the fall of Jerusalem in AD 70, and with it the collapse of institutional Judaism based on the temple, is never once mentioned as a past fact.
Redating the New Testament
By John A.T. Robinson, fellow and Dean of Chapel, Trinity College, Cambridge, and Assistant Bishop of Southwark. His other works include The Human Faces of God and Honest to God.
“One of the oddest facts about the New Testament is that what on any showing would appear to be the single most datable and climactic event of the period – the fall of Jerusalem in AD 70, and with it the collapse of institutional Judaism based on the temple – is never once mentioned as a past fact. “
J.A.T. ROBINSON’S REDATED NEW TESTAMENT CHRONOLOGY
|LISTED BY DATE
||LISTED BY TITLE
|I||Dates & Data|
|II||The Significance of 70|
|III||The Pauline Epistles|
|IV||Acts & the Synoptic Gospels|
|V||The Epistle of James|
|VI||The Petrine Epistles & Jude|
|VII||The Epistle to the Hebrews|
|VIII||The Book of Revelation|
|IX||The Gospel & Epistles of John|
|X||A Post-Apostolic Postscript|
|XI||Conclusions & Corollaries|
I really have no more to say than thank you — to my long-suffering secretary Stella Haughton and her husband; to Professor C. F. D. Moule from whose New Testament seminar so small a seed has produced so monstrous a manuscript, on which he gave such kindly judgment; to my friends, Ed Ball, Gerald Bray, Chip Coakley, Paul Hammond and David McKie, who advised or corrected at many points; and finally to Miss Jean Cunningham of the SCM Press for all her devoted attention to tedious detail.
Introduction: Dates and Data
WHEN WAS THE New Testament written? This is a question that the outsider might be forgiven for thinking that the experts must by now have settled. Yet, as in archaeology, datings that seem agreed in the textbooks can suddenly appear much less secure than the consensus would suggest. For both in archaeology and in New Testament chronology one is dealing with a combination of absolute and relative datings. There are a limited number of more or less fixed points, and between them phenomena to be accounted for are strung along at intervals like beads on a string according to the supposed requirements of dependence, diffusion and development. New absolute dates will force reconsideration of relative dates, and the intervals will contract or expand with the years available. In the process long-held assumptions about the pattern of dependence, diffusion and development may be upset, and patterns that the textbooks have taken for granted become subjected to radical questioning.
The parallel with what of late has been happening in archaeology is interesting. The story can be followed in a recent book by Colin Renfrew. [C. Renfrew, Before Civilization: the Radiocarbon Revolution and Prehistoric Europe, 1973.] As he presents it, there was in modern times up to about the middle of this century a more or less agreed pattern of the origins and development of European civilization. The time scale was set by cross-dating finds in Crete and Greece with the established chronology of the Egyptian dynasties, and the evidence from Western Europe was then plotted by supposing a gradual diffusion of culture from this nodal point of Aegean civilization, to the remotest, and therefore the most recent, areas of Iberia, France, Britain and Scandinavia. Then in 1949 came the first radio-carbon revolution, which made possible the absolute dating of prehistoric materials for the first time. The immediate effect was greatly to extend the time span. Renfrew sums up the impact thus [Ibid., 65f.]:
The succession of cultures which had previously been squeezed into 500 years now occupied more than 1,500. This implies more than the alteration of a few dates: it changes the entire pace and nature of the cultural development. But … it did not greatly affect the relative chronology for the different regions of Europe: the megalithic tombs of Britain, for instance, were still later than those further south. … None of the changes … challenged in any way the conventional view that the significant advances in the European neolithic and bronze age were brought by influences from the Near East. It simply put these influences much earlier.
There were indeed uncomfortable exceptions, but these could be put down to minor inconsistencies that later work would tidy up. Then in 1966 came a second revolution, the calibration of the radiocarbon datings by dendrochronology, or the evidence of tree-rings, in particular of the incredibly long-lived Californian bristle-cone pine. This showed that the radiocarbon datings had to be corrected in an upward (i.e. older) direction, and that from about 2000 BC backwards the magnitude of the correction rose steeply, necessitating adjustments of up to 1000 years. The effect of this was not merely to shift all the dates back once more: it was to introduce a fundamental change in the pattern of relationships, making it impossible for the supposed diffusion to have taken place. For what should have been dependent turned out to be earlier.
The basic links of the traditional chronology are snapped and Europe is no longer directly linked, either chronologically or culturally, with the early civilizations of the Near East. [Ibid., 105.]
The whole diffusionist framework collapses, and with it the assumptions which sustained prehistoric archaeology for nearly a century. [Ibid., 85.]
This is a greatly oversimplified account, which would doubtless also be challenged by other archaeologists. Nothing so dramatic has happened or is likely to happen on the much smaller scale of New Testament chronology. But it provides an instructive parallel for the way in which the reigning assumptions of scientific scholarship can, and from rime to time do, get challenged for the assumptions they are. For, much more than is generally recognized, the chronology of the New Testament rests on presuppositions rather than facts. It is not that in this case new facts have appeared, new absolute datings which cannot be contested – they are still extraordinarily scarce. It is that certain obstinate questionings have led me to ask just what basis there really is for certain assumptions which the prevailing consensus of critical orthodoxy would seem to make it hazardous or even impertinent to question. Yet one takes heart as one watches, in one’s own field or in any other, the way in which established positions can suddenly, or subtly, come to be seen as the precarious constructions they are. What seemed to be firm datings based on scientific evidence are revealed to rest on deductions from deductions. The pattern is self-consistent but circular. Question some of the inbuilt assumptions and the entire edifice looks much less secure.
The way in which this can happen, and has happened, in New Testament scholarship may best be seen by taking some sample dips into the story of the subject. I have no intention of inflicting on the reader a history of the chronology of the New Testament, even if I were competent to do so. Let me just cut some cross-sections at fifty-year intervals to show how the spanof time over which the New Testament is thought to have been written has expanded and contracted with fashion.
We may start at the year 1800. For till then, with isolated exceptions, the historical study of the New Testament as we know it had scarcely begun. Dating was dependent on authorship, and the authorship of the various New Testament books rested on the traditions incorporated in their titles in the Authorized Version – the Gospel according to St Matthew, the Epistle of Paul the Apostle to the Ephesians, the Revelation of St John the Divine, and so on. All were by apostles or followers of the apostles and the period of the New Testament closed with the death of the last apostle, St John, who by tradition survived into the reign of the Emperor Trajan, c. 100 AD. At the other end the earliest Christian writing could be calculated roughly to about the year 50. This was done by combining the history of the early church provided in Acts with the information supplied by St Paul in Gal. 1.13-2.1 of an interval of up to seventeen ‘silent’ years following his conversion, which itself had to be set a few years after the crucifixion of Jesus in c. 30. The span of time for the composition of the New Testament was therefore about fifty years – from 50 to 100.
By 1850 the picture looked very different. The scene was dominated by the school of F. C. Baur, Professor of Church History and Dogmatics at Tübingen from 1826 to 1860. He questioned the traditional attribution of all but five of the New Testament books. Romans, I and II Corinthians and Galatians he allowed were by Paul, and Revelation by the apostle John. These he set in the 50s and late 60s respectively. The rest, including Acts and Mark (for him the last of the synoptists, ‘reconciling’ the Jewish gospel of Matthew and the Gentile gospel of Luke), were composed up to or beyond 150 AD, to effect the mediation of what Baur saw as the fundamental and all-pervasive conflict between the narrow Jewish Christianity of Jesus’ original disciples, represented by Peter and John, and the universalistic message preached by Paul. Only a closing of the church’s ranks in face of threats from the Gnostic and Montanist movements of the second century produced the via media of early Catholicism. The entire construction was dominated by the Hegelian pattern of thesis, antithesis and synthesis, and the span of time was determined more by the intervals supposedly required for this to work itself out than by any objective chronological criteria. The fact that the gospels and other New Testament books were quoted by Irenaeus and other church fathers towards the end of the second century alone set an upper limit. The end-term of the process was still the gospel of John, which was dated c. 160-70. The span of composition was therefore more than doubled to well over a hundred years – from 50+ to 160+.
