Chapter 1 The history of the book: Augustine’s City of God and post-Roman cultural memory
1 Memory and expectation
Unlike Horace’s ideal epic poet, who hurried his readers “into the midst of things” (ars poetica 149), Augustine had a way of beginning at the beginning, then beginning again. In his other great narrative work, and his closest engagement with Roman epic outside City of God, he began the story of himself, if not ab ovo then at his mother’s breast (conf. 1.6.7). Ten books later, this long story told (conf. 11.1.1), he began again, as if where he had meant to begin all along, in order to (re)tell a greater story ab usque principio, “from the beginning in which you [God] made heaven and earth” (11.2.3). That greater story, extrapolated from Gen. 1:1–2:3, made a further triad of books, indeed a Christian trinity.1 Roughly two decades later, c. 417, upon completing another ten books in which he had reviewed the history of the civitas Romana from its beginnings (ab origine sua; civ. Dei 2.2) and “refute[d] the objections of the ungodly, who prefer[red] their own gods to the Founder of the Holy City of which [he had] undertaken to speak” (civ. Dei 10.32), Augustine renewed his undertaking in the preface to another supplementary triad, this time of four-book units, in which he would relate the “origin, progress and merited ends of the two cities,” one heavenly and the other earthly. The curious symmetry between Confessions and City of God was already recognized by Augustine’s first biographer, Possidius, who, embarking upon a “Life” that could not simply summarize the subject’s own account of himself in ten books, announced that he would narrate the latter’s “origin, career and appointed end.”2 In Augustinian narratology, as in evolutionary biology, ontogeny recapitulates phylogeny.
Although only half the length of the sentence addressed to Marcellinus at the beginning of City of God, whose sinuosity once deceived Henri-Irénée Marrou into declaring Augustine to be a poor stylist,3 the first sentence of Book 11 is just as enveloping (civ. Dei 11.1; Dyson 449):
The City of God of which we speak is that to which the Scriptures bear witness: the Scriptures which, excelling all the writings of all the nations in their divine authority, have brought under their sway every kind of human genius, not by a chance motion of the soul, but clearly by the supreme disposition of providence.
Almost all English translations stumble in their rendering of this pronouncement. Its first trick is to begin with the “City of God” as the subject and reiterated title of Augustine’s work, and then to begin again immediately with the “scriptures” – or rather the singular scriptura – that are (is) the grammatical subject of the main clause, and to which all writings and works of human genius, Augustine’s included, are said to have been subjected by divine providence. Dyson’s pluralizing of the noun and expansion of the syntax (“the Scriptures . . . the Scriptures”) instructively ruin one of Augustine’s more calculated compositional effects. The force of this second opening of City of God derives from the strict opposition that it establishes between the perfect singularity of God’s attributes (in the order of the Latin: civitas Dei, scriptura excellens, dispositio providentiae, divina auctoritas) and the plurality of each of the attributes of a fatally divided and distracted humankind (motus animorum, omnes gentes, omnes litterae, omnia genera ingeniorum humanorum). “The City of God of which we speak is one attested by that scripture . . . which, in its divine authority, excels all the writings (litteras) of all the nations.” More literally: “The City of God of which we speak is one attested by that writing which, in its divine authority, excels all the literary resources of all the nations.” “Literary” in this context should not be taken to imply the existence of any “literature” in the modern, post-Romantic sense of a more or less clearly delimited universe of mainly imaginative, aesthetically ambitious artifacts. Litterae in its extended, classical and post-classical Latin applications meant something like the written resources – both “archival” and “canonical” – of a human culture; the textual reserve and repertoire of a people’s “cultural memory.”4 Augustine’s City of God is a manifesto for the people or nation whose cultural memory – of past, present, and future – would be primarily encoded in the biblical–canonical books from Genesis to Revelation.
Like other Latin speakers and writers of his time, Augustine habitually spoke and wrote of the Christian biblical canon as scripturae (“writings”). To understand how much depends on the singular scriptura of the beginning of Book 11 of City of God, we must turn back to the final chapter of Book 10. In his work On the Return of the Soul the philosopher Porphyry claimed to have been unable to discover any “universal way of the soul’s deliverance,” despite his extensive researches into the traditions of various nations, which had included a process described by Augustine as “historical inquiry” (historialis cognitio). Augustine shows that Porphyry’s comparative and historical inquiries were incomplete. The universal way of the soul’s deliverance could in fact be discerned both in the historical records and in the contemporary practices of a particular people, the Hebrews or Jews, “whose commonwealth (res publica) was in a certain sense consecrated to foreshadow and herald, by its tabernacles and temple and priesthood and sacrifices, the City of God which was to be gathered in from all the nations (ex omnibus gentibus).”5 The conceptual and narrative relationship here sketched between the cultural institutions of a particular people and the promise of a universal way of salvation already points towards a resolution of the dialectic of the “all” and the “one” inscribed at the beginning of Book 11. Without yet mentioning the key notion of “scripture(s),” Augustine articulates two others essential to it, namely “history” and “authority” (civ. Dei 10.32):
Porphyry says that the universal way of the soul’s deliverance has never come to his knowledge through historical inquiry. But what more illustrious history can be found than that which has taken possession of the whole world (universum orbem . . . obtinuit) because its authority is so eminent? Or what history could be more faithful than that which narrates past events and foretells future ones, so many of the predictions of which have been fulfilled that we are enabled to believe without doubt that the rest will be fulfilled also?
