THE BIRTH OF THE NEW TESTAMENT by C. F. D. Moule, Lady Margaret's Professor of Divinity in the University of Cambridge. First published A & C Black Ltd 1962. This edition prepared for katapi by Paul Ingram 2013.



THE story of the formation of what is known as the New Testament 'canon' is a story of the demand for authority. The Christian Church set out with a preposterously unlikely tale: that a person who had recently been executed by the Romans at the instigation of the Jewish religious authorities had been restored to life and was the very corner-stone of the entire building called 'Israel'. Where was their evidence for this, and what was their authority? Probably the most immediately cogent evidence was the very existence of this group of' Nazarenes', so confident and convinced in their witness to the events, bound together in such transparently sincere brotherhood, attended by such unmistakable signs of God's power and presence. The early chapters of the Acts strikingly suggest (in a narrative of which there is no reason to doubt the substantial veracity) the impact of such a community suddenly making itself felt in Jerusalem. The whole population was shaken: a large number were convinced. It was difficult to believe that such results sprang from hallucination or a huge mistake.

But, more particularly defined, the authority for the Christian statements about Jesus was the eye-witness of twelve men. According to the Gospels, Jesus himself had in his lifetime chosen and commissioned them to be with him and to be sent out as heralds of the Kingdom of God (Mk iii. 13 ff. and pars.). One of them, Judas Iscariot, had turned traitor; but the remaining eleven are claimed, in the Acts, to have received a further express commission from the risen Master to be his witnesses, and a new twelfth, Matthias, was afterwards selected by lot as between two chosen from among the other eye-witness disciples. It was agreed (Acts i. 21 f., and cf. John xv. 26) that a condition for election should be that a new member must have been with the Twelve all the time from John the Baptist's mission until the ascension – that is to say, he must have been an intimate participant in the whole ministry of Jesus and a witness of the resurrection and its sequel. His evidence must cover the scope of the kerygma – the Christian proclamation. The use of lots (Acts i. 23 ff.), representing divine guidance, was evidently intended to be the equivalent of the express commissioning of Jesus. Thus was repaired the body of the Twelve, all intimate participants in the ministry of Jesus and witnesses of the resurrection, and all authoritatively commissioned to give such witness, 1 John i. 1 again alludes to the eye-witness evidence (cf. 2 Pet. i. 16). It is natural to see in the number twelve a deliberate allusion to the tribes of Israel: it was as much as to say 'Here in nuce is the real Israel's witness to the world'. It is possible also to see in the Acts narrative of the solemn restoration of the eleven to twelve a hint of the intention to underline the true, inner Israel's appeal to the wider, less loyal Israel in the Church's mission: here is a concentrated, twelvefold appeal to the twelve tribes themselves. [See further K. H. Rengstorf, 'The Election of Matthias (Acts i. 15 ff.)', in Festschrift for O. Piper, edd. W. Klassen and G. F. Snyder (1961). ] In this respect it is comparable to the mission of Yahweh's Servant to the rest of Israel in Isaiah.

However that may be, the Twelve evidently constituted the earliest Christian 'canon' or measuring-rod – the standard by which the authenticity of the Church's message was to be gauged, for the duration of their lifetime. They are the pillars of the whole structure (cf., though with wider reference, Gal. ii. 9, 1 Tim. iii. 15), and to such one must refer one's preaching (Gal. ii. 2.) But there is no sign that the Twelve were intended to be perpetuated by succession. Here was no caliphate: if there was a caliphate anywhere in the Christian Church, it was in the line of James the Lord's brother, not of the Twelve. [See A. A. T. Ehrhardt, The Apostolic Succession (1953), and E. Stauffer, 'Zum Kalifat des Jakobus', Zeitschrift für Religions- und Geistesgeschichte 4 (1952), 210 ff.] They were regarded as essentially a dominically chosen and commissioned foundation body, expressly authorized to give eye-witness evidence of the decisive events. As such they were, by definition, irreplaceable in any subsequent generation. Whereas the removal of Judas by apostasy was met by the special lot-casting for Matthias, none of the subsequent depletions, by martyrdom or natural death, were made up. The Twelve were no self-perpetuating body: they were simply the initial authority for the Christian claims about Jesus. [W. L. Knox's speculations (St. Paul and the Church of Jerusalem, 1925, 169) about James the Lord's brother in some sense taking the place of his namesake the martyred apostle are thus misleading. 'The mere death of an apostle need not have created a vacancy ... but apostasy is a different matter.' – C. H. Dodd, According to the Scriptures (1952), 58, n. i (on p. 59). The whole note is very instructive.]

Alongside this authority, and indeed as an integral part of it, there ran also the authority of the Jewish scriptures. The ' argument from Scripture', that is, the demonstration that what the apostles bore witness to was no isolated phenomenon, but could be shown to be the culmination and fulfilment of God's design for his People already sketched in scripture, has already been discussed in some detail (Chapter IV). It is enough here to recall that what the apostles had seen was claimed to be ' according to the Scriptures', and that the very form in which they framed their evidence was itself sometimes influenced by scripture. The witness of the Twelve was at once a confirmation of scripture and was confirmed by scripture. The way was already being paved for the recognition of the apostolic witness as itself material for inspired writing.

St Paul's efforts to establish his own claim to apostolic status seem to imply the same criteria as those defined in Acts i. 23 ff., in however abnormal a form. In 1 Cor. xv. 8 he reckons the appearance of the Lord to him on the Damascus road among all the other post-resurrection appearances, though at a different time; in 1 Cor. ix. 1 again he appeals to the fact that, in this sense, he is an eye-witness of the resurrection; and in Gal. i. 1, 11 f . he makes it clear that his commission as an evangelist came directly from the Lord. The peculiarity of his commission is that it is expressly to the Gentiles (Gal. i. 16, ii. 2; cf. Acts xxii. 21, xxvi. 17), and that he is outside the body of the Twelve – a thirteenth [See A. Richardson, An Introduction to the Theology of the New Testament (1958), 314, n. i.] – and never knew Jesus intimately during his earthly life. But his claim remains that he is an eye-witness of the risen Lord and personally commissioned by him to bear that witness.

Thus, to challenge the Christian message was to doubt a body of living eye-witnesses authorized by the Lord himself; and, so long as they or even their close followers still lived, one can readily understand the preference expressed by Papias (about AD 150) for the living voice rather than for writings. ' For I did not suppose', said he, 'that information from books would help me so much as the word of a living and surviving voice ' (ζώσης φωνῦς καὶ μενούσης) , Euseb. H.E. in. 39. 4 (Loeb translation). Admittedly, Papias himself nevertheless valued at least three of the written Gospels (for Matthew and Mark, Euseb. H.E. iii. 39. 15 f.; for John, an argument prefixed to a ninth-century Latin MS. of the Gospels in the Vatican library [See F. V. Filson, Which Books Belong in the Bible? (c. 1956), 152, for further reservations on Papias' evidence for the valuation of oral tradition. For the MS. in question, see J. B. Lightfoot in Contemporary Review, October 1875 (=Essays on Supernatural Religion (1889), 210 ff.); also Harnack, Geschichte der altchristlichen Litteratur (1893), 68. For Papias' knowledge of the Gospels, see R. M. Grant, 'Papias and the Gospels', Ang, Theol. Rev. xxv (1943), 218 ff.]); but it remains true that the Christian community was in essence not 'bookish': it had been called into existence by a series of events well remembered; it lived under the continued personal guidance, as it believed, of the central figure of those events; and the time would not be long, so it imagined, before he would return to sight. Its authority was 'the Lord and the Apostles'. [Whatever may be intended by the controversial phrase τὰ βιβλία καὶ οἱ ἀπόστολοι (if indeed this is the right reading) in a Clement xiv. 2 (see the discussion in H. Köster, Synoptischer Überlieferung bei den apostolischen Vätern (1957), 67 ff.), this homily certainly uses λέγει (εῖπεν) ὁ κύριος elsewhere (iv. 5, v. 2, etc.): so that ὁ κύριος and οἱ ἀποστολοι may justly be said to be its two sources of authority. For the date of 2 Clement (first half of second century?), see B. Altaner, Patrologie (5 1958), 82 f. (Eng. trans. (1960).] The only book it needed was the collection of scriptures already recognized by the Jews, in which the Christians now found explanation and confirmation of their own convictions, while, conversely, they found the scriptures explained and "confirmed in an entirely new way by the recent events.

