Jump to ContentJump to Main Navigation
Biblical Myth and Rabbinic Mythmaking$

Michael Fishbane

Print publication date: 2003

Print ISBN-13: 9780198267331

Published to British Academy Scholarship Online: January 2005

DOI: 10.1093/0198267339.001.0001

Show Summary Details

(p.325) APPENDIX 2

(p.325) APPENDIX 2

Biblical Myth and Rabbinic Mythmaking
Oxford University Press

(p.325) APPENDIX 2

The Term kivyakhol and its Uses

Appendix 2 The Term kivyakhol and its Uses

This appendix1 on the term kivyakhol (⩵ kibyakhol)2 draws upon my collection over many years of the available examples to be found in classical and related rabbinic sources: in published and critical editions; in published and manuscript variants; published and manuscript fragments from the Cairo Geniza and other library resources; and in data-base files in electronic form and data collected for historical research, such as the word files of the Israel Academy of Language. Several photocopies of unpublished manuscript fragments were made available to me through the courtesy of Prof. M. Bregman.

I have utilized all this material (as pertinent) in the categorization, analysis, and evaluation of ancient rabbinic and early medieval texts in which the term kivyakhol occurs. Almost universally, the term is found in the context of theological and homiletical teachings, and evincing a strong anthropomorphic and anthropopathic character in the framework of midrashic exegesis found in the great midrashic and Talmudic compendia (or extracts derived therefrom). Several examples are found in old legal sources (like the Mishnah and Tosefta), but the issues remain of a theological or homiletical sort and are exegetical in character. The organization and interpretation of the data is intended to bring some conceptual and thematic clarity to a sea of citations, and to elucidate opaque or otherwise elusive examples. Overall, this material supplements the examples of rabbinic mythic theology discussed in Part II. Hence, cases considered at length in the body of this book are only treated to brief annotations in this appendix. On the other hand, topics and texts that either extend or elaborate the categories and themes taken up earlier, are dealt with more fully here. In all instances, close attention has been given to the placement and apparent function(s) of the term kivyakhol. Considerable attention will be given to the techniques and character of the exegetical theology involved.

Because of inconsistencies in parallel versions, and the ongoing impact of scribal practices and pieties into the Middle Ages (long after the primary production of the exegeses themselves), it is not always possible to determine the authentic exempla of the term kivyakhol or estimate the precise stratification of the evidence. Hence, I have attempted to register all pertinent variations, and to highlight the function(s) of the term in the context of the hermeneutical and literary aspects of the passages. Similarly, the various difficulties and ambiguities involved in evaluating these midrashic teachings have been acknowledged and spelled out, to the extent that this is feasible or instructive. Thus I have sought to make the complexity of analysis and evalution as transparent as possible. Nevertheless, a number of practical determinations had to be made. For example, there are a large number of examples in the late midrashic collections (like Midrash Tanḥuma, Midrash Shoḥer Ṭov, and Aggadat Bereishit), (p.326) where the term kivyakhol is often used in ways inconsistent with the patterns isolated and evaluated below. I have judged many of these instances to be the result of later scribal whims, introducing new euphemistic attitudes into the data, and have collected the majority of them at the end with no detailed analysis. Only when these cases preserve arguably authentic or characteristic usages (no matter how weak), or provide instructive examples (no matter how tenuous), some analysis or reference has been provided in the main body of this appendix. Similarly, parallels from the late compendium Yalkut Shim͑oni are not given̶save for those special cases where notable variations or unique (or uniquely formulated) traditions have been preserved.

The following appendix is composed of three main parts: (1) an overview of the main interpretations of the term kivyakhol, and a statement of the considerations that will guide this inquiry; (2) a presentation and analysis of the textual evidence where the term kivyakhol occurs, with particular attention to its perceived functions; and (3) a final evaluation of the material assembled, with brief conclusions.

The Term

The following is a conspectus of explanations of the term kivyakhol, formulated by influential medieval and modern authorities.

1. Rashi (1040̶1105) gives four distinct comments on the term. In BT Yoma 3b, he says that ‘it is said against our wishes, as if it were possible to say so; and similarly every (case of) kivyakhol in the Talmud’. Thus in his view, the term serves as a euphemistic or apologetic qualification of certain remarks about divinity. More specifically, in BT Meg. 21a Rashi notes that ‘it is said of the Holy One, blessed be He (when one speaks of Him) as like (ke-) a person, of whom it is possible (yakhol) to speak thus’. This comment is similar to his point in BT Sanh. 97a. In BTḤag. 13b Rashi also notes the use of the term with regard to the angelic host.

2. R. Judah Halevi (c.1075̶1141) in the Kuzari, Book III (73), treats the term kivyakhol as marking a figure of speech about God, employed to strengthen or enjoin a theological point. Indeed, in his understanding, the term highlights a putative assertion that means: ‘if it were possible for the matter to be thus and so, it would then be so’.3 Thus Halevi regards the term kivyakhol as underscoring certain anthropomorphic features of midrashic rhetoric.4

3. R. Yom Ṭov b. Abraham Ishbili (Riṭba; c.1250̶1330), in his annotations to BT Yoma 3b, adduces Rashi's above-cited opinion, and adds: ‘But because it is written in Scripture (namely, the words that justify R. Jonathan's midrashic inference regarding divine preference for an individual offering) it is possible (yakhol) to say it.’5 This (p.327) statement appears to be more than a gloss that justifies the comment, and may also be construed as an explanation of the term kivyakhol, taken to mean something like ‘it is possible (to state the midrash)’ since it has scriptural warrant or support. The term is thus a hermeneutical marker, not a euphemistic apology (Rashi).

R. Yeshu͑a b. Joseph of Tlemcen (15th c.) in his Halikhot ὉOlam (2.1) follows this tack, but adopts a more constrained position in his gloss on the term.6 Basing himself on a Tosafist comment on BT Meg. 21a not found in our printed editions, which states that ‘it (the term kivyakhol) is often used with regard to theological teachings (haggadot debarim) that are not respectful (derekh kabod) of God (lit. ma‘alah),‘ R. Yeshu‘a drew the conclusion that the initial letters (kv; i.e. kaf + bet) of the term (which are numerically ‘22’) are to be distinguished from the verb yakhol (‘it is possible’). The result in his view is a hybrid or compound form, with kv functioning as a ‘notarikon’ (or abbreviation) that denotes the number of letters in the alphabet.7 Accordingly, k(i)v + yakhol is taken to be a composite construction that paraphrastically indicates that ‘it is possible for the Torah which was given through the twenty-two letters (of the alphabet) to speak so, but it is impossible for us to say it’. This explanation thus interprets the term kivyakhol as cautioning human exegetical restraint, while simultaneously acknowledging the scriptural basis for the comment.8 However, R. Elijah Levitas (1469̶1549) in his rabbinic dictionary Sefer Ha-Tishbi rebutted this explanation, and suggested that the term indicates that Scripture speaks about God ‘as by (kemo) a possibility (be-yakhol), specifically “as for one who is able (yakhol) to accept this teaching.”’9 This explanation treats the term as marking a rhetorical possibility for the individual—and not a feature of the comment per se.

4. Similar euphemistic and rhetorical explanations of the term are highlighted by many modern interpreters.

N. Bru¨ll understood kivyakhol as a term used to mark theological teachings in which something common or worldly is used (‘was gewo¨hnlich geschieht, in dem was in der Welt gebra¨uchlich ist’). In his view, the meaning of the term is literally ‘as in the possibility’ (‘wie in der Mo¨glichkeit’), which he further explains as ‘als ob es mo¨glich wa¨re’.10 This view basically follows the older opinion of A. Geiger, who apparently understood the term as marking a rhetorical construction, since he paraphrases (p.328) kivyakhol with the words ‘als spra¨che man von Einem, bei dem so etwas mo¨glich wa¨re’11 (though the phrase ‘von Einem’ may hint at a euphemistic component, insofar as the dominant subject of these interpretations is God and His anthropomorphic and anthropopathic nature).

W. Bacher gives an explicit euphemistic explanation of the term, when he refers to it as ‘a fixed expression used where there is a daring teaching about God, functioning to beg forgiveness (for the formulation)’.12 He reinforces this view by a construal of kivyakhol as a ‘notarikon’ for k(e͗ilu ne͗emar) b(e-mi she-) yakhol (㕲attah lomar zeh), ‘as if it were said concerning one of whom you could say this’13 (cf. Geiger). This rendition provides a corollary explanation, and is essentially taken over by A. Kohut.14 Bacher's initial characterization (following Rashi) is reaffirmed by E. E. Urbach.15

5. Finally, A. Marmorstein parts company from the tradition that interprets kivyakhol as some kind of terminological disclaimer, and gives a new emphasis to the scriptural basis or legitimacy of the reading at hand. In his view, the term kivyakhol ‘does not stress the idea or weaken the grossness of the conception, but indicates the scriptural basis of the exegetical teaching or homiletical truth’—since ‘the term is either preceded or followed by a scriptural reference’.16 He reinforces this positive view of the word by the proposal of another ‘notarikon’ of k-b-y-kh-v-l: k(a-yotzei) b(a-dabar) y(esh) k(oaḥ) v(e-͗efshar) l(omar).17 This proposal is a compound solecism which, at best, can be rendered ‘(and) similarly, there is foundation (for the view) and one can adduce (Scripture in support)’. Despite the oddity of this explanation, Marmorstein's observation that the term kivyakhol is regularly found with a scriptural citation, and that Scripture has a positive function in such cases, is an important addition to the discussion and deserves further investigation.

