Dionysius of Halicarnassus compares the two passages (D.H. Comp. 4: 18-19 Usener & Radermacher); but a feature of them which he does not focus on is that, appropriately for passages which introduce a new section of text, they are not linked to what precedes them by a connecting particle. Modern scholars have found a Homeric precedent for that, the sentence which begins 'There is a city Ephyre . . .' in Il. 6.152, and have suggested, or at any rate implied, that Thucydides on Epidamnus is deliberately echoing Homer on Ephyre.[2] Logically, however, there are three possibilities: that Thucydides was indeed deliberately echoing Homer; that he was not deliberately echoing Homer but, since he knew his Homer as all Greeks knew their Homer, he wrote as he did because a memory of Homer was at the back of his mind; or that this simple and not uncommon form of expression is just a natural way of beginning a new section of his text and cannot reasonably be regarded even as an unconscious echo of an earlier text. It seems to me that the third possibility is as likely as the second or the first.
In recent years deconstructionists have been eager to insist that we cannot recover either the intentions of an author or the reactions of the author's first readers, that we have only The Text and our own response to it. I do not want to champion that approach, to Thucydides or to any other author, but I do feel that students of Thucydides have sometimes been tempted too far in the opposite direction, and have tended to be so impressed by Thucydides that they assume that any effect which they can detect in his writing was an effect which he consciously intended. The examples I shall discuss have been remarked on by scholars for whom I have a high regard; the phenomena in question are certainly worth studying; my worries are only about the extent to which they were consciously intended by Thucydides.
Connor in his Thucydides does not argue for the unitarian view of the composition of the history, but he bases his approach on that view and says that he 'treat(s) the Separatist hypothesis as the last refuge of the philologists'.[3] Formally, I think, he might say that his interpretation is simply one possible reading of Thucydides; but he certainly assumes that the text as we have it is the text which Thucydides intended his readers to have, and he seems to me to imply that the effects which he can detect in the text are effects which Thucydides intentionally put there.
Connor suggests that the documents quoted in books IV-V are not a sign that in this part of the work the material is not yet fully digested (or, as Hornblower now suggests, a new technique with which Thucydides was experimenting at a late stage in his work),[4] but a way of emphasising the discrepancy between promises of stable relationships and the instability of the real world.[5] The second preface in book V, he thinks, is not the result of Thucydides' coming to realise that a war which he once thought had ended had not ended after all, but a deliberate device of Thucydides to destroy the illusion that the war was at an end. Similarly. the fragmentary nature of book VIII is not a sign that this part of the work is less finished, but a disintegrating world is deliberately represented in a disintegrating form of narrative.[6]
Connor may well regard it as liberating to set aside the problems about composition and simply read through the text from beginning to end as if that were how it was written. But one cannot solve a problem by refusing to face it, and it is not at all certain that the text was written as Connor assumes. I am among those who are convinced that the existence of a few identifiable early and late passages, of a few passages which ceased to be true, and of a few passages which are not wholly compatible with one another, makes it virtually certain that Thucydides' history was not finally revised and is not the unitary result of a single spell of concentrated work.[7] The contrast between what was promised and what was done at the end of the Archidamian War is indeed brought out effectively by Thucydides' quotation of the documents; and there is a kind of appropriateness in the disintegrating narrative of book VIII; but I am seriously afraid that by banishing separatism as the last refuge of the philologists Connor may have been led to see in the text intentions of Thucydides which never existed.
