Notable Markan Variants in Codex W
In earlier postings I’ve noted that major variants continue to appear, and in some cases “succeed” in gaining widespread acceptance in the subsequent manuscript tradition as late as the fifth century and thereafter. There are also examples of interesting variants that first appear in the extant Greek manuscripts that late, but didn’t “succeed”. Having done my PhD thesis on the text of the Gospel of Mark in Coldex W (the Freer Gospel codex), I’ll cite a few interesting variants in this manuscript as examples.[1]
I’ll begin with a variant that is more well known, the so-called “Freer Logion.” This variant appears only in Codex W among extant Greek manuscripts after what we know as Mark 16:14 (in the “long ending” variant). Here is an English translation:
And they excused themselves, saying, “This age of lawlessness and unbelief is under Satan, who does not allow the truth and power of God to prevail over the unclean things of the spirits [or, does not allow what lies under the unclean spirits to understand the truth and power of God.] Therefore reveal your justice now”— thus they spoke to Christ. And Christ replied to them, “The term of years of Satan’s power has been fulfilled, but other terrible things draw near. And for those who have sinned I was delivered over to death, that they may return to the truth and sin no more, in order that they may inherit the spiritual and incorruptible glory of justice which is in heaven.”[2]
A version of this variant is attested as known to Jerome in some Greek manuscripts in the fourth century.[3] But Codex W is the only extant example. Metzger opined that the logion was probably first added to the text of the “long ending” of Mark by a “scribe” sometime in the second or third century “to soften the severe condemnation of the Eleven in 16.14.”[4] The ascription of the motive for the variant seems cogent, but, as with other “intentional” variants, we should more plausibly ascribe the Freer Logion to some unknown reader/user of a copy of the Gospel of Mark.[5] But my main point is that, whatever its point of origin, this variant didn’t “succeed” in being accepted into the manuscript tradition of Mark beyond Codex W.
A far less well-known variant in the text of Mark in Codex W appears at Mark 1:3. Whereas the best-attested text in vv. 2-3 is a slightly variant quotation of Exodus 23:20 and Isaiah 40:3, in Codex W the quotation of Isaiah includes the whole of 40:3-8. Again, Codex W is the only witness in the Greek manuscript tradition for this variant. And again I think it likely that the variant arose in the course of someone reading the text and adding the words of Isaiah 40:4-8, perhaps originally as a marginal note. In any case, this is another variant of significant size that just didn’t “succeed”.
I cite now a third instance of a variant apparently unique to Codex W. This one is much smaller, but I judge it clearly intentional. I’ll have to set the scene in Mark where the variant appears. In Mark 3:21, the best attested text refers to Jesus’ “associates/friends” or (more likely here) his “relatives” (Greek: οι παρ’ αυτου) setting out to take Jesus in hand for they (or others?) judged him to have lost his senses (Greek: ελεγον γαρ οτι εξεστη).[6] In Codex W (and also Codex Bezae), however, those who sought to take Jesus in hand in v. 21 are “those around him, the scribes and the others” (Greek: ακουσαντες [οτε ακουσαν Codex D] περι αυτου οι γραμματεις και οι λοιποι). Note that this shared variant involved “correcting” the preposition, changing παρα to περι, and disambiguating the text further by specifying “those around him” as “scribes and the others.” The effect (and the purpose) of this variant was clearly to avoid (“correct”) the idea that Jesus’ family could have regarded him as needing to be seized.
But in Codex W there is a further, and apparently distinctive, variant.[7] In place of the report that some were saying Jesus had lost his senses (εξεστη), Codex W has “they were saying ‘he has made them his adherents’.”[8] This actually involved another minor change of two words. Instead of εξεστη (“he is beside himself”), Codex W has εξηρτηνται αυτου, which means something like “they have become his adherents.” The effect (and the purpose) was to avoid the suggestion that Jesus was thought by some to have become somewhat unhinged. It seems to me likely that an early reader of the passage, troubled by what the more familiar wording says, and thinking that it must be some mistake, sought to put things right. Note that this involved replacing εξεστη (exestē) with a word that has a slight phonetic similarity, εξηρτηνται (exērtēntai).
It’s pretty obvious that both of the variants in Codex W (the one shared with Codex Bezae and the one unique to Codex W) were motivated by a pious desire to avoid what seemed to some readers an embarrassing passage that could reflect badly either on Jesus’ family or Jesus himself.[9] So it’s all the more interesting that neither one “succeeded” in getting adopted into the subsequent textual tradition. Instead, the potentially “embarrassing” text was preferred! In short, if the variants in Codex Bezae and Codex W are examples of an attempt at what Ehrman colourfully termed “orthodox corruption” of Mark, the attempts failed, and what is likely the originating, and more “difficult,” form of the text was preferred and preserved.
These interesting variants in Codex W illustrate both the creation of variants, likely by devout readers seeking to amplify (as in the variant in Mark 1:3) and/or to correct the text (as in the “Freer Logion” and the variants in Mark 3:21), and also that such attempts didn’t by any means always “succeed” in the subsequent manuscript tradition. As I’ve emphasized repeatedly, the forces that generated such variants and that made for either their subsequent adoption or rejection deserve more attention.
