¹²⁰ DF iv.171; cf. iv.105, Letters, 638,677.

¹²¹ Essai, ch. XLVII.

¹²² Gibbon was unaffectedly modest – and clear-sighted – in matters of intellect, and never aspired to place himself on a level with ‘philosophers’ such as Hume or Adam Smith, cf. Letters, 227, 335, 402. (Even in the case of historical writing, Gibbon did not coin the image of a ‘triumvirate’ grouping him with Hume and Robertson, and when that association was made, he declared himself the Lepidus, Letters, 592, cf. 389.) Those who, since Guizot, accuse Gibboa's lack of ‘une grande élévation d'idées’ (Preface, History of the Decline and Fall (1828), 5, cf. H. Walpole to W. Mason 18 Feb. 1776, Correspondence xxviii.244), may be right, but they ignore both his own conception of philosophic truth (cf. p. 143 and n. 82 above) and his real, historiographical achievement (n. 75 etc.)

¹²³ cf. p. 151 below.

¹²⁴ Gibbon was explicit about the special status allotted to chapter endings: DF vi.531. He had previously made use of them in chs 4–7, as a forewarning of ‘the general irruption of the barbarians’ in ch. 10.

¹²⁵ DF iii.506; cf. Essai 1758 MS. cancel, ch. 56 n. ‡, Add. MSS. 34880 f. 151b.

¹²⁶ Other chapter endings make points about decline at the narrative or contingent level, directly or indirectly, by pointing out unrepresentative individuals, Marcian ch. 34, Claudian ch. 30; ch. 28 deals with the tongue durée in religion. I do not of course wish to suggest that, except in a formal sense, chapter endings supply the only ‘conclusions partielles’ in the History.

¹²⁷ Compare (e.g.) DF iii.506, iv.172–3. Ranging outside the (immense) time span of the History is, of course, a common procedure for Gibbon, particularly in pursuing the timeless bases of human nature through the fashionable Enlightenment (or Plutarchian) device of the historical parallel, which knows no chronological or geographical limit.

¹²⁸ In the 1781 edition, the ‘General Observations’ start on a new page (iii.629); there is no chapter number in the margins; and their italicized title is explained by the italicized short contents summarizing the main sections of chapters prior to the detailed contents linked to page numbers (iii.[i–viii]). The ‘Observations’ thus represent an independent, unnumbered sub-chapter, i.e. a formal anomaly – powerful evidence of their extraneous origin. Resolving the anomaly by making them part of ch. 38, as is done by Bury (in his contents, page layout and marginal chapter numbers), is illegitimate and misleading.

¹²⁹ i.e. that of Volumes V and VI, as distinct from Volumes I–IV; cf. p. 145 above.

¹³⁰ Womersley claims that the end of Volume I represents ‘the conclusion of a work’ because in chs 15–16 ‘the real villain of the fall of the Roman empire is revealed’, op. cit. (n. 32), 101, 102. But no evidence is cited to support this contention. In fact the narrative of Volume I breaks off midway through the reign of Constantine, whilst its Preface proclaims it is not independent (i.xxxix–xl) – serving to confirm that chs 15–16 are not a conclusion. That, nonetheless, Gibbon felt it desirable to end a volume published in isolation in a striking way may be readily conceded; chs 15–16 are paralleled by ch. 47, also a major religious statement, at the end of Volume IV.

¹³¹ DF iv.31 [ch. 36]; cf. iv.105 [ch. 37], 171 [ch. 38] which accord with the 1776 Preface (i.xxxix–xl).

¹³² cf. above p. 140.

¹³³ In its new context in the History, the paragraph in the ‘Observations’ on Rome and Constantinople, whilst focusing on the theme of decay, can also be read, in parallel with the strands in the main text (n. 131 above), which point to the future continuation of the Eastern narrative: ‘The foundation of Constantinople more essentially contributed to the preservation of the East than to the ruin of the West’ (DF iv.174–5). This is another justification for this passage to set off against the positivistic criticism given below (Appendix).

¹³⁴ Above p. 149.

¹³⁵ Regardless of the presence or absence of the ‘Observations’, there were good reasons why Gibbon should not descant on Roman decay here: neither ch. 37 nor ch. 38 is Rome-centred, whilst in ch. 36 the narrative closes not with the insignificant Augustulus but with the subsequent reign of Odoacer, making a forcible point about the unimportance of high political chronology (cf. 1776 Preface, DF i.xxxix), and the extreme difficulty of separating Roman from barbarian which had obtained since the early third century at least (i. 182–3); and such is the burden of Gibbon's general reflection when he reaches 476 (iv.58).