I now turn to part of an interpretation of book I which has been advanced by Badian. In 101.1-2 Thucydides states that the Thasians, when besieged by Athens in the 460s, appealed to the Spartans to distract the Athenians by invading Attica; the Spartans promised to do this, keeping their promise secret from the Athenians, and they intended to do it, but were prevented by the earthquake and the subsequent helot revolt. Given that the Athenians besieging Thasos were commanded by the pro-Spartan Cimon, and that the Spartans were soon afterwards to ask for Athenian help against the helots, it is likely that Sparta's unfulfilled promise, if not Thasos' appeal, is not authentic but was invented later, when Athens and Sparta had become enemies - and on that I agree with Badian. However, Badian argues that 'Thucydides himself is unlikely to have believed the story that he told', and that this is a 'clear example of what must be called disinformation'.[8] But this episode occurred probably before Thucydides was born, and I am not happy to assume that, if we can realise that a story is untrue, Thucydides must have realised it too: it seems more likely to me that he actually believed what his Spartan informants told him, as he believed some improbable things which his Spartan informants told him about the career of Pausanias.
Hornblower has recently taken an interest in narrative devices in Thucydides, a fascinating topic which I should like to consider at some length. One of the features which he has discussed is narrative dislocation, the mention of an item at a point in the narrative other than the obvious point.[9] Some instances Hornblower regards as innocent, but others have been regarded as calculated, and not only by Hornblower. The Peloponnesian invasions of Attica in the early years of the Archidamian War are mentioned separately as they occur or fail to occur, year by year. Athens' regular invasions of the Megarid are mentioned once for all at the time of the first, in 2.31; they are then not mentioned again until 4.66.1, and it is only there that we are told that they occurred twice a year.[10] I think the difference between the treatment of the invasions of Attica and that of the invasions of the Megarid is intended; it is probably to be connected with Thucydides' view that the war was a war about Athenian power, not a war about the particular grievances which preceded it (and especially not a war about Megara). But I am less happy with the suggestion that Thucydides has intentionally postponed the detail that the invasions of the Megarid occurred twice a year, that here he is cleverly using a technique of increasing precision. It had already been claimed by Wick that by means of this postponement Thucydides 'has easily and effectively insured that the information as a whole has minimum impact on the reader';[11] but I am not sure that invasions of the Megarid twice a year look much more serious than invasions once a year, and I am not sure that Thucydides deliberately kept the additional detail out of book II.
Another example in book IV is the delayed explanation of Brasidas' expedition to the north-east. In 70.1 Brasidas was able to go to support Megara against Athens because he 'happened to be in the region of Sicyon and Corinth, preparing for a campaign towards Thrace'. In 74.1 'Brasidas himself returned to Corinth and continued to prepare for the expedition towards Thrace, which had been his original objective'. The narrative of that campaign begins in 78.1, but after the bare introduction, 'Brasidas at the same time in the summer was journeying with one thousand seven hundred hoplites to the region towards Thrace', Thucydides launches directly into an account of his passage through Thessaly. It is only in 79.2, after Brasidas has reached Perdiccas of Macedon and the Chalcidians, that Thucydides explains that they had asked for a force from the Peloponnese. He then digresses, in 80-1, on Sparta's fear of the helots and consequent willingness to send some of them with Brasidas, and on Brasidas' enthusiasm for the campaign and the good impression which he created, before returning to the narrative.
Hornblower regards 70.1 (and, I assume, 74.1) as an innocent solution to the problem of arranging events in a linear sequence, but he thinks the withholding of the explanation for the expedition from 78.1 to 79.2 is an intentional device, 'to render Brasidas' arrival in Thrace highly dramatic'.[12] Certainly it is dramatically effective that our appetite is whetted and we are given a piece of vivid narrative before the explanation is supplied, but I am not sure that Hornblower's diagnosis is right. Gomme remarked that 'the narrative might almost be said to follow the order of appreciation of the events in Athens. . . . This order of narrative is not uncommon in Thucydides . . . It is a way, so common in Greek, of letting the narrative tell its own tale.' [13] With Thucydides we are not yet far from an oral culture in which cross-referencing is difficult and when possible is avoided. I should say not that Thucydides has deliberately postponed his explanation of Brasidas' campaign for dramatic effect, but that it seemed natural and economical to him to give this explanation, involving Perdiccas and the Chalcidians, at the point in the narrative where Brasidas reached Perdiccas and the Chalcidians.
