In this left column is Chapter VII of A Study of Origins, The Manuscript Tradition, Sources, Authorship, & Dates. By B.H. Streeter (1924). This chapter is called "The Fundamental Solution" and outlines Streeter's theories on the Q document. This chapter is reproduced entirely without permission, using the text as prepared by Paul Ingram in 2004 for Some of the formatting isn't intact, and I've not included the chapter synopsis or the additional notes. In this right column are the notes I (Luke Burrage) made while reading this chapter, which I then expanded so they could be understood by someone not-me (December 2012). Please read the introduction to this commentary before continuing.



historians and their sources

THE mechanical invention of printing has reacted on the methods and conventions of authorship itself in more ways than we are apt to imagine.
When books were copied by hand, COPYRIGHT had no commercial value;
no kind of injury could be done either to author or publisher by any one who made and sold copies.
But in the setting up of a printed book capital is sunk; work has been done and a risk has been incurred, in return for which it is reasonable that the publisher should enjoy such legal protection against unauthorised reproductions as will enable him to derive a fair profit.
Again, in antiquity an author, unless, as most commonly happened, he was a man of inherited wealth, lived on the bounty of some noble patron of letters.
Printing has enabled a modern Horace to live, not by flattering a Maecenas, but on the profits of his books.
For both these reasons the conception of property in literature has arisen.

Hence there has gradually grown up an entirely different convention as to the manner and conditions under which it is legitimate to make use of what others have written.

This opening section clearly states that considerations of authorship are different in the modern world compared ancient world. Despite claiming differences in technique and method, Streeter decides to immediatly impose the modern understanding of "historian" onto ancient writers. This is premature. Instead, a critic or historian should first ask the question: "What genre am I reading?" And, importantly, "Has the conception of this genre changed since the ancient times?"

The change is one that affects historical more than any other kind of writing. And so it begins: let's assume the study of the Synoptic Problem starts with treating the Gospels as history.
Whenever a historian is not an actual eyewitness of the events he records, or the first to write down a living tradition, he is bound to depend to a large extent on the works of previous historians.
The modern convention requires that when this happens he shall either quote the exact words of his authority or entirely rewrite the whole story with some general indication of the source from which it comes.
Here again the printing press has made a difference.
It has facilitated the development of inverted commas, footnotes for reference, and other such devices unknown to the scribes of Classical Antiquity, which make it easy for an author to indicate without clumsy circumlocutions the exact extent of his debt to predecessors.
The conventions of every art are determined by what is mechanically possible;
it is not, therefore, surprising that these inventions have reacted on actual methods of composition employed by the modern author in so far as these entail a use of previous writers.
In antiquity, however, and in the Middle Ages, only the writings of a few outstanding men like Thucydides are wholly original;
more commonly the historian pursued what we should call a method of "scissors and paste."
Here Streeter states that the ancient historian used this "scissors and paste" method, and the rest of this chapter pretty much argues that the Gospel writers were not using this technique. At all.
Without any acknowledgement, he will copy page after page from his source, omitting passages that for his purpose seem irrelevant, adding here and there material from some other authority.
What he copies he frequently gives almost word for word, but he will often abridge, and occasionally paraphrase, in order to elucidate some difficulty or to preclude what he would regard as a false impression which the language of the original might convey.

Here we see that a "historian" had no problem changing what he wanted about his source. The same is true, or more so, for an author not primarily concerned with historical accuracy.

This kind of editorial adaptation of earlier sources can be traced in all the historical books of the Old Testament, and in many classical and mediaeval writers.
I would call attention to one example in each of these fields where the survival of the original sources, the nature of the subject matter, and the accessibility to the ordinary reader of the relevant literature, combine to make a study of ancient methods and their bearing on our present investigation both exceptionally profitable and relatively easy.
Turn to the books of Chronicles in a reference Bible.
It is clear that, from 1 Chron.x. on, almost everything is an abridgement, with trifling modifications, of the narrative in the books of Samuel and Kings.
[These and other O.T. analogies may most conveniently be studied in Deuterographs by R. B. Girdlestone (Oxford, 1894), where the relevant passages are printed in parallel columns with the differences indicated by italics.].
Consult the appendices dealing with the earliest accounts of St. Francis of Assisi, either in the Life by Sabatier or in that by Father Cuthbert, and you will see a "synoptic problem," explicable on these lines.
Lastly, compare the fragment of the Greek historian Ephorus lately discovered at Oxyrhynchus with the account of the same events in Diodorus [Cf. Oxyrhynchus Papyri, i. p. 102 ff.], and you will find an illustration in a Greek writer practically contemporary with the authors of the Gospels of Matthew and Luke.
You will notice, and the analogy is important, that the Greek writer, in contrast to the Hebrew, makes many more little alterations of phrase so as to leave upon all that he has incorporated the impress of his own style.

the priority of mark

This is all great... but guess how long it will be before Streeter gives the reason that Q might be needed?

Really, wait and see how much reasoning he gives before even telling us WHY he is setting out all this introductory material. And, if you're anything like me, you'll laugh at the reason.

Such being the almost universal method of ancient historians, whether Jewish or Greek, it is natural to ask whether the remarkable resemblance between the first three Gospels, which has caused the name Synoptic to be applied to them, would not be most easily explained on the hypothesis that they incorporate earlier documents.

Immediately Streeter introduces new material, before seeing if the problem can be solved with the current data. Why not first see if the documents themselves contain the information we need to explain the similarities?

A century of discussion has resulted in a consensus of scholars that this is the case, and that the authors of the First and Third Gospels made use either of our Mark, or of a document all but identical with Mark.
The former and the simpler of these alternatives, viz. that they used our Mark, is the one which I hope in the course of this and the following chapters to establish beyond reasonable doubt.

I've no problem with the Priority of Mark, but it turns out that this chapter is more about the need for Q.

The attempt has recently been made to revive the solution, first put forward by Augustine (cf. p. 10), who styles Mark a kind of abridger and lackey of Matthew, "Tanquam breviator et pedesequus ejus."
But Augustine did not possess a Synopsis of the Greek text conveniently printed in parallel columns.
Otherwise a person of his intelligence could not have failed to perceive that, where the two Gospels are parallel, it is usually Matthew, and not Mark, who does the abbreviation.
For example, the number of words employed by Mark to tell the stories of the Gadarene Demoniac, Jairus' Daughter, and the Feeding of the Five Thousand are respectively 325, 374 and 235;
Matthew contrives to tell them in 136, 135 and 157 words.
[Cf. J. C. Hawkins, Horae Synopticae2, p. 159 (Oxford, 1909).]

Now there is nothing antecedently improbable in the idea that for certain purposes an abbreviated version of the Gospel might be desired; but only a lunatic would leave out Matthew's account of the Infancy, the Sermon on the Mount, and practically all the parables, in order to get room for purely verbal expansion of what was retained.
The phrase "only a lunatic" is indicative of Streeter's lack of scholarly rigor. Instead of trying to image WHY someone might abridge Matthew in this manner, he declares only a lunatic would do it. Maybe Mark had good reasons to do what he did. As an ancient author, we'll never be sure of his reasons for doing anything... unless we find evidence in the text itself. As it happens, we can find evidence that Matthew used Mark, not the other way round, but that doesn't mean that just because we don't know, or don't agree with, or don't understand an author's creative decision... it doesn't mean we can call them a lunatic and move on!

This point, and Streeter's disregard for possible motives of ancient authors', is going to be crucial later on.

On the other hand, if we suppose Mark to be the older document, the verbal compression and omission of minor detail seen in the parallels in Matthew has an obvious purpose, in that it gives more room for the introduction of a mass of highly important teaching material not found in Mark.

Further advance, however, towards a satisfactory solution of the Synoptic Problem has been, in my opinion, retarded by the tacit assumption of scholars that, if Matthew and Luke both used Mark, they must have used it in the same way.

To Professor Burkitt, I believe, belongs the credit of first protesting against this assumption:

Matthew is a fresh edition of Mark, revised, rearranged, and enriched with new material; ...
Luke is a new historical work made by combining parts of Mark with parts of other documents.

[The Earliest Sources for the Life of Jesus (Constable, 1922).]

Here Streeter acknowledge's different approaches by different ancient authors. However, he's fine with using modern understanding of labels and genre's to describe ancient works. Did ancient author's make "new editions" of previous works? Did the author's think they were writing history, or did they produce their works for liturgical, teaching, or devotional purposes?

The distinction thus stated by Burkitt, I shall endeavour to justify and to elaborate in a new direction in Chap. VIII. 
I conceive it to be one of fundamental importance in any attempt to estimate the value of the Third Gospel as an historical authority for the life of Christ.

Partly in order to clear the way for a more thorough investigation of this point, partly because this book is written for others besides students of theology, I will now present a summary statement of the main facts and considerations which show the dependence of Matthew and Luke upon Mark.
Familiar as these are to scholars, they are frequently conceived of in a way that tends to obscure some of the remoter issues dependent on them.
They can most conveniently be presented under five main heads.


The authentic text of Mark contains 661 verses.
Matthew reproduced the substance of over 600 of these.
Mark's style is diffuse, Matthew's succinct; so that in adapting Mark's language Matthew compresses so much that the 600 odd verses taken from Mark supply rather less than half the material contained in the 1068 verses of the longer Gospel.
Yet, in spite of this abbreviation, it is found that Matthew employs 51% of the actual words used by Mark.
[Oxford Studies in the Synoptic Problem, ed. W. Sanday, pp. 85 ff. (Clarendon Press, 1911).]

The relation between Luke and Mark cannot be stated in this precise statistical way--for two reasons.
First, in his account of the Last Supper and Passion, Luke appears to be CONFLATING--
to use the convenient technical term for the mixing of two sources--
the Marcan story with a parallel version derived from another source, and he does this in a way which often makes it very hard to decide in regard to certain verses whether Luke's version is a paraphrase of Mark or is derived from his other source.

This is an early example of circular reasoning by Streeter. That is: in laying out the evidence for Q (this data will be used to support Q later), he uses the Q source, this "other source", to explain Luke's differences compared to Mark.

Maybe this is why Streeter leaves it so late before explaining why Q is needed, because by then he'll have hoped you've forgotten that Q is needed to explain why Q is needed.

Indeed there are only some 24 verses (cf. p. 216 f.) in this part of Luke's Gospel which can be identified with practical certainty as derived from Mark, though it would be hazardous to limit Luke's debt to Mark to these 24.
Secondly, there are also, outside the Passion story, a number of cases where Luke appears deliberately to substitute a non-Marcan for the Marcan version of a story or piece of teaching.
Thus the Rejection at Nazareth, the Call of Peter, the parable of the Mustard Seed, the Beelzebub Controversy, the Great Commandment, the Anointing, and several less important items are given by Luke in a version substantially different from that in Mark, and always, it is important to notice, in a context quite other from that in which they appear in Mark.

Instead of entertaining the notion that Luke had no problem simply rewriting and moving a story from Mark, Streeter only considers that Luke must have got it from a prior source.

Another striking feature in Luke's relation to Mark is his GREAT OMISSION, so called, of a continuous section of 74 verses,
Besides this he omits several shorter sections, which added together amount to 56 verses.

If we leave out of account all passages where there is reason to suspect that Luke has used a non-Marcan source, it appears on an approximate estimate that Luke has reproduced about 350 verses (i.e. just over one half of Mark).
Translation: if we call material different but similar to that found in Mark "material from Q" it will give us more data to prove Q later.
When following Mark, Luke alters the wording of his original a trifle more than Matthew does;
on the other hand he retains many details which Matthew omits, and he does not compress the language quite so much.
The result is that on an average Luke retains 53% of the actual words of Mark, that is, a very slightly higher proportion than does Matthew.

From these various figures it appears that, while Matthew omits less than 10% of the subject matter of Mark, Luke omits more than 45%, but for much of this he substitutes similar matter from another source.

Is there a criteria for deciding what Luke is simply rewriting based on a reading of Mark as opposed to something he is copying exactly from another source? This is a tricky question because, two paragraphs ago, Streeter admits that Luke alters the wording of Mark when he copies it. Where is the dividing line between "copies from Mark but with modifications" and "not close enough to Mark, so he must have copied it from a non-Markan source"?
Each of them omits numerous points of detail and several complete sections of Mark, which the other reproduces;
but sometimes they both concur in making the same omission.
The student who desires to get a clear grasp of the phenomena would do well to prepare for himself, by the aid of the lists in the Additional Note A at the end of this chapter, a marked copy of the second Gospel, indicating by brackets of four different shapes or colours--

(Minor annoyance of mine: people who say they are doing historical criticism, but use terms like "the second Gospel" to refer to Mark's Gospel, in the context of talking about the priority of Mark's Gospel. If Mark's Gospel came first, why identify it by its canonical position in the New Testament?)
  1. passages peculiar to Mark;
  2. those reproduced by Luke, but not by Matthew;
  3. those reproduced by Matthew, but not by Luke;
  4. those which Luke omits, but for which in another context he substitutes a parallel version.
Again, no criteria for deciding what Luke changed about Mark or what he got from a different source. Appeals to Q can't be used as an explanation in this, because we're currently looking for a reason to posit Q. Streeter leaves this question, not only unanswered, but unacknowledged.


