Hostname: page-component-77f85d65b8-zzw9c Total loading time: 0 Render date: 2026-04-17T22:03:40.678Z Has data issue: false hasContentIssue false
  • English
  • Français

Dionysius of Halicarnassus and Roman chronology

Published online by Cambridge University Press:  28 February 2013

C. E. Schultze
C. E. Schultze
Affiliation:
Department of Classics and Ancient History, University of Durham
Get access Rights & Permissions [Opens in a new window]

Extract

For Dionysius of Halicarnassus, an important part of the historian's task is the gathering and analysis of material. The present article is concerned with one particular aspect of this, namely, the investigation of chronology. It has two aims: first, to defend the accuracy of Dionysius' chronological system against the sometimes unfair criticisms of modern scholars; second, to assess how, in his perception, the role of chronographical research related to a historical work on a noncontemporary subject.

The research qualities Dionysius singled out for praise in the writings of Theopompus – a historian he much admired – were care, effort, preparation, and autoptic investigation. Not all of these were applicable to a historian who chose a non-contemporary theme. However, within the limits determined by his subject – Rome's origins and history to 264 B.C. – Dionysius similarly attempted preparatory work and investigation of data for his history.

Information

Type
Research Article
Information
Proceedings of the Cambridge Philological Society (Second Series) , Volume 41 , 1996 , pp. 192 - 214
Copyright
Copyright © The Author(s). Published online by Cambridge University Press 1996

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.)

Article purchase

Temporarily unavailable

References

Notes

1. Tim Cornell, Raymond Davis, Chris Kraus, Peter Rhodes and Tony Woodman kindly read this paper, and I am grateful for many suggestions. Earlier versions were also given at seminars in Belfast and Newcastle: I should like to thank the participants for much useful discussion. Any errors which remain are my own.

2. Other aspects are the choice and the handling of subject, and the moral attitude adopted by the historian.

(Pomp. 6, UR 2.245.1–5)

… As one who concerned himself with history he is worthy of praise … in particular for the care and effort in his writing. For it is clear, though he made no mention of it, that he undertook a great deal of preparation for this work, expending large amounts on the collection of materials, and in addition becoming an eyewitness of many things …

3. Known as , or more usually Antiquitates Romanae (AR). All references of the form 1.70.1 are to this work.

4. On autopsy, see Nenci, G., ‘Il motivo della autopsia nella storiografia greca’, SCO 3 (1955) 16–46Google Scholar; Schepens, G., ‘Ephoros sur la valeur de l'autopsie’, Ancient Society 1 (1970) 163–82Google Scholar.

5. Andrèn, A., ‘Dionysius of Halicarnassus on Roman monuments’, Hommages à L. Herrmann (1960) 88–104Google Scholar; Dubourdieu, A., ‘Denys d'Halicarnasse et Lavinium’, Pallas 39 (1993) 71–82CrossRefGoogle Scholar.

6. Musti, D., Tendenze nella storiografia romana e greca su Roma arcaica. Studi su Livio e Dionigi d'Alicarnasso (1970)Google Scholar; Gabba, E., Dionysius and the history of archaic Rome (1991) 98–118Google Scholar.

7. See Untersteiner, M., ‘Dionisio di Alicarnasso, fondatore della critica pseudepigraphica’, AFC 7 (1959) 72–93Google Scholar; Marenghi, G., Dionisio di Alicarnasso: Dinarco (n.d., c. 1970) 36–42Google Scholar.

8. See Table 1 for a summary of the significant dates upon which Dionysius based his chronological calculations.

9. Jacoby, Thus, FGH Komm. 2D 826Google Scholar.

10. FGH 251 Fl (Clement of Alexandria), F4 (Suda). Jacoby, (Komm. 2D 826–7)Google Scholar considers it unlikely to have been a major source of Clement's.

11. F1 (Inachus), F3 (Thasos), F5 (dynasties), F4 (Euripides). However, in the case of F4, confusion with a work by Dionysius the mousikos may have occurred (Jacoby, , Komm. 2D 826Google Scholar).

