Hostname: page-component-586b7cd67f-2plfb Total loading time: 0 Render date: 2024-11-30T16:08:35.143Z Has data issue: false hasContentIssue false
  • English
  • Français

THE HISTORIA AVGVSTA BEFORE MS PAL. LAT. 899: LOST MANUSCRIPTS AND SCRIBAL MEDIATION

Published online by Cambridge University Press:  13 April 2021

Martin Shedd
Martin Shedd*
Affiliation:
Hendrix College
*
[email protected]
Get access Rights & Permissions [Opens in a new window]

Abstract

This article re-evaluates the role of the manuscript tradition of the Historia Augusta in debates over the original contents and authorship of the text. Evidence for physical disruptions to the text before our oldest surviving manuscripts points to an earlier manuscript distributed across multiple codices. A multi-volume archetype eliminates critical arguments against the author's claims about lives missing before the Life of Hadrian as well as in the lacuna for the years a.d. 244–260. Other multi-volume codices of the eighth and ninth centuries show that loss of an initial volume would have disrupted the textual tradition for the index, titles and authorial attributions. Comparison of our most complete early witness, Pal. lat. 899, to the independent branches of the textual tradition shows discrepancies between these paratextual elements as expected in a disrupted tradition. Ultimately, this article concludes that the current debates on authorship and the original scope of the Historia Augusta rest on paratextual elements from a single branch of the manuscript tradition, raising doubts about the centrality of these controversies to understanding the work.

Keywords

authorshipcodexHistoria Augustalacunamanuscriptparatext
Type
Research Article
Information
The Classical Quarterly , Volume 71 , Issue 1 , May 2021 , pp. 402 - 421
Check for updates
Copyright
Copyright © The Author(s), 2021. Published by Cambridge University Press on behalf of The Classical Association

Access options

Get access to the full version of this content by using one of the access options below. (Log in options will check for institutional or personal access. Content may require purchase if you do not have access.)

References

1 These contradictions and their implications were first expressed in Dessau, H., ‘Über Zeit und Persönlichkeit der Scriptores Historiae Augustae’, Hermes 24 (1889), 378–92Google Scholar.

2 Most recent monographs and studies on the HA take these two positions: Burgersdijk, D.W.P., Style and Structure of the Historia Augusta (Amsterdam, 2010), 73–4Google Scholar and 79–81; Alan Cameron, ‘The Historia Augusta’, in id., The Last Pagans of Rome (Oxford, 2011), 743–82, at 745; M. Thomson, Studies in the Historia Augusta (Collection Latomus 337) (Brussels, 2012), 20–36 and 101; S. Ratti, Polémiques entre païens et chrétiens (Paris, 2012), 12 and L'Histoire Auguste: les Païens et les Chrétiens dans l'antiquité tardive (Paris, 2016), 229; and D. Rohrbacher, The Play of Allusion in the Historia Augusta (Madison, 2016), 3–10. See also the introductions to the Collection Budé editions cited individually throughout this article. E. Savino, Ricerche sull'Historia Augusta (Naples, 2017), 69–76 defends the authenticity of the lacuna, but accepts the fabrication of the authorial names.

3 Key studies on the textual transmission of the HA have focussed on its use by later writers: Bertrand, C., Desbordes, O. and Callu, J.P., ‘L'Histoire Auguste et l'historiographie médiévale’, Revue d'histoire des textes 14–15 (1984–5), 97–130Google Scholar; Callu, J.P., ‘L’“Histoire Auguste” de Petrarque’, Antiquitas 4 (1987), 81–115Google Scholar; and Callu, J.P. and Desbordes, O., ‘Le “Quattrocento” de l'Histoire Auguste’, Revue d'histoire des textes 19 (1989), 253–75CrossRefGoogle Scholar. O. Pecere, ‘Il codice Palatino dell'Historia Augusta come “edizione” continua’, in O. Pecere and M.D. Reeve (edd.) Formative Stages of Classical Traditions: Latin Texts from Antiquity to the Renaissance (Spoleto, 1995), 323–69 examines the early textual and scholarly tradition of MS Pal. lat. 899.