From the foregoing explanations of the term kivyakhol, two considerations stand out. (1) The term is understood as qualifying the bold content of certain rabbinic teachings. This content characteristically deals with strong mythic portrayals of divine actions or emotions. In these cases, the term is said to inject a qualification or a hesitation before such material. (2) The term is understood as indexing the scriptural basis of the teaching. Such references appear as direct citations. In these cases, the term is said to indicate that the content may be condoned because it is grounded in Scripture itself. The first consideration thus refers to the theology itself; the second to its textual authority.

In effect, the two types of explanation highlight the two typical literary features of teachings where the term kivyakhol is found. In some cases, it occurs with the theological teaching (leaving the scriptural proof-text unqualified); whereas in others (p.329) it occurs with the scriptural proof-text (leaving the theological teaching unqualified). The reason for the preference for the one slot or the other is not always evident or certain, and often appears to be totally arbitrary. Presumably, such single qualifications were deemed sufficient for the teachers or tradents (including copyists) involved, even though the pertinent materials have the double feature of a theological teaching plus a proof-text.

One is contrained to ask: What does it mean to qualify the teaching, but not the scriptural support; and, conversely, to qualify the citation, but not the teaching itself? The matter is exacerbated by the fact that the qualification of the content often appears to undermine the rhetorical effect of the teaching from the outset, or even qualify a normative feature of rabbinic theology (like divine mercy or dominion). Similarly, the qualification of a scriptural citation must also be explained, especially as that affects the divine authority involved.

The complexity of the evidence does not always or easily allow us to get back to the original formulations of the teachings involved or to determine the meaning of the term kivyakhol for the teachers or tradents themselves. Undoubtedly, the term served various functions for various persons at various times; and some of these could even be totally disruptive additions to received materials, and devoid of any consideration as to the effect these interventions must have upon the meaning of the passages involved. Where stylistic or text-critical or manuscript evidence permits evaluations and judgements, such have been proposed. But overall it has seemed beneficial to focus primarily on the hermeneutical function of the term kivyakhol in its two contexts—a decision that yields the following additional perspective on the qualifications involved. Where the term occurs with the theological content, it qualifies the latter by marking the presumptive character of the teaching derived from Scripture (duly reinterpreted); and where the term occurs with the proof-text, it qualifies the latter by marking the presumptive reuse of the citation for the construction of the theology just proclaimed.

Examples for these assertions shall be found below. At this point it may suffice to add that on this perspective both the theological content and the proof-text are qualified in ways that help us attend to the positive constructions involved, as against negative constrictions placed upon the teachings and texts per se. Moreover, the chief value of this perspective lies in its heuristic potential for elucidating the dynamics and dimensions of the passages themselves—not its capacity to construct original intent. Withal, in some cases this viewpoint may also provide the primary intent of a given usage (particularly when the term qualifies the scriptural citation). But such determinations are not crucial, and remain secondary to the main purpose of this appendix—which is to elucidate the remarkable mythic content and exegetical theologies of rabbinic passages where the term kivyakhol is found.

Lexicographically, the term kivyakhol may be considered a counterpart to the word yakhol, regularly used in old midrashic discourse (especially legal) to propose an inference or possibility derived from Scripture—but one that is peremptorily rejected on the basis of another construal of the words of the text. In these cases, one exegesis trumps another. For its part, the term kivyakhol presents a certain theological teaching in aggadic or homiletical discourse ‘as a possibility’ (or ‘as by a possibility’), which then proceeds to support that proposal by a scriptural citation—whose relevance (p.330) must be midrashically construed.18 As we have suggested, none of this requires that the term kivyakhol is original in any given case. It only stresses that the term serves as a way of marking the hermeneutical and constitutive character of the midrashic discourses involved, as well as their presumptive boldness.

With these considerations in mind we turn to the midrashic evidence itself, arranged by subject and theme.


A. Divine Supremacy

1. Dominion.

(i) MidPs 114.3 (B, 471).19 A comment on be-YaH shemo (literally, a call to extol the Lord ‘whose name is YaH’, or ‘by His name, YaH’).

R. Judah the Prince asked R. Samuel bar Naḥman: What does be-YaH shemo mean? R. Samuel bar Naḥman answered: Every place has a superior designated over its biyya͗ (dominion);20 and who is designated biyya͗ of the world?—kivyakhol the Holy One, blessed be He; as it says, be-YaH shemo! Do not read be-YaH but biyya' shemo (i.e. ‘His Name is Dominion’; that is, ‘Lord’ or ‘Master’).

The chronological sequence of Judah the Prince (ha-nasi͗) (T4) asking a query of R. Samuel (A3) is impossible; hence one must prefer the version of this teaching found in GenR 12.10 (T–A, 108), where R. Yudan Nesiy͗ah III (A3) poses the query. R. Samuel's comment presupposes knowledge of Greek bia (dominion); and in it the term kiv. does not qualify the reference to the Holy One as the Lord of the world, but rather marks the reference to Him through the ensuing interpretative gloss on the divine Name. Notably, the term kiv. does not occur in either GenR 12.10 or JT Ḥag. 2.1, 7c, where the reference is to the biyah of the world. The bold transformation of the divine Name through the use of the ͗al tiqre (do not read) exegetical form may account for the occurrence of kiv. in MidPs 114.3.21

2. Duration.

(i) Tan Tazri͑a 2. On the verse ‘There is no holy one like YHWH, truly, there is no one beside You’ (1 Sam. 2: 2), the query is asked: (p.331)

What does ‘Truly, there is no one beside you (biltekha)’ mean?—A king of flesh and blood builds palaces (and) they outlast (meballim) him, but the Holy One, blessed be He, outlasts (meballeh) His world—kivyakhol, ‘Truly, there is none beside You (biltekha).’

This is a highly condensed teaching, utilizing the common trope which juxtaposes a human king to God in order to make a theological point. But what is it? The theology of supremacy or exclusivity is plain enough (though it is unclear whether this verse from Hannah's prayer is giving hyperbolic praise but not denying other divinities, or is making a categorical assertion of uniqueness). The comparison refers to duration (royal and divine), using a verb that alludes to biltekha in the lemma—thus indicating where the point of emphasis lies. But after the comparison, what is the force of the term kiv., followed by the lemma? Surely, the purpose of kiv. is neither to deny the previous theologoumenon (about God's everlastingness) nor the ensuing proof-text. However, the sense of this proof in relation to the opening query and the explanation is obscure.

The solution lies in perceiving that the midrashist presumes a construal of biltekha as meaning ballotekha, so that the final phrase asserts of God that ‘there is nothing that outlasts You’. That this is the intended hermeneutic is clear from the parallel version in TanB Tazri͑a 3 (B, 33). After the assertion of kiv. and the proof-text, there follows the epexegetical comment: ‘Truly, there is nothing that outlasts You (ballotekha).’ Similarly, in a variant formulation in BT Meg. 14a, the exegetical technique is explicit: ‘Do not read biltekha but ballotekha’—and then goes on to compare God to humans, the work of His hands. In this case, kiv. is unnecessary and not used; in Tan Tazri͑a (and Ca), the term marks the presumptive and implied rereading of the lemma.

In BT Meg. 14a, the teacher is given as R. Judah b. Menashia. This is probably a copyist's error for R. Judah Nesia (A1 or A3—assuming an original reading like R. Judah mi-Nishia; namely, ‘from N.’).

B. Divine Combat

1. Against Sea.

(i) PesR Addendum 1.1 (MIS, 183a). A tradition about the vanquishing of (the) Sea (Yam) at the time of creation. The rabbis give a brief allusion to this event through an interpretation of Ps. 93: 3, ‘The seas arose, O YHWH…the seas arose dokhyam.’ They explained:

‘The seas arose’…What is dokhyam? darakh yam (i.e. ‘He (God) trod over Yam’). (At the beginning) the waters of creation filled the entire world. What did the Holy One, blessed be He, do? He pressed (kabash) upon them—kivyakhol, and caused them to be gathered into ‘Ocean’.

This is a mythic account of events presumed to have occurred at the gathering of the waters at the beginning of the world (stated neutrally and without mythic drama in Gen. 1: 9). It depicts a scene of upswelling and flowing waters, but notably no rebellion is indicated. In order to contain Sea, God pressed the waters into fixed domains. Scriptural support is found in the word dokhyam, which is presumed to be a contraction of the verbal element dokh plus the noun yam. By this means, the word dokhyam was midrashically interpreted to yield the mythic event narrated briefly. (p.332) Presumably, the verbal element dokh was construed as if from the verbal stem d-kh-kh, with the sense of suppression, and this meaning was explicated and rendered by the verb darakh (tread upon). Choice of this latter verb indicates the influence of mythic motifs in Scripture (cf. Job 9: 8), as well as other midrashic traditions concerning this mythic event which employ it (see above, pp. 116–17). For its part, the verb kabash also retains a trace of an old mythic battle or theomachy (see above, pp. 117–18). However, in this setting, the term kiv. signals it as an explication of the terse midrashic exegesis at the outset. But this second account is weaker, mythically speaking; the initial aggression against Yam being rendered here more impersonally.


