Part 5
The earlier analysis of Greek traditions will have made some points clear. First, the investigation should not start from the historical reliability of the traditions available to Herodotus, let alone from the truth or falsehood of single statements or episodes; these are secondary questions, which can only be considered after the types of tradition have been established. This is of course a fundamental principle of all forms of source criticism, not one peculiar to oral traditions, though it tends to be forgotten more often in the oral context. Second, on the model proposed above, we should think especially in terms of the preservation of tradition and of channels of information: what types of logioi andres were available and recognizable to the more or less conscientious Greek enquirer?
I begin with a negative proposition: it is important to remember what was not available to or not used by Herodotus. There is no sign that Herodotus had access to a priestly tradition, or written. It is not possible to analyze the Persian logoi in the way that has become accepted for his account of Egypt;(1) Herodotus’ lack of understanding of Persian religion and superficial account of the Magi are enough to demonstrate that he had no direct contact with a Persian priestly class who may well have possessed an oral tradition of some interest.(2) One type of tradition is thus ruled out for Persia as for Mesopotamia.

The question of Mesopotamian traditions raises a wider question about Herodotus’ contacts: the most important groups of logioi andres in the Near East belonged essentially to a literate culture, some at least of whose main literary forms are known to us through written records. One of the most obvious characteristics of Herodotus’ accounts of eastern societies is that they show no sign of any influence from the known literary or historical genres preserved in writing, such as royal inscriptions, priestly chronicles, law codes, or sacred texts: in this he contrasts very strongly with the Jewish historical tradition both before and after the exile. This suggests, not only that Herodotus’ historical methods and literary techniques are independent of eastern written traditions, but also that he did not even have extensive access to the guardians of those traditions as oral witnesses; for their modes of thought would surely have been marked by the influence of their status and their skills as a literate caste. Herodotus’ accounts of eastern events are not patterned in the same way as his account of Egyptian history, by the influence (however mediated) of a priesthood whose skills as storytellers reflect their activities as guardians of a written tradition.

We must admit one significant exception. There is no doubt that documentary models lie behind three of the most famous Persian passages in Herodotus, the satrapy list (3. 89-97), the description of the Persian royal road (5. 52-3), and the Persian army and navy lists (7. 61-98). That is not, of course, to say that these passages rest on the documents: the notion of an army list left behind is even less plausible than the theory that Herodotus lifted this or that entire passage from Hecataeus of Miletus. These are not documents either in our modern sense or even in the contemporary Jewish sense. They are lists created under the influence on documentary models. Literacy, as Jack Goody has demonstrated, encourages certain mental forms, the most common of which, the table and the list, belong especially to bureaucratic practices.(3) In the case of the two main passages of Herodotus, the problems involved in detailed analysis of the information, and the uncertainty about a possible date or function for the alleged underlying ‘documents’, suggest that we should emphasize the aspects of orality and written model. But however that may be, clearly involved in their transmission or their creation is a documentary mentality which is not usual to Herodotus. It is this phenomenon of documentary orality which attracts me to the hypothesis of David Lewis, that one source for Herodotus’ information on Persia was the Greek element in the Persian imperial bureaucracy.(4) Belonging to at least the fringes of a highly specialized literate culture, in their organization of material they would naturally follow the scribal mental forms of the table and the list: asked for information, they would reply, not with a logos, but with an ordered ‘documentary form’.

To consider Momigliano’s comparison, this is one step short of the Jewish historian’s practice of actually quoting ‘documents’, since in that case the historian himself takes on elements from the scribal culture, but it can involve much the same potential danger of misleading us by suggesting the existence of an independent document behind what is in fact a form more or less consciously created or manipulated by the historian; yet both traditions rest on an acceptance of scribal practice and the scribal mentality. The attractions of this hypothesis as a way forward are obvious: it enables us to relate our two main bodies of evidence, the Persian documentary archives, both those surviving at Persepolis and those to be supposed elsewhere, and the Greek literary tradition; and it postulates a type of tradition which is likely to possess a relatively high level of detailed factual accuracy.