¹³⁶ For the other, p. 143 above.

¹³⁷ P. 149 above.

¹³⁸ DF i.47, cf. i.23, 422 etc.

¹³⁹ DF vii.317–20. Gibbon thus vindicates the original inclusion of ‘Les Sciences Naturelles’, one of the main headings of classical study in the Essai chs XXXIX–XLIII, as part of a unity. The patent connection between this strand of Enlightenment historiography and the French Annalistes of this century is curiously ignored; indeed ‘lay’ opinion seems to two are diametrically opposed, e.g. Stone, L., ‘The revival of narrative’, in The Past and the Present Revisited (1987), 74–06, at 74Google Scholar.

¹⁴⁰ However, following in the wake of Bury (DF i.vii), J. G. Pocock appears to reduce the History to the unitary theme of just one sentence, extracted out of context from these later ‘observations’ – ‘In the preceding volumes of this History, I have described the triumph of barbarism and religion…’ (DF vii.321) – and foreshadows an entire book based upon it (‘Gibbon … and the Late Enlightenment’, in Virtue, Commerce and History, 143 n. 1, 146 and n. 8). He would overlook: (1) that this is not all Gibbon has described; (2) that, in the context of ch.71, he allocates a subordinate role to these factors; (3) that he uses the term ‘triumph’ in its technical sense, i.e. as a spectacle; (4) that by departing from his plan to write the history of the City (and of the Papacy) he failed to complete the account of the triumph of religion.

¹⁴¹ Although this is logically tautologous, it must be remembered that moral ésprit did not exist apart from its historically specific embodiment in social or private practices, ‘manners’; the specificity of the latter saves the former from vapidity.

¹⁴² Mem. E, 338, my emphasis, 346. These expressions are of a piece with the idea of reading as a dialogue or converse, n. 106.

¹⁴³ There is danger in any analogy of this sort, given Gibbon's lack of interest in the visual arts, but it is a good deal more promising than those hitherto pursued (notably by Michel Baridon, e.g. ‘Le style d'une pensée: politique et esthétique dans le Decline and Fall’, in Gibbon et Rome a la lumière de l'historiographie moderne (1977), 73–101). In terms of chronology (his principal works appeared 1750–78), historicism, Romanitas, international rather than merely national orientation and reputation, and what one may call a distinctively neo-classical evocation of the power of ruins, Piranesi and Gibbon have much in common; cf. Wilton-Ely, J., The Mind and Art of Giambattista Piranesi (1978)Google Scholar. That Gibbon was working within a cultural mainstream is also evident from his dependence on Muratori's Antiquitates Italiae Medii Aevi (1738–42) for his pungent treatment of the ‘strong towers’ of medieval Rome (DF vii.326 n.48), and on Maffei's Verona Illustrata (1731–2) for material on Italian amphitheatres (vii.329 n. 59). On his orthodoxy in aesthetic taste, cf. ‘Gibbon's Dark Ages’, 5, n. 26.

¹⁴⁴ The other principal physical survivals Gibbon identified were geographical and topographical, and secondly, genealogical or ‘the history of blood’ (vii.1). Both broad types of survival are equally important to the History, though the latter is less recognized. The History (like its predecessor, the History of uris Roma) represents an attempt at uniting these two historical roots which Gibbon had treated discretely in the Essai and the Receuil Geographique of 1763–4.

¹⁴⁵ (1) Gibbon's farewells to Athens and Constantinople involve allusion to their buildings or ruins in a sentence and a page respectively (DF vi.507–8, vii. 140, cf. vii.210) – the former a remarkable comment on his lack of interest in the nascent ‘Greek revival’ in architecture. Given their unfamiliarity (not least to himself), Gibbon describes the buildings of Constantinople at length in DF iv.258–70, in his first chapter (ch.40) on Justinian. (2) The last page of the History (on the connection between Roman ruins and its conception), is repeated in one of the most famous passages of the Memoirs (Mem. C, 270; see also DF vii.235 n. 47); given the elemental logic it embodies, verifiable through the entire History, the quibbling of commentators on the Memoirs alone may be seen in its true proportions; cf. ‘Gibbon's Dark Ages’, 5–6.

