For whom were Gospels written

For whom were Gospels written?

© Richard Bauckham

This paper, which was published in Hervormde Teologiese Studies 55 (1999) 865-882, is a summary of the argument presented in more detail in chapter 1 of Richard Bauckham ed., The Gospels for all Christians: Rethinking the Gospel Audiences (Grand Rapids: Eerdmans/Edinburgh: T. & T. Clark, 1997). The paper presented here lacks footnotes and references: these can be found in the longer version in The Gospels for all Christians.

For the SBL session on this subject I hope to be able to contribute also some further reflections, in the light of what has been written since The Gospels for all Christians was published. But since ill-health has so far prevented me from doing so, this paper is offered for the time being. Please look at the web-site again later, when I hope I shall have been able to add my more up-to-date contribution.

My title - For whom were Gospels written? - could be analysed into two distinct questions, only one of which I intend to tackle this morning. One could ask: Were Gospels written for Christians or for non-Christians? This question has sometimes been discussed, particularly in the case of the Gospels of Luke and John, since a minority of scholars have argued that those Gospels were written as apologetic or evangelistic works, not for Christians but for outsiders. On this question I shall go with the general consensus, that all Gospels were intended primarily for Christians, without arguing that point. It does deserve to be argued, but I have another agenda this morning. I will only say that it seems to me that, if any of the evangelists did envisage reaching non-Christian readers, they would have to have envisaged reaching them via Christian readers, who could pass on copies of Gospels to interested outsiders through personal contact. So the Christian audience would in any case remain primary.

The second question one could ask, and the one I invite you to ask this morning is: Were the Gospels written for a specific Christian audience or for a general Christian audience? Was, for example, Matthew written for Matthew's own church, the so-called Matthean community, or was it written for the purpose of circulating widely around the churches? Are a Gospel's implied readers a specific Christian community, or are they the members of any and every Christian community of the late first century to which the Gospel might circulate? Whereas my first question has sometimes been discussed, with some substantial arguments deployed in its discussion, this second question is remarkable for having never, so far as I can tell, been discussed. No space remotely approaching even the scope of this lecture this morning has ever been devoted in print to arguing the case one way or the other.

The point is not of course that this question is not relevant to the concerns of current or recent Gospels scholarship. Quite the opposite. One of the two possible answers to this question - the option that each Gospel was written for a specific Christian community - has been taken entirely for granted in Gospels scholarship for some decades now. As an assumption on which arguments about the Gospels are based, it has come to play a more and more dominant role in Gospels scholarship, which since the late 60s has become increasingly interested in reconstructing the circumstances and character of the community for which, it is assumed, each Gospel was written. Almost all contemporary writing about the Gospels shares the unargued assumption that each evangelist, himself no doubt a teacher in a particular church, wrote his Gospel for that particular church, with its particular situation and character and needs at the forefront of his mind. The so-called Matthean, Markan, Lukan or Johannine community (or for that matter, Thomas community) may be understood as, not just one church, but a small group of churches, but in that case it is axiomatic that this group of churches be homogeneous in composition and circumstances. The unargued assumption in every case is that each Gospel addresses a localized community in its own, quite specific context and character.

Nearly all the literature of the last few decades which makes this assumption and increasingly builds large and highly sophisticated arguments upon it seems to regard this assumption as completely self-evident, as though no alternative could ever have occurred to anyone. There is, of course, a perfectly obvious alternative possibility: that an evangelist writing a Gospel expected his work to circulate widely among the churches, had no particular Christian audience in view, but envisaged as his audience any church (or any church in which Greek was understood) to which his work might find its way. This is the possibility which I hope to convince you deserves at least to be given serious consideration. The purpose of my paper is not simply to challenge the established consensus but to open up a discussion that has never so far taken place. Not only has noone apparently ever (all these statements, of course, have 'so far as I know...' as their implicit qualification - if there are debates about this in literature I've not yet discovered, I shall be delighted to know of them.... But in the present state of my knowledge I have to say:...) Not only has noone apparently ever argued for the consensus view in more than a few sentences; it is also the case that noone has ever argued for the alternative view, which I shall propose as more plausible. There has never been any debate. If I can only convince you that there ought to be a debate, I shall be happy to have accomplished that much.