A. Human Creation

1. Angelic Involvement.

(i) GenR 12.1 (T–A, 97–9; Lo). In the context of Gen. 2: 4, R. Ḥuna explicates Job 26: 14 to indicate the limited nature of human knowledge of God's ways. In the rhetorical crescendo he says that even the nature of the thunder is beyond man, ‘how much the more so the nature (or order) of the world itself (seder ͑olam)!’ To support this point, Eccles. 2: 12b is adduced as double proof (‘for what can the man do who comes after the king?—even that which has already been done’). The first phrase was glossed by the comment, ‘the King of the world; the Supreme King of kings, the Holy One, blessed be He’, which suggests that the lemma was construed to mean something like, ‘for what is man, since he comes (into existence) after God?’; or, ‘that he might come to know (the ways of) God?’ Several parables follow; and then the second phrase is cited, with the following comment:

‘Even that which has already been done (͑asuhu).’ It is not written here ͑shw (namely, ͑asahu, ‘he made him’) but ͑swhw (namely, ͑asuhu, ‘they made him’)—kivyakhol, the Supreme King of kings, (the Holy One),22 blessed be He, and His court (beit dino), who are appointed23 over every one of your limbs, and set you upon your created form (tikkunkha); ‘He made you and established you (hu͗ ͑asekha va-yekhonenekha)’ (Deut. 32: 6).

(ii) EccR 2.12. A variant of the preceding teaching. The final part reads:

‘and set you upon your completed form (tiqqunkha)’. And if you should say that there are two powers, Scripture already says, ‘He made you (͑asekha) and established you’ (Deut. 32: 6).

The comment on Eccles. 2: 12 emphasizes the orthography of the verb ͑swhw, whose plural form leads to the conclusion of a conjoint creation of man (namely, by God and His court). This is the primary and original purpose of the comment. The subsequent use of a verb in the singular form, in the statement about the establishment of man in the proof-text (Deut. 32: 6), ostensibly contradicts the initial exegetical point. This new emphasis may have entered in the course of the text's reception history, and the need to promote another theological sensibility. The version found in EccR (p.333) exemplifies this process, by raising the new possibility that the plural verb ͑asuhu indicates two divine powers involved in man's creation—this being a dualistic danger or heresy.24 Such a reading is immediately countered by citing Deut. 32: 6, with its clear use of the verb ‘to make’ in the singular. The technique of rebuttal employed here was one of several means of countering problematic passages of this sort;25 but this whole matter is clearly secondary to the original midrash, whose whole point is to prove a conjoint creation in valid theological terms.26

The term kivyakhol marks this theological teaching of man created by God and the angels—whose proof rests on the orthography used (‘it is not written x but y—); and since the whole point of the midrash is to demonstrate this fact from Scripture, it seems that a principle function of the term is to highlight the relationship between the proof-text and the mythologoumenon—indicating that this teaching is only imputed, but not directly stated in Scripture.

In GenR, the tradent is R. Ḥuna (T5); in EccR, it is R. Simeon (ben Pazzi; A3)—a frequent carrier of kiv. traditions.

B. Other Involvements

1. The Heavenly Court.

(i) ExodR 4.3 (ShR, 148). Citing Jer. 10: 10, ‘Va-YHWH is a God of truth.’

R. Eleazar says: Wherever Va-YHWH is said—(this means) kivyakhol He and His court. And the paradigm case (binyan ͗ab) for all others is ‘Ve-YHWH spoke evil against you’ (1 Kings 22: 23).27

An exegetical norm is given by R. Eleazar (ben Pedat; A3). The term kiv. connects the scriptural term to its meaning, and marks the presumptive nature of the exegetical claim. It does not qualify the existence of a divine court, which is presupposed by the plain sense of the paradigm case (1 Kings 22: 23). That instance contextualizes the use of the letter vav with the divine Name, understood to portray God in a court session. The letter vav is thus not only a lexical sign indicating a conjunctive particle but, when followed by a divine Name, is also a midrashic sign implying the co-involvement of the heavenly court with God.

(p.334) For this type of exegesis by the same sage (called R. Lazar), see GenR 55.4 (T–A, 587), citing Gen. 22: 1, ‘Ve-ha-͗Elohim (And Elohim) tested Abraham.’ The hermeneutical principle thus includes the DN ‘Elohim’. T–A ad loc. follows most manuscripts against Lo, which gives ‘Ve-YHWH’ as the principle. For the latter, see also GenR 51.2 (T–A, 533) regarding Gen. 19: 24, which deals with the punishment of Sodom. These two cases are without kiv.

(ii) LevR 24.2 ͗Emor (MM, 552). This brings R. Lazar's principle with the term kiv. to explain Job 1: 21 (‘Ve-YHWH took’).

(iii) NumR 3.4 Bemidbar. Citing Num. 3: 11, ‘And I (ve-͗ani) indeed have taken the Levites,’ extends the use of R. Eleazar's principle:

Our sages say: Wherever it is said of Him, ve-͗ani—(this means:) kivyakhol the Holy One, blessed be He, and His court. And the paradigm case for all others is ‘Ve-YHWH spoke evil against you’ (1 Kings 22: 23).

As above, kiv. marks the presumptive nature of the exegetical assertion. The principle of vav plus DN is extended to a personal pronoun.

For the cases in ME Pet 34 (B, 38) and PdRK 13.9 Dibrei Yirmiyahu (BM, 232), where vav plus the pronoun ͗ani (I) and hu͗ (he) refers to God alone, see discussion below, IV.D.4(i). This is also the way R. Eleazar is reported to have understood ‘Ve-hu' is one’ in Job 23: 12 (cf. ExodR 3.2).

C. Size and Characteristics

1. Size.

(i). SifNum 42 Naso͗ (H, 47). In an apparent addition to earlier versions,28 we read at the beginning of a collection of contradictions and their resolutions:

One verse (katub ͗e͗ad) says, ‘He makes peace in His heights’ (Job 25: 2); and one verse says, ‘Is there any number to His troops?’ (ibid, v. 3); and one verse says, ‘Thousands upon thousands serve Him; myriads upon myriads attend Him’ (Dan. 7: 10). How can these passages (all) be maintained (yitqayyemu)? Before they (the Israelites) were exiled from their land, ‘Is there any number to His troops?’ (applies); after they were exiled from their land, ‘Thousands upon thousands served him’ (applies)—kivyakhol, the (size of the) heavenly retinue was reduced (nitma ͑a͑ah).

An anon. teaching in a tannaitic corpus. The central contradiction is between Job. 25: 3 and Dan. 7: 10, as is clear from the resolution. The contradiction between two passages is in the standard format of establishing a distinction—here, the ante quem and post quem of the exilic event.29 The term kiv. marks this exegetical explanation and the mythic solution. To assume otherwise is to regard the resolution as rhetorical and without any mythic validity (see also b, below).30 A second resolution attributed (p.335) to Rabbi (Judah the Prince) (T4) in the name of Yose ben Dosta͗i (T3) is more practical and less bold: each troop has ‘thousands upon thousands’, but there is ‘no number to His troops’.

(ii). BT Ḥag. 13b. A variant of SifNum 42, also anon., using the formula ‘one verse’.31 However, in this case the Hebrew resolution formula, ‘How can the passages be maintained,’ is not used; rather, the Aramaic phrase la͗ qashia͗ (‘no problem’) appears—thus indicating the Babylonian redaction of the EI tradition. The resolution given is:

One (namely, the first case) refers to when the Temple was standing; the other, to when the Temple was not standing32kivyakhol the (size of the) heavenly retinue was reduced.

The resolution here refers to the Temple; the one above, to the exile. The term kiv. is used as above. Rashi's comment, ‘Kivyakhol: even with regard to the host of the Shekhinah (!) we must speak thus, just as one (yakhol) may speak about diminishment with respect to humans’, presumes that we are dealing with a euphemistic qualification. But the issue is manifestly related to the contradiction in Scripture, which is hereby resolved. The solution is hermeneutical, and the term kiv. marks this point and the mythic assertion adduced. Otherwise, the solution would be mere rhetorical play (also above). A second resolution follows this one, as in SifNum (above), but now ‘in the name of Abba Yose ben Dos[t]a͗i’.33

(iii). PesR 21 ͑Aseret Ha-Dibrot (A) (MIS, 103a). In this version, the second resolution comes first, attributed to ‘Rabbi in the name of Abba ben Yosef’ (a corruption of Abba Yose (T3)?). The second tradition is reported in the name of ‘Rabbis’. The resolution formula used is ͗ela͗, ‘however’—

Before the Temple was destroyed, the praise (shibḥo) of the Holy One, blessed be He, was complete (mushlam); (but) after the Temple was destroyed—kivyakhol the Holy One, blessed be He, reduced its nature. How so? He said, ‘My House is (now) destroyed, so (should) My praise (qillusi) amount to a complete sum (͑oleh mushlam)?!’

