REDATING THE NEW TESTAMENT. by J. A.T. Robinson SCM Press Ltd London.  First published 1976 by SCM Press Ltd 58 Bloomsbury Street, London Second impression 1977 © J. A. T. Robinson 1976. - Prepared for katapi by Paul Ingram 2006.

XI. Conclusions and Corollaries

Home | Contents | Dating: fixed points | External evidence... ancient testimony | 1st attestation | Quotation | Prophesy | dating  estimates | Internal evidence... recognition | elimination | development | assumptions | Conclusions | Dating: Table 1 | Table 2 | Dynamic Table | >

BEFORE summarizing our own conclusions and looking to the consequences from them, it will be useful to set down certain general observations that have emerged from the survey of the evidence and of other positions built upon it.

1. We may start with the fact, which I confess I did not appreciate before beginning the investigation, of how little evidence there is for the dating of any of the New Testament writings. Moreover, there are no fresh facts - like the introduction of carbon-14 datings into archaeology - which have clearly changed the picture or which have caused me to reopen the question. It is surprising to be made to realize that there is only one reasonably secure absolute date (and that within a year or so either way) in the life of St Paul, which in turn can be used to fix the chronology of his writings. And this - that of the pro-consulship of Gallic in Achaia - relates not to any statement of Paul himself but to a minor incident recorded of him in Acts. There are other events, such as the famine under Claudius, or the deportation of Jews from Rome, or the arrival of Festus in Judaea, or the alleged execution of Paul under Nero, which can provide very approximate supports. But the evidence for their dating is extraordinarily elusive, and none again turns upon anything that Paul himself wrote. The chronology of his life and letters has to be pieced together from a large number of statements and inferences - though the material for relative dating, both in the epistles and Acts, is far richer than for any other part of the New Testament literature. Yet at the end we have to confess that we cannot settle with any precision or finality the date of his birth, his conversion, his visits to Jerusalem, his various missionary journeys, his arrival in Rome, his death - or any of his letters. And if we know so little about Paul, how much less can we say about Peter or John? There is not a single book of the New Testament that dates itself from the internal evidence. And important recent discoveries - e.g. of new papyrus fragments or the Dead Sea scrolls or the gnostic library at Chenoboskion - have done little more than tilt the balance against guesses which rested in any case upon very questionable judgments. In the case of the fourth gospel, which they chiefly affect, they do not of themselves require any change in the estimates made, for example, by Lightfoot and Westcott a hundred years ago.

The conclusion to be drawn from this first point is not that there is nothing, or nothing new, to be said. It is that the consensus of the textbooks, which inform the student within fairly agreed limits when any given book of the New Testament was written, rests upon much slighter foundations than he probably supposes.

2. When we turn to the external evidence in the testimony of the early church the situation is not very different. Compared with the plethora of ancient tradition, good, bad and indifferent, with regard to authorship, it is surprising to discover, as we have seen, that only one book of the New Testament, the Apocalypse, is dated in early Christian writings. Irenaeus sets it 'towards the end of the reign of Domitian', a statement which, if one combines it (as Irenaeus does not) with what evidence there is for a Domitianic persecution, puts it at about 95. Yet we have found no more reason to accept this statement than most scholars have found to accept the other two with which Irenaeus associates it, namely, that the book of Revelation was composed by John the apostle and by the same man who wrote the fourth gospel. For the rest, the traditions (to take a selection) that the gospel of John was written when the apostle was a very old man, or that Mark was written during the lifetime - or after the death - of Peter, or that Matthew was written 'while Peter and Paul were preaching and founding the church in Rome', have been shown to be worthless, self-contradictory or ambiguous. Moreover, such state-ments as that John wrote his gospel 'last of all' or 'after the Apocalypse' or that the Pastoral Epistles come from a period following Paul's first Roman imprisonment or that the first epistle of Clement was composed during his episcopate turn out to be little more than guesses. The conclusion must be that, as with authorship, the external evidence is only as good as the internal, and cannot prevail over it. Indeed in contrast with the evidence for authorship, which sometimes, I believe, has to be taken seriously (e.g. on Mark, Luke-Acts, the gospel and epistles of John, and, in one instance, Hebrews), the external testimony on dating, with the single exception of the Apocalypse (where it is significant though far from unanimous), is virtually worthless.

3. Closely connected with the last is the evidence of first attestation by name to the existence of a New Testament book in the early church. The first thing that needs to be said is that one is dealing here almost totally with an argument from silence. The one exception that can be dated within the first century is the explicit reference in I Clement to I Corinthians (though described simply as Paul's 'epistle' to that church). This does nothing to help with the dating of I Corinthians, which if it is genuine (as no one now doubts) must in any case have been written years earlier. When later the Apocalypse is mentioned by Justin or the gospels by Papias or Irenaeus, we have, for what it is worth in any given instance, a certain ceiling for dating purposes. But the absence of such mention is a very different matter. The small quantity of early Christian literature and its occasional character make the argument from silence, that such and such a book of the New Testament was not in existence or was not known, precarious in the extreme. We should not even guess from Acts that Paul wrote any letters, but at whatever date we put Acts - from the early 60s to the mid-second century - it would be highly hazardous to conclude that its author did not know of them - let alone that they did not exist. The argument from attestation, whatever its weight in regard to authorship, is relevant for dating only if there is ground for supposing that the book in question was written so late that its first mention provides a terminus a quo and not merely a terminus ad quern. This could be so with the Shepherd of Hermas, if we can trust the Muratorian Canon, and it has in the past been supposed to be so with regard to the fourth gospel, first cited by name only c. 180. But the more reason there is for pushing the date of a book back, the less relevant becomes the argument from its earliest attestation. That the Apocalypse is first mentioned by Justin in 150 does nothing to help us decide whether it was written in the late 60s or the mid-90s. And the same applies to the dating, say, of the gospels and epistles of John, or even of Jude and II Peter. The gap is usually in any case so great - and the bridge so thin - that an extra thirty years or so can make little difference.