This hypothesis serves to highlight a quite different type of patterning in Herodotus’ Persian account, which, if it is related to less reliable types of information, is nevertheless more dominant. The main Persian narrative of Herodotus is organized in two great blocks. The first gives a description of the fall of the Median empire and the rise of Cyrus, centered on the figure of Harpagus the Mede (1. 73-4, 95-130). The account uses a number of stories of different origins, most notably the narrative of the birth and upbringing of Cyrus, which is a Mesopotamian foundation legend going back to Sumerian times, adapted to become part of the official Achaemenid dynastic myth.(5) But despite its use of disparate elements, the narrative possesses a unity and a number of recurrent explanatory motifs (such as the eating of human flesh-1. 73, 119.),(6) which suggest a single non-Greek reworking of more varied traditions; and, given the Median slant to the story, it is likely enough that its basic form represents a Median aristocratic version of events. The further theory that it came to Herodotus from the family tradition of Harpagus himself is less likely, given the way he is characterized (if only in a speech) at the end of the story, as ‘at once the silliest and the most unjust of men: the silliest, if when it was in his power to put the crown on his own head … he had placed it on the head of another; the most unjust, if on account of that supper he had brought slavery on the Medes’ (1. 129).(7)

The second great block of Persian narrative describes the episode of the Magian usurpation and the revolution by which Darius came to power, again from a distinctive viewpoint (3. 30, 61-88). The official version of these events was of course at least potentially widely available in the Persian empire, since Darius had ordered it to be circulated and published in the various languages of the empire (though it may be doubtful whether these would have included Greek).(8) But while Herodotus’ account corresponds closely with this version, it is not derived from it: it is rather a telling or retelling of the alleged events from the point of view of the small group of Persian conspirators who included Darius not as a leader but merely as one of their number. Here the combination of a close relationship to the official royal version promulgated by Darius with the non-royal viewpoint makes it very likely that we are dealing with an account derived from oral tradition within one of the great families involved; and J. Wells long ago identified the most likely source for this, as for the account of the siege of Babylon (3. 153-60), in the family traditions of Zopyrus, great-grandson of the conspirator, who deserted to Athens in the lifetime of Herodotus.(9)

These generally-accepted conclusions establish two blocks of historical narrative, one Median, the other more strictly Persian, which are perhaps as close as we are ever likely to get to what might be called a Persian historiography. It is worth therefore considering their characteristics and limitations.
In both cases the narrative is concerned with high politics and events that shaped world history; in both cases it is closely related to an official royal version of those events. But despite that relationship, in both cases we are offered not the official version itself, but a variant of it, related to the interests of a more or less precisely identifiable non-royal ruling group. Thus Herodotus had access, not to an official royal version of Persian history, but to variants of it current in the high aristocracy: paradoxically it was always easier for Greeks to make contact with the ruling classes in the Persian empire than with the imperial bureaucracy. Here, then, were men well qualified to stand among the normal types of Herodotus’ logioi andres.

The accounts that these groups could offer fall short of being historical in important respects. First, they seem to be episodic, rather than continuous or biographical.(10) We are not offered a coherent narrative or biography of any eastern king; rather Herodotus relates within a regnal framework a series of isolated but detailed stories. Second, the narrative itself and the elements of which it is composed seem to be fundamentally oral in form: it is patterned as a succession of stories independent of each other and often without obvious connections; the resonances and repetitions give the impression of being folk-tale motifs, traditionally accepted devices to explain motivation or actions. This is what we would expect from an aristocratic society which, for all its use of a literate bureaucracy, remained fundamentally illiterate.

Two points may make us hesitate. First, whence the regnal framework, which covers in formulaic phrases both the Median and the Persian royal houses: ‘having reigned three and fifty years Deioces was at his death succeeded by his son Phraortes’ (1. 102); ‘Cyrus himself fell after reigning nine and twenty years’ (1. 214)? But since this characteristic formula is also used by Herodotus in relation to Lydian and Egyptian kings, it is scarcely possible to claim it as a sign of the influence of Mesopotamian royal chronicle; it may be borrowed from these other cultures but it is anyway independent of the main Median-Persian narrative, with which it does not entirely fit.(11) Whatever its origin, it should not, I think, mislead us into claiming the existence of a continuous Persian account of each king, either biographical or in chronicle form.