Challenging a scholarly consensus is always, of course, rather fun - it has a whiff of intellectual excitement about it and more than a little danger of intellectual arrogance. It also encounters an obstacle: that only those hearers who have a naturally iconoclastic attitude to these things will be already disposed to favour it. Most of you, being yourselves immersed in the consensus, will be more disposed to think that a consensus which is not only so universally accepted but which has proved so fruitful in generating exciting and interesting work on the Gospels must be right. Any argument against this kind of consensus has an uphill struggle merely to gain an unprejudiced hearing, if there were such a thing.

So I shall begin by offering, as it were, a warm-up argument, whose function is merely to sow an initial seed of possibility that there might perhaps be something to be said for the view I shall propose. I put this argument in a form which presupposes the most widely accepted view of Synoptic relationships, but it could easily be restated to accommodate any theory of Synoptic relationship (none of the argument of this lecture depends on any particular theory of Synoptic relationships). But the present argument has to be stated in one form or another. So, assuming Markan priority, how is it that Matthew and Luke both had Mark's Gospel available to them? Noone imagines all three evangelists belonged to the same local Christian community. So the usual view (I have never seen any other suggestion) is that by the time Matthew and Luke wrote, Mark's Gospel had already circulated quite widely around the churches and was being read in the churches to which Matthew and Luke respectively belonged. This is a very reasonable view, since we know quite certainly that at a slightly later date Mark's Gospel was known in churches other than Mark's own, wherever that was. Matthew and Luke, in other words, knew Mark as a Gospel which had in fact circulated quite widely among the churches, and was proving to be useful and valued in many Christian churches. Whatever Mark had meant his Gospel to be, his work, when Matthew and Luke knew it, had already in fact come to be used and valued, not as a work focused on highly particular circumstances in Mark's own community, but as a work generally useful to various different churches. Surely Matthew's and Luke's model for what a Gospel was must have been Mark as actually circulating and used in the churches. They must have expected their Gospels to circulate at least as widely as Mark's had already done. They must have envisaged an audience at least as broad as Mark's Gospel had already achieved. Most likely Matthew and Luke each expected his own Gospel to replace Mark's. To suppose that Matthew and Luke, knowing that Mark's Gospel had in fact circulated to many churches, nevertheless each addressed his own Gospel to the much more restricted audience of his own community, seems to me prima facie very improbable. Such a view would need rather careful argument, and certainly should not be treated as a self-evident axiom.

This warm-up argument is a very simple one, and no doubt its simplicity is the reason why the literature, so far as I can tell, never deigns to notice it. Maybe there is an equally simple refutation of it, which you will put to me when I've finished. Meantime, I shall move on, first of all to some brief remarks about the history of scholarship, before sketching my own argument for the Gospels as literature written for all the churches.

The way the current consensus on this issue has come about, without anyone ever having seriously argued the case for it, would make a most interesting topic for study in the history of NT scholarship. It could also provoke reflections, perhaps rather disturbing reflections, about the sociology and psychology of NT studies as a discipline. I have done only a little research on the history of views about the audience of the Gospels; so what I say now is fairly provisional. I haven't time to give you any of the detail, but I'll make a few general points which I think have some importance for my argument.

First, it seems to me that the view that each evangelist wrote for his own community is an old view in British scholarship, going back at least to the end of the nineteenth century, though it was not the only view in older British scholarship. My guess is that this view is much more recent in German scholarship, though I could be wrong. However, this old notion about the implied audience of the Gospels only began to matter and only became influential when some of the redaction critics of the late 60s began actually to read the Gospels as addressed to specific circumstances of the each evangelist's community. At that point, the assumption, previously confined to discussions of introductory questions, became the basis for interpretative strategies which found the specific circumstances and needs of a particular community addressed in a Gospel. More recent social-scientific studies of the Gospels are in this respect directly continuous with redaction criticism. Though asking different questions about the relationship between a Gospel and its original audience, they have taken over without question the same assumption about the definition of the implied audience.

