Farmer's Argument For Matthean Priority

Portions ©1964 William R. Farmer

[Ed. note - The sixteen main points of Farmer's argument, as well as many of the examples and elaborations, have been lifted verbatim from Farmer's The Synoptic Problem, Mercer Univ. Press, 1981. However, I have freely reworded and reworked many details where appropriate for the sake of brevity. For a full appreciation of Farmer's work, consult the aforementioned book, or his more recent The Gospel Of Jesus, Westminster John Knox Press, 1994 - G.T.]

I. The similarity between Matthew, Mark and Luke is such as to justify the assertion that they stand in some kind of literary relationship to one another.

This is the very definition of the Synoptic Problem. 91% of Mark is reproduced in Matthew or Luke. Likewise, Matthew and Luke are 50% and 41% consistent respectively with the other synoptics. The nature of this similarity is such as to warrant the judgement that the literary relationship between these gospels could be one involving direct copying. It is highly recommended to consult the original Greek to appreciate the full degree of similarity. These examples are by no means exhaustive:

AGREEMENT BETWEEN MATTHEW AND MARK AGREEMENT BETWEEN MARK AND LUKE AGREEMENT BETWEEN MATTHEW AND LUKE AGREEMENT BETWEEN ALL THREE SYNOPTICS II. There are eighteen and only eighteen fundamental ways in which three documents, among which there exists some kind of direct literary dependence, may be related to one another.

III. While it is possible to conceive of an infinite number of variations of these eighteen basic relationships by positing additional hypothetical documents, these eighteen should be given first consideration.

This does not mean that the investigator should assume that there were no additional hypothetical documents. But a critic should not posit the existence of hypothetical documents until he has made an attempt to solve the problem without appeal to such documents and finds them necessary to explain phenomena otherwise inexplicable.

IV. Only six out of eighteen basic hypothetical arrangements are viable.

This follows from the circumstances that there are agreements between any two of the Synoptic Gospels against the third. Only the six cases where the second writer copied the first, and the third had direct access to both the first and second, do they afford the opportunity for any two of the Synoptics to agree against the third.

V. There are isolable and objectively definable categories of literary phenomena which have played a prominent role in the history of the Synoptic Problem which, when properly understood, are more readily explicable when Mark is placed third than when either Matthew or Luke is placed third.

These phenomena will be discussed in detail in steps VI-VIII.

VI. The phenomena of agreement and disagreement in the respective order and content of material in each of the Synoptic Gospels constitute a category of literary phenomena which is more readily explicable on a hypothesis which places Mark third with Matthew and Luke before him than on any alternative hypothesis.

Matthean and Lukan order very often deviate from Mark. However, when Markan order deviates from Matthew, it always agrees with Luke, and when Markan order deviates from Luke, it always agrees with Matthew. (The single exception is the Cleansing of the Temple, where Matthew and Luke agree against Mark by placing it on the same day as the Triumphal Entry.) Given the frequency with which Matthew and Luke deviate from Markan order, it is unreasonable to accept that they would never independently introduce variations to the Markan order in the same areas unless Mark is secondary to both of them and chose to consistently accept the order in one or the other. The only possible explanation under the Markan hypothesis is that Matthew and/or Luke had knowledge of the other's work and consciously chose to support Mark when the other did not.

VII. The minor agreements of Matthew and Luke against Mark constitute a second category of literary phenomena which is more readily explicable on a hypothesis where Mark is regarded as third with Matthew and Luke before him than on any alternative hypothesis.

These agreements between Matthew and Luke against Mark can be categorized as agreements of omission, agreements of addition, or agreements of modification.


On the Markan hypothesis, many of these common omissions by Matthew and Luke can be attributed to mere coincidence, though a few examples merit attention:


These passages are too numerous to mention here, most of them traditionally hypothesized to have been derived from the document "Q." Outside of those, however, it is worth noting that both authors opted also to include a genealogy, a nativity account, and an appearance after the resurrection. Additionally, there are certain key phrases to be found in the Passion accounts (see below).



