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.

X. A Post-apostolic Postscript

Home | Contents | Introduction | Dating: fixing points | Epistle of Barnabas | Shepherd of Hermas | Didache | 1 Clement | Timetable | >

If the canonical books of the New Testament are all to be dated before 70 the question naturally arises: What happens to the space in the last third of the first century previously occupied by so much Christian literature? Is there not an unexplained gap between the end of the New Testament writings and the first productions of the sub-apostolic age? And does not history, like nature, abhor a vacuum?

The possibility, if not the probability, must indeed be faced that there was not a steady stream of early Christian writings but that an intense period of missionary, pastoral and literary activity, culminating in the desolation of Israel and the demise of all the 'pillars' of the apostolic church except John, was followed by one of retrenchment and relative quiescence. A 'tunnel period' in which there was no evidence of literary remains would therefore be perfectly explicable - in fact more explicable, and less extended, than that which the traditional dating has presupposed prior to the emergence of the gospels in written form.

Yet it may also be that the gap to be accounted for is largely artificial. It may have been created by pushing the sub-apostolic literature late so as to leave room for meeting the supposed requirements of New Testament development. In other words, because the latter part of the first century is already occupied, other documents must belong to the second. Remove the initial presupposition and what happens? A look at the dating of some of the earlier sub-canonical literature will help to test and to set in perspective our previous conclusions.

The first thing that strikes one is the still greater lack in this twilight area of any fixed points or solid obstacles. Indeed there can really only be said to be two which are generally accepted, and they are by no means as secure as is usually assumed.

The first is the first epistle of Clement to the Corinthians, which is regularly dated in 95 or more often 96. Clement's episcopate at Rome, it is agreed, roughly coincided with the nineties of the first century, [Eusebius, HE 3.15 and 34, dates it as 92-101, but Hippolytus, probably more correctly, 88-97. Cf. Lightfoot, API. 1, 67 and 343; Harnack, Chron., 718. There is unammity among the various episcopal lists that it lasted nine years.]
and the assumption is that the opening reference to 'the sudden and repeated calamities and reverses which have befallen us' (1.1) and the admonition 'we are in the same lists [as the martyrs Peter and Paul], and the same contest awaiteth us' (7.1) refer to persecution of Christians at the close of Domitian's reign. Yet this is an assumption, however widely accepted. For the time being, however, let us leave it in possession of the field.

The second fixed point is the martyrdom of lgnatius and dependent upon it the dating of his epistles shortly before, and the epistle of Polycarp (at any rate in part) shortly after. [Cf. Lightfoot, AF II. 1, 583; and also P. N. Harrison, Polycarp''s Two Epistles to the Philippians, Cambridge 1936.]
Lightfoot was able to place this event 'with a high degree of probability ... within a few years of ad 110, before or after', [AF II.1,30.]
and except for those who question the genuineness of the entire Ignatian corpus there is no serious disagreement with this estimate. Harnack [Chron., 406.]favoured 110-17, Streeter [PC, 273-6.] 115.

A third possible stable point of reference is provided by the so-called Epistle of Barnabas, which is held by the majority of scholars, with greater or lesser assurance, to be datable around 130. But here there is much more dispute, and a consideration of it will introduce discussion of a number of factors relevant to the larger scene.

This Epistle is noteworthy as the first Christian document explicitly to mention the fall of Jerusalem in the past tense:

Because they [the Jews] went to war it [the temple] was pulled down by their enemies (16.4).

This is the kind of statement conspicuously absent from the New Testament and it clearly dates the Epistle after 70. But since there is no mention of the final Jewish rebellion and the reconstruction of Jerusalem as a pagan city under Hadrian (132-5), it is generally agreed that it is to be placed somewhere between these (wide) limits. But where?

Some have seen in the context of the same passage a reference to proposals by the Romans to rebuild the temple in c. 130:

Furthermore he saith again: 'Behold they that pulled down this temple themselves shall build it.' So it cometh to pass; for because they went to war it was pulled down by their enemies. Now also the very servants of their enemies shall build it up (16.3f.).
[There is a textual variant, 'they and the servants of their enemies', but the text given is preferred by almost all editors.]

But there is much uncertainty about such a plan, if it existed. Some [Cf. L. W. Barnard, 'The Date of the Epistle of Barnabas: A Document of Early Egyptian Christianity', JEA, 44, 1958, 101-7.]
refer it to a promise in the early days of the emperor Hadrian (117+) to rebuild the Jewish temple.
[Epiphanius, De mens. et pond. 14. But the account is highly unreliable.]
Yet the evidence for this is extremely sketchy [Cf. Gen. Rabbah, 64.10: 'In the days of Rabbi Joshua ben Hananiah the [Roman] State ordered the Temple to be rebuilt. Pappus and Lulianus [sic?] set tables from Acco as far as Antioch and provided those who came up from the Exile [i.e., Babylon] with all their needs'; H. Freedman and M. Simon (edd.), Midrash Rabbah: Genesis II, 1939,579f.]
and to see a reference to the Jews themselves in 'the servants of their [the Jews'] enemies' is very difficult. Schurer concludes, [HJP I,535.]'The historical value of the legend is nil', and Prigent, 'One must abandon this explanation and its promoters.' [P. Prigent, Les Testimonia dans Ie christianisme primitif: I'Epitre de Barnabe I-XVI et ses sources, Paris 1961, 76.] Others
[E.g. Harnack, Chron., 423-7; Windisch, 'Der Barnabasbrief in Lietzmann's HNT, Erganzungsband III, Tubingen 1920, 388f.; Schurer, HJP I, 536.]
see an allusion to the building of a pagan temple on the site in Hadrian's new city of Aelia Capitolina, which is said by Dio Cassius to have been planned in 130 before the Jewish revolt and indeed to have been its occasion.
[Hist. 69.12.1f.; contrast Eusebius, HE 4.6.4, who places it after the rebellion.]
Again, the evidence for the site of this temple is very doubtful, [P. Prigent, La fin de Jerusalem, Neuchatel 1969, 121f.]
and it is surely incredible that if this is the reference it should not be seen by a Christian writer as a sign of judgment on the infidelity of the Jews. [So Prigent, Testimonia, 78.]

But in fact all such speculation is beside the point. It is clear from the subsequent context (16.7-10) that the new temple that is being built is a spiritual one in the heart: its agents are Christians, viewed as the loyal subjects of the Roman empire. As Lightfoot argued long ago, [AF 241.] 'the passage has no bearing at all on the date' of the Epistle, and this is agreed by the two latest commentators. [R. A. Kraft in R. M. Grant (ed.), The Apostolic Fathers III, New York 1965, 42f.;
and P.
Prigent, L'Epitre de Barnabe, Paris 1971, 191.]
Naturally the Jews hoped all along that the temple would be rebuilt physically, [Cf. e.g. I Bar.4.21-5.9; II Bar.6.9; 68.4-6; II Esd.10.41-55; Orac.Sib. 5.414-33.]
but the response of this writer is to see all the ordinances of Judaism fulfilled in Christ in a spiritual manner (6-17). And this is true whatever the date.

The other reference to a possible dating is in 4.4-6, where there is an obscure allusion to contemporary history in the traditional apocalyptic mode:

Ten reigns shall reign upon the earth, and after them shall arise a little king, who shall bring low three of the kings under one. In like manner Daniel speaketh concerning the same: And I saw the fourth beast to be wicked and strong and more intractable than all the beasts of the earth, and how there arose from him ten horns, and from these a little horn, an excrescence, and how that it abased under one three of the great horns. Ye ought therefore to understand.

