Generating Bible: Many Scriptures into One Canon, One Codex from Many Scrolls

for Wistar Institute meetings in Berkeley CA, Nov 2011, by Robert A Kraft


Currently, the oldest complete Bibles that have survived come from the 4th century of the common era – from the time of the Roman emperor Constantine (died 337), the church historian Eusebius (died about 339), and the Alexandrian bishop Athanasius (died 373). The best known of these are two damaged Greek books called “Codex Vaticanus” (also here) and “Codex Sinaiticus” in which the various sections of handwriting by the various copyists have been dated to the mid 300s by paleographers and attributed to Egyptian origin – unfortunately, such manuscripts do not bear composition or “copyright” dates. Here is a page of Vaticanus; and here a page of Sinaiticus.

These codices once included, with some loose ends (and damaged portions), both the collection of Greek Jewish scriptures known now as “the Septuagint” or “the Greek Old Testament,” and the collection called “the New Testament,” although the exact contents and order of books differ in interesting ways  between the manuscripts and with the "canon" list produced by Athanasius in the year 367. Unfortunately, no table of contents has survived from either codex. Vaticanus does not include the books of the Maccabees, and the last part of the codex is missing where we might expect the Pauline Pastorals and/or Revelation, and any additional works. Sinaiticus is lost for most of the first half, but does include 1 and 4 Maccabees, and contains the Epistle of Barnabas and Shepherd of Hermas at the end of the "New Testament" collection. Athanasius explicitly rejects Wisdom of Solomon, Wisdom of Sirach, Esther, Judith, and Tobit, and does not mention any books of the Maccabees, or Barnabas (but knows of Hermas and the Didache as useful but not top level), and in general follows the Vaticanus order for the rest, with some interesting differences. It is clear that despite significant progress in formation of "the Bible," such matters as order and extent were not universally "fixed" in Christian usage by that time, even in Egypt. 

The fourth century represents the climax of a major technological shift in Christian bookmaking, from the earlier period of horizontally oriented scrolls containing individual works (or small collections such as the “Minor Prophets” or the letters of Paul) through the transition to small codex books of about the same length as a single scroll (e.g. Genesis, Matthew), or as a few scrolls together (e.g. the four gospels), to the “mega codices” capable of holding large numbers of works. This sort of shift was in some ways analogous to what happened with the invention of the printing press in the 15th century, or to what is still happening with digitization of texts and expansion of data storage in our own day. More people had easier and quicker access to larger amounts of material. With the development of mega codices, an entire library of scriptural writings could be consulted in a short amount of time and with manageable effort. Accessories such as indexing and cross-referencing (e.g. the "Eusebian Canons" to locate parallel materials in the gospels) were also developed, along with other such special “apps.”  Conceptually, it then became possible to point to a single codex book as the locus of scriptural authority, and to speak of “the Bible” as a tangible physical object that could be viewed and moved from place to place with relative ease.

 Prior to this technological marvel of the biblical mega codex, Christians – and Jews as well – were not without the ability to construct and discuss authoritative collections of “scriptures” (almost always in the plural). Indeed, the Greek and Roman scroll worlds were full of authored writings comprised of collections of multiple scrolls [#25], such as Homer’s Iliad (24 volumes/scrolls) and Odyssey (24 volumes/scrolls), as well as other collections of various sorts (e.g. treatises, letters). This situation even caused some concern in legal discussions [#30]. In Jewish circles, the “Pentateuch” ("five books of Moses”) is an obvious example of several scrolls under one title, as are the four books of Samuel-Kings, and the five “books” that make up the Psalms. That it was difficult to preserve textual integrity in such conditions is clear from the results of textcritical study of the extant manuscripts, where “text types” often switch between related books, or even within them (Samuel-Kings is a notorious example). Not that it was impossible to maintain physical unity of a sort, on library shelves, for example, as imagined at Alexandria . But the challenge of maintaining such unity over time, and space, was complicated, as the extant manuscripts attest. What Christians now call "the Bible" would have required at least 46 such individual scrolls.

