Return to Main Page -O- Order The Case Against Q (US) / UK

On Dispensing With Q


A. M. Farrer

This article originally appeared in D. E. Nineham (ed.), Studies in the Gospels: Essays in Memory of R. H. Lightfoot (Oxford: Blackwell, 1955), pp. 55-88, © 1955 Basil Blackwell, reproduced here with permission.
Web version by Mark Goodacre. The numbers in bold refer to the original page numbers.


WHY dig up solid foundations, why open questions long taken for settled? Much critical and expository work rests squarely on the Q hypothesis, and if the hypothesis loses credit, the nuisance will be great. The books we rely upon to guide our thought about the history of Christ will need to be read with painful and unrelaxing re-interpretation. Nor is it only the effect on past studies that disquiets us. We want an accepted foundation for our present studies, and it seems a grievous thing that we cannot proceed with them until we have re-investigated what was unanimously settled by a previous generation. Is there to be no progress in learning? Now that criticism is a science, are we not to hold any established positions as permanent conquests, from which a fresh generation can make a further advance? Have we always to fight the old battles over again? Minds of high ability and scrupulous integrity were brought to bear on the Q question in the great days of source-criticism. They sifted to the bottom, they counted every syllable, and they agreed in the substance of their findings. Is it likely that we, whose attention is distracted by the questions of our day, can profitably do their work again? And what reason have we to trust our judgement against theirs, if we find ourselves dissenting from their conclusions?

It would certainly be impertinence to suggest that the scholars who established the Q hypothesis reasoned falsely or misunderstood their own business; no less an impertinence than to talk of the great Scholastics so. St. Thomas understood the business of being an Aristotelizing Augustinian, and if I am not his disciple, it is not because I find him to have reasoned falsely. It is because I do not concede the premisses from which he reasoned. And if we are not to be Streeterians, it will not be because Dr. Streeter [56] reasoned falsely, but because the premisses from which he reasoned are no longer ours.

I take the situation to be this. Since Dr. Streeter wrote, our conception of the way in which the Gospels were composed has gradually altered; so gradually, that we have not observed the extent of the alteration. Nevertheless the change that has taken place removes the ground on which the Q hypothesis stood. For the hypothesis wholly depends on the incredibility of St. Luke's having read St. Matthew's book. That incredibility depends in turn on the supposition that St. Luke was essentially an adapter and compiler. We do not now, or ought not now, so to regard him. And being once rid of such a supposition, we can conceive well enough how St. Luke could have both read St. Matthew's book as it stands, and written the gospel he has left us. Then at one stroke the question is erased to which the Q hypothesis supplied an answer. For the hypothesis answered the question, "From what does the common non-Marcan material of Matthew and Luke derive, since neither had read the other?"

If there is no difficulty in supposing St. Luke to have read St. Matthew, then the question never arises at all. For if we find two documents containing much common material, some of it verbally identical, and if those two documents derive from the same literary region, our first supposition is not that both draw upon a lost document for which there is no independent evidence, but that one draws upon the other. It is only when the latter supposition has proved untenable that we have recourse to the postulation of a hypothetical source. Now St. Matthew and St. Luke both emanate from the same literary region-both are orthodox Gentile-Christian writings-composed (let us say) between A.D. 75 and A.D. 90, in an area in which St. Mark's Gospel was known. Moreover, St. Luke's own preface informs us that he writes "in view of the fact that several authors have tried their hands at composing an account of the things fulfilled among us". He claims to know, and, one would naturally suppose, to profit by, more than one gospel-narrative other than his own. By all agreement he knew St. Mark's, but what other did he know? It would be natural for him to know St. Matthew's, supposing always that it had been in existence long enough.

The point we are making is that the hypothesis of St. Luke's using St. Matthew, and the hypothesis of their both drawing independently from a common source, do not compete on equal terms. The first hypothesis must be conclusively exploded before we obtain the right to consider the second at all. Such is the [57] actual case. There are, of course, possible cases in which the hypothesis of a lost and unevidenced source might compete on equal terms with the hypothesis of simple borrowing. Suppose, for example, that the passages common to A and B have a strong distinctive flavour, unlike the remaining parts of either A or B. Suppose further that the common passages, once we have extracted them, cry aloud to be strung together in one order rather than in any other; and that being so strung together they make up a satisfyingly complete little book, with beginning, middle and end. Then indeed we might postulate the existence of a common source, without waiting to prove that B cannot derive directly from A, nor A from B.

But in the case before us neither supposition holds good. To begin with the second -- it is notorious that Q cannot be convincingly reconstructed. No one reconstruction, to say the least of it, is overwhelmingly evident, and no proposed reconstruction is very firmly patterned. It is fair enough to object that Q may in fact have been a somewhat shapeless writing. It may indeed, but if it was, then no positive argument can be drawn from its shapeliness or cohesion to its existence as a single distinct work. Then to take the other supposition. Can we say that the Q sections of St. Matthew's Gospel have a strong distinctive flavour, marking them off from the rest of his writing? We cannot. They have a special character of a sort, but a character which can be plausibly enough described as Luke-pleasingness. It seems a sufficient account of them to say that they are those parts of St. Matthew's non-Marcan material which were likely to attract St. Luke, in view of what we know about the general character of his Gospel, or can conjecture about his aims in writing it. For example, St. Luke was not interested in the detail of the anti-Pharisaic controversy and neglects much teaching of Christ which attacks the Pharisees on their own ground. Must we therefore distinguish in Matthew two elements, M and Q, M rabbinic in tone, Q popular and nonrabbinic, of which St. Luke knew Q, but not M? Will it not do as well to say that St. Luke let alone what he did not care for, viz., the rabbinic parts of Matthew?

There is another supposition which, if we could make it, might raise the hypothesis of a lost and unevidenced common source to something like a priori equality with the hypothesis of direct borrowing by one of our documents from the other. And that would be, if the lost source we proposed to postulate were a sort of book known to have been plentiful at the time. For example, suppose we were struck by certain resemblances between two [58] Victorian novels, suggestive of actual literary affinity. Then there would be scarcely any a priori disparity between the two hypotheses (a) that one borrowed direct from the other (b) that both were indebted to some third novel unknown to us. For there were a great number of novels published at the time, and many of them have since sunk into oblivion.

But unhappily the postulation of Q is quite the opposite of such a case. We have no reason to suppose documents of the Q type to have been plentiful. It is vain to cite Streeter's M and L, for the M and L hypotheses are corollaries to the Q hypothesis and have no independent standing. No, in postulating Q we are postulating the unique, and that is to commit a prima facie offence against the principle of economy in explanation. St. Luke's preface is evidence that several authors earlier than himself had undertaken the composition of an account of the things fulfilled in the Christian dispensation. But Q does not answer to the description. The "things fulfilled" are, in St. Luke's view, the death and resurrection of Jesus above all. Q is not supposed to have contained an account of them, and therefore Q is not covered by St. Luke's words. He was talking about gospels, about the sort of book he himself proposed to write. And Q was not a gospel.

There was a time when appeal was made from the silence of St. Luke to the supposed informativeness of the elder whom Papias cited. "Matthew arranged the revelations (logia) in Jewish speech." Had we not here, perhaps, a reference to the Aramaic original of Q? Our Gospel of St. Matthew was certainly not written by the Apostle whose name it bears, nor was it written in Aramaic. Perhaps, then, what St. Matthew really did compose was the Aramaic Q, and it was to this that the elder referred. Such was the suggestion. I do not propose to deal with it in this essay, partly because it has now been generally abandoned, and partly because I have written what I have to say about Papias's elder in the first chapter of my book called "A Study in St. Mark" [Note (MSG): Austin Farrer, A Study in St Mark (Westminster: Dacre Press, 1951)].

So there is no independent evidence for anything like Q. To postulate Q is to postulate the unevidenced and the unique. But there is worse yet to come. For it may seem tolerable to postulate even the unique and the unevidenced if the circumstances of the time were such as (in our judgement) to call for its production. "We have no evidence that the primitive Christians ever put together a Q or anything like it. Never mind; we can see that a Q is just what they would have wanted to produce towards the year 60." Can we indeed? I am afraid we cannot. But let us look once more at the familiar story. "In the middle of the first century [59] men recited the saving Passion as a set piece. Its dramatic quality made it easily memorable and the need to commit it to writing was not early felt. But the teaching of Jesus Christ was another matter. It was miscellaneous and not easy to hold in one's head as a whole body of doctrine. Nor was there any occasion for the continuous recitation of the whole teaching, in the way in which we presume the whole Passion to have been recited. And so it was natural that the Christian teacher should be equipped with a written manual of the teaching, and no less natural that the narrative of the Passion should be omitted from it. And such a manual we take Q to have been".

It is a well-sounding story, but unfortunately it does not square with the Q which the gospel facts require. For Q has to be allowed to possess a strongly narrative exordium, not to mention narrative incidents elsewhere interspersed. It is no simple manual of Christ's teaching. It tells us with considerable fullness how John Baptist preached before the public manifestation of Jesus, and how Jesus, appearing in fulfilment of John's prophecies -- and, it would seem, undergoing baptism at his hands -- endured a threefold temptation in the wilderness, after which he ascended a mountain, and was joined by disciples there. Having delivered beatitudes and precepts of life, he "concluded his words" and presently made his way into Capernaum, where his aid was invoked by a centurion on behalf of his servant.