1) Mk 2:1-12 // Mt 9:1-8 // Lk 5:17-26

Mark—2:1 And again He entered Capernaum after [some] days, and it was heard that He was in the house. 2 Immediately many gathered together, so that there was no longer room to receive [them], not even near the door. And He preached the word to them. 3 Then they came to Him, bringing a paralytic who was carried by four [men]. 4 And when they could not come near Him because of the crowd, they uncovered the roof where He was. So when they had broken through, they let down the bed on which the paralytic was lying. 5 When Jesus saw their faith, He said to the paralytic, "Son, your sins are forgiven you." 6 And some of the scribes were sitting there and reasoning in their hearts, 7 "Why does this [Man] speak blasphemies like this? Who can forgive sins but God alone?" 8 But immediately, when Jesus perceived in His spirit that they reasoned thus within themselves, He said to them, "Why do you reason about these things in your hearts? 9 "Which is easier, to say to the paralytic, `[Your] sins are forgiven you,' or to say, `Arise, take up your bed and walk'? 10 "But that you may know that the Son of Man has power on earth to forgive sins"--He said to the paralytic, 11 "I say to you, arise, take up your bed, and go to your house." 12 Immediately he arose, took up the bed, and went out in the presence of them all, so that all were amazed and glorified God, saying, "We never saw [anything] like this! (NKJV)

In the parallel passages of Matthew and Luke, they show the following series of agreements against Mark, the bracketed words being agreement by omission:

"Behold! ... on a bed ... [who was carried by four men] ... He said ... and ... [in His spirit] ... he said ... [to the paralytic] ... [take up your bed] ... on Earth to forgive sins ... departed to his house ..."

2) Mk 1:40-42 // Mt 8:2-3 // Lk 5:12-13

Mark—40 and there doth come to him a leper, calling on him, and kneeling to him, and saying to him—`If thou mayest will, thou art able to cleanse me.' 41 And Jesus having been moved with compassion, having stretched forth the hand, touched him, and saith to him, `I will; be thou cleansed;' 42 and he having spoken, immediately the leprosy went away from him, and he was cleansed. (NKJV)

Matthew and Luke agree against Mark as follows:

" ... and behold ... [to Him] ... Lord ... [having been moved with compassion] ... saying [to him] ... immediately ..."

In the Greek text, the above section includes eighteen consecutive words of verbatim agreement between Luke and Matthew against the parallel in Mark.

3) Mk 14:65 // Mt 26:67 // Lk 22:63-64

Matthew and Luke both include the question asked by Jesus's tormentors, "Who is the one who struck you?" not found in Mark. A theory of Markan priority affords no reasonable explanation as to how this phrase would have been added by both Matthew and Luke (the Passion being beyond the scope of Q), unless the phrase originally existed in Mark and was removed by a later editor. Once again, Markan redaction is the simpler and more reasonable theory.

VIII. There exists a positive correlation between agreement in order and agreement in wording among the Synoptic Gospels which is more readily explicable on the hypothesis that Mark was written after Matthew and Luke and is the result of a redactional procedure in which Mark made use of both Matthew and Luke.

When Matthew and Mark follow the same order, but Luke exhibits a different order, the texts of Matthew and Mark tend to be very close to one another. And when Luke and Mark are following the same order, but Matthew exhibits a different order, the texts of Luke and Mark tend to be very close to one another. If Mark copied Matthew and Luke, it is logical to assume Mark to have given some preference to the text of Matthew when he had deliberately chosen to follow Matthew's order instead of that of Luke, and conversely, it would not have been unnatural for him to have given some preference to the text of Luke when he had deliberately chosen to follow Luke's order in preference to that of Matthew.

IX. It is possible to understand the redactional process through which Mark went, on the hypothesis that he composed his Gospel based primarily on Matthew and Luke.