The form of resolution in BT Ḥag. 13b is apocopated and narrativized, obscuring the explanatory force of kiv. Thus the need for a second explanation using a different word for praise. It is introduced by minalan (here ⩵ ‘how so?’),34 a Talmudic formula that normally requests the scriptural source of an assertion—rather than serving as a neutral query, as in this teaching. The change in the number of the heavenly retinue was due to God's decision to reduce their chorus of praise ‘after the Temple was destroyed’.

(iv). LevR 31.6 ͗Emor (MM, 723 f.). This version is apparently derived from a tradition like that found in PesR 21.35 This is evident from continuities, changes, (p.336) and abbreviations. The first resolution, attributed to ‘Rabbi’ (pleonastically) cites Job 25: 3, Dan. 7: 10, and Ps. 68: 18. The second resolution, attributed to ‘Rabbis’, begins with ͗ela͗. The first part is like 1c, above.36 The divine justification here reads:

He said: It is not right that My praise37 should amount to its former sum (͑oleh kemo she-hayah ͑oleh).

This final phrase somewhat implies the comment found in PesR 21. Both LevR 31.6 and PesR 21 explain the reduction of the divine retinue as due to God's decision to reduce His praise.38 The tone of the decision is more judgemental in LevR 31.6. A different explanation for the diminution of praise is in 2b, below.

2. Characteristics.

(i) BT Ḥag. 13b. An anon. tradition prior to that cited above in C. 1 (ii) presents another instance of two contradictory texts about the cherubim. In Isa. 6: 2, ‘Each one had six wings’; whereas in Ezek. 1: 6, ‘Each one had four wings.’ This difference is resolved:

The one (Isa. 6: 2) refers to when the Temple was standing; the other (Ezek. 1: 6) to when the Temple was not standing—kivyakhol, the wings of the cherubim were reduced.

As in 1b, the term kiv. introduces and characterizes the exegetical solution. Rashi's comment applies to this case, as well. The Talmudic tradition adds a query, ‘Which (wings) were taken away?’ Two opinions are given. R. Ḥananel reported in the name of Rab (BA, 1) that they were ‘those with which they utter song’; whereas ‘our rabbis’ said they were the ones that covered their feet. Proof-texts are provided. This elaboration suggests that the tannaitic resolution had mythic validity, and was not deemed mere rhetoric by the above-noted Babylonian Amoras.

(ii) LevR 31.6 ͗Emor (MM, 723 f.; M2). In this MS version, a striking synthesis of teachings is found. First, the contradiction regarding the number of angels (here in the name of R. Isi b. Dusta'i) is given and resolved with references to the exile, and explained (with kiv.) as due to the reduction of the heavenly retinue. Then a tradition about the wings is given, which presupposes but does not give the contradictory verses.

Our rabbis say: Before the Temple was destroyed, the ministering angels would utter praise (meqallesin) before the Holy One, blessed be He, with two wings; after the Temple was destroyed, the Holy One, blessed be He, said: My House is (now) destroyed, so (should) my Glory be praised?—kivyakhol the praise was reduced; as it is written, ‘The pastures of peace (shalom) are destroyed’ (Jer. 25: 37),39 (viz. reduced) from the two wings of the complete (shillum) praise.

Components of this version are influenced by a tradition like PesR 21, especially the reference to a complete (mushlam) praise and the divine query—though now the reduction of song is based on Jer. 25: 37, where the phrase ‘pastures of peace’ (na͗ot ha-shalom) is taken as a reference to the Temple of Jerusalem, whose name (p.337) midrashically encodes the ‘complete’ praise (shillum);40 and the verb ve-nadammu, ‘destroyed’, was midrashically construed as if it was ve-niddamu, ‘became silent’. This tradition is apparently influenced by BT Ḥag 13b, where the issue of divine praise is linked to the wings of the cherubim. However, the presentation of the subject is different; and the contradictory verses about the wings are not given. The term kiv. now pertains to a deft reinterpretation of Jer. 25:37.

D. Agency

1. The Power of Peace.

(i) SifNum 42 Naso͗ (H, 46). In a collection of teachings on peace, each introduced with the formula ‘great is peace’, we read:

R. Eleazar son of R. Eleazar ha-Qappar, says: Great is peace, for even if Israel worships idols but there is peace among them—kivyakhol the Omnipresent said, Satan does not harm them; as it is said, ‘Ephraim is bound to idols; leave him alone’ (ḥabur ͑atzabim ͗Ephraim hanaḥ lo; Hos. 4: 17). But when they are divided, what is said of them? ‘Their heart is divided, let them now be punished’ (Hos. 10: 2).41

The initial assertion and proof-text appear as a bound unit, with no reference to peace in the latter. R. Eleazar presumably parsed Hos. 4: 17 thus: ‘(If) he is connected (in fellowship), (then, even though) Ephraim has idols—leave him alone’; and he presumably also construed the imperative (‘leave him alone’) as God's command to a heavenly being. The term kiv. precedes this presumptive exegesis of the imperative and marks it as a special exhortation. Given the structure of the teaching, the statement ‘Let them now be punished’ is similarly a directive to the divine agent. So also in NumR 11.7. In both cases, the tradent is R. Eleazar b. R. Eleazar ha-Qappar (T5).

In Derekh ͗Eretz Zuṭa 9.2 (DS, 53 f.), the divine instruction is obscured, as God says to himself, ‘It is not my will to harm them.’ Cf. similarly in GenR 38.6 (T–A 355 f.; Lo), ‘I shall not have power over them.’42 Finally, there is no divine word in TanB Tzav 10; ḥabur is glossed as ‘make a fellowship (ḥaburah)’; the agent is the attribute of justice. The term kiv. does not occur in these cases.


A. Abraham

1. Circumcision.

(i) TanB Lekh 24 (B, 80). When Abraham began to circumcise himself, he started to shake and said, ‘I am old.’

What did the Holy One, blessed be He, do?—kivyakhol He sent forth His hand and held (it) with him (͗a͗az ͑immo), and Abraham cut until he was circumcised. Thus did Ezra give praise, saying: ‘You alone are the Lord…You chose Abram…and made a covenant with him (ve-karat ͗immo berit)…’ (Neh. 9: 6–8). It is not written, ‘made a covenant for him (lo)’, but ‘…with him (͑immo)’.

(p.338) Remarkably, God is portrayed here as participating in Abraham's circumcision. The term kiv. precedes this mythologoumenon, which is exegetically established by a hyper-specification of the preposition used in Neh. 9: 8. The technical formula, ‘It is not written x but y’, underscores the hermeneutical character of the teaching. A variant occurs in AgBer 16 (B, 34 f.), also with kiv.43

Variants of the formulation in (i) occur in ed. prin. (Venice, 1545) and MSS of GenR 49.2—but without the term kiv. Note inter alia MH1, Ox, Ox4, St and Y ( >  Lo; see T–A, 498). The ͑Arugat ha-Bosem cites a similar GenR tradition;44 and Hadassi in ͗Eshkol ha-Kopher 36 adduces this exegesis in the name of GenR and Tan, with the variant that God ‘seized (tafas) his hand and sustained him (heḥeziqo), and the two of them’ performed the rite. Cf. also TanB Vayera͗ 4 B, 86), where Abraham says, ‘“Your right hand sustained me” (Ps. 18: 36)—when You held (it) with me (͗a͗azta ͑immi) the foreskin, and I cut’ (cf. AgBer 16 (B, 35) ). Unique is the citation of GenR regarding ‘This is My covenant’ (Gen. 17: 10) brought by Sefer Milḥamot ha-Shem: ‘This teaches that the Holy One, blessed be He, showed our blessed ancestor Abraham his circumcision with His finger.’45 This tradition is not found in GenR 47.9 (T–A, 476). Given that the pedagogical topos of divine demonstrations using a demonstrative pronoun (‘this’) uses heavenly prototypes,46 it cannot be excluded that the first pronoun (his) may not refer to Abraham.

Overall, the tradition of God's participation in the circumcision rite is ancient. It was also known to the early synagogue poet Yannai (5th–6th c.), whose language may reflect an unknown variant: ‘You held him (tamakhta) by his right hand.’47

2. God's Attributes Affected

(i) SifDeut 311 (F, 351). Citing Deut. 32: 8.

Before (͑ad shelo͗ ba͗) our father Abraham came—kivyakhol the Holy One, blessed be He, judged the world with the attribute of severity (͗akhzariyut)48…But after our father Abraham came into the world, it (the world) merited to receive sufferings; and they (the sufferings) began to manifest themselves, as it is said, ‘And there was famine in the land and Abram went down to Egypt’ (Gen. 12: 10).

The occurrence of the term kiv. seems without any exegetical basis here. Possibly, the verb be-hanḥelin the lemma (‘When (the Most High) gave inheritence (to the nations)’) was interpreted as ‘cause a flood’ (R. David Pardo);49 cf. Ps. 124: 4. The flood is, in fact, mentioned in the midrash as such a cruel punishment of extinction. Hence, with the advent of Abraham and ‘the number of the children of Israel’ (Deut. 32: 8) that were his descendants, God fixed the number of the other nations, and thus their continued (p.339) existence. Alternatively, kiv. simply qualifies the strong statement about the cruelty of divine judgements in the pre-patriarchal past.