Secondly, the context in which might expect to find arguments for the view which has become the consensus is therefore discussion of the conventional set of introductory questions about Gospels: date, place, readership and so on. In fact, one soon discovers that the tradition of discussing such questions has inherited and deploys an assumption that the question about the context in which a Gospel was written and the question about the audience for which a Gospel was written are the same question. Such discussions therefore regularly and systematically confuse the evidence for these two different questions. Good recent examples are Fitzmyer on Luke (AB), Davies and Allison on Matthew (ICC). This kind of confusion of issues on the basis of assuming precisely what needs to be proved goes back through, for example, B W Bacon's book on the origins of Matthew's Gospel (1946) to B H Streeter's The Four Gospels (1924).

Thirdly, I need to address this question: Even if I am right that the assumption that each Gospel was written for the evangelist's own community has come to be widely accepted largely without having been argued, might one not suppose that this assumption has been confirmed by the results which Gospel scholarship has built upon it? A large body of literature has been devoted to reconstructing each Gospel's own community and illuminating each Gospel by reading it as addressed to that reconstructed community, with its particular theological views and debates (the main concern of earlier redaction criticism), its particular social composition and social context (the concern of more recent study with social-scientific ingredients), even its own history (elaborately reconstructed in Johannine scholarship especially). Now a properly argued case for the view I am disputing would have to draw on this work, but the work itself does not constitute such a case. With very occasional exceptions in detail, this body of scholarship does not proceed by arguing that certain features of a Gospel text are explicable only if understood as addressed to a specific Christian audience rather than to a general Christian audience. Its results are the results of applying to the text a particular reading strategy, not of showing that this reading strategy does better justice to the text than another reading strategy.

Let me illustrate my point by observing what goes on in typical instances of this reading strategy. One form of it consists in applying to a specific Christian community textual implications which would readily apply to a very large number of Christian communities. Take, for example, J. Louis Martyn's classic argument that chapter 9 of the Fourth Gospel should be read, on one level, as a narrative of the Johannine community's expulsion from its local synagogue. Does this constitute evidence that the Gospel addresses the specific situation of the evangelist's own community? No, not even if one wholly accepts Martyn's account of when and how the expulsion of Jewish Christians from synagogues occurred. Precisely Martyn's own argument, that the introduction of the Birkat ha-Minim into synagogue liturgy late in the first century had the effect of forcing Jewish Christians out of synagogues, is an argument for a general process which, if he is correct, must have been going on in many diaspora cities where Jewish Christians had previously attended synagogue. If John 9 addresses that situation, it addresses, not a circumstance peculiar to the Johannine community, but a circumstance which would have been common in the churches of the late first century. Only because Martyn starts with the presupposition that the Fourth Gospel was written for the Johannine community, and has no intention of trying to prove this point, can his argument function for him to characterize only the Johannine community's relationship to the synagogue.

The same consideration applies to many such arguments. Probably most Christian communities in the period when the canonical Gospels were being written were located in cities, contained both Jewish and Gentile members, including Gentiles who had been attached to the synagogue, and included some people, even if not many, from both ends of the socio-economic spectrum.

If it is objected that such features, while not confined to one specific community, would still not have been true of every Christian community, then it is time to introduce the second aspect of the reading strategy which I observe in such arguments. This consists in supposing that all textual indications of the character and circumstances of the audience must all apply to the whole of the implied audience. Then one need only compile all such indications in order to produce an identikit description of the evangelist's community. However, supposing the Gospels were written for general circulation and therefore envisage the range of audiences their authors might expect them to acquire in the churches of the late first century, then there is no reason at all why every aspect of a Gospel should be equally relevant to all readers or hearers. An evangelist might well address features of Christian life and social circumstances which he knew to be fairly widespread in his time, without supposing his Gospel would therefore have no appeal or use in churches lacking some of these features. If so, he was right: the four canonical Gospels survived precisely because within a fairly short space of time they did prove relevant enough to most churches to come to be used very widely.