The point here is that it is possible to proceed through the Gospel of Mark on the hypothesis that he based his gospel on Matthew and Luke, and not encounter redactional problems which create serious and peculiar difficulties for that hypothesis. Farmer gives a nearly complete analysis of Mark's redaction of synoptic tradition in The Synoptic Problem, Chapter VII. Here, we shall simply address two common examples of Markan redaction which strengthen the argument against Markan priority:


Mk 1:32 - "And evening having come, when the sun did set, they brought unto him all who were ill, and who were demoniacs" (YLT)

Matthew and Luke tend to contain abbreviated forms of Mark's duplicate expressions by lacking one member of the pair, and most often Matthew and Luke each lack the expression used by the other:

Mt 8:16 - "And evening having come, they brought to him many demoniacs"
Lk 4:40 - "And at the setting of the sun, all, as many as had any ailing with manifold sicknesses, brought them unto him"

It is quite simple to see how Mark redactively combines these two verses to form a single sentence. If, on the other hand, we accept Markan priority, we must also accept that Matthew or Luke knew the work of the other, and consciously used only the Markan phrases the other had omitted.


The unusual Greek phrase 'opsías dè genoménes ("and evening having come"), common to Matthew and Mark in the above example, can be considered a "favorite expression" of Matthew, for the exact wording appears six times in his gospel. In Mark, however, the parallel verses use this wording only the first time; each subsequent passage shows some variation:

On the Markan hypothesis, the critic is asked to imagine that Matthew copied this expression exactly as Mark had it the first time he encountered it, but thereafter whenever he followed Mark in the use of this rare expression he consistently deviated from Mark and rigidly restricted himself to this particular grammatical form, even introducing it in 14:15 where Mark had `óras instead of 'opsías, and in 20:8 where there was no Markan parallel. If Mark were secondary to Matthew, however, the critic would only be required to imagine that Mark tended to modify this Matthean expression quite freely whenever he found it in his text of Matthew.

In Rev. Sir John C. Hawkins's Contributions to the Study of the Synoptic Problem, he lists 19 "favorite expressions" of Matthew that are not unique to Matthew—14 of these are found at least once in Mark, and 10 also in Luke. Likewise, of 7 Lukan expressions, 5 are paralleled verbatim in Mark and 2 also in Matthew. There seem to be no such expressions characteristic of Mark, however, which show up in either Matthew or Luke. There is only one solution that affords a ready explanation; namely, that Mark copied Matthew and Luke.

X. The most probable explanation for the extensive agreement between Matthew and Luke is that the author of one made use of the work of the other.

Not only do we find extensive agreement in content between the works of Matthew and Luke, but also remarkable similarities in literary form, much more so than either of them have with Mark. Both begin with birth narratives, contain genealogies, begin Jesus's ministry with his baptism by John, record the teachings of John, picture Jesus tempted by Satan in the wilderness, describe Jesus's ministry in Galilee, introduce into their narrative framework large collections of Jewish gnomic and parabolic materials - including primitive sayings reflecting the historical solidarity between Jesus and John, narrate Jesus's journey to Jerusalem with his disciples, his triumphal entry, his cleansing of the temple, the Last Supper with his disciples, the arrest in Gethsemane, the trial before the high priest and before Pilate, and finally the crucifixion, death, burial and resurrection of Jesus.

The fact that in some parts of this outline Luke introduced different content only indicates that he had ample reason to write a new gospel. Since it is not possible to explain the similarity in content and form between Matthew and Luke without making appeal to one or more hypothetical sources, we are led to posit direct literary dependence between Luke and Matthew.

XI. The hypothesis that Luke made use of Matthew is in accord with Luke's declaration in the prologue to his Gospel concerning his purpose in writing.