4 These are the Lives of Hadrian, Antoninus Pius, Marcus Aurelius, Verus, Commodus, Pertinax, Didius Julianus, Severus and Caracalla. The classifications were proposed by Mommsen, T., ‘Die Scriptores Historiae Augustae’, Hermes 25 (1890), 228–92Google Scholar, at 246, and refined by R. Syme, Emperors and Biography: Studies in the Historia Augusta (Oxford, 1974), 56–8 and Barnes, T.D., ‘Hadrian and Lucius Verus’, JRS 57 (1978), 65–79Google Scholar. Burgersdijk (n. 2), 30–4 treats in detail the scholarship on the categorization principles.

5 D. den Hengst, Emperors and Historiography (Leiden, 2010), 178–9. D. den Hengst, ‘The discussion of authorship’, in G. Bonamente (ed.), Historia Augusta Colloquium Perusinum (Bari, 2002), 187–95 questions whether the discussion of homogeneity and heterogeneity should be restricted to the options of six authors or one. The most recent computer study—J.A. Stover and M. Kestemont, ‘The authorship of the Historia Augusta: two new computational studies’, BICS 59 (2016), 140–57—indicates that the Primary Lives have an authorial style distinct from the later Lives.

6 This proposition was advanced by R. Syme, ‘Ignotus, the good biographer’, in A. Alföldi and J. Straub (edd.), Bonner Historia Augusta Colloquium 1966/67 (Bonn, 1968), 131–53; id., ‘Not Marius Maximus’, Hermes 96 (1968), 494–502; and T.D. Barnes, The Sources of the Historia Augusta (Brussels, 1978), 99–107, who identify the source as an unknown biographer (‘Ignotus’) from the third century. J. Schlumberger, Die Epitome de Caesaribus (Munich, 1974) and more recently A.R. Birley, ‘Marius Maximus, the consular biographer’, ANRW 2.34 (Berlin and New York, 1977), 2679–757 suggest that Marius Maximus was the primary source, which remains the dominant theory—see Burgersdijk (n. 2), 40–1. Kulikowski, M., ‘Marius Maximus in Ammianus and the “Historia Augusta”’, CQ 57 (2007), 244–56CrossRefGoogle Scholar identifies at least two prior sources, including Maximus.

7 In addition to the traits identified by Dessau (n. 1), both White, P., ‘The authorship of the Historia Augusta’, JRS 57 (1967), 115–33Google Scholar and Adams, J.N., ‘On the authorship of the Historia Augusta’, CQ 22 (1972), 186–94CrossRefGoogle Scholar identify unifying linguistic features shared by all six pseudonyms. The presentation of their data does not differentiate between the categories of Lives.

8 Thomson (n. 2), 37–53 contains a synopsis of the anachronisms. Most scholars see Prob. 24.2–3 as an allusion to the consuls of a.d. 395, although Cameron (n. 2), 772–8 argues for a date in the 370s.

9 Ratti (n. 2), passim. Ratti follows a tradition that sees the HA as pagan countercultural literature, a proposition championed by J. Straub, Heidnische Geschichtsapologetik (Bonn, 1963). See also F. Paschoud, ‘L'auteur de l'Histoire Auguste est-il un apostat?’, in J.P. Callu, F. Chausson and É. Wolff (edd.), Consuetudinis amor: fragments d'histoire romaine (II–VI siècles) offerts à Jean-Pierre Callu (Rome, 2003), 357–69 and F. Paschoud, Histoire Auguste: Vies d'Aurélien et de Tacite (Paris, 1996), xv.

10 Rohrbacher (n. 2), 21–9.

11 Savino (n. 2), 89.

12 Cameron (n. 2), 780–2.

13 The foundational theoretical study treating paratext as an element of a literary work is G. Gennette (transl. J.E. Lewin), Paratexts: Thresholds of Interpretation (Cambridge, 1997).

