Thursday, May 19, 2011

Is Text-Critical Work Actually Significant?

Often, when dealing with the minutiae of text-critical research, my pragmatic friends ask me if it really matters. Normally, I would delve into a list of significant passages where textual criticism has solved difficult problems. But today I would like to approach it from another angle. Based on my research on the text of the Genesis Flood narrative, the differing dates in the ancient witnesses demand explanation, and the results of this text-critical work will have significant ramifications in a number of areas:

1) They will help us understand the series of events of the Flood better with regard to the dates on which events occurred and the extents of time they lasted.

2) They will help us understand the theology of the narrative better. Most scholars see symbolic significance (or at least analogical connections) in the dates of the narrative, such as events occurring on significant days of the week, seasons, and/or later festival dates. Firm dates will help confirm or deny these allegations.

3) They will either support or contradict the various theories on Pentateuchal origins, as the coherence or lack thereof of the Flood chronology has been a central battleground in the history of biblical criticism. The significance of the debates over the composition of the Pentateuch is obvious to anyone familiar with the issue.

4) They will help us better understand the use of calendars in the Hebrew Bible. One particularly unexpected point of contact was brought to my attention by a friend of mine. He mentioned that the interpretation of Daniel's 70th week, upon which Dispensationalism largely rests, is heavily dependent upon the type of calendar used by Daniel. Most dispensational writers, in turn, point to the Flood narrative to confirm their calendar. But, rather ironically, the calendar of the Flood is one of the most hotly contested calendrical issues in the OT! It might be too much to say that dispensationalism hangs in the balance of the text-critical results of the Flood narrative, but I thought it was an interesting connection that shows how often the significance of text-critical results can go far beyond what we might naturally expect.

Thus, textual criticism is indeed a significant work.

Studies at the University of Birmingham, UK


I just received word that I was awarded the Doctoral Researcher Elite Scholarship for studies at the University of Birmingham in the United Kingdom. The scholarship covers tuition and research and living expenses for a full three-year Ph.D. program. Starting in October, God willing, I will spend the next three years researching the history of transmission of the text of the Hebrew Old Testament in light of the Dead Sea Scrolls. Charlotte Hempel, with a specialty in Qumranic and Second Temple Jewish studies, will be supervising my dissertation. I also hope that New Testament textual critic David Parker will be a valuable resource for interdisciplinary dialogue on issues of manuscripts and methodology. My family and I are very excited about these future studies and the adventure they are sure to bring as we move to a foreign country and culture. Your prayers would be greatly appreciated as we try to figure out the logistics.

A Synoptic Problem in Genesis 10 and 1 Chronicles 1

The concept of biblical authors using sources (particularly other biblical sources) is disconcerting to some people. This is not an altogether irrational fear, given the way liberals have often misused the idea to set biblical authors against each other, fragment the text to death, and ignore the final literary products of the various books of the Bible. That said, biblical authors did at times use other biblical works as sources, and careful analysis of apparently synoptic passages can be of great text-critical value.

My recent reading in 1 Chronicles 1 brought this consideration to the forefront of my mind, as I consistently found very precise correspondence between the genealogies of 1 Chronicles 1 and Genesis 10, not just in the names used, but even in very specific examples of precise correspondence of wording and syntax in narrative comments that seem to demand literary dependence. A few indicative examples will suffice to show that the Chronicler was using the text of Genesis as his source:

