I. Early Development

Rather than try to commit to specifics here, I will just give the possible ranges of dates that have been argued and which are at least possible. The material for this section is taken from my own survey of scholarly consensus found in numerous sources. It is believed that Jesus died c. 30 A.D. Specifically, if he died under Pontius Pilate, the date must have at least been between 26 and 36, the ten years we know Pilate to have served in Judaea.[1] Whatever the date, Paul's conversion follows one to three years later. The earliest known Christian writings are the epistles of Paul, composed between 48 and 58 A.D. Some of these are of doubted authenticity (and were even in antiquity), but the debate is too complex to summarize here. The other letters, and the Revelation (a.k.a. the Apocalypse of John), are of even more uncertain authorship and date. They are presumed to have been written in the same period or later (1 Peter, for instance, may have been written, some scholars say, as late as 110 A.D.).

The Gospels cannot really be dated, nor are the real authors known. Their names were assigned early, but not early enough for us to be confident they were accurately known. It is based on speculation that Mark was the first, written between 60 and 70 A.D., Matthew second, between 70 and 80 A.D., Luke (and Acts) third, between 80 and 90 A.D., and John last, between 90 and 100 A.D. Scholars advance various other dates for each work, and the total range of possible dates runs from the 50's to the early 100's, but all dates are conjectural. It is supposed that the Gospels did not exist before 58 simply because neither Paul nor any other epistle writer mentions or quotes them, and this is a reasonable argument as far as things go. On the other hand, Mark is presumed earlier, and the others later, because Mark is simpler, and at least Matthew and Luke appear to borrow material from him (material that is likely his own invention, cf. my review of The Homeric Epics and the Gospel of Mark).

All the Gospels except John contain possible allusions to the destruction of Jerusalem, which was destroyed by the Romans in 70 A.D., and thus it is likely they were all written after that date.[2] But that assumes the statements attributed to Jesus are apocryphal--they may have been genuine, the usual doom and gloom apocalyptic fantasizing, and then confirmed only by accident (or, if one is a believer, divine destiny) when the city and its temple were actually destroyed. They could also have been added to the text later. On the other hand, it has been argued with some merit that Luke borrowed material from Josephus, and if so that would date his Gospel (and Acts) after 94 A.D.[3] Finally, there are good arguments for the existence of a lost source-text called Q which was used by Matthew and Luke to supplement their borrowing from Mark, and this has been speculatively dated as early as the 50's A.D.[3a]

This is only an example of the state of ignorance we are in whenever scholars try to debate the dates of these writings. Although it remains possible that all the Gospels were written after 100, those rare scholars who try to place all Christian writings in the 2nd century have nothing to base such a position on. At least some of Paul's epistles can be reasonably taken as dating no more than 16 to 32 years after the oral tradition had begun to flourish after the death of Jesus, although adulteration of those letters by later editors remains possible, and it is also possible that even in Paul's day forgeries were being made and circulated (cf. 2 Thessalonians 2:2). The Gospels were not likely to have been written down so soon, and we have clear evidence, in numerous variations, that they were altered at various points in their transmission, and scholarly work in the last two centuries has gone far to get us to the earliest versions possible.

Nevertheless, any number of unknown alterations could still have been made that have not been detected (a great many have been--both errors and deliberate alterations or omissions), and it is important to note that the ancients did not have at one glance the scope of manuscript data we have, nor did they (with a few exceptions) even have the analytical and palaeographical skills now employed to derive a reliable manuscript archetype from a scientific collation of numerous exemplars. In other words, no one in antiquity ever saw a completely accurate collection of what would eventually become the 27 New Testament books, until perhaps the time of Origen or Clement of Alexandria (see XII and XIV), and even then most likely only those few scholars would have enjoyed the privilege. But this is still doubtful--it does not appear that either man went out of his way to find and trace the history of all existing manuscripts, in all churches, and in all translations, yet that is what would have been required to decisively collate a close approximation to the original texts (and with regard to facing an even worse problem today, cf. M 267ff.; and for an example, see Bible).

II. Ambiguous Pre-Canonical References

The first Christian text that did not become canonized but was respected as authentic is the first epistle of Clement of Rome, reasonably dated to 95 A.D. (M 40), and contained in many ancient Bibles and frequently read and regarded as scripture in many churches (M 187-8). This is relevant because even at this late date two things are observed: Clement never refers to any Gospel, but frequently refers to various epistles of Paul. Yet he calls them wise counsel, not scripture--he reserves this authority for the OT ("Old Testament"), which he cites over a hundred times (M 41-3). On a few occasions he quotes Jesus, without referring to any written source. But his quotations do not correspond to anything in any known written text, although they resemble sayings in the Gospels close enough to have derived from the same oral tradition. This suggests that the Gospels were not known to Clement. Yet he was a prominent leader of the Church in Rome. If they had been written by then, they must have not made it to Rome before 95. It is possible that they had not been written at all. In the case of Mark, for example, it is often thought that he was writing for an audience in Rome, thus it is most remarkable that Clement would not know of this, supposedly the earliest, Gospel. But it is also possible that he simply chose not to quote Mark, though knew the book--although why he would ignore Mark (even in his quotations of Jesus) and yet refer to numerous epistles of Paul is difficult to explain.