Not only is the narrative character of such an opening strongly marked; it further betrays a vigorous symbolical interest in the order of the events. It treats the Lord's temptations in the wilderness as the manifest antitypes to the temptations of Israel in the wilderness, three times citing the appropriate verses of Deuteronomy. Then it proceeds to bring Christ, as Israel was brought, to a mountain where divine teaching of special weight is delivered. If Q would have to be credited with a narrative of Christ's baptism immediately preceding his temptations (and it seems that it would), then another piece is added to the symbolical pattern. For, says St. Paul, it was after Israel had been "baptized unto Moses in the cloud and in the sea" that the people underwent their several temptations, in trial of their steadfastness in the grace they had received (1 Cor. I0.1-11). So Christ's baptism in Jordan and the descent upon him of the Spirit will answer to the passage of the Red Sea and the descent of the Shekinah.

This pattern of symbolism and narrative finds a natural place in St. Matthew's text, where, in our opinion, it indubitably originated. But what sort of place would it find in the imaginary [60] Q? After an exordium so full of dogmatic weight and historical destiny, is it credible that the book should peter out in miscellaneous oracles, and conclude without any account of those events which, to a Christian faith, are supremely significant? A primitive Christian writer might well string together the teaching of Christ and leave it at that. Or again, he might despair of the attempt to describe the ministry historically, and treat it as simply the field of a teaching activity, but provide it nevertheless with a historical exordium and a historical conclusion. What is hard to believe is that he should supply the exordium, while omitting the conclusion; that he should set in train the only story of unique importance, and break it off.

It can fairly be said that it took time for the whole body of Christian teaching to be brought into relation with Christ's redemptive acts. Men who knew themselves to be saved through Christ alone might make homilies on duties and virtues, citing the Old Testament, citing examples from common life, and making no mention of redemption through Christ. St. James's Epistle is not, perhaps, one of the earlier pieces in the New Testament, but it is arguable that it represents the survival of an early attitude. Such an attitude might find expression in the composition of a collection of Christ's sayings, without any narrative of his passion. But for an author to set about the narrative of Christ's life, and never conclude it with his death, is another thing.

Appeal has been made to the example of Old Testament prophecies. The call of an Isaiah, Jeremiah or Ezekiel is carefully narrated, and so are the acts in which the prophet begins to fulfil his calling. But the conclusion of the book is not the conclusion of his life, but (it may be) certain of his weightiest oracles. Isaiah was supposed to have suffered under Manasseh as Christ suffered under Pilate, but the book of Isaiah does not record his death. Why should not the author of Q follow the scriptural example, and write a "prophetical biography" of Christ beginning with history, proceeding to discourse, and ending with eschatological oracles? Why should he not? Because Christ was no mere prophet. Isaiah was no more than a prophet, an instrument of the Lord's word. It concerns us to know the history of his call, and how it was obeyed, for therein his authentication lies. The story of his end might be edifying, but it would be irrelevant to his message. The divine act in Isaiah is his call, not his death. It is otherwise with Christ.

It has sometimes been supposed that there was a primitive Christianity -- perhaps, indeed, the most primitive of all -- which [61] attached no positive value to Christ's death and resurrection, nor believed Christ to have attached any such value to these events beforehand. Christians of this school had only one concern about the Passion -- to palliate with scriptural excuses a disconcerting interlude between the coming of Messiah and the Kingdom of God. I have yet to be convinced that there were such Christians, or that their existence in the first days was a psychological possibility. They were presumably Jews, and no Jew could hold a negative attitude to Messiah's suffering an accursed death. No Jew could apologize for the cross unless he could glory in it. Yet Jews of a kind (I take it) are credited with having composed Q, and passed it current upon the Gentile churches.

It is sometimes thought that a decent agnosticism about the shape and nature of Q is a safe and honourable position. Why not be content to say that our two evangelists drew from a common written source, or sources, may be, but that we are in no position to decide what sort of writing, or writings, they drew from? Very well; but if so, the Q hypothesis must be allowed to lose heavily in a priori probability. The postulation of unevidenced writing of an indeterminable sort is a hazardous proceeding. If we were dealing with a rich and various literature it might be tolerable. If, for example, we return to our imaginary case of the two Victorian novels. Then we might say, "The common source may be another novel, or a magazine story, or a newspaper report of a law-court drama, or one or more of several other things". But what did the primitive Christians write, beside letters and homilies and gospels? Q was neither a letter nor a homily, nor was it a gospel. "Some writing or other, never mind what" will scarcely pass.

So far we have said nothing new. The difficulties of the Q hypothesis have been fully canvassed by its candid admirers, and subsidiary hypotheses have been introduced to meet them. A good deal of such hypothesis may be found in Dr. Streeter's 'Four Gospels' [Note (MSG): B. H. Streeter, The Four Gospels: A Study of Origins (London: Macmillan, 1924)], all of it developed with undeniable care and skill. But the palm should surely be awarded to his management of "the agreements of Matthew and Luke against Mark".

The difficulty Dr. Streeter has to face is that St. Luke, in a fairly large number of places, makes small alterations in the wording of his Marcan original which St. Matthew also makes. Now this is just what one would expect, on the supposition that St. Luke had read St. Matthew, but decided to work direct upon the more ancient narrative of St. Mark for himself. He does his own work of adaptation, but small Matthaean echoes keep appearing, because St. Luke is after all acquainted with St. Matthew. Such [62] is the apparent evidence against Dr. Streeter; such is the single hypothesis which springs immediately to our minds and covers all the facts.

What does Dr. Streeter do? He divides the evidence into several groups and finds a distinct hypothesis for each. In some cases he supposes that scribal error has assimilated St. Luke's text to St. Matthew's where no such similarity originally stood. In other cases it will be St. Matthew's text that has been assimilated to St. Luke's. In a third set of cases St. Matthew and St. Luke may really have coincided, but the original of their coincidence stood in St. Mark's text, from which scribal error has subsequently effaced it. In a fourth group of instances stylistic, and in a fifth doctrinal interests may have suggested the same emendation of St. Mark to St. Matthew and St. Luke independently. There remains a sixth group, where the coincidences are coincidences of substance, not amenable to any of the five methods hitherto advanced. In these cases it will be fair to suppose in Q itself a parallel to that Marcan paragraph upon which St. Matthew and St. Luke are both principally working. They both happen to incorporate the same Q features in their Marcan transcripts --that is all.

Thus the forces of evidence are divided by the advocate, and defeated in detail. His argument finds its strength in the fewness of the instances for which any one hypothesis needs to be invoked; but the opposing counsel will unkindly point out that the diminution of the instances for each hypothesis is in exact proportion to the multiplication of the hypotheses themselves. One cannot say that Dr. Streeter's plea is incapable of being sustained, but one must concede that it is a plea against apparent evidence, and that, other things being equal, we should accept the evidence and drop the plea. Of course, on Dr. Streeter's view, other things are by no means equal. There are solid grounds for denying that St. Luke can have known St. Matthew. Here is the heart of the matter. It is these grounds that we have to examine. But before we proceed to do so, let us sum up our preliminary survey.

The Q hypothesis is not, of itself, a probable hypothesis. It is simply the sole alternative to the supposition that St. Luke had read St. Matthew (or vice versa). It needs no refutation except the demonstration that its alternative is possible. It hangs on a single thread; cut that, and it falls by its own weight. [63]


Why is it said that St. Luke cannot have read St. Matthew? Five reasons may be considered.

  1. There are texts in St. Matthew which St. Luke would not have omitted, had he been acquainted with them.

  2. Where St. Matthew and St. Luke give the same saying of Christ, St. Luke's wording sometimes has the more primitive appearance.

  3. Our indubitable evidence for St. Luke's manner of using a written source is his use of St. Mark, whom he follows in continuous order over considerable stretches. Whereas if he used St. Matthew we should have to suppose that he treated him in a quite different way, dividing his text into small pieces and making a fresh mosaic of them.

  4. The order in which St. Luke places the material common to himself and to St. Matthew is mostly less appropriate and less coherent than the order it has in St. Matthew.

  5. In St. Matthew much of the material common to him and to St. Luke alone is placed in the context of Marcan paragraphs. St. Luke, even when he reproduces the same Marcan paragraphs, does not place the material we are speaking of in them, but somewhere else.

I shall make immediate comments on these five considerations, and afterwards proceed to a more systematic argument.

  1. No one has ever attached decisive importance to St. Luke's unexplained neglect of certain Matthaean texts, and whatever importance it ever had derived from an antiquated view of St. Luke's attitude to his work. If we regard him as essentially a collector of Christ's sayings, then the omission of some particularly striking blossom from his anthology may seem incompatible with his having known it. But if he was not making a collection but building an edifice, then he may have omitted what he omitted because it did not seem serviceable to his architecture nor come ready to his hand in the building of it.

  2. The suggestion appears to be that we should take separate units of discourse in isolation and pronounce on their degree of nearness to the spoken words of Christ. And where we find greater primitivity of form in this sense we are to impute literary priority. If the more primitive form were always St. Matthew's, then we might suppose that St. Luke had used, and in using modified, him. But since (it is alleged) the more primitive form is sometimes St. Matthew's and sometimes St. Luke's, it is more [64] reasonable to suppose that they used a common source, which now the one modified, and now the other.

