Perhaps the most obvious and fundamental characteristic of oral tradition is the importance of the group which preserves it. The old romantic belief in the accuracy or at least the symbolic significance of folk memory has been replaced by the realization that ‘accurate transmission is more likely if a tradition is not public property but forms the esoteric knowledge of a special group’ (Vansina, 31). Group memory is more accurate because it is more continuous and more cohesive than the general recollections of the past. Of course, in this context the question of accuracy must be distinguished from the question of truth. We are here only entitled to assert that the group memory ensures accuracy of transmission: it does not ensure truth, for a fiction or false story is just as capable of being transmitted accurately or inaccurately as a true story.
In contrast to many other societies, ancient and modern, the Greeks do not seem to have possessed a class of professional remembrances: once again their interest in the past was scarcely more than average. The occasional evidence of professional keepers of records, like the Pythioi at Sparta, the Kerykes perhaps at Athens, or the hereditary Cretan poinikastai who presumably had possessed a similar function and privileges before the introduction of ‘Phoenician’ writing to the city records,(1) serves only to demonstrate how limited and random such potential sources of tradition must have been. Jacoby’s refutation of Wilamowitz’s account of the origins of Greek and Attic historiography must stand, and the exegetai will never regain their former prominence.(2)
Herodotus recognizes the presence of such a professional tradition when he meets it; and he also recognizes in the same context the difference just mentioned between accuracy of transmission and truth. The Egyptians ‘who live in the cultivated parts practice mneme and are by far the logiotatoi that I have put to test’ (2. 77). But such a tradition has limitations: for he is anxious to distinguish that part of his account which is corroborated by Greek informants or depends on his own observation from that part for which he has relied on the Egyptian priests alone )2. 99, 142); and the consequence of the arrival of Greeks in Egypt is that from this point ‘we know all subsequent events accurately [atrekeos]’ (2. 154). The exact significance of this last assertion is not clear, but it refers to the fact that for the Saite period Herodotus could claim the agreement of Egyptian and Greek traditions, as well as his own opsis (2. 147).
This limitation to Herodotus’ respect for logioi andres should not obscure the fact that in general his work is explicitly founded on the testimony of such men. And though they do not normally constitute a professional class, one of whose chief duties is the preservation of tradition, the narrative of Herodotus shows that in each case they are chosen by him because they seem likely to possess an authoritative version of the past.
It is characteristic of Herodotus, and fortunate for us, that he at least appears to represent each tradition separately: he doe not seem to seek systematically to contaminate or to rationalize his sources. Instead, he gives one account from each place: when variants occur, they are normally derived from different localities. In this he approaches the ideal of the modern observer, who is expected to record each tradition separately. In principle we must assume that Herodotus wishes us to believe that each account is drawn from those whom he regards as logioi andres. The model is impeccable, however faulty the execution.(3)
The group memory is not only longer-lasting that folk tradition; it is also likely to be more limited and more liable to bias, for it reflects the interests of the group rather than those of the society as a whole. It often seems to be thought that this question in Herodotus and for early Greek tradition in general can be answered fairly simply by describing Greek oral tradition as generically ‘aristocratic’. Thus, for instance, Moses Finley asserts:
“In my judgment for the post-heroic period well into the fifth century, the survival of the sort of tradition I have been discussing must be credited largely to the noble families in the various communities, including royal families where they existed, and, what amounts to the same thing in a special variation, to the priests of such shrines as Delphi, Eleusis, and Delos.”(4)
And other scholars are fond of asserting in detail that the weaknesses of Herodotus’s account of particular episodes, e.g. Polycrates or Cleomenes, or Solon, are due to his reliance on an often undifferentiated ‘aristocratic tradition’. It seems to me on the contrary that the analysis of the structure of Herodotus’ logioi suggest strongly that, so far from his sources being as homogeneous as this account supposes, for different cities and different areas they have markedly different characteristics and interests. And more specifically it seems to me that the importance of an aristocratic tradition for the narrative of Herodotus has been mush exaggerated: with the somewhat surprising exception of Athenian history, there are very few of the typical signs of an aristocratic or family tradition in Herodotus.
