1. E.g. Plut. Mor. 346f: (‘Simonides, however, calls painting silent poetry and poetry articulate painting’). I am grateful to Dr Jaś Eisner without whose constant support I had not even thought about writing this paper; its argument had a considerable input from Prof. Glenn W. Most, and Jacob Goldstein kindly smoothened the English.

2. Most recently, an interpretation of this ‘story’ has appeared: Benediktson (2000).

3. To make it speak is likely to mean to make it written as well. For the habit of reading aloud texts at the period before the invention of an ‘inner voice’ cf. Svenbro (1993), 44-63 and 163-66.

4. Morrison (1988), 269.

5. Morrison (1988), 271.

6. Prudentius’ text is quoted from Cunningham (1966).

7. This is especially likely if one takes Peri. 9, 11 and 12 as a triptych of pilgrimage poems (cf., e.g., Roberts [1993], 131ff.).

8. For the poem’s dating cf. most recently Trankle (1999) who dates Prudentius’ visit to Rome in 395-397; assuming that the narrator of Peri. 9 wants to evoke in the poem Prudentius’ own visit and that the poem has been written shortly after this visit we can date it in this period; in any case it must be written between 392 and Prudentius’ death: cf. Palmer (1989), 23f.

9. Translations of Prudentius’ text are taken’sometimes adapted—from Thomson (1953).

10. Placuit picturas in ecclesia non esse debere, ne quod colitur et adoratur in parietibus depingatur (‘it was agreed that there shall be no paintings in a church, so that what is worshipped and venerated be not depicted on walls’). For the Council, cf. the references given by Trout (1999), 182 n.135, and also Murray (1977), 317f.

11. Cf., e.g., Metzler (1973) and Murray (1977).

12. Cf. Morrison (1988), 288: ‘In the emptiness of their two-dimensionality and silence, paintings lacked the power of discourse to command, criticize, and, above all, question.’

13. This is not to suggest that Prudentius knew this text, though how well-known it was is attested by quotations not only in Greek pagan authors (D.L. 2.125; Luc. Merc. Cond. 42 and Rh. Pr. 6), but also in the Latin Christian Tertullian (Praescr. 39.4). A summary of the Tabula’s influence in antiquity is given by Fitzgerald and White (1983), 7f.; for two more recent discussions of this neglected text which in greater detail discuss matters here only touched upon cf. Trapp (1997) and Eisner (1995).

The Tabula Cebetis is not the only text which displays the scheme I am going to set out. It also appears in Lucian’s Hercules, ps.-Lucian’s Amores, the gallery-scene in Petronius’ Satyrica (82-91), in Callistratus Descriptions of Statues 6 and, somewhat transformed, in the opening of Longus’ Daphnis and Chloe and the opening of Achilles Tatius’ Leucippe and Cleitophon. Predecessors are found in tragedy and Hellenistic poetry: the parodos of Euripides’ Ion, Theocritus Idyll 15 and Herodas Mime 4; a real-life counterpart can be grasped in Pausanias’ exegetes (e.g. 6.24.9). I hope to assemble all these texts and discuss the relevance of the scheme of the ‘exegetic triangle’—as I call the topos—elsewhere (cf. also Schissel von Fleschenberg [1913] who assembles and discusses some of these texts, though in a very formalistic way).

14. The pinax described in the Tabula was dedicated in a sanctuary of Kronos.

15. The text quoted is from Praechter (1893); the translations are, sometimes adapted, those of Fitzgerald and White (1983).

16. It does not matter much whether this pinax actually existed (for the discussion of this question cf. Fitzgerald and White [1983], 8 n.49).

17. For the entire complex, cf. Mercogliano (1987); for recent literature see Trout (1999), 44 n.129.

18. The text quoted is from Ritter von Hartel and Kamptner (1999).

19. Cod. Paris. NAL 1443. For this codex and its readings cf. Lehmann (1992).

20. For in situ remains of the apse mosaic cf. Lehmann (1990), 81. Apparently a new reconstruction of the painting which makes use of these remains has not yet appeared. Mercogliano (1987), 175-79, reprints attempts to reconstruct the apse painting from Paulinus’ poetic description and compares it with the mosaic of S. Apollinare in Classe (Ravenna). An attempt to reconstruct both the mosaic’s and the inscription’s respective position is found in Lehmann (1992), 281.