The next such text is the collection of letters by Ignatius. However, these were added to and redacted in later centuries, making the reliability of even the "authentic" letters uncertain. Ignatius wrote while on the road to his trial in 110 A.D. and it is important to note that he appears not to have had references with him, thus any allusions or quotations in his work come from memory alone (M 43-4). Thus, he borrows phrases and paraphrases from many Pauline epistles, yet never tells us this is what he is doing (he probably could not recall which letters he was drawing from at the time). Likewise, he borrows phrases or ideas which are found in Matthew and John, and on one occasion something that appears to be from Luke, but again he never names his sources or even tells us that he is drawing from a source at all (M 45-7). In no case does he name or precisely quote any NT ("New Testament") book, but again this may be due to the unusual circumstances in which he was writing.

Despite the difficulties, it seems plausible that the Gospels had been written by this date, although it is remotely possible that Ignatius is simply quoting oral traditions which eventually became recorded in writing, and also possible that this material was added or dressed up by later editors. Of greatest note is that in his letter to the Philadelphians, Ignatius recounts a debate he held with Judaizing Christians in which it is clear that only the OT was regarded as an authority (M 48-9). Instead of referring to any NT writings as evidence, he simply says that Jesus Christ is the witness to the authority of the tradition. This suggests that none of the NT was regarded even then as an authority. Like Clement, Ignatius and other Christians probably regarded these texts as wise counsel or useful collections of their oral traditions, and not as "scripture" per se.

Next comes the Didakhê (did-a-KAY), a manual of Christianity, which cannot be certainly dated, though it is believed to follow 110. Some scholars have weakly tried to place it much earlier, even to the time of Paul [4]--others have proposed a much later date for the existing text, as late as the 4th century (though it existed in some form without a doubt before the 3rd century). Its detailed account of a church hierarchy and rituals and the text's unusual organization into "The Way of Life" and "The Way of Death," among other details, likely suggest a 2nd century date (M 49). It does not name any written sources, but quotes exactly the Gospel of Matthew as just the "Gospel" of Jesus (M 50). No references are made which show any clear connection with the epistles, but the OT is quoted a few times. It is worth noting that the book attributes its ultimate source to unnamed itinerant evangelists (M 51), showing that anonymous oral tradition was still king when the Didakhe was written. It is also worth noting that this text was regarded as canonical scripture by Clement of Alexandria and Origen, and perhaps in the Egyptian churches for quite some time (M 187).

Unfortunately, we cannot date this text well enough for it to be helpful, and the same problem is faced by the Epistle of Barnabas (M 56-9), which cites many OT books by name and uses many phrases which appear in the Gospels, but never names any NT book--and the allusions are of the sort that could merely reflect common oral traditions. The date of this letter is unknown and could be anywhere from 70 to 130 A.D. (Barnabas was supposedly a companion of Paul), and it was for a long time actually a part of the NT canon itself, appearing at the end of the oldest surviving complete Bible, the Codex Sinaiticus (transcribed in the 4th century A.D., possibly based on a text produced by imperial commission, cf. XVI).

III. Clear Pre-Canonical References

The first author who shows a more concerted interest in textual sources is Papias. We do not know when he wrote, but presumably it was between 110 and 140, and most likely 130 or later (M 51-2). What he wrote has not survived, apart from fragmentary quotations in other works of his "Expositions of the Sayings of the Lord" which purported to be a collection of things he had actually heard said by the students of elders who claimed to have known the first disciples (yes, this sounds a lot like "a friend of a friend of a friend"), since he specifically regarded this as more useful than anything written, according to a quotation of his preface by Eusebius (History of the Church, 3.39.4), where Papias says "I did not think that information from books would help me so much as the utterances of a living and surviving voice" (M 52). Thus, Papias reveals the early Christian preference for oral rather than written tradition. It was only in the later 2nd century that this preference began to change. Other quotations of his work show how destructive this 'preference for oral tradition' was, since Papias apparently recorded the most outlandish claims as if they were true, such as the fact that Judas' head bloated to greater than the width of a wagon trail and his eyes were lost in the flesh, and that the place where he died maintained a stench so bad that no one, even to his own day, would go near it (from book 4 of the Expositions, quoted by Apollinaris of Laodicea, cf. footnote 23 in M p. 53).