3. Divine Rule Affected

(i) SifDeut. 313 (F, 354 f.). Deut. 32: 10 is applied to Abraham. The verb yeboneneihu is construed to mean, ‘He (God) caused him (Abraham) to know Him.’ The point is further elaborated.

Before our father Abraham came into the world—kivyakhol the Holy One, blessed be He, was only king over the heavens; as is said, ‘YHWH, the God of the Heavens, who took me’ (Gen. 24: 7). But after Abraham came into the world, he made Him King over the heavens and the earth; as is said, ‘I will make you swear by YHWH—the God of the heavens and the God of the earth’ (v. 3).

The term kiv. emphasizes a presumptive interpretation of the difference between the divine epithets in Gen. 24: 3 and 7.50 Were its function merely euphemistic, the force of the midrashic teaching would be lost. The phrase ‘until…came’ is formulaic, as in 2(i) above. The anon. unit is condensed in a version in GenR 59.8 (T–A, 636): ‘Before I made Him known to His creatures, (He was) “the God of the heavens”; since I made Him known ot His creatures, (He is also) “the God of the earth”.’51 This instance is without kiv., and less bold than SifDeut 311 (which has himlikho, ‘made Him king’, or ‘caused Him to be king’).52

4. God's Reward

(i) SER 13 (MIS, 60).

As a reward for our father Abraham, who accompanied the ministering angels, the Holy One, blessed be He, accompanied his descendants for 40 years in the desert; as is said, ‘And YHWH went before them by day’ (Exod. 13: 21). Were the matter not written, one could not say it—kivyakhol, as a father before his son, (or) as a master who carries a lantern before his servant…

This tradition presupposes the rabbinic understanding of ‘sending’ as ‘accompaniment’ (in general and specifically regarding Gen. 18: 16; cf. MdRI Beshalaḥ, Pet. (H–R, 81); but the reward is different). Though the matter is written in Scripture, kiv. marks an exegetical extension with an emphasis on the hierarchical reversal. A variant occurs in TanB Shelaḥ 11 addendum (B, 80), but kiv. is used oddly and euphemistically with the analogy; cf. the formulation in ExodR 25.6.

5. Messianic Age

(i) MidPs 18.29 (B, 157).

R. Yudan said in the name of R. Ḥama: In the future, the Holy One, blessed be He, will have the King Messiah sit at His right hand; (as is said,) ‘YHWH said to my lord, sit at My right hand’ (Ps. 110: 2), and Abraham at His left. Abraham's face fell and he said: The son of my descendants sits at the right side, and I at the left?! The (p.340) Holy One, blessed be He, appeased him, and said to him: Your descendant is at My right, and I am at your right—kivyakhol ‘The Lord at your right’ (v. 5)…53

A bold reading of the sequence of verses (vv. 2 and 5) as a dramatic dialogue between God and Abraham resolves the apparent contradiction (if the ‘lord’ is at YHWH's right, how is ‘the Lord’ at your right?). The term kiv. bridges the resolution and the scriptural proof (v. 5), which is contrued in terms of Abraham.54 Given these factors, and the messiology involved, it is unlikely that this forceful teaching would be undermined by a euphemistic qualification. It seems more probable that kiv. marks the presumptive reinterpretation of v. 5 as an expression of divine humility before Abraham.

As this divine exaltation of Abraham qualifies the Messiah's status, it may even have a polemical force. Cf. Matt 26: 64, where Jesus announces: ‘Hereafter you shall see the son of man sitting at the right hand of the (divine) Power (dynamis).’

R. Yudan (A 4) transmits the teaching of R. Ḥama (b. Ḥanina?; A 2).

B. Jacob

1. Image on the Throne

(i) TanB Num Bemidbar 22 (B, 19). Following the lemma (Num. 3: 40) are two teachings.

(1) This is what Scripture says: ‘Because you are precious to me, you are honoured (nikhbadeta) and I love you’ (Isa. 43: 4). How so?—kivyakhol I have set your image (͗ikoniyan) on My throne of Glory (kabod);55 and by your name the angels praise Me, and say: ‘Blessed is YHWH, the God of Israel’ (Ps. 41: 14). Truly, ‘Because you are precious to Me, you are honoured’ (ibid.).

The verbal link between Num. 3: 40 and Isa. 43: 4 is submerged (both deal with substitutions; cf. the occurrence of taḥat in Num. 3: 41 and Isa. 43: 4b). The explicit teaching cites God's love for Jacob in Isa. 43: 4 and its cosmic consequence. The fixing of Jacob's image on the heavenly throne is an established topos (GenR 68.12; 78.3). The term kiv. does not render this in doubt or hypothetical (‘as it were’). It rather indicates a striking use of Isa. 43: 4 in exegetical support of the teaching; namely, because God loves Jacob, He makes him part of the throne of Glory (nikhbadeta meaning here, ‘you have become glorified’; namely, become an aspect of the throne of Glory). The repetition of Isa. 43: 4a at the conclusion underscores this point. Presumably v. 4b, ‘I have put a man in your stead’, also refers to this iconic substitute for Jacob upon the throne.

(p.341) The angelic praise noted here emphasizes that YHWH is the God of Jacob. Elsewhere, there are references to Jacob himself in godly terms. In one instance, Gen. 33: 20 (‘And he (Jacob) built an altar and called it “El, God of Israel”’) was interpreted to mean that the Holy One, blessed be He, called him (Jacob): ‘El (God), the God of Israel’ (BT Meg. 18a). Even more remarkable is the passage in GenR 79.8 (T–A, 949 f.) where this verse was construed to say that Jacob called ‘himself’ God, and said to the Lord: ‘You are God on high and I am God below.’ According to the citation of this midrash by R. Baḥye ben Asher (ad loc.), God Himself said this to Jacob! R. Baḥye took this to mean that Jacob's heavenly image is the counterpart to God's Shekhinah on earth.56

The second teaching follows:

(2) Another interpretation. ‘Because you are precious to Me.’ The Holy One, blessed be He, said to Jacob: You are precious in My sight—kivyakhol, because I and My angels stood over you when you left Padan-Aram, and when you returned; as is said, ‘Jacob left…and behold YHWH stood over him, etc.’ (Gen. 28: 10–13).

It is unlikely that the term kiv. serves as a euphemistic qualifier here, since Scripture itself explicitly speaks about God standing (nitzab) over Jacob. More probable, therefore, is that kiv. marks the exegetical extension of the verse, whereby the angels who were on the ladder ‘set’ (mutzab) upon earth and ascending to heaven were also understood as standing watch over Jacob. A var. occurs in NumR 4.1.

This double teaching (with kiv.) also occurs in Tan Bemidbar 19; similarly in Ca.

2. Future Fate

(i) PdRK 23.2 Rosh Hashanah (BM, 334 f.). R. Samuel b. Naḥman (A3) interpreted the divine assurance to the nation in Jer. 30: 11, ‘Do not fear, My servant Jacob, be not dismayed, O Israel’, with respect to the patriarch Jacob who, in his dream, saw divine beings ascending and descending the ladder to heaven (Gen. 28: 12). The sage explained these beings as ‘the tutelary powers of the nations’, and said that God allotted fixed terms for their ascension or dominion over Israel—after which they would wane or descend in power. However, in the dream, the power of Edom (i.e. Rome) kept ascending and Jacob did not know how long this nation would rule.

At that moment, Jacob was afraid and said: Will you say perchance that this nation will not decline? The Holy One, blessed be He, answered: ‘Be not dismayed (͗al teiḥat), O Israel’ (Jer. 30: 11)—kivyakhol, even if you see it (Edom) sitting by Me, I shall bring it down; as is written, ‘Should you ascend like an eagle, or set your eyrie among the stars, even from there I shall bring you down’ (Obad. 4).

The preacher deftly uses Jer. 30: 11 as a divine response to Jacob's fear, for the words ͗al teiḥat (be not dismayed) were presumably construed to suggest (or imply) the sense ͗el taḥat (‘beneath’)—and thus convey the divine assurance that Edom would ultimately be brought ‘beneath Israel’ (i.e. be subservient to then). The term kiv. mediates between this citation and the ensuing explanation by God, thereby alerting the reader to a downfall exegetically encoded in Jer. 30: 11. It is less likely that kiv. introduces the second citation (from Obad. 4), since this is an explicit divine promise (p.342) addressed to Edom, and there would be no reason to qualify it. In fact, it would appear that this explicit quote reinforces the implicit promise marked by the word kiv.

Similarly, the same formula with kiv. occurs in Tan Vayeitzei͗ 2 in Col2.

3. God's Promise

(i) TanB Vayishlaḥ 10 (B, 168).

‘And Jacob returned safe’ (Gen. 33: 18). This confirms what Scripture says: ‘YHWH will guard your departure and return’ (Ps. 121: 8)—‘your departure’: ‘and Jacob departed’ (Gen. 18: 10); ‘and return’: ‘And Jacob returned’. ‘And you decree a matter, and it is fulfilled for you’ (Job 22: 28). The Holy One, blessed be He, says to the righteous one (tzaddiq): If you have done My will, I shall do yours more than Mine—and ‘You decree a matter, and it is fulfilled for you.’ R. Berekhia the Priest said: What does ‘and it is fulfilled for you’ mean? The Holy One, blessed be He, said to the righteous one: I say something and you say something—kivyakhol, I annul Mine and fulfil yours; indeed, ‘He fulfils yours’ (ve-yaqam lakh).