21. Verse numbers are given in accordance with the line numbers of Ritter von Hartel and Kamptner (1999), i.286; translations are, sometimes adapted, Goldschmidt’s (1940).

22. The respective quotes run: absidem solo et parietibus marmoratam camera musiuo inlusa clarificat, cuius picturae hi uersus sunt (‘The apse, which has a floor and walls of marble, is made bright by a vault pleasantly decorated with mosaics; the painting of which bears the following scenes’, ch.10 for the apse in Nola) and ideo super hocpraeter picturam gratia geminatus est titulus (“Therefore there is a separate second inscription with regard to this mercy which is detached from the painting’, ch.17 for the apse in Fundi).

23. I doubt that this translation of Goldschmidt’s (1940) is correct. Cf. p.165 below.

24. Paulinus in this passage discusses representations of scenes from the Old and New Testament which were also part of the church’s iconographic program. These representations are, of course, not figurative, and one might regard it as a loose plank in my analysis if in what follows I use the passage quoted to analyse the way Paulinus wants the tituli of the allegorical apse-painting to be understood. However, Paulinus does not exempt these tituli; rather, he wants Nicetas to see omnia in the church (27.513), so that I see my approach justified.

25. Trout (1999), 182.

26. Indeed, in the following passages of the poem, Paulinus lengthily describes how pilgrims preferred to have meals and drinks rather than devote themselves with uota to God (cf. esp. 558ff.).

27. The translation is taken from Conybeare (2000), 95; cf. also ibid. n.21.

28. It seems as if Paulinus’ uacuae figurae is just another term for the ‘Hermeneutic Gap’ with which I have started this essay. As we shall see, it is Paulinus’ aim to remove this misunderstanding.

29. Cf. Eisner (1995), 279-87, esp. 283 for ‘typological’ exegesis.

30. Conybeare (2000), 97 (my italics).

31. Goldschmidt (1940), 63.

32. Conybeare (2000), 95.

33. Cf. Harris (1989), 314f. Harris’s views, however, have been repeatedly challenged: cf., e.g., Bowman (1991); and esp. for Central Spain Churchin (1995).

34. A later parallel comes from Sil. 12.569; legeret uisu cuncta (‘he would have read it all by looking’).

35. Conybeare is right if she argues for the viewer’s active response as a most important feature in Paulinus’ work; it meets not quite the truth, however, if she illuminates this fact by a passage from Augustine’s Confessions which argues not for restricting, but for multiplying meaning: et si quid tertium et si quid quartum [Augustine had discussed before two divergent interpretations of Moses] et si quid omnino aliud uerum quispiam in his uerbis uidet, cur non illa omnia uidisse credatur (‘and if someone sees a third truth and a fourth and indeed any other truth whatever in these words, why should he [sc. Moses] not be thought to have seen all of them?’, Conf. 12.31). See Conybeare (2000), 99-102.

36. At the end of my section on Paulinus it might be worth remembering that Prudentius himself composed tituli for paintings and was well aware of the methods applied in such an undertaking. However, we cannot be sure whether the texts in the Dittochaeon were intended for existing or fictive paintings. Cf. for these texts Pillinger (1980) and esp. 12-18 on the question whether these poems were intended for existing paintings.

37. 1; we should note that in the Tabula there is no hint whatsoever as to where the sanctuary of Kronos in which the narrator encounters the pinax is to be located.

38. Both usages are well established in Prudentius’ period: cf. ThLL iv.584 and 585 respectively. Thomson (1953), 223, takes consultus as a participle.

39. Cf. pp.165f. above.

40. As historia, not as fabula: one notes the contrast between the first word of 19 (historiam) and the last word of 18 (fabula) which one might take as implying different levels of trustworthiness as well.

41. Roberts (1993), 140, however, arrives at the conclusion that ‘the Latin would bẹar the sense that the libri referred to followed, and were based on, the painting.’ Yet referre aliquid means in this context to represent, ‘to set forth anew’ (LS), and this means that the thing represented has been there before.