Of note in the surviving quotations of this same work are his claims about the writing of Mark and Matthew. The latter, he claimed, was a collection by Matthew of the sayings of Jesus in Hebrew, which several others had translated "as best they could" (M 54). This is the origin of the belief that the Gospel of Matthew was originally written in Hebrew, but there are three points against such a belief. First, we have seen that Papias is hugely unreliable. Second, he is not describing a Gospel at all, but a collection of sayings. He is thus describing some other book now lost (some have suggested it was the Q document), or that had never existed in the first place. Third, it is distinctly possible, since the text is vague, that instead of the "Hebrew sayings of Jesus" this book contained the "Hebrew (i.e. OT) prophecies aboutJesus," which curiously fits the fact that the Gospel of Matthew is the one to include many of these prophetic claims and allusions. Moreover, the word for "translated" may mean "interpreted," in which case what Papias is describing is perhaps a proto-Matthew containing a bare collection of OT prophecies, from which were drawn a few by the later author of the Gospel of Matthew, who had done his own "interpreting" of how they applied to Jesus. But this is speculative. At any rate, Papias only hints at a possible name for a possible Gospel author. And this reference is most likely to a different, now lost, work. This remark of Papias thus could have become an inspiration for naming a certain Gospel after the same man. So this is not entirely helpful.

Papias's account of Mark is stranger still. He says Mark was Peter's secretary (perhaps getting the idea from 1 Peter 5.13), and though he had never known Jesus, he followed Peter around and recorded everything he said, leaving nothing out and changing no details (M 54-5). However, he did not "set in order" the sayings of Jesus. It is hard to tell what he means, but scholars see in his account a growing apologetic in defense of Mark: Mark was regarded as unreliable because he did not know Jesus, and he was attacked for being incomplete and disorderly, and so on, so Papias defends him by putting him in the entourage of Peter and asserting that he faithfully recorded what Peter said, and so on. What is evident is that this, the first historical thinking about Christian literary traditions, shows a possible corruption of reliability by oral transmission and a readiness to engage in apologetic distortions.[5] This does not create much confidence in later reports, and raises the real possibility that other claims to authority are rhetorical rather than genuine (such as that made in the closing paragraphs of the Gospel of John). But at least we now discover (perhaps), between 110 to 140 A.D., the first definite name of a Gospel author: Mark. There is one outstanding problem for these references to Mark and Matthew in Papias: they appear only in Eusebius, who is notorious for reporting (if not creating) forgeries.[6] We cannot establish whether this has happened in this case, but there must always remain a pall of suspicion. Even if accurate, there is another side of the story: the situation evident in Papias is that there is little regard for any written Gospels, in contrast with nearly complete faith in oral tradition, with little critical thought being applied.[7] More importantly, the context seems to be one where there were perhaps no set written Gospels in his day, but an array of variously-worked texts. And this picture is somewhat confirmed by the remarkable discovery of fragments dated c. 130-180 A.D. from a lost synoptic Gospel, the composition of which has been dated "not later than A.D. 110-30" (M 167). In this text, there are echoes from all four Gospels, but also miracles and sayings of Jesus found nowhere else, and it appears the author was working not from textual sources but from memory, and composing freely in his own style (M 168). It is likely that this, in part, is how all the Gospels were written. Moreover, it is possible that the canonical Gospels did not achieve their final (near-present) form until during or shortly after the time of Papias.

In the same period, Polycarp wrote a letter which cites "Jesus" for certain sayings a hundred times, and the sayings match closely those appearing in the Gospels (and even things written in numerous Epistles, which were not originally attributed to Jesus), but he does not name any sources (M 59-61). We see the authority of oral tradition is again elevated above the written--like all the previous authors, no NT text is called scripture, though many OT texts are, and the only cited source for NT information is the report of 'unnamed' evangelists (Epistle of Polycarp, 4.3). However, a sign of a change lies in the very purpose of the letter: it is a preface to a collection of letters by Ignatius which another church had requested be copied and sent on to them. The interest in written documents is thus rising among Christian congregations in this period (unfortunately, this could also be a source of interpolated Gospel quotations in Ignatius). And so it is in this milieu, between 138 and 147 A.D., that the first philosophical defense of Christianity addressed to an Emperor (Antoninus Pius) appears, written by Aristides of Athens, in which there was vaguely mentioned "what [the Christians] call the holy Gospel writing" (M 127-8) which is alleged to be powerful in its effect on readers.

As all this is going on, however, one of the first written texts to become universally popular and an object of praise among Christians is none other than the book of Hermas, a.k.a. "The Sheppherd," an unusual (to us) collection of "visions, mandates, and similitudes" (the names of the three books that comprise it). This was written at some time in the 2nd century, and we have papyrus fragments from that very century to prove it (M 63-4). It may date even from the 1st century (cf. op. cit. n. 1), but references inside and outside the text create likely dates ranging from 95 to 154 A.D. (both Origen and Jerome thought the author was the very Hermas known to Paul, i.e. Romans 16.14), but it is probably more likely later than earlier in that range.

So popular the Sheppherd was that it was widely regarded as inspired--it was actually included, along with the Epistle of Barnabas, as the final book in the oldest NT codex that survives intact, the Codex Sinaiticus (c. 300 A.D.). But even the book of Hermas never names or quotes exactly any NT text. It contains many statements which resemble those in various NT books, but this could just as well reflect a common oral tradition. It is noteworthy that the only book actually named by Hermas is an apocryphal Jewish text, the Book of Eldad and Modat. In contrast, it is notable that none of the Gospels or canonical Epistles ever name any book of any kind apart from Jude--which cites another apocryphal text, the Book of Enoch (vv. 14-15).

