Tag Archives: textual criticism

Texts of the Greek New Testament

In a previous post I discussed the greek word translated “daily” in the Lord’s prayer. In the comments I got asked about the categories that we use to talk about the greek text. I promised to post more on that, so here goes. I don’t have long this evening to edit this thoroughly, so I may do that more over the next few days.

So, we have a lot of different Greek texts of the New Testament. How are these classified, grouped and understood? Well aside from dating, which is (obviously) a continuum, there are two important ways to group them:

The First Division — Type of Text

The most obvious division is between three groups of text that have different forms. They are the Papyri, the Uncial Texts and the Minuscule Texts. The papyri were written on papyrus (obviously), the uncial and minuscule texts were typically written on vellum, parchment or paper, and bound into codices (in book form), although in many cases we have only individual leaves. The difference is in the style of writing. The uncial manuscripts are written in capitals, the minuscules in lower case.

1 John 5:7-8 in the Codex Vaticanus, a Uncial Codex of the Alexandrian text-family.

1 John 5:7-8 in the Codex Vaticanus, an Uncial codex of the Alexandrian text-type

This division is roughly chronological – papyrus manuscripts (c. C2 – C8) tend to be earlier than uncials (c. C3-C10), which are in turn earlier than minuscules (c. C9-C16). This isn’t rigid, however. There is a lot of overlap between papyri and uncials in particular.

When you see a NT manuscript discussed, it will be referred to by a code. Papyri have a prefix written as a Blackletter P (e.g. 18, a papyrus of Rev 1 found in Oxyrhynchus, Egypt and now held at the British Library). Minuscules are written as numbers (e.g. 44, a complete  C12 text of the gospels also in the British Library). Uncials have a broader range of symbols, the most important are represented by single letters, in a range of different scripts (e.g. א, Codex Sinaiticus, or B, Codex Vaticanus), often with superscripts (so Ha is distinct from He – the stories behind these letters are complex and for another day). These letters are called ‘Sigla’, and there are a little less than 50 of them. Other uncials are represented by their number, which is always given with a leading zero (to differentiate from Minuscules), so 068 is a C5 uncial of John 16, once more at the British Library [I’m picking the British-housed texts deliberately – there are plenty of these texts not housed in the British Library!]. All Uncials have a numerical value, but those that also have a letter are rarely referred to by their number (א is 01, for example).

The Second Division — Text Family

Scholars also divide the texts according to the text tradition they come from. A text tradition is a group of related texts. Because corrections, changes, editing and additions all tend to be copied from one text to another, the premise is that, texts that all display the same features, are likely to trace back to common ancestry. This isn’t always a perfect premise, but it is normally a good one.

So we divide the texts into three again. The Alexandrian text tradition, the Byzantine text tradition and the Western text tradition.

These are obviously named for places, but don’t let the names fool you. Each of these traditions may be clustered in a particular area, but they are all relatively geographically spread. The names are names of convenience rather than being hard-and-fast geographical distinctions.

The Alexandrian text is evidenced in the earliest manuscripts we have. Most of the Papyri are Alexandrian (those that aren’t are either later, or else are unclassifiable because they contain too small a fragment, or contain sections without significant variation among the text-types). The Alexandrian texts are clearly the earliest we have. Whether they are better evidence of the original text, is a more theological debate. As I’ve said before I have little patience for the obsession with ‘original’ texts, whatever that means. Recently critical editions of the NT, and most new translations, have methodologically preferred the Alexandrian texts because of their demonstrable age, and the fact that they lack what look like interpolations and decorations of later texts.

The Byzantine text tradition is also sometimes called the Constantinople tradition (Constantinople was the capital of the Byzantine empire – the Greek speaking part of the Roman empire that survived until the C15 – 1000 years after you were probably told the Roman empire fell in your western-biased high school!). But it is also more problematically called the Majority text. This name is somewhat politically charged, it is the remnant of a bad scholarly program that tried to prioritize this text over the Alexandrian tradition because of the sheer number of manuscripts we have. It doesn’t take a scholar to notice that the quest for authenticity is not democratic – there are more extant copies of the LOLcats bible than the Byzantine text tradition – that doesn’t make the LOLcats bible more authentic. So I’d recommend staying clear of “Majority Text” (and its paranym “Received Text”). The Byzantine texts are much later, and tend to be more verbose than the Alexandrian. It is in the Byzantine family that we have the bulk of texts with the post-resurrection narratives in Mark, for example, or the woman caught in adultery, or the end of the Lord’s Prayer. We also have a lot of Byzantine-style texts in the form of Lectionaries – compilations of bible verses used for daily prayer or ritual. This is not surprising, the Byzantine empire was a Greek-speaking Christian state. Byzantine texts often have more fluent, liturgy-friendly, wording. You can think of the Byzantine texts as being equivalent to the various versions of the Latin Vulgate used in Catholic Europe.

