Part II - New Texts

5 New Bookforms

The Code(x)

I began Chapter 4 with the “aesthetic of accumulation” – the scholarly description of Theodosian Age scholastic production as both novel and in some ways peculiar, in which scholarly productions across domains came to be bound up in the “codification craze” in the period when Christians first became a ruling elite. I argued that the choices we see reflected in the scholastic production of the period are not purely or even largely aesthetic predilections, but rather that they signal the proliferation of a set of intellectual practices across diverse domains during the fourth and fifth centuries, namely the creation of aggregative codices with the potential to produce authoritative knowledge or which presented that universal authoritative knowledge itself. This is the sense in which the codes of the Theodosian Age differ from those which precede it and those which follow. Parchment codices had existed since at least the turn of the common era, but they were not “codes” in the sense that the Theodosian Code is a code, or the Christian bible is a code, or even the works of Macrobius can be understood as a code in the peculiarly Theodosian sense – as an aggregative work which presents the reader with an opportunity to grasp universal knowledge.

The codes engaged here all share a bookform: the codex. But homology obscures more than it enlightens. Etymologically “codification” refers to the transfer of a text into the codex format, but in contemporary usage it means something akin to “authorization”: the “codified” rules of football, for instance, are not simply those which are recorded in a codex, but those promulgated by the relevant authority and binding on sanctioned games. While the Theodosian Code and the Christian bible both circulated in codex form, the format did not itself lend authoritative status to a writing, even though, as I argue, it may have signaled a text’s authoritative status. This chapter attempts to untangle the reticulated categories of code and codex, and offer an account of how “codification” came to signify both “transfer to the codex format” and “authoritative promulgation of sources.” I argue that a confusion between the categories of code, codex, and codification have hamstrung attempts to understand some of the codes of the Theodosian Age, including the Theodosian Code and the fourth/fifth-century biblical pandects. This chapter addresses the confusion in preparation for a wider discussion to follow.

From an early period, Christians preferred the codex format for scriptural texts – mostly Greek editions of the Hebrew Bible – and for texts that would later be understood as scripture. The earliest fragment of the Gospel according to John, for instance, is a small scrap of papyrus from a codex leaf that was copied somewhere in the second or early third century.¹ While Christians preferred the codex format for some of their texts, they did not invent the format nor were they primarily responsible for popularizing it.² Parchment and papyrus codices are first extant from the second century ce, though literary attestation of the format begins somewhat earlier, with Martial’s Epigrams.³ The format appears to be Roman in conception, and its spread through the empire in the second through the fourth century has been proposed as a serviceable index of Romanization.⁴ The parchment codex, and its less prestigious cousin in papyrus, are modeled on an earlier instantiation of the form: they are plastic approximations of the wooden tabella, famed in applications ranging from legal promulgations (the so-called XII Tabulae) to the ritual inscription of temple boundaries.⁵ Elizabeth Meyer has demonstrated that the tabella was central to republican Roman ceremonial protocol and was involved in “the ordering of state, religion, magic, legal procedure, and some legal acts” and possessed “certain performative, almost magical, powers.”⁶ I have written elsewhere about the durability of what Matthew Larsen and I called “generic expectations”; ancient and late ancient readers considered the codex form to signal certain genre-inflected features of the writings contained.⁷ The “codification craze” of the late fourth century cannot be considered without acknowledgment of the historical table upon which it played out; the tabella was an object invested with potentiality for power in the Roman republic, and in the later empire the codex format retained some aspect of that ancient usage in the estimation of its users.

In their first three centuries of use, codices were not prestige objects. The format was primarily reserved for provisional writings, para-literary texts, and for work that was not yet “finished” and thereby ready for transfer to a prestige format like a bookroll or a bronze slab.⁸ By the Theodosian Age, however, the codex had been imbued with new associations. Not only was it the dominant format for all literary writing, but it was a prestige format associated with universal statements of truth that was often used to effectuate them. Christians stand in the gap between the use of the codex format for provisional and para-literary texts and the use of the codex for such monumental productions like the Theodosian Code(x).

Our earliest evidence suggests that Christians preferred the codex for their scriptural texts even while other texts such as homilies continued to be copied and circulated in roll format. For instance, we have P. Michigan 18.763, a homily containing significant New Testament quotations that is nevertheless preserved on a roll that was copied between 150 and 250 ce (TM 63857). On the other end of the temporal spectrum we have Princeton Garrett 24, a palimpsested rotulus whose undertext is a mēnaion with a Christmas homily dated paleographically to the eighth or ninth century (TM 63857). Throughout this period, Christians preferred the codex format for scriptural materials without any significant counter example, while less authoritative material enjoyed more flexibility when it comes to bookform.⁹

We can trace the Christianization of structures of power by following closely the shifting material expressions of power. When Nicene Christians came to widespread power in the Theodosian Age, armed with a novel set of scholastic practices and a canon of scripture that circulated in codex format as universally true, the peculiar Christian perspective on the codex format transferred to other universal statements of truth that took the same shape. The Christian practices of scholastic production that I have traced thus far can be followed further, into the physical instantiations of Theodosian Age works. “Code” and “codex” came to mean the same thing during this period; in the words of Martin Wallraff, “the utilization of this term [code(x)] is widely known as a story of great success – it caught on, and it led to an almost breathtaking semantic expansion.”¹⁰ The coalescence of “code” and “codex” into a single signifier is another effect of Christian ascendancy in the Theodosian Age.

Christians and the Codex

The “rise” of the codex was a slow process. Data from the Leuven Database of Ancient Books shows a slow debut starting in the late first century ce among extant manuscripts, most of which were found in Egypt. The format saw rapid adoption over the course of the third century, and the early fourth century witnesses two dramatic shifts: for the first time books on codices outpace rolls both in proportion and in total number of extant exempla (Figures 2 and 3).¹¹

Figure 2. Relative proportion of book formats, 350 bce–800 ce.

Figure 3. Books extant by format, 350 bce–800 ce.

The codex format was traditionally associated with para-literary texts: medical treatises, astronomical books, and provisional writing. By the fourth century it was also traditionally associated, among Christians, with scripture. It is precisely the moment of overlap, when extant codices overtake books on rolls, that ancient readers began to use the codex for a new purpose: not everyday writing or provisional texts but for deluxe editions and presentation copies.

The earliest attested deluxe parchment codices were both created for Constantine: one on his request and another as a gift. The gift was a presentation copy of poems by Optatian.¹² As literature, Optatian’s poems are altogether unremarkable, but his collection made ample use of the codex format to offer pictorial poems in various visual forms: an altar (Carmen 27), an organ (Carmen 20), etc. One poem dedicated to the Emperor Constantine bears a christogram across the center of the work along with “IESUS” outlined in red across the composition (Figure 4).

Figure 4. Bern, Burgerbibliothek, Cod. 212, f. 113v – composite manuscript: artes et carmina (www.e-codices.ch/en/list/one/bbb/0212). The complexity and visual nature of Optatian’s composition obviate any concern that this ninth-century copy is significantly different from the edition presented to Constantine in the early fourth century.

Constantine himself requested the other earliest attested deluxe edition of a parchment codex, and sometime after 335 ce, Eusebius’s Caesarean scriptorium carried out the work. In a letter to Eusebius, the emperor requested “fifty volumes with ornamental leather bindings, easily legible and convenient for portable use, to be copied by skilled calligraphists well trained in the art, copies that is of the divine scriptures.”¹³ Much has been made of this passage, especially since two fourth-century deluxe pandects remain to this day: the so-called codices Sinaiticus and Vaticanus.¹⁴ It is certainly possible that these texts are related to the imperial order for “sacred scriptures”; Theodore Skeat argued persuasively that a single scriptorium produced both manuscripts, and plausibly that they were produced under the direction of Eusebius for this very purpose.¹⁵ I find this analysis to be wishful, but only note here that Constantine’s order did not request pandects of the “sacred scriptures” like we have in these two manuscripts – comprising the entire canon, and then some – and there is some reason to believe that the pandect form was not typical, especially for bibles that were supposed to be “convenient for portable use” as Constantine’s letter requests.¹⁶ Sinaiticus, at least, is not.¹⁷

Canon and Codex

“Canon” and “codex” are easy to mix up. In fact, most scholars do mix them up, presuming that it is the act of binding texts together – codification – that confers “canonical” status on material between the covers. This causes problems, both conceptual and interpretive, that need to be addressed before I can offer an account of the material aspects Theodosian scholarship in Chapters 6 and 7.

The use of “canon (κανών)” language to delineate a group of authoritative Christian texts is likely a legacy of Eusebius of Caesarea, who invoked the term in both of its common Roman usages to mean both a measuring stick (and as such a Greek translation of the Latin regula), as well as in its technical sense to mean a set of tables.¹⁸ Tables, or tabellae in Latin, also enjoy dual usage, meaning either a codex format book, as discussed earlier, or precisely “tables” in the modern English sense: aligned lists of information relevant to a particular topic.¹⁹ The confusion among modern scholars between canon and codex arises out of these two words and their flexible usage in antiquity. We should not, however, presume that the modern confusion existed in the ancient world: it did not.

In antiquity, the canon of scripture was not a codex; it was a list. Consider Athanasius’s famous 39th Festal Letter from 367 ce, which delineated for the first time the precise bounds of Christian scripture that came to dominate Catholic and Orthodox Christianity in the Middle Ages. Athanasius did not offer to his fellow clergy a codex authoritative texts, but rather stipulated a list of books that are “canonized (κανονιζόμενα)” – that is, listed as authoritative – and another list of books that are not “canonized” but nevertheless may “be read outloud (ἀναγινωσκόμενα)” without objection.²⁰ Similarly, Canon 24 of Carthage, originally from the Council of Hippo in 393 ce, says that only texts from the canon can be read in churches “under the name of divine scripture.” Like Athanasius’s letter, it does not stipulate that only canonical materials can be read during services, but rather Canon 24 delineates the relative status of Christian documents that may well be used in preaching and catechesis.²¹ There is one dissenting voice: Canon 59 of the Council of Laodicea (ca. 365 ce) stipulates that only canonical texts can be read in a church setting. The following text from Laodicea, Canon 60, defines the bounds of the scriptural canon, but it comes with its own set of interpretive issues: its authenticity is dubious, at best. Further, at least according to Athanasius, books of the “Old Testament (παλαιὰ διαθήκη)” are intended to circulate in a particular “order (τάξις),” while the books of the New Testament are an unordered collection.²² The canon of the Hebrew Bible, according to Athanasius, was a pre-ordered list, while the canon of the New Testament was a collection of titles. Across the fourth century sources disagree on the extent, import, and implications of the “canon,” and the confusion did not let up in the fifth.

The slippage between categories of “codex” and “canon” so common in modernity does not occur in antiquity. As Martin Wallraff has persuasively argued, in antiquity writ large “where the bible was depicted, the thing depicted was not a book, but rather a bookshelf.”²³ Even Codex Amiatinus, produced around the turn of the eighth century, portrays the scribe Ezra rewriting the scriptures after the collapse of the Jerusalem Temple and specifically depicts the canon of the Hebrew Bible as a bookshelf of individual books rather than as a single codex.²⁴ It was perhaps Eusebius, “a Christian impresario of the codex,” who first extracted the medium of the codex from its common association with provisional and everyday writing (Fachliteratur) and ennobled it to use in prestige projects.²⁵ Grafton and Williams conclude: “If the chronological questions Eusebius and his anonymous helpers put were traditional, the answers he found glittered with methodological and formal novelty.”²⁶ Similarly, Wallraff argues, “With his magnificent staging of the Gospels Eusebius ennobled the medium of the codex, which had begun as a simple notebook and a shabby scratch pad, and definitively raised it to the rank of an archetype. A carefully produced gospel codex of Eusebius’s work shies away from no comparandum – in every respect: that of the sacred, the scholarly, or the aesthetic.”²⁷ No example of this transformation of the codex from “shabby scratch pad” to prestige object is more striking than the authority and pride of place afforded to the biblical codex at the councils of Ephesus (431) and Chalcedon (451), where the presence of a gospel codex was the sine qua non of valid proceedings.

At the Council of Ephesus in 431, the gospel codex was considered to stand in for Christ himself. At the beginning of the acta from this council, a gospel book is presented and the bishops in attendance come together “where the holy gospel lay in the midst of the throne, and presented Christ appearing among us.”²⁸ At the council of Constantinople in 449 litigants swore on the gospel book itself,²⁹ while at the Council of Chalcedon in 451 no session could commence without the presence of a gospel codex.³⁰ The acta of Chalcedon attend to the placement of gospel books with some regularity, in fact, and often repeat the requirement that a scriptural codex be present before proceedings. The fourth session begins typically, with a list of participants followed immediately by a description of the setting: “And when all had been seated before the railings of the most holy sanctuary, with the holy and undefiled gospel [book] having been brought to the center, the most glorious officials and the exalted assembly said: ‘So that we may decide what is to be done, let the decisions made in the previous hearings be read out.’”³¹ At the same session, confessions of faith could only be made in the presence of the “divine gospels.”³² During the tenth session, the gospel book itself signifies the authority of the speaker,³³ while during the eleventh session, Bassianus recounts a fight that he had at the altar in his episcopal see with Memnon the bishop of Ephesus that led to blood being shed on the gospel book itself because of its placement on the altar.³⁴ Again during the twelfth session, the acta record that the gospel book must be brought in before the session can commence.³⁵ The presentation of these copies of the sacred scripture at Theodosian imperial councils makes clear that the codex of scripture was an object of power itself, and it was emphatically not coterminous with the canon of scripture – everyone in the assembly agreed that the canon of scripture included books beyond the gospels, and yet it was a gospel codex which presented the power of the deity of material form.

In the Theodosian approach to scriptural codices we find a theology of bookish incarnation. In the words of Epiphanius of Salamis: “The acquisition of Christian books is necessary for those who can use them. For the mere sight of these books renders us less inclined to sin, and incites us to believe more firmly in righteousness.”³⁶ The codex had become a prestige object, capable of presenting the deity itself in time and in space. But the codices in these examples are not pandects like Sinaiticus or Vaticanus. Rather, they are gospel codices, containing presumably the four “canonized” gospels in a single codex, apart from the rest of the scriptural canon. Modern scholars confuse the data when we collapse canon and codex into a single signifier. And, as Wallraff notes, “nobody in Antiquity would have considered a gospel codex as a ‘partial edition’ of the New Testament.”³⁷ The examples here give voice to the fissure between canon and codex that must be appreciated before the great codices of Late Antiquity can be properly understood. The canon was a list of books – books that could be codified – but that was not specifically defined as that-which-lays-between-the-covers-of-a-codex. Any discussion which collapses the two categories will necessarily run into methodological and interpretive dead ends.

Sinaiticus, perhaps the most famous codex from antiquity, has itself suffered the conflation of codex with canon in its interpretation. In his influential article “The Codex Sinaiticus, the Codex Vaticanus, and Constantine,” Theodor Skeat jumps right over the question of whether these pandects are intended to be presentations of the canon of scripture, assuming that the covers themselves signal the canonicity of the books between. The article undertakes a long discussion of Athanasius’s canon list in order to justify the presence of two noncanonical works (the Epistle of Barnabas and the Shepherd of Hermas) in Sinaiticus alongside a relatively standard “canonical” collection of scriptural texts.³⁸ For Skeat, the presence of Barnabas and Hermas in a codex with the “canonical” texts suggests that, for the user of Sinaiticus, Barnabas and Hermas were “canonical” too. But this is to confuse the issue, and to presume wrongly that Athanasius’s canon was a codex. It was not, nor can the same be said for any other Christian of the fourth or fifth century. The presence of noncanonical material between the same covers as canonical material, even among Orthodox Christians in antiquity, was no cause for compunction. Collections of “Christian” and “classical” material are known in manuscripts as early (or late, depending on one’s perspective) as the fourth century – including, famously, the Bodmer Thucydides, an intact bifolium with a section of the biblical book of Daniel copied just before the beginning of book 6 of Thucydides.³⁹ This codex also included material from the biblical book of Susannah. Neither did ancient Christians display any concern about the status of the pandect’s conceptual opposite – namely, codices of scriptural texts that did not include the entire canon between its covers. Most late ancient scriptural texts were transmitted piecemeal. Consider, for instance, P. PalauRib Inv. 181–183, a Coptic codex of the late fifth century containing the Gospels of Luke, John, and Mark (TM 107760, 107904, 107905). Should we assume that this codex attests to a Christian community where only those three gospels were “canonical”? Athanasius wouldn’t make such an assumption, and neither should modern scholars. Likewise P. Bodmer 3 is a fourth-century Coptic codex containing only the Gospel of John and Genesis (TM 107758), while P. Bodmer 18 contains parts of the Gospel according to Matthew and Paul’s Letter to the Romans, and was copied in the second half of the fourth century or the first half of the fifth (TM 107759).

From the fourth and fifth century, not one example survives of a codex of Christian scripture that contains only the texts listed in any known canon from the period, including the famous fourth-century pandects that are so often hailed as ancient Christian bibles.⁴⁰ This insight necessarily complicates accounts like that of Robert Kraft, who claims that in Late Antiquity, “‘biblical canon’ took on a very concrete meaning in the shadow of the appearance of the Bible as a single book in codex form.”⁴¹ There is an added issue, discussed already in the late nineteenth century by Theodor Zahn: it is likely that the great fourth- and fifth-century pandects such as Sinaiticus and Vaticanus survive precisely because they were exceedingly difficult to use and, as a result, they were not.⁴² As Robbins argues, in the fourth century, pandects “were never more than curiosities.”⁴³

In the fourth and fifth centuries codices commanded respect and power, but they were not coterminous with canon. Skeat’s assumption of Athanasius’s “two categories” (κανονιζόμενα and ἀναγινωσκόμενα) itself breaks down, and Skeat admits as much, though without allowing that his analysis itself may be at issue. “Sinaiticus includes some which Athanasius does not include in either of his two categories, viz. 1 and 4 Maccabees in the Old Testament and the Epistle of Barnabas in the New.”⁴⁴ The fact is that most biblical books that we know from the fourth and fifth centuries are not pandects such as Sinaiticus or Vaticanus. And yet, most analyses of these pandects, and of Constantine’s request to Eusebius for fifty copies of “the sacred scriptures,” presume precisely that “scriptures” are those which are contained in a codex.⁴⁵ Skeat hastily jumps from the list of “holy scriptures” in Athanasius to the presumption that any request for such books would necessarily include all within the covers of a single codex.⁴⁶ Even Harry Gamble confused the issue, by justifying that the books dispatched to Constantine likely contained the four canonical gospels alone, on the basis that “the scope of the Christian Bible was still variable in the early fourth century.”⁴⁷ It may well be the case that the “sacred scriptures” dispatched to Constantine contained only gospels.⁴⁸ But the reason that this is possible is not because the canon was underdetermined. The reason is that a codex did not contain the canon, and the bible was not a book.