IX. Tatian

Curiously, the first "orthodox" Christian move toward canonization begins outside the Roman Empire, in the Syrian church. Moreover, this canon was ultimately not in Greek, but was a Syrian translation (M 114-7). The single man responsible is Tatian, who was converted to Christianity by Justin Martyr on a visit to Rome around 150 A.D. and, after much instruction, returned to Syria in 172 to reform the church there, banning the use of wine, the eating of meat, and marriage (M 115). At some point in all this (it is suggested c. 160 A.D.) he selected four Gospels (the four we now know as the canon, and which no doubt supported his own ideology and that of his tutor, Justin) and composed a single harmonized "Gospel" by weaving them together, mainly following the chronology of John. This is called the Diatessaron("That Which is Through the Four") and it became for a long time the official Gospel text of the Syraic church, centered in Edessa. The Syriac Doctrine of Addai (c. 400 A.D.) claims to record the oldest traditions of the Syrian church, and among these is the establishment of a canon: members of the church are to read only the Gospel (meaning the Diatessaron of Tatian), the Epistles of Paul (which are said to have been sent by Peter, from Rome), and the Book of Acts (which is said to have been sent by John the son of Zebedee, from Ephesus), and nothing else (M 113-4). This tradition is traced back to Tatian.

Unfortunately we lack any complete versions of this, the first Christian canon outside of the Gnostic tradition (see XVIII). We do not know which Epistles he accepted as authentic, yet we know he rejected some (cf. Jerome, "On Titus," pr.), including 1 Timothy because it allowed the taking of wine, meat, and marriage. Other references allow us to guess at some of those he thought authentic. But of the original Diatessaron we only have one fragment and a few quotations, although the fragment is very close to the original--within eighty years (M 115). The fragment matches the narrative just after the crucifixion and just before the body of Jesus is taken down, with verses mainly from the three synoptic Gospels, and one from John. However, in other quotations of the Diatessaron (and in late copies in Syriac and Armenian, which are not securely reliable) there are phrases which seem to come from other sources, such as the Gospel of Hebrews and the Protoevangelium of James, suggesting that the four Gospels at that time may have contained verses now missing or altered. The only complete work of Tatian's that survives is his "Oration to the Greeks" which is a scathing attack on Greek culture. We know he wrote books prolifically on a number of other topics. He was probably the first Christian to do so, apart from Justin.

What is significant is that it is shortly after Tatian and Justin's contributions that we discover the first instance of organized action against authors of new Christian source-texts. Although such action is necessary for there to be any hope of control over a reliable textual tradition in a milieu of wanton invention and combative propaganda, the fact that it only begins at such a late date is another blow against those who set their hopes on having complete confidence in the present canon. It means that a century of prolific writing went largely unchecked before the church took any concerted action to stop it. This first case is reported by Tertullian (On Baptism, 17): the story goes that a well-meaning priest in Asia Minor wrote the Acts of Paul to honor the Apostle, sometime around 170 A.D. He was brought before a church council, convicted of falsification, and removed from office (M 175). Nevertheless, though universally condemned by church leaders, it remained lastingly popular among educated church laymen, and one section of this text remains a part of the Armenian Bible to this day. But disturbingly, Tertullian attacks the book primarily because it depicts a woman (Thecla, a disciple of Paul) preaching and administering baptism. Thus, as we will see more than once, doctrine, not objective concern for history, loomed large behind the charge of falsification--so we are faced with uncertainties all over again. Indeed, Tertullian, as a hostile witness set on abolishing the text, might not be telling us the truth about its author or date of origin. This is all too likely, since there are indications that Tertullian was not an honest man (see XV).

It must also be noted that our evidence for church reactions to texts is incredibly scarce. For there were books that were extant in the 2nd century yet never mentioned and thus entirely unknown until recovered in more recent times. How many other Christian writings are we completely ignorant of? For instance, traces of a forged Epistle survive in the Coptic (Egyptian) and Ethiopian manuscript traditions: the almost ridiculous Epistle of the Apostles, a semi-apocalyptic text written by the "eleven disciples" after the resurrection "to the churches of the East and the West, the North and the South" (M 180-2), even though there could not have been any such churches at that time. This text has been plausibly dated to c. 180 A.D. (it does fit the mystical orthodoxy of Irenaeus), and even earlier than 120 A.D. by some scholars. It is too derivative and fantastic in my opinion to come so early, but redaction evidence points strongly to a middle date: the End Times is placed at 120 years after the Resurrection in one redaction, and this was altered to 150 years in another--a possible sign that the text was written shortly before 150 and then amended when the End did not come. Yet no extant Christian writer even took notice of this book--not even to denounce it. As another example, we have already discussed above the "lost synoptic Gospel" recovered in a 2nd century papyrus fragment.