“History” in this use of the word stands for a narrative and cognitive unity rendered through writing. As surely as this history has taken possession of the world, the full mental possession of it is the soul’s assurance of eternal life.
Augustine’s mature conception of the Jewish–Christian scripture/s was the product of complex processes in his intellectual life and environment, but probably no single factor was more important than his response to Manicheism. For this ex-Manichee, saving God’s justice and saving human freedom – that is, saving the two of them together – meant saving the overall unity of the scriptures by then held as canonical in the “catholic” church. In the first place, the writings of St. Paul (crucially, his Letter to the Romans) and the writings traditionally attributed to Moses (crucially, his account of Adam’s sin in Genesis) had to be shown to explicate each other. The Apostle had coordinated Adam’s original fault with the current condition of humankind and, through the person of Christ, with its future prospects. He had done so scripturally, by referring to the written history of God’s past dealings with the Jews and extrapolating from it, in the light of the “gospel,” to his present and future dealings with other nations. From Paul, Augustine derived two guiding principles: first, that God’s long-term purposes for humankind were securely encoded in the texts of his prophets, apostles, and evangelists; second, that in this life those purposes – and hence those texts – were indecipherable beyond a certain point (Rom. 11:33). The history that mattered was plotted in divine scripture; scripture was partly opaque; the clarity and opacity of scripture defined the human condition as, for the time being, a “textual” one.
Other Christian exegetes and theologians, working with the same or similar materials, had appealed before Augustine to the marvelous unity of the scriptural dispensation. None, however, had tackled the task of biblical interpretation with the same urgent desire to locate the predicament of the individual human being within a narrative of the origin, descent, and future of the species. Absent evolutionary biology, Augustine’s theological notions of the relation between human ontogeny and human phylogeny were worked out on the basis of the books that came to be known in the Middle Ages as “The Bible.” The Bible was his genetic code, his book of the generations and of the history of nations, the narrative structure in reference to which all other histories or stories, collective and individual, would make whatever sense they could ultimately bear.
Vital encouragement for the “historical” biblicist that Augustine would become by the time he composed City of God was supplied by the Book of Rules of Tyconius, a fellow North African writer and former Donatist whose work was in his hands by the early 390s. The Liber regularum furnished Augustine with the means for an anti-Manichean reading of Paul: “By insisting that the dynamics of Law and faith, will and grace were constant across nations, times, and individuals, from Abraham to Jacob to David to the apostles and on into the present life of the church, Tyconius disclosed in the entire double canon of scripture a continuous and consistent record of God’s saving acts in history.”6 It was partly on the strength of the guidance provided by Tyconius that Augustine conceived the project of De doctrina Christiana (“On Christian Instruction”), a work in which he intended to set out the principles for an effective exegesis of the “holy scriptures.”7 Like other works of his begun in the early and mid 390s, this one was abandoned incomplete.8 Not until the late 420s, around the time he finished the City of God and as he was reviewing his entire published output for the Retractations (retr.), did Augustine pick up his hermeneutical treatise again in the middle of Book 3 and relaunch it with an annotated digest of Tyconius’s Liber regularum. In the meantime, he had made a fresh assault on the problem of discerning and expounding God’s saving acts in history, with direct reference to his own case, in Confessions.
Augustine divided the Confessions as a whole between Books 10 and 11, describing the first decade as “about me” and the final triad as “about the holy scriptures” (retr. 2.6). The work can usefully be seen as a set of transactions between three kinds of memory: the natural memory of an individual human mind (Book 10); the prosthetic memory of a life humanly transcribed (Books 1–9); and the providential memory divinely encoded in the Bible as a whole and, on Augustine’s reading, already epitomized in the Genesis narrative of the seven days of creation (Books 11–13).9 To complete this analysis, while also extending it on the one hand to Christian Instruction and on the other to City of God, we should allow for a further repository of accumulated knowledge: the collective or social memory transmitted in the texts and practices of a given culture and acquired by individuals in the course of their education and wider experience (“cultural memory”). Augustine had traversed the “places” of such a cultural memory, in its late “classical” Graeco-Roman form, in Book 2 of Christian Instruction (doc. Chr. 19.29–42.63). He would review them again, according to a different taxonomy and with a sharper polemical intent, in Books 1–10 of City of God. In the interim, in Confessions, he narrated a process by which the inherited and assimilated resources of this same “classical” culture, now his own, were exposed to the radical critique of a biblical counter-memory.10 “Not in vain have you willed so many pages to be written, pages deep in shadow, obscure in their secrets,” Augustine assured himself, addressing God as usual (conf. 11.2.3, Boulding). His prayer for access to the scriptures echoes the idioms of his earlier exploration of the recesses of his own memory (conf. 10) and those of the prologue to Tyconius’ Book of Rules (quoted at doc. Chr. 3.30.43). “Let me confess to you,” he went on, “all I have found in your books (quidquid invenero in libris tuis) . . . from the beginning when you made heaven and earth to that everlasting reign when we shall be with you in your holy city.”
