f22

1. Introduction: Form-Criticism

Biographies usually contain anecdotes; but biographies and anecdotes are not the same thing. Jewish histories and folk-tales contain vignettes of kings, prophets, holy men, pious women; but Jewish histories and personal vignettes are not the same thing. The canonical gospels, which I have argued are a unique combination of Hellenistic biography and Jewish history, contain anecdotes and vignettes, almost all about Jesus. We cannot presume that these are the same sort of thing as the gospels themselves. Having examined (some of) the large stories in early Christianity, we must turn our attention to the smaller, and in many cases earlier, ones, to see whether they reflect the same patterns and concerns.

The study of these smaller and earlier stories is therefore clearly demanded by the logic of the argument within the present Part of the book. It is also highly desirable in view of one aspect at least of the larger project of which this book forms a part. Without a serious consideration of the forms taken by Jesus-stories prior to their inclusion within the gospels, we are at least open to the charge of ignoring crucial parts of the evidence when it comes to thinking about Jesus himself, as we intend to do in the next volume. Rather than take space, in a book about Jesus, to write about the context and content of the stories which circulated about him during the subsequent generation, it is far more natural to discuss these stories at this point, where they really belong. Form-criticism, despite often being treated simply as a tool for discovering about Jesus, is designed primarily to shed light on the early church. Since that is our present topic, it is appropriate to address these issues here.

Study of the early history of stories in the gospels has traditionally gone under two names: tradition-criticism (or tradition-history) and form-criticism. The two are sometimes used interchangeably; properly speaking, ‘tradition-criticism’ is the wider term, dealing with all early traditions, while ‘form-criticism’ is the more focused, concentrating on those traditions that have specific and recognizable ‘forms’. The principle behind all such activity is sometimes obscured in a maze of technicality, and it may be worth spelling it out at this stage.

Traditions do not lie around unshaped. As we saw in Part II, all history involves interpretation. In order for something to be told, it has to be put into some kind of form. There are few things more frustrating than having to listen—as one sometimes does with a child, or a drunk—to a story that remains unshaped and unsorted, with the salient facts that might have made the tale worth telling being hidden behind the fog of irrelevant information and comment. But there are different sorts of shaping. When my wife and I tell our children the story of their births, as we sometimes do on their birthdays, we naturally minimize the medical details, and highlight the sense of parental excitement, the delight at discovering a new member of the family, that accompanied the event. The story often concludes by retelling a remark made by one of us at the time, or by a nurse in the hospital, epitomizing the feelings of those present. If, however, we tell the same story to a doctor, especially if there is concern about the present health of the child, we select different information, highlight different things. Our feelings at the time are of little significance; whether the child began to breathe at the appropriate moment is far more important. It would thus be possible in principle to deduce, from the form of the story (its emphases, highlights, and perhaps concluding quotations), the context and purpose in and for which it was being told.

The basic insight of the form-critics was to apply this fairly obvious point to small units (pericopae or paragraphs) of material in the gospels.¹ Here is a story about Jesus. He performs a healing, enters into controversy, utters a memorable saying. The entire episode, if acted out using the gospel narrative as script, would take less than a minute. We may assume without more ado that it has been compressed, as all stories are, in the telling; that the story will not contain details of every single thing that happened, or that was said. We may assume, further, that it was told in a variety of forms, now lost, by the people involved, to highlight a multiplicity of things—their own feelings and emotions, what some other bystander said, and so on. Anecdotes are told for a variety of reasons, and the reasons dictate the form. What we have in the gospels are stories in which the telling and retelling has been made to focus on a particular point, one of many that could in principle have been made. This can tell us, again at least in principle, what sort of situation the teller was addressing, or was involved in. It is not ‘neutral information’: there is, as we have seen, no such thing. It is directed at particular needs, and makes a particular point.

So far, we have simply followed the implications of common sense. Serious study of forms begins, however, when we discover that some stories regularly fall into a more or less clearly defined pattern, suggesting that we might be able to construct a grid to determine, from the form of the story, the kind of setting it may have had. This is the sort of analysis of the stories in the gospels offered by the classic form-critics earlier this century. As the work progressed, ‘tradition-criticism’ set it on a broader canvas, attempting the second-order activity of studying the way in which particular traditions, particular tellings of one or more stories, developed over time and through different phases of the life of the early church.

The task of form-criticism is both important and difficult. Important, because the small stories came before the larger ones that we now possess, and with them we move back into the period about which we know virtually nothing except through Paul and Acts. Difficult, not least because of the intrinsic difficulties, which should not be minimized, in reading between the lines of a later document to find traces of earlier non-documentary storytelling. I suspect that even the most confident form-critic would become anxious if asked for proveable hypotheses about the pre-literary life of anecdotes which now form part of the biography of some twentieth-century figure. To attempt to reconstruct the equivalent phenomenon with first-century documents seems even more presumptuous. Difficult too, however, precisely because confident critics, undeterred by this problem, have left a tangled web of theories, hypotheses, misunderstandings, unproven speculations and downright bad guesses strewn around the subject, so that anyone wanting to come to it fresh finds a prickly hedge barring the way.²

Among the misunderstandings, we may here mention three. First, when form-criticism burst upon the scene in the years after the First World War, it was not designed primarily as a tool to find out about Jesus. In the hands of Rudolf Bultmann in particular, it was a tool to find out about the early church. Bultmann assumed that we could know certain things about Jesus—not very much, but enough to know that most of the gospel stories could not have taken place as narrated. He therefore looked for possible situations within the early church within which stories like these could have been told to express some aspect of the church’s faith and life. For Bultmann, as we saw earlier, Jesus within his historical context was not the focus of Christianity. Far more important was the faith of the early Christians. Once this is realized, a good deal of the reaction to form-criticism, not least in England, is shown to be off the point. The tool was not originally designed as a means of finding Jesus; that it proved unsuccessful in this task is not a telling criticism.