Let the student take a few typical incidents that occur in all three Synoptists--
I would suggest Mk.ii.13-17 and xi.27-33 to begin with--
and, having procured a Synopsis of the Gospels, underline in red words found in all three, in blue words found in Mark and Matthew, in yellow words found in Mark and Luke.
If this is done throughout the Gospels it will appear that a proportion varying from 30% to over 60% of the words in Mark are underlined in red, while a large number of the remainder are marked either blue or yellow.
[The happy possessor of W. G. Bushbrooke's magnificent Synopticon will find the work done for him by the use of different types and colours.
Of Greek Synopses on a smaller scale, the most conveniently arranged are A. Huck's Synopse (Mohr, Tubingen) and Burton and Goodspeed's Harmony of the Synoptic Gospels (Universities of Chicago and Cambridge).
For those who have little or no knowledge of Greek an admirably arranged Synopsis based on the English of the Revised Version is The Synoptic Gospels by J. M. Thompson (Clarendon Press).]

What is still more significant, if the collocation of words and the structure of sentences in Matthew and Luke be examined, it will be found that, while one or both of them are constantly in close agreement with Mark, they never (except as stated p. 179 ff.) support one another against Mark.
This is clear evidence of the greater originality of the Marcan version, and is exactly what we should expect to find if Matthew and Luke were independently reproducing Mark, adapting his language to their own individual style.

The words "never (except...)" is yet another example of special pleading. Streeter dedicates a section of this chapter, and a whole chapter later in The Four Gospels, repeatedly demonstrating that the Q document is able to be falsified by agreements between Matthew and Luke in their changes of Mark's material, and then by repeatedly denying these agreements either existed in the original texts, or if they did, can't be used as evidence that Luke knew Matthew. He literally argues for an unfalsifiable theory, discarding any evidence that can be brought against it.

Like I said, it's special pleading all the way down.


The order of incidents in Mark is clearly the more original;
for wherever Matthew departs from Mark's order Luke supports Mark, and whenever Luke departs from Mark, Matthew agrees with Mark.
The section Mk.iii.31-35 alone occurs in a different context in each gospel;
and there is no case where Matthew and Luke agree together against Mark in a point of arrangement.

A curious fact, of which an explanation is suggested later, p. 274, is that, while in the latter half of his Gospel (chap. xiv. to the end) Matthew adheres strictly to the order of Mark ( to end), he makes considerable rearrangements in the first half.
[A convenient chart showing Matthew's rearrangements of Mark's order is given in the Commentary on Matthew by W. C. Allen in the "International Critical" series (T. & T. Clark, 1907), p. xiv.
The discussion of the relation of Matthew and Mark in this work, pp. i-xl, is the most valuable known to me;
I cannot, however, accept the theory of Matthew's second main source, p. xli ff.]
Luke, however, though he omits far more of Mark than does Matthew, hardly ever departs from Mark's order, and only in trifling ways.
[These are enumerated and discussed in Oxford Studies, p. 88 ff.]

On the other hand, wherever Luke substitutes for an item in Mark a parallel version from another source, he always gives it in a different context from the item in Mark which it replaces.
This, as we shall see later, is a fact of very great significance for the determination of the source of Luke's non-Marcan material.

Is this the criteria I wanted before? I understand that changing some of Mark's material more than other parts, and positioning them out of order, makes these passages different. It raises a question as to why Luke might have done so, and we might find the answer to this question by studying the content of those passages. However, is this uniquely handled material different enough to propose a previously-unknown source document to explain it?

We note, then, that in regard to

  1. items of subject matter,
  2. actual words used,
  3. relative order of sections,

Mark is in general supported by both Matthew and Luke, and in most cases where they do not both support him they do so alternately, and they practically never agree together against Mark.
This is only explicable if they followed an authority, which in content, in wording, and in arrangement was all but identical with Mark.

Note that I actually agree with Streeter's conclusion about the priority of Mark. I just don't think his methods are solid, and it seems he is using them because, by applying the same floppy logic later in this chapter, he can "prove" an otherwise unsupported hypothesis.


A close study of the actual language of parallel passages in the Gospels shows that there is a constant tendency in Matthew and Luke--
showing itself in minute alterations, sometimes by one, sometimes by the other, and often by both--
to improve upon and refine Mark's version. 
This confirms the conclusion, to which the facts already mentioned point, that the Marcan form is the more primitive. 
Of these small alterations many have a reverential motive. 
Thus in Mark, Jesus is only once addressed as "Lord" (κύριε), and that by one not a Jew (the Syrophoenician). 
He is regularly saluted as Rabbi, or by its Greek equivalent
διδάσκαλε  (Teacher). 
In Matthew
κύριε occurs 19 times;
in Luke
κύριε occurs 16, ἐπιστάτα (Master) 6 times. 
In the same spirit certain phrases that might cause offence or suggest difficulties are toned down or excised. 

Thus Mark's

he could, do there no mighty work (vi.5)

becomes in Matthew (i.58)

he did not many mighty works;

while Luke omits the limitation altogether. 

Why callest thou me good? (Mk.x.18)

reads in Matthew (xix.17)

Why askest thou me concerning the good? 

Here Streeter finally touches on some solid reasons why Luke or Matthew wanted to change the content or meaning of Mark's Gospel when incorporating material into their own "histories". And yet, again, he provides no way to distinguish between this kind of changed material, and material that is so different from Mark that it "obviously" comes from a different source. Luke and Matthew are fine with changing Mark's material... until Streeter considers them historians, and as historians, they must work from a previous source... which in turn mustn't be changed... until one of them is fine with changing it...

Much more frequently, however, the changes merely result in stylistic or grammatical improvements, without altering the sense.

But the difference between the style of Mark and of the other two is not merely that they both write better Greek.

It is the difference that always exists between the spoken and the written language.
Mark reads like a shorthand account of a story by an impromptu speaker--
with all the repetitions, redundancies, and digressions that are characteristic of living speech.
And it seems to me most probable that his Gospel, like Paul's Epistles, was taken down from rapid dictation by word of mouth.
Ironically, in this section Streeter gives a very good counter argument that Mark must have been a lunatic if he used Matthew as a source but left out certain material (material that Streeter considers important):

Imagine Mark, over the course of a few dozen church meetings, hearing many readings from Matthew's Gospel. Later he recounts what he remembers to a scribe. If he missed certain readings, they'd be left out of his recounting of the story of Jesus. Mark would create his own order for material he doesn't remember so clearly, from longer ago (this would account for the differing order compared to shared material in Mark and Matthew at the beginning of Matthew, but that by the end matches up far better). Mark missed some of the key parables, and the Sermon on the Mount. He's played whole story over in his mind a few times, and knows how he's going to connect the material he remembers together, and also thinks up new facts and material to include.

The result: the Gospel of Mark, a Gospel that omits material from its source (Matthew) but also expands on what it does include.

I don't think this view of Mark's writing is correct, not even slightly. It does, however, show how someone who wasn't a lunatic could do something that Streeter claimed that only a lunatic would do... and I've used the later words and observations of Streeter himself as the basis of this very speculation!

The Mark to whom tradition ascribes the composition of the Gospel was a Jerusalem Jew, of the middle class
[His mother had a house large enough to be a meeting-place for the church, and kept at least one slave girl (Acts .12 f.), and his cousin Barnabas had some property.];
he could speak Greek fluently, but writing in an acquired language is another matter.
I'm not going to bother addressing this claim to church tradition, nor Streeter's use of the name "Mark" as anything more than a handy shortcut to refer to an otherwise unknown author.
Matthew and Luke use the more succinct and carefully chosen language of one who writes and then revises an article for publication.
This partly explains the tendency to abbreviate already spoken of, which is especially noticeable in Matthew.
Sometimes this leads to the omission by one or both of the later writers of interesting and picturesque details, such as

in the stern ... on a cushion (Mk.iv.38),


they had not in the boat with them more than one loaf (Mk.viii.14).

Usually, however, it is only the repetitions and redundancies so characteristic of Mark's style that are jettisoned.
Sir John Hawkins [Hor. Syn2 p. 125] collects over 100 instances of "enlargements of the narrative, which add nothing to the information conveyed by it, because they are expressed again, or are directly involved in the context," which he calls "context-supplements."
Matthew omitted the majority of these and Luke also omitted a large number; though Luke sometimes omits where Matthew retains, as well as vice versa.
Again, Mark is very fond of "duplicate expressions" such as "Evening coming on, when the sun set" (i.32).
In these cases one or other of the later Evangelists usually abbreviates by leaving out one member of the pair;
and not infrequently it happens that Matthew retains one and Luke the other.
Thus in the above example Matthew writes "evening coming on," Luke "the sun having set."

Matthew and Luke regularly emend awkward or ungrammatical sentences;
sometimes they substitute the usual Greek word for a Latinism;
and there are two cases where they give the literary equivalent of Greek words, which Phrynichus the grammarian expressly tells us belonged to vulgar speech.
Lastly, there are eight instances in which Mark preserves the original Aramaic words used by our Lord.
Of these Luke has none, while Matthew retains only one, the name Golgotha (xxvii.33);
though he substitutes for the Marcan wording of the cry from the Cross, "Eloi, Eloi ..." the Hebrew equivalent " Eli, Eli ..." as it reads in the Psalm (Mk.xv.34 = Mt.xxvii.46 = Ps.x.1).

The examples adduced above are merely a sample given to illustrate the general character of the argument.
But it is an argument essentially cumulative in character.
Its full force can only be realised by one who will take the trouble to go carefully through the immense mass of details which Sir John Hawkins has collected, analysed and tabulated, pp. 114-153 of his classic Horae Synopticae.
How any one who has worked through those pages with a Synopsis of the Greek text can retain the slightest doubt of the original and primitive character of Mark I am unable to comprehend.

But since there are, from time to time, ingenious persons who rush into print with theories to the contrary, I can only suppose, either that they have not been at the pains to do this, or else that--
like some of the highly cultivated people who think Bacon wrote Shakespeare, or that the British are the Lost Ten Tribes--
they have eccentric views of what constitutes evidence.

As you can tell from this commentary, I'm starting to think that Q and Proto-Luke are ingenious but as unfounded in evidence as these same theories.


An examination of the way in which the Marcan and non-Marcan material is distributed throughout the Gospels of Matthew and Luke respectively is illuminating.
The facts seem only explicable on the theory that each author had before him the Marcan material already embodied in one single document; and that, faced with the problem how to combine this with material from other sources, each solved it in his own way --
the plan adopted by each of them being simple and straightforward, but quite different from that chosen by the other.

Again, without even bothering to explain a reason to think these other sources might exist, Streeter presumes they do exist. And before looking at documents for which we have evidence (as in: Luke used Matthew's Gospel as a source), Streeter decides that each of them had a hypothetical and previously unknown document.

And, even as admitting they had different reasons and methods in their treatment of a previous source, he presumes they were both aiming for the same thing: a history book. If both were writing history, either one is a reliable historian, or the other is a reliable historian. It's disingenuous to say you can have it both ways.

Certain elements in the non-Marcan matter clearly owe their position in the Gospels to the nature of their contents.
For example, the two first chapters of Luke, with their account of the Birth and Infancy of Christ, differ so much in style and character from the rest of the Gospel that they are almost certainly to be referred to a separate source, whether written or oral we need not now discuss;
and the same remark applies to the first two chapters of Matthew.
Obviously, however, these stories, whencesoever derived, could only stand at the beginning of a Gospel.
Similarly the additional details, which Matthew and Luke give in their accounts of the Temptation and the Passion, could only have been inserted at the beginning and at the end of their Gospels.

But the greater part of the non-Marcan matter consists of parables or sayings, which do not obviously date themselves as belonging to any particular time in the public ministry.
It would appear that the Evangelists had very little to guide them as to the exact historical occasion to which any particular item should be assigned.
The phrase "exact historical occasion" is very slippery.

It seems that this is the only thing Streeter is willing to say might not be historically accurate... and yet doesn't understand that, for a historian, this would be the most important element. Time and time again, if the words of Jesus are removed from a (double tradition) story, the story has no reason to exist. The "occasions" of Jesus's story, where he went and what he did, are not important, and only exist to frame the teaching. You can't put words together with narrative and claim them both as history when, from all other evidence, neither of them have historical basis, nor the need to to posit a historical basis.

A frame without a picture is not evidence that it ever contained a picture, just as a picture without a frame is not evidence that it once had a particular frame. The only frame that we know once must have existed (in this case) is not that a guy called Jesus existed, but that this teaching material was used by Matthew and Luke (or by their leaders) in church, either before or after they were written in their Gospels.

That, at any rate, seems to be the only explanation of the curious fact (to which my attention was drawn by Sir John Hawkins) that, subsequent to the Temptation story, there is not a single case in which Matthew and Luke agree in inserting a piece of Q material (the meaning of the symbol Q will appear later) into the same context of Mark.
Q! Well done for bringing up Q in support of Q!
The way, then, in which materials derived from the Marcan and from non-Marcan sources are combined must have been determined mainly by literary considerations, and very little, if at all, by extrinsic historical information.

Finally Streeter admits that literary concerns might guide the writing of the Gospels... but only in the case of the context of otherwise unquestionably historical events! And how do we know they are historical? Because Luke and Matthew hypothetically read about them in a hypothetical scroll!

Excuse my sarcasm.

The student who wishes to get a thorough grasp of the facts is advised to mark off in blue brackets--
in a New Testament, not in a Synopsis of the Gospels--
all passages of Matthew and Luke which appear to be derived from Mark.
For this purpose the list of parallels in Additional Note B will be of assistance.
He will then see clearly the difference in the methods adopted by Matthew and by Luke.

Matthew's method is to make Mark the framework into which non-Marcan matter is to be fitted, on the principle of joining like to like.
That is to say, whenever he finds in a non-Marcan source teaching that would elaborate or illustrate a saying or incident in Mark, he inserts that particular piece of non-Marcan matter into that particular context in the Marcan story.
Sometimes he will insert a single non-Marcan verse so as most appropriately to illustrate a context of Mark, e.g. the saying about faith (Mt.xvii.20), or about the Apostles sitting on twelve thrones (Mt.xix.28).