The last three books of Confessions model a process of “finding” (invenire) and of “proclaiming” (proferre) the sense of the scriptures, which Augustine had lately theorized (doc. Chr. 1.1.1). The exegetical–confessional text becomes a medium between God’s understanding of all things, as selectively imparted in the books of his appointed authors, and the understandings of Augustine’s present and future readers. Like the other confessions that make up the Confessions, Augustine’s act of confessing what he knows and does not know about God’s “law” in the scriptures (conf. 11.2.2) is essentially an act of recollection. It is a making-present, in the extended “now” of a humanly composed text, of things previously discovered by the author in other texts that had been humanly composed but divinely disposed (cf. doc. Chr. 2.2.3). From the providential memory of the scriptures those things pass via the natural memory of Augustine to the prosthetic memory of his text and so to the natural memories of his readers. Those “things,” as the above-quoted passage from conf. 11.2.3 already suggests, were identical with the history of the “city of God” hymned by the psalmist. Another poem, Ambrose’s hymn Deus creator omnium (“God, creator of all things”), gave the cue for the famous analogy by which, later in Book 11, Augustine would attempt to account for the general human experience of life in time (conf. 11.28.38):
Suppose I have to recite a poem I know by heart. Before I begin, my expectation is directed to the whole psalm, but once I have begun, whatever I plucked away from the domain of expectation and tossed behind me to the past becomes the business of my memory, and the vital energy of what I am doing is in tension between the two of them . . . What is true of the poem as a whole is true equally of its individual parts and syllables. The same is true of the whole long performance, in which the poem may be a single item. The same thing happens in the entirety of a person’s life, of which all his or her actions are parts; and the same in the entire sweep of human history (in toto saeculo), the parts of which are individual human lives.
God’s knowledge of “all past and all future things,” Augustine ventured, might be comparable with a human being’s knowledge of a single poem, due allowance made for the critical fact that the mind of God, being outside time, could never be pulled between memory and expectation as human beings were in the saeculum or “historical time.”11
That predicament of suspense (distentio), emblematized in Confessions by the experience of reciting a memorized text, would for Augustine be a defining trait of the “city” or community of God in its earthly wayfaring. It is strikingly enacted in the first sentence of City of God, which mimics by its own drawn-out periodic structure the processes of deferral and distraction that are natural to human life “in the course of these succeeding times”, (in hoc temporum cursu; civ. Dei 1.1.1). No less than all other human beings, the godly are forever being pitched in medias res, even when beginning again with their eyes set on that “most glorious City of God.” The pathos of Augustine’s prayer for entry into the hidden sense of God’s scriptures stems from a recognition that those inscribed places of divine mind held out rare opportunities of stability, self-collectedness, and “continence” (cf. conf. 10.9.16, 29.40) in the midst of a world set whirling by the same God: “Yours is the day, yours the night, a sign from you sends minutes speeding by; spare in their fleeting course a space for us to ponder the hidden wonders of your law” (11.2.3). If the Psalms were a primary inspiration for both Confessions and City of God, that was no doubt partly because, as readily memorizable lyric texts of impressive narrative compass, they had the power not only to represent but also to relieve a secular and textual condition of life whose highest felicity could – on scriptural authority – be identified with meditation on the scriptures (Ps. 1:1–2).
2 The writings of all the nations
2.1 Varro’s pen
Immense as it is, City of God is held together by a tissue of actual or implied cross-references, many of them turning on individual figures from biblical or Roman history. One such figure is that of Rome’s second king after Romulus, Numa Pompilius. It was thanks to Numa, Augustine reports midway through the synoptic–historical Book 18, that Rome acquired many of its (false) gods – so many, indeed, that apparently there was no space left in heaven for Numa himself, who unlike Romulus was not deified after his death (civ. Dei 18.24). Numa’s role in establishing Roman religious rites (sacra) is signaled early on and frequently underlined (civ. Dei 2.16, 3.9, 3.12, 4.23). The main treatment of it is reserved for a prominent passage at the end of Book 7, as Augustine wraps up his critique of the demonic religious practices current in the more popular forms of Roman culture, both “poetic” and “civic,” and prepares to engage with the “natural” theology of the philosophers in Book 8 (following a triple division of theologia that he took from Varro; civ. Dei 4.27, 6.5). In his De cultu deorum (“On the Worship of the Gods”), Varro (116–27 BC) had specified a class of “select gods” whose functions he identified with aspects of the natural world. Augustine rejects out of hand Varro’s “naturalistic” defense of the core elements of the pagan pantheon. These supposed deities of nature, like the deities popularly worshipped in place of long-dead human beings, were merely the imposture of demons (fallen angels) who liked to be taken for gods (civ. Dei 7.33). The example of Numa Pompilius clinches the argument (7.34–35).
Augustine inferred from Varro that Numa had communed with demons who had revealed to him the origins of the rites of Roman religion. Afraid to disclose what he had learned, yet not wishing to offend his informants, Numa had committed these mysteries to books that, after his death, were buried with him on the Janiculum at Rome. Years later, by Varro’s account, a farmer accidentally turned up Numa’s secret books with his plow, and took them to the praetor. The praetor referred them to the Senate, and leading senators read far enough to appreciate why Numa had acted as he did, whereupon they ordered the books burnt. Augustine assumes that Numa must have discovered how execrable were the demons whom the Romans mistakenly venerated as gods, or else that all the reputed “gods of the nations” were merely dead human beings. Either way, Augustine’s case was made. But he does not rest it there. Varro’s story of Numa’s books was evidence, for the author of City of God, not only of the demonic nature of traditional Roman religion but also of the sublime workings of divine providence: the demons had not been permitted to warn Numa that he should burn his books before he died, “nor could they prevent them becoming known by resisting either the plow that uncovered them or the pen of Varro (stilo Varronis) through which the record of these events has come down to our time” (in nostram memoriam pervenerunt; civ. Dei 7.35).
