THE FOUR GOSPELS - A Study of Origins, The Manuscript Tradition, Sources, Authorship, & Dates. By B.H. Streeter, Hon.D.D.Edin. Fellow of Queen's College, Oxford; Canon of Hereford. First published by MacMillan & Co Limited 1924. - Prepared for katapi by Paul Ingram 2004.

Chapter X


Home | Diagram – synoptic sources | synopsis | Q: original order | textual assimilation & the Lord's Prayer | 5 blocks of Q | overlapping of Q & M | scattered fragments | omissions from Q | reconstructed Q | list of MSS | (pages 271-292)


the original order of Q

IF attention is concentrated on the larger groups of sayings, to the exclusion of scattered fragments, there is considerable agreement between Matthew and Luke in the relative order in which they arrange material from Q.

The diversity in order results chiefly from the incorporation by Matthew into the three great discourses, v.-vii., x., xi.-xxv., of material disposed of by Luke in different parts of his "Central Section."

Generally speaking, Luke seems to preserve the original order of Q.

textual assimilation and the lord's prayer

The Synoptic critic must be on the look-out for the possibility that even the best MSS. have been corrupted by assimilation from parallel passages in another gospel.

Evidence that the true text of Luke xi.2 read, "Thy Holy Spirit come upon us and cleanse us," which has been replaced by the words "Thy Kingdom come" from Matthew. If so, the Lord's Prayer was not found in Q; but two different versions of it were found by Matthew and Luke in M and L respectively.

five blocks of Q

In Luke's "Central Section" Q and L material is arranged, roughly speaking, in alternate blocks. A study of some shorter passages, within the blocks derived from Q, which are not found in Matthew, suggests that these passages may have stood in Matthew's copy of Q, but that he preferred and substituted for them the parallel versions of the same item which he found in Mark or M.

overlapping of Q and M

Five passages considered, in which Luke appears to have preserved the version of Q with but little modification, while Matthew has conflated Q with similar sayings from M.

scattered fragments

Though Luke normally gives Q material in its original context, he occasionally allows himself to depart from this usage in order to secure appropriateness of connection of subject matter.

The saving "Ye shall sit on thrones ..." and some editorial formulae.

omissions from Q

The probability that neither Matthew nor Luke made any considerable omissions from Q.

the reconstructed Q

List of passages from Luke probably derived from Q.

The document so reconstructed is one whose purpose and character is intelligible, in spite of its not having contained an account of the Passion.



the original order of Q

THE critic who wishes to reconstruct Q can start off at a run.
It having already been decided (p. 188) that Q contained an account of the Baptism and Temptation, we find at once five items that occur in the same order in both Matthew and Luke —John's Preaching, the Baptism, the Temptation, the Great Sermon, and the Centurion's Servant. Those portions of the Sermon on the Mount, which, though not contained in the Sermon on the Plain, are, yet from closeness of parallelism obviously derived from Q, must, we have already seen, have occurred in that document in some later context. In Luke the next Q item after the Centurion's Servant is John's Message,

"Art thou he that should come?"

This occurs somewhat later in Matthew; but the motive for postponement is obvious. Jesus refers John's disciples to the evidence afforded by certain miracles (Mt.xi.5); Matthew postpones the incident until he has had time to give an example, taken from Mark, of each of the miracles mentioned. Luke solves the same problem in another way, by inserting (vii.21) a statement, 'In that hour healed he many,' etc., which he doubtless supposed was implied in the context.
We infer that Luke's order is original.
Both Matthew and Luke then concur again in the relative order in which they introduce 'Foxes have holes,' 'The harvest is plenteous,' and the Mission Charge (which Luke gives as the Address to the Seventy, but which Matthew conflates with Mark's Charge to the Twelve); but it is noticeable that Matthew places the Charge much earlier than does Mark, a rearrangement of Mark which is probably due to the influence of the order in Q. Matthew also expands the discourse with Q material found elsewhere in Luke, as well as with material from other sources.

If we ignore the Q matter added to the Mission Charge, and also the section Lk.xi.9-13, which Matthew has given already in the Sermon on the Mount, the coincidence in order is interrupted by the occurrence in Luke of two verses (Lk.x.23-24), which Matthew postpones to a later context. But of the next five items, Woes to the Cities, "I thank thee, Father," Beelzebub Controversy, Parable of Unclean Spirit, and Sign of Jonah, the relative order (except that the last two are transposed) is the same in both Gospels. The next item in both is the pair of parables, the Mustard Seed and Leaven. This brings us into Mt.i. Now we have observed (p. 161) that up to this point Matthew seems to have rearranged the materials he took from Mark with the greatest freedom; but that from chap.xiv. onward he never departs from Mark's order. We seem to have lighted on the explanation. Matthew's rearrangement of Mark has been, at any rate partly
[The endeavour to group together representative miracles seems to have been another motive for rearrangement. Of. W. C. Allen, Commentary on Matthew, p. xiv ff.]
, determined by the necessity of combining Mark with Q. Thus the order of Q has evidently suggested to him to anticipate the place of the Mission Charge in Mark; and the late occurrence of the Mustard Seed and Leaven in Q has led him to postpone Mark's collection of parables of the Kingdom, among which he desired to include this pair from Q and others from M.

