Volume 6 - Issue 2

The Q Debate since 1955

By Howard C. Bigg

‘Q is and remains a hypothesis.’ So wrote Austin Farrer in 1955 in an article,1 which may fairly be regarded as initiating the developments in the Q debate over the past two decades or so. Farrer’s view was that Q is an unnecessary hypothesis and should be dropped. Recently, articles and studies touching on all aspects of the Q question have appeared regularly. Further, the scope of the question has broadened into the area of redaction criticism, in an effort to show Q as having particular theological motifs and interests. In this respect, it is important to recognize that studies in the theology of Q are part of the wider search for the theology of the evangelists, away from ‘solutions’ of the synoptic problem in purely literary terms. Nevertheless, Q regarded as an hypothesis necessary for a proper solution of the literary relationship between the synoptic gospels is still very much part of the contemporary debate. This article will take a look at Q from three angles. First, we must see how Q is faring as one of the main planks in the classical two (or four) documents solution as set out by Streeter in 1924.2 This will lead to a look at the revival of hypotheses which wish to dispense with Q and seek a solution in terms of direct borrowing between Matthew and Luke, or more complex solutions involving multiple sources. Finally, we will give some account of recent attempts to establish Q as a theological document.

For an exposition of the traditional solution of the synoptic problem, one of the best available is that of W. G. Kümmel in his Introduction to the New Testament3 pp. 52ff. It is vital to understand that Q does not stand on its own, but alongside its twin hypothesis of the priority of Mark. Thus Kümmel sets out the reasons for accepting Markan priority before going on to the Q question (pp. 63ff.). Indeed he notes that C. H. Weisse in 1838 linked the idea of a common sayings source for Matthew and Luke to the priority of Mark, thus carrying on the line of research opened up by Schleiermacher and Lachmann. H. J. Holtzmann in 1863 took up Weisse’s proposals in their modified Ur-Markus form to provide an important milestone in the development of the two-document hypothesis, eventually abandoning the Ur-Markus theory in favour of Matthew’s and Luke’s knowledge of canonical Mark.

Kümmel, with admirable clarity, rehearses the reasons why Q must be a written source copied independently by Matthew and Luke and not oral tradition.

  1. Verbal agreements. These are so close as to compel reliance upon a common written source in some instances (e.g. Mt. 3:7–10; 7:7–11; 11:4–6; 12:43–45; 24:45–51 par). Elsewhere, such agreement is ‘rather slight’ (e.g. Mt. 10:26–33; 25:14–30 par) and Kümmel offers no explanation for the wide variations in verbal correspondence. He contents himself by remarking that ‘the common vocabulary in all the sections … is over fifty percent, which can hardly be accounted for by simple oral tradition’.4
  2. The argument from order. This is an important line of investigation which for reasons of space it is not possible to develop and the diagram is not very helpful. The force of the case may however be gauged by referring to V. Taylor’s presentation of it in his New Testament Essays,5 his arrangement of the Matthean material in five columns corresponding to the discourses. The result does indicate some clear sequences and leads Taylor to claim that ‘the manifest signs of a common order in Matthew and Luke raise the hypothesis to a remarkable degree of cogency short only of demonstration’. This is extravagant and Taylor’s case is slightly weakened by the omission of several passages where verbal resemblance is slight. Taylor justifies these omissions by explaining that ‘if in these passages, Q and another source overlapped, it is reasonable to expect that the order of Q as reflected in Matthew and Luke may be obscured’.6
  3. The presence of doublets and double traditions (double traditions are verses presented by both evangelists, but in different forms; doublets are verses which one evangelist presents twice). In addition to passages like the mission charge which Luke reports twice, Luke 9 and 10, in parallel with Mark 6:7–12 and Matthew 10 respectively, there are a string of sayings of Jesus which appear twice in Matthew and Luke, once in a setting parallel to Mark and again in a sayings setting found only in Matthew and Luke, e.g. ‘He who has, to him will be given’ (Mt. 13_12=Mk. 4_25=Lk. 8:18, cf. Mt. 25_29=Lk. 19:26).

Compared to this evidence of doublets and double traditions Mark presents a single doublet (Mk. 9:35; 10:43f.). From this, Kümmel deduces that Matthew and Luke must have used a second source in addition to Mark, the linguistic agreements demonstrating that this source was in Greek.

We find a similar defence of the traditional Q in J. A. Fitzmyer’s contribution in Jesus and Man’s Hope,7where he rejects A. M. Farrer’s theory of Luke’s dependence on Matthew. He makes the following points:

  1. No reason is given for Luke failing to utilise Matthew’s additions (e.g. Lk. 5_3=Mt. 4:18; Lk. 5_27=Mt. 9:9).
  2. Why does Luke break up Matthew’s sermons, incorporating some of it in his own sermon on the plain, and scattering the rest in different contexts in his travel narrative? However we account for Luke’s redactional activity, the explanation that he has rearranged material from Matthew is the least convincing.
  3. Apart from 3:7–9, 17 and 4:2–13 Luke never puts the material from the ‘double tradition’ (Q) with the same Markan context as in Matthew. If such material derives from Matthew, it is surprising that no other Q material occurs in the same Matthean context, which are frequently very appropriate. Despite the widely differing contexts, Fitzmyer notes the underlying order of Q as shewn by Taylor.

The difficulties in delineating the extent of Q are recognized by both writers, but such lack of agreement does not justify the abandoning of Q as a written source. But of course the problem does not end there. We have already noted that Taylor excludes certain passages from his reconstruction of Q because of the possibility that Q and another source overlapped. The extent of such overlap cannot be arbitrarily limited to those passages which weaken a particular hypothesis as Taylor does here. We will have occasion to refer to this again later. Again, scant regard has been paid to the wider variations in verbal resemblance. Taylor appeals to conflation of M (special Matthean material) and Q. At different times, translation variants from Aramaic, different Q recensions, as well as Q being an oral deposit, have been proposed to account for these variations. The problems are as difficult as ever and Kümmel is over-confident in regarding certain points as ‘decisive’ for the existence of a definitive Q.

