Mark 2:26 and the Problem of AbiatharRelated Media
Evangelical Theological Society
Southwest Regional Meeting, Dallas Theological Seminary
March 13, 2004
Bultmann was not right about everything, but he was certainly right when he recognized that presuppositionless exegesis was not possible. There are few texts where an exegete’s presuppositions can cloud his interpretation more than Mark 2.26. The issue here is not simply a conservative vs. liberal debate. Of course, battle lines are drawn by one’s bibliological convictions, but the tapestry of this passage is richer than that. Source criticism (specifically, whether one holds to Markan priority or Matthean priority), tradition criticism, textual criticism, and christological constructs are also lurking in the background here, to name a few. We will have a chance to explore these issues only briefly in the time allotted.
In Mark 2.26, as found in Nestle-Aland27, Jesus is reported as saying: πῶς εἰσῆλθεν εἰς τὸν οἶκον τοῦ θεοῦ ἐπὶ ᾿Αβιαθὰρ ἀρχιερέως καὶ τοὺς ἄρτους τῆς προθέσεως ἔφαγεν, οὓς οὐκ ἔξεστιν φαγεῖν εἰ μὴ τοὺς ἱερεῖς, καὶ ἔδωκεν καὶ τοῖς σὺν αὐτῷ οὖσιν… Or, in English, “Haven’t you ever read what David did when he was in need and he and his companions were hungry? How he entered into the house of God when Abiathar was high priest and ate the sacred bread that is not lawful for anyone but priests to eat, and also gave it to his companions?” (Mark 2:25-26). The fundamental problem with the phrase “when Abiathar was high priest” is that this incident in David’s life is recorded in but one passage in the OT, 1 Sam 21.1-7. But there, Ahimelech is mentioned as the priest; Abiathar, his son, would later become high priest, but he is not introduced into the narrative for another chapter (22.20).
On the one hand, the prepositional phrase, ἐπὶ ᾿Αβιαθὰρ ἀρχιερέως, has caused some angst for evangelicals because it ostensibly is a historical error. And if so, whose error is it? Did some early scribe corrupt his copy of Mark, which then influenced other witnesses and became the predominant text? Or did Mark add this as an editorial comment on his own? Or did he copy down accurately what his source said (which, according to patristic writers at least, would have been the apostle Peter)—a source that created the historical discrepancy? Or is it possible that Mark’s source repeated Jesus’ words accurately, but that Jesus made a mistake? Or did Jesus summarize the OT text accurately, but the OT was in error? Assigning error to someone is one route that is taken today in dealing with this problem. What I wish to contend, however, is that several presuppositions are at work in assigning blame; the matter cannot simply be isolated to a bibliological problem. Yet even here, there are rather different approaches to the problem by evangelicals.
In addition to the bibliological issue is the question of which Gospel came first. Those who embrace Markan priority tend to argue for an error on Mark’s part that would have been detected and eliminated by Matthew and Luke. Those who embrace Matthean priority tend to downplay any error on Mark’s part by various, although rather brief, explanations.
Then there is the christological issue. Very few scholars even entertain the notion that Jesus could have had a mental lapse. Here is where both liberal and conservative scholars are usually in agreement, but for different reasons: the more conservative scholars, because of their high christology and high bibliology, almost never raise the possibility that Jesus could have erred for that would apparently impugn the character of both the Lord and the Bible. Less conservative scholars (moderate as well as liberal) often see only part of the pericope going back to Jesus, and v. 26 is sometimes relegated to a later source. But Jewish scholars have no problem seeing this pericope going back to Jesus and attributing error to him!
Textual criticism also plays a role in this passage. There are variants that either alter the prepositional phrase and its subsequent translation or eradicate it altogether. But one’s text-critical theories inform his decision here—or at least they should!
This is just the tip of the iceberg. Unfortunately, the interpretations of this text are so vast and our time so short that we will have to park ourselves on that part of the iceberg that is above water. Perhaps that is the safest place to be though after all.
The fundamental problem in this text is that Abiathar was not the high priest when David went into the sanctuary and ate the showbread. This raises several questions; in the least, someone or something seems to be wrong. Here are the facts: (1) 1 Sam 21.1-7 mentions Ahimelech as the priest when David entered the sanctuary; (2) Abiathar was Ahimelech’s son; although he was a priest when this incident occurred, he was not the high priest but would become so later (after Saul murdered his father and eighty-four other priests); (3) Ahimelech’s ministry was in Nob, while Abiathar’s would especially be in Jerusalem; (4) except for the possibility of text-critical solutions, Mark’s gospel has the words ἐπὶ ᾿Αβιαθὰρ ἀρχιερέως, normally translated “when Abiathar was high priest.” In addition, there are several other, less significant differences between the dominical version of this story and that found in 1 Sam 21.1-7 (Gundry lists seven).
In addition to the differences between Mark 2.26 and 1 Sam 21.1-7, there are differences between Mark 2.26 and the parallel accounts in Matthew and Luke. The parallel in Matt 12.3-4 reads, “Haven’t you read what David did when he and those with him were hungry—how he went into the house of God and they ate the bread of presentation, which was not lawful for him or those with him to eat, but only for the priests?” And Luke 6.3-4 has, “Haven’t you read what David did when he and those with him were hungry, how he entered the house of God, took and ate the bread of the Presence (which is not lawful for any but the priests to eat) and gave it to those with him?” Except for a few stylistic changes between Mark 2.26 and the parallels in Matt 12.4 and Luke 6.4, the only difference is the omission of Mark’s “when Abiathar was high priest” (ἐπὶ ᾿Αβιαθὰρ ἀρχιερέως) by both Matthew and Luke. It is hard to resist the notion that Matthew and Luke deliberately expunged this line from their respective copies of Mark so as not to impugn the character of Jesus. But if one holds to Matthean priority, then a softer explanation for the differences must usually present itself.
What possible avenues for a solution do we have for the Abiathar problem? The leading contenders1 are as follows:
- Text-critical: the text is wrong and needs to be emended;
- Hermeneutical: our interpretation is wrong and needs to altered;
- Dominical: Jesus is wrong (or intentionally midrashic) and this needs to be adjusted to;
- Source-critical: Mark’s source (Peter?) is wrong (or intentionally midrashic);
- Mark is wrong (or intentionally midrashic).