X. Theophilus and Serapion

Near Tatian's Syrian church, but across the border in Roman territory (and amidst a decidedly Greek culture) flourished bishop Theophilus at Antioch, around 180 A.D. (M 117-9). Theophilus is important for a variety of reasons: he was the second, very shortly after Athenagoras (below), to explicitly mention the Trinity (Ad Autolycum 2.15); he may have composed his own harmony and commentary on the four Gospels chosen by Tatian; and he wrote books against Marcion and other heretics. He is also a window into the thinking of converts: he was converted by the predictions concerning Jesus in the OT (ibid. 1.14), perhaps the weakest grounds for conversion. But most of all, he routinely treats Tatian's Gospels as holy scripture, divinely inspired, on par with the Hebrew prophets (M 118). He also refers to John's Revelation as authoritative.

Theophilus' successor, Serapion, reveals the next stage in the process in 200 A.D. While touring churches in Asia he came upon a dispute in a village in Cilicia about whether the Gospel of Peter could be read in church. He tentatively agreed, but after reading it he closely instructed them not to use it anymore because it supported the Docetic heresy--the belief that Jesus only "seemed" to be a man, and was not really flesh (M 119, cf. 77)--so he concluded on this ground alone that it was falsely ascribed to Peter. Thus, doctrine more than objective evidence of historicity was driving the selection of canonical texts. This despite the fact that this Gospel may have been written as early as 100-130 A.D. (M 172, n. 18), again if not earlier, although a later date is still possible, especially if the four canonical Gospels are likewise given later dates than usual, since Peter may have drawn on them.[8]

Unfortunately, we do not know if this Gospel of Peter was the same as the surviving Gospel of that name, but if it is, Jürgen Denker has shown that, for instance, "almost every sentence in the passion narrative of this gospel was composed on the basis of Scriptural references in the Old Testament, particularly in Isaiah and the Psalms" (M 172), which makes this text much more like what we would expect a Jewish messianic sect to compose, even more than what we find in the canonical Gospels, and this allows that Peter may in fact have preceded the canonical Gospels: we know that the Gnostics began canonization and reliance on writings first, and it is possible that "less Gnostic" texts (the present Gospels) were then created in response, even if by drawing on other, no longer surviving works. There really is no way to resolve this question. At best (or at worst, depending on your point of view), it remains only a faint possibility. If the Gospels are believed to date according to simplicity and lack of embellishment, then Peter must come later than the canonical four--but this method of dating texts is not always sure to be correct. Sometimes simpler redactions follow, rather than precede, the originals.

XI. Dionysius, Athenagoras and Irenaeus

In this same period we know these books were being doctored and battles were being fought over authenticity along ideological lines. In the letter from Dionysius cited earlier, where he informs his readers that even his own letters are being cut up and added to, he quotes the curse for such people in Revelation, which reveals that this sort of license was being exercised widely enough even when the Revelation was written that a curse had to be reserved for it (M 123-4, cf. Rev. 22.18ff.). He notes that this was being done even with the "scriptures of the Lord," thus recognizing certain NT books as scripture, but also that their integrity was being compromised at the very same time (and probably for the same reason). Dionysius is also notable for having tried to resolve doubts about the authenticity of the Apocalypse of John by ascribing it to a John other than the Apostle.

In 177 A.D. Athenagoras of Athens composed a lengthy philosophical Defense of the Christians addressed to the emperor Marcus Aurelius in which the first articulation of a theory of the Trinity appears. He quotes the OT and NT several times, but does not name his sources from the NT. The quotes or paraphrases that he uses happen to come from a few Epistles of Paul, and from all the Gospels in a mishmash (M 125), suggesting a harmonic source like the Diatessaron. But the respect that this defense, and others like it, earned among orthodox Christians contributed to forming decisions on canonicity based on whether they accorded with works like it.

Shortly after 177 A.D. Irenaeus was asked to compose an account of the persecutions in Lyons for the churches in Asia, and this letter is preserved by Eusebius (History of the Church 5.1ff.). This text quotes or paraphrases various NT books without naming them. Some years after this he composed a mighty treatise Against All Heresies and a Demonstration of the Apostolic Teaching. In these he quotes exactly almost every book of the NT, numerous times, demonstrating that the orthodox canon, though not established officially, was by this time generally accepted in practice (M 154). And in an account of the martyrs of Scillium (in Numidia, i.e Tunisia) who were tried in Carthage in 180 A.D., we find an overt mention of Christians carrying around "the books," including Paul's letters. Metzger reasons this as proof that Latin translations of the letters and Gospels existed by then, though this is a shaky argument at best (cf. M 156-7).

However, Irenaeus, whose voice is as close to official as any of the time, has this to say about the selection of the four Tatian Gospels (which he calls "the four-form Gospel" or "one Gospel in four forms"):

It is not possible that the Gospels can be either more or fewer in number than they are, since there are four directions of the world in which we are, and four principal winds...the four living creatures [of Revelation 4.9] symbolize the four Gospels...and there were four principal covenants made with humanity, through Noah, Abraham, Moses, and Christ. (Against All Heresies 3.11.8; cf. M 263)

Clearly, less than scholarly reasoning was affecting the canonization process (and Irenaeus' reasoning may reveal an earlier, more profound basis for the four Gospels: cf. Darek Barefoot, "The Riddle of the Four Faces: Solving an Ancient Mystery"). Moreover, Irenaeus includes the book of Hermas as holy scripture, a part of the NT (Against All Heresies 4.20.2).