The mega codex solution was also of limited applicability, given the costs involved in producing and preparing such a volume. Papyri was too brittle and bulky to be practical for such extended service, while parchment was suitable but also quite expensive – how many animal skins and how many copyists over how much time would be needed to produce something like those famous 4th century codices? Originally Vaticanus probably contained almost 800 double sided folia with three columns on each side (1588 pages, probably requiring more than 350 animals for the skins), and Sinaiticus about the same, mostly in 4 columns per page (400 folia of which are extant). It is no wonder that even in subsequent centuries, such copies of the entire Bible (called “pandects”) were rare, and Jewish and Christian scriptural materials tended to circulate in more limited forms such as the Pentateuch, the Psalter, the Prophets, the Gospels, and the like.  Book shelves continued to function and maintain some sort of organization, as we can see from various representations that have survived -- e.g. in 5th century Ravenna, or in this 8th century Latin pandect. Just who would have owned mega codices, or even extensive book cabinets, is no longer clear. Did every church or synagogue possess its authoritative literature in one of these forms? We do not know, but I suspect not. Indeed, as we will see, it may have been due to the government sponsorship of Constantine that massive biblical anthologies in codex format came into existence.

Still, a concept of authoritative scriptural literature was already present and continued to function even without the physical realities of a biblical codex or a cabinet full of scriptural works. The determinative device used was the list. We know of formal lists, such as reportedly used by Melito of Sardis around the end of the second century, or by Origen of Alexandria and Caesarea a generation or two later. Even earlier, at the end of the first century, the Jewish historian Josephus knows of a concrete number of books and various categories into which they fit, although he does not give exact individual titles for everything. This resembles ways of cataloguing known from other Greek examples, by type of literature, as seems to have been done with the great Alexandrian library (e.g. Callimacus). Whether Josephus, with his priestly connections in Jerusalem, is telling us how things were actually organized and stored in an official Jewish library associated with the Jerusalem Temple cannot be determined. Similar categories to those he used are also found in the prologue to the translation of Sirach, from around the end of the 2nd century before the common era, in some of the Dead Sea Scroll texts, as also in some early Christian works. The list, like the book cabinet (scrinium) could function independently of any specific book format (scroll, codex) for the relevant materials. And the list doubtless sometimes functioned in connection with tags or labels placed on the individual scrolls or codices as in this reconstruction.

So, while the concept of “the Bible” as a single physical book depended on the technological development of the mega codex (which brought with it specificity of content, with reference both to titles and to order, and also the possibility of identifying and labeling the development of “text types”), the idea of an authoritative canon of scriptural writings could develop as well through lists and/or through physical storage options such as the cabinet (scrinium, also here) or even the earlier and more portable scroll box (capsa), which is widely attested [#15.2], even when codices are also in use [#16] [#17] -- all of these formats survived for a long time alongside each other.  While the idea of canon (rule, measurer) implies some sort of authority behind the standard, it is seldom clear how such authority was determined and exercised. Lists and collections could be constructed by anyone, and while mega codices required special economic circumstances combined with appropriate writing skills, it is not usually clear whose authority they represent.

In the early 4th century the emperor Constantine requested of Eusebius 50 copies of the “divine” writings (θείων ... γραφῶν) for use in churches of his realm (often built with encouragement from his mother Helena), although the exact contents of such codices is as unclear as the actual authority structure involved – who determined the details, Constantine or Eusebius, or someone else? We have already noticed that significant differences in content and order existed between the two surviving mega codices of that period, which suggests that considerable latitude was present in their construction. Later in the 4th century, we also saw that there was a list attributed to Athanasius, off and on bishop of Alexandria, instructing those in his charge what scriptural books to read as authoritative, and which not  –  which suggests that even then, there were reading practices that needed correction. We also noted that Melito, bishop of Sardis in Asia Minor (west Turkey) around 180, writing as a person of authority in a world still dominated by scrolls and small scale codices, has not only left us a list of Jewish scriptures but had investigated in what order those books should appear. Apparently these matters of scope and order were problematic in the churches of his jurisdiction. In the early 3rd century, Origen with his list has a different claim to authority, as a teacher and scholar more than as an ecclesiastical cleric. It is even less clear for whom Josephus speaks as he lists the special writings in his defense of Judaism to attacks from outside. Discussions of the rise of the Rabbinic movement(s) in Judaism tend to ascribe to "the rabbis" an authority that may not have been present in the earlier period, of which we can speak only tentatively due to the nature of the sources. The situation is extremely complex. The evidence from the fragmentary Dead Sea Scroll materials for the question of Jewish scriptural “canon(s)” probably will never be established with any clear consensus, so we muddle along, struggling for clearer definitions and usually concluding with whatever suits, or at least does not overturn, our often unexplored assumptions.