The third, fourth, and fifth responses especially need to be examined more carefully, as they are usually rejected by those who embrace both a high bibliology and a high christology. It is imperative that we do not allow our presuppositions to preclude a solid historical investigation. The problem is that some evangelicals—especially members of this society—frontload their investigation with the explicit premise that the scriptures cannot err.2 Ironically, by starting with this presupposition, they may inadvertently pit Christ against the Bible. The incarnation demands that we do careful historical work, for God became man in time-space history. As such, he invites us to examine the data about his life and death, rather than take a fideistic stance of naïve, uncritical acceptance. As painful as it may be to think about some of these possibilities, if we do not wrestle with them then we will be dishonest in our handling of the text. Each of these approaches will now be examined; our order of investigation will be 1, 3, 4, 5, and 2. That is, we will leave the hermeneutical solution till the end.
(N.B. We are renumbering everything according to the new scheme below. This will be referred to later in the paper.)
1. Text-Critical: The text as it stands is incorrect and needs to be emended.
There are two basic alterations in the ancient witnesses here: D W 271 Itala Syriacs and a few others omit ἐπὶ ἀρχιερέως, no doubt in conformity to the parallels in Matthew and Luke. This is thus almost strictly a Western reading. Those who adopt this these textual variant are, generally speaking, more inclined to embrace Matthean priority. For example, in William Farmer’s The Synoptic Problem: A Critical Analysis3, the author enlists the help of V. H. Stanton4 in treating the minor agreements between Matthew and Luke.5 Farmer quotes from Stanton’s volume, apparently with approbation, that ἐπὶ ᾿Αβιαθὰρ ἀρχιερέως in Mark 2.26 is “erroneous” and that it “may have been an addition by a ‘badly informed copyist.’”6 Mann also entertains the possibility of scribal corruption, as does Sanders.7
As a sidenote, it is interesting that the Western scribes expunge the wording here. In the least, this seems to be evidence that they were concerned about protecting the Lord’s reputation when citing scripture. It is texts such as this (and there are hundreds of them, and in all text-types) that reveal early scribal piety across the board, suggesting that Dean Burgon’s condemnation of the early uncials as products of wicked men was unfounded.
A C Θ Π Σ Φ 074 1 131 209 f13 and many others add τοῦ before ἀρχιερέως. The significance of the article is that it turns ἀρχιερέως into an appositive, while the anarthrous noun remains a predicate genitive to ᾿Αβιαθάρ. (This will be discussed in some detail later.) The addition of the article gives the meaning “in the days of Abiathar the high priest,” suggesting a more general time-frame.8 This reading thus has a mixture of some Byzantine, Caesarean, and even semi-Alexandrian support. Neither reading has significant external support and both are obviously motivated by scribal piety toward the text. It is difficult to imagine scribes intentionally creating a problem by adding ἐπὶ ᾿Αβιαθὰρ ἀρχιερέως to Mark’s gospel and only to Mark’s gospel. Though perhaps easier to understand, the omission of τοῦ before ἀρχιερέως would hardly have occurred intentionally. And there is little good reason for it to occur accidentally as well. Thus, when it comes to determining which reading gave rise to the others, ἐπὶ ᾿Αβιαθὰρ ἀρχιερέως clearly is superior and obviously authentic.
A general caution about textual criticism I tell my students is that one should not use this discipline as a way out of a difficult problem, but as the means to determine the wording of the original. The biggest danger in textual criticism is to choose a reading that agrees with the interpreter’s preconceptions instead of choosing the reading that best explains (both internally and externally) the rise of the other readings. One has to wonder whether some Griesbachians need to heed that advice.
2. Dominical: Jesus himself made a mistake or was intentionally midrashic (i.e., he embellished the OT story to make his point).
There are two distinct options here: some think that Jesus may have erred; others think that Jesus embellished the OT text to make a point.
2.a. Jesus erred.
It might not surprise us to learn that Jewish interpreters have no problem seeing Jesus committing a historical mistake here. Thus, D. M. Cohn-Sherbok, a rabbi, argues that “though Jesus seems to have been familiar with rabbinic hermeneutics, the arguments he employs are invalid from a rabbinic point of view.”9 But what about Christian scholars? Indeed, there are some who entertain this view.
Brown argues that Jesus may have erred here10:
In Mark 2:26 Jesus says that David entered the house of God when Abiathar was high priest and ate the loaves of the presence. The scene is found in I Sam 21:2-7; there, however, the high priest is not Abiathar but Ahimelech. Matt and Luke seem to have noticed the difficulty, for their accounts of this saying of Jesus omit any mention of the high priest (Matt 12:4; Luke 6:4). Abiathar was better known than Ahimelech and more closely associated with David in later life, so that popular tradition may have easily confused the two. But if the reading is genuine, Jesus shows no awareness that he is following an inaccurate version of the story.
Brown is quick to note that he is not altogether comfortable making such pronouncements; indeed, he has been one of the strongest defenders of the deity of Christ in the 20th century, a factor which may give him pause here. This is an intriguing though disturbing option to consider. Yet few scholars give this more than a glance. Nevertheless, some of the rationale for considering this option is as follows:
- Several verses in the NT seem to indicate that Jesus’ humanity was no different from ours, except that he did not sin. Cf. Heb 4.15 (“tempted in every way just as we are, yet without sin”; Heb 5.8 (he “learned obedience”); Luke 2.52 (he “increased in wisdom and in stature, and in favor with God and with people”11).
- These texts seem to indicate that Jesus’ growth as a human being was along essentially normal lines. Thus, a part of this almost surely involved such things as the following: (1) he probably stumbled and fell the first time he stood up to walk. Would we really expect him to rise and walk without stumbling on his first attempt? That view of our Lord seems to be more docetic than orthodox. (2) He probably hit his thumb with a hammer working for Joseph (though he didn’t swear when he did it!); (3) He probably made Hebrew (or Aramaic or Greek) grammatical, pronunciation, and syntactical mistakes as he learned the language, being gently corrected by his mother. This is almost surely the case, for to learn a language well requires interaction, trial and error, correction, and instruction. If Jesus did not say anything until he was sure he was right, then his childhood would be marked out as both ostentatious and exceedingly quiet; but if his childhood proceeded along normal lines, and if he always used the correct grammatical forms, then he would most likely have had to learn at least some of those forms supernaturally. But if that is the case, then his childhood was anything but normal, and would seem to be an implicit denial of the principles taught in Luke 2:52. And (4) he probably made a whole host of other mistakes that would not be classified as sin.