14 Bertrand et al. (n. 3), 126.

15 S. Ballou, The Manuscript Tradition of the Historia Augusta (Leipzig and Berlin, 1914), 41–2.

16 The main witnesses for reconstructing P's original readings are Codex Bambergensis (B), copied before major emendation to P, and Vat. 5301, the manuscript used for the first print edition. For an updated stemma, see Callu and Desbordes (n. 3), 274–5.

17 E. Hohl, Scriptores Historiae Augustae, vol. 1 (Leipzig, 1927; rev. 1997), viii–x.

18 Ballou (n. 15), 62–74. On the identity of the first corrector, see Ballou (n. 15), 5–6 and Hohl (n. 17), viii.

19 On the complex interrelationship between P and the other surviving manuscripts, see especially Pecere (n. 3) and Callu (n. 3).

20 Boyer, B., ‘Insular contributions to medieval literary tradition on the Continent’, CPh 43 (1948), 31–9Google Scholar demonstrates the independence of Π, following W.M. Lindsay, Palaeographia Latina, Part 3 (Oxford, 1924), 25. Manuscripts P and Π are viewable at the Digital Vatican Library: https://digi.vatlib.it/

21 Pecere (n. 3), Callu et al. (n. 3) and Callu (n. 3).

22 Callu et al. (n. 3), refuting Ballou (n. 15), who maintained that Σ depended on Petrarch's emendations of P.

23 See also the opening (Gall. 1.1): Capto Valeriano, enimuero unde incipienda est Gallieni uita, nisi ab eo praecipue malo, quo eius uita depressa est … This Life suffers from several short lacunae and textual deficiencies, although the Σ-branch preserves a better text than P.

24 See above, n. 2, as well as J.P. Callu, O. Desbordes and A. Gaden, Histoire Auguste: vies d'Hadrien, Aelius, Antonin (Paris, 1992), xlvii–xlix. S. Ratti, Histoire Auguste: vies des deux Valériens et des deux Galliens (Paris, 2000), vii–xxviii repeats the case for religious motivations behind the lacuna, against which see Rohrbacher (n. 2), 10. Recent scholarship on the lacuna largely follows the work of A.R. Birley, ‘The lacuna in the Historia Augusta’, in G. Alföldi and J. Straub (edd.), Bonner Historia Augusta Colloquium 1972/74 (Bonn, 1976), 55–62 and D. den Hengst, The Prefaces in the Historia Augusta (Amsterdam, 1981), 71–2. J. Fündling, Kommentar zur Vita Hadriani der Historia Augusta (Bonn, 2006), 10–13 reviews the history of the debate and takes an agnostic position.

25 Savino (n. 2), 70–1.

26 Dessau (n. 1), 363–7 identified two borrowings from Aurelius Victor accepted by subsequent scholarship. Information from the Caesares appears throughout the HA: Burgersdijk (n. 2), 16.

27 E.g. notice that Maximinus was declared a public enemy appears at Maximin. 15.2, Gord. 11.1 and Max. Balb. 1.4.

28 Gall. 1.2–3.5 describes the usurpations of Macrianus, Ballista, Quietus, Domitianus, Aureolus, Odenathus and Piso, each of whom receives individual treatment in the Thirty Tyrants.

29 These parallels have been used to speculate about lost historical works available to Zosimus and the HA author. See J. Schwartz, ‘Á propos des données chronographiques de L'Histoire Auguste’, in G. Alföldi (ed.), Bonner Historia Augusta Colloquium 1964/65 (Bonn, 1966), 197–210; F. Paschoud, Zosime: Histoire nouvelle (Paris, 1971), xxxvii–xl; Barnes (n. 6), 111.

30 den Hengst (n. 24), 71, restated in Rohrbacher (n. 2), 10.

31 den Hengst (n. 24), 71 notes that internal lacunae in Tacitus display such fraying but neglects the issue of Tacitus’ lost books.

32 All codices from the St. Gallen collection referred to in this article can be viewed at: http://www.e-codices.unifr.ch/en/list/csg. Multi-volume sets can be identified throughout the lists in G. Bekker, Catalogi bibliothecarum antiqui (Bonn, 1885).

