The Synoptic Problem

Whole Incidents—When Mark Is Made Secondary

In the foregoing, there has, I think, been developed a thoroughly reasonable purpose for the Markan writer when he is conceived as a compiler working with the Gospel of Matthew before him. This purpose has been considered under two conditions—(1) when no account is taken of external evidence; and (2) when the internal indications are supplemented by the ancient tradition as to the connection between the Apostle Peter and the Gospel of Mark. Let us now go on with our inquiry into the acceptances, omissions and additions of whole incidents and the character of these, the assumption being set up that Mark was derived from Matthew.

Acceptances Of Whole Incidents

The acceptances amount to about 92 per cent, of the total table of contents of Mark. This is so nearly all of the Markan presentation of whole incidents that it suggests that perhaps the writer was dependent for substantially his entire framework upon the First Gospel. We are confirmed in this view by an examination of the seven whole incidents which together make up the remaining 8 per cent. With the possible exception of the Appointment of the Twelve (Mk. 3:13–15), all of the incidents are rather closely linked with parallels of Matthaean events. At all events, we must conclude that the hypothesis of a dependent Mark requires us to conceive the writer either as unable from want of material or literary ability to plan and write a Gospel with a different framework or else as disinclined to do so. If Peter is not brought in, these alternatives appear to create no difficulty. Nor is the situation made difficult, even if we consider the apostle and John Mark as jointly the producers of the Second Gospel from the First. Peter doubtless had other material, material concerned with tours unreported in Matthew, sufficient in amount to have made possible a narrative with a framework decidedly different from that disclosed in Mark and in the Matthaean parallel. But we do not know that he and John Mark together had the necessary literary ability nor that either of them had the ambition to become an original author. The fewness of the added incidents and the almost complete absence of changes and additions to the framework do not encourage us to believe it necessary to attribute any especial literary capacity or desire. In short, whether we view Mark as a compilation made by some unknown persons or as the product of Peter and his interpreter’s activity in abridging the Matthaean table of contents and in supplying additional details and a few fresh incidents, there appears to be no difficulty in understanding that the secondary author or authors produced a book 92 per cent, of whose succession of events and in fact practically the whole of whose framework are derived from Matthew. There is no embarrassment for the hypothesis of a dependent Mark.

It is of importance to note the character of the acceptances. The whole incidents taken over are concerned principally with miracles and with the forward movement of the history. A heavy proportion of the miracles are told with additional details. On the other hand, incidents which in Matthew are narrated with a great wealth of didactic matter are generally omitted altogether—as The Sermon on the Mount—or seriously cut down—as The Twelve sent forth. That the purpose of the secondary writer, when Mark is made a derivative of Matthew, must be conceived as having mainly in view occurrences having to do with the historical advance and events associated with miracles, is confirmed upon noting that the aggregate of Markan incidents unparalleled in Matthew is confined pretty closely to members of these two classes. Of the seven whole incidents added by the Markan writer, three are miracles (The man with the unclean spirit, Mk. 1:21–28; The deaf-mute, Mk. 7:31–37; The blind man of Bethsaida, Mk. 8:22–26), two are concerned with the movement of the history (Tour through near-by places, Mk. 1:35–38; Appointment of the Twelve, Mk. 3:13–15), one is an associated incident (The young man who follows, Mk. 14:51–52), and one only is occupied with teaching (The widow’s two mites, Mk. 12:41–44).

In view of what has been set forth, it can hardly be said that either the number or the character of the Markan acceptances of whole incidents imposes any difficulty upon the hypothesis that Mark was derived from the Gospel of Matthew.

Omissions Of Whole Incidents

Let us now attend to the matter of the Markan omissions of whole incidents. These events have already been listed—in the seventh installment—as Matthaean additions. We are now assuming Mark as dependent, and this requires us, for the time being, to view them as omissions. I eliminate those which are concerned principally with discourse and the two which occur at the end of Matthew beyond the point in the history at which our present Mark terminates.

No. 5 is inclosed in brackets. This may be taken to signify that the propriety of its presence in the table is uncertain. Mt. 15:29–31 and Mk. 7:31–37 are only partially parallel, though in each case the introductory words indicate that the event is to be taken as the narrative which comes next chronologically in the series presented by the particular Gospel.

At most, then, we have eight whole incidents omitted by the Markan compiler. Four of these—Nos. 1, 2, 3 and 6—are miracles described in detail. They are thus of the kind suited to the probable purpose of the Markan writer. And, perhaps, Nos. 7 and 8 may also be consistent with that purpose. As to Nos. 4 and 5, their omission can be explained as in harmony with the omissions of a number of similar compendious references to groups of miracles. That is to say, there are six or seven such Matthaean statements, including Nos. 4 and 5, which are not present in Mark. The whole of these may very well have been rejected by the Markan writer because of the lack of detail.

In fact, whatever difficulty there may be for the hypothesis of a derivative Mark lies in the absence of the four detailed accounts of miracles and of the two passages which deal with the thirty pieces of silver and with the guard—six omissions of narratives of whole incidents. Over against this difficulty, I place a similar one which we have seen to exist when Matthew is made the derivative document and is then found to have omitted three detailed accounts of miracles in Mark together with four narratives of a non-miraculous character. Five of the seven describe undoubtedly the same kinds of events as are elsewhere recounted in Matthew. If it be asked, Why did the Markan writer omit the six?, the question may also be raised, Why did the Matthaean writer leave out the seven?

If it be replied that two of the miracles described in Mark are of the class which may be denned as consisting of miracles performed in association with some physical act, and that their omission in a derivative Matthew is explained as due to an unwillingness of the compiler to include accounts of this class of miracles, then I respond that this point has already been treated. If it nevertheless be said that, despite the fact that the Matthaean writer has been shown willing to describe miracles which were wrought in association with a certain amount of physical activity, two of the Markan miracles (Mk. 7:31–37 and 8:22–26) are involved in so great an amount of such activity that there is a residual difficulty in believing that the Matthaean compiler would accept them, I am ready to acquiesce. Let us grant, then, that it is somewhat harder to believe that the Markan compiler left out the six than that the Matthaean compiler omitted the seven.