¹⁴⁶ Though the progression from the Preface of 1781 to that of 1782 (DF i.xli), from a tentative to a definite resolution to proceed through to 1453, may seem smooth, it was not so in fact – as Gibbon candidly recognized once sure of completing his project (DF vii. 1 n. 1; 308 n. 102). Simply put, we are lucky to have the last two volumes of the History: these owe their existence to (a) the surmounting of major intellectual obstacles – it was one thing to preach abjuration of prolix narrative, another to achieve it (Mem.E, 332); (b) the contingency that Gibbon went to Lausanne, rather than trying to earn his living as a Commissioner of Customs or Secretary at the Paris Embassy, which must have put an end to writing – Letters, 570–611 passim; Add. MSS 34882 f.256 ‘Reasons for and against accepting [the Paris embassy]’.

¹⁴⁷ The note to the Memoirs explaining the point (pr. p. 133 above) was no accidental insertion. Draft C of the Memoirs, though it stops c. 1770, was divided up into sections, and Gibbon's content headings for the subsequent sections survive; ‘Sect. IV’ covering 1776–82 includes the heading ‘Louis XVI’, a reference to this note, Add. MSS 34882 f. 253b.

¹⁴⁸ Mem. B, 195–7; cf. P. 142 and n. 75 above.

¹⁴⁹ Note three examples of this change in context, from the original essay of 1772 – in some sense self-sufficient, yet coming at the end of the ‘draught’ of 1771 – to the insertion of 1781: p. 142 on ‘simple and obvious’; n. 133 on Rome and Constantinople; p. 155 on ‘this history’.

¹⁵⁰ Milman, op. cit. (n. 79); Morison, J. Cotter, Gibbon (1878)Google Scholar.

¹⁵¹ e.g. Momigliano, op. cit. (n. 67), 49; Baridon, M., Edward Gibbon et la mythe de Rome (1975), 656–8Google Scholar; Pocock, J. G., ‘Gibbon as civic humanist’, Daedalus (Summer 1976), 103–19, at 115–16Google Scholar. Womersley is exempt from the criticism, but occupies the still less tenable position that the ‘Observations’ and Volumes II and III of the History form part of a progressive sequence – n. 1 above, Transformation of the Decline and Fall, op. cit. (n. 32), 188–91.

¹⁵² e.g. Cotter Morison, op. cit. (69), chs 7, 9, esp. p. 116: ‘From Constantine to Augustulus Gibbon is able to put forth all his strength’; Bury in DF i.x–xxi – representing ‘positivism’ both in its stricter, Comtean sense and its broader historiographical or evidential one. Bury in turn was greatly influenced by a linear historical, though not philosophical, descendant of Gibbon, Freeman, E. A., who came to the same conclusion: thus his famous declaration that ‘Whatever else we read, we must read Gibbon too’ (The Methods of Historical Study (1886), 104Google Scholar) is offset by stringent criticism of Gibbon on Byzantium (‘The Byzantine Empire’, in Historical Essays (3rd Series, 1879), 235–47).

¹⁵³ cf. Coleridge, Porter, op. cit. (n. 82); Bury's argument that Gibbon was trying ‘to prove a congenial thesis (DF i.ix) – a thesis which Bury had composed and, predictably, regarded as unanswerable (i.vii–viii) – is another vent to the same frustration.

¹⁵⁴ cf. ‘Gibbon's Dark Ages’, 21. Among many indications, we may note that, from the famous footnote on the Canary Islands (DF i.28 n. 94) to the end of Volume VI (vii.211 n. 114), Gibbon is waging a long-distance competition with Voltaire. The latter's great compilation the Essai sur les Moeurs (revealingly cited by Gibbon as Histoire Générate) held the field in 1772 as the first philosophical universal history; a central, though implicit, ambition of the History was to supersede it, which it did with great success, even in France, though cf. Walpole to Lady Ossory 8 Nov. 1789, Horace Walpole's Correspondence, xxxiv.79. Sniping at a lesser rival, the Universal History (1747–54), is also obvious: e.g. iii.58 n. 142; v. 43 n. 3, 339 n. 24, 422 n. 204, 484 n. 145, cf. Mem. F, 56.