By 1900 this schema had in turn been fairly drastically modified. The dialectical pattern of development had come to be recognized as the imposition it was [For the story, cf. W. G. Kummel, The New Testament: The History of the Investigation of its Problems, ET 1973, 162-84.]. A major factor in the correction of Baur’s picture of history was the work of J.B. Lightfoot, who was appointed a professor at Cambridge in 1861, the year following Baur’s death [Lightfoot’s achievement is particularly well brought out by S. C. Neill, The Interpretation of the New Testament, 1861-1961, Oxford 1964, 33-60.]. By the most careful historical investigation he succeeded in establishing the authenticity of the first epistle of Clement, which he dated at 95-6, and of the seven genuine epistles of lgnatius, between no and 115. In each of these both Peter and Paul are celebrated in the same breath without a trace of rivalry [I Clem. 5; Ignatius, Rom. 4.3.], and he demonstrated how groundless were Baur’s second-century datings. This achievement was acknowledged by the great German scholar Adolf von Harnack (1851-1930), who in 1897 published as the second volume of a massive history of early Christian literature [A. Harnack, Geschichte der altchristlichen Litteratur bis Eusehius, Leipzig 1893-7, vol. II (cited hereafter as Chron.).] his Chronologic der altchristlichen Litteratur bis Eusebius. Harnack’s survey, which has never been repeated on so comprehensive a scale [For a survey of surveys, cf. 0. Stahlin in W. Schmid and 0. Stahlin (cdd.), Geschichte der griechische Literatur, Munich 1961, 11.2, esp. 1112—1121.], gives a good indication of where critical opinion stood at the turn of the century. It still carried many of the marks of the Tiibingen period and continued to operate with a span of well over a hundred years. Isolating the canonical books of the New Testament (for Harnack covered all the early Christian writings, a number of which he placed before the later parts of the New Testament), we have the following summary [Chron.717-22. A comparable picture is to be found a few years earlier in A. Julicher’s Einleitung in das neue Testament, Tubingen 1894, though he put Mark after 70 and the Pastoral Epistles (I and II Timothy and Titus) at I25+.] (ignoring qualifications and alternative datings at this point as irrelevant to the broad picture):
|48-9||I and II Thessalonians|
|53||I and II Corinthians, Galatians (?)|
|57-9||Colossians, Philemon, Ephesians (if genuine), Philippians|
|59-64||Pauline fragments of the Pastoral Epistles|
|81-96||(‘under Domitian’) I Peter, Hebrews|
|80-110||John, I-111 John|
|90-110||I and II Timothy, Titus|
It is to be observed that the gospel of John has reverted to somewhere around the turn of the first century and no longer represents the terminus ad quern. Mark and Acts have been set much further back, and Harnack was subsequently to put them a good deal earlier still.
A similar but slightly more contracted scheme is to be found in the article on New Testament chronology by H. von Soden in the contemporary Encyclopaedia Biblica [Encyclopaedia Biblica, edd. T. K. Cheyne and J. S. Black, 1899-1903, I, 799-819.] His summary dates are:
|50-60+||The Pauline Epistles|
|93-96||Hebrews, I Peter, Revelation|
|-100||Ephesians, Luke, Acts, John, I-III John|
|100-33||Jude, Matthew, the Pastoral Epistles|
The individual articles in the same Encyclopaedia reveal however how volatile opinion was at that time. Acts is still put well into the second century and John shortly before 140. No date for II Peter is given, but even I Peter is put at 130-40. Above all, while I and II Corinthians are set in the mid-50s, Romans and Philippians are put in 120 and 125! But the articles on the latter two were written by the Dutch scholar W. C. van Manen (1842-1905), who regarded all the Pauline epistles (and indeed the rest of the New Testament literature) as pseudonymous, or written under false names.
Yet while the radical critics were still oscillating wildly, conservative, yet still critical, opinion of the period was content to settle for a span of composition between 50 and 100+, with the single exception of II Peter at c. 150. This was true both of English scholarship reflected in Hastings’ Dictionary of the Bible [Dictionary of the Bible, ed. J. Hastings, Edinburgh 1898-1904.] and of American represented by B. W. Bacon’s Introduction to the New Testament [B. W. Bacon, Introduction to the New Testament, New York 1900.]. Indeed the most conservative dating of all was by the German Theodore Zahn (1838-1933) whose Introduction to the New Testament [T. Zahn, Introduction to the New Testament, originally Leipzig 1897-9, ET Edinburgh 1909.] a monument of erudition and careful scholarship, set all the books between 50 and 95, including II Peter.
By 1950 the gap between radical and conservative had narrowed considerably, and we find a remarkable degree of consensus. There is still marginal variation at the upper limit, but the span of composition has settled down to a period from about 50 to 100 or no, with the single exception again of II Peter (c. 150). This generalization holds of all the major introductions and comparable surveys, English, American and Continental, Protestant and Catholic, published over the twenty years following 1950. [R. G. Heard, An Introduction to the New Testament, 1950; H. F. D. Sparks, The Formation of the New Testament, 1952; A. H. McNeile, An Introduction to the Study of the New Testament, revised by C. S. C. Williams, Oxford 1953 (cited henceforth as McNeile-Williams); W. Michaelis, Einleitung in das neue Testament, Bern 1954; A. Wikenhauser, New Testament Introduction (Freiburg 21956), ET New York 1958; A. Robert and A. Feuillet, Introduction to the New Testament (Tournai 1959), ET New York 1965; D. Guthrie, New Testament Introduction, 1961-5, 31970; Peake’s Commentary on the Bible, revised, ed. M. Black, 1962; The Interpreter’s Dictionary of the Bible, New York 1962; R. M. Grant, A Historical Introduction to the New Testament, i963;W. G. Kummel, Introduction to the New Testament (Heidelberg i963),ET 1966; 21975; W. Marxsen, Introduction to the New Testament (Gutersloh 1963), ET Oxford 1968; E. F. Harrison, Introduction to the New Testament, 1964; R. H. Fuller,A Critical Introduction to the New Testament, 1966; W. D. Davies, Invitation to the New Testament, New York 1966; A. F. J. Klijn, An Introduction to the New Testament, ET Leiden 1967; D.J. Selby, Introduction to the New Testament, New York 1971.]
The prevailing position is fairly represented by Kummel, who tends to be more radical than many Englishmen and more conservative than many Germans. His datings, again omitting alternatives, are:
|50-1||I and II Thessalonians|
|53-6||Galatians, Philippians, I and II Corinthians, Romans|
|90-5||I Peter, Revelation|
|100+||I and II Timothy, Titus|
In this relatively fixed firmament the only ‘wandering stars’ are Ephesians, I Peter, Hebrews and James (and occasionally the Pastorals and Jude), which conservatives wish to put earlier, and Colossians and II Thessalonians, which radicals wish to put later. So once more the span (with one exception) is back to little more than fifty years.
But before closing this survey I would draw attention to the latest assessment of all, Norman Perrin’s The New Testament: An Introduction [N. Perrin, The New Testament: An Introduction, New York 1974.], since it could suggest a return to a wider spread. His approximate datings are:
|50-60||I Thessalonians, Galatians, I and II Corinthians, Philippians, Philemon, Romans|
|70-90||II Thessalonians, Colossians, Ephesians, Mark, Matthew, Luke-Acts, Hebrews|
|80-100||John, I-III John|
|90-140||I Peter, James, Pastoral Epistles, Jude, II Peter [The order of this last group is only a guess. No dates are given, except that I Peter is about the end of the first century and II Peter c. 140.]|
Perrin represents the standpoint of redaction criticism, which goes on from source criticism (dealing with documentary origins) and form criticism (analysing the formative processes of the oral tradition) to emphasize the theological contribution of the evangelists as editors. There is no necessary reason why its perspective should lead to later datings. Indeed other representatives of the same viewpoint who have written New Testament introductions, Marxsen and Fuller, have taken over their precursors’ datings. Moreover, the gospels, with which the redaction critics have been most concerned, all remain, including the fourth, within what Perrin calls ‘the middle period of New Testament Christianity’, ‘the twenty-five years or so that followed the fall of Jerusalem’. Yet subsequent to this period he sees a further stage, extending into the middle of the second century, in which the New Testament church is ‘on the way to becoming an institution’. If we ask why it is only then becoming an institution, the answer is bound up with his ‘theological history of New Testament Christianity’ [Op. cit, 39-63.]. The course of this he traces from ‘Palestinian Jewish Christianity’, through ‘Hellenistic Jewish Mission Christianity’, ‘Gentile Christianity’ and ‘the apostle Paul’, to ‘the middle period’, and finally into ’emergent Catholicism’. Yet these categories, taken over from Rudolf Bultmann and his successors, have of late come in for some stringent criticism not only from England [I. H. Marshall, ‘Palestinian and Hellenistic Christianity: Some Critical Comments’, NTS 19, 1972-3, 271-87; ‘Early Catholicism’ in R. N. Longenecker and M. C. Tenney (edd.), New Dimensions in New Testament Study, Grand Rapids, Michigan, 1974, a 17-31.] but from Germany itself [M. Hengel, ‘Christologie and neutestamentliche Chronologic’ in H. Baltens-weiler and B. Reicke (edd.), Neues Testament und Geschichte: Oscar Cullmann zum 70. [Geburtstag, Zurich and Tubingen 1972, 43-67; Judaism and Hellenism, ET 1974.], none of which Perrin acknowledges. The entire developmental schema (closely parallel to the ‘diffusionist framework’ in archaeology), together with the time it is assumed to require, begins to look as if it may be imposed upon the material as arbitrarily as the earlier one of the Tiibingen school. It is premature to judge. But certainly it cannot itself be used to determine the datings which are inferred from it. It must first be submitted to a more rigorous scrutiny in the light of the independent data.
