Only recently, in a centenary review of 2005, Andrew F. Gregory and Christopher M. Tuckett have picked up the often used, but rarely diligently read slender work of the Oxford Society of Historical Theology of 1905, entitled The New Testament in the Apostolic Fathers, brought together a group of specialists and reviewed these earlier findings. William L. Petersen gives a short summary of what had been achieved at Oxford over a hundred years ago:
The charge given the committee [of the Oxford Society of Historical Theology] was ‘to prepare a volume exhibiting those passages of early Christian writers which indicate, or have been thought to indicate, acquaintance with any of the books of the New Testament’. The committee limited itself to the so-called Apostolic Fathers, examining eight authors (and/or texts) [i.a. Barn., Did., 1Clem., Ign., PolPhil., Herm., 2Clem.]. … The 1905 researchers ranked the likelihood that a specific Father demonstrated knowledge of a given book in the New Testament by assigning each possible intersection a letter grade from ‘A’ to ‘D’. ‘A’ designated ‘books about which there can be no reasonable doubt’ that the Father knew it; ‘B’ referred to books where there was a ‘a high degree of probability’. ‘C’ referred to a ‘lower degree of probability’. And ‘D’ meant that the evidence was ‘too uncertain to allow any reliance to be placed upon it’. A table on page 137 summarized the results. Out of a total of 216 possible intersections between a Father and a specific book, conclusions were possible in only eighty-five of the intersections, 39 per cent. Out of those eighty-five places where it was possible to assign a letter rank, we find forty-three Ds and twenty-two Cs. There are fourteen Bs (eight of them, however, come from a single source: Polycarp), and six As. … The most remarkable aspect of the 1905 volume is the fact that now, a century later, the significance of the ‘formal’ results achieved by the committee … pale into insignificance when compared with the notes the researchers offered on the passages they examined. … The 1905 researches … were well aware of the multiplicity of possible explanations for the evidence they found in the Apostolic Fathers; they were also acutely aware of their inability to reach definitive judgements on the basis of the evidence. All they could do was follow the via negative: the source(s) used in about three-quarters of the passages in the Apostolic Fathers with a parallel in the New Testament … ‘affords no evidence for the use of either of our Gospels in its present form’; that being the case, one had to consider … ‘the direct use of another [viz. non-canonical (W.L. Petersen)] source altogether, whether oral or written’.
W.L. Petersen concludes his summary pointing out that ‘their empirical, textual observations were devastating for the idea of a ‘standard’ or ‘established’ text of the New Testament in the first half of the second century. And he specifies the results from his own reading of them:
First, it is clear that the vast majority of passages in the Apostolic Fathers for which one can find likely parallels in the New Testament have deviations from our present, critically reconstructed New Testament text. It must be emphasized that the vast majority of these deviations are not minor (e.g., differences in spelling or verb tense), but major (a completely new context, a substantial interpolation or omission, a conflation of two entirely separate ideas and/or passages).
Second, harmonization is a surprisingly common phenomenon. Sometimes the harmonizations are (almost) entirely composed of material found in our modern editions of the New Testament; more often,however, they contain material which we today classify as extra-canonical.
Third, the Apostolic Fathers often reproduce, without remark, material that we, today, call extracanonical. Sometimes this extra-canonical material is introduced with the quotation formula – such as, ‘the Lord says’, or ‘the Gospel says’. The obvious inference is that the Faher considered this extra-canonical source as authoritative as any other. … Some might wonder if the disagreements would disappear if the basis for comparison were changed from our modern critically reconstructed text to the texts of the ‘great uncials’ of the mid-fourth century (Codex Sinaiticus and Codex Vaticanus). They do not. Even if the basis for comparison is changed to the text of our oldest continuous-text manuscripts of the New Testament documents (P64+67 + P66 [both of which date from ‘ca. 200’, W.L. Petersen]), the differences remain. One simply must admit that the passages found in the Apostolic Fathers are different from the texts found in our oldest New Testament papyri, from the texts of the ‘great uncials’, and from the text of our modern editions.