It must be conceded at once that it is hazardous to build anything firm on this. [Cf. the discussion in Harnack, Chron; 418-23, whose conclusion is 'non liquet'.]
The reference is not as clear as it is even in the comparable passage Rev.17.7-18. But it is evident that the fourth beast of Daniel stands here for the Roman empire, and the 'little horn' who is 'from' the ten kings is probably again Nero redivivus. Prigent [Testimonia, 151f.; Epitre Barnabe, 97.]supports Lightfoot [AF 240f.; 1.2, 509-12.] in saying that the most likely reference of 'the three kings under one' is to the three Flavian emperors, Vespasian and his two sons Titus and Domitian, who shared the rule even during Vespasian's lifetime.
[Cf. the very similar expressions in II Esd.11.29f.; 12.22f.]
The passage is therefore to be dated before the death of Vespasian in 79, since he has still to meet his doom, with his sons, at the hand of the returning Nero (which he did not). [For another such unfulfilled prophecy in regard to Titus, who it was predicted would die as soon as he set foot on Italian soil after sacking Jerusalem, cf. Orac. Sib. 5.408-11.]
Whereas, as we have seen, in Revelation in 68 the sixth emperor is on the throne, and by 100 in II Esdras twelve have already reigned, here the tally to date is ten. [In the latter two cases Julius Caesar is already included in the reckoning, as in Orac. Sib. 5.1-51.]

Prigent is at pains to stress that this tells us no more than the date of this particular passage, which indeed he thinks goes back to c. 70. Yet there is no real evidence for supposing that it is not homogeneous with the rest [Contrast, for instance, II Esd.1-2 and 15-16 which are fairly evidently separate from the main body of the book.]
and it fits with what can be gleaned from contemporary Jewish apocalypses. A brief comparison with these will be instructive.

I Baruch (in the Apocrypha) claims to be written in 'the fifth year after the Chaldeans had captured and burnt Jerusalem' in 586 BC (1.2; cf. II Kings 25; Jer.52). Yet it is clear that this is but a thin disguise for the similar action of the Romans in AD 70, and the book, whatever earlier material it may incorporate, thus dates itself in 75. The Jews are urged to 'pray for Nebuchadnezzar king of Babylon and for his son Belshazzar' (sc. Vespasian emperor of Rome and his son Titus) (1.11).
[Contrast the very different attitude towards Domitian in II Esd.11.36-46.]

There are allusions to recent calamities notably absent from the New Testament apocalypses - to parents eating their children in the extremities of the siege (2.3), [Cf.Josephus, BJ 6.201-3. There is no such reference in our accounts of the events of 586.] to the burning of the city (1.2), and to the deportation of captives to Rome (4.6,15f.,31f.; 5.6). The references to the doom of 'Babylon' in 4.30-5 are strikingly similar to those in Rev.18, but here the fall of Rome is seen as direct retribution for the sacking of Jerusalem ('The same city that rejoiced at your downfall and made merry over your ruin shall grieve over her own desolation', 4.33) in a way that we should expect but significantly do not get in Revelation. There is possibly also a reference to the Christians in 4.3, 'Do not give up your glory to another or your privileges to an alien people', corresponding to the reference to the Jews in the Epistle of Barnabas.

Again in parts at any rate of II Baruch (the Syriac Apocalypse of Baruch)
[Charles, AP II, 470-526, from which the translation is taken.]
we seem to be in the same period. There are similarly circumstantial references to the overthrowing of the walls of Jerusalem and the burning of the temple (7.1; 80.3), and to the despoiling of the sanctuary:

And I saw him descend into the Holy of Holies, and take from thence the veil, and the holy ark, and the mercy-seat, and the two tables, and the holy raiment of the priests, and the altar of incense, and the forty-eight precious stones, wherewith the priest was adorned, and all the holy vessels of the tabernacle (6.7). [Cf.Josephus, BJ 148-151, 161f.]

There is also the same prediction of the reversal of judgment upon Rome and particularly upon Vespasian:

The king of Babylon will arise who has now destroyed Zion, and he will boast over the people, and he will speak great things in his heart in the presence of the Most High. But he shall also fall at last (67.7f).

If in the passage from the Epistle of Barnabas we examined earlier there were a Christian riposte to Jewish hopes of a literal restoration of the temple and its worship, it could equally come from this same period. Witness the muted promise given to the apocalyptist:

After a little interval Zion will again be builded, and its offerings will again be restored, and the priests will return to their ministry, and also the Gentiles will come to glorify it. Nevertheless, not fully as in the beginning (68.5f.).

One could go on citing parallels. Thus there is the passage in the Apocalypse of Abraham 27: [The Apocalypse of Abraham, ed. and tr. G. H. Box, 1918. Box contends (xvi) that the description suggests that the events are fairly recent. Similarly D. S. Russell, The Method and Message of Jewish Apocalyptic, 1964, 37, dates it between 70 and 100.]

And I looked and saw: lo! the picture swayed and from it emerged, on its left side, a heathen people, and these pillaged those who were on the right side, men and women and children: some they slaughtered, others they retained with themselves. Lo! I saw them run towards them through four entrances, and they burnt the temple with fire, and the holy things that were therein they plundered.

And there are comparable references in the somewhat later book of II Esdras, especially in 10.21-3:

You see how our sanctuary has been laid waste, our altar demolished, and our temple destroyed. Our harps are unstrung, our hymns silenced, our shouts of joy cut short; the light of the sacred lamp is out, and the ark of the covenant has been taken as spoil; the holy vessels are denied, and the name which God has conferred on us is disgraced; our leading men have been treated shamefully, our priests burnt alive,
[Cf. Josephus, BJ 6.280, who specifically mentions two leading men among the priests who threw themselves into the fire and were burnt together with the holy house.]
and the Levites taken off into captivity; our virgins have been raped and our wives ravished, our god-fearing men carried off, and our children abandoned; our youths have been enslaved, and our strong warriors reduced to weakness. Worst of all, Zion, once sealed with God's own seal, has forfeited its glory and is in the hands of our enemies.

I quote these passages as the contrast with the New Testament is so glaring, and it is surely incredible that if parts of it too came from the same period nothing of the kind is reflected in it. But that the Epistle of Barnabas should come from these traumatic years following the fall of Jerusalem is entirely possible; and several of those who put it in the reign of Hadrian or suspend judgment admit that the internal evidence would naturally suggest an earlier dating. [Cf. Eltester, IDB I, 358: 'Even though the letter suggests an earlier dating'; J. Lawson, A Theological and Historical Introduction to the Apostolic Fathers, New York 1961, 201: 'The evidence would well accord with the early date [70-79].']
Indeed there are many other pointers to this.