At the moment, the matter stands thus: It is somewhat more difficult to view Mark as the derivative document than it is to consider Matthew as such. I propose to ease the difficulty of making Mark the secondary writing. That is to say, I advance the consideration that the Second Gospel may very well have contained, in the autograph, three of the four accounts of miracle. These three are Nos. 1, 2 and 3 of the table. They may have been lost incidentally to a mechanical derangement of the first third of the autograph or an early copy.¹ That these three miracles were narrated in the autograph of Mark is a proposition that cannot actually be proved, but its tenability is such that it cannot be asserted with any considerable assurance that the Markan writer did indeed fail to include narratives of them. That is to say, when it is urged that Mark could not have originated as a compilation, because of the omissions of whole incidents which it would be necessary to assume, it is possible at once to reply that the three miracles with which we are now concerned cannot confidently be included amongst the events omitted. The result is to reduce the list of whole incidents omitted to Nos. 6, 7 and 8. And, it is very uncertain that No. 6 should be considered an event of the kind the Markan writer would wish to include, since it is an incident in which Peter is brought into favorable prominence. Accordingly, the objection against the derivative character of Mark that is based upon the omissions of whole incidents is reduced to very small proportions. Whether No. 6 is included or not in the list of omissions, the resultant objection is more than offset by the argument against the derivative character of Matthew which may be maintained upon the basis of the seven incidents which it would be necessary to view as having been omitted when the Matthaean writer was compiling from the Gospel of Mark.

Additions Of Whole Incidents

Having taken account of the situation created when the Markan writer is conceived as a compiler making omissions of certain whole incidents and having given attention both to the number and to the character of these events, let us now consider the additions of whole incidents which the hypothesis of a derivative Mark requires us to assume. These have already been mentioned more than once, but it will be safe to give a tabulation.

If we take the number of distinct incidents described in the Second Gospel as 83, then the proportion of items in the table of contents that have been added by the compiler is 8 per cent. Now, a programme for the writing of a book that contemplates a list of topics only 8 per cent, of which are new cannot be regarded as the product of a highly original purpose. There is too much asquiescence in what another has planned and wrought. But, even though the programme for the carrying out of the project may nevertheless be a conceivable one, there is no occasion to cry, Why should the Markan writer seek to compose a book which should add so little to what was already available to the reading world? As a matter of fact, the new material added in Mark to the history accepted from Matthew is not 8 per cent., but round about 30 or 32 per cent. Eight per cent, represents the additions of whole incidents—there are in Mark new topics which amount, in number, to 8 per cent, of the whole table of contents. Are these new topics, these new sections, the material that the Second Gospel has added to what it obtained from the First? By no means. These 7 sections which constitute 8 per cent, of the total of 83 sections do represent part of the additional matter the writer of Mark had to tell. But only a part, and a small part at that. There is a great deal more. This additional matter consists of fresh details supplementing the information taken over from Matthew when the Markan writer accepted from Matthew 76 topics and found in Matthew varying amounts of information upon these several topics. The new and supplementary matter which the Markan writer had to add in the narration of events also recounted in Matthew plus the 7 complete accounts now found in Mark, but not in Matthew—this fresh material amounts to something like 30 or 32 per cent, of the whole text of the Second Gospel. Was the fact that he proposed to produce a book 70 or 68 per cent, of which should come from a work already extant and 30 or 32 per cent, of which should be fresh a sufficient reason to warrant a new Gospel? Of course it was. Probably, the Markan writer made no precise calculations—70 and 30, or 68 and 32, were never numbers which he considered in connection with the material which he proposed to take over from Matthew and that which he proposed to add from some other source or sources—say, from the memory of Peter. All the same, whether he formulated a purpose in definite and precise terms or whether he looked at the matter roughly, the new information was considerable in amount. Or, if the plan grew for the most part without direct attention and the purpose to publish came only after he had observed that the new material, consisting in part of fresh supplementary matter concerned with topics already treated in Matthew and in part of accounts of new incidents, amounted to a considerable total, the intention to publish is nevertheless sane and understandable. As a matter of fact, Mark does contain not far from one-third of itself in the form of new information. And whether the Markan writer saw this before composition or after composition, but before publication, or whether he saw it dimly or clearly, the proportion of 30 or 32 per cent, is substantial enough to justify either the composition for publication or the publication after composition. He had a good deal to tell which is not to be found in Matthew. Here is sufficient reason to justify the addition of another narrative to the already available First Gospel.

Nevertheless, the fact that of his entire table of contents only 8 per cent, of the items are new topics makes the Second Gospel one whose purpose is not highly original. And this lack of high originality is seen still more clearly when it is realized that, with the possible exception of the Appointment of the Twelve, the new topics all belong to the framework taken over from Matthew. But high originality is not a necessary ingredient in a reasonable plan and purpose for a book.

Our Second Gospel is, indeed, an abridgment, and this indicates a purpose wanting in high originality. Its plan is the plan of Matthew after certain subtractions are made. The abridgment is limited pretty closely to an account of the public works of the Saviour and to the events of the Last Days; and great amounts of material, partly narrative and partly discourse, having close reference to Jewish matters are omitted. Here we have indications of a purpose beyond that of a mere abridgment. While Mark is a shortened edition of Matthew, it is nevertheless more than just that. The considerable total of eliminations of Jewish things serves to show that Gentile readers were in mind. The purpose underlying the Second Gospel must, accordingly, be conceived as sufficiently different from that upon which the First is based. Mark is an abridgment, unoriginal in structure, containing a large proportion of new material and directed to readers predominantly Gentile. Here, in this one sentence, is a sufficient answer to any who may ask why the writer of Mark should compile his work when the Gospel of Matthew was already available.

Details Of Acceptance, Omission And Addition

Hitherto, our attention has been directed, though perhaps not exclusively, to the matter of whole incidents, their number and character. We are now to consider details. These will consist of the details of acceptance, of omission, and of addition. It will be convenient to alternate between the two hypotheses of a dependent Matthew and a dependent Mark, and it will be our duty to take account of favorable and unfavorable inferences as to these rival hypotheses that may be drawn from the facts as to details.