Indeed what one looks for in vain in much recent scholarship is any serious wrestling with the external or internal evidence for the daring of individual books (such as marked the writings of men like Lightfoot and Harnack and Zahn), rather than an a prioripattern of theological development into which they are then made to fit. [Perrin’s particular schema is in itself fairly arbitrary. It is hard to see by what criteria of doctrine or discipline I and II Peter are both subsumed under the heading of ’emergent Catholicism’; in fact in the analysis of the marks of this phenomenon (op. cit., 268-73) I Peter is scarcely mentioned. Moreover, while he acknowledges his deep indebtedness to E. Kasemann for his estimate of II Peter (‘An Apologia for Primitive Christian Eschatology’, Essays on New Testament Themes, ET (SBT 41) 1964, 169-95), he ignores Kasemann’s equally strong contention (‘Ketzer und Zeuge’, ZTK 48, 1951, 292-311) that III John reflects a second-century transition to Ignatian monepiscopacy. (Of the Johannine epistles he merely says, 249: ‘We are now in the middle period of New Testament Christianity.’) He does not explain why I Clement’s concern for apostolic succession and Ignatius’ plea for unity around the monarchical bishop (quintessential interests, one would have thought, of ’emergent Catholicism’) receive no mention in New Testament documents supposedly later than they are.] In fact ever since the form critics assumed the basic solutions of the source critics (particularly with regard to the synoptic problem) and the redaction critics assumed the work of the form critics, the chronology of the New Testament documents has scarcely been subjected to fresh examination.
No one since Harnack has really gone back to look at it for its own sake or to examine the presuppositions on which the current consensus rests. It is only when one pauses to do this that one realizes how thin is the foundation for some of the textbook answers and how circular the arguments for many of the relative datings. Disturb the position of one major piece and the pattern starts disconcertingly to dissolve.
That major piece was for me the gospel of John. I have long been convinced that John contains primitive and reliable historical tradition, and that conviction has been reinforced by numerous studies in recent years. But in reinforcing it these same studies have the more insistently provoked the question in my mind whether the traditional dating of the gospel, alike by conservatives and (now) by radicals, towards the end of the first century, is either credible or necessary. Need it have been written anything like so late? As the arguments requiring it to be set at a considerable distance both in place and time from the events it records began one by one to be knocked away (by growing recognition of its independence of the synoptists and, since 1947 by linguistic parallels from the Dead Sea Scrolls), I have wondered more and more whether it does not belong much nearer to the Palestinian scene prior to the Jewish revolt of 66-70.
But one cannot redate John without raising the whole question of its place in the development of New Testament Christianity. If this is early, what about the other gospels? Is it necessarily the last in time? Indeed does it actually become the first? – or are they earlier too? And, if so, how then do the gospels stand in relation to the epistles? Were all the Pauline letters penned, as has been supposed, before any of the gospels? Moreover, if John no longer belongs to the end of the century, what of the Johannine epistles and the other so-called Catholic Epistles which have tended to be dated with them? And what about the book of Revelation, which, whatever its connection with the other Johannine writings, everyone seems nowadays to set in the same decade as the gospel?
It was at this point that I began to ask myself just why any of the books of the New Testament needed to be put after the fall of Jerusalem in 70. As one began to look at them, and in particular the epistle to the Hebrews, Acts and the Apocalypse, was it not strange that this cataclysmic event was never once mentioned or apparently hinted at? And what about those predictions of it in the gospels – were they really the prophecies after the event that our critical education had taught us to believe? So, as little more than a theological joke, I thought I would see how far one could get with the hypothesis that the whole of the New Testament was written before 70. And the only way to try out such a hypothesis was to push it to its limits, and beyond, to discover what these limits were. Naturally, there were bound to be exceptions – II Peter was an obvious starter, and presumably the Pastorals – but it would be an interesting exercise.
But what began as a joke became in the process a serious preoccupation, and I convinced myself that the hypothesis must be tested in greater detail than the seminar-paper with which it started would allow. The result is that I have found myself driven to look again at the evidence for all the accepted New Testament datings. But so far from forcing it to a new Procrustean bed of my own making, I have tried to keep an open mind. I deliberately left the treatment of the fourth gospel to the last (though increasingly persuaded that it should never be treated in isolation from the other three, or they from it) so as not to let my initial judgment on it mould the rest of the pattern to it. Moreover, I have changed my mind many times in the course of the work, and come through to datings which were not at all what I expected when I began. Indeed I would wish to claim nothing fixed or final about the results. Once one starts on an investigation like this one could go on for years. Problems that one supposed in one’s own mind were more or less settled (e.g. the synoptic problem) become opened up again; and almost all the books or articles that have been written on the New Testament (and many too on ancient history) threaten to become relevant. But one has to stop somewhere. I am much more aware of what I have not read. But this will have to do as a stone to drop into the pond, to see what happens.
Naturally if one presumes to challenge the scientific establishment in any field one must be prepared to substantiate one’s case in some detail. So I have tried to give the evidence and provide the references for those who wish to follow them up. However, short of making it one’s life’s work (and frankly chronology is not mine), one must delimit the task. I have not attempted to go into the theoretical basis of chronology itself or to get involved in astronomical calculations or the complex correlation of ancient dating systems. [Cf.J. Finegan, Handbook of Biblical Chronology, Princeton 1964, for the single most useful survey; also T. Lewin, Fasti Sacri: A Key to the Chronology of the New Testament, 1865; J. van Goudoeuver, Biblical Calendars, Leiden 21961; A. K. Michels, The Calendar of the Roman Republic, Princeton, NJ, 1967; E. J. Bickermann, Chronology of the Ancient World, 1968; A. E. Samuel, Greek and Roman Chronology, Munich 1972; E. Schiirer, The History of the Jewish People in the Age of Jesus Christ, revised ET, Edinburgh 1973, vol.1. Appendix III (‘The Jewish Calendar’).] These things are too high for one who finds himself confused even when changing to summer time or crossing time zones! Nor have I entered the contentious area of the chronology of the birth, ministry and death of Jesus, since it does not seriously affect the dating of thebooks of the New Testament.
Nor have I found it necessary to be drawn into the history of the canon of the New Testament, since, unless one has reason to suppose that the books were written very late, how long an interval elapsed before they became collected or acknowledged as scripture is but marginally relevant. Above all, I have not ventured into the vast field of the non-canonical literature of the sub-apostolic age, except to the extent that this is directly relevant to the dating of the New Testament books themselves. Without attempting to survey this literature, both Jewish and Christian, for its own sake (which would have taken me far beyond my competence), I have simply devoted a postscript to it, in so far as by comparison and contrast it can help to check or confirm the conclusions arrived at from the study of the New Testament.
Finally, in a closing chapter I have sketched some of the conclusions and corollaries to be drawn – and not to be drawn – from such a study. My position will probably seem surprisingly conservative – especially to those who judge me radical on other issues. But I trust it will give no comfort to those who would view with suspicion the application of critical tools to biblical study – for it is reached by the application of those tools. I claim no great originality – almost every individual conclusion will be found to have been argued previously by someone, often indeed by great and forgotten men – though I think the overall pattern is new and I trust coherent. Least of all do I wish to close any discussion. Indeed I am happy to prefix to my work the words with which Niels Bohr is said to have begun his lecture-courses: ‘Every sentence I utter should be taken by you not as a statement but as a question.’ [Quoted by J. Bronowski, The Ascent of Man, 1973, 334.]