A second misunderstanding, once that one is out of the way, is the assumption that the discipline of form-criticism necessarily belongs with one particular hypothesis about the origin and development of the early church. Since the major practitioners of the discipline had a fairly clear idea of how Christianity grew and changed, it was comparatively easy for them to assign gospel anecdotes, or fragments of them, to different periods or stages. As Bultmann assumed that he knew a certain amount about Jesus, so he assumed that he knew a certain amount about the early church: that it began as a kind of variant on Gnosticism, though using some Jewish language; that it developed in two strands at least, one of which carried on the Gnostic or ‘wisdom’ tradition, while another interpreted Jesus within a more Jewish-style development; that these two were combined in the writing of the first canonical gospel; that Christianity quickly spread beyond its original base, which merely happened to have been Jewish, and that it translated the language of its earliest expression into Hellenistic thought-forms as it did so, which was all the easier to do since the Jewish thought-forms had in any case been mere accidental features of a message that was more akin to Hellenism.³ The Bultmannian paradigm, showing just how much it remained rooted in that of F. C. Baur, thus envisaged Jewish Christianity and Hellenistic Christianity existing more or less independently, side by side, until they were combined, some time in the second generation, to form the beginnings of early catholicism.

These were the assumptions of the pioneers of form-criticism. Their analysis of forms, and their hypothetical history of traditions, is heavily dependent on this picture. In consequence, it is often assumed that to practise form-criticism is to accept this view of Christian origins. But by itself the idea of examining the form of individual pericopae, and searching for their probable setting in the life of the early church, need not be committed to accepting one view of early Christian history rather than another.

A third misunderstanding concerns the belief of many early form-critics that the stories in the early tradition reflected the life of the early church rather than the life of Jesus, in that the early church invented (perhaps under the guidance of ‘the spirit of Jesus’) sayings of Jesus to address problems in their own day. The main problem with this assumption is that the one fixed point in the history of the early church, i.e. Paul, provides a string of good counterexamples, which work in two directions.⁴

On the one hand, as is often pointed out, Paul regularly addresses questions of some difficulty, in which he does not even quote words of Jesus, in the synoptic tradition, which could have been helpful to him. Still less does he appear to attribute sayings to Jesus which were not his.⁵ Why was he so reticent, if ‘words of Jesus’ were regularly invented by Christian prophets, of whom Paul was assuredly one, to address problems in the early church?

On the other hand, as is not so often noted, Paul provides evidence of all sorts of disputes which rocked the early church but left not a trace in the synoptic tradition. From Paul, we know that the early church was torn in two over the question of circumcision. There is no mention of circumcision in the whole synoptic tradition.⁶ From Paul, we know that some parts at least of the early church had problems in relation to speaking in tongues. There is no mention of this in the main stream of synoptic tradition.⁷ From Paul, it is clear that the doctrine of justification was a vital issue which the early church had to hammer out in relation to the admission of Gentiles to the church. The only mentions of the admission of Gentiles in the synoptic tradition do not speak of justification, and the only mention of justification has nothing to do with Gentiles.⁸ In Paul it is clear that questions have been raised about apostleship, his own and that of others. Apostleship is of course mentioned in the synoptic tradition, but so far is the tradition from addressing post-Easter issues here that it does not discuss the question of subsequent apostolic authority except for one passage—and in that passage it still envisages Judas sharing the glorious rule of the twelve.⁹ In Paul we meet the question of geographical priority: does the church in Jerusalem have a primacy over those working elsewhere? In the synoptic tradition the criticisms of Jerusalem have to do with its past and present failures, and with its wicked hierarchy, not with the place of its church leaders within a wider emerging Christianity. So we could go on: slavery, idol-meat, womens’ headgear, work, widows; and, perhaps above all, the detailed doctrines of Christ and the divine spirit. The synoptic tradition shows a steadfast refusal to import ‘dominical’ answers to or comments on these issues into the retelling of stories about Jesus. This should put us firmly on our guard against the idea that the stories we do find in the synoptic tradition were invented to address current needs in the 40s, 50s, 60s or even later in the first century.

Conversely, it has been shown often enough that the synoptic tradition has preserved material which is not so relevant to, or so obviously taken up by, the first-generation church. Well-known examples include the concentration on Israel;¹⁰ Jesus’ attitude to women;¹¹ and many other features. As Moule concludes, ‘Aspects of Jesus’ attitude and ministry have survived in the traditions, despite the fact that the early Christians do not seem to have paid particular attention to them or recognized their christological significance.’¹²

Recognizing these points does not, however, mean that we must abandon the discipline of form-criticism. On the contrary. There is every reason to study the early stories and their forms. A good case can be made out for saying that oral history within the early church was likely to have been strong and formative.¹³ Three basic moves need to be made here: one about Jesus, one about his first followers, and one about the meaning of ‘oral history’.

First, unless we are to operate with a highly unlikely understanding of Jesus and his ministry, we must assume some such picture as we find in Gerd Theissen’s brilliant work, The Shadow of the Galilean. Jesus was constantly moving from place to place, working without the benefit of mass media. It is not just likely, it is in the highest degree probable, that he told the same stories again and again in slightly different words, that he ran into similar questions and problems and said similar things about them, that he came up with a slightly different set of beatitudes every few villages, that he not only told but retold and adapted parables and similar sayings in different settings, and that he repeated aphorisms with different emphases in different contexts.¹⁴ Scholars of an older conservative stamp used to try to explain varieties in the synoptic tradition by saying cautiously that ‘maybe Jesus said it twice’. This always sounded like special pleading. Today, once a politician has made a major speech, he or she does not usually repeat it. But the analogy is thoroughly misleading. If we come to the ministry of Jesus as first-century historians, and forget our twentieth-century assumptions about mass media, the overwhelming probability is that most of what Jesus said, he said not twice but two hundred times, with (of course) a myriad of local variations.¹⁵