    There is a deceptive simplicity about the proposed method of argument which evaporates as soon as we try to apply it. There is scarcely an instance in which we can determine priority of form without invoking questionable assumptions. "If I by the Spirit of God cast out devils" writes St. Matthew, and St. Luke, "If I by the finger of God". St. Luke's version contains a forcible allusion which St. Matthew lacks (Exod. 8.19, Matt. 12.28, Lk. 11.20). Is such an allusion more likely to be original, and later effaced by a more commonplace substitute, or adventitious, and due to our evangelist's Bible learning? "Until heaven and earth pass away, not one jot nor one tittle shall pass away from the law, until all be fulfilled". So writes St. Matthew, and St. Luke, "It is easier for heaven and earth to pass away, than for one tittle of the law to fall". Who can say whether the rhetorical fulness of St. Matthew, or the pointed brevity of St. Luke is more likely to be original? Is the copiousness of St. Matthew that of the Galilean gospel, or that of (say) the Antiochene pulpit? If we look at the context, we observe that St. Matthew is developing a flowing discourse (5.17-48), whereas St. Luke is giving us one of those short paragraphs packed with gnomic sentences which are an occasional feature of his style (16.15-l8, cf. 12.49-53, 16.8-13, 17.1-6) We are left in complete indecision. Either could be adapting the other's text to his own purpose.

    Even the apparently plain cases turn out to be not plain at all. We all agree at first sight that Christ is more likely to have blessed the poor, than the poor in spirit. "In spirit" looks like an editorial safeguard against misunderstanding: to be in lack of money is not enough. St. Luke's phrase, then, is the more primitive. But on the other hand St. Luke's eight beatitudes-and-woes with their carefully paired antitheses are not a more primitive affair than St. Matthew's eight beatitudes, but very much the reverse. And the phrase "in spirit" cannot stand in St. Luke's beatitudes-and-woes without overthrowing the logic of the paragraph. The poor are opposed to the rich. The poor in spirit would challenge comparison with the rich in flesh, but that does not mean anything. Thus St. Luke may well have read "in spirit" in St. Matthew, and dropped it in obedience to the logic of his own thought.

    The case of the Lord's Prayer is equally inconclusive. Here we may hesitate to attribute the greater bareness of the Lucan version either to editorial economy or to the logical requirements of the context. For surely the words of the Lord's Prayer must be [65] sacred to a Christian. But if they are sacred to him, it is because they are hallowed in usage, not because they happen to turn up in a book from over the sea. The presence of the Lord's Prayer in St. Matthew's Gospel may suggest to St. Luke the appropriateness of placing that prayer in his own, but he may nevertheless write it in the form familiar to those for whom he writes. Now it may be true that the prayer current in (let us say) Achaea towards the end of the first century was more primitive than the prayer current in Antioch at the same time and even a decade earlier. But that casts no light whatever on the literary relation between St. Luke and St. Matthew.

    We must content ourselves with these few examples of an enquiry which yields no decisive results. To express my own opinion, I agree with the findings of Harnack and of Loisy, rather than with those of Dr. Streeter. For much the most part the Matthaean forms look the more original. But I would not base any argument on such grounds.

  3. The suggestion that St. Luke might be expected to use St. Matthew as he uses St. Mark sounds reasonable on a first hearing but it will not bear examination. To follow two sources with equal regularity is difficult. Anyone who holds that St. Luke knew St. Matthew is bound to say that he threw over St. Matthew's order (where it diverged) in favour of St. Mark's. He made a Marcan, not a Matthaean, skeleton for his book. But as to the clothing of the skeleton, was not St. Luke going to do that according to his own wisdom, or where was the peculiar inspiration God had given him to operate? Is it surprising that he should lay his plan on Marcan foundations, and quarry St. Matthew for materials to build up his house?

  4. It may well be that we shall have to accuse St. Luke of pulling well-arranged Matthaean discourses to pieces and re-arranging them in an order less coherent or at least less perspicuous. St. Luke would not be either the first planner or the last to prefer a plan of his own to a plan of a predecessor's, and to make a less skilful thing of it. We are not bound to show that what St. Luke did to St. Matthew turned out to be a literary improvement on St. Matthew. All we have to show is that St. Luke's plan was capable of attracting St. Luke. You do not like what I have done to the garden my predecessor left me. You are welcome to your opinion, but I did what I did because I thought I should prefer the new arrangement. And if you want to enjoy whatever special merit my gardening has, you must forget my predecessor's ideas and try to appreciate mine. [66]

  5. It is largely true that St. Luke does not give non-Marcan material the same Marcan setting as St. Matthew gives it. But that is not to say that he transfers it to other Marcan settings. He does not incorporate it in Marcan episodes at all. What we have to explain is the single fact that St. Luke disencumbers the Marcan narrative of St. Matthew's additions to it, and puts them by themselves. The fact is striking enough, and certainly requires explanation. But it is capable of being explained, as we will proceed to show.


The Q hypothesis is a hypothesis, that is its weakness. To be rid of it we have no need of a contrary hypothesis, we merely have to make St. Luke's use of St. Matthew intelligible; and to understand what St. Luke made of St. Matthew we need no more than to consider what St. Luke made of his own book. Now St. Luke's book is not a hypothetical entity. Here is a copy of it on my desk. Let me consider what kind of a book it is.

Dr. Streeter says that St. Luke wrote his book in alternate Marcan and non-Marcan strips. That is, roughly speaking, true, but it casts at the best an indirect light on what St. Luke was trying to do. "Strip-formation" was not his formula for writing a gospel, especially as he was at pains to make the strips invisible. It is only by a tedious comparison of his text with St. Mark's that we establish the division into strips at all. The strip-formation is the by-product of something St. Luke really was trying to do, and it is this that we have to find out. Dr. Streeter's observation is exterior and diagrammatic, like the observation that my journeys to Paddington bunch together in certain months of the year, with wide gaps between the bunches. It is not my purpose to spend a good part of the months of March, May, July and October on the Oxford-Paddington line, while keeping off it in the intervening months. My doing so is incidental to the execution of more intelligible projects.

St. Luke's non-Marcan strips are very far from equal. One of them, in fact, is out of all proportion to the others (9.5l - 18.l4) and it alone corresponds (very nearly) to a single striking and visible feature of this gospel. No one, reading St. Luke for his own sake, would notice the discrepancy between Marcan and non-Marcan strips in 4-9, but every attentive reader observes that 10.25 - 18.30 constitutes a prolonged lull in the progress of the action, and that St. Luke uses it to set before us the greater part of the teaching of Christ. [67]

Surely this part of St. Luke's plan is immediately intelligible. If you or I attempted an account of Christ's life, we might do worse than finish the history of the Galilean ministry, and then break off to give an account of our Lord's teaching, illustrated, perhaps, by anecdotes. Then we might resume the narrative style to describe the visitation of Jerusalem, the passion and the resurrection. And that is what St. Luke does, except that he does not formally abandon narrative style in his middle section. It would, of course, be quite alien from the ways of a primitive Christian evangelist to do that. What St. Luke does is to have a sort of narrative stand-still. A period in which nothing of decisive historical importance happens provides a setting for the exposition of the teaching.

Such an arrangement is natural in itself, but more particularly it commends itself to a writer who has St. Mark's and St. Matthew's gospels both before him. He is struck by the special excellence of each and would be happy, if he could, to combine them. St. Mark has narrative vigour and rapidity of movement. St. Matthew has fullness of doctrine and exhortation. St. Mark is deficient in discourse, St. Matthew, by constantly exploiting the occasions for discourse which St. Mark supplies, somewhat muffles the action: the discourses run so long that we lose sight of the narrative situation altogether. An obvious way of keeping abundance of doctrine without allowing action to be muffled is to put doctrine in a place by itself. In nine and a half chapters of lively narrative St. Luke gives us the nativity and childhood, the relations with John Baptist, and the great events of the Galilean ministry: the works of power, the appointment and mission of the Twelve and the Seventy, the feeding of multitudes, the confession of Peter, the Transfiguration. In eight chapters more he gives us the teaching and in the remaining six and a half returns to unencumbered narrative for the events at Jerusalem.

The plan is a happy one, and in its narrative parts it is an undisputed success. It is only in respect of the teaching part that we can find a shadow of justification for Dr. Streeter's boutade, that if St. Luke did what he did after reading St. Matthew, he behaved like a madman. St. Luke's teaching section is not so complete a literary success as St. Matthew's great discourses. But then what St. Luke attempted was, on any showing, an awkward task. One great Sermon on the Mount covering eight chapters instead of three was not to be thought of, and three Sermons on the Mount, one after another, would be scarcely more thinkable. It is a paradoxical truth, but a truth nevertheless, that an evangelist who proposed to himself a long continuous teaching was bound to [68] carve it up. The discourses of Christ in St. Luke's middle part are conceived in episodes of moderate length, one following another. And it must be difficult to employ such a method without seeming somewhat monotonous and somewhat miscellaneous. Fresh episodes arise, but nothing much happens; the teaching is the thing, but the teaching is unsystematic because episodic.

