The Repeated Use of “Vayo’mer” Within One Character’s Speech

This article is dedicated to the memory of my mother, Ada Landau z“l.

The Tanakh often divides one character’s remarks into parts by repeating the word vayo’mer (“he said”) to introduce subsequent parts, even though there has been no change in speaker and no comment by the narrator between the statements which would justify (at least in our view) the repeated use of the word.¹ Except for the rare instances which I will point out, the traditional Jewish commentators in Miqra’ot Gedolot don’t comment on the reason for the repetition of the word. We shall also see that modern commentators sometimes view the repeated use of vayo’mer as a sign that the parts of the character’s remarks originated in different source documents. However, they don’t invariably explain this phenomenon in this way, leading to an inconsistent treatment of the phenomenon as such.

In a 1948 article dealing with the question of the ancient pronunciation of the divine name YHWH, G.J. Thierry had occasion to briefly remark that in biblical narratives, the “old oriental narrative style” often involves the repeated use of a superfluous (in our view) term like vayo’mer to stress the importance of each utterance.² He brings numerous examples from Genesis and Exodus to support his point, but he simply cites them without commenting on them (except in one case, which we shall note).

In this article, I will analyze in detail the examples Thierry cites, add to them, assess the extent to which modern commentators actually see this phenomenon as a sign of the combination of different sources, and characterize the functions which it serves.

Division of a Character’s Speech Into Two Remarks or Parts, With the Second Part Introduced by (a Repeated) “Vayo’mer”

Nine of Thierry’s examples are speeches that almost always begin with the word vayo’mer, and are divided into two parts by a second vayo’mer. I will attempt to propose a reason for each such division, and will cite any relevant comments made by the traditional Jewish commentators in Miqra’ot Gedolot that appear to bear on this issue. I will also indicate whether, in the eyes of three modern commentators who use source criticism (S.R. Driver,³ Martin Noth⁴ and R.E. Friedman⁵), the appearance of such a division by the word vayo’mer is an indication of different sources.

In two of Thierry’s examples, the two remarks or parts of God’s speech apparently address two different subjects; both are introduced by vayo’mer (lahem/lo) ’Elohim (“God said to them/him”):

In Genesis 1:28-29, God’s first statement constitutes His blessing of fertility and dominion to the new human couple; His second statement is His grant of the plant kingdom to the humans, as well as to all the other newly-created animals, for food.

In Genesis 35:10-11, God’s first statement changes Jacob’s name to Israel; in His second statement He blesses Jacob with fertility and with the land He promised to Abraham.

In these two examples, all three of our sample modern commentators agree that both parts of each speech belong to the P source. They thus don’t see the repetition of “God said” as a sign that different sources have been used. However, none of them comment about the reason for the repetition.

In two additional examples, the character speaking is not God/YHWH but a human being. In each example, the character’s speech is divided into two parts by a repeated vayo’mer (in Sarah’s case, vato’mer, “she said”). In each case, each part presents a different view of or different information about the same subject:

In Genesis 21:6-7, Sarah’s first remark serves to introduce the root tz.ḥ.q. (“laugh,” “joke”), which ties in with the child’s name Yitzḥaq, into her assessment of her having a child. Her second remark expresses her surprise and pleasure at having had a child with Abraham. The commentary Da’at Zeqeinim tries to find a reason for the repeated vato’mer. There its purpose is an indication that praise will follow; in this view, in v. 7 Sarah praises the God who told Abraham that Sarah would nurse children. This commentary seems to be implying that vato’mer is repeated so as to indicate a change in the tenor or tone of Sarah’s words from happiness to praise.

In Genesis 31:48-53, Laban’s two sets of remarks explain the purpose of the mound of stones which he and Jacob have erected. In vv. 48-50 he refers to it as showing that God/YHWH would be a witness if Jacob mistreats Laban’s daughters, while in vv. 51-53 he refers to the mound and to “the” pillar (which was mentioned in v. 45) as marking a border between himself and Jacob, enforced by “the gods of their fathers” Abraham and Nahor. Notice that Laban’s two sets of remarks refer to God in different terms, portray different functions of the mound, and differ as to whether or not a pillar is involved.