Attempts are made from time to time to establish the existence of Q by refuting hypotheses which seek to shew direct contact between Matthew and Luke. This was Farrer’s purpose in the article cited at the outset. He wrote that the hypothesis of Luke using Matthew must be explored before the hypothesis of a common source be considered.

In 1965, the challenge was taken up by F. G. Downing in an article8 which examined some key passages to test Farrer’s proposition that Luke used Mark as a framework with which to fit material he had quarried from Matthew’s additions to Mark. The passages selected are those where Matthew has apparently conflated Markan material with similar, but distinct material of his own. He considers in detail the notoriously difficult Beelzebub controvery (Mt. 12:22–45; Mk. 3:20–29; Lk. 11:14–26; 12:10; 6:43–45) with shorter studies of the Baptism narrative (Mt. 3:1–4:11 par); the sending out of the twelve (Mt. 9:35–10:16 par); and The Synoptic Apocalypse (Mt. 24:4–26 par). In the case of the Beelzebub controversy, Downing notes that Luke does not use Mark where Matthew has taken him over more or less intact, but only follows Matthew where he adds new material to Mark or largely alters Mark. From this it is concluded that Luke did not know Matthew’s use of Mark for no convincing reason can be suggested for Luke’s rejection of material taken over intact from Mark. A similar pattern is found in the other passages.

Downing’s treatment of Farrer must be pronounced successful in refuting the idea that Luke used Matthew in the manner suggested, but he does not thereby prove that Matthew’s extra material without the Markan addition is Q or Q + M. He merely shews that on Farrer’s argument, Luke has for some inexplicable reason decided to ignore pure Mark, which is most unlikely from what we know of Luke’s high regard for Mark as a source.

We have already taken note of the fact that the priority of Mark and Matthew’s and Luke’s use of Q are twin hypotheses. With regard to the former of these it has long been held as axiomatic that the order of the whole of Mark except what is peculiar to him is confirmed by either Matthew or Luke and the greater part of it by both. The corollary to this is with regard to Q that a parallel passage in the triple tradition is never immediately followed in both Matthew and Luke by a separate incident or discourse common to these two alone. Streeter9 used the phenomena of order to shew not only that Mark was the best representative of the original gospel, but also that Matthew and Luke used Mark and Q independently.

The data relating to order have had a stronger influence on the two-document hypothesis and therefore on Q than any other. It is this plank in the traditional approach which has been under strong attack recently. In an article entitled ‘The argument from order and the relationship between Matthew and Luke’,10 E. P. Sanders, following up his earlier book on The Tendencies of the Synoptic Tradition,11 subjects this axiom to minute examination. He points out that any agreements, even minor ones between Matthew and Luke on a point of order which cannot be attributed to their independent use of Mark and Q will raise the possibility of some contact between Matthew and Luke. Verbal agreements between Matthew and Luke against Mark are treated separately.

The crucial question is what constitutes an agreement in order? Earlier statements of the argument from order dealt only with full pericopes as found in Tischendorf’s synopsis. These corresponded with the essential biographical outline of Jesus’s life. But once the question becomes the literary one of possible contact between Matthew and Luke it is clear that these arbitrary divisions are no longer warranted. Obviously, once the larger pericopes are broken up, small points where Matthew and Luke agree are as difficult to explain as large ones. Here is one example: Matthew 11:10 par Luke 7:27 place the idou apostellōquotation concerning John in the context of John’s question and Jesus’ testimony to him. Mark 1:2 places it in the context of John’s preaching.12

Sanders also rejects the suggestion that in some instances where Matthew and Luke place Q material in the same place in the Markan outline, this can be explained if the preceding Markan passage was also in Q. Agreements between Matthew and Luke under these circumstances would be caused by their following the order of Q. (e.g. Mt. 3:7–10 par Lk. 3:7–9). Sanders lists the following passages in Mark which are inexplicably unsupported by either Matthew or Luke 1:4; 11:11; 11:15–19; 13:33–37; 3:13–19; 6:1–6a; 4:23; 12:34c; 11:25; 1:4–6; 9:41; 6:34b.13 Sanders’ conclusion is worth quoting: ‘The assurance with which it is usually said that Matthew and Luke were independent of each other rests on the assertion that they neveragree together in such a way that it cannot be explained by reference to their independent use of Mark and Q. When we note the number of instances where they do, the assurance we have felt in the traditional hypothesis must be correspondingly weakened.’14

With this brief discussion of Sanders’ work, we have reached the point where we must examine the question of whether the revival of alternative hypotheses offers better possibilities for the solution of the synoptic problem.

It is worth noting here that there has always been an undercurrent of opposition to the two-document hypothesis by defenders of rival hypotheses which place Matthew first.15 Early in this century E. W. Lummis and H. G. Jameson16 sounded an alarm that the reigning two-document hypothesis ignored important data. Jameson wanted to return to the Augustinian solution, which has been given a further airing in recent years by B. C. Butler.17 The hypothesis propounded by Augustine held that the order of the gospels was Matthew, Mark, Luke, John, that no gospel was written in ignorance of the others and that Mark was the epitomiser of Matthew. Recently, dissatisfaction with Q, fed by a growing quantity of data which seriously undermines it, has led to a considerable revival of interest in alternative hypotheses, chiefly the Griesbach hypothesis, which sees Matthew as a source for Luke with Mark using both.