But even if St. Luke was going to give the teaching in episodes, not in great discourses, might he not have profited more from the preparatory work St. Matthew had done for him? Could not he have broken the Matthaean discourses as they stood into two or three parts each at the points of logical division, provided each part with a distinct narrative setting, and left it at that? Has he not given himself unnecessary trouble in his handling of Matthaean material, and trouble worse than vain, if the Matthaean paragraphs are better than St. Luke's mosaics?

To ask such a question is to misunderstand St. Luke's task in 10-18. He is not dividing and re-arranging existing material, he is presenting his vision of the gospel according to his inspiration. And inspiration works in such a field as this by novelty of combination. Every episode in these chapters puts together two texts at the least which had not been combined before, and the new combination reveals the point that St. Luke is specially inspired to make. To say that St. Luke's points are less natural or less well made than St. Matthew's is irrelevant. St. Luke was not re-writing, still less abolishing, St. Matthew: St. Matthew remained to teach the Church St. Matthew's lesson. St. Luke was bound to write what was committed to him, and he was not free to cross it out afterwards even if the excellent and candid Theophilus found it inferior to St. Matthew in literary skill.

Every one of the short episodes in Luke 10.25-18.11 is composite. This fact, so far from being a scandal, so far from making St. Luke's handling of St. Matthew incomprehensible, is our best clue to what St. Luke was doing. It was the standing method of the Jewish preacher to seek his inspiration in the drawing together of old texts into fresh combinations: the striking of the flints brought forth the spiritual fire. The preacher would not merely juxtapose his texts, he would put in his own words what issued from their juxtaposition. St. Luke perhaps adds little of his own except by way of setting and suggestion. He puts the texts down side by side, and leaves them to speak for themselves, like the texts combined in the liturgy for a feast day.

But surely, it will be said, St. Luke was no Jew; it is not permissible to invoke the methods of the Jewish pulpit to explain him. [69] We must answer that the "Jew or Greek?" issue is not so simple as that. No New Testament writer was all that Jewish, and none of them was all that Greek. Let St. Luke have been a Greek; that is to say, an uncircumcised man. That will not have prevented him from standing, year after year, among God-fearing gentiles in the local synagogue, storing his mind with the Septuagint (what primitive Christian knew it better?) and accustoming himself to the methods of the rabbinic expositor. And when he adhered to the Church he would find nothing different. There were the same Greek scriptures, as soon as the congregation had contrived to get a set; and there was the Christian preacher, using the same weapons to vindicate a fuller truth.

What strikes us about St. Luke is not his hellenism but his versatility. His history unfolds in the bosom of Jewish piety and works its way out into the hellenistic agora. The infancy of Christ is written in the spirit of Tobit, the tumult at Ephesus almost in that of Lucian. The appropriate manner comes ready to the matter. The preaching of Jesus Christ is Jewish preaching, and St. Luke becomes the Jewish preacher in delivering it. We must not first assign St. Luke the Grecian label and then argue to the contents of the parcel. We must study to unfold just how Greek and just how Jewish he was.

A few examples of St. Luke's method in 10-18 will have to suffice here.

In the Sermon on the Mount, St. Matthew attaches the Lord's Prayer somewhat loosely to the second paragraph on the unostentatious performance of the three good works, almsgiving, prayer and fasting (6.7-l5). A couple of pages later, in what appears the most miscellaneous part of the Sermon, he has the paragraph "Ask, and it shall be given you, seek and ye shall find" (7.7-12). The Lord's Prayer and the "Ask" paragraph surely demand to be put together. At 11.1-13 St. Luke in fact joins them in a single episode. There is no doubt about its singleness. It is marked off from what precedes by its own narrative introduction, "And it came to pass that, as he was praying in a certain place". It is similarly marked off from what follows by the introduction of the next paragraph "And he was casting out a dumb demon, and it came to pass that . . ." Let us see how, in the area thus delimited, the new Lucan combination handles the old Matthaean material.

St. Matthew's paragraph on the Lord's Prayer ends with a comment: "For if ye forgive men their trespasses, your heavenly Father will forgive you, but if ye forgive not men their trespasses, neither will your Father forgive you your trespasses". The comment [70] fixes our attention on one clause of the prayer particularly, "Forgive us our debts, as we have forgiven our debtors". But if we want to pass on from the prayer to the "Ask and it shall be given you" paragraph, this is not the clause of the prayer to be kept specially in mind, but "Give us this day tomorrow's bread". For that paragraph continues "What man is there of you, of whom his son shall ask bread, and he will give him a stone?" St. Luke smooths the transition by omitting the comment on "Forgive us our debts".

That omission once made, the transition from the one Matthaean paragraph to the other could perfectly well be immediate. But St. Luke prefers to embellish the transition with a parable from his own store, preached (as it might seem) on three phrases of the second Matthaean paragraph, "Ask, and it shall be given you" -- "Knock, and it shall be opened to you" -- "If ye, being evil, know how to give good things" . . . . A man knocks up a friend at night to ask for the loan of three loaves. He is not a good friend; he yields to the other's importunity, not to his own good nature; but he yields. After the perfect introduction which such a parable affords, the second Matthaean parable follows with redoubled force. And who will hesitate to say that in the episode taken as a whole St. Luke has put an aspect of Christ's true teaching in a fresh and clear light, by means of the combination he has made?

St. Luke gives a twist to the last phrase of the discourse, when he particularizes the "good thing" which above all we should ask of our Heavenly Father. It is "Holy Spirit". By means of this particularization the evangelist eases the transition to his next episode, in which Christ will cast out an unclean spirit "by the finger of God", and give a warning against leaving the room vacated by the demon empty. It need not be empty, if the Heavenly Father only awaits our prayer to garrison it with Holy Spirit. But the special twist St. Luke gives to the termination of his episode not only opens the way to the next episode, it also echoes the termination of the episode preceding. It is not bread after all (the evangelist is telling us) that we should make most work about, but a diviner gift. And so, to go back a paragraph, Martha had been mistaken in being so preoccupied with the preparation of a meal. There was one thing needful and Mary had chosen the good part in seeking it at Jesus' feet.

To proceed with the next Lucan episode, the Beelzebul sayings (11.14-28). The divination on which St. Luke built that episode was a perception of the relation between two Matthaean parables, "Who can enter into the strong man's house and spoil his goods, unless he first bind the strong, and then he will spoil his house" -- [71] "When the unclean spirit goes out of a man he wanders through waterless places seeking rest, and finds none. Then he says, I will return to my house whence I came forth," etc. (Matt. 12.29, and 12.43-46). St. Luke perceives that it is actually the same house in the two parables, here despoiled of the gear of devilry, there found swept and garnished and re-occupied by the demon. Not content with juxtaposing the two parables, St. Luke equalizes them, writing up the first in the style and almost to the scale of the second. "When the strong man in armour guards his house, his goods are in peace; but when the stronger than he comes upon him, he prevails over him, and takes his armour wherein he trusted, and divides his spoils. He that is not with me is against me, he that gathereth not with me scattereth. When the unclean spirit goes out of a man, he passes through waterless places, seeking rest; and finding none, he saith, I will return to my house whence I came forth. And coming, he finds it swept and garnished. Then he goes and takes seven other spirits more wicked than himself, and entering they dwell there, and the last state of that man is worse than the first."

In Matthew the long discourse which ends in the parable of the house swept and garnished has for its pendant the visit of Christ's mother and brethren. That is a Marcan episode, and St. Luke has already reproduced it in a Marcan setting (Lk. 8.19-21). He now writes an evident equivalent for it as a pendant to the episode of the disputed exorcism. Not the womb or the paps of Mary are so blessed as they who hear God's word and do it. Would St. Luke have taken the hint from St. Matthew and repeated his Marcan theme here unless it had served him to bring the conclusion of the exorcism episode into line with the conclusions of the two previous episodes? Not Martha, who prepares nourishment for Christ, is so blest as Mary, who listens to his word. It is good to ask daily nourishment from God, but above all it is good to ask for Holy Spirit. Not the womb that bore Christ or the paps that nourished him are so blest as they who hear the word of God and keep it.

In joining the house swept and garnished to the house defended in arms, St. Luke has omitted two intervening paragraphs, the blasphemy of the Holy Ghost and the sign of Jonah. He takes the sign of Jonah for the beginning of his next episode. But he links the new episode to the old in a peculiar way, which clearly betrays dependence on St. Matthew, or (if you will have it so) on a Q which was virtually identical with St. Matthew for a couple of pages.

St. Matthew has two connected episodes, each with its own [72] narrative occasion. (a) The accusation "By Beelzebul" led Christ to give the Beelzebul parables and to add a warning against blaspheming the Holy Ghost. (b) The demand for a sign occasioned Christ to speak about the sign of Jonah and to give the parable of the house swept and garnished. As we have seen, St. Luke forms a single episode from the head of (a) and the tail of (b), the Beelzebul parables and the parable of the swept and garnished house. Then he begins a fresh episode with the head of (b), the sign of Jonah. But instead of giving each episode its own narrative occasion, he puts together both occasions into a joint occasion for the beginning of the first episode. "Some of them said, he casts out devils by Beelzebul the prince of devils, and others tempting him, asked of him a sign from heaven." By the time we reach the end of the episode the malice of the some has been fully answered, but the temptation from the others has not been further alluded to. Christ addresses himself to it in the next episode (11.29ff). "And as the crowds gathered about him, he proceeded to say: This generation is an evil generation; it seeketh a sign, but there shall no sign be given it save the sign of Jonah."