The opening context sets up a pious confirmation of divine aid for Jacob through a correlation of scriptural passages. Nothing, however, anticipates the shift to a righteous person's decree. Moreover, the first explanation skirts the matter and only deals with God fulfilling the will of His obedient one. R. Berekhia's teaching explicates the verse, with the term kiv. marking that God is construed as the subject of ve-yaqam lakh. However, any link to Jacob or the incipit of the verse is missing. As rendered here, the unit is a sequence of non-sequiturs and utterly opaque.

R. Berekhia is A5.

Traces of the foregoing midrashic tradition occur in YalShim, Job §908, which is an otherwise unknown homily on Job 22: 28, interpreted in terms of Jacob and Gen. 33: 18. S. Buber suggests that this unit is the missing opening portion of the TanB, cited above.57 It reads:

‘And you decree a matter, and it is fulfilled for you’ (Job 22: 28)—this refers to Jacob; ‘and on your paths a light shines’ (ibid.). Two paths (are implied), since it is written ‘If Elohim be with me and guard me on this path’ (Gen. 28: 20)—(and correspondingly God said) ‘And surely I shall be with you’ (ibid. v. 15); (and Jacob also said) ‘And I shall return in peace’ (be-shalom; v. 20)—(and correspondingly it says) ‘and Jacob returned safely (shalem)’.

In this midrash there are verbal tallies between Gen. 28: 20 and Job 22: 28, as well as Ps. 121: 8 (guard); and it also clarifies Jacob (the righteous one's) decrees, reading the two subjunctive clauses as asseveratives.58 R. Berekhia's foregoing explication may also have Jacob's words in mind, and understand God as suppressing His general promise in favour of Jacob's specifications. The concluding phrase of his teaching emphasizes that ‘indeed, “God will fulfill”’ the patriarch's decree. The term kiv. arguably marks this exegetical reading, and not merely a pious qualification of God's suppression of His own word.

(p.343) 4. Jacob, Leah, and God

(i) JT Soṭah 3.4, 15a. After Rachel makes a deal with Leah, giving her permission to lie with Jacob that night in exchange for Leah's mandrakes (Gen. 30: 15), the episode continues with Leah informing Jacob of her conjugal rights that evening (having ‘hired’ him with the plants). The narrator concludes, ‘And he lay with her ba-laylah hu͗’ (v. 16). What does this mean? Proper grammar should have yielded ba-layha ha-hu͗, ‘on that night’. Ever intent to find deep layers in Scripture, this odd formulation evoked a striking exegesis.

R. Abbahu said:—kivyakhol, he (Jacob) had Him (hu͗) alone in mind; (for) he knew that this (plan) did not enter her mind save for the purpose of establishing the tribes (of Israel).59

In this bold exegesis, R. Abbahu construes the pronoun hu͗ as indicating God Himself—upon whom Jacob focused his mind while lying with Leah. The term kiv. precedes this striking exegetical presumption. The sage further justifies her act by stating that Jacob knew her noble purpose, and thus acceded to the deed. Another version of R. Abbahu's teaching only deals with this latter point, and gives it a theological justification in saying that ‘the Holy One, blessed be He, foresaw that her sole intent was to establish the tribes—therefore Scripture had to say “come in to me” (v. 16)’ (GenR 72.5 (T–A, 841) ). In this version, the first part of the teaching is omitted.

The particular point made by R. Abbahu (A3) can be appreciated in comparison with the exegesis of this passage given by his teacher R. Yoḥanan in the preceding generation (A2). His teaching is preserved in BT Niddah 31a.

What does Scripture mean by ‘He lay with her ba-laylah hu͗?’—It teaches that the Holy One, blessed be He, helped him in that act, since He foresaw that her sole intent was for the sake of Heaven, in order to establish tribes in Israel. As it says, ‘Issachar is a strong-boned ass (ḥ‰amor garem)’ (Gen. 49: 14)—The ass of Issachar caused (garam) him ‘to Issachar’ (liYssakar).

Two components comprise this teaching. The first is the bold inference that the phrase va-yishkab ͑immah ba-laylah hu͗ means that ‘he (Jacob) lay with her in the night (with) Him (God)’. This would be the only way to understand the reference to divine help in relation to the words of the verse; and there is no doubt that the ma͑aseh involved is the sexual act itself. Confirmation of this perspective is found in a comment on this passage by R. Abraham Saba͑, who quotes the otherwise unknown Midrash Hagalui to the effect that Gen. 30: 16 means: ‘That the Holy One, blessed be He, helped in her impregnation—for He is called hu͗.’60 Hence we may say that R. Yoḥanan (in the above passage) is far bolder than R. Abbahu (who only speaks of Jacob's divine intention); and note also that the inference given is without the term kiv.

(p.344) It appears that later tradition found the formulation in BT Niddah 31a too bold and compromised it by adding a proof-text and dense explanation which is required neither by the opening form nor content. Indeed, the second part of the comment is confusing, and presumably suggests that the ass of Issachar caused Jacob ‘to be hired’ (construing the preposition + personal name as the verbal form liyssakher—not ‘(to go) to Issachar’). According to Rashi (ad loc.), who had this full version before him, the braying ass of Issachar caused Jacob to swerve by the tent of Leah on his way in from the field, and thus be engaged by her.61 Such a midrash is clearly based on the first part of v. 16, but transforms and dislocates the opening teaching. It transforms it by requiring the sexual ma͑aseh to be understood as the ‘event’ of encounter aided by God. But it distorts it, as well, since the clear intent of va-yishkav is the act in which God partook. The pious result leaves the meaning of the whole unit in shambles. R. Abbahu, who softened the divine involvement, does not use the final proof, nor is it found elsewhere. This is also true of the version found in Bereishit Zuṭa (ad loc.),62 in which the first part of R. Yoḥanan's teaching found in BT Niddah 31a is joined to the second part of R. Abbahu's teaching, as reported in GenR 72.5. One may speculate whether this was an original reading or a conflation.

C. Joseph

1. Divine Promise

(i) MidPs 80.2 (B, 362). The lemma, ‘O Shepherd (ro͑eh) of Israel, give ear; who leads (noheg) Joseph like a flock’ (Ps. 80: 2) is variously interpreted as a plea for God to deal mercifully with Israel, just as Joseph requited his brothers with favour. A named teaching concludes the series.

R. Menaḥama taught in the name of R. Abin: Just as Joseph's brothers requited him with evil deeds (ra͑ot), but he requited them with good ones, so we (Israel) have requited You (God) with evil deeds (ra͑ot), and You have requited us with good ones—kivyakhol, for (though) we have transgressed Your commandments, truly You (are He) ‘who leads Joseph like a flock’.

The exegesis is compact and the function of kiv. obscure. The analogy to Joseph's acts of kindness, and their application to God, suggests, first, that the noun ro͑eh in the first clause is a double entendre. It explicitly calls upon God, the superior Shepherd of Israel, to heed the psalmist's plea, and implicitly calls upon God (implied) to give (favourable) ‘ear to the evil (ra͑ah) of Israel’. In the same vein, the verb noheg in the second clause not only refers to God ‘leading’ His flock, but also to His ‘treating’ them in the kind manner Joseph requited his brothers. Viewed thus, the term kiv. marks the exegetical reading of ro͑eh with respect to the evil deeds (ra͑ot) of the people Israel, and the praise of God's providential treatment of them. Remarkably, in this homily Joseph is presented as the precedent for favourable divine behaviour, and thus the appeal for its continuity (this being the force of the final clause).63

(p.345) The tradent given here is R. Menaḥema, an obscure figure. To be preferred is the name R. Tanḥuma (A 5), found in the ed. prin. His source is R. Abin I (A4) or his son, R. Abin II (A5). The latter is frequently mentioned in Tanuma midrashim.

D. Moses

1. Intercession with God

(i) SifNum 84, Be-ha͑alotekha (H, 83). This passage presents a striking interaction between Moses and God read into Num. 10: 34.

When it (the ark) rested (ube-nuḥo), he (Moses) would say (yo͗mar): ‘Dwell (shubah), O YHWH, among the myriads (rebabot) of the tribes (͗alfei) of Israel’ (Num. 10: 34). Scripture hereby indicates (maggid) that when Israel was travelling, they were thousands (͗alafim), and when camping they were myriads—kivyakhol Moses said before the Omnipresent: I shall not allow (meiniaḥ) the Shekhinah to dwell64 until You make Israel's thousands, myriads; since from the answer (teshubah) given you can comprehend (teida͑)65 that He (God) said to them: ‘YHWH, God of Israel, will increase you a thousandfold’ (Deut. 1: 11).

This passage is exceptionally dense and elusive. In it, the term kiv. is apparently used to soften the bold condition set by Moses before God.66 However, on the basis of the implication derived from the lemma, and its subsequent elaboration (with terms like meiniaḥ and teshubah alluding to ube-nuḥo and shubah in the lemma), it would seem that kiv. also serves to mark the exegetical innovation regarding Moses' statement to God—and God's putative reply. On the basis of the tally of terms in the lemma and interpretation, as well as the development of the homily, the following explication may be proposed.