Second, those who heard Jesus even on a few of these occasions would soon find that they remembered what was said. We do not even have to postulate a special sort of oral culture to make this highly likely; even in modern Western society those who hear a teacher or preacher say the same thing a few times can repeat much of it without difficulty, often imitating tones of voice, dramatic pauses, and facial and physical mannerisms. Moreover, when there is an urgent or exciting reason for wanting to tell someone else what the teacher has said and done, a hearer will often be able to do so, in summary form, after only one hearing; then, once the story has been told two or three times, the effect will be just as strong if not stronger as if it had been heard that often. This is a common-sense point, which would not need spelling out, were it not so often ignored. When we add to this the high probability that Palestinian culture was, to put it at its weakest, more used to hearing and repeating teachings than we are today, and the observation that much of Jesus’ teaching is intrinsically highly memorable, I submit that the only thing standing in the way of a strong case for Jesus’ teaching being passed on effectively in dozens of streams of oral tradition is prejudice.¹⁶ The surprise, then, is not that we have on occasion so many (two, three, or even four) slightly different versions of the same saying. The surprise is that we have so few. It seems to me that the evangelists may well have faced, as a major task, the problem not so much of how to cobble together enough tradition to make a worthwhile book, but of how to work out what to include from the welter of available material.¹⁷ The old idea that the evangelists must have included everything that they had to hand was always, at best, a large anachronism.¹⁸

The material available would, then, have been ‘oral history’, that is, the often-repeated tales of what Jesus had said and done. This is to be distinguished from ‘oral tradition’ proper, according to which a great teacher will take pains to have his disciples commit to memory the exact words in which the teaching is given.¹⁹ If that had been Jesus’ intention, and the disciples’ practice, one might have supposed that at least the Lord’s Prayer, and the institution narrative of the eucharist, would have come out identical in the various versions (in Paul as well, in the latter case) that we now possess.²⁰ Jesus, it seems, did not act as a rabbi, saying exactly the same thing over and over until his disciples had learned it by rote. He acted more as a prophet, saying similar things in a variety of contexts; not only his disciples, but most likely many of his wider circle of followers, would have talked about them, in their own words, for years to come. And it is morally certain that they would, without even thinking about it, have cast this material into a variety of forms, which we can now observe in their eventual literary contexts.

There remains, therefore, a valid and indeed vital task for form-criticism to perform, once it has shed unnecessary assumptions. The critics of form-criticism have not, to my knowledge, offered a serious alternative model of how the early church told its stories.²¹ Here there has been a hiatus in gospel study. The heyday of form-criticism coincided with the heyday of a Hellenistic history-of-religions ‘explanation’ for the New Testament. When the latter gave way, after the Second World War, to a Jewish history-of-religions hypothesis, enthusiasm for form-criticism was on the wane in any case. Neither the redaction-criticism of the 1950s and 1960s, nor the serious study of Jesus of the 1970s and 1980s, needed it; indeed, it has sometimes been pointed out that if the redaction-critics were right, that is, if the evangelists really took as much liberty with their material as they seem to have done, the chances of finding pre-literary forms in their ‘pure’ state is fairly limited.²² The revival of interest in form-criticism in recent years has taken place, perhaps not surprisingly, within the revived Bultmann school, offering updated versions of the Hellenistic hypothesis.²³ Without wishing to prejudge the issue, it seems to me that a prima facie case could be made out for a different approach.

One final preliminary issue must be faced in relation to form-criticism. It has often been supposed that the best word to describe what the early tradition produced was ‘myth’. The reason for this is clear: communities, as we have often remarked, tell stories, often about remote antiquity, as a characteristic way of articulating their worldview and maintaining it in good repair. Subversive groups and individuals within societies tell variants on these myths as a way of advancing their modification of the worldview or, more radically, their replacement for it. It is quite clear that the stories about Jesus which circulated in the early church functioned in some such ways vis-à-vis the early Christian communities and the Jewish communities from which, initially, they sprang. Therefore, if that is what we mean by ‘myth’, that is indeed what these stories, quite manifestly, are.

Unfortunately, things are not so straightforward. As is now often pointed out, Bultmann confused this sense of ‘myth’ (legitimating quasi-historical stories) with several others, notably the ‘myth’ that primitive people use to explain ‘natural’ phenomena (e.g. ‘Thor is hammering’ in reference to thunder). He also added the notion of ‘myth’ as the projection of an individual human consciousness on to reality.²⁴ The first of these additional senses he was able to relativize as ‘primitive’; the second (projection), as claiming a spurious objectivity. But this does no justice to the reality. First, many societies, modern as well as ancient, have held worldviews in which, however distasteful to Enlightenment thought-forms this may be, the idea of divine activity and that of space-time events are somehow held together. Describing the language-systems of such worldviews as ‘mythological’ may help to alert us to the way the systems work; it cannot function of itself as a critique. Second, we spent some time in Part II demonstrating that the empiricist critique of the possibility of knowledge of extra-mental, and extra-linguistic, worlds cannot carry the day. Humans do ‘project’ on to reality, but not everything that they say, not even when they are articulating worldviews, can be reduced to terms of that projection.

There is, in fact, an essential irony to Bultmann’s analysis of the material in the gospels. He was right to see apocalyptic language as essentially ‘mythological’, in that it borrows imagery from ancient near eastern mythology to clothe its hopes and assertions, its warnings and fears, in the robes of ultimacy, seeing the action of the creator and redeemer god at work in ‘ordinary’ events. But he was wrong to imagine that Jesus and his contemporaries took such language literally, as referring to the actual end of the space-time universe, and that it is only we who can see through it and discover its ‘real’ meaning. This is the mirror-image of the mistaken idea that the stories about Jesus, which are prima facie ‘about’ Jesus himself, were really, in the sense described above, ‘foundation myths’ and nothing more. Bultmann and his followers have read metaphorical language as literal and literal language as metaphorical. We must note once more that almost all language, and especially that dealing with those things in which human beings are most involved at a deep personal level, is metaphorical, or at least laden with metaphors. That last clause is itself a metaphor, suggesting that the abstract entity ‘language’ is like a cart, piled high with another abstract entity, ‘metaphors’. Specifically, the language of myth, and eschatological myths in particular (the sea, the fabulous monsters, etc.), are used in the biblical literature as complex metaphor systems to denote historical events and to invest them with their theological significance (see chapter 10 above). The language functions as a lens through which historical events can be seen as bearing the full meaning that the community believed them to possess. However foreign to post-Enlightenment thought it may be to see meaning within history, such language grows naturally out of Israel’s basic monotheistic and covenantal theology. To fail to see this—to imagine, for instance, that the New Testament writers were the prisoners of a primitive, literalistically circumscribed supernaturalist worldview—is simply a gross distortion.²⁵