The argument that not everything in a Gospel need be there for all readers applies also to other types of material. When John finds it necessary to explain what the words Rabbi and Messiah mean (explanations which I cannot believe even diaspora Jews would need), this need only imply that some of his readers would need such explanations, not that all or even a majority would need them. When Mark tells us that Simon of Cyrene was the father of Alexander and Rufus, he need only be supposing that these persons would have been known in a significant number of churches, which is entirely possible, not that every church to which his Gospel might circulate would have heard of them. Knowing these names already would give added significance to Mark's narrative for those who did know them, but not knowing them would be no impediment to other readers' reading of Mark.

In conclusion, therefore the relative success of a reading strategy based on the assumption that a Gospel addresses a specific community is no proof at all that a reading strategy based on the contrary assumption would not be equally or even more successful.

III

The rest of my argument - really no more than a sketch of the argument which I think needs developing - aims to establish the antecedent probability that someone writing a Gospel in the late first century would have envisaged the kind of general Christian audience which the Gospels in fact very soon achieved through circulation around the churches.

The first stage of the argument consists in contrasting Gospels and Pauline epistles. This stage is important because what the consensus I'm attacking has in effect done is to attempt to treat Gospels hermeneutically as though they were Pauline epistles. In other words, scholars have sought to see the audience and therefore also the message of the Gospels as just as local and particularized as those of the major Pauline letters, which certainly are addressed to specific Christian communities and envisage the specific needs and problems of those communities. The fact that our reading of 1 Corinthians, for example, is therefore illuminated by our attempts to reconstruct the specifically Corinthian situation which Paul addressed, has led Gospel scholars to seek the same kind of illumination of Gospel texts by reconstructing the specific church context in which they originated. However, Gospels are not letters, and to appreciate the crucial difference we need to put together two issues.

The first is the question of genre. It is a special quality of the letter genre that it enables a writer to address specified addressees in all the particularity of their circumstances. Even if other people read 1 Corinthians (as they fairly soon did), the genre encourages them to read it as a letter addressed to the Corinthians. To some extent every attentive reader of 1 Corinthians has always felt obliged to imagine what the specifically Corinthian situation Paul addressed was. This is not the case with the Gospels. From the second century to the mid-twentieth century noone ever supposed that the specific situation of the Matthean community was relevant to reading the Gospel of Matthew.

Of course, the genre of the Gospels is debated. It seems to me that recent discussion has all but conclusively established the case that contemporaries would have recognized them as a special category of the Graeco-Roman bios (which we can translate biography provided we understand the term in the sense of ancient, not modern biography). Although the implied readership of the ancient biography is a topic which might repay investigation, it seems to me unlikely that anyone would expect a bios to address the very specific circumstances of a small community of people. A bios certainly aimed at relevance to its readers. Its subject could be depicted as a moral or religious inspiration to its readers. It could be highly propagandist literature, recommending a political, philosophical or religious point of view. But its relevance would be pitched in relatively broad terms for any competent reader.

But the full force of the difference of genre will come home to us only if we add a second consideration. We need to ask, about both an apostolic letter and a Gospel, the question: Why should anyone write it? - by which I mean, why should anyone put this down in writing? In the case of 1 Corinthians, for example, the answer is clear: Paul could not or preferred not to visit Corinth. Paul seems only to have written anything when distance required him to communicate in writing what he would otherwise have spoken orally to one of his churches. It was distance that required writing whereas orality sufficed for presence. So the more Gospels scholarship envisages the Gospels in terms approximating to a Pauline letter, addressing the specific situation of one community, the more odd it seems that the evangelist is supposed to be writing for the community in which he lives. An evangelist writing his Gospel is like Paul writing 1 Corinthians while permanently resident in Corinth. Paul did not do this, so why should Matthew or the other evangelists have done so? Anyone who wrote a Gospel must have had the opportunity of teaching his community orally. Indeed, most Gospel scholars assume that he frequently did so. He could retell and interpret the community's Gospel traditions so as to address his community's situation by means of them in this oral context. Why should he go to the considerable trouble of writing a Gospel - since even Mark's Gospel was certainly not tossed off in an afternoon - for a community to which he was regularly preaching? Indeed, why should he go to such trouble to freeze in writing his response to a specific local situation which was liable to change and to which he could respond much more flexibly and therefore appropriately in oral preaching?