33 A six-volume set of Gregory the Great's Moralia in Iob has volumes that range from 139 to 236 folios containing five to eight chapters each (St. Gallen, Stiftsbibliotek, Cod. Sang. 206–9). A ninth-century copy of Augustine's Commentary on the Psalms creates groupings of twenty-four to thirty-five psalms on 187 to 250 folios per volume (St. Gallen, Stiftsbibliotek, Cod. Sang. 162–6).

34 L.D. Reynolds, ‘Nonius Marcellus’, in ibid., Texts and Transmission (Oxford, 1983), 248–52.

35 Savino (n. 2), 69–76.

36 For example, Savino (n. 2), 101 notes that the Didius Julianus was switched with Avidius Cassius owing to spatial constraints in his first proposed volume, yet elsewhere proposes volumes longer than the chronological order would have created. The seven-thousand-word variance in proposed volumes undermines the proposition of strict spatial constraints without specifications for the manuscripts’ dimensions.

37 A. Chastagnol, Histoire Auguste: les empereurs romains des IIe et IIIe siècles (Paris, 1994), xxxv simply calls the existence of these lives ‘proprement invraisemblable’. Rohrbacher (n. 2), 5 echoes this sentiment. den Hengst (n. 24), 14–16 leaves room for a Suetonian series, ruling impossible only minor figures before Aelius.

38 Callu et al. (n. 24), xvi and xlviii–xlix proposes that the Lives of Nerva and Trajan were intentionally omitted. Paschoud (n. 9), xxvii suggests that the work once contained these Lives, now lost.

39 See the introduction, as well as Callu et al. (n. 24), xvii–xix; Birley (n. 6), passim; Cameron (n. 2), 778; Thomson (n. 2), 93.

40 In addition to the sources in n. 6, see A. Molinier, ‘Marius Maximus source latine de la Vie de Commode?’, in G. Bonamente, F. Heim and J.P. Callu (edd.), Historia Augusta Colloquium Argentoratense (Bari, 1998), 223–48.

41 These passages only conflict if in the phrase omnes qui post Caesarem dictatorem … uel Caesares uel Augusti uel principes appellati sunt the word post is interpreted as excluding its object from the series. Usage suggests that inclusivity is possible, especially in a sequence of related figures; TLL 10.2.169.80–170.7 and 174.59–175.26.

42 The character counts for each are presented without counting spaces as characters. Both texts were checked against one another and against the Teubner edition for major omissions of text and edited to remove extraneous numbering and footnotes.

43 Scriptores Historiae Augustae, The Latin Library, accessed 8 August 2020, http://thelatinlibrary.com/sha.html

44 B. Thayer, ‘The Historia Augusta’, Lacus Curtius: Into the Roman World, accessed 25 July 2019, http://penelope.uchicago.edu/Thayer/E/Roman/Texts/Historia_Augusta/home.html Text from D. Magie (transl.), Scriptores Historiae Augustae (Cambridge, MA, 1921, 1924, 1932).

45 See n. 33.

46 S.J. Tibbetts, ‘Suetonius’, in L.D. Reynolds (ed.), Texts and Transmission (Oxford, 1983), 399–404.

47 G. Regenos, The Letters of Lupus of Ferrières (The Hague, 1966), 52.

48 Kaster, R.A., ‘The transmission of Suetonius's Caesares in the Middle Ages’, TAPhA 144 (2014), 133–6Google Scholar.

49 M. Meckler, ‘The beginning of the Historia Augusta’, Historia 45 (1996), 364–75 and Savino (n. 2).

50 Savino (n. 2), 65–7.

51 Savino (n. 2), 65 n. 23.

52 Meckler (n. 49), 372.

53 Burgersdijk (n. 2), 54–8.

54 On the author's prefatory habits, see den Hengst (n. 24). The question of a lost introduction was raised by H. Dessau, ‘Über die Scriptores Historiae Augustae’, Hermes 27 (1892), 561–605, at 578–9 and R. Syme, Ammianus and the Historia Augusta (Oxford, 1968), 206–7.

