The Structure of Repetition

I titled last week's shiur "Twice Told Tales" and will now expand on that topic, relying upon an outstanding section of Meir Sternberg's The Poetics of Biblical Narrative (Chapter 11, "The Structure of Repetition: Strategies of Informal Redundancy," 365-440). Turning to Sternberg will also allow us to pause mid-series and reflect on the specific contributions of the modern literary approach to Biblical interpretation. In the series introduction, I wrote of midrashim anticipating the insights of modern scholars; for instance, Bereishit Rabba notes how the repeated use of the unusual phrase "haker na" in chapters 37 and 38 indicates that the Tamar episode punishes Yehuda for selling Yosef. Subsequent shiurim provide several examples of Rishonim also preceding the moderns regarding particular interpretations; for example, R. Eliezer Ashkenazi explains Lavan's clever usage of bekhira and tzeira as a criticism of Yaakov's stealing the blessing from his older brother, Abravanel identifies purposeful ambiguity in the phrase "and Ninveh will be overturned," and Rashi notes that the servant's account of meeting Rivka differs from the Torah's earlier narration of the event in order to convince her family to agree to the match. Apparently, the medieval commentaries were already on to these techniques. If so, wherein lies the innovation of Alter, Sternberg, and Grossman?

First, some of the techniques, such as type-scenes and key words, were not truly identified by earlier voices. Second, when earlier commentators did notice these literary elements, they often addressed them on an ad hoc basis and not in a systematic way. Even the introduction of terminology such as type-scenes and intertextuality can help us map out what Tanakh attempts to convey. The terminology forces us to define categories and analyze multiple possible variations within those categories.

With that introduction, we can return to Sternberg, who contrasts his approach with that of earlier commentators. On the topic of why repetitions of stories frequently involve variation, Radak downplays the significance of the changes.

And when these things get repeated, there occurs a change of words but the meaning is one, for such is Scripture's way with repetition: it preserves the meaning but not the words. (Commentary on Bereishit 24:39, trans. Sternberg)

For R. David Kimchi, Tanakh simply cites the same idea with different terminology. Umberto Cassuto saw the phenomenon as part of a larger Ancient Near Eastern love of repetition, which we find in Ugaritic as well. While in oral discourse this would mean exact repetition, as ancient listeners appreciated hearing familiar passages over again, the written word allows for some slight variation to maintain readers’ interest.

An ancient alternative approach is found in the Septuagint and the Samaritan Bible, both of which try to erase the variety in retellings by what Sternberg calls foresmoothing or backsmoothing. If Yosef's brothers tell Yosef after Yaakov's death that their father commanded him to forgive them (Bereishit 50:16-17), but we never hear such a conversation while Yaakov was alive, these texts will insert it into an earlier section to avoid any discrepancy. Sternberg cleverly writes: "Insofar as the Samaritan and Septuagint readings follow a version independent of the Masoretic, they establish a poetics of their own, whose compulsiveness bespeaks a form of narrative neurosis" (Sternberg 373). Chazal, in contrast, suggest that the brothers are fabricating to promote peace (Yevamot 65b). Where these other texts attempt to cancel discrepancy, our Sages, like modern literary commentators, often found meaning in the changes.

When it comes to Radak and Cassuto, who acknowledge the repetitions in our Masoretic text, including their variations, but downplay their significance, Sternberg claims they cannot account for the many diverse ways in which Tanakh manifests variation in repetition. Furthermore, they cannot explain situations where exact repetition slips through. Sternberg admits that some earlier voices, such as Abravanel and Nechama Leibowitz, did pick up on the artistic use of variation in repetition (as did some midrashim, such as the Gemara mentioned above), but they did not make it into a general interpretive methodology the way he and other modern scholars have. In terms of his disagreement with Radak and Cassuto, the greatest proof will be in advancing convincing, meaningful explanations for the changes. Last week, we saw what I think are some very convincing interpretations, including Eshet Potiphar modifying her description of Yosef based on her audiences. 

Another strong example helps highlight the difference in approaches. When Pharaoh and the Egyptians close in on the Jewish people stopped before the Red Sea, the people panic and complain:

And they said to Moshe, "Was it for lack of graves in Egypt that you took us to die in the wilderness? What is this you have done to us, to bring us out of Egypt? Isn't this the word we spoke to you in Egypt, saying, 'Leave us alone, that we may serve Egypt, for it is better for us to serve Egypt than for us to die in the wilderness?’" (Shemot 14:11-12, based on Alter's translation)

Now, the prior accounts never place such words in the mouths of the people. Ibn Ezra smoothens out the discrepancy.

“Isn't this the word” – This is not stated. However, we know that this was the case, for how could they say to his face something that was not true? This remark was included in “but they hearkened not unto Moses” (Shemot 6:9).

Ibn Ezra assumes that the people would not lie to Moshe, and he finds an earlier general statement within which to insert this previously unheard-of complaint. As expected, the Samaritan Pentateuch backsmooths the disparity by inserting just such a complaint into the mouths of the Israelites in Egypt. Modern scholars, however, note the possibility that, in their frightened fury, the people exaggerate or fabricate – a possibility that does not occur to those (including Ibn Ezra) who are committed to removing variation.

