At the court of the Phaeacians, Demodocus sings of the quarrel between Odysseus and Achilles and delights his listeners, all except the still-unrevealed Odysseus who covers his head and weeps. During the feast that follows, Odysseus, despite his grief, sends the singer a rich portion of meat and salutes him, praising how well he sang ‘all that the Achaeans did and suffered and toiled, as if you were present yourself, or heard it from one who was’. In this simile, Odysseus anticipates the twin methods of validation for contemporary historians: eyewitness (autopsy) and inquiry of the participants in events. In ancient historiography, professions of autopsy and inquiry are found from Herodotus to Ammianus, and they serve as one of the most prominent means of claiming the authority to narrate contemporary and non-contemporary history. In this chapter, we shall survey some of the issues revolving around inquiry for ancient historians, treating the theoretical observations of the historians on the difficulties and problems raised by inquiry, as well as the explicit claims made by historians in the course of their narratives.

EYES, EARS AND CONTEMPORARY HISTORY

In Greek historiography, reliance on autopsy and inquiry is first found in developed form in Herodotus but, not surprisingly in view of the Greek capacity for examination, a long tradition of reliance on, and questioning of, the validity of this type of knowledge lies behind the first historian.

It is characteristic of the Roman historians to use the first-person plural frequently when referring to the Roman state or to Roman soldiers in battle. Cato may have been the first to do this, since one fragment of the Origins speaks of ‘our commander’. By Sallust's time, the convention is already fully developed. He speaks of ‘our ancestors’, ‘our state’, ‘our army’, and he refers to Romans in battle as ‘our men’, ‘our forces’ or simply ‘nostri’. Caesar as well in the Gallic War refers to ‘our men’.

Sallust does not use this convention, however, in the Catiline, nor do the Romans use this designation when writing about civil war, where it might seem manifestly inappropriate. A striking exception is Caesar, who even in the Civil War uses ‘nostri’ — but as a reference to his own men, not the Roman soldiers of his enemies. This must have been a bold step, and, if not pure arrogance, may perhaps be taken as another indication that the work lacked his ultima manus. Of Caesar's continuators the authors of the Alexandrian War and the Spanish War use ‘nostri’ of Caesar's men, the author of the African War does not.

Velleius too speaks of Roman soldiers in the first person plural, the more appropriately in that he himself was a soldier and participant; but he also follows the convention where he is not involved, referring to ‘our empire’ and ‘our slaughter’ by Ariminius.

Two passages of Isocrates are sometimes cited to argue that the orator inverted the usual preference of eyes over ears, and in doing so encouraged the writing of non-contemporary over contemporary history. In the former (Paneg. 7–10), Isocrates' exhortations not to avoid what has been said before but rather to attempt to say it better than has previously been said is seen both as a rejection of contemporary history that relies on eyewitness accounts, and a preference for non-contemporary history whose main goal is stylistic excellence and a lack of concern with accuracy. This view is then supported by the second passage (Panath. 149–50) where Isocrates says that all men have greater knowledge ‘from hearsay rather than from sight’, διὰ τῆς ἀκοῆς ἢ τῆς ὄΨεως. It would indeed require a great orator to demonstrate that ears are more trustworthy than eyes, and an examination of the full context of each passage shows that Isocrates in fact said nothing of the sort.

(i) Paneg. 7–10: Isocrates has announced at the beginning of this work that he will give his counsels on the war against the Persians and on domestic concord; he knows it is a much-worked theme, but he hopes to treat it differently (and, of course, better).

I began this work with a simple question: what do ancient historians tell us about themselves? Like many simple questions, it proved rather difficult to answer. Ancient historians give us a great deal of information about themselves, whether in their choice of subject, or their language of implied judgement, or what incidents they choose for elaboration, or how they see human actions and motivation. Indeed, it is only a slight exaggeration to say that for most ancient historians almost every sentence in their work tells us something about its creator. As I further refined the topic, I saw that I wanted to examine the Selbstdarstellung, the self-presentation or self-display, of the ancient historian, that is, how he claimed to be a trustworthy and authoritative narrator. When I saw that the same or similar claims reappeared so often, I realised that I also needed to consider the importance of tradition in ancient literary criticism and creation. And I began to see, moreover, how the two issues were closely related.

My goal in this book has been to examine the creation of authority as an aspect of the ancient historiographical tradition, as it developed from Herodotus and Thucydides. In doing this, I have treated many topics in ancient historiography that have been intensively studied for over a century.

Our study has concentrated on the explicit attempts by the ancient historians to convince the reader of their authority to narrate the deeds, and to portray themselves as believable narrators of those deeds. We have seen how the dictates of ancient literary criticism enjoined authors to work within a tradition, and to show their innovation within that tradition. As certain historians became accepted models for imitation, their concerns and approach dictated for those who followed the proper way to write history. As the earlier historians were ‘authoritative’, so their followers sought to imitate the manner by which those predecessors had constructed their own authority. It was in this way that authority and tradition were closely related, and so long as the belief in imitation held sway, there could be no authority outside of tradition.

In concluding, it may be useful to organise our results according to three categories. In the first we shall summarise our findings by topic, following the order of the chapters in the book. Then we shall compare the procedures of contemporary and non-contemporary historians. Finally, we shall look at some of the differences between Greek and Roman historians.

In presenting his work to the public, the ancient historian portrays it as something of great importance both in the doing and the telling.

Can a revised history of exegesis now be sketched? Origen, who really did build an alternative paideia based on the alternative biblical literature, pirating all the methods used in the Hellenistic schools for the exegesis of barbarian books and defending that position against the criticisms of Celsus, will clearly provide an important pivot, but the attempt here is to offer a chronological account.