A further point about myths also tells quite heavily against the theory that much of the gospel tradition consists of them. Myths of the basic kind Bultmann envisaged (quasi-folk tales, articulating the worldview of a people) characteristically take a long time to develop, at least in a complex and intricate form. But the first generation of Christianity is simply too short to allow for such a process. This point has been made often enough, but it is still necessary to repeat it. The hypothesis about the early church necessary to support the idea that the first Christians told ‘foundation myths’ to legitimate their faith and life is far too complex to be credible. On Bultmann’s view of Mark, two strands of thought were developing independently. On the one hand there was the early Christian experience, orientated away from the past—including the past of Jesus—and towards the present and the future. This was quickly translated into Hellenistic categories, becoming the Hellenistic kerygma so famous in the twentieth century and so unknown (perhaps) in the first.²⁶ Within this kerygma, ‘Jesus-stories’ were invented or possibly adapted for the needs of the community. Meanwhile, there were a few real ‘Jesus-stories’ still floating loose in the memory of some early Christians. What Mark has done is to produce a combination of these, expressing the Hellenistic kerygma in terms of Jesus-stories, i.e. using material which might conceivably be historical but without actually intending to refer to Jesus himself. (For Bultmann, this was a brilliant move on Mark’s part: for Mack in our own day, a disaster.) Such a scheme is incredibly complex, and looks very much like a reconstruction designed to save the phenomena of the gospels without damaging a hypothesis (that Jesus was a certain type of person and that the early church was uninterested in him) which is threatened by the actual evidence. When it is suggested that these developments took place within forty years at the outside, it becomes not just incredibly complex but simply incredible.

The gospels, then, are ‘myth’ in the sense that they are foundational stories for the early Christian worldview. They contain ‘mythological’ language which we can learn, as historians, to decode in the light of other ‘apocalyptic’ writings of the time. But they have these features because of their underlying, and basically Jewish, worldview. Monotheism of the creational and covenantal variety demands that actual history be the sphere in which Israel’s god makes himself known. But this means that the only language in which Israel can appropriately describe her history is language which, while it does indeed intend to refer to actual events in the space-time universe, simultaneously invests those events with (what we might call) trans-historical significance. Such language is called ‘mythological’, if it is, not because it describes events which did not happen, but because it shows that actual events are not separated from ultimate significance by an ugly ditch, as the whole movement of Deist and Enlightenment thought would suggest, but on the contrary carry their significance within them.

(i) Introduction

All theories about the history of tradition are parasitic upon assumptions about Jesus and the early church, and we must say something about each in order to establish some ground rules. To begin again with Jesus: Jesus was born, lived, worked and died within a Jewish environment. This world was, as we saw, permeated with Hellenistic influence, but that is not a reason to marginalize the deep and rich Jewishness of his context. Further, a good deal of Jesus’ teaching, on almost anybody’s showing, had to do with the coming kingdom of Israel’s god. I shall argue in the next volume that this Jewish context makes sense—more sense than the currently fashionable Cynic alternatives—of Jesus’ agenda and of the reasons for his death; until very recently, all the main work on Jesus simply assumed this. Second, we have seen that the major books written about Jesus later on in the first century all took the shape of the Jewish story for granted, and interpreted Jesus within that framework. So did Paul. Significantly, even when Christians of a somewhat later date (Ignatius, Justin, Polycarp) were facing their own world, that of paganism, they held tenaciously to a form of Christianity which remained recognizably Jewish. It is true that, at least by the mid-second century, quite different strains had appeared. We shall look at some of them presently. But the case needs to be spelled out for seeing the whole of the first generation of Christianity as essentially Jewish in form, however subversive of actual Judaism it was in content.²⁷ If Jesus was Jewish, and thought and acted within a world of Jewish expectations and understandings of history; if Paul did the same; if the synoptic evangelists and even John retold the Jewish story so as to bring it to its climax with Jesus; and if even in the second century, even out in the pagan world, Christianity still bore the same stamp—then it seems highly likely a priori that the early tellings of stories about Jesus would also carry the same form. What we need, and have never had in the history of the discipline, is a hypothesis that would show at least the possibility of a Jewish form-criticism of the synoptic tradition, a reading of the stories which did justice to the high probability that their earliest form was Jewish, and that Hellenistic features may be signs of later development.

The question may be sharpened up as follows. Did the evangelists ‘Judaize’ a tradition which, until Mark (or whoever) got hold of it, did not have the Jewish story-line as a basic component? This has been the hypothesis of the Bultmann school all along, supported in our own generation by a good many studies which first strip off the ‘Jewish’ elements from a passage and then purport to show its ‘original’ Hellenistic, perhaps Cynic, meaning.²⁸ I suggest that this makes no sense historically. The sayings do not for the most part exist in that simple form, except sometimes in the Gospel of Thomas, which ought to be sub judice at this point.²⁹ To put it crudely, if Christianity had started off as a not-very-Jewish movement, it is extremely implausible to suggest that it suddenly developed a penchant for rereading Jesus within a Jewish framework, just when, because of AD 70, that Jewish framework must have seemed singularly unattractive to Christians who had up till then had little time for Jewish tradition anyway. I suggest, on the contrary, that the high historical probability is the other way about, namely that (i) Christianity began within a firmly Jewish context; (ii) its earliest tellings of stories about Jesus naturally fell into recognizably Jewish forms; (iii) it was only when the stories began to be told in other contexts, where recognition of or familiarity with Jewish storytelling modes was not so likely, that they began to take on more obviously non-Jewish forms. That such a development is intrinsically likely is clear from the parallel with Josephus, who, as we saw, transforms Jewish traditions and ideas at many points into apparently Hellenistic ones. This hypothesis of development, like most others, is obviously too simplistic. There was certainly movement and change in all sorts of directions. But if we may hypothesize a general drift, it is far more likely to be from Jewish to Greek, not vice versa.