What do we know for sure about the steps along the way from early hints at scriptural authority to the full blown collecting of scriptures into mega codices with canonical implications? Within Jewish and Christian scriptural writings themselves, there are sometimes references to earlier works that presumably lend authority to what is being presented [also here and here]. There are also claims that divine authority lies behind certain words or activities. Of course, anyone can say – and many do say – that they represent the voice or will of deity. It is when such claims get preserved and embedded in an ongoing influential context (such as a socio-religious community) that they can become authoritative for those in that context who submit to the authority. And there can be many such contexts, both simultaneously and sequentially. The models of Jewish or Christian “orthodoxy” that emerge from the middle ages and then splinter in various directions developed their own institutional lines of authority “in the day,” that permit us, with appropriate caution, to speak in deceptive generalities of “the classical Jewish” or “the classical Christian” view or position on this or that. But there is no reason to believe that such observations applied to earlier times or to everyone we might want to include under the label “Jew” or “Christian.”

What sort of “Jews” produced and left us those “Dead Sea Scrolls”? Clearly there were rules and structures of authority as reflected in the “sectarian” writings such as the "Manual of Discipline" and the "Community Rule," but do we learn anything from them to help us with constructing pictures of the Judaism (or Judaisms) from which Christianity (or Christianities) developed? Making scriptural excerpts was practiced, as later with Melito and others. Psalms of various sorts were a big item. Obviously the Enoch cycle and the book of Jubilees were popular in that Dead Sea Scroll setting and elsewhere. Does that mean they were considered authoritative (i. e. “scriptural”) at that time? If so, does that suggest that a Philo or a Josephus necessarily would have known such works? Or a Jesus or a Paul? This sort of approach, straining to fill in gaps, interesting though it may be, is full of historical dead ends and non sequiturs. A basic question that always needs careful consideration is “why do I (or why should I) care?” as well as "what did I learn?"

Similarly with the development of early Christianity – or perhaps, Christianities: Looking back from the classical constructions of the middle ages, we are faced with claims and questions that are quite anachronistic and misleading. “What was Jesus’ Bible?” or “Paul’s?” In the fourth century, as we have seen, Athanasius of Alexandria (and Egypt) had a Bible (at least conceptually) as did the constructors of the great Bible codices, but in the first century, neither Jesus nor Paul could have had a Bible in the same sense: nor Philo (despite his many quotations and essays), nor Josephus (despite his list and narrations), nor the earlier translator of the Sirach preface (despite his categories), nor the DSS depositors (with their plethora of known and previously unknown writings). We learn [#20.2] of a wide variety of writings, both Jewish and Christian, from the references in the preserved literatures of the pre-4th century Christians (e.g. Justin, the two Clements, “gnostic” texts, Origen), and even of some efforts to preserve or restore textual integrity (e.g. Justin on Aquila, Origen's Hexapla), as had been true in Alexandrian study of Homer for generations past. We even have various fragments of known and hitherto unknown Jewish and Christian writings  preserved in the Dead Sea Scrolls and the papyri. But connecting those dots does not produce a coherent picture, largely because we usually are not dealing with the sorts of institutional authority that came later on, with its attempts to control such flights of variety (see, e.g. the Priscillian controversy in the 4th/5th century).

From scroll to codex: Among the factors that set the stage for the concretization of “holy scriptures” as “the Bible” in the 4th century churches of Constantine and his Roman successors are two that I would single out for closer attention: the development of the codex, and the formalization of the scriptural canon, in addition to the development of effective authority structures.

With regard to the formation of literary works in the Greco-Roman worlds, it could be said that “in the beginning was the scroll.” That may be too simple a statement, insofar as writing and record keeping also took place in other formats [#28], such as on wood and stones and other hard surfaces as well as in “notebook” form with pages of wax covered wood (for easy correcting and/or reuse) or more flexible materials (papyrus, leather, cloth). But for “published” as well as less formal records and private writings, the roll format seems to dominate in the hellenistic period, from the time of Alexander the Great in late 4th century before the common era through to the turn of the eras. We first learn of the use of the codex format from a Roman satirical poet, Martial, in the last quarter of the first century of the common era, and a few fragments from codices have survived from that period or slightly later. For whatever reasons – convenience, cost, marketing, novelty, etc. – the codex format caught on in the Greek and Latin worlds over the next four centuries, finally replacing the scroll except for special purposes in the civil and religious worlds, and even then, in a different format called the "rotulus," which the medieval artists knew and thus depicted anachronistically for the earlier periods.

