NUMBERS RABBAH, aggadic Midrash to the Book of Numbers, also called Va-Yedabber Rabbah in medieval literature. (For the name "Rabbah" see *Ruth Rabbah.)
The book is divided into 23 sections. The Midrash on chapters 1–8 of the book of Numbers, which are the first two weekly portions as read today – Ba-Midbar and Naso – is two and a half times longer than the remaining Midrash on chapters 9–36, which cover the eight remaining portions. This disproportion
Numbers Rabbah I
This appears at first sight to be an exegetical Midrash, since (with certain omissions) it forms a kind of consecutive interpretation to Numbers 1–8, chapter by chapter and verse by verse. Nevertheless, many of its long expositions deal with one single theme and are typical of homiletic Midrashim. The division into sections is at times determined by the open and closed sections of the Torah (see *Masorah) and at times by the weekly division of the reading of the law according to the triennial cycle once customary in Ereẓ Israel. In general each section begins with an anonymous proem, either an imitation (not always successful) of the classical proem typical of the amoraic Midrashim (see *Midrash; *Homiletics), or of the type combining halakhah and aggadah common in the Tanḥuma Yelammedenu Midrashim. Some sections have epilogues of consolation or of future destiny. The language of the Midrash is Hebrew, in part mishnaic and in part of the early medieval period. It contains a little Galilean and also Babylonian Aramaic and a few Greek words.
In the light of the many parallels between Numbers Rabbah I on the one hand, and *Genesis Rabbati and Midrash Aggadah (see Smaller *Midrashim) which are of the school of Moses ha-Darshan, the 11th-century scholar of Narbonne, on the other, it seems that Numbers Rabbah I is also based on Moses ha-Darshan's Midrash to the Pentateuch, of which it preserved not only the contents but even the terminology. This conclusion follows also from the fact that quotations by medieval scholars from the work of Moses ha-Darshan are found in Numbers Rabbah I. Since, however, the parallel part of Midrash Aggadah (to Num. 1–8) contains many homilies not in Numbers Rabbah I, it is obvious that Numbers Rabbah I is not an actual part of the work of Moses ha-Darshan, but his book served as the main source for its editor and compiler. The basis of Numbers Rabbah I was a Midrash of the Tanḥuma Yelammedenu type (which is the reason for the many parallels to these Midrashim and for the homilies which mix halakhah with aggadah), but the late compiler broke down and reconstructed its homilies, changing its character by greatly enlarging it (particularly in the case of the homilies to Naso, sections 6–14, which themselves constitute four sevenths of the whole Midrash Numbers Rabbah), adding to it from various sources, especially from the work of Moses ha-Darshan. That work was a combination of biblical commentary, aggadot and homilies, and halakhic topics, and included old and new sources (the greater part of which had been revised), together with original novellae. Among the works utilized by Moses ha-Darshan were the *Apocrypha and *Pseudepigrapha of the Second Temple period, especially those of the *Enoch circle (see also *Jubilees; the *Testaments of the Twelve Patriarchs), of which he seems still to have had Hebrew versions. He used all the tannaitic literature, the Jerusalem Talmud, the early amoraic Midrashim, the Tanḥuma Yelammedenu, Midrashim (including *Pesikta Rabbati), *Seder Eliyahu Rabbah and Seder Eliyahu Zuta, the Babylonian Talmud, and even late Midrashim like the Midrash Tadshe; his work also contains pseudepigraphic material. Numbers Rabbah I also makes use of the piyyutim of *Kallir and of Sefer *Yeẓirah, and contains topics of esoteric lore, mysticism, and combinations of numbers and calculations. Hence its comparatively late Hebrew is understandable. As the compiler was apparently acquainted with the Midrash *Lekaḥ *Tov, which like the work of Moses ha-Darshan dates from the end of the 11th century, the middle of the 12th century seems to be indicated as the earliest possible date for the compiling of Numbers Rabbah I. It is of interest that the Paris manuscript (no. 149) of 1291 only includes sections 1–5 of Numbers Rabbah (on the reading of the law for Numbers), while the Munich manuscript (97, 2) of 1418 includes the whole of Numbers Rabbah I but not Numbers Rabbah II.
Numbers Rabbah II
This homiletical Midrash of the Tanḥuma Yelammedenu type is identical in all respects with the part parallel to it in the printed Tanḥuma and in Buber's edition of the Tanḥuma. Moreover a much better version has at times been preserved in Numbers Rabbah II than in the parallel passages of both the above-mentioned Tanḥuma Midrashim. Thus, for example, it has not the Babylonian She'ilta, added to the Tanḥuma Midrashim Ḥukkat, 2. Instead of the expression Yelammedenu Rabbenu ("teach us, our master") found in the Tanḥuma, it has halakhah (in the manuscripts, however, Numbers Rabbah II too has Yelammedenu Rabbenu). Many of the halakhic proems found in the Tanḥuma have been abridged in Numbers Rabbah II, as for example the first. However, they are found in full in the manuscripts. The division of Numbers Rabbah II into sections, not found in the manuscripts, is, as in the Tanḥuma, almost identical with the division of the triennial cycle. The view accepted by the majority of critical scholars is that Numbers Rabbah II, which is apparently the second half of a complete Midrash whose first half, which served as the original basis, was lost, was compiled in the ninth century, like most of the Tanḥuma Yelammedenu Midrashim. It has, however, also some late additional interpolations from the book of Moses ha-Darshan (18:15–18; 20:5–6, lacking in the Tanḥuma; and 18:29 found also in the printed Tanḥuma).
The union of Numbers Midrashim Rabbah I and Rabbah II is the work of a copyist of the beginning of the 13th century. The complete Midrash was not yet known to the author of the Yalkut Shimoni; it seems that the first to cite it was Naḥmanides. The earliest manuscripts of the whole of Numbers Rabbah date only from the 15th century, but they are nevertheless much better than the printed versions.
Zunz-Albeck, Derashot, 125–7, 397–400.