12. Compare Livy's procedure: Luce, T. J., Livy. The composition of his history (1977) 185–229Google Scholar.

13. There is no indication that Dionysius shared Varro's tripartite division into the adelon, mythikon and historikon periods (Censorinus, , DN 21Google Scholar). However, note the mention in 1.73.3–4 of the three possible foundations of Rome: . Cf. 1.74.1.

14. Note also the discussion of Numa and Pythagoras in 2.59: Olympiad dates, generations and reference to universal history (2.59.2). See also FGH 251 F1, F3–F5.

15. See below (pp. 206) for examples of these.

16. The provenance of these fasti will be considered below (pp. 204–5); the present aim is to determine what the requirements of the system were.

17. Gabba, E., ‘Un documento censorio in Dionigi d'Alicarnasso 1.74.5’, in Synteleia Vincenzo Arangio-Ruiz I (1964) 486–93Google Scholar. Gabba convincingly argues that Dionysius had seen this particular document. In brief: that there is no need for doubt on account of cognomina, which were in older Fasti too; that the census itself is most likely authentic; that Papirii were prominent in this period; for censorial records cf. Varro, , LL 6.86Google Scholar; Gell. 3.10; that private archival material is quite feasible; and that on another occasion (4.22.1) when merely citing census figures, Dionysius makes no claim to have seen the document.

18. Leuze, O., Die römische Jahreszählung (1909) 183Google Scholar; cf. 6 n. 12; 111 (on Polybius).

19. Leuze (n. 18) 184; Mommsen, T., Römisches Staatsrecht I 3 (1887) 598–9Google Scholar; see however Leuze 342 on the difficulty of dating entry into office by the Fasti Triumphales data.

20. Note however that he apparently knew other arguments: ‘this is shown by many other things’ (1.74.5), which were no doubt included in the chronological treatise. For the procedure of working backwards cf. Leuze (n. 18) 181; Perl, G., Kritische Untersuchungen zu Diodors römischer Jahrzählung (1957) 4–17Google Scholar; Piérart, M., ‘Les dates de la chute de Troie et de la fondation de Rome: comput par génération ou comput à rebours?’, in Historia testis. Mélanges d'épigraphie, d'histoire ancienne et de philologie offerts à Tadeusz Zawadzki, ed. Piérart, M. and Curty, O. (Seges n.s. 7) (1989) 1–20Google Scholar. Most recently, de Cazenove, O., ‘La determination chronographique de la durée de la période royale à Rome’, in La Rome des premiers siècles. Légende et histoire. Actes de la table ronde en l'honneur de Massimo Pallottino (Biblioteca ‘Studi Etruschi’ 24) (1992) 69–98Google Scholar (especially 96–7), criticises modern views that backward reckoning was the primary mode of ancient chronological calculation.

21. Dionysius included a third Decemviral year; added to the 118 colleges (as in Varro), this makes 119 years: Gabba (n. 17) 487.

22. See Table 2 for the calculations based upon the date of Troy's fall.

23. Aeneas 6, Ascanius 37, Silvius 29, Aeneas 31, Latinus 51, Alba 39, Capetus 26, Capys 28, Calpetus 13, Tiberinus 8, Agrippa 41, Allodius 19, Aventinus 37, Proca 23, Amulius 42, Numitor 1 (see words quoted in text, showing Numitor was thought of as reigning a year before the foundation occurred in his second year). Thus 16 kings, 431 years.

24. Werner, R., Der Beginn der römischen Republik (1963) 117CrossRefGoogle Scholar. Werner goes on (118) to use this calculation to conclude that Dionysius really agrees with a duration of 243 years for the Roman kings, not the 244 years achieved by adding up reign-lengths: however, given Dionysius' consistent expressions, this cannot be correct.

25. (1.63.1); (1.63.3); (1.64.1); (1.64.3); (1.65.1).

