The quirkiness of the comment made by a character in Nabokov (quoted above) makes an apt beginning for an essay that will have much to say about meaningful coincidences. A reader who encounters Nabokov’s novel without knowing Herodotus’ story about Polycrates may be momentarily puzzled, though amused. Those who do catch the allusion will note the modernizing variant (a cuff-link, not a ring), and the added detail of its being the anniversary of the loss, a point that reinforces the unlikelihood of a parallel outcome; but above all such a reader will recognize that the climax has become an anti-climax: nothing happens, and the fish is just a fish. In a sense there is no story. Or rather, the story becomes effective here only through the contrast, because in Herodotus Polycrates did recover his ring from the belly of a fish. Nabokov’s outcome is, of course, overwhelmingly more plausible, but accustomed as we are to narratives in which loose ends are tied up, we feel that something is missing. In Herodotus, however, the general tendency is for items and episodes to connect, for links to be made, and for meaning to emerge from these connections. The following pages will explore how far that tendency extends, and the final section is intended to place Herodotean practice in a larger context.

John Marincola’s remarkable monograph Authority and Tradition in Ancient Historiography (1997) was inspired by a simple question. What do ancient historians tell us about themselves?1 This culminated in a study illuminating the complex and evolving processes whereby historical writers sought to imitate and manipulate traditions established by their predecessors as a strategy to underpin their own authority – and ultimately to persuade their readers.

You do not hold an inquest into a book more than twenty years later unless it was a gamechanger. That is not just because John Marincola’s Authority and Tradition is where you go first to find out what the historians said about what they were doing. It is also because it ranged so surefootedly over such a wide range that it opened many new perspectives and corrected many old mistakes – whether Caesar’s use of the third-person in the Commentarii was just the sort of thing one did, for instance, or a bold unobvious choice (H. Flower,in this volume); or what difference Rome’s preoccupation with social status made to the way historians framed their projects.

Looking back on the year 44 from the comparative calm of March 43 BC, Cicero described it as a caecum tempus (Fam. 12.25.3). But in addition to being a time of confusion for those who lived through it, the assassination of Caesar and the ensuing eight months brought transformations in fortune that produced the men who would dictate events in Roman history over the next dozen years.

This essay grows out of an ongoing interest in the first autobiographies written in Latin, the remains of which date to the early first century BC.2 We have only small surviving fragments of the memoirs composed by four leading Roman senators in the 90s, 80s, and 70s BC. They are Quintus Lutatius Catulus (cos. 102), Marcus Aemilius Scaurus (cos. 115), Lucius Cornelius Sulla Felix (cos. 88, 80), and Publius Rutilius Rufus (cos. 105).3 These men seem to have been the first Romans to write about their own lives in the first-person singular.4 Their writings were circulated either in their own lifetimes or immediately after their deaths. They knew each other and were linked by complex networks of competition, mutual influence, and enmities sharpened by a harsh environment of political disintegration and civil war. In other words, this little group represents an intellectual milieu of sorts, operating at a rather specific time of political and military crisis.

The study of traditional narrative originated largely as a result of the combined influence of two of the most momentous historical phenomena of the nineteenth century, namely, romanticism and the rise of the nation-state. The first modern collection of folktales, the Kinder- und Hausmärchen, edited by the brothers Jacob and Wilhelm Grimm, started coming out at the beginning of 1813, months before the Völkerschlacht in Leipzig in October 1813, which marked the end of Napoleon’s Empire. It is more than a mere coincidence that the ‘battle of the nations’ received its nickname from the German romantic literate Achim von Arnim, who was also one of the original sponsors of the Grimm enterprise.1 The ultimate goal of the collection was transparent in the very technique adopted by the brothers: by gathering stories directly from the voice of oral informants whenever possible and including data on the informant and on the time and place of the recording, they intended to document with philological accuracy the spirit of the German people in the most pristine and least adulterated form possible. Here, however, comes a striking paradox, for soon the Grimm brothers themselves realized that many of the stories they had collected found close parallels far away in time and space – so much for the German spirit. Their response to this finding, in extreme simplification, consisted in connecting folktales to an original corpus of Indo-European myths that had devolved into folktales over time and spread all over Indo-European cultural areas.

On the flyleaf of his personal copy of The Red and the Black, Stendhal wrote: ‘I believe that the truth in small as in large things, is almost unattainable – at least a truth that is somewhat circumstantial. Monsieur de Tracy used to say to me: truth can be found only in novels.’1 Ever since the emergence of ‘scientific’, evidence-based history during the second half of the nineteenth century,2 the discipline of ancient history has set itself the task of establishing what actually happened in the past. In this day and age such positivism may well seem naïve to many, yet the ancient Greek and Roman historians themselves deserve much of the responsibility for giving the overall impression that their narratives correspond closely to events as they in fact happened. This is because they have the power to cast a spell over their readers by creating narratives that are so intensely vivid that the events seem to be taking place before our very eyes.3 And that raises the question of what kind of truth they were attempting to reveal. Is it a truth that today can be found only in novels?