In each example, two of our three modern commentators explain the character’s two statements as originating in different source documents. For Genesis 21:6-7, Driver assigns both verses to E, but Noth and Friedman both assign v. 6 to E and v. 7 to J. For Genesis 31:48-53, Friedman assigns all the verses to E, but Driver assigns vv. 48-50 to J and vv. 51-53 to E, and Noth assigns all the verses to J except vv. 50 and 53b, which he assigns to E.⁶

Notice that in each example, one of the three commentators finds no evidence of multiple sources in the repetition of vayo’mer. On the other hand, the view that multiple sources are involved seems to be based on the fact that each part of the character’s speech presents a different view or different (even conflicting) information about the same subject.

Two more examples exist in which a human character’s speech is divided into two parts by a repeated vayo’mer. In these examples, the two parts of the character’s speech are repetitive, stating the same subject matter first more extensively and perhaps using more flowery language, and then in simpler words:

In Genesis 15:2-3, both verses of Abram’s speech are introduced by vayo’mer Avram (“Abram said”). Abram’s first statement is more flowery; it uses the word ’ariri for “childless” and points out with a play on the sound of the words that ben-mesheq beiti (“the steward of my house”) is (since Abram is childless) dameseq ’Eliezer (“Eliezer of Damascus”). His second statement states in simpler language that he has no son and his servant will be his heir.

In Genesis 41:39-41, both parts of Pharaoh’s speech are introduced by vayo’mer Par’oh ’el-Yosef (“Pharaoh said to Joseph”). Pharaoh’s first statement in vv. 39-40 gives his positive assessment of Joseph and indicates in somewhat flowery language that therefore Joseph will be his second-in-command: ’al-picha yishaq kol-‘ami / raq hakisei’ ‘egdal mimeka (“by your word shall all my people act / only the throne will I exalt above you”). However, his second, shorter statement in v. 41 sounds like a formal statement of appointment, i.e. “I hereby appoint you…” This explanation is proposed by Rashi who cites Onkelos’ Aramaic translation ḥazay maneiti yatach (“see, I appoint you…”). Rashbam goes further along this line, commenting that Pharaoh’s statement in v. 41 accompanied Pharaoh’s giving Joseph his ring in v. 42; in other words, it was all part of the same formal, official act.

Our sample modern commentators are quite divided as regards these two examples.

As to Genesis 15:2-3, Friedman assigns both verses to J (albeit with some possible outside influence), but Driver assigns v. 2 to E and what he calls the “parallel” v. 3 to J, and Noth assigns vv. 2a to J and v. 3a to E, calling v. 2b a non-source addition and assigning v. 3b to J. Again, one of our three commentators has no problem seeing these verses as the product of one author. Driver (and Noth) presumably base their views on the fact that the two verses (or their first halves) are “parallel,” i.e. provide the same information using different words.

As regards Genesis 41:29-41, two of our three commentators (Driver and Friedman) assign all the verses to the one source E, but Noth assigns vv. 39-40 to E and v. 41 to J. The line of reasoning indicated by Onkelos, Rashi and Rashbam would argue that there is no need to see v. 41 as a “parallel” to vv. 39-40 arising from a different source.

In Genesis 37:21-22, Reuben’s first statement begins with vayo’mer and his second with vayo’mer ‘aleihem Reuven (“Reuben said to them”). V. 21 actually begins with the narrator’s comment that Reuben saved Joseph from the other brothers, which is then followed by Reuben’s first brief statement lo’ nakenu nafesh (“let’s not kill him”). His second statement in v. 22 ’al-tishpechu-dam (“don’t spill blood”) repeats this in different words and then proposes an alternative (“throw him into this cistern”); the narrator then repeats that Reuben intended to save Joseph from the other brothers. Genesis 37:21-22 is thus constructed in chiastic form: narrator’s explanatory comment, Reuben’s statement, Reuben’s expanded statement, and a repetition of the narrator’s explanatory comment. The repeated vayo’mer may be serving to highlight Reuben’s alternative proposal.