The name which is probably most familiar to students of the synoptic problem in current debate is William R. Farmer whose book The Synoptic Problem—a critical analysis18 sought to reopen the whole question of synoptic relationships with the aim of establishing the Griesbach hypothesis as the only one which satisfied the requirements of the data. But, anybody who reads Farmer hoping for an exhaustive discussion of Griesbach will discover that the first 200 pages are taken up with a critical history of the treatment of the synoptic problem from the 18th century to the present. Farmer then leads us to ‘a New Introduction to the Problem’ covering some 84 pages along Griesbachian lines including a necessarily abbreviated chapter entitled ‘Notes for a History of the Redaction of Synoptic Tradition in Mark’ (50 pages). The reader must judge for himself, but in our judgment the value of Farmer’s book lies in the fascinating historical section. Farmer skillfully, but without rancour, exposes the foibles and lack of objectivity at crucial turning points in the debate. In this writer’s view, the best two chapters are 3: ‘The English Endorsement and Modification of the Two-Document Hypothesis’ and 4: ‘An Analysis of Streeter’s contribution to the Two-Document Hypothesis’.

One of the main reasons why the Griesbach hypothesis has aroused such interest in recent years is the failure of the two-document solution to deal with the problem of the so-called ‘Minor Agreements’. These are the many instances where Matthew and Luke agree against Mark. C. H. Turner in 1924 declared that ‘so long as it is supposed that there is a residuum of agreements between Matthew and Luke against Mark in matter taken from Mark … so long will research into the synoptic question be hampered and a final solution be delayed’.19

The study of the minor agreements has become something of a ‘growth industry’. Recently a full-scale study of this question has appeared edited by F. Neirynck.20 It is a book which provides complete data relating to the minor agreements with an introductory essay on the way the problem has been treated in the past.

One of the interesting facts which emerges from the survey of the minor agreements is that some scholars are able to combine an acceptance of the priority of Mark with the abandonment of Q. This was Farrer’s position,21 but he was vulnerable to attack because he was proposing a general hypothesis which was exposed at many points. But he was followed by others who looked at particular instances of agreement between Matthew and Luke. N. Turner22 in a contribution to the Oxford Congress of 1957 replying to an article by E. L. Bradby23 discusses some Markan passages produced by Bradby (Mk. 2:26; 4:4, 9, 19, 20; 6:7; 8:31, 34, 35, 36). Turner is relying upon arguments from style when he writes: ‘There cannot be any other reason than literary dependence to explain these apparently irrelevant agreements of style and grammar.… Because these agreements are so often inconsistent with St Luke’s style elsewhere, it is more likely that Luke depends on Matthew than vice-versa’ (p. 234).

R. T. Simpson in his article ‘The Major Agreements of Matthew and Luke against Mark’24 discusses three passages: Mark 1:1–13; 12:28–31; 3:22–27. As a general proposition Simpson writes: ‘The more strongly we plead the case for Q as a means of explaining all those resemblances between Matthew and Luke which are not attributable to their common use of Mark, the more we undermine the theory of the priority of Mark and the more we stress the importance of the Matthean and Lucan “improvements” of Mark, the more the significance of those minor agreements which are such a difficulty for the defender of Q will be enhanced’.25

The crux of Simpson’s argument is that in the passages referred to, Matthew and Luke do not simply add to Mark, but also improve it, often in precisely the same way. The choice of words carried with it a value judgment, but the essential point still stands when it is noted that there is a close connection between the Q material and that taken from Mark. In fact the Q material cannot stand on its own.26 Simpson notes the volte face by Streeter on this point between the publication of the Oxford Studies (1911) and The Four Gospels(1924). Simpson’s other axiom is the priority and independence of Mark and his conclusion that ‘once this is conceded, then the case for believing that the major agreements could have been produced only as a result of St Luke’s use of an edited version of Mark is entirely convincing’.27

Simpson’s analysis of the three passages cited and the conclusions drawn are not in fact equally convincing. The best evidence for Luke’s knowledge of Matthew is found in the pericopes dealing with the appearance and preaching of the Baptist. As an example, Luke 3:16–17 contains two improvements of Mark which are substantially paralleled in Matthew. Mark’s rather clumsy phrase egō ebaptisma hymas hydati is replaced by the more stylish egō men (en) hydati baptizō hymas and the saying is moved to a more emphatic position. Again the Q material in Luke 3_17=Matthew 3:12 is an expansion of the Markan narrative. Simpson believes that the independent alteration of Mark in this way must be more than coincidental. More striking however is the fact that the material appended to Mark 1:8 begins with the pronoun which refers back to the subject autos taken from Mark.

Simpson’s final conclusion is that ‘either … Mark knew and modified Q (or even Matthew) or … Luke knew Matthew as well as Mark. In every case we have considered, it is Mark who appears to give the more primitive version and it is obvious that the simplest solution to the problem is that Luke made use of Matthew.’28 He is careful to say that on the evidence adduced, the probability is greatly strengthened that Matthew was one of Luke’s sources rather than offering an overall solution. However, Simpson did not consider the older view of overlapping of sources, and further, by insisting that in each case Mark offered the original version, he ruled out the Griesbach solution. It is to this that we must briefly now turn.