The lemma was read atomistically, and its exegesis builds on such linguistic features as (1) the archaic orthography of ube-nuḥoh (where the final heh marks the third person masculine singular suffix, and refers to the ark); (2) the occurrence of the imperfect form yo͗mar (used to mark a future conditional); (3) the cohortative form shubah; and (4) the sequence of rebabot before ͗alfei (the latter means ‘tribes’; but it is construed here to be the numerical ‘thousands’). These features are homiletically exploited to yield a statement by Moses to God to the effect that he (Moses) would permit the Shekhinah to alight (taking ube-nuḥoh as if it refers to Moses' relation to the Shekhinah, which is putatively marked by the final letter heh, now taken as a third-person feminine singular suffix (namely, when Moses let the Shekhinah alight) ) only if God ‘would say’ something about the increase of the people of Israel. Since the terms of this request are not explicitly given, the homilist infers the details from the trace of a divine ‘answer’ (the verb shubah is apparently construed as a bi-form of the noun teshubah) that is presumed to be the reference to ‘myriads’ and ‘thousands’ in the (p.346) lemma. Support for the latter interpretation derives from Deut. 1: 11, understood here as the words that Moses tells God to say (yo͗mar is now construed as a cohortative, ‘let Him say’).

If this reconstruction of the midrash is accurate, the term kiv. serves a hermeneutical function and marks the exegetical construction of events purported to occur in the lemma—regarding a coercive condition set upon God by Moses, and God's compliance. The concluding phrase, rebabot ͗alfei yisra͗el, is thus taken as evidence of God's blessing; namely, that the ‘thousands of Israel’ will become ‘myriads’ (that is, the rebabot will become ͗alfei yisra͗el).

A version of the homily occurs in Sif Num Zuṭa (H, 267) in the name of R. Judah (bar Ila͗͑i; T3). But in this instance there is no initial statement of inference; the explication is presented as a statement by Moses,67 without the term kiv. and without a condition put to God; and there is no presumed divine answer. Hence the exegetical link to the lemma is utterly opaque.

In other midrashic sources, a righteous person (tzaddik) can effect the descent of the Shekhinah because of his pious behaviour (PdRK 1.1 (BM, 2) ). However, the present case is particularly notable since a person sets a condition for this divine descent, and even coerces blessings for the nation. Overall, this phenomenon is a variation of the theme of a tzaddiq whose decrees are obeyed by God, or who otherwise interferes with or overrules divine actions (see in (ii), 2(i), and E, below).

(ii) Tan Vayera͗ 19.68 This teaching presents an explication of the power of the tzaddiq, who may countermand a divine decree.69 Eccles. 8: 4 is cited (‘A king's command is authoritative, and none can say, “What are you doing?”’). The special case of Moses is adduced.

Know that when (Israel) did that (mis)deed (with the Golden Calf), the Holy One, blessed be He, wanted to destroy them. Moses arose (and), kivyakhol, seized the Holy One, blessed be He, like a person seizes his fellow70—for Scripture states, ‘And now, leave Me be…(that I may destroy them)’ (Exod. 32: 10); and you may also learn this from another verse…(Deut. 9: 14).

Scripture does not indicate an act of physical restraint by Moses. But just this interpretation is the result of the concrete construal of the verb haniḥah li (‘leave Me be’) in this midrash. Hence, the term kiv. does not so much undermine the new bold assertion (which would weaken the rhetoric) as underscore the exegetical presumption at play, whereby God putatively begs Moses to ‘let go of Me’. In this account, Moses appears to have acted boldly and aggressively as an intercessor—trying to stay the divine wrath. Notably, Moses' prior act is presumed, but not indicated. For a different rendition of the divine–human psychodynamics of this episode, see V. B.2(i) below.

(p.347) (iii) MidPs 90.5 (B, 388 f.). Another teaching dealing with Moses' influence upon God, is based on an interpretation of the reference to Moses as an ͗ish ha-͗elohim (‘man of God’, Deut. 33: 1) in terms of his being the husband (͗ish) of God.71

R. Judah b. R. Simon said in the name of R. Simeon b. Lakish. Why is he (Moses) named ͗ish ha-͗elohim?—Just as when a man (͗ish) wishes to annul his wife's vow he may do so, or if he wishes to uphold it (le-qayyem) he may do so; as it says (in Scripture:) ‘her husband (͗ishah) may uphold it (yeqimennu) or her husband may annul it’ (Num. 30: 14)—kivyakhol, (with respect to) the Holy One, blessed be He; (for) Moses said to Him: ‘Arise (qumah), YHWH’ (Num. 10: 35), and ‘Return, YHWH’ (v. 36).

The imperatives used by Moses to activate God's arousal to lead the nation, or His repose (Num. 10: 35 ff.), are interpreted in terms of the efficacy of a husband's word to uphold or annul his wife's vow. A series of verbal tallies are involved: (1) the word ͗ish links Moses to a husband; and (2) the verb qum (‘get up’) links Moses' divine exhortation (qumah) to a husband's ability to sustain his wife's vow (yeqimennu). The bold application of the marital analogy to Moses' relationship to God is thus an exegetical tour de force, hermeneutically marked by the term kiv. The discretion used by the tradents in formulating this teaching underscores its daring presumptions.

R. Simeon b. Lakish (A 2) and R. Judah b. Simon (A 4) are often linked in kiv. traditions, and constitute a chain of transmission.

2. Issues Decreed before God

(i) ExodR 21.2. A recurrent proof-text used to mark a tzaddiq's powers to command God is Job 22: 25, ‘You decree a word and it is fulfilled for you.’ This is also the text used here (regarding God's word to Moses in Exod. 14: 15).

R. Levi said: Just as the Holy One, blessed be He, commands Moses and speaks to him, so did Moses issue commands kivyakhol before the Holy One…for indeed the Josephites say to him, ‘YHWH commanded my lord (͗et ͗adoni tzivvah YHWH)… and my lord was commanded by YHWH’ (Num. 36: 2).

As formulated, the text adduced from Num. 36: 2 does not appear to prove the point, and is apparently redundant. This leaves the homily confusing, since the lemma only deals with God's command to Moses. Hence one must assume that the first clause of the verse was construed to indicate that ‘my lord (Moses) commanded YHWH’. This exegetical presupposition, which treats ͗adoni as both the subject and object of the verse,72 provides the right symmetry to the opening analogy. The term kiv. thus marks the hermeneutical turn that links Job 22: 25 to Num. 36: 2, and anticipates the bold rereading of the latter verse.

The teacher is R. Levi (A 3).

3. God Elevates Moses

(i) ͗Eileh Debarim Zuṭa. Adduced by this name in YalShim, Pinḥas, §776. According to this citation, R. Judah said that in 175 portions the DN precedes the name of Moses; whereas in Num. 27: 16,

(p.348) Moses precedes the (divine) Name—kivyakhol, the Holy One, blessed be He, exalted Moses to His Name.

This is a presumptive explanation of the change in precedence; the term kiv. marks this theological assertion. It is a weak usage.

E. Other Personalities

1. The Tzaddiq

(i) DeutR 10.3. The scriptural phrase, ‘He who rules justly (tzaddiq), he who rules (moshel) [in]73 awe of God’ (2 Sam. 23: 3) is radically reinterpreted by a redivision of the words.

What does ‘tzaddiq moshel (in the awe) of God’ mean?—(It means that) tzaddiqim (righteous persons) rule, kivyakhol (like) the Holy One, blessed be He. ‘Rule’ in what way?—Whatever the Holy One…does, the tzaddiqim do.

The term kiv. mediates the analogy between the rule of tzaddiqim and God. It also marks the exegetical presumption of the comparison, which depends on making the concluding adverb of one clause (tzaddiq, justly) into the subject of the second (a tzaddiq, righteous person). The sentence is thus construed to refer to human decrees. In the process, the phrase ‘awe of God’ (yir͗at͗elohim) is transformed from the quality of a just ruler to a quality of God Himself. A series of similar acts performed by God and a tzaddiq follows.


A. The Covenant Relation

1. Several Teachings Stressing the Intimate Relationship between God and Israel

(i) ExodR 33.1. The presence of God is given to Israel through the Torah itself, by a paradoxical reading of the phrase ve-yiqḥu li terumah (Exod. 25: 2). This divine behest that the Israelites ‘bring Me a gift’ for the building of the Tabernacle is homiletically linked to Prov. 4: 2, ‘For I have given you a good teaching (leqaḥ), (therefore) do not abandon My instruction (torah)’. The homily turns on the fact that the word leqaḥ can mean ‘purchase’ as well as ‘teaching’, and thus raises the paradoxical possibility that God gave Israel a ‘purchase’.

Now is there a purchase (leqaḥ) about which one can say that the seller is sold (along) with it?! Said the Holy One, blessed be He: I sold you My Torah (torati)—(and) kivyakhol I am purchased with it; as it says, ve-yiqḥu li terumah.

