Wednesday, May 3, 2017

Corrupted Texts and Manuscript Edits

If you spend time reading through academic commentaries on biblical texts, you often will come across individuals who claim that the text is probably corrupted and should be read in a different way. A good example of this kind of academic analysis is in the works of Frank Moore Cross (Canaanite Myth and Hebrew Epic) where he analyzes both Az Yashir (the song of the sea) and Psalm 29. In both cases he proposes edits to restore the text to what he thinks is its original version.

As someone with a reasonably good background in Talmudic argumentation, I often felt that these arguments were a bit too similar to the common Gemara phrase, "Don't read x, read y" often substituting an entirely different word in for the original one in order to justify some law or custom. I have several questions on this topic, I probably won't be able to answer them all, but I will provide what I think is a compelling example of what textual corruption might look like.

Here are the questions I have:

When the Talmudic authors say, "don't read x, read y" did they actually mean that there were versions of the text with y, or that the text might have been corrupted from y to x?

What types of textual corruption are likely, what kinds are unlikely?

What kind of manuscript edits are likely, what kinds are likely?

Can we distinguish between a text that has been corrupted and one that has been edited?

Which texts are the most likely to have been corrupted or edited?

Textual Corruption/Editing in Early Sources

The best way to figure out what corruption or editing would look like is to find examples of texts that were edited. To date we have several early versions that provide exactly this kind of analysis. By comparing the Masoretic Text (currently used by nearly all Jews) the Septuagint (Early Greek Translation), the Samaritan Bible, and the Dead Sea Scroll fragments, we can piece together what an original text might look like. If three sources say one thing but the fourth differs, then probably the fourth is edited. As a side note, this is precisely how the Talmud comes to conclusions about what the actual text should read. (See Masechet Soferim 6:4 for example).

While this approach works well for academics, and we've used it in this blog in the past to argue for different readings of specific verses, if you come from a religious background, this analysis won't carry much weight. After all, the Masoretic text is "obviously" the correct one, and all the others are corruptions.

Luckily the Tanach itself supplies us with some examples of the same text repeated in different places, but with significant edits. The most glaring example is the comparison between Divrei Hayamim (Chronicles) and Melachim (Kings). This was touched on in this post. But while this tells us a lot of what changes later editors were willing to make, it doesn't answer too much about textual corruption. Instead we look at two Psalms.

Psalms 14 and 53

Here are the texts of Psalms 14 and 53, in Hebrew and English.

תהילים פרק יד


א  לַמְנַצֵּחַ, לְדָוִד:
אָמַר נָבָל בְּלִבּוֹ,    אֵין אֱלֹהִים;
הִשְׁחִיתוּ, הִתְעִיבוּ עֲלִילָה--    אֵין עֹשֵׂה-טוֹב.
ב  יְהוָה--    מִשָּׁמַיִם, הִשְׁקִיף עַל-בְּנֵי-אָדָם:
לִרְאוֹת, הֲיֵשׁ מַשְׂכִּיל--    דֹּרֵשׁ, אֶת-אֱלֹהִים.
ג  הַכֹּל סָר, יַחְדָּו נֶאֱלָחוּ:    אֵין עֹשֵׂה-טוֹב--אֵין, גַּם-אֶחָד.
ד  הֲלֹא יָדְעוּ, כָּל-פֹּעֲלֵי-אָוֶן:    אֹכְלֵי עַמִּי, אָכְלוּ לֶחֶם; יְהוָה, לֹא קָרָאוּ.
ה  שָׁם, פָּחֲדוּ פָחַד:    כִּי-אֱלֹהִים, בְּדוֹר צַדִּיק.
ו  עֲצַת-עָנִי תָבִישׁוּ:    כִּי יְהוָה מַחְסֵהוּ.
ז  מִי יִתֵּן מִצִּיּוֹן,    יְשׁוּעַת יִשְׂרָאֵל:
בְּשׁוּב יְהוָה, שְׁבוּת עַמּוֹ;    יָגֵל יַעֲקֹב, יִשְׂמַח יִשְׂרָאֵל.

1 For the Leader. [A Psalm] of David.
The fool hath said in his heart: 'There is no God';
they have dealt corruptly, they have done abominably; there is none that doeth good.
2 The LORD looked forth from heaven upon the children of men,
to see if there were any man of understanding, that did seek after God.
3 They are all corrupt, they are together become impure; there is none that doeth good, no, not one.
4 'Shall not all the workers of iniquity know it, who eat up My people as they eat bread, and call not upon the LORD?'
5 There are they in great fear; for God is with the righteous generation.
6 Ye would put to shame the counsel of the poor, but the LORD is his refuge.
7 Oh that the salvation of Israel
were come out of Zion! When the LORD turneth the captivity of His people, let Jacob rejoice, let Israel be glad.