The Greek in Luke 1:1 is best translated as "many have undertaken to compile a narrative." The usual interpretation—that at the time Luke wrote there were many gospels in existence—is linguistically possible though unnecessary. Luke likely refers to a single gospel whose compilation, arrangement, and development of tradition can be viewed as the work of "many," perhaps even a school; the traditions being "delivered to us by those who were from the beginning eyewitnesses." It is reasonable, if not compelling, to identify this narrative with the Gospel of Matthew.

There is no doubt that from the point of view of readers who were sensitive to the prevailing standards of Hellenistic historiography that Matthew's gospel had difficulties. It was not set within an adequate chronological framework, so that readers acquainted with world history could view it within the context of that history. It contained duplicate accounts of certain events, and reported these as if they were completely separated in time and circumstance, when it was clear that they were but different accounts of the same matter. There were defects in the order in which certain material appeared in Matthew. Thus no sufficient reason is given for Jesus to have left Nazareth of Capernaum at the beginning of his ministry in Galilee, though a story of his being rejected by the people of his native place is included later in the narrative. And the call of the disciples after his great Sermon on the Mount is subject to the criticism of being anachronistic, in view of the indications that the sermon was for the disciples. At all such points, Luke's gospel seems to reflect the results of a prolonged and careful study of Matthew, with a view to the creation of a new narrative which would be free of such defects.

XII. Assuming that there is direct literary dependence between Matthew and Luke, internal evidence indicates that the direction of dependence is that of Luke upon Matthew.

Most passages found in Matthew which express a point antithetical to the mission to the Gentiles (i.e., 10:5) or the continuing importance of Jewish practice (i.e., 24:20) are not found in Luke. Words and customs which would have become increasingly less intelligible as the frontiers of Christian movement expanded farther from its place of origin (i.e., "raca" in 5:22, "phylacteries" in 23:5) are also excluded from Luke. Matthew's Semitic parallelism, used during the period of oral tradition for easier transmission, is frequently broken in the parallel passage in Luke (compare Mt 7:24-27 // Lk 6:47-49). All these instances support the hypothesis that Luke has altered Matthew and not vice-versa.

XIII. The weight of external evidence is against the hypothesis that Matthew was written after Luke.

The unanimous agreement among the writings of the early Church fathers supports the theory that Matthew was the earliest of the canonical gospels. The reliability of this testimony is always a matter of debate. However, given that the Church and its canon were a result of the mission to the Gentiles in the apostolic period, it is highly unlikely that they would all independently support the priority of the more Judaized Matthew unless the tradition represents the true chronological relationship.

XIV. The weight of external evidence is against the hypothesis that Matthew was written after Mark.

The same considerations set forth in support of the thesis in step XIII argue in favor of this thesis.

XV. That Mark was written after both Matthew and Luke is in accord with the earliest and best external evidence on the question.

The earliest statement on the question of order in which all the gospels were written is by Clement of Alexandria, who stated that the gospels with genealogies were written before the gospels without them. In view of the tradition that Mark founded the Church in Egypt, it seems highly unlikely that Egypt would have been the place for a tradition to develop placing the Gospel of Mark after Matthew and Luke unless there was some foundation for this tradition.

The earliest extant citation of the Gospel of Mark is found in Justin Martyr's Dialogue With Trypho. By way of contrast, clear citations exist of gospel texts distinctive of Matthew and Luke, not only in Justin, but in much earlier writings.

XVI. A historico-critical analysis of the Synoptic tradition, utilizing both literary-historical and form-critical canons of criticism, supports a hypothesis which recognizes that Matthew is in many respects secondary to the life situation of Jesus and the primitive Palestinian Christian community, but that this Gospel was nonetheless copied by Luke, and that Mark was secondary to both Matthew and Luke, and frequently combined their respective texts.