Two of our three modern commentators see these verses as originating in different sources: Driver and Noth see v. 21 as J and the “parallel” v. 22 as E, presumably because they say the same thing in different words. Friedman, however, sees both verses as E because they both see Reuben as the brother who saved Joseph; he argues that Genesis 37:26-27 give J’s version in which Judah saved Joseph. The chiastic structure of vv. 21-22 may be an additional argument in favor of these verses being the product of one author.

In Exodus 32:7-10, YHWH first tells Moses (introduced by vayedaber YHWH ’el-Moshe, “YHWH spoke to Moses”) what the people have done with the Golden Calf (vv. 7-8); in vv. 9-10 (introduced by vayo’mer YHWH ’el-Moshe, “YHWH said to Moses”), He draws a more general conclusion that the people is hopeless and proposes a solution: destroy them and make Moses into a new nation. The commentator Or Hachayyim seems to take the new vayo’mer in v. 9 as indicating a change in focus; vv. 7-8 deal with YHWH’s focus on what Moses needs to do and why, whereas in vv. 9-10 YHWH changes His focus to Israel and their sin.

In this example, all our sample modern commentators see these verses as belonging to one source. Driver and Friedman see all the verses as E (Driver feels that the JE redactor may have influenced them somehow), while Noth sees them as J.

Finally, in Genesis 20:9-10, Abimelech’s first statement is introduced by the narrator’s comment that Abimelech summoned Abraham and then continues vayo’mer lo (“and said to him”); his second statement is introduced by vayo’mer Avimelech ’el-Avraham (“Abimelech said to Abraham”). The commentator Or Hachayyim notes the repetition of vayo’mer and explains that the first statement is an expression of outrage, of reproof and complaint, whereas the second statement is a demand for an explanation of Abraham’s action. Or Hachayyim writes explicitly that the text therefore separates the two statements (which have two different tones and purposes) by using a second vayo’mer. All of our three modern commentators assign these verses to E.

Although Thierry doesn’t cite this text, Exodus 3:5-6 is yet another example. God speaks to Moses at the burning bush, making two consecutive remarks each introduced by vayo’mer. In v. 5, He tells Moses to remove his sandals because the ground upon which he is standing is holy; in v. 6, without any intervening change of speaker or narrator’s comment, He introduces Himself to Moses as “the god of your ancestors.”⁷ The repeated vayo’mer may serve to separate the warning/prohibition from the main focus, God’s introduction of Himself, as well as to highlight the importance of the second remark.

In commenting on these verses, William H.C. Propp⁸ writes that the repetition of vayo’mer within God’s address to Moses has sometimes been taken as a sign that the statements originated in different source documents. Driver sees both verses as E. However, Noth (and Friedman agrees) assigns v. 5 (and vv. 4a and 7-8, all of which contain the name YHWH) to J, and v. 6 (and vv. 4b and 9ff, all of which use the divine name ’Elohim) to E; in their view, these fit into the rest of the J and E continuous versions of the same story which are combined in this chapter.

Summary of the Examples in Which a Character’s Speech Is Divided Into Two Parts by “Vayo’mer”

First, the division of a character’s remarks into parts separated by vayo’mer is a phenomenon that occurs repeatedly, and is not limited to the divine character but is used with humans as well. Second, in four of the examples our sample modern commentators assign all the parts of the character’s speech to the same source; in one additional example, two out of the three commentators do so. Thus, modern commentators don’t invariably explain this phenomenon as resulting from the splicing of two sources; at least implicitly, it seems to be understood by these commentators as a stylistic technique. In these five examples, we have seen that there are several possible explanations for the division of the character’s remarks into separate parts:

Each remark relates different information, and the repeated vayo’mer marks off and highlights each separate piece of information (Gen. 1:28-29 and 35:10-11).

The remarks vary in tone or focus. For example, the second remark constitutes an “official” statement based on the first remark (Gen. 41:39-41, per Onkelos, Rashi and Rashbam), or the first remark is an expression of outrage while the second demands an explanation from the other person (Gen. 20:9-10, per Or Hachayyim).

The first remark gives specific information, while the second presents the character’s general conclusion and proposal based on that specific information (Ex. 32:7-10).