How can we account for these findings, if the later canonical Gospels were written before 100, or around 70, or, as some opt, in the early 40th and had become authoritative readings in the community? Petersen sees only two models of explanation, one which is ‘profoundly flawed’ and another one which seems the only way to make sense. The first, flawed model would assume that the ‘deviations’ from the established and fixed texts of the Gospels and Letters, dated to the first century, ‘would be – as suggested by many Victorian (and even contemporary) scholars – due to citation from memory or adapting the text to the purposes of the moment (e.g., preaching, evangelizing, teaching, disputing):
In the first half of the second century – that is, in the age of the Apostolic Fathers – and even later, into the time of Tatian and Clement of Alexandria (near the end of the second century), there was neither a fixed canon nor a fixed text for any of the New Testament documents. Rather, ‘clusters’ of sayings/episodes/parts of (what later became our canonical) gospels and epistles circulated, initially (for the gospels, at least) probably without a title, and then, later, with a title. But the contents of the ‘cluster’ bearing the title ‘Mark’ or ‘Romans’ was still very much in flux and subject to change. Additions were still being made, as were deletions; the sequence of the text was still being modified. … Subscribing to this model has certain consequences. It means that scholars must be very circumspect about attributing anything to the first-century church. And what evidence we have from the second century – in the Apostolic Fathers, for example – hardly inspires confidence. The problems are not confined to the liberties taken with the texts …, but also extend to the matter of the boundary between what would later be called canonical and extra-canonical texts, and the citation of extra-canonical material as ‘gospel’ or logia Iesou during the age of the Apostolic Fathers. The issue, then, is not just one of the texts being unsettled, but also one of which documents (or, more properly, clusters of material) and which traditions were authoritative, and which were not.
W.L. Petersen is joined in his judgement by J. Keith Elliott: ‘As far as the Apostolic Fathers are concerned, we may well agree that they, like Justin and most other early writers, are unlikely to have had access to the “published” documents’. As has been shown by Elliott, W.L. Petersen’s view is materially embedded in the critical apparatus of both the Nestle-Aland and the United Bible Societies’ Greek Testament hand editions which with the exception of Did. 8,2 (// Matth. 6:9) in Nestle-Aland ignore even the very few ‘A’-rated citations in their critical apparatus.
That the Gospels are not quoted or referred to in our early Christian literature prior to Marcion is clouded by the boundaries between the disciplines of New Testament Studies and Patristics. The well-known reference work, Biblia Patristica, for example, covers texts ‘from the origins to Clement of Alexandria and Tertullian’ in its first volume, but excludes all writings that can be found in the New Testament. If these were included, it would become even more apparent that pp. 223-415 of this volume listing over 10,000 quotes (!) from Matthew, Mark, Luke and John for the period from Marcion onwards, does not provide a single one (!) for the time before Marcion. The first arguable cases are those authors who are sometimes dated to the beginning of the second century (Ignatius, Papias, Polycarp, Hegesippus), which, however, many scholars rather date contemporary or later than Marcion. But even, if we date them early and include them into the comparison with the Gospel, we come to the same conclusion as shown by the findings of the 1905 Oxford researchers and the revisiting of these findings in 2005.
 W.L. Petersen, ‘Textual Traditions Examined’ (2005), 29-32.
 W.L. Petersen, ‘Textual Traditions Examined’ (2005), 33f.
 W.L. Petersen, ‘Textual Traditions Examined’ (2005),42-5 gives a fourfold account why he finds this model untenable.
 J.K. Elliott, ‘Absent Witnesses?’ (2005), 53.
 See J.K. Elliott, ‘Absent Witnesses?’ (2005), 48.
 A fact that is, for example, clouded by the boundaries between the disciplines of New Testament Studies and Patristics. The well-known reference work, Biblia Patristica, for example, covers texts ‘from the origins to Clement of Alexandria and Tertullian’ in its first volume, but excludes all writings that can be found in the New Testament. If these were included, it would become even more apparent that pp. 223-415 on quotes from Matth., Mark, Luke and John resulted in entries only that derive from the time of Marcion onwards.