26. Jacoby, F., Atthis (1949) 371 n. 99Google Scholar.

27. It is even easier to see the result in the following way – which may approximate to Dionysius' method of working. He ceases to mention ordinals for the kings after Ascanius: their reigns (given individually) total 387. Adding this composite reign to year T + 43 (Ascanius' last regnal year), we get T + 430 for Amulius' 42nd and last regnal year. The year for part of which he reigned, when the fatal Lupercalia occurred, and for the latter part of which Numitor reigned, is the next one (T + 431), and represents Numitor's first regnal year. Numitor's second regnal year is of course the next one – T + 432 – and in accordance with Dionysius' expression in which T + 1 is protos and so on, T + 432 is ‘the 432nd’. There is no reason why it should be the 433rd. The expression is equivalent to that used of Cato in 1.74.2 : one adds 432 to T (the, as it were, ‘valueless’ year of the war – not year 1, but 0) and what happens in that year (0 + 432 =) 432 happens in the 432nd year. See also Samuel, A. E., Greek and Roman chronology (1972) 198, 251–2Google Scholar on ἐνεστῶτος.

28. I owe this point to Professor P. J. Rhodes.

29. The question of generation reckoning, and the length accorded to generations, is very vexed: see Mosshammer, A. A., The chronicle of Eusebius and Greek chronographic tradition (1979) 101–5Google Scholar; and, most recently, Cazenove (n. 20) 86–90, with bibliography.

It does however seem suspicious that the figure 432 is one so replete with significant numbers: 4 × 4 × 3 × 3 × 3. This must suggest reckoning based on a 27-year generation: 16 kings × 27 years = 432. Note however that tradition did not make each king the son of his predecessor: (half-)brothers – Ascanius and Silvius, Amulius and Numitor – mean that there are actually only 14 (or 2 × 7) generations.

30. The death of Numitor's son (not named as Aegestus) while hunting receives mention in Origo gentis Romanae 19.3, while 19.4 on Rhea Silvia is from Valerius Antias. Livy says stirpem fratris uirilem interemit (1.3.11). See also Dio fr. 4.11; App. Reg. 1.2; Plut. Par. min. 36.

31. Counting complete reigns only (i.e. not the one year that Numitor ruled before Romulus and Remus departed) the Alban figures range from 8 to 51 years. Roman reigns cover periods from 24 to 44 years.

32. Trieber, C., ‘Zur Kritik des Eusebios’, Hermes 29 (1894) 124–42Google Scholar, with table.

33. Perl (n. 20) 11–17.

34. Liv. 1.3.6–10, with Ogilvie, R. M., A commentary on Livy, Books I–V (1965) 43–6Google Scholar.

35. For the other king-lists see Varro, , De gente populi Romani (ed. Fraccaro, ), frr. 8–16, 19, 27, 29, 30Google Scholar. His Alban list had the same order because Aventinus falls at position 13 (fr. 31 Fraccaro = Aug. CD 18.21). Reign-lengths: Cens. DN 21.1, with Fraccaro 255–7.

36. FGH 273 Fl 10; see also Komm. 3A 282–3Google Scholar on F70: he is not Livy's source; cf. Ogilvie (n. 34) 44.

37. Origo gentis Romanae 10.2; 13.8; 18.3, with Richard, J.-C., Pseudo-Aurélius Victor. Les origines du peuple romain (1983)Google Scholarad locc. Epitomarum Pisonis secundo (OGR 18.3) must be incorrect: Rawson, E., ‘The first Latin annalists’, Latomus 35 (1976) 689–717 (see 704)Google Scholar.

38. Frr. 8–13 Peter; cf. Ogilvie (n. 34) 34.

39. . See Leuze (n. 18) 171 n. 213. Note also Schröder, W. A., M. Porcius Cato: das erste Buck der Origines. Ausgabe und Erklärung der Fragmente (1971) 170–1Google Scholar; Cornell, T. J. ‘Aeneas and the twins: the development of the Roman foundation legend’, PCPS 21 (1975–1976) 1–32Google Scholar (see 4 n. 3).

40. Manganaro, G., ‘Una biblioteca storica nel ginnasio di Tauromenion e il P. Oxy. 1241’, PP 29 (1974) 389–409Google Scholar; Manganaro, G., ‘Una biblioteca storica nel ginnasio a Tauromenion nel II sec. a.C.’, in Alföldi, A., Römische Frühgeschichte: Kritik und Forschung seit 1964 (1976) 83–96Google Scholar (and see also 50–1). Musti (n. 6) 31 n. 4 claims too much in saying the dipinto shows Pictor gave a consistent list.