[1] A lightly revised version of my 1973 PhD thesis was published later: Larry W. Hurtado, Text-Critical Methodology and the Pre-Caesarean Text: Codex W in the Gospel of Mark (“Studies and Documents,” 43; Grand Rapids: Eerdmans, 1981).
[2] Translation from Bruce Metzger, A Textual Commentary on the Greek New Testament, Second Edition (Stuttgart: Deutsche Bibelgesellschaft, 1994), 104. The Greek text (adapted slightly from the Nestle-Aland 28th edition apparatus): κακεινοι απελογουντο λεγοντες οτι ο αιων ουτος της ανοµιας και της απιστιας υπο τον σαταναν εστιν ο µη εων τα υπο των πνευµατων ακαθαρτα την αληθειαν του θεου καταλαβεσθαι δυναµιν. δια τουτο αποκαλυψον σου την δικαιοσυην ηδη εκεινοι ελεγον τω χριστω και ο χριστος εκεινοις προσελεγεν οτι πεπληρωται ο ορος των ετων της εξουσιας τον σατανα αλλα εγγιζει αλλα δεινα. και υπερ ων εγω αµαρτησαντων παρεδοθην εις θανατον ινα υποστρεψωσιν εις την αληθειαν και µηκετι αµαρτησωσιν ινα την εν τω ουρανω πνευµατικην και αφθαρτον της δικαιοσυνης δοξαν κληπονοµησωσιν.
[3] Jerome, Dialogues Against the Pelagians 2.15.
[4] Metzger, Textual Commentary, 104.
[5] In my own earlier work, such as my 1981 study, I unthinkingly echoed the ascription of such intentional changes to “scribes”. I have come to see, however, that we should grasp and distinguish the processes of copying and reading/using texts more carefully. Basically, copyists copied; and “intentional” changes tended to be produced by users/readers who took the time to study and puzzle over the text. As Zachary Cole has shown in his study of how the copyist of Codex W handled the designation of numbers, he rather mechanically reproduced his exemplar: Zachary J. Cole, “Evaluating Scribal Freedom and Fidelity: Number-Writing Techniques in Codex Washingtonianus (W 032),” Bulletin of the American Society of Papyrologists 52 (2015): 225-38.
[6] Those mentioned here are likely the family members who re-appear slightly later in Mark 3:31-35, which explains Jesus’ response in vv. 33-35.
[7] I discussed this variant in Hurtado, Text-Critical Methodology, 77, and I also make reference to a couple of earlier discussions.
[8] Codex Bezae has εξεσταται αυτους (“he has maddened them”) instead of εξεστη.
[9] There are actually a number of other variant readings in the manuscript tradition that show that various readers found Mark 3:21 a troubling text.
Comments are closed.
Larry Hurtado's Blog 
Dear Larry,
Thank you for this post. Do you think there is any significance that the Freer Logion is at the end of the manuscript? Perhaps as some sort of canonical close?
No, Benjamin. The Freer Logion seems to have been intended to soften the critical tone of Jesus’ remarks in v. 14. There’s no “canonical” issue.
The interesting thing for me in the variants you cite is the need felt to comment on “unclean spirits”. Of course we are reminded of the accusation against Jesus, for they had said, “he has an unclean spirit” (Mark 3:30). Any thoughts to share on what Mark thinks an unclean spirit is? And if we say “demon”, what might Mark think a demon is?
John: “Unclean spirit(s)” is a Markan expression, referring to what are also termed “demons”. These are spirit-beings associated with Satan in ancient Jewish and Christian thought.
If your footnote [6] is correct that they are the same family members as in Mk 3:31 this would mean that in Mk 3:20 Jesus went into a house (or his home?), in Mk 3:21 his family/relatives went out of the house to restrain him, but then in Mk 3:31 he is back inside what we assume is the same house and his mother and brothers (and sisters?) are outside trying to get back in. As this is all part of one sequence of events in Mark then it makes no sense, and either it was not his family in Mk 3:21, or it was the family/friends of a different person. As the translations vary considerably here (those belonging to him, his own people, his friends, his relatives, his family, his friends and family, Jesus’ family) it seems more likely that the people in Mk 3:21 were not Jesus’ family, and that W and D are more likely to reflect the original meaning.
David: The verb in 3:21, “exelthon” seems to mean they “set out” or “went out” (from their home). They “heard” of Jesus’ activities described in 3:20, so they obviously weren’t then in the “house” where he was teaching. Mark 3:20-35 is an example of a well-known Markan device involving the “sandwich” of two stories. He begins one, inserts another, and then completes the first one. So, here, the family set out to seize him in 3:20, and reappear in the narrative at 3:31 to do so. In between, there is the story of opponents who accuse Jesus of being in league with Satan.
Two questions: (1) Is there any evidence that Codex W was used as the source for any subsequent manuscript copying? If so, how were these intentional insertions treated? If not, does this help explain why the insertions were not “successful”? (2) Is there any evidence for combination “copyist/readers” who were pious Christians with scribal ability, and were on the look-out for “corrections” that needed to be made to their source manuscripts?