It is to be noted that details of acceptance are altogether the same, or nearly so, whether we make Matthew or Mark the dependent Gospel. If a word, or a phrase, or a thought is to be viewed in Matthew as an acceptance of a parallel in Mark, under the hypothesis of a dependent Matthew, then when the case is reversed and Mark made dependent the parallel in that Gospel becomes a Markan acceptance. Similarly, a detail of omission assignable to either Gospel under the hypothesis of that Gospel’s dependence becomes a detail of addition for the other Gospel when dependence is assumed for that other document. It will be desirable to bear this principle in mind and thus facilitate the rapid consideration of details now as omissions and immediately afterwards as additions.

In each Gospel, we will have an aggregate of details of parallelism, which may, on the hypothesis of that document’s dependence, be regarded as acceptances taken over with or without modification from the other writing. The parallels become, in turn, acceptances made by the other Gospel when it is assumed to be the dependent composition.

The two aggregates of parallelisms cannot be precisely defined, partly because of uncertainties as to equivalence, partly because of the existence, particularly in Matthew, of doublets. When something occurs twice in one Gospel and once in the other, are we to match the occurrence in the latter document with one or with both members of the doublet in the former? And, if with one member, with which of the two? And sometimes, no doubt, a parallel detail in one Gospel will be an interpolation from the other Gospel of the thing paralleled. Our four Gospels as we know them are, if we except misplacements and some losses through accidents, doubtless in substantially the condition in which they left the hands of the several authors. At the same time, there have certainly been interpolations from one to another which textual criticism, in consequence of the lack of MSS. of the highest antiquity and perhaps because of other reasons, has been unable to remove. In view of considerations such as these, we can scarcely hope so to match up the two Gospels as to determine with exactitude the two aggregates of text, each of which duplicates the import of the other.

But we can get an approximate view. And there is some importance in doing so, for the reason that a substantially correct realization of the amount of duplication assists us in comprehending the extent of the dependence between the two Gospels, once we have assumed that one or the other is a derivative of the remaining document. Let us insist, then, upon ascertaining what proportion of Mark is contained also in Matthew. On the other hand, let us not be misled. The mere fact that Mark is largely duplicated in Matthew tells us nothing as to priority. Even if the proportion of the Second Gospel paralleled in the First amounted to 100 per cent., that would not inform us which document is parent and which offspring. Grant that the whole of Mark is to be found again in Matthew—what of it? It may mean that the writer of the First Gospel used everything contained in the Second. But, it may just as well mean that the compiler of the Second did nothing more than abridge the First. There is, in fact, nothing substantial and direct as to priority to be gained by following up this matter of the proportion of Mark to be read again in Matthew.

Errors Of Fact And Interpretation

Prominent writers on the Synoptic Gospels have, however, committed themselves to a different statement of the case. It seems fairly clear that they have grasped

1. Neither the facts

2. Nor their significance

But they knew some of the facts and part of the significance. It is understandable that from incorrect data and unscientific procedure a wrong result might emerge. I venture the foregoing statement despite the confidence, expressed by some, that the hypothesis which derives Matthew from Mark has adequately been established. I give two excerpts, one from H. J. Holtzmann and the other from Adolf Harnack.

“Two results have been realized. In the first place, this, that if there is a common source at all it can only be conceived as either identical with our second canonical Gospel or much more similar to it than to the two others.

“This principal result of critical labor has, after a transitory prelude at the close of the eighteenth century, required about sixty years for its establishment; but in the course of this period it has been elevated to such a firm position that, at least when expressed with such generality, the Markan Hypothesis is no longer to be called an hypothesis.” H. J. Holtzmann, Hand-Commentar turn Neuen Testament. Die Synoptiker (3 Ausgabe, 1901), S. 15 (§6).

“The Markan problem, which has been treated with scientific thoroughness.” A. Harnack, The Sayings of Jesus (translated from the German) (1908), Introduction, p. xiii.

That this confidence is quite unwarranted may be gathered in part from what follows. I raise first a question in connection with the total amount of duplication between Matthew and Mark—What proportion of Mark is to be found again in Matthew?

The modern discussion of the Synoptic Problem is now about one and one-half centuries old. One would think that within this very long period someone at some time would have established a fairly accurate statement of the facts, and that his results would be well known to writers engaged in the investigation of Synoptic matters. I have at hand a very recent work, The Four Gospels, a Study of Origins (1925) by B. H. Streeter. Let me quote from this book:

“Matthew reproduces 90 per cent, of the subject matter of Mark in language very largely identical with that of Mark.”—B. H. Streeter, The Four Gospels (1926), p. 151.

“The authentic text of Mark contains 661 verses. Matthew reproduced the substance of over 600 of these.”—Ibid., p. 159.

“Matthew’s supposed omissions from Mark shrink on examination to very small dimensions. Matthew reproduces the substance of all but 55 verses of Mark.”—Ibid., p. 169.

Are these statements even roughly in accordance with the facts? Let us see. It will be sufficient for present purposes to tabulate merely the whole verses in Mark which are unparalleled in Matthew. The inclusion of portions of verses would nearly double the amount of text. Let the unparalleled whole verses suffice for the time being. The table is not to be regarded as necessarily exhaustive.

The tabulation shows that the material in Mark which is unrepresented in Matthew amounts to a much different total than that set forth by Dr. Streeter. One hundred and fifteen unparalleled verses do not constitute, by any means, the whole of the narrative and didactic matter present in Mark but absent from Matthew, as the whole would include the many unparalleled fragments of verses; but taking 115 verses as the amount, we will readily note that this total is more than double one of 55 verses. If Matthew be assumed as the derivative document, then it has absorbed a good deal less than the whole substance of Mark; and if Mark be viewed as the secondary writing, then the compiler included a very considerable amount of fresh material.

But Dr. Streeter does not appear to be the only one who has—what shall I say?—deluded himself or been deluded by others. So far as I know, the question as to what proportion of Mark is to be found again in Matthew has always been, not without an answer, but without a correct or reasonably correct answer.

Consider the following:

“This [dependence of Matthew upon Mark] is proved not only through the fact that the entire contents of it [i.e., of the Second Gospel], with insignificant, easily explained exceptions, have been taken over into the First [Gospel] . . .”—B. Weiss, Die vier Evangelien im Berichtigten Text (1900), Einleitung, S. 2.