The idea that Matthew is chopping up another source document and inserting it where appropriate is based on the theory that Matthew had another source document. This arrangement of material from a hypothetical document can in no way be used as evidence for that hypothetical document.

If, on the other hand, Matthew developed this extra material himself (or was recording new teaching of the leader of his church), and this teaching had been developed in the context of readings from Mark's Gospel in church services, the logical place for this teaching to be placed in the new Gospel is alongside the same material from Mark.

Without even considering another explanation for the same arrangement of additional material, Streeter skips directly to a circular argument for a hypothetical document.

Sometimes he expands a piece of teaching in Mark by the addition of a few verses from another source on the same subject;
e.g. the non-Marcan saying on divorce, Mt.xix.10-12, is appropriately fitted on to Marcan discussions of the same theme.
So the Marcan saying, repeated in Mt.xix.30,

The first shall be last and the last first,

suggests to him the addition in that particular context of the parable of the Labourers in the Vineyard which points the same moral.
Similarly the moral of the Marcan parable of the Wicked Husbandmen, Mt.xxi.33 ff. (which is directed against the Jewish authorities), is reinforced by the addition immediately before and after it of the anti-Pharisaic parables of the Two Sons and the Marriage Feast.

Examples of this kind of adaptation of non-Marcan matter to a Marcan context could be indefinitely multiplied.

But it is worthwhile to call special attention to the bearing of this process on the longer discourses in Matthew.
All of them are clear cases of AGGLOMERATION, that is, of the building up of sayings originally dispersed so as to form great blocks.
Four times, starting with a short discourse in Mark as a nucleus, Matthew expands it by means of non-Marcan additions into a long sermon. 
It is worthwhile to call special attention to these longer discourses in Matthew. If, as proposed above, Matthew was recording new teaching material that used Mark's Gospel as the starting point, the most natural place to record this new teaching would be, obviously, by having Mark's material lead directly into Matthew's material.
Thus the 7 verses of Mark's sending out of the Twelve ( ff.) become the 42 verses of Mt.x. 
The three parables of Mk.iv are made the basis of the seven parable chapter, Mt.i.--one only being different. 
The 12 verses, Mk.ix.33-37, 42-48, are elaborated into a discourse of 35 verses in Mt. xviii. 
The LITTLE APOCOLYPSE (Mk.i.) is expanded, not only by the addition of a number of apocalyptic sayings (apparently from Q), but also by having appended to it three parables of Judgement, Mt.xxv. 
Streeter can't help hinting at the importance of Q, even before he's shown why Q is going to be so important!
To some extent analogous is the way in which the Sermon on the Mount, far the longest and most important block of non-Marcan matter, is connected with the Marcan framework. 
It is inserted in such a way as to lead up, and thus give point, to the Marcan saying,

And they were astonished at his teaching:
for he taught them as one having authority,
and not as the scribes.
Cf. Mk.i.22; Mt.vii.29.  

That the Sermon on the Mount is itself an agglomeration of materials originally separate will be shown later (p. 249 ff.).

Translation: "Meanwhile, keep in mind that Q existing is going to come in really handy for proving that Q existed."
Luke's method is quite different and much simpler. 
There are half-a-dozen or so odd verses scattered up and down the Gospel in regard to which it is disputable whether or not they are derived from Mark. 
Translations: "When there is evidence against our hypothesis, we can ignore it for now, and use special pleading to show how it won't effect our reasoning later."

This is another case of the "never (except...)" concept I noted earlier. What criteria is used to decide either way on a disputable verse? Streeter typically takes them on a case by case basis, and, using a variety of different techniques and arguments, proclaims each one of them "non-problematic" for his (eventually) stated hypothesis.

Multiplying explanations and special pleading? Good science, this does not make.

Apart from these, we find that, until we reach the Last Supper (Lk.x.14), Marcan and non-Marcan material alternates in great blocks. 
The sections, Lk.i.1-iv.30 (in the main); vi.20-viii.3; ix.51-xviii.14, and xix.1-27 are non-Marcan.
The intervening sections, iv.31-vi.19; viii.4-ix.50; xviii.15-43; xix.28-x.13, are from Mark, with three short interpolations from a non-Marcan source. 
From x.14 onwards the sources, as is inevitable if two parallel accounts of the Passion were to be combined, are more closely interwoven. 
Yes, this would be inevitable. What is this non-Markan source Streeter is talking about? More to the point, has he yet shown any reason to posit such a source? I'm not sure, up until this point, that Streeter has given a single reason for a non-Mark-non-Matthew source at all!
This alternation suggests the inference that the non-Marcan materials, though probably ultimately derived from more than one source, had already been combined into a single written document before the author of the Third Gospel used them. 
Here we set up more circular reasoning. Using the existence Q and Proto-Luke to explain why Q and Proto-Luke are needed is not evidence for the existence of either Q or Proto-Luke.

To be clear: Streeter has not yet provided a reason why Q is needed, nor what problems it solves.

The further inference that this combined non-Marcan document was regarded by Luke as his main source and supplied the framework into which he fitted extracts of Mark is worked out in Chap. VIII. of this volume.

The net result of the facts and considerations briefly summarised under the foregoing five heads is to set it beyond dispute that Matthew and Luke made use of a source which in content, in order, and in actual wording must have been practically identical with Mark.
Can we go a step farther and say simply that their source was Mark?

Again, while Streeter here tries to say that the chapter so far has been the establishment of the Priority of a document exactly like Mark, he's been slipping in information about the Q source, laying the groundwork for the importance of the Q source... It's almost as though he considers Q to be foregone conclusion.

To the view that their common source was exactly identical with our Mark there are two objections.

  1. If the common source used by Matthew and Luke was identical with our Mark, why did they omit some whole sections of their source?
  2. How are we to account for certain minute agreements of Matthew and Luke against Mark in passages which, but for these, we should certainly suppose were derived from Mark?

It has been suggested

  1. that the omissions of material found in Mark would be explicable on the theory that the document used by Matthew and Luke did not contain the omitted items --
    that it was an earlier form of Mark, or " Ur-Marcus," of which our present Gospel is an expanded version;
  2. that if the text of Ur-Marcus differed slightly from that of Mark, the same theory would account for the minute agreements of Matthew and Luke.

Clearly a decision as to the merits of an Ur-Marcus hypothesis can only be made after a study of the actual passages omitted by Matthew and Luke respectively, and a careful scrutiny of the so-called "Minor Agreements."
But there is one preliminary consideration that ought not to be overlooked.

In estimating the probability of Matthew or Luke purposely omitting any whole section of their source, we should remember that they did not regard themselves merely as scribes (professedly reproducing exactly the MS. in front of them), but as independent authors making use, like all historians, of earlier authorities, and selecting from these what seemed to them to be most important.

When it suits Streeter to have Luke and Matthew be "merely scribes", he assigns to them the motives and techniques of mere scribes. When it suits his theory to have Luke and Matthew hold non-historian motivations, that's fine too. When he wants them to be literary authors, that's fine too.

But, most importantly, they were careful historians in all regards, respecting their sources totally (except for order of events).

Excuse me if I remain skeptical.

Moreover, for practical reasons they probably did not wish their work to exceed the compass of a single papyrus roll.
If so, space would be an object. As it is, both Matthew and Luke would have needed rolls of fully thirty feet long;
and about twenty-five feet seems to have been regarded as the convenient length [Cf. Oxford Studies, p. 25 ff.].
And, when compression of some kind is necessary, slight reasons may decide in favour of rejection.
Very often we can surmise reasons of an apologetic nature why the Evangelists may have thought some things less worthwhile reporting.
But, even when we can detect no particular motive, we cannot assume that there was none;
for we cannot possibly know, either all the circumstances of churches, or all the personal idiosyncrasies of writers so far removed from our own time.
These are Streeter's wisest words in the entire chapter. It's a great pity, then, that he forgets these words when convenient: If in doubt, Luke and Matthew were historians, concerned with recording the true facts about Jesus, and if that is in doubt, say their motivations are unknowable.

matthew's omissions

Matthew's supposed omissions from Mark shrink on examination to very small dimensions.
Matthew reproduces the substance of all but 55 verses of Mark: of these 24 occur in Luke, a fact that creates a strong presumption that these at any rate were in the original source.
But Mk.iv.21-24, and i.33-37, which account for 9 of the 55 verses, are really cases, not of omission, but of substitution;
for in other contexts Matthew has sayings equivalent to, and usually more elaborate than, those which he here omits
[See footnote, p. 196.].
It is usually said that Matthew's omissions include three miracles of healing--
a Demoniac (Mk.i.23 ff.), a Dumb man (vii.32 ff), and a Blind (viii.22 ff.).
In the first of these the demon, as if by way of protest, "rent" the patient before coming out, and in the other two the cure is a gradual process with the use of a medicament like spittle instead of by a mere fiat.

Such details obviously would make these three healings less miraculous, less "evidential" of supernatural power, and, therefore, from an apologetic point of view, less worth recording, than others.

Note that doesn't say these miracles didn't happen, or that Matthew thought they were untrue. Instead, these miracles are merely "less worth recording". See what I mean about Streeter assigning motives to the actions of Matthew, despite saying those motives can't be known for sure?

But is it correct to say that Matthew has "omitted" these three incidents?
In his account of the Gadarene Demoniacs (viii.29) he modifies the words of the demoniac so as to combine the cry, as given in his immediate source (Mk.v.7), with that of the demoniac as given in the apparently omitted section (Mk.i.24).
This proves that Mk.i.24 stood in the copy of Mark he used.

Moreover, Matthew makes the demoniacs two in number, instead of one as in Mark.
This change also downplays Mark's Homeric source for this story as recounted in his Gospel.
Taken together, these phenomena suggest that Matthew considers himself to be, not omitting one, but, as it were, telescoping two healings of demoniacs which he found in Mark.
Or, to put it bluntly: he's a really shitty historian.
Again, Mark's cure of the dumb man is not "omitted," for Matthew substitutes in the same context as Mark a general statement that Jesus healed various sick persons, including dumb and blind, and calls attention to the impression produced on the multitude in words that appear to be suggested by the omitted section in Mark (cf. Mt.xv.31 = Mk.vii.37).
Also he inserts in another context (Mt.ix.32-33) a healing of a dumb man.
Here we have an example of the importance of textual criticism for the Synoptic Problem;
verse 34, which says that Jesus was accused of healing by the prince of devils, is omitted by D, a, k, Syr. S., and is a textual assimilation to the almost verbally identical passage in Lk.xiv.15;
it is a "Western non-interpolation" with more than ordinarily good MS. support. 
Read without this verse, the story in Mt.ix.32-33 looks like an abbreviated version of Mk.vii.32 ff. (with the "offending" details excised), transferred after Matthew's manner to another context. 
In that case one would be inclined to think that Matthew originally intended the healing of two blind men--
which he inserts immediately before this (Mt.ix.27-31)--
as another telescoping of two Marcan miracles into one (i.e. Mk.viii.22 ff. and Mk.x.46 ff.), for the detail "touched their eyes," ix.29, may well have come from Mk.viii.23, the other apparently omitted miracle. 
When, however, in copying Mark he actually reached the story of Bartimaeus, Mk.x.46 ff, he preferred to retell it in its original context, but forgot to delete it in the earlier part of the Gospel.

The rebuke of John for forbidding those who cast out devils in Christ's name but do not follow with the disciples (Mk.ix.38 ff.) is a passage which would so readily lend itself to being quoted in favour of the Gnostics who were already, when Matthew wrote, beginning to demoralise the Church, that its omission can occasion no surprise.
[Cf. Matthew's significant addition to Mark, "By reason of the spread of antinomianism (ἀνομία) the love of the many shall wax cold," xxiv.12. N.B. also Matthew elsewhere records a condemnation of some who profess to cast out devils in Christ's name, Mt.vii.22.]
Again, the attempt of our Lord's relatives to arrest Him (Mk.iii.21) and the incident of the young man with a linen cloth in Gethsemane (xiv.51 f.) are both cases where it is harder to explain why Mark thought it worthwhile to record than why Matthew (and Luke also) omitted.
Both Matthew and Luke also omit the parable of the Seed growing secretly.
In favour of its originality in the text of Mark is the fact that, with the Mustard Seed, it forms one of those pairs of twin parables illustrating different aspects of the same idea which are a notable feature of the tradition of our Lord's teaching (cf. p. 189 f.).

I think one must seriously consider the possibility that this had accidentally dropped out of the copy of Mark used by one or both of the other Evangelists owing to HOMOIOTELEUTON.  The eye of the scribe might very easily pass from the first to the third of the three successive paragraphs, each of which open with the words καὶλεγεν (Mk.iv.21, 26, 30).
While this is possible, before appealing to textual criticism, and suggesting a change to a hypothetical copy of Mark that no longer exists, it's probably best to consider and discard other explanations for Matthew and Luke agreeing against Mark. Namely, Luke knew Matthew.

Again, Streeter has yet to give any reason why this couldn't be the case!

