² Doenges, N.A., The Letters of Themistocles: A Survey (Diss. Princeton, 1953), p. 22.Google Scholar

³ See e.g. Lenardon, R.J., ‘Charon, Thucydides and “Themistokles”’, Phoenix 15 (1961), 35:CrossRefGoogle Scholar ‘If the letters of Themistokles represent a kind of historical novelette, their artistic construction … must be considered a dismal failure.’ Cf. Nylander, C., “ΑσσύριΛ Γράμματα: Remarks on the 21st “Letter of Themistokles”’, Opuscula Atheniensia 8 (1968), 131:Google Scholar ‘The theory of the “historical novelette” or “historische Roman” has had to face so many contradictions and oddities within the collection that the general reaction has been a very strong criticism of the author’s — or authors’ — ineptness and complete lack of talent.... What we have here is clearly not a structural whole and it seems useless to look at these letters as a unit in a more than formal and superficial sense.’ However, Lenardon is now prepared to grant the Letters a measure of artistic respectability. See ST, p. 155: ‘The epistolographer works very much in the manner of a good historical novelist who uses his evidence to create his drama, to weave his artistic design and to elucidate his political or philosophical points of view … Underlying all is a sincere personal and tragic note and a dramatic thrust…’

⁴ Podlecki, A.J., The Life of Themistocles (Montreal and London, 1975), p. 131:Google Scholar ‘New items of information for Themistocles’ career [contained in the Letters] must be treated with the utmost circumspection.’

This present essay seeks to offer a literary interpretation of the Letters. Since for such an interpretation the historical accuracy or otherwise of the facts presented in them is irrelevant, I have sought as far as possible to avoid discussion of historical problems. Historical accuracy is not the artist’s first concern; my reading of the Letters suggests that their author was no more a purist in this matter than Lucan, Shakespeare or Kubrick. Historical and chronological problems will only be raised where they affect our understanding of the Letters as a work of literature.

⁵ Cf. e.g. the remark of Bentley, quoted also by Nylander, op. cit. (above n.3), 119: ‘It will be no unpleasant labour … to pull off the disguise from those little pedants, that have stalked about so long in the apparel of heroes’ (Dissertations upon the Epistles of Phalaris, Themistocles, etc., ed. Dyce, A. [London, 1836] ii, p. 182).Google Scholar Critical appraisal of the Letters reached its nadir with Jackson, J., ‘The Text of the Epistles of Themistocles’, CQ 19 (1925), 167,CrossRefGoogle Scholar whose odiously supercilious remarks regarding both the Letters and its manuscript do not merit quotation.

⁶ So e.g. Doenges, op. cit. (above n.2), p. 5.

⁷ There is a slight difficulty here in that Letter 11 appears to indicate that T.’s wife is still in Athens (752.7). This conflicts with the tradition recorded by Plutarch from Stesimbrotos that T.’s wife and children were smuggled out of Athens and conveyed to him in Epeiros by Epikrates of Acharnai (Plut. Them. 24.4), a tradition which seems to be accepted by modern historians (cf. Podlecki, op. cit. [above n.4] , p. 41 and Lenardon, ST, p. 136). And certainly, if my interpretation of Letter 21 is correct (see below p. 99), the author of the Letters must have taken it that T.’s wife and children were with him in Magnesia. However, (i) the author is certainly not aware of the Epikrates story — in Letter 4 it is Sikinnos who is entrusted with contingency plans for getting T.’s family out of Athens (see 744.24–48), and the actual performance of this exploit is never narrated; (ii) the text at 752.7 is corrupt — κατα was inserted before ττ)ς άΰλίας μου ó by Caryophilus in order to make some sort of sense of the passage (for without it the implication is that T.’s wife is plotting against him rather than being plotted against along with him), but it is quite possible that more words have dropped out (e.g. κατά τν); (iii) in any case, even if one accepts Caryophilus’ reading, it need not necessarily mean that T.’s wife is in Athens — if T. can be ‘plotted against’ in his absence, so can she.

⁸ Needless to say I do not accept the fictive date proposedfor Letter 21 by Nylander, op. cit. (above n.3), 130. See below n.34. It is strange in fact that Nylander fails to perceive the pattern in the collection so strongly suggested by the diagram on p.133 ofhis article — a diagram which does not at all support his earlier contention that ‘the order of the letters is, from the point of view of contents, chronology and logic, so confused thatitis even difficult to understand what the author … wanted to achieve’ (op. cit. 119).