- If Jesus made mistakes in certain areas as a child—even in the area of knowledge—why should we suppose that he did not do so as an adult? Did he stop learning as an adult? Hebrews 5.8 says no; he continued to learn obedience through the things he suffered. Although the learning here is related to experiencing obedience as a human being, it is still learning and it takes place preeminently in Jesus’ adult life, reaching its climax in the crucifixion.12
- As well, there is clear evidence in the gospels that Jesus’ omniscience was not always on a human conscious level. Many texts address this, but chief among them is Mark 13:32: “No one knows about that day or hour, not even the angels in heaven, nor the Son, but only the Father” (NIV).
- In sum, although we may well feel uncomfortable with this approach, it must be admitted that to see Jesus err in Mark 2.26 is not to deny his deity, for a mistaken identification is not the same as sin. We will revisit this approach later.
2.b. Jesus embellished.
Gundry suggests the following13:
To strengthen his argument, Jesus adds a number of features not found in the OT passage: (1) David’s having companions with him (contrast 1 Sam 21:2-3 [1-2]); (2) his having need; (3) his and his companions’ being hungry; (4) the house of God and David’s entering it rather than merely asking for bread; (5) Abiathar’s being a “high priest,” not just a “priest”; (6) David’s eating the loaves of presentation, either while he is still inside the house of God or after he has come out; and (7) his giving some of the loaves to his companions. Moreover, the OT text speaks of Ahimelech, not of Abiathar. … Apparently, then, Jesus not only adds a number of features. He also replaces Ahimelech with Abiathar the son of Ahimelech for a link with the added house of God, which for Jesus and his audience stands in Jerusalem, where Abiathar officiated (2 Sam 15:24, 35; 17:15; 19:12 ), not in Nob, where Ahimelech gave bread to David.
This line of argument is in keeping with Gundry’s earlier (and infamous!) commentary on Matthew, in which he argued that Matthew’s gospel finds its closest genre parallels in Jewish midrash. He now sees Jesus following the same hermeneutical method. In his conclusion to this problem, Gundry forcefully argues: “The fact that when Abiathar does appear in 1 Sam 22:20-23 he does so in connection with the foregoing incident at the house of God makes it easy for Jesus to use his name in blotting out Ahimelech for the sake of a link with Jerusalem.”14 What is interesting is that Gundry apparently changed his views from his doctoral days at Manchester; his doctoral thesis, published under the title, The Use of the Old Testament in St. Matthew’s Gospel with Special Reference to the Messianic Hope15, argued essentially that Matthew’s and Jesus’ use of the OT was hermeneutically quite different from rabbinic exegesis. For example, on 215 he says, “Most of all, the theological depth and coherence of the hermeneutical principles (in sharp contrast with Qumran and rabbinic exegesis) demand the unique genius of the kind of man Jesus must have been—they cannot reasonably be set down to Gemeindetheologie.” Thus, it might not be unfair to ask ‘Which Gundry?’ when wrestling with his recent views on Mark 2.26.
Nevertheless, his rabbinic views are generally found in much older, and sometimes even conservative, literature. In the last two hundred years, such notable scholars as Christopher Wordsworth, James Morison, A. E. J. Rawlinson, Hugh Anderson, and J. Bowman have held to one form or another of the midrashic approach.16
What shall we say about Gundry’s treatment? At least in comparison with Brown’s approach, Gundry is more conservative than his Roman Catholic counterpart; he is an evangelical who embraces inerrancy, though his definition allows for quite a bit of latitude. Nevertheless, Gundry’s approach may well be more troubling for many evangelicals than Brown’s, for it may seem to some to be a case of scripture-twisting on Jesus’ part. Mistaken identification is one thing; intentional alteration is another. Perhaps more important is the criticism that Cohn-Sherbok leveled against a midrashic Jesus: “though Jesus seems to have been familiar with rabbinic hermeneutics, the arguments he employs are invalid from a rabbinic point of view.”17 One wonders, along these lines, why Matthew—whose gospel is surely the most Jewish of the synoptics—omits “when Abiathar was high priest” if he learned his hermeneutics from Jesus (as Gundry earlier affirmed). Why would a midrashic Jesus here cause problems for Matthew?
3. Source-critical: Mark’s source (Peter?) made a mistake in reporting Jesus’ words, or else was intentionally midrashic.
Several scholars indicate that the problem in Mark 2.26 may have been due to the evangelist or to his source(s).18 Now if Peter was the source behind Mark’s gospel, as early patristic writers suggest,19 it is possible that he added to the dominical saying as he recalled Jesus’ teachings. If so, Mark could have faithfully copied down what Peter wrote, even to the point of recording his mistaken identification of the priest at Nob. This is not much different from what some scholars argue regarding Luke’s recording of Stephen’s speech in Acts 7: there may be historical errors in the speech, but Luke was faithful to record his speech, warts and all.20 There is of course a certain attractiveness to this view: It absolves both Jesus and Mark from error, and by so doing maintains both a high bibliology and a high christology. Even though Peter was an apostle, in this instance he would not have written scripture. Thus, his oral sermons could hardly be viewed as inspired.
The problem with this view is that if Peter is the source, then that means that he would have most likely given this sermon on dozens of occasions. Surely someone would have corrected him on his historical blunder long before Mark ever wrote down Peter’s memoirs. It is quite different if Jesus or Mark is the source for the ἐπὶ ᾿Αβιαθὰρ ἀρχιερέως. If Jesus said it, the phrase could easily have remained in the oral tradition out of respect for Jesus’ words, even if there were questions as to what was meant by them. However, if Peter is the source of the phrase, in the least one of the apostles would surely have pointed out the error of his ways. An intermediate source, especially if it is Peter, then, remains one of the least likely options.
On a midrashic approach, the same problems as are mentioned with #3 also are raised here, only more so!
