The Synoptic Problem

Facts and Conclusions as to the Synoptic Orders of Events

THE reader will perhaps be greatly astonished at the state of affairs I am now about to disclose. I refer in particular to the argument for the priority of Mark that has been based on a comparison of the progressions of events in the three Synoptic Gospels. One would naturally expect that an adequate and correct statement of the facts as to agreement and disagreement in respect to order would long ago have come into existence. Interest in this matter of order dates back to the time of Lachmann,¹ but so far as I know no sufficient and true statement of the facts has ever been made. One would also expect that a rigid inquiry would have been made into the logical support properly belonging to the inference as to priority that has been based on the matter of order. So far as I know, this has never been done.

The reader may think that this is an incredible state of affairs. It ought to be incredible. Of course, it is very unscientific to go ahead without making sure of the facts. But let us see how things stand. I begin by adducing statements as to the fundamental facts with respect to order, and then proceed to show that even in such an elementary matter they are incorrect. The statements begin in 1882 and extend to 1916. Their import is to the effect that the Markan order is always supported, when one of the other Synoptic Gospels deviates, by the remaining document.

[Commending the Griesbach Hypothesis, Dr. Davidson wrote: ] “The following positions seem safe. . . .

“5. Mark’s arrangement is always the same as that of Matthew or Luke.” Samuel Davidson, An Introduction to the Study of the New Testament, 2d ed. 1882), vol. i, pp. 549, 551.

[Dr. Salmon, in an excerpt given later in connection with the argument for priority based on corroboration, states in effect that in every case where the order of whole incidents is the matter in hand and where two Synoptic orders agree against the third, the Markan order is one of the two.]

“The order of the whole of St. Mark, excepting of course what is peculiar to that Gospel, is confirmed either by St. Matthew or St. Luke, and the greater part of it by both.”

“. . . no portion of St. Mark of which the order can not be traced either in St. Matthew or St. Luke.” F. H. Woods, Studia Bibliea, vol. ii (1890), The Origin and Mutual Relation of the Synoptic Gospels (1886), pp. 62 and 63.

“It may be taken as a prima facie argument in favor of St. Mark’s order that it is ‘confirmed either by St. Matthew or St. Luke, and the greater part of it by both.’² Moreover, when one of the other Synoptists strikes out a path peculiar to himself, his order usually has less verisimilitude, and is open on internal grounds to suspicion.” H. B. Swete, The Gospel According to St. Mark (1902), Introduction, p. lxx.

“Luke sometimes deserts the order of Mark, and Matthew often does so; but in these cases Mark is always supported by the remaining Gospel.” F. C. Burkitt, The Gospel History and its Transmission (1906), p. 36.

“The common order is S. Mark’s order, and when one [of the other Synoptic Evangelists] departs from it, the other follows it.” J. R. Cohn, The Gospels in the Light of Modern Research (1909), p. 213.

“Next, as to the order in which this matter is arranged, we find S. Luke sometimes, and S. Matthew rather more frequently, presenting a different order from S. Mark, but never at the same point; so that whenever S. Luke deserts S. Mark’s order S. Matthew supports it, and where S. Matthew deserts it S. Luke supports it.” E. R. Buckley, An Introduction to the Synoptic Problem (1912), pp. 77f.

“A sixth line of evidence is that wherever Matthew drops Mark’s order of events, Luke resumes or retains, so that, as was first pointed out by Lachmann in 1835, Mark’s order is always supported either by Matthew or by Luke.” W. R. Smith, American Journal of Theology, October, 1913, article Fresh Light on the Synoptic Problem, p. 616.

“Its sufficiency [i.e., the sufficiency of the Markan order] was evidently recognized at once: Matthew follows it, and so does Luke, and though each of them deviates from it somewhat, yet they never deviate from it altogether. Mark always has the support of either Matthew, or Luke, or both. We never have to balance the order of Matthew and Luke against that of Mark.” A. Plummer, The Gospel According to St. Mark (1915), Introduction, p. xxiii.

“Again, not only have St. Matthew and St. Luke preserved practically all the sections of the second Gospel, they have arranged them in the same order. The order of St. Mark is always confirmed either by St. Matthew or St. Luke, and for the most part by both.” H. H. B. Ayles, The Interpreter, January, 1916 (vol. 12), article Origin and Date of the First Gospel, p. 171.

The following tabulation will disclose that there are quite a number of instances where the Markan order is deviated from by one of the other Gospels and remains nevertheless without corroboration by the other.

In studying the foregoing list, it is to be borne in mind that we are concerned only with sequences of whole incidents, and not with the order of details all of which belong to single events. The list presents the Markan instances each as a standard. In either the Matthaean column or the Lukan or else in both are presented deviations from this standard sequence. In all cases where a Matthaean or Lukan sequence is given as deviating, the Lukan or Matthaean column will show a failure of the other to support. And there are several cases in each of which both Matthew and Luke may each in turn be regarded as deviating from Mark and Mark remaining without support from the other. These are the instances where the Markan sequence is affirmatively opposed by both the remaining Synoptic Gospels. They are Nos. 4, 5 and 6.