55 C.E. Murgia, ‘The textual transmission’, in V.E. Pagán (ed.), A Companion to Tacitus (Chichester, 2012), 13–22.

56 On the creation of titles by booksellers and manuscript users, rather than by authors, see especially N. Horsfall, ‘Some problems of titulature in Roman literary history’, BICS 28 (1981), 103–14 and B.-J. Schröder, Titel und Text (New York, 1999).

57 L. Jansen (ed.), The Roman Paratext: Frame, Texts, Readers (Cambridge, 2014). Cases for authorial paratext are discussed in R. Gibson, ‘Starting with the index in Pliny’, 33–55 and S. Butler, ‘Cicero's capita’, 73–111, which note the consistency across manuscript branches, as well as in R. Rees, ‘Intertitles as deliberate misinformation in Ammianus Marcellinus’, 129–42, which draws on external confirmation about the paratext preceding the earliest manuscripts.

58 Example from Vat. lat. 1898, viewable at the Digital Vatican Library (n. 20).

59 Sedulius Scottus’ excerpts similarly give ex uita Caesarum: J. Klein, Über eine Handschrift des Nicolaus von Cues (Berlin, 1866), 94–9.

60 J.P. Callu, ‘La prémièr diffusion de “l'Histoire Auguste”’, in J. Béranger and J. Straub (edd.), Bonner Historia Augusta Colloquium 1982/83 (Bonn, 1985), 89–129. Against, see Pecere (n. 3), 338 n. 41 on the significance of the term Caesarum to later audiences.

61 Pecere (n. 3), 329; Savino (n. 2), 68. Opil. 1.1 Vitae illorum principum seu tyrannorum siue Caesarum … M. Thomson, ‘The original title of the Historia Augusta’, Historia 56 (2007), 121–5 reaches a similar conclusion.

62 This comparison may help explain the glossing of principum as imperatorum, the former preserved in P, the latter possibly original to the Σ family.

63 Pecere (n. 3), 337–8 discusses wilful modification in the context of Π and M.

64 profectus dehinc ad bellum parthicum edito gladiatorio munere & congiario paulo (Πa)/pplo (Πb) dato.

65 See n. 56.

66 F. Eyssenhardt and H. Jordan, Scriptores Historiae Augustae ab Hadriano ad Numerianum (Berlin, 1864), iv. The scribe also seems to believe that Spartianus was the master compiler.

67 Hadr. 2.10, a Life attributed to Spartianus, with whom Maximus is paired in the index. Marius Maximus reappears at Hadr. 12.4, 20.3, 25.4; Heliogab. 3.9, 5.5; Ant. Pius 9.3; Aurel. 1.5, 25.10; Avid. Cass. 6.6–7, 9.5; Comm. 13.2, 15.4, 18.2; Pert. 2.8, 15.8; Sept. Sev. 15.6; Alb. 3.4, 9.2, 9.5, 12.14; Geta 2.1; and Heliogab. 9.6.

68 Pecere (n. 3), 329–30.

69 Quatt. Tyr. 1.4, 15.10; Carus 4.3, 19.1. Carus is listed after Numerianus only at Carus 18.3.

70 C. Bertrand-Dagenbach, Histoire Auguste: Vie d'Alexandre Sévère (Paris, 2014), xvii.

71 Aurelius Alexander: Aur. Vict. Caes. 24.1, Eutr. 8.23; Alexander: Cass. Dio 80.17.3, Hdn. 5.7.4, Euseb. Hist. eccl. 6.28, Jer. Chron., Amm. Marc. 26.6.19, Julian. Caes. 313a; Severus Alexander: Epit. de Caes. 24.1.

72 Neither the initial scribe nor the first two correctors provided a title for this Life.

73 Cf. CIL 7.965, Imp Caes M Aurelio Seuero Alexandro Pio Fel aug pont maximo trb pot cos p p coh i ael hispanorum eq devota numin

74 Index and 199r.

75 Hohl (n. 17), vol. 2; Paschoud (n. 9 [1996]). It is only by considering these sections as a single Life that scholars who see the number of thirty Lives as a significant structuring principle can reach that number: e.g. Thomson (n. 2), 97–100.