4. Mark erred in reporting what his source said, or was intentionally midrashic.
That Mark may have been midrashic is not very likely; such would have been lost on his audience. But that he could have created the error here, or have gotten it from a written source (as opposed to oral tradition), is more likely. This is the view that is probably the most popular among critical scholars. Many commentators simply assume this is the case, without much comment beyond mere assumption. So Meyer, Wendling, Hultgren, Tolbert, O’Connell, Turner, Morgan, Kiilunen, and Pesch,21 among others, take this route. Kiilunen is representative: he unceremoniously calls the ἐπὶ ᾿Αβιαθὰρ ἀρχιερέως phrase “das Fehlen.”22 Others spend much more time on the issue but come to the same conclusion. So Lagrange, Swete, Guelich, Hawkins, and Casey.23
A point often put forth in this connection is that the OT seems to confuse the two names at times. Hurtado, for example, notes: “It is possible that the Markan account is confused here, for the OT itself is not easy to follow in its references to Ahimelech and Abiathar. In 1 Sam. 22:20, Abiathar is described as son of Ahimelech; whereas 2 Sam. 8:17 and 1 Chron. 24:6 refer to Ahimelech as son of Abiathar and as priest under David.…”24 This is an old view, with a long list of patristic writers and later authorities embracing some spin on it. Chrysostom, Victor of Antioch, Euthymius Zigabenus, Theophylact, Beza, Heumann, Kuinoel, Garland, Hurtado, Guelich, and many others mention it. It takes two forms. First, the OT is confused, or at least the copies are confused. Second, the OT is correct and both men shared the same name. 25 If the OT erred, this would hardly absolve Mark. As the adage goes, two wrongs don’t make a right. But what of the likelihood that both men shared the same name? In his defense of Matthean priority, Buchanan went so far as to suggest that “According to the LXX, well-known to all three evangelists, the priest who gave David the Bread of the Presence was Abiathar. It is not likely that either Matthew or Luke would have omitted the LXX account just because it did not agree with the MT.”26 But Morgan proved that Buchanan was wrong: no extant LXX MSS in 1 Sam 21 read “Abiathar.”27 Even if these two men’s names were sometimes confused, they are not so in 1 Sam 21. Only Ahimelech is seen there, as far as any extant witnesses reveal. And since that is the story that Jesus is referring to, the difficulty of the mention of Abiathar remains.
5. Hermeneutical: The interpretation that “when Abiathar was high priest” is incorrect.
Several solutions present themselves here, but two predominate. First, it is possible that the reference to Abiathar is not to the person per se but to the section of scripture that is being alluded to—thus, “in the portion relating to Abiathar.” Michaelis (in his Introduction to the New Testament) seems to have been the first to propose this view. So Mark 12.26: οὐκ ἀνέγνωτε ἐν τῇ βίβλῳ) Μωυ¨σέως ἐπὶ« τοῦ βάτου (“have you not read in the book of Moses, in [the passage about] the [burning] bush”). Robinson28 finds a parallel in the Mishnah: “Whoever confesses his guilt shall have a portion in the world to come; for so we find in Achan [i.e., in the section of the book of Joshua about Achan] that Joshua said to him, ‘My son, give glory to the Lord…” Though apparently dormant for some time, Wenham resurrected the view in the 20th century, and has been followed by Lane, Roure, and a few others.29
The major problem with this solution is that it requires linking ἐπὶ ᾿Αβιαθὰρ ἀρχιερέως with οὐδέποτε ἀνέγνωτε at the beginning of v. 25. But there is too much distance between the words to do this naturally. Such a reading, precisely because it is not natural, would probably never have presented itself except for the historical problem of the text. Interestingly, Lane, who apparently embraces this view, nevertheless offers an excellent critique of it:
The objections which may be raised against this proposal are that ἐπὶ ᾿Αβιάθαρ [sic] ἀρχιερέως is considerably separated from ‘have you not read,’ unlike Ch. 12:26; that Abiathar is by no means the central element in this section of I Samuel; that the introduction of Abiathar first in Ch. 22 constitutes it unlikely that his name would be given to the section; and that numerous instances in Tannaitic documents indicate that a section was usually designated by a term which occurs early, not late, in the section. The strongest argument for this proposal is the undoubted use of ἐπί cum genitive in Ch. 12:26 to indicate a section of Scripture.30
The second possible hermeneutical solution is that ἐπὶ ᾿Αβιαθὰρ ἀρχιερέως could possibly be translated “in the days of Abiathar the high priest.” This was the view of Grotius, Wetstein, Wordsworth, Scholz, and many others. It is the wording of the KJV as well, though the KJV is based on a different text here (which has τοῦ before ἀρχιερέως). Thomas Fanshaw Middleton, in his still unexcelled treatment of the article in the Greek NT, spends much time on this interpretation, but he bases his views on the articular reading.31 Indeed, Middleton provides the basis for this view’s rejection: “That reading [the one without the article which is adopted in NA27]… would indeed mean, that Abiathar was actually High Priest at the period in question.32 Middleton cites several classical references to back up his statement. In grammatical terms, we could say ἐπὶ ἀρχιερέως involves a predicate genitive (“when Abiathar was high priest”) while ἐπὶ ᾿Αβιαθὰρ τοῦ ἀρχιερέως involves an appositive to ᾿Αβιαθάρ (“in the time of Abiathar the high priest”).33
Nevertheless, several modern scholars adopt this view. Standard works such as BDAG and BDR argue this without much fanfare—and, unfortunately, without much basis!34 Edwards makes a stronger case: “Mark’s wording… employs epi technically to mean ‘in the time’ (so 1 Macc 13:42; Luke 3:2; Acts 11:28; Martyrdom of Polycarp 21).”35 But these texts do not help the case as much as he would suppose: they were cited by Swete and Middleton on behalf of the opposite view, viz., that “when an anarthrous title is added to the personal name, the period is limited to the term of office.”36 I do not yet know of any texts in which the construction ἐπί + genitive of personal name + anarthrous title indicates the general “in the time of,” though one or two seem to come close (see appendix). To be sure, I have not searched very diligently for this construction. But I am not satisfied that BDR and BDAG have supplied sufficient evidence on its behalf.37
In 1883, Thomas M. Lindsay could write about the Abiathar problem: “Various explanations of the difficulty have been given, none very satisfactory.”38 It’s one hundred and twenty-one years later and you may feel, as do I, that if Lindsay were to rise from the dead he’d repeat his complaint verbatim!