These examples show that in antiquity, the act of binding texts together into a codex did not render them “canonized.” Further, these examples demonstrate that scriptural codices possessed an excess of vitality, beyond their function as inert vehicles for the transmission of text.⁴⁹ In the impossibly elegant words of Martin Wallraff:

The late-antique book thus achieves a depth of meaning that extends far beyond the function of writing and reading. It is more than text carrier. There is an excess of meaning, of effort, of medial impact, which transcends the contained and transmitted text and does not exhaust itself through reading. The book not only contains letters (Zeichen), but it becomes a sign (Zeichen) itself.⁵⁰

For many Theodosian Age productions the codex form itself signaled the authority of the materials presented within. Speaking about the Theodosian Code, Serena Ammirati argued that during the Theodosian Age, “both the law of God and the law of people need to be put into writing, and their ‘scriptural’ authority receives external confirmation from the idea of authority intimately connected with the new format [of the codex].”⁵¹ Ammirati goes further, arguing that even the choice of the uncial script – the same as was used in contemporary scriptural codices – signaled to the reader the universalizing aims of the Theodosian legal codification.⁵² Nowhere is the material expression of power in codex form more clearly visible than in the Acts of the Roman Senate Concerning the Theodosian Promulgation (Gesta senatus Romani de Theodosiano publicando). The Acts record the presentation of the Theodosian Code in the West by Faustus the Pretorian Prefect, during a meeting of the senate at his private residence in Rome late in 438.⁵³ The details of this fascinating document cannot hold me here, except to say that the Roman senate met continuously for almost a thousand years, and this is among the only transcripts of actual senatorial proceedings that remain extant.⁵⁴ As we saw earlier at the Council of Chalcedon, the meeting began with a call to order and a reading from what is consistently referred to as a “consecrated” book given by the emperor’s “divine hand (manu divina).”⁵⁵

The consecrated book in question was the Theodosian Code. The Gesta reads: “The Code(x) was received into our hands, as directed by the order of both emperors … they ordered that this undertaking should be performed in order that we may obey with proper devotion the most carefully considered precepts of the immortal emperors.”⁵⁶ One section of the book was read – Theodosian Code 1.1.5 – in order that the assembly might know the intention of the codification program, namely the creation of an aggregative scholarly resource which could serve as the basis for a further, universal “guide to life,” about which I have written more in the Appendix.⁵⁷ A collection of forty-six exclamations follows, ranging from the general (“May it please our Augusti to live forever! Repeated 22 times”)⁵⁸ to the specific (“Let the codes be copied and dispatched to the provinces! Repeated 11 times” (5, 3.16); “We request that the codes be kept in the public archives! Repeated 15 times” (5, 3.20); “We request that you make a report to the emperor about the desires of the senate! Repeated 20 times” (5, 3.31). Many exclamations confirm the extraordinary status of the object of the Code(x) itself – it was to be emblazoned with the seal of the prefects in whose office copies are kept, and many copies of the codices are to be made “in order that the established laws may not be changed” (5, 3.8).

And yet, while each copy of the Codex was intended to stand in for the divine authority of the emperor himself, the prestige of the object diminished as its text was transmitted – copies of the manuscript, even if identical, did not retain the special status as the original codex. As child nodes receded further from the original product presented at the wedding of Western emperor Valentinian III to Licinia Eudoxia (daughter of Eastern emperor Theodosius II), the status of the object changed.⁵⁹ When presenting the Western senate with this prestige object from the Eastern court, Faustus ordered copies of the codex to be made in three distinct groupings (corpora): the first was a copy brought from the East and presented to the Senate, which was to stay under lock and key in the archives of the Pretorian Prefect. Another copy, part of a different corpus, was to be sent to the archives of the Urban Prefect, while a third copy comprising a third corpus was to be entrusted to two specially chosen constitutionarii who were tasked with personally transcribing every published copy of the Code, including one to be sent to the province of Africa (Figure 5).⁶⁰ Faustus’s declamation is clear: each copy of the codex has a different, and diminishing, status, though they are all equally authoritative and though all copies are to be made by the constitutionarii “in their own hand (eorundem manu).”⁶¹ Our earliest surviving manuscript of the Code is Vat. Reg. Lat. 886, from the late fifth or early sixth century, and it shows clear signs both of being descended from an exemplar of “corpus 3,” as well as being a private copy. The inclusion of explanatory marginal notes in particular suggests that this cannot have been an official copy of the Code, and further that this manuscript did not command the awe evidenced in the senatorial reaction to the Code’s presentation in 438; it wouldn’t have been annotated otherwise.

Figure 5. Stemma of Theodosian Codices described in Gesta Senatus 7.

Chart adapted from Matthews, Laying Down the Law, 51.

The Code presented to the Senate in 438 was sacred. Its status was reiterated in a number of novellae promulgated by Theodosius II after the publication of the Code, and collected in a dossier dispatched to the Western court in 447.⁶² In his first novella (or “new law”), Theodosius II made provision for subsequent additions to the body of imperial general law, but always with reference to the version of the Code that was “kept in the sacred imperial archives (sacris habentur in scriniis).” Laws not officially added to the Code(x) were to be considered forgeries.⁶³ In other words, the Theodosian Codex itself, and not strictly the text that it contained, was an object of power and the singular locus of authority in the later Theodosian empire. Copies could be made and the content of the Theodosian Code could morph as necessitated by the continuing needs of a functioning imperial apparatus. But the object itself – its very materiality in physical form as a codex – remained the central focus of authority.

Conclusion

In antiquity the codex was pluripotent: it could heal the sick, drive away sin, invite Christ incarnate to an imperially sanctioned debate, and present the authority of the emperor at a distance. Codices were utilitarian receptacles of information, but to view them solely as such is to fail to grasp the profound political and cosmic significance that became attached to the objects themselves.⁶⁴ By the Theodosian Age the codex had completed its metamorphosis from caterpillar to butterfly. It was no longer a “shabby scratch pad.” It was capable of any number of miraculous deeds, and it was a sign itself of authority and religious sanction. Its larval stage can be seen in the exaltation of a lowly form by Christians, beginning with Constantine himself. By the time that Christians stalked the halls of power and created new, universal legal regimes as we find in the Theodosian Code, the codex had become the code – a symbol in and of itself.

6 New Texts

The period between the modern imperial scramble for Africa and the strongest phase of its decolonization spanned the lifetime of a single man: Winston Churchill (1874–1965).¹ The greatest part of the European colonial enterprise, and the most thoroughgoing reaction against it, occurred in a period roughly the same duration as the Theodosian dynasty. Profound societal change in the modern era does not occur only in the longue durée, and late ancient life sometimes proceeded at a breakneck pace.

In February of 380 ce, the emperors Gratian, Valentinian II, and Theodosius I decreed that “all peoples who are ruled by the administration of our forbearance shall practice that religion which the divine Peter the Apostle transmitted to the Romans.”² According to a constitution preserved in book 16 of the Theodosian Code, this “religion” was preached by the Pontifex Damasus and by Peter the Bishop of Alexandria, and the constitution stipulates that “we shall believe in the single Deity of the Father, the Son, and the Holy Spirit, under the concept of equal majesty and of the Holy Trinity.”³ According to the emperors, this particular statement of was more than just something that “we must believe (credamus)”: it was a law (lex), and obedience to the law granted one the right to be called by “the name ‘Catholic Christian.’”⁴ Dramatic and systemic changes characterize the Theodosian Age, beginning with the elevation of the Nicene confession of faith to the status of law.

Historians attempting to trace the rise of Christianity in the second and third centuries necessarily take a somewhat speculative approach to the evidence, given its relative scarcity and the inherent difficulty in relating fragmentary data to a political and social environment that was intermittently hostile. The late fourth-century revolution in governing ideology and in scholastic methodology, on the other hand, is well documented and was active throughout the period under discussion. New forms of knowledge production arose rapidly in the Theodosian dynasty, and scholars reacted and responded to these new forms of knowledge production in real time. In Chapters 3 and 4 I investigated literary sources in order to trace the proliferation of Nicene Christian methods through Theodosian Age scholarly productions. This chapter turns to the material evidence for many of those same literary sources: to manuscripts, and to the ways that norms, creeds, and laws were aggregated, distilled, and promulgated. By the time that scribes copied most of our earliest extant manuscript evidence – by the late Theodosian Age – new forms of argumentation and compilation had already suffused a scholastic landscape over which the predominance of codes cast a long shadow. This chapter picks up there, where I left off at the end of Chapter 5.

Gratian’s Talisman

In the summer of 378 the emperor Gratian was nineteen or twenty years old, and he was going to war. Gothic tribes had invaded Thrace, a “countless horde that had taken possession of the mountain heights as well as the plains,” and before leaving the Western court for the field of battle the young emperor requested a talisman in the form of a codex from Ambrose, the bishop of Milan.⁵ Ambrose warned that the “book about the faith (fidei libellum)” which he delivered to the emperor would not be up to the task of “acting as an adjudication of the faith (de fidei disceptandi),” but rather was intended as a “collation of a multitude of opinions (de testimoniis plura contexam),”⁶ and would satisfy the emperor’s needs for a talisman even though, he added, the Nicene Creed itself already was “just like a trophy (tropaeum), raised to proclaim victory over faithless ones throughout the world.”⁷ As Ambrose well knew, Constantine himself marched to battle with a tropaeum – a “trophy of a cross in light (σταυροῦ τρόπαιον ἐκ φωτός)” that god revealed to Constantine before his battle at the Milvian Bridge, in Eusebius’s account.⁸ Gratian did not request a gem for his dangerous journey, similar to what many Christians wore in this period for health and for safety,⁹ nor did he order his soldiers to affix a Christian symbol to their shields to invoke divine power and protection, as Constantine had done some fifty years before.¹⁰ The young emperor asked for a codex of scholastic opinions on a theological question that had been adjudicated thirty-four years before his birth.¹¹

Gratian’s request for a book of opinions as a protective amulet is utterly bizarre from the point of view of even the earlier fourth century. Ambrose regarded the Nicene Creed itself as a “trophy,” but the young emperor requested a book – a book of scholarship – for his talisman instead. His request betrays the extent to which the aggregative codex had become a symbol of power and divine guidance for Nicene Christians who were, as I have argued, peculiarly bookish, and invested in the production of textualized, aggregative truth.

It is the historian’s good luck that two copies of Gratian’s talisman remain extant, both produced during the Theodosian Age. The literary qualities of this text are remarkable in and of themselves; in the earlier quotation Ambrose suggested that the collation of opinions was only the first step toward the production of universal truth. In other words, he takes part in a structure of knowledge known from a wide variety of other Theodosian Age productions that I detailed in Chapters 3 and 4.¹² Yet the material qualities of these two manuscripts demonstrate most vividly the extraordinary coherence of the Theodosian structure of knowledge in which Ambrose’s book takes part. Ambrose indicated that aggregation was the first step toward the production of universal truth, but he was also concerned with the effects of such aggregation. We learn from these two manuscripts that his earliest readers held similar concerns about the effect of placing heretical and orthodox opinions side by side.

Ambrose: Concerning the Faith

In Concerning the Faith, Ambrose implements an aggregative structure of knowledge, but he also concerns himself with the problem that this manner of argumentation poses for a reader trying to extract truth from the fray; he concerns himself with what I call the “problem of discernment.” A structure of knowledge in which any truth claim must be based on an aggregation of the sources also must employ a manner of deciding between opposing sources. For Ambrose, this problem was acute. In Concerning the Faith he claims that his predecessors, and scripture itself, demanded that “impious doctrines should be included in the record of their decrees,” but Ambrose laments that a credulous reader may accidentally stumble into heresy on account of this requirement:

So of course our fathers spoke following the guidance of the Scriptures, holding that impious doctrines (sacrilega dogmata) should be included in the record of their decrees in order that the unbelief of Arius should discover itself, and not so as to hide itself with red-blush (fucis), or with dye. Those who don’t dare explicate what they think are in fact carrying out a fraud (fucum). The impiety of Arius is not propagated through exposition, like in the censor’s books. Rather it is exposed [as heresy] through condemnation, such that the curious person eager to hear won’t fall into error, because he knows already that it is condemned, before he hears, in order that he might believe.¹³

This passage immediately precedes Arius’s heretical statement of faith. Ambrose admits that the dominant scholarly method requires him to include heretical statements in his work, and that those who fail to say openly what they think thereby render their thoughts “deceitful” – playing in a particularly Ambrosian manner with the dual meaning of fucus, as both “red blush” and “deceit.” In order to expose heresies as such, and not to allow them to hide as if wearing makeup, Ambrose names and condemns heretical opinions even before they appear in his text “such that the curious person eager to hear won’t fall into error, because he knows already that it is condemned, before he hears” (1.18.119). Ambrose was willing to hew to the Nicene manner of argumentation but he wanted to save his reader from falling into the trap of heresy when they read impious doctrines as part of his text. In order to combat heresy, Ambrose amply warns the reader that what is to come should not be trusted.

The earliest known reader of Ambrose’s Concerning the Faith was a scribe working from Italy in early to mid-fifth century. The scribe must have found Ambrose’s warning compelling, and perhaps insufficient, because while copying this text the scribe employed another set of mechanisms to make abundantly clear to any reader that the text they were about to read was dangerous. Perhaps taking a cue from Ambrose’s warning that heresy is sometimes disguised “with red blush” (fucis), the scribe of Stiftsbibliothek Lavanttal 1 marked out the heresy that follows with the addition of two words in red ink: Expositio Arii – “Arius’s statement of faith” (Figure 6).

Figure 6. Stiftsbibliothek Lavanttal 1, 34r (TM 67637, CLA 1450). The scribal addition to Ambrose’s text, reading “Expositio Arii,” is rendered in red ink, different from the brown used for the base text.

Images graciously provided by the Stiftsbibliothek St. Paulus in Lavanttal, Austria.

This is the earliest extant manuscript of Ambrose’s Concerning the Faith, and aside from incipits and explicits, its scribe uses red ink only to mark off heresy. In other words, this deluxe Theodosian manuscript uses red ink to deal with the problem of discernment occasioned by an expectation of aggregation as the proper form of scholastic knowledge production. As in many other Theodosian Age manuscripts, the incipits of each section are written in three lines of red. Such use of red ink was well established by the fifth century, as we see in, for instance, a (palimpsested) scholion on Cicero for which the scribe used red for the first three lines of each book and for words rendered in Greek.¹⁴ Red ink was commonly used for identifying to the reader something in the text requiring emphasis. In this fifth-century copy of Ambrose’s Concerning the Faith, it was used to mark off heresy. As is clear in Figure 6, a later corrector (likely Abbot Hartmut of St. Gall) crossed out the warning because, strictly speaking, it is not part of Ambrose’s text.¹⁵ Rather, I argue that it is an addition made by this Theodosian scribe in response to the problem of discernment.

In this manuscript all heretical statements are indicated in a similar manner, with material included for the purpose of aggregation marked off on either side by red uncial lettering. See, for instance, Figure 7, where the scribe warns that the following statement comes from arch-heretic Arius by writing Expositio dogmatis Arriani in red ink. See also Figure 8, where the scribe ends another heretical statement with a red uncial title reading “On the Eternal Son of God (De Sempiterno d(e)i Filio),” indicating to the reader that each of the heretical statements in the previous section would be refuted thereafter.¹⁶ Again, this title appears to be a scribal gloss and not part of Ambrose’s initial text. This scribe appears to have added these additional warnings for the same reason that Ambrose added the initial ones: to save the credulous reader from falling into heresy.

Figure 7. Stiftsbibliothek Lavanttal 1, 10v. The scribal gloss “Expositio dogmatis Arriani” appears in red ink, different from the brown used for the base text.

Figure 8. Stiftsbibliothek Lavanttal 1, 15r, with the scribal gloss “De Sempiterno d(e)i Filio” in red ink.

I have argued that across the scholarly landscape of the Theodosian Age we see a dominant method featuring arguments based on aggregation. This earliest manuscript of Ambrose shows that already in the Theodosian Age, concerns stemming from this method made the leap from text to paratext, and influenced the way that aggregative scholarship looked on a manuscript page. New readers make new texts; here we have one such new text. Throughout the fifth century, however, scholars and scribes continued to engage the problem of discernment, and scribes responded to the same problems in divergent ways. We can glimpse the variety of responses with high fidelity when contemporaneous manuscripts survive of the same text. In the case of Ambrose’s Concerning the Faith, we are lucky to have two manuscripts from the Theodosian Age, both produced in Italy, so far as paleography and codicology can attest.¹⁷ It appears that both scribes were cognizant of the issues related to aggregation and discernment, but that each responded to the issue in a slightly different manner. While the scribe of the Lavanttal manuscript uses red uncial lettering only to mark off heresy, the scribe of Paris Latin 8907 supplemented Ambrose’s text in their own way, using red throughout the manuscript to mark off heretical and orthodox creeds alike.¹⁸ For instance, in the transition between Ambrose’s preface and the Orthodox creed that follows, the scribe of the Paris manuscript added Expositio Fidei (“statement of faith”) in red letters, making it abundantly clear that what follows is an authorized creed (Figure 9). We can see the Paris scribe’s agency in this incursion into Ambrose’s text, and guess at their intention, by comparing it with the same passage in the Lavanttal copy. The Lavanttal manuscript witnesses the same base text as the Paris copy, but only the Paris scribe added Expositio Fidei in red ink to indicate that the pietatis exemplum which follows is an orthodox creed (Figure 10).

Figure 9. Paris, Bibliothèque Nationale Lat. 8907, 298v.

(source: gallica.bnf.fr / BnF). The scribal gloss “Expositio Fidei” appears in red ink, different from the brown ink used for the base text

Figure 10. Stiftsbibliothek Lavanttal 1, 4r, which notes the coming pietatis exe(m)plum without recourse to red ink, which is reserved in this manuscript for heretical creeds and the incipits and explicits of books.

Look carefully at Figures 9 and 10. Notice in both manuscripts the rounded D of a single strike, the high hasta and closed eye of the E, the L that rises above the line, and the calligraphic, oversized A with a pointed bow. These two scribes received remarkably similar training, and the hands must be dated to the same paleographic period. The scribes received similar training, but each responded in a distinctive way to the problem of aggregation that placed orthodox and heretical statements side by side.

Between chapters 17 and 18, too, the scribe of the Paris manuscript has inserted Definitio patrum de fide in red uncial before the Orthodox statement of faith (Figure 11). On the other hand, the Lavanttal scribe left a blank space at the same point in the text (Figure 12). Perhaps the Lavanttal scribe left this gap so that a later reader could add the “Definitio patrum de fide” witnessed in the Paris copy if they so desired. The introduction of a space between these chapters in the Lavanttal manuscript suggests that the scribe knew of the clarifying addition but did not find it necessary to identify orthodox creeds with clarifying additions in red ink, as they did for heretical creeds.

Figure 11. Paris, Bibliothèque Nationale Lat. 8907, 315r.