From early times the Epistle achieved near-canonical status, being included with the Shepherd of Hermas immediately after the book of Revelation in Codex Sinaiticus. Yet it makes no claims to apostolic authorship characteristic of later pseudepigrapha. In fact the writer disavows even the authority of a 'teacher', addressing his audience simply as 'one of yourselves' (1.8; 4.6,9). There is no reference to any specific order of ministry apart from that of teacher - merely to 'every one who speaks the word of the Lord to you' (19.9) and to those 'in higher station' (21.2), which, however, almost certainly refers in the context to those who are economically better off ('Keep amongst you those to whom ye may do good'). He calls his readers on their own initiative to 'assemble yourselves together and consult concerning the common welfare' (4.10). The whole approach is strikingly different from the second-century appeal in the Ignatian epistles to the authority of the bishop. And unlike these, and still more the epistle of Polycarp, this epistle makes no reference to any other Christian writing, [There are phrases in the 'two ways' material at the end which have been said to reflect knowledge of the New Testament, but these are also in the Didache and are much more likely to have come from the source behind them both (see below). Again, Barn.15.4 represents not a quotation from II Peter 3.8 but a common use of Ps.90.4.]
not even to the epistle to the Hebrews with whose argument it has so many affinities. Its appeal is to the Old Testament and Jewish tradition. [Despite assertions to the contrary, Barn.11.9 cannot establish literary dependence on II Bar.61.7, nor Barn.12.1 on II Esd.5.5. The phrases are part of the common stock of late Jewish imagery.]
The sole apparent reference to any saying of Jesus, though not by name, is in 4.14:
'As it is written. Many are called, but few are chosen.' [Cf.Matt.22.14.] But since this is entirely isolated the commentators are rightly inclined to doubt whether it is a citation from the gospel tradition, seeing in it rather 'a popular Jewish apocalyptic saying also known to (Jesus and) the author of Matthew'.
 [Kraft, ad loc.; similarly Windisch and Prigent. For the idea in Judaism, cf. II Esd.8.1-3. I Tim.5.18 similarly cites as 'scripture' not only Deut.25.4 ('You shall not muzzle a threshing ox') but, apparently. Matt. 10.10 = Luke 10.7 ('The labourer is worthy of his hire'); but the latter too is probably a proverbial Jewish saying rather than an original word of Jesus.]
There are indeed allusions to the gospel tradition about Jesus (5.81.; 7.3; 8.3) (as well as to an unwritten saying) ['Thus, he saith, they that desire to see me, and to attain unto my kingdom, must lay hold on me through tribulation and affliction' (7.11; cf. Acts 14.22).]
but nothing that demands dependence on our written gospels. In 15.9 there is a reference to the ascension having taken place on Easter day, contrary to the tradition in Acts. The epistle contains no developed doctrine of the person of Christ, still less of the Spirit, and remains within the purview of Jewish-Christian theology. The 'gnosticism' of the author is a naive and primitive one, exegetical, ethical and eschatological [Cf. Kraft, op. cit., 22-9.] rather than systematic, heretical or polemical. It stands in strong contrast to the gnostic Gospels of Thomas and Philip and the Gospel of Truth, which really do seem to belong to the mid-second century from which the Epistle of Barnabas has been supposed to come.

In sum, there is nothing here that could not have been written, as Lightfoot said, about 75. It does not begin to reach the heights of the New Testament, and the church was obviously right to exclude it from the canon. But in date there is no reason to think of it coming far behind.

With the Epistle of Barnabas must be considered its nearest associate, the Shepherd of Hermas. This again has regularly been placed in the middle of the second century, but solely on the ground of one piece of external evidence, the Muratorian Fragment on the Canon: [There is a similar reference in the fourth century Liberian Catalogue (cf. Lightfoot, AF I.I, 254) but it evidently goes back to the same common source.]

Very lately in our times Hermas wrote 'The Shepherd' in the city of Rome while his brother Pius, the bishop, was sitting in the chair of the Church of the city of Rome, and therefore it ought to be read; but it cannot, to the end of time, be placed either among the prophets who are complete in number, nor among the Apostles, for public lection to the people in church.

Pius was bishop of Rome from c. 140-155.

The Muratorian Canon is usually held to be the work of Hippolytus and to come from Rome c. 180-200, [Cf. Lightfoot, AF 1.2, 405-13, who dated it before 185-90; Hennecke, NT Apoc. 1,42-5.]
though recently it has been asserted to be not a second-century Roman product but a fourth-century eastern list. [A. C. Sundberg, 'Canon Muratori: A Fourth Century List', HTR 66, 1973, 1-41. His argument in questionable at many points.]
In any case for no other book should we take its unsupported evidence seriously, and it is full of palpable mistakes. With regard to Hermas in particular there are good grounds for questioning its statements. Thus Irenaeus, who resided in Rome less than twenty years after the death of Pius, quotes the opening sentence of the first Mandate of the Shepherd as 'scripture', [Adv. haer. 4.34.2.] which would scarcely be likely if it was known to have been composed within living memory. Not much later Tertullian [De pudic. 20.] strongly disparages Hermas in contrast with Hebrews and it seems improbable that he would not have deployed against it the argument of its late composition. Origen, who freely cites the Shepherd as scripture, attributes it indeed in his Commentary on Romans to the first-century Hermas greeted by Paul in Rom.16.14. In his early work on the Shepherd [Der Hirt des Hermas, Gotha 1868.] Zahn seriously challenged the evidence of the Muratorian Canon, and Edmundson argued that its attribution to the bishop's brother arises from a sheer blunder.
[The Church in Rome, 208-15. Cf. Streeter, PC, 202-13, who however detects anti-Montanist polemic at work. But if the Shepherd was favoured by the Montanists, why does Tertullian slate it as lax in its attitude to post-baptismal sin in comparison with Hebrews?]
It is on the face of it highly unlikely that one who tells us he was a foster-child sold into slavery in Rome (Vis.1.1.1), probably from Arcadia in Greece (Sim.9.1.4), [Cf.J. A. Robinson, Barnabas, Hermas and the Didache, 1920, 27f.]
should have had a brother in Rome called Pius who was head of the church there at the time but whom he never mentions, despite several references to his family. But elsewhere [The Acts of Pastor and Timothy. For the detailed evidence, see Edmundson, Op.cit.,210-2.] we are told that this Pius was 'the brother of Pastor' and it looks very probable that the Shepherd of Hermas, which in its Latin version, possibly dating from the end of the second century and therefore perhaps contemporary with the Muratorian Fragment, is called 'Liber Pastoris' (or the Book of the Shepherd), has by a natural confusion been attributed to the brother of the bishop.

But the external evidence can in any case only be as strong as the internal, and this latter suggests a considerably earlier date. In Vis.2.4.2f. the seer is told:

When then I shall have finished all the words, it shall be made known by thy means to all the elect. Thou shalt therefore write two little books, and shalt send one to Clement, and one to Grapte. So Clement shall send to the foreign cities, for this is his duty; while Grapte shall instruct the widows and the orphans. But thou shalt read (the book) to this city [Rome] along with the elders that preside over the Church.

There is general agreement that unless this reference is a pseudonymous fiction (which there is no other reason to suppose) it must be to the Clement who was bishop of Rome in the last decade of the first century. But Edmundson argues cogently that it relates to a time before Clement held that office. He seems to have an appointment which, in Lightfoot's words 'constituted him, as we might say, foreign secretary of the Roman Church'. [AF I. 1, 348.] But, says Edmundson, [Op. cit., 203f.]

such a description surely implies that at the time Clement was occupying what can only be described as a subordinate position, since he was charged with secretarial duties entrusted to him by others. The particular charge was one that might very well be assigned to a younger member of the presbyterate distinguished among his colleagues for wider culture and greater familiarity with literary Greek. The mere fact that his name is here coupled with that of Grapte, apparently a deaconess, is of itself a proof that the Clement of Hermas' second Vision had not yet become at the close of a long and honoured career the venerated bishop of 96 ad.