Other points may, perhaps, best be treated by means of illustrative examples. Thus, in considering whether Matthew deviates from Mark in No. 1, we are not concerned with the incident in Mk. 1:21–28 as it is unparalleled in the First Gospel. The differing sequence in the Matthaean column consists in part of a parallel to one of the members of the Markan sequence. The other Matthaean incident is necessarily a parallel to some incident in Mark (namely, 1:39). The Lukan column is not marked as supplying or deviating from the Markan sequence, but only as failing to support that sequence. The only possibility of furnishing support would be by considering Lk. 5:1–11 parallel with Mk. 1:14–20. Even under this assumption, Luke fails to support Mark in connection with the Matthaean deviation from the Markan sequence. In instance No. 4, the notations “Deviate from Mark” in the Matthaean column and “Fails to support Mark” in the Lukan column are to be considered together, the sense being that the Matthaean sequence deviates from the Markan and the Lukan column fails to support the latter. After this, No. 4 may be considered as a separate instance, disclosing a Markan sequence, deviated from by Luke and left unsupported by Matthew. This is the sense of the Lukan notation “Deviates from Mark” when considered along with the Matthaean “Fails to support Mark.” In No. 3, we have in the Matthaean column a blank for one member of the sequence. This would be improper if the sequence were claimed as divergent from the Markan. But that is not the case, the notation in the Matthaean column merely stating “Fails to support Mark.” That is, there is failure to support the Markan sequence in connection with the Lukan deviation from it. This failure is due to the fact that Matthew contains no incident paralleling the Markan and Lukan ones concerned with the Appointment of the Twelve. 

It may be gathered from the foregoing “List of Opposed Markan Sequences” that the order of the Second Gospel is by no means always corroborated by one or both of the other Synoptic Gospels. In fact, three instances are given where the Markan sequence is opposed both by Matthew and by Luke. And in one of these, Matthew and Luke are in agreement against Mark.

There are indications in the writings of some of the advocates of the priority of Mark over the two others that the writers knew of something wrong in the claim of unvarying support, or suspected something wrong. I cite the following passages, italicizing words of qualification.

“When the sequence of narratives in St. Matthew or in St. Luke differs from that in St. Mark, the other one agrees with St. Mark. In other words, St. Matthew and St. Luke do not, save in one or two instances, unite against St. Mark as to order. When all three do not agree in respect to it, we have the same sequence in St. Matthew and St. Mark, or in St. Luke and St. Mark.” V. H. Stanton, The Gospels as Historical Documents, Part II, The Synoptic Gospels 1909), p. 34.

“Sometimes all three agree in order, but where two agree Mark is practically always in the majority. Mark and Matthew may agree against Luke, or Luke and Mark against Matthew, rarely, if ever, Matthew and Luke against Mark.” A. S. Peake, A Critical Introduction to the New Testament (1916, prefaces 1909), p. 102.

Statements such as these are, however, insufficiently qualified to render them adequate reflections of the facts. There are, as Dr. Stanton implies, one or two cases where Matthew and Luke agree in order against Mark. No. 5 in the “List of Opposed Markan Sequences” is an instance of this character. But, as the same list shows, there is much more disagreement with the Markan order in cases where Mark has no corroboration whatever. Prof. Peake, similarly to Dr. Stanton, allows that perhaps there may be a case or two where Matthew and Luke agree in their order against Mark. The statement is perhaps sufficient for this particular class of divergence. The preceding qualification, “practically always,” gives its clause an equivalent sense, since it makes the statement mean that very infrequently is there a dual agreement in order that does not involve Mark. Such a dual agreement would accordingly be one between Matthew and Luke.

We have, then, on the part of some writers the concession that perhaps in a case or two Matthew and Luke may diverge in the same way from Mark. This is all very well, as far as it goes. The list which has been given exhibits eight instances not of this class.

We have in the list a total of nine cases, in each of which the Markan order is deviated from by one of the two other Gospels and remains uncorroborated by the remaining one. We have now to consider whether these nine instances constitute a considerable part of the entire class of instances where Markan sequences are deviated from by one or both of the remaining Gospels, Matthew and Luke. If the entire class is so considerable in number that the nine instances constitute only an insignificant part of the whole, then the claim for the corroboration of the Markan order may still be set up, though in a qualified form. That is to say, the statement might be urged that the order of Mark is, except for a number of cases relatively few, always supported by one or both of the remaining Synoptic Gospels. But, are the exceptions relatively few?

What Are The Facts?

In order to get the matter of deviations squarely before us, it will be helpful to introduce a tabulation showing (1) all the incidents of Mark arranged in the Markan order; and (2) all the Matthaean and Lukan parallels, also in the order of Mark. In constructing a table of this kind, some difference of opinion may arise as to details. I call particular attention to the fact that I reject Mt. 9:27–31 and 9:32–33 [34] as parallels of Mk. 8:22–26 and 7:31–37 and Lk. 7:36–50 as a parallel of Mk. 14:3–9, and make Mt. 15:29–31, Lk. 4:14–15 and Lk. 5:1–11 only doubtfully parallel to Mk. 7:31–37, 1:14–15 and 1:16–20. On the other hand, I make Lk. 4:16–30 parallel with Mk. 6:l-6a.