The Griesbach solution to synoptic relationships can on the face of it eliminate just those difficulties which are so troublesome to defenders of the two-document solution, especially of course the minor agreements. Hardly less important for advocates of Griesbach, however is the claim that there is no need to have recourse to any theory of overlapping sources. This they regard as a desperate expedient to explain passages in Mark which appear to contain secondary features. Thus D. L. Dungan writes: ‘The existence of Q has always been essential to the argument for Mark’s priority—precisely as the loophole to invoke any time one finds a pericope that is more primitive in Matthew and/or Luke when they are supposedly using Mark: the blessed overlap’.29 This is not quite accurate. Overlap is more usually postulated to explain different features or versions in Mark, not necessarily secondary ones. That is a separate and further problem. Further, advocates of Griesbach sometimes appeal to an overlap of sources in passages where Luke’s version appears more primitive than Matthew’s. Farmer, for instance, says that Luke used parallel traditions for his versions of the parables of the lost sheep, the talents/pounds, and the wedding feast, in addition to parts of the apocalyptic discourse.30 Thus defenders of Griesbach cannot claim to have escaped from this problem. The difficulty of establishing proper criteria for determining when a Mark-Q overlap exists is one that cannot be discussed here, but enough has been said to show that the existence of parallel and overlapping sources is not peculiar to Marcan priorists and defenders of Q.31

Even if it is granted, however, that the possibility of overlapping sources does not present so much of a problem to advocates of the Griesbach hypothesis and further, when the difficulty of the minor agreements are dealt with by placing Mark third, can we see the revival of Griesbach as offering a way forward? The only way to arrive at a satisfactory answer to the question is to examine in outline a typical passage where Farmer claims that Mark is the result of direct redactional activity on the text of Matthew and Luke.

The parable of the mustard seed is an excellent example since it has always presented problems, whichever view one takes as to the relationship between the three versions. Farmer’s basic axiom in describing Mark’s method and purpose is put as follows: Mark was ‘guided by the literary purpose not to deviate from the text to which his predecessors bore concurrent testimony’.32

Firstly, on Farmer’s principles, it is not at all clear why Mark has included this parable at all. According to Farmer, Mark has just stopped using Luke as his source (4:21–25=Lk. 8:16–18) and gone over to Matthew. He then omits Matthew’s parable of the tares, but inserts his own parable of the seed growing secretly (4:26–29). It could be that Mark decided to follow Matthew at this point because there is a Lucan parallel, though in a different Lucan context (Lk. 13:18f). Why then does Mark not include the parable of the leaven which follows in both Matthew and Luke? The verbal agreements are so great as to compel some literary relationship. We are left in some doubt as to what Mark’s redactional procedure really was!33

However, granted that Mark has chosen to include this parable, his choice of wording presents problems for the Griesbach hypothesis. A few examples must suffice.

  1. Having reverted to Matthew as a source, he immediately changes the Matthean introduction and substitutes a simple kai elegen, although this is a feature of Markan style and may not be significant.
  2. Mark now follows with a double question taken from Luke, but not exactly. Luke’s homoioō is moved to the first half of the double question, preceded by pōs rather than the more usual construction using tini. Even if it could be granted that the changes are due to Markan style, lack of close verbal similarity makes the theory of direct literary dependence difficult.34
  3. The next passage makes very difficult reading in Mark. Both Matthew and Luke have the same construction homoia estin followed by the dative, but precisely where on Farmer’s thesis Mark should be following his common source, he ignores it and substitutes his own hōs plus dative. The whole of Mark’s sentence is odd here, since he has left it without a proper main verb and he makes other changes which make the Griesbach solution look decidedly shaky. In fact, it looks as if, far from Mark following his sources where they bear concurrent testimony, he has painstakingly avoided the coincidences of Matthew and Luke. Since the result is grammatically baffling, we have to place a large question mark against the hypothesis.35

It seems, then, that insuperable difficulties lie in the way of accepting the Griesbach solution, but if we take seriously the evidence often presented that in this and other passages36 there are two separate versions, independent of each other, we are back to discussion of overlapping sources and in the present writer’s opinion, the existence of an independent source drawn upon by Matthew and Luke which overlapped with Mark is still the most probable explanation of the phenomena.37

One thing at least is clear from our discussion up to this point and that is that the Q question is as controversial as ever. Moreover, it is probably true to say that the central question is whether we may speak meaningfully of Q as a documentary source at all or whether we must rest content with the vaguer idea of Q as a strand of tradition, the origin of which can never be established—it is merely a label for common material. Some assessment of the status question is necessary before moving on.

Beneath the swirling waters of controversy, it appears to us that certain factors persist rock-like which in our view tip the balance in favour of Q as a written source or sources used by Matthew and Luke independently in addition to Mark.

  1. The existence of common material requires some explanation. F. G. Downing has given convincing reasons why Farrer’s solution is unsatisfactory and this in our view puts the onus on those who wish to dispense with Q to come forward with good grounds for doing so.
  2. Sir John Hawkins at the beginning of this century in his contribution to Oxford Studies in the Synoptic Problem divided up the Q material into three classes. Class A—passages very probably derived from Q are almost twice as long as classes B and C put together (i.e. those passages where derivation from Q is less likely). It seems to us that the weight of probability is still on the side of an independent use of Q by Matthew and Luke when the data given by Hawkins is taken in conjunction with Luke’s odd editorial policy if we wish to make Matthew Luke’s source. This latter point is still a most serious obstacle for supporters of Griesbach.
  3. The oral hypothesis is no longer taken seriously and Hawkins in the article mentioned suggests powerful reasons why this is so. Thus, we are concerned with some kind of literary dependence and whilst the revival of the Griesbach hypothesis is welcome in that it draws attention to weaknesses in the traditional Q hypothesis, it raises too many fresh problems of its own to provide a decisively convincing alternative.

The debate is refreshingly vigorous and shews no signs of abating as yet. In particular, the Baptist material poses real difficulties for traditional Q supporters.