Edmundson himself dates the Shepherd of Hermas in the first decade of the reign of Domitian (81-91), [Ibid., 203f., 215-21. W.J. Wilson, 'The Career of the Prophet Hermas', HTR 20, 1927, 21-62, while agreeing that Zahn discredited the testimony of the Muratorian Canon for a date c. 140, opted with him, Salmon and Bigg for one c. 95, but only because of the reference to Clement, who was simply assumed to be bishop at the time. Similarly Streeter, FG, 528, put it c. 100.]
pointing out that the allusions to past sufferings correspond closely with the records of the Neronian persecution (Vis.3.2.1; Sim.8; 9.19.1; 9.28). A fair amount of time has elapsed, which now makes possible a forgiving attitude towards previous betrayals (Vis.2.2.4; Sim.9.26.6). Yet the references to the Christian ministry still presuppose a relatively early period. Thus Vis.3.5.1 speaks of

the apostles and bishops and teachers and deacons, who walked after the holiness of God, and exercised their office of bishop and teacher and deacon
[Note that he does not say bishop, presbyters and deacons, as in Ignatius.]
in purity and sanctity for the elect of God, some of them already fallen on sleep, and others still living.

This passage appears to imply that some of the original generation of church leaders were still alive. [Cf. Sim.9.16, which does not say that all the apostles and teachers had fallen asleep, but speaks of those who had.]
In Sim.9.15.4 there is a distinction made between the first 'foundation' generation, who are represented by ten stones (not, be it noted, as in Revelation, twelve), a second generation of 'righteous men', represented by twenty-five stones, and a third group of thirty-five 'prophets of the Lord and his ministers'. Yet the 'apostles and teachers of the preaching of the Son of God' are not, as we might expect, identified with the first, but are yet a fourth group, forty in number. Except in Sim.9.17.1 the term 'apostles' is still being used at this stage, as in some passages of the New Testament, in the wider sense of missionaries. As in the Didache, to be discussed below, 'prophets' for this writer appear to be in a category apart. Though they do not feature in the list of ministries in Vis.3.5. 1, Mand.11 gives careful criteria for distinguishing true prophets from false. He himself claims the gift of prophecy and with it the authority, like the seer of Revelation, to deliver charges and admonitions to the church and its rulers (Vis.2.2.6; 2.4.21.; 3.8. 11; 3.9.7-10; Sim.9.31.3-6). These are called the 'chiefs' or 'leaders' of the church, the same terms that are used in Hebrews and I Clement. [Cf. p. 209 n. 45 above.] He speaks of 'the elders that preside over (προΐσταμένων) the church' (Vis.2.4.3) in exactly the same way as the Pastoral Epistles (I Tim.5.17; cf. I Thess.5.12; Rom.12.8), and the qualities commended in such 'bishops' are again the same (Sim.9.27.2; cf. I Tim.3.2-7; Titus 1.6-9). There is no sign yet of a monarchical episcopate, even in Rome (Vis.2.4.3), such as would have been enjoyed by Pius I in the mid-second century, though there are indications of struggles for 'first places and a certain dignity' (Sim.8.7.4; cf. Vis.3.9.7). Lightfoot himself recognized that these references suggested an earlier date. [AF 294.; cf. Lawson, op. cit., 224f.] Still there are no direct quotations from or references to Christian books, and its 'spirit' or 'angel' Christology remains within the limits of primitive Jewish-Christianity. [Cf. A. Grillmeier, Christ in Christian Tradition, ET 1965,41-66.]

There thus seems nothing against and everything in favour of the sort of date proposed by Edmundson. In his chronological table [Op. cit., 241.] he finally plumped, without having argued it, for c. 90. But this depends on his belief that Clement did not become bishop of Rome till 92. If with Lightfoot we put that back to 88 or even 86 [AF 1.1.343.] then perhaps c. 85 would be a better estimate. Indeed this is an upper limit; it could be earlier, and may well have been composed over a period. [W. Coleborne, 'A Linguistic Approach to the Problem of Structure and Composition of The Shepherd of Hermas', Colloquium (The Australian and New Zealand Theological Review} 3, 1969, 133-42 (especially 141f.), thinks that it was written, by several hands, between 60 and 100 and that the older parts could well go back to the Hermas at Rome mentioned in Rom. 16.14.! doubt the spread at either end, but the authorship is not impossible chronologically.]
This would allow a twenty years' interval after the Neronian persecution, and put the Shepherd a decade later than the Epistle of Barnabas.

So we turn to the third of three writings that have been closely linked and indeed held to be mutually dependent - the Didache, or the Teaching of the Twelve Apostles, discovered in 1875 and published in 1883. Of no other Christian book have the dating-estimates shown a wider or a wilder swing - ranging between 50 and the fourth century. It is significant that Edmundson, [Op. cit., 187.] who opts for an early dating of everything else, is inclined, though without any adequate discussion, to concur with Bigg (who thought II Peter apostolic!) in placing it at the latter extreme. [C. Bigg, The Doctrine of the Twelve Apostles, 1898.] In Armitage Robinson's words,

It does not seem to fit in anywhere, in either time or place. The community which it presupposes is out of relation to all our knowledge of Church history.... We still ask, Was there ever a Church which celebrated the Eucharist after the manner here enjoined? Was there ever a Church which refused to allow Apostles more than a two days' stay?
[J. A. Robinson, 'The Problem of the Didache', JTS 13,1912,340, reprinted in Barnabas, Hermas and the Didache, 86.]

His conclusion was that it was an artificial and imaginative construction of an ideal apostolic era which affords no reliable historical information of that or any other time. But his own question, 'What after all was the writer's object in composing the book?', remained unanswered. [Barnabas, Hermas and the Didache, 103. Streeter, PC 283, says that his theory is 'one that I cannot bring myself to take seriously'. Yet the theory persists in one form or another. F. E. Vokes, The Riddle of the Didache, 1938, similarly regards it as a fictitious reconstruction, but thinks its object is to present and defend the 'new prophecy' of the Montanist movement as 'apostolic'. He places it at the end of the second century or the beginning of the third (216-20). W. Telfer, 'The "Plot" of the Didache', JTS 45, 1944, 141-51, thinks it is a pseudepigraph which is 'supposed to be the work of the apostolic council of Jerusalem, narrated in Acts 15' (142). C. G. Richardson, Early Christian Fathers (Library of Christian Classics I), 1953, 165, holds it to be 'the first of those fictitious Church Orders which edit ancient material and claim apostolic authorship'. He dates it c. 150.]

But, if we cannot fit it into any period of liturgy or ministry for which we have written evidence, is it possible that it belongs to a period before such documentation? This is the thesis that has boldly been advanced by the massive recent commentary by the French Canadian J.-P. Audet, who concludes that it was composed, almost certainly in Antioch, between 50 and 70.
[La Didache: Instructions des Apotres, 219.]
Coming to this only after reaching my own conclusions on the chronology of the New Testament, I cannot but concur with the remarkably sympathetic review by Kelly [JTS n.s.12, 1961,329-33.] in regarding it as a most persuasive thesis argued in a masterly manner.