XII. Pantaenus and Clement: the Seminary at Alexandria

The first Christian seminary was established in Alexandria around 180 A.D. by Pantaenus, whom Eusebius claims had found a Hebrew version of the Gospel of Matthew in India, or perhaps Ethiopia or Arabia (M 129). Although the story is not entirely vouched for even by Eusebius (he presents it only as something that "is said"), it no doubt contributed to the belief that Matthew had originally been composed in Hebrew. Pantaenus is also the first to defend the Epistle to the Hebrews as authentic (this had long been in dispute even by his time), on the argument that Paul wanted to compose it anonymously for that particular audience (M 130), and this opinion is generally carried as authoritative (M 134). Naturally, the formation of a 'school of Christian teaching' is a decisive moment in compelling the selection of a canon, if nothing else as a textbook, and fittingly for this task the next head of the school (in 190), Clement of Alexandria, represents the most scholarly Christian to date: in his surviving works, he cites other written sources about 8000 times, over 2500 of those citations being of works outside the Christian or Jewish tradition (M 131), the mark of a true scholar.

By about 200 A.D. we find from Clement's researches that he regarded the Tatian selection as being the primary source of the Gospel tradition, and that he believed that the chronological order of the books was Matthew and Luke, then Mark, and finally John, but he also acknowledged as authentic the Gospel of the Egyptians (not the same as the one recovered from Nag Hammadi, but still likely a Gnostic text; M 171), the Gospel of the Hebrews, and the Traditions of Matthias (M 132), as well as Hermas, the Epistle of Barnabas, the Apocalypse of Peter (M 134), and the Didakhe (M 187). But he also recorded certain additional oral traditions that he thought were authentic as well, including three sayings of Jesus that did not make it into any known written text (M 134, cf. Stromata 1.28: 94.5, 158.2, and 177.2). The rest of his loose "canon" included the fourteen Pauline Epistles, Acts, 1 Peter, 1 John, and Revelation.

The Apocalypse of Peter is especially worth mentioning here. It was probably written between 125 and 150 A.D. and remained in various church lists as a canonical text for centuries (M 184-6). It gives detailed accounts--in the words of Jesus as he instructs Peter after the Resurrection--of signs of the End Times and then of the various kinds of punishments awaiting sinners in Hell, and the pleasures of Heaven. This text was not only popular and often treated as the genuine work of Peter (even by the very scholarly Clement), but its influence on the Christian religion as a whole is profound: this is the first text to introduce detailed pagan ideas of heaven and hell into Christian belief (drawing on Homer, Virgil, and Plato, as well as Orphic and Pythagorean traditions), and the popular view of these destinations, adopted and embellished by Dante centuries later, as in all Medieval art, is a direct outcome of this early Christian book and its widespread influence in the church for many centuries. What is of interest to us is that this book should have been regarded as obviously false, yet instead was generally received as genuine--its eventual omission from the canon was more the result of the fact that some church leaders did not want it read aloud (according to the Muratorian Canon, cf. M 184), perhaps because its descriptions were so disturbing.

In addition, thanks to Clement, we know of a Gospel written some time in the first half of the 2nd century (if not earlier) that did not make it into the final canon despite having been held as canonical by Clement, and many others (including Jerome): the Gospel of the Hebrews (M 169-170). This was used as an authority in Syria even as late as the 4th century, and might have originally been composed in Aramaic. Though it only survives in a few quotations, we know it was only slightly shorter than the present Gospel of Matthew. It has some interesting features, such as Jesus calling the Holy Spirit his "mother." And it clearly presents James, not Peter, as the first to see Christ risen. In fact, James is depicted as having expected and anticipated the resurrection, even fasting until it should occur. If true, this is an excellent starting point for possible hallucinations of a risen Christ: deprivation and expectation. The loss of this text, and thus our inability to assess its merit, is another fact that greatly obscures any attempt to get at the historical truth behind the origins of Christianity.

Finally, Morton Smith discovered a very late copy of a certain letter by Clement which has something unusual to say about the Gospel of Mark. Though Smith's scholarship has been questioned, the find has been accepted as authentic (M 132-3). The letter asserts that there are three versions of Mark: a shorter one written in Rome based on Peter's teaching, a longer "more spiritual" (more Johanine?) one written in Alexandria by Mark after Peter's death, and a "secret" version left by Mark before he died, carrying on a tradition of Christian mysteries initiated by Peter. The last was supposedly smuggled out of the library at Alexandria and "corrupted" by the heretic Carpocrates between 100 and 125 A.D. Whether the letter is authentic or not, this betrays a problem for current scholars: secret traditions. To what extent were these lost, or incorporated in orthodox or heretical writings? Since secret traditions are the easiest to lose or corrupt, there may be a lot to the Christian creed in the 1st century that is lost to us today and that would, if found, radically change what we think about Jesus or the first evangelists and their beliefs. There may also be doctrines or sayings among the Gnostics which are authentic, but indistinguishable from others that are not.