42. This statement—to turn away one’s attention from the painting—might find a parallel in Philostr. Imag. 1.1.1: (‘turn your eyes away from the painting itself so as to look only at the events on which it is based’) if one follows Eisner (1995), 30 n.28, and the Loeb translation (Fairbanks [1931], 7) I have just quoted in understanding as ‘to look away’ and not as ‘to gaze at’, as suggested by Bartsch(1989), 20n.19.

43. Roberts (1993), 140f.

44. Cf. Opelt (1967), who establishes that Prudentius’ poems also with a few exceptions noticeably stick to that scheme. Typical motifs in Prudentius are traced most recently by Evenepoel (1996).

45. Their typical scheme is conveniently summarised by Opelt (1967), 243: ‘1. Die “Exposition”: Einsetzen einer allgemeinen oder auch nur partiellen Christenverfolgung. 2. Aufforderung an den künftigen Märtyrer, er solle ein Opfer darbringen. 3. Seine Absage und Confessio. Sie erwächst meist aus dem dialogischen Wortgefecht mit dem Verfolger. Verhaftung des Bekenners. 4. Die eigentliche Passio ist der Wettkampf im Zufügen und Ertragen von körperlichem Schmerz. 5. Die standhaft ertragene Folter wird schließlich mit dem Tod gekrönt’ (added numbers 1-5 are mine).

46. It might be worth noting that in another case in which a martyrdom is related by the description of a painting, a cycle of paintings is described which quite closely resembles all the five points: Asterius’ story of the martyr Euphemia (Patrologia Graeca xl.333-37; cf. Grabar [1946], ii.72-74). Noteworthy is also, that at the beginning of this account a contrast is drawn to reading by the viewer relating that before he read Demosthenes and Aeschines.

47. It is highly likely, though not conclusively provable, that this interpretation is strongly supported by another fact. We do not know of any account of Cassian’s martyrdom before Prudentius’ poem; furthermore, all later accounts seem to be dependent upon Prudentius’ poem (cf. Palmer [1989], 242). It might well be that there was no written account at all before Prudentius had written his poem. In this case, the reader inevitably is led to the conclusion that Prudentius’ poem is what finally proves the credibility of the story.

48. Cf. Tab. Ceb. 3.1-4, quoted p. 162 above.

49. The same point, slightly differently argued, has been made by Eisner (1995), 42f.

50. Cf. Peri. 1.74; a passage which already has been used by Jager (2000), 88-90, to understand Peri. 9. (I gratefully owe the reference to Jager’s book to Prof. Philip Hardie.)

51. In such an undertaking Prudentius might find a parallel in the inscriptions Pope Damasus put up at tombs of martyrs: cf. for these tombs and the inscriptions the edition of Ferrua (1942) and, most recently, Carletti (1997), 158f.

52. Cf. also praefuerat (21).

53. Cf. p. 167 above.

54. Jones (2000), 9.

55. Jones (2000), 12; Vegetius 2.5 provides a testimony that is contemporary with Prudentius (Jones [1987], 149).

56. Gustafson (2000), 18ff. However, it needs to be noted that voluntary tattooing seems to have taken place among Christians as well; the earliest example cited by Gustafson (2000), 29f., stems from the mid-fourth century. It is interesting that Claudian once (Adv. Eutr. 2.342-45) calls a tattoo a titulus, notably using the same word Paulinus applies to the explanations in his church.

57. Jer. 31.33. Paul alludes to this passage in 2 Cor. 3.3.

58. As translated by Betz (1979), 324.

59. Cf. also Gustafson (2000), 29, on this passage: ‘[Paul] is deliberately invoking the degrading practice of punitive tattooing.’

60. Cf. on this passage in more detail Malamud (1990), 76f.

61. More recently, writing on the body as an instrument of exerting power has found instantiations in two famous writers: in its literal sense, it has been used by Franz Kafka in his ‘Penal Colony’; the metaphorical sense has been explored, e.g., by Foucault (1979), 25f.