Classical form-criticism began with forms taken from the world of Hellenistic literature: ‘apophthegms’, wonder-working miracle stories, and so forth. (An ‘apophthegm’ is a short story leading up to a pithy saying; another, perhaps better, word for it is ‘chreia’.³⁰) There are parallels to these in some Jewish writings, for instance in the rabbinic sources, but these are generally much later. To these, the classic form-critics added ‘parable’, of course, and other less well-defined categories such as ‘legend’ and the now notorious ‘myth’. There was, and still is, much debate as to whether the ‘pure’ form of a story was likely to be original, and a more complex form a later development (as Bultmann thought), or whether to the contrary the ‘pure’ form represented a smoothing off, over time, of an originally rough-hewn tradition (as Taylor suggested). There is a lot to be said for reckoning with the possibility that traditions both expand and contract, and for refusing to accept simplistic developmental theories either way.³¹

From our earlier study of stories within the Jewish world of the first century, a case can be made out that we know what sort of stories Jesus’ contemporaries regularly told and retold, forming a grid through which they perceived the whole of reality.³² They told stories of Israel’s suffering and vindication; of exile and restoration; of Passover, exodus, wandering and settlement. They told stories about Israel’s god coming to redeem his people; about prophets and kings whose mighty acts were signs of this divine liberation; of the biblical stories coming true, whether secretly or openly. These are, of course, descriptions of content, not of form; yet in this case content powerfully suggests form. An excellent example of this occurs in 1 Maccabees’ account of the state of things under the rule of Simon (140–34 BC). Instead of saying ‘Israel lived in peace and prosperity’, the writer chooses to tell the story in words which awaken all sorts of prophetic echoes:

They tilled their land in peace;

the ground gave its increase,

and the trees of the plains their fruit.

Old men sat in the streets;

they all talked together of good things,

and the youths put on splendid military attire.

He supplied the towns with food,

and furnished them with the means of defense,

until his renown spread to the ends of the earth.

He established peace in the land,

and Israel rejoiced with great joy.

All the people sat under their own vines and fig trees,

and there was none to make them afraid …

He made the sanctuary glorious,

and added to the vessels of the sanctuary.³³

Thus, in typical Jewish manner, the story was invested with its full significance.

Moving back to the gospel traditions, our earlier remarks about oral history suggest that the place to start a form-critical investigation ought to be with the questions: how did Jesus’ contemporaries perceive him?; how would people of that background tell stories about someone whom they perceived in that way?; and, what forms would such stories naturally take? If we proceed by this route—which has an excellent prima facie claim to be taken seriously from a historical point of view—the answers are striking.

(iii) Controversies

If Jesus was perceived as a prophet, he was also perceived as the focal point of some sort of new group, movement or perhaps sect within Judaism. From our earlier study of sects and groups it is clear that such bodies would tell and retell the biblical stories in which a righteous Jewish remnant stands up for Israel’s god against pagan (or even Jewish) authorities, and is vindicated.³⁹ A group that had already begun to regard itself like this, and which was fortifying itself with such retellings, would naturally interpret opposition to itself or its leader in terms of biblical stories that provided a precedent. One way of doing this is clearly visible in the Scrolls: the book of Habakkuk is mined for secret clues about the battle between the Teacher of Righteousness and the Wicked Priest on the one hand, and between the community and the Kittim (i.e. the Romans) on the other.⁴⁰ This gives a parallel in content, though not in form, to the synoptic stories; the form of the Habakkuk commentary reflects a quasi-academic exercise arising out of an incident, rather than, as with the synoptic material, a retelling of the incident itself.

I suggest that the controversy stories in the synoptic gospels may well have started life in a similar setting—not the academic exercise, but the sense of a small group perceiving opposition to itself in the light of biblical precedent. It is highly likely that Jesus met with opposition during his work; this cannot be substantiated in detail here, but will be spelled out in the next volume. It is also extremely probable that such opposition, when it occurred, did not take the form of a single question or challenge, answered by a single remark from Jesus—which is of course the form of most controversy stories in the gospels. Debate is likely to have raged to and fro, more like the protracted and frankly rambling discussion in John 6 than like any of the well-known stories in Mark 2 and 3. It is, further, highly probable that a group like that around Jesus, regarding itself as the nucleus of a divinely called remnant, as some sort of spearhead of renewal, would instantly ‘read’ any opposition in terms of the battle, with clear scriptural precedent, that all such groups would expect to fight. It is therefore to be expected that, as they told and retold the story of this or that brush with opposition, whether official or self-appointed, they would naturally both shrink the scene to its bare essentials and choose as those bare essentials the elements which highlighted the group’s identity as the beleaguered renewal movement awaiting vindication. If a group has fuelled its hopes on biblical tales in which the true Israelites are called to account by the authorities, and subsequently vindicated, and if this group then perceives itself to be in an analogous situation, it will ‘see’ the events in question through this grid of expectation, and will be highly likely to tell the story of the events in a way which reflects the same form.⁴¹

I suggest, therefore, that the ‘apophthegms’, ‘paradigms’, ‘pronouncement stories’, ‘chreiai’, or whatever we like to call them, should have their form-critical history reassessed, and that the normal (i.e. Bultmannian) way of reading them should be stood on its head.⁴² These stories are of course normally regarded as having started out life as detached individual sayings, which gradually acquired a story-line in order to provide a more fitting showcase for the telling rejoinder of Jesus to his opponents. The early church is supposed to have remembered (or created) isolated sayings for isolated needs, and only gradually thought of giving them a narrative framework.⁴³ I suggest that, within the original contexts, the most natural Jewish way for supporters of Jesus to tell stories about his controversial actions and words was to tell them in the form of Jewish controversy-stories, such as we find (for instance) in the book of Daniel. These contexts can be explored one by one. At one level, there is the context of Jesus’ ministry itself. Stories would be told quickly, excitedly, when Jesus was still in the same village. At another level, there are memories of events concerning him which were retold during his lifetime when he had moved on elsewhere. At another level again, there are the memories of him which were cherished after Easter. In each case, the pattern of the story would be the same, and would carry the same echoes of Jewish tradition. The true Israelites act boldly; they are challenged by real or self-appointed authorities; they stick to their guns, often with a well-chosen phrase; and they are vindicated. In the greatest Danielic controversy-and-vindication story of all, it is ‘one like a son of man’ who is vindicated. The gospel stories have exactly this form, and as often as not end with a reference to Jesus as the ‘son of man’.⁴⁴