P37, from the University of Michigan library. An example of a Papyrus containing the Western text of Acts.

The Western text tradition is more eclectic. Originally named for a group of texts associated with the Western Mediterranean, this group of texts has now been found in a much wider spread. From France to Switzerland to Syria. It is characterized by a freer style, bordering on paraphrase and retelling at points. There are more additions (some the same, some different to the Byzantine texts), although in Luke there is a section that is shorter in the Western text (proving that all rules have exceptions). Unlike the Byzantine texts, which were highly influential in the production of early protestant translations, the Western texts haven’t had much of an impact on our English bibles, and so their alterations are less well known. There are relatively few Western texts, in contrast to the other categories. Though a couple of Papyri in this text-type are very early (C2-3 for P37, for example), most scholars consider their obvious paraphrasing and changes in wording to be evidence that they witness a later stage of the text than the Alexandrian text-type. With a much smaller number of texts, we get more patchy coverage. We have no copies of this tradition for the non-Paul epistles and Revelation, for example (and it could be that these books were not canonical for the communities who originated the Western text-style). Scholars of the Western text tradition often use non-Greek sources as primary evidence, particularly Syriac versions of the text, which appear to have a lot of similarities.

There are plenty of other texts that aren’t obviously thoroughly in one category or the other, or else are so fragmentary as to have no clear indication of their type. Various other text-types have been proposed, although none are unequivocally accepted. The Ceasarean text tradition, for example, is useful in categorizing some patterns of change in the Gospels, but on close inspection tends to fragment into sub-families and more nuanced arguments.

And that, phew, is a whirlwind tour of NT text classification. I hope it wasn’t too tedious. I find this geektastic, but I’m aware that there were plenty of my fellow undergrads asleep in these lectures!



Filed under Uncategorized

Critical Editions

In my previous post, I gave an example of using a critical edition of the Old Testament, the Biblia Hebraica Stuttgartensia. In the comments I mentioned that it would be worth talking a bit about critical bibles generally, in particular the difference between eclectic and diplomatic texts.

A critical edition of the bible isn’t so much the bible text itself, but the footnotes. A critical edition is designed to tell you what variations there are in the text, and how those variations are commonly approached. It should provide a way of jumping off into other resources to find the in-depth debate on a particular verse and, using your personal criteria for how you evaluate the various source texts, allow you to come to a conclusion about how to treat the passage. This is all about the footnotes (called the ‘apparatus’). The text itself should only be there to provide the bits where there is no reasonable disagreement, and it obviously provides one of the alternatives when there are disagreements. Normally critical editions don’t log every spelling mistake or obvious error in every text, but they may point out such problems in the most important text, if the editor thought it worth the note.

The difference comes in how you decide what goes in the main block of text and what goes in the footnote. In a diplomatic text, the main text follows almost exactly some specific manuscript. So BHS is a diplomatic edition, because its main text is the Leningrad Codex. The Hebrew University Bible is another diplomatic critical text (not yet published in full), but it follows the Aleppo Codex. In a diplomatic text, the main text isn’t necessarily the ‘best’ text, you have to read the footnotes to get a sense of what form of words you might want to use. For translation (particularly the kind of translation you do at graduate school) diplomatic texts can be harder work.

In an eclectic text, the editors synthesize the main text from their judgements on the variations. So the total effect is a text that does not reflect any extant manuscripts, but picks and chooses from moment to moment. This puts the ‘best’ form of words in one place (as long as you agree with the editors on what constitutes ‘best’). The main NT critical text: Nestle Aland v.27 is such an eclectic text. There is a project underway to produce an eclectic Hebrew Bible, published by the OUP. For detailed scholarship, eclectic texts can be fiddly, because you have to continually unwind the decisions made by the editor.