We now turn to the third main area of importance in Q research, which concerns the status of Q as a theological work. Such an undertaking is beset with difficulties, not the least of which is scholarly agreement on the extent of Q. The danger of circularity is never absent. Nonetheless, the work done in this area does suggest that in broad terms the material labelled Q possesses certain distinctive characteristics and motifs. T. W. Manson in his important book The Sayings of Jesus38 believed that Luke preserved to a fair degree the order of Q and further that it is possible to uncover an outline of Pre-Q (and Pre-M) tradition consisting of the following elements. (1) Jesus’ Preaching; (2) Mission charge; (3) Against the Pharisees; (4) Eschatological Speech.

The consideration of Q as a distinctive theological work is of course part of the larger enterprise called ‘Redaction Criticism’39 itself an aspect or rather a stage in the traditio-historical study of the gospels. For a useful introduction to and survey of redaction criticism, readers are referred to J. Rohde’s Rediscovering the Teaching of Jesus.40 For the purposes of this article, it is assumed that the reader is familiar with the developments in Form and Redaction Criticism and we will proceed straight to the application of Redaction Criticism to Q.

Reference has already been made to the work of T. W. Manson who pointed out that Q carries certain emphases. His explanation as to why Q lacked a passion account was simple—the community which used this collection already knew it. This is a neat solution and Manson makes a considerable point of the small amount of the polemical material in Q (ten per cent). He explains this by saying that it is better to be positive than controversial.

Can this classic kerygma-plus-didache solution be accepted? Unfortunately it cannot if only because of the presence in the Q material of such passages as the Beelzebul controversy, the John sayings and the prophecy about Jerusalem. It was H. E. Tödt in 1959, who first focused attention on Q as a ‘controversy’ document, as a by-product of his study of the Son of Man sayings.41 Tödt picked up Bultmann’s observation, that Jesus’ teaching was not simply gathered, but also proclaimed.42 Q ceases to be thought of simply as a moral guide, supplementing the kerygma, but rather Q was produced by a community not centred on the kerygma at all. This community saw its task as continuing to proclaim the message which Jesus proclaimed. The words of Jesus now have significance and not just his death.

What were the motives that resulted in the Q collection? Tödt starts with the mission discourse in Matthew and Luke, which is based on Q, and compares it to Mark. The influence of Q is seen primarily in Matthew 10_7=Luke 10:9 where the disciples are told to pass on the message preached by Jesus—the imminence of the Kingdom of God. Mark 1:1 states that the message is about Jesus the Christ.43

It is clear, according to Tödt, that the Q community believed the announcement of the Kingdom of God to be as relevant now (in the post-Easter situation) as it was in Jesus’ lifetime. Hence the character of Q is determined by the imminence of God’s reign and not by the passion kerygma. Tödt accepts the future Son of Man sayings as authentic and his point is that the expectation of the Q community is found in its use of the Son of Man title and the related statements of the nature of the disciples’ mission. Another point worth noting is Tödt’s treatment of the resurrection. Although of crucial significance for the Q community, the resurrection was not a subject of its early preaching. The resurrection affirms Jesus’ authority, and that of his teaching. Tödt calls this ‘Christological Cognition’.44 The continuity between Jesus and the risen Christ lies in equating of Jesus and the Son of Man.

Tödt’s book is a long one and we have done little more than point to its significance in relation to a redaction-critical approach to Q, although in substance his work probably has more christological significance than methodological. He does not for instance suggest anything more original than a Palestinian provenance for Q, and the idea that the Q community gathered the teaching of Jesus with a view to its continued contemporary proclamation is taken from Bultmann.45

It is worth mentioning here that W. D. Davies also presented a critique of the view that Q was preserved primarily for hortatory reasons in his book The Setting of the Sermon on the Mount,46 quite independently of Tödt. Among the reasons given for denying the hortatory character of Q is the fact that, as recent studies have shown, catechetical material in the New Testament offers few, if any, parallels to Q. For example, Davies does not see the section on John the Baptist as an exhortation to avoid the sect of his followers (Mt. 3:11 = Lk. 3:16) but as a proclamation of the impending crisis which John had announced: ‘The original Q form … had no reference to the Spirit. It ran “He will baptize you with fire”.… The sense of the saying is not that John’s baptism is the preliminary to something better, but that it is the last chance of escaping something very much worse, namely, the coming of judgment.’47 Davies observes this note of crisis through much of the Q material and concludes that the ethical teaching in Q ‘expresses the total final demand that God lays upon men in Christ’.48 It therefore follows ‘that the teaching of Jesus in Q cannot with justice be characterized as catechetical, if by catechetical is meant elementary instruction given to candidates for admission into the Church at baptism’.49

These works of Tödt and Davies mark the beginnings of what was soon to become a stream of studies in which redaction-critical methods were to be brought to bear on Q.50 Amongst the important names in the late 60’s and early 70’s are those of D. Lührmann51 and P. Hoffmann.52

Lührmann’s point of departure is the distinction made by Bultmann between collecting and editing. Is it possible to discern within the Q material a purpose which goes beyond merely collecting material according to certain laws of transmission? Lührmann believes that such a redactional purpose is discernible. The two-source hypothesis is assumed as is also the form-critical axiom that the synoptic tradition originally consisted of independent units circulating according to local need and circumstance. He also accepts that there were community formulations (inventions!?) at all levels of the tradition.53 As an indication of the provenance of Q, he believes that Q presupposes the Gentile mission and that the final editing cannot be earlier than ad 50 or 60 in the Hellenistic community. As regards the form of Q, Lührmann agrees with Robinson’s logoi sophōn, but not that the form of Q as such is gnosticising.54

What then of the theology of Q? Here, Lührmann believes that for the Q community the most important redactional motif is the contrast between Jesus and the disciples on the one hand and Israel on the other. The separation is absolute; for Israel there remains only judgment. Christologically, the leitmotif is that the community continues the proclamation of Jesus, in particular the judgment. ‘Jesus is not the one proclaimed, but on the contrary, the content of the proclamation is the coming judgment in which Jesus as Son of Man will save his community.’55 Hyios and Kyrios occur as Christological titles in addition to Son of Man. There are few explicit Old Testament citations but instead we find frequent allusions to Old Testament narratives, particularly those which accompany minatory words (Drohworten), underlining Q’s purpose of admonition to watchfulness as judgment approaches.