4. More relevant, and much more difficult to decide, is the question of quotation for establishing literary dependence and therefore temporal posterity. This applies both to literary dependence within the New Testament itself and to its subsequent citation. Though practically no one would question the fact of literary interrelationship between the synoptists, it is less clear than it was fifty years ago that the first three gospels can be set in a simple chronological series or that we know what the order of the sequence is. Equally it is much less evident than it once seemed that John is dependent upon, and for that reason later than, the synoptists. Confident assertions too that James quotes Paul or Matthew, or that there is a direct literary connection, whichever way, between Ephesians, I Peter and Hebrews, or that the author of Revelation knows and uses most of the other books of the New Testament, are more muted than they were. The work of the past two generations has made us far more conscious of the common tradition both of preaching and teaching within the apostolic communities and sensitive to the processes of oral transmission. Direct citation from previous documents known to be available requires to be argued with much greater precision than earlier scholars who spotted similar phraseology assumed, particularly in the first part of this century - the heyday of source criticism. And this applies equally to the sub-apostolic period. The readiness to assert specific quotation from canonical books of the New Testament in the Apostolic Fathers has been much chastened and modified. Particularly in relation to the gospel traditions, it is becoming clearer that other channels of trans-mission, both oral and written, continued to function well after the latest probable date of any of our gospels. Studies of the quotations in I Clement and even in the Didache far from compel the conclusion that their authors used our synoptic gospels, while the discoveries of the Egerton papyrus of an unknown gospel, or of the Gospel of Thomas, or even of the 'Secret Gospel' of Mark (with its witness to a non-Johannine (?) version of the raising of Lazarus), testify to parallel traditions in written form that may go back earlier and certainly go on later than our canonical gospels.

All this suggests a much greater rigour and reserve in the use of the argument from quotation as an indicator of dating. This applies both positively and negatively. We cannot say positively that on these grounds James must be later than Romans, or Ephesians than I Peter (or vice versa), or the Apocalypse than Luke, or the Didache than Matthew. Nor can we say, negatively, that the gospel of John was unknown to Ignatius or Justin because they do not specifically quote it. Since both of these arguments have, paradoxically, been used in the interests of late dating, it is relevant to remind ourselves how precarious is their foundation. Indeed I would think it safe to say that there is no certain argument for dating to be drawn from the use of any one New Testament book by any other - and this applies even where there is undoubted literary interrelationship, e.g., between the synoptic gospels or between Jude and II Peter. Moreover, with the exception of the clear allusions to Hebrews and I Corinthians in I Clement (36,47,49), I doubt whether any of the references in the four sub-apostolic writings which we have ventured to set in the first century can unquestionably be said to show dependence on any of our canonical New Testament books or on each other. This does not of course prove that the apostolic or later writers wrote in mutual ignorance or isolation (which is highly improbable), nor is it in itself any argument for early datings. It is merely a salutary warning against misplaced dogmatism based on arguments from literary dependence.

5. A similar chastening would seem to be appropriate in the assurance with which scholars have pronounced on prophecy after the event. That such activity was a stock-in-trade, especially of apocalyptists, cannot be doubted. Daniel's 'prophecy' in ch.7 of the four beasts, to depict the progress of the Babylonian, Median, Persian and Greek empires, is an obvious paradigm. And the later apocalypses of Baruch and Ezra (ostensibly set in the period after the capture of Jerusalem by Nebuchadnezzar) describe the capture of Jerusalem by Vespasian and Titus, while the Sibylline Oracles use the device of the Sibyl to 'predict' the detailed pattern of world history to date (and where they go beyond that they start getting it wrong). But the very detail of these, which could have deceived or been intended to deceive no one, must make us pause before assuming that every prophecy in the gospels and elsewhere, and particularly of the fall of Jerusalem and the destruction of the temple, must come into this category. On the contrary, we have seen reason to question very closely the prevail-ing assumption that this is so in regard to the gospels of Matthew and Luke (the nearly parallel statements in the Markan apocalypse have not, rather surprisingly, led most scholars to the same conclusion). At any rate prophecy ex eventu has to be demonstrated, and demonstrated by minute and strict criteria, rather than simply assumed.

This is not of course to say that subsequent reflection on events or the later experiences of the church have not shaped or conditioned the gospel tradition as we have it. John's looking back on the manner of Jesus' death, or of Peter's, obviously presupposes the former and in all probability the latter. Equally the predictions of the rejection, crucifixion and resurrection of the Son of Man in the synoptic tradition are clearly influenced in lesser or greater degree by the knowledge of what happened. Again, the synoptic apocalypses and the Johannine last discourses have evidently placed onJesus's lips warnings to the church that have been conditioned by the church's own sufferings. Indeed there is not a saying or a story in the gospel tradition that has not reached us through the sieve of the community's needs and uses. Yet it is quite another matter to say that these sayings or stories have simply been created by the history of the church and then put back into the mouth or the life of Jesus, or to say that Jesus could not have foretold what would befall his followers or his nation. Moreover, in Christian apocalyptic, whether set on the lips of Jesus or of John, there is no hint of the convention of pre-casting predictions so as to make it appear that occurrences within the readers' time were fore-known from the distant past. While the Christian prophet might indeed shape his oracles, as John evidently did, out of his own experiences, the very limits of those experiences indicate where events had not yet reached. Thus, there is nothing in Revelation that speaks of the fall of Jerusalem or that certainly reflects anything beyond the late 60s, just as there is nothing in the predictions of Paul in Acts that certainly reflects the situation beyond the point at which its story ends – or the subsequent organization of the Roman Empire or of the Christian church. Whether the gospels, Acts, or the Apocalypse were written after the fall of Jerusalem must be assessed on the merits of each case. The argument from prophecy, like the argument from quotation, must take its place within the larger context and must in each instance be deployed with the most exacting critical discrimination.