1) וְכ֖וּשׁ יָלַ֣ד אֶת־נִמְר֑וֹד ה֣וּא הֵחֵ֔ל לִהְי֥וֹת גִּבּ֖וֹר בָּאָֽרֶץ׃ "And Cush became the father of Nimrod; he began to be a mighty one in the earth." (1 Chr 1:10 = Gen 10:8)
2) וּכְנַ֗עַן יָלַ֛ד אֶת־צִיד֥וֹן בְּכֹר֖וֹ וְאֶת־חֵֽת׃ "And Canaan begat Zidon his firstborn, and Heth," (1 Chr 1:13 = Gen 10:15)
3) וּלְעֵ֥בֶר יֻלַּ֖ד שְׁנֵ֣י בָנִ֑ים שֵׁ֣ם הָאֶחָ֞ד פֶּ֗לֶג כִּ֤י בְיָמָיו֙ נִפְלְגָ֣ה הָאָ֔רֶץ וְשֵׁ֥ם אָחִ֖יו יָקְטָֽן׃ "To Eber were born two sons: the name of the one was Peleg (for in his days the earth was divided), and his brother's name was Joktan." (1 Chr 1:19 = Gen 10:25).

If the Chronicler was indeed using Genesis as his source, this means that 1 Chronicles 1 can be used as a textual witness to the text of Genesis 10 at points where the latter reproduces the former (while, of course, taking into account the possibility that the Chronicler could have adapted the material to fit the context of his own work). Thus, to the traditional witnesses to Genesis (MT, SP, G, Targums, Vulgate, Syriac, etc.) must be added 1 Chronicles. Admittedly, the text of Chronicles itself has undergone over two millennia of transmission and so must likewise be critiqued before being useful for reconstructing the text of Genesis, but an ancient witness it still is.

This brings us to one example textual problem in 1 Chronicles 1:7 || Genesis 10:4, where one of Yavan's sons (or tribe of descendants) is named variously Dodanim דֹדָנִֽים and Rodanim רוֹדָנִים, where the variants obviously arose from graphic confusion of the very similar first letter ד/ר. And given the literary dependence and content parallels, both obviously refer to the same person, so only one of them can be correct. The textual evidence among the major sources looks as follows:

Dodanim = דֹדָנִֽים MT(Gen 10:4)

Rodanim = רוֹדָנִים MT(1 Chr 1:7) SP (Gen 10:4) G (Ρόδιοι, Gen 10:4 and 1 Chr 1:7)

The majority Masoretic witness is divided in the two places. It is unlikely that the Samaritan Pentateuch and Septuagint would have assimilated the Genesis text to fit the 1 Chronicles reference. Thus many have concluded that the text of Genesis 10:4 originally said Rodanim רֹדָנִים instead of Dodanim דֹדָנִֽים and have emended the text accordingly (e.g., NRSV, NIV, NLT, and also I suspect is correct). Others have chosen Dodanim as the proper name in both cases (e.g., KJV). But many translators have chosen to represent the discordant majority MT readings in both cases, saying in Gen 10:4 that his name is Dodanim and in 1 Chr 1:7 Rodanim (e.g., RSV, NAS, NKJV, ESV, NET Bible). In my view, this latter option is entirely irresponsible and shirks the translators' responsibility to present a critically sound and coherent text. The man's name (or tribe's name) is either Dodanim or Rodanim, and not both, and the Bible translators who maintain the majority MT reading in both cases do a disservice to their readers out of a misguided faithfulness to the (in this case) textually corrupt MT. The church is better served with an accurate text than with a traditional text, and textual criticism, taking into account synoptic passages, can help us get there.

Merib-Baal or Meri-Baal in 1 Chronicles 9:40

My wife and I read an interesting note about the tiqqune sopherim ("emendations of the scribes") in my wife's archaeological study Bible today, and they mentioned that often scribes changed the names of individuals to remove the name of a pagan deity (for instance Ish-Baal "man of Baal" becomes Ish-Boshet "man of shame," and likewise Merib-Baal becomes Mephi-Boshet). Irony of ironies, in my personal reading tonight I was in 1 Chronicles 9 and came across exactly this genealogy of Saul. I noted a textual problem beyond the question of what the article in the study Bible mentioned, however, in the name of Merib-Baal.

1 Chronicles 9:40 gives the name of Jonathan's son Meri(b)-Baal twice:

"And the son of Jonathan was Merib-Baal (מריב בעל), and Meri-Baal (מרי בעל) begot Michah."