(source: gallica.bnf.fr / BnF)

Figure 12. Stiftsbibliothek Lavanttal 1, 33v. A corrector has struck out the M on the end of disimile in Lavanttal, and added an N at the end of lumen in what appears to be a contemporary half-uncial hand. The thinner brown lines filling the blank space are reminiscent of the strikethrough on 34r (Figure 6), and are likely the result of a ninth-century reader intending to clarify that the extra space should not be used for a clarifying insertion – perhaps precisely the clarifying insertion found at this point in BnF Lat. 8907, 315r reading definitio patrum de fide (Figure 11).

The scribes of Paris Latin 8907 and Stiftsbibliothek Lavanttal 1 both copied Ambrose’s Concerning the Decrees in Italy at about the same time,¹⁹ and both scribes deal with the same problem: the problem of including heretical material in an orthodox production.²⁰ Yet each arrived at a slightly different solution. The difference between these solutions suggests that these scribes were aware of the issue of discernment and that the issue remained unresolved in late ancient scriptoria, just as it was in scholarly salons of the Theodosian Age.

Hilary: Concerning the Synods

Hilary of Poitiers was exiled to Phrygia in 356. In 358 he wrote back to his colleagues in the West about the controversies embroiling the Greek-speaking empire and the creed that had been decided twenty years earlier as an attempt to settle the matter. Hilary’s letter is remarkable evidence that the Nicene Creed, a central focus of theological dispute in the Eastern empire almost since its creation in 325, was largely unknown in the West until at least the 360s. The earliest manuscript of Concerning the Synods was copied in 509. In this manuscript, too, scribes introduced paratextual solutions are introduced to deal with the problem of discernment.

Hilary presents a heretical creed in chapter 11 of Concerning the Synods with an explicit, textual notification: Exemplum blasphemiae apud Syrmium per Osium Potamium conscriptae (Figure 13). This part of the text is marked out with ekthesis, as is the heretical creed on the next page, beginning with Unum constat (Figure 14).

Figure 13. Vat. Arch.Cap.S.Pietro.D.182, 303r, where a heretical creed is signaled with ekthesis and a textual note, reading Exemplum blasphemiae apud Syrmium per Osium Potamium conscriptae.

Figure 14. Vat. Arch.Cap.S.Pietro.D.182, 303v, where a heretical creed is identified with ekthesis, an obelus in the upper left, and the text Exemp(lum) blasph(emiae).

The same scribe identified the end of the heretical creed with another paratextual feature, this time writing finit blasphemia in oversized capitals; the scribe changed scripts to clarify to the reader that what precedes is heretical and what follows is approved (Figure 15). A later reader (though likely one and the same scriptorium as the original scribe) added an arrow at the beginning of the heretical statement with a note reiterating that the following was an Exemp(lum) blasph(emiae): “a sample of heresy” which is present only for the purposes of aggregation. This marginal note transparently mirrors Finit blasphemia on the following page, and reiterates, yet again, that the intervening text should not be mistaken for truth.

Figure 15. Vat. Arch.Cap.S.Pietro.D.182, 304r. The end of a heretical creed, identified by the initial scribe with Finit blasphemia in oversized capitals, to which a later scribe added an obelus on the right, and the word pessima in the margin.

Apparently, however, even this much paratextual forewarning was eventually deemed insufficient, because a later reader added an obelus at the beginning and the end of the passage, in addition to the word pessima: “totally wicked.” This further addition apparently updated the manuscript for different conventions of notification, but it carried the same message: what is here is heresy.²¹ Already during the Theodosian Age, the problem of discernment led writers and readers to take prophylactic steps to stem the effect of heresy that was present in orthodox manuscripts. This manuscript shows that the problem remained in the minds, and the marginalia, of later readers too.

Manuscripts from the fifth century show a variety of textual and paratextual solutions to the problem of discernment. There is another early manuscript of Hilary’s Concerning the Synods, copied likely during the Ostrogothic period and thus shortly after the end of the Theodosian Age (LDAB 7924). This manuscript shows a different method of dealing with the inclusion of heretical materials in Hilary’s text. According to Lowe, the manuscript was “written doubtless in Italy by a master scribe in a scriptorium maintaining high standards,”²² and it marks out the heretical creed from Sirmium in 357 as we saw earlier: with ekthesis in addition to the text Exemplum blasphemiae apud Sirmium per Ossium et Putamium conscribtae, “a copy of the heresy of Sirmium written by Ossius and Potamius” (Figure 16).²³

Figure 16. Paris, Bibliothèque Nationale Lat. 2630, 320r. Concerning the Synods 11.

(source: gallica.bnf.fr / BnF)²⁴

After this textual warning, the same scribe identifies heresy by indenting (eisthesis) heretical statements further than the rest of the text (Figure 17). Orthodox statements, on the other hand, are not indented in this manuscript. See, for instance, the orthodox statement given at Sirmium against the heresies of Photius. It is introduced as a statement of faith in the same way as the heretical ones (in this case, with Exemplum fidei Sirmio ab Orientalib(us) contra Fotinum scribtae), but it is not indented (Figure 18). In this manuscript, extra indentation is reserved for heresy, alerting the reader to the status of the text through the material form of its presentation.

Figure 17. Paris, Bibliothèque Nationale Lat. 2630, 322r. Concerning the Synods 15.

(source: gallica.bnf.fr / BnF)

Figure 18. Paris, Bibliothèque Nationale Lat. 2630, 335r. Concerning the Synods 38.

(source: gallica.bnf.fr / BnF)

Before the fourth century, paratextual intervention in Christian scholarship is vanishingly rare. The earliest Christian manuscripts have little by way of paratextual markup, and nearly all that do have paratexts indicating something about the status of the text date to the period after the Council of Nicaea. A curious exception is P.Oxy 3.405 , a late second- or early third-century copy of Irenaeus’s Against Heresies with “wedge-shaped signs in the margin similar to those employed for filling up short lines.”²⁵ In this case, the scribe used wedges typically employed by Egyptian scholiasts on the right margin of the recto, which appear to indicate a quotation from the Gospel according to Matthew 3:16–17.²⁶ The relative paucity of evidence means that it is hard to say anything conclusive about the extent of this particular paratextual feature in Christian theological scholarship before the fourth century. The scribe of this manuscript, however, found value in using small wedges to mark out at least this one quotation in Irenaeus’s text. We can presume that other places in this manuscript would have quotations similarly marked out, though it is impossible to say whether the motif was used for texts that were considered authoritative, for all quotations, or in some other capacity.

Already in the late second or early third century, Tertullian suggested paratextual solutions to problems occasioned by his own form of theological scholarship, particularly when discussing Greek cosmological texts and ideas in Latin. In Against the Valentinians, he notifies readers that “for some of the names, a translation from the Greek does not bring out the appropriate force of the name. For others, the gender of the word in the two languages does not match. Finally, we are more used to the citation of others untranslated.”²⁷ Tertullian offers his solution in the next sentence: “For the most part, then, we shall use the Greek names; their meanings will be noted on the margins of the page. Nor will the Greek be unaccompanied by Latin equivalents. Rather, such will be marked with supralinear strokes – because explaining the personal names is made necessary by the ambiguities of some of them, which suggest some different meaning.”²⁸ Because the earliest extant manuscript of Tertullian’s text is medieval, we do not know the precise form of that his paratextual markup took in third-century manuscripts, nor the extent to which ancient copies employed these tools.²⁹ What is clear, however, is that he used paratextual solutions to engage a theological problem – in this case, the problem of explicating Greek cosmology in Latin. Apparently Tertullian did not expect his reader to recognize these paratextual features, or how they were employed, without some explanation. Without further manuscript evidence we cannot tell whether he was using common in-text signs to signify paratextual features (perhaps a generalized version of the Aristarchan system) or whether he invented a new group of signs himself. Tertullian’s use of paratextual features in Christian scholarship substantially predates the Theodosian Age, but his rationale for marking up his margins is dramatically different from what we see, for instance, in the work of Jerome, to which I now turn. Tertullian needed to clarify issues of translation. Jerome was concerned with discernment.

Jerome’s Obelus

Jerome was prolific. He wrote commentaries on nearly every part of the canonical bible and left a vast trove of letters and other theological works to set alongside his most enduring accomplishment: a full translation of his bible from Hebrew and Greek into Latin. He was intimately familiar with his sources, with the problem of textual variation within authoritative texts, and he was regularly in contact with Roman Traditionalists in the imperial capitols as well as Jewish teachers in Palestine. And yet, for all his learning and famed reclusive irascibility, Jerome possessed a single point of scholarly humility. Before offering the final word on any particular topic, he would first aggregate all of the relevant sources, regardless of whether they were capable of edification.

In the preface to his translation of the Book of Job, Jerome claims that aggregation is the proper form in which to present scholastic arguments, and also that paratextual accommodations can deal with problems of discernment. He undertook to study the various Greek translations included in Origen’s Hexapla and made a text critical investigation into the relationship between Hebrew and the Greek manuscripts, which differ by “almost seven hundred or eight hundred verses.”³⁰ Confident in the quality of his work, Jerome went so far as to boast, “let whoever wishes keep the old books, either written on purple skins with gold and silver, or in uncial letters (as they commonly say: ‘loads of writing’ rather than books), while they leave me and mine to have poor little leaves, and not such beautiful books as correct ones (non tam pulchros codices, quam emendatos).”³¹ The boast might seem hollow, or at least exaggerated, were it not for a manuscript of precisely this type and time period remaining extant: the so-called Codex Veronensis, a fifth-century codex that fits Jerome’s mocking description of the most gauche Theodosian Age bibles.³² The fifth-century codex comprises 388 folia dyed purple, with Latin uncial of a Theodosian type written in silver and gold ink: silver for the text, gold for the first lines of each gospel, nomina sacra, and the Lord’s prayer.

Jerome offered a translation from the Septuagint and the Hebrew in his Old Testament, and from the Greek in the New. In his estimation the rendering was of the highest quality, yet his translation is not meant to stand alone, nor to be read exclusively at the expense of “loads of writing”: deluxe copies of lesser scholastic productions. Rather, Jerome purposefully left it to the studious reader to decide the best reading in each case: “Each edition – both the Septuagint according to the Greeks, and mine according to the Hebrews – has been translated into Latin by my labor. May each one choose what he will, and prove himself studious rather than malevolent.”³³

His uncompromising text critical work brought Jerome into contact with Origen’s Hexapla, which offered two columns of Jewish scriptures in Hebrew and four in later Greek translations, including a second century translation by Theodotion.³⁴ Origen and Theodotion both repurposed two Aristarchan signs – the asterisk and the obelus – to offer a transparent collation of translations, indicating places where the Hebrew base text did not match up with Greek translations. Jerome was familiar with Origen’s polyglot edition and explicitly claims to benefit from the addition of asterisks and obeli, which Sebastian Brock points out were not used in a precisely text critical sense.³⁵ Origen was not concerned with establishing the “original text”; his interests were apologetic, “providing the Christian controversialist with a text that would be acceptable in the authoritative eyes of contemporary Jewish scholars.”³⁶ For his own work, Jerome did not simply pick up the Aristarchan asterisk and obelus and employ them in his translation to clarify the text, either. Rather, he resuscitated generally disused tools that he knew from Greek and Latin poetry and repurposed them to apologetic ends – to deal with the problem of Christians entering into theological debates with other biblically minded communities whose texts differ from Jerome’s own, superior edition.³⁷

Already in the early second century, Suetonius made clear that Roman scholars had employed the Aristarchan signs for quite some time and to various ends.³⁸ Even in antiquity, the Aristarchan sigla were generalized tools without strictly circumscribed significations. As shown by Kathleen McNamee, “none of these sigla had a tightly restricted significance, and (outside Oxyrhynchus and the second and third centuries) the same jobs were also done by various other sigla. The most useful reminder, for editors, that the meaning of these signs did vary is inconsistent use by scribes of even the very specialized sigla of the system of Aristarchus – and the toleration of those inconsistencies by readers.”³⁹ In the Herculaneum and Oxyrhynchus papyri, Aristarchan signs generally function as rudimentary hypertexts, pointing simply to the existence of a separate commentary or hupomnēma.⁴⁰ Jerome, too, picked up old tools and employed them to solve new problems, motivated by a new focus on aggregation.⁴¹

The Problem of Discernment in Nontheological Texts

The problem of discernment was not restricted to Christian scholastic productions during the Theodosian Age. The Theodosian Code was meant to be an aggregative work, bringing together both in-force laws and laws “which had fallen into disuse” as the basis for a future code that could serve as a “guide to life (magisterium vitae)” – one worthy to bear the name of the emperor from whose court it arose (1.1.5). Contemporary scholars of Roman law question whether the Theodosian Code as we have it – that is, roughly the project as proposed in CTh 1.1.6 – also includes disused laws, as the initial codification was intended to have. As promulgated, the Theodosian Code apparently did include disused laws, but a full discussion is beyond the scope of this chapter.⁴² My interest is not in the precise legal force of the Code but in its stated intentions, framing, and reception. Early commentaries on the Theodosian Code clearly stress that the collection was intended as a scholarly resource, and explicitly included laws that were no longer in force: as I suggest, laws that were present for the scholastic purpose of aggregation.

The so-called Summaria antiqua codicis Theodosiani is a marginal commentary on the Theodosian Code, written sometime in the fifth century and extant in the margins of a sixth-century manuscript of the Code now housed at the Vatican Library.⁴³ The semi-cursive uncial of the base text gives the impression that it was a private copy,⁴⁴ and the presence of staurograms at the beginning of each fascicle, along with the Latin letters rendered in with a stylus cut for Greek, suggests strongly that the text was copied in the Eastern empire, likely Constantinople (Figure 19).⁴⁵ The base text and the marginal commentary were copied in the early sixth century, but both are products of the fifth.⁴⁶ The marginal commentary reflects an attempt to categorize and interpret the Theodosian Code, with over a thousand numbered and cross-referenced entries which have the effect of negotiating the fraught relationship between the base text as an imperially promulgated code of law alongside its creation in an aggregative format, as a scholarly resource for diligentiores: “more industrious people.”⁴⁷

Figure 19. Vat. Reg. Lat. 886, 17r, with a staurogram at the upper left denoting the beginning of the fascicle and a marginal note on the right, appearing darker due to use of reagents to reveal the faint brown ink, likely by Angelo Mai. The base text is CTh 9.10.2–3. The strange appearance of the Latin text is due to the letters being rendered with a pen cut for Greek.⁴⁹

Throughout Vat. Reg. Lat. 886, the only ancient copy of the Theodosian Code in books 9–16, a sixth-century scribe copied clarifying notes out of the fifth-century Summaria antiqua, over the entire length of the 448 folia noting important details such as: “this is no longer in force (haec inutilis est),” or “this is ancient, and does not hold in this [current] period (haec antiqua est et non tenet his temporibus)” (Figure 20). The most common type of marginal note in this manuscript indicates that the statute is no longer in use, that it is similar to another constitution (similis followed by the number of the note, or superiori similis), or that it contradicts another constitution (generally contaria superiori or contraria, followed by the number of the note). Of the 1,230 scholia on these 8 books of the Theodosian Code, fully 25 percent are of this final type, indicating a law in the collection supervened by another in the collection.⁴⁸ For instance, the scholion on CTh 16.10.1 reads “this is no longer in force (haec inutilis est)” next to Constantine’s famous provision that in the event of lightning strikes on public buildings, “the observance of the ancient custom shall be retained,” namely that a haruspex should be consulted to interpret the portent. By the time of the Theodosian Code’s promulgation, this law, and its institutional support for Traditionalist haruspicy, had been long abrogated. This copy of the Code makes clear to a reader that the law is included in the collection for the purpose of aggregation, and not because it is still enforceable.

Figure 20. Vat. Reg. Lat. 886, 244r, note 90 on CTh 12.1, reading haec antiqua e(st) et n(on) tenet his temp[or]ib(us).

The fifth-century scholia on Vat. Reg. Lat. 886 are those of a scholar – likely a lawyer interested in clarifying which laws are in use, which are not, and what the relationship is between constitutions that have been aggregated for the use of “more industrious types” but which nevertheless present an authoritative promulgation of law.⁵⁰ The centrality of aggregation as a practice in the Theodosian Age offered scholars in different disciplines the opportunity to innovate because it necessitated new tools of discernment. Jerome used the obelus and asterisk to offer insight into biblical variance. Scribes transmitting Hilary’s theological works employed ekthesis and indentation to warn readers when heresy was in their midst. The fifth-century scholiast responsible for the Summaria antiqua created a corpus of marginal notes, complete with idiosyncratic shorthand and cross references between notes to clarify the status of laws promulgated as authoritative, but unequally so. In these and other works of Theodosian era scholarship we see the downstream effects of aggregation as a central scholarly practice. We see fifth-century readers responding to a new environment, writing new texts.

New Texts

Kathleen McNamee undertook the most extensive study of the history of paratextual markup in ancient manuscripts. In a magnum opus of careful, detailed papyrological scholarship, McNamee demonstrates that the late fourth century was a seminal moment in book history. It was the reign of Theodosius I when “for the first time in the history of the book … books were regularly laid out with the intention that they should include extensive exegesis in the margins.”⁵¹ McNamee demonstrates a clear link between the reforms of Theodosius and changing bookforms, and she specifically argues that the teaching of Latin legal texts in the Greek East after 395 ce necessitated wide margins which allow for glosses and commentary. “The fashion [of producing scholastic work with wide margins] quickly spread to literary productions … Like scholia – and unlike the ad hoc notes in other ancient books – the marginalia in these manuscripts were planned from the books’ inception.”⁵² But the known scholia in the margins of juristic texts, beginning with the Summaria antiqua discussed earlier, all respond to the aggregative format. The Summaria antiqua presupposes an aggregative codex; its commentarial format cannot precede it.

Texts with an aggregative format, which “show signs of having been compiled from multiple commentaries,”⁵³ appear only in the fifth century, and the paratextual features that I described earlier arise in nontheological works during this same period. McNamee argues persuasively that the Theodosian Age gave rise to an almost complete overhaul in bookforms across scholastic disciplines, because it was the Theodosian Age in which annotation and commentary on primary texts was so foundational as to precipitate a wholesale changeover in the format of books themselves. I argue, however, that the shift in format of Theodosian books did not begin in the law schools of Beirut and Constantinople, being quickly picked up in other scholastic domains. It was Nicene Christians, and not lawyers, who explicitly discuss aggregation, annotation, and commentary as central scholastic operations in the fourth century. And it was Nicene Christians who prized book formats such as the wide-margin codex which most readily invited commentarial intervention in the margins, such as the Codex Sinaiticus discussed earlier. Christians are the proximate source for this innovation, and not lawyers, who arrived at the aggregative party fashionably late.

