F.W. Walbank

Review of John Marincola Authority and Tradition in Ancient Historiography

Cambridge: University Press, 1997. Pp.xvi+361. ISBN 0 521 48019 1. £45.

1.

In an appraisal of Arnaldo Momigliano's Contributi, published in his collection of essays The Use and Abuse of History,[1] Moses Finley singled out a recurrent theme: 'Why do historians write histories in the way they do?' He then went on to observe that in trying to answer this question the reader can expect little help from ancient manuals -- not to mention the 'massive silence' of Aristotle on the subject. It has sometimes, but wrongly, been argued that a remark put into the mouth of M.Antonius in Cicero's De oratore (ii.62) implies that there were no Greek treatises on the theory of history; but what Cicero is saying here is simply that in rhetorical handbooks history is nowhere treated separately from the general precepts of the ars ('neque eam reperio usquam separatim instructam rhetorum praeceptis'). There were certainly Greek treatises Peri historias - those, for example, of Theophrastus, Praxiphanes and later Caecilius of Caleacte and Theodorus of Gadara. But all are lost and their contents unknown. True, we possess the short, somewhat superficial work of Lucian, How must one write history?, which H.Homeyer[2] classified as a sort of diatribe; but, though it hardly deserves Finley's harsh condemnation as 'a shallow and essentially worthless pot-boiler'[3], it is of limited use as a serious contribution to the subject.

'It is a commonplace', continues Finley in the essay just mentioned, 'that every historian's notion (conscious or subconscious) of his function is based on both the social and political situation in his own world and the literary and moral tradition he has inherited.' The present book by John Marincola (henceforth M.) is concerned primarily with the second half of this proposition and in particular with the literary tradition (though both the moral tradition and - especially where discussion centres on the difference between writing in a free and in a closed society - the social and political situation also find a place). M.'s subject is 'authority and tradition' as they affect the historian. By 'authority' he means the rhetorical and other compositional devices which the historian uses to establish his competence in his field and by 'tradition' all those precepts, exempla and modes of operation handed down from his predecessors which determine what he himself is likely to adduce or employ to support that claim. The book is therefore largely concerned with 'validation', which broadens out however into a learned and sensitive discussion of the variety of ways in which the historian relates to his subject-matter, his predecessors and his public and how he contrives to innovate, while still maintaining the criteria established by earlier writers. It draws on (but also looks beyond) the individual authors in order to identify the general aims of Greek and Roman historians (they were not always identical) and their assumptions about what constituted a satisfactory history. It should be noted that M.'s argument rests on the assumption (which some would challenge[4]) that by and large historians were familiar with their predecessors and wrote with them in mind. I think he establishes this, at any rate for what we may call the 'major figures'. As for the many who are merely names to us, one can usually only speculate about their connections one with another.

One will look in vain here for a comprehensive study of any one writer. Ephorus'.writing methods and treatment of particular technical problems, for instance, are spread out over all the chapters. But it is precisely in this that the merit of M.'s method lies, since it lets us identify similar overall problems and the successive attempts by historians to solve these within a developing historiographical context. The separate trees are not allowed to obscure our vision of the wood. In the following discussion I shall stick to this pattern and comment successively on the topics raised in the separate chapters.

In his introduction M. sets out his general programme and touches on several issues relevant to the main theme - to what extent ancient historians felt themselves to be that and no other kind of writer, the various audiences addressed by particular historians (increasingly, M. argues, the interested observer rather than the politically or militarily active public figure) and the changed circumstances which arose under Roman domination. Already under the republic, if we can believe a remarkable assertion by Cicero (de fin.v.52), there was an interest in historia (books or just the past generally?) among homines infima fortuna, including artisans. But if this was so, we have no inkling how far historians catered for it, unless it is reflected in the growing popularity of compendiums and a greater attention to moral issues. As time went on there was also a widening of the area from which Roman historians were drawn, though indeed from the days of the middle republic they tended to come from outside the capital. An example is M. Cato from Tusculum - whom M. (p.30) ought not to have referred to as a provincial.

2.