76 See n. 60.

77 R.W. Burgess, ‘Eutropius v.c. magister memoriae?’, CPh 96 (2001), 76–81, at 80: ‘It must be noted that incipits, explicits, dedications, colophons, headings, and subscriptions, in fact any short texts that stand separate from the beginning or end of an ancient author's work, are the most fragile parts of the text, often failing to be treated with the same consideration as the works themselves.’

78 Dessau (n. 1), 378–92.

79 Pesc. 9.3, attributed to Aelius Spartianus, and Clod. 1.4, attributed to Julius Capitolinus.

80 Amm. Marc. 28.4.14 cites Marius Maximus as one of two authors who receive attention from the aristocracy of his day. On the identity and works of Marius Maximus, see especially R.P.H. Green, ‘Marius Maximus and Ausonius’ Caesares’, CQ 31 (1981), 226–36; A.R. Birley, ‘Indirect means of tracing Marius Maximus’, in G. Bonamente and G. Paci (edd.), Historia Augusta Colloquium Maceratense (Bari, 1995), 55–74; and Birley (n. 6). See also the relevant passages in J. den Boeft, J.W. Drijvers, D. den Hengst and H.C. Teitler, Philological and Historical Commentary on Ammianus Marcellinus XXVIII (Leiden, 2011).

81 In the same passage, the author attributes the Life of Trajan to Marius Maximus and the fictionalized authors Aurelius Verus and Statius Valens.

82 G.B. Conte, Latin Literature: A History (Baltimore, 1994), 617 trusts the HA and includes a Life of Alexander in his entry on the historical Gargilius Martialis. Against this, see Syme (n. 54), 100.

83 Thomson (n. 2), 21 n. 11 agrees that this is a list of spurious authors but maintains that their appearance in the paratext was intended by the author.

84 Thomson (n. 2), 100–2 accuses the index of making a mistake here because of the ambiguity of the colophons. Ratti (n. 24), xiv–xv n. 15 traces the history of this emendation, which relies on conjecture rather than on manuscript-based evidence for the originality of an attribution to Pollio.

85 Burgersdijk (n. 2), 107–9 identifies Nepos as the inspiration for the combined biography.

86 The Life of Aurelian features prominently in discussions of the HA for its description of the author's agenda and the chronological inconsistency created by the character of Tiberianus. For analysis of the introduction, see den Hengst (n. 24), 94–110, Pausch, D., ‘libellus non tam diserte quam fideliter scriptus?’, AncNarr 8 (2010), 115–35Google Scholar, and the relevant passages in Paschoud (n. 9 [1996]). On the chronology, A. Chastagnol, Les Fastes de la prefecture du Rome au Bas-Empire (Paris, 1962), s.v. ‘Junius Tiberianus’ and Rohrbacher (n. 2), 7.

87 R. Syme, ‘Bogus authors’, in id., Historia Augusta Papers (Oxford, 1983), 103–5.

88 E.g. Maximin. 29.10 reliqua qui uolet nosse de rebus Veneriis et amatoriis, quibus eum Cordus aspergit, eundem legat. Cordus appears at Alb. 5.10, 7.3, 11.4; Opil. 1.3; Maximin. 4.1, 6.8, 27.7, 28.10, 29.10, 31; Gord. 4.6, 5.6, 12.1, 14.7, 17.3, 19.9, 21.3–4, 22.2, 26.2, 31.6, 33.4; Max. Balb. 4.2, 12.4.

89 Cameron, Alan, ‘Literary allusions in the Historia Augusta’, Hermes 92 (1964), 363–77Google Scholar.

90 A.R. Birley, ‘“Trebellius Pollio” and “Flavius Vopiscus”’, in G. Bonamente and F. Paschoud (edd.), Historia Augusta Colloquium Perusinum (Bari, 2002), 33–47. This interpretation has found favour also in Rohrbacher (n. 2), 21–2.