I have argued that the focus on aggregative scholarship arose under the influence of a particular set of intra-Christian, theological arguments. These scholarly practices proliferated through other domains when Christians came to significant political power – only during the Theodosian Age. McNamee’s analysis already presupposes the cross-disciplinary interaction that I explore in Chapter 1:

Let us review the situation. The scholastic model for books, involving compilations of commentaries (labeled or not) and written professionally in very broad margins surrounding the text they explain, is likely to have originated in the context of legal education. It must have quickly been adopted for works of scripture like catenae and for works of the classical authors, which all were extensively read and studies and for which large quantities of exegetic writings existed. Once the prototype of scholia had been established, its point of origin – law schools or sacred scriptoria? – and its very point of entry into scriptorial practice – Beirut, Constantinople, or Gaza? – were forgotten. At the time, these were details of minimal importance.⁵⁴

The details of the origin of new practices of textual production, spurred on by new readers with new expectations, were perhaps of minimal importance in antiquity. But they are of great importance to contemporary historians trying to understand the central question that animates this book: what does it matter that Christianity came to Rome? McNamee’s analysis pinpointed the shift in bookforms, but she did not connect it to a previous shift in scholarly expectations among Christians. It is not surprising that McNamee did not notice that Christians came to the aggregative and commentarial format first, however, because her analysis explicitly excludes both Christian materials and literary sources: the two archives that might have suggested an underlying rationale for the Theodosian Age revolution in book forms.⁵⁵

The look and layout of scholastic books changed in the late fourth century. Scholars have suggested a number of possible explanations for this shift. The new readers making new texts may have been jurists, as McNamee argued. They may have been scholiasts of the fourth or fifth century, interested in bringing together commentaries that were traditionally transmitted separately from the lemmatic text, as John Williams White argued.⁵⁶ The problem with this theory is that it must be divined in the silence between the rather cursory marginalia on known ancient poetic texts and the full-blown scholiastic tradition received in medieval manuscripts.⁵⁷

Günther Zuntz put forward an interesting proposal, namely that rabbis inspired the late ancient interest in wide-margin codices with paratextual markup:

There is, in fact, a catena from the fifth century, and its form confirms our solution. It is the Talmud. In the middle of Talmudic manuscripts is a section with the oldest biblical interpretation (the “Mishna”), outside of which stands a collection of exegeses from different interpreters (the “Gemara”). The whole thing is not rare, for instance in von Strack’s facsimile, framed by the “outside commentary,” a rich collection of later explanations, but which, as their text demonstrates, originally stood in different editions. The Mishna (first recorded in the second century) was bound together with the Gemara in the fifth century; the outside commentary comes from the eleventh and later centuries. The rigidity of Jewish tradition makes it certain that the Mishna and the Gemara in the fifth century did not look different as in the twelfth. Thus, philologically speaking, the Talmud represents a fifth century “catena with lemmata,” written in the normal form of hupomnēmata. And since the eleventh century, these “text catenae” were presented with a “border catena,” no different from how Christian catenae looked during this period.⁵⁸

It is hard to overstate the extent to which Zuntz’s proposal is misguided. There are no Talmud manuscripts from the fifth century, and his proposal for the similarity of fifth and eleventh century is based entirely on an unconsidered stereotype about Jewish tradents and their texts (“die Festigkeit jüdischer Tradition”). Beyond this, there is no reason to think that Talmudic material was committed to writing during the fifth century, or in any of the centuries immediately following. Add to this the fact that Zuntz explicitly writes of the Babylonian Talmud, which did not coalesce until around two centuries after period under discussion, and it is easy to put the suggestion aside on account of his failure to grasp basic facts about the tradition.⁵⁹

Nevertheless, Zuntz’s proposal is not altogether mistaken. While there is no reason to think that any of the evidence that he offered bears on the question posed about the birth of the scholia/catenae tradition, Zuntz is nevertheless quite right to point out a set of coincidences in the fifth century which demand explanation. During the Theodosian Age a peculiar form of literary/commentarial production, based on the aggregation of sources, appeared in a wide variety of traditions. In the case of the Talmudic material that was beginning to coalesce in the late fourth century (which is to say the Palestinian rather than the Babylonian Talmud), there is no reason to think that the rabbis responsible explicitly modeled their work on biblical catenae or florilegia, nor vice versa. But the Palestinian Talmud nevertheless does model some of the features of the Theodosian Age structure of knowledge known from other sources, such as the Theodosian Code and the Christian theological scholarship discussed in Chapters 2 and 3. To the extent to which the Palestinian Talmud takes part in the Christianized structure of knowledge, it may be said to be influenced by Christian ways of knowing, even (and perhaps especially) when the scholars quoted in the Talmud reject the preceptual truths held by members of the Theodosian court. And, to the extent to which the Palestinian Talmud differs from the Sassanian recension of the tradition while correlating with features of Theodosian scholarly practices, we can perhaps further situate it as provincial Roman literature. I return to this argument in Chapter 8.

For the moment it is important simply to point out that the physical form of books, and the intellectual projects contained within their leaves, changed dramatically during the late fourth and fifth centuries. Scholars have proposed a wide variety of unsatisfactory solutions to a set of changes visible across the Roman scholastic spectrum, but all agree that the situation cannot be reduced to mere coincidence. In the Theodosian Age a variety of new readers made new texts. I have argued that the common denominator among these new readers is their individual reactions to Christian ascendance and to the centrality of new scholarly practices inflected by a century of doctrinal dispute. The intention and methods of jurists and scholiasts must be read from the details of their literary productions, and the reason that jurists in the fifth century, and scholiasts of Late Antiquity (perhaps) shifted to preferring wide-format books capable of receiving significant paratextual markup is far from clear.⁶⁰ Nicene Christians, on the other hand, tell us explicitly why it is that such new book forms as define the Theodosian Age should arise, and their own internal shift began already during Constantine’s reign. Their changing scholarly predilections, I have argued, best explain those which followed in other scholarly domains.

7 Christian Tools in Traditionalist Texts

Vat. Reg. Lat. 2077 is a sixth-century codex comprising Eusebius of Caesarea’s Chronicon, Gennadius of Massila’s supplement to Jerome’s On Eminent Men, and Vegetius’s Epitome of Military Science: an encyclopedic handbook written sometime in the late fourth or early fifth century “for the Emperor Theodosius.”¹ It is an odd collection of materials. The codex includes works ranging from historical chronicle to military history, framed by an opening request and closing invocation visible today only under ultraviolet light. On the first sheet of the manuscript, in the upper right margin before the first text’s incipit, the scribe responsible for the body text wrote a small note that reads “Christ give help (Christe adiuba)!”² and at the end, trailing the final piece of scholarship in the codex, “Christ give help to the one desiring to know you (Christe adiuba desiderantem te nosse).”³ This scribe responsible for collecting these works of Theodosian Age scholarship together into a single codex began and concluded the work with an invocation to the deity and an invitation to the reader: read these texts with the help of and desire to know Christ. My suggestion – that texts of Theodosian Age scholarship are rightly read within a Christian scholastic context – is not an etic heuristic: this manuscript of Vegetius quite literally has a Christian frame. The manuscript is not unique, either; as mentioned in Chapter 5, the so-called Bodmer Thucydides from the fourth century ce is in fact part of a larger codex including material from the biblical books of Daniel and Susannah, and apparently originating in a Christian monastery.⁴ Even for elite and theologically interested Christians during the Theodosian Age and after, the collocation of Christian and Traditionalist materials, or biblical and secular, was no apparent cause for concern.⁵

Apart from being bound together, manuscripts of Theodosian Age scholarship show signs of production by and for Christians, using tools, framing devices, shortcuts, and notational forms known only from Christian scribal practices. Later I discuss a codex of Livy that boasts all of these, along with one copy of Vergil that uses peculiarly Christian formulae for writing the name of the deity, and another which was apparently copied in an Italian scriptorium that produced Traditionalist classics such as the Aeneid alongside one of the most exquisite biblical manuscripts to survive from antiquity. Our earliest extant copy of the Theodosian Code, too, uses staurograms as binders’ marks, and a papyrus with quotations from the jurists Papinian, Ulpian, and Paul employs scribal tools known only from Christian manuscripts.

This chapter investigates manuscripts in which scribes copied non-Christian works using Christian scribal tools. I describe the proliferation of Christian scribal practices through products of Theodosian Age scriptoria in order to trace the influence of Christianity in a manner that does not involve speculation about the faith of the scribes of these texts or these texts’ users. One main argument of this book is that argumentative tools which were initially devised for internal use in Christian theological disputation came uncoupled from the ideology of their producers. Legal scholars, miscellanists, and historians used tools from Christian disputation in a manner that concealed the tools’ history. I have argued that during the Theodosian Age argumentative forms such as aggregation and distillation were generalized, designified, and reused. Manuscripts from the Theodosian Age show that originally Christian scribal tools such as nomina sacra and even peculiarly Christian symbols such as christograms did not long remain uniquely Christian. Scholastic exchange did not occur solely in the heady, refined space of argumentative forms. It also happened on the space of the page. The generalization of originally Christian scribal tools, and their reuse in works of no obvious theological import, is another important aspect of Christianization in the Theodosian Age that can help us to describe what it means for a society to “become Christian” without recourse to spiritual renewal, moral change, or demographic flux.

Nomina Sacra and Nomina Vulgaria

Ancient scribes employed a range of tools to simplify their texts, to remove extraneous verbiage, and to save space on parchment, papyrus, or stone. Often final a N or M in Latin manuscripts, or final Nu in Greek, is indicated simply with a short supralinear stroke.⁶ Especially in late ancient legal manuscripts, common words are often abbreviated with a stroke across the descender: for instance Ꝑ for “per.”

Scribal tools utilizing supralinear strokes fall into two broad categories: abbreviations and contractions. “Supralineate abbreviations” simply omit letters from the word, generally those letters after the first one or two, and indicate the omission with a small stroke above the word in question.⁷ The other broad category comprises “supralineate contractions,” in which letters are omitted from the middle of a word that remains inflected and identified with a supralinear stroke.⁸

The most recognizably Christian scribal practice is the use of so-called nomina sacra: supralineate contractions that are traditionally restricted to a relatively circumscribed set of lemmata. The peculiarly Christian nature of nomina sacra in literary texts has been widely recognized since the pioneering work of Ludwig Traube in the late nineteenth century, and a number of studies have followed up on Traube’s conclusions about the Jewish origin of this scribal practice and its expression in early Christian manuscripts.⁹ Arthur E. Gordon traced the use of supralineate abbreviations and supralineate contractions in the CIL,¹⁰ and concluded that the corpus leads one to “observes how late contraction is in beginning and how few there are in comparison with [abbreviations]; also how preponderantly Christian it is in its application.”¹¹ The earliest securely dated use of nomina sacra in a literary context occurs in P. Dura 24,¹² and they arrive in the Latin epigraphic record only with an epigram of Damasus from the late fourth century.¹³

Yet, during the Theodosian Age and immediately thereafter, this markedly Christian scribal practice found its way into wider usage among nontheological works, as did the practice of contracting words such as deus when the term refers transparently to traditional gods of Rome and not to the Christian god. The so-called Roman Vergil dates paleographically to the later fifth century: perhaps as early as the late Theodosian Age, though more likely in the decades following.¹⁴ It is among the most beautiful illustrated manuscripts of Late Antiquity, and along with the Vatican Vergil is one of only two illustrated manuscripts to survive of antiquity’s most famous poet.¹⁵ The first folio of this fifth-century luxury copy includes a beautiful miniature depicting the two main characters in Vergil’s first Eclogue, and underneath a striking scribal form: a nomen sacrum in line 6, which reads “Oh Meliboee, a g(o)d has created this leisure for us” (Figure 21).¹⁶

Figure 21. Vat. Lat. 3867, 86r. Aeneid 1.303, where $\overline{\mathrm{DO}}$ expands to d(e)o in corda volente deo; in primus regina quietum.

Traube rightly recognized that a contracted, supralineate form of deus is a remarkably odd usage for this text. It seems that the scribe thought so, too; when the copyist transcribed the next line of their exemplar, they chose a more obvious form: deus in plene form, without contraction. In fact, only twice in this codex of 309 folia does the single scribe use a nomen sacrum.¹⁷ In his own analysis of this strange usage, Traube remarked that “[the scribe’s] intention was to keep the classical text free from Christian abbreviations, but in these two places the habitual form has escaped his stylus,” presuming that in this instance, a slip of the pen betrayed the scribe as a Christian: “Perhaps it was a monk.”¹⁸

Perhaps it was a monk. But such a presumption is just that: something that the historian might assume based on the scribe’s use of a form typically reserved for biblical manuscripts and theological tractates. Because it seems certain that this scribe had used nomina sacra before taking up the task of copying a luxury edition of Vergil’s works, the most likely explanation is surely, per Traube, that “in these two places the habitual form has escaped his stylus.” But suggestions of a theological commitment underlying this bit of scribal somnambulism are less secure and less plausible. Biblical transcription does not a Christian make, just as the fifth-century philosopher Marius Victorinus argued that presence or absence from a Church building did not reveal interesting information about an individual’s beliefs.¹⁹ Instead, I suggest and argue at length later, this manuscript was produced in an environment so thoroughly Christianized that scribal practices that were once the strict purview of Christian texts had become a generalized tool of the trade.²⁰

Literature is not the only domain where Christian scribal tools were reused during the Theodosian Age. There is, in fact, a juristic papyrus from the period that uses the tool of supralineate contractions – nomina sacra – in a rather less spiritual manner. It employs, one might say, nomina vulgaria. P. Haun III 45, along with fragments belonging to the Arangio-Ruiz private collection (CPL 73 A, B), comprises an ancient handbook on the topics of legacies (legati) and trusts (fideicommissa): a work of scholarship bringing together opinions of the five jurists mentioned in the Law of Citations under useful thematic groupings.²¹ The compiler of this text is unknown, but there is reason to believe that it was originally put together in the late third or early fourth century, and that this copy was produced in the late fourth or early fifth.²²

The juristic opinions included were previously lost to posterity; they were not transmitted in the Digest or any of the late antique compilations.²³ For this reason, the legal content of the papyrus has received the vast majority of critical attention, and its form and scribal peculiarities have gone largely unremarked upon.

The scribal peculiarities of this papyrus are astonishing. The manuscript was likely that of a scholar,²⁴ and incorporates both traditional supralineate abbreviations, for instance P͞P for propter, Q͞A for quia, and N̅ for non, as well as other typical juristic abbreviations such as P with an ascending stroke across the descender for per (Figure 22).²⁵ But this papyrus contains not only traditional scribal abbreviations that we might cluster under the loose heading notae iuris, it also includes supralineate contractions of common words.²⁶ It contains, in other words, the type of scribal tool that papyrologists typically cluster under the heading nomina sacra. Supralineate contractions in this papyrus has gone largely unnoticed by the broader public because no editor – and there have been five – offers anything but the most cursory remarks on them.²⁷ Nevertheless, there they are. Consider, for instance, Figure 23, in which the scribe of this papyrus abbreviates fideicommissarius as $\overline{\mathrm{FC}}$RIUS. Throughout this papyrus a specific set of lemmata relating to the subject under discussion (testamentum, fideicommissum, and heres) are consistently contracted, marked by a supralinear stroke, and inflected. Figure 24, for instance, shows the contraction of testamentum inflected in the ablative to read TT(AMENT)O.

Figure 22. P. Haun III 45, selection from lines 85 and 86, infrared photograph. In the center of the upper line we see a P with an ascending line across the descender indicating per, and on the second line $\overline{\mathrm{PP}}$, $\overline{\mathrm{QA}}$ and N̅ for propter, quia, and non, respectively. Line numbers are according to Larsen and Bülow-Jacobsen.

Photos courtesy Adam Bülow-Jacobsen.

Figure 23. P. Haun III 45, selection from line 65, infrared photograph. The line reads $\overline{\mathrm{FC}}$RIUS EO ꝗ, with a supralinear stroke over the FC and an ascending stroke through the descender of the Q. Expanded, the phrase is f(idei)c(ommissa)rius eo q(uod).

Figure 24. CPL 73 B recto, detail reading SECUNDO $\overline{\mathrm{TTO}}$ RẸ. From line 70 as published in Nasti, corresponding to the lacuna in line 60 of Larsen and Bülow-Jacobsen. This fragment is in the Arangio-Ruiz collection and the photo is from CLA Supplement 1756.

This papyrus presents the earliest example of supralineate contractions in a juristic context (Figures 25 and 26). In fact, it presents the earliest use of supralineate contractions in any Latin manuscript that does not present explicitly Christian content. In addition, the closest paleographical parallel to this papyrus is not another juristic fragment but P. Rylands Greek 472: among the earliest known Latin Christian papyri.²⁹ Serena Ammirati argues persuasively that these two manuscripts must be understood as arising out of a similar, bureaucratic – to which I would add scholastic – context. “Books of law represent the specific professional interests of individuals who are simultaneously producers and users of Latin books with literary content. If the users of books containing literary and juristic content belong to the same professional category, it is reasonable to expect that these books would share formal characteristics.”³⁰

Figure 25. Supralineate abbreviations in P. Haun III 45 identified by Larsen and Bülow-Jacobsen. Line numbers follow their edition.²⁸

Figure 26. Supralineate contractions in P. Haun III 45 identified by Larsen and Bülow-Jacobsen. Line numbers follow their edition.

Christians are the ultimate source for such a thoroughgoing and standardized use of supralineate contractions; by the time that the scribe of P. Haun III 45 put pen to papyrus they had been used in biblical manuscripts for over 200 years. A full account of this papyrus, however, will identify the proximate source for this scribe to import the technology of supralineate contractions into the juristic domain. The cluster of coincidences – a Theodosian date, the closest parallel being a Latin Christian liturgical fragment, and the bureaucratic or scholastic environment of both papyri – suggests that this papyrus presents precisely the reuse of the technology of nomina sacra in a juristic context.³¹ The distinction that I draw here, between traditional juristic abbreviations and these new contracted forms, may seem to be inconsequential, or at least too arcane to offer fruitful insights into Late Antiquity. Quite the opposite is true. The use of this peculiarly Christian tool in a context so remote from theological study shows that in the Theodosian Age, what used to be the oddities of Christian scribal practice were no longer odd, nor were they particularly Christian in implementation or meaning. A scribe implemented a tool invented for biblical manuscripts to simplify a legal handbook. We cannot see into the mind behind the pen, nor can we probe the propositional truths held by these scribes. What they believed is inaccessible, but perhaps it is also not particularly interesting. What is clear is that the scribe of P. Haun III 45, and the scribe of the Roman Vergil mentioned earlier, appropriated a tool that was once the solely purview of theological works and applied it in a new context with new aims.