The inscription (following the text of Manganaro (1976) 87) reads:

41. Thus Ogilvie, R. M., CR n.s. 24 (1974) 65Google Scholar, reviewing Schröder (n. 39).

42. Strasburger, H., ‘Zur Sage von der Gründung Roms’, SHAW 5 (1968) 1–43 (see 17)Google Scholar.

43. Pearson, L., The Greek historians of the west. Timaeus and his predecessors (1987) 47Google Scholar n. 40: ‘Timaeus' figures seem not to have been reached by generation-counting, and there is no evidence that he had any faith in this method of establishing dates, though it may have been used by earlier historians of Sicily.’

44. Cf. Ogilvie (n. 34) 44, 47. See also Cornell (n. 39) for the twins' story as the early, native element.

45. It is scarcely possible to identify the inventor, but Piso was someone who showed concern for what might be called genealogical probability: see below, p. 206. Alternatively, the whole episode may be due to Valerius Antias (cf. n. 30 above).

46. For Tarquinian genealogy and Dionysius' preference of Piso over Gellius and Licinius, see 4.6–7; 6.4.1; 6.11.2. Tanaquil as Superbus' mother (5.3.1) is a slip by Dionysius (contrast 3.46.5; 3.47.4; 4.6.3). The problems and sources are discussed by Martin, P. M., ‘Le souci chronologique dans la tradition sur la généalogie des Tarquins’, in Caesarodunum 10 bis (1976) 55–64Google Scholar; Gantz, T. N., ‘The Tarquin dynasty’, Historia 24 (1975) 539–54Google Scholar; Bessone, L., ‘La gente Tarquinia’, RFIC 110 (1982) 394–415Google Scholar; Cazenove (n. 20) 86.

47. de Cazenove, O., ‘La chronologie des Bacchiades et celle des rois étrusques de Rome’, MEFRA 100(2) (1988) 615–48CrossRefGoogle Scholar.

48. Cazenove (n. 47) 619.

49. Cazenove (n. 47) 636.

50. Cazenove (n. 47) 622.

51. Pol. 1.5.1 with Walbank, F. W., Historical commentary on Polybius I (1951)Google Scholarad loc.

52. Mommsen, T., Römische Chronologie (1859) 122ffGoogle Scholar.; see Leuze (n. 18) 180; and for a discussion of the derived theories, and objections, Werner (n. 24) 140 n. 1.

53. Zon. 8.8.6: . Cf. Pol. 1.10.9–11.1 for the delay, which is quite compatible with an appeal arriving in Ap. Claudius' year of office – Werner (n. 24) 142 n. 1. As parallels, Leuze (n. 18) 5 cites Pol. 3.6.3 but this is not valid (see Walbank ad loc.) nor is D. S. 12.84.4 (Werner (n. 23) 143 n. 1 cont.); better is D. S. 1.4.7.

54. There is of course no suggestion that he covered any of the events of the war itself: Leuze (n. 18) 186.

55. I owe this suggestion to the late M. W. Frederiksen.

56. The years 333, 324, 309, 301; probably devised by Varro. Dionysius' system did not require them: essential to it was the synchronism with the Gallic Sack, and the dictator years break that synchronism (386 B.C. in reality, 390 according to Varro.) The dictator years are lacking in the annalistic sources: Drummond, A., ‘The dictator years’, Historia 27 (1978) 550–72Google Scholar (see 551). There is no sign that Dionysius knew the Varronian chronological system, though Drummond points out (556 n. 47) the similarity of 1.74.1 and Solin. 1.27.

57. Leuze (n. 18) 187.

58. As derived from Pol. 1.6.1–2 (with Walbank ad loc.); see tables in Leuze (n. 18) 147, 179.

59. Leuze (n. 18) 188: ‘rationeller’.

60. See Ogilvie (n. 34) ad loc.; and Ogilvie, R. M., ‘Livy, Licinius Macer and the libri lintei’, JRS 48 (1958) 41–6Google Scholar.

