In this important and wide-ranging book, Helmut Koester returns to the subject suggested to him by Rudolf Bultmann in 1950: the gospels in the second century. In approach and content it reflects his own research in the intervening years, and the continuing influence of both Bultmann and the Religionsgeschichte school. This book is intended to be a thorough introduction to most aspects of the transmission and collection of gospel traditions from the first into the second century. Although full of informed discussion, numerous insights and careful attention to detail (from which all will surely learn), there are, in my opinion, serious weaknesses in the approach taken, weaknesses of both form and content.
In form, the book consists of separate chapters on (i) the term ‘gospel’; (ii) ‘Sayings Collections’; (iii) dialogues, narratives and the Johannine traditions; (iv) the synoptic gospels; and (v) the harmonization of the canonical gospels (this includes a most helpful introduction to Tatian’s Diatessaron by W. Petersen). There is no general introduction (although the preface, to which we shall return, is important) and no conclusion; thus there is no overall synthesis and few summaries of the argument (and no subject or author index!). These features, combined with a style which tends to pronounce on any number of disputed claims in a forthright manner, rather disguises the tendentious (!) nature of many of Koester’s arguments.
We might summarize the content of Koester’s position along the following lines (noting that his detailed and suggestive discussions cannot be adequately summarized or responded to in a short review): the words of Jesus were collected in different forms by many (and diverse) groups of followers, some of whom emphasized his wisdom teaching, others added apocalyptic elements, others created a passion narrative to explain the death of the Messiah, others created birth legends under the influence of Egyptian mythology. Several of the non-canonical gospels (Koester concentrates on Gospel of Thomas. Dialogue of the Savior, Apocryphon of James, Papyrus Egerton 2, Gospel of Peter) are not dependent upon the canonical gospels (as Christian scholars have generally insisted) but are often prior to, and sometimes even function as sources for, the canonical gospels (or share common sources with them). The distinction between canonical and non-canonical, orthodox and heretical, is a product of dogmatic bias and is an obsolete notion, unsupported by historical evidence.
This assertion, defended elsewhere by Koester and applied to the literature of early Christianity in his two-volume Introduction to the New Testament, is vital: ‘Only dogmatic prejudice can assert that the canonical writings have an exclusive claim to apostolic origin and thus to historical priority’ (p. xxx). But in response to Koester we must note the following: first, his argument that the canonical gospels are no better attested than most non-canonical ones often disallows contradictory evidence: Didache 1.3–5, which certainly quotes from the canonical gospels, is attributed to a later interpolation (although the only basis for this is its use of gospel material!); the Muratorian Canon and the Anti-Marcionite Prologues which attest a four-gospel canon are dated by Koester in the fourth century (rather than in the second century with the scholarly consensus). Secondly, he ignores the geographical and linguistic spread of attestation for the canonical gospels (as others have pointed out). Thirdly, he is overdependent on the seminal (but now outdated) work of W. Bauer for the assertion that there was no ‘orthodoxy’ in the earliest period.
It is quite clear that the notion of a ‘canon’ is a dogmatic one, delimiting the content of the church’s authoritative rule of faith. The process of canon formation is also clearly a historical phenomenon, and in the early discussions it was those documents which exhibited longevity and wide attestation (in addition to apostolicity) which were included. But canonicity cannot demand unique priority for the canonical gospels (cf. Lk. 1:1–4). So, while Koester’s assumption that there is no basis for the canon can, and must, be challenged, it is fair to agree with his statement: ‘whether my own reconstruction of the development of this literature is plausible should be argued on historical and source-critical grounds’ (preface).
The key word here is ‘reconstruction’, and this might be applied both to his approach to more mainstream positions, and his approach to the documentary evidence. Thus Koester makes no attempt to interpret documents (whether canonical or not) in a holistic way, his sole preoccupation being with source criticism. This provides him with explanations for the secondary elements in, for example, the Gospel of Peter and the Gospel of Thomas: they are later additions to earlier (but hypothetical!) sources, and it is the earlier source which often stands behind the canonical gospels. Scholarship can only proceed by the making and defending (or challenging) of hypotheses, but at least two things are crucial for this advance: (i) a methodological defence of each suggested hypothesis; (ii) a continued awareness that however many scholars are persuaded by a hypothesis, the solution remains hypothetical. Koester’s weakness is not in the making of hypotheses (by no means!), but he never provides a defence of his method. In particular, the priority of source criticism over an assessment of the whole document cannot be assumed, especially when a holistic literary approach may help explain the relationship between canonical and non-canonical gospel literature (cf. my article ‘On the Christology of the Gospel of Peter’, forthcoming in Vigiliae Christianae).
There are other particular problems in Koester’s procedure when the Coptic Gospel of Thomas from the fourth century is used as a primary source (from the first century), even though the Greek fragments from Oxyrhynchus, the citations in Hippolytus, and the use of Gnostic technical terms suggest that it has undergone a thorough redaction between the second and fourth centuries. One might also question the emendation of the Markan text in order to make it fit Koester’s theories (e.g. pp. 275ff).
More importantly, Koester rejects the view that Papyrus Egerton 2 (now attested by another manuscript from the Köln collection), the Secret Gospel of Mark and the Gospel of Peter could be second-century redactions of the canonical gospels because it would require authors harmonizing material from all three gospels (see e.g. pp. 214f.). This stands in some degree of tension with his final chapter which demonstrates the influence of harmonizing tendencies on 2 Clement, Justin and Tatian. This tendency is also apparent (in material most of which is not discussed by Koester) in Didache 1.3–5, the Gospel of the Ebionites, the harmonies of Ammonius and Theophilus, and the well-known tendency of scribes to harmonize the gospel accounts. This widespread harmonistic tendency provides, in my opinion, a more plausible rationale for the development of second-century (non-canonical) gospel traditions than the arguments of Koester. Readers might also be referred to David Wright’s review of Crossan’s Four Other Gospels in Themelios 12.2 (1987), pp. 56–60, for a useful discussion of a similar position.
The NT suggests that the apostles of Jesus functioned as guardians and interpreters of the traditions of his words and deeds (e.g. Jn. 14:26; 16:13; Acts 1–21f.; 1 Cor. 2:13; 4:1f.; Eph. 2:20). It also suggests that this apostolic deposit would be rejected by many (why should the followers expect a more positive response than their master?), some of whom would reinterpret the content of the message (e.g. 2 Tim. 2:17f.; 3:8; 2 Pet. 3:16; Rev. 22:18f.). Patristic evidence suggests that the process of heterodox (or non-apostolic) reinterpretation of the gospel traditions continued into the second century and beyond (the earliest references are Polycarp, Phil. 7.1; Dionysius according to Eus, Eccl. Hist. IV.23.12). The evidence discussed by Koester has not persuaded me that his hypothesis is more plausible than that suggested by the NT witness itself. Admittedly I have a dogmatic bias in favour of a hypothesis which takes the NT as its starting point, makes sense of the existence of the diverse traditions, treats documents as wholes rather than quarries, and enables me to think that Christianity is true; but it is always possible that dogma could be true.