6. It is sobering too to discover how little basis there is for many of the dates confidently assigned by modern experts to the New Testament documents. The argument advanced by so great an authority as Harnack that when Matthew says that the coming of the Son of Man would occur 'immediately' upon the tribulation in Judaea this means that his gospel could not have been written more than five years after 70 is, to say the least, a disconcertingly tenuous deduction. We have observed also how not only Harnack and Lightfoot but the vast majority of scholars take over the assumption that the Neronian persecution (and therefore, if it is apostolic, I Peter) is to be dated in the year 64 - when the sole piece of evidence for its association with the fire of Rome (in the Annals of Tacitus) clearly points to its being at the earliest in the spring of 65. We have noted too how incredibly limited is the evidence (depending on the passing reference in its opening sentence to our recent local difficulties) on which I Clement has almost universally been assigned to the year 95-6. Similarly the Domitianic persecution of the church which has been the basis for dating so much (either because the writing in question is held to reflect it or because it must ante-date or post-date it) has itself turned out to be virtually a non-event. In the same way, the dating of the gospel of Mark, when it has not merely been held to reflect the progress of the Jewish war, whether finished or unfinished, has been made to turn on the report that it was written after the death of Peter - or even on the a priori assumption that the death of Peter would (only then?) have led to the need for a written record - whereas the tradition that it was written during the lifetime of Peter is just as strong, if not stronger. Finally, as examples of how much has been built on so little - yet constantly reiterated by commentators till their weaknesses were exposed - we may mention the alleged use of Josephus by Luke to put Acts after 93, or the mention of the founding of the church of Smyrna in Ep.Polyc.11.3 to date Revelation late, or the reference in John 5.43 to 'another coming in his own name' to the revolt of Bar-Cochba in 132. Dare we think that the allusion to the issuing of the twelfth of the Eighteen Benedictions in 85-90 still found in scholar after scholar to explain the language of expulsion from the synagogue in the same gospel will turn out to be a similar curiosity of criticism?

7. Then there is the apparently almost wilful blindness of investigators to the seemingly obvious. Thus Harnack chided scholars - and himself - for failing to take seriously the explanation which stared them in the face of why Acts ends where it does. They could not believe it, because of previous assumptions about the dating of Mark and therefore of Luke. In the same way, the absence of all reference in the gospels, Acts, epistles or Apocalypse to the fall of Jerusalem in the past tense is a phenomenon that should have raised far more questions than it has. Even - or especially - a deliberate rewriting of his-tory to conceal it must surely have left more trace than it has. It is not enough to say, however loudly, that its 'importance ... is impossible to exaggerate', and that 'much of the subsequent literature both of Judaism and Christianity took the form it did precisely in an attempt to come to terms with the catastrophe of ad 70', [Perrin, NTI, 40.] and then to give no specific evidence. The assumption that the epistle to the Hebrews could have been written as it is without reference to this cataclysmic event (so specifically mentioned in the comparable Epistle of Barnabas) is surely quite astonishing. So too is the blindness to the obvious interpretation of the statement in Rev.17.10 that it was written during the reign of the sixth Roman emperor. While certainly not ruling out more complicated explanations or resort to purely symbolic interpretation, the consistent evasion by modern commentators of a solution they have already prejudged to be impossible contrasts strikingly with the openness of an earlier age (of whatever school of thought). And of the Apocalypse in general the marked contrast with the post-70 Jewish apocalypses, not least in their prediction of the doom of 'Babylon' as direct revenge for its sack of Jerusalem, is passed over extraordinarily lightly. Another blind spot, on which we have already touched, is the way in which commentators have blandly assumed that Matthew is deliberately writing for the interval between the fall of Jerusalem and the parousia when he alone inserts the statement that the latter will subvene 'immediately' upon the former. Finally I believe one must insist that most liberal scholars have allowed themselves to be insensitized, whether by the climate of critical opinion or for other reasons, to the very considerable strength of the external - and internal - evidence for the apostolic authorship of the fourth gospel, and that all students of it (including myself) have until very recently too hastily assumed that the relations it depicts between Judaism and Christianity can only reflect a situation after 70.

8. We have drawn attention to the way in which so much dating of New Testament books has been determined more by a process of elimination than by positive indications. This is particularly true again of St John's gospel. There is really not a single positive reason for associating it with the years 90-100. Indeed the reign of Domitian has, as we pointed out, become the repository, for one investigator or another, of every book of the New Testament (and several outside it) with the exception of the undisputed epistles of Paul: even the gospel of Mark has recently been put there by Trocme and II Peter by Reicke. But there cannot be that much which this period has in common - except our convenient ignorance of it. For here we are in an age with few firm landmarks and with correspondingly few objections to almost any dating. Writing of the book of Revelation in relation to the gospels of Matthew, Luke and John, Austin Farrer has said with characteristic charm:

The datings of all these books are like a line of tipsy revellers walking home arm-in-arm; each is kept in position by the others and none is firmly grounded. The whole series can lurch five years this way or that, and still not collide with a solid obstacle.
[A. M. Farrer, The Revelation of St John the Divine, Oxford 1964, 37.]

And to show how fluid the situation is he himself proceeds to ignore the only remotely fixed point (the association of the Apocalypse with the last years of Domitian) and to set it in the reign of Trajan! The received schema of New Testament chronology scarcely convinces by its own internal logic. As Farrer says, each book is held in place by the other, yet negatively rather than positively. They coexist rather than cohere. There is no compelling reason why Matthew or John, James or Jude, should belong where they do. Perhaps it is wrong to ask that there should be. But the accepted pattern is the outcome of a protracted period of shunting and jostling until the pieces have settled in a certain contiguity. It is remarkable how little it itself tells us about the course of church history - unless with a Baur or a Perrin we superimpose an a priori pattern upon it and force the pieces to fit.

