Text Note: Luke 23:17

Image:  Screenshot of online Codex Sinaiticus of Luke 23:17 (with first word "anagken" in red box; notice its similarity to the first word in v. 18 "anekragon" and you see how an error of homoeoarchton could easily have occurred in other copies). This 4th century manuscript provides one of the earliest witnesses for the inclusion of Luke 23:17 in the text of Scripture.  This is unusual, since Sinaiticus does not typically support the traditional text. Also interesting is the fact that the verse in Sinaiticus ends with the conjunction "hina" rather than the masculine accusative adjective "hena," a variant that is not noted either by Metzger in his Textual Commentary or in the critical apparatus of the NA 28.

Here is another text issue I ran across last week when preparing to preach on Luke 23:13-25.

I.  The Issue:

KJV Luke 23:17 (For of necessity he must release one unto them at the feast.)

Should Luke 23:17 be included as part of the text of Scripture?  It is included in the traditional text and omitted in the modern critical text.

II.  External Evidence:

The traditional text is supported by the following:   Sinaiticus, W, Gamma, Delta, family 1, family 13, 565, 700, 2542, and in the vast majority of extant manuscripts.

The verse also appears with minor variations in:  N, Theta, Psi, 579, and the margin of 892.

Finally, the verse appears in D but after v. 19.

The modern critical text is supported by the following:  p75, A, Vaticanus, K, L, T, 070, 892 (txt), 1241.

Comments:  Here is an example where the witnesses divide on unusual lines.  Sinaiticus, which normally supports the modern critical text, includes the verse.  Alexandrinus, which normally supports the Byzantine text, omits it.

III.  Internal Evidence:

Metzger, in his Textual Commentary, begins:  “The secondary character of the verse is disclosed not only by its omission from such early witnesses as p75 A B ….but also by its insertion in slightly different forms , either here or after ver. 19 (where codex Bezae agrees in wording with the reading of Theta Psi)” (p. 179).  His conclusion:  “The verse is a gloss, apparently based on Mt. 27:15 and Mk 15:6” (p. 180).  So, for Metzger, the verse is a harmonization to its Synoptic parallels.

Metzger does, however, acknowledge the possibility that the omission of the verse came through the scribal error of homoeoarcton (an error due to a similar beginning).  The first word in v. 17 is anagken and of v. 18 is anekragon.  Thus, one can easily see how the verse might have been accidentally omitted or even omitted and replaced after a later verse (as in D).  Metzger dismisses this possibility, however, as “unable to explain its widespread omission and its presence at two different place” (pp. 179-180).  In response, however, one might observe that if it was possible for scribal error to occur in one family of witnesses it is also possible that it occurred in others as well, and thus appear widespread.  In addition, Metzger seems to overplay the variations.  It occurs after v. 19 in only one witness (D).  According to NA 28 the variations in Theta and Psi are minor (the exact wording of the variant is not cited either by Metzger or NA 28).  Also, NA 28 does not record that the variant in D is the same as the variant in Theta and Psi, as Metzger suggests.

It might also be observed that if there was a scribal assimilation of Luke to Matthew and Mark, it is not an exact verbal parallel to either.  Compare:

KJV Matthew 27:15 Now at that feast the governor was wont to release unto the people a prisoner, whom they would.

Kata de heorten eithei ho hegemon apoluein hena to ochlo desmion hon ethelon

KJV Mark 15:6 Now at that feast he released unto them one prisoner, whomsoever they desired.

Kata de heorten apeluen autois hena desmion hon paretounto

KJV Luke 23:17 (For of necessity he must release one unto them at the feast.)

Anagken de eichen apoluein autois kata heorten hena.

If Luke 23:17 was a harmonization to Matthew and Mark, why does it not begin as the others do with Kata de heorten?  Why does it not mention, as Matthew and Mark do, that the released prisoner was given up at the will of the people?  The Lukan passage is also different in that it introduces the concept of the necessity of Pilate releasing a prisoner.  Rather than being a textual harmonization, it appears that the mention of Pilate releasing a prisoner at the feast was a deeply imbedded part of the early Christian tradition which Luke conveys in a typically distinctive manner.

III.  Conclusion:

The external support for Luke 23:17 is early and weighty.  It is even included in Codex Sinaiticus, an authority normally appealed to for support of the modern critical text.  It is entirely plausible that the verse was accidentally omitted in some witnesses due to its similar beginning with v. 18.

It is interesting to note how even a very able exegete like Leon Morris seems to have adapted the assessment of the modern text critics in his commentary on this verse:

In some mss. the words which appear in AV as verse 17 are found, but they are inadequately attested and appear to be an importation from Mark 16:5.  Such an insertion would be favoroured by the fact that verse 18 does not follow very smoothly from verse 16 and a scribe might well to improve the connection (Luke, Tyndale Commentaries, p. 323).

As shown above, the external evidence for v. 17 is hardly “inadequately attested.”  The speculation that a scribe attempted to “improve the connection” between vv. 16 and 18 is speculative and subjective. 

At what price is this verse removed?  To remove the verse would deny that Luke was aware of an important early Christian tradition relating to the trial of Jesus before Pilate which he conveys in a manner that is both compatible with the Synoptics but also distinct.  We lose the insight not only that Pilate chose to offer the release of a prisoner at the feast but that he was compelled to do so.  Luke 23:17 should, therefore, be retained.   

JTR