The foregoing tabulation exhibits the division of the whole of Mark into 83 sections each constituting an incident or event. In the lists of Matthaean and Markan parallels, the letter D is used to indicate an incident immediately after which begins a deviation from the Markan order. The bracketed events are entirely disregarded in inserting this indication of deviation, no bracketed event being marked and none allowed to influence a marking. There are, it may readily be seen by counting the Ds in the proper columns, 12 instances where Matthew deviates from Mark and 14 where Luke deviates from Mark. A second letter, f, is employed to indicate failures to support Markan sequences. It is placed alongside a Matthaean or Lukan parallel or blank which, when taken with the next incident or blank, in the column, in the proper textual order, shows the failure to support the Markan sequence beginning abreast the f. By noting those Ds which have abreast of them either another D or else an f, we may

determine the earlier member of all Markan sequences which are opposed by Matthew or by Luke, lor by both, and which are left without support. In each case, the earlier incident of the Markan sequence will lie between two Ds or between a D and an f. In this way, we may readily single out the following- nine Markan sequences which are deviated from and at the same time left without support:

It may readily be seen that there are 7 instances of Markan sequence where Matthew deviates and where Mark is left without the support of Luke; and similarly 5 instances of Markan sequence deviated from by Luke and unsupported by Matthew. Of the total of 12, we may note 3 duplications, cases where a Matthaean D and a Lukan D are abreast of each other. These duplications indicate the three points where the Markan order is affirmatively opposed by both the other Synoptic Gospels. In the last table, these are Nos. 4, 5 and 6. In the case of No. 5, Matthew and Luke not only oppose the Markan order together, but agree with each other in respect to order at the moment of deviation from Mark.³

There are, as we have seen, 12 Matthaean deviations from Mark and 14 Lukan. Adding these, we have a total of 26. But there are 3 duplicates, so that the net number of deviations from the Markan order, deviations involving Matthew, Luke, or Matthew and Luke, amounts to 23. We have also noted that there are 7 Matthaean deviations where Mark has no support from Luke and 5 Lukan deviations where Mark has no support from Matthew. Adding these we get a total of 12. There are, however, 3 duplicates, so that the net number of cases of deviation from Mark where Mark remains without support is just 9. We are now prepared to compare the 9 and the 23 cases. That is to say, we may now state that Matthew, or Luke, or both together, deviate from Mark at 23 points; and that in 9 of these instances the Markan order remains unsupported. The failure to support thus extends to nearly 40 per cent, of the total number of deviations.

Certain of the foregoing numerical results are of great advantage if we desire to form a just view of the argument for Markan priority based upon the Matthaean and Lukan support of the order of events in Mark. That argument may be put in the following form: Mark is the primitive document because it discloses the primitive order; and the Markan order is to be adjudged primitive because, of the three Synoptic orders of events, that of Mark is the only one which, when deviated from, has in general the support or corroboration of one or both of the remaining two documents. Supplied with the actual figures, we may now note that the deviations from Mark total 23. Of these 23, only 14 have the support of at least one of the other Synoptic Gospels. In nearly two-fifths —that is, in 9 cases out of the 23—there is no corroboration whatever. From this we may gather that, even if we allow that the argument from corroboration is logically sound, we cannot grant that it is especially formidable when the corroboration has to be reduced nearly two-fifths. Mark is not generally corroborated unless we allow that failure in nearly two-fifths of the cases is no bar to a statement as to general corroboration. As a matter of fact, this amount of failure is too considerable. The disclosure of the actual figures enables us, accordingly, to conclude that we are not entitled to single out Mark from the three Synoptic Gospels as the one whose order is so much better corroborated that it may fairly be claimed that in those cases where there is a break in the triple agreement as to order the Markan order is generally corroborated by one or both of the others. If the advocates of the priority of Mark think it worth while to continue a presentation of the argument from corroboration, it would seem that, in view of the numerical results I have set forth, a good deal of modesty should characterize their future statements of this argument.

The fact that one of the 9 failures in corroboration is an instance where Matthew and Luke agree in their opposition to the Markan order, is a consideration that will, if properly taken into account, not have the effect of lessening that proper modesty which should accompany future presentations of the claim for the primitiveness of the Markan order that may be based on corroboration. When Matthew and Luke agree, in opposition to the Markan order, in placing their parallels of Mark so that the combined incident of The ruler’s daughter and the woman with the issue of blood is immediately followed by the incident of The Twelve sent forth, we can hardly view this agreement in order other than a reflection of the facts. Those who see in Matthew a truly chronological presentation of events will surely do so; and those who regard Matthew and Luke as documents neither of which was directly derived from the other are apparently required to concur by the logic of the situation.

Three out of 9 cases of failure to corroborate are instances of joint opposition to the Markan order. They have already been pointed out, and the preceding case where Matthew and Luke agree in the deviating incident creating the opposition is one of these three. The remaining 2 are instances where the failure in corroboration is not merely a case of absence of support but an instance where an actual deviating order is disclosed. Both Matthew and Luke affirmatively deviate. They agree to deviate, though they deviate with different incidents. The fact, then, that 2 of the seven failures in corroboration are instances of affirmative opposition is an additional reason why modesty should attend the pressing of any future claim that the primitiveness of the order of events in Mark is shown by the way in which this order is corroborated.