This is not to say that eclectic texts bear no relationship to existing manuscripts. Typically the editorial team will favour a particular text tradition, and so the eclectic text will closely follow one particular set of manuscripts. NA27, for example, follows the Alexandria text tradition quite closely (for which it is often criticised by conservative evangelicals who prefer the Constantinople tradition used in their beloved KJV). And on the other hand, editors feel free to correct obvious mistakes even in diplomatic texts, removing obvious scribal errors. So you could say that the diplomatic-eclectic distinction is actually a continuum. So far, however, critical editions do tend to cluster at either end.

But notice, in both cases the text itself is ideally irrelevant. The text, combined with the footnotes, should give you the whole picture. If you just isolate the text itself, you don’t have a critical edition any more, you just have a particular original language version, one that more or less follows some set of manuscripts.

I say ideally, because, in practice the base text does influence the footnotes to some extent. It is difficult as an editor not to be drawn into the mindspace where the base text is the foundation and the footnotes are merely variations. Treating the base text itself as just another variation is a challenge. But in my experience (which by far is with the eclectic NT editions) the editors of critical editions do a great, and unsung job*. Without their hard and grinding work, engaging with the bible would be orders of magnitude more difficult.

Of course, the end reader of an English bible doesn’t see any of this. Most bibles have footnotes that indicate the odd major variation, but you miss out on the decades, or centuries of work compiling variations that went into the critical edition that your translator worked from. It couldn’t be otherwise, of course, but behind the seeming homogeneity of your english translation it is worth keeping in mind the thousands of texts and millions of variations that the critical editors have wrestled with.

I have a soft-spot for folks who do this kind of work. The professor who taught me Hebrew as an undergrad is on the BHQ editorial team. I got a glimpse of the painstaking effort that lies behind a few words of abbreviated latin.


Filed under Uncategorized

Reading the Resources — Ecclesiastes 2:25

Over on Sabio’s blog, he’s been blogging through Hector Avalos’s book “The End of Biblical Studies”. Now I confess that, like a lot of people who actually quite like biblical studies, I was prejudiced by the title. Also, I haven’t read it, so I’ve been happily spouting off in ignorance over there.

I don’t want to bring that discussion here. But I did think it was interesting talking through a bit about the resources one can bring to bear on these kind of linguistic issues. Because I think those readers of both blogs who’s resources are mainly web-based or study bibles might be interested in having a peek at the two main critical resources: the critical edition of the Old Testament, and the granddaddy of Hebrew lexicons.

Ecc 2:25 in the BHS - the standard critical edition of the Hebrew Bible.

First the text. Shown here is Ecc 2:25 in the Biblia Hebraica Stuttgartensia. The latest complete critical edition of the Hebrew bible*. The text is split into three. The body of the text itself (our verse highlighted – remember Hebrew reads right to left); the masorah parva in the outside margin; and the critical notes at the bottom of the page. The masorah parva is half of a traditional system of commentary and analysis – the larger half (the masorah magna) isn’t in modern critical editions. The parva contains checksums to help scribes make sure their copying is correct, links to other uses of a phrase, or indicates where what is read should be different to what is written (another tradition in Hebrew writings). The critical footnotes are written in highly abbreviated Latin with a good number of special abbreviations and symbols.

The phrase that Hector Avalos focussed his rebuttal of my off-the-cuff translation in Sabio’s blog is the two words at the end of the verse. The text reads CHUTZ MIMENI (tr. “except I”) It has a footnote. The footnote reads “sic etiam TV, sed pc MSS GS, MIMENU, l?” (where T,V,G and S are fancy symbols). Expanded this reads “sic etiam TV, sed pauci manuscripta GS MIMENU, legendum?”, or in english:

“thus, and so [i.e. as it was printed in the main text] are the Targumim [Aramaic translations and commentaries] and the Vulgate [the Latin version], but a few manuscripts, including the Septuagint [the Greek OT] and the Syriac have MIMENU, and that should be what is read? [the question mark indicates that this view is not settled]”

An excerpt from Brown Driver Briggs on Ecc 2:25

Now, onto the lexicon. I’ve shown here the relevant page from the Brown Driver Briggs translation and expansion of Wilheim Gesenius’s German lexicon (always just BDB). There are others, but this is the undisputed grandfather of Hebrew language resources. The relevant bit, to us, starts at “f. (late)” in the middle of the image. Up to this point, the lexicon has been discussing the various usage of the word CHUTZ. It has meanings around “outside”, but so far we’ve had more or less literal meanings. Here the indication changes. It says, there is a late use, in Ec 2:25, where it has the sense “outside of, i.e. except me”: a more metaphoric meaning. But, it goes onto say, it should be read, (along with G, the Septuagint, S, the Syriac, and the work of Delitzsch [De], Nowack [Now] and others [al]) “apart from him (i.e. God)”. It then goes on to say that there’s a similar Aramaic idiom which is used in (at least) two places in the Targums (the commentary on Ex 20:3, and Is 36:10) and a similar Syriac phrase with the same meaning.