But we must put this problem in perspective. What is at stake? Is the deity of Christ at stake? Apparently not, for two of the leading advocates of the “Jesus erred/midrashed” view embrace the deity of Christ. Is the inerrancy of scripture at stake? Possibly so, for if either option 3(a), or 4(a) is adopted, inerrancy cannot hold up. Is the infallibility of scripture at stake? Ironically, it seems to be so only if Gundry’s view is given full force and if Jesus’ use of scripture would have been perceived as self-serving and as eisegetical, for Jesus’ invoking of scripture here is directly related to a matter of faith and practice.
Second, what options seem to be excluded by the evidence? Option 1 (text-critical) is clearly out. The others all have some merit on the basis of evidence.
Third, how and when should our theological presuppositions enter the picture as we try to handle the data of the text honestly? In the least, it is imperative that we not frontload our presuppositions to such an extent that we don’t listen to the text. Evangelicalism is populated with all sorts of academic gatekeepers whose theological a priori drives their investigation and determines its results. The tragic irony is that such people never really learn from the text, for they have already decided what it will tell them. At the same time, we must not think that exegesis can ever be presuppositionless. That notion went out with the demise of historical positivism. There is a difference between giving generally reliable witnesses—reliable as determined by a sound historical method—the benefit of the doubt and assuming that the biblical writers cannot possibly err.
Fourth, this leads to a taxonomy of the doctrine of scripture. With reference to bibliology, I believe we should first and foremost embrace the Bible as a witness to the great acts of God in history, especially to the Christ-event. This is enough for salvation. Second, we should recognize it as an infallible guide in matters of faith and practice. This is needed for sanctification as well. These two pillars seem to be the hallmark of the Church throughout its history, until recently. And third, we should see it as inerrant—true in what it touches. This basically is a safeguard for infallibility, but must never supersede the first two credos about scripture. For when it does, then the incarnation is dishonored. Thus, inadvertently, when we frontload inerrancy and refuse to really probe the tough historical questions, we end up betraying our commitment to the incarnation. The deepest tragedy along these lines is when someone never differentiates doctrinal commitments, for this leaves him wide open to chucking his entire belief system when the weakest link is broken. From experience, I can tell you that this “domino view of doctrine” is altogether too prevalent and has been the ruin of a great many evangelical doctoral students.
Fifth, how can we assess these various options? It must be admitted that views 2-5 all have a certain plausibility.39 If you were to decide to opt for 2, 3, or 4, I would simply plead with you not to abandon Christ. If your bibliology goes down a notch or two because of this problem, the deity of Christ in the least should still be insulated—unless of course you hold to a domino view of doctrine! Further, if I were to decide that view 3 or 4 was the most compelling, and that this decision resulted in my abandoning inerrancy, it would be a gross distortion to call me liberal! At the same time, one must be very careful about making major theological shifts, especially before the data have been sufficiently examined. Caution is needed when examining material that could change your theological commitments.
Along these lines, I am reminded of what a sage wrote nearly one hundred and fifty years ago. J. A. Alexander concluded, concerning this passage, “It is best, however, as in all such cases, to leave the discrepancy unsolved rather than to solve it by unnatural and forced constructions. A difficulty may admit of explanation, although we may not be able to explain it, and the multitude of cases in which riddles once esteemed insoluble have since been satisfactorily settled, should encourage us to hope for like results in other cases…”40
Appendix: my present preference
My own preference is for view 5: I believe that we have interpreted the text incorrectly. I am least comfortable with Gundry’s view (2.b., and by implication 3.b. and 4.b.): to see Jesus’ use of the OT as midrashic is to overturn all the work that Gundry had done earlier; further, his view of Matthew’s hermeneutic as essentially midrashic seems to ignore the ostensibly far closer parallels of Mark and Luke. If Gundry argues that Mark is also midrashic, then Luke must surely follow, for Luke does not differ too significantly from his source. But Luke can hardly be midrashic, for he opens his gospel by echoing Thucydides’ historical principles. Further, if Luke is midrashic, then virtually everything is up for grabs in the gospels, with all the historical and archeological spadework of two millennia being tacitly ignored.
But I am also not particularly comfortable with Brown’s approach, for two reasons: (1) Although Jesus certainly displayed ignorance on occasion by way of omission (e.g., not knowing the date of his return), that seems to be qualitatively different from a statement that involved error. Further, although he almost surely made mistakes as a child while in the process of learning, I tend to view Luke 2.52 as indicating the growth that produced the adult and mature man (thus making his adulthood on a different plane than his childhood). (2) Although Brown defends the deity of Christ, as a Roman Catholic his view of God is not the same as the Protestant view, especially the Reformed Protestant view. Catholic perceptions of God’s sovereignty and majesty tend to be semi-Pelagian, while Protestant views range from Arminian to Calvinistic. Thus, in Brown’s view, it might possibly be easier to affirm an error in Jesus’ statement because such does not impugn his doctrine of God. (For example, although Catholics embrace the omniscience of God, their very doctrine of conditional election seems to presuppose a growth in God’s knowledge and on that basis he chooses.) Views 3 and 4 simply push the issue away from Jesus, making Peter or Mark the errant party. But there is actually little to commend the notion that they did not get this statement about Abiathar from Jesus.