If the earliest form of the controversy-stories is therefore likely to have been that of the Jewish stories of the struggles and vindication of the little remnant or renewal movement, it is not difficult to see how these stories could have become smoothed down over time into something more like Hellenistic chreiai, especially as the news of Jesus passed beyond the area where Jewish-style controversy-and-vindication stories would be an expected form. This, I suggest, is the most likely explanation for works like the Gospel of Thomas. So far from isolated sayings becoming short chreiai, and then longer stories, the opposite process seems to me historically far more probable. The greater number of such isolated sayings in Luke may well serve as initial confirmation of this suggestion.⁴⁵

3. Stories but no Story? Q and Thomas

For fifty years, covering the middle of the present century, more or less all New Testament scholars believed that the main sources of Matthew and Luke were Mark and a lost document which may for convenience be called Q.⁵⁷ The hypothesis has declined in popularity over the last twenty years or so, but despite strong attacks is still equally strongly defended.⁵⁸ But its role in scholarship has undergone, for the most part, a radical and fascinating change.

At the same time, it has long been the received wisdom among students of early Christianity that the so-called Gospel of Thomas, a collection of sayings of Jesus which was found among a collection of Coptic codices at Nag Hammadi in Upper Egypt in 1945, belongs to a comparatively late stage in the development of the Christianity.⁵⁹ It is first mentioned by Hippolytus and Origen in the early years of the third century. It is a random collection of short sayings, almost all introduced simply with the phrase ‘Jesus said …’, and without any attempt at a connecting narrative, even at the beginning or the end. Like Q, its role in New Testament scholarship has changed dramatically within the last few years.

The original proposers of the Q hypothesis came to it by a detached, almost mathematical argument. Matthew and Luke overlap in a good many places where there is nothing in Mark. It makes sense to suggest that in these passages they were following a common document, now lost. ‘Q’ was a way of producing a simple hypothesis which appeared to fit the evidence. Q clearly pre-dated Matthew and Luke, and so was early. The Oxford scholars whose work on the synoptic problem was eventually crystallized in Streeter’s major work The Synoptic Gospels were for the most part interested in Q as a means of gaining access to the historical—by which they would have meant the ‘real’—Jesus. Mark, they thought, gives us an outline of the ministry and the main events; Q, a solid collection of Jesus’ sayings. Both can be accepted confidently as going back, more or less, to Jesus himself. The additions in Matthew and Luke are less certain, but not so much hinges on them if the agenda of the moment is (as it was for Streeter and his colleagues) the defence of something like traditional Christianity against the ravages of David Friedrich Strauss and his successors.⁶⁰

The Q hypothesis held sway, virtually unchallenged, for half a century. Solid doubts were expressed from outside the mainstream guild, and small pinpricks were aimed at it from within;⁶¹ however, the consensus was not disturbed. Farmer’s book (1964) was the first to shake the foundations, and they are still regarded as shaky in many quarters today.⁶² But, just when debate on the synoptic problem looked like becoming simply an irrelevant crossword-puzzle, a new motive has emerged for the study of Q, strong enough to give fresh impetus to what had seemed almost a dying cottage-industry.⁶³

Q, it is believed in some quarters today, not only existed as a document, but developed in a way which can be plotted with some accuracy. Moreover, its original existence does indeed take us quite close to Jesus, but it is a Jesus quite unlike the Jesus that Streeter and his colleagues thought they would find by this route. In its original form, Q reflects a very early Christian community for whom the Jewish stories, both in form and in content, were not particularly important. The focus, instead, was on a different style and content of teaching: the Hellenistic philosophy known as Cynicism, on the one hand, and, on the other, a tradition of teaching which offered a secret wisdom, a secret Gnosis. It was, in fact, a community that would have been just about as happy with the Gospel of Thomas. Jesus was a teacher of aphoristic, quasi-Gnostic, quasi-Cynic wisdom; his first followers collected his sayings in the way that one would expect with that sort of teacher—as, for example, Arrian did with Epictetus. Q is one result of this process. Its later use by Matthew and Luke represents an attempt, which may have already begun in the later strands of Q, to blend these non-historical, aphoristic and potentially Gnostic teachings with a Jewish-style telling of the Jesus-stories, and the Jesus-Story, that offers a fundamental change of direction and emphasis to that embodied in at least the early strands of Q itself.⁶⁴ It should be stressed that by no means all contemporary supporters of the Q hypothesis would take this line. Some continue to insist that the document(s) must be seen as prophetic and Jewish-Christian, rather than within a Cynic, Stoic or Gnostic matrix.⁶⁵ But the majority of recent Q scholarship is firmly within the tradition I have just described.

This Q-and-Thomas hypothesis belongs closely with, and both reinforces and is reinforced by, a way of telling the story of early Christianity which runs as follows.⁶⁶ The earliest stage, dated within ten to thirty years of the death of Jesus, is represented by the first version of Q. This is basically a ‘sapiential’ document, offering its readers the choice of two ways, the way of wisdom and the way of folly. At this stage, Q knows nothing of an apocalyptic future expectation, and is silent about the ‘coming of the son of man’: the End, in the only sense that matters, has already come with the teaching of Jesus, specifically with the purveying of a special and hidden wisdom which sets his hearers apart from the rest of the world. This is, as it were, a ‘vertical’, rather than a ‘horizontal’, eschatology. The ‘End’ has nothing to do with the events for which Israel was waiting, but rather with a fresh, secret divine revelation. No connection is made with Jewish expectations: there is no controversy between the followers of Jesus and those of John the Baptist.