If it be said that probably Dr. Weiss has in mind the fact that of the topics treated in Mark only a few are without representation in Matthew, then I ask, Where is the recognition of the heavy total amount of Markan details which are without Matthaean parallels? Where is the evidence that Dr. Weiss was correctly informed as to the extent of non-Matthaean matter in Mark? The argument for the dependence of Matthew which he briefly states in the excerpt requires, on the assumption that it is a valid argument, that there should be but little in Mark that is not in Matthew. Apparently, he entertained the delusion that the fragmentary matter occurring in Mark as unparalleled details was inconsiderable in amount. As a matter of fact, it is five times as great in amount as the total of material occurring in the form of unparalleled narratives of whole incidents.

This matter of the amount of the “absorption” of Mark was under consideration a good while ago, as may be gathered from the following excerpt from the same Dr. Weiss, which dates back about half a century:

“Accordingly, with the exception of some very small portions, whose omission is very simply explained, the entire contents of Mark have been transferred to our Gospel [Matthew] and in fact in the same order and grouping.”—B. Weiss, Dan Matthausevangelium und seine Lucas-Parallelen (1876), S. 13.

And the same idea is expressed in still another work from the same pen:

“As a matter of fact the entire substance of Mark is taken into our First Gospel even down to small unimportant parts whose omission explains itself on the clearest grounds.”—B. Weiss, A Manual of Introduction to the New Testament, vol. ii (1889, from the German), p. 263.

I think it would be very hard to maintain, in the face of such statements, that Dr. Weiss had any conception of the true state of affairs. But, let us turn to an English writer who has become prominent in Synoptic discussion. I give the following:

“Almost the entire substance of the second Gospel has been transferred to the first. The only omissions of any length are the following.” He then proceeds to enumerate these passages: Mk. 1:23–28; 35–39; 4:26–29; 7:32–37; 8:22–26; 9:38–40; 12:41–44.— W. C. Allen, A Critical and Exegetical Commentary on the Gospel according to S. Matthew (1907), Introduction, p. xiii.

These Markan portions may be regarded as whole incidents. They total but 33 verses. Naturally, this result is far short even of the aggregate of 115 verses, which itself is considerably less than the whole. What can be said of this writer’s conception of the proportion of Mark that is to be found again in Matthew?

But, let us come on down to a point of time sixteen years later. I subjoin the following from a Central European writer:

“In the case of Matthew, the result is such that he [Matthew] presents almost the entire Markan material, with the exception of eight short sections, between 3:1 and 28:8.”—Rudolf Knopf, Einführung in das Neue Testament (1923), S. 105.

Does this writer set forth anything but a seriously misleading view of the proportion of Markan material to be found again in Matthew?

I have now reviewed a period of nearly fifty years— 1876, 1889, 1900, 1907, 1923, 1925—giving excerpts from such writers as B. Weiss, Allen, Knopf and Streeter— but have found nowhere any sufficient appreciation of the proportionate part of Mark which duplicates Matthew. All appear to be under the delusion that it is nearly the whole of Mark that is involved in the duplication. This conception is absurd, in view of the fact that a good deal more than 115 verses out of 661 are not thus involved. About a half century of study and an absurd result as to a mere matter of easily ascertainable fact!

Let us now turn our attention to the question as to whether the “facts” have been understood in respect to their significance. The proposition that the material of Mark is nearly all to be read again in Matthew has been considered as affording a substantial basis for claiming the dependence of Matthew upon Mark. Thus, Dr. B. Weiss, in a passage to be found a page or so back, states that one of the proofs of this dependence consists of the reappearance in the First Gospel of nearly the entire contents of the Second. This passage dates from 1900.

A quarter of a century later, Dr. Streeter gives five reasons for accepting the priority of Mark. With him, “priority” signifies priority over both Matthews and Luke. He gives as the first of the five reasons the following:

“(1) Matthew reproduces 90 per cent, of the subject matter of Mark in language very largely identical with that of Mark; Luke does the same for rather more than half of Mark.”—B. H. Streeter, The Four Gospels (1925), p. 151.

Presumably, the argument here is a double one: (1) Mark is prior to Matthew because the latter contains 90 per cent, of the former; and (2) Mark is prior to Luke because in the Third Gospel is to be found more than half of the Second.

It may well be asked, in view of the passages from Weiss and Streeter a quarter of a century apart, wherein is the cogency of the argument. In what way is Mat-thaean dependence shown, once we grant the proposition that nearly the entire contents of Mark are contained in Matthew? How is it that the premise which asserts that 90 per cent, of Mark is to be found again in Matthew brings us to the conclusion that the Second Gospel was written prior to the First?

Apparently, the logical attitude of these investigators may briefly be stated as follows:

1. Mark is a simple abridgment, or else it originated prior to Matthew.

2. A simple abridgment is something impossible.

3. Therefore, Mark originated prior to Matthew. Attention has, in effect, already been directed to the fact that it is not permissible to regard the Second Gospel as a simple abridgment. It contains too much information absent from the exemplar. At the moment, however, I propose to let proposition No. 1 stand, and to consider proposition No. 2. Has anyone ever established this statement? I know of nothing of the kind. And I am quite sure that no one else knows of such a demonstration, either by Dr. Weiss, Dr. Streeter or any other writer, whether engaged upon Synoptic matters or not. I do not propose, at this juncture, to enter upon a discussion as to whether there are or are not now extant examples of simple abridgment which have come down to us from ancient times. The thing as to which I am concerned is the emphasis of the logical duty of those who wish to urge the priority of Mark. It appears to me that it is incumbent upon them to establish the proposition which asserts the impossibility of simple abridgments. This proposition is by no means axiomatic. Proof is necessary for its establishment. In the meantime, I throw back into the lap of those who maintain Markan priority and Matthaean dependence, upon the basis of the “absorption” of Mark, the inference that their conclusion would be warranted by the truth of the proposition that Mark consists almost entirely of Matthaean material.