As for view 5, my preference right now is to take the prepositional phrase as meaning “in the days of Abiathar the high priest.” Although Mark apparently does not employ the temporal use of this preposition elsewhere, he almost surely does so here—for both “when Abiathar was high priest” and “in the days of Abiathar the high priest” are temporal expressions. Further, the construction ἐπί + genitive noun is frequently used with a temporal sense outside of Mark—with a meaning similar to ‘in the days of…’ BDAG lists numerous biblical and patristic references under ἐπί with a genitive for time, all in the sense of “in the time of, under (kings or other rulers).” Cf., e.g., Luke 4:27 (‘in the time of Elisha’), Luke 3:2 (‘in the time of the high priest, Annas and Caiaphas’) and even Mark 2:26 (‘in the time of Abiathar the high priest’). Two questions remain: (1) Can any of these texts mean ‘in the time of’ as distinct from ‘when’? That is, can they mean something like “the 1990s will forever be linked to Clinton’s presidency,” even though he was not president for the whole decade? (2) If so, do any of them have ἐπί + genitive proper noun, followed by an anarthrous common noun? Without examining all the data supplied by BDAG, Luke 3:2 looks to be the closest parallel to Mark 2:26, even though ‘high priest’ comes before the two names (the grammatical meaning differs when the proper name comes second; no article is required). But if these two men did not function as high priest simultaneously—and since the singular event of the word of the Lord coming to John the Baptist was during their high priesthood, then this seems to be a clear text in support of the general time frame of ‘in the days of.’ More work certainly needs to be done, but suffice it to say that this view has a certain plausibility and cannot be hastily rejected.
1 There are several other interpretations that have been put forth besides these five broad categories. James Morison, Mark’s Memoirs of Jesus Christ: A Commentary on the Gospel according to Mark (London: Hamilton, Adams, & Co., 1873) 67-70, gives the most comprehensive discussion I have found, listing ten different interpretations, some of which are still popular today. As well, others have offered idiosyncratic views (or at least views that never commanded much of a following). For example, Lightfoot in 1658 argued that “Abiathar” = Urim and Thummin (John Lightfoot, A Commentary on the New Testament from the Talmud and Hebraica: Matthew—I Corinthians (reprint; Grand Rapids: Baker, 1979) 2.402): ““It is well enough known what is here said in defence of the purity of the text; namely, that Ahimelech the father was called Abiathar, and Abiathar the son was called Ahimelech. But I suppose that something more was propounded by our Saviour in these words. For it was common to the Jews under Abiathar to understand the Urim and Thummin. Nor without good reason, when it appears, that under the father and the son, both of that name, the mention of inquiring by Urim and Thummin is more frequent than it is ever anywhere else; and, after Abiathar the son, there is scarcely mention of it at all. Christ therefore very properly adds, ἐπὶ ᾿Αβιάθαρ [sic] ἀρχιερέως, in the days of Abiathar the high priest, therein speaking according to a very received opinion in the nation: as though he had said, ‘David ate the shewbread given him by the high priest, who had the oracle by Urim and Thummin present with him, and who acted by the divine direction.’“ This novel view has had little following. Lenski held a view that was only slightly less improbable (R. C. H. Lenski, The Interpretation of St. Mark’s Gospel (Columbus, OH: Wartburg, 1946) 127-8: “Another solution is that the father and the son were both present when David came to Nob, and both gave the bread to David. Ahimelech, the father, soon died, and Abia- thar, the son, became high priest and made a record of the facts, which are thus rightly said to have taken place in his day.” Alexander criticized a variant of this interpretation in his day as follows (J. A. Alexander, The Gospel according to Mark (New York: Scribner, 1858) 54: “Another explanation of the discrepancy is that the Greek phrase means in the presence of Abiathar, although Ahimelech performed the act. But even if that were so, which is assumed without the slightest proof, why should a person merely present have been named, when the act in question was performed by another?”
2 Illustrations of this mentality are not difficult to find. J. A. Alexander, The Gospel according to Mark (New York: Scribner, 1858) 53: “Even if no solution could be given of this discrepancy, it would be absurd to let it shake our faith in the substantial truth of either narrative. … Even if the passage be retained, and in its ordinary form, there are several possible solutions, any one of which is far more likely than the supposition of a contradiction or a blunder, which would certainly have been detected and expunged, instead of being cherished and transmitted to posterity.” But Alexander concludes his discussion with the sober advice (54): “It is best, however, as in all such cases, to leave the discrepancy unsolved rather than to solve it by unnatural and forced constructions. A difficulty may admit of explanation, although we may not be able to explain it, and the multitude of cases in which riddles once esteemed insoluble have since been satisfactorily settled, should encourage us to hope for like results in other cases…”
3 William R. Farmer, The Synoptic Problem: A Critical Analysis (Macon, GA: Mercer University Press, 1976).
4 The Gospels as Historical Documents, 1909. Although he does not indicate which page is quoting from, it is 145.
5 Farmer, Synoptic Problem, 110.
6 Stanton did not embrace Matthean priority, but Farmer uses this argument of Stanton’s to bolster his own case. He says nothing more about the matter.
7 Both men embraced Matthean priority. See C. S. Mann, Mark, AB 27 (Garden City, NY: Doubleday, 1986) 238; E. P. Sanders, “Suggested Exceptions to the Priority of Mark,” in The Two-Source Hypothesis: A Critical Appraisal, edd. Arthur J. Bellinzoni, Joseph B. Tyson, and William O. Walker ([Macon, GA:] Mercer University Press, 1985) 203. Others who entertain the textual solution include McNeile, Bartlett, Branscomb, and Sherman Johnson. But apparently so does Taylor (Vincent Taylor, The Gospel according to St. Mark, 2nd ed. [London: Macmillan, 1966] 217).
8 Several modern translations have a reading that seems to be a translation of ἐπὶ ᾿Αβιαθὰρ τοῦ ἀρχιερέως, especially the more evangelical translations (cf. NIV, ESV). In the least, an alternative rendering or a text-critical note would perhaps be warranted here. See later discussion.
9 D. M. Cohn-Sherbok, “An Analysis of Jesus’ Arguments concerning the Plucking of the Grain on the Sabbath,” JSNT 2 (1979) 31-41; here quoting from 31.