But the last of the Twelve died before 'the consummation of all things', and it began to become evident that the Church must continue for an indefinite time in an imperfect world. Where were the guarantees to be found for the authenticity of its claims, after the accredited eye-witnesses had ceased to be available, and when even their immediate followers were growing scarce? The answer lay inevitably in written records. With the appeal to 'the Lord and the Apostles' begins an inevitable process of development leading to accredited writings. [Cf. W. Schneemelcher in E. Hennecke, Neutestamentliche Apokryphen 3, i (1959), 9.]

What did books look like at this period, and how general was literacy? The answer to the first question seems to be that important, and especially sacred, books had usually been in roll form, until the Christians adopted the codex form (essentially the form of a modern book) for their own sacred writings. Codex (caudex ) is Latin for a tree-stump, and then a slab of wood; and hence a slab or tablet for writing – the wax-coated board for note-taking; and then, eventually, the term was transferred to the flat, rectangular piles of folded paper, or, in some cases, parchment, constituting what we now call a book. By association and derivation, therefore, the codex tends to be used for the more incidental, transitory jottings, as contrasted with the biblion (or Latin volumen ) , the roll, containing the treasured, sacred writings. And although the word biblion is merely derived from the word for the papyrus reed, from the pith of which paper was made, in fact it came to denote the form (roll) rather than the material. And although rolls were sometimes written not on papyrus but on animal skin – parchment or vellum – while many a codex was made of paper (that is, papyrus) leaves, yet, while biblion tended to mean a roll, membranae (skins) tended to mean a codex. Hence it is plausibly suggested [See C. H. Roberts, 'The Codex', Proc. Brit. Acad., xl (1954), 169 ff.] that when, in 2 Tim. iv. 13, Timothy is instructed to bring the biblia, but particularly the membranae, what is meant by the biblia is the Jewish scriptures (in roll form), while the membranae meant the apostle's own notes (in codex form, whether in loose sheets or stitched together) – perhaps his sermons and disquisitions, perhaps copies of letters, perhaps simply notes. If this is correct, the biblia and the membranae stand (in this particular context) for Jewish and Christian writings respectively (though the latter would not at this stage have attained to 'canonical' standing or been regarded as among ἱερὰ γράμματα, sacred writings, 2 Tim. iii. 15); and it appears that it was in the Christian communities that the codex came into its own, so that eventually it superseded the roll as the recognized form for permanent – even sacred – writings. [See P. Katz, 'The Early Christian Use of Codices instead of Rolls', J.T.S. xlvi (1945), 63 ff.] That this took place earlier than used to be thought is important for New Testament criticism. The displacement of leaves in a codex is a possibility, as it is not in a roll; and the date of the use of the codex is relevant, therefore, to the discussion, for instance, of the possibility of displacement in St John's Gospel. On the other hand, as we shall see (below, pp. 201 f.), certain theories about the order of books in the New Testament canon depend upon the assumption of roll form.

As for literacy, Paul, the possessor of those ' books and parchments', was of course well educated, and so were his assistants who acted as his amanuenses; and indeed most religious Jews would be at least literate, for they would at least have had the training of the local synagogue school, and would be used to reading the scriptures in translation if not in Hebrew (see above, p. 157). Jesus, it is true, is contemptuously described in Jn vii. 15 as never having learnt letters, but that would only mean that he had never been through the full course of training for a rabbi. [J. Jeremias, Jerusalem zur Zeit jesu (2 1958), 104.] Gentiles within the Christian Church would probably include among their numbers a good many (especially among the slaves) with little or no education; but a high proportion, of St Paul's converts at least, had already come into contact with the synagogue and had presumably begun to be familiar with the sound, if not the sight, of the Old Testament writings. Although it is impossible to go far beyond guess-work, for lack of data, one may guess, then, that most Christians of the New Testament period could read. That, however, is not the same as being wealthy enough to possess books; and presumably the average synagogue congregation would depend largely on remembering the scriptures and hearing them read Sabbath by Sabbath, while Christians, if they did not attend synagogue, might largely have to learn by listening to what was read at assemblies for Christian worship. In discussions and disputes, in the defence and confirmation of the Gospel, they would have to quote what they could remember of the Old Testament scriptures and of the sayings and writings of apostolic men. Comparatively few would be able to refer to writings in their own possession.

Scarce and precious, then, were Christian books and writings, and very long would Christian memories need to be. And it is against this mainly unliterary (though not illiterate) background that we have had to picture the rise of a Christian literature and must now consider the ultimate selection from it of authoritative writings. One of the obvious consequences of such a situation is the natural leadership that would be acquired in a given community by anyone sufficiently well equipped both with good character and learning to hold what few books might be available and to communicate their contents. (In 1 Tim. iv. 13, when Timothy is bidden to attend to reading this means, no doubt, reading aloud to his congregations.) Another consequence is, of course, that Christian assemblies for worship and religious instruction would be the most natural seminaries for such communications. Thus we may confidently assume that the messages of Christian writings took root largely through the ears of the faithful and of enquirers and catechumens, and in the context of worship and religious instruction.