A late ancient reader may well have approached these manuscripts with the same historical incredulity expressed by Traube and others regarding the use of a “Christian” tool in a “Pagan” context. An ancient reader might also have passed over these nomina vulgaria without giving them a second thought, as has been the case for most modern editors of the Haun papyrus. But there is another way to read these manuscripts. If we assume that the scribe was in fact a Christian, and purposefully used a theological tool while copying a Traditionalist text, then we can speak of ideological and textual “Christianization” happening in late antique scriptoria. If, on the other hand, these scribes made casual mistakes or technological transpositions, unintentionally inserting tools from Christian scribal practice into nontheological texts, then we can speak of the technological “Christianization” of late antique scriptoria still. In the latter case the point is doubly made: during the Theodosian Age, in nontheological manuscripts, scribes began to use tools that were forged in scriptural fires and they applied these tools without obvious implication. The fact that nomina sacra and nomina vulgaria appear at all attests to the thoroughgoing Christianity of the scholarly and scribal context, quite apart from the beliefs of any of these texts’ producers. Scribes reusing Christian tools and symbols in nontheological contexts is interesting if it is value laden – if the producers of texts intend to “Christianize” manuscripts of non-Christian texts. But it is perhaps more interesting if the importation of nomina sacra, and the other symbols of Christianity discussed later, are employed completely devoid of ideological meaning.

By way of analogy, imagine that the fascist era in Italy had lasted as long as the period between the conversion of Constantine and the end of the Theodosian Dynasty – around 140 years, from 1912 to 2057 – rather than the twenty-one years that it lasted in reality. If, in the twenty-teens, the symbol of the fasces began to be used as arrows on highway signs, pointing the way for travelers to the closest fuel station or roadside motel, we could not responsibly presume that the maker of the sign was a supporter of the long-dead Mussolini’s policies. Instead, the most natural interpretation would be that the sign of the fasces, which was reintroduced a hundred years earlier as a symbol of military might and political ascendancy, had become so naturalized in the social environment that its meaning was no longer inextricably connected with the ideology that it was originally intended to signify. Much the same happened in the Roman empire of the fourth and fifth centuries: symbols of military might and political ascendancy such as the Constantinian staurogram, as well as scribal tools such as nomina sacra, came to be used in dramatically new ways. It would be historically irresponsible to interpret such usages as indicating something about the faith of the user, but they may evince something about the culture in which the user lived.

Consider, for instance, Vat. Urb. Lat. 1154, a late fifth-century copy of the grammatical work of (pseudo-)Probus.³² The Proban tradition was already complex in Late Antiquity, and at least three different hands supply additions in the margins of this manuscript. A number of markers are used in late ancient manuscripts to indicate the place where text should be inserted, and the text that should be inserted. Often, hs is inserted in line with the base text, indicating the location of an insertion, and hd is written in the margin next to the supplementary material. One corrector of Vat. Urb. Lat. 1154 uses the hs/hd method elsewhere in the manuscript,³³ but on 20v, they chose a somewhat different tack; the corrector inserted hs in a half-uncial hand contemporary with that of the base text, but instead of the correlating hd in the margin, this scribe chose a staurogram, with an alpha and omega underneath, to alert the reader of this grammar that an insertion should be made (Figure 27).

Figure 27. Vat. Urb. Lat. 1154, 20v. The staurogram is repeated as well in the bottom margin to indicate that the lower text continues what is above.

The use of the same Latin uncial for the text of the correction as well as the alpha of “α:ω” (along with the colon in identical dark brown ink) make clear that this corrector’s sign belongs with the marginal note, rather than having been added subsequently. In other words, here we have perhaps the most banal use of the staurogram surviving from antiquity: pointing a reader to a textual variant in a grammatical treatise. I return to this point later.

In the Roman context, christograms, of which the staurogram is one type, were associated initially with Constantine’s victory at the Milvian Bridge in 312. In the early years of the fourth century, the christogram was a potent symbol of political domination under the aegis of a new god: the Christ to which Constantine had allegedly sworn his allegiance the night before marching on Rome. This category of scribal symbols that overlaps with nomina sacra came to symbolize Christ, Christian faith, and eventually, Constantinian dominance.³⁴ Early on the symbol was most common on dynastic coinage. For instance RIC VII Constantinople 19 depicts Constantine laureate on the obverse and on the reverse a military standard, topped with a christogram, and the legend SPES PUBLIC (“the safety of the republic”). This coin was struck in 327 in a variety of denominations and seems to refer to Constantine’s victory over Licinius and assumption of sole rule over the East and the West.

The staurogram also functions as a part of a nomen sacrum, and appears supralineate in some early New Testament manuscripts including a fourth-century codex containing the Gospels according to Luke and John.³⁵ It was a potent enough symbol to be a significant part of Constantine’s program of visual propaganda, and it continued to appear on coinage throughout the Theodosian Age to symbolize the orthodox Christianity of the regent. Its use across media from manuscripts to coins indicates the ubiquity of the staurogram as a symbol, but it does not indicate how that symbol was used or what it meant: the insertion which it signals in Vat. Urb. Lat. 1154 has no dynastic, military, or theological valance whatsoever. It is a rather bland grammatical note.

Images of the goddess Victoria succumbed to a similar process of resignification in the Theodosian Age. Consider, for instance, RIC IX Cyzicus 26a, a coin of Valentinian II depicting on its obverse the goddess Victoria, with a trophy in her right hand, dragging a captive in her left and standing next to a staurogram. Like the Constantinian coin, from some sixty years before, the legend reads SALUS REIPUBLICAE (“the health of the republic”). Images of Victoria signaled Roman might and subduing of foreign peoples since at least the time of Augustus, when a statue and altar for the god were installed in the Senate curia. Despite a few brief removals, the altar remained in the Roman senate chambers well past the reign of Theodosius I, and despite its clear Traditionalist associations, many Orthodox Christians were willing to deploy the image of Victoria devoid of any overt religious meaning.³⁶

Many regarded the staurogram in a similar way, though the change occurred on a significantly shorter timescale. By the time of the Theodosian empire, the staurogram could be used to signify the Christian deity’s protection of the “health of the republic,” but it could also be used to indicate the presence of a textual variant in a grammatical treatise. As I read the evidence, a scribe’s use of the staurogram as a corrector’s symbol in Vat. Urb. Lat. 1154 need not indicate anything about the faith or political inclinations of the corrector. Instead, what the choice indicates is that by the later fifth century, the semantic range of the staurogram, once a sign of imperial power most commonly associated with military equipment, had expanded to encompass any number of applications that have neither imperial nor theological relevance.

The staurogram was also reused as a multipurpose symbol in the Greek East of the early sixth century. I wrote in Chapter 6 about the literary qualities of the so-called Summaria antiqua that fills out the margins of Vat. Reg. Lat. 886. The scriptorium that produced this manuscript of the Theodosian Code, however, repays further attention. The binder, whose job it was to keep the original quaternions of this substantial manuscript in order and to stitch them together after the scribe had finished their work, used the same symbol – a staurogram – to indicate the beginning of each gathering, as is visible in the upper left corner of 17r (Figure 19), and throughout the manuscript: on 9r, 25r, 33r, 41r, 47r, 55r, 61r, 69r, 77r, 84r, 92r, 100r, 108r, 115r, 123r, etc.

None of these examples appear to be an ancient attempt to cast otherwise dry, pre-Christian scholarship within a Christian frame. Rather, in these manuscripts we see scribes reusing symbols that originated in Christian contexts as ideological blank slates. The fact that a late fifth-century scribe could use a staurogram as a corrector’s symbol or a binder’s mark suggests nothing credible about the faith of the scribe (or that of a reader), but such uses say a great deal about the culture within which these texts were transcribed and read. The signs are not innocuous or irrelevant pious ephemera. The recasting of even such tedium as marginal notes quite literally under the rubric of Christian symbolism indicates the thoroughgoing extent to which an ideology had been generalized through the remobilization of its operative symbols. Writing in the seventh century about various forms of critical signs that his readers might find in manuscripts, Isidore of Seville agreed that the christogram had lost all inherent meaning. These examples demonstrate that Isidore was not remarking on a novelty when he offers to his reader, “Chrisomon: this is placed according to the interest of the individual to mark something out.”³⁷

Analysis of such clear instances of reuse – a sort of scribal spoliation – may help to clarify the scribal and ideological context in other, less clear cases. Vienna Österreichische Nationalbibliothek Latin 15, for instance, is an early fifth-century deluxe codex of Livy’s History of Rome that uses a supralineate $\overline{\mathrm{PR}}$ to stand for various inflected forms of populus romanus.³⁸ This contraction is not known from the epigraphic corpus,³⁹ and the other known copy of this text from the Theodosian Age uses the same abbreviation but without supralinear strokes.⁴⁰ Even within the same section of text, the scribe indicates the abbreviation sometimes with a supralinear stroke, in the manner of a nomen sacrum, and sometimes without (Figure 28).

Figure 28. Vienna, Österreichische Nationalbibliothek Latin cod. 15, 162r. (TM 67658). From Livy, History of Rome 45.28. The PR abbreviation is used three times in this section, with only one instance supralineate, on the seventh line from the bottom.

The scribe copying this manuscript apparently uses the tools of Christian manuscript production to indicate to a reader the presence of an abbreviation. The inconsistent use within this manuscript, as well as the comparison with a contemporary manuscript of the same text that does not utilize supralineate abbreviations, suggests possibly that the scribe in question is not taking over usage from an exemplar, but rather reused tools known from a different scribal domain in their work on this manuscript of Livy. It is of course possible that this scribe implements an epigraphic practice when employing supralineate $\overline{\mathrm{PR}}$, but it is not probable: the abbreviation is otherwise unattested. The most proximate context in which to understand this manuscript’s form of abbreviation is a that of a Christian, biblical scriptorium. Another late fourth- or early fifth-century manuscript of North African origin uses the same form – $\overline{\mathrm{PR}}$ supralineate – to indicate presbutero in a copy of Cyprian’s Letter 54 (Figure 29).⁴¹

Figure 29. Vat. Lat. 10959, 1r. (TM 66155). Incipit of Cyprian Letter 54 with PR supralineate to indicate presbutero as well as ITE superlineate to indicate ite(m). The supralinear stroke, in other words, is used for different purposes in successive words. The first indicates the contraction, while the second indicates a suppressed M, as is common in fourth- and fifth-century majuscule manuscripts (though more common at the end of lines). The same calligraphic supralinear stroke is used in line 12 of the same column to indicate a suppressed M at the end of laetatu(m).

In line 18 of the right column of this leaf of Cyprian’s letter, the scribe used ΧΡΣ supralineate to indicate Christus, indicating their familiarity with nomina sacra. Thus, the same scribe’s use of supralineate PR in line two of the left column should be understood as a reuse of that same scribal tool, even though presbuteros is not a traditional nomen sacrum. It is reasonable to assume that the scribe copying Cyprian used a supralinear stroke as a scribal tool to indicate, simply, “this is a contraction.” My argument is that the same assumption is reasonable in the terms of the roughly contemporary scribe copying Livy. In both manuscripts new contractions are indicated with the same tool, and yet scholars are only comfortable calling one a nonstandard nomen sacrum, while the other is simply a scribal oddity. A responsible historical methodology requires us to consider that these coincidences may not be accidental, and that they may say something about the ideological context in which each text was transcribed even though they do not speak to the ideology of the scribes themselves.

Such correspondences in scribal practice, and in the use of seemingly “Christian” tools in nontheological contexts, are so common in fifth-century manuscripts that the trend cannot be reduced to training or local peculiarities. A juristic fragment in Berlin dated between the mid-fourth and mid-fifth century uses $\overline{\mathrm{PR}}$ supralineate to indicate praetor.⁴² It is unclear whether this supralineate $\overline{\mathrm{PR}}$ is intended as a contracted or an abbreviated form – that is, whether the $\overline{\mathrm{PR}}$ stands for pr(etor) or p(reto)r. If the former, then there is some classical precedent for such abbreviation. If the latter, then this contraction is more easily placed squarely in a Christian scribal context. The ambiguity itself is telling, and a Theodosian reader of this text might reasonably come to either conclusion.

Similarly, Codex Puteanus uses a supralinear stroke over CN for Gaius (22v, Figure 30), COS for consul (22v and very often elsewhere, Figure 30), M for Marcus (29r), PR for praetorum (31v), and SC for senatus consultis (31v).⁴³ The four distinct uses of supralineate PR mentioned here alone suggest that what we are dealing with is not a standardized set of abbreviations but rather that the supralinear stroke is deployed as a common tool: an aid to readers whom the scribe expected to be familiar with such indications.⁴⁴

Figure 30. Paris, Bibliothèque Nationale Lat. 5730, 22v, “Codex Puteanus.”

Note the supralinear abbreviations for “Gaius” (line 2) and “consul” (line 3), as well as a supralinear stroke at the end of line 1 noting the suppressed final M of “idem.” The text is Livy From the Founding of the City 22.

(Source gallica.bnf.fr / BnF.)

Examples could be multiplied almost $\overline{\mathrm{AI}}$. Among the closest paleographic parallels to Codex Puteanus is the Lavanttal Ambrose, discussed earlier. Clear links are visible among the two manuscripts in paratextual features (running titles, binder’s marks) and ligatures (Figures 31–34). Given the overwhelming similarities in script, paratext, and codicology, it is near certain that these manuscripts come from the same period, and perhaps from the same scriptorium.⁴⁵ Both are dated to the mid-fifth century and, I argue, we must bring an analogous set of assumptions to understanding the particular scribal features found in both. In both we find the same tools – Christian scribal tools – used throughout.

(Source gallica.bnf.fr / BnF.) Stiftsbibliothek Lavanttal 1, 54v (right).

Figure 31. Running titles. Paris, Bibliothèque Nationale Lat. 5730, 28r (left).

(Source gallica.bnf.fr / BnF.) Stiftsbibliothek Lavanttal 1, 54v (right).

Figure 32. Binder’s marks. Paris, Bibliothèque Nationale Lat. 5730, 25v (left).

(Source gallica.bnf.fr / BnF.) Stiftsbibliothek Lavanttal 1, 8v (right).

Figure 33. NS ligature. Paris, Bibliothèque Nationale Lat. 5730, 2v (left).

(Source gallica.bnf.fr / BnF.) Stiftsbibliothek Lavanttal 1, 7r (right).

Figure 34. NT ligature. Paris, Bibliothèque Nationale Lat. 5730, 9r (left).

The phenomenon is not merely literary either, nor is it solely found in manuscripts. As I mentioned earlier, supralineate abbreviations arrive remarkably late as a standard tool in the epigraphic record, and as Nicoletta Giovè Marchioli has demonstrated, there is a surprisingly close connection in scribal habits between epigraphic and literary materials, particularly when it comes to abbreviations.⁴⁶ In the classical period a supralinear stroke was typically used only to indicate numbers, and occasionally (though inconsistently) consular dates.⁴⁷ By the mid-fifth century, however, supralinear strokes were used to identify even the most common contractions in public inscriptions. For instance, a dedicatory inscription from 468/469 for Flavius Eugenius Asellus records only three lines but boasts four supralinear strokes, indicating v(ir) c(larissimus), praef(ectus) urb(i), and v(ice) s(acra) i(udicans).⁴⁸ By the early sixth century, these indications were used even in funerary contexts. For instance a funerary inscription from June 23, 525 ce for a certain Maxima, “enslaved attendant of Christ (ancilla Cristi),” uses supralinear strokes in every possible place: eight times in an unimposing inscription of seven lines.⁴⁹ One supralinear stroke, carved over the word “in” on line 7, appears to be completely extraneous – so much so that Ernst Diehl excluded it from his edition of the text. Perhaps the scribe got carried away with all the abbreviating.

This Christian tool found its way from biblical manuscripts to other literary texts, and eventually to dedicatory and even funerary inscriptions. There is a classical precedent for the use of supralinear abbreviations: they occur in the epigraphic record from the period of the early empire, though rarely. The dramatic increase of attestations of this tool, and its use to identify both contractions and abbreviations in manuscripts and inscriptions occurred only after the period of Christian ascendancy. Supralineate contractions and abbreviations are not uniquely Christian, but their thoroughgoing use is, and they came to be used widely only when Christians became politically and scholastically dominant. I argue that this state of affairs is no coincidence.

I want to discuss one more pair of Theodosian manuscripts which appear to have the same provenance but which scholars typically class differently because of the content of their leaves. The so-called Vatican Vergil is a deluxe manuscript from the late fourth or early fifth century that contains portions of the Georgics and Aeneid.⁵⁰ The seventy-six surviving leaves have been the subject of hundreds of paleographic, text critical, and art historical studies.⁵¹ Though it uses no nomina sacra of any sort, nor contains any obvious paratextual features of note, we can be relatively certain that the Theodosian, Italian center responsible for the production of this manuscript produced at least one other manuscript which remains extant: a deluxe Latin Bible known as the Quedlinburg Itala.⁵²

The Vatican Vergil is not complete: it originally contained the entirety of Vergil’s work in what must have been around 440 folia like its cousin the Roman Vergil discussed earlier, and it is written in “an old type of rustic capital.”⁵³ The text is of remarkable quality but is unremarkable otherwise; the single scribe responsible for the entirety of the text was well trained (and perhaps suppressed some forms that they knew⁵⁴), but based on the text alone little can be said about the intellectual or ideological environment in which the scribe worked. More remarkable are the fifty surviving illustrations, which comprise two-thirds of many pages, and in places take up entire leaves of the codex. David H. Wright’s reconstruction suggests that originally there must have been approximately 280 illustrations, and it is the art that has been used most often to situate the work. Commentators have focused on these illustrations to make arguments about the codex’s intellectual context, and in turn about the ideology of its owner. Wright’s approach is characteristic, and echoes the majority of critical opinions. He claims that in Late Antiquity, “Christians continued to read and admire Vergil, both for the exemplary qualities of the poetry and as the embodiment of a national tradition, but no Christian is likely to have commissioned a fine illustrated edition, especially one containing many scenes of pagan sacrifice.”⁵⁵ Unsurprisingly, given his dating of this text to “the time around 400, meaning probably within two decades on either side of that date,”⁵⁶ Wright presumes that this text should be understood as arising under the patronage of “an associate of Quintus Aurelius Symmachus,” the renowned Roman Traditionalist of the Theodosian Age, and proposes that the most reasonable understanding of the context in which this codex was produced is the flourishing intellectual life at Rome around 400, specifically in the absence of significant Christian voices at Rome when compared with the rest of the empire.⁵⁷ “The power of the church was growing rapidly,” he writes, “but the most important Christian intellectuals were not in Rome: Ambrose was in Milan, Jerome in Bethlehem, and Augustine in Hippo. The pagans were on the defensive politically, but because of the diffusion of authority, in an important sense Rome was still theirs.”⁵⁸

Wright’s contention, that the text cannot have been commissioned or owned by a Christian, is unconvincing a priori because it confuses the stressed and stressful orthodoxy of “the most important Christian intellectuals” like Ambrose, Jerome, and Augustine with the interests of Christians writ large. Jerome was at least as hated as he was loved in the late fourth century, and Augustine was almost entirely unknown outside of a dramatically circumscribed group of clerics and governmental officers.⁵⁹ In Wright’s work, as well as studies by other specialists on the topic, the backward gaze of orthodoxy has produced a blurred picture of Theodosian Age Christianity. Ambrose, the only cleric mentioned by Wright who wielded significant political power, regularly peppered his writings with Vergilian quotes and reminiscences, and such a proposal forgets Christians at least as powerful as Ambrose whose confessional commitment comes across hardly at all in their surviving literary work.⁶⁰ One thinks of Macrobius, who was likely Praetorian Prefect in Rome in 430, or Vegetius, who was a vir inlustris comes – both of whom have been discussed in this book already.⁶¹ Vergil was so untainted by paganism for the Theodosian Age Christians of Cuicul that they constructed a magnificent baptistry to serve the metropolitan basilica, complete with a quote from Eclogue 3 at its center, over which initiates received the Christian rite of baptism. The baptismal fount reads [Gentes t]empus erit omnes in fonte [lavari] – “There will come a time for all people to be washed in the fount.”⁶² North African Christians baptized Vergil, quite literally.