Augustine’s mention of the providential disclosure of Numa’s books by the farmer’s plow and Varro’s pen is matched by his emphasis a few pages earlier, in a passage that exceeds the requirements of its immediate argumentative context, upon the providential preservation and worldwide dissemination of Jewish writings prophesying “the eternal salvation which was to come through Christ” (civ. Dei 7.32). That passage is a prelude to the elaboration, between Books 10 and 11, of Augustine’s main argument for the supreme importance of the Jewish scriptures, understood in a Christian sense, as revelation to all nations of the soul’s deliverance (see Section 1 above). There is thus in view at this point of City of God a double dispensation in writing: on the one hand, a set of texts that, right down to the details of Jewish worship, ceremonies “and whatever else pertains to that service which is due to God and properly called latreia,” symbolize those things which “the faithful in Christ . . . believe to be fulfilled, or which we see in the course of fulfilment, or the fulfilment of which we trustingly wait”; on the other, the ultimately illegible, non-extant but conjecturally recoverable records of a religious cult erroneously rendered to demons. Varro’s story of Numa’s sacrilegious books would have been attractive to Augustine for the opportunity it afforded him of drawing negative inferences about the bases of Roman religion. Its prominence in City of God owes at least as much to his desire to enhance the authority of God’s special textual provision for his people, now embracing all nations, by juxtaposing it with the written records of the gens populi Romani, “the nation of the Roman people” (the title-phrase of one of the works of Varro that he relied upon for his representation of traditional Roman religious observance).
While Augustine might claim that divine scripture outranked the literary resources (litterae) of all the nations, the only comparison that he was competent and concerned to pursue in any detail was with the “letters” or writings belonging to the Roman nation and the Latin language. The four main curricular authors of the Latin grammatical and rhetorical schools – Virgil, Cicero, Sallust, and Terence – are all to the fore in City of God and play vital roles in defining its polemical targets. None of them, however, competes in eminence with Varro.12 These other authors all excelled in their respective genres or disciplines. Varro was the great Roman polymath, hailed by one poet as “a man most learned in all domains” (quoted by Augustine, 6.2). He was the author of the encyclopaedic Libri disciplinarum that Augustine had once meant to transpose into a Christian key.13 More significantly, for City of God, he was an outstanding researcher of Rome’s antiquities and tireless compiler of the records of her “culture” in the widest sense of that term. “When we were wandering and roaming like strangers in our own city,” Cicero had written of Varro in the Academica, “it was as though your books led us back to our home” (quoted by Augustine, civ. Dei 6.2).14 Cicero’s allusion here to Varro’s books (tui libri) is doubled and expanded in the phrase with which Augustine introduces the quotation, which refers to the same author’s litteraria opera – not “literary” works in the modern sense but “written works” intended for publication in the form of books (i.e., papyrus book-rolls, in Varro’s time). A few lines later, Varro’s efforts to preserve the memory of the Roman gods are recalled from a passage in eo ipso opere litterarum suarum, “from that same work of his writings/letters.” These are not ordinary locutions. The materiality of Varro’s texts has ideological value for Augustine. Associated with their extent and durability, it is an aspect of their still commanding authority. A little later on, the Varronian oeuvre is designated, even more unusually, as omnis vestra litteratura, “all your writing” (civ. Dei 6.6). Here the word litteratura is as good as synonymous with scriptura. Almost from the start, and uniquely among the Latin authors cited by Augustine in City of God, Varro cuts the figure of a writer in the most literal sense, one who sets things down in a written form that will stand the test of time. (Augustine would have relished the irony that most of what now survives of the works of Varro quoted by him in City of God does so only through his quotations.)
After a fleeting appearance at civ. Dei 3.4, Varro is formally introduced at 4.1. Augustine looks back on his summary of Roman history in Books 1–3: “I have drawn partly upon my own recent memory . . . and partly upon the writings (conscripta) of those who have left accounts of these matters to posterity.” He adds: “Varro, who is esteemed among our adversaries as a most learned man (vir doctissimus) and the weightiest of authorities, is a case in point. He compiled separate books (dispertitos faceret libros) concerning things human and divine, assigning some books to the human and some to the divine, according to the dignity of each.” Varro’s sense of the boundaries between divine and human things will be disputed in City of God. What is perhaps most noteworthy in this formulation, however, is the precision of Augustine’s sense of Varro as a redactor of Roman cultural memory, organizing his material by means of a division between sets of book-rolls. The use of the verb conscribere (literally, “gather up, compile in writing,” “write or copy out in full”), here and elsewhere, with respect to Varro (cf. 6.2–3), points in the same direction. Even when it serves merely for rhetorical variation and rhythmical effect, the frequent substitution of scribere by conscribere with reference to Roman texts in City of God emphasizes the material solidity of the cultural archive that Augustine was aiming to represent in its most imposing form, in order then to invalidate it. In Varro’s case, that solidity is quickly dismissed as a distraction. His forty-two books of Antiquitates, twenty-five devoted to “things human” and sixteen to “things divine,” were a wonder to contemplate: “But in this whole series of distributions and distinctions,” observes Augustine, “it is vain to seek eternal life” (civ. Dei 6.4). Watching Augustine wander the imaginary library of Roman culture in City of God is like watching him investigate the places of his own memory in Book 10 of Confessions, this time secure in the knowledge not only that he will not be able to find what he is looking for but also that the thing in question is nowhere there to be found.