But what were the factors that controlled the writing, and then the selection or rejection, of the various Christian books? Such evidence as we possess suggests that before the last of the apostles had died, there were already in existence various written documents containing sayings of Jesus and perhaps certain anecdotes in his life, and at least one full-length Gospel – St Mark's. It is difficult to establish more than this. The earliest (and most plausible) tradition about Mark places the writing even of his Gospel after Peter's death (though that would not mean after the last surviving apostle's death): the traditions which put it within Peter's life-time look suspiciously like 'improvements'. The direct apostolic authorship of St John's Gospel, maintained by tradition, has been very widely disputed. And Matt., in its present form, can hardly have been written by an apostle. [See the evidence for the traditions about Mk and Matt, in Huck-Lietzmann, Synopse 9 (or Eng. ed. by F. L. Cross), or in any good N. T. Introduction (e.g. A. Wikenhauser, Eng. trans. (1958), 160 f., 179 ff.); and see B. H. Streeter, The Four Gospels (4 1930) passim. See also pp. 89 ff., above.] So far as full Gospels go, therefore, we cannot be certain of more than that Mark's Gospel was written well within the natural life span of the apostolic generation – though in fact the most plausible dating of Matthew, Luke, and John places them also within this era. But it is not unlikely that Mark used already existing written sources besides spoken traditions. Of the existence of widely current traditions, whether written or spoken, at a very early stage there can be no doubt. One may already detect in the Pauline or near-Pauline Epistles echoes of just such traditions about Jesus as later find their way into the Gospels (see pp. 144 ff. above). Thus a missionary like Paul already had access to traditions, oral or written, such as the evangelists ultimately drew upon. Even much later, in the sub-apostolic writers known as the Apostolic Fathers, there are traces of similar traditions still running free, and possibly running parallel to the Gospels we know rather than through them. [See H. Köster, Synoptische Überlieferung bei den apostolischen Vätern (1957). ] Thus is built up a picture of a great reservoir of traditions, some spoken, some already written down, on which the early preachers are already drawing and from which ultimately the full-length Gospels are going to crystallize. Once again, it is the assemblies for worship and religious instruction that are the most likely reservoirs. The use of a writing in worship is the antecedent stage to its recognition as canonical.
Once the first full-length Gospel appears, a new genre of writing is in being. As we have seen, it is in essence only a fuller, more circumstantial, more pictorial form of the basic kerygma, the Christian proclamation. If we accept the priority of Mark over the other canonical Gospels – and despite the challenges to this view from some of the Roman Catholic scholars, it is soundly based [J. Chapman, Matthew, Mark and Luke (1937) and B. C. Butler, The Originality of St Matthew (1951) are comparatively recent Roman Catholic representatives of the arguments for the priority of' Matthew' (in some form or other). But it has to be added that this position is not by any means maintained by all R.C. scholars, and an instance of a Roman scholar abandoning it is W. Trilling, Das wahre Israel (1959). For some discussion of the Chapman-Butler position, see B. T. D. Smith (reviewing Chapman), J.T.S. xxxix (1938), 170 ff., A. M. Farrer (reviewing Butler), J.T.S., n.s., iii (1952), 102 ff. See further Excursus IV: The Priority of Mark, kindly contributed by G. M. Styler.] – we do not know positively of any full-length Gospel earlier than Mark's, though it is possible that he had precedessors, perhaps in Aramaic, or even Hebrew. [For Hebrew, not Aramaic, as a medium for early Christian writing, see M. H. Segal, A Grammar of Mishnaic Hebrew (1927), H. Grimme, 'Studien zum hebräischen Urmatthäus", Bibl. Zeitschr. 23 (1935 – 36), 244-265, 347-357, apud P. Nepper-Christensen, Das Matthäusevangelium etc. (as in next note); and for the possibility of some form or forms of spoken Hebrew side by side with Aramaic, see the interesting discussion (concluding in favour of the possibility), id. 101 ff., with special reference to H. Birkeland, Språk og religion hosjøder og arabere (1949), whose view is criticized by J. A. Emerton, 'Did Jesus speak Hebrew?', J.T.S., n.s., xii (1961), 189-202. Further, C. C. Torrey, 'The Aramaic Period of the Nascent Christian Church', Z.N.T.W. 44 (i952), 205 ff. (cf. id. The Apocalypse of John (1958)), suggests that the list of canonical scriptures discussed by J.-P. Audet in J.T.S. n.s. i (1950), 135 ff., lying as it does in the manuscript between 2 Clement and Didache, is probably of Christian origin, and therefore that Aramaic was the religious language even of Greek-speaking Christians before AD 70; he alludes (as does Audet, loc. cit.) to the lists in Epiphanius, De mens. et pond. 23, and Origen, In Ps. i, cf. Euseb. H.E. vi. 25, I f. ] After him, in due course, there came to be a great spate of Gospels. But it is noteworthy that of these the now canonical Gospels of Matthew, Luke, and John are still plausibly dated earlier than almost any others we know of (whatever may be implied by Luke i. 1 with its reference to 'many' predecessors); and many of the others, [Jerome alludes to an evangelium juxta (or secundum) Hebraeos (Adv. Pelag. iii. 2 and De vir. illust. ii.), and in the former passage says that many affirm that it is juxta Matthaeum. But it is notoriously doubtful how far (if at all) there really was a Semitic antecedent to Matt.: and in any case P. Nepper-Christensen, Das Matthäusevangelium: ein judenchristliches Evangelium? (1958), 64 ff. has cast grave doubts on almost every sentence of these two testimonia from Jerome. There is but slender evidence, then, that even this Gospel (if it ever existed) was prior to the canonical ones. When Luke refers (i. i) to 'many' predecessors, does he mean more than Mk and whatever other component parts he employed?] whether still extant or known through references and quotations, are not Gospels in the same sense at all, but are either anecdotes about Jesus or mythological romances, or sayings-collections. [It is striking that comparison of the Coptic Gospel of Thomas with the Greek Oxyrhynchus sayings suggests that the Greek sayings may represent earlier sayings-collections used by the writer of the Gospel of Thomas. If so, we are presented with an instance of non-canonical sayings-collections circulating and being absorbed into a subsequent writing, rather as we may assume happened earlier in the compilation of the canonical Gospels. See H.-W. Bartsch,' Das Thomas-Evangelium und die synoptischen Evangelien", J.N.T.S., 6 (1960), 249 ff. But there is the important difference (noted by Bartsch) that the Synoptic antecedents are not mere collections of sayings detached from their narrative settings. The sayings-collections behind Ev. Thom, are not therefore closely comparable. G. Miegge, Gospel and Myth (Eng. trans, by S. C. Neill, 1960, from the Italian of 1956), 121 f. points out that, whereas orthodox Christian writing moved in the direction of 'theological meditation', the apocryphal writings with Gnostic leanings moved in the direction of 'mythological imagination'.] Perhaps in the end it was really not so much a matter of selecting as of recognizing that only four full-length Gospels were available from within the apostolic period. And if it be asked why these maintained their independence, instead of suffering fusion (as in Tatian's Diatessaron) or instead of one alone coming out as sole survivor, the answer may be found in the authority of local churches or in some other prestige. The one of the four that most nearly went under was Mark, because of its brevity and because of the fact that its substance was so largely included in Matthew; but, perhaps because it was connected with Peter or because it was connected with Rome or because of both, it held its own. And each of the other three evidently represented an influential centre of Christianity. Not that they all immediately gained universal recognition. Towards the end of the second century, Luke was still only hesitantly recognized, and John was to meet much opposition until as late as about AD 22O. [See Schneemelcher, after Bauer, in Hennecke 3 (as on p. 182, n. i), i. n.] But the point is that what ultimately emerged was not a single Gospel but four – neither more nor less. Marcion and Tatian both tried, in their different ways, to establish a single Gospel, but did not carry the whole Church with them. [According to W. Bauer's thesis (in Rechtgläubigkeit u. Ketzerei im ältesten Christentum, 1934) the Roman Church and other Churches in its sphere of influence, for long recognized only Matt, and Mk, accepting only with hesitation Lk. (discredited by heretical use) and abandoning direct opposition to Jn only c. 200. He thinks Papias recognized only Mk and Matt.; Justin did not regard Jn as authoritative; Ignatius used not Jn but Matt.
This differs from Harnack (see, e.g., The Origin of the New Testament, Eng. trans., 1925), despite many points in common. Harnack too saw the fixing of a four-Gospel canon as a compromise, and saw Jn as playing a decisive role in the process. But he thought that the opposition to the 'Alogi' led to the championing of the Fourth Gospel in the East at a time when the other Gospels were so well established that they could not be ousted.]
The process of selection was well under way before ever it began to be consciously reasoned about or rationalized: owing to a variety of causes, some of which have just been mentioned, the four-Gospel canon slid into existence almost furtively. It was certainly not the arbitrary decision of a single Christian body, still less of an individual. Its formal declaration, when it was made, [For the ultimate determination of the Canon of the New Testament, see the bibliography for this chapter. The date usually given is that of Athanasius" 39th Paschal letter, AD 367.] was only the recognition, by the Church collectively, of a conviction that had long been silently growing on their consciousness. Perhaps at least one or two of these writings had been regularly used in assemblies for worship long before they were officially described as authoritative.

As it happens, much the same could be said of the recognition of the Jewish scriptures also. We have to remember that there is no known official pronouncement embodying a Jewish' canon' of scripture until near the end of the New Testament period (although there is evidence of long debate before this). The Synod of Jamnia (Jabne) in AD 90 is usually claimed as the occasion of this pronouncement; and a very interesting discussion of it by P. Katz [P. Katz, 'The Old Testament Canon in Palestine and Alexandria', Z.N.T.W. 47 (1956), 191 ff. ] attributes to this synod also the present arrangement and division of the Hebrew Bible, arguing that this was not an ancient order at all, but that (contrary to generally held opinion) it is the arrangement of the books in the Septuagint that represents the earlier Hebrew order. However that may be, the Synod of Jamnia seems to represent the first official Jewish canon – and even that was only official for a section of Judaism: [katapi ed: but read also this WIKI article! ] there was no such thing as an ecumenical organ of Jewish opinion, and doubtless the Jews of the Alexandrine and other dispersions continued to be without a defined 'canon' of scripture. [One wonders what relation, if any, to Jamnia may be assumed for the use by a Christian of the phrase 'every inspired scripture' in 2 Tim. iii. 16. Was it known by this time precisely what 'scripture' comprised?] The ' Sanhedrin' of Jamnia (and thus of the Mishnah Tractate Sanhedrin) is not the lineal descendant of the Jerusalem Sanhedrin of the pre-70 period; it consisted not of priests and elders but of rabbis. Tractate Sanhedrin is not to be incautiously used as evidence for the Jerusalem Sanhedrin. [Lohse in T.W.N.T., s.v. συνέδριον.] Jamnia may well, however, have added some momentum to the corresponding Christian process.

If we ask what criteria the Church consciously applied to test the authenticity of its writings, we shall find that they are criteria dictated by controversy with heretics or disbelievers. The most obvious one is 'apostolicity'. If not actually written by one of the Twelve, a Gospel (to confine our enquiry for the moment to this category) must at least have some kind of apostolic imprimatur: it must be shown to come from some close associate of an apostle and, if possible, with the apostle's express commission. Consequently, it must necessarily belong to an early period, and would be expected (one may suppose) to show signs of at least derivation from the primitive Aramaic-speaking Church. A corollary of this, itself constituting another criterion, was that no genuinely apostolic Gospel could contain an interpretation of the incarnation contrary to that orthodoxy which (however difficult to define exactly) undeniably belonged to the communis sensus fidelium of these early decades, despite the great range of differences already described. For by the time the last of the Twelve had died, there was a sufficiently powerful uniformity, so far as the basic convictions of the Church's leaders go, running throughout the Christian centres all over the empire, to detect and extrude' heresy' – that is, any opinion incompatible with the apostolic witness. Finally, no doubt a Gospel needed to be a Gospel. There were evidently a number of fragmentary documents in circulation then, as there continued to be later – sayings-collections, collections of miracle stories, perhaps, and of other anecdotes; but these could not stand on their own feet as Gospels proper, comprising the kerygma.