The sign of Jonah is only the beginning of the new episode. What makes the episode is the inspired juxtaposition of the sign of Jonah (Matt. 12.38-42) and the lamp of the body (Matt. 6.22-23). The lamp of the body is the eye; the body is enlightened if the eye is good. The "good eye" signifies generosity, and St. Matthew is attacking miserliness in the Sermon on the Mount when he records how Jesus had declared that the good eye lets the light into our own person; it does not merely direct the beam of favour upon our neighbour. It is St. Luke's inspiration to see the connexion between the evil eye's exclusion of light, and the evil generation's blindness to a more than Solomon, a more than Jonah in their midst. In divining this connexion, St. Luke sees what is particularly characteristic of his own vision of the gospel. What shuts out the light of supernatural revelation is the refusal of a moral demand, and primarily the demand of generosity. Hearing Christ's teaching the Pharisees mocked him, because they were lovers of money (16.14).

The parable of the good and evil eye, if it is to have its full effect, must stand between matter explicitly concerned with failure to see divine signs on the one side and matter explicitly concerned with the denunciation of covetousness on the other. St. Luke makes a further divination no less brilliant than the last, when he passes on from the evil eye to the woes on the Pharisees. The evil eye darkens the whole man within, and Jesus had called the [73] Pharisees blind, because they cleansed the outside of the platter, when they should have taken thought lest what was inside it might be impoverishing the needy. He had proceeded to transfer the outside-inside antithesis from the platter to its owner, a sepulchre whitewashed without, but full of dead men's bones within. So in the episode of the Pharisaic lunch-party (11.37-52) St. Luke goes on to give a carefully arranged anthology of texts from the woes on the Pharisees (Matt. 23). He begins from the topic of miserliness and works round again to the rejection and suppression of divine truth. The Pharisaic brotherhoods are covetous and hypocritical (11.39-44), their scribal teachers are the enemies of God's word (11.46-52). They reject more than Jonah; their fathers killed the prophets, they complete their fathers' work and bring all the blood of God's messengers on their own generation.

We will turn back and pick up a couple of small points. (a) St. Luke simplifies the sign of Jonah by omitting the distracting allegory on the whale's belly and the Easter sepulchre (Matt. 12.40). Our attention is left free to concentrate on the perversity, more than that of Nineveh, which rejects a more than Jonah, and we are ready to be taught the cause of it in the parable of the evil eye. (b) The parable of the eye itself receives a convenient introduction in the form of a sentence culled from the beginning of the Sermon on the Mount: "No man lighteth a lamp and putteth it in the closet or under the bushel, but on the lampstand, that those who come in may see the light".

Nothing but a complete exposition of St. Luke's gospel could provide a complete refutation of the Q hypothesis, and, conversely, when such an exposition had been made, no further arguments in refutation of Q would be required. We have merely attempted a specimen of St. Luke's working from St. Matthew in 10-18. So far from his possession of St. Matthew making what he does a mystery, his possession of St. Matthew is the indispensable explanation of what he does. Let us follow St. Luke's eye and memory as they run up and down St. Matthew's pages under the direction of his own inspiration. To enter into the mind of St. Luke at work would be to dissolve the mystery, and, in the nature of the case, nothing else can possibly dissolve it.


We have been discussing the teaching section (Lk. 10.25 - 18.30). This section, we have said, is roughly equivalent to the widest by [74] far of St. Luke's non-Marcan strips. But what are we to say about the contents of the other strips? About those of them that consist of incident we need say nothing at all. That St. Luke should intersperse his Marcan narrative with non-Marcan incidents or versions of incidents is the most natural thing in the world. So we find him giving the Matthaean account of John Baptist's preaching and of the Lord's temptations; having his own views about the migration from Nazareth to Capernaum, which St. Matthew has mentioned in the same place (Lk. 4.16-29, Matt. 4.13) and about the call of Simon Peter (5.1-10); paraphrasing the Matthaean story of the centurion (7.2-10), and adding from his own store the widow of Nain (711-17) and the sinful woman (7.36-50); adding Zacchaeus to the story of Jericho and adding a parable to Zacchaeus (19.1-27). There is nothing surprising about such embellishments of the Marcan story, nor is it at all surprising that they tend to come in groups. St. Luke is following St. Mark as his main narrative guide, and feels the spell. When he has once turned his eyes away from the Marcan text he is open to think about his other stores of knowledge. When at length he returns to St. Mark the spell reasserts itself and he follows his written guide for some distance before digressing again.

But since we have said that St. Luke's plan assembles the Lord's sustained teaching in a single place (10-18), we may be expected to show why the evangelist gives a Sermon on the Mount (or under it, rather) in 6.20-49, and why Christ's sayings about John Baptist are recorded in 7.18-34. The placing of the sayings about John presents no difficulty on any showing. They are inseparable from their Matthaean introduction, the message from John in prison. By the time St. Luke's long teaching section begins, John has been already reported dead (9.9). An incident from his life in prison could scarcely come later than 7. But constant to his purpose of reserving the bulk of Christ's teaching for the central section, St. Luke detaches all he can from the Lord's discourse upon this occasion according to St. Matthew (Matt. 11.12-15, 11.20-30, cf. Lk. 16.16, 10.12-25).

St. Luke's sermon at the mountain is also vastly shortened from the Matthaean form. It has the same beginning and the same end as its original. But in the body of the sermon St. Luke, with a skill from which no one can withhold the praise, extracts a single essence from the wide range of the Matthaean sermon, renunciation seen as humility and generosity. Everything which does not belong to the chosen theme is left for a more convenient occasion. But why (it has still to be asked) should St. Luke give us even a [75] short sermon at the mountain, if he has resolved to keep Christ's teaching for the middle part of his book? If the sermon at the mountain is not a formal declaration of the teaching, then what is it?

Why St. Luke did what he did, rather than anything else, cannot be the question. He did what he was moved to do. It is enough if we can see what he did, and what he meant by it. At an earlier point in this essay we imagined St. Luke coolly resolving to put the mass of Christ's teaching where it would least impede the action. And that is what St. Luke did in effect resolve to do, and we may believe that he was not insensitive to the purely literary advantages of the choice he made. But it is not very likely that the choice would present itself to him as a mere point of literary craft. Let us endeavour to give a more plausible story of how a first-century evangelist arrived at such a decision.

We will suppose that St. Luke has St. Matthew before him. Now St. Matthew is a forerunner in the course which St. Luke is about to take: he first has written a new Mark with the Lord's teaching more fully embodied in it. What path has St. Matthew taken? How closely will St. Luke wish to follow it? St. Matthew has not been content simply to exploit such opportunities for the development of discourse as St. Mark happens to afford. He has so arranged his matter as five times to bring the teaching to a head in a set discourse, and in case we should fail to distinguish the five discourses from other passages of dialogue, he has concluded each discourse with an identical phrase: "And it came to pass when Jesus had finished these sayings . . ."

It has been suggested that St. Matthew's five set pieces (1) have something to do with the five books of Moses, as though the evangelist were presenting his gospel as a new Pentateuch. The suggestion, in that form, remains sterile. We are disappointed to discover that the first set piece has nothing to do with Genesis nor the second with Exodus. We have made a mistake somewhere. Our mistake was to miss the first set piece of all, the genealogy, with which the Gospel opens. If that is not a set piece, what is? It cannot, of course, have the set conclusion "When Jesus had finished these sayings", for it does not consist of his sayings but of his ancestors. In any case the set conclusion has not yet been set, [76] and the only reader who looked for it at 1.17 would be the reader who read the book backwards. He who takes it as it comes is put on the right track from the very first moment by a different and far more explicit indication. "Book of Genesis of Jesus Christ" is the title to the genealogy and the first line of the Gospel. The new "Book of Genesis" derives the legal ancestry of Jesus from the hero of the old Genesis, Abraham. Having done with genealogy, St. Matthew resumes: "Now the genesis of Jesus Christ was thus . . ." So much, then, for Genesis.

The Exodus set-piece is identified neither by heading nor by termination, but by context and character. That the Sermon on the Mount stands out as a formal unity scarcely needs to be said. It is a new law from the mountain, like the law of Sinai, and the setting is strikingly reminiscent. Jesus passes the waters and undergoes forty days" temptation in the wilderness after the pattern of Israel at the Red Sea and in the desert. Then he comes to the mountain of revelation. By using the formula "When Jesus had finished these sayings" by way of conclusion to what is obviously an Exodus discourse in its own right, St. Matthew first gives it significance as the termination to a "set scriptural piece". When it recurs we shall know what to make of it.

St. Matthew's Leviticus is the mission-charge which is also, in his Gospel, the institution of the Twelve (10). The example "Book of Genesis" in 1.1 (cf. 1.18) has already shown us that our author is sensitive to the prima facie meaning of a book-title. Now "Leviticus" means "The Book about Levites" and the Apostles are the corresponding ministry of the New Covenant. Similarly, if we are to go by titles, "Numbers" is the muster of the host. St. Matthew's Leviticus (10) sends forth "labourers into the harvest", the Parables which compose his Numbers (13) show how plenteous the human harvest is, how numerous the catch to which fishers of men were previously called (4.19); they deal with the criterion according to which some pass the muster and are admitted to the promised land, while others are rejected.