Most of the remaining Q material Matthew disposes of by working it into one of his great blocks of discourse, the Sermon on the Mount, v.-vii., the Mission Charge, x., or the Denunciation of Judgement, xi.-xxv. The problem, then, of the original order of Q resolves itself into the question. Are the additions in Matthew's versions of the Great Sermon and Mission Charge, the half-dozen or so scattered sayings and the Q matter in Mt.xi.-xxv., in a more original context in Luke or in Matthew? If, however, we consider
(1) Matthew's proved habit of piling up discourses compacted from Mark, Q and M;
(2) the fact that sayings like "Blessed are your eyes," Mt.i.16-17, concerning Offences, Mt.xviii.7 — being embedded in extracts from Mark — cannot possibly be in their original context as they occur in Matthew, the presumption is plainly in favour of the view that Luke's order is the more original.

This conclusion is important for the light it throws on the problem of the original extent of Q, for this, it will appear, is more closely bound up than one would suppose with the question of the original order. Luke in his use of Mark and Proto-Luke differs from Matthew in three ways.
(1) He as a rule avoids conflating his sources.
(2) He usually gives them in approximately their original order.
(3) He has a tendency to follow one source at a time. It looks as if the person who combined Q and L so as to form Proto-Luke, whether that person was (as I believe) Luke himself or someone else, adhered to the same methods.

textual assimilation and the Lord's Prayer

At this point Textual Criticism must be summoned to the rescue of the puzzled Synoptic critic. The evidence accumulated in Chap. XI. shows that assimilation between the texts of the Gospels in parallel passages has operated along every line of textual transmission, and that, though B has suffered less in this way than any other MS., it has not entirely escaped. That chapter being a study of minor agreements of Matthew and Luke against Mark, the passages examined are necessarily ones in which three Gospels were involved; but obviously assimilation would not be less likely to operate in passages contained only in Matthew and Luke. If an example is required of the avoidance by B of an assimilation found elsewhere, I would instance the omission by B Syr. S. of the words "bread, will he give him a stone, or," Lk.xi.11—an interpolation from the parallel Mt.vii.9. As an example of an assimilation that has infected B א, τασσόμενος, Mt.viii.9—an intrusion from Lk.vii.8.
["After these things," Mt.viii.5, Syr. S. k, instead of "He having entered into Capernaum " (= Lk.vii.1), and the omission by the same authorities of καὶ πτωχοὶ εὐαγγελίζονται, Mt.xi.5 (= Lk.vii.22), are probably original readings which have been altered through assimilation in all other MSS.]

But there are two cases where assimilation has affected B א etc. that are of real importance in the attempt to reconstruct Q. The first is the version of the Voice from Heaven at the Baptism, "this day have I begotten thee," Lk.iii.22, which has already been fully discussed, pp. 143, 188. All I need do is to recall the fact that the acceptance of the Western reading (in preference to the reading of B א, which has been assimilated to the other Gospels) proved that Luke derived his account of the Baptism, not from Mark, but from Q. The second occurs in the Lord's Prayer. The liturgical use of the Lord's Prayer in the form in which Matthew gives it would make assimilation of the shorter form in Luke to that of Matthew more inevitable than in any other passage in the Gospels. And the great MSS. and early versions show it at work in a very varied way.

In A D and the T.R. the assimilation of Lk.xi.2 ff. to ff. is almost complete.
Syr. C. has effected it, but for the clause "Thy will be done, etc." which it omits. א curiously enough inserts this clause, but leaves out "Deliver us from evil," which Syr. C. contains; א also joins B L Syr. S. etc. in omitting "our, which art in heaven" after "Father" in the opening address. B L Syr. S., and apparently the ancestor of fam. Θ, agree in all the omissions.

Here we find that B, as usual, has been less affected by assimilation than most other MSS.; but here also there is evidence that B has not entirely escaped. For 700, 162, instead of "Thy kingdom come", read "Thy holy spirit come upon us and cleanse us" (ἐλθέτω τὸ πνεῦμα ... ἐφ' ἡμᾶς).