Now, when the Church came consciously to apply these tests, it was sometimes one of them, and sometimes another which was uppermost. The original apostolic contact was clearly a primary demand; but it was not always possible to test this as rigorously as might have been desired; and alongside came the additional test of usage: had the book proved its worth? had it survived the critical sense of the Christian tradition? – for it is possible that certain writings had already asserted themselves as eminently useful and sound before evidence for apostolic contact was discovered. In some few instances it may even be that the latter was a post-hoc rationalization. But in such cases the communis sensus fidelium had already been so soundly informed by authentic tradition that its own imprimatur was in fact sufficient.

It must be added, in all fairness, that we have, in Christian antiquity, at least one instance of an author being very severely penalized for a work which, though perfectly orthodox (at any rate by the standards of his period), turned out to be a fiction. This is the story told by Tertullian in De Baptismo 17 of the priest of Asia who admitted to having written the Acts of Paul (which includes the Acts of Paul and Thecla) and was deprived of office (presumably on the charge of passing off fiction as though it had been history). This gives the lie to the idea that the early Church never exercised any historical criticism, but relied entirely on its sense of what was orthodox and edifying. The unsuspecting priest had done it in all good faith, meaning to enhance the honour of the apostle; and he had written anonymously, not attempting to assume the authority of another. How much severer (it might have been supposed) would the Church have been towards one who was caught impersonating (say) an apostle! On the other hand, it is one thing to write fiction as though it were history and quite another to communicate teaching in someone else's name. This latter was a time-honoured technique and scandalized nobody unless the teaching was either maliciously intended (cf. 2 Thess. ii. 2, iii. 17) [But 2 Thess. iii. 17 represents the stress of dangerous circumstances: by the time of the Pastorals, things may have been different. ] or heretical (cf. the case of the Gospel of Peter, below, p. 193). Yet even in the realm of teaching, if they did indulge in what has just been called a post-hoc rationalization, one may assume that it was not consciously thought to be a rationalization. The acceptance of Hebrews as Pauline is an instance. There were at first some doubters in the East and many in the West. But it was indispensable; it was compatible with Pauline doctrine; and the prevailing opinion of the Eastern Churches was eventually accepted – that it was Pauline (Moffatt, I.L.N.T., 431). And it must be remembered that Hebrews was not accepted without challenge. As late a writer (in our time scale) as Tertullian, De Pud. 20, attributes Hebrews to Barnabas, and, though preferring it to Hermas, does not treat the work as scripture (Gregory, Canon and Text, 222 ff.; see further Moffatt, I.L.N.T., 437).

Moreover, it must be remembered that Hermas, the writer of the popular Shepherd, did not, even as a teacher, try to write under another name; and it is noteworthy, further, that in the ante-Nicene era, forgeries were relatively rare and often detected. ' No one in this period made much use of most of the forgeries we later encounter in such quantities.' (R. M. Grant, 'The Appeal to the Early Fathers', J.T.S. n.s. xi (1960), 23.)

Nevertheless, even though the early Church was much more alert against 'forgery' than is sometimes supposed, we have to recognize a two-way traffic: the living community was indeed constantly subject to check and correction by the authentic evidence – by the basic witness, first of accredited eye-witness apostles and later of the written deposit of that witness; yet also the documents which soon began to circulate in considerable numbers were themselves in some measure subject to check and correction, whatever their origin, by the living community – simply because it by that time contained within it, or among its leaders, a sufficiently firm and uniform tradition to constitute it corporately a preserver of tradition. One may see the living community's control operating in the Johannine Epistles, where appeal is made to authentic, early, eye-witness tradition against the opinionated interpretations and assertions of men who were running contrary to these. Something similar, though without the allusion to eye-witness, is to be seen in the Pastoral Epistles. The further one goes in time from the original sources of evidence ('that which was from the beginning', 1 Jn i. 1, etc.) the more precarious becomes the claim of the Church, in its own right, to test the evidence. But until there is a recognized body of authentic documents to appeal to, there is no other way; and this is how we see the Church proceeding before the establishment of the canon of Christian scriptures. It is certainly a fact that the claims of documents were, until this stage, checked by that tradition of orthodoxy which was generally diffused through the living, worshipping communities. [Very important here is the fact that Gnostic writings, such as The Gospel of Truth, are now available to demonstrate what sort of doctrinal criteria the Church must have used in excluding them. See W. C. van Unnik, Newly discovered Gnostic Writings (Eng. trans., 1960, of Openbaringen uit Egyptisch Zand, 1958), 68, 90, 91. Note, further, that the inclusion of the Paulines and the exclusion of (e.g.) Clement shows that the Church had not lost its appreciation of what Paul stood for.] That tradition, in its turn, however, was subject to the check of apostolic eye-witness as long as the apostles were available; and before the death of the last apostle, as we have seen, there had already begun to exist at least some soundly attested apostolic documents.

At this point it is worth while to observe that a distinction has been drawn, by K. Aland [K. Aland, ' The Problem of Anonymity and Pseudonymity in Christian Literature of the First Two Centuries', J.T.S., n.s., xii (1961), I ff.] between an earlier period (to about AD 150) and a later. In the earlier period, the prophetic afflatus was recognized, and a teacher could stand up and speak in a Christian assembly in the name of the Spirit and in the name of some great apostolic leader, and be accepted. This (Aland suggests) is how some writings that we should be tempted to call pseudonymous came to be openly accepted: they were bona fide utterances of formerly known (though now anonymous) speakers in the name of apostolic men. But when the afflatus waned, and the Church became conscious of living in an era separated from that of the apostles, the more literary tests of authority began to come in, as did also deliberate forgeries. [There is a well-documented essay oz. 'epistolary pseudepigraphy' in D. Guthrie, New Testament Introduction: the Pauline Epistles (1961), 282 ff.]

Such extra-canonical documents styling themselves 'Gospels' as we actually possess are all of a suspect character. The two most recently recovered – the probably Valentinian Gospel of Truth and the Gospel of Thomas – must certainly have been rejected for their contents, even if on other grounds they had seemed to carry impressive claims. As a matter of fact, neither of them is a Gospel in the sense of comprising the kerygma, and the Gospel of Truth has not even the semblance of apostolic authority either. There is no known extra-canonical Gospel material which is not (when it can be tested at all) in some way subject to suspicion for its genuineness or its orthodoxy: many of the recovered fragments are in some sense 'gnostic' in tendency. [See, e.g., J. Jeremias, Unknown Sayings of Jesus (Eng. trans. (1957) of Unbekannte Jesusworte (1948)); B. Gartner, The Theology of the Gospel of Thomas (Eng. trans. (1961) from Swedish); E. Haenchen, Die Botschaft des Thomasevangeliums (1961).]