The obvious function of writing was its capacity to communicate widely with readers unable to be present at its author's oral teaching. Oral teaching could be passed on, but much less effectively than a book. Books, like letters, were designed to cross distances orality could not so effectively cross. But whereas letters usually stopped at their first recipients (though not invariably), anyone in the first century who wrote a book such as a bios expected it to circulate to readers unknown to its author. That small circle to which the author might initially read it or those friends to whom he might initially give copies were the merely first step to wider circulation. Once there was a copy outside the author's possession, he would expect others to make copies for their own use, and his book to have embarked on a journey into the world beyond his control. This was true even of the religious literature of a minority culture such as the Jews, probably the most obvious model for the Christian author who wrote the first Gospel. Jewish religious literature in Greek, wherever it might have been written, circulated among the communities of the western diaspora, presumably by the normal channels of personal contacts and travelling which account for the circulation of most literature in the period. Why should Mark, if Mark was the first evangelist, have written merely for the few hundred people, at most, who composed the Christian community in his own city, when the very act of writing a book would naturally suggest the possibility of communicating with Greek-speaking Christians everywhere?

Now for the major stage of my argument for the likelihood that Gospels would be written for general circulation we must turn to a crucial feature of the general character of the early Christian movement. The early Christian movement, I want to argue, was not a scattering of isolated, self-sufficient communities with little or no communication between them, but quite the opposite: a network of communities with constant, close communication among themselves. Therefore the social character of early Christianity was such that the idea of writing a Gospel just for one's own community is unlikely to have occurred to anyone.

The consensus I'm challenging seems to depend on a view of an early Christian community as a self-contained, self-sufficient, introverted group, having little contact with other Christian communities and little sense of participation in a world-wide Christian movement. Identity, issues and concerns, it seems to be presupposed, are thoroughly local. Andrew Overman's recent book on Matthew, for example, contains no reference at all to a Christian world beyond Matthew's own community (which consists of a small group of churches). That Matthew even knew about other Christian communities, still less that his community had any kinds of relationships with them, is never suggested, despite the notably universal thrust of the Gospel itself, with its strong indications of a world-wide Christian mission. Overman discusses the Matthean community's theological and social self-understanding as though the Matthean community were the only Christian community in existence. Although leadership in the community is envisaged as itinerant, the possibility that itinerant teachers would travel between the Matthean community and other Christian communities is never mentioned. Even the role of Peter is discussed as though it related solely to the Matthean community. Such a picture of isolated and inward-looking parochialism is both generated by and then serves to reinforce the notion that a Gospel has only a particular community in view. But it is in serious conflict with all the real, concrete evidence we have about early Christianity. So in this section I want to indicate, by sampling only, the large amount of relevant information we have in the sources - information which deserves precedence over tenuous inferences drawn from the Gospel texts on the basis of an already assumed model.

First, we should note that mobility and communication in the first century Roman world were exceptionally high. Unprecedentedly good roads and unprecedentedly safe travel by both land and sea made the Mediterranean world of this time more closely interconnected by constant travel and communication than any such large area of the ancient world had ever been. People travelled on business as merchants and traders and bankers, on pilgrimage to religious festivals, in search of health and healing at the healing shrines and spas, to consult the oracles which flourished in this period, to attend the panhellenic games and the various lesser versions of these all over the empire, as soldiers in the legions, as government personnel of many kinds, and even as holidaymakers and sightseers. It was certainly not only the wealthy who travelled. Quite ordinary people travelled to healing shrines, religious festivals and games. Slaves and servants frequently accompanied their masters on journeys. Runaway slaves, freed slaves returning home, people in search of work, soldiers and sailors and brigands all travelled. Travel, after all, was usually by foot and so was cheap. So people quite typical of the members of the early Christian churches regularly travelled. Those who did not, if they lived in the cities, would constantly be meeting people passing through or arriving from elsewhere.