2.2 Comparative “literature”
Augustine’s turn to Varro at the beginning of Book 4 is echoed by his return to him at the beginning of Book 18. Again, the accent is on written delivery. Augustine had “promised to write (me scripturum esse promisi) of the origin, progress, and proper ends of the two cities” (civ. Dei 18.1). The unfolding of the promised contents of his own writing in City of God becomes a theme almost as insistent as the fulfilment of the prophecies of Jewish–Christian scripture. At the beginning of this new book, but not for the first or last time, he inventories the contents of the entire work, Varro-like in his care to apportion his subject matter into its respective containers. In Book 15 the stories of the two cities down to Abraham had run concurrently “in [Augustine’s] writing as in real time” (sicut in temporibus, ita et in nostris litteris). Then, in Books 16 and 17, the city of God had taken its course alone in his text (in meo stilo). Now it was time to put this other city back in the picture, so that the fortunes of both could be held together for comparison “in the overall view of readers” (consideratione legentium; civ. Dei 18.1). Book 18 will provide a synoptic history of the two cities from the time of Abraham to the time of Christ, beginning with a synchronism between Abraham’s birth and the accession of Ninus as king of Assyria, which Augustine found in the Chronicle of Eusebius of Caesarea, as translated into Latin and expanded by Jerome, c. 380.15 Straightaway he coordinates those data with a pre-Roman chronology supplied by Varro – hailed again as “most learned in all domains” – in his De gente populi Romani (civ. Dei 18.2). Only Assyria/Babylon could bear comparison with Rome as an empire, and, since the rise of the latter began around the time of the eclipse of the former, the combined history of these two powers would properly constitute the main axis of any narrative of the earthly city. Yet Augustine will have little to say about the Assyrians. The culture of another people (or confederation of peoples) claimed his interest instead, namely that of the Greeks, and not only because of their closer historical relationship with Rome. As representatives of the earthly city, the Athenians in particular punched far above their weight, and for a reason already given by Sallust: the “genius of their writers” (scriptorum magna ingenia) had made this people famous “throughout the world” (civ. Dei 18.2). The renown that Athens had acquired “from its learning [i.e., as committed to and transmitted by writing] and from its philosophers (ex litteris et philosophis)” ranked it with Rome for the purposes of the present comparison of the histories of the earthly and heavenly cities (propter comparationem civitatis utriusque; civ. Dei 18.2).
From the outset, then, the main section of Augustine’s parallel narrative of the two cities (Book 18) is organized partly as a comparison between two “literary” histories, one relating to the authors and texts of the “scriptural” dispensation, the other to those of “classical” Greek culture (or, speaking somewhat anachronistically, “literature”) as it had been assimilated by the Romans. The focus on textual canon-formation as a dimension of culture was all but dictated to him by his primary historical source, the comparative chronological tables (Chronici canones or Chronicle) of Eusebius–Jerome. Eusebius had followed earlier Greek chronographers in including notices on eminent Greek poets, philosophers, orators, and historians amid the records of political history. In a preface, moreover, he had made much of the fact that, on his dating, Moses not only came before Homer and Hesiod but even before the so-called “theological poets,” Linus, Musaeus, and Orpheus. And since Homer came long before Thales and the other “Seven Sages” (sophoi, sapientes), who in turn preceded Pythagoras, Socrates, Plato, and their fellow self-styled “philosophers,” the priority of Hebrew over Greek cultural tradition was assured; the tabulation of the Chronicle, which presented the records of the various “nations” in parallel columns, was designed to make that assurance visible. In Book 18 Augustine faithfully rehearses the Eusebian comparative cultural chronology (civ. Dei 18.14, 18.24–25, 18.37, 18.39), reconciling it where possible with information from Varro’s De gente populi Romani. At the same time, he adds his own flourishes. Thus he makes a pointed distinction between the myths purveyed by Greek poets about their gods and the things that can truly be ascribed to “history” (civ. Dei 18.8, 18.12–13, 18.16); recalls with relish how Athens, “the mother or nurse of liberal studies and of so many great philosophers,” owed its name to a trick of the demons (civ. Dei 18.9); and throws up his hands in dismay at the irreconcilable differences between the philosophical schools that flourished in that “demon-worshipping city” (civ. Dei 18.41; see further Book 19). To the false fictions and discordant doctrines of such a demonically possessed culture, he naturally opposes the unified and historically reliable divine discourse of the Jewish–Christian textual dispensation (civ. Dei 18.40–41 and passim).