In 1 Chronicles 8:34, an identical verse, both instances are spelled Merib-Baal מריב בעל, strongly arguing that this is the correct spelling. It is interesting though, that the LXX (according to Rahlfs) has in all cases Μεριβααλ without two betas, which may mean that מרי בעל was read. HALOT mentions that an ostracon was found with the name מרבעל. Holladay actually argues that מריב בעל "antagonist of Baal" may even have been "reconceived" from מרי בעל.  In the end, it is clear that one of these two spellings put side by side is textually corrupt, either by haplography or dittography of the additional ב beth.

Friday, May 13, 2011

The Vocalization of Proverbs 31:21 and the Rule "Lectio Difficilior Potior"

My wife and I have a family tradition of reading Proverbs 31 at dinner on Friday nights. Today I read from the New English Bible for the first time and was alerted to an exegetical issue in Proverbs 31:21 that I had not seen before. This issue is a question of vocalization from a like consonantal text, so it is not properly text-critical in my view, but a note in the NET Bible seemed to me to make it so.

MT vocalizes Proverbs 31:21 as:

לֹא־תִירָ֣א לְבֵיתָ֣הּ מִשָּׁ֑לֶג כִּ֥י כָל־בֵּ֜יתָ֗הּ לָבֻ֥שׁ שָׁנִֽים׃
"She is not afraid of snow for her house, because all of her house is clothed in scarlet."
"Scarlet" here is שָׁנִים shanim.

But the LXX and Vulgate understand the same consonants differently.

The LXX reads it as the number "two" שְׁנַיִם shnayim and attaches it to the "coverings" at the beginning of the next verse: δισσὰς χλαίνας ἐποίησεν "she makes two coverings."

The Vulgate also reads it as the number "two" שְׁנַיִם shnayim, but it includes it at the end of verse 21: omnes enim domestici eius vestiti duplicibus "for her whole house is doubly clothed." The NEB and NLT agree with the Vulgate in reading "two garments" or "warm clothes."

The difference between the readings is one of interpretation of the same consonantal text, but the reason I bring it up is that the note in the NET Bible that tries to use a text-critical principle to resolve an exegetical problem. The NET Bible note says that the reading of the numeral "two" is the "easier reading and therefore suspect." The note is alluding to the often-abused text-critical canon of lectio difficilior potior, or "the more difficult reading is superior." While there is nothing wrong with the principle in general, wrongly applied it can lead scholars in the wrong direction. In particular, the NET Bible is using the principle to determine the correct vocalization of an ambiguous consonantal text. But the question of vocalization is not properly a text-critical decision, and it is questionable whether the text-critical principle preferring the more difficult reading can be made to apply in these situations.
The vocalization of the Hebrew text was not indicated in the original texts of Scripture (with the possible exception of original vowel letters). This means that many passages of Scripture, even when the consonantal text is agreed upon, are ambiguous and allow for multiple readings. The readings one chooses, then, are actually exegetical decisions rather than textual ones. In exegesis, I question whether the lectio difficilior canon is applicable. What it is essentially saying is that the most difficult interpretation is likely correct. Thus, the numeral "two" cannot be correct, because it fits too neatly into the context and practically begs to be read. But in exegesis, contextual suitability is the primary canon. What exegete in his right mind would look for the most unlikely, hard-to-understand, stretched, contextually unsuitable interpretation? Vocalizations should be chosen based on which reading best fits in the context, not based on which one makes the least sense! This kind of thinking only makes sense if the vocalization of MT is assumed to be original and authoritative and alternative readings are seen as secondary corruptions of the MT. This is a methodologically flawed approach. In reality, the consonantal text is vocalically ambiguous and requires independent interpretive disambiguation by the reader.