This (hypothetical) first stage of Q is therefore very close in substance, as well as quite close in form, to the Gospel of Thomas. Thomas knows nothing of an apocalyptic future: the hidden wisdom in the present is what matters—a ‘realized eschatology’ that actually has nothing to do with Jewish ‘eschatology’ per se, and everything to do with a secret or hidden revelation that Jesus imparts to his followers in the present. So close are Thomas and Q that a saying which occurs in only one of either Matthew or Luke can be confirmed by a parallel in Thomas as a Q saying, omitted by the other evangelist, rather than special material of Matthew or Luke.⁶⁷

Q then undergoes a redaction. At this stage the motifs previously absent are introduced: ‘the most obvious signs of a secondary redaction of Q can be found in the apocalyptic announcement of judgment and of the coming of the Son of man which conflicts with the emphasis upon the presence of the kingdom in wisdom sayings and prophetic announcements’.⁶⁸ This leaves some traces at the level of form: a collection of sapiential speeches becomes a wisdom-plus-apocalyptic book. But, despite the protests of its proponents, it is clear that the basic difference is theological. The original Q had a realized eschatology, i.e. a ‘kingdom of god’ in the here and now. The redacted Q has a future-orientated, much more Jewish, eschatology.⁶⁹

At about this stage, some little while after AD 70, Luke makes use of Q. It has long been believed by those who support the Q hypothesis that Luke is closer to the supposed ‘original Q’ than Matthew: where these two canonical gospels overlap with ‘Q material’, Luke’s version is regularly supposed to be the earlier and less developed.⁷⁰ But this does not mean that the ‘developments’ evidenced in Matthew are simply the work of the evangelist himself. A further version of Q seems to have been produced between Luke’s use of it and Matthew’s: at this stage the (second) Q redactor, reflecting perhaps a decision by his community, comes down in favour of Christians continuing to observe the Jewish Torah, something that was not envisaged in the early period. Finally, Matthew makes use of this doubly-redacted Q, weaving it into new patterns of his own devising.⁷¹

What is to be said about this detailed and intricate hypothesis? Some might find it easy to pour scorn on it. ‘Q’ is no more, after all, than a figment of scholarly imagination (i.e. a hypothesis). Not one scrap of manuscript evidence has turned up which can plausibly be thought of as part of this document, in any of its recensions. The three supposed stages by which it came into its final form, visible in Matthew, reflect suspiciously closely the theological and history-of-religions predilections of one strand within modern New Testament studies, rather than any hard evidence within the first century. The neat ‘coincidence’ whereby the earliest form of Q happens to have so much in common with the Gospel of Thomas could conceivably be due to (a) the theory’s having made a brilliant guess at how to strip off redactional layers and (b) a very early date for Thomas. Almost anything is possible in history, and one cannot rule out this solution from the start. But equally—some would say, far more likely—it may be due to (a) a desire on the part of some modern readers to imagine early Christianity as very similar to the religion of Thomas and (b) the consequent critical activity of pulling apart what is already a purely hypothetical document, Q, in order to reveal an ‘early version’ which just happens to fit the history-of-religions theory. Frankly, some of the arguments advanced along the way within the suggested model look suspiciously naïve; others, manifestly and damagingly circular.⁷²

These general problems about the Q-plus-Thomas hypothesis lead to some more particular ones, to do with the categories involved. First, by what stretch of the imagination can we insist, in relation to a hypothetical document whose origin is supposedly found in otherwise unknown Jewish-Christian groups in first-century Palestine, that a firm distinction could be made between ‘sapiential’ and ‘prophetic’ traditions? Koester wisely attempts to hold them together;⁷³ yet elsewhere in recent discussion they are firmly set over against one another.⁷⁴ But once we admit ‘prophetic’ material in Q, it becomes extremely difficult, given the high profile of apocalyptic language and imagery in first-century Palestine, to keep ‘apocalyptic’ out as well.⁷⁵ Q looks more and more like a combination of wisdom-tradition and prophetic discourse; and there is every reason, within the actual known history of first-century Palestinian religion, as opposed to the mythical model constructed by some scholars, why such a combination should be perfectly natural.

Likewise, there is every reason not to split apart ‘realized’ and ‘future’ eschatology. To drive a wedge between the two in the interests of an Early Q that looks to the present and a Late Q that looks to the future is totally unwarranted. At this point we may refer to a well-known Jewish book never mentioned by Koester in his recent and full discussion of Q. The Community Rule from Qumran is, like the hypothetical Q, a book which sets out the ‘two ways’ that one may follow, the way of wisdom and the way of folly (3:13–4:26). The same book makes it quite clear that the community is already living in the time of salvation: ‘realized eschatology’ is evident in every column, since the basis of the community’s very existence is its belief that Israel’s god has re-established his covenant with precisely this group. There is even some reason to understand this ‘realized eschatology’ in a vertical as well as a horizontal sense, since the wonderful hymns with which the book ends contain several passages which, if detached from their contexts, would look perfectly at home in the Gospel of Thomas.⁷⁶ The same book, equally, sees a future still before the community: the Messiahs of Israel and of Aaron are yet to come (9:11), and the Day of Revenge is still awaited (9:23). Here we have, I submit, a document which is far closer to Q ‘as we have it’ (i.e. at its most obvious reconstruction out of the non-Marcan portions of Matthew and Luke) than is the Gospel of Thomas. Moreover, it is manifestly a piece of Jewish sectarian literature, as Q must have been if it existed; it is the manifesto, the community rule, of its group, enshrining the teachings of its founder on the one hand and the present ordering, experience and hope of the community on the other: again, Q, if it existed, must have been this sort of book. The Qumran Community Rule offers excellent evidence that the divisions made in the Q material by Kloppenborg and others have no real basis in the actual history of first-century religion, but belong rather to the world of contemporary mythologizing, projecting unwarranted distinctions on to the screen of speculative history.