The core of the study is a detailed analysis of Q passages classified under three major headings: (1) ‘Jesus and “this generation” ’ subdivided into: (a) the question of John the Baptist (Lk. 7:18–35 and 3:7–9, 17); the Beelzebul controversy and request for a sign (Lk. 11:14–23; 24–26; 29–32); discourse against the Pharisees (Lk. 11:39–52). (2) ‘The Community’ (Lk. 12:2–9; 6:20–49; 7:1–10; 9:57–60; 10:2–12, 13–15, 21–24). (3) ‘Eschatology’ (Lk. 12:39–40, 42–46; 19:12–27; 17:24, 26–30, 34–35, 37 [see pp. 24–83]).

Luke 7:18–35 is given extended treatment by Lührmann, which is worth sketching as an illustration of his method and conclusions. Basically, Lührmann concludes that the redactor of Q has joined together three pericopae all dealing with Jesus and John the Baptist with the hand of the redactor seen most clearly in verses 31–35. This is where he introduced the reference to ‘this generation’ and provides an interpretation for the parable of the Children in the Market Place, in which the point is no longer Jesus versus John, but rather Jesus and John versus ‘this generation’. Lührmann finds the same concern elsewhere in Q. Lührmann with others believes that, the parable of verses 31f. probably referred to people’s reactions to Jesus himself—they liked neither his call to repentance nor his invitation to joy. Hence the application in verses 33f. to John and Jesus is secondary. It is argued that there is some discrepancy between the parable and its interpretation, that elsewhere parables are given without interpretation and that the reference to the Son of Man and Wisdom reflect the concerns of the early Church (p. 29f.). However, the parable as parable (rather than viewing it as an allegory) fits quite well with the explanation and it cannot be ruled out that the parable had an application from the start.56

Space does not permit further comment on Lührmann’s work, but since redactional studies of a postulated source are such risky and uncertain undertakings, it is of some comfort to know that the contemporary study of P. Hoffmann referred to above supports some of Lührmann’s main conclusions, particularly with regard to eschatology and the pronouncement of judgment. The motifs identified by Lührmann are undoubtedly present, but the question remains open as to whether such motifs are primary and what is there relation to other possible motifs in Q.

Before leaving the German scene, no survey would be complete without an acknowledgment of the great commentary of S. Schulz.57 Here, we can do no more than note that Schulz takes account of recent studies in German by Tödt, Hoffmann and Lührmann. His interest is the light thrown on the development of early Christianity by a study of Q. He sees the sayings source used by Matthew and Luke as itself the product of two stages of development. The first is marked by post-Easter apocalyptic enthusiasm and by charismatic eschatological intensification of the Torah. Jesus was the exalted Son of Man and expected judge. To this earlier stage are assigned the Beatitudes, the Lord’s Prayer, the injunction to love of enemies and the prohibition of divorce and judgment of others. The early Q community is regarded as a type of ‘Hebrew’ Christianity and draws attention to passages upholding the Torah and disparaging Gentiles and tax collectors. No future study of Q can safely ignore Schulz, and it seems churlish to criticize a work which is by any standards monumental. But, from a conservative standpoint, its sheer massiveness makes one ask whether in the end the author has paid sufficient attention to the question of the creative role of Jesus himself. We are back with the crucial question of criteria. This question becomes very pressing when dealing with the authenticity of the sayings of Jesus and is therefore relevant to the study of Q. Whilst Schulz does not employ the absurdly stringent criteria of N. Perrin and others (dissimilarity, coherence and multiple attestation), there is always the danger that the sheer impact of Jesus himself will not be given sufficient weight in his desire to locate Q in its appropriate post-Easter milieu.

The French have not been idle in the field of Q studies and we must now look briefly at some recent contributions in French. To maintain continuity, we start with a most useful article by M. Devisch in a volume published in 1972 on Matthean questions.58 His article entitled ‘Le document Q, source de Matthieu. Problématique Actuelle’ is in itself a sound review of existing redactional studies of Q dealing with method, theology and Sitz-im-Leben. Devisch identifies with great charity the dangers and risks in identifying successive phases of the Q tradition in which the theology is supposedly located. He summarises the approaches of Polag, Lührmann and Hoffmann and concludes as follows: ‘To sum up, the method can be reduced principally to isolating by means of the Formgeschichte the pre-existing units in the great discourses of the Q source so as to describe in this way the redactional activity of the composer of these discourses and the theological motifs expressed there. When we can discover the same motifs for several discourses, we can, in agreement with Lührmann, justly characterize the document Q as a work of editorship or of theology.’59 We will leave our general criticism of this until the end.

It is convenient at this point, however, to mention the question of the literary genre to which Q belongs. One of the difficulties faced by supporters of Q in the past has been the fact that the idea of a sayings collection of the kind represented by Q is without parallel. In 1964, however, J. M. Robinson in an influential article60 put forward evidence to suggest that Q belongs on a ‘trajectory’ stretching from the Jewish wisdom literature to the full gnosticised Pistis Sophia in the second century ad. He cites in particular the Gospel of Thomas and says that Q and Thomas represent particular moments in the development of a literary Gattungwith a history in Judaism and in ‘orthodox’ and ‘heterodox’ Christianity.