The foregoing exposition makes clear the following conclusions:

1. The Markan order is deviated from at 23 points.

2. In 9—that is, in nearly two-fifths—of these instances of deviation, Mark is uncorroborated.

3. In 3 of the uncorroborated instances of deviation, we have cases of affirmative opposition.

4. In 1 of these cases of affirmative opposition, we have Matthew and Luke in agreement in the sequence that is in affirmative opposition to the Markan order.

In view of the foregoing presentation of facts respecting the order of events as disclosed in the three Synoptic Gospels and in view of the language of the excerpts cited, we have just cause, I think, for surprise that so many prominent writers, covering a very extended period of time, entertained so incorrect an apprehension of the relations of the three documents in respect to their disagreements in the matter of sequence. The reader should, by now, begin to perceive that there may very well be sound reasons for challenging claims that the investigations into the Synoptic Problems have proceeded along scientific lines. One of the basic rules of scientific procedure is to make sure of the facts.

But, let us go on. Just as there has been misapprehension of some of the main facts pertaining to order, so there has been a misconception of the consequences of the data as to sequence that have been assumed. That is to say, the logical situation has been misapprehended.

1. It has been argued that the Markan order is, because of its relatively strong corroboration, the most primitive. Here we have the argument from corroboration.

2. It has been argued that the Markan order is basic because it is possible to interpret the orders of Matthew and Luke as based upon it. That is, the Markan order may be used to explain the orders of Matthew and Luke. Here we have the explanatory argument.

In both cases, the logical procedure is wrong. It will naturally be in order to justify this statement.

The Argument From Corroboration

I begin with the argument that the order of Mark is the most primitive because of the superiority of its corroboration over that of Matthew or of Luke, and proceed at once to adduce citations which show how representative writers have presented the argument. The reader will please bear in mind that we are at present concerned not with the question of fact which relates to the amount of corroboration possessed by the Markan order but with the question whether the proposition that the order of Mark is much superior to the other orders in respect to the matter of corroboration can justly be used as a basis for the inference that Mark antedates the other documents. We are now concerned as to the correctness of a logical process. Is it good logic to argue for the primitiveness of the Markan order from the superiority of the support possessed by that order?

“There are a few cases, of which I will speak presently, where no two of these Evangelists agree in their order. But in many cases two of them agree against the third; and it is natural to adopt the principle that the order in which two agree is likely to have been the original order. Now it will be found on examination that the adoption of this principle practically amounts to recognizing Mark’s order as the original; for in every case (except one to be mentioned presently), where two agree against the third, St. Mark’s is one of the two.” George Salmon, A Historical Introduction to the Study of the Books of the New Testament⁴ 9th ed. (1899), p. 574.

“Moreover, there are only two sections in which the order followed by Mark is different from both the other Synoptics, and the other Synoptics in these two cases disagree with each other. That is to say, Matthew’s order is peculiar in eleven cases, Luke’s in fourteen, Mark’s in two. This certainly seems a strong proof that Mark’s is the original order.” C. E. Scott-Moncrieff, St. Mark and the Triple Tradition (1907), p. 27.

“And it should be noted that where Mt. deviates from the order of Mk., Lk. commonly follows it. Mk. is nearly always supported by either Mt. or Lk. or both: his is the original order. A. Plummer, An Exegetical Commentary on the Gospel According to S. Matthew (1909), Introduction, p. xviii.

The argument that the assumption of the Markan order as one that is always, or nearly always, corroborated by the other Synoptic Gospels combined with the assumption that this order is the only one so corroborated—the argument that these assumptions conjointly require that this order be viewed as the most primitive—is unsound, because it overlooks the fact that we have to do with documents that are assumed to be dependent upon one another. Here is the root of the great fallacy. If the three Gospels be viewed as mutually independent, then the Markan order would, because of the relatively strong corroboration by one or two independent witnesses, be the best attested order and so probably the most primitive of the three. The assumption of mutual independence is, however, inadmissible. I am investigating the consequences of these facts as to order under the hypothesis that Matthew and Mark are two documents such that one or the other is a derivative of the remaining one. The corroboration afforded by Matthew to the order of Mark would then be no appreciable corroboration at all, whether we make Matthew or Mark the derivative writing. In fact, the introduction of the conception of dependence disturbs the whole situation. If we assume that the three Synoptic Gospels are mutually interdependent, we thereby give up pretty much all the support that is given by corroboration. An order of events that is essentially a copy of another order of events affords no proper corroboration of that order. But let us attend to the several ways in which the three Gospels may conceivably have been derived and note whether the relatively strong corroboration of Mark has any real significance in connection with the priority of this Gospel.

The assumption that Mark was derived both from Matthew and Luke accounts very well for their continued corroboration of the Markan order. That is to say, if we conceive the Second Gospel as produced by selections made now from Matthew and now from Luke, it is very easy to perceive that such a procedure might result in a very general agreement in order with one or the other or with both. In fact, the assumption that in respect to order Mark is always or generally supported by Matthew or Luke or both is a situation explicable by the old Griesbach Hypothesis. Again, if we assume derivation in the order, Luke-Mark-Matthew, we can give another explanation of the fact of the corroboration of the Markan order. The derivation of Mark from Luke would account for the general correspondence in order between the two; and, similarly, the derivation of Matthew from Mark would account for the general sameness of sequence between them. That is, we may thus explain the similarities of the First and Third Gospels to the Second Gospel.