II. The Significance of 70
ONE of the oddest facts about the New Testament is that what on any showing would appear to be the single most datable and climactic event of the period – the fall of Jerusalem in AD 70, and with it the collapse of institutional Judaism based on the temple – is never once mentioned as a past fact. It is, of course, predicted; and these predictions are, in some cases at least, assumed to be written (or written up) after the event. But the silence is nevertheless as significant as the silence for Sherlock Holmes of the dog that did not bark. S. G. F. Brandon made this oddness the key to his entire interpretation of the New Testament: [S. G. F. Brandon, The Fall of Jerusalem and the Christian Church, 1951; 2I957; ‘The Date of the Markan Gospel’, NTS 7, 1960-1, 126-41; Jesus and the Zealots, Manchester 1967; The Trial of Jesus, 1968.] everything from the gospel of Mark onwards was a studied rewriting of history to suppress the truth that Jesus and the earliest Christians were identified with the revolt that failed. But the sympathies of Jesus and the Palestinian church with the Zealot cause are entirely unproven and Brandon’s views have won scant scholarly credence. [Cf. the devastating review of Jesus and the Zealots by Hengel, JSS 14, 1969, a 31-40; and his .Die Zeloten, Leiden 1961; Was Jesus a Revolutionist?,‘ ET Philadelphia 1971; Victory over Violence, ET 1975; also W. Wink, ‘Jesus and Revolution: Reflection on S. G. F. Brandon’s Jesus and the Zealots’, USQR 26, 1969, 37-59; O. Cullmann, Jesus and the Revolutionaries, ET New York 1970; and especially the forthcoming symposium edited by C. F. D. Moule and E. Bammel, Jesus and the Politics of his Day, Cambridge 1977(?). P. Winter makes the important point that ‘nothing that Josephus wrote lends any support to the theory that Jesus was caught up in revolutionary, Zealotic or quasi-Zealotic activities. … The relatively friendly attitude of Josephus towards Jesus contrasts with his severe stricture of the Zealots and kindred activist groups among the Jews responsible for encouraging the people to defy Roman rule’ (Excursus II in Schurer, HJP I, 441).] Yet if the silence is not studied it is very remarkable. As James Moffatt said,
We should expect … that an event like the fall of Jerusalem would have dinted some of the literature of the primitive church, almost as the victory at Salamis has marked the Persae. It might be supposed that such an epoch-making crisis would even furnish criteria for determining the dates of some of the NT writings. As a matter of fact, the catastrophe is practically ignored in the extant Christian literature of the first century. [J. Moffatt, Introduction to the Literature of the New Testament, Edinburgh 31918, 3. This is quoted by L. H. Gaston, No Stone on Another: Studies in the Fall of Jerusalem in the Synoptic Gospels (Nov Test. Suppl. 23), Leiden 1970, 5, who continues: ‘There is no unambiguous reference to the fall of Jerusalem anyplace outside the gospels.]
Similarly C. F. D. Moule:
It is hard to believe that a Judaistic type of Christianity which had itself been closely involved in the cataclysm of the years leading up to AD 70 would not have shown the scars – or, alternatively, would not have made capital out of this signal evidence that they, and not non-Christian Judaism, were the true Israel. But in fact our traditions are silent. [C. F. D. Moule, The Birth of the New Testament, 1962, 123.]
Explanations for this silence have of course been attempted. Yet the simplest explanation of all, that perhaps … there is extremely little in the New Testament later than AD 70 [Moule, op. cit., 121.] and that its events are not mentioned because they had not yet occurred, seems to me to demand more attention than it has received in critical circles.
Bo Reicke begins a recent essay with the words:
An amazing example of uncritical dogmatism in New Testament studies is the belief that the Synoptic Gospels should be dated after the Jewish War of AD 66-70 because they contain prophecies ex eventu of the destruction of Jerusalem by the Romans in the year 70. [B. Reicke, ‘Synoptic Prophecies on the Destruction of Jerusalem’, in D. W. Aune (ed.), Studies in New Testament and Early Christian Literature: Essays in Honor of Alien P. Wikgren (NovTest Suppl. 33), Leiden 1972, 121-34.]
In fact this is too sweeping a statement, because the dominant consensus of scholarly opinion places Mark’s gospel, if not before the beginning of the Jewish war, at any rate before the capture of the city. [Cf. the summary of opinions in V. Taylor, St Mark,21966, 31. He himself opts, with many others, for 65-70. Kummel, INT, 98, hedges his bets: ‘Since no overwhelming argument for the years before or after 70 can be adduced, we must content ourselves with saying that Mark was written ca. 70.] Indeed one of the arguments to be assessed is that which distinguishes between the evidence of Mark on the one hand and that of Matthew and Luke on the other. In what follows I shall start from the presumption of most contemporary scholars that Mark’s version is the earliest and was used by Matthew and Luke. As will become clear [Cf. pp. 92-4 below.], I am by no means satisfied with this as an overall explanation of the synoptic phenomena. I believe that one must be open to the possibility that at points Matthew or Luke may represent the earliest form of the common tradition, which Mark also alters for editorial reasons. I shall therefore concentrate on the differences between the versions without prejudging their priority or dependence. The relative order of the synoptic gospels is in any case of secondary importance for assessing their absolute relation to the events of 70. Whatever their sequence, all or any could have been written before or after the fall of Jerusalem.
Let us then start by looking again at the discourse of Mark 13. It begins:
As he was leaving the temple, one of his disciples exclaimed, ‘Look, Master, what huge stones! What fine buildings!’ Jesus said to him, ‘You see these great buildings? Not one stone will be left upon another; all will be thrown down.’
When he was sitting on the Mount of Olives facing the temple he was questioned privately by Peter, James, John, and Andrew. ‘Tell us,’ they said, ‘when will this happen? What will be the sign when the fulfilment of all this is at hand?’ (13.1-4).
The first thing to notice is that the question is never answered. In fact no further reference is made in the chapter to the destruction of the temple. This supports the judgment of most critics that the discourse is an artificial construction out of diverse teachings of Jesus, with parallels in various parts of the gospel tradition, and linked somewhat arbitrarily by the evangelist to a subsequent question of interest to the church, such as Mark regularly poses by the device of a private enquiry by an inner group of disciples (cf. 4.10; 7.17; 9.28). We need not stop to wrestle with the complex question of how much goes back to Jesus and how much is the creation of the community. That Jesus could have predicted the doom of Jerusalem and its sanctuary is no more inherently improbable than that another Jesus, the son of Ananias, should have done so in the autumn of 62 [Josephus, BJ, 6. 300-9. In citing Josephus I have followed the notation and, unless otherwise indicated, the translation in the Loeb Classical Library.]. Even if, as most would suppose [Josephus, BJ, 6. 300-9. In citing Josephus I have followed the notation and, unless otherwise indicated, the translation in the Loeb Classical Library.],0 the discourse represents the work of Christian prophecy reflecting upon the Old Testament and remembered sayings of Jesus in the light of the church’s experiences, hopes and fears, the relevant question is, What experiences, hopes and fears ?
The mere fact again that there is no correlation between the initial question and Jesus’ answer would suggest that the discourse is not being written retrospectively out of the known events of 70. Indeed the sole subsequent reference to the temple at all, and that only by implication, is in 13.14-16:
But when you see ‘the abomination of desolation’ usurping a place which is not his (let the reader understand), then those who are in Judaea must take to the hills. If a man is on the roof, he must not come down into the house to fetch anything out; if in the field, he must not turn back for his cloak.
It is clear at least that ‘the abomination of desolation’ cannot itself refer to the destruction of the sanctuary in August 70 or to its desecration by Titus’ soldiers in sacrificing to their standards [Josephus, BJ 6. 316.]. By that time it was far too late for anyone in Judaea to take to the hills, which had been in enemy hands since the end of 67 [Brandon, who argues for this, JTS 7, 133f., merely omits any reference to the injunction to take to the hills.]. Moreover, the only tradition we have as to what Christians actually did, or were told to do, is that preserved by Eusebius [HE 3. 5.3. Quotations from this work are from the translation and edition by H.J. Lawlor and J. E. L. Oulton, 1927-8.] apparently on the basis of the Memoirs of Hegesippus used also by Epiphanius. [Adv. haer. 29.7; 30.2; de mens. et pond. 15.2-5. For the case for a common source in the Hypommmata of Hegesippus, cf. H.J. Lawlor, Eusebiana, Oxford 1912, 27-34, who prints the full texts (101f.).] This says that they had been commanded by an oracle given ‘before the war’ to depart from the city,15 and that so far from taking to the mountains of Judaea, as Mark’s instruction implies, they were to make for Pella, a Greek city of the Decapolis, which lay below sea level on the east side of the Jordan valley. [According to Epiphanius’ version, the flight was made just before the beginning of the siege of Jerusalem itself. At that stage escape was indeed still possible. Speaking of November 66 Josephus says: ‘After this catastrophe of Cestius many distinguished Jews abandoned the city as swimmers desert a sinking ship’ (BJ 2.556). But an earlier reference (Ant.20.256) to the period between the arrival of Gessius Florus as procurator in 64 and the beginning of the war in 66 fits better a popular exodus and the Eusebian dating: ‘There was no end in sight. The ill-fated Jews, unable to endure the devastation by brigands that went on, were one and all forced to abandon their own country and flee, for they thought it would be better to settle among gentiles, no matter where’. If the Christian Jews were among them, then the λησταί (Josephus’ word for the Zealots) would have been the cause for the Christians’ dissociation from the revolt rather than, as Brandon thought, their attachment to it. This seems altogether more likely.] It would appear then that this was not prophecy shaped by events and cannot therefore be dated to the period immediately before or during the war of 66-70. [This point is made strongly, perhaps over-strongly, by Reicke, op. cit., 125. For a defence of the Pella tradition, against the criticisms of Brandon, Fall of Jerusalem, 168-78, cf. S. S. Sowers, ‘The Circumstances and Recollection of the Pella Flight’, TZ 26, 1970,305-20.]