In 78.1 Thucydides gives simply the total size of Brasidas' forces, 1,700 men; in 80.5 he says that 700 were helots. Hornblower suggests that this is a further instance of the technique of increasing precision.[14] I prefer to think that, in the manner of oral narrative, Thucydides has allowed one thing to lead to another: the Spartans were glad to accept the invitation and to send out some helots because they had been particularly afraid of the helots since their disaster at Pylos (80.1-2); then comes a digression on an (undated) occasion when the Spartans eliminated two thousand helots (80.3-4); after that Thucydides mentions that Brasidas' force included 700 helots (80.5); having thus returned to Brasidas, he adds that Brasidas was himself eager to go, and then digresses on the good impression which Brasidas created (81); after which he resumes the narrative of the expedition (82 sqq.).
The invasion of Attica by Plistoanax in 446 is mentioned in 1.114.2, 2.21.1 and 5.16.1-3. The fact that he was exiled after his withdrawal is mentioned in the second passage but not in the first; the fact that he was later recalled from exile appears in the third but not in the second. His return must have occurred in 427/6, but the only mention of him in book III is the statement in 26.2 that the invasion of Attica in 427 was commanded by his brother because his son was too young to take the command. Hornblower again sees the technique of increasing precision in use.[15] I should say that Plistoanax' return is mentioned in 5.16 because that is where it is most relevant. In 3.26 we have Thucydides in austere mood, formally mentioning the invasion of Attica in 427 as he does in other years, and not explaining either why the young Pausanias was king or why, uniquely in this year, the Agid house had to provide the commander for the invasion (perhaps the Eurypontid Archidamus was ill but not yet dead). It is characteristic of Thucydides in the Pentecontaetia that in 1.114 he gives the bare minimum of information, that Plistoanax went as far as the Thriasian Plain but then withdrew. 2.21 does not at first sight need the parenthesis on Plistoanax' exile, but there is perhaps a reason for its inclusion there, if Thucydides means to suggest that the Athenians not only remembered that the invasion of 446 had been cut short but also wondered if Pericles would again try to bribe the Spartans. I do not think we have to suppose that Thucydides kept these widely separated passages together in his mind, and deliberately released more information as he progressed from one passage to the next.
Thucydides' austerity in the Pentecontaetia involves one particularly startling instance of narrative dislocation. Notoriously, the fact that Sparta proposed to support Samos against Athens in 440, and Corinth prevented this, is mentioned not in his account of that episode (1.115.2-117) but in the Corinthian speech in Athens' debate on Corcyra and Corinth (1.40.5, cf. 41.2), and both Badian and Hornblower are among those who believe that the fact is omitted from the Samian narrative to make it less conspicuous and to avoid emphasising that Sparta was willing to go to war against Athens so soon after 446.[16] Even here it is possible that in his account of the episode Thucydides concentrated on the actual campaign, and was aware that he had already mentioned Sparta's proposal to intervene and thought it unnecessary to mention it again, but I grant that it is harder to be sceptical in this case than in others.
In one passage Hornblower claims to detect a deliberate preparation for something that is to follow later. In 4.3.2, giving a note on the situation of Pylos, Thucydides adds: 'It is called Coryphasium by the Spartans'. The site is called Coryphasium in the year's truce of 423/2 (4.118.4) and in the Peace of Nicias (5.18.7), and Hornblower suggests that the name Coryphasium is introduced in 4.3.2 to prepare the reader for its occurrence in the later documents.[17] But, as he acknowledges himself, it is entirely characteristic of Thucydides to give information of this kind with no ulterior purpose; there is no exact parallel to 4.3.2, but, for instance, in 4.76.3 and elsewhere we are given old and new names for the same place.