The central chapters of the book deal with various themes which come up as the historian sets out to justify his work. In each case the evidence is derived from a detailed examination of the main Greek and Roman historians, taken chronologically. Chapter one asks what drew the historian to history and what specific aspects of the past aroused his interest. M. shows how, once his subject has been chosen, the historian tends to magnify it (by auxesis = amplificatio) to make it redound to his credit. More than one historian (here following in the footsteps of the poets) seems to have been led to his topic by a dream (as when, for instance, a vision of Drusus Caesar bade the Elder Pliny write on the German Wars). Once adopted, the subject was lauded for its magnitude, its uniqueness or its importance - though Livy, in this as in so much else the odd man out, claims merely that his work is a kind of escapist self-amusement (praef.5), which is clearly not the full story. Appian (praef.45-52), rather exceptionally, emphasises not his subject, but rather his new approach to it, and so does not quite fit into M.'s picture here.

A device, first found in the Hellenistic period, when it is commoner in local than in general history, is the 'dedication'. A dedication can draw on the prestige already attached to a distinguished dedicatee, but at the same time (and this especially in the writing of contemporary history) risks attracting a charge of bias. For that reason dedications are more often to be found associated with literary and antiquarian works than with straight history. M. is anxious to establish that this pattern holds good for Rome too and to restrict the use of dedications there to autobiographies, memoirs and scholarly or antiquarian books. There appear, however, to have been dedications in Coelius, Claudius Quadrigarius and Velleius. These M. explains away with separate, not implausible, hypotheses. But an equally valid explanation may be that at Rome the distinction between antiquarian and historical works was in this context less rigid than M. would allow and that where (as his pp.55-60 show) the available information is quite flimsy, one should be cautious about assuming rules.

Historians were led to write history in various ways. But what did they hope to get out of it? The answer, already formulated by Polybius (xvi.14.3) in connection with the Rhodian historians Zenon and Antisthenes, is 'glory and renown'. It is through the historian that the statesman and general acquire glory (vixere fortes ante Agamemnona!) and part of this glory rubs off on the writer. In Rome this creates a problem, for there gloria has a special connotation as something won in public service. Sallust had to argue that it could also be within the grasp of historians. In this context M. makes the interesting point that the historian usually discusses the question of glory in general terms, not specifically claiming it for himself, but displaying a reticence, which is indeed wholly understandable.

3.

Chapter two discusses the means historians adopted to achieve their ends; it falls into two halves. In the first M. deals, especially in relation to contemporary history, with the historian's treatment of evidence based wholly on autopsy ('the eyes') and that coming from the accounts of others ('the ears'). This is a distinction familiar in the philosophers, begining with Thales, but the first historian to enunciate it clearly was Herodotus, with his contrast between opsis and akoe. Non-contemporary history here constitutes a special case, since for this only 'the ears' are available and, as Ephorus observed (FGH 70 F 9),the nature of that evidence is such that, whereas a detailed account of a contemporary situation is felt to command belief, for the past the reverse is true.

Any narrative carries a special significance if the historian or his informant has access to the great and powerful; and under autocratic regimes this factor grows in importance - as does the problem of how to validate reports, which is inherent in that situation (cf.Dio liii.19). For this category of information M. adopts the convenient phrase 'privileged access' and he emphasises the frequency with which it was exploited. He is, however, perhaps too ready to identify examples of it where the evidence is short of conclusive. On pp.87-8 I noted seven instance of phrases such as 'we ought not to doubt', 'it is not to be doubted', 'nor is it unlikely' (here in relation to Alexander-historians), 'the trend may well have continued under the diadochi'; and, when we come to imperial Rome, though we are without the works of early historians, here too 'it can hardly be doubted' that they draw validation from proximity to the ruler. Well, perhaps; but it is safer to restrict one's examples to what is attested rather than accumulate cases that are merely probable. When we come to the actual details of history written in closed societies, however, M. sketches a variety of devices adopted by historians to secure validation for their narratives, of which 'privileged access' was only one. Alternatives were to limit one's account to events with which one was personally concerned or, in imperial Rome, to write 'senatorial history', thereby avoiding dangerous areas of enquiry.