If one thing is now probable it is that the material on 'the two ways' which comprises the first half of the Didache (1.1-6.2) is not, as Armitage Robinson, Vokes and others argued, dependent upon the Epistle of Barnabas (18-20) with which it has many close parallels, but that both go back to common Jewish sources. The evidence of the Qumran Manual of Discipline, which preserves very similar material, [1 QS 3.18-4.26; cf. also Test. Asher 1.3-6.6 and, behind all, such passages as Deut.30.15-20; Ps.1; and Prov.2.9-22; 4.18f.]
has tilted the balance again in favour of the latter view. [Cf.J.-P. Audet, 'Affinities litteraires et doctrinales du Manuel de Discipline', RB 59, 1952, 219-38;
122-63; Kraft, op. cit., 4-16; L. W. Barnard, 'The Dead Sea Scrolls, Barnabas, the Didache and the Later History of the "Two Ways" ' in his Studies in the Apostolic Fathers and their Background, Oxford 1966, 87-107.]
The same applies to the much weaker case for the Didache's dependence on Hermas. [Cf. Audet, Didache, 163-6.]
More contentious is the relationship between the Didache and the New Testament. It was characteristic of an earlier period to see every echoed phrase as denoting direct citation and literary dependence. Thus, even the 'Amen' in Did.10.6, says Armitage Robinson, 'doubtless comes from I Cor.14.16', [Barnabas, Hermas and the Didache, 96.]
and Vokes holds that the Didache is based on 'the whole of our New Testament, with the possible exception of the very late II Peter and the unimportant Mark and Philemon'. [Op. cit., 119.]
But there is an increasing tendency to recognize that apparent quotations in this period are far more likely to reflect oral tradition, [Cf. The Oxford Society of Historical Theology, The New Testament in the Apostolic Fathers, Oxford 1905, especially 24-36; Koester, Symptische Uberlieferung bei den apostolischen Vatem, especially 159-241; E. P. Sanders, Tendencies of the Synoptic Tradition, 361.]
and Audet argues that the Didache is completely independent of our written gospels.
[Didache, 166-86. Similarly R. Glover, 'The Didache's Quotations and the Synoptic Gospels', NTS 5, 1958-9, 12-29. To the contrary: Streeter, FG, 507-11; B. G. Butler, 'The Literary Relations of Didache Ch. XVI', JTS n.s.11, 1960, 265-83; J. P. Brown, 'The Form of "Q" known to Matthew', NTS 8, 1961-2, 41f.] Though he believes it to have been written at two stages (by the same hand), even the allusions at the second stage to a written 'gospel' do not, he contends, refer to our Matthew but to a sayings-collection of ethical teachings. Moreover, the passage in 1.3b-5, which contains the closest parallels of all and which with most others he agrees to be an interpolation, [Cf. B. Layton, 'The Sources, Date and Transmission of Didache 1.3b-2.1', HTR 61, 1968, 343-83, who puts it at c. 150 while conceding with Audet that the rest could be early. Glover, NTS 5, 12-29, denies it is an interpolation at all.]
still, he believes (unlike Koester), represents common oral tradition rather than a conflation of Matthew and Luke. The Didache, in other words, is valuable evidence for the prehistory of the synoptic tradition, and particularly of the Matthean: it does not reflect later quotations from it.

None of this can be more than a matter of probability. It is impossible to be dogmatic about the source of quotations. But I find the presumption against literary dependence to be strong. Yet, though dependence could knock out a very early dating (depending of course on the date of the gospels), independence cannot establish it. [Any more than the fact that the Gospel of Thomas may contain parallel tradition independent of our gospels proves that it was written early - though I should be prepared to see the date-span of that, as of a good deal else, reopened.]
The case must rest on the genuine primitiveness of the many indications in the Didache which point to a stage in the life of the church which is still that of the New Testament period itself.

Audet examines these at length [Didache, 187-206.] and we cannot go over his arguments in detail, some of which are more convincing than others. [I cannot see, for instance, that the expression 'hosanna to the house of David' (even if the correct reading in 10.6) is 'almost unthinkable after the events of 70' (Didache, 1891.).]
The prayers and thanksgivings are full of archaic terminology, echoing not only the servant (παῖς) Christology of the early speeches of Acts (Did.9.2f.; 10.sf.; cf. Acts 3.13,26; 4.27,30), later abandoned, but what I have ventured to call 'the earliest Christian liturgical sequence' (Did.10.6; cf. I Cor.16.22-4). ['The Earliest Christian Liturgical Sequence?', JTS n.s.4, 1953, 38-41; reprinted in Twelve NT Studies, 154-7.] In Did.9.1-9 the eucharistic cup still precedes the bread, as in I Cor.10.16 and Luke 22.17-19. Audet argues that the terminology relating to baptism (7.1; 9.5) is similarly primitive, and that the regulations about food (6.3) presuppose a period and a milieu where the dietary question is still genuinely posed:

We are in the first Christian generation born of the Gentile mission, at little distance, it seems, in time if not in space, from I Cor.8-10; Rom.14; Col.2.16, 20-3; and I Tim.4.3.
[Didache, 199.]

Above all, we are in an age of itinerant apostles, prophets and teachers (11-13), where 'apostles' designate not a closed body but any men commissioned as missionary preachers and 'prophets' exercise a high charismatic ministry (10.7; 13.3) more honoured than that of local appointments. It is still the world reflected in such incidents as that of Acts 19.13-20, where strolling Jewish exorcists might be encountered by any congregation. But we are also 'at a point of transition from the ministry of prophets and teachers to that of bishops and deacons' [Ibid., 195. Similarly Streeter, PC, 149-52.] when the former are not available for regular ministry in the local church:

Appoint for yourselves therefore bishops and deacons worthy of the Lord, men who are meek and not lovers of money, and true and approved; for unto you they also perform the service of the prophets and teachers. Therefore despise them not for they are your honourable men along with the prophets and teachers (15.1f.).

This is not the later transition from a presbyteral to a monepiscopal ministry but the much earlier one from the primacy of the charismatic to the recognition (and that by congregational appointment) of an established ordained ministry. It is a transition already presupposed by Philippians (1.1) and the Pastorals in the later 50s. In an astonishingly percipient review-article of Harnack's original edition of the Didache, first published in the Church Quarterly Review of April 1887, C.H.Turner said:
[C. H. Turner, 'The Early Christian Ministry and the Didache', reprinted in his Studies in Early Church History, 1-32 (31).]

The 'Teaching', then, represents a stage or organization intermediate between the Corinthian and the Ephesian letters: parallel, let us say roughly to the Epistle to the Philippians with its earliest mention of episcopi and deacons. It follows from this, that, if the 'Teaching' is to be a factor in the series of the full current of Church development, it ought to be placed about the year 60.

He hastened to guard himself by saying that 'it does not follow that so early a date is inevitable' but said 'a date between 80 and 100 ad is as late as we are prepared to admit'.