At no point in this critical–comparative review of cultural traditions are the material media of textual transmission far from the surface of Augustine’s own text. The legend of the invention of writing (litteras) by Io, daughter of Inachus, also venerated as Isis, is skeptically rehearsed early on (civ. Dei 18.3), to be revived some chapters later in rebuttal of the claim that Egypt could have had wise men in pre-Abrahamic times (civ. Dei 18.37). For this and other synchronisms, Augustine never tires of observing, we are reliant on the texts of those who “have committed these ancient things to writing,” the scriptores historiae, as already laid under contribution by “our own people who have written chronicles – first Eusebius and afterwards Jerome” (civ. Dei 18.8; cf. 23). Varro, in this perspective, does not merely tell a story, he “inserts [it] into his writings” (suis litteris inserit; civ. Dei 18.10), just as Lactantius “inserts” certain Sibylline prophecies “into his work” (civ. Dei 18.23). Other facts and fictions are said to be “published in writing,” “committed to writing,” and so on. The event of the Fall of Troy is common knowledge “because of both its own magnitude and the outstanding skill of the authors who recorded it” (civ. Dei 18.16). The Seven Sages “did not leave behind them any memorials, at least not in writing” (quod ad litteras adtinet; civ. Dei 18.25). The Hebrew prophets, by contrast, can often be dated on the basis of their own texts (litterae, volumina, scriptura, scripta; civ. Dei 18.27, 18.29, 18.31, 18.33). Israel had had prophets in earlier times too but “those prophets who, not only in speech, but also in their writings bore witness to [Christ],” belonged to the period of Rome’s rise. As Rome was the second Babylon and future ruler of many nations, so from the moment of her foundation (ab urbe condita) the countervailing textual resources of the City of God were providentially reinforced by the “founding of a writing that prophesied more openly” (scriptura manifestius prophetica condebatur; civ. Dei 18.27). The synchronism is later repeated with a poetic twist: it was during the reign of Romulus that “the river of prophecy burst forth from its source in Israel, embodied in those texts (in eis litteris) which were to spread throughout the world (toto orbe)” (civ. Dei 18.37).
Divinely not demonically inspired, manifesting a doctrinal unity that put the “philosophers” to shame, diffused as wide in the world as Rome’s empire, predating at their origins the earliest of the Greek or Egyptian sages – the “holy writings called canonical” by Christians (civ. Dei 18.36) are characterized in Book 18 of City of God chiefly by their capacity to match and exceed the most that could be claimed for the writings of the “nations.”
At a critical juncture, however, Augustine’s argument for the supremacy of the Jewish–Christian scriptures goes beyond the literary–comparative frame that he has constructed for it on the basis of Eusebius–Jerome. Among Jerome’s most salient additions to the Chronicle was a record, for the year 283 BC, of the commissioning of the “Septuagint” (Greek translation of the Hebrew Bible) by King Ptolemy Philadelphus. Jerome’s notice concluded: “And these divine scriptures [the king] held in the Alexandrian library that he had procured for himself out of every kind of writing (ex omni genere litteraturae).” The translation of the Septuagint, which Augustine, unlike Jerome, was inclined to regard as a miracle (cf. doc. Chr. 2.15.22), provides the climax of the literary–historical narrative of Book 18 and at the same time marks the limit of the comparative model underlying it. Though included by Ptolemy in his universal library, the Septuagint is to be seen as sui generis, beyond comparison, an exorbitant textual universe unto itself. The inspired consensus of its translators ensured that “the authority of those scriptures was commended not as human, but as divine (which it was), for the benefit of the nations who were to come to believe in Christ” (civ. Dei 18.42). Unified by the spirit, dictated at a single sitting, disseminated far and wide, the Septuagint instantiated the idea of a perfectly singular scripture of divine dispensation (civ. Dei 11.1, Section 1 above). Not even Jerome’s impressively ‘literate labor’ (tam litteratum laborem) as translator of the Old Testament from Hebrew could undermine its authority (civ. Dei 18.43). In promoting the Septuagint, Augustine was careful, nonetheless, not to lower the status of the Hebrew books. In his imagination, the latter were as likely as any Greek version to be in the hands of the wandering Jews whose continuing possession of scripture was the Christians’ most practical guarantee of the reliability of the source-texts for their narrative of the City of God (civ. Dei 18.43, 18.46–47).
3 History and the books
A certain Firmus was so enthralled by Book 18 of City of God when he heard it recited over three successive afternoons, in Augustine’s presence, that he immediately asked the author for a copy of the complete work in twenty-two books (ep. 1A*.1; cf. 2*.3). Nearly every one of the other twenty-one books could have been recited comfortably in a single afternoon. But Book 18 was a monster, accounting for an eighth of the whole. It had to be that long; it is the nearest Augustine came to writing a “very short history” of the world, and also the clearest account he ever gave of the temporal and textual dispensation under which Christians held their scriptures. It is at once a self-contained historical epitome and an opening into the larger synoptic history of the two cities presented by Books 15 to 18, and by the City of God as a whole.
We have seen how, in Book 11 of Confessions, Augustine toyed with the idea that God’s comprehension of all reality might be analogous to a human being’s perfect mastery of the text of a single poem, only timeless. Poems in the classical world were traditionally imagined as sung or spoken aloud, even when they were written. When Virgil appeared with a scroll in late classical art, it was usually rolled up; the poet’s performance did not depend upon a script.16 Owing no allegiance to such traditions of high “literary” culture, Christians as a collectivity were no more obliged to keep their texts scrolled than they were to keep them all singable. In practice, they preferred the spine-hinged book or codex.17 Codex technology was mainstream book technology under the Roman Empire, relied upon by traders, lawyers, teachers, soldiers, civil servants, and by anyone else with a useful occupation requiring ready and secure access to written data. Although scrolls or rolls were in daily practical use too, those of the fancier, more polished, and calligraphic kind were mainly reserved for a niche market of leisured consumers of poetry, history, oratory, and philosophy – and, on somewhat different terms, for the sacred texts of the Jews.18 Augustine used the words liber and volumen with the same freedom as other Latin authors, in senses ranging from “book as object” through “unit of text” to “literary work.” However, when he wanted to focus attention on the bodily and cognitive–affective relationship between a reader and the book that he or she took in hand or held in sight, codex was generally the word he chose.19 The distinctive functionality of the codex, as both a highly capacious container for text and as a device for bringing the complex relationships between textually represented phenomena into a unified and cognitively impactful view, is exploited to special advantage in City of God.