Finally, we may assume the order of derivation to be Matthew-Mark-Luke. This also is competent to explain a constant, or nearly constant, corroboration of Mark by the others. Mark derived from Matthew would naturally exhibit Matthew’s order as a general rule. Luke derived from Mark would also be largely in the same order. That is, Mark would be generally like Luke.

Accordingly, the bare assumption that Mark is always, or nearly always, corroborated by one or both of the remaining two Synoptic Gospels may be explained by any one of the following modes of derivation:

We thus have four alternatives. These likewise account for the considerable difference in order between Matthew and Luke, provided we limit or exclude direct connection between these two.

So, then, even if the facts permitted the assumption that Mark, and Mark alone, is always, or nearly always, corroborated by one or both of the remaining Synoptic Gospels, we do not get an indication that points in a single direction. We get, in fact, a sign-post that points four ways at once. The priority of Mark to both the other Synoptic Gospels is by no means singled out for favor. It is merely one possibility out of four.

As a matter of fact, Mark is posterior to one or both the others in three of the four alternatives. That is, this Gospel is so far from being indicated as prior to the two others that it is three times as probable that it followed one of them. However, we are only mildly concerned with Luke at the present juncture. It is important, then, to call attention to the fact that, if the four alternatives be examined, it will be found that Matthew precedes Mark as often as Mark precedes Matthew.

The Explanatory Argument

I turn now to the view that Mark may be claimed as prior to Matthew and Luke on the ground that the orders of these latter Gospels may be satisfactorily explained on the assumption that the Markan order lay before the two writers. Let us consider the following excerpts, which set forth the views of prominent writers.

“But if one takes the succession of individual incidents in Mark and places that in Matthew to one side and that in Luke to the other side, then one may prove step by step that each of the two others presupposes precisely this succession as the primitive one. Upon this consideration is founded very especially the strength of the Markan hypothesis; here it has never been shaken, indeed scarcely noticeably attacked. Still less then has it been refuted.” H. J. Holtzmann, Einleitung in das Neue Testament (1892), S. 359.

“When two auothors present almost the whole matter of a third 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.

“Now none of the First Three Gospels have disposed their contents in precisely the same order. But it may be said broadly that, if Mark’s order be put in the middle, with Matthew and Luke on either side, it will serve as a standard of comparison explaining them both. The divergences of each can be referred to this as the original type. If this be so, Mark must have preceded the other two.” J. E. Carpenter, The First Three Gospels—Their Origin and Relations (4th ed., 1909), p. 204.

Learned and scientific circles seem to be considerably befuddled over what is really a simple logical fallacy. That is to say, many appear captivated with the idea, that if an hypothesis can be shown to be compatible with the facts, therefore the hypothesis is true.

In illustration of this statement, I may reproduce from a recent book notice a quotation from a work by J. H. Beibitz:

“The test of a hypothesis is whether it is capable of giving a coherent and rational explanation of observed facts.” Theology, March, 1923, p. 174.

Another illustration may be cited from a recent work dealing in part with Synoptic matters:

“Consideration of the second alternative may prove superfluous, if the third be examined first and found to explain the facts adequately.” G. W. Wade, New Testament History, 1922, p. 155.

As a matter of logical fact, when we find that an hypothesis constitutes an adequate explanation of the known facts, we are still far short of a proof that the hypothesis is true. What we have done is merely to ascertain that the hypothesis is tenable. That is all. In order to prove our case, it is necessary to go on and show that the hypothesis, in addition to being an adequate explanation, is the only tenable one possible. The alternatives—that is to say, all conceivable alternatives—must be eliminated. This can not be done by showing that a certain favorite is adequate and therefore tenable. In short, hypotheses cannot be proved by showing that they afford adequate explanations of the known facts. Viewed as a means of establishing hypotheses as reflections of the truth, the argument from explanation is a logical fallacy.

The reasoning is invalid which would conclude that the priority of Mark is a fact merely because it is a tenable proposition which asserts that the Markan order may be conceived as a “standard of comparison” or may be viewed as the “original type” from which the order of Matthew and Luke may be derived, “step by step.” Even if the view be allowed that this may be done, the only thing proved thereby is that Mark might perhaps have been primary to both the others. It is not an argument that singles out Mark as the only possible prior document.

The logical error here is based, not upon a failure to recognize that derivation breaks up independence of evidence, but upon the fallacy that consistency amounts to proof. The mere fact that we can rig up an explanation of certain phenomena, or show that these phenomena are consistent with a given hypothesis, is insufficient to establish that explanation or that hypothesis. Nor will even an argument be furnished tending to such establishment, if another and equally good explanation of the phenomena can be adduced or if the phenomena are just as consistent with an alternative hypothesis.

In the present case, several alternatives will explain the facts as to order and will constitute hypotheses with which the phenomena of sequence will be consistent. We may view the orders of Matthew and Luke as both available to the writer of Mark. The Markan order would then be understandable as originating in part from one, in part from the other, and in part from both. Or, the Lukan order may be set up as the source of the Markan progression of events, and this progression of events may in turn be made the basis of the Matthaean sequence of incidents. Or, we may adopt the succession of documents which corresponds with that disclosed by most ancient MSS.—Mathew-Mark-Luke, understanding that Mark got its order from Matthew and Luke got its order from Mark. In fact, we get the same group of four alternatives as when we considered the matter of corroboration. Similarly, also, we find that it is three times as probable that one of the other Gospels preceded Mark as that it preceded both of them.