XVIII. The Eastern Canons

For centuries the Diatessaron of Tatian, along with Acts and the Pauline Epistles (except Philemon), comprised the only accepted books in the Syrian churches, meaning that Tatian's stricter views, resulting in the rejection in 1 Timothy, did not win out. Moreover, after the pronouncements of the 4th century on the proper content of the Bible, Tatian was declared a heretic and in the early 4th century Bishop Theodoretus of Cyrrhus and Bishop Rabbula of Edessa (both in Syria) rooted out all copies they could find of the Diatessaron and replaced them with the four canonical Gospels (M 215). Thanks to them, no early copies of the Diatessaronsurvive--although a very early fragment suggests it would have been crucial evidence for the true state of the early Gospels (see IX).

By the fifth century the Syrian Bible, called the Pe****ta, became formalized somehow into its present form: Philemon was accepted, along with James, 1 Peter and 1 John, but the remaining books are still expelled (2 and 3 John, 2 Peter, Revelation, and Jude). After the Council of Ephesus in 431 A.D., the Eastern Syrian church, in turn divided between the Nestorian and the Syrian Orthodox Churches, broke away, and retained this canon of only 22 books (the Pe****ta) until the present day. However, to confuse matters, a monument erected by a Nestorian in China in 781 A.D. states that there were 27 holy books (the number in the standard Western Bible of today), although they are not named and there is debate over what books are meant. Meanwhile, the Western Monophysite Syrian church, at the urging of Bishop Philoxenus in 508 A.D., abandoned the Pe****ta altogether and adopted a new Syriac translation of the Catholic Bible, yet the Harcleans still insisted on including 1 and 2 Clement in their Bible, the last surviving copy of which dates to 1170 A.D. (M 218-22).

Then there is the Armenian Church, significant not only in being a breed apart, but also in being the first "national church" in Christian history--the royal family, and thus at their behest the rest of the nation, converted to Christianity a few years before Constantine. The Armenian Bible is essentially the same as ours, with one addition: a third letter to the Corinthians, actually taken directly from the Acts of Paul (M 176, 182, 219, 223; cf. IX), became canonized in the Armenian Church and remains a part of the Armenian Bible to this day. Revelation, however, was not accepted into the Armenian Bible until c. 1200 A.D. when Archbishop Nerses arranged an Armenian Synod at Constantinople to introduce the text. Still, there were unsuccessful attempts even as late as 1290 A.D. to include in the Armenian canon several apocryphal books: Advice of the Mother of God to the Apostles, the Books of Criapos, and the ever-popular Epistle of Barnabas (M 224).

Then there are the African canons. The Coptic Bible (adopted by the Egyptian Church) includes the two Epistles of Clement, and the Ethiopic Bible includes books nowhere else found: the Sinodos (a collection of prayers and instructions supposedly written by Clement of Rome), the Octateuch (a book supposedly written by Peter to Clement of Rome), the Book of the Covenant (in two parts, the first details rules of church order, the second relates instructions from Jesus to the disciples given between the resurrection and the ascension), and the Didascalia (with more rules of church order, similar to the Apostolic Constitutions).

XIX. The Western Canons

Bishop Hilary gained respect and authority among the Western orthodoxy for his clever and impassioned attack on Arianism at the Council of Seleucia in 359 A.D. Since he had an affinity for some of the books accepted in the East but rejected in the West, this had the effect of turning the tide of opinion in the Western Church (M 232-3). The great scholar Jerome was influenced by this, and by his Eastern education, and when he decided to replace the numerous conflicting Latin translations of the New Testament texts his choices would be decisive for the rest of Western Christendom, for his translation would become a monumental masterpiece in its own right, winning respect for its literary competence and unity. Once it won the endorsement of the pervasively-influential Augustine of Hippo in the early 5th century, its authority would never be questioned. This was the Latin Vulgate Bible, of which the Gospels were completed and delivered to the Pope in 384 A.D. It is debated whether Jerome himself finished the rest, or if it was completed by others, but the finished project contained all the 27 books now found in the Bible, no more and no less, and the choice was almost certainly Jerome's, since his many other works all speak of these books as authoritative, and a letter he wrote listed all these books as his ideal canon--a letter which was placed as a preface to many of his Latin Bibles (M 234-5).

We have some insights into his thinking. On several occasions he makes statements that entail the belief that those books were to be accepted which had gained authority merely by having been long held in respect by the churches (M 235). Not an objective criterion, this is a vote by fatigue, a tacit acceptance of argumentum ad nausium. Yet this manner of thinking has resulted in a certain contradiction in thinking about Biblical canonicity that remains to this day: Jude was accepted as canonical simply because it was long held in respect. But Jude quotes the book of Enoch as an authority (vv. 14-15)--yet the book of Enoch was rejected because it was not so widely respected. Curiously, Jude is the only book in the NT that actually cites any other book outside of the OT, and such a citation by its force and uniqueness should have won Enoch a place in the NT. For if Enoch is false, so must be Jude, at least in part--it makes no sense to call Jude an authority, and yet reject his sources. Similarly, Jerome fully believed that the Epistle of Barnabas was authentically written by the companion of Paul (M 236), a fact that surely should have won it a place in the NT (Luke and Acts, even in tradition, have no better authority than this), yet because it was not universally popular it was not accepted into his idea of the Bible. In direct contrast, though he declared that no one really knew who wrote Hebrews, he still accepted it as an authority. This is a method that contradicts all objective sense. Yet thus came the Bible.