So the context in which the early Christian movement developed was not conducive to parochialism; quite the opposite. Frequent contact between the churches scattered across the empire was natural in such a society, but in addition to Christian participation in the ordinary mobility of this society, much communication was deliberately fostered between the churches, as we shall see in a moment.

For, secondly, the evidence of early Christian literature (not least the Gospels) is that the early Christian movement had a strong sense of itself as a world-wide movement. For Jewish Christians who made up most of the early Christian leadership, this must have come naturally, since the communities of the Jewish diaspora were used to understanding themselves in terms of their common membership of a people scattered across the world. But Gentile converts were inculturated as Christians into a new social identity which was certainly not purely local. Paul's letters, for example, are constantly relating the churches he addresses to other churches and to the Christian movement as a whole, even to the churches of Judaea and other non-Pauline churches. The language of fictive kinship encouraged converts to replace their natural ties of family loyalty with new Christian ties that encompassed brothers and sisters throughout the world. Such ties could be important. A small minority group experiencing alienation and opposition in its immediate social context could compensate for its precarious minority position locally by a sense of solidarity with fellow-believers elsewhere and a sense of being part of a world-wide movement destined to become the world-wide kingdom of God. 1 Peter, for example, encourages its readers by reminding them that 'your brothers and sisters in all the world are undergoing the same kinds of suffering' (5:9), while the book of Revelation enables potential martyrs to see themselves as belonging to an innumerable company drawn from every nation on earth. Why, I ask myself, has social-scientific study of the NT not given an account of the functions which belonging to a world-wide movement performed for early Christians, instead of constructing such artificially isolated communities as Overman's Matthean community?

Thirdly, we should note that most of the Christian leaders of whom we know in the NT period moved around. They were not all as constantly on the move as Paul and those missionary colleagues who travelled with him were, but most of them are to be found in several locations at different times in their careers. This is true of Peter, Barnabas, Mark, Silas/Sylvanus, Apollos, Philip the evangelist and his prophet daughters, Aquila and Priscilla, Andronicus and Junia, Agabus, the brothers of the Lord, and others. Even the unknown author to the Hebrews, writing from one location, expects to be visiting his addressees in another (13:23). The importance of this is that surely these are the kind of people we should take as models for the kind of person who might have written a Gospel. Why do we so readily assume that the author of a Gospel would be someone who had spent all his Christian life attached to the same Christian community, when the evidence we have about Christian leaders suggests that he might well be someone who had spent much time travelling around various churches or someone who had spent some time established as a teacher in more than one church successively? In that case, his own experience of the Christian movement could well be far from parochial. And since the writing of a Gospel could well have taken several years, why should we even assume that even the writing of a Gospel took place in the context only of one community?

Admittedly, the leaders I have just mentioned all belonged to the first Christian generation, and specific information about named Christian leaders from the later part of the first century is much more scarce. But there is no reason to suppose that Christian leaders became more static. Itinerant teachers travelling from one church to another were still common up to the end of the century: we find them in Revelation, the Johannine letters and the Didache. And as we move into the second century, while it is true that the leadership of travelling missionaries, teachers and prophets gradually gave way to the leadership of local bishops, it is worth noticing that these bishops maintained the habit of quite extensive travelling and visiting of other congregations, while prominent Christian teachers in the second century seem, almost as a rule, to have taught for a period in more than one major Christian centre. I mention this second-century evidence (I haven't time for the detail of it), because it helps to establish a pattern of mobility in early Christian leaders which is continuous from our earliest evidence in the time of Paul through to the late second century. This is a constant feature of the early Christian movement, which means we must reckon quite seriously with the chances that some, if not all of the evangelists were people whose own experience was far from limited to a single Christian community.

Fourthly, another feature of the early Christian movement which we can establish as a continuous practice from the time of Paul through to the mid-second century is the sending of letters from one church to another. We find, for example, he leadership of the Roman church writing a letter of pastoral concern to churches scattered over a wide area of Asia Minor (1 Peter) and another to the church of Corinth to deal with the problems and disputes in that church (1 Clement). From the early second century we have the letter of Polycarp of Smyrna to the church at Philippi, and the letters of Ignatius of Antioch to six different churches. These surviving letters are the tip of an iceberg.