There is no credible reason to think that this copy of Vergil could not be owned or commissioned by a Theodosian Age Christian, and as Inabelle Levin has demonstrated, there is strong reason to think that the object itself was produced in the same scriptorium as produced the Quedlinburg Itala, the oldest illustrated biblical manuscript extant.⁶³ The illustrations of this bible share significant stylistic and thematic overlap with both the Vatican Vergil as well as with the mosaic cycle found in the nave at the basilica of Santa Maria Maggiore, dated to the papacy of Sixtus III between 432 and 440.⁶⁴ All three of these objects – the bible, the Vergil, and the mosaic cycle – are likely products of the flurry of activity in Rome after the sack of 410. Striking similarities in so many details among the illustrations suggest the manuscripts were also made in Rome, and that they were all products of the same generation.⁶⁵ Whether the Vatican Vergil and the Quedlinburg Itala were products of the same scriptorium is impossible to say. It is clear, however, that both arise from the same milieu, and we cannot dismiss the very real possibility that Christians are responsible for all three.

The historical interpretation of this cluster of materials is difficult, but it is not intractable. If the Quedlinburg Itala and the Vatican Vergil came from the same scriptorium, they might have been delivered to the same household; or, perhaps not. It is perfectly reasonable to think that the Vergil codex found its first home on the bookshelf of an elite Roman Christian, but it is just as reasonable to think that the bible came to rest in the scrinium of a Roman Traditionalist interested in having a copy of texts so central to many in their social network. We simply don’t know. Clearer is that the distinction between “Pagan” and “Christian” manuscript production does not hold in the Theodosian Age. If the same iconographic language is visible in the mosaics of Santa Maria Maggiore, the Vatican Vergil, and the Quedlinburg Itala, then the visual language is either meaningless when trying to assign ideological commitments to the producers, or the ideological commitments of the producers have no clear bearing on the visual language used. In either case a picture emerges of elite society in which ideology and materiality are separable, even when not always separate; a society in which the tools of scholastic production are influenced by the ideological context, but in which the tools selected by any particular user do not identify their theological or ideological commitments. This is a society with thoroughgoing Christian influences, but a society in which one’s choice to use “Christian” tools speaks to the context of production rather than to the commitment of the producer.

Conclusion

This chapter, and Chapter 6, investigated the new texts produced by Theodosian Age writers and scribes against the background of the dominant scholastic argumentative framework. I have tried to explain why books from the Theodosian Age look so dramatically different from those which preceded them, and why books across genres began in this period to look so remarkably similar: with shared formats, features, and scribal interventions. Why is it that, for instance, the closest paleographical and codicological parallel for a late fourth-century juristic fragment such as P. Haun III 45 is not another juristic fragment, but rather one of the earliest known Latin fragments of a Christian text: a liturgical codex fragment in the Rylands collection?⁶⁶ It is because, so far as we can tell, the producers and users of these sort of texts were one and the same.⁶⁷

The fifth century, from which the vast majority of our relevant manuscripts survive, was a time of extraordinary innovation. Approximately 40 percent of all surviving Latin literature produced before the seventh century was composed between 400 and 499 ce.⁶⁸ Literary production flourished, and old tools were reused in new ways in order to accommodate a set of novel scholarly practices. The penetration of peculiarly Christian tools and symbols into nontheological domains does not speak to the Christianity of writers producing our extant manuscripts as much as it speaks to the dominant ideological framework in which these individuals worked. The weight of the evidence, I suggest, demonstrates that a shift in scholarly practices caused a shift in material production that is visible in manuscripts and inscriptions from the period under discussion. From an expectation of aggregation came problems of discernment and new strategies for making clear what material is included for edification, and what is included because scholars expected that a good argument was one based in aggregation. Christian tools such as supralineate contractions came to be widely used throughout the literary and documentary landscape even as symbols such as the christogram were rendered ideological blank slates, capable of signifying everything from imperial triumph to textual variation. Professionals trained in Christian scribal practices and iconography took their talents to work on nontheological texts, such that the contemporary scholarly separation of “secular” from “sacred” Theodosian Age material becomes a distinction without a difference.

A book that looks different elicits a different interpretation, and new reading strategies commensurate with the change in form. New bookforms and textual features influenced the meanings read into and out from books during the Theodosian Age. Chapter 8 returns to literary evidence, to explore the new meanings that readers found within the pages of these Theodosian productions.

8 New Meanings

Reading practices have always responded to changing media landscapes, as new formats and writing practices demand novel strategies of understanding. Today, the simple presence of hyperlinks in online news stories results in significantly increased recall of details and assessments of credibility, whether or not the information is credible, in point of fact.¹ Media bandwidth crescendoed over the past thirty years as new and social media were added to traditional outlets; in turn, all media have adapted to the new landscape or ceased operations. The expansion of outlets has led many to bemoan the downfall of credibility accorded to mainstream news organizations, as has the apparent proliferation of “fake news” from outlets with lower editorial standards than traditional media organizations, but near-equal reach.² Modern news consumers have adapted: educational institutions teach classes on “media literacy” to young students and even major, multinational corporations have begun to team up with governmental agencies to “empower[] young people with the critical thinking skills necessary in today’s digital age.”³ Changing technology creates new subjects with novel reading strategies calibrated to exigencies of our media environment. We read differently than we once did, and when the social and political landscape shifts again, so too will our practices.

The same can be said of Late Antiquity: social, political, and material shifts necessitated novel interpretive strategies for scholastic productions. When Nicene Christians first came to significant political power in the late fourth century, they brought with them a structure of knowledge that gave pride of place to projects of aggregation, distillation, and promulgation. Chapter 4 detailed the proliferation of these tools from the margins to the center of the Theodosian scholarly landscape, while Chapters 5, 6, and 7 surveyed the effect of new scholarly practices on the production of manuscripts. My final chapter explores the net effect of the changes described: the new reading strategies that Theodosian Age readers implemented in response to new scholastic forms transmitted in novel formats. I am interested in what might generally be called “interpretive strategies,” but which are more precisely modes of “actualization,” to use Michel de Certeau’s term.⁴ As Roger Chartier writes:

To reconstruct this process of the ‘actualization’ of texts in its historical dimensions first requires that we accept the notion that their meanings are dependent upon the forms through which they are received and appropriated by their readers (or hearers). Readers and hearers, in point of fact, are never confronted with abstract or ideal texts detached from all materiality; they manipulate or perceive objects and forms whose structures and modalities govern their reading (or their hearing), thus the possible comprehension of the text read (or heard). Against a purely semantic definition of the text … one must state that forms produce meaning and that a text, stable in its letter, is invested with a new meaning and status when the mechanisms that make it available to interpretation change.⁵

At stake for the Theodosian reader was not just understanding what the text in front of them said; they must discern the proper response to those words. The issue was rendered acute in a world of aggregation, when materials with different status stand on a single page, side by side. In nearly every instance, a reader must ask, “do these words express something true, and what is their value in relation to other assertions visible on the page?”

The rise of aggregation as a central facet of scholastic work necessitated the development of corresponding strategies of discernment. Chapter 6 detailed simple paratextual strategies through which Theodosian readers indicated heretical opinions, disused laws, and the like. Here I explore the intellectual strategies that scholars and readers implemented in order to retrieve truth from works whose format placed truth and falsity, heresy and orthodoxy, good law and supervened law, side by side. Of course, algorithms for determining truth within multivocal literary productions predate the Theodosian Age. Yet the sophistication and widespread implementation of these particular strategies across so many scholastic genres and linguistic divides mark the creation of these new reading strategies as particularly embedded in their intellectual environment. Such strategies were not new, but they were newly necessary across scholastic domains.

I argue that Theodosian Age readers responded to the literary scholastic environment of the late fourth and fifth centuries along two central trends. The first that I discuss, “rules for deciding,” deal with the problems inherent in reading works of aggregative scholarship. The second, which I group under the heading “institutionalized suspicion of documents,” deals with the stresses involved in compiling such works. The extraordinary preeminence of archival sources in Theodosian scholastic work led to new ways of approaching material handed down by tradition and to an invigorated suspicion of archives and documents. Simply put: if one is to authorize, codify, and promulgate a particular historical opinion, one must be certain that the source for that opinion has not been tampered with. Pressure to create monumental, universalizing works of final authority such as the Theodosian Code or official ecclesiastical pronouncements such as the acta of councils required certainty about the precise wording of archival sources. If conciliar proceedings held no intellectual weight, there would be little reason to certify the contents of acta with anything like the rigor brought to bear on the documents from Ephesus (431) or Chalcedon (451). During the Theodosian Age, however, when conciliar proceedings gained the patina of patristic authority and when the documents themselves were bestowed imperial backing, it mattered what they said. The centrality of documents to conciliar dispute appeared late in the fourth century, and was already firmly seated in 381, at least to judge by Palladius’s exasperated response to Ambrose’s insistent questioning at the Council of Aquielia: “You’re the judge, [on account of the fact that] your note-takers are here! (Tu iudex es, tui exceptores hic sunt).”⁶ Conciliar acta became theologically dispositive only in the period of the Theodosian empire, and in this period we see an institutionalized suspicion of their production and of the production of documents that underlie the final, authorized codex.⁷ So it was in the domain of Theodosian patristic theology, but such concerns echoed across the scholarly landscape.

This chapter thus situates Theodosian Age readers with respect to the Theodosian writers detailed earlier. In Chapter 4, I argued that scholars produced aggregative works with an eye toward how they were to be used. I argue now that Theodosian readers used these sources with cognizance of – and concern over – how the collections were produced. The resulting dialectic, visible perhaps only from the outside and in retrospect, defines the new order of books in the Theodosian Age that is my central focus. This chapter presents the last piece of the puzzle, placing textual producers and receivers together into a single frame. The story could not be told in a linear fashion because each Theodosian producer was also a receiver; there is no single way into or out of this labyrinth. But the effect of the analysis should be a sense of coherence visible even among fragmentary evidence.

Rules for Deciding

The Theodosian Code was a universal statement of law, but it was compiled from constitutions that revised earlier legal practice, in most cases. Supervening laws were placed next to the laws they supervened, with little attention paid to the state of law prior to 312 ce. In consequence, the Theodosian Code is not a handbook of law; it would be nearly impossible to learn legal praxis simply by reading through the content of the Code without the framing offered in works by the great classical jurisconsults: Gaius, Ulpian, Papinian, and the like.⁸ There are, however, some constitutions preserved that deal directly with legal praxis, and with the selection and weighing of sources by lawyers in the Theodosian legal framework. I turn to one such example now.

The so-called “mini-code of 426” was promulgated at Ravenna under the authority of Theodosius II and Valentinian III (though presumably the constitution reflects Galla Placidia’s wishes rather than those of her son Valentinian, who was seven years old at the time), including both reforms to inheritance law and clear statements about the sources of law that could be legitimately cited in court as precedential. The “mini-code” was excerpted into five extant constitutions, of which one survives in the Theodosian Code and four survive in the Justinianic Code.⁹ The portion of this “mini-code of 426” that survives in the Theodosian Code is perhaps the most interesting, as it deals both with the issue of validating sources and the problem of discernment among competing authorized voices. The constitution is often referred to as the Law of Citations, and it is perhaps the purest example of the dangers involved in producing and using aggregative scholarship in service of a universalizing knowledge regime. In it we see a clear illustration of both facets of Theodosian Age textual practice under scrutiny in this chapter: an institutionalized suspicion of documents and rules for deciding.

On its face, the Law of Citations provides for the authorization of a collection of Severan juristic texts as holding an equal standing as those of earlier Republican and Imperial jurists. It reads:

We confirm every writing of Papinian, Paul, Gaius, Ulpian, and Modestinus, such that the same authority shall attend Gaius as Paul, Ulpian, and the others. Additionally, passages from the whole body of his work may be offered [as evidence]. We also decree to be valid the learning of those persons whose treatises and opinions all the aforesaid jurists have incorporated in their own works, such as Scaevola, Sabinus, Julianus, and Marcellus, and all others whom they cite, provided that, on account of the uncertainty of antiquity, their books shall be confirmed by a collation of the codices (propter antiquitatis incertum, codicum collatione firmentur).

(CTh 1.4.3)

The Law of Citations authorized the texts of Papinian et al. for use in Roman courts. Codification and authorization of previous authorities, like we see here, was right at home in the courts of Theodosius II and Valentinian III, and recalls any number of parallel scholastic productions of the Theodosian Age detailed earlier. However, the law provides for more than just the authorization of certain sources. It imposes a condition as well: juristic opinions shall be authorized only after they are confirmed through a collation of the books – or, roughly, a confirmation of the wording of the text through multiple independent witnesses. According to the law, this verification is to take place “on account of the uncertainty of antiquity (propter antiquitatis incertum)”; because the authors of these texts were long dead, their opinions had been transmitted and commented upon repeatedly in the centuries intervening, and as a result the precise wording had a distinct capacity for instability. In the Law of Citations we see a Theodosian attempt to aggregate and authorize the work of a scholastic patrimony, and we find a correlated concern for the purity of the textual tradition involved. The law states that the textual tradition must be confirmed because the resource produced will become a codified authority.¹⁰ This is not a garden variety concern for textual purity: the drafters of this constitution, working in the court of Galla Placidia, were concerned with scribal or editorial incursions into these sources precisely because the end point of the project was legal authorization and promulgation of a certain set of juristic texts; it needed to be right.

The Law of Citations does not authorize ancient legal opinions themselves but rather certain texts produced by ancient legal thinkers, and specifically the original wording of those texts. For instance, Paul’s, Ulpian’s, and Papinian’s opinions are equally authorized, but not Paul’s and Ulpian’s commentaries on Papinian: the law explicitly states that the markup/commentary (notae) of Paul and Ulpian on the text of Papinian (in Papiniani corpus) are not to be considered valid, precedential opinions. So, while the law orders that opinions of Paul are to be certified and authorized, even his authentic comments on the text of Papinian are not. In other words, the concern was to establish Papinian’s words as he wrote them, without corruption from later commentary or editorial incursions – even when the incursions in question are the product of another authorized jurist like Ulpian, whose opinions are acknowledged as equally authoritative by the very same statute.¹¹ The law stands within a tradition of the increasing textualization of legal praxis during the Theodosian empire. With the Law of Citations, the court of Galla Placidia placed ancient works of scholarship – and specifically an authorized version of the texts – as the final arbiter of legal orthodoxy and orthopraxy. Suspicion of the documents and archives in question is motivated by the fact that these books will become the final word on legal matters. There is, then, a sense in which I agree with Oronzo Pecere’s suggestion that the act of textual verification and emendation changed in a Christian imperial context:

In a culture which has deconstructed the literary institutions of classical society, replacing the consolidated hierarchies of its authors/authorities (auctores) with biblical texts and commentaries (the center of which, as a result, is the belief that every earthly event or human action carries out a divine plan) it is not surprising that the correction of a book is not simply a technical-scholarly operation, but is conceived according to a theological perspective: in fact, in it the tradition of Alexandrian philology is merged with that of biblical criticism, which had refined the methods and forms of reading texts by experimenting with complex questions of authenticity (Echtheitsfragen). Moreover, it should be noted that for Christians, writing itself, traditionally considered a lowly technical craft (opus servile), becomes a means for moral and spiritual elevation.¹²

Pecere is right to note that in the Theodosian Age the tradition of textual emendation came to be spiritually significant, and significantly concerned with questions of textual and archival veracity. I hope that this chapter and Chapters 6 and 7 serve to demonstrate that the change, both ideological and material, is not unique to Christian books.

The Law of Citations witnesses another significant facet of the Theodosian order of books: problems of discernment, and rules for deciding between authorized voices. The law continues from above:

Moreover, when conflicting opinions are cited, the greater number of the authors shall prevail, or if the numbers should be equal, the authority of that group shall take precedence in which the man of superior genius, Papinian, shall tower above the rest, and as he defeats a single opponent, so he yields to two. … Furthermore, when their opinions as cited are equally divided and their authority is rated as equal, the regulation of the judge shall choose whose opinion he shall follow.¹³

Here the Law of Citations formulates a set of rules and an order of operations through which a scholar can decide between conflicting legal opinions – an eventuality necessitated by the confirmation of a discursive commentarial tradition as legally binding. The solution reached is this: should opinions be evenly split, Papinian’s opinion prevails over the others. However, when two authorized jurists rule together against the opinion of Papinian, their collective opinion shall be judged as superior.¹⁴ The decision falls to judicial discretion if and only if opinions are equally split and Papinian has not commented on the matter at hand.

The Law of Citations circumscribes a judge’s creativity when interpreting legal opinions. A. H. M. Jones famously called it “the low-water mark of Roman jurisprudence” for precisely this reason: because it apparently reduces the resolution of complex legal questions to the counting of heads.¹⁵ But these rules are a solution to a problem of the Law of Citations’ own making: they were necessary because the law identified, verified, and authorized a contentious and multivocal body of scholarship. The reiteration of these rules just three years later in the Theodosian Code, the great high-water mark of post-classical law, suggests either that the depth of the water has been overstated or that the tide of juristic excellence turns on a remarkably short period.

The Law of Citations was promulgated three years before the first constitution calling for the creation of the Theodosian Code, and it stood as binding juristic praxis. In the context of book one of the Code, however, the Law of Citations takes on an even more comprehensive meaning: because of its inclusion, the binding nature of the collected juristic opinions could only be abrogated through a novella. The Law of Citations was no longer read simply in the context of the “mini-code of 426,” but as a programmatic statement for the entire body of law. Its inclusion in the Theodosian Code reiterates the validity of the concerns and the solution reached, but the same problems that attended the “mini-code of 426” – the possibility of forged archival sources and problems of discernment between competing voices – attended the Theodosian Code. Unsurprisingly, the Roman senate instituted similar solutions to these problems.