61. Leuze (n. 18) 191–4. This argument is accepted by Werner (n. 24) 109–10.

62. See Ogilvie, R. M., CR n.s. 15 (1965) 86Google Scholar, reviewing Werner (n. 24).

63. When an Olympiad number alone is mentioned, the first year is meant – e.g. 2.25.7, and the ‘formula’ dates given below. Dates in the second, third or fourth years are occasionally given: 1.3.4; 1.8.2; 6.34.1; 11.62.1; cf. 1.7.2. The abbreviations O = Olympiad, V = victor, A = archon, C = consuls are here used to indicate order, with subscript figures when an Olympiad year other than the first is specified. The formula OVAC predominates (10 uses): 5.1.1; 5.37.1; 5.50.1; 8.77.1; 9.56.1; 9.61.1; 10.1.1; 10.26.1; 10.53.1 (= 301 A.U.C.); 11.1.1 (the Decemvirate). Then COVA (5 occasions: 6.1.1; 6.49.1; 8.1.1; 9.18.1; 9.37.1. Once COA (victor omitted, synchronism with Xerxes instead: 9.1.1). See also, for years which are not the first of an Olympiad but are specially marked, 6.34.1 – CAO4V, 260 A.U.C.; 8.83.1 – CA, 270 A.U.C.; 11.62.1 – (first military tribunes) O3A.

64. As at 5.59.1 cf. 5.49.1.

65. There have been seventeen Olympiads in all; the text is lost for O1. 84.1.

66. Leuze (n. 18) 189 offers an alternative: that Dionysius included the college L. Quinctius, A. Sempronius found in Diodorus (12.77.1 (between 428 and 427 B.C.). However, as he points out, other colleges found in Diodorus are omitted in the systems of Polybius, Varro and Dionysius, so it is not easy to see why Dionysius should have included this one. Note however that Livy seems at 4.30.15 and 4.31.1 to have a trace of this tradition (see Ogilvie (n. 34) ad locc. and ad 4.20.8).

67. See Broughton, T. R. S., Magistrates of the Roman republic I (1951) s. aa. 392–390 B.C.Google Scholar

68. See e.g. Liv. 9.44.3. For the sort of variants possible, see Pinsent, J., Military tribunes and plebeian consuls: the Fasti from 444 V to 342 V (1975) 6–9Google Scholar.

69. Note that other near-contemporary users reached a different conclusion regarding the foundation date: Nepoti et Lutatio, opiniones Eratosthenis et Apollodori comprobantibus, olympiadis septimae anno secundo (Solin. 1.27 = FGH 244 T 7).

70. On chronology as ideological, and Apollodorus' decision to avoid mentioning Rome's foundation, see Gabba (n. 6) 198.

71. Leuze (n. 18) 159, 200–1.

72. Werner (n. 24) 163–5. Moreover, as he points out, Piso cannot be positively shown to have had a foundation date of O1. 7.1: the implication of Cens. DN 17.13 is that he had it as 757 B.C. (165 n. 2).

73. Werner (n. 24) 165–6 (cf. 111). He thinks Dionysius worked out his own tables on the basis of Macer's eponym lists; the frequency of citation is for him a corroborative indication (165 n. 1).

74. Cf. Ogilvie (n. 62).

75. Ogilvie (n. 34) 286; Ogilvie, R. M. ‘A correction’, Gymnasium 75 (1968) 505–9Google Scholar. (The point was in reply to Werner, R., ‘Die Auseinandersetzung der frührömischen Republik mit ihren Nachbarn in quellenkritischer Sicht’, Gymnasium 75 (1968) 45–72Google Scholar; Werner's subsequent rejoinder (‘Stellungnahme’, 509–19: see 512–13) is unconvincing.)

76. Walbank (n. 51) 663 on Pol. 6.11a argues that only the first part of the sentence refers to Polybius, and οὐδ᾽ ἐπὶ ϰτλ. to a different source: this is clearly correct. Musti, D., ‘Polibio e la storiografia romana arcaica’, in Polybe (Entretiens Fondation Hardt XX) (1974) 103–43Google Scholar, adds (127–8) that Cic. Rep. 2.18 gives Ol. 7.2 as foundation date on the authority of Graecorum annales. If based on Polybius, this would further tend to exclude any reference to a tabula by Polybius. For an application of the pinax reference to Polybius, and a summary of earlier views, see Steinby, T., ‘A pontifical document’, Arctos 2 (1958) 143–51Google Scholar, criticised by Gabba (n. 17). See also Frier, B. W., Libri Annales Pontificum Maximorum: the origins of the annalistic tradition (1979) 111–12Google Scholar.