It remains that the next set piece (Matt. 18) should be a Deuteronomy. The Marcan original is already so Deuteronomic at this point that there is little left for St. Matthew to do. The Transfiguration has already brought Moses to witness the divine repetition of his Deuteronomic testimony about his great Successor, "Hear ye him" (Deut. 18.13, Mk. 9.9, Matt. 17.5). In the discourse at Capernaum St. Mark, and St. Matthew following him, proceed to take up the next preceding paragraph of Deuteronomy, the Law of the Kingdom (Deut. 17.14-20, Mk. 9.32-37, Matt. 17.25 - 18). [77] The princes in God's kingdom are not to exert privileges or make exactions like Gentile kings, but to humble themselves among their brethren. St. Matthew goes further than St. Mark by going one paragraph further back in Deuteronomy (Deut. 17.2-13, Matt. 18.15-20). The Israelite who has a grievance against his neighbour must be prepared first to call two or three witnesses, then to have recourse to a higher court: the decree of ultimate authority must be enforced. The sequel to the Matthaean set discourse rejoins St. Mark, and remains in step with Deuteronomy. It is the question of divorce (Deut. 24.1-4, Mk. 10.1-l2, Matt. 19.1-12). The next paragraph, the embracing and blessing of the children, simply repeats the Deuteronomic theme of princely humility, while the paragraph after that carries us to the very heart of Deuteronomy (Deut. 5-6, Mk. 10.l7-3l, Matt. 19.16-30). For the episode of the rich man's question associates the keeping of the decalogue with the Oneness of God, the attainment of "life", and "inheritance". The exhortation to make distribution and to shun the snare of riches is no less Deuteronomic.

We have had five Matthaean "Books of Moses". There remains one "book" (24-25), the "Book of Jesus" (Joshua) without a doubt. The new Jesus comes through Jericho, indeed, but it is Jerusalem he condemns to utter overthrow, so that not one stone shall remain upon another. The fall of the city is the sign and the condition of the gathering of Israel into the true land of promise under the leadership of Jesus (23.37 - 24.2, 24.l5-3l).

Such in outline is the structure of St. Matthew's hexateuch, and if we are allowed to reason a priori at all, we must suppose it to have been as evident to St. Luke as it is to us, for he was a next-door neighbour and we are visitors from a far country. Supposing then that St. Luke understood it, what did he do with it? Did he adopt it, or reject it? He did neither. He allowed the general pattern to stand, but he redistributed the weight of the teaching, placing as much of it as he could in the Deuteronomic position. Shall we allow the question "Why?" to be asked once more? We have answered it already in terms of literary propriety and of respect for Marcan narrative. Must we answer it over again in terms of scriptural typology? Among all the books of Moses why should Deuteronomy appeal to St. Luke as specially typical of Christ's doctrine? We are not bound to find certain answers to such a question, probable answers will do. If there are still more probable answers than those we find, why, so much the better.

First, then, the primitive Christian saw the Law reasserted and yet transformed in the Gospel, and it would easily strike him that [78] a model for such a relationship was to be found within the Law itself. In his Deuteronomy Moses reasserted his Protonomy, that is, the Law from Exodus to Numbers, and in reasserting it illuminated it. Had not St. John this example in mind when he meditated on the commandment which in being new is also old (1 Jn. 2.7-8)? The very occasion upon which Moses gave his Deuteronomy enforces the same point to the Christian mind. For it was in his last hours and in connexion with his giving place to the Jesus who could alone fulfil his words, and who was the first to be designated by that promise on which we have already dwelt: "The Lord will raise up unto you a prophet from among your brethren like unto me: to him hearken".

Such considerations, being formal and typological, make less appeal to us, perhaps, than they did to the first-century mind. But there are more material considerations with which our sympathy will be as great as St. Luke's own. Deuteronomy is the book which adds the spirit to the observance, it is the law of love towards God and man, and especially of humility, generosity, and compassion. It is well indeed if these virtues are as dear to us as they were to St. Luke.

St. Luke might desire, therefore, on some such grounds as these to place the weight of Christ's teaching in what his predecessors had already marked out as the Deuteronomic position. But logic forbade him to gather the whole of it there. The Deuteronomy will not stand out as Deuteronomy unless there is some semblance of a Protonomy; without a first law the second law will be second to nothing. The recapitulation on the plains of Moab presupposes a first statement at the foot of Sinai. And so St. Luke gives us a short sermon beneath the Mountain in 6 as well as the long discourses of 10-18.

The Deuteronomic passage in which Moses most clearly draws a new command out of the old is to be found in Deut. 5-6. Here the Lawgiver first recapitulates the decalogue from Exodus and then adds the Shema as the heart of the matter. The Lord is One Lord, and is to be loved with entire devotion. Now we have already seen that the passage is commented upon by the paragraph of the Rich Man, in which St. Matthew's Deuteronomic section culminates. St. Luke allows his own Deuteronomy to run out into the same conclusion (18.18-30). But he is not content to conclude with the Shema, he must begin from it too (10.25-28). His Deuteronomic exordium anticipates the explicit discussion of the Shema between Jesus and the Pharisaic scribe in the temple court, according to Matt. 22.35-40. (That the Matthaean version rather [79] than the Marcan is St. Luke's model is the natural conclusion to draw from a comparison of the texts. Streeter's plea to the contrary is a necessity forced upon him by his general position.)

The Scribe's Question and the Rich Man's Question are the twin pillars which mark out the extent of St. Luke's Deuteronomy, and the fact is made more evident by the evangelist's assimilation of the one to the other. A doctor of law is the questioner in 10, a ruler of synagogue in 18. Both ask the same question, the Deuteronomic question, "What must I do to inherit eternal life?" Both are credited with a knowledge of the formal answer which the old law supplies. It is the new Deuteronomy, the life-giving exposition of the old precept, that is reserved for Christ.

It seems, then, that St. Luke consciously regarded what he wrote in 10.25-18.30 as a Christian Deuteronomy. How far can we say that the contents of this Deuteronomy are Deuteronomic in order or in detail? They range over the field of human duty as Deuteronomy does, and in a Deuteronomic spirit. But do they follow the topics of the fifth Mosaic book with any particularity? Here is a complicated enquiry, and it is fortunate for us that Mr. Evans has undertaken it in this volume. We need only refer the reader to what he has written. [Note (MSG): C. F. Evans, "The Central Section of St Luke's Gospel", the preceding article in the volume, op. cit., pp. 37-53]

So much for St. Luke's Deuteronomy. But what, if anything, has he made of St. Matthew's Genesis, Exodus, Leviticus, Numbers and Joshua? He has denuded them of prolonged discourse; but has he entirely effaced them? Not entirely, but he has rubbed them faint.

As to the Genesis, St. Luke has his own infancy narratives, and their patriarchal, especially their Abrahamic, flavour is unmistakable. He has his own genealogy too, though he places it differently: after the end of his Genesis, not at the beginning. At first sight we are struck by the differences between the two evangelists in their opening chapters; their genealogies are not the same genealogy nor their narratives in any particular the same narrative. On second thoughts we observe the points of identity. The Matthaean genealogy has an artificial structure and an openly symbolical value: the Lucan genealogy develops the symbolical architecture of the Matthaean to a further pitch of elaboration, as the reader may see by referring to the note appended to this essay. The Matthaean narratives are made to revolve round two principal points: Jesus, by domicile a Nazarene, was a Bethlehemite by birth; Jesus, by family a descendant of David, was Son of God by supernatural generation. St. Luke's narratives present a story which a man who had it to tell might [80] surely prefer to the Matthaean form, even if he knew it. But it is to be remarked that he so tells it as to cover the two principal Matthaean points. What shall we say? We used to say: "His genealogy is a different genealogy, his infancy narratives are different narratives, he had not read St. Matthew". But now we shall say: "St. Matthew's early chapters define a task, which St. Luke takes up and deals with from his own resources, and with his own improvements. It is most unlikely that he had not read St. Matthew's." So much, then, for St. Luke's Genesis.

St. Luke's Exodus chapters preserve the most striking Matthaean feature, the temptations which Christ, after the example of ancient Israel, endured in the wilderness. They add two distinctively Lucan developments -- the rejection at Nazareth, embodying the principal discourse of St. Luke's Exodus; and St. Peter's confrontation with the supernatural in the miraculous fishing. The Scriptural typology of these two episodes can be studied in St. Stephen's Speech, Acts 7.23-35. They are antitypical to Moses' rejection by his brethren on his first appearance, and to the vision at the Bush (Ex. 2.l1 - 4.17).

After Exodus, Leviticus. St. Matthew's Leviticus is the institution, mission and mission-charge of the Twelve (10) with which the embassy from John Baptist is associated (11). St. Luke holds over the mission and mission-charge for the enrichment of his own "Numbers", but he still is able to present the institution of the Twelve and the embassy from John in close succession (6.12ff, 7.18ff). St. Matthew's Sermon on the Mount, i.e. his Exodus, becomes St. Luke's Sermon after the Mount, i.e. his Leviticus: it loses its character of being a comment on the Ten Commandments and becomes the ordination sermon of the new Levites ("Lifting up his eyes upon his disciples he began to say, Blessed are ye poor", etc.).