If there are 48 examples in the Gospels of omission through homoioteleuton in א alone
[For the whole N.T. the number is 115. Cf. Scrivener's collation of א, p. xv.],
it would be odd if there were none in the first copy of Mark that went to Antioch.
Or again, either Matthew or Luke may have omitted it because he preferred to reproduce the Mustard Seed along with the Leaven (its twin parable in Q),
and having already a pair to illustrate the idea of the Kingdom as a gradual growth, thought a third with the same moral superfluous.
Since Matthew's personal predilections are all on the side of the more catastrophic apocalyptic conception of the Kingdom;
and since Luke, as we shall see, inclines to prefer his non-Marcan source (which gives the pair in another context), this may seem to some a more probable explanation.
But it is quite possible that the omission of the parable by Matthew may be due to one of these causes, and its omission by Luke to the other; both, at any rate, are causes that we can verify as operating elsewhere.
This is pretty much the definition of an unfalsifiable hypothesis or theory. That is, in every case where it helps our theory, further disparate and/or hypothetical occurrences will fall in line, and all on our side of the line.
In fact the only omission by Matthew for which it is hard to find a satisfactory explanation is the story of the Widow's Mite, Mk..41-44.
But here considerations of style almost guarantee the section as original in Mark.
In four verses we find no less than four examples of the most characteristic features of Mark's style--a "context supplement,"
[On the significance of this and the following expression the student is referred to Hawkins' Hor. Syn. pp. 125, 139; cf. also pp. 34, 132.]
a "duplicate expression," the idiom ὅ ἐστι, and the Latinism κοδράντης--
all of which we may note Luke is careful to revise away.

luke's great omission

It would seem, then, that there is no sufficient reason for supposing that any substantial passage in our present text of Mark was lacking in that known to Matthew.
When, however, we turn to Luke, the case is more debateable.

Luke frequently omits a section of Mark, but substitutes for it in a different context another version of the same saying or incident--apparently derived from the source which, as will appear in Chap. VIII., he on the whole preferred to Mark.
In many cases Streeter is happy to admit Luke rewrote Mark's material. In some cases, when it suits his own theories, he decides that Luke got the similar but different material from a different source to Mark. The only difference is the extent of the rewriting and its placement in Luke's Gospel.

It's important to point out that this is a description of this material, not evidence for it coming from a non-Markan source. Look up the "no true Scotsman" fallacy.

Obviously where this has occurred, though we cannot prove that the omitted passages stood in his copy of Mark, there is not a shadow of a reason for supposing that they did not.
The real problem arises from Luke's one "great omission" totalling some 74 consecutive verses (
[If vii.16 is genuine (om. B א L 28) the number is 75.].
Apart from this, his omissions are few, short, and easily accounted for.
But the absence from Luke of the equivalent of is, prima facie, evidence that at any rate the greater part of this section was absent from his copy of Mark, although it was indubitably present in that used by Matthew.

Internal evidence also is, up to a point, favourable to the theory that the section is a later insertion into the text of Mark, provided we suppose the opening and concluding paragraphs of it to be original.
In Jesus sends the disciples on ahead by boat to Bethsaida, while He Himself stays behind to dismiss the crowd.
He rejoins them, walking on the water during the storm, vi.51;
but the arrival at Bethsaida, the destination for which they set out, is not mentioned till viii.22. 
That is to say, the omission, not of the whole section omitted by Luke, but of vi.53-viii.21, viz. all of it except the first and last paragraphs, would make, superficially at any rate, a more coherent story.
Curiously enough, some critics who wish thus to connect the start for and arrival at Bethsaida have failed to notice that the Walking on the Water, which tells how Jesus rejoined the disciples, is needed to make the narrative cohere.

On the hypothesis that the original Mark omitted, not the whole section, but vi.53-viii.21, it could be argued that there are three paragraphs in the inserted section, which might very plausibly be regarded as parallel versions or "doublets" of matter occurring in the uninterpolated edition.
These are

  1. the Feeding of the Four Thousand, viii.1 ff., cf. Feeding of the Five Thousand, vi.30 ff.;
  2. the gradual cure of a deaf man by means of spittle, vii.31 ff., cf. the similar use of spittle to cure a blind man, viii.22 ff.;
  3. a voyage across the lake, immediately following a feeding of a multitude, in which the failure of the disciples to understand about the loaves is specially emphasised,, cf. viii.17.

Further, if only the Walking on the Water and the gradual Cure of the Blind Man, which are the first and last paragraphs of the "great omission," had stood in Luke's copy of Mark, it would not be hard to explain his electing not to reproduce them.
The gradual cure by means of spittle may have seemed to him a miracle lacking in impressiveness, while the story of the Walking on the Water might appear to play into the hands of the Docetae, who asserted that Christ's human body was a phantom, and were already beginning to cause trouble in the Church before the end of the first century.

Lastly, the retirement of Jesus into a mountain alone after the Feeding of the Five Thousand (, and the Walking on the Water must have stood in the copy of Mark used by John.
For John's version of this has (p. 410) conspicuous agreements with Mark against Matthew.
Again, if, as many think, the healing of a blind man with spittle, Jn.ix.6-7, implies a knowledge of the similar story, Mk.vii.32-34, this too must have stood in John's copy of Mark.
Thus, though John's copy cannot have lacked the whole of Luke's "great omission," it may have omitted all but the first and last paragraphs.

Nevertheless, to the attractive hypothesis that the original Mark lacked the section vi.53-viii.21, there are two very formidable objections.

  • There are some remarkable facts to which Sir John Hawkins [Oxford Studies, p. 64 ff.] first drew attention. 
    By a careful tabulation of minute linguistic peculiarities he has shown that in style and vocabulary the section resembles Mark in no less than eleven striking points in which Mark's usage differs conspicuously from that of Matthew and Luke, and, indeed, from all other New Testament writers.
    In fact, the style and vocabulary of this section are, if anything, more Marcan than Mark.
    Sir John's argument, being cumulative in character and dependent on a statistical comparison of minute details, cannot be summarised without weakening its force;
    but to my mind it is all but unanswerable.
  • The difficulty in the way of supposing that the passage was absent from the original text of Mark is enormously enhanced by the fact that it was present in that used by Matthew.
    For, once postulate two editions of Mark--a shorter edition known to Luke and a later longer edition known to Matthew--
    and the question of the lost end of the Gospel cannot be excluded from consideration.
    It is incredible that the editor of a second edition, whether it was Mark himself or some other, who was prepared to take upon himself to add as much as a couple of chapters in the middle, should have left the Gospel without an end--
    supposing the first edition had already lost it.
    But if the first edition had not already lost its end, how explain Luke's desertion of Mark's narrative at Mk.xvi.8, viz. at the exact point at which later on an accidental injury was to cause a mutilation?
    There are, moreover, further reasons (cf. p. 338 ff.) for supposing that Matthew and Luke both used a text of Mark, which, like ours, ended at xvi.8.
    It is very remarkable that any edition should have circulated which broke off short without giving an account of the Resurrection Appearances;
    but that a second and greatly enlarged edition should have been published without an ending is quite incredible.
Putting aside whether his conclusion is correct or not, here is another example of Streeter's argument from personal incredulity. He find something incredible, and so assumes there is no possible reason that someone could make that decision for any rational reason.

He also assumes Mark's ending is "lost" or "missing". Why? The only reason being that there are no resurrection appearances. Mark obviously had no problem with the disciples not seeing Jesus in Galilee after he rose, as he clearly states that the women at the tomb told nobody. The text itself tells a complete story, and needs to go no further. Different scholars practicing reader response criticism have come up with various reasons for Mark to end his Gospel like this, and each has its merits:

In Let the Reader Understand, Robert M. Fowler shows how the treatment of the disciples gets worse and worse throughout Mark, their status and understand of Jesus diminishing at every step. It gets to the point where they desert and deny Jesus. Jesus (via the young man in the tomb) offers them forgiveness, but even this isn't taken up. Mark's reasoning? The Jerusalem church fathers/the followers of Peter (in Rome?) become negative examples of discipleship, with authority transferred directly to Jesus (following the teaching of Paul).

Darrell Doughty suggests that the Gospel of Mark has a cyclical structure. As in, it was used in chruch meetings, and when the church leader got to the end, he would continue again at the beginning. Jesus says "meet me in Galilee" and then, in the following readings, the Disciples meet Jesus in Galilee. This doesn't make sense if Mark was writing history, but it makes perfect sense if he is writing (or recording) liturgical material. Robert M. Fowler also posits this cyclical nature of Mark, with the reader being encouraged to start reading the Gospel again as soon as it is over.

In The Homeric Epics and the Gospel of Mark, Dennis R. MacDonald brings out the links between Jesus's death and the destruction of Jerusalem. Leaving Jerusalem is a good idea, and many Jews who fled the city before its destruction went to Galilee. The words of the young man are a warning and a suggestion to the Jerusalem Christians. Is Mark explaining why some Christians didn't make it out of the city in time?

A possible reading (which I've yet to write about extensively) is that Jesus meeting someone in Galilee is a reference to Paul's vision of Jesus on the Damascus road. The disciples (representing the Jerusalem church fathers) do not see Jesus in Galilee... but Paul does, and authority is transferred to him. This reading relies on Mark, and his intended audience, already knowing the location of Paul's vision, which could have been possible, though I'm not going to explain the evidence or reasoning here.

What is very clear is that, whatever reason Mark had for ending his Gospel how he did, Streeter isn't the only one who decides that it needed expanding. Luke and Matthew both add more material, and Mark's Gospel was added to later too. The reasons for all these longer endings? Theology, and the rehabilitation of the Disciples.

We must not assume Mark had the same theology, nor the same view of the disciples, as Matthew, Luke, Streeter, nor anyone else, especially when, if Streeter would care to look, it's possible to find Mark's opinions in his text itself. While it is possible parts of Mark might be missing from all known copies, we must first look at Mark in its earliest knowable form (knowable to us, and to Luke and Matthew), and see if it can explain all the features within it, and its use and modifications by later writers.

Before that, we can't use hypothetical differences in Mark to both create a reason for Q or other hypothetical documents to exist and as evidence that they existed.

The precise weight to be attached to these two objections will be estimated differently by different people.
But at least they are serious enough to compel us to ask whether Luke's "great omission" can be explained by any other hypothesis than the absence of this material from his source.
Now it is a fact that plausible reasons can be produced why most of the contents of this particular section of Mark would not have appealed to Luke.
Motives which might have induced him to omit each separate item are put forward by Sir John Hawkins
[Oxford Studies, p. 67 ff.]
moreover, if, as I argue in Chap. VIII., Luke regarded Mark, not as his main authority, but as a supplementary source, the hypothesis of intentional omission cannot be ruled out.

My own mind has of late been attracted by a third alternative, that Luke used a mutilated copy of Mark. 
The case for this I state, but merely as a tentative suggestion.

There are four features in Luke's narrative that cry out for an explanation.

  1. Why does he place the Feeding of the Five Thousand at a "village
    eading κώμην for πόλιν with D Θ, discussed in Appendix I.]
    called Bethsaida," ix.10, when Mark, his source, expressly says that it was in a "desert place"?
  2. Why does he omit the place-name Caesarea Philippi as the scene of Peter's Confession (ix.18)?
  3. Why does he say that Jesus was "praying alone" on that occasion, while Mark distinctly says that the incident occurred "in the way"?
  4. How is the reading of B in Lk.ix.18, which on transcriptional grounds looks the more original, to be accounted for?
    B is supported by 157 f. Goth, and three other cursives in reading συνήντησαν (f. occurrerunt) for συνήσαν --
    "they met" for "they were with."
Feature 1 is a good example of a context dependent error or incongruity which is actually a very powerful critical tool that can prove one author used and modified the material from a previous author. Unfortunately for Streeter, Mark Goodacre has shown that the very same tool (in this case "editorial fatigue") proves that Luke used and modified material found in Matthew's Gospel.

All these questions receive a completely satisfactory answer if we suppose that Luke's copy of Mark included merely the beginning of the "great omission," as far as the words αὐτὸς λόκος; in vi.47, and then went straight on to ἐπηρώτα τοὺς μαθητὰς, viii.27.

Now, if a piece is torn out of the middle of a roll the mutilation is not likely to begin and end exactly with a paragraph which opens a new section; an accidental loss is far more likely to cut across the middle of a sentence at both ends.
Let us for the moment assume just such a mutilation. Luke's MS. of Mark would have run as follows (words in italics are specially significant; asterisks indicate where the break in the papyrus occurred):

Notice Streeter doesn't have any problems assuming Luke's copy of Mark was a papyrus scroll, not a codex, not parchment. Codices lost pages, usually from the beginning and end. Likewise scrolls would have missing sections, usually from the start. Would sections usually go missing from the middle of a scroll? I guess we'll have to assume so, as it handily provides more "evidence" for Streeter's upcoming hypothetical source documents. I have no problems with missing sections of documents, but using hypotheticals to show reasons for hypothetically to later prove other hypotheticals? Here I have some problems.

Italics are missing in this reproduction of Streeter's text, unfortunately.

And straightway he constrained his disciples to enter into the boat,
and to go before him unto the other side unto Bethsaida,
while he himself sendeth the multitude away.
And after he had taken leave of them, he departed into the mountain to pray.
And when even was come, the boat was in the midst of the sea, and he alone...***
and in the way he asked his disciples, saying unto them,
Who do men say that I am?

( ... viii.27b).

Granted such a text, what would Luke make of the story?
What he actually does (in the B text) is to write,
immediately after the account of the Feeding of the Multitude,

And it came to pass,
as he was praying alone,
the disciples met him:
and he asked them, saying,
Who do the multitudes say that I am?

And he inserts the place-name Bethsaida into the opening sentence of the Feeding of the Multitude,
though in other respects he closely follows Mark's version of the story.
A study of the passage shows that this procedure is of the most natural and reasonable kind.