Letters establish more than literary connexions between churches. Letters imply messengers. The messenger would either be a member of the sending church who was in any case travelling through or near the church addressed, or a member of the sending church who travelled specifically to carry the letter. Messengers stayed in the homes of members of the church, met with the whole church for worship, conveyed orally news not included in the letter, received news to take back home, and surely forged warm personal contacts with their hosts. Because of the role of messengers, a letter is merely the formal, surviving element in a two-way communication with wider oral and personal dimensions. Messengers were one way in which personal links between churches were created, which must have given the even most untravelled Christian a strong sense of participation in something much broader than his local church. But messengers carrying letters are only one example of the kind of informal contact which must have been constantly created by members of one church, travelling for all kinds of reasons, passing through and enjoying the hospitality and fellowship of other churches. (Of course, they also clashed and quarrelled, as we shall note in a moment.)

Fifthly, let me briefly allude to some of the kinds of concrete evidence we have for close contacts between churches in the period around or soon after the writing of the Gospels. (a) The famous fragment of Papias's prologue affords us one glimpse of what happened. Though writing in the early second century, Papias was recalling a time in the late first century (precisely the time when Matthew, Luke and John were being written). As a young man in Hierapolis, he was an avid collector of oral traditions. He collected them not by travelling himself but by quizzing anyone who happened to pass through Hierapolis who had heard the teaching of personal disciples of Jesus at either first or second-hand. Hierapolis is a little off the much travelled route that ran east from Ephesus through Laodicea, so Christians travelling that route must sometimes have turned aside specifically to visit the church at Hierapolis. (b) The letters of Ignatius, written two or three decades after Matthew, Luke and John, give us a remarkably detailed picture of an active communication network among the churches of the area from Syrian Antioch to Philippi, as well as between these churches and Rome. Letters, delegates and even bishops travel back and forth between these various churches for a variety of purposes, all in the time it took Ignatius and his guards to travel from Antioch to Italy. In other words, the same kind of frequent and vigorous communication, by travel and letter, as we see in the Pauline letters is still happening. We can also observe, even within the period covered by the letters of Ignatius and Polycarp, that letters of Ignatius are already being copied and circulated around churches other than those to which Ignatius addressed them. In other words, we can observe the ease and speed with which copies of Christian circulated around the churches. (c) When the Roman Christian prophet Hermas received his visions, he was told not only to read them to the Roman church, but also to make a copy for Clement, who 'will send it to the cities abroad, because this is his job.' Clement, as the Roman church's secretary responsible for communication with other churches, has the job of making multiple copies of literature written in Rome and sending out copies by messenger to other churches. He probably had this job in the late first century, just when Matthew, Luke and John were being written. Other churches might not have been quite so well organized, but just a few copies sent to other churches would be quite sufficient to launch a piece of Christian literature on a journey round the rest of the churches.

Sixthly, and finally, the evidence for conflict and diversity in early Christianity supports my picture of the early Christian movement as a network of communities in constant communication. I stress this point because, when I've presented this view of early Christianity on previous occasions, I've sometimes been misunderstood as portraying the Christian movement as entirely harmonious and homogeneous, playing down conflict and diversity. This is not at all my intention. The point is that this network of communication is the vehicle for conflict and disagreement, as well as for fellowship and support. All the evidence we have for rivalry between Christian teachers or conflict between different versions of the Christian message, from Paul's letters through to Revelation and the letters of Ignatius, shows us that conflict operating across the network of communication I have depicted. Teachers of one version of Christianity do not keep to a little patch of like-minded churches. On the contrary, itinerant teachers of any persuasion are always liable to turn up in any church. Congregations divide. Leaders from elsewhere write to support one faction or another. Much as some leaders strove to get teachers they strongly disapproved of excluded from churches where they had influence, clearly they constantly failed. None of this evidence for conflict and disagreement suggests that any version of Christianity formed a homogeneous little enclave of churches, out of communication with other churches and renouncing any interest or involvement with the wider Christian movement. Quite the opposite: all such evidence confirms my picture. Churches take an intense interest in conflicts happening elsewhere. Leaders and teachers actively promote their versions of the Gospel anywhere and everywhere in the Christian world. These are not the introverted communities and teachers which would produce written Gospels purely for home consumption.