And D has ἐλθέτω βασιλεία φ' ἡμᾶς, where, as Rendel Harris pointed out, ἐφ' ἡμᾶς is only explicable as a remainder of the other reading which a corrector of some ancestor of D omitted to strike out. And this reading was in the text of Luke used by Gregory of Nyssa in Cappadocia in 395; he says so plainly twice, and moreover gives no hint that he had even heard of any other reading. It is also quoted by Maximus of Turin, c. 450. So the reading was current both in the East and in the West to quite a late period. But it also stood in the text of Marcion (AD 140), and from Tertullian's comment on this it is not at all clear that his own text was in this respect different from Marcion's. Now in view of the immense pressure of the tendency to assimilate the two versions of this specially familiar prayer, and of the improbability that various orthodox Fathers should have adopted (without knowing it) the text of Marcion, the probability is high that the reading of 700, 162, which makes the Gospels differ most, is what Luke wrote. Matthew's version is here more original.

Now, even if we accept the reading of B, the difference between the two versions of the Lord's Prayer, Lk.xi.1-4 and, is so great as to put a considerable strain on the theory that they were both derived from the same written source. But, if we accept the reading of 700 and its supporters, that theory becomes quite impossible. [The rare word ἐπιούστιος remains as a remarkable point of contact between the two versions. I think it not impossible that its presence in Luke is due to an assimilation to Matthew which has infected all our authorities.] We next notice that in neither Matthew nor Luke are the sayings in the immediate context derived from Q; the Lord's Prayer in Matthew is in the middle of a block of M, in Luke in the middle of a section of L, material. The natural inference is that the respective versions belong to these two sources. I am aware that to some who have fallen into the habit of regarding Q as the sole citadel of authenticity it will be something of a shock to realise that the document did not include the Lord's Prayer. I would suggest, however, that the real effect of the discovery is to enhance our conception of the value of the special sources of Matthew and Luke.

five blocks of Q

We now turn to the text of Luke and discover that the assignment of the Lord's Prayer to L, instead of to Q, makes x.25-xi.8 a single block of L. Then we perceive that in Luke's "Central Section" the Q and the L material tends to sort itself out into alternate blocks — the five blocks ix.57-x.24, xi.9-.10, .22-.59, and perhaps i.18-i.35, being in the main derived from Q, while the intervening blocks are mainly L. There are never more than four, and rarely more than two, consecutive verses in any of these sections, which do not also occur in Matthew. We may conjecture, then, that they are really solid blocks of Q, from which Matthew, in the course of rearranging to fit into his great discourses, has omitted a few odd verses. This gives us a working hypothesis with which to start our quest.

The provisional hypothesis that the five passages just indicated are solid blocks of Q receives a good deal of confirmation from a closer scrutiny of a number of the short passages within these blocks, which Matthew omits, but for which he substitutes something derived either from Mark or M which might well be regarded as an equivalent. These are as follows: (a) Lk.x.16 ἀκούων ὑμῶν. For this Mt.x.40 substitutes ὁ δεχόμενος ὑμᾶς from Mk.ix.37. (b) Lk.xi.27-28.
An unknown woman cries, "Blessed is the womb that bare thee ..." Our Lord replies, "Nay, blessed are they that hear the word of God ..." In Luke this follows immediately after the parable of the Unclean Spirit.
Matthew immediately after this same parable (Mt..46-50) inserts our Lord's reply to the announcement, "Thy mother and brethren without seek thee," from Mk.iii.31, a different incident but one of which the moral is the same, "whoso will do the will of my Father ... the same is my brother and sister and mother." As there is no connection of thought between this and the preceding parable, the position of Matt..46-50 can only be due to a deliberate substitution of the Marcan for the Q version of the saying as to what constitutes true relationship to Christ. (c) Lk.xi.37-38, on not washing before meat.
Matthew omits this, but in xv.2 ff. has a much longer discussion of this point derived from Mk.vii.1 ff., which Luke omits. (d) The same thing probably applies to the "Leaven of the Pharisees" (Lk..1b).
The phrase stood both in Q and Mark. Matthew omits it where it occurs in Q, because he has it in a more meaningful context, xvi.6, from Mk.viii.15. [The Marcan equivalent of both (c) and (d) occur in Luke's "great omission" of Mark which I have argued (p. 175 ff.) did not stand in the copy of Mark which Luke used.] (e) Lk..35-38, "loins girded, lights burning."
Matthew leaves this out; but immediately after the (Q) paragraph which follows in Luke (Lk..39-46 = Mt.xxiv.43-51) he inserts the parable of the Wise and Foolish Virgins, which contains the same point as Lk..35-38, but considerably amplified. Accordingly "substitution" rather than "omission" again seems the proper description of his procedure. (f) Lk.i.25-27.
The main ideas, and even the more striking phrases "open," "I know you not," "depart from me," occur in the Apocalyptic parables of Matthew; cf. Mt.xxv.11-12, 41. (g) Lk.i.30, "the last shall be first."
There are two reasons—in addition to the fact that it occurs in Luke in connection with Q sayings—for referring this saying to Q. Mark x.31, followed by Matthew in the same context, Mt.xix.30, has the saying, but in the reverse order, "the first shall be last." But Matthew repeats the saying in another context, xx.16, but this time he gives the words in the same order as Luke. This then will be the Q order. Luke has it in a Q context (i.30); but note that, when he is copying the context in which it occurs in Mark (Lk.xviii.30), he leaves out just this one sentence— evidently because he has already recorded the saying in its Q form.