As Vansina says, ‘every testimony and every tradition has a purpose and fulfils a function. It is because of this that they exist at all. For if a testimony had no purpose, and did not fulfil any function, it would be meaningless for anyone to pass it on, and no-one would pass it on’ (77). It is the investigation of the purpose of the logoi in Herodotus which reveals the milieu or group within which each of them was preserved and repeated, and the purpose reveals itself in the process  of selection and reorganization which the logos has undergone. In this discussion I would prefer to avoid using words which suggest deliberate intent to mislead or deceive; this may of course be present; but often the factors which have caused a particular tradition to take on a particular shape are not reasons of self-interest or conscious political distortions, but aesthetic or moral considerations. Words like bias, Tendenz, or prejudice have the wrong connotations; we need a more neutral word, covering both conscious and inconscious self-interested distortion and literary or aesthetic distortion, as they operate over time within a tradition. The world I would offer is ‘deformation’.
Conscious political deformation of course exists. One of the best non-aristocratic examples is the tradition of the Greeks in Egypt. A balanced account of their presence would have drawn on two sources: the merchants of Naucratis, and the descendants of the Greek and Carian mercenaries; the continued presence of the latter is known from Herodotus himself (2. 61: Carians at the festival of Isis slashing their faces with knives, thus proving that they are foreigners and not Egyptians), from Hellenistic evidence, and from archaeological finds of the Persian period.(5) But there is no sign that Herodotus met a mercenary, though he visited their deserted stratopeda: his information about their activities and their way of life is general and imprecise. His Greek sources for Egyptian history lay in Naucratis, and surely within a particular group in that town. Modern writers have commented on the peculiar nature of his account of Naucratis and the way it ignores the early history of the town. One passage seems to reveal why: Herodotus describes the largest temple, the Hellenion, and lists the nine city groups who control it. He continues, ‘the shrine belongs to these people, and these cities are the ones who appoint the prostatai tou emporiou: and any other cities that lay a claim to do so claim falsely’ (2. 178). The city groups thus excluded consist of three largest and oldest trading communities in Naucratis, those of Aegina, Samos, and Miletus, whose independent sanctuaries are shown by archaeological evidence to be earlier than the Hellenion and to ante-date the reign of Amasis. The history of Naucratis as told by Herodotus has been shaped by the claim of one political group, that centered on the Hellenion, to control the city magistracy: it is not surprising that such a tradition records nothing before the reign on Amasis, when this group seems first to have achieved separate status in the town.(6)
This is a tradition of a merchant class with political pretensions; it is scarcely aristocratic in any normal sense, if what is meant by aristocratic tradition is the persistence within particular important families of a set of traditions concerning members of the family. We might (as Vansina and others do) prefer to call them family traditions; but with the proviso that any such tradition which survives to impose itself on a wider public is likely to come from an important family. Such aristocratic or family traditions have particular characteristics. They concern primarily one family and its exploits; their purpose is through the justification and repetition of these exploits to enhance the present standing of the group. Their deformation tends therefore towards political apologia and exaggeration through biography; and they are essentially rationalistic, for they lack any religious or moral purpose. Unlike Finley, I think with most anthropologist that it is in fact useful to distinguish such aristocratic family traditions from a type of tradition in many ways similar, royal family traditions. For royal traditions concern the status not just of a particular family, but of an institution and often of the people as a whole. The Macedonian royal tradition of the activities of Alexander during Persian wars, and his claim not just to be philhellene but hellene in every respect, are perhaps so clearly represented in Herodotus because they concern not one family, but the Macedonian people as a whole.