There has been, however, a disposition in some quarters to see importance in the fact that the Second Gospel contains but a small amount of material that cannot be found in the combined texts of the First and Third Gospels. One leading writer of those who advocate the dependence of Matthew sees in the smallness of the amount of peculiar matter in Mark a criterion which limits the solutions of the Synoptic Problem to two alternatives—the Second Gospel prior to or subsequent to the First and Third. Later statements by other advocates of the secondary character of Matthew conjoin the almost complete duplication of the Markan material in the consolidated texts of Matthew and Luke with their agreements with the order of Mark and view the resultant combination as affording a substantial argument for the dependence of the remaining Synoptic Gospels upon the Second or for the limitation of the Synoptic Question to the two alternatives mentioned. The writers referred to are H. J. Holtzmann, P. Wernle and A. J. Maclean. I excerpt from a work by the first the following statement:

“The entire material of the Second Gospel, within about 30 verses lost in the combined text of the whole, is to be found again in the two other works. This indicates either an excerpt (the Griesbach Combination Hypothesis) or the common exemplar (the Markan Hypothesis).”—H. J. Holtzmann, Hand-Commentar zum Neuen Testament, Die Synoptiker (3. Aufl., 1901), S. 6.

The significance seen by Holtzmann in the fact that Mark contains very little that is to be found neither in Matthew nor in Luke is to the effect that the smallness of the amount of matter peculiar to Mark means that the possibilities are cut down to just two alternatives— either Mark is a composite excerpt from Matthew and Luke or else the Second Gospel was exemplar for both of the others.

This is a gross blunder. If the statement were true, it would be exceedingly important, because in order to prove that Mark was a parent writing for both the other Synoptic Gospels it would then only be necessary to find a criterion competent to exclude the hypothesis that Mark is a composite excerpt from Matthew and Luke. However, the statement of Holtzmann is by no means a true one. Apparently this writer was unequal to the task he had in hand and perhaps he did not know, in particular, just how to go about discovering the remaining explanations of the fact that Mark contains only a small amount of peculiar material. That there are other explanations may be perceived by taking into account the following considerations. If we assume that Mark was compiled from Matthew and that Luke in turn was compiled from Mark, we will be able to supply an explanation. We have only to suppose that the Lukan writer accepted from Mark the bulk of the fresh information added when this Gospel was compiled from Matthew. The small residue unaccepted by the author of the Third Gospel would then constitute the peculiar material of Mark. Another explanation may be seen when the order of derivation in this one is reversed and Mark is understood to have been compiled from Luke and Matthew from Mark. If now we assume that the Matthaean writer took over from Mark pretty much all that the Second Gospel added to the Third when this Second Gospel was compiled, there will remain of the Gospel of Mark only a small total of material peculiar to itself. That Mark contains 30, or 50, or 70 verses represented neither in Matthew nor in Luke may accordingly be readily explained by either of the following assumptions: Matthew parent to Mark and Mark parent to Luke or Luke parent to Mark and Mark parent to Matthew. Whatever measures, if any, Holtzmann took to discover additional hypotheses capable of accounting for the small amount of material peculiar to Mark must have been very unproductive, as he does not seem, in this connection, to have uncovered these two.

In the conduct of investigations in accordance with scientific procedure—that is to say, under regulations which have themselves been rigorously established—it is often necessary to develop not some but all of the alternative hypotheses in order to exhaust the possibilities. The object here in view is to continue the investigation by a process of exclusion and thus eventually effect a proof of the truth of that particular one of the alternatives which remains. In carrying out this indirect method of proof, it is essential to develop alternatives until it is ascertained conclusively that there are no further possibilities. Investigators of the Synoptic Problem who are not alive to these elementary principles of logic are unfitted, however learned they may otherwise be, to prosecute successfully the work they have in hand. Sometimes, in the development of alternative explanations, even those who are relatively competent fail to exhaust the possibilities, but their competence will appear in the fact that they recognize the incompleteness of their results and that they then exercise reserve in what they claim.

In view of the foregoing discussion it would appear that the Holtzmann criterion is quite unimportant, having little or no selective power. In fact, if we are going to solve the Synoptic Problem, it will be necessary to discover facts of far more significance than the comparatively negligible consideration that almost the whole of the Markan matter may be found again in the combined text of Matthew and Luke.

I proceed to give excerpts from the two remaining writers already mentioned. Here is the first:

“When two authors [as the authors of Mt. and Lk.] present almost the whole matter of a third [the author of Mk.] in the order which this third writer has given to it, the most convincing proof is thereby furnished that this third author is a source for the other two.”—P. Wernle, The Sources of Our Knowledge of the Life of Jesus (from the German, 1907), p. 68.

That this is fallacious reasoning, despite the enthusiastic “most convincing proof,” may be seen by observing that there are at least four alternatives—Mt. and Lk. the source of Mk.; Mk. a source for Mt. and Lk.; Lk. a source for Mk., and Mk. in turn a source for Mt.; Mt. a source for Mk., and Mk. in turn a source for Lk. Why give the prize to No. 2?

The remaining excerpt is as follows:

“Except about thirty verses all the narrative that we find in Mk. we find also (and in the same order) in either Mt. and Lk., or in both. This might tell both ways.” [l. e., Mk. either a source for the others or else a product of those others.]—A. J. Maclean, A Dictionary of Christ and the Gospels, vol. ii (1908), article “Mark, Gospel according to,” p. 125.

We find here the argument that the small total of matter peculiar to Mark justifies the inference that Matthew and Luke are the source of Mark or else Mark is a source for Matthew and Luke. There are four or more alternatives. Four have just been enumerated. So that we are not entitled to conclude that there are just two—Nos. 1 and 2.

The reader may, perhaps, be ready to conclude that neither Holtzmann, Wernle nor Maclean is a safe reasoner in connection with the significance properly to be given to the fact that Mark contains but little that cannot be found in the others, whether the matter of order is taken into account or disregarded. If so, I think there is ample justification for such a conclusion.

What Are The Facts As To The Non-Matthaean Content Of Mark?