(h) Lk..11-12. Mk.i.11. Mt.x.19-20.
And when they bring you before the synagogues, and the rulers, and the authorities, be not anxious how or what ye shall answer, or what ye shall say:
for the Holy Spirit shall teach you in that very hour what ye ought to say.
And when they lead you to judgement, and deliver you up, be not anxious beforehand what ye shall speak: but whatsoever shall be given you in that hour, that speak ye:
for it is not ye that speak, but the Holy Ghost.
But when they deliver you up, be not anxious how or what ye shall speak: for it shall be given you in that hour what ye shall speak.
For it is not ye that speak, but the Spirit of your Father that speaketh in you.

We note Mark and Luke are furthest apart. Matthew is almost verbally identical with Mark. [In Lk.xxi.14-15, which is the actual parallel to Mt.i.11, there is a kind of paraphrase,

"Settle it therefore in your own hearts,
not to meditate beforehand how to answer:
for I will give you a mouth and wisdom,
which all your adversaries shall not be able to withstand or to gainsay

I suggest that Luke, recollecting that he had already copied from Proto-Luke a sentence practically identical with that in Mk.i.11, purposely paraphrased Mark's wording here so as to avoid tautologous repetition.] The only verbal agreement of Matthew and Luke against Mark is "how or what"; but nothing can be inferred from this since πῶς ἤ is omitted in Mark by a b k Syr. S. Cypr. and ἤ τί in Luke by D 1 157 Old Lat. Syr. S. and C. It is, however, noticeable that in both Matthew and Luke the saying occurs in the same discourse as, though separated by a few verses from, "there is nothing hidden which shall not be revealed," etc. (Mt.x.26 ff. = Lk. 2 ff.).


As there is no obvious connection of thought to suggest bringing the two together, the view that Lk..11-12 stood in Q and formed part of the block of Q material, .2-10, would explain the collocation in both Gospels. The saying will then be one of those, which in a slightly different form, occurred in both Mark and Q.

The reader who has followed the above with a Synopsis and a marked copy of the Gospels will see that the facts noted all tend to justify our provisional assignment to Q of the five specified blocks. I proceed to show that, if we invoke the principle of parallel versions worked out in the previous chapters, we discover phenomena, which not only add further confirmation to the above hypothesis, but also help to identify as Q certain passages of Luke that are outside these blocks.

overlapping of Q and M

There are certain cases where the parallels between Matthew and Luke are not close enough to make derivation from the same written source probable, but where the hypothesis which most easily explains the phenomena is that the saying stood in both Q and M—Luke reproducing the version of Q, Matthew con­flating Q and M.
(a) The saying about Forgiveness is perhaps the clearest example.
In Lk.xvii.1-4 this saying follows immediately after one about Offences, a version of which seems to have stood in both Q and Mark
[Mk.ix.42 has only one member of the double antithetical saying which occurs Lk.xvii.1-2, Mt.xviii.6-7.]
.  In Mt.xviii.15 ff. also the saying about Forgiveness follows that about Offences in the same discourse—only with half-a-dozen verses (from Mark and M) intervening. Seeing there is no very obvious connection of thought between the two topics, the connection (Offences—Forgiveness) must have been made in the common source Q. How, then, are we to explain the fact that, while the Offences saying is virtually identical in Matthew and Luke, that on Forgiveness appears in versions exceptionally diverse? I suggest that M also contained a version of the latter saying, which Matthew on the whole prefers; and this is not pure conjecture, for in the fragments of the Gospel according to the Hebrews we have evidence that this saying was in circulation in more than one version. It will be instructive to set the three versions side by side.

Mt.xviii.15-16, 21-22. Lk.xvii.3-4. Ev. Heb. (quoted by Jerome).
And if thy brother sin against thee,
go, shew him his fault between thee and him alone: if he hear thee, thou hast gained thy brother. But if he hear thee not, take with thee one or two more, that at the mouth of two witnesses ...
Then came Peter, and said to him, Lord, how oft shall my brother sin against me, and I forgive him?
until seven times?
Jesus saith unto him, I say not unto thee, Until seven times;
but, Until seventy times seven.
Take heed to yourselves:
if thy brother sin, rebuke him;
and if he repent, forgive him.
And if he sin against thee seven times in the day, and seven times turn again to thee, saying, I repent; thou shalt forgive him.
If thy brother shall have sinned in word and given thee satisfaction, seven times in a day receive him.
His disciple Simon said unto him, Seven times in a day?
The Lord answered and said unto him, Yea, I say unto thee until seventy times seven; for even in the prophets after they had been anointed by the Holy Spirit, there was found sermo peccati.
" (probably an Aramaism = "matter of sin," not merely = "sinful speech")