1. See most systematically A. B. Lloyd, Herodotus Book II, esp. Introduction (Leiden, 1975).
2. See most explicitly the claim of Strabo about the Magi, 15.3.18. If such a tradition existed, it could of course have influenced indirectly Herodotus’ logoi: for this possibility see esp. P. R. Helm, ‘Herodotus’ Medikos Logos and Median History’, Iran, 19 (1981), 85-90. I am not, however, clear, when Helm talks of ‘Iranian popular saga’ and ‘independent heroic sagas’ as source for Median and Persian history, whether he is seeking to revive the theory of A. Christensen of the existence of fixed texts is the form of heroic poetry, or whether he is merely postulating free prose tales.
3. The Domestication of the Savage Mind (Cambridge, 1977), esp. chs. 4 and 5. O. K. Armyor, ‘Herodotus’ Catalogues of the Persian Empire in the Light of the Monuments and the Greek Literary Tradition’, TAPhA 108 (1978), 1-9, criticizes the passages as if they were documents, and inevitably finds them unsatisfactory; better Briant, ‘Sources grecques et histoire achemenide [n. 42], 495-500.
4. See ‘Persians in Herodotus’ in M. H. Jameson (ed.) The Greek Historians: Literature and History. Papers presented to A. E. Raubitscheck (Saratoga, 1985), 101-17, and D. M. Lewis, ‘The King’s Dinner (Polyaenus IV 3.32)’, in Sancisi-Weerdenburg and Kuhrt (eds.), Achamenid History [p. 16, pref. note], ii, 79-87; also M. A. Dandamayev, ‘Herodotus’ Information on Persia and the Latest Discoveries of Cuneiform Texts’, Storia della storiografia, 7 (1985), 92-9.
5. We are fortunate in knowing something about both myth and ritual: see A. Alfoldi, ‘Konigsweihe und Mannerbund den Achameniden’, Schweizerisches Archiv fur Volkskunde, 47 (1951), 11-16; G. Binder, Die Aussetzung des Konigskindes (Meisenheim a. G., 1964), with my review, CR, NS 17 (1967), 329-93. R. Drews, ‘Sargon, Cyrus and Mesopotamian Folk History’, JNES 33 (1974), 387-93, has some interesting observations on the version of the Cyrus legend derived from Ctesias, which suggest that it is closer to the Sargon story, and therefore perhaps a ‘Mesopotamian’ version rather than a Persian one.
6. Not in itself of course unknown to the Greeks, but treated by them rather differently; compare Thyestes. For this theme see W. Burkert, Homo Necans (Berlin, 1972), 108-25; M. Detienne, Dionysos mis a mort (Paris, 1977), ch. 3. In connection with p. 43, I note that the motif is transferred to Lydia by Xanthus, FGrHist 765 F. 18.
7. The Median origin is generally accepted, e.g. J. M. Cook, Cambridge History of Iran, ii (Cambridge, 1985), 203-4; the family tradition of Harpagus is an idea that goes back to the 19th cent: see the references in J. V. Prasek, ‘Hekataios als Herodotus Quelle zur Geschichte Vorderasiens’, Kilo, 4 (1904), 199-200. But ‘there must have been some Greek reworking of the story. The H in Harpagus seems to refer to popular etymology and can only have been attached to the Iranian name Arbaka in Greek surroundings; cf. R. Schmitt, ZDMG 117 (1967), 113 n. 103; M. Mayrhofer, Onomastica Persepolitana (Vienna, 1973), 154’ (letter from H. Sancisi-Weerdenburg).
8. Behistun inscription col. 4 paras. 60-1 and 70 in R. G. Kent, Old Persian, 2nd edn. (New Haven, 1953), 131-2.
9. J. Wells, ‘The Persian Friends of Herodotus’, in id., Studies in Herodotus (Oxford, 1923), 95-111. I agree with Lewis, ‘Post-script 1984’, in A. R. Burn, Persia and the Greeks (London, 1962, znd edn. 1984), 105-6, that Zopyrus is not likely to have been a source both for this (often tendentious and unreliable) narrative and for the more ‘documentary’ elements discussed earlier.
10. It is for this reason that I do not discuss the question of biography raised by Momigliano (above, p. 35).
11. H. Strasburger, ‘Herodotus Zeitrechnung’, in Marg (ed.), Herodot [n. 13], 688-736; cf. R. Drews, ‘The Fall of Astyages and Herodotus’ Chronology’, Historia, 18 (1969), 1-11.



Your Name Email Comment Type this number