  1. From the mutilated text before him he might infer that Bethsaida was only a short way off, so that the disciples would be able to land and come back to meet our Lord by road, after He had dismissed the multitude.
    It would follow that both the Feeding of the Five Thousand and Peter's Confession took place near Bethsaida.
    That being so, if the story is to be clear to the reader, the proper place to insert the name is obviously before the Feeding of the Five Thousand, not in between the two incidents.
    Luke, therefore, inserts the name Bethsaida in the most appropriate place, ix.10.
  2. Luke's omission of the name Caesarea Philippi has been quoted as evidence of his indifference to geographical detail.
    But the whole case for this indifference rests on his supposed omission of the geographical details contained in this section of Mark.
    And if the mutilation in his MS. of Mark included the half of verse viii.27, then Bethsaida was the only place-name he had in his source; and he does the best he can with that.
  3. The incident of Jesus "praying" and being "alone" is not an "editorial addition" directly contradicting Mark, but a reproduction of what in Luke's text of Mark was the immediate introduction to Peter's Confession.
  4. The reading of B (συνήντησεν = occurrerunt = "go to meet") is, as so often, shown to be original.
    [Probably the original reading was ἤντησαν = "met." συνῆσαν = "were with," the reading of most MSS., is a very early scribe's emendation.
    Someone then tried to correct an ancestor of
    B by this text and wrote συν over the ἠν, but the next copyist combined the two. 
    Prof. Burkitt has pointed out a similar reading of
    B. In Lk.xix.37 D has πάντων (neut.), א πασῶν δυναμέων; B (supported curiously enough by 579) has πάντων δυναμέων, a false concord explicable if πάντων  was original, δυναμέων an addition from the margin.]
    It translates Mark's ἐν τῇδῷ in the only meaning that could be given to it, if it followed just after

If Luke wrote at some distance from Rome, no difficulty is presented by the hypothesis that the only copy of Mark, which had reached him, was a mutilated one.
Speculation, however, as to how the mutilation occurred is not very profitable.
A papyrus roll was a very fragile thing, and the number of accidents that could happen to it was very large.

All I submit is that, in view of such a possibility and of the difficulties of supposing the section was not in the original copy of Mark, its absence from Luke constitutes quite insufficient ground for postulating an Ur-Marcus.

When it suits Streeter to disprove a hypothetical source document he'll lay out some pretty solid arguments. But then, when laying the groundwork for his own hypothetical source document (in this case Proto-Luke), instead of acknowledging arguments or evidence against is, he calls them "minor issues" and, with a variety of explanations, rationalizes them away.

But if the theory of an older and shorter edition of Mark is not needed to explain Luke's Great Omission, it is certainly not called for to explain his shorter omissions.
Several of them only amount to one or two verses, and there are obvious reasons why Luke should have left out the others.

Three passages, for instance (ix.28-29, x.35-41, xiv.26-28), reflect some discredit on the Apostles, and Luke always "spares the Twelve"--
"Apostles" is reading into Mark terminology from the Epistles and other Gospels. Mark always calls refers to the twelve as "disciples". The only time Mark uses the word "apostles" is after the person the disciples follow has died. In this case (Mark 6:30) Mark is specifically talking about the followers of John the Baptist, who when John was still above ground were called John's "disciples", and after John was in his tomb were they called "apostles", and were a different group to the Disciples of Jesus.

However, in this case it is good that Streeter is looking for the motives of a Gospel author, and not just assuming they are brainlessly copy and pasting by predetermined methods.

omitting the rebuke "Retro Satanas" (Mk.ix.33), and excusing the slumber in Gethsemane as due to sorrow (Lk.x.45), and only recording one of the three lapses.
The dancing of Salome ( has little value for edification.
The pith of the long discussion on Divorce (x.1-12) is given in the last two verses, for which Luke has an equivalent in another context (Lk.xvi.18).
The Cursing of the Fig Tree (xi.12-14, 20-22) might seem a harsh act for the Great Healer;
besides, Luke has the parable of the Fig Tree (Lk.i.6 ff.), which may be the origin of the story, and at any rate contains all the moral that can be drawn from it.
Mark ix.43-47 may already, for all we know, have been seized upon by certain over-zealous believers as an exhortation to self-mutilation of the kind which others justified from Mt.xix.12.
Finally, some of the omitted passages must have stood in Luke's copy of Mark, for Luke reproduces some verses which in Mark are intimately connected with others which he omits.
Thus Lk.ix.36 is an abbreviation of Mk.ix.9, but in Mark this verse forms the introduction to the four verses that follow.

These facts must be considered in the light of the evidence to be submitted in the next chapter that Luke regarded his non-Marcan source as primary, and conceived himself as producing a new and enlarged edition of that work, incorporating what seemed most important in Mark.
In that case passages of Mark not included in Luke must be regarded, not so much as "omissions" as "non-insertions," and the absence of any particular passage from Luke creates no presumption that it was absent from the copy of Mark that he used.

Before considering that Luke's primary source was his own teaching, Streeter considers the non-Mark material to come from another documentary source. Why? Streeter admits that it is for purely apologetic reasons, and to strengthen evidence for (or about) a historical Jesus.

minor agreements of matthew and luke

Accordingly it is clear that the only real difficulty in accepting out of hand the conclusion that the document used by Matthew and Luke was identical with Mark lies in the occurrence of agreements between Matthew and Luke against Mark consisting either in minor omissions or in some minute alterations in a turn of expression.
A full discussion of this subject is attempted in Chap. XI. But for the benefit of the reader who is not conversant with the Greek language, I will briefly sum up the conclusions there reached.

So far, if the reader did not already accept that Q existed, there is absolutely no reason for them to consider this a problem. Why is Streeter talking about a common non-Markan source used by both Luke and Matthew? He's not yet given a reason for Q to exist! Why can't Luke have known Matthew's Gospel? We still don't know!
  1. Such agreements are only significant in contexts where there is no reason to suppose that the passage also stood in Q.
  2. Most commonly these agreements result from Matthew and Luke changing a historic present in Mark into an imperfect or aorist tense, in their substituting a participle for a finite verb with "and," or in using a different conjunction or preposition from Mark.
    In every instance the change is, from the stylistic or grammatical point of view, an improvement.
    And as both Matthew and Luke continually make this kind of improvement independently, it is not surprising that both sometimes concur in doing so in the same place.
  3. If the agreement consists in an omission it is almost invariably of the unnecessary or unimportant words, which are characteristic of Mark's somewhat verbose style.
    Matthew and Luke both compress Mark;
    it would be hard to find three consecutive verses in the whole of his Gospel of which either Matthew or Luke have not omitted some words, apparently with this object.
    Since, then, both Matthew and Luke independently compress Mark by the omission of unnecessary words or sentences, and since in any sentence only certain words can be spared, they could not avoid frequently concurring in the selection of words to be dispensed with.
    Under such circumstances, coincidence in omission calls for no explanation.

  1. To help us remove evidence against Q, we will say that this hypothetical document was written in a way to make this evidence disappear. As it is a hypothetical document, and we have no evidence of what it actually contained. Which means we can use the hypothetical contents of Q to strengthen the claim that it existed! Very handy!
  2. Coincidences!
  3. More coincidences! Every non-significant agreement is a coincidence. Remember "always (except...)" means we forget about the "excepts".

There are, however, three instances where the agreement of Matthew and Luke against Mark amounts to five consecutive words;
and there are perhaps thirty of an agreement in one or two words.
These agreements are all of a kind which, if there were fewer of them, could easily be attributed to accidental coincidence.
But there are just too many of them to make this at all a plausible explanation.

Except... when coincidences can't easily explain away the evidence. This is a pity.

But though some explanation is required, a study of the phenomena reveals the fact that the hypothesis of an Ur-Marcus is of no service to us whatever for that purpose.
The essential point that emerges is that in the great majority of cases where Matthew and Luke agree against Mark, the existing text of Mark seems the more primitive and original. If, then, the document used by Matthew and Luke was not identical with our Mark, so far from being an earlier form of Mark, it must have been a later and more polished recension, all copies of which have since disappeared.
This is the explanation of the phenomena which was adopted by Dr. Sanday in the Oxford Studies (pp. 21 ff.) and is, I believe, accepted by the majority of authorities as the most probable.
It involves no a priori difficulties.
There would have been several copies of Mark at Rome at a very early date; and it is quite likely that one copyist would have felt free to emend the style a little.
From this copy those used by Matthew and Luke may have been made, while the unrevised copies, being in the majority, may yet have determined the text that has come down to us.

The essential point that emerges is... what? That Luke probably knew Matthew? That Luke couldn't have known Matthew? That Luke and Matthew both knew Mark and Q?

Or is it that Streeter still isn't arguing for Q, but against Ur-Marcus... and now he's introduced a new hypothetical version of Mark's Gospel that was known by both Luke and Matthew, but unknown to anyone else... and that Luke's copy of this hypothetical version of Mark's Gospel also had a part missing?

Actually, no, he's not arguing for that position, but it seems like he has to give it lip service. Let's see what he really thinks is going on...

Personally, however, I am inclined to seek a different explanation of that residue of the agreements between Matthew and Luke which cannot naturally be ascribed to occasional coincidence in the type of improvement in Mark which they constantly make independently.
The Synopses of the Gospels in Greek most widely used by scholars give the text either of Tischendorf or of Westcott and Hort which are based on the text of Alexandria as preserved in B א.
But in nearly every case where a minute agreement of Matthew and Luke against Mark is found in
B א it is absent in one or more of the other early local texts;
though, on the other hand, these other texts frequently show such agreements in passages where they do not occur in B, while quite a different set of agreements is found in MSS. which give the Byzantine text.
Indeed, even as between א and B there is a difference in this respect;
there are agreements of Matthew and Luke against Mark in the B text, which are not in א, and vice versa.
A careful study of the MS. evidence distinctly favours the view that all those minute agreements of Matthew and Luke against Mark, which cannot be attributed to coincidence, were absent from the original text of the Gospels, but have crept in later as a result of "assimilation" between the texts of the different Gospels.
Detailed evidence for this conclusion is submitted in Chapter XI.;
and, if it is correct, the one objection to the view that the document used by Matthew and Luke was identically our Mark completely disappears.
If, however, that evidence be deemed inconclusive, then Dr. Sanday's hypothesis best explains the facts.
But in any case, as I have already urged, they offer no support to the hypothesis of an Ur-Marcus.

Again, Streeter is having no problems multiplying explanations many times over. The two competing views go like this: to explain why Luke and Matthew sometimes agree in wording against Mark let's posit that:
  • Luke knew Matthew's Gospel.
or that:
  • Luke didn't know Matthew, and that:
    1. Both had access to a document (Q) which left no evidence of its existence outside of Luke and Matthew's writing.
    2. That Q had material that overlapped with some passages of Mark.
    3. That Luke and Matthew made many improvements to Mark that coincided.
    4. That Luke and Matthew omitted the same material from Mark in some cases.
    5. And that, in the remaining agreements, one of the following hold true:
      • A different version of Mark existed that left no evidence of its existence outside of Luke and Matthew's writing.
      • Taking a disparate set of scribal errors in our earliest extant texts, and combining the "correct errors" into a single set (though no single text contains them all), we can construct evidence to make the last evidence against the argued position to disappear.
To be clear, these multiplied explanations are not for the existence of Q (indeed, Q is needed for these explanations to work), but merely to explain "minor agreements of Matthew and Luke against Mark".

Why are these a problem? Again, we still don't know!

the document Q

Finally! I hope Streeter has a really good reason why Luke couldn't have known Matthew...

Although Matthew embodies about eleven-twelfths of Mark he compresses so much that the Marcan material only amounts to about half of the total contents of his Gospel.
It is remarkable that the additional matter consists preponderantly of parable and discourse.

This is not remarkable if, in fact, Matthew developed this extra material in a church meeting context. Teaching, not history, would be Matthew's main focus.
[Narratives peculiar to Matthew, apart from generalised statements of healing like xv.30 and xxi.14, are as follows:
the Infancy, i.-ii.; Peter walking on the water, xiv.28 ff.; the coin in the fish's mouth, xvii.24 ff.; various small additions to Mark's story of the Passion (i.e. xxvi.52-54; xxvii.3-10, 19, 24-25, 51b-53, 62-60); the Resurrection Appearances.
The two miracles, ix.27-34, are possibly intended to be the same as two recorded by Mark, which otherwise Matthew has omitted. Cf. p. 170.]
I'd regard Peter walking on the water as a small addition to story found in Mark, rather than an all-new narrative.
Of Luke rather less than one-third appears to be derived from Mark, though owing to the greater length of his Gospel--1149 verses as compared with 661-- and to some compression of Mark's style, this one-third of Luke includes the substance of slightly more than half of Mark.
Luke's additional matter includes both more narrative and more parables than Matthew's, but not quite as much discourse.
The discourse occurs in shorter sections, and is not to the same extent as in Matthew collected into large blocks.

Again, Luke could be arranging discourse into shorter sections to fit into a church meetings. A lectionary is a book that contains, not the whole Bible, but passages from it, arranged into lessons for reading in churches. Luke could be doing something similar with material from Matthew when including it in his Gospel: that is, breaking up long sermons of Jesus into smaller sections for easier teaching.

We notice that, of this large mass of material that must have been derived from elsewhere than Mark, a certain amount, approximately 200 verses, appears in both Matthew and Luke.
This matter, which they have in common, includes most of John the Baptist's Preaching, the details of the Temptation, the Sermon on the Mount, the Healing of the Centurion's Servant, John's Message,

Art thou he that should come,

Be not anxious for the morrow,

and many more of the most notable sayings in the Gospels.

If Luke used Matthew, we would notice this, but it wouldn't be a problem. Instead we could use it to study the different theologies and messages and audiences of the Gospel writers.
But there are two facts of a puzzling nature.