(b) The Parable of the Pounds = Talents.
A glance at a Synopsis shows that in the latter part of this parable the verbal agreements between the two versions are such as to favour, though not actually to compel, the assumption of a common written source. But the divergences between the versions in the first half are so great as to make this assumption highly improbable. Here again the Gospel according to the Hebrews may help us. Eusebius tells us that the Parable of the Talents stood in this Gospel but

"told of three servants, one who devoured his Lord's substance with harlots and flute girls, one who gained profit manifold, and one who hid his talent; and then how one was accepted, one merely blamed, and one shut up in prison."

Is it not possible that M had a version something like this and that Matthew has conflated Q and M, following M more closely at the beginning and Q at the end? Luke, then, preserves approximately the Q form. Only the very daring, nowadays, venture on speculations in regard to the Gospel according to the Hebrews. [The theory that the Gospel written in Chaldee characters here quoted by Eusebius was not the same as that quoted by Jerome seems to me unproved; but in any case the evidence for the existence of divergent traditions of our Lord's sayings is in no way impaired.] Nevertheless, I would in parenthesi throw out the suggestion that the same Jerusalem tradition that we have postulated as the source M used by Matthew was incorporated in, or in some other way affected the text of, that lost Gospel. At any rate we have here evidence, outside the Synoptic Gospels, for the contention of the previous chapter that there were in circulation divergent versions of exactly those of our Lord's sayings in which the versions in Matthew and Luke differ too widely to be referred to a common written source.

(c) The two versions of the saying about "the strait gate," Lk.i.23-24, cf. Mt.vii.13-14, cannot reasonably be referred to a common source.

Lk.i.23-24. Mt.vii.13-14.
And one said unto him,
Lord, are they few that be saved?
And he said unto them,
Strive to enter in by the narrow door:
for many, I say unto you, shall seek to enter in, and shall not be able.
Enter ye in by the narrow gate:
for wide [is the gate], and broad is the way, that leadeth to destruction, and many there be that enter in thereby.
For narrow [is the gate], and straitened the way, that leadeth unto life, and few there be that find it.

But Luke's version comes in the middle of a section of which the beginning (Mustard Seed and Leaven), i.18-21, the middle, i.28-29, and the end, i.34-35 ("Jerusalem, Jerusalem"), are certainly Q, and of which, as we shall see later, much of the rest is probably Q; Q. But the words ἡ πύλυ ("is the gate") are omitted in Matthew on their second (א Old Lat.) and third (544 Old Lat.) occurrence. If this reading is original, Q had the Lucan saying about "the narrow gate," M had one quite different—the antithesis between the "broad and the narrow ways."   
(The contrast of the Two Ways occurs in the Didache and elsewhere.)    
Matthew has conflated Q and M.
[Prof. Dodd pointed out the relevance of the textual variants to me.]

(d) A fourth example of the same kind may be Lk..32-34, "Fear not, little flock ..." cf., " Lay not up for yourselves treasures on earth. ..." In the first two verses the differences between Matthew and Luke are considerable, but the third verse in each is practically identical, "Where your treasure is there will your heart be also. " Matthew has disconnected this verse from the discourse, "Be not anxious" (, of which it forms the concluding sentence in Luke. Here the combination of variation in order with diversity of wording suggests that Matthew is conflating Q and M—in which case Luke may be presumed to follow Q.

(e) Mk.xi.22-23. Mt.xvii.19-20. Lk.xvii.5-6.
And Jesus answering saith unto them,
Have faith in God.
Verily I say unto you, Whosoever shall say unto this mountain, Be thou taken up and cast into the sea; and shall not doubt in his heart, but shall believe that what he saith cometh to pass;
he shall have it.
Then came the disciples to Jesus apart, and said, Why could not we cast it out?
And he saith unto them,
Because of your little faith:
for verily I say unto you, If ye have faith as a grain of mustard seed, ye shall say unto this mountain, Remove hence to yonder place;
and it shall remove;
and nothing shall be impossible unto you.
And the apostles said unto the Lord,
Increase our faith.
And the Lord said,
If ye have faith as a grain of mustard seed, ye would say unto this sycamine tree, Be thou rooted up, and be thou planted in the sea;
and it would have obeyed you.

