The question here is not about trying to isolate which particular manuscript (extant document) has the most variations from some standard determination of the text. Rather, it is about which particular text (i.e. passage from Scripture) has the most documented variations among the over 5,000 extant Greek manuscripts.

Personal examination is fine (and if used, please list the variant readings and their source), but I'm particularly interested if anyone knows of published research backing up what passage has the most variants—since personal research would also have to technically analyze a number of other variant passages to determine which has the most.


For purposes here:

  1. A passage is roughly limited to a clause or small set of related contiguous clauses (i.e. they belong to the same paragraph so to speak, and reside next to one another—see "Clarification" below).
  2. A variant can be any standard variant (spelling, word order, etc.), but consider the omission of a contiguous group of words as a single variant (not a variant for each word missing).
  3. Only Greek manuscript evidence (not a variant that only exists within a Latin, Coptic, or other translation), and limited to attestation prior to the use of printing for Greek texts (so target 1500 A.D.; I realize research found may have lower limits in dating for their study, but this would be the maximum date limit).

For a baseline to consider variation by, use the NA28 text.


While I realize some will disagree with me, I am of the opinion that the original reading of every text is preserved within the extant manuscript tradition. My personal position is a nuanced majority text position for isolating the correct reading, but for purposes here, that is irrelevant (other than to help you know where I am coming from).

So my intent from this question is to isolate which "passage" (which is somewhat loosely defined) has historically shown the most variations from which one might possibly choose from if confronted with trying to determine the original reading. This would certainly consider variations found across manuscripts, but some manuscripts may include a few variations in themselves (in side margin notes, etc.), which may match to another manuscript (in which case the count would still only be one extra variant).


Trying to define "passage" here is somewhat challenging. That is why I attempted to limit it generally to a clause, or small contiguous set of clauses. The idea is some small "thought unit" for which one might be trying to determine the "meaning" of the statement (generally this is going to be a sentence, which may contain a single or multiple clauses).

I realize some textual problems, such as the Pericope Adulterae, have quite a range of both "locations" it is found (if it is found in a text), as well as generally recognized longer and shorter versions. The common variants are noted in some modern translations, and there is a contiguous relation to its set of verses, such that each likely has some variant. This pericope might qualify as a candidate (I really don't know how many actual variants exist in its readings),1 however that grouping is still larger than what I was envisioning—more sentence level than paragraph level, and the pericope is multiple sentences. So perhaps a particular sentence within the pericope is the most variant in attestation (though I am not opposed to considering the whole as a candidate for a "thought unit").

An answer here will certainly be limited to the scholarship available (since I do not really expect someone to research out the vast number of passages themselves, but perhaps someone has, or been involved in a project that did). I realize there has also been a serious lack of work done with actually examining all the Byzantine texts, so some undiscovered variants may exist in those (though whether that would push one passage over another as having the most variants seems unlikely).

So there are some practical limits to what may be available, but the question here is to isolate what research is presently available that shows a particular "passage" to have the widest attestation of variation in the manuscripts (X number of different readings across Y number of manuscripts).


1 If one takes just the summary of manuscript evidence noted in Wikipedia (accessed 10-16-2014), then there are six variants regarding the presence of the passage in whole or in part:

  1. Omitted entirely (#1)
  2. Contains John 7:53-8:2 (#2)
  3. Contains John 8:2-11 (#3 part)
  4. Contains John 8:3-11 (#3 part, #6 [questioning it does not make it a variant])
  5. Includes all John 7:53-8:11 (#4, #5, #8 [though #8 might be considered two variants, the original reading and the addition, though the original, if omitted, simply matches my 1. above])
  6. Relocation to a different place (#7)—In this case I would count this as an omission (1.), and so does not add a "variation" to the John passage here, but would count as an inclusion in whatever passage it was "moved" (if one were examining textual variations for say Luke ch.24 or 25 (since one place it is moved to is after 24:53/before 25:1).

Yet within those broad concepts of inclusion, there are further variations (of which the wiki only notes "some"), so clearly more than six variations exist for the entire passage when one starts examining sentence details of each of the broader categories of "longer" and "shorter" versions.

  • Do you mean: which single textual problem has shows the most/widest variations in the manuscript tradition? or which "passage" (as in your #1) has the most problems, no matter how wide the attestation? I'm guessing the latter, but not 100% sure!
    – Dɑvïd
    Oct 15 '14 at 21:00
  • 1
    After a few minutes reflecting ... I'm not sure how answerable this question is. Can you focus it? give it some context? It feels like something is at stake for you, but even so has something of "the beauty context" about it! :/
    – Dɑvïd
    Oct 15 '14 at 21:07
  • 2
    Please don't take this personally - I'm a big fan of your contributions here on BH, but I see this as a 'shopping' question and not well suited to the Q&A model here. Oct 15 '14 at 21:23
  • 1
    @JackDouglas: I see a "shopping" question as desiring a list, and not having a single answer (similar to what you mention here). My question should have but a single correct answer, technically a verse reference with some documentation to back it up (so I also would not classify it as "too broad," like one close vote has). I'm seeking an answer based on statistics, somewhat like this question.
    – ScottS
    Oct 16 '14 at 0:48
  • 1
    not sure that the amount of information you're requesting would even be comprehensible. there'd be a lot of noise to filter which is why the apparatuses to na28,ubs5, sblgnt are so helpful - they get to the ones that seem to have the highest s/n ratio. if you're really interested in getting everything, perhaps reach out to dr. dan wallace at cnstm
    – swasheck
    Apr 29 '15 at 3:25

Browse other questions tagged or ask your own question.