Like Athanasius’s Concerning the Decrees some two generations before, the Minutes of the Roman Senate Concerning the Theodosian Promulgation (Gesta Senatus) is the cover letter for a dossier. As discussed in Chapter 5, it comprises minutes from the Roman senate in 438 detailing the reception of the Theodosian Code in the West, along with the text of a rescript given by Valentinian III in 443. The rescript is referred to as the Constitution concerning constitutionaries (Constitutio de constitutionariis), and it legislates duties of the prefect of Rome regarding the publication of the Theodosian Code. Here, as part of the proceedings of the Roman senate, we see that the emperor had concerns similar to those visible in the Law of Citations; he was worried that the text of the law would be liable to falsification unless its publication and circulation was tightly controlled.

Therefore, the illustrious prefect of the city (our kinsman and friend, whose duty it is to enforce quite diligently what the Senate has decided for the security of all), shall know that the license to publish copies has been assigned to you; that the production, also, of copies of the aforesaid body of law [the Theodosian Code] shall be provided for at the risk of you alone; that those persons may have no traffic in either the publication or production of copies, since it is certain that the hazard of falsification falls upon you.

(Gesta Senatus 7)

At issue here is not solely the initial editorial work involved in producing the Theodosian Code, but control over reproduction and distribution networks which are particularly vulnerable to obtrusion. If Valentinian’s rescript appears pessimistic about the conduct of scribes and tradents in legal material, his concern merely reflects what is found in the Law of Citations and reauthorized in the Theodosian Code itself. The Minutes of the Senate of Rome, in turn, reiterates the concern for editorial intrusion again, as the collected senators cry out acclimations aimed at preventing falsification of the authorized codex:

“Let many copies of the Code be made to be kept in the governmental offices!” Repeated 10 times.

“Let them be kept under seal (sub signaculis) in the public bureaus!” Repeated 20 times.

“In order that the established laws may not be falsified, let many copies be made!” Repeated 25 times.

“In order that the established laws may not be falsified, let all copies be written out in letters (litteris)!” Repeated 18 times.

“Let no annotations upon the law (notae iuris)¹⁶ be added to this copy which will be made by the constitutionaries!” Repeated 12 times.

“We request that copies to be kept in the imperial bureaus shall be made at public expense!” Repeated 16 times.¹⁷

The two parallel concerns which animate this chapter – suspicion of documents and rules for deciding between authorized voices – reverberate through Theodosian scholarship because they proceed logically from a structure of knowledge in which collections of traditional material are authorized and promulgated in view of universal assent. The fact that Theodosian writers and readers dealt with the same problems in different domains stems from the coherent set of aims and expectations from which each proceeds: universality by way of aggregation, distillation, and promulgation. Different readers and writers dealt with the exigencies the process in different ways, just as the scribes we encountered in Chapter 6 dealt with the peculiarities of aggregative scholarship in a variety of manners. The range of answers that we encounter, however, are all predicated on roughly the same question: “If a multi-vocal tradition is to be transformed into an authorized, aggregative product, how do we know what texts to authorize, and what should we do when they disagree?” The multiplicity of solutions speaks to the coherence of the problems introduced by new dominant scholastic practices in the Theodosian Age. I now turn to the multiplicity of those solutions.

Institutionalized Suspicion of Documents and Archives

By the time he began writing commentaries on the Christian scriptures, Jerome had reached an impasse. He wanted to hold to an ideal: hebraica veritas (“the Hebrew is the truth”). In Jerome’s estimation, the final authoritative version of scripture should be a faithful rendering of the Hebrew Bible into the vernacular. But he had a problem: Jesus in the gospels, as well as Paul in his letters, quote verses and stories from the “Old Testament” that do not exist in Hebrew, and sometimes they build theological scaffolds around Greek translations that are not faithful to the original. He laments that “[t]he evangelists – and even our lord and savior, and the apostle Paul, also – bring forward many citations coming from the Old Testament which are not contained in our manuscripts … but it is clear from this fact that the best copies are those which agree with the authority of the New Testament.”¹⁸ That is, the “best copies” from a theological perspective were at odds with the “best copies” from a philological perspective. Jerome could either censure Jesus or he could dispense with the ideal of the primacy of the Hebrew scriptures. In the end he did neither. Rather, he used the tools of aggregation and suspicion of documents to justify that the New Testament was true even when it expanded falsely on the Old Testament.

Jerome’s solution is visible throughout his body of work, but it is stated perhaps most succinctly in the preface to his Book of Hebrew Questions on Genesis, composed in the early 390s and intended “to refute the mistakes of those who suspect some fault in the Hebrew scriptures (qui de libris hebraicis varia suspicantur), and to correct the faults which appear to abound in the Greek and Latin codices by reference to the [Hebrew] authority.”¹⁹ Jerome attests suspicion of the biblical text in the minds and work of others, and is skeptical himself of the veracity of the Greek and Latin translations available on the late fourth-century book market: translations known as the Old Latin and the Septuagint, and those by Josephus, Theodotion, Aquilia, Symmachus, and Origen. In cases of faulty transmission or imprecise translation of scriptural texts – that is, in the case of competing codes – Jerome’s answer was not further archival work; it was linguistic work.

In Chapter 6 I detailed Jerome’s aversion to offering a single, final, authoritative version of the scriptural truth. Like many adherents of the late fourth-century version of Nicene orthodoxy, Jerome thought that scripture and tradition dually undergird the final statement of truth found in the Nicene Creed even when they are faulty. As a result, he was unwilling to offer anything more than a better translation of the Hebrew along with a more accurate commentary, and to place his work beside the deficient efforts of lesser scholars: an aggregative compendium in codex form.

To enable the student more easily to take note of an emendation, I propose in the first place to set out the witnesses as they exist among us, and then, by bringing the later readings into comparison with it, to indicate what had been omitted or added or altered. It is not my purpose, as jealous people pretend, to convict the seventy translators of error, nor do I look upon my own work as a censure of theirs.²⁰

Unlike other attempts to purify the original text of scripture from corruptions, Jerome’s own approach, partially adopted from Origen and partially created in response to his own scholastic environment, was to lay bare for his reader the fact of the variation and to encourage them to remain skeptical of the ability of translators and the trustworthiness of scribes. Thus he repeatedly offers two versions of the same text and does not offer an opinion on which is correct, for instance at 49:27, where he says: “Although this is a most clear prophecy of Paul the Apostle … nonetheless in the Hebrew it is read as follows.”²¹ Jerome was committed to placing the sum total of the scholarly tradition together.²² In his Preface to the Book of Job, Jerome explicitly claims that scriptural texts – and especially those with fraught transmission histories – should be transmitted with asterisks and obeli intact.²³ In his Preface to Ezekiel, Jerome goes so far as to prescribe scribal practice for copies of his translation: the text is to be written with spaces between words so as to avoid confusion and interpretive failures. His Preface to the Gospels takes a somewhat different tack: he claims that he has attempted to change little from current Latin translations unless the translations reflected a corruption in the underlying Greek or the Latin fails to render its sense. He does this, apparently, because the gospel texts were translated directly from the Greek, and not from Hebrew to Greek and then to Latin.²⁴ Even the text of scripture is liable to censure, but Jerome claims that even faulty witnesses possess authority of one sort or another.

Jerome is perhaps the scholar best equipped to discuss the relationship of Theodosian era scholars to the work of their predecessors because he dealt with a long tradition of translation, and with at least five competing versions of the same text to which he was asked, again and again, to return and translate anew. At the behest of dozens of different patrons he thought constantly about the relationship between his own scholarly output and the work of his disciplinary elders. Though they were written over the course of many years in a number of different locations and institutional contexts, Jerome’s prologues all speak to a singular ideology of scholarship; his commitment to a suspicion of documents and archives was thoroughgoing and ongoing. In his case, the archives are scriptures that are both true and incontrovertibly faulty.

The Proceedings of the Council of Chalcedon

The Proceedings (acta) of the Council of Chalcedon are not user-friendly documents. All told, they comprise nearly 1,000 pages, preserved mostly in Greek, with lacunae filled by reference to the ancient Latin translation. They are not verbatim transcripts of the proceedings of the council of 451 but rather a collection of notes (ὑπομνήματα/commentarii), petitions, libelli, and letters that were edited together with an eye toward validating the case of the prevailing (by definition, “Orthodox”) side of the dispute at hand.²⁵ However, the compiled acta were imperial documents, produced in the court of the Eastern emperor in the immediate aftermath of the council, and circulated along with supporting documents by early 455.²⁶

The Proceedings’ peculiar structure allow insight into the documents brought before the council along with the interpretation of those documents in a cross section of the later Theodosian Age. Each council read out and copied into its records some portion of the records from the previous meeting. At the first session of Chalcedon (451), for instance, acta from the previous council (Ephesus II, 449) were read out before the assembly and entered into the official record. In turn, bishops in Ephesus read out acta from the Home Synod of Constantinople (448) and entered those into the conciliar record. Through this process multiple successive layers came to be embedded within a single document. We are left with a textual nesting doll, where the oldest documents in the Proceedings of the Chalcedon, held in 451, stem from the Council of Ephesus, held twenty years earlier. The Proceedings of Chalcedon thus contain the proceedings of previous councils and information about the way that these documents were read and interpreted. Within the acta we see clerics reading and assenting to records from previous councils, along with bishops resisting the authority of these documents, denying their veracity, and questioning their validity as records of the past. The Proceedings of Chalcedon are valuable because they allow historians to look over the shoulder of bishops as they interpreted imperially authorized documents, and to make inferences regarding the guiding principles of their interpretive gaze. In the acta of Chalcedon we see a fully crystallized suspicion of documents and archives, one that has become part of the institutional framework of interpretation. A few examples will suffice to bear this out. Consider a statement of Basil, bishop of Seleucia, preserved within the acta from Ephesus II:

“This statement that they say I made I did not make in these words (ταύτην ἣν λέγουσίν με εἰρηκέναι φωνὴν ἐγὼ οὐκ εἶπον αὐταῖς λέξεσιν). I am not aware of having said this … ”

Juvenal bishop of Jerusalem said:²⁷ “So, was your statement altered (αὕτη οὖν ἡ φωνὴ παραπεποίηται)?”

Basil bishop of Seleucia said: “I have neither memory nor knowledge of having made it.”²⁸

Here, Basil of Seleucia defends himself against acta from a council that was held just one year prior. The reader may reasonably infer (and perhaps is supposed to infer) that the bishops in the room also attended last year’s council, and that many attendees remembered what Basil said. Nevertheless, as we saw in the Law of Citations, it was the text of the council that was authorized and not the events which the text relays. Witnesses are required to respond to the imperially sanctioned account of the council rather than to any living witness. The locus of truth is textual, and the text’s authority does not lie in its referentiality – in the fact that it points to the moment of actual import, which happened in the past. Rather, the document itself is the authority and it exists separate from the events which it narrates, even when it is faulty.

This startling centrality of documents to conciliar dispute appeared early in the Theodosian Age, as seen for instance in Palladius’s exasperated response to Ambrose quoted earlier: “You are judge, [on account of the fact that] your note-takers are here!” Everyone in the room knew that at future events, human witnesses would be required to answer to the imperially sanctioned, authorized codex of the proceedings and decisions of the council rather than to anyone’s recollection of the event or any other account. Consider a charge of editorial forgery in the statement of Theodore of Claudiopolis at Chalcedon, made while discussing the Synod of Ephesus II in 449 ce:

“Let him bring in his notaries, for he expelled everyone else’s notaries and got his own to do the writing. Let the notaries come and say if this was written or read in our presence, and if anyone acknowledged and signed it.”

The most glorious officials and the extraordinary assembly said: “In whose hand are the notes written?”²⁹

We see here again that the acta themselves serve as the authorized account; accusations of malfeasance must be made on the basis of that codified document rather than against other people who were present at the document’s creation.

The Proceedings of Chalcedon are shot through with a concern, from parties on all sides of the dispute, that official documents have been the victim forgery and editorial malfeasance. As I have argued, they echo rhetoric we see throughout the Theodosian Age. The rhetoric and concerns are shared across corpora, but the solutions sometimes diverge even within a scholastic tradition. For instance, the legal proceeding of Catholics against “Donatists” held at Carthage in 411 show a similar centrality of import given to documents, but in this corpus, attendees built mechanisms of verification directly into the production of acta.³⁰ In fact, the production and verification of documents was so important in 411 that the first session begins with a detailed discussion of the method of transcription and identification of the functionaries called on to perform the task. The solution agreed was as follows. Six functionaries were tasked with recording the proceedings: one scribe (scriba) from the legislature’s office, one scribe from the curator of Carthage, two clerks (exceptores) from the office of the proconsul, one clerk from the office of the vicarius, and one clerk of the legate. These bureaucrats from the governmental apparatus were assisted by dueling secretaries (notarii) – two each from the Catholic and Donatist factions – intended to take down statements in duplicate.³¹ At the conclusion of every statement, the speaker proceeded to the workspace of a notary for each side and signed the statement in his own hand, writing recognovi (“I have inspected”) or subscripsi (“I have undersigned”), often with his full title included.³² While the Proceedings of the Council of Chalcedon admit continually to their own faulty transmission, the Proceedings of the Council of Carthage in 411 insist upon their own verbatim account, and derive their authority therefrom. As Brent Shaw put it with characteristic verve, the document from 411 is “a real gem of hard reportage.”³³

The Proceedings of the Council of Carthage in 411 are among the most self-consciously authoritative texts surviving from antiquity. The importance of documents produced by the council is confirmed from the first page, and quite literally reinscribed on each subsequent sheet. Perhaps most interesting is that scribes transmitting copies of the Proceedings show extreme sensitivity to the materiality of the methods of verification instituted at the council. In copies of this text, each statement is followed by more than just the statement recognovi or subscripsi, as autographs of the proceedings would originally have read. Manuscripts of this text attest to their own status as secondary copies by preceding each mark of verification with the words et alia manu (“and, in a different hand”), indicating that the attendee named wrote recognovi or subscripsi personally, rather than leaving it to the scribal stenographer.³⁴ Subsequent copies of these documents, in other words, attest to their derivative status, similar to the derivative status of copies of the Theodosian Code discussed in Chapter 5, by indicating that the original edition was composed by multiple different hands – that the scribal multivocality which was intended as a mark of authenticity has been lost in transmission.³⁵

Scholars in the Theodosian Age were not the first to show concern for the purity of textual transmission: neither in the domain of Christian theological dispute nor anywhere else. I am not arguing that suspicion of documents is a Theodosian or a Christian innovation. It is not. Rather, the case that I present here regards the relationship between the institutionalization of suspicion of documents and prevailing scholastic practices in the Theodosian Age. People have been skeptical of documents as long as there have been people and documents. The change that I document is this: the centrality of traditional documents for the creation of reliable knowledge led to the necessity of scrutiny. The prevalence of this trope across the Theodosian scholastic landscape results from practical shifts in the way that most scholars went about their tasks in the late fourth and early fifth centuries. I argue that historians can and should take note when precedented actions are undertaken for unprecedented reasons. There are limitless reasons to be skeptical of documents, and yet Theodosian readers and writers, by and large, were skeptical of documents because their scholastic projects were built on the aggregation of archival sources. Mine is not a whig history of practice, in which all of the features of the Theodosian Age order of books necessarily flow from a single, motivating shift. Nevertheless, there are certain scholarly practices that imply others. A structure of knowledge in which archival documents are central to the production of truth invites, or perhaps demands, scrutiny of those sources.³⁶

Some scholars during the Theodosian Age adjudicated disputes between opposing authorities on a case-by-case basis. Some scholars, such as those involved in the production and dispute over the documents before the Council of Chalcedon in 451, engaged in intensive archival work and interpersonal dispute to adjudicate problems of transmission of documents manufactured by the imperial chancery, sometimes as little as twelve months prior. Scholars at the Conference of Carthage in 411 knew that the Proceedings of the council would become part of a tradition obsessed with archives, so they implemented unprecedented mechanisms of verification from the first moment of the document’s production. In all of these examples we see scribes and writers self-consciously paying attention to the problems caused by work predicated on archival sources that themselves have unclear transmission histories.

As I showed at the beginning of this chapter, Jerome was unwilling to make a final, one-size-fits-all rule about which version of the scriptures was true and therefore authoritative. It would be hard for Jerome to stick both to his ideal of hebraica veritas as well as to accede to the authority of the words of Jesus and the Apostles by saying that anything that is not found in the Hebrew is false. He would end up censuring either the Hebrew Bible or Jesus; neither is a good look. Some scholars, however, had no such compunction about strict rules for deciding between opposing authorities. We saw one example of such an algorithm in the Law of Citations. I now turn to another.

The Theodosian Talmud

The rabbis of Late Antiquity were scholars, and they worked and lived within a self-aware system of scholastic disputation. This is, perhaps, the full extent of clear similarities between rabbis and the other scholars engaged in this book. Among the greatest challenges in studying rabbinic literature is that basic questions remain unanswered by the tradition itself. As Yitz Landes observes:

Of the various difficulties facing the student of classical rabbinic literature one immediate one, that is for the most part unsolved by the evidence provided in the corpus itself, is what this corpus even is and how it came into existence … No classical rabbinic text ever discusses its origins. At best, the Talmuds offer sporadic statements concerning the authorship of the Mishnah, Tosefta, and tannaitic midrashim, but without offering any explanation as to why they were compiled.³⁷

It was not until the Geonic age, around the turn of the millennium, that rabbis began to engage in sustained historical theorization as to the “when and the why” of the rabbinic corpus. The tradition was completely devoid of the sort of programmatic statements that I have engaged in this book thus far: texts like the constitutions calling for the creation of the Theodosian Code or the explicit theorization as to the “how” and the “why” of theological disputation penned by Athanasius, Ambrose, and Jerome. The rabbinic corpus offers no leg-up to understand its intellectual project, and at times the text seems to be purposefully obtuse.³⁸ Any contextualization of rabbinic materials will necessarily be speculative – if the question was interesting in antiquity, we have no record of anyone asking it.

And yet, it is the rabbinic material most clearly situated in a Roman province of the Theodosian Age that glimmers with intriguing tendencies that are both singular among rabbinic texts and conceptually similar to the developments that I have traced in the wider realm of Theodosian Age scholastic production. Namely, the two facets of Theodosian codification engaged in this chapter appear in the Palestinian Talmud as well; there, too, scholars deal with the effects of textual authorization and answer the same concerns seen in other Theodosian corpora with similar intellectual strategies. By placing the Palestinian Talmud in its Theodosian scholastic context, we may recognize it as a particularly Roman and Theodosian project. The correlation suggests that practices developed within a Christian empire, proffering Christianized intellectual practices across the scholastic landscape, came to inflect even the scholarly production of “rabbis [who] proclaimed their alienation from normative Roman culture in every line they wrote,” as Seth Schwartz rightly argues.³⁹ A full discussion of the ways in which a peculiarly Theodosian structure of knowledge inflects the Palestinian Talmud is beyond the scope of this book. Here I offer here just two examples, which I argue are illustrative of the place of the Palestinian Talmud among Roman provincial literature.

