The Salamis Epigram Boegehold, Alan L Greek, Roman and Byzantine Studies; Fall 1965; 6, 3; ProQuest pg. 179 The Salamis Epigram Alan L. Boegehold HBRODOTOS (8.94) reproduces and deplores an Athenian canard that Adeimantos led the Corinthians in craven re­ treat from Salamis. Plutarch sees malice in its mete repro­ duction, and in repudiating the story says that the Athenians let the Corinthians bury their dead from that battle, since they had acted as brave men, by the dty of Salamis. He adds that the following epigram was inscribed on the grave monument: 'f'n l:. I "~I ", " V I 8 ~,: !:. E"VE" E"VVUpOV 7'fO'T E"V(tLOfLE"V /Illtognll'h i,y \F. K. Pritchett) ALAN L. BOEGEHOLD 181 PLATE 6, figures 3 and 4 show that no top diagonal stroke was ever cut for a kappa, and that the line which might seem to be a lower diagonal stroke for a kappa does not join the vertical stroke of qoppa. Someone (the idle hand discussed infra?) may have started to change it into a kappa, but only one short diagonal stroke was added near the tail of qoppa, and so the putative change, if indeed the stroke is not accidental, was never effected. On the other hand, if kappa had been intended originally, the circle of qoppa, added as an afterthought, would surely have crowded the upsilon to its left. The single, visible, diagonal stroke is consequently an irrelevance. Rho, if its full form has been preserved,4 is idiosyncratic and does not identify any particular epichoric alphabet. Dragoumes noted its unusual form (I» and cited as a parallel an inscription found near Megara which contained the word or name, EYPONE:L has the same form and has long been regarded as distinc­ tively but not uniquely Megarian. Dragoumes concluded that the similar rho in the Salamis inscription only showed that this form of the letter was not alien to Corinth, and his conclusion finds abundant support in later publications.6 Nor is the triangular, tailless rho a stranger elsewhere in the Greek world. It is found scratched into potsherds, painted on vases, and cut into stone at Athens, on Chios, and possibly on Crete.? , Certainty is impossible. A tail would have extended below the line formed by the bottoms of letters in line 1 (ef. qoppa) and into an area that is today a shallow depression. See PLATH 6, fig. 3. & Jeffery, op.cit. 136, 138 no. 9. • H. Payne, Necrocorinthia (Oxford 1931) 160 said that the form is found early and late on Corinthian vases. I note in Necrocorinthia the following: p. 161 no. 4; p. 163 no. 11; p. 165 no. 37; p. 167 no. 61. Cf. Jeffrey, op.dt. 142, and in T. J. Dunbabin et al., Perachora II (Oxford 1962) 397 no. 55; 398 no. 109; and perhaps 401 no. 167. 7 Athens, ostraka: Kallias, son of Kratios in W. B. Dinsmoor, Hesperia Suppl. 5 (1941) 140 fig. 68; some of those prepared by hand K against Themistocles in O. Broneer, Hesperia 7 (1938) 239 fig. 68; another against Themistocles in O. Kern, Inse.Graee. (Bonn 1913) pI. 13; painted on vases: Perichthonios, Hermes, Aristaios in Graef-Langlotz, Die antiken Vasen von der Akropolis VI Athen I (Berlin 1925) plates 33, 35 (Beazley ABV 107); Artemis in E. Gerhard, Auserlesene grieeh. Vasenbilder I (Berlin 1840) pI. 21 (Beazley, ABV 333); Paris, Patroklia, Chiron, Gerhard, op.dt. III (1847) pI. 227, cut in stone: four times in IG 12 2 (see S. Dow, AJA 65 [1961] 352 and pI. 110); A. E. Raubitschek, Dedications from the Athenian Akropolis (Cam­ bridge [Mass.] 1949) nos. 53, 147,283 (illustrated in J. Kirchner, IIAB, no. 21; Chios, on stone: W. G. Forrest, BSA 58 (1963) 54f, pI. 15. Gortyna, on stone: Inscr.Cret. IV, p. 76 no. 22B (the tail is minimal). Eleutherna, on stone: Inscr.Cret. II, p. 147 no. 1. The radical changes proposed by R. Carpenter in AJP 84 (1963) 81ff for the date and origin of the Salamis stone rest in part on the shape of rho (ef n.4 supra) and on the supposition that [his particular shape of rho was totally unknown outside Megara. I have felt it necessary to enumeraTe 182 THE SALAMIS EPIGRAM The upsilon of KoplvOov, which Dragoumes printed, is not on the stone and has been left out by subsequent editors. There is, however, an omicron which has not been properly reported. Its center can be seen ·025 m. above and a little right of the center of theta. Dragoumes included it in his account of the letters that were scratched into the stone by some idle hand, but the size of this