So, it seems like everybody agrees that the text says “except me (i.e. Qohelet)”, but both resources agree that this was probably best read “except God”. Interestingly BDB changes the comparative when it changes the pronoun here, from “except, or outside” to “without”: a subtle change of gloss, but one which changes the feel of the passage.

Anyway, as I’ve promised to read Hector’s book, I can’t comment on whether I agree or not with his reading of this passage yet (I’m a bit confused what he’s saying, after hearing it second hand through Sabio). But I thought it might be interesting to have a peek at a couple of the most common resources used for looking at these questions.

* There is a more recent version of the Biblia Hebraica, called Quinta (its the 5th edition, after all). It isn’t completed yet, and isn’t expected until 2015. But Ecclesiastes has been published in this edition, in an initial fascile containing the “5 scrolls”. I don’t have a copy of that, however, so if anyone does and wants to scan in the relevant page, it would be informative to see what has changed.

Below Sabio asks about the Septuagint. Here is the greek text tradition, evidenced in the Codex Sinaiticus:

It has ‘AUTOU’ at the end of the first paragraph (line 6, which is the end of 2:25, in capitals this looks like AYTOY – see my previous post on original manuscripts for more details on reading these texts). AUTOS is ‘him’. As per the BHS and BDB, both of which said that ‘G’ (i.e. the Greek) had the ‘except him’ reading.


Filed under Uncategorized

The story of how a bible verse is read — 1 John 5:7-8

Annotated exceprt from the Codex Sinaiticus

1 John 5:7-8 highlighted in the text of the Codex Sinaiticus, with various textual features signified.

Following on from yesterday’s post on the Johannine Comma, I was enjoying looking again through the online copy of the Codex Sinaiticus. It is both more difficult, and more rewarding than reading the reprinted greek critical text. I thought it would be interesting, for those who don’t know greek, ancient texts, or have never used their greek on primary documents, to point out a few interesting features. There’s loads I could say, and far far more I’m ignorant of, but I’ll limit myself to the text I posted yesterday, the text surrounding the Comma in 1 John 5:7-8.

As a brief reminder, an english translation would be:

For there are three that bear witness: the Spirit, and the water, and the blood, and these three agree.

– 1 John 5:7-8

I’ve indicated five things in the document.

  1. ΟΙ (“oi” – a form of “the” or “these”) here is unusual and isn’t in other versions of the text. It is probably a mistake, because later in the highlighted section we get ΟΙ ΤΡΕΙΣ again (“oi treis” – these three). When copying, it is very easy to have your eye drawn to the wrong occurrence of a word and write what comes before or after. So the extra ΟΙ before this first use of ΤΡΕΙΣ is considered a copying error.
  2. A pair of characters (ΣΕ) gets added into a small gap. Maybe there was not enough room? The two letters mark the end of the previous word and the start of the next, respectively.
  3. We’re missing an Ν from this word (ΜΑΡΤΥΡΟΥΝΤΕΣ “marturountes”- bear witness to). The overbar indicates this abbreviation, it can also be seen in the same word in the Codex Vaticanus, below.
  4. This is the word for ‘spirit’ (ΠΝΕΥΜΑ ,”pneuma”, spirit is also the word for breath, and this is where we get the word pneumonia). It was conventional from very early on to write certain words as acronyms. We call these ‘nomina sacra’* – sacred names, because most commonly they are names (although words like ‘cross’ are also abbreviated in this way). Here we get it written ΠΝΑ. The bar over the top of the word is often used to indicate abbreviation.
1 John 5:7-8 in the Codex Vaticanus

The same text 1 John 5:7-8 highlighted in the Codex Vaticanus, for comparison. Vaticanus is a very similar text to Sinaiticus, from the same text family. But still there are thousands of differrences. See if you can spot the differences in just this passage. Apologies for the quality of this scan, I'm using the Vatican's own published scans, which are pretty low quality.

And finally you notice that the words are all written into one another. There are no word spacings and words are free to continue over the end of the line without this being indicated. The red squares show word boundaries.