We do not know how the codex came to be introduced into the world of Christian literature, but it happened fairly quickly. I suspect that it may have been another adopted child from Greek Judaism, but at this point the evidence is entirely circumstantial. Other early uses of the codex in the Greco-Roman world at large include especially the fields of astrology and law, although scroll use also survives alongside. Christians also continued the use of scrolls, even while the codex grew in popularity especially for “scriptural” texts. Even in the 4th century, literary works such as Eusebius’ Church History were probably published on scrolls (that work was made up of 10 “books” of similar size) while at the same time, virtually all surviving scriptural texts were in codices. We lack sufficient evidence to determine how this process occurred in Greek (or Latin) Judaism, although there is some early evidence for Jewish scriptural codices, along with scriptural scrolls. The situation in Syriac and other eastern forms of Judaism and Christianity remains to be researched with more precision.

By the last part of the 3rd century of the common era, the contents of surviving codices had increased from the earlier practice of pretty much replicating what a scroll could hold to a more intermediate size, equal to several scrolls. In terms of technology, this set the stage for the development of the mega codex, with hundreds of pages and multiple columns per page. Even so, as already noted, the mega codex was the exception rather than the rule from that time onward until the invention of the printing press. Intermediate sized codices containing sections of scriptural writings predominated for centuries, sometimes as sections of a multi-volumed edition. Constantine's contribution to the process is noteworthy -- he commissioned multiple well written copies on parchment for use in the churches of his realm. This may have served like a government grant to develop a new industry that was otherwise unaffordable. Some scholars think that the mega-codices Vaticanus and Sinaiticus were results of the actions of Constantine and Eusebius, although there are serious problems with this view.

From scriptures to a canon: The development of the mega codex set the stage for firming up the contents and textual homogeneity of subsequent scripture copies. Earlier lists set the stage for determining contents, while physical collections of scriptural units (e.g. in cabinets) could attempt to control textual integrity. Still, even in the early 4th century, Eusebius was aware of “disputed” works on the edge of his listings, and even the mega codices did not agree completely on what to contain or in what order. Even with the growth in power and authority of the “great church,” whether focused on Latin Rome or on Greek Constantinople, official decrees concerning the contents and limits of the scriptural canon were rare – it seems to have been more a matter of church usage (depending, perhaps, on “official codices” used in the various churches) than of general legislation. For Jewish scriptures, the materials adopted from Greek synagogues seem to have won the day, although exactly under what conditions is not clear. As we have noted, Melito, in Asia Minor at the end of the 2nd century, seeks information from eastern (Jewish?) sources when questions about scripture arise. A few decades later, Origen’s sources may have been similar when he attempts to create a tool for assessing the extent of textual variation in available manuscripts. Exactly what formats of manuscripts were used by Origen (scrolls? intermediate sized codices?) and what format he used to construct his famous six columned Hexapla (presumably in dozens of separate volumes) or how he determined the order of the works represented (probably at least with lists) remain mysteries. But he, and the libraries with which he came in contact in Alexandria, Jerusalem, Caesarea, and perhaps elsewhere, doubtless could provide important evidence relative to the situations under discussion.

From texts to the (or at least, a) text: Origen is also a key player in the process of attempting to create and preserve textual integrity of the sort sought in Alexandrian scholarly circles for the Homeric corpus and other revered works. But Origen’s efforts proved to complicate matters more than to solve problems. His system of notation was too complex for those unfamiliar with such tools, and his full text of all the Jewish scriptural corpus too large and disjointed to copy as a convenient unit, so his work was sometimes corrupted in transmission, sometimes transmitted only in part, and, in short, his original goals of textual reform went unrealized. After him, Jerome attempts to apply Origen’s work to textual problems in which he and presumably also his Latin audience were interested, with some success, although Jerome is also quite aware of the existing textual diversity in his own day. And although Origen was perhaps the most ambitious repairer of Greek Jewish scriptural texts of whom we know, he was not the first. Justin the martyr in the mid 2nd century comments on what he perceives to be mistranslations and textual corruptions in the scriptural sources available to him, and Josephus a century earlier is aware of problems of textual transmission and of differences between texts available to him. With the development of the mega codex, it would have been possible to solve the problem of textual diversity as well as canonical extent and order if there had been a desire to do so from a sufficiently authoritative Christian voice, such as the Council of Trent for Latin orthodoxy a millennium or so later. The Greek and eastern churches never did standardize canon in such a formal manner. But these factors involving canon, text, and format did not come together at that earlier time, leaving us with many complex and hopefully interesting issues to contemplate.The path from ancient scrolls to mega codices to fixed canons is not well mapped and is replete with dead end side paths, detours, as well as stumbling blocks of various sorts. Let the traveler beware. /end/