In all of these cases, the repetition of vayo’mer seems to function as a stylistic technique which highlights the individual importance of each separate remark and serves as a transition between subjects, tones, or information and logical conclusion.

Third, in the remaining five examples two of our sample modern commentators see the parts as arising from different sources while only one assigns all the parts to one source. There are several justifications for seeing the parts as indicating different sources; we can challenge the justification in at least two of the five examples:

Each part expresses different (Gen. 21:6-7) or even conflicting views (Gen. 31:48-53). We should note, however, that Genesis 21:6-7 does not seem different from the examples above in which different information is given in each remark and yet the commentators see both remarks as arising from the same source. As we have seen, the commentary Da’at Zeqeinim sees this as signaling a change in tone, just as Or Hachayyim sees vayo’mer in Genesis 20:9-10.

Each part says virtually the same thing in different words (Gen. 15:2-3 and 37:21-22). As to Genesis 15:2-3, an alternative explanation is theoretically possible, namely that the second remark was added by a later glossator to explain the first remark in simpler words; if this were so, we should note that the glossator made use of a characteristic biblical narrative stylistic technique to attach his gloss. As to Genesis 37:21-22, we have seen that the two verses have a chiastic structure which would argue for them having been written by the same author.

In only one example, Exodus 3:5-6, two of the commentators argue that Exodus 3 in general is composed of a combination of two continuous sources, and each of vv. 5 and 6 can be smoothly integrated into one of those two versions.

Division of a Character’s Speech Into Three Parts by the Repeated Use of “Vayo’mer”

Thierry cites four examples in which a speech (in all four cases the speech is by God/YHWH) is divided into three successive statements, with at least the second two each introduced by the repeated use of vayo’mer. As in the examples above, the appearances of vayo’mer are neither preceded by any change of speaker nor by any narrator’s comment which would (at least to us) justify such repetitions. Although it would seem that the appearance of three successive statements might more easily be explained as either arising from different sources or from a later scribal gloss or glosses, in these four cases our sample modern commentators all assign all three statements to one source. We should also be aware that biblical narrative often uses groups of three repetitions of an act to heighten the drama and importance of both the process and of the final result.⁹

In Genesis 9:8-17, God makes three consecutive statements, each introduced by vayo’mer ’Elohim (’el-Noah) (“God said (to Noah)”). God’s first statement (vv. 8-11) is His promise (covenant); the second (vv. 12-16) gives the sign of the promise (covenant). Thus the repeated use of vayo’mer in vv. 8 and 12 serves to distinguish between the revelation of the covenant and the revelation of its sign. God’s third statement (v. 17) restates v. 12, repeating the phrase zo’t ’ot-habrit (“this is the sign of the covenant”) but using the phrase ’asher haqimoti (“which I establish”) instead of ’asher-’ani noten (“which I place”) and using the phrase beini uvein kol-basar ’asher ’al-ha’aretz (“between Me and all life on the earth”) instead of beini uveineichem uvein kol-nefesh ḥayah ’asher ’itchem (“between Me and all of you and all living creatures who are with you”). Because of the similarities between vv. 12 and 17, it is possible that vv. 12-17 comprise a unit with v. 12 as an introduction, vv. 13-16 as the content, and v. 17 restating v. 12 in different words as a wrap-up or summary.¹⁰ All of our three modern commentators assign all of these verses to P.

Genesis 17:9-16, which Thierry cites, should be considered together with 17:3-8 which is also part of God’s speech. The first part is introduced by vayedaber ’ito ’Elohim lei’mor (“God spoke with him as follows”) and the second and third parts are each introduced by vayo’mer. In His first statement (vv. 3-8), God declares that He is making a covenant with Abram, changes his name to Abraham, promises to give him many descendents, to be his and their God and to give them the land of Canaan. This first part thus includes several different elements as part of the covenant. God’s second statement (vv. 9-14) specifies that the (sign of the) covenant which Abraham must keep is circumcision of all males. His third statement (vv. 15-16) relates to Sarah; God changes Sarai’s name to Sarah and promises to give Abraham a son through her. In short, God’s speech is divided into a main section which includes several different subjects, followed by two additional sections which each address an additional, different subject. All three of our modern commentators assign all of these verses to P.