The stock example of the application of the test of orthodoxy, as against that of apostolic attribution, is the story of Serapion, bishop of Antioch in about AD 200, quoted by Eusebius (H.E. vi. 12). The Greek of the passage is by no means lucid; but this much is clear, that Serapion had, without reading it himself, sanctioned the use, in the Christian community of Rhossus (a small town in Cilicia), of a writing calling itself the Gospel of Peter. Subsequently, however, he discovered (still possibly not at first-hand – the Greek is very odd) that, while the greater part of this Gospel was 'in accordance with the true teaching of the Saviour', there were heretical additions which had apparently been used by heretical teachers to lead the Christians of Rhossus astray. Against these he warns them, since, whereas ' we receive both Peter and the other apostles as Christ' (cf. Matt. x. 40, Gal.iv. 14, 1 Clement xii. 1), 'the writings which falsely bear their names we reject, as men of experience, knowing that such were not handed down to us'. Serapion's treatise from which Eusebius quotes was called περὶ τοῦ λεγομένου κατὲ Πέτρον εὐαγγελίου ('concerning the so-called Gospel according to Peter'), and it seems to be clear that the grounds for this stricture, 'so-called', were entirely its heretical character, not any research into its origin. The fragment of this very Gospel which was discovered at Akhmim in the winter of 1886-87 confirms Serapion's judgment: or, if it is impossible to prove that it is strictly docetic or otherwise heretical, at least its extravagances mark it as spurious and well on the way to heretical fancies. It breathes an entirely different atmosphere from that of the canonical Gospels. [See Chr. Maurer in Hennecke 3 (as on p. 182, n. i), 119 f.]
Thus, while the earliest Church was shaped and controlled by the evidence of all the eye-witnesses, and especially the authenticated Twelve, there came a brief period when this evidence had become so entirely a part of the life and thinking of the leaders of the Church that they automatically refused to assimilate into their system what was contrary in doctrinal tendency to the now indigenous standards. This brief transitional period, between the earliest stage, when presumably the eyewitness test was constantly applied, and the later stage of confidence when even what claimed to be apostolic witness was itself subjected to the doctrinal test, may perhaps be illustrated by parts of the Pastoral Epistles. These betray an awareness of 'orthodoxy'; and although the 'faithful sayings' cited in the Pastorals are not sayings of Jesus and do not in any sense represent a ' canon', yet the very phrase shows an instinct for classification into true and false. [See Excursus III: ΠΙΣΤΟΣ Ο ΛΟΓΟΣ. ] Moreoever, a good deal of prominence is given in these Epistles to the need for careful transmission of the apostolic teaching: it is a precious deposit, entrusted by God to the apostle, and by the apostle to his chosen disciple, to be handed on by him to carefully chosen men. The 'pattern of teaching' (τύπος διδαχῆς) of Rom. vi. 17, and the 'traditions' (παραδόσεις) of 2 Thess. ii. 15, iii. 6 (cf. 1 Thess. iv. 1 f.), are on their way, via the 'sketch' or 'outline' (ὑποτύπχσις) of sound teaching (2 Tim. i. 13) and the παραθήκη or 'deposit' (1 Tim. vi. 20, 2 Tim. i. 12, 14; cf. ii. 2), into the 'canon' of approved writings. The recognition in the whole Church of the four Gospels as alone authentic is difficult to date. There are scraps of evidence that they circulated independently for a considerable period after their first appearance. Thus, to judge by the textual history of the manuscripts of Mark, this Gospel was by far the most heavily corrected by scribes; which probably means that it had a longer independent history than the others. [See G. D. Kilpatrick, The Transmission of the New Testament and its Reliability (Victoria Institute Lecture, 1957), 96; and note (with S. E. Johnson, Mark (1960), 30) the interesting case of the solitary Mark in the University of Chicago Library (Chicago MS. 972 = Gregory-Eltester catalogue Codex 2427), described by H. R. Willoughby in Munera Studiosa (ed. M. H. Shepherd Jr., and S. E. Johnson, 1946), 127 ff. and R. P. Casey in Journal of Religion xxvii (1947), 148 f.] Indeed, there are signs that it was early recognized as the First Gospel, despite the later traditions which placed Matthew on the pedestal of primacy, [3 F. C. Grant, The Gospels (1957), 64 ff.] and, if so, it may well have circulated independently. But as soon as Matthew did rise to the highest place in popularity, then, for the same reason, it may be presumed to have been often copied by itself (though, as has just been implied, less often than Mark had been separately copied). Thus, 'for the Syrian Church the (written) Gospel long continued to be that of Matthew, as it had been elsewhere'. [B. W. Bacon, Introduction to the New Testament (1900), 38.]

As for Luke and John, there is much a priori likelihood that they both circulated for some time among the particular persons or groups for whom, respectively, they were first intended. Many of the Gospel echoes in the Epistles seem to be exclusively Lucan, suggesting, perhaps, that Luke alone was the Gospel for the Pauline circle; and R. G. Heard, in a posthumously published article, [R. G. Heard, J.T.S., n.s., vi (1955), 3.] points out the evidence afforded by a comparison among themselves of the 'old Gospel Prologues' for the independent circulation of the Prologue to Luke even in the fourth century. With regard to John, it is a well-known matter for discussion whether even Justin Martyr (circa AD 150) knew it. [On Justin's sources, see W. Bousset, Die Evangetiencitate Justins (1891); J. M. Heer, Römische Quartalschrift 28 (1914), 97 ff.; E. R. Buckley, J. T.S. (1935). 173 ff-> A. Baumstark, Biblica 16 (1935), 292 ff. B. F. Westcott, A General Survey of the History of the Canon of the New Testament (4 1875), 166, n. i, cites as the chief passages: Jn iii. 5-8, Apol. i. 61; Jn i. 13, Trypho Ixiii; Jn i. 12, Trypho cxxiii; Jn xii. 49, Trypho Ivi; Jn vii. 12, Trypho Ixix. Of these, Jn i. 12, 13, vii. 12 are very faintly echoed, if at all; and xii. 49 seems to be cited by Westcott in error (? v. 19). J. N. Sanders, The Fourth Gospel in the Early Church (1943), examines a wider range of passages, concluding (31): ' The most reasonable conclusion from this examination of the passages ... which appear to show traces of the influence of the Fourth Gospel seems to be that certain passages are most naturally explained as reminiscences of the Fourth Gospel, while there are few, if any, which can be certainly said to be dependent upon it'. ] There is an echo of what we know as a Johannine saying ('unless you are born again, you will certainly not enter the Kingdom of the heavens') in Apol. i. 61, but if Justin did know the Gospel, it has not influenced his theology. On the other hand he cites a 'Q' saying ('All things are committed to me ...', Matt. xi. 27, Lk. x. 22) with the formula ἐν τῷ εὐαγγελίῳ γέγραπται, 'it is written in the Gospel' (Dial. 100).

Rather earlier (? circa AD 130) the Epistle of Barnabas (iv. 14) uses γέγραπται, 'it is written", to introduce a phrase known to us as Matthean ('many are called, but few chosen'). But it must be admitted that γέγραπται need not imply the concept of authoritative scripture; and Köster (op. cit., p. 185, n. 2, 126) questions whether even this saying can be pinned down to an exclusively Matthean origin. However, the main point is that here are sufficient hints to suggest a period during which the four Gospels were in existence separately and were used, perhaps, only locally.

But if there was a period when each group or area recognized only one or two Gospels in writing, there came a time when the four emerged into equal recognition. Cullmann ['Die Pluralität der Evangelien als theologisches Problem im Altertum', Th. Zeitschr. i (1945), 21 ff., = The Early Church (1956), 39 ff.] argues that there were two conflicting motives for the fourfold canon: (i) the conviction that no one Gospel presented the fullest possible witness to the Incarnate; (ii) the desire of each Evangelist to make a single Gospel which should include his predecessors' witness. The result was reduction – though not reduction to only one. In briefest summary, Cullmann's argument may be outlined as continuing thus: in view of the spate of manifestly gnostic and false writings, it was natural to treasure all of the few really apostolic ones. Then (unlike the attitude of the earlier period) came the desire to champion one versus the rest. Mar-cion's is the best known attempt. But it is a docetic tendency to conflate into one, or to choose one against the rest: it is a genuinely historical insight to recognize a plurality in human witness. Irenaeus himself, claiming that four is a divinely rather than a humanly chosen number, really looks in the same (mistaken) direction as docetism. He ought to have seen that the scandalan of human diversity should be accepted as such! The earlier Church was right – to accept all that was thought to be truly apostolic, and to see it as mediating through human diversity, the one divine event. (The Muratorian Canon itself is not far from such a recognition when it says '... though various ideas are taught in the several books of the Gospels, yet it makes no difference to the faith of believers, since by one sovereign Spirit all things are declared in all of them ...' [Translation from Stevenson.]; yet, in the preceding section the concept of a single, supreme Gospel has already appeared in the legend of John, of whom it was revealed to Andrew that John was to write all things in his own name.) It was probably from the middle of the second century that it became usual to speak of 'the Gospel according to Matthew' etc. (reflecting what have recently become form-critical findings about a single kerygma, repeated in various ways).