⁹ For a longer list of discrepancies cf. Nylander, op. cit.(above n.3), 132 n.55.

¹⁰ μή áyvóei 6τι ατεχυώς [ώ Παυσανία] τοίς άνρώπους τα ατυχήματα τρέφσνσιν at etinpaylai (742. 15–16).

¹¹ So P (= Palatums 398, the only MS. in which the Letters are preserved). Hercher’s έπί νόμοκ reκαι δήμοις seems unnecessary and gives inferior sense.

¹² 741. 22. I have restored P’s φύσω for Caryophilus’ φυyήν — the expression may be awkward, but I find it hard to conceive how φχητν could have become φύσιν. A clear case for the lectio difficilior.

¹³ αλλούχ οϋτως ήδη μιχτητέον Αθηναίους ημΐν, στ€ και τοιαΰτά τίνα περί αύτύν ύπολαμβάνειν ούδε èyèvero πώποϋ ύπ αύτων (744. 48–51).

¹⁴ èyùi δε αίδοΰμαι μέν, ώ ΦίΚοστέφανε, όμοΧοεΧν Sri λίαν ήδομαι τοιαΰτα σοΟ φρονοΰντος μη τοϊς τετταράκοντα μάλλον ταλάντοις f¡ τ μη διεσφάλϋαι τν ύπέρ σον έλπίδων Ιήδόμεα] (747. 5–8). Hercher omits λίαν ήδομαι and keeps ήδόμε#α: I prefer the other solution to the textual difficulty.

¹⁵ At least according to Westermann’s emendation: the personae according to P are Αεώτης Άρχιεύς, Λύσανδρος Σκαμβωνίδης and Πρινάπης Πρασιεύς.

¹⁶ ούχ οίον at scripsi; ούχ (sic) οίονταιP, ούχ δτι Hercher, ούχ οίον Jackson (op. cit. [above n.5], 176).

¹⁷ ημαί και προεδρίαι scripsi; πμφ και προεδρείαι Ρ, τιμή και προε P, ρία Westermann, Hercher, Jackson (loc.cit).

¹⁸ τι δράσων ¥¡v δύνωμαι, αίσχύνομαι μεν Xéyeiv, άλλ' Sri βούλομαι πράττει» ήν δύνωμαι (749. 52–4).

¹⁹ άπόλλυςήμδς scripsi; άπολλυς (sine accentu) ήμων Ρ, εΊάλλη ήμΤν Caryophilus and Hercher, απόλαυε ημών Jackson (‘The Text of the Epistles of Themistocles’, Part 2,CQ 20 [1926),27).

²⁰ ούδεν μέντοι των Αδόκητων, εΐ και ημίν ποτέ hi yobv έμπνέουσι και την φυχηνϋχουσιδώσει θεός ύπομνήσαι ταύτης τής δημηγορας (750. 51–751. 1).

²¹ Could this be the κοινον δικαοτήριον των Ελλήνων mentioned in Letter 18 (757. 36 — cf. Diod. xi 55. 4–6, Plut Them. 23. 4)? If so, it has met and condemned T. in his absence. Certainly Letter 11 emphasizes the fact that prominent Hellenes from cities other than Athens joined in this new condemnation; and if both Letters do refer to the same event it forms another instance of complementary information being supplied in each sequence (cf. below pp. 101 f.). Letter 18 complains that Dorians were to have more seats on this tribunal than lonians (757. 37); the πολλούς άλλους of Letter 11 (751. 34) would therefore be Peloponnesians.

²² Cf. above n. 7.

²³ όίδα Sè δτι κατά σαντον ΰαυμάσψ τοις δέ (προς τό πλήθος λόηοις προσποίηση μη παρ êXnlSa σοι τούτον τον λάγον λελίχδαι (752. 13–15). I have supplied the words in brackets to fill the obvious lacuna, but the sense is in any case clear enough.