Exodus 3:13-15 is the only example which Thierry discusses. He cites Eerdmans (1910) as an example of those who see the three uses of vayo’mer in vv. 14-15 as a sign of different authors. According to Thierry, Eerdmans sees v. 14, which intervenes between Moses’ question in v. 13 and YHWH’s answer in v. 15, as a “marginal annotation, made by two ancient scholars” to comment on the phrase ki-’eheye ’imach (“I will be with you”) in v. 12. The repeated use of vayo’mer in v. 14 shows that the verse was “refashioned” and is a gloss.

However, Thierry disagrees. He notes that several examples exist in Tanakh in which God appears to a human character, that character asks for God’s name, and He replies evasively.¹¹ Exodus 3:14 is one such case, in which two apparently evasive answers are given to the question Moses asks in v. 13; however, they are answers which focus attention on the connection between God’s name YHWH in v. 15 and the Hebrew verb “to be/exist.” This God, says v. 14, is the one who is, is there with you, and the implication is that this is why His name is YHWH. Eerdmans may thus have a point when he sees v. 14 as responding to ki-’eheye ’imach in v. 12.

Rashi cites TB Berachot 9b, which proposes essentially the same explanation of ’eheye ’asher ‘eheye (“I will be who I will be”) in v. 14a: “I am the one who will be with them in this time of oppression, and I am the one who will be with them when future kingdoms oppress them.” Rashi then makes use of Shemot Rabbah’s explanation of the second vayo’mer in v. 14. After God’s remark in v. 14a, Shemot Rabbah adds that Moses told YHWH: “Don’t have me mention future oppressions – the people have enough on their plate with the current one!” YHWH agreed and therefore in v. 14b He issues a revised answer to Moses’ question which omits ’asher eheye: “‘I will be (with you)’ sent me to you.” According to Shemot Rabbah, then, the second vayo’mer hints that there is an unstated change of speaker between v. 14a and 14b. If, however, we don’t assume a response by Moses which doesn’t appear explicitly in the text, then the first vayo’mer appears to introduce an evasive response, the second one introduces a remark that converts that first evasive response into a “name” which answers Moses’ question, and the third introduces God’s “real” answer: YHWH. There is thus a logical sequence set up by the three uses of vayo’mer: evasion using the verb “to be,” creation of a “name” using that verb, and the answer – God’s “real” name is a form of that verb. This appears to be an application of the principle of using a group of three to heighten the importance of the group and of the final result. All of our three modern commentators assign all of these verses to E.

In Ex. 33:19-23, YHWH addresses Moses’ request to see Him, with each of His three statements introduced by vayo’mer. In His first statement (v. 19) YHWH reveals His intention to grant Moses’ request. In the second (v. 20), He raises the problem of Moses dying if he sees His face.¹² In the third (vv. 21-23), He reveals His planned solution: placing Moses in a cleft in the rock and shielding him from seeing His face. Thus, these are three statements which are a logical progression. Driver sees these verses as J (with perhaps some JE redactor influence); Noth sees them as a non-source addition; Friedman sees them as E.

To these examples, I would add a fifth: Genesis 16:9-12. Here a manifestation of YHWH (an “angel”) speaks to Hagar telling her three brief things, each introduced by vayo’mer lah mal’ach YHWH (“An/the angel of YHWH said to her”). First, in v. 9, He commands her to return to her mistress and submit herself to her. Second, in v. 10, He makes a general promise to give Hagar uncountable descendents. And third, in vv. 11-12, He tells her specifically that she is pregnant with a boy, what to name him, and what his nature will be. In short, a command, a general promise, and the specific way in which that promise will begin to be fulfilled. Rashi, basing himself on the rabbinic principle that an angel can only execute a single mission and on the possibility of understanding the text as referring to “an” angel each time, explains each vayo’mer as indicating that each statement was said by a different angel; Rashi’s explanation implies that each statement deals with a separate subject, which perhaps implies that the repeated use of vayo’mer functions to separate and highlight each. However, these three remarks can also be seen as a sort of logical progression: go back, you have a great future, and this is how that future will begin to come to pass. Driver and Friedman assign all the verses to J; Noth calls vv. 9-10 a non-source addition and assigns vv. 11-12 to J.