9. Another factor which we have observed is the subjectivity in assessing the intervals required for development, distribution or diffusion. There is a close parallel here, as we saw, with what has been going on in archaeology. What Renfrew calls the 'archaeological bellows' can be moved in or out at will. And there is a kind of Parkinson's law that takes over: the intervals will contract or expand to fill the time available. Thus, we have watched Ogg stretching the intervals in Paul's activities (even by having him off sick for a year), with Barrett compressing them equally violently - all to fit the space they regard as available between two points. While indeed certain points may be determinable (within whatever margin of error), the intervals tend to be indeterminable - and can therefore be treated freely. One is reminded of the dumpy lady who concluded that she had the same vital statistics as Marilyn Monroe - apart from the spaces in between: and these no one thinks to measure or to mention.

Estimates of how long it takes, say, for doctrines to develop or structures to become institutionalized, for a golden age to decline into a silver, or for documents to travel, vary wildly. Some can squeeze the whole of church history up to the conversion of Paul into a single year, others require decades for the emergence of the conditions in I Peter. And nothing is so slippery as the relation between Christology and chronology. The greater the stress on the part of the early Christian communities in the formation of the tradition, the longer the 'tunnel period' tends to be. Yet the procedures to which the form critics and redaction critics draw attention carry in themselves no implications for dating. The whole process could be speeded up or slowed down, as in a film, without being essentially affected. The assumption that it was slow or steady is purely arbitrary. In fact it is inherently probable that, as in the creative ferment of any new movement, it was swift and took place in spurts - with periods of retrenchment and consolidation in between. The only reliable canon or measure of development is the Pauline corpus. For there within a testable period of less than a decade we can see something of what could go on in one creative mind and within a shorter space still what could change in the condition of a single congregation at Corinth. Moreover, if we accept the basic reliability of the Acts outline - which for Corinth certainly holds - there is a great deal more to go on. For if all the developments described in Acts - in theology, organization and ethos - could come into being between 30 and 62, it is difficult, on developmental grounds alone, to demand more time for the formation of the gospel tradition. Moreover, if the Didache is really also evidence for the stage that discipline and liturgy had reached in this period, practically nothing is foreclosed.

10. Closely connected with the supposed requirements of development is the manifold tyranny of unexamined assumptions. Even (perhaps most of all) in their reactions against each other, different schools of critics take these over from their predecessors, and of course individual commentators and writers of introductions take them over from each other. Fashions and critical orthodoxies are established which it becomes as hard to go against in this field as in any other. A notable instance of this was the swing, about the turn of the century, among New Testament (but not classical) scholars in the dating of the Apocalypse; but the fashions on the fourth gospel - or the epistle of James - have not been far behind. Again, it seems to have been 'respectable' in critical circles till recently to put Hebrews late - with few appearing to question how improbable this is. Solutions of the synoptic problem (including the relation of John to the synoptists) have tended to become accepted for extended stretches as assured -and therefore reassuring - results. Some of this is sheer scholarly laziness. We all respond to the urge quieta non movere; and I confess to a long reluctance to reopening the synoptic problem for what it might force one to reconsider. There is also an understandable temptation to depreciate or lose patience with the lower reaches of 'mere' introductory questions of date and authorship. Those who press on to the more constructive work of building theologies of the New Testament tend to be content to assume and incorporate the foundations laid by others. It is noticeable as one visits the literature of the past hundred years how much more thoroughly grounded in these questions was the work of the older generation - most of whom were brought up on the classics - and how much more rigorous in the dating of evidence, as well as attentive to the evidence of dating, than some of their successors. This, one has to say it, has been true of many, though not all (R. H. Lightfoot was an honourable exception), of the form critics and redaction critics. Their world has been a world without fences, where words and ideas, myths and movements, Hermetic, gnostic, Mandean and the rest, have floated freely with no very noticeable tethering to time or place. Many of the circles and communities of the early church with their tensions and tendencies are frankly creations of the critics or highly subjective reconstructions.

Yet this has not prevented the fixing and indeed the freezing of a number of powerful assumptions. We may instance three.

One has been that the period of oral tradition preceded, and was in turn succeeded by, the period of written tradition. In a broad sense this is obviously true. Where it becomes dangerous is when it hardens into two presumptions, (a) The first is that the writing down of traditions did not begin until after a considerable stretch of oral transmission - the transition being marked, it is also often assumed, by the passing of the first apostolic generation or by the fading of the hope of an early parousia. Ellis in an important and refreshing article on 'New Directions in Form Criticism' [E. E. Ellis in G. Strecker (ed.), Jesus Christus in Historic und Theologie: Neutestamentliche Festschrift fur Hans Conzelmann zum 60. Geburtstag, Tubingen 1975, 299-315 (304,306,309).]
observes that the supposition that

writing would begin only when the expectation of an imminent end of the age subsided foundered with the discovery of the Dead Sea Scrolls: the Qumran sect viewed itself to be in the 'last generation' (1 Q, pHab.2.7; 7.2), expected an imminent end, but, nevertheless, produced a large body of literature.

Moreover he makes the obvious point, when one comes to think about it, that

the circumstance that gave rise to written teachings in early Christianity was not chronological distance but geographical distance. This is evident in the case of Paul's letters and of the Jerusalem Decree (Acts 15), but a similar situation on a smaller scale was also present in the mission of Jesus.

Indeed he argues for 'a considerable degree of probability for some written transmission of Gospel traditions from the time of Jesus' earthly ministry'. (b) The second presumption is that once the period of writing did begin the traditions were transmitted, and mutually influenced, almost exclusively by the processes of literary dependence, as one writer 'used', 'copied' or 'altered' another. On the contrary, there is every reason to think that both oral and literary processes went on concurrently for most of the first hundred years of the Christian church. The writing was earlier and the reign of the 'living voice' longer than we have tended to suppose. If the preaching material of Peter or the eucharistic prayer of the Didache came at a certain point to be written down, that per se is to say nothing about dating.