Here Matthew agrees with Mark in speaking of the "mountain," with Luke in "the grain of mustard seed." The most natural conclusion would be that Luke gives the saying as it stood in Q, while Matthew, as usual where Mark and Q overlap, conflates the two. In further confirmation of this we note that in Luke the saying occurs immediately after xvii.1-4, which we decided above was from Q; also that Matthew inserts it in a context, derived indeed from Mark, but occurring much earlier (Mk.ix.28), which is the first occasion appropriate for the insertion, from another source, of a saying on faith.

scattered fragments

The observation that Luke in general seems to follow one source at a time and to reproduce it in its original order has proved a valuable clue. But a generalisation of this kind must not be made a fetish. There is always an incalculable element in the working of the human mind; and there is no reason to suppose that in following one source at a time Luke was adhering to a consciously formulated principle. It was merely that he refrained from rearranging or conflating sources where there was no special reason for doing so. But to any rule of average human behaviour there are always some exceptions; and there are several sayings in Luke which there is good reason to assign to Q, although they are not found embedded in a mass of other material from that source.

(a) "No man can serve two masters," Lk.xvi.13.
This is very close in wording to, and is therefore to be referred to Q; but its context in Luke, immediately following the parable of the Unjust Steward, is obviously suggested by the accidental occurrence of the rare word Mammon both in this saying of Q and in the parable.

(b) The saying "Whoso exalteth himself shall be humbled ..." occurs twice in Luke, xiv.11, xviii.14. As it occurs in Mt.xi.12, it probably stood in Q; although, for a short proverbial saying of this kind, there is really no need to postulate a written source at all.

(c) The pair of sayings, Lk.xiv.26-27, "If any one cometh unto me and hateth not his father, etc." and "Whoso beareth not his cross," occur together and in the same order in Mt.x.37-38. Hence, though the wording differs to a certain extent, they are probably to be referred to Q. Here again their present position in Luke seems to be due to their eminent fitness as an introduction to two parables—the Tower Builder and the King making War—, which emphasise the same idea of counting the cost. I would remark that, if they are Q, Luke's harsh-sounding "hateth his father" obviously preserves the wording of the original more closely than Matthew, "He that loveth father, etc., more than me." The saying about carrying the cross occurs also in Mark (viii.34) in a slightly different version—which is copied by both Matthew and Luke in the same context (Mt.xvi.24, Lk.ix.23)—in spite of the fact that they give another (? from Q) version elsewhere, Mt.x.38 = Lk.xiv.27. [The passages are printed in parallel columns in Hawkins, Hor. Syn.2 p. 86.]

(d) The saying about Salt, Lk.xiv.34-35, shows several agreements between Luke's version and that in Mt.v.13 against the version of Mk.ix.50, and also Matthew and Luke agree in omitting it in the context parallel to Mark's, which looks as if both had already extracted it from Q. Luke separates it from the saying about carrying the cross by the Tower Builder and its twin parable. But if, as it stood in Q, it followed immediately after Lk.xiv.27, its meaning would be quite clear. In that connection, "Salt is good, but if the salt have lost its savour ..." would naturally mean, "Disciples are good, but if they have lost the power to carry the cross they cease to be a leaven to the lump of humanity." It would seem to follow that the words "Salt is good," καλὸν τὸ λα, stood in Q as well as in Mark; but that Matthew in the Sermon on the Mount has altered them to "Ye are the salt of the earth" in order to make the interpretation quite clear, the wording of the alteration being suggested to him by that of the M saying "Ye are the light ..." to which he has prefixed it.

(e) The three sayings, Lk.xvi.16-18, are perhaps from Q.  
Q, so far as one can make out, was a collection of the "Wise Sayings" of Christ, comparable to a book like Proverbs or the Pirqe Aboth, with very little attempt at arrangement. And what we have in the passage of Luke is three Separate aphorisms (the Law and the Prophets until John, the passing of the Law, Divorce), the only connection between which is that they are epigrammatic rulings on disputed points connected with the Law. In Matthew they appear in a somewhat modified form and in three quite different contexts, worked into connected discourses. The Lucan version looks on the whole more primitive [Not always, e.g. Lk.xvi.17 ἡ βασιλεία τοῦ θεοῦ εὑαλλελίζεται is a Lucan phrase—an attempt to explain an extremely obscure saying.]; but the modifications of it in Matthew do not go beyond what an editor, who, like Matthew, evidently felt that he was also to some extent an interpreter, might consider legitimate.

