Anyone interested in the Historical Jesus of course needs to have some theory of the pre-synoptic tradition, that is, the sources (oral and/or literary) that lie behind Matthew, Mark, and Luke. For several decades, the dominant paradigm has been that originating with the form critics, with Rudolf Bultmann's History of the Synoptic Tradition being perhaps the classic statement of this approach. This paradigm, however, has recently been challenged by several scholars, notably James Dunn and (even more so) Richard Bauckham.
Dunn and Bauckham argue that the understanding of the pre-synoptic tradition derived from the form critics is defective (and one consequence of this is that it hinders our study of the Historical Jesus). Their main bone of contention is that the form critics have failed to appreciate the importance of distinctively oral modes of tradition transmission, by examining the pre-synoptic tradition through the lenses of a 'literary' paradigm. Dunn argues, instead, that the community as a whole would keep a close guard on the original tradition, whilst allowing limited variety within original pattern; Bauckham, of course, argues that this regulation comes from eyewitnesses.
The strengths and weaknesses of their positive proposals have been much debated, and I do not intend to engage them here. The question that I do want to engage is whether a broadly 'form critical' approach is really so unhelpful in understanding the pre-synoptic tradition.
The basic proposal which I would like to advocate is that the tendencies we see operating between the synoptic gospels provide us with evidence of the kinds of tendencies that would have been operative in the pre-synoptic period. This is what I take to be at the heart of the 'form critical' approach. Consider an example: in the synoptic gospels, it is widely agreed that there is a tendency to make more explicit and/or pronounced the subordination of John the Baptist to Jesus (seen, for instance, in the accounts of Jesus's baptism). From this, it is suggested, we can infer that in the pre-synoptic tradition, there may have been a similar tendency to increase subordination of John the Baptist to Jesus. One practical outcome of this approach might be that where, in the Markan tradition, we see subordination of the Baptist to Jesus (e.g. Mark 1:7), we might suspect (though not necessarily affirm) that this subordination is itself the result of a tendency that was also found in the pre-synoptic tradition. (I probably should emphasize that this is only a possible example of the kind of thing I am talking about; I am not especially interested here in making a strong claim about this particular tradition!)
As I see it, there are two possible objections to this line of reasoning. The first is that it is inappropriate to speak of 'tendencies' of the synoptic tradition at all. This criticism, if made, would probably be justified with reference to E.P. Sanders' Tendencies of the Synoptic Tradition (which of course was the inspiration for the title of this post!). However, the kinds of 'tendencies' of which Sanders is rightly critical are not those I am suggesting we should detect in the synoptic and (hence) pre-synoptic tradition. Sanders criticizes detection of 'mechanical' tendencies such as length of pericope, addition of names, and so on. It is not clear, however, that his criticism would apply to a tendency such as 'subordination of the Baptist to Jesus', which does not require a dubious regularization of the development of the tradition.
The second objection is this: that we cannot move from characterization of the synoptic tradition to characterization of the pre-synoptic tradition, because the pre-synoptic tradition was in some way 'fundamentally different'.
I would like to argue, however, that this objection does not go through. Firstly, this objection is often made from the perspective of 'oral vs. literary': the synoptic tradition (as we have it in our gospels) is of course written, and therefore it is illegitimate to infer from it the tendencies of the pre-synoptic tradition, which was oral. Even if this were true of oral traditions (see below), this objection would not apply to any of the written sources that Mark and the other evangelists used. So wherever we postulate a written source, it does not seem unreasonable to see some similarity between the tendencies of the synoptic tradition and the tendencies of the pre-synoptic tradition.
What, then, of 'oral' sources (to which Mark especially presumably had some access)? If we are to make any claims about developments of the tradition in cases where that tradition was orally transmitted, we need a theory of oral tradition, which in turn needs to be based on evidence, and I would suggest that the best evidence we have for how oral traditions were treated and developed is in fact the synoptic gospels themselves (rather than theories drawn from totally different geographical and/or temporal contexts). I am not saying, as does, for instance, Dunn, that the variant synoptic traditions themselves directly attest to oral traditioning (though this may be true in some cases). All I am suggesting is that the closest evidence we have to the treatment of Jesus traditions in the pre-synoptic period are the synoptic gospels themselves, and therefore, despite being written documents, the synoptic gospels should be given pride-of-place as evidence in our discussion of pre-synoptic tradition.
In conclusion, then, if the synoptic gospels are our most temporally and geographically proximate evidence for the pre-synoptic tradition, then in our discussion of the pre-synoptic tradition (whether oral or literary) the best place to start is with an examination of the tendencies that can be detected between these documents. From there, we can move on to tentatively identify possible tendencies in the pre-synoptic tradition. This is not an illegitimate construction of non-existent 'tendencies', nor is it a failure to understand 'oral tradition': it is simply an employment of the best evidence we have for the pre-synoptic tradition, namely, the synoptic gospels themselves. Therefore, recent criticisms of this sort of approach (e.g. from Dunn and Bauckham) are arguably misguided.