Or, if we confine our attention solely to Matthew and Mark—as our present interest dictates that we should do—we find, upon considering the four alternatives, that Mark is prior to Matthew twice, and that Matthew is prior to Mark twice. As between the two Gospels, the matter of priority is thus left equally balanced.

I give the four alternatives again, which will enable the reader conveniently to verify the foregoing statements:

It will now be clear, perhaps, that whether we consider the matter of the corroboration of the Markan order, or whether we weigh the possibility of assuming this order as primary to the two other Gospels, we derive no argument for the priority of Mark over Matthew. It is just as probable that Matthew preceded Mark.

Assumption That Mark Is Chronological

Let us now proceed to a further consideration of the deviations between Matthew and Mark, one or the other of these two gospels being assumed as secondary to the other. I raise the question: What is the logical effect in respect to the matter of priority when we assume Mark to be chronological?

The first thing to note is the opposition that is thus almost certainly set up against the testimony of an exceedingly ancient witness, one who was either the Apostle John himself or else a person who lived at a period overlapping the days of this Apostle. Eusebius has preserved the testimony in the form of an excerpt from the writings of Papias himself, a bishop flourishing in the early part of the second century.

“And the Elder [or Presbyter] John said this also: Mark having become the interpreter of Peter, wrote down accurately everything that he remembered, without, however, recording in order [οὐμέντοι τάξει] what was either said or done by Christ.” Fragments of Papias, translation in J. B. Lightfoot’s work, The Apostolic Fathers (1907), p. 529. Or Eusebius, Historia Ecclesiastica, 3:39.15. (See, in English, Nicene and Post-Nicenc Fathers, Second Series, vol i).

Whether the Elder John, who said that Mark did not record the sayings and deeds of Christ in order, was the Apostle, the son of Zebedee, or whether he was a sub-contemporary, we have here a statement dating from about 100 A. D. As this testimony is apparently in such sharp conflict with the hypothesis of a chronological Mark, it constitutes a formidable difficulty in the way of this subsidiary hypothesis as to priority. The difficulty is in fact in the way either of a prior Mark or of a prior Matthew.

Let us now revert to our question and consider some consequences besides that of almost certain opposition to this most venerable testimony, consequences which result when we persist in asking what is the logical effect of the assumption of a chronological Mark upon the matter of priority? We shall find that for the hypothesis of a prior Matthew the assumption is inconvenient and damaging, and for the hypothesis of a prior Mark disastrous consequences result.

Immediately we assume Mark as chronological, we assert the Matthaean deviations to be non-chronological. This results from the consideration that, under the assumption of derivation, there is, as between the first two Synoptic Gospels, but two alternatives. Either Mark is a derivative of Matthew or Matthew is a derivative of Mark. Whether we elect the one alternative or the other, the deviations in the order of events call for explanation.

If now we make Matthew the prior writing and concede that its deviations from Mark are departures from the true chronology, then we have to explain how it came about that the secondary document, so dependent upon the primary for its language, even in the narrative framework, should nevertheless be able to correct the chronological errors. Perhaps this embarrassment may be relieved by having recourse to Peter. With Mark the secondary writing, there appears no reason why this Apostle may not have supplied guidance for the Markan author— John Mark, no doubt—along the line of the exact chronology. That is to say, there appears to be no reason except this, that the maintenance of Matthaean priority would then be tied up with the view that the Second Gospel was put into definite shape during the lifetime of Peter. This may or may not be a liability.

Let us now consider the effects upon the hypothesis of a prior Mark that result from the assumption that Mark is chronological. This assumption, as has already been pointed out, involves another to the effect that the Matthaean deviations from Mark are non-chronological. Consequently, the advocates of the priority of Mark that view this Gospel as a chronological record are involved in the necessity of explaining what they are thus obliged to conceive as the non-chronological deviations of Matthew.

Consider now the following excerpts which sets forth the view that Matthew deviates from Mark in the process of carrying out a literary plan and in so doing breaks away from the chronological order.

“For the purpose of setting up the Gospel of Mark as the first [Luke having already been put last], it is at once permissible to bring into the field the fact that the Gospel of Matthew has a literary plan. In Matthew, everything is rounded off. In this Gospel, the numbers three, seven, ten control the Genealogy, the Temptation, the miracles, the parables and the Woes. In Matthew, everything is arranged in order and is marshalled in rank and file in accordance with rules. Matthew gathers sayings and small discourses in order to form the Sermon on the Mount into a rather considerable discourse and unites miracles of every sort into a comprehensive picture of the miracle-working activity of Jesus. But precisely because of this, because of this order, conditioned and defined in accordance with the content of the material, Matthew becomes diverted from the natural development and completely destroys the chronological order, so that Jesus is just about the same at the beginning as at the end, the result being that the Matthaean writer carries back into the beginnings of the activity of Jesus his [own] view of the Messianic dignity of Jesus.” C. R. Gregory, Einleitung in das neue Testament (1909), S. 758.