The Palestinian Talmud (sometimes referred to as the “Yerushalmi”) is structured as a commentary on the Mishnah and reached its final form sometime early in the Theodosian Age. The text is layered, woven together in a mixture of Hebrew and Jewish Palestinian Aramaic. While some statements in the text are attributed to named authorities, others remain anonymous. The largest part of the anonymous material comes from the general narrative voice of the Palestinian Talmud – known to medieval and contemporary scholars as the “Stam” (סתם, translated “anonymous,” though literally “stop” or “seal” — a metonymic use to refer to the final layer editors that “seal” or “close” the book).⁴⁰ The last generation of scholarly sources named in the Palestinian Talmud come from the so-called fifth generation: rabbis who lived and worked in the second half of the fourth century. Whether the Stam of the Palestinian Talmud should be attributed to the final generations of the named sages, or whether it is a subsequent redactional layer, its temporal context is squarely Theodosian; more precision is impossible given the current evidence.⁴¹ This is a work of aggregative scholarship that crystallized into its current form during the Theodosian Age and, as I argue later, the suspicion of documents and rules for deciding that we know from other Theodosian Age works are unequivocally attested only in the latest layer of the Palestinian Talmud.

In its tractate on Heave Offerings (תרומות), the Palestinian Talmud offers a set of rules for deciding between authorized voices.⁴² The rules are given by Rabbi Zeira, a resident of Roman Palestine during the third generation of rabbinic scholars:

R. Yose said in the name of R. Johanan that [in disputes between] R. Yose and his colleagues (רבי יוסי וחביריו), practice (הלכה) follows R. Yose … R. Zeira [and] R. Jacob bar Idi in the name of R. Johanan [say: in disputes between] R. Meir and R. Simeon, practice follows R. Simeon. [In disputes between] R. Simeon and R. Judah, practice follows R. Judah. One need not mention that [in disputes between] R. Meir and R. Jehudah, practice follows R. Jehudah. R. (Ab)ba bar Jacob bar Idi in the name of R. Jonathan: [between] R. Meir and R. Simeon, practice follows R. Simeon. [In disputes between] R. Simeon and R. Judah, practice follows Rebbi Judah, and one need not mention that (ואין צריך לומר) [in disputes between] R. Meir, R. Jehudah, and R. Simeon, practice follows R. Jehudah. And from this you infer that (ומינה את שמע) [in disputes between] R. Judah and R. Simeon, practice follows R. Judah.⁴³

For the first time in the rabbinic tradition, these rules offer an internally consistent algorithm for deciding between scholarly opinions of the Tannaim (“Repeaters”) who lived many generations before – mostly in the Antonine Age. The rules are particularly interesting in their Theodosian context because the solution reached to the problem of codified authorities who occasionally disagree is remarkably similar to the solution reached in the Law of Citations mentioned earlier. Yehudah Brandes has analyzed this passage in the context of the Palestinian Talmud as a whole and has found that not a single sugya (“section,” roughly) contradicts these simple rules for deciding between conflicting opinions that were initially proposed by R. Yohanan in the late third century.⁴⁴ Building on Brandes’s work, Richard Hidary demonstrated that it is likely only the Stam – the anonymous, Theodosian Age layer of the Palestinian Talmud – that unequivocally endorses rules for deciding between voices that were first suggested over a hundred years before.⁴⁵ The Theodosian layer of the Palestinian Talmud embraces a common answer to an obvious problem of codification, one known from other Roman sources of the period. Rendering the insight more interesting is this: the Sassanian recension of the same text takes a radically different approach.

The Babylonian Talmud (sometimes referred to as the “Bavli”) was compiled in Sassanian Iraq some two centuries later and it certainly has a number of rules for deciding, as noted by Dov Zlotnick, including a principle of הלכה כסתם משנה (“the law is according to the anonymous Mishnah”) by which the anonymous voice has final say when rabbis disagree. Even later commentators such as Rashi (late eleventh century ce) saw principles in the Bavli such as the notion that contradictory regulations should be preserved even when they come down in the name of a single teacher, or from a teacher who changed his mind – a principle that “would have been understood, if not praised, by Roman jurists.”⁴⁶ But the Babylonian rabbinical community had a different approach than the rabbis of late Roman Palestine to algorithmic rules for deciding like we see in the Law of Citations.

The Babylonian Talmud contains a parallel to the Palestinian Talmud’s rules on deciding in its own tractate on Communal Mixing (ערובין). The sugya comprises two parts. First, it repeats the same rules for deciding recorded in the Palestinian tractate on Heave Offerings, quoted earlier. In the Babylonian Talmud’s version of this tradition, the “rules for deciding” offered and embraced by the Palestinian Talmud are followed by a sustained discussion, including six countervailing cases where the rules are shown to be riddled with exceptions – with the implication being that the rules themselves are useless. As Hidary concludes:

Thus, the two parts of the sugya actually stand in tension with each other, one stating the rules and the other questioning them. The first part is parallel to the Yerushalmi presentation, while the second part is unique to the Bavli. The Yerushalmi never doubts the authority of the rules, not in [Heave Offerings] nor anywhere else … The Bavli, however, does quote the controversy and leaves it open-ended, suggesting that the Bavli editors themselves saw reason to doubt the categorical application of these rules.⁴⁷

While there is reason to believe that the compilers of the Bavli’s anonymous redactional layer knew and responded to its redactional counterpart in the Palestinian Talmud, the Bavli as a whole arose out of and crystallized in a remarkably different intellectual and political milieu.⁴⁸ Here we see one vestige of the intellectual contexts of these two remarkable projects; while the Roman provincial compilers embraced rules for deciding, Sassanian rabbis were – at the very most – ambivalent about them and inconsistent in their application.

The Stam of the Palestinian Talmud unequivocally supports the rules for deciding and applies them uniformly. As Richard Hidary argues, it is only the Theodosian stammatic layer which considers these rules to be ironclad. Parallels with broader Roman scholastic aims and methods are not restricted to the invocation of rules, however.⁴⁹ Like the Law of Citations, the Yerushalmi’s Stam also witnesses a suspicion of authorized traditions, and offers creative solutions to the problem of intermittently unreliable transmitters. The tractate Fast Days (תעניות) undertakes a discussion of private fasts and their relationship with a particular fast day, the Ninth of Av:

It has been taught, “The Ninth of Av that coincided with the eve of Sabbath, a person eats even an egg and drinks even a cup so that this person should not enter the Sabbath while fasting,” so the words of R. Yehudah. R. Yose says, “He should complete the fast.” R. Zeirah in the name of R. Yehudah, R. Ba, R. Imi bar Ezekiel in the name of Rav, “The halakha is in accord with him who says that he should complete the fast.” Why did he not simply say, “Practice follows R. Yose?” There are reciters who recite and swap the words of the sages.⁵⁰

For my purposes here, the problem under discussion in this sugya is less relevant than the solution adopted by the Yerushalmi’s redactional layer. The core issue is that two authorities disagree. In normal circumstances, a rabbinic student following along with the discussion, regardless of the particular issue at stake, should be able to predict the solution reached by the text: namely, that “The halakha is in accord with him who says that he should complete the fast.” This is, after all, simply an application of the rules for deciding laid out in the passage of Heave Offerings discussed earlier: “[In disputes between] R. Yose, and his colleagues, practice follows R. Yose.”⁵¹

The Palestinian Talmud is famously terse, and the anonymous redactional layer (Stam) rarely offers clarifying information that can be gleaned from the discussion or that should be known by the rabbinic student already. Accordingly, the Stam clarifies why it is that such an obvious answer, in this case, is worth recording in full. An explicit ruling is necessary because some scholars “swap the words of the sages,” such that the wrong name might be attached to halakhic guideline leading to an improper judgment made on the basis of the rules for deciding.⁵² Here, in the voice of the Palestinian Talmud’s Theodosian redactor, we see the complexities of applying rules for deciding even in a tradition of tightly controlled legal recitation where editorial obtrusion is difficult, to say the least. The application of such rules is intimately bound up with suspicion of the tradition itself, and extra care is taken in this and other sugyot to ensure that rules are applied to the correct tradition, because the existence of any instability in the tradition renders the rules for deciding essentially worthless. In this case, suspicion of documents caused the famously terse redactor of the Palestinian Talmud to become uncharacteristically loose lipped.

Hayim Lapin is right to stress that “what makes the rabbinic movement so striking is the juxtaposition between an apparently thoroughgoing Romanization of the subject population and the emergence, precisely where we expect Romanization to be most effective, of groups of men who organized to express their non-Romanness.”⁵³ And yet we glimpse glimmers of regional variability in these divergences between Sassanian and Palestinian rabbis; we start to see Roman ways of knowing finding expression even in a corpus of material that disclaims its Romanness at every turn. Amit Gvaryahu argued recently that rabbinic laws on usury show that Palestinian rabbis held a shared concept of the scope and definition of a “loan” with the Roman jurists, even while they rejected the substantive law embraced in Roman courts:⁵⁴

The difference in the substantive law … is what enabled the rabbis to say, and in all likelihood to sincerely believe, that their law and the Roman law were “different:” that they did not follow the laws of the Romans, they did not avail themselves of their courts, they did not abandon the laws of the Torah and of their ancestors. They were upright in following the commandment, “and you shall not follow their laws.” But at the same time, some rabbis at least needed to be able to say that the Torah was, in fact, a law, and as such it shared much in discourse, scope, heuristics, and definitions, with the law of the Romans … But we should also bear in mind that for its adherents, rabbinic law was at its core not “Roman.” Borrowing and structural similarities were, for the rabbis, a way to effect distinctiveness.⁵⁵

My argument here is similar to Gvaryahu’s, but I see borrowing on a scholastic level in addition to a structural one. The Palestinian Talmud is thoroughgoingly out of step with the other works of Theodosian Age scholarship engaged in this book. For one, it wasn’t written down. The idea of an authoritative oral tradition that stood beside scriptural material was, quite literally, anathema to Nicene Christians. Yet the Yerushalmi’s compilers were part of the Roman world that they rejected in part.⁵⁶ Bertyus (modern-day Beirut), the epicenter of Roman law in Late Antiquity, lies just over 100 miles north of Caesarea Maritima, on the Levantine coast. The form and content of the Palestinian Talmud exclaims its singularity on the Roman scholastic landscape, but some of its underlying assumptions about the production of authoritative knowledge betray the tradition as arising partially within a Theodosian Age scholastic framework, facets of which we have seen time and again over the last six chapters. This scholastic framework has a provenance, too: it is inflected by Christian ways of knowing, forged in the fires of doctrinal controversy. The effect of my analysis, too, is to help situate the Stam of the Palestinian Talmud in its Theodosian context, to stress both the coherence of some facets of its method within the Roman scholastic context of the late fourth century, and to help differentiate ideologically, and perhaps also temporally, between the stammatic layer of the Yerushalmi and the last generation of named sages.

Conclusion

This chapter has attempted to demonstrate the ways in which Theodosian Age scholars read and interpreted differently from their predecessors because of exigencies related to the aggregative format. From the Law of Citations to the acta of church councils and even in the Palestinian Talmud we see Theodosian readers grappling with the fact of codification and employing remarkably similar intellectual strategies in reading and interpreting intellectual products of the late fourth and early fifth centuries. Rules for deciding and an institutionalized suspicion of documents and archives proliferated through the scholastic landscape in response to the changing formats and aims detailed in Chapters 5 through 7: “new readers of course make new texts, and their new meanings are a function of their new forms.”⁵⁷

Conclusion

It is a characteristic of human thought that our concepts do not stay put behind the neat logical fences philosophers like to erect for them. Like sly coyotes, they slip past these flimsy barriers to range far and wide, picking up consorts of all varieties, and, in astonishingly fecund acts of miscegenation shocking to conceptual purists, leave offspring who bear a disturbing resemblance to the wayward parent and inherit the impulse to roam the old territory. The philosophical guardians of these offspring, trying to shake off the taint of sexual scandal but feeling guilty about the effort, don’t quite know whether to cover up a concept’s pedigree or … deny that it matters.¹

I have attempted to trace a constellation of ideas about truth, and how a variety of late ancient scholars thought about, and went about, bringing it to light. Even if truths are unchanging, there is a history to the way that people have sought to access it. That history is obscured when modern disciplinary boundaries become wardens of historical imagination, limiting our estimation of ancient networks of influence. I have argued that the rise of Christianity in the Roman empire caused a revolution in meaning-making, and that as Nicene Christians came to hold positions of imperial power, their argumentative methods and aims found expression in domains of knowledge production far removed from theology.

I argued that Christians were not always “people of the book” – that, instead, antiquity witnesses a spectrum of Christian approaches to finding truth. Some preferred to understand truth as something latent in textual traditions: letters and “memoirs of the apostles” whose text will yield an abundance of universally binding precepts if read with the right set of assumptions and hermeneutical strategies. Others, such as Tertullian, saw truth as fundamentally pre-textual, while others still found textual interpretation to be an impotent distraction; the author of the Gospel of Truth asserted that truth could not be contained in language, let alone on parchment.

But a group of textually interested Christians were the recipients of imperial largesse from an emperor who was, after all, Roman, and concerned with the same “peace of the gods” that had preoccupied emperors before him. Constantine’s obsession with unity, and with the relationship between doctrinal harmony and heavenly favor, led him to demand a solution to a theological problem roiling the clerical elite: a problem predicated on the idea that scriptural texts held cosmic truths and that those truths were accessible through close scrutiny. But, beginning already in the 320s, Constantine and his advisors found that the underdetermined nature of scripture itself frustrated any attempt to divine universal doctrine solely through textual interpretation. Factions arose, each claiming different interpretations of the same text. A group of clerics debating the relationship of the Christian Father to the Son found scriptural interpretation incapable of answering the question with satisfactory finality, and disputants on either side of the debate invented new tools to answer the question that traditional methods were unable to adjudicate.

Theological scholars conceived and refined these tools during a generation spanning the middle decades of the fourth century, while Christians gained stature and their numbers swelled across the empire. By the time that Theodosius I ascended to the purple and instituted a violent purge of anti-Nicene voices,² the ground rules of theological discourse had fundamentally shifted; Christian scholars of the late fourth century went about producing knowledge differently from their predecessors, and it was these same Christian scholars who came to hold the reins of power across the empire under the aegis of Theodosius I and his dynastic offspring.

Ideas, including ideas about how one might get at truth, are remarkably fecund. I have argued that Nicene Christian scholars came to power in the Theodosian empire armed with scholastic practices inflected by doctrinal controversy, but that this peculiarly Christian structure of knowledge did not long remain solely the purview of theologians. A manner of thinking about truth – including a fundamental interest in universal truth itself as a worthwhile pursuit – found its way from the rarified air of theological disputation into other domains of knowledge. Across the ideological and intellectual landscape of the Theodosian empire, scholars searched for universal truths in their own areas of expertise, and they did so using a method of aggregation, distillation, and promulgation that was initially conceived to settle a thorny theological dispute. Christian and Traditionalist scholars alike took up this method in works of law, history, and miscellany. Glimmers of it can even be seen in the Palestinian Talmud, helping us to situate that production as particularly Roman provincial literature.

The proliferation of a scholastic regime that began as a theological tool through “secular” domains is an aspect of Christianization. It shows us how dominant modes of thought can be ported from one field of inquiry to another in the same way that, for instance, the earliest critical scholars of the bible used advances in genetic and evolutionary theory to understand the relationship between texts and the proliferation of “heresies” in the early Jesus movement.³ In Late Antiquity, legal scholars used the dominant scholastic framework to craft the Theodosian Code. Given the Christian foundations of that framework, we could conclude that, therefore, the Theodosian Code is a Christian production. Alternatively, we could say that the Christian/non-Christian distinction fails in this context. We could contend that, if the adjective “Christian” is to have any analytical purchase, it must be capable of making a distinction; because the methods used to produce the Theodosian Code were dominant, we might argue that it doesn’t mean anything – it doesn’t make a difference – to say that the use of a “Christianized structure of knowledge” in the framing of the Theodosian Code serves to categorize the work as Christian.

What I want to say is that the answer to the question depends on the analytical interests of the person asking. When describing the great scholarly productions of the Theodosian Age, the Christian/non-Christian distinction may be a distraction, or a distinction without a difference. At the same time, there is value in understanding the history of practices which inflected the production of the first universal codification of Roman legal truth, or a great late ancient work of bookish antiquarianism such as Macrobius’s Saturnalia. As I have told it, that history is inflected by doctrinal disputes of the early fourth century, and in this sense the history of juristic practice, antiquarian method, and Christian theological disputation are intimately intertwined – not to mention historiography, military history, or any of the other domains of Theodosian knowledge production detailed in this book. I have tried to trace the inter-implication of Christian ways of knowing and Roman modes of knowledge production, and to show that Christian doctrinal disputes affected ancient people even when those ancient people did not know, or care, about the theological truths under discussion.

Historians can ply their trade without detailed knowledge of the history of method. Countless scholars of antiquity write beautifully compelling, methodologically sound historical accounts without knowing the ins and outs of Prussian academic culture and nationalist fervor that initially animated the methods that we currently employ. Historians can perform intensive, virtuosic post-structuralist analyses deeply indebted to the “literary turn” without any knowledge of what happened in Paris, California, and elsewhere during the 1960s and 1970s.⁴ So, too, could a Theodosian Traditionalist, or Christian, write a miscellany, history, military handbook, or code of law that employed Nicene Christian methods even if they had no knowledge of the contours of the Nicene controversy itself. Nevertheless, the history of method matters. This is the argument that I have made: that there is a history to how people think about producing valid knowledge, and in this instance, understanding the theological disputes of the fourth century helps us to contextualize the scholastic field of the fifth.

I could have told the story in any number of ways. I have chosen to tell this story in this manner because I think that it helps to elucidate a number of fascinating shifts in Late Antiquity that reverberate even today. My major focus, on theologians and jurists and the shared methods between them, is not exclusive of other scholastic network entanglements during the Theodosian Age. Rather, theologians and jurists present a potent test case, helping to clarify the extent of methodological exchange across ancient disciplines that today are studied in very different corners of the academy. The Theodosian Age reverberates in contemporary society most potently, perhaps, from the epistemic overlap in juristic and theological scholarship. We clarify the notion of law as a fundamentally textual and interpretive discourse, for instance, by understanding a time when it was not, and by investigating the circumstances in which law codes first started to look like bibles, and vice versa. The strange, fetishistic power of books in contemporary American discourse, in which the final act of presidential investiture is accomplished with a politician’s hand on a bible, has part of its roots in the conflation of code, codex, and codification explored here, and the institutionalization of material, biblical power that spread through the Roman empire of the late fourth century. The extraordinary durability of these ideas has obscured their complex genesis in Christian Rome of the fourth and fifth centuries. By diving deep into the literature and material of the period, we may yet uncover some pearls of great price that help to understand what it means for a society itself to “become Christian.”