So far, our investigation has not developed the exact total amount of material contained in the parallel aggregate of the first two Gospels. As to the proportion of the Matthaean text whose import is paralleled in Mark, it does not appear necessary to determine it with any great precision. It is probable that something like 40 per cent. of the First Gospel is duplicated in the Second. However, it is fairly clear that it is unimportant to know the percentage with exactness. Forty per cent, of Matthew paralleled in Mark means on the hypothesis of direct derivation, if we disregard doublets and the like, that the Matthaean author added fresh material to the amount of 60 per cent, of his Gospel or that the Markan writer omitted 60 per cent, of what he had before him. Even so, the situation is substantially the same, if instead of 40 per cent, of Matthew, we substitute 35 or 45 per cent, of this book. The Matthaean compiler would then have added an aggregate of new matter equal to 65 or 55 per cent, of his entire work or the Markan secondary writer would have omitted 65 or 55 per cent, of the material in his exemplar. On the other hand, when we turn our attention to Mark, it becomes highly desirable to ascertain, more precisely than has hitherto been done in our inquiry, the proportion of this Gospel which is to be found again in Matthew. It has already been shown that the amount of unparalleled material is not less than 115 verses out of a total of 661. As this summation of the unduplicated text takes no account of fragments of verses, it is too small. We are, accordingly, not permitted to regard the difference between 661 and 115 verses—that is, 546 verses—as the amount of Markan text paralleled in Matthew. This statement would be much beyond the truth. The proportion of Mark which parallels Matthew is by no means so large as 546/661—that is, 83 per cent. To some, it has seemed important to claim that pretty much the whole of Mark reappears in Matthew, or— what is substantially the same thing—that Mark contains but little that is not to be read again in the First Gospel. In view of the fact that some think they see in a small relative amount of independent material in the Second Gospel evidence of the dependence of the First, it is desirable to ascertain with some exactitude what proportion of Mark is unparalleled in Matthew. The amount has been minimized by one author and another. We have already seen that B. H. Streeter regards the fresh matter as totalling no more than 55 verses. It may not be inadvisable to mention B. W. Bacon in this connection. In An Introduction to the New Testament (1900 to 1924), he appears to think that if we view Mark as a derivative from Matthew we must conclude that the fresh material consists merely of “a few unimportant embellishments and additions.” He says:

“But what could be more absurd than an evangelist who attempts to improve upon a gospel already current by simply extracting a part of its contents and that part the smaller and less authoritative, and touching it up with a few unimportant embellishments and additions! Could he imagine that his readers would prefer plagiarising extracts to the rich source from which he drew? If his source were Matthew, either he would not have written at all, or he would have re-edited the larger work; most of all, if it was then regarded as the work of an Apostle.”—B. W. Bacon, An Introduction to the New Testament (1924), p. 181.

Such a statement as this could hardly be made by one who knew the facts—by one who knew that the additions of new material totalled up to 30 or 32 per cent, of the entire Gospel of Mark as we now have it. The Second Gospel was not written by an author who had no new material. The purpose of one who proposes simply to abbreviate is quite understandable; and, in the present case, it is not at all difficult to understand that John Mark may very well have wished to compose a Gospel for Gentile readers, which should omit the introductory matter of the first two chapters of Matthew and the greater part of the discourse material. If his purpose had stopped there, it would be understandable. But, when we add the intention of expanding the accounts of most of the miracles and of otherwise making additions to the history and discourse, we have a still more comprehensible purpose, especially if we take into account the fact that the fresh material was so considerable in amount that, when the book was finished down to 16:8, the new matter totalled 30 or 32 per cent, of all that had been written.

At another point in the same work, Bacon again assumes that the material in Mark unparalleled in Matthew is quite inconsiderable. He says:

“It being insupposable that the author of Mark should have composed his work by merely subtracting the narrative element of Matthew.”—Ibid., p. 187.

I am not at all willing to grant that it is highly improbable that a secondary write would not have a thoroughly good purpose even through his object were merely to excerpt narrative and decline discourse. At the same time, it is unnecessary to stress this point. The book of Mark cannot be regarded, by those who know the facts, as a transcript of the narrative portions of Matthew. The independent material is too considerable in amount.

I propose to tabulate the portions of Markan text that are, upon the hypothesis of a Matthew derived from Mark, to be regarded as omissions made by the compiler, and upon the assumption of a Mark dependent upon Matthew, as additions made by the secondary writer. The table lists textual matter which gives, in verses and fragments of verses, various details. The unparalleled passages and aggregates are all estimated in size by the number of contained Greek words, WH text. Thus, in the portion Mk. 1:1–8, verse 1 entire is unparalleled in Matthew, and there are fragments of verses 4 and 7 which are also unparalleled. Altogether, the unparalleled text is estimated as consisting of 13 Greek words.

The total of unparalleled material in Mark is, accordingly, an aggregate of 3,579 words of text. This aggregate represents, if we assume Matthew an immediate derivative of Mark, the omissions, whole incidents and fragments, made by the Matthaean compiler. Why did he reject so considerable an amount of informative matter? The incidents and parts of incidents are, on the whole, just the kinds of things favored by the writer of Matthew. There is some discourse and a great deal of miracle. To say that he was seeking to save space is to say something unconvincing. What if the First Gospel had included, in addition to its 18,000 or 19,000 words, these 3,579? The total, even so, would have amounted only to a book of say, 22,600 words. It would still have been a shorter document than is any one of half a dozen writings of the LXX.

There are many notices, occurring in Mark but not in Matthew, which supply information as to geographical details, which give numerical precision, which mention the names of persons directly or indirectly involved and which furnish details of descriptive value. I refer to small additions present in the Second Gospel but not in the First. These cannot be explained as omissions made by a Matthaean compiler for the purpose of saving space.