What apparently the instruction is shaped by (whether in the mind of Jesus or that of a Christian prophet speaking in his name) is, rather, the archetypal Jewish resistance to the desecration of the temple-sanctuary by an idolatrous image under Antiochus Epiphanes in 168-7 BC. This was ‘the abomination of desolation … set up on the altar’ (I Mace. 1.54) referred to by Daniel (9.27 [LXX]; 11.31; 12.11), and it was in consequence of this and of the local enforcement of pagan rites that Mattathias and his sons ‘took to the hills, leaving all their belongings in the town’ (I Mace. 2.28). It is here that we should seek the clue to the pattern of Mark 13.14-16. Moreover the influence of the book of Daniel is so pervasive in this chapter [As well as in this passage, it is echoed in 13.4 (Dan. 12.7); 13.7 (Dan. 2.28); 13.19 (Dan. 12.1); and 13.26 (Dan. 7.13).] that it is hard to credit that what is regularly there associated with the abomination of desolation, namely, the cessation of the daily offering in the temple (Dan.8.13; 9.27; n.3i; 12.11) would not have been alluded to if this had by then occurred, as it did in August 70. [Josephus, BJ 6.94.]
It is more likely that the reference to ‘the abomination of desolation standing where he ought not’ (to stress Mark’s deliberate lack of grammatical apposition) is, like Paul’s reference to ‘the lawless one’ or ‘the enemy’ who ‘even takes his seat in the temple of God’ (II Thess.2.1-12), [There is here the same transition between neuter and masculine: τὸ κατέχον (v.6), ὁ κατέχων (v. 7).] traditional apocalyptic imagery for the incarnation of evil which had to be interpreted (‘let the reader understand’; cf. Rev. 13.18) according to whatever shape Satan might currently take. It is indeed highly likely that such speculation was revived, as many have argued [E.g. B. W. Bacon, The Gospel of Mark, New Haven, Conn., 1925, 53-68.], by the proposal of the Emperor Gaius Caligula in 40 to set up his statue in the temple (which was averted only by his death). [Josephus, Ant. 18. 261-309. For the horror and alarm which this raised among Jews, cf. Philo, Leg. Ad Gaium, 184-348.] Paul was evidently still awaiting the fulfilment of such an expectation in 50-1 (to anticipate the date of II Thessalonians), where ‘the restrainer’ holding it back is probably to be interpreted as the Roman Empire embodied in its emperor (ὁ κατέχων being a play perhaps on the name Claudius, ‘he who shuts’). His expulsion of the Jews from Rome in 49 could be reflected in the phrase of I Thess. 2.16 about retribution having overtaken them εἰς τἐλος (‘with a view to the end’?). [A suggested interpretation I owe to Dr E. Bammel.] The only other datable incident to which ‘the abomination’ might conceivably refer in retrospect is the control of the temple not by the Romans but by the Zealots temporarily in 66 and permanently in 68, which Josephus speaks of in terms of its ‘pollution’. [BJ 2.422-5; 4.147-92; 5.IQ. So M.-J. Lagrange, S. Matthieu, Paris 1927, 462; R. T. France, Jesus and the Old Testament, 1971, 227-39; W.J. Houston, New Testament Prophecy and Christian Tradition, unpublished D.Phil, thesis for the University of Oxford, 1973. Cf. F. F. Bruce, ‘Josephus and Daniel,’ ASTI 4, 1965, i53f.] This would be the very opposite of Brandon’s thesis, with the Zealots filling the role of antichrist. But it does not explain the masculine singular (as a vaticinium ex eventu should require) and again it is too late for a pre-war flight, and perhaps for any.
One is forced to conclude that the reference in Mark 13.14 to ‘the abomination of desolation standing where he ought not’ is an extremely uncertain indicator of retrospective dating. G. R. Beasley-Murray ends a note on the history of the interpretation of this verse with the words:
It would seem a just conclusion that the traditional language of the book of Daniel, the Jewish abhorrence of the idolatrous Roman ensigns, attested in the reaction to Pilate’s desecration, [The reference is to an incident in Caesarea in a6 (Josephus, Ant. 18. 55-9; BJ 2.169-74; Philo, Leg. ad Gaium 299-305) and therefore well before Jesus’ supposed utterance. Cf. P. L. Maier, ‘The Episode of the Golden Roman Shields at Jerusalem’, HTR 6a, 1969, 109-21.] and Jesus’ insight into the situation resulting from his people’s rejection of his message, supply a sufficient background for this saying. [G. R. Beasley-Murray, A Commentary on Mark Thirteen,1957, 72 (cf. 59-72).]
Marxsen, writing from a very different standpoint, regards the phrase as a vague reference to the forthcoming destruction of the temple and is forthright in saying: ‘From Mark’s point of view, a vaticinium ex eventu is an impossibility.’ [W. Marxsen, Mark the Evangelist, ET Nashville, Tenn., 1969, 170 (cf. 166-89); similarly E. Trocme, The Formation of the Gospel according to Mark, ET 1975, 104f., 245. He thinks Mark 1-13 was written c. 50 (259).]
With regard to Mark 13 as a whole the most obvious inference is that the warnings it contains were relevant to Christians as they were facing duress and persecution, alerting them to watchfulness against false alarms and pretenders’ claims, promising them support under trial before Jewish courts and pagan governors, and assuring them of the rewards of steadfastness. Doubtless the phrasing has been influenced and pointed up by what Christians actually experienced, but, as Reicke argues in the second half of his essay [‘Synoptic Prophecies’, 130-3.], there is nothing that cannot be paralleled from the period of church history covered by Acts (c. 30-62). As early as 50 Paul can say to the Thessalonians: ‘You have fared like the congregations in Judaea, God’s people in Christ Jesus. You have been treated by your countrymen as they are treated by the Jews’ (I Thess. 2.14). Unless the flight enjoined upon ‘those who are in Judaea’ is purely symbolic (of the church dissociating itself from Judaism) – and with the detailed instructions and the prayer that it may not be in winter (Mark 13.18) there is no reason to assume it is figurative any more than the very literal dissolution of Herod’s temple – then the directions for it must surely belong to a time when there still were Christians in Judaea, free and able to flee. Finally, we are in a period when it could still be said without reserve or qualification on the solemn authority of Jesus: ‘I tell you this: the present generation will live to see it all’ (13.30).
In fact there is, as we said, wide agreement among scholars that Mark 13 does fit better before the destruction of the temple it purports to prophesy. This is relevant as we turn now to Matthew and Luke. What will be significant are differences from Mark: otherwise the same presumption will continue to hold.
We will take Matthew first, since he is closest to the Markan tradition. But the first relevant passage in his gospel is not in fact in Markan material but in that which he has in common with Luke, the parable of the wedding feast (Matt.22.1-10 = Luke 14.16-24), where Matthew inserts the following:
The others seized the servants, attacked them brutally and killed them. The king was furious; he sent troops to kill those murderers and set their town on fire (22.6f.).
