Saturday, November 11, 2017

Where does the Parable of the Sower begin? (Mark 4:3)

In the Greek New Testament, Produced at Tyndale House, Cambridge (THGNT) we made the decision to begin the Parable of the Sower in Mark’s gospel with the second, not the first, word of Jesus’s speech. In Mark 4:3 we have ἀκούετε ‘listen’ and then a new paragraph beginning Ἰδοὺ ἐξῆλθεν ὁ σπείρων σπεῖραι· ‘Behold the sower went out to sow’.

I don’t know that we could reasonably have done anything else.

To make the point, I’ll just paste a series of pictures of manuscripts below with brief comments.

Vaticanus, fourth century, marks the new paragraph with the paragraphos above ιδου (credit Vatican Library).

Sinaiticus, fourth century, leaves space to the end of the line after ακουετε and has ekthesis before ιδου (credit British Library).

Alexandrinus, fifth century, ends the column with ακουετε and begins a new page with ιδου beginning with a littera notabilior (credit British Library).

Ephraemi Rescriptus, fifth century, does not use ekthesis with ιδου. Though ιδου does begin a line this is probably just a result of where it naturally falls within the paragraph. This manuscript therefore goes against the trend of the others (credit Bibliothèque Nationale de Paris)

Codex Bezae, fifth century, ends a line with ακουετε when there’s plenty of space for more. It then has ekthesis with ιδου (credit Cambridge University Library).

1424, 9th-10th century, is included here as illustrative of a later manuscript. There’s now a gap before ακουετε and another between ακουετε and ιδου. It’s a sort of intermediate form evolving from the earlier pattern of paragraphing to the more recent system of having the main break before ακουετε (credit

Of course the beauty of the old system, restored now in the THGNT, is that it separates the command to use one’s hearing from the command to use one’s imagination (or mind’s eye).

‘Listen up’
‘Now imagine you can see ...’


  1. There's a gap there in GA 2474, too.

  2. Do we suppose then that Mark originally wrote it that way (with some kind of indicating marks or spacing), or that these scribes merely interpreted it that way?

  3. Since these manuscripts are from a long time after composition, it is difficult to be confident of any conclusion. Most ancient texts of length show some unit division and most NT manuscripts do. P90 (John) is often dated to the second century and has paragraph marks. P46 (Paul) is early, but doesn't.

    Here we note that 4 of our 5 earliest manuscript witnesses converge on dividing ακουετε and ιδου, despite substantial textual differences between the same witnesses elsewhere in Mark (e.g. presence/absence of 16:9-20). The view that the division in 4:3 is a shared inheritance rather than a common innovation seems preferable to me. However, we're still at a very early stage in studying paragraphs.

    As far as I know, the THGNT is the first printed edition systematically to use manuscripts as a basis for paragraphs, but our edition is just a first attempt which others will be able to improve on. One of the main benefits of the edition will be to raise awareness of the issue, so that more people will work on the subject and perhaps one day we'll have a grand database of all the manuscript unit divisions in existence and thus be better placed to express an opinion on their origin.

    However, this paragraph division, even if not Marcan, is at least ancient and gives us the best insight we can have into how an ancient reader/ hearer/ performer might have divided the text.

    In the case of the paragraph division after John 1:5, I think that the convergence of our early witnesses is so strong that the division can be judged more probably belonging to the archetype of the tradition than having entered post-archetype.

  4. Not so sure about Vaticanus here. Wouldn't the paragraphos be in the same place even if it was supposed to mark after AUTOU? Also paragraphoi are secondary in B. Gaps are original indications of paragraphing in B.

  5. I was thinking that the paragraphoi were later, but hopefully not too much later. I guess that at least the first hand did not leave a gap after αυτου. 2:14 and 4:18 show paragraphoi without gaps within the preceding line and the syntax indicates that the break should be interpreted as at the line end. 2:25 is a possible counter example if you think there's a gap after εξεστιν. Obviously B is less certain here than א A D.

  6. Thanks, Peter. I and my class at APTS Baguio will be looking at this parable this week.

  7. Pete, I was reading on the parables last night and came across this pertinent argument from Klyne Snodgrass’s (excellent) big book on parables and it reminded me of this post. Snodgrass says, “Mark ties the parable to the quotation of Isa 6:9-10 more strongly than the other Evangelists, even though his connection is lost in translations. Mark starts the parable with ‘hear, see’ (akouete, idou), both of which occur in v. 12 in the quotation of Isaiah” (160-161).

    If this is so, then the scribes you cite here all seemed to have missed it. In any case, a good example of how formatting affects our reading.

  8. I'm not sure that my view necessarily is in conflict with his. Both 'listen' and 'behold' are highlighted by their being at the extremities of paragraphs and this fits with the background in Isaiah. I don't think that a paragraph is a 'connection severer', but something that highlights the new onset. It may signal change, but that's not the same as discontinuity. I don't think that the modern definition of a paragraph as a set of grouped/related ideas is very helpful in dealing with ancient text since typically ideas continue across paragraph boundaries.

    1. Hi Peter, interested observer here. What did the paragraphs communicate, if not a set of grouped/related ideas? Does it signal a new direction of thought, at least?
      I'm pondering how paragraph breaks in the THGNT might shape preaching pericopes for pastors. For example, starting a new paragraph at Eph 5:22 rather 5:21, or Romans 13:7 marking a new paragraph (rather than the conclusion of the text beginning at 13:1).