A different problem arose for the authors of non-contemporary history. For remote happenings what evidence was actually available (an acute problem for historians of ancient Rome)? And, as a corollary, how was the historian to present (or evade) this difficulty? As regards the first question, W.K.Pritchett in his Dionysius of Halicarnassus on Thucydides (1975) 54 n.20 has listed a varietyof documents - archon-lists, leukomata in archives, deme-records, axones, pedigree-lists and Craterus' collection of inscriptions - available at Athens. The existence of these must certainly be taken into account; nevertheless, along with Jacoby and now M., I still believe[5] that taken together these can have given only the bare bones of a historical record. The same problem existed of course at Rome. To the problem of how to discuss this hiatus in knowledge different historians found different answers. Some simply quoted different variants in the tradition without deciding between them. Others referred to some source assumed to carry conviction, such as the priests in Egypt. Herodotus employs both devices. Alternatively one might emphasise that one was oneself a priest, like Manetho, 'a certain Ptolemy' (FGH 611), Chairemon (FGH 618) or Berossus. M. makes a novel and not unconvincing point, when he adduces their perceived prestigious occupation as a validation of these men as reliable historical authorities.

At this point M. raises the relevant question: why did historians feel the need to attempt a new version of the past - and how did they justify their attempt? Sometimes by saying it had not been done before; or that earlier versions were incomplete; or that previous historians were biased; or that one's own version was more accurate. Syme would make an addition[6]: 'the justification for a new history, so Livy avers, is greater accuracy or a finer style'. But M. is, I believe, right in querying this, for Livy (praef.2) only says that historians believe they will achieve the one or the other. His antithesis is merely descriptive, neither a rule nor a prescription. In fact, M. claims to have found no example of a historian alleging 'a finer style' as a justification for writing history.

M. concludes that there was no single recognised methodology for dealing with early history - which was, moreover, both in Greece and to a lesser extent at Rome, complicated by the problem of how to deal with myth. Here again historians adopted a variety of approaches - reporting without comment, exclusion, rationalisation or the mere juxtaposition of 'mythical' and 'historical' versions. Fortunately this is a field in which one can observe both historians' practice and such rhetorical theory as existed about it in writers like Asclepiades.[7]

4.

In chapter three M. discusses the use of the historian's character to validate a narrative; and, for Roman historians and their sources, this necessarily includes dignitas and social status. The term 'character' is in fact here given a wide reference, since, besides its obvious .meaning, M. takes it to involve experience, effort and fair-mindedness. Here he pin-points a distinction between Greek and Roman historians. The former might quote experience (of various kinds) and established rank as evidence for their status as researchers, i.e.their claim to accuracy (as we might say) as scholars. No Greek historian before the period of the Roman empire quotes social status as, in itself, a qualification for an author. That is a Roman tradition and for it M. assigns a crucial role to Cato as the historian who alleged auctoritas as an important (though not of course the only) validation of a writer's competence. Such a claim, he asserts, would have been incomprehensible to his Greek contemporaries. That is indeed possible. But here, as elsewhere, M. seems to be treating a possibility as a virtual certainty. In fact his case rests on an argumentum ex silentio and is not water-tight.

The evidence for this is Polyb.iii.9.1. Here the Greek historian, writing primarily for a Greek public, asserts that the fact that Fabius Pictor was a senator (as well as a contemporary of the Hannibalic War) has led some people to regard him as wholly trustworthy.[8] Clearly that statement cannot have been meaningless to Polybius's readers. Hence there is no reason to assume that Fabius, though also writing in Greek and for a primarily Greek public, could not himself have referred to his rank as a reason for confidence in his reliability. He was a Roman and his work had also a pronounced Roman character.[9] One cannot, therefore, I suggest, rule out at least the possibility that Cato may have been anticipated in his claim to validation through status. One should not perhaps be too sure about what was, or was not, incomprehensible to Greeks.

In any case, however, the situation changed under the empire, when Greek historians frequently appeal to their dignitas as a means of self-validation. Arrian (i.12.5) discusses this in his 'second preface'; it is, he says, a Roman convention, which he chooses to regard as irrelevant to his own work. Appian (praef.15.62), by contrast, quotes his offices proudly. It is generally assumed that the two passages have some connection. Usually Arrian is dated first, which would mean that he cannot here be replying to Appian.[10] But neither here nor in an earlier article[11] does M. come out clearly on this question of priority. On p.146 he argues convincingly that Appian is using his offices as an indication of social status, but then speaks ambiguously of Arrian's 'echoes withAppian' - leaving open who was echoing whom. It seems clear that Arrian was attacking the Roman convention - but not necessarily in the person of Appian!