[Quick note, in my modern greek font, the capital sigma appears as Σ, but you’ll notice in the text it looks more like a C – letter forms change over time and between regions]

I know that’s a very short passage, but I hope it gives the sense of how dense these texts are, and how much effort it takes to engage with them as a textual critic. I don’t have the patience to do this kind of work, although I do love the manuscripts themselves. I’m very much glad I can just buy a critical copy of the NT and have the hard work done for me!

* A challenge for anyone who knows a bit of greek: You can see two more ΠΝΑ abbreviations on the same page – one on the top line (overbar is half-clipped off the image) and one at the start of the third line. There is another nomina sacra at the bottom of the page: ΘΥ for “God”. But 10 points for any readers who can spot the other two in the image.


Filed under Uncategorized

The Johannine Comma – The story of how a bible verse was written — 1 John 5:7-8

The Johannine Comma in the Codex Sinaiticus

A version of the greek text (The Codex Sinaiticus - part of the Alexandrine text tradition) without the comma in place. The highlighted area shows 1 John 5:7-8. The small text above the following line shows a correction to a scribal error. The full photographic facsimile of this text is available online.

One of the hard things about studying the bible is working out what version of the bible to study. There are thousands of copies, all written by hand, and all different. Reconstructing what might have been the original text is the job of ‘textual criticism’, and it isn’t an easy job.

One example (which is much easier to figure out than many) is the so-called Johannine Comma. In the first epistle of John, chapter 5, there is a suspect phrase spanning the end of verse 7 and the start of verse 8. Those two verses say:

For there are three that bear record in heaven, the Father, the Word, and the Holy Ghost: and these three are one. And there are three that bear witness in earth, the Spirit, and the water, and the blood: and these three agree in one.

– 1 John 5:7-8

This section seems to read pretty well, and the two parallel sections fit together. In wider context, however, the passage does jar a little. And it becomes really suspicious when you realise that it seems to be slam-dunk evidence of the doctrine of the trinity, but that doctrine wasn’t formulated for more than a century after 1 John was composed. The trail is even more intriguing when you look at the earliest manuscripts we have, and the textual traditions from those early manuscripts.

[An aside – a textual tradition is what we call the set of copies that are made in one community or region from an initial set of texts. So we have an ‘Alexandrine’ textual tradition which is the set of texts copied in Alexandria, derived ultimately from some original set of texts (original for Alexandria, not necessarily first copies of the original works). Because of the copying within a textual tradition, you get consistent errors showing up through the family of copies, and conversely by looking at what is consistent and what is different within a tradition, you can reconstruct what the original texts in that tradition might have looked like.]

We find that the section about the trinity (the section in italics, above) doesn’t exist in early manuscripts. In fact it only appears in one textual tradition. That textual tradition is a tradition built on an early Latin translation of the Greek text.

What we think happened is this:

  1. The original text looks like it has a trinity in it (spirit, water and blood). But after the doctrine of the trinity was decided upon, some scholar felt that this alternative trinity needed explanation. So they wrote a marginal note in their copy of the latin version of the text, showing the parallel between John’s trinity and the trinity of God: one heavenly, one earthly.
  2. Some later point this text was copied. A scribe seeing this marginal note mistook it for an omitted part of the text (it wasn’t uncommon to add accidentally omitted parts of the text as marginal notes, much as we’d insert a bit more content into a handwritten text by writing it in the margin and adding a line, arrow or caret). He merged it into the text as best he could.
  3. At a later date again, a translation into Greek was made of this Latin text, and so the appropriate section was then turned into greek along with its surrounds.
  4. The textual tradition that grew out of these texts is the one that ultimately was used in the writing of the King James Version of the bible. So it ended up in the KJV.

Most recent bibles will omit the Johannine comma, with the text placed in a footnote. The evidence is so clear  that you might think it would be worth removing entirely. It is testimony to the great weight of the KJV of the text in English that most bible editors won’t do this. They think (rightly, I say) that people will compare their translation against the KJV and wonder why they are missing chunks out. So the footnote has to be there to explain.

This chunk of text is just representative of thousands upon thousands of variations in the early text. Almost every verse in the New Testament has several alternate renderings. Some are obviously incorrect (such as spelling mistakes), others dramatically change theological implications, and others (like this one) invent evidence for things that the New Testament writers knew nothing about.

It is that texture: interdependent, complex and contingent, that makes it such a fascinating text, I think.


Filed under Uncategorized