omicron, identical with that of others in line 1, the skill with which it was cut, and its position associate it beyond question with the original letters of the inscription. Dragoumes noted its superior execution without further remark (op.cit. 53). It is, in fact, the final omicron of fJoplvOo. To the right of theta there are marks which have suggested the presence of omicron in the place one would normally expect, but close study of those marks reveals a high place on the stone that excludes the possibility of an omicron to the right of theta. If a start was made, the circle was never completed.s The resulting genitive in -0 is unusual in a Corin­ thian inscription. M. N. Tod in GHI 12 no. 16 prints Koplv8w without comment, but a poet who was not himself a COlinthian could in fact have written f)oplvOo as a genitive since that form was generally acceptable in the early fifth century. It was only around Corinth that special care was taken to write such genitive endings in full, vi{. -OY. But other explanations are also possible. Of the second line of the distich, tau, although faint, is indubitable, and enough of it can be measured to give assurance that these letters were originally as large as those in the first line. The short, diagonal stroke flOm which one infers nu could also serve for upsilon. Omicron and san, to the right of tau, are legible under strong, raking sunlight, and then they are unmistakable. Where other letters to the right were once cut, there remains now only worn stone. The other letters on the stone need a word. Dragoumes reported above letters 14-18 of line 1: IN(f)OOA (and above letter 23 a similar omicron which, as noted supra, really belongs to the original inscrip­ tion). These six letters he correctly identified as attempts by an idle hand to copy the old letters of the monument. Two other letters, (though without an exhaustive search) so many instances of tailless, triangular rho being written elsewhere because of Carpenter's own eminence and because the theory is but­ tressed by J. and L. Robert's approval ("hypothese raisonnable sur les circonstances," REG 77 [1964] p. 130 no. 14); authorities so impressive could inspire general belief despite the fact that the theory is controverted by the data on which it is built. 8 The last two letters of the inscription on the base of a statue by Phaidimos can be com­ pared. See Gisela M. A. Richter, The Archaic Gravestones of Attica (London 1961) fig. 200. ALAN L. BOEGEHOLD 183 I M (presumably epichoric IL), which are barely visible just below letters 14 and 15 of line 1, he also saw were not genuine.9 It seems, however, mOle than coincidence that the person who scratched these last two letters on the stone chose a collocation of letters that occurs twice in the second line of the epigram that literary tradition has preservedt and it may be justifiable to regard them as testimony of a witness who saw the second line when it was still visible.10 Physical Characteristics No detailed description of the physical characteristics of the stone exists elsewheret and yet some are directly relevant to a proper reconstruction of the text, and an enumeration of others will make dearer the difficult problem of restoring the original design of the monument. The stone is rectangular, 0·455 m. high and at present 0·79 m. wide, although it must once have been ca 1·10 m. The ten or eleven letters that are missing from the first line would have required about another 0·31 m. Its thickness is 0·06 m. The break at the left side which took roughly a third of each of the (four) lines with it was not accidental, for marks of the chisel that broke the stone can stilI be seen. It follows that whoever shaped the stone to his own uses would not have scrupled to use the smooth, inscribed face where it would often be stepped on. And indeed a shallow, irregular depression, deeper on the right side than on the left, extends through the whole width of the stone from under line 1 almost to the bottom. A straight­ edge, when laid on the inscribed face, perpendicular to the top. makes contact with the face 0·175 m. down from the top near the left edge of the stone, 0·124 m. down near the right edge. Below a line that passes with minor deviations between these two points, the hollow is deep enough to exclude any possibility of recovering letters that once may have been inscribed there. The meaning of this observation for the text of the epigram is elaborated infra. 