Other aspects of what M. refers to as 'character' are effort and impartiality. The former can include both the preliminary investigations (research, travel etc.) and also the actual burden of writing; later historians, indeed, took to listing the number of years so spent. Writing was burdensome not only because of the time it took, but also through the need to achieve the appropriate level of vividness. And here I have some difficulty with M.'s translation of Sallust, Cat.3.2 on p.152: 'inprimis arduum uidetur res gestas scribere: primum, quod facta dictis exaequanda sunt.' This M. translates: 'because the deeds must be made equal with the words.' Ernout has: 'son récit doit être à la hauteur des faits', which seems to imply: 'the deeds must be equalled by the description of them'. This I take to be correct, but the opposite of M.'s version. I would translate: 'the deeds must be equalled by the words', i.e. the words must match the deeds.

On 'impartiality' M. has some interesting and original comments. For ancient historians, he observes, the opposite of 'true' is 'biased'; and bias is seen as specifically occasioned by favours or injustices (past or anticipated). Even patriotic bias is linked with what one's patria gives one. The first recorded claim to impartiality occurs in Polybius (xii.5.1-5, discussing Epizephyrian Locri); an example appears later in Sallust's Catiline and the theme becomes increasingly common as an accompaniment to the Roman claim to validity from dignitas (since such a claim implies the danger of bias). M. finds no evidence for a claim to impartiality in earlier historians such as Ctesias or Theopompus; but here again we should remember the dangers of the argumentum ex silentio.

5.

Chapter four discusses a special problem, that of the historian as himself a participant in the events he is describing. Obviously this can only arise in contemporary history. How should the historian present himself so as to avoid reprehensio (the danger of which is so apparent to Cicero in his letter to Lucceius)? M. observes that the question of whether one should use the first or third person is not a very important issue. Nevertheless he feels it necessaryto discuss it at some length and he notes, significantly, that Thucydides uses the first person for anything affecting the historian's work (and research done for it) and the third person for formal openings or when he appears as a participating character.

The case of Polybius is interesting and here M. has a new point to make. Up to book xxxvi Polybius follows the Thucydidean norm (third person as a participant, first person when commenting as a historian). But at xxxvi.11-12 there is a change. In the middle of this passage, which describes how Polybius was summmoned to Lilybaeum by the consul, he suddenly switches to the first person; and in xxxvi.12 he comments on his procedure. M. points out (as Polybius does not) that this is an innovation and he links it convincingly with the more personal role (and style) of the author in books xxxv-xxxix (xl is a kind of index). This fits in very well with my own views on Polybius' last books,[12] but whereas I discussed the change in these books partly in terms of Polybius' own situation and partly as the solution of the problem how to incorporate material from 168 to 146 within a universal history, M. discusses it as a change from a history-style to a memoirs-style work.There is no contradiction between the two approaches; on the contrary, M. has made a valuable additional point, for Polybius may well have thought about the composition of his additional books in those terms. In making the change, was he, M. asks, thinking ofAratus' Hypomnemata or of Roman Commentarii? The latter seems to me unlikely, for the first commentarii de vita sua were (as far as we know) those of Aemilius Scaurus, written after his censorship of 109, when Polybius was already dead. Earlier Roman commentarii, as M. points out (p.181), were completely non-literary productions.

6.

Chapter five deals with problems arising out of two partially contrasted practices employed by historians, viz. the long-established custom (going back to the attacks made on Homer by the early poets) of denigrating one's predecessors in order to portray oneself as the one praiseworthy seeker after truth and, on the other hand, that of setting oneself in a historical tradition, by starting out at the point where an earlier historian left off and frequently (though not invariably) holding him up for praise. M. has a long discussion of the role of polemic in Greek historiography, especially as a means of self-definition. It ws less common at Rome, though Livy as usual is the exception, with his attack on levissimi ex Graecis, who had had the audacity to favour Parthia against Rome (ix.17-19) - this in the course of an oddly placed digression on what would have happened, had Alexander moved west to attack Rome[13]. M. here shows clearly the various ways in which polemic could serve the historian and the factors, traditional and other, which encouraged it.

In contrast, the role of the 'continuator' presents contradictions. This practice is linked to, but does not always coincide with, the adoption of an 'approved model'. Nor does an 'approved model' necessarily attract his successor's undiluted praise, for the latter may aim to surpass him, as Polybius did Ephorus, of whom he generally approves. Moreover, in 'continuing' an author one may be merely putting forward a claim to write a particular sort of history or to develop a theme already present in the author one is 'continuing'. The earliest and most striking example of this is the 'continuation' of Thucydides in Xenophon's Hellenica; but Thucydides was also continued by Cratippus, the Oxyrhynchus historian and Theopompus. The same motive is probably also present in Polybius, whose early introductory books followed on after Timaeus, but who subjected the latter to the most virulent and abusive criticism. Here the link probably lay in Polybius' determination to oust Timaeus as the historian of the west.