91 Birley (n. 90), 36; Asinius appears as a critic of Caesar's writings in Suet. Iul. 56.4.

92 Birley (n. 90), 37.

93 Honoré, T., ‘Scriptor Historiae Augustae’, JRS 77 (1987), 156–76Google Scholar, at 170–6.

94 Savino (n. 2), 91–5.

95 Thomson (n. 2), 30–1.

96 Thomson (n. 2), 90–3 considers this as evidence that the author intentionally sequenced the Lives out of order.

97 Savino (n. 2), 76–8 argues that the disorganization of the Lives occurred later in the transmission and suggests that we can resolve the Pescennius and Albinus attribution issue by rearranging the index chronologically without altering the titles. This would assign both to Spartianus but requires prioritizing the evidence of the index over the incipit, which attributes the Life of Albinus to Capitolinus.

98 For a detailed treatment on the types of liberties taken by manuscript scribes, see D. Wakelin, Scribal Correction and Literary Craft (Cambridge, 2014) and M. Fisher, Scribal Authorship and the Writing of History in Medieval England (Columbus, 2012), 16 n. 5.

99 On the titles, see n. 56.

100 F. Clover, ‘The Historia Augusta and the Latin Anthology’, in E. Birley and K. Rosen (edd.), Bonner Historia Augusta Colloquium 1986/89 (Bonn, 1991), 34–9 and Schröder (n. 56), 293–6. Similar issues plague the Greek Anthology: see A.S.F. Gow, The Greek Anthology: Sources and Ascriptions (London, 1958).

101 Sweeney, R.D., ‘The ascription of a certain class of MSS. of the “De viris illustribus” of the Pseudo-Aurelius Victor’, RhM 111 (1968), 191–2Google Scholar and Sage, M., ‘The “De viris illustribus”: authorship and date’, Hermes 108 (1980), 83–100Google Scholar. The most recent critical edition is P.M. Martin, Les hommes illustres de la ville de Rome (Paris, 2016).

102 Shortened from magistri mem<oriae> et <rhetoris latini>: see C.E.V. Nixon and B.S. Rodgers, In Praise of Later Roman Emperors: The Panegyrici Latini with Latin Text of R.A.B. Mynors (Berkeley, 1994), 9–10, at the suggestion of Seeck, O., ‘Studien zur Geschichte Diocletians und Constantins’, Neue Jahrbücher für Philologie und Pädagogik 137 (1888), 713–26Google Scholar.

103 Rees (n. 57) begins from an assumption of authorial control over the transmission of the paratext to conclude that the author could use these cues to mislead the reader. This proposition requires a reading public that is not an editing public or a public that can interact with the author. The book numbers in Ammianus are, however, attested prior to the earliest manuscripts. The same is not so for the HA.

104 Marriott, I., ‘The authorship of the Historia Augusta: two computer studies’, JRS 69 (1979), 65–77Google Scholar claimed to prove single authorship through computer analysis. Sansone, D., ‘The computer and the Historia Augusta: a note on Marriott’, JRS 80 (1990), 174–7Google Scholar outlined the methodological issues with Marriott's study, which was further discredited in B. Frischer, ‘How to do things with words per strong stop’, in H. Rosén (ed.), Aspects of Latin (Innsbruck, 1993), 585–99. Three related articles likewise tested the single authorship hypothesis against the six named authors and found that results slightly favoured multiple authorship. These are P.J., and Gurney, L.W., ‘Authorship attribution of the Scriptores Historiae Augustae’, Literacy and Linguistic Computing 13 (1998), 119–31Google Scholar; P.J., and Gurney, L.W., ‘Subsets and homogeneity: authorship attribution in the Scriptores Historiae Augustae’, Literacy and Linguistic Computing 13 (1998), 133–40Google Scholar; and Tse, E.K., Tweedie, F.J. and Frischer, B., ‘Unravelling the purple thread: function word variability and the Scriptores Historiae Augustae’, Literacy and Linguistic Computing 13 (1998), 141–9CrossRefGoogle Scholar. See also Stover and Kestemont (n. 5).