²⁴ δεοϊς evxov σωτηρίαν και νόστον ημΤνάπήμονα όπιο ω πέμψαι και τό τέλος των έγχεφουμένων μη μόνον ώς έλπΙΧομεν αλλά και ώς Αξιοι έομεν ούτως ημΧν έπιϋάσϋαι (749. 37–40). Lenardon’s ‘make my journey safe’ for νόστον ημΤν άπημονα πέμψαι (ST, p. 170) robs the expression of its undoubtedly intentional sinister implication.

²⁵ The role of friend and enemy in this last group serves to make its structure yet more complex and sophisticated. Letter 8 contains the announcement to T.’s friend Leagros, while 9 is a dissembling letter to his enemy Rallias; 11 is a dissembling letter to his friend Ameinias, while 12 contains the announcement to hisenemy` Aristeides.

²⁶ The disparaging comparison made between the ancient Briefroman and the eighteenth-century examples of the genre on the basis of relative psychological depth and subtlety by Sykoutris, J. (art. ‘Epistolographie’, RE Suppl. v [1931], 214) is thus not altogether justified.Google Scholar

²⁷ Doenges, op.cit. (above n. 2), p. 4: ‘The letters… are so arranged as to tell the story of Themistocles most effectively and dramatically’, and p. 22: ‘The letters were intended to do what they in fact do: tell the story of Themistocles’ banishment and flight to Persia.’ Needless to say I do not agree with this view, which would be more apt for Plutarch’s Life of Themistocles or (mutatis mutandis) a ghosted autobiography such as Robert Graves’ I Claudius than for the Letters.

²⁸ αλλα σχέτλιά ε καί. ταΰθ imo της φύγης ήνα-ίκάσμςϋα βίπεΐν,ώσπερ μανίας (753. 23–5).

²⁹ καί φνγάας μέν ημάς δύναιντ' άν ol κάκιστοι Αθηναίων noielv, κακούς Sè (753. 35–6).

³⁰ The letter opens with the striking expression απολβλάγησαι ùnep της τύχης, ώ Παυσανία. Pausanias has indeed ‘provided an apologia for Tyche’ because his fate shows her to be a moral rather than an amoral force.

³¹ For details see Plut. Them. 24 .1.

³² It is wrong to maintain as some have (e.g. Lenardon, op.cit. [above n. 3], 34) that the attitude to Aristeides of 18 and 19 is identical to that of 3. In 3, it is true, T. acknowledges (with surprise) the stand taken by Aristeides in the case — but it is just one of several occurrences alluded to in this brief letter and is certainly not there seen as the basis for a reconciliation. The relationship with A. gets steadily worse as the first sequence progresses. The reason for the inconsistency between the two sequences in this and other matters is outlined below pp. 99 ff. Likewise Doenges is wrong when he implies (op.cit. [above n. 2], p. 20) thatthe attitude to A. in 19 is hostile. The point of this letter is that Antagoras and Autolykos are misrepresenting the situation in Athens; what they should be doing is following the lead of his ‘έχθρός’ Aristeides (cf. και éneíodwe ημών τω έχνρω, 758. 6) instead of implying, falsely, that this influence is declining.

³³ Cf. Plut. Them. 30.6.

³⁴ This letter and its position have puzzled most commentators — cf. Lenardon, op.cit. (above n. 3), 35: ‘Epistle 21 as the conclusion of a literary exercise I find incomprehensible’; and Nylander, op.cit. (above n.3), 121: ‘It is wholly unsuited to round off the collection and rather looks like some misplaced footnote or appendix, inserted among the letters to give by its sheer triviality a feeling of authenticity and detailed knowledge.’ The items in question are 4 large silver kratïres, a number of golden censers inscribed with the old Assyrian script and a number of iron corslets (762. 16–20). Nylander takes it that this letter was written from Ephesos before the journey to Susa; it would on this view be classed as a financial letter like 6 and 7 (op.cit. 130). However, if 6 and 7 show anything, it is that T. was not short of cash, and there seems little reason for him at that stage in his travels to request burdensome items of property as well; moreover, Nylander’s interpretation would put Letter 21 out of chronological sequence. It seems much better to ascribe it to the Magnesia period. Precious vessels — especially ones inscribed with an archaic script — look more like material for furnishing a temple than a simple cash requirement; hence the interpretation offered here.

Temenidas is also instructed to avoid sending the consignment via Korinth or Kenchreai (762. 22–5); Nylander’s explanation of this, that it relates to T.’s role in an earlier conflict between Korinth and Kerkyra (Plut. Them. 24. 1), is undoubtedly correct: see Nylander, op.cit.131 and Lenardon, ST, pp. 127, 192.