As the Leviticus begins with the institution of the Twelve, so the Numbers begins with their mission (9.l-10). The "Numbers" typology of this section stands out clearly. When our evangelist is simply following a source (say St. Mark) it is unsafe to attribute to him a conscious interest in every symbolical feature already embedded in the text he reproduces. But where he introduces his own additions and modifications, as he does here, it is reasonable to make him responsible for their more evident symbolical bearings. We observe the following facts. St. Luke so abbreviates St. Mark as to bring a certain sequence of events into close proximity: the commission of the Twelve (9.1-10), Jesus' reception of them on their return from mission (9.10ff), the disclosure of the [81] Divine Son to the Twelve and the leaders of the Twelve (9.18-45). After a few short incidents, Marcan and non-Marcan (9.46-62), St. Luke supplies a parallel cycle: the commission of the Seventy (10.1-l6), Jesus' reception of them on their return from mission (10.17-20), and the disclosure of the Divine Son to his disciples (10.21-24). Now that the Divine Son has been twice testified to as the sole revealer (9.35, 10.21-24) his law, his new Deuteronomy, most fitly follows (10.25ff).

St. Luke is himself responsible for placing the commission of the Seventy in striking and elaborate parallel with the commission of the Twelve. But to do so is undisguisedly to invoke the example of Moses, and of Moses in Numbers. For in Numbers the commission of the Twelve (1-2, cf. 7) is succeeded by that of the Seventy (11), not immediately, but only after the solemn setting forth of Moses and Israel for the promised land (10). St. Luke, too, places Christ's solemn setting forth between the cycles of the Twelve and of the Seventy (9.51-62). "It came to pass, as the days of his Assumption began to be fulfilled, he set his face to go to Jerusalem". "Assumption" is a word commonly used of the ends of Moses and Elijah. When we hear it, we still have those two saints' voices ringing in our ears. We have just heard them conversing with Christ on the mount of Transfiguration about the exodus he was to complete et Jerusalem (9.31).

In thus developing the theme of Numbers St. Luke lays the appropriate foundation upon which to raise his great Deuteronomic superstructure. For Deuteronomy itself opens with a recapitulation of precisely those incidents in Numbers to which St. Luke has supplied the antitypes (Deut. 1.6-8, the setting forth from Sinai; 1.9-18, appointment of ministers; 1.19ff, the sending of men to prepare the way whither Israel was to come). The Deuteronomic setting of Lk. 10.25ff could, in fact, scarcely be more strongly marked than it is. To ask for more would be blank ingratitude. It is hardly necessary to say anything about St. Luke's Joshua. For in any case the triumphant passion and resurrection compose the "Book of Jesus" par excellence. If the birth is a Genesis and the ministry a Lawgiving, then the death and resurrection are a Conquest. But we may anyhow observe that St. Luke shows himself fully alive to the Jerusalem-Jericho paradox. It is Jerusalem, not Jericho, that the new Jesus is called upon to overthrow by the trumpet of his prophecy. Jericho, once the city of the repentant harlot (Hebr. 11.31), is now the city of the repentant publican, and Jerusalem that of the proud Pharisee. That is the impression which we form, if we read the story of [82] Zacchaeus upon its Lucan background (19.1-27 and 19.41-48, cf. 18.9-30 and 10.30).


It is alleged by those who deny the credibility of St. Luke's having used St. Matthew, that St. Luke never places Matthaean material (in their language, Q material) in the Marcan place which St. Matthew assigns it; and that the fact is very surprising, if St. Luke knew St. Matthew's book. The allegation is not wholly true, to begin with; and what truth it has is no cause for surprise. Have we sufficiently considered the bewildering way in which Mk. 1-6 is used in Matt. 3-14? To find the "Marcan place" of any one paragraph in these chapters may be a teasing puzzle. If St. Luke began with the best will in the world to use Matthew as a direct comment on Mark, is it surprising that he gave it up in the maze of Matt. 3-4, simply followed Mark through, and dealt with the Matthaean additions afterwards on a system of his own?

What, in fact, had St. Matthew done in these chapters? Four times he skipped selectively over the same Marcan ground, each time making a fresh selection until the material was exhausted. In 3-7 he covered Mk. 1-3.13, recounting the teaching of John, the baptism and temptations of Jesus, Jesus' coming into Galilee and fixing upon Capernaum, his calling of the four; how a mission throughout Galilee (Mk. 1.39ff) led to the collection of a vast crowd from all the quarters of Palestine, which are named (Mk. 3.8) and how, in face of the crowds, Jesus ascended the mountain and his chosen disciples came up to him. So far St. Luke follows St. Matthew and refers all St. Matthew's special material to the corresponding Marcan places, including the Sermon at the Mountain. But when, in 8, St. Matthew jumps back to the scene in St. Peter's house, according to Mk., St. Luke deserts him, and is he to be blamed? St. Matthew, unaccompanied by St. Luke, flies over Mk. 1.29 - 5.21 in 8, leaps back to Mk. 2.1 and flies forward as far as 6.15 in 9-11, returns to Mk. 2.23 and reaches as far as 6 again in 12-14, after which he goes on steadily forward. But we are not concerned to unravel St. Matthew's doings ourselves; we are merely excusing St. Luke for not making the attempt.

We will clear up an allied difficulty, and so make an end. It is common form to say: If St. Luke drew the so-called Q material from St. Matthew, and yet did not produce it in the Marcan settings St. Matthew had given it, we must suppose that he went [83] carefully through his text of Matthew blocking out the Marcan parts, before he could see what was available for his own Q passages. And it is unlikely that St. Luke did this.

It is more than unlikely, but then there is no need to suppose it. Up to the point at which St. Luke makes his great desertion of St. Mark (9.50 = Mk. 9.40) the issue does not arise at all. The Matthaean material in Lk. 3-6 has its Marcan place; in 7.1 - 9.60 there are two Matthaean episodes, the centurion's message (7.2-10) and the Baptist's message (7.18-34). Each is already a distinct and self-contained episode in St. Matthew, wholly unconfused with its Marcan context, and St. Luke could be in no hesitation at all where to draw the boundaries round either. He shortens both, and makes internal rearrangements in the second, but that has no bearing on the point.

When, on the other hand, St. Luke laid St. Mark aside at 9.40 and took up St. Matthew for the composition of his long teaching section, he had already made such use as he wished to make of the Marcan elements in Matt. 3-18. And so, when he set about quarrying these chapters, all he needed to do was to bear in mind what elements in them he had used already. St. Matthew's Marcan material was marked off for him by the mere fact that he had just been using it in its pristine Marcan form. Equally, of course, he had already used some of the Matthaean material, for example in the Sermon and in the reply to John's disciples. He had no difficulty in letting alone what he had used, and picking up what he had neglected. He has no strict rule against Marcan material in his teaching section, but only against used material. He is perfectly ready to transcribe unused Marcan sentences embedded in Matthaean discourses, for example, in the Beelzebul controversy (Lk. 11.15-22 and 12.10) or in the sermon on the little ones (17.2).

There is no difficulty, then, about the selection of non-Marcan material from Matt. 3-18 for incorporation in Lk. 10-18. If there is a difficulty, it will concern the incorporation in these Lucan chapters of material from the Matthaean chapters which St. Luke has not yet skimmed of their Marcan elements, that is to say, from Matt. 19-25. For here we can no longer invoke the explanation we have given for the ready discrimination of Marcan from non-Marcan in St. Luke's use of Matt. 3-18.

The difficulty melts away on examination, because the anticipations of Matt. 19-25 which St. Luke does make in 10-18 are, with one exception, massively simple and not such as to lay up trouble for the future. There are six in all, and five of them are so [84] whole and single, that they come away clean from their settings. Here is the list:

  1. The lawyer's question (Matt. 22.35-40, Lk. 10.25-28).
  2. Woes on scribes and Pharisees (Matt. 23.1-36, Lk. 11.39-52).
  3. Servants watching (Matt. 24.42 - 25.12, Lk. 12.35-46).
  4. Jerusalem that slays the prophets (Matt. 23.37-39, Lk.13.34-35).
  5. Invited guests (Matt. 22.l-l4, Lk.14.16-24).

The sixth, and exceptional, case is the apocalyptic cento in Lk. 17.22-37, put together from non-Marcan details of the augmented Marcan apocalypse in Matt. 24.23-41. Here, and here only, St. Luke must be credited with measuring the Marcan text against St. Matthew's augmented version of it, before he reaches the place. But there is no great difficulty in believing that St. Luke already knew that he meant to give the substance of the Marcan apocalypse in its Marcan place. Its Marcan place is also, of course, its Matthaean place: and if we are right in supposing that St. Luke was ranging forward through St. Matthew's text when he composed Lk. 17, we may reasonably suppose also that he saw the Marcan apocalypse through its Matthaean wrappings and realized that he would need it later in its own position. To suppose this is further to suppose that St. Luke had the Marcan apocalypse virtually by heart. But there is no text he is more likely to have had by heart than that.