There can be little doubt that these verses are secondary to the parable. [Matthew has also tacked on the (originally separate) parable of the wedding garment (22.11-14).] They form part of an allegorical interpretation of the successive servants (Luke has one only) in terms of the prophets and apostles sent to Israel, as in the immediately preceding parable of the wicked husbandmen (Matt. 21.33-45). [Cf. especially 22.4, 6 with 21.35f.] The introduction of a military expedition while the supper is getting cold is particularly inappropriate. Luke has also allegorized the parable, to match the Jewish and Gentile missions of the church, by introducing two search-parties, first to the streets and alleys of the city and then to the highways and hedgerows. The secondary character of all these features is now further established by their absence from the same parable in the Gospel of Thomas (64). This version also supports the supposition, which we should independently deduce from his usage elsewhere (Matt.18.23; 25.34, 40), that it is Matthew who has brought in the figure of the king as the subject of the story: Luke and Thomas both simply have ‘a man’. It is therefore as certain as anything can be in this field that the crucial verse, ‘The king was furious; he sent troops to kill those murderers and set their town on fire’, is an addition, probably by the evangelist. The sole question is, When was it added and does it reflect in retrospect the destruction of Jerusalem (to which it must obviously allude)?
It has to be admitted that this is the single verse in the New Testament that most looks like a retrospective prophecy of the events of 70, and it has almost universally been so taken. It is the only passage which mentions the destruction of Jerusalem by fire. Yet, as K. H. Rengstorfhas argued, [K. H. Rengstorf, ‘Der Stadt der Morder (Mt 22.7)’ in W. Eitester (ed.), Judentum-Urchristentum-Kirche: Festschrift fur Joachim Jeremias (ZNW Beiheft 26), 1960, 106-29 (especially 125f.).] the wording of Matt. 22.7 represents a fixed description of ancient expeditions of punishment and is such an established topos of Near Eastern, Old Testament and rabbinic literature that it is precarious to infer that it must reflect a particular occurrence. He concludes that it has no relevance for the dating of the first gospel. And this conclusion is borne out in a further study by Sigfred Pedersen [S. Pedersen, ‘Zum Problem der vaticinia ex eventu (eine Analyse von Mt 21.33-46 par; 22.1-10 par)’,.ST19, 1965, 167-88.], who believes that this and the preceding parable of the wicked husbandmen are fundamentally shaped by material from the Old Testament, especially Jeremiah. The most he will say is that if Matthew is writing after 70, then we must see this as a contributory occasion for the addition (which of course no one would deny).
Moreover, if Matt. 2 2.7 did reflect the happenings of 70 one might expect that it would make a distinction that features in other post eventum ‘visions’, namely, that while the walls of the city were thrown down, it was the temple that perished by fire. Thus the Jewish apocalypse II Baruch clearly reflects the fall of Jerusalem to the Romans, though it purports to be the announcement to the prophet Baruch of a coming Chaldean invasion. It recognizes that the city and the temple suffered separate fates:
They delivered … to the enemy the overthrown wall, and plundered the house, and burnt the temple (80.3).
If one really wants to see what ex eventu prophecy looks like, one should turn to the so-called Sibylline Oracles (4.125-7):
And a Roman leader shall come to Syria, who shall burn down Solyma’s [Jerusalem’s] temple with fire, and therewith slay many men, and shall waste the great land of the Jews with its broad way. [Tr. R. H. Charles, The Apocrypha and Pseudepigrapha of the Old Testament II, Oxford 1913,395.]
It is precisely such detail that one does not get in the New Testament.
Finally, in Matthew’s parable the king clearly stands for God. In the war of 66-70 the king who sent the armies to quell the rebels was Nero, followed by Vespasian. Reicke says:
The picture of God sending his armies to punish all guests not willing to follow his invitation was in no way applicable to the war started by Nero to punish the leaders of rebellion against Roman supremacy. [Op.cit., 123.]
He argues indeed that there is every reason to assume that the final redactor of the parable would have altered the reference if he had been writing after 70. This, I believe, is putting it too strongly, since undoubtedly Christians came to see the destruction of Jerusalem as God’s retribution on Israel, whoever the human agent. [Cf. later (c. 300) Eusebius, HE 3.5.3: ‘The justice of God then visited upon them [the Jews] all their acts of violence to Christ and his apostles, by destroying that generation of wicked persons root and branch from among men’; also (c. 400) Sulpicius Severus, Chron.2.30. But evidence for this is remarkably absent from earlier writings where one might expect it, e.g. the Epistle of Barnabas or Justin’s Dialogue with Trypho.] Yet the correspondence does not seem close enough to require composition in the light of the event.
Nevertheless, the conclusion must, I think, stand that on the basis of Matt. 22.7 alone it is impossible to make a firm judgment. It could reflect 70. [R. V. G. Tasker, St Matthew (Tyndale NTC), 1961, 206, suggests that the verses may have been marginal comment (subsequently embodied in the text) added after 70 to draw attention to the judgment on Israel for persecuting the Christians. The weakness in this suggestion is of course the lack of any textual evidence.] On the other hand, it need not. One must decide on the evidence of the distinctive features in Matthew’s apocalypse in chapter 24.
The first observation to be made is how few these are. As K. Stendahl says, ‘He does not have any more explicit references than Mark to the Jewish War or the withdrawing of the Christians from Jerusalem’. [PCB, 793. Cf. Marxsen, Mark the Evangelist, 198, who himself has no doubt that Matthew is later than 70: ‘If we begin by inquiring into the time of Matthew’s composition, we encounter the startling fact that chap. 24 is scarcely ever used in evidence. It is rather on the basis of 22.7 that the Gospel is assumed to have originated after AD 70.’] Apart from minor verbal variations he follows the tradition common to Mark, with only the following differences of any significance:
1. In 24.3, the purpose of the discourse is broadened to answer the disciples not merely on the date of the destruction of the temple (‘Tell us, when will this happen ?’) but on the theme to which the chapter (and the one following) is really addressed: ‘And what will be the signal for your coming and the end of the age?’ It is significant, however, that the former question does not drop out, as might be expected (especially since Matthew has no more answer to it than Mark) if at the rime of writing it now related to the past whilst the parousia was still awaited.
2. In 24.9-14, the prophecies of persecutions ahead found in Mark 13.9-12 are omitted, being placed by Matthew in Jesus’ mission charge to the disciples during the Galilean ministry (10.17-21). Whatever the motives for this, the effect is to see the predictions fulfilled earlier rather than later, and evidently they are not intended by Matthew to have any reference to the sufferings of the Jewish war. In their place Matthew has warnings against division and defection within the church, which are presumably relevant to the state of his own community but have no bearing on the question of date.
3. In 24.15, the cryptic reference to ‘the abomination of desolation’ is specifically attributed to the prophet Daniel (which was obvious anyhow), and Matthew has the neuter participle ἑστος for the masculine ἑστηκ ὀτα (as the grammar demands), and ἐν τλοπῳ ἁγλιῳ for the vague ὅπου οὐ δεῖ. Despite the lack of article, ‘(the) holy place’ must mean the temple (evidently intended by Mark’s allusion), and the choice of phrase may again reflect the scriptural background already referred to:
How long will impiety cause desolation, and both the holy place and the fairest of all lands be given over to be trodden down? (Dan. 8.13)
They sat idly by when it [Jerusalem] was surrendered,
when the holy place was given up to the alien (I Mace. 2.7).
Yet none of Matthew’s changes affects the sense or makes the application more specific (in fact the neuter participle does the opposite). Again he does not mention the reference in Daniel to the cessation of the daily sacrifices. If Matthew intended the reader to ‘understand’ in the prediction events lying by then in the past he has certainly given him no help. Moreover, as Zahn said long ago [INT,571.], in view of Matthew’s appeal to conditions in Jerusalem ‘to this day’ (27.8; cf. 28.15), one would have expected him of all people to draw attention to the present devastation of the site.
4. In 24.20, there occurs the only other change in the decisive paragraph about Judaea, with the addition of the words in italics:
Pray that it may not be winter when you have to make your escape, or Sabbath.
‘When you have to make your escape’ merely specifies what must be meant in Mark. The reference to the sabbath could again contain an allusion back to the fact that when the faithful of Judaea took to the hills after the original ‘abomination of desolation’ their first encounter with the enemy was on the sabbath and because of scruples which they later abandoned they were massacred without resistance (I Mace. 2.29-41). But it is more likely to refer to the obstacles to movement on the sabbath for Jewish Christians who were strict observers of the law. In any case it bespeaks a primitive Palestinian milieu and a community-discipline stricter than that recommended in Matthew’s own church (cf. Matt. 12.1-14). It is certainly not an addition that argues for a situation after 70. Indeed it is one of those points of difference where, unless one is committed to over-all Markan priority, it looks as though Mark has omitted an element in the tradition no longer relevant for the Gentile church.