תהילים פרק נג

א  לַמְנַצֵּחַ עַל-מָחֲלַת, מַשְׂכִּיל לְדָוִד.
ב  אָמַר נָבָל בְּלִבּוֹ,    אֵין אֱלֹהִים;
הִשְׁחִיתוּ, וְהִתְעִיבוּ עָוֶל--    אֵין עֹשֵׂה-טוֹב.
ג  אֱלֹהִים--    מִשָּׁמַיִם, הִשְׁקִיף עַל-בְּנֵי-אָדָם:
לִרְאוֹת, הֲיֵשׁ מַשְׂכִּיל--    דֹּרֵשׁ, אֶת-אֱלֹהִים.
ד  כֻּלּוֹ סָג, יַחְדָּו נֶאֱלָחוּ:    אֵין עֹשֵׂה-טוֹב; אֵין, גַּם-אֶחָד.
ה  הֲלֹא יָדְעוּ, פֹּעֲלֵי-אָוֶן:    אֹכְלֵי עַמִּי, אָכְלוּ לֶחֶם; אֱלֹהִים, לֹא קָרָאוּ.
ו  שָׁם, פָּחֲדוּ פַחַד--    לֹא-הָיָה-פָחַד:
כִּי-אֱלֹהִים--פִּזַּר, עַצְמוֹת חֹנָךְ;    הֱבִשֹׁתָה, כִּי-אֱלֹהִים מְאָסָם.
ז  מִי יִתֵּן מִצִּיּוֹן,    יְשֻׁעוֹת יִשְׂרָאֵל:
בְּשׁוּב אֱלֹהִים, שְׁבוּת עַמּוֹ;    יָגֵל יַעֲקֹב, יִשְׂמַח יִשְׂרָאֵל.

1 For the Leader; upon Mahalath. Maschil of David.
2 The fool hath said in his heart: 'There is no God';
they have dealt corruptly, and have done abominable iniquity; there is none that doeth good.
3 God looked forth from heaven upon the children of men,
to see if there were any man of understanding, that did seek after God.
4 Every one of them is unclean, they are together become impure; there is none that doeth good, no, not one.
5 'Shall not the workers of iniquity know it, who eat up My people as they eat bread, and call not upon God?'
6 There are they in great fear, where no fear was;
for God hath scattered the bones of him that encampeth against thee; Thou hast put them to shame, because God hath rejected them.
7 Oh that the salvation of Israel were come out of Zion!
When God turneth the captivity of His people, let Jacob rejoice, let Israel be glad.

So what's going on here? The Psalms are almost identical, in fact we can sum up all the differences. They are:


  1. The introductory verse is different, although both are attributed to David
  2. The verse breakdown is different (from here on, I'll use v. x(y) where x is in psalm 14 and y is in psalm 53
  3. Some words are replaced with synonyms v. 1(2)
  4. Some words look as if one verse has a typo v. 3(4)
  5. In some places different names of God are used v. 2(3) 4(5) 7(7) 
  6. Minor grammatical differences, probably for clarity in reading v. 4(5), 7(7)
  7. One verse is completely different 5-6(6). 
Most of these are pretty much what you'd expect from scribal emendations, changing a word here and there, making things easier to comprehend. Perhaps a typo snuck in over the years. But overall, if you were to hand someone both of these they would probably say they are slightly different versions of the same Psalm.

But there's really one point of difference that is more striking, and that is difference 5. One Psalm (14) uses the tetragrammaton 3 times, and Elohim 3 times. The other only uses Elohim. Again, it is hard to imagine that this is a coincidence. Either the author of Psalm 53, didn't want to use the tetragrammaton and changed all the instances to Elohim, or the author of Psalm 14 changed some of the instances of Elohim to the tetragrammaton perhaps to ensure that the Psalm equated the two names. Either way, there is a clear theological motivation for the change, although we're not sure which one represents the original.

Answering the Questions

Now maybe we can look at the questions, and provide some semblance of answers to them, using these psalms as a guide.

When the Talmudic authors say, "don't read x, read y" did they actually mean that there were versions of the text with y, or that the text might have been corrupted from y to x? Answer: They probably did not have a version that said y. Although we can't rule it out. The types of changes often mentioned in the Talmud are not similar to the changes we see here. They tend to fall more in the "wordplay" category.

What types of textual corruption are likely, what kinds are unlikely? Answer: Single character typos are very likely, especially for letters that are similar. In this example 4 we have a Resh exchanged with a Gimel, which seems very plausible in the old Israelite writing system.

What kind of manuscript edits are likely, what kinds are likely? Answer: Replacing words with synonyms are plausible. Fixing up sentences grammatically are also plausible. Most importantly, larger theologically motivated edits are definitely plausible, especially when the name of God is concerned.

Can we distinguish between a text that has been corrupted and one that has been edited? Answer: Corrupted texts probably have more typo corrections than grammatical or theological edits.

Which texts are the most likely to have been corrupted or edited? Answer: There isn't enough sample size here to really answer this. But if you start looking at some of the other old sources, you can find out that the most questionable texts tend to be the ones with the most archaic language and poetry.