I don’t know of evidence indicating whether Codex W was itself copied. But it is likely that these variants did not originate with Codex W. So a failure to copy this manuscript wouldn’t fully account for the failure of the variants to become “successful” thereafter.
The evidence of individuals both copying and amending a text would, I suspect, most readily be given in the copies that seem to have been prepared by individuals for their own usage. I’ve studied one, “P22”, which seems to be a personal copy of GJohn.
As you say Larry.
The variants at 3:21 certainly did not originate in Codex W, as they are substantially found in Old Latin witnesses of much earlier date. We can be confident that Augustine read Mark 3:21 as saying that those seeking to take Jesus in hand were the ‘scribes and others’ not his relatives. Similarly throughout Africa, Italy and Ireland at this date.
I suppose one possibility is that this reading originated in the Old Latin and so became ‘successful’ in all Latin traditions; maybe being selectively admitted later to some Greek traditions through Greek/Latin diglot manuscripts?
Tom: To my knowledge the variant in Codex W at Mark 3:21, εξηρτηνται αυτου, is unique to this manuscript. The other variants in 3:21, essentially shifting the reference from Jesus’ family to others (scribes, etc.) are found also in Latin, etc.
Not sure it is that straightforward Larry
The Old Latin form of the second variant at Mark 3:21 is “quoniam exsentiat eos” meaning something like ‘he has removed their reason’. The sense of the counterpart phrase in Bezae could well be a back translation into Greek “he is bewitching them”; while the sense in W, “they have become his followers” could not. Or So Burkitt argued. But then ‘exērtēntai’ in Greek could be argued as ‘phonetically similar’ to ‘exsentiat’ in Latin (following your perception above). I suggest that we cannot rule out the texts in W and Bezae both being (differently formed) back translations of “quoniam exsentiat eos”. If you did not know that the original Greek read ‘exestē’, could you get back to that reading from its Latin translation ‘exsentiat’?
Tom: The large number of variants at 3:21 suggests to me that various readers found the text difficult and tried to “improve” it. Some of these changes could as easily have been coincidental.
Indeed Larry
The presumed original text for both problematic phrases in Mark 3:21 are circumlocutions; “those of his” and “he is beside himself”. So several readers may well have been motivated to expand and clarify. But for translators, this issue is arguably more pressing, as not wanting to retain such circumlocutions unresolved ; ‘those of his friends’ or ‘those of his relatives’ or ‘those of his antagonists’? It would not be surprising if the creator of the Old Latin version plumped for the third expansion. Which then implies a counterpart clarification of the second expansion; should it be read as ‘he is mad’, or as ‘he is maddening’? Again it would appear possible that the Old Latin translators went for the latter clarification.
The pattern of variant successes, and failures, in the fourth century, and later, doesn’t really tell us anything about the pattern of success in the first three centuries, does it? I mean logically, it is perfectly possible that new variants from the fourth century onward tended to fail, whereas a large number in the first three centuries succeeded, is it not? In fact the circumstantial evidence would tend to suggest as much, when you have testimony about leaders such as Marcion editing the text, the community being younger and norms not established. The fact is that the manuscript traditional itself cannot settle this issue because the evidence doesn’t go back far enough, and drawing inferences from later patterns is dubious.
Donald: On the one hand, yes, as I’ve repeatedly indicated, there were apparently factors (as yet to be adequately identified) in the 4th/5th century and later that made for the “success” of certain major variants that prior to that did not enjoy such “success”. So, these factors aren’t relevant to the earlier period. But, on the other hand, the 3rd century manuscripts (and the few 2nd century ones too) clearly reflect other factors in the transmission of the text. And the chronological proximity means that it is (despite your stubborn refusal to recognize this) valid to make inferences about the immediately preceding decades and how the texts of the Gospels were handled. Nothing dubious about it, just (apparently) unsettling to you?
Donald, it is perfectly permissible to use later evidence in order to make inferences about an earlier period of copying. Suppose that for the next 200 years people make copies of my phone number. And suppose that someone in 200 years wants to know how accurately that was done. The trouble is that he only has evidence of the copying done between 2118 and 2218. Therefore he can’t say anything about the standard of copying done in the first 100 years.
Not true. If the copying was done very carelessly during the first 100 years, there won’t be anything left of my phone number. Instead, there will be random sequences with no agreement amongst them. Even if people start copying accurately after that, the damage will already have been done.
So if we find uniformity in texts at the earliest point at which we have enough evidence to make an assessment, that tells us something about the period for which we don’t have evidence.
A mass of textual variants might seem disconcerting but it is the price we pay for more copying and more widespread copying. And widespread copying is the best insurance we have against textual corruption. A similar situation obtains with discrepancies in the Gospels. If there was only one Gospel, there would be no contradictions. But it is better to have four accounts which sometimes seem at odds with each other than to have only one.
It might also seem desirable to have just the four canonical Gospels and no rival accounts of Jesus’ life. But the apocryphal works also serve a useful purpose. By comparing them with the four Gospels we can clearly see the difference between accounts written by authors who have an understanding of first-century Palestine and accounts whose authors have no such knowledge. Perhaps we should be grateful for the messiness of history.