5. Matthew’s material without parallel in the Markan tradition (24.26-8; 24.37-25.46) has no reference to the fall of Jerusalem but, like the additional signs of the parousia in 24.30f., solely to ‘the consummation of the age’. Yet his version of the ‘Q,’ material in 24.26, ‘If they tell you, “He is there in the wilderness”, do not go out’, clearly shows that in his mind the scene is still in Judaea (in the Lukan parallel in 17.23 it could be anywhere). It is significant therefore that in 24.29, ‘the distress of those days’ (i.e., on the assumption of ex eventu prophecy, the Judaean war) is to be followed ‘immediately’ (εὐθέως) by the coming of the Son of Man, whereas in Mark 13.24 it is promised vaguely ‘in those days, after that distress’. Normally Matthew edits out (if this is the relationship between them) Mark’s incessant use of εὐθύς. Never elsewhere does he alter a Markan phrase to εὐθέως. [Though he adds the word, without significant change of sense, in 27.48. B. W. Bacon, ‘The Apocalyptic Chapter of the Synoptic Gospels’, JBL 28, 1909, a, argued (without a shred of evidence) that εὐθύς could ‘easily’ have been in the original text of Mark 13.24 – though this would still not explain why Matthew retained it.] This makes it extraordinarily difficult to believe that Matthew could deliberately be writing for the interval between the Jewish war and the parousia. So conscious was Harnack [Chron., 653f.] of this difficulty that he insisted that the interval could not be extended more than five years (or ten at the very most), thus dating Matthew c. 70-5. He would rather believe that Matthew wrote before the fall of Jerusalem than stretch the meaning of εὐθέως further. It seems a curious exercise to stretch it at all! Even E. J. Goodspeed, [E. J. Goodspeed, An Introduction to the New Testament, Chicago 1937, 176.] who put Luke at 90, said of Matthew, ‘A book containing such a statement can hardly have been written very long after AD 70’ (though his elastic was prepared to extend to 80). The only other way of taking this verse retrospectively is to say that ‘the coming of the Son of Man’, though not ‘the consummation of the age’, did occur with the fall of Jerusalem. [Cf. A. Feuillet, ‘La synthese eschatologique de saint Matthieu’, RB 55-6, 1949-50, 340-64, 62-91, 18o-211 (especially 351-6); ‘Le sens du mot parousie dans l‘evangile de Matthieu’ in W. D. Davies and D. Daube (edd.), The Background of the New Testament and its Eschatology: In Honour of C. H. Dodd, Cambridge 1956, 261 —80; Gaston, No Stone on Another, 484; also (somewhat differently) France, Jesus and the OT, 227-39; and G. B. Caird, Jesus and the Jewish Nation (Ethel M. Wood Lecture), 1965.] But it is a fairly desperate expedient to seek to distinguish these two (joined by Matthew by a single article in 24.3) in face of the usage of the rest of the New Testament.
Finally, Matthew retains unaltered Jesus’ solemn pronouncement, ‘The present generation will live to see it all’ (24.34), preserving also (as the equivalent of Mark 9.1) the saying, ‘There are some standing here who will not taste of death before they have seen the Son of Man coming in his kingdom’ (16.28). Most notoriously of all, he has, alongside the apocalyptic material from the Markan tradition which he sets in his mission charge, the promise, ‘Before you have gone through all the towns of Israel the Son of Man will have come’ (10.23). [This again could well be a saying which Mark has omitted from the common tradition as irrelevant to his Gentile readers.] If, on the usual reckoning, the evangelist is writing some 50-60 years after the death of Jesus, it is surely incredible that there are no traces of attempts to explain away or cover up such obviously by then unfulfillable predictions. One would equally expect modifications to prophecies after the non-event.
Indeed, I think that it needs to be asked much more pressingly than it is why warnings and predictions relating to the crisis in Judaea should have been produced or reproduced in such profusion after the events to which they referred. Just as Jesus’ parables were reapplied to the life of the church and to the parousia when their original setting in the crisis of his ministry was no longer relevant [Cf. C. H. Dodd, The Parables of the Kingdom, 1935, and J. Jeremias, The Parables of Jesus, ET 3I972.], so one might suppose that instructions given, or pointed up, for earlier situations would, if remembered at all afterwards, have become related more timelessly to the End. Alternatively, if subsequent occasion required, they might have been brought out and subjected to recalculation (the way that Jeremiah’s unfulfilled prediction of the seventy years’ duration of the exile is reapplied ‘on reflection’ in Dan. 9.1-27). But the period of composition commonly assigned to both Matthew and Luke (80-90) was, as far as we know, marked by no crisis for the church that would reawaken the relevance of apocalyptic. [B. H. Streeter, The Four Gospels, 1924, 516-23, associated it with the rumours of the return of Nero redivivus. But there is no other evidence connecting this myth with the gospel tradition, even if we could date it with certainty (see pp. 245f. below). Moreover Streeter’s argument depends on his omission (with the Sinaitic Syriac) of ‘standing in the holy place’ from Matt. 24.15.] I fail to see any motive for preserving, let alone inventing, prophecies long after the dust had settled in Judaea, unless it be to present Jesus as a prognosticator of uncanny accuracy (in which case the evangelists have defeated the exercise by including palpably unfulfilled predictions). It would seem much more likely, as the form critics have taught us to expect, that these sayings, like the rest, were adapted to the use of the church when and as they were relevant to its immediate needs.
There is one other passage common to Matthew and Luke which it will be convenient to mention briefly before turning to Luke. This refers to the murder of Zechariah ‘between the sanctuary and the altar’. In Matthew (23.35), but not Luke (i 1.51), he is called ‘son of Berachiah’, and this has been held [E.g. by J. Wellhausen, Einleitung in die drei ersten Evangelien, Berlin 2I9II, 118-23. To the contrary, Zahn, INTII, 589f.] to contain an allusion to the murder by two Zealots ‘in the midst of the temple’ of a certain Zacharias, son of Baris (v.L, Beriscaeus) in 67-8. [Josephus, BJ4, 334-44.] But the identification rests on a rather remote resemblance of names, and this Zacharias, not being a priest, would have been unlikely to have been ‘between the sanctuary and the altar.’ On Jesus’ lips it makes entirely good sense to interpret the reference, with the Gospel according to the Hebrews [According to Jerome, in Matt. 23.35.], as being to the murder of Zechariah son of Jehoiada the priest (II Chron. 24.20-2), whom Matthew, like some of the rabbis, has evidently confused with Zechariah son of Berechiah, the prophet (Zech.i.i). [So e.g. A. H. McNeile, St Matthew, 1915; J. M. Creed, St Luke, 1930; H. St J.Thackeray, Josephus, Loeb Classical Library, 1928, ad locc.] In any case it is far too uncertain a piece of evidence to carry any weight by itself.
Finally, then, we turn to Luke. His parallel to the Markan apocalypse must be taken closely with another earlier passage relating to Jerusalem and it will be convenient to set them out together.
When he came in sight of the city, he wept over it and said, ‘If only you had known, on this great day, the way that leads to peace! But no; it is hidden from your sight. For a time will come upon you, when your enemies will set up siege-works against you; they will encircle you and hem you in at every point; they will bring you to the ground, you and your children within your walls, and not leave you one stone standing on another, because you did not recognize God’s moment when it came’ (19.41-4).
But when you see Jerusalem encircled by armies, then you may be sure that her destruction is near. Then those who are in Judaea must take to the hills; those who are in the city itself must leave it, and those who are out in the country must not enter; because this is the time of retribution, when all that stands written is to be fulfilled. Alas for women who are with child in those days, or have children at the breast! For there will be great distress in the land and a terrible judgment upon this people. They will fall at the sword’s point; they will be carried captive into all countries; and Jerusalem will be trampled down by foreigners until their day has run its course (21.20-4).
The latter passage replaces, and at some points echoes, that in Mark 13.14-20 beginning, ‘But when you see “the abomination of desolation” …’. Its relation to it must be considered shortly. But first let us look at what Luke himself actually says.
At first sight it seems clearly to be composed (or at any rate pointed up) in the light of the siege of 68-70. For here indeed is the greater specification we expect but fail to find in Matthew. The details, says Kummel, ‘correspond exactly to the descriptions which contemporary accounts offer of the action of Titus against Jerusalem’.[INT, 150. Similarly, among many others, R. Bultmann, The History of the Synoptic Tradition, ET Oxford 1963, 123.]
