The last post begins by considering whether Rousseau believes in a legitimate state—political authority—and finishes by questioning whether Dylan believes in legitimate interpretation—textual authority. By examining an old puzzle in Biblical textual criticism, I’ll connect these two problems, suggesting that hermeneutics can never be entirely divorced from questions of community—politics in its broadest sense.
Let’s start with some genealogy. Moses’ father married his aunt: “And Amram married Jochebed his aunt; and she bore him Aaron and Moses” (Exodus 6:20). We know the meaning of the Hebrew, דודה, in part because it reoccurs in Leviticus 18:14, which says, “You shall not uncover the nakedness of thy father’s brother, you shall not approach his wife: she is your aunt” (דודתך). This verse comforts the lexicographer, but it may trouble the rabbi: should Amram have married his aunt? Sure, the laws of Leviticus had not yet been given. Still, this is Moses’ parentage, and it’s an uncomfortable juxtaposition.
Luckily, there are multiple texts of the Bible: the Septuagint, a Greek translation dating from between the third and first centuries BCE, calls Jochebed Amram’s “θυγατερα του αδελφου του πατρος αυτου,” the daughter of the brother of his father. That is, Moses’ parents are kissing cousins. The line is no anomaly: the Syriac Bible, translated around the second century CE, also reads “cousin,” and at Numbers 26:59, the Septuagint has a parallel variant reading which records Jochebed’s lineage. Cousin marriages, while illegal in much of the US today, were not a problem for Leviticus.
So which text is correct? Was Jochebed originally Amram’s aunt or his cousin? An Orthodox Jew can answer that question easily, since it is an article of faith that the Masoretic text (the one Jews use today) is totally accurate. For me, it’s a bit trickier. The Masoretic Hebrew text is in some ways very recent: the earliest full extant manuscripts of it date from the eleventh century CE, whereas there are manuscripts of the Septuagint and Syriac Bible from the third and fourth centuries CE. The Masoretic text, a massive editing project by a group of sectarian Karaites at (hermeneutic) war with Rabbinic Judaism, has its own biases and agenda.
Luckily, there’s a handy principle of textual criticism, lectio difficilor potior, which translates from Latin as “the more difficult reading is stronger” and means exactly what it says: given two plausible readings, both supported by reliable manuscripts, the more difficult reading, which makes less sense, is likely to be the earlier, original. This might sound crazy. Wouldn’t you think the most reasonable reading is likely to be the author’s version, and the more difficult reading a later corruption?
You would, and you’d be wrong. To understand why, imagine you are a scribe, and ask yourself why you change a text. Do you transform Jochebed from Amram’s cousin to his aunt? Probably not. You might easily, however, smooth over a difficulty. Thus, the difficulty has to be older, because no scribe in his right mind would ever change an easy passage into a hard one. Of course this reasoning only applies with variants that are well established, and unlikely to be single errors or bloopers. Still, I think it is very likely that in this case, the Septuagint and Syriac Bible are later; they reflect someone’s discomfort with the original, earlier Hebrew.
Dylan uses “reading” to mean “a fleshed-out interpretation of a text”; I use it more technically to mean “the actual words one reads in a particular version of a text.” Still, the two meanings are clearly connected: the Septuagint reading (narrow sense) of “cousin” reflects its authors’ reading (broad sense) in which the Hebrew “aunt” is problematic. Lectio difficilor assumes that scribes interpret, and that interpreters follow the principle of generosity, that is, that they try to understand the text in a way that makes it seem coherent, reasonable and sensible. The principle of generosity is an important part of a certain approach to the humanities. I was reminded regularly in philosophy and literature classes not to dismiss Nietzsche or Faulkner because at first glance, their ideas seemed obviously stupid. Indeed, canonical texts are canonical just because people have been reading them generously for a long time, starting from the assumption that they have something to say.
But as Dylan’s reading of Rousseau or Strauss’s of The Republic illustrate, the principle of generosity can lead us to think that the text always agrees with us. After all, who is to say what’s reasonable, true, or coherent? In many cases, those standards are contested in the first place. If, noting the totalitarian government proposed in The Republic, I conclude that Socrates must be writing ironically or esoterically, perhaps I am just refusing to take seriously the possibility that a great thinker disagrees with me about totalitarianism. Lectio difficilior works because over time, canonical texts come under pressure to conform to “generous readings”; the text is constantly being remade in its ideal image. Nor does that process stop when the letters and words of text are fixed: Genesis contains numerous anthropomorphisms and descriptions of God as having a physical body, but eight hundred years after Maimonides argued that God can have no body and suggested metaphorical readings of this language, the average Jew would deny vociferously that the Biblical God has a body.
Now, modern scholars read with lectio difficilor as if they were getting a “better” reading; actually, they are just getting a different reading, in particular, a reading that has been divorced from the history of interpretation, commentary, and transmission. It’s possible that such a reading reflects the “author’s original intent,” but maybe the text was meant to be read in terms of later commentaries and interpretations: Michael Fishbane argues that biblical narrative, which constantly circles back to and revises older biblical narratives, is thus continuous with the later Rabbinic tradition of reinterpretation. Further, why should we privilege the “divorced” meaning of the text? Religious Jews, for instance, are primarily concerned about the text’s role within a community, just as the philosophy professor may be primarily concerned with the text’s role in getting students to think deeply clearly together. Both these functions require ignoring “divorced” meaning. That’s not to say historical methods don’t have a role in communities, but that role has to be articulated in terms of the effect on the community—not in terms of “truth” as something entirely outside of community. A reader attuned to how attitudes towards homosexuality have changed can prevent us from over-reading the prohibition in Leviticus. Generous reading can be narcissistic reading, closing us off to difference; when we try too hard to see ourselves in the text, we can lose sight of others. “Divorced” readings can help us see that difference again. But such a reading is ultimately worthwhile (true, if you like) because of what it can do for us.