With the state of the ministry goes the general theological character of the book. It is content (like the epistle of James) to leave doctrinal issues on one side. There is no polemic (as, for instance, in the Pastorals) against heterodox or gnostic tendencies within the church - merely a concern to maintain a practical mark of difference between Christians and Jews. [Cf. 8.1: 'Let not your fastings be with the hypocrites, for they fast on the second and fifth day of the week; but do ye keep your fast on the fourth and on the preparation (the sixth) day'.]
The final chapter on eschatology breathes much the same apocalyptic atmosphere as I and II Thessalonians (with which it has many parallels) and may represent one of the many fly-sheets of this kind, combining dominical and traditional Old Testament materials, which seem to have been produced by the early church between 40 and 70. [Cf. my Jesus and His Coming, 118-27. I did not at that time recognize that Did.16 might be another such example. It reflects many of the same common features that I noted between I and II Thessalonians and Matt.24 and looks like an important clue which I missed to the development of the parousia doctrine.]
Yet in contrast with the synoptic apocalypses (but not Thessalonians), there is no attempt to fuse this material with predictions of the destruction of the temple or the fall of Jerusalem. This suggests that it is composed well before or well after these events. But, in notable distinction from the Epistle of Barnabas or the Jewish apocalypses of Baruch or II Esdras, there is no hint of any such event lying in the past. It seems much easier to see it as early rather than late. Indeed of the book in general I would agree with the assessment of J. A. Kleist:

If we admit an early date of composition, all the evidence is in favour of it; if we insist on a late date, we have to face a mass of conjectures and hypotheses.
[J. A. Kleist, The Didache, etc., in Ancient Christian Writers 6, eddJ. Quasten and J. C. Plumpe, Westminster, Md., and London 1948, 10. He puts it 'before the end of the first century'; but this is early on the usual New Testament chronology. Cf. H. Chadwick, The Early Church, Harmondsworth 1967, 46f.: 'The situation regarding Church order presupposed in the Didache makes it hard to find any plausible niche for it in early Christian history other than the period between about 70 and 110. It may be odd there, but it is much odder anywhere else.' Similarly Bartlet, HDB V, 449, opted for 80-90. Streeter, PC 279-87, argued that it could not be later than 100 nor earlier than 90; but the lower limit derived from his dating of the gospel of Matthew, on which he held it was dependent.]

In conclusion, I believe that we are here in a thoroughly primitive situation and though the Didache, as Audet says, was probably formed, like the gospels, over an extended period, I should be inclined to put it between 40 and 60 rather than between 50 and 70. For there is little or nothing of the signs of persecution or 'falling away', and with it the concern for consolidation in doctrine and structure, so characteristic of the 60s. If this is its period, then there are a number of features in the New Testament itself which cannot be argued, as they usually are, to demand a date in the latter part of the first century (if not later). Among these may be mentioned the instruction to 'baptize in the name of the Father and the Son and the Holy Spirit' (Matt.28.19; cf. Did.7.1,3); the doxology to the Lord's Prayer (Did.8.2) later incorporated into Matthew (6.13, marg.); the qualifications of bishops and deacons in the Pastorals (I Tim.3.2-13; Titus 1.5-9; cf. Did. 15.1); the instructions about Christian hospitality in the Johannine epistles (II John 10f.; Ill John 8-10; cf. Did.11-12); the use of the term 'the Lord's day' (Rev. 1.10; cf. Did. 14.1); and perhaps the phrase 'the apostles and prophets' in Ephesians and Revelation (Eph.2.20; 3.5; Rev. 18.20; cf. Did. 11.3). In general, if the Didache is really to be set before 60 then the placing of the whole of the New Testament before 70 may turn out not to be the wild hypothesis that at first sight it appeared.

Finally, I return, with some hesitation, to the first epistle of Clement. The consensus for a date of 95-6 is so strong, backed by the magisterial authority of Lightfoot's arguments, [AF I. 1, 346-58.]
that it might seem temerarious merely to question it. 'It has even been said', writes Cullmann, 'that it is the document of ancient Christianity which can be dated with the greatest certainty.' [Peter, 90.]
Yet in fact its basis is a great deal weaker than it appears and the case against it has been powerfully stated by Edmundson, [The Church in Rome, 188-202.]
whose book seems to have been ignored at this point as at others. It is particularly remarkable that he is nowhere referred to in The Primitive Church by Streeter, who would have been at Oxford during his Bampton Lectures.

He begins by agreeing that this epistle, though anonymous, is genuinely by the Clement who became bishop of Rome in the last decade of the century. The sole question is whether he wrote it when he was bishop or at an earlier stage. Edmundson argues strongly that the evidence points to the latter alternative.

At no point in the epistle is appeal made to episcopal authority. Indeed Lightfoot himself says:

Even the very existence of a bishop of Rome itself could nowhere be gathered from this letter. Authority indeed is claimed for the utterances of the letter in no faltering tone, but it is the authority of the brotherhood declaring the mind of Christ by the Spirit, not the authority of one man, whether bishop or pope.
[AF 1.1, 352.]

Not only is the author not writing as a bishop, but the office of bishop is still apparently synonymous with that of presbyter (42.41.; 44.1,41.; 54.2; 57.1), as in the New Testament and all the other writings we have examined. As Streeter says, [PC, 215.]

As in Philippians, bishops and deacons are the names of two kinds of officers. These two offices are spoken of by Clement in a way which excludes the possibility that presbyters is the name of a third and intermediate office. ... There is nothing to call forth surprise in this evidence that in Rome and Corinth a system still prevailed not very far removed from that established by Paul.

If this is really the state of affairs in Rome in 96, then we are faced with a very remarkable transition within less than twenty years to that presupposed by the epistles of lgnatius. For he, while addressing the church of Rome in the salutation of his epistle to it with the utmost veneration, says elsewhere that apart from the three orders of bishop, presbyters and deacons 'there is not even the name of a church' (Trail. 3), and he speaks of bishops, in his sense, as being by then 'settled in the farthest parts of the earth' (Eph.3; cf. Eph.4i.; Magn. 3, 6f.; Trail. 21.; Philad.4; Smyrn.8). It is easier to believe that I Clement, like the Shepherd of Hermas, reflects an earlier period.

The main reason for placing it in the 90s is the assumption that the opening words refer to the persecution of the church under Domitian:

By reason of the sudden and repeated calamities and reverses which have befallen us, brethren, we consider that we have been somewhat tardy in giving heed to the matters of dispute that have arisen among you (1.1).

Often indeed the opening words of I Clement have actually been cited as evidence for a Domitianic persecution. Yet, as Merrill says, [Essays in Early Christian History, 161. Merrill himself argues for a much later date. Cf. p. 334 below, n. 107.]

It is quite preposterous to claim that the innocent sentence with which it starts bears manifest and conscious witness to a persecution of the Church in Rome by Domitian.

The evidence for any such persecution at all is, as we have seen, extraordinarily thin. [Pp. 231-3 above.] But even supposing Clement had just passed through a persecution in which Christians of illustrious rank had suffered, and with whom as bishop he must have had intimate relations, is it conceivable, Edmundson asks, that

none of their examples should have been brought forward, but only those of an already distant persecution, whose memory more recent events must have tended to throw into the background?
[Op. cit., 191.]

Rather, he contends, 'the sudden and repeated calamities and reverses' which have befallen 'us' refer to the chaotic political situation in Rome during the year 69. He quotes again Philostratus' Life of Apollonius of Tyana for the impact of the successive shock-waves of that fateful year:

Galba was killed at Rome itself after grasping at the Empire; Vitellius was killed after dreaming of empire; Otho, killed in lower Gaul, was not even buried with honour, but lies like a common man. And destiny flew through all this history in one year.
[Vit. Apol. 5.13 (tr. Phillimore, II, 58).]