(f) Where Matthew has fitted an isolated saying into a new context—whether in Mark or in a larger block of Q—some modification of its wording might be required to make it harmonise with its new context. In such cases, therefore, we must recognise that a greater divergence than elsewhere between the parallels in Matthew and Luke is compatible with derivation from Q. In the light of this reflection we may consider certain sayings found in Mt.xxiv.25-28, 37-41, ingeniously fitted into the "Little Apocalypse" of Mk.i, in such a way as to amplify certain ideas taken over from Mark. All these sayings occur in the discourse Lk.xvii.22-37, but somewhat differently worded. Can we, in spite of the considerable verbal differences, hold that the whole section, Lk.xvii.22-37, which has been described as "the Apocalypse of Q" has been legitimately so named? I think so. For if Q had contained an Apocalypse, Matthew would certainly have conflated it with the Apocalypse of Mark. If it stood in Q at all, Lk.xvii.22-37 must have stood at, or very near, the end. That is an additional reason for supposing the passage stood in Q; for, in view of the absorbing interest of the early Church in the subject of the Parousia, we should naturally expect to discover a quantity of Apocalyptic "Gospel." For that same reason I am inclined to think that Q not only contained, but actually ended with, the Parable of the Pounds, the moral of which—"Occupy till I come"—would be so extraordinarily appropriate to the hopes and circumstances of the time. This presumption is distinctly strengthened by the fact that the Didache, a first-century manual of Christian instruction, ends with an Apocalyptic passage.

There remains to be considered one other saying which occurs in both Matthew and Luke.

Mt.xix.28. Lk.x.29-30.
And Jesus said unto them,
Verily I say unto you, that ye which have followed me, in the regeneration when the Son of man shall sit on the throne of his glory,
ye also shall sit upon twelve thrones, judging the twelve tribes of Israel.
And I appoint unto you a kingdom,
even as my Father appointed unto me,
that ye may eat and drink at my table in my kingdom;
and ye shall sit on thrones judging the twelve tribes of Israel.

Observe that, apart from the words in italics, there are no points of contact between these parallels. No doubt the words found in both are the most striking, but to assume that these alone stood in Q and that all the rest in both Matthew and Luke is "editorial" is a reductio ad absurdum of the theory of a written source, only possible under the distorting influence of an a priori Two Document Hypothesis. Rather, this is a good example of the currency of widely different versions of the same saying; and since neither in Matthew nor Luke is it found in a Q context, we naturally assign the two versions to M and to L.

In the above survey no notice has been taken of short passages, evidently inserted in order to break the monotony of discourses following one another with no obvious connection. Phrases like "He said to the disciples," "He said to the multitude," are evidently merely inferences from the context of the sayings, made explicit in words to improve the literary form. Probably the same thing applies to some longer phrases like Lk.xiv.25, "There were journeying with him great crowds; and he turned and said unto them," or the triple "He said to the guests," "He said to the host," "One sitting at meat said to him;" Lk.xiv.7,12,15. The occurrence of stylistic improvements of this kind has prejudiced many critics against Luke's version of Q as compared with that of Matthew. But these are obvious and superficial adornments easily separable from the actual saying. And when one studies the actual Logia of Christ, the advantage in respect of accuracy of preservation of the original form is, I believe, more often with Luke. "Blessed are ye poor, ye that hunger" is surely more original than "poor in spirit" or "hunger after righteousness"; "hateth father and mother" is more original than "loveth father and mother more than me."

omissions from Q

The question must now be raised, Have any sections of Q which have been completely ignored by Matthew been preserved in Luke and vice versa? To this I am inclined to answer, Very few. Before I had disentangled myself from the Two Document Hypothesis I used to suppose that the more Judaistic sayings in Matthew were probably in Q, but omitted by Luke. But reasons have been given above for assigning these to M. Of his other source, Mark, Matthew omits very little, so the probability is that he would omit very little of Q, unless to substitute for what was omitted something which he regarded as a superior version of the same thing. Nevertheless, as he seems completely to omit a few items in Mark, which Luke retains, the same thing has probably happened in regard to Q. To identify these, we return to the examination of those passages in Luke's great blocks of Q that have no sort of equivalent in Matthew. But we cannot assume that all of them stood in Q. We have seen above that Lk.xiv.34-35 probably stood in Q followed immediately by xiv.26-27. If so, then we have evidence that Luke (or the compiler of Proto-Luke) sometimes allowed himself to interpolate into a Q section highly appropriate matter (e.g. xiv.28-33) from L. That shows that, on occasion, he would break away from his general rule of following one source at a time. Accordingly we cannot be sure that Lk.ix.61-62, "No man putting his hand to the plough," and x.18-20, "I saw Satan fall," .47-50, .54-57, have not similarly been interpolated from L into what looked like highly appropriate contexts of Q. In the Oxford Studies I endeavoured to find reasons why, even if they should be in Q, Matthew should have omitted them. I now feel less confidence in their validity. To feel confidently that Matthew omitted any notable saying at all in Q one must see clearly that the saying would lend itself to an interpretation by the faithful, which he definitely disliked. Two such passages there are—Lk.ix.51-56, the Samaritan village, and xvii.20-21, "The Kingdom of God cometh not with observation ... the Kingdom of God is within you." Both of these stand at the head of a block of Q material, and both are passages which Matthew would have had good reason to omit. The one involves a rebuke to the Apostles, and Matthew elsewhere tends to tone down or omit such. [Hawkins, Hor. Syn.2 p. 121.]
The other suggested a view of the Kingdom which Matthew, who more than any other evangelist emphasises the objective catastrophic side of the Apocalyptic hope, believed to be incorrect.