9 Both are visible in the photograph (PLATE 5). Some editors have been led to print san as a dotted letter in the second line of the text, but it is patently not original. It crowds letters of the first line, while genuine letters of the second line are well separated (0·024 m.) from those of the first. W. Peek in Griechische Vers-Inschr!{ten I (Berlin 1955) no. 7 and in Griechische Grabgedichte (Berlin 1960) no. 4 prints "auo, €X]E' E[ ail.afLl" but after the most painstaking examination, neither R. S. Stroud nor I have been able to find the faintest trace of any of these letters. The stone where they ought to be was worn away long ago. If Peek is reading later, added letters as genuine, he is mistaken. 10 Where the two letters can be seen, the stone makes contact with aT-square laid on the top edge and inscribed face. The copied letters consequently were not scratched into an eroded surface from which earlier letters had vanished. 184 THE SALAMIS EPIGRAM The left side is broken; the right so chipped and worn that no original surface on the edge can be identified with confidence, although the last omicron would not have been inscribed above theta if theta were not very near the edge. Top and bottom edges are smoothly finished and original. The top edge meets the inscribed face at a right angle, the bottom edge meets it at an acute angle. The effect is that the stone, when standing on its bottom edge, cants imperceptibly away from a person who is facing it. The back is levelled to a rough, quarry finish. Along top and bottom, a band 0·06 m. wide has been chiselled, presumably to help the mason get true edges. Up the right edge (still on the back), there is another trimmed band, which is 0·04 m. wide. This band does not create a slight bevel as do those along top and bottom. It is level and can be interpreted as preparation of the back for a narrow, smoothly finished band which was once a resting surface but has now completely disappeared. No taper in any dimension can be ascertained. There is no cutting or projection to show how the stone was secured, and it is this lack especiaU y that makes difficult the most natural interpretation of what clues to its original setting the stone preserves. It could have been one of four or more similar plaquesll which revetted a square or rec­ tangular core of stones or rough masonry but for the absence of any traces of clamps or dowels, which should have fixed the revetment against its core. The stone does not seem thick enough to have stood by itself or to have borne weight without some form of fastening, yet another, more likely solution does not suggest itself. The Authenticity of Lines 3 and 4 The second distich of the traditional epigram has been regularly denounced as false. Its quality may not excite admiration, but the purportedly objective data which have led to its condemnation need to be re-examined. G. Kaibel12 quarreled with II/pacts scanned as a trochee in Doric fashion, and with the distinction expressed between Persians and Medes, which he thought wrong for the years directly after the battle of Salamis. He, however, would consign the whole epigram in subditiciorum numerum, an impulse that could no longer 11 U. von Wilamowitz, Sappho und Simonides (Berlin 1913) 193, thought there must have been other stones on which were inscribed the names of the fallen, but cf F. Jacoby, Kleine philologische Schriften I (Berlin 1961) 476 n.64. 12 "Quaestiones Simonideae," RhM N.P. 28 (1873) 442. ALAN L. BOEGEHOLD 185 be indulged after publication of the venerable stone with its archaic letters. Eminent authorities then and thereafter welcomed the stone as unimpeachable evidence that the second distich was a forgery,13 although Simonidean authorship of the first distich must remain a question. There was space on the stone for two more lines, but the lines, they asserted, had never been inscribed there. Now the importance of the observations reported supra becomes apparent. By the time Dragoumes found the stone, a hollow had been worn where two additional lines could have been inscribed. Conse­ quently the stone in no way proves that the original epigram con­ sisted of a single elegiac distich. To the contrary, the text on the stone compromises the validity of Kaibel's objection to the scansion of