I begin by calling attention to the following geographical notices. It is in Mark, not in Matthew, that we learn that Jesus came to His baptism from Nazareth (1:9); that we find it stated that those who identified the Saviour’s power over demons with that of the prince of the demons had come down from Jerusalem (3:22); that we note that, immediately after the feeding of the five thousand, the disciples are directed not merely to cross the Sea of Galilee to the other side but to proceed to Bethsaida (6:45); that we are informed that, as the Saviour departed from the district where He met the Syro-phoenician woman, He went to the Sea of Galilee by way of Sidon and that He also went through the midst of the borders of Decapolis (7:31); that we are told that, upon the final journey to Jerusalem, the Saviour and those with Him drew near to Bethany as well as to Bethphage (11:1). I pass now to Markan additions and substitutions which give numerical precision to matters less closely defined in Matthew or not defined at all. It is in the Second Gospel, not the First, that we read that the great herd of swine consisted of about two thousand head (5:13); that we learn that the daughter of the ruler of the synagogue was twelve years of age (5:42); that we find it related that the disciples were sent forth two by two (6:7); that we get the information that Herod swore to honor the dancer’s request to the half of his kingdom (6:23); that we get an estimate, made at the time, of the value of the amount of bread necessary for the feeding of the five thousand (6:37); that we find it described how the multitude sat down by fifties and by hundreds (6:40); that we are informed not merely that the disciples forgot to take a supply of bread but that there was in fact only a single loaf in the boat (8:14); that we obtain a statement that the value of the ointment exceeded three hundred denarii (14:5); that we read that the number of disciples who were sent to prepare the Passover was two (14:13); that we learn that Peter’s denials would be complete before the cock would crow twice (14:30) and that the cock-crowing which did immediately follow the third denial was the second (14:72); and that we are told that the Crucifixion occurred at the third hour (15:25).

Again, in the matter of names, it is the Markan writer, and not the Matthaean, who tells us that the house where Peter’s mother-in-law was cured was the house of Andrew as well as the house of Peter and also that James and John accompanied the Saviour as He entered into the home where the sick women lay (1:29); who informs us that Levi, the publican, was a son of Alphaeus (2:14); who narrates that the Saviour gave to the sons of Zebedee the surname Boanerges, and at the same time interprets to us its meaning (3:17); who tells us not merely that the father of the little girl who was at the point of death was a ruler but that he was one of the rulers of the synagogue and that his name was Jairus (5:22); who mentions the names of Peter, James and John as those of persons who accompanied the Saviour into the house of the ruler (5:37); who states that one of the blind men whom Jesus cured as He left Jericho was a beggar and that his name was Bartimaeus (10:46); who informs us that the names of the disciples who came to the Saviour as He sat on the Mount of Olives and desired to know when the Temple would be razed to the ground and what would be the sign were Peter, James, John and Andrew (13:3); who tells us not merely that it was Simon of Cyrene whose service was commandeered in the matter of bearing the cross, but also that he was the father of Alexander and Rufus (15:21); and who informs us that, when the women at the tomb were directed to give the message to the disciples that they should see the Saviour in Galilee, Peter was especially singled out (16:7).

Then there are a number of instances where the Markan text gives an additional piece of information of descriptive value. Thus, we find the Second Gospel stating that the place where the great herd of swine were feeding was towards the mountain (5:11); that the grass upon which the five thousand sat was green (6:39); and that the colt upon which Jesus was to ride on His entrance into Jerusalem was tied and that it had never yet been ridden (11:2; see also 11:4).

The foregoing flood of short but informative notices constitutes a serious embarrassment for the hypothesis which makes Matthew a derivative of Mark. Why should the Matthaean compiler disregard the considerable total of important information? We cannot very well answer that he was seeking to save space, since the amount of text necessary in order to include the entire import of the whole list would have been well nigh a negligible matter. It will be well for those who are going to maintain the derivative character of Matthew to face this problem, and see what they can do in the way of discovering an answer to the question, Why should the writer of Matthew omit so many valuable notices?

But, as a matter of fact, the foregoing small notices are by no means all there are. There remain twenty-eight —perhaps more—that may be classified as chronological and circumstantial. I give the twenty-eight in tabular

form, using the renderings of the American Revised Version.

It appears to be more or less characteristic of the Markan style to use double determinations of time and place—to define broadly and then more precisely. I cite: Mk. 1:32; 4:35; 5:2; 6:2; 6:54; 8:10; 9:33; 14:30; 16:2.

Upon comparing Mk. 1:32 and its parallel Mt. 8:16, it may be noted that both have identically the same Greek expressions back of the renderings “when even was come” and “at even,” and that in Mark there is added a more precise determination of the time consisting of the Greek clause rendered “when the sun did set.” That the word rendered “even” did not always define the time with precision is suggested by a comparison of Mt. 14:15 and 14:23. In these passages, the word is used both for a point of time preceding the feeding of the five thousand and for a point of time subsequent to that event. If, as seems probable from Mk. 1:21, 23, 29 and 32, what the Markan writer had in mind was to define the time in 1:32 in such manner as to exclude the thought that the sick were brought to the Saviour while a Sabbath was still continuing, then we can readily understand why he should add to the Matthaean notice of the time words which would put the matter beyond doubt. There was, in Matthew, no necessity for a precise determination of the time, since the Sabbath had not been mentioned.

However, in all the instances listed, the Markan text presents a more precise definition of the time or the circumstances. So, then, even though no necessity requires, it is not easy to understand why, if the Matthaean writer had Mark before him, he so often rejected the exact notices in his exemplar. The text of Matthew is characterized by frequent indications of time and circumstance. And this is the basic reason for difficulty in viewing the list as consisting of Matthaean omissions. There is no objection to the instances when they are conceived as Markan additions, since it would be natural for a writer to use any information he possessed that would enable him to give further precision to the movement of the history.

Reference has now been made to a very considerable aggregate of small notices present in Mark but absent from Matthew. They have been classified as follows:

How are we to reconcile the presence of these 57 notices in Mark and their absence from Matthew, when we assume the latter document as the secondary one? In fact, these notices stand as a substantial objection to the hypothesis of a derivative First Gospel. It will be fair, naturally, to consider what may be said in connection with similar notices which inquiry may show are to be found in Matthew but not in Mark; and this matter the reader may expect to be treated in due course. In the meantime, let him ponder the fact that, when we assume a Matthaean compiler as having omitted 57 such notices, we are taking a step hard to justify.

On the other hand, it is easy to allow that a sufficiently informed Markan compiler would naturally be quite ready to add such notices to the information obtainable from his exemplar. It is unnecessary to cast about for a suitable person who might have been informed as to the foregoing details. Ancient tradition tells us of Mark, the interpreter of Peter, and the NT itself (1 Pt. 5:13) gives support to this assertion. In fact, the maintenance of the view that the Second Gospel was derived from the First is much facilitated by the ease with which the 57 Markan additions may be accounted for once we accept the testimony of antiquity that Mark, an associate of Peter, was the writer of the book assumed to be derivative.