Our earliest extant manuscripts – the Chester Beatty papyri and the Sinaitic Syriac – take us back to AD 250 or earlier for the four together; but the Gospel of Peter, the fragment of which, already mentioned, shows signs of the use of all four, cannot be much later than 150, since it was well established at Rhossus when Serapion found it; and Tatian's famous Diatessaron – his conflation of the four into a single Gospel – gives us a similar date for the joint recognition of at least the four (even if there are obscure traces of a fifth or of an independent tradition in his conflation – unless these be due simply to the use of a different text). [See B. M. Metzger, art. Evangelienharmonie in R.G.G. Tatian was a pupil of Justin's. Hitherto, only one tiny fragment in Greek of his Diatessaron has been found (publ. by C. H. Kraeling, 1935). Otherwise, our sources are (for the East) Ephraem's commentary on the Diatessaron (c. 360) in Armenian translation; Arabic translations from Syriac; some later versions; and scattered quotations in the Syriac and Armenian Church Fathers; (for the West) the harmony, with Vulgate text, in codex Fuldensis (before 546); later versions; scattered quotations. There are apocryphal additions and the canonical material is occasionally coloured with ascetical touches. It is these phenomena that G. Messina, S.J., Diatessaron Persiano (1951), and others are trying to account for. See an interesting discussion in H. W. Montefiore and H. E. W. Turner, Thomas and the Evangelists (1962), forthcoming. Jerome (Epist. cxxi. 6. 15) alludes to Theophilus of Antioch's similar work, but of this we have no other-trace.] By about AD 185, Irenaeus is proclaiming that it is as inevitable that there should be four Gospels as that there should be four winds and four corners of the earth. If this is polemic, as it well may be, the target of his attack is probably heretical opinion: there is nothing to suggest that any considerable section of the orthodox Church would have wished to deny his assertions. After Irenaeus follows Clement of Alexandria and thereafter plenty of evidence to confirm the recognition of the four alone. But it is to be noted that until this mid-second century date there is remarkably little sign of this. Hermas, Visions, iii. 13, misses a golden opportunity of mentioning the four (unless one can indeed believe that the four legs of the bench are meant to symbolize the Gospels [Strongly denied by H. Köster, Synoptische Überlieferung bei den apostolischen Vätern (1957), 254. ]), and it is a matter of considerable uncertainty whether the recognition of the canonical Gospels can be detected in the Apostolic Fathers generally. But at least it can be said with confidence that the four-fold canon is well established before our earliest official lists of accredited books. Also (to judge by Justin's attitude) Gospels generally (no matter how many) tended to carry more weight, as the apostles' 'reminiscences', ἀπομνημονεύματα, [See R. G. Heard, 'The ἀπομνημονεύματα in Papias, Justin, and Irenaeus", J.N.T.S. I. 2 (Nov., I9S4), 122 ft, with some discussion of the meaning of the word and of Heard's article by N. Hyldahl, 'Hegesipps Hypomnemata', Studio Theologica, xiv (1960), 70 ff. Of earlier writers, note E. Lippelt, Quae fuerint Justini M. ἀπομνημονεύματα (1901).] than other writings. Before Justin, Ignatius had said that he took refuge in the Gospel as in the flesh of Jesus and in the apostles as in the presbytery of the Church (Philad. v. 1). But the margin of time between such vaguely defined independent or partially independent circulation of Gospel material and the solid unity-in-quaternity [There may be something to be learnt from the parallel and contrast offered by the convergence and conflation of various accounts of Israel's 'Gospel' – the story of the exodus and the covenant – within the Pentateuch.] of the fourfold Gospel canon appears to be slender. It looks uncommonly as though something which we do not know about acted as a rather sudden incentive to the collection.

This brings us to Marcion. Did that extremely interesting heretic find four Gospels already recognized together by about AD 140, and did he deliberately drop off Matthew, Mark, and John (as well as the unacceptable parts of Luke)? Or was it rather that the catholic Church, after seeing what havoc Marcion wrought by his one-sided use of documents, brought the four Gospels together to restore the balance and make a fourfold harmony? This is the same problem as confronts us for the whole New Testament canon: was Marcion's the first canon, and is the orthodox canon the catholic Church's subsequent reply? Or did Marcion play fast and loose with an already existing canon? There is at present no absolutely conclusive evidence for the existence of a pre-Marcionite catholic canon. Marcion may have been the catalyst we have already hinted at. We cannot be certain.

Even the evidence of the famous Gospel prologues [See D. de Bruyne, 'Prologues bibliques d'origine marcionite', Revue Benédictine, xxiv (1907), i ff.; 'Les Plus Anciens Prologues latins des évangiles', ibid. xl (1928), 193 ff.; also Huck-Lietzmann, Synapse 9 (or Eng. ed. by F. L. Cross), viii; W. F. Howard, E.T. xlvii (1936), 535 f.; R. M. Grant, 'The Oldest Gospel Prologues', Ang. Theol. Rev. xxiii (1941), 231 ff. ] is inconclusive. These are to be distinguished from the brief descriptive prologues to the Epistles which, since de Bruyne's brilliant observations, are believed to have originated with Marcion himself. The oldest Gospel prologues, by contrast, were believed by de Bruyne to be anti-Marcionite. The late Mr R. G. Heard, [R. G. Heard, 'The Old Gospel Prologues', J.T.S., n.s., vi (1955), 1 ff., and further literature there cited.] however, questioned this conclusion, at any rate for the prologue to Luke, which, he argued, is independent of the prologues to Mark and John (that to Matthew is not extant), [The 'Monarchian Prologues' (for description and references, see F. J. Foakes-Jackson and K. Lake, The Beginnings of Christianity, ii (1922), 242 ff.), which do include Matt., are not to be confused with the ancient, anti-Marcionite ones.] and is free from anti-Marcionite material. Once again, then, little if any light can be shed on the relation of the canon to the heretic.

But meanwhile, how had the epistles and other writings fared? One of the most elusive problems in the history of the New Testament canon concerns the origin of the collected letters of Paul, the Pauline corpus. Even when it did first emerge, it was not complete: some Pauline writings had already been lost (or at any rate were omitted), 1 Cor. v. 10 appears to refer to a previous letter of which, at most, a fragment may be detected in 2 Cor. vi. 14-vii. 1. In 2 Cor. vii. 8 allusions are made to a severe letter which, though it may conceivably be i Cor. itself, is more likely identified either with 2 Cor. x ff. or with a completely lost letter. In Col. iv. 16 a letter 'from Laodicea' is mentioned, which evidently means a letter from Paul which is to be passed on to Colossae from Laodicea in exchange for the letter to the Colossians. This 'Laodiceans' is also either totally lost, or is to be identified with what we now know as 'Ephesians' (or else, as has been conjectured by E. J. Goodspeed, with 'Philemon'). [E. J. Goodspeed, New Solutions to New Testament Problems (1927), The Meaning of Ephesians (1933), The Key to Ephesians (1956).
For theories finding a plurality of letters in Phil., see F. W. Beare, Philippians (1959); B. D. Rahtjen, 'The Three Letters of Paul to the Philippians', J.N.T.S. 6. 2 (January 1960), 167 ff., replied to by B. S. Mackay, 'Further thoughts on Philippians", J.N.T.S. 7. 2 (January 1961), 161 ff.]
Conversely, that some of the letters ultimately included in the canon as Pauline are not by Paul falls only just short of demonstration. The Epistle to the Hebrews, though eventually included in the Pauline canon, was regarded as not of Paul's workmanship by several writers of antiquity (cf. p. 191 above), and now is widely acknowledged to be un-Pauline; and a large number of scholars regard the Pastoral Epistles as at least in part post-Pauline; while varying degrees of suspicion are cast on other Epistles, especially on 2 Thess. and Ephesians. Thus, some were irretrievably lost, some may have eventually been added from the Pauline circle rather than from Paul himself. It was a deeply ingrained tradition in Jewish circles that certain genres of writing should, as a matter of course, be written under the name of their representative authors: Law was by Moses, Wisdom by Solomon, Psalms mostly by David; [See L. H. Brockington, 'The Problem of Pseudonymity', J.T.S., n.s., iv (1953), 15 ff.] and at least during K. Aland's 'earlier period' (see above, p. 192) this may have held good for Christians also. But the question still remains: Who or what prompted a collection of Pauline Epistles in the first instance?

As with the four Gospels, so with the Pauline Epistles, we know that they existed before ever they were presented in a single collection. Not only so, but there is some evidence in the manuscript tradition, comparable to that mentioned above for the Gospels, suggesting that they actually circulated for a time separately (and this, it is to be noted, applies to Eph. i). [See a suggestive footnote in F. W. Beare's commentary on Ephesians in The Interpreter's Bible x (1953), 601. ] And the appearance of the collection can be plausibly placed within a fairly short period. Admittedly, we have insufficient data for certainty: hence the word 'plausibly'. But the point is that Acts shows no trace of a knowledge of the Pauline Epistles, whereas 1 Clement (generally dated about AD 95) does, and thereafter there are enough echoes to show that they were at least beginning to be known. Yet, even so, evidence for the knowledge of one or two Pauline Epistles is not evidence for the existence of a collection, a corpus; and as with the Gospels, so with the Pauline corpus, Marcion is the really important land-mark. He, we know on the evidence of Tertullian, used a collection of Pauline Epistles. Was it then he, in fact, who created it?

Two particularly interesting things we learn from Tertullian about Marcion's Apostolicon or collection of the apostle's writings: that he knew our Ephesians as 'Laodiceans', and that his collection contained the following letters (probably in this order): Gal., 1, 2 Cor., Rom. 1, 2 Thess., Laodiceans, Col., Philem., Phil. That is, Marcion used the nine great Paulines (and Philem.).