Let us make no difficulty over the hypothesis that Matthew discloses topical groupings of incidents into clusters of three, seven and ten each. Perhaps five may also be admitted to have been a number having a degree of favor. The text itself points out the numerical structure of the Genealogy. That there were more than seven parables actually spoken upon the day of the visit of the Savior’s mother and brethren is indicated by the additional one given in Mk. 4:26–29; so that we may conclude that the fact that just seven are given in Matthew is due to a choice exercised by the writer. We may grant, too, that the existence of just five considerable sections of discourse, each concluding with a formula more or less standardized is due to a literary plan—in the mind of the Matthaean writer. There is doubtless danger of pushing the numerical idea too far. We can hardly view the fact that we have just three petitions in Gethsemane, three denials of Peter, and three petitions and three aspirations in the Lord’s Prayer as something over which the author exercised any control. Then there are cases where the matter hangs in doubt. There are perhaps just seven Woes; but is this due to choice on the part of the speaker or to a reduction from a larger number made by the writer? But, the fact that we are willing to concede that Matthew discloses numerical groupings of topically related incidents and parts of incidents does not involve us in the further concession that the Matthaean writer tampered with the chronology in carrying out his plan. We might, I think, grant that perhaps he would cut down our actual group of eight or more parables of the kingdom of Heaven to a group of seven, or the accounts of four or five successive miracles of healing to a cluster of three. Such changes, effected by simple elimination, need not disturb the chronology. But it is not necessary that we view it as at all possible that he would dislocate the chronological succession of events in carrying out his topical and numerical plan. Thus, because we grant that he had such a plan, it is not therefore necessary that we allow the possibility of the transfer of the account of the incident of Touching the leper from its true chronological position to one quite different for the purpose of creating a group of three narratives all dealing with the healing of disease. Of course, one might be expected to yield to evidence. But, I may say that I know of no single instance where it can be shown that the Matthaean author has placed an event in a wrong historical position. I do not wish to use here evidence of an affirmative character tending to show that the First Gospel was written with a chronological plan and that this plan was successfully carried out, notwithstanding other possible purposes and plans, preferring to set it forth at a different stage of the argument against the priority of Mark. At the present moment, however, it is desired nevertheless to point out the weakness of the claim that when Matthew deviates from Mark it departs from the true chronology. We have no right to assume without evidence that the Matthaean author was willing to disturb the historical progress of events for the sake of a literary plan. I know of but three instances where the chronology in Matthew is probably at fault. The passage Mt. 10:17–23 belongs doubtless in ch. 24, perhaps between verses 9 and 10; ⁵ Mt. 26:6–13 should, I suppose, stand between chh. 20 and 21; ⁶ and Mt. 26:59–66 and 26:69–27:1 are probably in correct position when interchanged. The last two are not deviations from, but conformities with, the Markan order. None of the three needs be referred to the author. All are explicable as misplacements of the text due to accidents to MSS. The order of the three parts of the Temptation differs in Matthew from that disclosed in Luke. This deviation occurs not in connection with whole incidents but within a single event. And yet, even here, the First Gospel is evidently right in putting Satan’s demand for worship last, because of the words, “Get thee hence, Satan.”^{7 8} With no instances to cite for a Matthaean departure from the chronology, it is, of course, absurd to claim that he did deviate in order to further a literary purpose. The claim is a mere conjecture.

However, particular instances have been cited, in support of wrong chronology in Matthew. Consider the following statement:

“Moreover, when one of the other Synoptists [i.e., other than Mark] strikes out a path peculiar to himself, his order usually has less verisimilitude, and is open on internal grounds to suspicion.

“Thus (1) when Mt. places the gathering of crowds from Decapolis and Judaea at the very outset of the Ministry (Mt. 4:25), there can be little doubt that he antedates a state of things which Mc. rightly places at a later stage (Mc. 3:7 ff). (2) The crossing to the Gadarene (Gerasene) country, if preparatory to an evangelistic tour in the Decapolis, seems to come too early in Mt.’s order, and on the other hand he places the calling of the Apostles too late; in Mc. both incidents occupy places which accord with what appears to be the natural course of events. . . .

. . . whilst in Mt. the incidents have sometimes fallen into new surroundings which are inconsistent with those assigned to them in Mc. or Lc. or in both; comp., e. g., Mt. 8:1 Καταβάντος δὲ αὐτοῦ ἀπὸ τοῦ ὂρους (Lc. ἐν τῷ εἶναι αὐτὸν ἐν μιᾷ τῶν πόλεων), 1x, 18, ταῦτα αὐτοῦ λαλοῦντος (Mc. and Lc. place the preceding parallels in other contexts).” H. B. Swete, The Gospel according to St. Mark (1902), Introduction, pp. lxx-lxxi.