Here it is! We're finally going to be told the reason for Q to exist!
  1. The common material occurs in quite different contexts and is arranged (cf. p. 273 ff.) in a different order in the two Gospels.
  2. The degree of resemblance between the parallel passages varies considerably.
    For example, the two versions of John the Baptist's denunciation, "Generation of vipers ..." (Mt.iii.7-10 = Lk.iii.7-9), agree in 97% of the words used;
    but the two versions of the Beatitudes present contrasts as striking as their resemblances.
  1. Is this a puzzling problem? We already know Matthew and Luke weren't primarily concerned about keeping the order of material from their sources.
  2. Is this a puzzling problem? We already know Matthew and Luke weren't primarily concerned about retaining the exact wording or meaning of passages from their sources.
The simplest formulation to account for this "puzzle" is that Luke's audience already knows Mark's Gospel well (probably from repeated readings in church meetings). Mark's material would then be considered "canonical", not just by Luke, but by his audience. Luke didn't consider Matthews expansions with the same weight, maybe because he himself considered Mark's Gospel as especially authoritative, but more likely because he could change Matthew's material without having to justify or explain his actions.

How are we to account for this common matter?
The obvious suggestion that Luke knew Matthew's Gospel (or vice versa) and derived from it some of his materials breaks down for two reasons.

Here it comes...
Sir John Hawkins once showed me a Greek Testament in which he had indicated on the left-hand margin of Mark the exact point in the Marcan outline at which Matthew has inserted each of the sayings in question, with, of course, the reference to chapter and verse, to identify it;
on the right-hand margin he had similarly indicated the point where Luke inserts matter also found in Matthew.
It then appeared that, subsequent to the Temptation story, there is not a single case in which Matthew and Luke agree in inserting the same saying at the same point in the Marcan outline.
If then Luke derived this material from Matthew, he must have gone through both Matthew and Mark so as to discriminate with meticulous precision between Marcan and non-Marcan material;
he must then have proceeded with the utmost care to tear every little piece of non-Marcan material he desired to use from the context of Mark in which it appeared in Matthew--
in spite of the fact that contexts in Matthew are always exceedingly appropriate --
in order to re-insert it into a different context of Mark having no special appropriateness.
A theory that would make an author capable of such a proceeding would only be tenable if, on other grounds, we had reason to believe he was a crank.

Streeter, so lacking in imagination, can't think of a single reason Luke might use the arrangement of material he did. And so, he concludes, to arrange Matthews material in such a way, he must have been a "crank".

"he must have gone through both Matthew and Mark so as to discriminate with meticulous precision between Marcan and non-Marcan material" - What is so unbelievable about this? It's exactly the kind of study that Streeter has just cited, via anecdote, from Sir John Hawkins!

If Luke knew and had used Mark's Gospel for many years, and then came across a copy of Matthew's Gospel, he'd recognize it right away as an expansion of Mark. What is so strange about him going through Matthew's Gospel and (in the margin) indicating what was also found in Mark and what was new. He could then decide if he wanted to use this new material for teaching in his own church and, later, in his own Gospel.

Finally, once he wrote his Gospel, Luke used Mark as a framework, and then inserted his own material, and the extra (already delineated) material by Matthew as and where he wanted.

The appropriateness of this placement might not be immediately clear to us. But is no attempt going to be made to look at Luke's arrangement of Matthew's material? Other scholars have done so, and while it isn't immediately obvious, there is a discernible pattern.

As an analogy, if Streeter looked at a lectionary, would he be able to discern the arrangement of the teachings within? These readings match up with holy days and festivals. If you didn't know Easter came in the spring, wouldn't you assume that the death of Jesus should come at the end of the year? And the new year would begin with the birth of Jesus? But there are many months in between!

Instead of admitting Luke might have his own reasons for cutting Matthew's material into shorter passages and arranging them in a different order, Streeter declares that only a crank would have done so. However, after diagnosing the mental health of Luke, Streeter declares Luke was not a crank.

To be clear:

  • I don't understand a decision by Luke.
  • Luke did something I didn't understand, so he must be a crank.
  • Luke wasn't a crank.
  • Therefore: Q.
This is the number one argument against Luke knowing Matthew's Gospel!!!
Sometimes it is Matthew, sometimes it is Luke, who gives a saying in what is clearly the more original form.
This is explicable if both are drawing from the same source, each making slight modifications of his own;
it is not so if either is dependent on the other.
Are we going to get any examples of these original readings? How do we know what the original reading would have been if, as is the case, we don't have access to this hypothetical source document?

What would be more useful is, where the wordings differ, we could try to work out why Luke changed the wording of Matthew. We can then use the documents as historical records of, not the words of Jesus, but the different theology between two geographic regions of the ancient world (that is, if we go with Streeter's investigations in the first five or six chapters of this book).

A second explanation of the phenomena that has been suggested is that Matthew and Luke had access (in addition to the written Gospel of Mark) to different cycles of oral tradition, or to documents embodying such, and that these cycles, though in the main independent, overlapped to some extent.
For those cases where the degree of verbal resemblance between the parallel passages is small I myself believe that some such explanation is a true one.
For the more numerous examples where the verbal resemblances are close and striking it is far from convincing.

This is another case, in fact the main case, of Streeter multiplying explanations. Why do we need to posit an additional source of oral tradition when we already have two known sources of new material?

Accordingly a third hypothesis, that Matthew and Luke made use of a single common document that has since disappeared, has secured, if not quite universal, at any rate an all but universal, assent from New Testament scholars.

Here we could play "spot the logical fallacy", but as an introductory text, it's fine to lay out the state of scholarly consensus. But, you know, only if this isn't taken as evidence for the theory in question.
This hypothetical source is now by general consent referred to as Q, though in older books it is spoken of as THE LOGIA or THE DOUBLE TRADITION
Technically, "Double Tradition" is a description of the non-Markan shared material of Matthew and Luke, not the hypothetical document that contained it. By claiming older books mean "Q" when the author actually writes "Double Tradition" is a way to claim support for Q where, in fact, none might be intended.
Seeing that Q, if such a document ever existed, has disappeared, the hypothesis that Matthew and Luke used it cannot be checked and verified as can the hypothesis that they used Mark.
Despite this admission, Streeter has used the "fact" of Q extensively in his setup for the introduction of Q.

Most importantly, Streeter thinks the most important thing is that a hypothesis can be "verified". Logically, the opposite is true. There are many ways we can verify a false claim. Creationists verify the age of the Earth as 6,000-10,000 years old all the time. This doesn't mean they are correct.

What is more important is that a claim can be shown to be false. The age of the Earth can be verified at 4 billion years, yet if we found evidence that it was built as by superintelligent mice as a computer to solve the question of meaning of life, the universe, and everything, that verified fact would be falsified.

A hypothesis isn't incorrect if it can't be verified, but is incorrect if the evidence points to an incompatable hypothesis.

But it explains facts for which some explanation is necessary, and it has commended itself to most of those, who have studied the subject minutely in all its bearings, as explaining them in a simpler and more satisfactory way than any alternative suggestion that has so far been put forward.
I find it unbelievable that Streeter could write this with a straight face. Really? Simpler and more satisfactory?

What facts need explaining that can't be explained more simply by "Luke knew Matthew"? Streeter spent about 12,000 words setting up a (hypothetical) situation that would allow Q to exist (using Q to do so) and explaining away all the evidence against Q. Then Streeter writes TWO PARAGRAPHS outlining the problems Q solves.

And I, and many others, in using only two paragraphs more, can explain how they're not real problems at all.

Does Streeter think that, once we got to this point in his chapter explaining Q, we'd be bowled over by this? I am honestly astounded that anyone can take this chapter seriously at all. Are they all cranks and lunatics? (Actually, I've already mentioned their reasons, and Streeter will soon explain them more clearly.)

We are justified, then, in assuming the existence of Q, so long as we remember that the assumption is one, which, though highly probable, falls just short of certainty.

Q itself is based on a huge pile of assumptions. Why not use this assumption-based hypothetical document as the launch pad for loads more assumption-based speculations and study?

But it does not follow, because we accept the view that Q existed, that we can discover exactly which passages in Matthew and Luke were, and which were not, derived from it.
Nearly all writers on the Synoptic Problem have attempted to do this.
I have done so myself.
[Oxford Studies, Essay VI. On the hazards of reconstructing Q there are some valuable warnings in Burkitt's review of Harnack's attempt, J.T.S., Ap. 1907, p. 454 ff. I cannot, however, accept his own suggestion that Q contained an account of the Passion.]
But, for reasons which will be developed in Chap. IX., I now feel that most of these attempts to reconstruct Q have set out from false premises.

Here Streeter admits that the hypothetical Q document is unknowable, and makes no predications that can be confirmed or denied through further studies. "And so, in lacking any further use, let's study the reason why Q is useless for study. And then, at the end, still affirm its existence and use."
  1. Critics have underestimated the probability that in many cases slightly differing versions of the same sayings or parables would be in circulation.
    They have therefore been unduly anxious to extend the boundaries of Q by including passages, like the Lord's Prayer and the parable of the Lost Sheep, where the parallelism between Matthew and Luke is not exact enough to make derivation from a common written source its most likely explanation.
    Even if items like these stood in Q, it is probable that one or other of the Evangelists also had before him another version as well.
    Further study of the facts convinces me that a substantial proportion of the 200 verses in question were probably derived from some other source than Q.
  2. On the other hand, since Matthew and Luke would presumably have treated Q much in the same way as they treated Mark, it is fairly certain that some passages which are preserved by Matthew only or by Luke only are from Q;
    but I feel less confidence than heretofore in the validity of some of the principles by which it has been sought to identify them.
  3. Not enough allowance has been made for the extent to which sayings of a proverbial form circulate in any community.
    One such,

    It is more blessed to give than to receive,

    which does not appear in any of the Gospels, is quoted by Paul (Acts xx.35).
    At the present day, at the Bar, in the Medical Profession, in every College in Oxford or Cambridge, professional maxims, or anecdotes and epigrams connected with names well known in the particular society, are handed down by word of mouth.
    The same thing must have happened in the early Church;
    and it does not at all follow that a saying of this character, even if it occurs in almost identically the same form in two Gospels, was derived from a written source.

  1. Streeter admits that Q need not contain all double tradition. And so, as well as Q, he adds lots of oral tradition to account for some of the vaguer similarities between Matthew and Luke. Oral tradition that is as hypothetical as Q, and diminishes Q.

    "Further study of the facts convinces me that a substantial proportion of the 200 verses in question were probably derived from some other source than Q." - And yet Streeter refuses to considers that non-Q source might be Matthew.

  2. Why would Matthew and Luke have treated Q in the same way as they treated Mark? We know for a fact that this statement is logically impossible!

    To be clear: Both Luke and Matthew usually follow Mark's order. The double tradition is in a different order in Luke and Matthew. This difference in order is the very issue that Streeter gives for the existence of Q. Now he ignores the fact that, even if Q did exist, either Luke or Matthew must have changed the order of the Q material completely. He says Matthew did so, which means Matthew treated Q in very much not the same way as Mark's Gospel.

  3. As pointed out by Earl Doherty (and I'm sure others), we have no evidence that there was any oral tradition of the kind in the double tradition material, at least not in the earliest Christian writing such as the epistles. See The Sound of Silence, 200 Missing References to the Gospel Jesus in the New Testament Epistles by Earl Doherty.

    To be clearer: If there was oral tradition about the life and teaching of a historical Jesus, no early Christian author until Mark mentioned any of it, and in the case of the Epistle writers, where you would expect to find tradition about a historical Jesus, and where it would be helpful to those writers, the Epistles are continually silent.

    In regards to all the teaching and miracle stories associated with Jesus, the first evidence we have of any of these existing, ever, is in the Gospel of Mark. When new tradition appears, it is in either Luke, or Matthew, or both. The only evidence we have is the Gospels themselves.

    And so we can't use the variation in the Gospels to multiply the sources (written or oral) of this tradition. More importantly, we can't use this confusion of sources as evidence of more sources. The confusion is not in Matthew or Luke. The confusion is by Streeter.

    "... it does not at all follow that a saying of this character, even if it occurs in almost identically the same form in two Gospels, was derived from a written source. " - For Streeter, two cases of oral tradition with identical words is both proof of a common source, and also proof of oral tradition. When it suits. So even if it is found in Mark, it could still be oral tradition.

Where, however, a number of consecutive sayings occur in two Gospels with approximately the same wording, or where a detached saying is not of a quasi-proverbial character, a documentary source is more probable.
Hence, while the phenomena make the hypothesis of the existence of a written source Q practically certain, its exact delimitation is a matter of a far more speculative character.  
A tentative reconstruction is essayed in Chap. X.
The written source of Q is practically certain... except for the content of this practically certain document. How can Streeter be fine with this kind of evidence?

Again, Streeter is fine with an unknowable hypothetical as the basis of further study. I don't mind speculation, not at all, but speculation stands or falls when tested against:

  • Evidence - and Q fails at every hurdle.
  • Problem solving - Q solves a very minor problem (a problem Streeter has no issue with in other instances).
  • Conciseness - 15,000 words to explain how Q can exist? And still more chapters to come?
  • Predictive power - and Q is useless.
  • Falsifiability - Q is amorphous enough to continually ellude evidence against it.

The overlapping of Q and Mark

But although it is impossible to determine exactly what was and what was not contained in Q, one fact cannot be disputed--
there is a certain amount of overlapping between Q and Mark. 

Handily, this indisputable fact is the very thing that explains away all the evidence against Q.

Instead of seeing "minor agreements between Luke and Matthew against Mark" I see "massive disagreements between Luke and Matthew against Q". But then, if your mind is already made up, you can pick and choose your evidence as you want.