My view on the question of whether Luke omitted any substantial amount of Q has been modified by three new considerations,
(a) The evidence submitted previously (cf. p. 175 ff.) that the long section did not stand in the copy of Mark used by Luke. If one supposed that Luke was capable of leaving out over 70 continuous verses in one source, he would be capable of making drastic excision in another. Apart, however, from this one passage, Luke's omissions—as distinct from substitution of parallel versions—of material found in Mark are on a very small scale,
(b) If I am right in supposing that Q and L were combined into one document by Luke himself before he came across Mark, he would not at that earlier date be embarrassed by the problem of getting all his material into a roll of manageable size, as probably was the case at the later stage. Hence he had a motive for omission at this later stage, when he expanded Proto-Luke with extracts from Mark (which, moreover, he seems to have regarded as a source of subordinate authority) which would have been inoperative at the time when he combined Q and L.
(c) If there is a shadow of ground for the guess that Q was the old Gospel of the pro-Gentile church of Antioch, it would not have contained Judaistic passages which Luke would have wanted to excise.
[This argument might appear to prove too much, for the passages do occur in the probably Antiochene Gospel of Matthew. But their presence in Matthew may well be due to their occurrence in a document already too ancient to be ignored. Matthew prefers to counteract, e.g. the prohibition to preach to Gentiles and Samaritans, x.5, is revoked by the command to preach to all nations, xxviii.19.]
Accordingly the probability is high that the passages of Luke that we can identify as Q represent that document, not only approximately in its original order, but very nearly in its original extent. And seeing that Matthew's method of rearranging sources led necessarily to considerable verbal modification, it is probable that, allowance being made for a slight polishing of the Greek, the form in which the sayings appear in Luke is also on the whole more original.

the reconstructed Q

For the convenience of the reader I append a list of the passages I should assign to Q. Brackets signify considerable doubt:
Lk.iii.2-9, (10-14), 16-17, 21-22; iv.l-16a; vi.20-vii.10; vii.18-35; ix.(51-56), 57-60, (61-62); x.2-16, (17-20), 21-24; xi.9-52; .1a-12, 22-59; i.18-35; xiv.11, 26-27, 34-35; xvi.13, 16-18; xvii.1-6, 20-37; xix.11-27.
Unbracketed verses = 272.

As thus reconstructed, Q is a document the purpose and character of which are perfectly intelligible. It is comparable to an Old Testament prophetic book like Jeremiah, consisting principally of discourse, but with an occasional narrative to explain some piece of teaching. The Baptism and Temptation are described because the author regarded these as virtually the "call" to the Messianic office. The author would regard them, like the "call" of the Prophet so often given in the Old Testament, as of great apologetic value as evidence of the Divine authorisation of our Lord's mission. The relatively large amount of space given to John the Baptist, and the emphasis on his relations with our Lord, suggest that Q was composed at a time and place where the prestige of John was very considerable. Q. [The point is elaborated in my Essay on "The Literary Evolution of the Gospel" in Oxford Studies, p. 210 ff.] In Q John's testimony to Christ is appealed to, because among those for whom it was written "all held John to be a prophet."
In Mark the apologetic motive for mentioning John is that he fulfilled the prophecy of the fore­runner; that is to say, it is not John's personal prestige which is appealed to, but the fact that his coming at all was part of that "programme," so to speak, of events, anciently foretold and in the career of our Lord recently fulfilled, which was the main plank of early Christian apologetic. It is the difference between the point of view of Rome c. AD 65 and Syria (where John's name and following were great) fifteen or twenty years earlier.

The absence in the Passion story of any substantial agreements of Matthew and Luke against Mark, in the view of most scholars, compels us to conclude that Q contained no account of the Passion.
We must ask, Why?
I think the answer must be sought in two directions.

(1) The Passion and its redemptive significance could readily be taught in oral tradition.
But ethical teaching implies detailed instruction, which sooner or later necessitates a written document. Such a document is found in the Didache, which obviously presupposes a general knowledge of the central facts of the Christian story. Similarly Q was probably written to supplement an oral tradition.

(2) Of less weight is the consideration that, while to Paul the centre of the Gospel was the Cross of Christ, to the other Apostles it was His Second Coming. Peter's speeches in the Acts show that to them, as to other Jews, the Crucifixion was a difficulty. The Resurrection had cancelled it, so to speak. The Prophets had foretold it, and this showed that it was somehow part of the Divine Plan; but it was still one of those calamities, which darken men's understanding of His Purpose, rather than the one act that has unveiled the mystery.