A second assumption has been that Aramaic-speaking Christianity was prior to Hellenistic Christianity. Again in a general sense this is true. But I believe it is misleading, if the deduction is then drawn that Greek was not spoken in Palestine from the very earliest days of the church and indeed that the spiritual majority was not in the first instance made up of those who also (or most naturally) spoke Greek, whether they were from Galilee, Jerusalem or the Diaspora. The mere fact that all the surviving Christian literature is in Greek while all the surviving Qumran literature is in Hebrew (or to a small extent in Aramaic) must say something about the relative provenance of the two movements. The testimony of the New Testament itself, not to mention the growing weight of contemporary evidence from outside it, suggests that the assumption that Hellenistic Christianity, with the use of the Septuagint, was a secondary phenomenon confined to the Gentile churches is far too facile. Certainly the assumption that Peter would have needed Greek only in addressing Gentiles, or James would not have been able to write it at all, or that the Johannine tradition must have passed through the medium of translation, demands challenge and scrutiny. This does not in itself affect the question of dating, but, unless questioned, the assumption tends to work uncritically in the direction of identifying what is Hellenistic with what is late. There is nothing inherently impossible about the notion that both the epistle of James and the first draft of the gospel of John could be very Jewish
and very
be written in Greek.

The third assumption I would mention primarily concerns authorship, though it regularly recurs in conjunction with dating. This is that there was an indefinite number of totally unrecorded and unremembered figures in the history of early Christianity who have left absolutely no mark except as the supposed authors of much of its greatest literature. This creates relatively minor problems when the writings in question are either anonymous or of secondary significance. Thus, who wrote the Epistle of Barnabas or II Clement or even the epistle to the Hebrews can ultimately be left unanswered without the overall picture being affected - though it is noticeable that all attempts to answer these questions in the early church turned upon names, like those of Paul or Luke, Barnabas or Clement, that we have already heard of: no one thought to postulate ex nihilo some forgotten spiritual genius. (Even the ghostly figure of John the Elder is not recorded in the tradition as the author of anything. Eusebius merely guessed that he might have written the Apocalypse: Dionysius was more judicious.) It is when we are dealing with the conjuring out of thin air of major theologians or spiritual giants, like the authors of the fourth gospel and the epistle to the Ephesians, who not only died as if they had never been but also claimed to be the apostles who over-shadowed them, that credibility begins to be stretched. Pseudonymity is invoked as if it were an accepted and acceptable way of life at a date and to an extent for which we simply have no evidence. Indeed the fact of pseudonymity is frequently just assumed: it is the explanation for it alone that is argued. I Peter, II Peter, Jude, James, the Pastoral Epistles and Ephesians, not to mention Colossians and II Thessalonians are freely, and for very diverse reasons, attributed to men who were well known to be dead - without this practice being remarked upon or noticed except negatively in any early Christian writers. The most that is suggested by way of acceptable literary explanation is that two Johns may have been confused, not that one was purporting to be the other.

None of this is to say that the question of authorship is to be settled in a simplistic or fundamentalistic way. For we know that there were from the middle of the second century pseudo-apostolic writings both among the heretics and within the main stream, which when exposed were rejected. The Christian fathers were well familiar with the categories of genuine, disputed and spurious literature. But there has been insufficient discrimination in modern scholarship between hypotheses of secretaries (like Tertius for Romans), ghost-writers (such as, supposedly, Silvanus for I Peter), agents (such as possibly for the Pastorals or II Peter) and impostors (on the usual explanation of II Peter). In particular, it is important from the point of view of dating to distinguish between theories which presuppose that the document is being written with the authority and within the lifetime of the apostle in question and those which presuppose a purely fictional and subsequent setting. The peopling of the sub-apostolic era with a penumbra of pseudo-Pauls, pseudo-Peters, pseudo-Johns (and even pseudo-Judes!) on no evidence which is not drawn out of the documents themselves is really an astonishing confidence trick. Surely a Clement or an Ignatius, a Justin or an Irenaeus, might have expected to refer to them either positively or negatively. But we hear nothing of them. And it makes one gasp the more that these same critics are entirely prepared to use the absence of external attestation to question the genuineness or even the existence of the same documents.