¹⁵⁵ These are mots justes; after moving to Lausanne in 1783, Gibbon was careful to reassure his domestic audience that ‘I shall ever glory in the name and character of an Englishman’ (1788 Preface, DF i.xlvi); but he gloried yet more in the philosophical character of ‘Citizen of the World’, e.g. Letters, 642 (20 Jan. 1787); cf. Essai ch. XXXIX, Mem. E, 335 etc.

¹⁵⁶ Marx (and Engels), The German Ideology Part I, ed. Arthur, C. J. (1970), 39Google Scholar; Marx individually was an (the ?) exception, but not his nineteenth-century followers. In Britain and France this pathology may be traced simultaneously but separately, with J. S. Mill and Comte attempting to embrace history in their broader philosophical schemes, but failing actually to write it. Conversely, historical authors such as Macaulay, though retaining distinct and interesting philosophical notions, had too little system or originality to make any impact in this guise.

¹⁵⁷ cf. nn. 79, in, 115. Neither the aim of reading itself, nor the reading and critical environment of the modern academic is remotely like that of the contemporary audience; the obstacles cited, for example, by Walpole (n.91) to reading Gibbon's second and third volumes hardly apply today, but it is this apparently helpful fact which has tended to deceive us.

¹⁵⁸ Barring his opening a priori, Womersley, 47.

¹⁵⁹ See n. 152 above.

¹⁶⁰ DF ii.181. This is a more sophisticated version of a view expressed by Gibbon, in his Essai of 1761, ch. LXXXII n. *; Le Journal de Gibbon a Lausanne 1763–4 ed. , Bonnard, 19 December 1763Google Scholar.

¹⁶¹ DF i.41–3, 56; ii.61–7; iii.379–80; iv.73–4, 104–5, v.52–6 and n. 22 etc.

¹⁶² Sometimes the mutual support of East and West is highlighted (DF iii.417, iv.2), sometimes the reverse (iii.243–4, 454–5); and some cases are enmeshed in Gibbon's ambiguities (iv.30–1). Territorial division within the Western empire is also of great significance, DF iii. 422.

¹⁶³ DF iii. 11; reiterated iii.229, 421. The confusion of this experience in detail corresponds to larger structural uncertainties as to the merits of ‘Eastern’ versus ‘Western’ history, and will also account for the declaration of 1790–1 – when considering revision of the History – that ‘The distinction of North and South is real and intelligible … But the difference of East and West is arbitrary, and shifts round the globe’, EE 339.

¹⁶⁴ The lack of any footnote cross-reference is a striking negative argument here, given the presence of such notes elsewhere in the ‘Observations’, DF iv.177 n. 7. On the inadequacy of mere positivistic criticism: above pp. 144–6 and n. 133.

¹⁶⁵ However, those not convinced by the reconstruction which follows may be attracted by this solution as the most free from other perplexities.

¹⁶⁶ On Hurd: ‘Gibbon's Dark Ages’, 18; on the footnotes, above p. 136. Further evidence will be found at p. 137 and n. 35; p. 142 (‘prophetic’); and n. 128 above.

¹⁶⁷ Mem. C, 284; on the distinction made there between Muratori's Annali and Antiquitates, as against the Rerum Italicarum Scriptores, cf. DF vi.174 n. 1, vii.311 n. 110. The last is not cited until the beginning of Volume IV, DF iv.218 n. 120 [ch. 39]. It should be noted that ‘the era of the Western empire’ is not precise; the ‘General Observations’ plainly take in Justinian's reconquest (DF iv.174), and this is mirrored by Gibbon's ‘reconnaissance’ of 1771–2, where his coverage of primary sources remained extensive into the sixth century, including all of Procopius, Agathias, Jordanes, Cassiodorus and Justinian's Code. This is not to be inferred directly from the Memoirs, but from Gibbon's command of these sources when writing the early chapters of Volume I, cf. Add. MSS 34882 ff. 108–115b, an index of first citations of sources in the History. Thus, so far as chronology goes, the ‘General Observations’ might as well have been inserted at the end of Volume IV as of Volume III – renewed indication of the plasticity of Gibbon's original structural outlines, and his particular difficulties in the years 1781–4, cf. Mem. E, 308, 325; ‘Gibbon's Dark Ages’, n. 109 and pp. 20–3.

¹⁶⁸ DF iv.21 n. 52; Mem. C, 270, 284.

¹⁶⁹ ‘Gibbon's Dark Ages’, 19.