Robinson’s view has had a mixed reception and Devisch is among those who find it untenable. He points to certain differences between Q and Thomas as refuting any connection between them. He points out (rightly) that Thomas has no Christology, where Q does, but when he says that ‘Q, actually is not a gnostic writing where the resuscitated Christ, placed outside the world and of time, reveals the gnosis’61 he is not understanding Robinson correctly. Thomas, according to Robinson, represents a movement away from the synoptic Jesus towards the gnostic revealer Jesus, but outwardly there is a similarity of form. In our view, this question requires more detailed attention before we can accept Kümmel’s conclusion that ‘the Gospel of Thomas can teach us nothing about the origin and the literary character of Q’.62

The French, it seems go in for solutions of the synoptic problem which can only be called elaborate, dealing in multiple sources. All that is possible here is a glance at one of the more involved, not to say eccentric, schemes described by P. Benoit and M. E. Boismard.63 This gives an outline of the pre-history of the gospels with a section-by-section commentary on the working out of the theory in practice. The complexity can best be illustrated by referring the reader to the review in JTS 25 p. 485 where an amazing diagram of interrelationships is presented!

Before drawing the threads together, we must mention the recent book by J. A. T. Robinson, Redating the New Testament64 insofar as it bears upon the question of Q. This book is worth reading as a brilliant tour de force, challenging as it does many sacred cows in the matter of relative dating. Bishop Robinson is found to be at many points a surprising ally of the conservative cause!

Robinson’s basic thesis is that the entire New Testament was completed by ad 70, which means, of course, that so far as the synoptic gospels are concerned, the time span for gestation and acceptance as source material is drastically reduced. The following brief extract will, I hope, whet the appetite: ‘My conclusion is that we must be open to seeing that the most primitive state of the triple … tradition is not consistently or exclusively to be found in any one gospel, to which we must then assign overall priority. Rather, I believe that there was written (as well as oral) tradition, underlying each of them, which is sometimes preserved in its original form by Matthew, sometimes by Luke, though most often I would judge by Mark. Hence the strength of the case for the priority of Mark, which is nevertheless overstated when this document is itself regarded as the foundation document of the other two. The gospels as we have them are to be seen as parallel, though by no means isolated developments of common material for different spheres of the Christian mission, rather than a series of documents standing in simple chronological sequence.’65

He sees Q then as a collection of sayings (rather than stories) but insofar as the gospels grew out of and with the needs of the communities, one must be prepared to allow for cross-fertilization between the on-going traditions.

To conclude this review of Q in recent debate, the following points emerge:

  1. The ‘fundamental solution’ proposed by Streeter, although not abandoned, is under considerable strain. We are witnessing the curious situation, in which, on the one hand, literary criticism is undermining the viability of Q as a distinct source and on the other, large edifices of theological construction are being erected on the assumption that Q is an ‘assured result’. This phenomenon demonstrates the inconclusive and indeed circular nature of Q studies. In this writer’s view, a Q hypothesis is still the best explanation of the large body of common material in Matthew and Luke, but the indeterminate scope and content of the hypothetical source Q remains a major problem. Further, such problems as the minor agreements of Matthew and Luke against Mark mean that the Griesbach challenge cannot be dismissed as a temporary aberration.
  2. Whilst we do not wish to detract from the solid achievements of practitioners of the theology of Q, this writer feels some disquiet at some of the results. Luhrmann for example stresses the Q distinction between Jesus (and John the Baptist) and ‘this generation’, but it seems clear that for Lührmann this reflects the situation of the Q community in its proclamation of judgment. The style of the proclamation is thus a community formulation and does not necessarily tell us anything vital about Jesus and his generation during the ministry. Can we, indeed visualize the kind of community delineated here by Lührmann and others, where in Lührmann’s words ‘continuity between Jesus and the Church is provided in eschatology and not in kerygma?66 What sort of gospel community is it which concerns itself with judgment and not with the saving events of the kerygma?67
  3. The third and final point arises directly out of the second. Devisch drew attention to studies like that of A. B. Polag who makes a distinction between the theology of the source and the theology of the editor who utilises the source. The identification of theological ‘fault lines’ (as Devisch calls them) is a very risky business, especially when we are working backwards to the supposed earlier form of a tradition, upon which late editors have imposed a fresh meaning, either by the addition of introductory matter, juxtaposition of content or use of Christological titles. As Devisch aptly points out, ‘what we are concerned with is the theology of the last editor,’68 but he implicitly concedes that to lay bare such final redaction is not possible for certain since the final editor may incorporate the theology of his source intact. What we must not lose sight of is the fact that Matthew and Luke have incorporated into their gospels material which prima facie has certain emphases and concerns, but as finished products they display a concern to incorporate all aspects of Jesus’s career and its final character in ushering in the day of salvation.

1 ‘On Dispensing with Q’—Studies in the Gospels, Ed. D. E. Nineham (Blackwell, 1955)

2 The Four Gospels (MacMillan).

3 ET 1975 (SCM).

4 Op. cit., p. 65.

5 Epworth (1970) pp. 90ff., also JTS NS 4 pp. 29ff. (1953)

6 Op. cit., p. 91.

7 Pittsburgh (1970).

8 ‘Towards the Rehabilitation of Q’ NTS 11 pp. 169–181.

9 Four Gospels, p. 161.

10 NTS 15 (1968–69), pp. 249–261.

11 CUP (1969).

12 See also Mt. 7:2 par Lk. 6:38 cf. Mk. 4:24; Mt. 3:2 par Lk. 3:3 cf. Mk. 1:4; Mt. 3:11 par Lk. 3:16 cf. Mk. 1:7f.

13 Cf. also points where Matthew and Luke agree to some extent in placing same material at same place in Marcan outline where such agreement not due to Q. Mt. 13:33 par Lk. 13:20f.; Mt. 18:10–22 par Lk. 17:3f.