Kitto claimed to detect an anticipation of a more substantial kind.[18] 3.39.3-4, in Cleon's speech in the Mytilene debate, accuses the Mytilenaeans of embarking on their revolt from excessive confidence, and proceeds from that to a generalisation about the hybris of cities to which success has come very suddenly and unexpectedly - which was not the case with Mytilene. Kitto saw the fulfilment of the generalisation in Athens in the sequence of events from the success at Pylos to the defeat at Amphipolis (4.1-5.13), and claimed that that was intentional, that it was a part of Thucydides' technique to introduce the generalisation first and let the reader find its fulfilment later. I fully accept his main thesis, that Thucydides like other early Greek writers did tend to construct his narrative so that readers would find points in it rather than to spell out his points explicitly, and that in the sequence of events from Pylos to Amphipolis Thucydides does present the Athenians as 'reaching out for more' with disastrous consequences. However, it is not uncharacteristic of Thucydidean speeches to proceed from a specific case to a generalisation which does not precisely fit the specific case; this generalisation is a natural development from the accusation of excessive confidence made against the Mytilenaeans, and I do not think Thucydides need have been thinking of a specific instance of it when he included the generalisation in Cleon's speech.
Some of the other features which Hornblower discusses involve not a possible connection between different passages but single passages in which he detects 'rhetorical devices for producing an emotionally and intellectually satisfying interaction between narrator and narratee'.[19] Some of these involve presentation through negation. I think he is right to suggest that when, in 5.70, Thucydides remarks that the Spartans march to the sound of pipes 'not for religious reasons but to keep in step', he means 'not - as the reader might think - for religious reasons'. Again, however, it is possible to be too suspicious. In 4.125.1, in their second attack on Arrhabaeus the Lyncestian, during the night 'the Macedonians and the mass of the barbarians immediately took fright, as a large force is apt to be terror-stricken for no obvious reason'. Hornblower translates, 'were instantly seized with one of those unaccountable panics to which large armies are liable', and he takes this to imply a negation, 'not, as you might superstitiously think, due to the intervention of Pan'.[20] However, there is nothing in the Greek text to make the reader think of Pan, and I am sure Gomme was right to reject the suggestion:[21] Thucydides is simply making one of his comments on the irrationality of large crowds (cf, e.g., 4.28.3, on the moods of a crowd within a city).
We must remember that Thucydides and all ancient authors lacked, and were writing for readers who lacked, the easy ways of checking a passage elsewhere in their own text or in another text which are available to us today. We must not over-react: there are some cases where I am as happy as anybody to believe that he did write one passage with another in mind. For instance, I have no doubt that 1.20.3 on Spartan kings and on the Pitanates lochos is mtended to correct Herodotus, and I have little doubt that Thucydides consciously intended to say simliar things about Athens in the first Corinthian speech at Sparta, in 1.70, and in Pericles' last speech, in 2.64.3. But I do think we need to exercise a degree of realism and caution in our detection of intended effects. To end with a celebrated pair of passages, I am sure that Thucydides knew that he was comparing the Athenian empire with a tyranny both in Pericles' last speech (2.63.2) and in Cleon's speech in the debate on Mytilene (3.37.2); but I am not quite so sure that he deliberately, in order to make a contrast, included the qualifying 'like' a tyranny in Pericles' speech but omitted it from Cleon's.[22]
Many more passages could be mentioned, but I hope I have discussed enough to make my point. Proof is impossible in this area, but I do think there are many places where scholars have supposed that an effect is intentional but a more satisfactory explanation can be given without that supposition.