I Clement, he argues, was written in the early months of 70.

I confess that when I first read that I thought that if he can persuade me of that he can persuade me of anything. But I am convinced that his case merits the most serious consideration.

The Epistle, he says, presupposes that the temple sacrifices in Jerusalem are still being offered:

Not in every place, brethren, are the continual daily sacrifices offered, or the freewill offerings, or the sin offerings or the trespass offerings, but in Jerusalem alone. And even there the offering is not made in every place, but before the sanctuary in the court of the altar; and this too through the high-priest and the aforesaid ministers (41.2).

Lightfoot [AFI. 2, 124f.] maintained that this provides no evidence of dating, since Josephus, writing in 93, also speaks of the sacrificial system in the present tense. [Ant. 3.224-57.] But Josephus is giving a summary description of the Old Testament ordinances contained in the Mosaic Law. Clement is appealing, like the author to the Hebrews, to actual practice. He claims its divine sanction for the good ordering of the Christian liturgy, and this could hardly fail to have been undermined by its total disruption. The parallel therefore is far from exact. And the same applies to the other passages that Lightfoot adduces: the Epistle of Barnabas 71., which is concerned with the typology of Old Testament sacrifice fulfilled in Christ, and the Epistle to Diognetus 3, which contrasts the presuppositions behind Greek, Jewish and Christian understandings of worship. Yet one must admit that this argument cannot in itself be decisive or so important as Edmundson claims.

More significant is his contention that Clement's references to the Neronian persecution point to events still fresh in the memory:

But, to pass from the examples of ancient days, let us come to those champions who lived nearest to our time. Let us set before us the noble examples which belong to our generation. ... Let us set before our eyes the good Apostles [Peter and Paul]. ... Unto these men of holy lives was gathered a vast multitude of the elect, who through many indignities and tortures, being the victims of jealousy, set a brave example among ourselves. By reason of jealousy women being persecuted, after that they had suffered cruel and unholy insults as Danaids and Dircae, safely reached the goal in the race of faith, and received a noble reward, feeble though they were in body (5.1-6.2).

He comments:

If anyone were to read those paragraphs for the first time without any presuppositions or arrière-pensées, would they doubt that they told of scenes of horror which not only the author but all those in whose name he wrote had literally before their eyes, and which still haunted the minds of the witnesses?
[Op. cit. 191.]

This, I believe, is a fair observation, though again it cannot be decisive.

Furthermore, the metaphor in the subsequent words, 'we are in the same lists, and the same contest awaiteth us' (7.1), which takes up that of the 'athletes' or champions of the faith in 5.1f., need have no reference to renewed persecution, whether in Rome or Corinth, but, as in the New Testament generally (I Cor.9.24-7; Heb.12.1f.; cf. II Clem.7, 20), may be a summons to the common Christian struggle. Indeed, in very similar words Paul had called the Philippians to 'contend as one man for the gospel faith', saying: 'You and I are engaged in the same contest: you saw me in it once, and, as you hear, I am in it still' (Phil. 1.27-30). But we do not conclude from that that they too are in prison. Similarly, the prayer in I Clem.59.4, 'release our prisoners', in which Streeter [FG, 528; PC, 201.] saw a reference to the Domitianic persecution, may, like the clauses on each side of it, 'feed the hungry', 'raise up the weak', be entirely general - or could equally well allude to the situation Edmundson envisages in early 70, when the author of Revelation was among those in detention.

There are, however, two main passages which have regularly been held to presuppose a later date. The first is 44.1-3:

Our apostles [I.e., in all probability, Peter and Paul, who were subsequently regarded as joint founders of the churches of Corinth and Rome.]
knew through our Lord Jesus that there would be strife over the name of the bishop's office. For this cause, therefore, having received complete foreknowledge, they appointed the aforesaid persons
[viz. bishops and deacons; cf. 42.4]
and afterwards they laid down a rule
[Following the Latin 'legem dederunt', which is probably the sense of ἐπινομήν (the reading of Codex Alexandrinus). Lightfoot amended to ἐπιμονήν, 'provided a continuance'. But it does not affect the argument here.]
that if these should fall asleep, other approved men should succeed to their ministration. Those therefore who were appointed by them, or afterward by other men of repute with the consent of the whole Church, and have ministered unblameably to the flock of Christ in lowliness of mind, peacefully and with all modesty, and for a long time have borne a good report with all - these men we consider to be unjustly thrust out from their ministration.

It is however a fallacy to suppose that a second- or third-generation ministry implies a span of two or three generations. The first presbyters (by definition 'elderly') could have been appointed by Peter in Rome in the mid-50s (if not the mid-40s) and by Paul in Corinth in the early 50s. Even by 70 there must have been many subsequent creations and some of these men could have been long established in office. (I recently took part in the consecration of a new bishop of Woolwich, and by the end of the service there were present four holders of that see, my predecessor and I and two successors, all within a span of less than twenty years!) Nor does the reference in 63.3 to the Roman delegates as 'faithful and prudent men that have walked among us from youth unto old age unblameably' necessarily mean that they had been Christians all that time - even though this would not have been impossible. For, according to Acts 2.10, there were converts from Rome on the day of Pentecost, and in Rom. 16.6f. Paul greets Andronicus and Junia(s) as eminent among the apostles, adding: 'They were Christians before I was.' The other passage is in 47.1-6:

Take up the epistle of the blessed Paul the Apostle. What wrote he first unto you in the beginning of the Gospel (ἐν ἀρχῆ τοῦ εὐαγγελίου)? Of a truth he charged you in the Spirit concerning himself and Cephas and Apollos, because that even then ye had made parties. Yet that making of parties brought less sin upon you; for ye were partisans of Apostles that were highly reputed, and of a man approved in their sight. But now mark ye, who they are that have perverted you and diminished the glory of your renowned love for the brotherhood. It is shameful, dearly beloved, yes utterly shameful and unworthy of your conduct in Christ, that it should be reported that the very steadfast and ancient (ἀρχαίαν) Church of the Corinthians, for the sake of one or two persons, maketh sedition against its presbyters.

This has been interpreted to mean that the church of Corinth was by the time of writing regarded as an 'ancient' foundation. But evidently in the context the meaning of ἀρχαίαν is determined by the phrase 'the ἀρχή of the Gospel', which is precisely that used by Paul to the Philippians of the period when he first preached to them - after an interval of only a decade (Phil.4.15; cf. also Luke 1.2; Acts 11.15; I John 2.7,24; 3.11; II John 6). Similarly, in Acts 15.7 ἀφ'ἡμέρων ἀρχαίων is used at the council of Jerusalem of 'the early days' less than twenty years previously, and Mnason, 'a Christian from the early days' is described already by Luke in the early 60s as an ἀρχαῖος μαθητής (Acts 21.16).

The objections therefore to placing I Clement in 70 cannot be regarded as decisive. Its references to Hebrews in the exhortation of ch.36, so far from arguing, as has been claimed, a late date for Hebrews, on the ground that I Clement quotes from a recent document, would be entirely natural if Hebrews had been addressed to the Roman church but two or three years earlier. And there are other positive indications which Edmundson adduces in favour of an early date:

1. The continued use in the liturgical passage of 59.2-4 of the primitive description of Jesus as παῖς, the servant or child of God, common to the Acts speeches and the Didache.