The fact that the evidence for the existence of family tradition in Herodotus seems to be strongest in the case of democratic Athens may lead us to speculate on the special status of the Athenian aristocracy. The Alcmaeonid tradition in Herodotus is the obvious example, because we know of a number of episodes in which this version of events differed rightly or wrongly from that which seems to have been more generally current in Athens.(7) Another example is perhaps the influence of Philaid tradition on the account of the career of Miltiades. And the importance of family tradition in Athens can be used to explain certain gaps in Herodotus’ Athenian history. Thus the weakness of his account of the Pisistratid tyranny, in contrast to that found in Thucydides, the Aristotelian Constitution of the Athenians, and the Plutarch, is no doubt partly due to the disappearance of a Pisistratid family tradition, and to the deliberate silence of their allies the Alcmaeonids on this aspect of the past. Similarly the flight of Themistocles and the disappearance from Athens of any family tradition related to him are perhaps responsible for the peculiar character of the tradition about him, from which he emerges as a culture hero of a particular type,(8) associated with many different popular rather than aristocratic traditions, the Trickster, well represented in most cultures and exemplified in Greek heroic myth by Odysseus. (9)

[This paper was originally published in H. Sancisi-Weerdenburg and A. Kuhrt (eds.), Achamenid History, ii. The Greek Sources (Leiden, 1987), 93-115, and is reprinted with kind permission of the Nederlands Instituut voor het Nabije Oosten. Additions to the original version are put in square brackets.] The first draft of section I of this paper was written for a seminar given by myself and Professor Arnaldo Momigliano in Oxford in Hilary Term 1977; it was later discussed with anthropologists and classical scholars on a number of occasions, before being presented at the Groningen workshop. As it represents the theoretical underpinning of my Early Greece (Brighton, 1980) (see briefly pp. 27-32=22-8 of the 2nd edn., 1993), it was perhaps time it was published. Section II was written in the light of the Groningen discussions. Thanks are due to my colleagues there, and to David Asheri and Heleen Sancici-Weedenburg, who made valuable comments on the paper in its later stages.
1. L. H. Jeffrey and A. Morpurgo-Davies, ‘∏oωικαvrάѕ and ποιvικάζєιv: BM 1969 4-2, 1, a New Archaic Inscription from Crete’, Kadmos, 9 (1970), 118-54; compare the remarks of Evans, Herodotus [n. 4], 149-50, on mnamones and hiaromnamones.
2. Atthis (Oxford, 1949).
3. To postulate deliberate and wholesale description (with Fehling, Herodotus and his ‘Sources’ [n. 1], rather than faulty execution, requires an answer to the question, ‘Who invented the model which Herodotus is thought to have abused?’ It implies a proto-Herodotus before Herodotus.
4. ‘Myth, Memory and History’ [n. 9], 27.
5. Simon Hornblower pointed out to me the significance of the Herodotus passage, other evidence in M. M. Austin, Greece and Egypt in the Archaic Age (PCPhS suppl. 2; Cambridge, 1970), 18-19; T. F. R. G. Braun, CAH iii/3 (Cambridge, 1982), 43-8.
6. See my Early Greece [p. 16 pref. note], 215-17 [-2nd edn. (1993), 228-31].
7. [See now Rosalind Thomas, Oral Tradition and Written Record in Classical Athens (Cambridge, 1989), ch. 5; but I remain unconvinced by her arguments that an ‘Alcmaeonid tradition’ is not dominant in Herodotus’ description of a number of key episodes in Athenian history.]
8. A different type of tradition was available to Plutarch in his Life of Themistocles, drawing on local historians in Magnesia: D. Asheri, Fra Ellenismo e Iranismo (Bologna, 1983), 52-3.
9. E. E. Evans-Pritchard, The Zande Trickster (Oxford, 1997); M. Detienne and J-P. Vernant, Les Ruses de l’intelligence (Paris, 1974; English trans. 1978).



Your Name Email Comment Type this number