I repeat: in view of all this evidence that the early Christian movement was a network of communities in constant communication with each other, by messengers, letters, and movements of leaders and teachers - moreover a network around which Christian literature circulated easily, quickly and widely - surely the idea of writing a Gospel purely for the members of one's own church or even for a few neighbouring churches is unlikely to have occurred to anyone. The burden of proof must lie with those who claim it did.

I conclude with a number of hermeneutical observations:

First, the attempt by the consensus to give the so-called Matthean, Markan, Lukan and Johannine communities a key hermeneutical role in the interpretation of the Gospels is completely mistaken. If the Gospels do not address those communities in particular, those communities have no hermeneutical relevance. I doubt very much whether we can in fact know anything worth knowing about them. If the Gospels were not written for specific communities, then the situation is quite different from that which enables us to know quite a lot about Paul's Corinthian church. Certainly it may be argued that the community in which a Gospel was written is likely to have influenced the writing of the Gospel even though it is not addressed by the Gospel. But it does not follow that we have any chance of reconstructing that community. We cannot, as I've suggested, even take it for granted that a Gospel was written in only one community, and certainly not that its author was influenced only by one community context. Anyone who finds my argument convincing should forthwith stop using the terms Matthean community, Markan community and so forth. They no longer have a useful meaning.

Secondly, my argument does not de-contextualize the Gospels. I am not treating the Gospels as autonomous literary works floating free of any historical context. (I say this because again I have been misunderstood on this point.) The context of the Gospels is the early Christian movement in the Roman empire of the late first century, and we can bring to their interpretation everything we know about that movement and its political, social, economic, religious, ideological contexts. This context is a good deal less specific than the consensus desires, but it is no more general than the context which most literature of that period addresses, or the context which most literature of any society in any period addresses. Literature addressing one tiny community in a specific locality is very rare, but to claim that most authors address wider contexts than that does not de-contextualize their work.

Thirdly, however, it is true that my argument smooths the hermeneutical path from the way the Gospels addressed their first readers to the way they have been read ever since (though this was not in my mind when I developed the argument). As I said earlier, no attentive reader can miss the hermeneutical relevance of the church at Corinth to the interpretation of 1 Corinthians. But everyone before the mid-twentieth century missed the hermeneutical relevance of the Matthean community to the interpretation of Matthew, and who can blame them?

Fourthly, the mistake of the consensus view derives from a misplaced desire for historical specificity. It has behind it that tremendous drive towards historical specificity which has fuelled a considerable part of the whole enterprise of modern biblical scholarship. The desire is to define the historical meaning of the text as specifically as possible by defining its historical context as closely as possible. Just as we know we've understood 1 Cor 8-10 better when we've studied pagan sacrificial meals in Corinth, so we think we shall know more precisely what Luke's teaching on wealth and poverty means if only we can define just where the dozen rich people in Luke's community belonged in the social hierarchy and exactly how they were actually treating the poor.

This is a hermeneutical mistake, but the mistake does not consist in thinking historical context relevant. It lies in failing to see that texts vary in the extent to which they are context-specific. Some texts (Umberto Eco calls them 'closed texts') define their implied reader very closely, and also have a determinate meaning which depends on knowing what the implied reader is supposed to know. If one does not know this, one can misunderstand badly. If we knew nothing at all about idol-meat in Corinth, we might well mistake Paul's meaning quite seriously. But other texts (Eco calls them 'open texts') leave their implied readership more open and consequently leave their meaning more open to their real readers' participation in producing meaning. The Gospels are relatively open texts, though not as open as some (a lyric poem, for example). For various late-first-century churches hearing Matthew's Gospel in differing situations Jesus' command to love their enemies would have meant rather different things. I do not think Matthew would have minded at all. To think we do not know what Matthew meant unless we can pin down what sort of enemies his community had is trying to read an open text as a closed one.