10 Raymond E. Brown, An Introduction to New Testament Christology (New York: Paulist, 1994) 37-38. In addition to Brown, others suggest this option. C. E. B. Cranfield, The Gospel according to St Mark (Cambridge Greek Testament Commentary; Cambridge: CUP, 1959) 116: “ἐπὶ ᾿Αβιαθὰρ ἀρχιερέως must mean ‘when Abiathar was High Priest.’ … A C Θ and a good many other MSS. insert τοῦ before ἀρχιερέως. The phrase then means ‘in the days of Abiathar the High Priest,’ which need not imply that he was actually High Priest at the time. The variant is probably due to a sense of the historical difficulty. The fact that D W it sys omit the phrase altogether—as do Mt. and Lk.—makes the suggestion that the whole phrase is a misguided gloss not unreasonable. But it is perhaps more likely that Jesus himself or possibly Mark mentioned Abiathar as the High Priest particularly associated with David, forgetting that at the time of the incident he was not yet High Priest.” In this discussion, Cranfield seems to entertain the notion that Jesus erred as the leading solution. So also Roger E. Van Harn, editor, The Lectionary Commentary: Theological Exegesis for Sunday’s Texts. The Third Readings: The Gospels (Grand Rapids: Eerdmans, 2001) 194: “But interestingly, Jesus’ reference is technically incorrect. 1 Samuel 21:1-6 tells us that Ahimelech was actually the high priest during the episode to which Jesus refers. In the parallel texts, neither Matthew (12:1-14) nor Luke (6:1-11) names the high priest, which could be their way of correcting Mark by silence. If, historically speaking, Mark quotes Jesus correctly, then Jesus was either wrong in his citation or intentionally ‘gets it wrong’ to tweak them in defiance of their authority standards for precision. Indeed, Jesus was not above deconstructing a text (Mark 12:35-37) in order to get a rise out of his opponents, which puts a different spin on our common perceptions of ‘What Would Jesus Do?’“
11 Translations in this paragraph are from the NET Bible.
12 LSJ gives as the primary definition of μανθάνω (the verb translated ‘learned’ in Heb 5.8) “to learn, esp. by inquiry.” LN give three definitions:  “to acquire information as the result of instruction, whether in an informal or formal context” (§27.12),  “to learn from experience, often with the implication of reflection” (§27.15),  “to come to understand as the result of a process of learning” (§32.14). They place Heb 5.8 under definition 2.
13 Robert H. Gundry, Mark: A Commentary on His Apology for the Cross (Grand Rapids: Eerdmans, 1993) 141.
14 Ibid., 142.
15 Leiden: Brill, 1967 (reprinted without substantial changes in 1975).
16 Morison, Mark’s Memoirs, 70, adopts the view that the prepositional phrase should be translated “in the days of Abiathar.” But he adds a midrashic twist, quoting from an unspecified source written by bishop Wordsworth: “If our Lord had mentioned Ahimelech, the Pharisees’ answer might have been that Ahimelech was punished by God for this profanation of sacred things; he and his were soon overtaken by divine vengeance and slain. But by specifying Abiathar, who was then with his father (1 Sam. xxii, 20), and who (we may reasonably infer from our Lord’s words, which are the words of Him who knows all history) was a party to his father’s act, and was afterwards blessed by God in his escape, and in a long and glorious priesthood, our Lord obviates the objection of the worldly-minded Pharisees, and strengthens his own argument, by reminding them that this action took place in the time and under the sanction of one whom they held in reverence as a venerable ornament of the pontifical family and dignity.” A. E. J. Rawlinson, St Mark, 3rd ed. (London: Methuen, 1931) 34: “Our Lord appears to follow a traditional Jewish ‘haggada’ or expansion of the O.T. story.” He adds nothing else to this point, however. So also Hugh Anderson, The Gospel of Mark (NCBC; Grand Rapids: Eerdmans, 1976) 110. J. Bowman, “Abiathar,” International Standard Bible Encyclopedia (Grand Rapids: Eerdmans, 1979) 1.7: “Jesus uses the incident in the manner of Haggadic midrash, i.e., with the aim of illustrating His message rather than recounting history. For this purpose it is more apt that Abiathar, the priest at Nob and later high priest at Jerusalem, should be the central figure in the story rather than his father Ahimelech; and it is imperative that he be styled high priest in spite of the mistaken, or deliberately altered, reading at 2 S. 8:17 and the derivative 1 Ch. 18:16; 24:6.” Perhaps most surprisingly, E. Schuyler English, a staunch conservative, adopts this position, apparently deriving his views from Wordsworth (though without any credit given): cf. E. Schuyler English, Studies in the Gospel according to Mark (New York: Our Hope, 1943) 63.
17 Of course, this argument might be countered in that Jesus was not using the example of David as a mere precedent that should be followed, even less as an example of illicit behavior. These are the issues that Cohn-Sherbok raises. If Jesus was hinting that one greater than David, greater than the high priest, and greater than the temple was in their midst, then his arguments would of course not be strictly rabbinic and would not be acceptable to the Pharisees. This would not make them any less legitimate.
18 These will be discussed in the following section.
19 The universal testimony of the early fathers connects this gospel with Peter. For example, Papias writes: “And the elder said this: ‘Mark became an interpreter of Peter; as many things as he remembered he wrote down accurately (though certainly not in order) the things said or done by the Lord. For he neither heard the Lord nor followed him, but he came later—as he said with reference to Peter who taught whenever the need arose, but he did not [teach] according to the arrangement of the oracles of the Lord, with the result that Mark did not err when he thus wrote certain things as he recalled them. For he planned out one goal ahead of time, namely, to leave out nothing which he heard and not to falsify any [of the words of Peter]” (my translation of Fragments of Papias 2.15 (also recorded in Eusebius, HE 3.39.15).
20 By this we are not implying that Luke recorded the ipsissima verba of Stephen’s or anyone else’s speech, nor that Mark did this with Peter (or Jesus). Rather, our point here is simply that faithful copying would get the gist of what the source had to say, even down to some particulars.
21 H. A. W. Meyer, The Gospels of Mark and Luke (Critical and Exegetical Commentary on the New Testament; Edinburgh: Clark, 1890) 1.45: “Mark has erroneously confounded these two…”; Emil Wendling, Die Entstehung des Marcus-Evangeliums (Tübingen, 1908) 11; Arland J. Hultgren, “The Formation of the Sabbath Pericope in Mark 2:23-28,” JBL 91 (1972) 38-43, argues simply that v. 26 does not go back to Jesus (40-41); M. A. Tolbert, “Is It Lawful on the Sabbath to Do Good or to Do Harm: Mark’s Ethics of Religious Practice,” PerspRelStud 23.2 (1996) 199-214, esp. 208, implicitly lays blame at Mark’s feet for the error; L. J. O’Connell, “Boismard’s Synoptic Theory: Exposition and Response,” Theology Digest 26.4 (1978) 325-42, esp. 335: “Lk omits Mk’s erroneous reference” (see also 336, 337); C. H. Turner, The Gospel according to St. Mark (London: SPCK, n.d.) l.c.; C. Shannon Morgan, “‘When Abiathar was High Priest’ (Mark 2:26),” JBL 98 (1979) 409-10; Jarmo Kiilunen, Die Vollmacht im Widerstreit: Untersuchungen zum Werdegang von Mk 2,1—3,6 (Helsinki: Suomalainen Tiedeakatemia, 1985) 200; Rudolf Pesch, Das Markusevangelium (HTKNT; Freiburg: Herder, 1976) 1.182, n. 15.