And, let it be particularly noted, these 57 small notices, constituting as they do a body of highly informative material, are to be regarded as evidence of what may very well have been an important reason why a Markan compiler should produce his compilation. Equipped with such a mass of significant details, he would be ready not merely to make an abridgment for Gentiles but to supply important additional information.

Viewing the total aggregate of material which must, under the general assumption of direct derivation, be considered as omissions made by a Matthaean secondary writer or as additions supplied by a Markan compiler, we have some 3,579 Greek words. Since the entire book of Mark as we now have it—that is, from 1:1 to 16:8— contains but 11,001 words,² the proportion of this document unparalleled in Matthew is to be estimated as 3,579/11,001, or as 32:5 per cent. There may be some difference of opinion at this and that point; so that it may be well for me to state the matter conservatively. I say, then, that a Matthaean compiler omitted 30 per cent, of the material before him; or else a Markan secondary writer added enough new material to account for 30 per cent, of his whole document.

We have already found that this body of matter, whether omitted or added, contains an aggregate of 57 small notices of a highly informative character. I proceed now to call attention to the fact that a large proportion of the remaining omitted or added material consists of words of the Saviour. That is to say, in the book of Mark is a total of unparalleled direct discourse, the Saviour being the speaker, amounting to 876 words—7.9 per cent, of the entire work as we now have it. These 876 words are, roughly, one-fourth of the 3,579 words omitted by a Matthaean compiler or added by a Markan. Why, on the assumption of a derivative First Gospel, would the writer omit so much direct discourse of the Saviour? It is easy

to answer the question as to why the Markan compiler, on the alternative assumption of a derivative Second Gospel, would add such precious material. That there are some 876 words of the Saviour’s direct discourse to be found in Mark but not in Matthew may be confirmed by the reader. I supply a tabulation exhibiting, for single verses and small groups of verses throughout the book, the number of words present in the Second Gospel, but absent from the First.

He who maintains that Matthew was compiled from Mark must now face the fact that his hypothesis brings him up against the improbability, not only that the compiler omitted a number of incidents, but also that he failed to include 57 notices of an informative character and a large total, amounting to 876 words, of the direct discourse uttered by the Saviour.

It will scarcely be possible for competent and informed experts in the future to claim that “the entire contents of the Second Gospel, with easily explained exceptions, have been taken over into the First,” nor that “Matthew reproduces 90 per cent, of the subject matter of Mark.” I am right now exploding this conception of the relation between the two Gospels.

On the other hand, the facts which, at the present moment, are being brought to light will serve to facilitate the maintenance of the view that, if we are going to consider that one or the other of the two documents was derived from the remaining one, Matthew is the independent writing and Mark the dependent. All this information as to the 57 small notices and the 876 words of the direct discourse of the Saviour tends to clear up the matter of purpose when John Mark is conceived as a compiler working with Matthew before him.

We have seen how Prof. Bacon exclaims when he contemplates a Mark derived from Matthew and says: “But what could be more absurd than an evangelist who attempts to improve upon a gospel already current by simply extracting a part of its contents, and touching it up with a few unimportant embellishments and additions!” Let me reply with the remark that if, in face of the evidence which has now been adduced, some future writer shall echo this exclamation, then absurdity will indeed exist. However, it will not be the Markan writer conceived as a compiler from Matthew who will be absurd, but the expert who nevertheless imagines that this compiler had nothing fresh to supply beyond “a few unimportant embellishments and additions.”

In concluding the present installment, it is pertinent to add that the embarrassment of the hypothesis of a derivative Matthew and the facilitation of the conception of a secondary Mark will be still further emphasized by considering the facts that may be brought forth relative to other details contained in the 30 or 32.5 per cent, of Mark that is unparalleled in Matthew.

(To be continued)

1) That a mechanical derangement of the first third of Mark occurred in ancient times is a thoroughly tenable proposition, as may be learned by an adequate study of my discussion of The Order of Events in Matthew and Mark, published in two instalments in Bibliotheca Sacra for April and July, 1922. On p. 145, April issue, is to be found the average unit amount of text that must be assumed as on each column portion of parchment or papyrus, in the case of a roll, or on each leaf, in the case of a codex. This was ascertained to be 111.6 words. It was also noted on p. 146, at the head of the third column of the table there printed, that the unit amount is to be understood as having varied from 101.3 to 129 words. Accepting these data, let us proceed to consider the possibility of the presence, in the very earliest MSS. of Mark, of the three accounts of miracle now found in Mt. 8:5–13; 9:27–33[34]. A principal matter to be determined concerns the probable size of the two blocks of text to be assumed as the Markan parallels of the Matthaean narratives. The textual space employed in Matthew is our guide. Mt. 8:5–13 occupies in WH, if we omit the bracketed word in verse 9, the space of just 165 words. However, there is considerable matter having reference to Jewish things which is to be omitted. The terminus ad quern is at the end of verse 12. But as to the terminus a quo there is room for difference of opinion. If we begin the elimination immediately after the Greek word translated “marvelled,” the amount of text to remove is 58 words, but if we commence with the first word of verse 11, the amount is 43 words. Accordingly, the block after the removal of the Jewish material would be 107 or 122 words. The Markan reproduction of these might be assumed as slightly in excess; but, even so, we have amounts between 101.3 and 129 words, which we have accepted as limits. The second block counts, if we omit the 34th verse which is bracketed by WH, of 100 words. Allowing a slight increase for the assumed Markan parallel, we have a proper amount of textual matter. And, if we elect to add verse 34, this will cause an increase of but 12 words, and so will not alter the situation. It may be concluded from the foregoing that it is quite possible to explain the original occurrence of the accounts of the three miracles in Mark and their subsequent loss through an accident.”

2) This is, in effect, the result obtained, after a slight correction, by Mr. Ralph Kamenoff and by my son, Mr. Arthur Springer. It has been corroborated from other sources. The figures 11001 represent the number of Greek words in WH’s recension of Mark when all bracketed words are omitted. There are six cases of elision (1:35; 5:28; 6:56; 9:30; 12:4, 5). In all of these, both words are counted.