Summary of the Examples in Which a Character’s Speech Is Divided Into Three Parts by Repeated Uses of “Vayo’mer”

First, almost all our modern sample commentators agree that in each of the above five examples, the entire speech is the product of one author (the exception is Noth’s view of Genesis 16:9-12). Thus, although they don’t explicitly say so, they must see this phenomenon as a stylistic technique that links three parts of one speech.

Second, we have seen three justifications for the division of a speech into three parts by repeated uses of vayo’mer:

The word is used to separate and highlight three elements of a logical progression or development of a subject (Ex. 3:13-15 and 33:19-23, and possibly Gen. 16:9-12).

The word is used after a main speech with numerous topics, to set off and highlight two additional topics (Gen. 17:3-16).

In Genesis 9:8-17, the first two uses of the word divide between two subjects: God’s explanation of His covenant/promise and His explanation of the sign of that covenant/promise. The third appearance of the word in Genesis 9:17, however, is difficult to explain.

In all cases, this stylistic technique lends increased weight not only to each separate statement but to the series as a whole. This is further strengthened by the fact of there being three statements involved.

Conclusion

Although in any given case it is theoretically possible that the different sections of a character’s speech may arise from different source documents, or from a scribal gloss or glosses, we have shown that the phenomenon of the repeated use of vayo’mer within one character’s speech appears most often to be a stylistic technique. This technique can function to separate subjects, indicate a change in tone or focus, or indicate the steps in a logical thought process.

Footnotes

1. All translations of biblical phrases are my own. ↩

2. Thierry, G.J., “The Pronunciation of the Tetragrammaton,” Oudtestamentlische Studien (OTS) 5:30-42, 1948, see especially pp. 37-38. ↩

3. For the examples from Genesis: Driver, S.R., The Book of Genesis, Fifth Edition, New York: Edwin S. Gorham, 1906. For the examples from Exodus: Driver, S.R., An Introduction to the Literature of the Old Testament, New York: Meridian Books, 1957. ↩

4. Campbell, Antony F. and Mark A. O’Brien, Sources of the Pentateuch, Minneapolis: Fortress Press, 1993. This book summarizes Martin Noth’s source division of the Pentateuch, which appeared in several of his works in the 1950s and 1960s. ↩

5. Friedman, Richard Elliott, Who Wrote the Bible?, New York: Harper & Row, 1987. ↩

6. For our purposes, in cases where a commentator splits a character’s speech between sources J and E, it is immaterial why the commentator assigns a particular piece to J or to E; our only interest is the fact that he splits the speech into two sources. ↩

7. See Ramban on Exodus 3:6, where he gives examples showing that ’elohei ’avicha refers not to “your father” (singular) but to “your fathers/ancestors” (plural). Note also that in v. 13, Moses echoes back the phrase to God but changes it to the plural ’avoteichem (“your fathers/ancestors”). ↩

8. Propp, William H.C., Exodus 1-18, Anchor Bible Vol. 2, New York: Doubleday, 1999, p. 192. On p. 193, he says that the three uses of vayo’mer in Exodus 3:14-15 “suggests editorial interference to most.” ↩

9. For example, Noah’s sending of three doves from the Ark in Genesis 8:8-12; Jacob sending three presents of livestock to Esau in Genesis 32:14-20; Elijah pouring water three times onto the altar on Mt. Carmel in I Kings 18:34. ↩

10. And yet, it seems odd that such a wrap-up statement would be introduced once again by vayo’mer, which would seem to grant it extra focus. I would suggest as another possible explanation that v. 17 was a different form of v. 12 from a different traditional rendition of God’s revelation of the sign. The author of this section may have wanted to include both forms of the verse and therefore placed the v. 17 form after vv. 12-16, introducing it with vayo’mer in accordance with biblical style to show that it was an alternative tradition and using it to wrap up or restate the contents of the section. ↩

11. For example, Genesis 32:30 and Judges 13:17-18. ↩

12. Ibn Ezra cites Judges 13:22 and Genesis 32:31 as examples of this belief; see also Judges 6:22-23. ↩