What may be learnt by comparing the order in other known lists? The best-known of these is the Muratorian Canon (Stevenson, No. 124), towards the end of the second century, whose catalogue is: Cor., Eph., Phil., Col., Gal., Thess., Rom. (in that order), and also (though not necessarily in this order) Philem., Titus, 1, 2 Tim. The Muratorian Canon goes out of its way here to mention also the Epistles to the Laodiceans and Alexandrines, falsely, it says, attributed to Paul 'in connexion with' (ad ) the heresy of Marcion. A brilliant conjecture by J. Knox [I. Knox, Marcion and the New Testament (1942).] relates Marcion's and the Muratorian orders as follows, in two steps: (a) Marcion, using an already established list with Ephesians at its head, transposed Eph. (Laodiceans) and Gal., bringing Gal. to the head because it was specially important for his own doctrinal purposes, (b) Assuming that the letters of the already established list were contained in two rolls of roughly equal length, [W. H. Brownlee, in a paper read to the Orientalists' congress (1960, cf. his The Meaning of the Qumran Scrolls for Bible and Religion, 1962), wrote: 'The gap between Chapters thirty-three and thirty-four in the complete Isaiah scroll (1Q Isa"), together with orthographic peculiarities of each half, points to the practice of bisecting the Book of Isaiah into two scrolls: (1) Chapters i-xxxiii and (2) Chapters xxxiv-lxvi. ... The ancient practice of bisecting books is well discussed by H. St John Thackeray in his Schweich Lectures of 1920, The Septuagint and Jewish Worship, Appendix IV, pp. 130-136. This was done, at times, for the convenience of handling by a purely mechanical division of large works into two scrolls; some of the better constructed books of antiquity, however, were so composed as to yield a natural literary division at about the mid-point of the work. ... In the scroll, each [half] consists of twenty-seven columns, and the point of division lies between two sheets of skin, so if it were not that they happen to be sewed together they could easily circulate as two separate scrolls. An unprecedented gap of three lines occurs at the bottom of Col. XXVII, separating the two volumes.' ] the two most substantial components, 1, 2 Cor. and Rom., will probably have been placed one in each roll. This might make one roll contain Eph. and Cor.; the other all the rest. And if one divides the Muratorian canon into two sections on this principle and reverses the orders of the two sections, one gets exactly Marcion's order (with the one alteration already noted in (a)). Is it possible, then, that these two lists are the result of rolling up the two rolls in opposite directions? This, though an exceedingly ingenious guess, is only a guess as the author himself is the first to admit. And, as has already been remarked (p. 183 above), it is by no means clear that the writings catalogued in the Muratorian canon would have been on rolls and not in a codex – in which case the theory would fall to the ground. If we deny ourselves the luxury of Knox's theory, we seem to be left without a clue as to the relation between the heretical and the (presumably) orthodox list: and in that case we still do not know which came first – whether Marcion first collected his apostolicon and later the orthodox Church decided to make theirs, or whether Marcion tampered with an already existing list. [See a discussion of this problem in E. C. Blackman, Marcion and his Influence (1948), ch. 2.] The situation here is the same as with the Gospels.

One other exceedingly ingenious attempt has been made to find an individual to whom to attribute the formation of the Pauline collection. This is E. J. Goodspeed's theory (see p. 199, n. 3 above) about Ephesians. Noting, as all serious enquirers must, the close relation between Eph. and Col., and, at the same time, the very remarkable differences, and convinced that Eph. was not Pauline, Goodspeed reached the conclusion that it must be the work of a student and admirer of the Pauline Epistles, who was particularly familiar with Col., but knew and echoed all the others also. Who might such an individual be, and what led him to know the whole Pauline corpus ? Might it not have been the reading of the Acts that led this person to go round the Pauline centres looking for the letters? It is well known that the Acts shows practically no trace of a knowledge of the Pauline Epistles; whereas Eph. (assuming its non-Pauline authorship) is the first writing to reflect them all, and thereafter they are often echoed. Moreover, a fashion for issuing collections of Epistles seems to have set in – witness the seven letters of the Apocalypse and the seven letters of Ignatius (note that the Pauline corpus can itself be seen as seven-fold plus one – Rom., Cor., Gal., Phil., Col., Thess., [Philemon] (six) plus Eph.). Of Goodspeed's theory that Philemon was ' Laodiceans' we need not here speak. What concerns us at the moment is that, on his showing, Eph. is, as it were, the covering letter to the collection. Instead of writing an introduction in his own name, as a modern editor would, this disciple and admirer writes a glowing recapitulation of the apostle's message, modestly concealing his own identity by the then familiar technique of writing as from the apostle himself. And who was he? If an identification were to be hazarded, Goodspeed (supported here by Knox) [J. Knox, Philemon among the Letters of Paul (11959), especially ch. 5.] would suggest Onesimus. Who would be more intimate with Col. (and Philemon) than the slave whose future depended so completely on the success of these letters? And who is so naturally associated with Ephesus as Onesimus who (according to Ignatius' letter to Ephesus) was in later years Bishop of that Church? There are many minor difficulties in the way of this extremely skilful solution of the Ephesians problem (for which the reader must be referred to works of introduction and exegesis); but one major problem is the position of Eph. in lists of the Epistles. In no known list does it occur either first or last (the only positions which seem to be appropriate to it on Goodspeed's theory of its origin and function). Only if we accept Knox's brilliant conjecture about the original order both of Marcion's and of the Muratorian list can Eph. be brought to the head of the list.

If we abandon the idea that the collecting of the Pauline letters was the work of an individual, such as Onesimus or Marcion, we are left with the time-honoured alternative – the slow, anonymous process of accretion, the snowball theory. We have to suppose, that is, that the intercourse between one Pauline centre and another gradually led to the exchange of copies of letters, until, at any given centre, there came to be not only the letter or letters originally sent to it, but also copies of certain others, collected from other Pauline Churches. Thus in each centre there would come to be little nests of letters, and gradually these would move into wider circulation and would be augmented, until the full number, as we know it, was reached. Then all that remained to be done was the making of a careful 'edition* of the whole corpus.

Such a theory ignores Acts as a precipitant; it also depends a good deal on the assumption of a live interest in Paul between his death and the decisive acceptance of his letters; and there is no evidence that Paul dominated the Christian world of those days as he has (for the most part) dominated it since the inclusion of his letters in scripture. Papias (to cite a famous example), although the friend of Polycarp, shows no evidence of the use of the Pauline writings. [See, e.g., B. F. Westcott, A General Survey of the History of the Canon of the New Testament (4 1875), 77 f. (attributing the silence to Papias' Judaistic sympathies). It must be remembered, however, that Papias' work only survives in scattered fragments, and the silence is not therefore specially significant.] Yet, even so, it is hard to believe that the Churches of Paul's own founding did not treasure his memory (cf. 1 Clement xlvii. 1 with its 'Take up the epistle of the blessed Paul the Apostle', implying that the Corinthians had treasured one letter at any rate), and that would be enough to start the process of gradual accretion. Communications were good between different centres, and a process of exchanging and transcribing is not difficult to imagine. Moreover, as against Goodspeed's theory, it must be remembered that Acts makes no mention whatsoever of Paul's letter-writing activities, in spite of its mention of other Epistles (see Acts xv. 23-29, xviii. 27, xxiii. 25-30, xxviii. 21). At best, then, Acts can only have provided the incentive in the sense that it recounted Paul's activity as a founder of various local Churches. But the Churches in question did not need Acts to tell them that.

Thus, although one-man theories are exceedingly attractive, they are highly speculative, and the anonymous, gradual evolution is not by any means ruled out. If one other individual name were to be suggested, however, might it not be that of Luke himself? It has just been pointed out that the Acts would provide no reader with any knowledge of the Pauline Epistles; but what if it was after the writing of the Acts, and after Paul's death, that Luke himself – who must have known about the letters although he had not written about them – began to revisit the Pauline centres which he had described, and to look for the letters there? No one knew better than he the fact that they were written. It is entirely in keeping with his historian's temperament to collect them. [H. Chadwick reminds me that Eusebius collected the letters of Origen, H.E. vi. 36. 2.] And the considerable link, in respect of vocabulary, contents, and outlook, between the Pastoral Epistles and Lk.-Acts lends some plausibility to the suggestion that Luke was the collector, editor, and augmenter of the Pauline corpus.[See Excursus II: Luke and the Pastoral Epistles.] Whether the Pastorals were in existence by the time of Marcion or not is still a matter of dispute.