77. FGH 566 F60, with Komm. 3B564–6. See above at n. 43.

78. ἀϱχιεϱεῦσι is Niebuhr's emendation for codd. ἀγχιστεῦσι, generally accepted (but see Cantarelli, L., ‘Origine degli Annales Maximi’, RFIC 26 (1898) 209–28Google Scholar, at 220 n. 2). Note Degrassi, A., ILLRP I no. 333Google Scholar for a second-century use of archiereus; cf. Pol. 22.1.2; 32.6.5.

79. Using pontifex as an example, Dionysius discusses at 2.73 the alternative ways of conveying Roman concepts to his Greek audience. He transliterates, gives the etymology, indicates the connotations of the term, and supplies several equivalents. See Mason, H. J., Greek terms for Roman institutions (1974) 115–16Google Scholar.

80. Frier (n. 76) 112 finds ‘nor does [Dionysius] assert that any earlier historian had actually used it’. This is too negative: if no–one had, there would be no point in denigrating its validity. He is right, however, to stress that Dionysius' fundamental objection is to ‘unexamined reliance on authority, no less to be shunned than the bald assertions of Polybius’ (cf. 113–14).

81. Holzapfel, L., ‘Zur römischen Chronologie’, Klio 12 (1912) 83–115CrossRefGoogle Scholar, suggested (99) Acilius, C.. Kornemann, E., ‘Die älteste Form der Pontifikalannalen’, Klio 11 (1911) 245–57Google Scholar pointed out the close agreement between Dionysius' expression and Cato's: tabula apud pontificem maximum (fr. 77P). Against the view that Dionysius could be referring to Cato see however Schröder (n. 39) 169. Leuze (n. 18) 197–8 suggested Piso, and also equated the pinax with the sacred books (11.62.3) and the libri lintei (Liv. 4.7.12): this is unconvincing. Gabba (n. 17) 486–7 also considers Piso likely. On balance, the view of Frier (n. 76) 276–7 that the reference is to Pictor or to Alimentus seems the most satisfactory.

82. See Frier (n. 76) 182–3; and on the hierai deltoi of 1.73.2, which again Dionysius does not claim to have seen, see 109–111.

83. Olympiad dates for kings: 2.58.3; 3.1.3; 3.36.1; 3.46.1; 4.1.1; 4.41.1. See also 4.62.6; 7.1.5; 7.3.1. Non-Olympiad: 5.37.1; 6.21.2.

84. Office entry: 6.49.2; 9.25.1; 10.59.1; 11.63.1. Tribunes: 6.89.2; 7.1.1 cf. 6.49.2.

85. For Dionysius, Tarquinius apparently dies in this actual year of Lake Regillus (496 B.C. according to him); Livy 2.21.5 has the death in 495 B.C., obviously connected with the date of 496 for the battle, cf. 2.21.3–4.

86. Other dates: 2.31.1; 2.50.4; 2.52.5.

87. In other instances, he dates by Roman festivals. The most striking instance of this – and one which strengthens the argument against his dependence on Varro – is the ‘birthday of Rome’, the festival of the Parilia (1.88.3). Dionysius leaves it an open question whether Romulus regarded the day as suitable for founding the city because it was already considered a day of rejoicing, or whether it became a festival because it was the birthday of the city. There is reference to the astronomical or astrological aspects of the day, or the ‘horoscope’ of Rome, which interested Varro (Plut., Rom. 12Google Scholar; Cic. Div. 2.98–9). See on this Grafton, A. T. and Swerdlow, N. M., ‘Technical chronology and astrological history in Varro, Censorinus and others’, CQ n.s. 35 (1985) 454–65CrossRefGoogle Scholar.

88. Cf. 8.41.5; 8.50.1. Note however 5.65.1.

89. See also 3.10.1; 3.23.19 (nearly 500 years – in speeches); 3.31.4 (487 years – in narrative).

90. 1.9.4; 1.11.2; 1.17.3; 1.22.3; 1.26.1; 1.44.3; 1.45.3. Year dating at 1.31.1; 1.44.3 is perhaps from Cato.