³⁵ Lenardon (op.cit [above n. 3], 35) would deny that the letters display a uniformity of style: cf. however the not altogether unjustified remark of Jackson (op.cit. [above n. 5], 167) that they all manifest a ‘lumbering, contorted diction’ and the arguments of Doenges on the subject (op.cit. [above n. 2], pp. 13 f.). To my mind the style of all the letters is characterized by tortuous and elliptical forms of expression, an excessive love of antithesis, and a predilection for metaphor and paradox. The stylistic variations noted by Niessing (De Themistoclis Epistolis, Freiburg, 1929; his argument is summarized by Doenges, loc.cit.) are quite consistent with the unitarian view, if account is taken of different subject matter and the fact that some letters are more narrative, others more reflective. (So Doenges, op.cit. p. 13.)

³⁶ On the way in which individual letters complement each other, see Doenges, op. cit. (above n. 2), pp. 9–22; though Doenges does not subscribe to the ‘two sequence’ theory, some of his arguments are pertinent here and provide further examples in support of my contention. On his treatment of the Admetos discrepancy, however, cf. below n. 39.

³⁷ Author’s note to Balthazar (London, 1958, p. 7), the second novel in the Alexandria Quartet, apropos its relationship to Justine, the first.

³⁸ In fact the opening antithesis of Letter 1 brings out a similar point (741. 2–6): απερχόμενοι μεν είς Δελφούς καταίρειν διεηνώκειμεν, ώς εν Αελφοκ, έφ' δσον δοκεϊ ΆθηναΙοις, βιωσόμενοι καδ όδόν δε των έξ Αργους μοι ξένων έντνγχάνουσι Νικίας και Μελέαγρος, ôf τε ού προ πολλού Άοήνησι διατρίψας Εύκράτης (‘When I left I decided to go down to Delphi, with the intention of living in Delphi for as long as the Athenians’ decree remained in force; but while on the road I encountered some friends of mine from Argos — Nikias and Meleagros, and also Eukrates, who not long before had spent some time in Athens’.) T.’s decision (διεγυώκειμει») to go and live in the sanctuary and political neutrality of Delphi is overturned by this chance (έντιτγχάνουσι) meeting on the road, and instead he is carried off (àmryov ημάς 22) to Argos. Tyche thus intervenes (note ευτυχία 18, ευτύχημα 19) in this instance just as she later does when T. is forced to leave Argos; in neither case can his decisions release him from the clutches of his past.

³⁹ In spite of the valiant and highly ingenious attempt of Doenges to do so (op. cit. [above n. 2], pp. 14–18). It is impossible to take Í¡επιουσί) ημέρα of759. 1 in any sense other than that the Athenians and Lakedaimonians arrived the day after T. did; this creates an unresolvable factual conflict with the οκτώ f¡ έννέα ήμέραι of 745. 9. The point is that the two outlooks give rise to two completely different accounts of the same incident.

⁴⁰ The collection thus joins at least one other proven example of this genre, the Letters of Chion (cf. I. During, Chion of Heraclea: A Novel in Letters [Göteborg, 1951]). Doenges suggests (op. cit. [above n. 2], 23) that the letters of Phalaris and Hippokrates may also be worth looking at in this way.

⁴¹ As Niessing rightly saw (op. cit. [above n. 35], pp. 5–12), the Letters of Themistokles has no connection with the Stoic/Cynic περί φύγης literature. The second (‘good’) T. has far too narrow a conception of patriotism to satisfy Stoic cosmo politanism (cf. e.g. Musonius ap. Stob. 40. 9), while the bon mot of the historical T. on the subject of his exile recorded by Teles and Plutarch, άπωλόμεϋαμη άπωλόμεϋα (Teles Ρ• p. 22. 14 Hense, Plut. Them. 29. 7, Mor. 185 f and 602 a), is alien to the spirit of both sequences. Rather it is a novelistic product of the Second Sophistic era (which also saw the flowering of the Greek Romance) and displays that interest in the classical past so typical of the period (cf. Bowie, E.L., ‘Greeks and their Past in the Second Sophistic’, Past and Present 46 [1970), 3–41).CrossRefGoogle Scholar