n,pua.s. The Doric shortening of the second syllable could reflect a learned, or a Doric, poet's careful observance of a characteristic feature of Doric verse,14 and the stone makes it clear that the epigram on the gravestone on Salamis was written in Doric Greek. The Doric Greek of the first line preserved on the stone, therefore, far from proving line 3 a forgery, actually establishes a presumption that line 3 is genuine. An ancient editor changed 7WK' and €va.tojL€S, but for some reason not vCiuos, to regular Attic-Ionic forms. Doric n,pua.s re­ mained, intractable, jarring in Attic-Ionic, but quite at home in the original Doric lines. The second distich cannot be judged false from the blank space on the stone. The quantity of the second syllable of n,pua.s is evidence for and not against the authenticity of line 3. What then of the assertion that Greeks shortly after Salamis were not aware of a distinction between Medes and Persians? No such pronouncement, regardless of how eminent or numerous the modern authorities who utter it, can be a positive criterion by which to judge the authenticity of an ancient text. We are not in a position to draw a map of the Greek mind as it existed during a given decade. Still less can we outline small islands on such a map and label them "Not Known." This is especially true in the present case, where information concerning the 13 Kaibel's objections to the second distich were accepted and repeated by e.g. U. von Wilamowitz, loc.cit. (supra n.11), A. Wilhelm, lOAI 2 (1899) 227, and C. M. Bowra, Early Greek Elegists (Cambridge [Mass.] 1938) 189. T. Bergk, PLG', Simonides no. 96 and T. Preger. IGM no. 6, had expressed reservations concerning them before publication of the stone. F. Jacoby, op.cit. 1.458, spoke of the epitaph as "preserved on the stone and in an enlarged form by literary tradition." Cf J. Geffcken, RE IlIA (1927) 194. 101 See D. L. Page, Aleman: The Partheneion (Oxford 1951) 132f. 186 THE SALAMIS EPIGRAM sometime distinction between Medes and Persians had been accessible to Greeks for a decade and more. It is not surprising, therefore, to observe a modification of learned opinion concerning this question. F. Jacoby in his study of the Marathon Epigram (which refers to "Persians") noted «the superstition that really old epigrams always speak of them as Mfj8o," but conceded that "the problem asks for special treatment."15 E. Lobel is willing, after necessary qualifications, to associate the name of Simonides with a fragmentary papyrus text which seems to name both Medes and Persians in succeeding lines,18 and C. M. Bowra17 agrees that the lines of the papyrus text may conceivably be the work of Simonides. Conclusion In summation: Plutarch and Favorinus knew the Salamis epigram as two elegiac distichs; Plutarch described the grave as being situated in the general area in which the gravestone was found; there is space for a second distich on the stone but no way to tell if it was ever engraved there or not; the epigram as written on the stone was in Doric Greek; the first line of the second distich, preserved only in literary tradition, has imbedded within its metrical scheme a charac­ teristic Doric shortening. In the absence of other specific objective evidence, there seems little reason to reject the tradition that the epigram was originally composed of two distichs, those which are preserved in substance in Plutarch and Favorinus.18 BROWN UNIVERSITY September, 1965 15 F. Jacoby, op.cit. 1.489 n.107. For a recent discussion of Mfj80s in Aesch. Persae 765, see ed. H. D. Broadhead (Cambridge 1960) comm. ad loco and pp. 278f. 18 POxy. 2327 fro 27, line 13: o.ppaTrop.€Vp.T]8[, line 14: Ka'Trfpawv· /)wpov8[. 17 Greek Lyric Poetry from Aleman to SimonidesS (Oxford 1961) 344 n.4. Bowra's citation should be corrected. 18 I am indebted to Professors F. Mitchel, W. K. Pritchett, R. S. Stroud, and E. Vanderpool, who examined particular letters of the inscription and improved the present study with criticism and suggestions. Mr Joseph Shaw, who drew the facsimile, and Mr Charles Williams were helpful on architectural questions. Professors R. Carpenter and S. Dow have enlightened me in various ways per litt. A grant from the Howard Foundation and a grant­ in-aid from the American Council of Learned Societies enabled me to do necessary work in the Epigraphical Museum in Athens.