Yet this is far from indisputable. In an article written now thirty years ago but strangely neglected, Dodd argued strongly and circumstantially that no such inference could be drawn. [C. H. Dodd, ‘The Fall of Jerusalem and the “Abomination of Desolation” ‘, JRS, 1947, 47-54; reprinted in his More New Testament Studies, Manchester 1968, 69-83.]
These operations are no more than the regular commonplaces of ancient warfare. In Josephus’s account of the Roman capture of Jerusalem there are some features which are more distinctive; such as the fantastic faction-fighting which continued all through the siege, the horrors of pestilence and famine (including cannibalism), and finally the conflagration in which the Temple and a large part of the city perished. It is these that caught the imagination of Josephus, and, we may suppose, of any other witness of these events. Nothing is said of them here. On the other hand, among all the barbarities which Josephus reports, he does not say that the conquerors dashed children to the ground. [The youths under the age of seventeen were sold into slavery (BJ 6.418).] The expression ἐδαφιοῦσιν σε καὶ τὰ τἐκνα σοῦ ἐν σοίis in any case not based on anything that happened in 66-70: it is a commonplace of Hebrew prophecy. [Op-cit.49f. (74f.)]
Dodd then proceeds to show in detail how all the language used by Luke or his source is drawn not from recent events but from a mind soaked in the Septuagint.
So far as any historical event has coloured the picture, it is not Titus’s capture of Jerusalem in AD 70, but Nebuchadrezzar’s capture in 586 BC. There is no single trait of the forecast which cannot be documented directly out of the Old Testament. [Ibid., 52 (79). Cf. earlier (though Dodd does not refer to it) C. C. Torrey, The Composition and Date of Acts (Harvard Theological Studies, I), Cambridge, Mass., 1916, 691., who concludes: ‘Every particle of Luke’s prediction not provided by Mark was furnished by familiar and oft-quoted Old Testament passages.’]
It has justly been said that if this article had appeared in the Journal of Theological Studies rather than the Journal of Roman Studies New Testament scholars would have taken more notice of it. It is still ignored in Kummel’s extensive bibliography, and no recognition is given to the case it argues. Interestingly, it had no influence on Reicke’s article cited above [Though it is cited with approval by Pedersen, ST 19, 168.], which independently reaches much the same position.
But the absence of any clear reference to 70 does not settle the question of what Luke is doing in relation to the Markan material. Indeed on this Dodd and Reicke come to opposite conclusions. Reicke, with the majority of critics, thinks that Luke 21.20-4 is an editing of Mark: Dodd holds that it is independent tradition into which the evangelist has simply inserted verbatim two phrases from Mark: ‘Then those who are in Judaea must take to the hills’ (21.21 a) and ‘Alas for women who are with child in those days, or who have children at the breast!’ (21.23a). [In 21.20 the reference to the ‘desolation’ of Jerusalem derives, Dodd argues, not from Mark (and Daniel) but from the frequent use of the word in this context by Jeremiah.] The latter alternative seems to me the more probable [Cf. my Jesus and His Coming, 1957, 122-4. Similarly T. W. Manson, The Sayings of Jesus, 1949, 328-37; Taylor, Mark, 512; Gaston, op. cit., 358.], if only because the introduction of ‘Judaea’ in 21.21a upsets the reference of ἐν μέσω αὐτῆς in 21b, which must be to Jerusalem αὐτῆς 21.20). But, whether or not this was material which Luke had prior to his use of the Markan tradition, he has clearly now united the two. Is the effect of their combination to suggest or to require a later date?
Luke has preferred to concentrate on the destruction of the city rather than the temple, the last reference, veiled or unveiled, to the sanctuary having disappeared, despite his retention of the opening question about the fate of the temple buildings (2I.5-7). [Luke broadens the audience (‘some people were saying’) but not, like Matthew, the question.] The answer therefore is even less precise, though there is now a definite reference to devastation and not simply to desecration. Reicke indeed argues that by replacing Mark’s ‘abomination of desolation standing where he ought not’ with ‘Jerusalem surrounded by armies’ Luke actually makes it more certain that he is not writing after the event. For
if the Gospel of Luke is supposed to have been composed after the historical siege of Jerusalem in AD 70, the evangelist must be accused of incredible confusion when he spoke of flight during that siege, although the Christians were known to have left Judaea some time before the war even began in AD 66. [‘Synoptic Prophecies’, 127.]
The last clause goes beyond the evidence, for Luke may not have known it.
Nevertheless the point stands against a vaticinium ex eventu. Things did not in fact turn out like that. Indeed they could not, for there was no escaping once the city had been encircled.
But the saying about getting out and not going back in, which in Luke 21.21 is applied to the city, has probably nothing in origin to do with a siege. In Mark and Matthew it relates to a man’s house, as in the closely parallel saying which Luke himself preserves in 17.31:
On that day the man who is on the roof and his belongings in the house must not come down to pick them up; he, too, who is in the fields must not go back.
As when Mattathias and his sons ‘took to the hills, leaving all their belongings behind in the town’, the context seems more likely to be local harassment than a military siege. If, as is entirely possible, Jesus himself did utter some such urgent exhortations to vigilance and rapid response, [Cf. the whole of Luke 17.2 0-3 7; also 12.35-13.9; Mark 13.33-6; Matt. 24.37-25.30.] they were almost certainly independent of any programme of future events. If subsequently they were incorporated by the church into instructions for Christians in Judaea and combined with other words of his about the desolation of the city, [Cf. Matt.23.37-9 = Luke 13.34f Without Mark’s story of the widow’s mite, Matthew makes this saying lead directly into the programme of ch. 24.] this does not mean that they were edited after or even during the war. In fact there is nothing that requires them to be restricted to the events of the latter 60s. The ‘wars and rumours of wars’ between nations ἔθνος ἐπ΄ ἔθνος and kingdoms (Mark 13.71. and pars) have no obvious reference to Vespasian’s campaign against the Jewish extremists. [Cf. Reicke, op. cit., 130f., who instances rather the wars of Rome against the Parthians in 36 and 55 which inspired the Jewish nationalists to violent activities.]In Luke this is ‘wars and insurrections’ (ἀκαταστασίας ) (21.9). The latter word appears here to have the same meaning as στἀσις, which is used by Luke (23.19, 25), as by Mark (15.7), of the Barabbas incident, and in the context (cf. Luke 21.8) seems to refer to risings led by messianic pretenders, such as he also records from the 40s and 50s in Acts (5.36f.; 2I.38). [στἀσις refers also, of course, to purely civil disturbances (Acts 19.40; 23.10; 24.5), as presumably do the ἀκατασταςίαι II Cor. 6.5.] There is no ground for assuming that he is alluding specifically to the Jewish revolt of 66-70, let alone writing after it.
None of this in itself decides the issue of when the synoptic gospels were written. In fact, despite the arguments he puts forward, Dodd (followed by Gaston and Houston) thinks that Luke and Matthew were composed after 70. Reicke, although regarding Luke 21 as secondary to Mark, concludes that ‘Matthew, Mark and Luke wrote their Gospels before the war began’. [Op. cit., 133.] That issue must be considered in due course on its own merits. The one conclusion we can draw so far is to agree with Reicke’s opening statement that it is indeed ‘an amazing example of uncritical dogmatism’ that ‘the synoptic gospels should be dated after the Jewish War of AD 66-70 because they contain prophecies ex eventu of the destruction of Jerusalem’. Indeed on these grounds alone one might reverse the burden of proof, and reissue Torrey’s challenge, which he contended was never taken up:
It is perhaps conceivable that one evangelist writing after the year 70 might fail to allude to the destruction of the temple by the Roman armies (every reader of the Hebrew Bible knew that the Prophets had definitely predicted that foreign armies would surround the city and destroy it), but that three (or four) should thus fail is quite incredible. [Wink, USQR 26, 48, poses a similar question to Brandon who wishes to put Mark after 70: ‘Is it really conceivable that Mark should fail to mention, even by allusion in a single instance, an event so traumatic that it is alleged to be the sole motification for his undertaking to write his gospel?’] On the contrary, what is shown is that all four Gospels were written before the year 70. And indeed, there is no evidence of any sort that will bear examination tending to show that any of the Gospels were written later than about the middle of the century. The challenge to scholars to produce such evidence is hereby presented. [C. C. Torrey, The Apocalypse of John, New Haven, Conn., 1958, 86, quoting his earlier book, The Four Gospels, New York 21947.]
But before we can even consider that piece of bravado it is necessary to establish some sort of scale of measurement by which the progress of affairs in the Christian church ‘about the middle of the century’ can be assessed. And the best, indeed the only, way of discovering any fixed points is to turn to the evidence provided by the life and writings of Paul.