This observation holds good in principle even if we think (with Prof. E. de W. Burton
[Principles of Literary Criticism and the Synoptic Problem, p. 41 ff. (Chicago, 1904).]
that the "Q material" was derived, not from a single document, but from two, or (like the late Mr. A. Wright) that it represents a cycle of tradition and was not derived from a document at all.
In other words, whatever the theory we accept as to the character of the source or sources of the non-Marcan matter common to Matthew and Luke, it is clear that certain items were known to Matthew and Luke both in Mark's version and also in another decidedly different.
In fact, to put it paradoxically, the overlapping of Mark and Q is more certain than is the existence of Q.

You know, I actually thought Streeter's logic hit a low point earlier in this chapter. He admits there is a paradox in Q, but somehow manages to convince himself that a paradoxical situation is a good place to be. Q can't exist unless Q is exactly as it needs to be for Q to exist, and we ignore or explain away all evidence against Q by postulating more facts about Q?

Circular reasoning and special pleading? More, please!

The student will find convincing proof of this, if, in his Synopsis of the Gospels, he will underline in red words found in all three Gospels, in blue those found in Mark and Matthew, in purple those in Mark and Luke, and, say, in yellow words found only in Matthew and Luke, in the accounts of John the Baptist's Preaching, the Baptism and Temptation, the Beelzebub Controversy, the parables of the Mustard Seed and Leaven, and the Mission Charge (, cf. Mt.x.l-16a = Lk.x.1-12).

The phenomena revealed are only explicable on the theory that Matthew and Luke had before them a version of these items considerably longer than that of Mark.
Yes. Matthew had longer versions that he made up, and Luke had Matthew. Except, when needed, evidence for Luke knowing Matthew becomes, in the hands of Streeter, evidence for Q, and by extension, against Luke knowing Matthew.
And it will be noticed that, while Matthew carefully combines the two versions, Luke prefers the non-Marcan, introducing at most a few touches from that of Mark.

Streeter has already admitted that Matthew's writing is better and more concise than Mark's writing. If Luke is wanting to leave room for more of his original material, would he pick the longer, clumsy telling of a story by Mark, or the more concise retelling by Matthew?

This overlapping of Mark and Q, found in the above sections and in a few other short sayings, covers about 50 verses of Mark.
[Most of the relevant passages are printed in parallel columns and discussed in my essay in Oxford Studies, p. 167 ff.
There is a valuable discussion in Sources of the Synoptic Gospels, p. 234 ff. (C. S. Patton, Macmillan Co., New York, 1915).]

And, wherever it occurs, we find that Luke tends to preserve the Q version unmixed, while Matthew combines it with that of Mark.
This, indeed, only means that Matthew and Luke differ in their treatment of Q in precisely the same way as in their treatment of Mark--
in both cases Matthew conflates his sources, Luke alternates them.

Maybe I misunderstood when Streeter said Luke and Matthew treat Q in the same way that they treat Mark. Not the same as each other, but that Luke treats Q the same as he treats Mark. This means that our reconstruction of Q can be based on using Luke's handling of Q as evidence of Q. Or maybe I'm misunderstanding Streeter yet again.
This difference, of which we shall see many examples, affords a valuable principle for distinguishing the Marcan and the Q versions in doubtful cases.

Many critics explain this overlapping of Q and Mark on the theory that Mark knew and made extracts from Q.
In favour of this view there is the fact that in many cases where Mark and Q overlap, the Q version is longer and also looks the more original.
In fact, as I put it in an Essay in the Oxford Studies, the Marcan often looks like a "mutilated excerpt" from the Q version.
[Oxford Studies, p. 171.]  

In that case the first difficulty would be to explain the very small amount of matter (not more than 50 verses) which Mark derives from Q.
The suggestion I then made was that Mark wrote for a Church in which Q was already in circulation, and intended to supplement rather than to supersede Q, and he therefore only drew from it some brief allusions to certain outstanding points which could not be altogether passed over in a life of Christ.
Here Streeter agrees with one of the concepts I mentioned earlier: a Gospel author writing for a church in which a previous scripture is known. Except, in this case, Streeter posits it to explain the absence of evidence for an overlap between a known document and a hypothetical document (or as evidence for the lack of greater overlap?).

In opposition to Streeter, I've used the idea to explain why:

  • both Matthew and Luke include most of Mark's Gospel - as if Mark's Gospel is known in their churches already, so both Matthew and Luke attempt to supplement it.
  • Luke felt free to change what he wanted from Matthew's Gospel - as if Matthew's Gospel wasn't known in his chruch, so Luke tries to preemptively supersede it.
Neither of these involve hypothetical documents, nor as a way to try to prove negatives.
But the net result of the discussion of the question among scholars during the last thirteen years has been to add weight to, rather than to detract from, the difficulty I even then expressed of supposing that Q lay before Mark in a written form.

Here Streeter admits that the only thing Q is good for is the study of the problem of Q. Get back to me when Q is useful for solving any other problem that is not related to the content of Q. Thanks!

In Mark's account of the Temptation there is no mention of the fast.
Indeed, if we did not unwittingly read into Mark's account what is so familiar to us from the other two Gospels, we should naturally interpret the imperfect tense of the verb in the phrase "the angels ministered to him" as meaning that Jesus was continuously fed by angels, as once Elijah was by ravens.

And "if we did not unwittingly read into Mark's Gospel what is so familiar to us from the other two Gospels..." for example: that Mark thought that Jesus might have appeared to his disciples after he rose from the dead?
Again, while in Matthew and Luke the emphasis is on the internal content of the various temptations of our Lord to a misuse of His lately realised Messianic powers, in Mark it is on the external fact that "he was with the wild beasts," which is not even mentioned in the other accounts. Mark's representation of this incident is so wholly different from that in Q that, if we were compelled to assume that he could have derived it from no other source, we must say that he had read Q long enough ago to have had time to forget it.

Mark knew Q... but read it so long ago he forgot it! Remember my hypothetical reconstruction of an enthusiastic church attendee dictating what he remembered from hearing from Matthew to his scribe? This kind of speculation is fine when it fit's Streeter's theories, but instead of Mark being called a lunatic or crank, Mark is merely forgetful. What is Streeter arguing for here?

John's Preaching, the Baptism, and the Temptation obviously form a single section, and a source, which contains the first and third, must have contained the second, which not only connects the other two but is the point round which they hinge.
Q, therefore, must have contained an account of the Baptism.
But whereas in Mark's version the voice from heaven is

Thou art my beloved Son, in thee I am well pleased,

in Q it must have read,

Thou art my beloved Son, this day have I begotten thee;

which is the reading of the Western text of Luke, and is undoubtedly right.

For Q, presume Q, because Q, therefore... Q!
As has been already pointed out (p. 143), it can be traced back to Justin Martyr; and since it is a reading which not only introduces a discrepancy between the Gospels but also seems to favour what was later regarded as the dangerous heresy that Jesus only became Son of God at His Baptism, its "correction" in the Alexandrian MSS. is easy to explain.
But if Luke wrote this, and that with Mark in front of him, it must have been because it stood in his other source.

Again, the picturesque details which Mark (i.6) gives as to John's dress and food look authentic, but there is no reason to suppose they stood in Q.
Here, and indeed wherever Mark and Q overlap, Matthew conflates the two versions;
Luke prefers that of Q.

Streeter's logic breaks my brain when I try to work out what he is taking for evidence of Q. If Luke likes the way Matthew tells a story better than Mark, why shouldn't he prefer the refined retelling?

This is a major agreement against Mark, not a "picking of Q over Mark."

But if we take Luke as on the whole representing Q, and consider the section John's Preaching, Baptism and Temptation as a whole, the differences between his version and Mark are far more striking than the resemblances.
It is only the fact that Matthew combines the two versions, and most people read Matthew first, that has concealed the extent of the contrast so long even from students. And it is only that most people hear of Q first which has concealed the problems with the Two Document hypothesis for so long.

The case for regarding Mark's version of the Beelzebub Controversy as an extract from Q is stronger.
But, again, if we realise that Matthew's version is partly derived from Mark, and therefore take Luke's version as on the whole nearer to Q, the verbal resemblances between the two accounts are no more than would be inevitable if they represent two quite independent traditions of the same original incident and discourse.

Again, it's difficult to tease apart Streeter's logic enough to understand what he is arguing for here. Did Mark know Q, and then write his own version of the story anyway? Did Mark know Q, but read it and forgot parts of it?

Why not just go the whole way and say that both Mark and the Q author were both working from yet another hypothetical document (let's call it Proto-Q), and that both of them read that so long ago that they forgot parts of it... and that the parts that they did remember, or have in front of them, they decided to re-write anyway. I'm sure we could cook up evidence for this... and we could ignore all the evidence against while we're at it.

Special pleading, circular reasoning... and hypothetical documents all the way down.

In this case, however, part of the argument that Mark derives from Q depends on the suggestion that the way in which the section appears in Mark is such that it looks as if it were an interpolation.
But this contention disappears on closer investigation.
The removal of Mk.iii.22-30 does not leave the smooth connection we should expect if it was really an interpolation.
On the other hand, if the words "they said" in Mk.iii.21 are interpreted as meaning "people were saying," on disait, the section reads, not like an interpolation, but as a digression intended to explain their action.
"They did so, for report said He was mad, and the scribes had gone so far as to say He was Beelzebub, but He made short work of them." Mark's phrase is ambiguous and not very good Greek, and, as usually happens with imperfectly educated writers, the digression is clumsily introduced.
Is this a good time to mention that Streeter's digressions are so clumsily introduced that I'm never sure if he is arguing for or against his main points?
But it is more likely that our Lord's relatives should have come to apprehend Him, because they had heard a report that He was beside Himself, than that they should have arrived at such a conclusion for themselves.
And it is by no means likely that Mark would have told the story at all, if he had meant what he is usually understood to mean.
I think this is the first time where Streeter's language unambiguously shows his scholarly position: that is, he is talking about "his Lord".

Just as Luke and Matthew have their characters refer to Jesus as Lord in their new material, Streeter falls directly in line. I have no problems with devout Christians doing critical or historical studies of the Bible, but trying to protect the reputation of authors and, in turn, trying to defend the historicity of Jesus, needs to be separated from critical analysis.

In all our sources we find the phenomenon of twin-parables [Oxford Studies, p. 173.] illustrating different aspects of the same idea--
the Hidden Treasure and the Pearl of Great Price (Mt.i.44-46), the Tower Builder and King making War (Lk.xiv.28-32), the New Patch and the New Wine (Mk.ii.21-22), the Lost Sheep and the Lost Coin (Lk.xv.3-10).

In the Oxford Studies I argued that the Mustard Seed in Mark was the mutilated half of the Mustard Seed and Leaven, which since they stand together in both Matthew and Luke must have formed such a twin pair in Q. Order or arrangement of material is evidence for one thing... until it is evidence for something else.
But Mr. E. E. Buckley [Introduction to the Synoptic Problem, p. 147 (Arnold, 1912).] acutely points out that in Mark the Mustard Seed does not stand alone;
it is paired with the parable of the Seed growing secretly, which is quite as appropriate a twin as the Leaven to illustrate the idea of the gradual growth of the Kingdom. It would seem, then, that the twin-parable argument really cuts the other way, and suggests that in Mark and Q we have two pairs which have descended along quite independent lines of tradition.
Two pairs of related tradition that are not connected... in this special case? When is pairing of traditions evidence for a common source, and when is it evidence that we have two different sources?
Mt.x.5-16 is clearly a conflation of the Q discourse, given by Luke as the Charge to the Seventy (Lk.x.1-12), with Mark's discourse on the Mission of the Twelve (
Matthew has additional matter both at the beginning and the end which may possibly come from a third source (cf. p. 255), but in the central part of his version of the discourse (Mt.x.9-16a) there is hardly a word which is not to be found either in or in Lk.x.1-12.
Extra sources when needed? Handy!
The five words, on the other hand, common to and Lk.x.1-12 (I do not count καὶ, ἄν and μή, the definite article and personal pronouns, which for this purpose are not significant), "wallet," "enter," "house," "remain," "feet," are such as must occur in any version of this discourse.
Assuming, then, that Luke x.1-12 (not being conflate with Mark) represents
Q, the differences between Mark and Luke are so great and the resemblances so few that they favour the view that Mark's version is independent, not derived from Q.
If Mark did use Q, he must have trusted entirely to memory, and never once referred to the written source. Again, I'm not sure if Streeter is arguing for or against Mark knowing Q, or for or against Mark knowing Q well or by distant memory. That this is even an issue, and that I'm unsure of Streeter's position on it, belies just how nebulous Q is in Streeter's hands.

There remain no other considerable passages where Mark and Q are parallel;
for only portions of Mk.iv.21-25 and Mk.ix.42-50 have their equivalents in Q, and that in scattered contexts.
The rest are all quite short, consisting of one or two verses.

Considerable is a handy term. I think it means "evidence that can't be ignored or argued away using a variety of explanations". Word of mouth and oral tradition are also handy here, as they are open to evidence-free claims and the development of ever more complex hypothetical situations.
Mostly they belong to the class of proverb-like saying, which, as has been argued above (p. 185), would be likely to be circulated in different forms by word of mouth. I like how, to explain different forms of the same proverb, Streeter introduces the idea of different types of oral tradition. "Proverb-like sayings" as opposed to... what? And we have another "mostly" which, as we know, means "never (except...)".
To the critic perhaps the most interesting examples are Mk.viii.34, cf. Mt.x.38 = Lk.xiv.27, "take up the cross," and Mk.viii.38, cf. Mt.x.33 = Lk..9, "denies me on earth."
A glance at the Synopsis will show that Matthew and Luke give these sayings twice over--
once in the context parallel to Mark and in a version very close to Mark's, and again in the quite different contexts to which the references are given above, but in a version much less close to Mark's.
This shows beyond doubt that Matthew and Luke had versions of the sayings in two distinct sources.
Assume Q, and all evidence can fit it. Assume the Gospel authors have brains, and Q is no longer needed, nor any overlap between Q and Mark.