The observations made so far may have appeared excessively negative and critical. Before passing to my own conclusions I should like to correct that impression. No one working in this field - least of all one like myself who came to it so ignorant - can fail to be conscious of the overwhelming debt one owes to those who have worked in it before and to the introduction to so many valuable studies one has received. There is nothing that I have put into the pool that I have not first taken out. And delving back into the history of the work has revealed labours and labourers all too easily forgotten or ignored by the modern surveyor of the scene. In particular, there was a generation of scholars astride the turn of this present century whose erudition was matched by a judgment which must still command immense respect even when one dissents from their conclusions. What for me has come from this work has been the rediscovery - or in some cases the discovery - of men like Lightfoot and Hort, Harnack and Zahn, C. H. Turner and Armitage Robinson, Mayor and Chase, Edmundson and Peake, to name but a few, and those chiefly from my own heritage. [Since the others have left their memorials behind them, the reader's interest may by now have been aroused, as has mine, to enquire what if anything is known about George Edmundson. In spite of holding no high academic or professional office he was sufficiently distinguished to rate a substantial entry in Who's Who. The son of a Yorkshire parson and lord of the manor, he was born in 1848. After taking a double first in mathematics at Magdalen College, Oxford, as well as a university scholarship in mathematics and prize in the Greek Testament, he was ordained to the title of a fellowship at Brasenose, of which College he was mathematical lecturer, tutor and junior bursar until 1880. Then at the age of 32 he left Oxford, to enter upon two incumbencies covering forty years, first at Northolt and then at St Saviour's, Upper Chelsea, which were to last until his retirement to the south of France, where he lived on till 1930. He received no visible recognition from the church, not even an honorary canonry, though at the very end of his career he reached the heights of becoming war-time Rural Dean of Chelsea. In the world however it was a rather different story. Apart from an Oxford D.Litt., he was a Fellow of the Royal Historical and of the Royal Geographical Societies, as well as being an honorary member of two Dutch learned bodies. He was twice employed by the Government in boundary-arbitrations in British Guiana and by the historical department of the Foreign Office. His literary productions were equally varied and included books on Milton, Archbishop Laud, Anglo-Dutch rivalry in the seventeenth century (the Ford lectures at Oxford), the chapter on Spain and Portugal in the eighteenth century in the Cambridge Modern History, a history of Holland, and the editing of the manuscript of an early Spanish journal from the Amazon. None of this or his mathematical training would prepare one for his Bamptons, at the age of 65, on The Church in Rome in the First Century, which disclose an extensive and intensive knowledge of classical and patristic learning, as of modern German, French and especially Italian literature and excavation - prompting one of its rare reviewers (TLS, 19 February 1914, 86) to the back-handed compliment: 'The learning and erudition shown on every page is quite unusual in an English work.' Yet it received no discussion in any classical, historical or theological journal that I have been able to discover. Shall we again see the like of such a clerical career?] I say this in no way to depreciate contemporaries - for who could be more richly endowed in their scholarly equipment than a Bultmann or a Dodd, a Jeremias or a Kasemann, a Kummel or a Hengel? - but because one cannot help observing how the caravan of New Testament scholarship has tended to move on to fresh sites, attracted by different and exciting questions, but leaving behind the other workings in the state in which they then were and taking over their results to date, without further re-examination, as the presuppositions of the new work. This I suggested happened, notably in regard to the synoptic problem, both when the form critics took over from the source critics and when the redaction critics took over from the form critics. These newer disciplines have had much illumination to shed. But they become dangerously unrelated to reality when they ignore detailed investigation of the 'introductory' questions - and still more when they claim that these can be settled by the answers to their own questions. Thus, it is frankly nonsense to say, as I have recently heard claimed, that the priority of the gospel of Mark can be demonstrated by its theology alone. In the study of chronology the argument from development may be a useful servant but it is a bad master. Used with great discrimination and delimitation - e.g. within the Pauline corpus - it can sometimes help towards establishing the sequence of the New Testament documents (though 'the argument from order', while it looks the most objective in synoptic studies, can be made to prove almost anything). What it cannot do, except within the very broadest limits, is itself to determine the span of time over which the development takes place. There is no more reason to suppose that the history of the synoptic tradition (or the Johannine) as delineated by the form critics requires a canvas of sixty years rather than thirty, or that the questions and answers of the redaction critics are not just as valid if the entire process took place before the fall of Jerusalem rather than after it.
These issues can only be decided by the historical evidence; and of course the signs of doctrinal or organizational development, in so far as we can pin them down to time or place, are one important part of this. Naturally, too, this evidence can never be investigated 'without presuppositions'. But the presuppositions must continually be re-examined and tested. The chronology of the New Testament documents, however fundamental to other studies, has tended to lie too long at the bottom of the agenda - in fact for the three-quarters of a century since Harnack. Each new student enters a field already marked out for him by date-lines which modesty as well as sloth prompts him to accept, and having accepted to preserve. The mere fact that 'New Testament introduction' tends to occupy his earliest and most inexperienced years has a formative effect, for good or for ill, on all his subsequent work.

But having questioned the basis of much that passes for assured results in this field, it behoves us now to state in summary form what we would put in its place. The arguments for the conclusions reached on each particular book of the New Testament have already been set out and it would be tedious to condense them here. All that is necessary is to bring together the results and make some comment on the inferences to be drawn, and not to be drawn, from them.

The dating of any document must stand or fall primarily on its own merits, though inevitably the strength of some datings will depend – on any chronology – upon how they cohere with others. Yet, if a coherent or more coherent pattern emerges, its truth cannot be made to rest simply on its self-consistency. For it would be possible to devise many such patterns. Each dating must be grounded in the requirements of the internal and external evidence in so far as this points to a particular period or even year. Again it has to be said in reminder that every date is more approximate than it looks - though relatively few (and those mainly for the gospels) require, I believe, a margin of error of more than two years either way. This may seem extraordinarily small, but with an overall span of not much more than twenty years to which we have found ourselves working for the completion of the New Testament documents, it is proportionately high.

First, then, I would set out the results in a simple chronological table, for comparison with the tables in the opening chapter. Yet there is an important factor which can only be brought out in the subsequent graph, namely, that while some books, notably epistles, are written over a short period for a specific occasion, others, notably gospels (but also the Didache), must be seen as the product of a much longer period of gestation in which there are at least three stages (represented in the diagram by dots, dashes and continuous lines), corresponding to the traditions behind them, the first drafts, and the final writing and rewriting. Obviously these stages, which run into each other, can be represented only very approximately, but the last term (which is itself only the starting-point for textual and other accretions) gives but an inadequate and misleading idea of the date of composition. We cannot simply say that Mark was written in the year 'x' as with fair precision we can say that I Corinthians was written in the year y.

So, bearing this in mind, I first list the books of the New Testament and of the immediate sub-apostolic age in what I believe (with qualifications and alternatives again omitted) to have been their approximate order of completion. Then I try to give a conspectus of their overlapping development in the order in which they appear in the New Testament canon and in Lightfoot's collection of the Apostolic Fathers.

James   c. 47-8
1 Thessalonians   early 50
2 Thessalonians   50-1
1 Corinthians   spring 55
1 Timothy   autumn 55
2 Corinthians   early 56
Galatians   later 56
Romans   early 57
Titus   late spring 57
Philippians   spring 58
Philemon   summer 58
Colossians   summer 58
Ephesians   late summer 58
2 Timothy   autumn 58
The Didache   c. 40-60
Mark   c. 45-60
Matthew   c. 40-60+
Luke   -57-60+
Jude   61-2
2 Peter   61-2
Acts   -57-62+
1, 2, 3 John   c. 60-65
1 Peter   spring 65
John   c. -40-65+
Hebrews   c. 67
Revelation   late 68 (-70)
1 Clement   early 70
Barnabas   c. 75
The Shepherd of Hermas   -c. 85

What corollaries then should be drawn from these conclusions?