One of the most telling weaknesses in the whole Q-and-Thomas hypothesis, it seems to me, is the presence within Thomas of sayings about the ‘kingdom of god’, or, as the book regularly calls it, the kingdom of the Father.⁷⁷ From our earlier study of the Jewish evidence, it is unthinkable that this motif should be introduced into a community from scratch with the meaning that it comes to have in Thomas, i.e. the present secret religious knowledge of a heavenly world. It is overwhelmingly likely that the use of this emphatically Jewish kingdom-language originated with an overtly Jewish movement which used it in a sense close to its mainline one, i.e. which spoke of the end of exile, the restoration of Israel, the rebuilding of the Temple, the return of YHWH to Zion, and so forth, however much these ideas were transformed within the ministry of Jesus and the lives of his first followers. If there has been a shift in the usage one way or the other, it is far more likely to have been from this Jewish home base into a quasi-Gnostic sense, rather than from a Gnostic sense, for which there is no known, or imaginable, precedent, to a re-Judaized one—a shift which, on the hypothesis, must have taken place somewhere between an early Thomas and a later Mark.

If this Q-and-Thomas hypothesis creates so many difficulties, does this mean that we should abandon the Q hypothesis altogether? By no means. There are clearly other ways of stating the hypothesis, which are not open to any of the objections just raised, and which focus simply on the attempt to solve the synoptic problem in its own terms. There are also completely different traditio-historical possibilities on offer in relation to Q in current scholarship; one need only cite again the work of Theissen.⁷⁸ There is no reason why some form of this hypothesis should not continue to be fruitful, though I suspect it will be more a matter of ‘Q material’, floating traditions to which Matthew and Luke both had access, than of a solid and reconstructible document. But it does seem to me that the more speculation there is about Q the less plausible the hypothesis appears overall.

This may be unfair: bad coin is perhaps driving out good. But some of the coin is bad indeed. So many unproveable and mutually contradictory things are said about Q that the non-specialist may well feel the whole discussion is a waste of time. In particular, as soon as the argument turns to the potential relationship between Mark and Q, the sceptic must protest: those who first hypothesized Q, after all, saw it simply and solely as consisting of those passages in Matthew and Luke which do not overlap with Mark. Of course, if Q existed, and if it extended beyond those overlaps, as most proponents of Q are now inclined to suggest, it may well have contained all sorts of things for which there is now no evidence. It may, in particular, have contained birth narratives, passion narratives, Peter’s confession, and all sorts of interesting christological, apocalyptic and other material. We simply do not know. Once we admit that any of the evangelists, or any transmitters of written tradition, chose to omit any scrap of evidence available to them, there is no means whatever of being certain about the extent of earlier hypothetical documents.⁷⁹ Supporters of Q should beware; if first-century editors were allowed to omit as well as to add material, the case for saying that Luke simply used Matthew looks more and more plausible.⁸⁰ It would therefore be as well to keep on a tight rein any theories which depend on the significance of, for instance, Q’s not having a passion narrative. Proceeding down that sort of road is like walking blindly into a maze without a map.

Finally in relation to Q: if some sort of Q existed, and if it contained roughly the material it was originally imagined to have contained, is there a more plausible life-setting for it than those sketched above? Here we rely again, as all searches for life-settings do, on assumptions about Jesus and about the subsequent church. These assumptions, to avoid the wrong sort of circularity, must stand up to being tested in the light of material other than simply that which is to be explained. In this process, we cannot ignore the evidence of Paul, or assume that Q belonged to a totally different branch of Christianity to that which he represents. Q belongs, if anywhere, within the early missionary community of Jesus’ followers, but how do we conceive of that group? The early Palestinian Christians, still with a strong awareness of Jesus as a prophet announcing the kingdom, and with a clear sense that this kingdom would subvert the existing Jewish structures, almost certainly announced to their contemporaries that the kingdom for which they had longed had indeed arrived in and through Jesus, even though it was not as they imagined it. However, the all-important difference between expectation and reality was not that the kingdom was Gnostic or Platonic rather than essentially Jewish and historical, but that it was based upon Jesus rather than upon a restoration of Jewish national and ethnic primacy, and that the former actually subverted the latter. This was a message which could naturally draw upon traditions containing ‘apocalyptic’ sayings as well as ‘wisdom’ sayings, embracing between those two many ‘prophetic’ aspects as well. The document, if it existed, may have been more a collection of preachers’ materials than any attempt at a total narrative. Yet if Catchpole and others are right, and Q began with the story of John the Baptist, it may be the case that some sense of the total story of Jesus, rather than simply an abstracted collection of sayings, was present from the beginning. If so, the story in question belongs clearly on the map we have already sketched of both the large and the small stories which characterized the early Christian movement. It will not have been primarily a Cynic-style document. It will have been a Jewish-style story capable of taking on the Hellenistic world. It will have seen Jesus as both the focal point of the Jewish sapiential, prophetic and apocalyptic traditions and the one who had inaugurated the worldwide kingdom of Israel’s god, the creator of the world.⁸¹

Finally in relation to Thomas: it seems to me equally clear that some of the sayings in Thomas are derived from the synoptic tradition and that some, which may well go back a very long way, are independent.⁸² But hardly any of them (except for the parables) contain stories in themselves, and the work as a whole has, at the level of poetic sequence, no plot whatsoever. This in itself is strong evidence of a matrix outside the entire context that we have been considering in the last three chapters.⁸³ If, granted this absence of narrative or plot at the level of poetic structure, we approach Thomas with the insistent question as to its implied narrative world, the reply we get is most instructive. It presupposes a story in which the traditional concerns of Judaism play no part at all. Its implicit story has to do with a figure who imparts a secret, hidden wisdom to those close to him, so that they can perceive a new truth and be saved by it. ‘The Thomas Christians are told the truth about their divine origins, and given the secret passwords that will prove effective in the return journey to their heavenly home.’⁸⁴ This is, obviously, the non-historical story of Gnosticism. As we conclude our discussion of stories in early Christianity it has to be said that Thomas stands out from everything else we have examined. It is simply the case that, on good historical grounds, it is far more likely that the book represents a radical translation, and indeed subversion, of first-century Christianity into a quite different sort of religion, than that it represents the original of which the longer gospels are distortions. Some of the technical terms, and recognizable versions of many sayings, are there, but the essential substance is altered.⁸⁵Thomas reflects a symbolic universe, and a worldview, which are radically different to those of the early Judaism and Christianity which we have studied. In particular, it marks itself out as different from the longer gospels and Paul; different also from the shorter units of which the longer gospels are mostly composed; and even (if indeed it ever really existed) different from Q.