Is there no other case where Matthew repeats four words in a row? No other case where Luke repeats four words in a row? I'd have to check to see, but I have a feeling that Streeter is giving these special attention.

Anyway, I've not studied these passages particularly, but from glancing at these two four-word phrases (and not even in their contexts in the Gospels) I recognize them as places where the disciples failed Jesus. "take up the cross" is something that Simon Peter didn't do in Mark, and Simon the Cyrenian did instead. "denies me on earth" is something that Simon Peter did do, which even Matthew found so bad that he doesn't mention Simon Peter by name again after his final denial, not even for the resurrection appearances. The rehabilitation of the Disciples after Mark's harsh treatment is something both Matthew and Luke attempt, though Matthew's motives where Simon Peter is concerned is slightly more nuanced than pure whitewashing (see Peter in Matthew by Arlo Nau).

That both Matthew and Luke might find need to repeat these two particular phrases comes as absolutely no surprise to me, and that is with no further study on my part. Streeter doesn't care to look at the reasons Matthew and Luke might do anything, sees them as nothing but scissor and paste historians, and in doing so only sees evidence of the material Matthew and Luke would need to scissor and paste. In other words, every slight anomaly becomes evidence for more hypothetical source documents, rather than as a way to determine the different theology or motives of authors we know for sure existed.

The two versions differ to an extent that makes it improbable that Mark's was derived from Q, unless his dependence on Q is held to be already securely established on other grounds. Not acknowledged: Every problem for Q makes it less secure, as it hasn't been established on any grounds.
[Prof. C. H. Dodd points out to me that in three cases (Mk.iii.28, iv.22, vi.8) the variations between the Marcan and the Q versions might be explained as divergent translations of Aramaic: iii.28 בר' נשא singular or collective; iv.22 ד = ἵνα or ὅ; vi.8, where לא for אלא would give instead μὴ αβδόν.]

Are we now appealing to a hypothetical original language and a hypothetical translation of a hypothetical document that might have been hypothetically used by Mark? How is this useful for anything except as extra fodder for the study of solution to a non-problem that achieves nothing but raising more problems?

On the whole, then, the evidence is decidedly against the view that Mark used Q.
In that case the general, though not invariable, superiority of the Q version remains to be accounted for.
This can only be done if we suppose that Q was a document of very early date and represents a peculiarly authentic tradition.

If Q consists of the Double Tradition (that is, material shared by Luke and Matthew but not in Mark), the fact that Luke and Matthew are better writers than Mark accounts for its superiority in terms of literary style and language. If Streeter is talking about the historical superiority of Q, then again, the definition of Double Tradition is material that Matthew invented that was good enough that Luke wanted to use too. If Luke liked a Mark version of a story better, he could use that. If (as seems common) he liked Matthew's material on the subject better, he used that instead.

To be clear, calling Q superior is the same as calling "the material that Matthew contributed anew or modified from Mark that Luke liked enough to include in his Gospel" superior. If Luke knew Matthew, what we would expect to find is the overlapping material to be the most superior. That is evidence that Luke used both Matthew and Mark, not that he was using Q.

And what would we expect to find if Q was indeed an earlier document than Mark, full of authentic tradition, and of superior quality compared to Mark? We would expect to find some or even any evidence of Q. But we don't. We have no copies, no mentions of it by the church fathers, no clues at all that anyone even suspected its existence for almost two thousand years. What happened to this superior and authentic source after two disconnected authors (Luke didn't know of Matthew's work, remember) individually found and used a copy? Do we have evidence that Q was considered heretical and all copies ordered destroyed? We have no evidence for such a decree, but that would explain the lack of copies... but why would a superior and authentic source be considered heretical by someone with authority to determine its fate?

So "Luke knew Matthew and used what he liked best" makes a prediction: that the Double Tradition is actually of higher quality than even the Triple Tradition (that is, Mark's material unmodified, or less modified, but still included in both Matthew and Luke). Streeter himself has shown that this prediction is true (though he attributes this superiority to a superior source document).

On the other hand, "Q is a superior source" makes no prediction. Q is a description of a class of material. If Q material is considered less superior, it fades into Markan material. Or if it is less superior, neither Luke nor Matthew include it, and all evidence for that less-superior material vanishes. Or, in the non-general cases when Q is said not to be superior, and Mark is used instead... well, let's just ignore those cases.

A modern illustration

Small things that fall within our own experience may often illuminate great things known to us only through books.
Says Gibbon, speaking of the insight into military system that he gained from the peaceful manoeuvres of his county militia:

"The captain of the Hampshire grenadiers (the reader may smile) has not been useless to the historian of the Roman Empire."

In a similar spirit it may be worthwhile to quote an experience of my own, which shows the psychological working out in practice of "conflation," "agglomeration," and other kinds of "editorial modification."
A prolonged study of ancient documents compels the professional critic to assume such processes, but to the plain man the hypothesis often seems over-ingenious, unverifiable, and unreal.

I will leave this "modern illustration" passage without much comment. However, I will note that Streeter assumes his own actions as the historian of a religious figure is a good illustration of Luke and Matthew's techniques as historians of a religious figure. And yet, when it comes to his own actions as someone motivated to study scripture, with all the notes in margins and comparing sources this entails, he can't see that as a good illustration of what Luke might have been capable of with the texts of Mark and Matthew's Gospels on a table in front of him.

In 1920 I undertook, in collaboration with a friend, to prepare for publication a sketch of the personality and teaching of Sadhu Sundar Singh, commonly known as "the Indian St. Francis."
[The Sadhu: A Study in Mysticism and Practical Religion, by B. H. Streeter and A. J. Appasamy (Macmillan, 1921).]
The Sadhu had left England, and any extensive correspondence with him seemed impracticable.
Hence we had to rely upon a collection of printed and manuscript material, and on the recollections of our own personal intercourse with him and that of some other friends.
That is to say, we were dependent on written documents supplemented by a certain amount of "oral tradition."

Our materials included two brief "lives" written in India, which to some extent overlapped.
We had also three different collections of addresses, given by him in India, Ceylon, and Great Britain respectively, and various newspaper reports.
Seeing that the Sadhu is in the habit of freely repeating the same story or parable on different occasions, the phenomenon of parallel versions of the same material frequently occurred.
Thus the problem we had to solve was essentially that of combining into a single whole materials derived from a number of disconnected, independent, but to a large extent overlapping, sources.
It was not, however, till the book was in proof that I realised that circumstances had forced us to devise ways of dealing with our materials having the closest analogy to those which criticism suggests were habitually employed by editors in antiquity.

  1. Our first step was to single out the central ideas and leading topics to which the Sadhu most frequently recurred; our next, to sort out roughly the materials from various sources under headings corresponding to these main ideas.
    Then we carefully rearranged and fitted together the sayings and parables collected under each several head in such a way as to present, in the Sadhu's own words, a coherent, connected, and as far as possible complete, account of his teaching on that particular topic.
    Thus almost every discourse in the book is an "agglomeration," containing material drawn from two or three different sources.
  2. The frequent occurrence in our sources of two and sometimes three versions of the same story or saying presented us with a problem we could not avoid facing.
    The solution that seemed obvious was to select what seemed the freshest and most original version.
    But wherever an otherwise inferior version contained a detail or a phrase which, from our knowledge of Indian conditions or our interpretation of the character and philosophy of the man, seemed to us specially interesting or authentic, we worked this detail or phrase into the substance of the selected version.
    In other words, we "conflated" two, and occasionally even three, parallel accounts.
  3. For a variety of reasons, one of which was the desire not to swell the size of the book in view of the high cost of production, we decided not to reproduce the whole of our materials.
    Inevitably, in considering what to jettison, we were guided by our own feeling, or by the opinion of friends, as to which sections were the less interesting, valuable, or characteristic, and decided to omit these.
  4. Since the Sadhu's knowledge of English was limited, we considered ourselves free to amend the grammar and style wherever it seemed desirable, so long as we did not alter the sense.
  1. Our first step was not to be impartial or objective historians, but to select the material that would best support our view of his teaching. We had no problem rearranging or recombining material into different orders to fit our own publication needs.

  2. Instead of using the least-fit or primitive material, we selected the best. If needed, we chose the best elements from more than one telling of the same story.

  3. We left out the material that was least suitable, based on feelings, interest, or value to our audience, to keep the book to publishable length.

  4. We fixed the primitive language of our source.
Why is it that I don't see the "best practices of a historian" in Streeter's example, but a pretty good guide to "making a book full of spiritual guidance and discourse"?

Were Luke and Matthew interested in the "best practices of a historian" or "making a book full of spiritual guidance and discourse"?

Modern devices like an introduction, footnotes, and inverted commas enabled us to take our readers into our confidence as to our method, and to give references to original sources where anything depended on it.
An ancient author was not expected to do this, and for purely mechanical reasons it would have been impossibly cumbrous with the ancient form of book.

Due to the lack of modern devices, Streeter is fine with assigning any technique and motive to Luke or Matthew, so long as it fits his preferred theory. But instead of modern devices, does Luke or Matthew use an ancient device to explain their methods?

I like the speculation that Matthew did hint at what he was doing with his Gospel. Matthew 13:52: "Then said he unto them, Therefore every scribe which is instructed unto the kingdom of heaven is like unto a man that is an householder, which bringeth forth out of his treasure things new and old."

Old and new. Bringing treasure from previous scriptures, and adding new material. Matthew the historian would bring forth treasure old and old.

But the point I wish to bring out is how a personal experience has caused me to realise that--
given a multiplicity of sources--
some such editorial methods are forced upon an author who wishes to present the reader with a biographical portrait rather than a chaotic mass of disconnected obiter dicta.

To hint at a comparison, though it be only in regard to the mere mechanism of editing, between a work of one's own and a book of the Bible will seem, I fear, to some readers, to border on indecency.

Streeter shows his hand here. Why should any speculation about any actions or motives or techniques of the Gospel writers be considered indecent? Only if you are practicing apologetics could this even be a concern.
I conceive, however, that it is worthwhile to incur the risk of such a criticism in order to be able more firmly to substantiate a proposition on which I laid some stress in the Oxford Studies.
How daring! Should the opinions of those who consider the Bible the inspired word of God even be a consideration when it comes to historical/source/textual/redaction/etc criticism of the Bible?
It is really important for the ordinary man to realise that the use by the authors of our Gospels of editorial methods like "conflation" and "agglomeration" of sources does not necessarily impair, indeed under the circumstances it may well have been the best way to secure, an effective presentation of the total impression of our Lord's teaching.
Our Lord's teaching is unknowable. All we will ever know is what the Gospel writers tell us Jesus said, and that will always tell us more about their theology and lives than Jesus's.
On this subject I venture to repeat some words I wrote thirteen years ago. [Oxford Studies, p. 226.]

I think Streeter is trying to appear devout and inspirational with this long quote to finish out this chapter, but to me it does nothing but reveal the reasons for his utterly shoddy scholarship, and his disregard for anything approaching sound logic.

"Insomuch as the loss of a single syllable which might throw a ray of light on any act or word of our Lord is to be regretted, we must regret that Q, and possibly some other early writings used by Matthew and Luke, have not been preserved unaltered and entire.
Yet perhaps the loss is less than we may think.
Who does not feel that St. Mark, the oldest of the Gospels we still have, is the one we could best spare?

Without him we should miss the exacter details of a scene or two, a touch or two of human limitations in the Master, or of human infirmity in the Twelve, but it is not from him that we get the portrait of the Master that has been the inspiration of Christendom. Without Mark we wouldn't have a glimpse at an important step in the historical development of the concept of Jesus.
A mechanical snapshot is for the realist a more reliable and correct copy of the original than a portrait by Rembrandt, but it cannot give the same impression of the personality behind.
The personality behind Matthew and Luke's work is not the personality of Jesus, it's the personality of Matthew and Luke, or the personalities of Matthew and Luke, as they each have their own styles. In knowing Matthew and Luke better we glimpse another few steps in the historical development of the concept of Jesus.
The presence of a great man, the magic of his voice, the march of his argument, have a mesmeric influence on those who hear, which is lost in the bare transcript of fragmentary sayings and isolated acts such as we find in Mark or Q. I find it sad that Streeter can't find a meaningful message in Mark, nor admire Mark's literary talent. I am glad that both Matthew and Luke did appreciate Mark though, as Mark inspired them both to develop and record their own versions of the same story.
Later on, two great, though perhaps unconscious, artists, trained in the movement begun by the Master and saturated by His Spirit, retell the tale, idealise--if you will--the picture, but in so doing make us to realise something of the majesty and tenderness which men knew in Galilee.

Really? Matthew and Luke weren't divinely inspired authors, but divinely inspired historians? They unconsciously wrote some of the best known and enduring stories in human history? Is Streeter going to strip them of all individual creativity, and also take from them their free will?

Can't we leave divine inspiration out of it?

The only Master whom Luke and Matthew followed was Mark, an author Streeter relegates to merely the provider of a source document, less superior than even the hypothetical Q.

An instance will make this clear.
The realist may object that the Sermon on the Mount is not the sermon there delivered, but a mosaic of the more striking fragments of perhaps twenty discourses, and may approve rather of St. Mark or Q because there we have the fragments frankly as fragments.
But on the hill or by the lake they were not listened to as scattered fragments but in the illuminating context, and behind the words was ever the speaker's presence.

'The multitude marvelled as they heard,'

says Mark in passages where Matthew's story leaves us cold.
We turn to the arresting cadence of the Sermon on the Mount and it is no longer the multitude but we that marvel."

Now shoot me.

Or go back to read my final comments.