In such a history, the New Testament's use of the Jewish scriptures cannot be passed over, for here the initial lines were laid down. The letters of Paul, our earliest Christian writer, are steeped in scriptural language and allusion. Paul was a Jew who knew his Bible at the deep, almost unconscious, level in which its language and perspective, especially but not solely that of Psalms and Prophets, created his world. His writing is in part collage, his argument is given authority by quotation. Along with some other New Testament writers who are on occasion contrasted with him – the authors of Hebrews and of Matthew's Gospel, and even those who produced the Johannine material – Paul evidences an exegesis of scripture which is unselfconscious, entirely taken for granted. There is, of course, a different perspective from most Jews, yet here we seem still to be reading the work of Jews with a new perspective on their own heritage. Their prophecies, they believe, have been fulfilled, and that gives a new focus, namely Jesus the Christ; their covenant has been renewed, and that gives a different approach to the true import of their law. But the crucial transformation, I suggest, does not lie here.

It might be said that paraenetic exegesis had primacy of place. The scriptures were always treated as the Word of God and as the guide to life. Even though Halakah was rejected by the increasingly Gentile Christian movement at a very early stage, the notion that moral teaching was enshrined in God's Word replaced it very rapidly. As we have seen, the pervasive exhortation not only compounded sayings from Psalms, Proverbs and the Gospels into new expressions of the character to be fostered by the Christian, but also drew on biblical models who exemplified particular virtues or lived a life worth imitating. Furthermore, the common assumption of the surrounding culture was that literature was read for the sake of moral improvement. It was characteristic of the Antiochene school to follow the practice of the rhetorical schools in seeking the moral import of the text. Generally, the expectation that a tale had a moral, that the text's intention was the improvement of the reader, encouraged ways of reading the text, especially in the homily genre, that had a practical outcome with respect to lifestyle, interior attitude and ethical choice.

It is against that background that we turn to examine the exegetical homilies of John Chrysostom. John drew huge crowds as the leading rhetorical speaker of the time, offending people in high places by his rigorous puritanism and moral outspokenness, but delighting the populace with the message of God's mercy and love and the call for repentance.

In addition to the importance of intertextuality, critical theory has latterly reminded us of the problematic relationship between language and its referent. For the Fathers, reference constituted meaning, and inappropriate interpretations could be challenged by consideration of that to which the language referred. Although language could have a ‘biblical meaning’ discernible by multiplying cross-references, text being interpreted by text, ascertaining reference was far more fundamental to patristic exegesis than determining the sense of the words. Discussion of the traditional ‘senses of scripture’ (literal, allegorical, tropological and so on) does not provide a key to method as such. A complex grid of ‘reading strategies’ plotted against a typology of multiple senses illuminates a fundamentally ‘sacramental’ understanding of the meaning of scripture: the linguistic sign represented the reality to which it referred, whether transparently or obliquely.

Towards the end of the last chapter, we began discussing how a ‘sacramental’ understanding of language might relate to the distinction between typology and allegory. Chapter 11 will show how ultimately typology and allegory contribute to Christian mimēsis, or figural representation, both being so interwoven that a firm differentiation is very hard to make. But meanwhile we must consider the explicitly exegetical debate which emerged in the fourth century, and which is often characterised in terms of a difference between typological and allegorical exegesis – I refer of course to the Antiochene reaction against Alexandrian allegory.

Northrop Frye, whose book The Great Code provided some useful clues for us in the last chapter, speaks of allegory arising when the story-myth finds its true meaning in a conceptual or argumentative translation. Allegory ceases to be story and becomes propositional; typology, on the other hand, retains the narrative and sequence. Building on Frye, we observed that by typological exegesis, meaning is discovered in ‘universal’ narrative patterns played out in past, present and future, the intersection of particular story and storytype. We also found non-narrative types, signs and symbols like the bird signing the cross as it flies, and Moses' arms doing likewise at the battle with Amalek. We decided that the important thing was the mimetic or ikonic quality of symbol or story. What I now propose is a distinction between ikonic and symbolic mimēsis, associating the first with Antiochene exegesis, the second with Alexandrian allegory.

The discussion so far, especially in Part III, surely raises some questions for standard accounts of patristic exegetical method. It is time we took stock and faced the implications.

Patristic method is commonly characterised in terms of ‘literal’, ‘typological’ and ‘allegorical’ exegesis, the Fathers being understood as the precursors of the mediaeval fourfold sense. Simonetti, for example, tends to work with two senses, material and spiritual, but subdivides the latter into moral and typological. This threefold division, found he thinks in Origen, is then developed into a fourfold pattern by John Cassian. Debate in the fourth century between the two great schools of Antioch and Alexandria is generally taken to focus on the merits of the ‘literal’ and ‘allegorical’ approaches, the ‘literal’ usually being equated with the ‘historical’. Some of the great exegetes, like the Cappadocians, Jerome and Cyril of Alexandria, since they do not fit into these categories, are described as ‘eclectic’ in their method. But not one of these conventional terms is univocal. A review of each will highlight the problems of treating them as methodological categories.

Literal meanings

The assumption that we know what ‘literal’ means conceals different usages, with different weightings in our world from the world of the Fathers. For us, ‘literal’ may mean the ‘plain sense’ of the words, taking full account of context and including metaphors such as ‘God is my Rock’; by contrast the Fathers distinguished wording from sense, and the normal sense of a word from its use as a metaphor, so that they would argue that ‘God is my Rock’ is an absurdity ‘according to the letter’, and so one must take it tropikōs, that is, metaphorically or tropologically