On the nature of the archetype of the Pauline corpus there is a valuable discussion in G. Zuntz's Schweich Lectures for 1946. [G. Zuntz, The Text of the Epistles: a Disquisition upon the Corpus Paulinum (1953).] A careful criticism of Goodspeed's hypothesis leads him to the conclusion (276 f.) that 'Whoever wrote Ephesians, it was not the editor of the corpus. ...' He then points out that 'faithfulness, completeness, and non-interference with the available material' are the characteristics of 'the traditions of editorship in antiquity generally'. As instances, he adduced the editing of Thucydides, Lucretius, and the Aeneid ; of the private diaries of the Emperor Marcus; of Plotinus' essays, with special reference to interesting information about methods in the 'life' of Porphyry, Plotinus' editor. He thinks, therefore, of many faults arising in the Pauline Epistles, individually and (perhaps) in earlier collections, in the fifty years between the autographs and the formation of the corpus – a period during which he believes there was some use and circulation – ergo copying – of the Epistles; and then he thinks of a corpus produced about AD 100, which, in the scholarly Alexandrine tradition (and perhaps at Alexandria) aimed at the qualities already mentioned, and thus (e.g.) produced an Ephesians with a blank in i. 1 and with ἐν Ἐφέσῳ in the margin. Thus there came to be a splendid variorum archetype, from which we may derive many of the later variants.

However and whenever the Pauline corpus emerged, at any rate we have Marcion's date as one landmark; and not many decades later, in about AD 180, we can listen to the martyrs of Scilli or Scillium (in Numidia, in Africa proconsularis ) being interrogated by the proconsul about the books in their possession: 'what is there in your capsa [book box]'? and replying Libri et epistulae Pauli viri iusti ['Viri iusti ' (as H. Chadwick comments) is meant to show that the books in question were not, as the proconsul might have expected, pornographic.] (i.e. probably the Gospels, and the Epistles of Paul ...).'The books, but especially the parchments' (2 Tim. iv. 13 discussed above, pp. 182 f.) has now taken on a new and wholly Christian significance! If we could date 2 Pet. with any certainty, it might prove (iii. 16) to contain the earliest reference to Paul as scripture. But who can say when it was written?

What, now, may be said of a Johannine corpus ? Extremely little, it must be confessed. One clear fact, however, is that there is a close connexion between the Fourth Gospel and the Johannine Epistles (whether or not they are actually by the same hand). [See P. Katz, 'The Johannine Epistles in the Muratorian Canon", J.T.S., n.s., viii. 2 (Oct. 1957), 273 ff. for an ingenious bid for the inclusion of 2, 3 John in the Muratorian Canon, by suggesting that duos (leg. duae ) in catholica habentur represents an original Greek such as δύο σύν καθολικῇ, two in addition to the Catholic (Epistle ). This seems to me to be in principle convincing; but I do not see why the original should not have had πρὸς καθολικήν – even more natural Greek for in addition to, and more easily translated in.] Both these groups, though anonymous, are associated by tradition with the name of John and with Ephesus. The Apocalypse, itself claiming to be written by one named John and also associated by tradition with Ephesus, is still held by some modern scholars to be by the same hand as some or all of the other Johannines: but despite certain contacts in vocabulary and thought, its style, and, still more, its theological outlook, are very different. However that may be, it is perhaps a more useful thing to think in terms not so much of a Johannine corpus as of an Ephesian tradition, and to speculate cautiously about the courses along which diverse streams of tradition flowed to that centre. The late T. W. Manson revived, hi a modified form, [See T. W. Manson,' The Life of Jesus: a survey of the available material, (5) The Fourth Gospel', B.J.R.L. 30 (1946-47), 312 ff., with W. Sanday, The Criticism of the Fourth Gospel (1903), 199.] Sanday's interesting suggestion that there was 'an anticipatory stage of Johannean teaching, localized somewhere in Syria, before the apostle reached his final home in Ephesus'. Manson proposed Antioch as a centre of 'Johannine' traditions en route to Ephesus. Manson did not discuss the relation of the Epistle to the Ephesians to the Johannine traditions; but it is well known that, whereas this Epistle (as has been already observed) does not fit beyond doubt into the typically Pauline mould, it does present certain affinities with Revelation (e.g., the Church as the Bride of Christ, and as founded upon the apostles and prophets). There is not here, as a matter of fact, any necessary contradiction to the Pauline phrases about Christ as the one foundation; but may not Ephesians represent the flowing together and fusing of Pauline and other types of thought, and may not Ephesus (possibly with Antioch as a kind of halfway basin or reservoir) be the centre of confluence – especially if much of the Johannine tradition really does go right back to Palestinian sources? [See p. 93, n. 3. ] Possibly it is worth while to throw into the discussion the fact that Matthew is the only Gospel that speaks of the Church and its (apostolic) foundation, and that there is something to be said for an Antiochene connexion for this Gospel. [For discussions of Mart's, provenance, see, e.g., T. W. Manson, 'The Life of Jesus: a Survey of the Available Material: (4) the Gospel according to St Matthew', B.J.R.L. 29 (1945-46), 392 ff.; G. D. Kilpatrick, The Origins of the Gospel according to St Matthew (1946); K. Stendahl, The School of St Matthew (1954); P. Nepper-Christensen, Das Matthäusevangelium: einjuden-christliches Evangelium? (1958); W. Trilling, Das wahre Israel (1959); E. P. Blair, Jesus in the Gospel of Matthew (1960). ] The letters of Ignatius of Antioch seem to contain nothing that can demonstrate any knowledge of the Fourth Gospel or of Ephesians. But (as H. Chadwick observes) it is not Ignatius' manner to cite, but he may allude. In these conjectures, however, we are on a quaking bog of uncertainty. Notions of ecclesiastical authority might seem to provide a more fruitful clue to classification: Matthew, Ephesians, the Fourth Gospel, and the Johannine Epistles are all very vigorously concerned with authority; but again it is impossible to group them together in an undifferentiated way, for, in a sense, the Fourth Gospel is remarkably non-ecclesiastical in its treatment of authority (as indeed throughout). Non-apocalyptic eschatology is, to some extent, common to Ephesians and the Fourth Gospel; but this does not apply to the Johannine Epistles – still less to the Apocalypse; and there is non-apocalyptic eschatology in Rom. viii also.

In short, we know pitifully little about the cross currents of Christian teaching and apologetic at Ephesus or elsewhere that led to the ultimate recognition of those Johannine writings by the whole Church. What we do know is that St John's Gospel was the latest of the four to gain this status and that it was opposed by a group contemptuously known for this reason as the 'Alogi' (the 'logos-less' in relation to the Fourth Gospel prologue and therefore 'logos-less' also in the sense 'stupid'); likewise that the Apocalypse was roughly handled by those who were opposed to millenarianism, of which it alone, within the New Testament, is the spokesman. [For anti-millenarianism, see Euseb. H.E. iii. 28; vii. 25; and (e.g.) I. T. Beckwith, The Apocalypse of John (1919), 340 ff. The 'Alogi' opposed the Johannine writings because they seemed to offer a handle to the charismatic anarchy of the Montanists.] Conversely, it is a familiar fact that certain other writings associated with the name of John never gained wide recognition at all; while writings widely read at one time for edification, such as the Apostolic Fathers, were ultimately excluded. The necessity for decision on authoritative books as against the false or the unauthoritative, was imposed by heresy from within and attack from without. Perhaps a fruitful line of advance may lie (cf. Chapters I and VIII above) in the direction of freshly investigating the purposes of some of the New Testament writings. 2 Pet. and Jude are manifestly attacks on perversions of Christianity, perversions which had arisen, at least in part, through the kind of misappropriation of Pauline doctrines which Paul himself attacks (e.g. in Rom. vi). The Pastoral Epistles are also concerned in part with correcting perversions of Pauline teaching, and it is notoriously possible that James may (though it is far from certain that Paul is presupposed) have something of the sort in view in Chapter ii. The Johannine Epistles, attacking a docetic type of misinterpretation of Christ, may similarly be viewed as a corrective to perversions of the teaching in the Fourth Gospel. And in addition to insidious dangers from within, necessitating a clear recognition of what was sound in Christian writing, there were attacks of opponents from without. In controversy with these, authoritative references were needed; while in times of persecution, it might be of vital importance (as we have seen at Scilli) to define the sacred Christian manuscripts. The story of the fluctuations on the fringe of the canon – on one side the ultimately excluded, on the other the doubtfuls ultimately included (2 Pet. and Jude, 2, 3 John, Hebrews, etc.) – is told in all books of introduction, and need not be repeated here, any more than the final settling down of the present canon, in Athanasius' 39th Paschal letter of AD 367 and (?) the Council of Laodicea. The purpose of this chapter has been not to retell the whole story but rather to throw into relief – so far as is possible in a realm too remote to focus clearly – some of the motives and principles, theological and disciplinary, behind the long process. All the time, moreover, it is viewed, so far as possible, as a human process. That it is also the tale of a divine overruling of the gropings and mistakes of men is here assumed without further discussion.