A. The canons of criticism utilized in the analysis of the Synoptic tradition referred to in this thesis are as follows:
  1. Assuming that the original events in the history of the Christian movement took place in Palestine within predominantly Jewish circles, and that by the time the Gospels were written, Christianity had expanded outside of Palestine and outside of circles which were predominantly Jewish in orientation: That form of a particular tradition found in the gospels which reflects an extra-Palestinian or non-Jewish provenance is to be adjudged secondary to a form of the same tradition which reflects a Palestinian or Jewish provenance.
  2. Assuming that the apocryphal gospel literature was written after the canonical gospels, and since, by comparison, this literature is characterized by a tendency to make the tradition more specific (by giving names to persons not named in the canonical narratives, etc.): That form of a particular tradition found in the gospels in which the tradition is more specific is to be adjudged secondary to a form of the same tradition in which the tradition is less specific.
  3. Assuming the redactional tendency to add explanatory glosses, and otherwise to expand tradition to make it applicable to new situations in the churches: That form of a tradition which exhibits explanatory redactional glosses, and expansions aimed to make the tradition more applicable to the needs of the Church, is to be adjudged secondary to a form of the tradition which is free of such redactional glosses and expansions.
  4. Assuming the tendency of all writers to use some words and phrases more often than is generally true for other writers when dealing with the same subject: That form of a tradition which exhibits words or phrases characteristic of a redactor whose hand is clearly traceable elsewhere in the same gospel is to be adjudged secondary to a form of the same tradition which is free of such words and phrases. And, as a corollary to this: That form of a tradition which exhibits words or phrases characteristic of a redactor whose hand is only traceable in another gospel is to be adjudged secondary to the form of the parallel tradition in the gospel where the redactor's hand can be clearly traced, provided the characteristic word or phrase occurs in the former gospel only in passages closely paralleled in the latter, where the verbatim agreement indicates direct literary dependence.
B. These four canons supplement the six set forth by Ernest DeWitt Burton in his monograph Some Principles of Literary Criticism and Their Application to the Synoptic Problem, which states that in questions of literary dependence between two documents, that document is to be adjudged dependent which contains features of a secondary character. The following are to be regarded as evidences of a secondary character:
  1. Manifest misunderstanding of what stands in one document on the part of the writer of the other;
  2. insertion by one writer of material not in the other, and clearly interrupting the course of thought or symmetry of plan in the other;
  3. clear omission from one document of matter which was in the other, the omission of which destroys the connection;
  4. insertion of matter, the motive for which can be clearly seen in the light of the author's general aim, while no motive can be discovered for its omission by an author who had it in his source;
  5. vise-versa omission of matter traceable to the motive natural to the writer when the insertion of the same matter in the other gospel could not thus be accounted for;
  6. alterations of other kinds which conform the matter to the general method or tendency of the author.
C. Considerations which sometimes have influenced students of the gospels in their statements about the Synoptic Problem, but which are either irrelevant or inconclusive, and therefore have little or no probative value in settling a question of literary dependence, may be listed as follows:
  1. The relative length of a given passage. Since writers sometimes enlarge and sometimes condense their sources, the relative length of a given passage by itself offers no criteria by which it may be adjudged primary or secondary to another.
  2. The grammar and style of a writer. Since sometimes writers improve the grammar of their sources while others spoil it, such considerations provide no objective basis by which one document may be adjudged primary or secondary to another. There is no provable correlation between style and chronology in matters involving the question of literary dependence between documents of the same general period and class of literature.
  3. The Christology of a given passage. Since the letters of Paul disclose the fact that Christology was already both complex and highly developed in some circles in the period before the gospels were written, and since our knowledge of Christological developments in the Churches in the post-Pauline period depends upon a correct solution to the problem of the chronological and literary relationship between the gospels, and not vice-versa, the Christology of a given passage in the gospels affords the critic no reliable criteria by which to adjudge it primary or secondary to its parallel in another gospel.

Return to the Introduction

Proceed to the Analysis of the Canonical and Apocryphal Scriptures.

Return to Geoff and Heidi's homepage