2. The fact that, as Lightfoot recognizes, [AF I.1,353.]the quotations from the gospel tradition 'exhibit a very early type'. The author does not introduce them (as he does citations of the Old Testament) with the words 'It is written' or 'The scripture says'. Indeed on the only two occasions (13.1f.; 46.7f.) he cites such material he employs precisely the same formula that Luke places on the lips of Paul in Acts 20.35: 'Remember the words of the Lord Jesus, which he spake.' And once more in all probability the quotations are not from our gospels but from oral tradition or 'some written or unwritten form of 'Catechesis" ... current in the Roman Church'. [NT in the Apostolic Fathers (see n. 70 above), 61. Similarly, W. K. Lowther-Clarke, I Clement, 1937, 1.1f.; Koester, op. cit., 12-19; Grant and H. H. Graham in Grant, The Apostolic Fathers II, ad locc.; D. A. Hagner, The Use of the Old and New Testaments in Clement of Rome, Leiden 1973, 171.]

3. In a later letter to Soter, Bishop of Rome, Dionysius, Bishop of Corinth, says:

This day, therefore, we spent as a holy Lord's day, in which we read your epistle; from the reading of which we shall always be able to obtain admonition, as also from the former epistle written to us through (διά) Clement.
[Quoted Eusebius, HE 4.23.11.]

Though Edmundson, following Bigg, thinks that this is parallel to I Peter 5.12, we have seen reason to doubt whether Silvanus is there designated as more than the carrier of the letter. The closest parallel would seem to be in the Martyrdom of Polycarp 20.1, where the church in Smyrna writes an account of Polycarp's death to the church at Philomelium 'through our brother Marcianus'. He is not simply the amanuensis (Euarestus is that; 20.2), but he is the church's agent. Similarly, says Edmundson, Clement is 'only the servant, not the head of the Church acting on his own initiative'. [Op. cit., 202.] In fact he is fulfilling precisely the role which Hermas (Vis.2.4.3) says was his assignment (ἐποτέτραπται), that of correspondent of the Roman church in its external relations (εἰς τὰς ἒξω πόλεις). He is not (yet) its bishop. The assumption that if I Clement is by Clement it must have been written during his episcopate,
[First asserted (though still not explicitly) by Irenaeus, Adv. haer. 3.3.3.]

that is, in the last nine years of his life, no more follows than it does of most bishops' literary productions, despite Lightfoot's fantastic achievement in working on the completion of his own revised edition of Clement up to within three days of his death as Bishop of Durham. [Cf. Westcott's preface to Lightfoot, AF L 2, v-viii.]

4. Finally, and of least importance, the concluding reference in 65.1 to Fortunatus, who, unlike Claudius Ephebus and Valerius Bito, appears not to be a Roman envoy but a member of the Corinthian church, would fit the Fortunatus whose coming from Corinth to Ephesus so relieved Paul in 55 (I Cor.i6.i7f.), if the epistle was written in 70. It is, however, as Edmundson says, [Op. cit., 199.]extremely unlikely that he 'was still active and travelling to and fro as an emissary between his native town and Rome in 96 ad, more than forty years later'. Of course it may not have been the same Fortunatus - though the fact that the only two we know of both came from Corinth looks more than a coincidence.

There are other points that Edmundson makes, including some intriguing speculation on the occasion of the Corinthian dissensions following the drafting by Nero of 6,000 Jewish prisoners to dig the Corinth canal in 67-8. [Op. cit., 195f. According to Eusebius, HE 3.16, who himself put I Clement under Domitian (following Irenaeus?), Hegesippus had evidence of a dissension which took place at Corinth at that time. But we know nothing else of this.]
None of his arguments is in itself decisive. The overall balance of probability will be assessed differently by different people. But if the case Edmundson makes is not proven, [His dating of I Clement, though largely ignored, was accepted, most notably, by Henderson, Five Roman Emperors, 45, despite his earlier adoption of Lightfoot's dating in Nero, 443, 484. It was also supported by Badcock, Pauline Epistles, 133, i86f., 208; and by L. E. Elliott-Binns, The Beginnings of Western Christendom, 1948, ioif., 225, as 'much more probable'. Lowther-Clarke, I Clement, 11f., while disagreeing, conceded that it was 'not impossible'. A. E. Wilhelm-Hooijbergh, 'A Different View of Clemens Romanus', HJ 16, 1975, 266-88, also contends (without any reference to Edmundson) for a date of 69. Some (but not all) of his arguments merit further consideration.]
it shows at least how fluid and uncertain is the dating of one of the so-called 'landmarks' of the sub-apostolic age. [All Lake was prepared to say for certain for the date of I Clement was 'between 75 and no' (The Apostolic Fathers, Loeb Classical Library, 1912, 5). My colleague J. V. M. Sturdy, in an acutely argued article, soon to be published, on 'Clement, Ignatius and Polycarp', also starts from the quite insubstantial basis of the traditional dating. But he proceeds, with Merrill, op. cit., 217-41, to put it much later, c. 140, arguing back from the Muratorian Canon's dating of the Shepherd ofHermas with its reference to Clement. The corollary of his position is that 'the Ignatian Epistles are pseudepigraphical, from a date late in the second century, and that the Epistle of Polycarp is also pseudepigraphical, from yet a later date, perhaps in the third century, but possibly even later', and that the Martyrdom of Polycarp is no longer a contemporary account but comes from later in the second century. I cannot possibly enter into the details of his arguments (which include the usual cumulative one from pseudonymity), but, as when Knox throws over most of the Acts evidence, it would seem a weakness of any position to be required to jettison so much. Like Merrill, he denies that Clement was ever bishop of Rome (Merrill also disputing, op. cit., ch.11, that Peter had ever been there!). But he does bring out very clearly the inconsistency we observed in Perrin (p. 9 n. 21 above), and which comes out also in Streeter's remark (PC, 108) that 'in Asia monepiscopacy antedates the writing of the Pastoral Epistles', of putting 'the "early catholic" books of the New Testament like the Pastorals' after or about the same time as 'the "definite catholic" books like I Clement and Ignatius'. With these last banished to the mid-second century and beyond, he observes that 'the strongest check on the dating of the New Testament books is removed' - and, apart from the genuine Pauline epistles, Mark and Colossians, he puts everything (even if only tentatively) after 110!]

To conclude, there would seem to be very little against the following sequence:

The Didache



I Clement


early 70

The Epistle of Barnabas


c. 75

The Shepherd of Hermas


-c. 85

But even if I Clement were still to be placed last, c. 96, we should have a perfectly intelligible series. The pressure to push any of them into the second century has, I believe, largely been created by their natural place having been usurped by books of the New Testament. Conversely, if these non-canonical documents do not belong to the second century, then their affinities with certain features in the canonical writings cannot be used to relegate the latter to the same period. The arguments for dating the Pastorals and II Peter, let alone I Peter and Acts, in the second century begin to look less and less substantial. Obviously there is a circularity here, and only if the chronology of the sub-apostolic literature as a whole, including that of the crucial Ignatian epistles, [Sturdy regards these with their evidence for mon-episcopacy in the early second century as 'the Piltdown man of the history of the Christian church'!]
were being established in its own right could this be used to argue for an early dating of the New Testament. All that I have attempted in this postscript is to remove some of the objections to such a dating arising from the vacuum it could appear to leave in the last quarter of the first century.