Here we have three citations of instances supposed to favor Matthaean departures from the historical order. The “gathering of crowds from Decapolis and Judaea” is not presented jn Matthew without adequate explanation. In 4:23, we are told that Jesus went about in all Galilee. He was engaged in teaching in the synagogues and in preaching the gospel of the kingdom. We are also told that He engaged in healing all manner of disease and all manner of sickness. In 4:24, we have in consequence the spread of His fame into all Syria. And we are told also how the people brought the sick, the diseased, the suffering, the demoniac, the epileptic and the palsied to Him, and how He healed them. It is upon this prefatory statement that 4:25 comes. “Crowds from Decapolis and Judaea” do not appear at all out of place, when such an adequate explanation for their gathering immediately precedes. The second matter advanced by Swete is the suggestion that the crossing to the eastern shore of the Sea of Galilee (Mt. 8:18–34) seems placed too early in Matthew. But this objection is conditioned by the words, “if preparatory to an evangelistic tour in the Decapolis.” But, why suppose that such a tour was in contemplation?

The third objection seems based on a misapprehension. The Apostles are listed, but not called, in Mt. 10:2–4. Their appointment, though not narrated in Matthew, must have come earlier as it was doubtless a distinct event, just as narrated in Mk. 3:13–15 and Lk. 6:12–13.

Other objections to positions of incidents in Matthew are based upon supposed contradictions between the setting in Matthew and that in the other Synoptic Gospels. The incident of Touching the leper is, in Matthew, placed next after the Savior’s descent from the mountain on which the Sermon was preached. It is not said whether the event occurred in a city or not. There appears to be no contradiction with the words in Luke which state that the incident occurred while he was in one of the cities.

In Matthew, the discourse consequent upon the question as to fasting (Mt. 9:14–17) is followed by an entirely different incident (Mt. 9:18–26) to that which follows in Mark and Luke (Mk. 2:23–28 and Lk. 6:1–5). That is to say, when Matthew here deviates from Mark, the latter is supported by Luke.

In fact, this instance adduced by Swete is but one of five that may be brought forward. Let us consider all of them. Matthew deviates altogether twelve times from the Markan order. In five of these, Mark has the support of Luke. But, for those who view Luke as deriving the greater part of its order from Mark, this corroboration is not the testimony of an independent witness, but the correspondence in sequence to be expected of a derivative writing. There are then five apparently supported Markan sequences from which Matthew deviates. As the support by Luke, under the hypothesis of derivation, counts for nothing, we have simply five cases where one writer is against another. This will not justify us in setting up Matthew as the one in error.

That Mark is a chronological document has been argued or asserted by upholders of its priority. Thus, we have the following passage from Menzies:

“While some connected groups of stories had been formed, many pieces of the tradition had nothing about them to show their connection; they were, as it were, loose leaves at the writer’s service, but not numbered nor provided with any reference to their proper position. How find the cord on which all these pearls were to be placed; how fix their proper position on that cord? What indeed was the story of which these were the incidents: of what nature was the central development around which they were all to fall into place? This Mark, alone of the Evangelists, was enabled clearly to make out and to record. It is here indeed that the observer most of all discerns that Mark must have been guided by one who knew the life of Jesus not only as a set of isolated stories but as a connected whole inspired by a growing purpose.” A. Menzies, The Earliest Gospel (1901), Introduction, p. 29.

“It is impossible to resist the impression that the writer who constructed this chain of sequence [disclosed by 33 indications of time in the section 1:14–9:50] believed himself to be presenting his facts upon the whole in the order of their actual occurrence. . . .

“But granting that the writer intended to follow the relative order of time, is there reason to suppose that he has succeeded? Can we recognize in this part of his work the steady and natural development of events which possesses historical verisimilitude?

“The answer makes itself distinctly heard by the careful student. He observes a progress in the history of the Galilean Ministry, as it is depicted by St. Mark, which bears the stamp of truth.” H. B. Swete, The Gospel according to St. Mark (1922), Introduction, pp. lviii-lix.

(To be continued)

1) C. Lachmann, De ordine narrationum in evangelicis synopticis, a paper in the periodical Theologische Studien und Kritiken, 1835, S. 570ff.

2) F. H. Woods.

3) This may be read from the long table by noting that the Matthaean parallel which, in the proper textual order of Matthew, immediately follows Mt. 9:18–26 is the incident 9:35–11:1; that the Lukan parallel which, in the proper textual order of Luke, immediately follows the incident Lk, 8:41–56 is the incident 9:1–6 and that these Matthaean and Lukan sequences are the same.

4) What Dr. Salmon conceived to be an exception is, in reality, none at all, for the reason that the deviation referred to by him concerns not the order in which incidents occur, but merely the order of presentation of the details in a single incident. His instance relates to the agreement of Matthew and Luke in mentioning the commencement of John’s activity before citing the passage from Isaiah and to the reversal of this order in Mark (Mt. 3:1–3; Lk. 3:1–6; Mk. 1:2–4).

5) This misplacement does not, in fact, involve the chronological succession of incidents, the passage being a piece of discourse which has been transferred from the midst of one long discourse to the midst of another.

6) This would bring the Matthaean order into agreement with that of John (12:1, 12). The present order in Matthew is readily explicable as due to a mechanical misplacement.

7) See the present writer’s article, The Lukan Narrative of the Temptation in The Methodist Quarterly Review, October, 1922, pp. 704-707.

8) The first two of the Matthaean group of instances of misplacement and the single Lukan instance are all treated, though briefly, in the present writer’s portion of the paper entitled A New Branch of Textual Criticism, presented at the annual meeting of the American Oriental Society held in Chicago in April, 1922, and published in The Augustana Quarterly, March, 1923, pp. 11-24.