Augustine all but codified this method, declaring without qualification that one is to "prefer those that are received by all Catholic Churches to those which some of them do not receive" (On Christian Doctrines 2.12). Of course, this whitewashes the fact that by "Catholic Churches" he means those whose opinion he accepts (a problem we have seen throughout the history of NT canonization), since many Eastern Churches rejected some of the very books Augustine upheld as universally received. In the same passage, Augustine allows these dissenting churches to be outweighed by the opinions of "the more numerous and weightier churches." Thus, this is a purely circular argument: those books are to be accepted by the Church that are accepted by the Church. This is not an objective methodology by any stretch, and is entirely driven by blind tradition and the demands of authoritarian dogma.

Augustine effectively forced his opinion on the Church by commanding three synods on canonicity: the Synod of Hippo in 393, the Synod of Carthage in 397, and another in Carthage in 419 A.D. (M 237-8). Each of these reiterated the same Church law: "nothing shall be read in church under the name of the divine scriptures" except the OT and the 27 canonical books of the NT. Incidentally, these decrees also declared by fiat that Hebrews was written by Paul, ending all debate on the subject. That may have been convenient for the Church, but it was hardly honest. Nevertheless, Hebrews continued to be excluded from many Bibles in the West, while the bogus Epistle to the Laodiceans (see XV) continued to be found in hundreds of Bibles in various languages until relatively recent times.

Strangely, this is essentially where the story ends. It is most curious that there was never any pronouncement by any central authority such as the Pope in all of Christian history as to which books belonged in the Bible, until 1443 A.D. at the conclusion of the Council of Florence--yet this only carried weight in the West. This pronouncement excluded Laodiceans and included Hebrews, thus effectively ratifying the 27 books that had been the staple of orthodox opinion since the 4th century A.D. (M 240). This no doubt arose because for the first time in almost a thousand years scholars were once again starting to question the authenticity of certain books in the canon, for example the authorship of Hebrews. A telling case is that of Erasmus, who, after being chastised by the Church, renounced his rational doubts about this and various Biblical books, on the ground that "the opinion formulated by the Church has more value in my eyes than human reasons, whatever they may be" (Response to the Censure of the Theology Faculty at Paris, 9.864; M 241). No freethinker he. No one can trust the opinions of such a man. Nevertheless, the canon of Florence was still not enforced by threat of excommunication until the canon was made an absolute article of faith at the Council of Trent in 1546 A.D. Almost all the Protestant churches followed suit within the next century with essentially identical conclusions (M 246-7), dissenting only by excluding the OT apocrypha held as canonical by the Catholics.

But it is worth adding an interesting irony: for with the Reformation the history of canonization came almost full circle. Luther wrote prefaces on the books of his Bible, and ordered the books consciously in descending degrees of credit (M 242-3), and his entire scheme reveals a pervasive criterion: everything that agrees with Paul and preaches Christ is a priori true and to be held in highest esteem, while everything else is to be doubted. And he repeats the argument from fatigue: though he explains why certain books like Hebrews and Jude are to be doubted--namely, they contradict the teachings of Paul--he goes on to declare that he does not want to remove them from so venerable a collection. Thus, not only dogmatic presupposition, but mere tradition wins the canon--not objective scholarship. The irony is that Luther is almost a twin of the heretic Marcion, who was, if you recall, the first man in Christian history to propose a canon. For Marcion believed that only Paul's doctrine was true--although he was in a better position to be more consistent about this by rejecting all books that contradicted Paul. And it is well known that Luther was rabidly anti-Jewish--as was Marcion. Though the two men differed on many key points, in a small sense the Reformation effectively re-launched the old Marcionite heresy, at the very end of the process of canonization that Marcion had begun.[9]

For more on the NT see: 

Everett Harrison, Introduction to the New Testament (1971)

Bruce Metzger, The Early Versions of the New Testament (1977)

William Wake and Nathaniel Lardner, The Apocryphal New Testament (1997)

James Robinson, The Nag Hammadi Library in English (1990)

W. Bauer, Orthodoxy and Heresy in Earliest Christianity  (2nd ed. 1996)

[for more recent books on the NT see the Secular Web Bookstore]

For more on the OT see: 

Gerald Larue, Old Testament Life and Literature, 1968 (online version 1997)

Earle Ellis, The Old Testament in Early Christianity: Canon and Interpretation in the Light of Modern Research (1991)

Roger Beckwith, The Old Testament Canon of the New Testament Church and its Background in Early Judaism (1986)

Annemarie Ohler, Studying the Old Testament from Tradition to Canon (1985)

Alberto Soggin, Introduction to the Old Testament, from its Origins to the Closing of the Alexandrian Canon (1999).