Bar-Ilan University's Parashat Hashavua Study Center
Lectures on the weekly Torah reading by the faculty of Bar-Ilan University in Ramat Gan, Israel. A project of the Faculty of Jewish Studies, Paul and Helene Shulman Basic Jewish Studies Center, and the Office of the Campus Rabbi. Published on the Internet under the sponsorship of Bar-Ilan University's International Center for Jewish Identity. Prepared for Internet Publication by the Computer Center Staff at Bar-Ilan University. Inquiries and comments to: Dr. Isaac Gottlieb, Department of Bible, firstname.lastname@example.org
The Controversy over Saadiah Gaon’s Interpretation of Genesis 2:6
Dr. Yair Haas
HaKeter Mikraot Gedolot
(4) Such is the story of heaven and earth when they were created. When the Lord G-d made earth and heaven – (5) when no shrub of the field was yet on earth and no grasses of the field had yet sprouted, because the Lord G-d had not [Heb. lo] sent rain upon the earth and there was no [Heb. ayin] man to till the soil, (6) but a flow would well up from the ground and water the whole surface of the earth – …
Saadiah Gaon (882-942 C.E.) translated Genesis 2:6 in a most surprising way: “Neither was there a flow welling up from the ground to water its surface.”  This reversal of the usual meaning was elegantly explained by Radak (1160-1235), in his commentary on this verse: “The Gaon, Rabbi Saadiah, interpreted the word “flow” as “neither was there a flow” – for the negative particles lo and ayin, previously mentioned, can serve doubly; thus he interpreted that when the world was first made their was neither man to sow and plant the soil, nor a flow that would well up and water it. In Saadiah Gaon’s view, the word ayin (= there was no) of verse 5 also applies to verse 6, in order to explain why “no shrub of the field was yet on earth…” In his biblical lexicon,  Radak repeated this explanation, adding: “Rather, the Holy One, blessed be He, created by fiat everything to perfection and this interpretation is the best of all possible.” Radak identified with Saadiah Gaon’s approach, adding a theological reason, namely, that the Holy One, blessed be He, created a world which was perfect from the moment of its creation and that the creation of vegetation was not dependent on a flow welling up from the ground, rather it was a separate and independent act of the Holy One, blessed be He.
Saadiah Gaon’s view was also mentioned by Abarbanel, although he flatly rejected it:
The Gaon, Rabbi Saadiah … interpreted the words, “a flow would well up from the ground,” as neither was there a flow, as if to say the negative statement that “there was no man” carries over to the next phrase, as if it had been written: there was no man, and there was no flow welling up from the earth. But the plain sense of Scripture does not tolerate such an interpretation.” 
Abarbanel did not take the trouble to explain why Saadiah Gaon’s interpretation is not possible within the rules of biblical language and style, perhaps because his opposition to Saadiah Gaon’s understanding stemmed from a matter of principle which made it inconceivable that this be what Scripture intended.
In our opinion, the controversy over Saadiah Gaon’s interpretation of Genesis 2:6 can be explained in the context of the medieval debate about whether one can have an ellipsis in which the negative particle has been omitted. The very first generation of Spanish exegetes made interesting comments on this issue. In his biblical lexicon,  the Spanish grammarian Menahem ben Saruk (900-970) explained such verses as Job 32:9 – “It is not the aged who are wise, the elders, who understand how to judge” – according to the principle that “one negative particle serves doubly” (neither are the aged wise, nor do the elders understand how to judge). However, aside from citing several binary verses consisting of two short clauses, Menahem ben Saruk also cited the verse, “Do not eat any of it raw, or cooked in any way with water” (Ex. 12:9) as meaning: “Do not eat any of the paschal offering raw, nor cooked in any way with water.” Since this verse contains but a single clause, this explanation evoked a sharp response from Dunash ben Labrat (920-990), Saruk's constant critic:
You wrote in your book: there are instances of a word serving doubly, such as, “It is not the aged who are wise, the elders, who understand how to judge,” which you claim should have been written as: it is not the elders who understand how to judge. In truth, I agree, for what you wrote was correct. But, when you went on to include such instances as “Do not eat any of it raw, or cooked in any way with water,” you destroyed your argument, for there is no need to add a negative particle.” 