So when determining which reading to read for שנים, we need to look at which fits the context best. In favor of the reading "two," we have the immediate context of the verse and common sense, which says doubling up garments is more fitting for keeping warm in snow than putting on extravagant scarlet garments. In favor of the reading "scarlet," we have in the next verse a reference to the extravagance of the woman's fine linen and purple clothes and the possibility of a somewhat stretched metaphor of scarlet clothing = richness and abundance of clothing = warmth. At first glance, it seems to me like the numeral "two" may actually be the better interpretation, but either way, the point is that the exegetical question must be resolved by appeal to the best fit in context, not the worst fit.

Thursday, May 5, 2011

Benjamin Kennicott

I just found an old series of blog posts that are quite interesting about Benjamin Kennicott and his collation of Hebrew variants here. There were also a number of links to digitized works that I included in my collection of online Hebrew editions here, including the collations of Kennicott and De Rossi.

Messiah: Stripe or Stripes?


I was recently helping a friend of mine translate Isaiah 53:5, and came across the phrase וּבַחֲבֻרָתוֹ נִרְפָּא־לָנוּ "and by his stripe we are healed." He mentioned that another friend of his had made a significant theological point on the MT reading of חֲבֻרָתוֹ "his stripe" being singular. I pointed out, however, that the consonants could also be easily emended to the plural חֲבֻרֹתָיו "his stripes."

Though in most cases (as far as I can tell, all cases) the plural would have an extra י as in חברתיו, the two forms could easily have interchanged in the tradition. Ironically, in my reading in Ezekiel immediately thereafter, exactly such an interchange presented itself repeatedly. In Ezekiel 43:11, MT twice reads the singular phrase צורתו "its plan" as the Kethiv and the plural צורתיו "its plans" as the Qere. Furthermore, in both Ezekiel 43:11 and 44:5 MT has תורתו "its law" as the Kethiv and תורתיו "its laws" as the Qere. The same phenomenon can be found in Exod 28:28; 37:8; 39:4; Deut 5:10; 7:9; 8:2; 27:10; Josh 16:3; Ruth 3:14; 1 Sam 10:21; 26:7, 11, 16; 2 Sam 12:20; 1 Kings 16:19; Ezek 31:5; 33:13, 16; 40:6, 22, 26; 47:11; Amos 9:6. It is interesting that only once that I found, in 1 Kings 16:26, is the situation reversed with the plural in the Kethiv and singular in the Qere, showing a definite trend towards the plural in the Qere. This may show an understanding that haplography in such examples was more common than dittography. Perhaps noteworthy, however, is that this phenomenon apparently never occurs in the book of Isaiah.

After examining these other occurrences of the interchange between the תו and תיו endings, I was even more suspicious of the theological point based off of the singular חֲבֻרָתוֹ in Isaiah 53:5. Brief consideration of the LXX showed the singular reading τῷ μώλωπι αὐτοῦ "by his bruise," which slightly alleviated my fears. I decided to check the Great Isaiah Scroll 1QIsa(a) for curiosity's sake, and much to my surprise, it reads the plural חבורתיו! My suspicions (viz., my diagnostic conjectural emendation) were correct! The plural was indeed an ancient Hebrew variant that did not make its way into the Masorah of MT. 1QIsa(b), normally much closer to MT, had the singular חברתו.
All this to say, the singular "stripe" in Isa 53:5 may not be as significant as we make it out to be, or it may not even have been original. Perhaps we were healed by Messiah's "stripes" after all? One would be hard pressed to overthrow the singular reading with the single Qumran scroll, but I think this example proves that my cautious skepticism about basing significant theological points on textually dubitable minutiae is indeed warranted. It is the nature of language and exegesis that often important information must be drawn from the most precise of details, but this just goes to show that the exegete cannot naively assume the Leningrad Codex as his text without critically evaluating its readings for the given passage. Criticism of the text must precede exegesis of its theology, and any given theological induction from a text is at least as dubitable as its text.