Following a survey of comments made by Roman historians from Sallust to Ammianus about their predecessors, M. has an interesting note on Ammianus xxxi.16.9, where the latter refers to himself as 'miles quondam et Graecus'. This phrase Fornara saw as self-definition, whereas Matthews declared that it was not a boast, since by that time history was normally written by civilians and soldiers were, and were seen as, boorish. Then what was it? M. puts the phrase in a new light. It is, he argues, a challenge to the reader inasmuch as it recalls the ancient tradition of the soldier-historian, like Xenophon or Polybius (the word 'graecus' points in that direction rather than to the Romans Caesar and Velleius). This is a good example of the fruitfulness of M.'s approach.

7.

The concluding chapter begins with a succinct review of the main argument. Briefly, this emphasises the importance of the rhetorical tradition, within the limits of which innovation could take place and also shows how that tradition had itself to a considerable extent been moulded and confirmed bythe procedures of earlier writers, seeking to validate their own writings. Next there is a useful summary of some of the compositional problems that arose in contemporary and non-contemporary history-writing and the different conventions prevalent in Greek and Roman historians respectively. There are five appendices: : 1. a list of the main ancient historians with dates; 2. the practice of Greek and Roman historians respectively in giving or witholding their names and place of origin; 3. a convincing argument that in Panath. 149-50 and Panegyr. 7-10 Isocrates was not rating the ears above the eyes generally, but merely enunciating the truism that we necessarily know about more historical events from the reports of others than we do from autopsy. (I was not alone in getting this wrong in my Commentary on Polybius, on xii.27.1); 4. a list of various criteria available for distinguishing between alternative versions presented by one's sources; 5. the Roman use of nos and nostri.

The above comments hardly begin to indicate the richness and variety of the subjects discussed in this admirable book. It rests on a thorough appraisal of an extensive range of Greek and Roman authors and constantly throws fresh light on what they were really up to and why. It will be widely acclaimed, read and referred to by all interested in ancient historiography.

I conclude with a few minor corrections:

p.82 (and index): for 'Chimara' read 'Chiomara'.
p.155: for 'chronical' read 'chronicle'.
p.179 n.17: for 'Alien Wisdom' read 'Greek Biography'.
p.199: the later books of Dio are here said to be preserved only in excerpts and an epitome 'like Polybius'; there was no epitome of Polybius.

F. W. WALBANK

Peterhouse, Cambridge


NOTES

[1] Penguin, 1975, 76. [Return to text]

[2] H Homeyer, Lukian: wie man Geschichte schreiben soll (1965). [Return to text]

[3] Op.cit. (n.1) 12. [Return to text]

[4] For example, S. Hornblower in his introduction to Greek Historiography (1996), 1-72, where he likens the succession of historians to 'a set of pigeon-holes' rather than 'an organically growing coral reef'. [Return to text]

[5] Cf. F.W.Walbank, Selected Papers: Studies in Greek and Roman Historiography (1985), 230. [Return to text]

[6] Tacitus (1984) i.138. [Return to text]

[7] ap.Sext.Emp.Math.i.263-4. [Return to text]

[8] Polybius uses two words about Fabius Pictor, pistis and epigrafe. The former will mean 'reliability', 'trustworthiness' and the latter 'his repute' (not, as Schweighaeuser, 'the title of his book'). The pistis, however, is based on his being a senator as well as a contemporary and his repute must come primarily from his rank. [Return to text]

[9] See F. W. Walbank, op. cit. (n.5) 94-8 = CQ 39 (1945) 15-18. [Return to text]

[10] Cf. J. Moles, JHS 105 (1985) 168. [Return to text]

[11] JHS 109 (1989) 186-9. [Return to text]

[12] Op.cit. (n.5) 338-43 = Historia antiqua: Commentationes Lovanienses in honorem W. Peremans septuagenarii editae (Leuven, 1977) 156-62. [Return to text]

[13] See my discussion in Ancient Macedonia: Studies in Honor of C. F. Edson ed H. Dell (1981) 344-56. [Return to text]


Back to HISTOS: Contents of the current volume

Back to HISTOS Main Menu

To Durham Classics Department Home Page