Let's temporarily put aside cases of variation in repetition and ask what Scripture accomplishes with precise repetition. Once exact replication is not an absolute standard, it becomes more meaningful when employed. Sternberg outlines several possible usages: The exactitude may indicate obedience, as when Noach carries out the divine command to the letter (Bereishit 6:18 and 7:7). It may justify a Biblical character's perspective, such as when we assume Sara reports correctly that a pregnant Hagar now treats her lightly since her complaint, "I became slight in her eyes" (16:5), exactly echoes the narrator's account, "and her mistress seemed slight in her eyes" (16:4). Finally, it conveys divine omnipotence and omniscience; "God said ‘Let there be light,’ and there was light" (1:3) serves for the former, and God asking Kayin "Why are you incensed and why is your face fallen," the exact words the narrator used to describe Kayin's reaction, serves for the latter. Similarly, God knows that Sara "laughed" even though she "laughed inwardly" (18:12-13).

Sternberg also adds some excellent explanations for variation, one of which I found particularly subtle. We have seen instances in which the second account adds a detail, subtracts a detail, or changes details. Sometimes, however, the variation is a simpler matter of sentence structure:

Now Yisrael loved Yosef more than all his children, because he was the son of his old age, and he made him a coat of many colors. And when his brothers saw that Yosef was more beloved to their father than all of the brothers, they hated him and could not speak peacefully to him. (37:3-4)

Two successive pesukim state that Yaakov loved Yosef the most, but the first one places Yaakov at the beginning of the thought ("Yisrael loved Yosef") while the latter puts Yosef front and center ("Yosef was more beloved"). The latter verse conveys the brothers’ reaction: they primarily blame Yosef, not their father, for their non-favored status; therefore, their bitter thoughts begin with their brother.

In another example, Sternberg concurs with Chazal. Sara is incredulous about the possibility of having a child: "after I have withered, will my skin grow gentle? And my husband is old" (18:12). In the next pasuk, God asks Avraham why Sara is doubtful, and says she said "and I am old." A Gemara (Yevamot 65b) suggests that God's report purposely deviates from the original, to avoid offending Avraham with the idea that his wife thought him elderly.

In the next instance, one of the medieval commentaries anticipated Sternberg. Pharaoh's two dreams appear several times, as we first have the narrator record them, then Pharaoh tells them over to Yosef, and finally, Yosef interprets them. Sternberg's thorough analysis deserves a close reading, but I will restrict our discussion to one point. According to Sternberg, the Egyptian magicians cannot interpret the dreams to Pharaoh's satisfaction because they think the dreams convey separate messages when they actually teach the identical idea. Seforno earlier saw this same point in a subtle shift in 41:8: "And Pharaoh told them his dream and there was no one to interpret them for Pharaoh." Pharaoh is certain that the twin dreams are one and is not convinced by any interpretation that relates to them in the plural form. Not surprisingly, Pharaoh employs only the singular when describing his experience to Yosef (42:15).  

Our next example relates to a crucial point in Jewish political philosophy. Despite the instructions in Devarim 17 about establishing a monarchy, Shmuel reacts quite negatively when the people request a king. What explains his negativity?

And they said to him: "Behold you have grown old, and your sons have not walked in your ways, and now place upon us a king who will judge us like all the nations." And the matter was evil in the eyes of Shmuel when they said "give us a king to judge us" and Shmuel prayed before God. (Shmuel I 8:5-6).

Many readers might think our prophet is upset about their desire to emulate the nations, but that is the specific part of the citation left out when we read about his anger. Apparently, Shmuel objected to their looking for a judge. Sternberg sees this as a self-centered thought on Shmuel's part, since Shmuel had hoped his sons would inherit his role as judge of Israel and took the people’s request for a monarch to judge as a personal rejection. His approach fits well with the next verse, in which God tells Shmuel, "they have not rejected you, but it is Me they have rejected."

However, alternative possibilities exist. Perhaps Shmuel does object to their seeking a judiciary king, but not out of selfish concerns. Rabbeinu Nissim develops a famous theory which finds two spheres of justice in Jewish law: that of the courts and that of the king. The courts create an idealized justice system meant to bring about the “divine overflow.” However, idealized justice does not always meet a people's practical needs, and the king steps in as judge to address more pragmatic concerns of a functioning society. For example, the ideal system makes it extremely difficult to punish criminals; the king will occasionally act as sole judge to administer punishments necessary to maintain a safer public square (Derashot Ha-Ran 11).

Rabbeinu Nissim explains that the people should appreciate the ideal justice system that brings down the divine influence. Asking for a human monarch to judge means that they think in purely pragmatic terms, devoid of idealism. Thus, Shmuel gets angry for reasons unconnected to his personal role and honor. In either case, both the Ran and Sternberg agree that the key to understanding Shmuel’s anger lies in the shift from "who will judge us like all the nations" to "to judge us."

Sternberg ends the chapter with a powerful warning not to force the issue every time we discover a variation, since some may not be meaningful. He cites the Talmudic story (Pesachim 22b) about a fellow named Shimon Ha-Amsoni who took on the audacious task of making a derasha every time the Torah uses the word et. According to Shimon, the word et always comes to include something more – as in a well-known teaching that “Kabed et avikha ve-et imekha, honor your father and your mother” (Shemot 20:12) alludes to an obligation to honor one’s step-mother as well as one’s mother (Ketubot 103a). However, when Shimon arrived at the pasuk "et Hashem Elokeha tira,” he was reluctant to include anything else along with God as an object of reverence. When his students questioned this move, given his life’s mission, Shimon responded: "Just as I received reward for the expounding (derisha), so should I receive reward for the desisting (perisha).” So too regarding our current category, all the insight that Sternberg and others have found in repetition with variation should not motivate us to advance unconvincing explanations. The modern literary approach, like all approaches, can be used poorly. Nevertheless, when used well, it produces glorious results.