1. There is, first of all, the observation that all the various types of the early church's literature (including the Didache, a version of its 'manual of discipline') were coming into being more or less concurrently in the period between 40 and 70. This, I believe, is what we should naturally expect. The notion that all the Pauline epistles, with the theology they imply, were prior to all the gospels, with the theology they imply, is not one that we should derive from the documents themselves. Laymen are always surprised to be told it, and I believe they are right to be surprised. I suggest that what has emerged is a more credible pattern.

CE: 30 40 50 60 70 80 90
Matthew          - - -- --- --- -- --- ---- -- -- --------- ----    
Mark          - - -- - - -- -- ------ - ----------      
Luke          - - -- --- - - - - - ------ -- -- --------- -    
John          - - -- --- -- -- -- ------ --------------- ---------    
Acts          - - -- --- - - - - - ------ - - - - -------- -----    
Pauline Epistles       ---------------      
Hebrews                   --    
James               --        
1 Peter                --    
2 Peter           ---    
Johannine Epistles            -----    
Jude          --    
Revelation                  ----    
1 Clement                      - --  
The Didache          - - -- - - - -- --- --- --------      
Barnabas                ---  
Shepherd of Hermas                     --- ------
CE: 30 40 50 60 70 80 90

2. The pattern of early church history suggested by the New Testament documents now reinforces that which we should independently deduce from the Acts narrative (up to the point that it takes us). This could approximately, decade by decade, be plotted as follows:

30-40   early mission in Palestine and Syria
40-50   consolidation of bases for the next thrust
50-60   rapid expansion into Asia Minor and Europe
60-70   testings from within and from without
70+   re-orientation and reappraisal

The epistle of James affords us a glimpse into the period prior to 50. But the really creative period of the primitive church, its 'Elizabethan era' from the point of view of literary output, was undoubtedly the 50s. These saw the full flowering of the preaching and teaching traditions in the gospels and the Didache and the creation of the Pauline corpus. The 60s mark the beginning of the silver age (already foreshadowed by the Pastorals), with its concern for the confirmation and defence of the faith against the threats of heresy and schism, persecution and defection (Acts, Jude and II Peter, the Johannine epistles, I Peter, the prologue and epilogue of the fourth gospel, Hebrews and Revelation). Yet the glow remains till the end of the first apostolic generation, fed by the fire of the martyrs and reflected in I Clement. Thereafter the spiritual temperature drops and the literary production of the church falls away, both in quantity and quality (Barnabas and Hennas), till it begins to emerge again with the martyrs and apologists of the second century (Ignatius and Polycarp, Justin and the Epistle to Diognetus). In the first century as in every other age of church history there were periods of rapid social change and of relative quiescence, of carefree ferment and of careful consolidation. The pattern of dating that has emerged helps us plot these without unnatural forcing. In particular, we can see a reason why the books we have assigned to the 60s belong together and belong there (in a decade full of events at each end of the empire that we know a great deal about) in a way in which there is no visible rationale for the books usually consigned to the reign of Domitian.

3. Thirdly, and most difficult to interpret aright, there are the deductions to be drawn from the contraction of the overall span. Just as the shrinking of the span from 50-150+ to one from 50-100+ resulted in discrediting some of the extremer forms of scepticism about the Christian tradition, so a further reduction in final datings by more than half from -50 to -70 must tend to reinforce a greater conservatism. Yet it is important to define this rather carefully. The last inference to be drawn is that it renders otiose or invalid the critical work done on the documents of the New Testament over the past two hundred years. For it is by applying the same critical methods and criteria that the conclusions have been reached. In particular the recent light thrown on the history of the tradition by form criticism and redaction criticism is, as we have stressed, by no means to be repudiated. It is merely that unexamined assumptions have tended to lead to the unwarranted conclusions that the more the documents tell us about the early church (a) the less they tell us about Jesus and (b) the longer they took to develop. But neither conclusion necessarily follows. As George Orwell showed in his Critical Essays, [George Orwell, Critical Essays, 1954.]
it is possible to put questions to all sorts and levels of literature - from Dickens to seaside postcards - to get it to yield information (in his case, about the socio-economic attitudes of its authors) which it was never written to provide. Yet it does not follow that the more it tells us of this, the less it tells us of what it is meant to be about. In fact the more we recognize the standpoint and prejudices of the writers, the more we are in a position to discount these when assessing their contribution to their subjects. Thus, in relation to the gospels, Jeremias has demonstrated in his Parables of Jesus and his New Testament Theology how it is possible to use all the tools and techniques of critical scholarship not to induce historical scepticism but by a gradual peeling back of the layers superimposed by the church to expose with the greater confidence what is likely not to come from it but from Jesus himself.

In the same way there is no necessary correlation between the wealth of knowledge the documents can be made to yield about their setting in the life of the church and the duration of the period for which these processes give evidence. In logic this is obvious. Yet the 'tunnel period' between the events of the life, death and resurrection of Jesus and the moment when, as it were, the train emerges, laden with ecclesiastical baggage, in our canonical documents has been viewed as so extended that almost anything could have happened on the way. To shorten the tunnel in principle changes nothing. For much can overtake and overlay a tradition (especially about a storied and creative character) in a remarkably brief time. But obviously there is less likelihood of distortion the shorter the interval. Moreover, there is a critical point of transition. If one is dealing with a gap, say, of thirty years (the distance that separates us, at the point of writing, from the end of the second world war), there is a good deal of built-in control in the form of living memory - whereas if the distance is doubled the controls are much less than half as strong. Without access to public records, when parents or grandparents die folklore takes over. And what applies to the gospel stories applies also to the history researched by the author of Acts. His claim to have 'gone over the whole course of events' for himself will obviously be affected by whether, as the 'we' passages imply, he has personally shared in much of it. This does not mean that we must simply take him at his word. The scholarly checks remain, to test whether in any given instance the tradition (of whatever date) is good tradition, or whether the docu-ments present us with a picture that is evidently unreliable or palpably anachronous. The results of such tests will continue to be a matter of degree and of judgment, on which scholars will properly differ. But it is worth reminding New Testament theologians of the friendly chiding they have received, for instance, from the classical historian Sherwin-White for not recognizing, by any contemporary standards, what excellent sources they have. This judgment does not depend on dating - as far as I know he would accept the traditional datings of the gospels and Acts - but it is obviously strengthened if the gap between the records and the events is that much shorter.