14 Art. cit., p. 261. But see the article in The minor agreements of Matthew and Luke against Mark (291ff., B. E. T. L. Louvain, 1974) for some modification of Sanders’ argument.

15 How Luke was written (Cambridge, 1915).

16 The origin of the Synoptic Gospels (Blackwell, 1922).

17 The originality of Matthew (Cambridge, 1951).

18 New York (1964), recently republished without alteration by Western North Carolina Press (1976).

19 JTS Vol. XXV, p. 377.

20 The Minor Agreements of Matthew and Luke against Mark (B. E. T. L. Louvain, 1974).

21 Art. cit.

22 ‘The Minor Verbal Agreements of Matthew and Luke Against Mark’—Studia Evangelica I (1959), pp. 223–234.

23 ‘In Defence of Q,’ ExpT 68 (1956–7), pp. 315–318.

24 NTS 12 (1965–6), pp. 273–284.

25 Art. cit., p. 274.

26 Cf. the remark of Downing Art. cit. p. 171 that Q material in the Beelzebul controversy formed a ‘Coherent Narrative.’

27 Art. cit., p. 275. Major agreements: the inclusion of Q material alongside certain of the minor agreements gives them an importance which distinguishes them from other agreements in Mark.

28 Art. cit., p. 282. Simpson effectively disposes of Hawkins’ suggestion in Horae Synopticae that some of Matthew and Luke’s agreements against Mark may be due to their use of a revised ‘deutero-Mark’. In a number of places Luke appears to conflate Matthew’s improved version of Mark with the original (e.g. Lk. 3:4a, 16) so that he must have known both. On Simpson’s general approach see now M. D. Goulder, ‘On putting Q to the test’ NTS 24, No. 2 (1978).

29 ‘Mark—the abridgement of Matthew and Luke’ in Jesus and Man’s Hope Vol. 1 (Pittsburgh, 1970), pp. 51–97.

30 Op. cit., pp. 248, 272. Also his article ‘A fresh approach to Q’ in Christianity, Judaism and other Graeco-Roman Cults—Studies for Morton Smith at Sixty Pt. 1 (ed. J. Neusner-Leiden, 1975), pp. 39–50.

31 Cf. Streeter op. cit., p. 306 and the criticism of Sanders in ‘The overlaps of Mark and Q and the Synoptic Problem’ NTS 19 (1973), pp. 453–65. His own criteria are open to criticism.

32 Op. cit., p. 217. Farmer is attending to E. D. Burton’s Principles of Literary Criticism and the Synoptic Problem (Chicago, 1904).

33 See Farmer’s comment op. cit. p. 248, ‘Whenever Mark undertook to conflate or combine material from one of his sources with parallel material from another, he tended to confine himself to literary units between which there already existed a close relationship of literary dependence’ (!).

34 See F. Neirynck Duality in Mark (Louvain, 1972), p. 56.

35 Lohmeyer in Das Evangelium des Matthaus (Gottingen, 1967), p. 216 put forward another theory of the Markan conflation of two sources.

36 E.g. Mk. 3:22–30 par Mk. 4:21–5 par Mk. 1:12–13 par.

37 For an assessment of the Synoptic situation by a classical scholar see J. M. Rist On the Independence of Matthew and Luke SNTS Monograph No. 32 (Cambridge, 1978).

38 London (1937).

39 An attempt to translate the German Redaktionis-geschichte.

40 SCM (1968).

41 The Son of Man in the Synoptic Tradition (SCM) ET, 1965.

42 Ibid., p. 247 and R. Bultmann Theology of the NT 1, p. 34.

43 This comparison is arbitrary since Mark goes on to record Jesus’ preaching of the Kingdom in 1:14f.

44 Ibid., p. 252.

45 Ibid., p. 247 (Citing Bultmann Theology 1, p. 34).

46 Cambridge (1963), pp. 366–386.

47 Ibid., p. 369 (Citing T. W. Manson).

48 Ibid., p. 385.

49 Ibid., p. 386.

50 See also particularly J. M. Robinson ‘On the Gattung of Q’ in The future of our Religious Past—Essays in honour of R. Bultmann SCM (1971), pp. 84–130.

51 Die Redaktion der Logienquelle (WMANT 33: Neukirchen-Vluyn: Neukirchener—Verlag, 1969).

52 Studien Zur Theologic des Logienquelle (NT Abh 8, Münster: Aschendorff, 1972).

53 See the valuable new study by D. Hill Prophecy in NT (MM & S, 1979) who levels very just criticism against this widely held view.

54 ‘Die Form Von Q als Solche gnostisierend sei’ (ibid., P. 91).

55 My translation.

56 See Jeremias Parables, pp. 160–162.

57 Die Spruchquelle der Evangelisten—Zurich, Theologischer Verlag (1972).

58 BETL (Leuven Univ. Press).

59 Art. cit., p. 89 (my translation).

60 See note 50 ante.

61 Art. cit., p. 85.

62 Kümmel Introduction p. 76. See also M. J. Suggs, Wisdom, Christology and Law in Matthew’s Gospel(Harvard, 1970), pp. 6ff.

63 Synopse des quatres Evangiles en francais—Tome II (Paris, 1972).

64 SCM (1976).

65 Op. cit., p. 93f.

66 Op. cit., p. 96.

67 See G. N. Stanton ‘On the Christology of Q’ in Christ and Spirit in the NT (Cambridge, 1973), pp. 27–42.

68 Art. cit., p. 90.

Howard C. Bigg