Gillian Clark has noted that “Augustine was particularly aware of City of God as a book, a physical object used by readers.”20 We know that by the time he began to dictate this work, in the year 413, he was already planning the general review of his literary output that would become the Retractations. We know, because he mentioned the project in a letter of the previous year to the same Marcellinus who would receive the dedication of City of God (ep. 143). More intently and with better reason than any previous Christian author, Augustine was by then contemplating the ultimate dimensions of his own published oeuvre and anticipating the judgments that it would incur in the sight of future readers.21City of God is not only a late work in the Christian apologetic tradition, one that summons its readers to act as judges in a forensic case.22 It is also a late work of its author, one that demands to be read from beginning to end, in order to be apprehended and appreciated as a whole – like the text of the poem that Augustine had used in Confessions as a paradigm for the entire life of a human being or the entire course of human history. Whatever the external impulses for its inception, City of God has the distinction of being one of the three truly monumental works that Augustine produced on his own terms, and over which he intermittently spent many years – the two others being De trinitate (“On the Trinity”) and De Genesi ad litteram (“On the Literal Sense of Genesis”). Although its germ is already visible in Confessions, City of God was by more than a decade the last of the three to be undertaken, and it is also the most architectonic in design. Augustine had measured his task carefully. Even if this opus magnum turned out in the end to be far longer than he had expected, its overall proportions remained as envisaged in his original plan (ratio operis; civ. Dei 1.pref.).
The very first sentence of the work, as we have seen (Section 1 above), is already mimetic of the compulsive logic of the twenty-two books. City of God is Augustine’s greatest work of writerly and “bookish” coercion. In a letter to Firmus (ep. 2*.2–3), he barely concealed his irritation that his correspondent, having pressured him to send the whole, showed no sign yet of having read for himself beyond the first ten books (libri). He reminded him, as he would remind readers of the Retractations (retr. 2.43), that these first ten “books” and the subsequent twelve belonged inseparably together, the two parts of the work completing, interpreting, and reinforcing each other. Already when dispatching the exemplar from which Firmus was to have further copies made, Augustine explained how the component sections of this huge artifact ought to be bound up (ep. 1A*.1). A single papyrus or parchment codex could if necessary hold considerably more text than a single papyrus roll, which beyond a certain length would become unwieldy and fragile. Still, Augustine recognized that the City of God would probably be too big to fit manageably in one codex. The first ten libri, he suggested to Firmus, might conveniently be combined in one codex, and the last twelve in another. Or, for greater portability, Part One (Books 1–10) could be divided after Book 5, where there was a break in the argument, while Part Two could be split into three codices of four books each; the first comprising the origins of the two cities (11–14), the second their historical development (15–18), the third their appointed ends (19–22).
The simultaneously authorial and bibliological sensibility on display in Augustine’s correspondence with Firmus is something of a novelty with respect to classical literary or textual culture. What we find Augustine doing here is of a different order from a Roman historian’s division of his narrative into ten-book packets (“decades”: Livy) or a Greek philosopher’s (or his editor’s) arrangement of his work into sets of nine chapters (“enneads”: Plotinus). Those were routine devices of ancient book-making as tomology, the slicing of literary compositions to fit convenient lengths of papyrus roll. They were procedures of publishing, rather than principles of composition. By contrast, Augustine’s codicology in City of God made publishing and composition aspects of each other. We have seen him review Varro’s procedure of dividing his “human” and “divine” subject matter between different sets of “books” (dispertitos faceret libros; civ. Dei 4.1). Varro would not have considered the codex as a possible container for each of these groupings, for the simple reason that in his time and milieu codices were generally not used for the calligraphically transcribed “works” or “monuments” of higher intellectual culture, the category to which his Antiquitates was clearly meant to belong. Augustine’s bibliological horizon was different from Varro’s. The increasing visibility, in the period now known as late antiquity, of Christian biblical and non-biblical religious texts in codex form – along with the many legal, administrative, technical, and professional works that conventionally also used that format – led to a situation in which the scroll or roll form of book (in Latin, volumen) was no longer felt to be obligatory even for reputed works of higher, “literary” culture (poetry, philosophy, oratory, history). The tendency of Christian writers to disparage such works in order to exalt the “holy scriptures” will have contributed further to the weakening of the ideological link between those genres and the roll format that had traditionally been used for them. It is telling that the Latin Christian text from late antiquity that most exhaustively contrasts the respective material–textual resources of classical Graeco-Roman and biblical Jewish–Christian culture – Augustine’s City of God – is also the first one for which we have the author’s exact instructions to the copyist for its presentation in codex form.