Perhaps I may be allowed to insert at this point a somewhat naughty comment, quoted by others before me, from a book by a layman on the fourth gospel which with a light touch takes the academics to task: [A. H. N. Green-Armytage, John Who Saw, 1952, 12f.]

There is a world - I do not say a world in which all scholars live but one at any rate into which all of them sometimes stray, and which some of them seem permanently to inhabit - which is not the world in which I live. In my world, if The Times and The Telegraph both tell one story in somewhat different terms, nobody concludes that one of them must have copied the other, nor that the variations in the story have some esoteric significance. But in that world of which I am speaking this would be taken for granted. There, no story is ever derived from facts but always from somebody else's version of the same story. ... In my world, almost every book, except some of those produced by Government departments, is written by one author. In that world almost every book is produced by a committee, and some of them by a whole series of committees. In my world, if I read that Mr Churchill, in 1935, said that Europe was heading for a disastrous war, I applaud his foresight. In that world no prophecy, however vaguely worded, is ever made except after the event. In my world we say, 'The first world-war took place in 1914-1918.' In that world they say, 'The world-war narrative took shape in the third decade of the twentieth century.' In my world men and women live for a considerable time - seventy, eighty, even a hundred years - and they are equipped with a thing called memory. In that world (it would appear) they come into being, write a book, and forthwith perish, all in a flash, and it is noted of them with astonishment that they 'preserve traces of primitive tradition' about things which happened well within their own adult lifetime.

Such a statement can be used - and has been used - to buttress, if not the fundamentalism of the fearful, at any rate the conservatism of the committed. [For an expansion of this and other attitudes to the Bible, I would refer to a forthcoming popular book of mine, Can We Trust the New Testament?, 1977.]
Yet it would be sad if the conclusion were to be drawn from this study that I was giving any comfort to an obscurantist or literalist approach to the New Testament. Since the passage quoted was written a propos the fourth gospel and since it is on this gospel above all that my argument for early dating and apostolic authorship could well be interpreted in this direction, it may be worth indicating very briefly what in my judgment this conclusion does, and does not, imply.

It does not imply a return to a position in which John is held to be reporting the ipsissima verba of Jesus (whether he catches what Jeremias distinguishes as the ipsissima vox is a very different matter). [For this important distinction, cf. his New Testament Theology I, ET 1971,37.] His theological purpose is unaffected by whether he is writing late or early, from sources or from source. If we conclude that 'his witness is true' we are not back at a purely physical understanding of witness or at a verbalistic understanding of truth - both of which are decisively repudiated by the gospel itself. Nor are we denying, in this gospel or any other, the processes of community tradition - fostered by the needs of apologetic and preaching, catechesis and liturgy - in favour of an individualistic understanding of the channel of transmission as the memory of one old man. Indeed, while nowhere more than in the Johannine corpus are we so aware of the authoritative note, 'La tradition c'est moi', [Cf. P. H. Menoud, L'evangile deJean d'apris les recherces recentes, Neuchatel 21947, 77.]
nowhere else is the 'we' of the community so explicit or the overtones of worship and the sacraments (to mention no others) so clearly to be heard. [Cf. Cullmann, Early Christian Worship, ET 1953 - however exaggerated on occasion.] Nor are we saying that the Johannine tradition, if or because it goes back to one who claimed to rest on his master's breast, is for that reason always reliable or in isolation the whole truth. In fact the more one becomes convinced of the complex and complementary nature of the synoptic relationships and dissatisfied with a simple pattern of literary dependence and temporal sequence, the more persuaded also one becomes of the distortion involved in the academic isolation of 'the synoptics' from 'the fourth gospel'. And this distortion has its effects in both directions. John has to be complemented by the other traditions and the other traditions by John for a 'stereoscopic' view. And nowhere, I believe, is this more true than in interpreting the literally crucial meeting between Word and flesh, theology and history, πνεῦμα and κόσμος, that comes to its climax in the trial and death of Jesus. [Cf. again for the outworking of this theme my article, 'His Witness is True' in Moule and Bammel, Jesus and the Politics of his Day.]
What is being asserted is that John has just as much right to be taken seriously on the history as well as on the theology, that his tradition reaches continuously back at least as far as that of the others, and that his claim to be heard, if indeed he is John the son of Zebedee, is certainly no less than that of Matthew, Mark or Luke. And yet, if Plato may be said to have understood his master best, the Socrates of Xenophon, or even of Aristophanes, is an indispensable supplement and indeed corrective to the portrait he paints.

In closing I would return to the position from which I began, that all the statements of this book should be taken as questions. It certainly makes no claim to represent a conclusive redating of the New Testament - if only because I am aware of how often I have changed my mind in the course of the work. It is an irritant and incentive to further exploration, and, I should like to think, to the opening up of fresh questions. For, as again in archaeology, settled positions, even if they prove to be vindicated, can by the very weight of their consensus deaden dissatisfaction and deter discovery. But if the chronology of the documents and the pattern of development should turn out to be anything like what I have suggested, then there will be scope for numerous new trajectories to be drawn and for the rewriting of many introductions to - and ultimately theologies of - the New Testament. For dates remain disturbingly fundamental data.