[1] A shorter version of this paper was written for the Second International Colloquium on Thucydides, organised by the deme of Halimous in September 1997, and is to be published in the proceedings of the colloquium. My thanks to the organisers of the colloquium, for their invitation and their hospitality, to Dr. S. Hornblower, for his generous reception of my scepticism, and to Dr. L. G. Mitchell, for helping me greatly to improve this paper. [Return to text]
[2] H. Lloyd-Jones, The Justice of Zeus (Sather Classical Lectures 41, U. of California P., 1971/1983) 223 n. 53; 5. Hornblower, Thucydides (London 1987/1994) 116; cf H. Strasburger, Studien zur alten Geschichte (Hildesheim 1982) chs. xv, xviii, xx. E. Fraenkel, De Media et Nova Comoedia Quaestiones Selectae (Diss. Goettingen 1912) 45-7, noted that this esti formula is common in Homer - sometimes without de but sometimes with it, e.g. Il. 11.711, 722 - and after Homer in later epic and in tragedy and new comedy; and M. L. West, The East Face of Helicon (Oxford U. P., 1997) 259, cites a couple of near-eastern instances (I thank Prof. J. S. Rusten for drawing my attention to the first and Prof. A. J. Woodman for drawing my attention to the second). [Return to text]
[3] W. R. Connor, Thucydides (Princeton U. P. 1984) 19. [Return to text]
[4] Homblower, A Commentary on Thucydides (Oxford U. P. 1991- ) 2.113-7. [Return to text]
[5] Connor, Thucydides, 144-7. [Return to text]
[6] Connor, Thucydides, 212-8. [Return to text]
[7] See especially the appendixes by A. Andrewes and K. J. Dover in A. W. Gomme et al., A Historical Commentary on Thucydides (Oxford 1945-81) v. [Return to text]
[8] E. Badian, From Plataea to Potidaea (Johns Hopkins U. P. 1993) 125-62 at 134-5. Gomme without stating his own opinion noted that the authenticity of the promise had been doubted, Homblower describes it as 'far from certain'. [Return to text]
[9] See in general Homblower (ed.), Greek Historiography (Oxford U. P. 1994) 139-48. [Return to text]
[10] See Greek Historiography, 145-6; Commentary, 2.230-1. According to Plut. Per. 30.3 the decree of Charinus added to the generals' oath of office an undertaking to invade Megarian territory twice a year. [Return to text]
[11] T. E. Wick, Hist. 28 (1979) 3. [Return to text]
[12] See Greek Historiography, 143 with n. 35; Commentary, 2.41-2, 238, 262-3. [Return to text]
[13] Hist. Comm. Thuc. 3.546. [Return to text]
[14] Commentary, 2.257. [Return to text]
[15] Commentary, 2.466. [Return to text]
[16] Gomme did not speculate about the reason for the displacement. For the argument that it is a device to make the fact less conspicuous by not mentioning it in the obvious place see Badian, From Plataea to Potidaea, esp. 141-2; Homblower, Commentary, 1.83-4; Greek Historiography, 144-5. Badian argues that the Thirty Years' Peace contained an autonomy clause which the Athenians were infringing, and that the Spartans would have been within their rights in going to war; G.E. M. de Ste Croix, who discusses this episode in The Origins of the Peloponnesian War (London: Duckworth, 1972) 200-3, believes that the Athenians were innocent and the Spartans would not have been within their rights. [Return to text]
[17] Commentary, 2.114, 154. [Return to text]
[18] H. D. F. Kitto, Poiesis: Structure and Thought (Sather Classical Lectures 36. U. of California P., 1966) 343-50. Gomme remarked that the generalisation of 3.39.4 does not apply to Mytilene, and explained it as I do; Homblower does not comment on it. [Return to text]
[19] Greek Historiography, 152-65: quotation from 152; 'as you might think':155-7. [Return to text]
[20] Commentary, 2.394. The suggestion was first made by W. Schmid, RM 1 (1895) 310-11. [Return to text]
[21] Hist. Comm. Thuc., 3.613. [Return to text]
[22] Gomme (Hist. Comm. Thuc., 2.175, 299) and Homblower (Commentary, 1.337) both are sure of this. [Return to text]
Back to HISTOS: Contents of the current volume
Back to HISTOS Main Menu
To Durham Classics Department Home Page