Late in his life and in his career as a Christian author, Augustine had conceived of a large-scale written work whose overall unity would be both articulated and apprehended according to the techno-logic of the codex as text-container. It has been astutely conjectured that Augustine “never saw a bible.”23 That is to say, he is unlikely ever to have set eyes on a “pandect” or single-codex exemplar of the canonical scriptures that contained, as he would have said, the whole story of the two cities. He did, however, have direct knowledge of a codex that presented the history of all kingdoms of the “known” world from the time of Abraham and Ninus down to the present, in a tabular form that highlighted synchronisms and graphically revealed the chronological succession of powers from ancient Assyria/Babylon to the Roman Empire under its Christian leaders. He had that book open before him as he composed Book 18 of City of God. The revolutionary cognitive–epistemological force of Eusebius’s Chronicle as an instance of codex-making has been demonstrated by Anthony Grafton.24 It was Jerome, as translator and expander of the Eusebian Chronicle, who imparted that force to Latin readers in late antiquity, and Augustine who first felt it in a way that we can still register today.
The layout of the Chronicle absolutely required the codex format, and Jerome in his preface admonished future copyists to respect the special logic of its mise-en-page, including the use of different colors of ink to distinguish adjacently tabulated world powers. These visual effects, he explained, had been contrived not merely to delight the eye but in order to clarify the “order of reading” (legendi ordinem).25 Augustine’s stated concern at the beginning of Book 18 to set the histories of the heavenly and earthly cities within a single frame, “so that the two may be compared with each other in the overall view of readers (consideratione legentium),” is the clearest sign of his indebtedness to the codex-dependent optics of the Chronicle.26 In an essay that claims Augustine’s Confessions as prototype of one of the four dominant forms of “prose fiction” in western “book literature,” the Canadian literary critic and theorist Northrop Frye once described the spine-hinged book as “a mechanical device for bringing an entire artistic structure under the interpretive control of a single person,” namely the reader.27 Augustine anticipates this insight in Book 20 of City of God, when he interprets the “book of life” of Revelation 20:12 as signifying “a certain divine power, by which it will be made possible for every man to recall to memory all his own works, both good and evil, and for the mind to review them all with miraculous speed, so that each man’s knowledge will accuse or excuse his conscience” (20.14). As Eric Jager has pointed out, “the same principle informs the Confessions”28 – just as it informs Augustine’s review of his life’s “literary” works in the Retractations, and the text-based synopsis of the collective “life” of humanity (cf. conf. 11.28.38) that he urged upon Firmus for the eternal good of the latter’s soul. It was for the sake of a reader thus mechanically empowered to appreciate and respond to the supreme non-fiction of God’s ordering of universal history that Augustine designed his books of the City of God.
1 O’Donnell Reference O’Donnell1992: 3.251.
2 Possidius, Vita Sancti Augustini, pref.
3 Marrou Reference Marrou1938: 66–67; but cf. Marrou Reference Marrou1949: 668–70, with Vessey Reference Vessey1999: 13–15.
4 For a model of the operations of “cultural memory” between “archive” or “reference memory” and “canon” or “working memory,” see Assmann Reference Assmann2008.
5 For Augustine’s thinking on this subject, and much of what follows in the next three paragraphs, see Fredriksen Reference Fredriksen2008.
6 Fredriksen Reference Fredriksen2008: 163.
7 Pollmann Reference Pollmann1996.
8 See O’Donnell Reference O’Donnell1992: xlii–xlv.
9 Stock Reference Stock1996: 207–12.
10 The expression “counter-memory” is used here in a sense nearly opposite to the Nietzschean one for which it was coined by Foucault 1977.
11 Marrou Reference Marrou1950; Markus Reference Markus1988; Markus Reference Markus2006: 31–48.
12 O’Daly Reference O’Daly1999a: 236–38 itemizes Augustine’s recourse to Varro in the City of God.
13 Shanzer Reference Shanzer2005.
14 The passage is highlighted by Clark Reference Clark2007 in the course of a valuable discussion of the “bookishness” of the City of God: Varro, according to Cicero, had made accessible the “‘societal archive’ of history and religion” that was constituted by the “buildings and monuments” of Rome (127).
15 For an introduction to the Chronicle in Jerome’s version, see Burgess Reference Burgess2002, and on the work as “literary history,” Vessey Reference Vessey2010. For the Latin text, see Helm (Reference Helm1984).
16 See further Vessey Reference Vessey2002: 59–64, 67–71.
17 Hurtado Reference Hurtado2006: 43–93 surveys evidence and interpretations for the earliest period.
18 Augustine, however, imagined the Jews of his time carrying their scriptures in codices; Petitmengin in AL 1: 1029–30.
19 See especially Jager Reference Jager2000: 33–38, with Holtz Reference Holtz1989, and Petitmengin in AL 1: 1022–32.
20 Clark Reference Clark2007: 120.
21 For the wider context in his thought and writing, see Vessey Reference Vessey1997.
22 Tornau Reference Tornau2006.
23 O’Donnell Reference O’Donnell1999: 99.
24 Grafton and Williams Reference Grafton and Williams2006: 133–77.
25 Jerome, Chron. pref. (ed. Helm) 5; Grafton and Williams Reference Grafton and Williams2006: 199.
26 Feeney Reference Feeney2007: 229, n. 110, cites this passage from civ. Dei 18.1 as evidence of Augustine’s “profound understanding of the mechanisms and implications of synchronism.”
27 Frye Reference Frye1957: 248; cf. 307.
28 Jager Reference Jager2000: 43.