In the treatise conventionally entitled De malignitate Herodoti, Plutarch charges Herodotus with systematic malevolence, claiming that his intent as an author was to tarnish the reputation of the great cities and individual heroes featured in his Histories. Plutarch lists eight different ways in which he thinks Herodotus deliberately presented information that would diminish reputations: that he used the harshest of descriptive adjectives, when softer ones were possible; emphasized personally discreditable gossip; omitted mention of noble deeds; sided with the worst version when two or more versions of events were available; attributed base and self-serving motives whenever possible; claimed the desire for money as a motive when it was unnecessary; used disreputable reports while claiming that he did not personally believe them; and, finally, used some weak words of praise in order to make his much stronger criticisms more convincing.1

John Marincola has defined Polybius as ‘a highly intrusive explicator’ of his own narrative.1 Polybius regularly interrupts the main narrative of events to explain and clarify what procedure he is following. Such intrusions always retain a historiographical flavour, and Polybius comes up with words or expressions used in a new way and with a new nuance, which I define as ‘historiographical neologisms’. This chapter will show how Polybius inserts himself into a tradition (which he criticizes as well) in order to establish his own authority, and will highlight two ways: borrowing and revisiting terms from other genres – the much-discussed apodeiktike historie is a famous example2 – to give them a historiographical nuance, or creating new ones.

‘At the very beginning of non-contemporary history there was myth.’1 Thus wrote John Marincola, in a much-cited section of Authority and Tradition in Ancient Historiography devoted to history and myth or, more precisely, the ways in which historians established authority by their exclusion, and occasional inclusion, of myth. One of the attractions of Marincola’s discussion is that he stresses not just the reaction against the ‘mythical’ that was one of the hallmarks of the historiographic tradition at least from Thucydides onwards, but also the continuing importance of myth for writers such as Diodorus Siculus and Dionysius of Halicarnassus. ‘The charms of myth, like those of the Sirens’, he concludes, ‘were simply too great to resist.’2

Six decades ago (my chronology is as exact as that of Thucydides’ ‘pentekontaëtia’), A. (‘Tony’) Andrewes, Oxford’s Wykeham Professor, laid down a number of markers for the ways in which a historian of late fifth-century BC Greece (such as he – and I) could or properly should make use of a famous pair of Thucydidean speeches, for purposes of historical reconstruction and analysis.1 Since 1962, there has been a plethora of scholarship published on Thucydidean historiography in general, and on the speeches/rhetoric in particular, some of it very good, including great work by Andrewes himself in Commentary mode.2 But the issues raised in and by the ‘Mytilenaean Debate’ and indeed by Thucydides’ inchoate History as a whole have not gone away, and they will not do so any time soon.3

As the terms ‘authority’ and ‘tradition’ in my title suggest, in this chapter I try to ask of Philostratus’ Lives of the Sophists some of the questions asked of Greek and Latin historians by John Marincola in his influential book Authority and Tradition in Ancient Historiography. These questions are related to ones that were asked by Thomas Schmitz in his chapter ‘Narrator and Audience in Philostratus’ Lives of the Sophists’, published in 2009 but drafted several years earlier. Schmitz made a pre-publication draft available to Tim Whitmarsh when he was writing his shorter but comparably illuminating scrutiny of the narrator of the Lives, published in 2004 as part of chapter 32 of Narrators, Narratees, and Narratives in Ancient Greek Literature.

I put forward a new suggestion here about what John Marincola in an important paper called ‘the narrator’s presence’ in Herodotus’ Histories.1 I reconsider, in particular, a number of claims of autopsy and suggest on that basis that it is sometimes illuminating to think of Herodotus as adopting the role of tour guide or, rather, virtual tour guide. My arguments involve a view about Herodotus’ ‘good faith’ or ‘trustworthiness’ that will not be agreeable to those who receive with (sometimes exasperated) indignation any suggestion that Herodotus knowingly said untrue things in the Histories. I hope, however, that what I have to say makes a coherent and convincing case, with a more persuasive explanation than has been offered by others,2 that false statements some scholars would condemn as mere lies make in their context in the Histories an effective and readily comprehensible contribution to Herodotus’ narrative aims, and that he would have seen them as harmless or, in our contemporary term, ‘victimless’ untruths.