22 Kiilunen, Vollmacht, 200.
23 M.-J. Lagrange, Evangile selon Saint Marc (Paris: Librairie LeCoffre, 1966) 53-55; Henry Barclay Swete, The Gospel according to Mark (London: Macmillan, 1913) 49; Robert A. Guelich, Mark 1:1-8:26 (WBC; Dallas: Word, 1989) 122; John C. Hawkins, Horae Synopticae, 2nd ed. (Oxford: Clarendon, 1909) 122. Casey’s treatment is perhaps the most elaborate: he argues that Mark’s Aramaic source was translated incorrectly: Maurice Casey, “Culture and Historicity: The Plucking of the Grain (Mark 2.23-28),” NTS 34 (1988) 1-23. P. 8: “This is one of the mistakes of the Marcan narrative. Its origin may be discovered by retroversion into Aramaic: ביומה אביתר כהן רב. Abiathar was much more important than Ahimelech, and his presence may reasonably by deduced from the narrative in 1 Samuel. כהן רב meant only that he was one of the most important religious authorities, the ἀρχιερεῖς of the later Marcan narrative.”
24 Larry W. Hurtado, Mark (NICBC; Peabody, MA: Hendrickson, 1989) 54.
25 2 Sam 18.17 calls “Ahimelech the son of Abiathar”; 1 Chron 18.16 speaks of “Abimelech the son of Abiathar” (MT, followed by NASB; the NIV and NRSV have ‘Ahimelech’ for ‘Abimelech’ [with the support of LXX, Syriac, Arabic, and Vulgate] and the REB both swaps out Ahimelech for Abimelech and reverses the order [‘Abiathar the son of Ahimelech’!], apparently without MS support, to conform it to 2 Sam 8.17. Here is an instance of the REB being more evangelical than the NASB!) 1 Chron 24.3 associates Zadok with Ahimelech, while 1 Chron 15.11 and 2 Sam 15.29, 35 associate Zadok with Abiathar.
26 G. W. Buchanan, “Has the Griesbach Hypothesis Been Falsified?” JBL 93 (1974) 550-72, quoting 562.
27 Morgan, “‘When Abiathar was High Priest’ (Mark 2:26),” 409-10.
28 Robinson, The Evangelists and the Mishna, 169-70.
29 J. W. Wenham, “Mark 2,26,” JTS n.s. 1 (1950) 156; William L. Lane, The Gospel of Mark (NICNT; Grand Rapids: Eerdmans, 1974) 116: “An attractive proposal is that Mark’s intention has been misunderstood in the translation of the passage. The same grammatical construction occurs in Ch. 12:26, where it must be translated ‘have you not read in the book of Moses, in the passage concerning the Bush, how God spoke unto him…?’ The construction is designed to call attention to the section of a biblical book where the reference is found… In Ch. 2:26 Mark may have inserted the reference to Abiathar to indicate the section of the Samuel scroll in which the incident could be located.” Damia Roure, Jesús y la Figura de David en Mc 2,23-26 (Analecta Biblica 124; Roma: Editrice Pontificio Istituto Biblico, 1990) 14, goes so far as to translate the phrase as “en el pasaje del sumo sacerdote Abiatar.” Perhaps also R. Alan Cole, The Gospel according to Mark, 2nd ed. (Tyndale New Testament Commentaries; Leicester, England: Inter-Varsity, 1989) 129, n. 1.
30 Lane, Mark, 116, n. 86.
31 T. F. Middleton, The Doctrine of the Greek Article, new ed. (London: Rivington, 1841) 188-90.
32 Ibid., 189.
33 Middleton, Swete, and others cite texts such as 1 Macc 13.42, Luke 3.2, Martyrdom of Polycarp 21, as evidence that the anarthrous construction meant “when so and so was such and such” rather than “in the days of so and so.”
34 BDAG 367, 18.a., s.v. ἐπί: ‘in the time of.’ Without further explanation. Several passages are cited in support, but they are either very general or suggest simply ‘when.’ BDR §234.5 (187): “Öfters temporal zum Ausdruck der Gleichzeitigkeit (klass.): Mk 2,26 ἐπὶ ∆Αβιαθὰρ ἀρχιερέως “zur Zeit.” The following texts are cited in support (n. 8): Matt 1.11; Eph 1.16; Heb 1.2; Acts 11.19 v.l. But these are rather tapered parallels. Curiously Rehkopf also says that perhaps (“vielleicht”) Mark 12.26 also should be included!
35 James R. Edwards, The Gospel according to Mark (Pillar; Grand Rapids: Eerdmans, 2002) 95, n. 42.
36 Swete, Mark, 49.
37 Gundry offers the curious argument that “Mark uses ἐπί in a temporal sense nowhere else, no matter what case follows” (Mark, 141). This may say too much, for “when Abiathar was high priest” (the translation that Gundry prefers) is also a temporal statement! Further, of the 21 instances of ἐπί + genitive in Mark, all but three or four have a geographical/place name as the object. Hence, the semantic situation is not the same as what we have in 2.26.
38 Thomas M. Lindsay, The Gospel according to St. Mark (Edinburgh: Clark, 1883) 91.
39 For convenience’ sake, the options are again listed below:
Text-critical: the text is wrong and needs to be emended;
Dominical: Jesus is wrong (or midrashic) and this needs to be adjusted to;
Source-critical: Mark’s source (Peter”) is wrong (or midrashic);
Mark is wrong (or midrashic);
Hermeneutical: our interpretation is wrong and needs to be altered.
40 Alexander, Mark, 54.
Related Topics: Textual Criticism