Dunash ben Labrat agrees with Menahem ben Saruk, viewing these verses as instances of elliptic texts omitting the negative particle. However he did not consider the verse from Exodus as relevant in this regard, since it contains a single clause and therefore the word “not” at the beginning of the verse naturally applies to the entire clause. It seems than Menahem ben Saruk was actually drawing a parallel between verses with two clauses and the verse in Exodus in order to clearly illustrate what he had in mind when he said that "a single ‘no’ can serve doubly.” Just as it is clearly true in the case of a single clause, so, too, in binary verses each clause relates to the negative particle at the beginning, and there is no omission. Dunash ben Labrat held that the negative particle at the beginning of the verse should also be added to the second part of the verse in order that it be understood correctly.
Most medieval commentators were of the opinion that the Bible contains instances of a negative particle being omitted, and as far as they were concerned the controversy between Menahem ben Saruk and Dunash ben Labrat had no exegetical implications. For an exegete like Ibn Ezra, however, who flatly rejected the possibility of an ellipsis which omitted the negative particle, the approach taken by Menahem had important implications. In his commentary on Deuteronomy 20:19, Ibn Ezra stated that “any word may be omitted in an ellipsis … save for the word ‘no/not,’ which cannot conceivably be omitted, for the meaning would be the opposite.” This, however, did not prevent Ibn Ezra from reading the verse in Job 32 (and other verses) similarly to Menahem ben Saruk: “‘It is not the aged who are wise’ – [the not] serves for this phrase and the next, thus: and it is not the elders who understand.” Without doubt Ibn Ezra was relying on Menahem ben Saruk’s approach, not Dunash ben Labrat’s, and surely he did not take the rule, “The first ‘no’ serves doubly,” to mean that the clause was lacking a word, i.e., that this was an ellipsis.
Abarbanel tended to minimize the number of instances of ellipsis in Scripture. For example, regarding the words of the chief cupbearer to Pharaoh, “And as he interpreted for us so it came to pass: I was restored to my post, and the other was impaled” (Gen. 41:13), Rashi commented: “Restored to my post [Lit. “Me, he restored to my post”] – by the person in whose power it was to do so, namely Pharaoh … This is an ellipsis, for it does not spell out who restored him, since there is no need to spell this out.” Abarbanel, however, commented: “Joseph, who is mentioned [in the preceding verse], restored me to my post through his solution of my dream, for he said that I would be restored to my post.” Abarbanel preferred to take a strained interpretation than say that the subject [who restored me?] had been omitted from the verse.
In view of Abarbanel’s care not to rely on ellipsis as a legitimate exegetical tool, it would not be unreasonable to assume that he, too, like Ibn Ezra, ruled out the possibility of Scripture omitting a negative particle and thus confusing the reader. Like Menahem ben Saruk and Ibn Ezra, so too Abarbanel was of the opinion that the verses in which “a single negative particle serves doubly” are not instances of ellipsis.  However, returning to our original verse, it is difficult to apply this rule to Genesis 2:6. First, because we do not have symmetric clauses with similar meters in each hemistich, in which case the negative particle must be carried over in order not to have an internal imbalance in the verse. Second, because in Abarbanel’s opinion the division of the Torah into verses reflects an exegetically binding tradition.  Hence, in Abarbanel’s view, carrying over the negative particle from one verse to the next (Gen.2:5-6) and treating it as an ellipsis of the negative particle is impermissible. Radak, in contrast, followed the approach of Dunash ben Labrat, as we see from the way in which he characterized verses such as Job 32:9: “Omission of words based on recalling the contiguous word.”  Thus we see that for Radak it was admissible to have elliptical expressions that omitted the negative particle.
 Kapah ed., p. 19.
 Sefer ha-Shorashim, root a-y-d.
 Commentary on Genesis 2, question 4.
 Ha-Mahberet, root h-y.
 Resp. Dunash ben Labrat, Philipovsky ed., p. 48.
 See his commentary on Num. 23:19, Isa. 10:26, 23:4, 42:8, Ezek. 16:47.
 Cf. Nahalat Avot, Zilberman ed., p. 30.
 Sefer ha-Mikhlol, Rittenberg ed., p. 51.