So much for the six anticipations in Lk. 10-18. They are neatly made, but they do not in fact altogether avoid trenching on Marcan material. When in due course St. Luke arrives at Mk. 12, he discovers that he has already used up the good scribe's question in the lawyer's question (Lk. 10.25-28), so he allows the merest vestige of it to appear in its own place (20.39-40). Mk. 13.21-23 is found to have been anticipated in the apocalyptic cento (Lk. 17.22-23) and so St. Luke omits it at 21.24f. Mk. 13.33-37 has been anticipated in the parables of the watching servants (Lk. 12.35-46). St. Luke substitutes a generalizing paraphrase for it in 21.34-36.

Besides these anticipations in the Lucan Deuteronomy, there is one which falls outside it. In the story of Christ at Jericho, St. Luke anticipates a piece of the Matthaean apocalyptic discourse, the parable of money on trust (Matt. 25.14-30, Lk. 19.11-27). This anticipation creates no kind of difficulty. The parable is a single unit and manifestly non-Marcan; it has not the least tendency to bring Marcan masonry away with it when it is pulled out of its Matthaean setting.

[85] Thus, when we come to look at the alleged mystery about St. Luke's wrenching of St. Matthew's non-Marcan material away from its Marcan contexts, it turns out to be no mystery at all. Everything that happens happens much as we might expect.

It is time that we concluded the whole discussion. Let us hope we have sufficiently stated the principles required for dispensing with the Q hypothesis, and done something besides to illustrate the application of those principles to the task. We have certainly not given a complete demonstration, for to do that would be nothing less than to write a complete exposition of St. Luke, beginning from the beginning and unfolding the movement of his thought as it comes. But, on the rash assumption that the fulfilment of such a labour would confirm our guesses, let us indulge ourselves a little here, and prophesy.

The literary history of the Gospels will turn out to be a simpler matter than we had supposed. St. Matthew will be seen to be an amplified version of St. Mark, based on a decade of habitual preaching, and incorporating oral material, but presupposing no other literary source beside St. Mark himself. St. Luke, in turn, will be found to presuppose St. Matthew and St. Mark, and St. John to presuppose the three others. The whole literary history of the canonical Gospel tradition will be found to be contained in the fourfold canon itself, except in so far as it lies in the Old Testament, the Pseudepigrapha, and the other New Testament writings.

The surrender of the Q hypothesis will not only clarify the exposition of St. Luke, it will free the interpretation of St. Matthew from the contradiction into which it has fallen. For on the one hand the exposition of St. Matthew sees that Gospel as a living growth, and on the other as an artificial mosaic, and the two pictures cannot be reconciled. If we compare St. Matthew with St. Mark alone, everything can be seen to happen as though St. Matthew, standing in the stream of a living oral tradition, were freely reshaping and enlarging his predecessor under those influences, practical, doctrinal and liturgical, which Dr. Kilpatrick has so admirably set before us in his book. (2) But then the supposed necessity of the Q hypothesis comes in to confuse us -- these apparently free remodellings of St. Mark cannot after all be what they seem, nor are they the work of St. Matthew in his reflection on St. Mark, for they stood in Q before St. Matthew wrote. And that is not the end of the trouble, for if the so-called Q passages were in a written source, so, we must suppose, were other Matthaean paragraphs which have the same firmness of [86] outline as the Q passages and are handled by the evangelist in the same way. They were not in Q, or St. Luke would have shown a knowledge of them, which he does not do. Never mind, we can pick another letter from the alphabet: if these are not Q passages, let them be M passages, or what you will. Once rid of Q, we are rid of a progeny of nameless chimaeras, and free to let St. Matthew write as he is moved.




The Matthean genealogy is commented on by its author. Three fourteens of generations correspond to three periods, before the kingdom, the kingdom, since the kingdom. The suggestion is, "And now the kingdom again" (2.2, 3.2, 4.17). That three spans should bring us to the kingdom of Christ, seems to inevitable to anyone acquainted with the Gospel tradition, "On the third day" or "After three days" (12.40, 16.21, 17.23, 27.63). The three spans are of fourteen each, and a fourteen strikes the Jewish mind as a fortnight, a double seven. Three fortnights -- otherwise put, six weeks, a working-week of weeks -- and then, of course, the Sabbatical week, the Messianic kingdom, must follow. The total number of generations contained in the six weeks has the same significance -- forty: "After forty years of wandering and temptation, the Promised Land" (4.2).

The number forty is not contained without art -- 7 x 6 = 42, but St. Matthew makes it forty by making David and Jeconias each do double duty: they end one fortnight and begin another. Such a reckoning may suggest a similar function for the name of Jesus -- he fulfils the working days and initiates the sabbath.

By noting the irregular marriages in the genealogy (Thamar, Rahab, Ruth, Bathsheba) St. Matthew shows that God can "of the stones raise up children to Abraham" (3.9) and in particular graft his Son into Abraham's stock by a virginal conception.

This genealogy has two formal faults:

  1. The artificial doubling of two names, as indicated.
  2. The omission of several generations from the biblical list between David and Jeconias.
Both faults are eliminated in St. Luke's rewriting.


The Lucan genealogy was conceivably written out by its author in groups of seven names each, a division disregarded by his copyists. However that may be, the clue for counting in seven remains embedded in the beginning of the list. Jesus is both the son of a Joseph, and the seventh descendant of another Joseph; and this remoter Joseph is himself both the son of a Mattathias and the seventh descendant of another Mattathias. For our present purpose it suffices to write down the beginnings and the ends of St. Luke's sets of sevens, leaving the middles blank.


  1. Jesus, Joseph . . . . . . . . . . . . . . . . . . Jannai
  2. Joseph, Mattathias . . . . . . . . . . . . . . Meath
  3. Mattathias . . . . . . . . . . . . . . . . . . . . Zerubbabel
  4. Shealtiel . . . . . . . . . . . . . . . . . . . . . . Er
  5. Jesus . . . . . . . . . . . . . . . . . . . . . . . . Judah
  6. Joseph . . . . . . . . . . . . . . . . . . . . . . . Nathan
  7. David . . . . . . . . . . . . . . . . . . . . . . . . Admin
  8. Arni . . . . . . . . . . . . . . . . . . . . . . . . Abraham
  9. Terah . . . . . . . . . . . . . . . . . . . . . . . . Shelah
  10. Cainan . . . . . . . . . . . . . . . . . . . . . . . Enoch
  11. Jared . . . . . . . . . . . . . . . . . . . . . . . . God.

The genealogy is written backwards. The name of Joseph in (2) suggests the family background of Jesus, Matthathias in (3), being the name of the father of the Maccabees, suggests the second Jewish kingdom, Shealtiel father of Zerubbabel in (4) brings us to the exile. The rhythm is then repeated: it runs through an earlier Jesus and an earlier Joseph to David, the father of the former kingdom, as Mattathias was of the later. And so we arrive with (8) at the previous exile -- Arni lived under Egyptian bondage, as did Shealtiel under Babylonish captivity.

By italicizing the exilic lines (4) and (8) as we have done, we reveal at a glance the meaning of the list. St. Matthew had a threefold division in his genealogy, of which Babylonish captivity marked the second period. St. Luke's system is likewise divided threefold, but now exile marks both the points of division. After a first captivity the Davidic kingdom arises, and in declining towards a second brings forth the name of Jesus, though not yet of the Jesus. After the second captivity the Maccabean state, declining towards a third captivity (the fall of Jerusalem, 21.24) brings forth Jesus Christ. But this coming of Jesus Christ closes no more than the eleventh "week" of generations, and the eleven "weeks" of St. Luke, like the six "weeks" of St. Matthew are an incomplete number (Acts 1.13-26). As the Matthean six point forward to a seventh, so the Lucan eleven point forward to a twelfth, the week of the fall of Jerusalem in which St. Luke lives, a week destined to last until the times of the Gentiles are fulfilled (Lk. 21.24). And that is the end (21.9, 27). The first advent occupies the seventy-seventh (11 x 7th) place, the eighty-fourth (12 x 7th) "year" is that perfect period at which the Son of Man, returning, finds faith in the "poor widow" who awaits him with constant prayer (2.35-38, cf. 18.1-8).

How does St. Luke obtain the liberty to construct so balanced a scheme as his genealogy? By deserting scriptural tradition from David to Jesus, he has the greater part of the list under his absolute control. He derives Jesus not from Solomon, but from his brother Nathan, whose descendants are nowhere listed in scripture.

A diagram of historical providence composed by the grouping of generations in "weeks" or sevens could be found by our Evangelists already standing in 1 Enoch 93.


(1) It is sometimes assumed that the set pieces are each the conclusions of whole "books" into which St. Matthew is divided. There is no obvious reason for that assumption. St. Matthew wrote his book in one continuous script, divided not by chapter headings, but by "stripes" in the subject matter, the "stripes" being these set pieces which carry the set terminations. The additional matter attaching to the "stripe" may be on either side of it, or on both sides.

(2) The Origins of the Gospel According to St. Matthew. [Note (MSG): the full reference is: G. D. Kilpatrick, The Gospel According to St. Matthew (Oxford: Clarendon, 1946).]

Return to Top

Return to Main Page -O- Order The Case Against Q (US) / UK

This file was last updated on 5 March 2002.

© 1997, 1998, 1999, 2000, 2001, 2002 Mark Goodacre