Based on reading Prov. 4: 2 as suggesting that God has ‘given’ (natati) Israel a ‘purchase’ (leqaḥ), which is the Torah-instruction, Exod. 25: 2 is construed to teach that with that ‘purchase’ Israel also ‘will acquire’ (ve-yiqḥu) God Himself (li, ‘Me’) as a ‘gift’. This bold theological conclusion is read into the first phrases of the command in Exod. 25: 2, and the term kiv. alerts one to interpret the proof-text in a manner consonant with both the homilist's rhetorical query (Now is there a purchase?) and the subsequent divine assertion (I am purchased with it). On this view, kiv. does not so much mark a (p.349) euphemistic weakening of the teaching (which would deprive it of motivation and force), as signal a daring theological interpretation of the ensuing proof-text. The power of the homily lies in the mythic concreteness of the exegesis involved, even if it was deemed necessary to mark its presumptive character.

A variant teaching in ExodR 33.6 is attributed to R. Berekhia (A 5). Here again normal purchase practices are juxtaposed to divine actions, but without any direct reference to Prov. 4: 2. In this case, the exegetical point is made explicitly: ‘But the Holy One, blessed be He, gave the Torah to Israel, and says to them—kiv. “(It is) Me (li) that you are purchasing (loqeḥim)”; thus (it says,) ve-yiqḥu li terumah (“and they shall acquire Me…”).’ The use of the verb loqeḥim indicates that a reference to Prov. 4: 2 has been lost here. As above, the term kiv. marks the presumptive relation between the bold theologoumenon and its scriptural proof. There would be no sense in underminding the former and then linking it to a proof-text.74

(ii) LevR 30.13 ͗Emor (MM, 710 f.). In this case, the presence of God is given the nation through the Sanctuary. A homily of R. Judah (bar Ila͗͑i, T3) transmitted by R. Simeon b. Pazzi (A3) notes that Scripture refers to commands where Israel was asked to ‘take’ something—for their benefit. On this basis, Exod. 25: 2 is interpreted as a divine request to the people.

Kivyakhol the Holy One, blessed be He, said: Take Me and I shall dwell among you. (Hence Scripture) does not say ve-yiqḥu terumah (‘And they shall take a gift’) but ve-yiqḥu li terumah—You are taking Me (namely, ‘And they shall take Me as a gift’).

The theological assertion is underscored by the specificity of scriptural langauge. The term kiv. thus marks the bold statement of God, which is based on a presumptive reading of the proof-text; for were its primary function to qualify this assertion, the effect of the homily would be thereby neutralized or undermined.

The same homily is found in Tan ͗Emor 17, though here with the explicit qualification ‘dabar qasheh (a difficult passage) kiv.’—marking the boldness of the exegesis.75 Similarly, in TanB ͗Emor 24 (p. 98) and Ca.

(iii) MdRI Yitro 2 (H–R, 208). Emphasizing the covenantal relationship between God and Israel, Exod. 19: 6 is explained:

‘And you shall be for Me’ (Exod. 19: 6)—kivyakhol I shall not set up or cause any others to rule (over you) but Me; as it says, ‘Behold, the guardian of Israel will neither slumber nor sleep’ (Ps. 121: 4).

The point made here is to stress the exclusive lordship and protection of YHWH for Israel; only divine sovereignty shall prevail, according to this divine promise. As this teaching of providence is not unusual, it would be hard to explain the use of the kiv. here as qualifying the teaching. One possibility is that it marks the categorical aspect of the use of Exod. 19: 6 here—in contrast to its original function as a result clause consequent upon Israelite compliance with covenantal demands (‘If you heed…then you shall be for Me’; Exod. 19: 5–6). The exegetical shift is thus to construe v. 6a in (p.350) apodictic terms, and not as an apodosis-clause. Alternatively, kiv. marks the emphasis on the exclusivity of God's relationship with Israel (i.e. ‘for Me’)—a point reinforced by the divine epithet in the proof-text.

2. Mighty Manifestations of Divine Aid

(i) SongsR 1.9. Songs 1: 9 is boldly interpreted by R. Akiba (T 2) in terms of God's appearance at the Sea in the image of a mare. This passage is cited and analysed above (pp. 233–4). The term kiv. marks the exegetical presumption involved (namely, that the scriptural verse indicates this appearance). Twelve other proof-texts are appended and deemed to indicate the variety of divine manifestations on Israel's behalf. The term kiv. accompanies each instance. The whole thrust of the teaching would be destroyed if kiv. merely functioned as a euphemistic qualification.

In MdRI Beshalaḥ 6 (H–R, 112) only the example from Songs 1: 9 is given, in the name of R. Pappius (in debate with R. Akiba). The swift censure of the remark by R. Akiba indicates that the exegetical proposal was taken seriously. In light of this debate structure, where R. Akiba takes a more moderate position in the other cases, R. Pappius may be the original proponent of this bold exegesis.76

B. Divine Participation in Miracles, Sorrow, and Joy

1. God and the History of Israel

(i) MdRI Beshalaḥ 2 (H–R, 186). After Moses built an altar, following the defeat of Amalek (Exod. 17: 15), the speaker of the next phrase Va-yiqra͗ shemo YHWH nisi is ambiguous. According to R. Joshua, Moses spoke it (Hence he construed the phrase: ‘And he (Moses) called its name (i.e. the altar): “YHWH is my banner”’); but according to R. Eleazar ha-Moda͑i, the speaker was God (hence, he construed it: ‘And YWHW called its name (i.e. the altar): “My banner”’). After this, an anonymous midrashic expansion follows. It picks up the term nes (banner), but now construes it as ‘miracle’ (another meaning of this word).

So also you find (in Scripture) that whenever Israel experiences a miracle—kivyakhol the miracle is His (lit. before Him); as it says, ‘YHWH is my miracle (nisi)’; (and whenever there is) trouble for Israel, (the) trouble is His (lit. before Him); as it says, ‘In all their trouble He was troubled’ (Isa. 63: 9); (and whenever there is) joy for Israel, joy is His (lit. before Him); as it says, ‘I rejoice in your salvation’ (1 Sam. 2: 1).

In each of these three teachings, God is presumed to be involved with Israel's fate, by virtue of an exegetical emphasis: in the first case, YHWH nisi is construed to mean ‘YHWH is my miracle’ or ‘(partakes of) the miracle that befalls me’; in the second, the sense is clearer that God partakes of Israel's trouble;77 and in the third instance, it would appear that the speaker is midrashically presumed to be God, who says, ‘For I (God) rejoice in your (Israel's) salvation.’ In all cases, the pertinent phrase is isolated from context and given a new sense. Only in the first case is the hermeneutical proof preceded by the term kiv.;78 and this instance may have been the original supplement, (p.351) with the other two subsequently added.79 The force of the supplement is to make a new theological point based on Scripture; its qualification would weaken this intent.

C. Divine Participation in Egyptian Bondage, Exiles, Return

1. Comprehensive Anthologies of Cases

(i) MdRI Bo͗ 14 (H–R, 51 f.). See the translation and discussion above (pp. 134–9). This collection is the most complete and nuanced of all exempla. Three general categories are used, followed by examples and proof-texts: (1) ‘Whenever (kol zeman) Israel is enslaved (meshu͑abadin)—kivyakhol the Shekhinah is enslaved (meshu͑abedet) with them.’80 The proof-text is Exod. 24: 10, whose two clauses are said to mark the enslavement and liberation, respectively. (2) ‘Everywhere (be-khol maqom) that Israel was exiled—kivyakhol the Shekhinah went into exile with them’; the exiles of Egypt (1 Sam. 2: 27), Babylon (Isa. 43: 14), Elam (Isa. 49: 38), and Edom (Isa. 63: 1) are adduced with proofs.81 (3) ‘And when (Israel) shall return—kivyakhol the Shekhinah returns with them’; Deut. 30: 3 and Songs 4: 8 are adduced in proof. The exegetical use of the proof-texts is analysed above; the term kiv. marks the hermeneutical presumptions about divine participation in Israelite history that are involved.

In (1) Isa. 63: 9 is adduced as another proof-text; but this passage only states that ‘He was troubled’ in all their troubles, without details. 2 Sam. 7: 23 is also cited, but only the liberation is specified. Both texts are discussed fully above (pp. 139, and 140). Breaking the overall editorial pattern, where the term kiv. heads up each subunit, kiv. is also mentioned in the bold interpretation applied to 2 Sam. 7: 23.

SifNum Beha͑alotekha 84 (H, 84) contains the same anthology of passages as in 1(i), but only category (1) is introduced by a category designator plus the term kiv. (Here the Shekhinah is said to be mishta͑bedet.)82

(ii) BT Meg. 29a is a condensed anthology of traditions about divine participation, in a beraitha attributed to R. Simeon bar Yochai (T3).83 A variant of the rubrics in (2) and (3) occur, but without the term kiv. This compilation may be the core form of the bi-polar anthology (exile-restoration) found in the sources. It is notable that late Babylonian tradition has preserved this old collection of materials (unknown as such in EI sources), and associates it with a named EI sage from tannaitic times.

ExodR 15.16. A variant of (2), with other nations introduced (Media and Greece). No comprehensive topic heading is used here. Rather, there is first the statement ‘When Israel was enslaved in Egypt, the Holy One, blessed be He, kivyakhol went into exile with them’; after this, the other cases are introduced by ‘so too’. This example is a later and weakened formulation of (2).

Ostensibly, the topic categories suggest that kiv. functions as an overall theological qualifier of the types of divine participation noted. However, the fact that proof-texts (p.352) follow, and almost all of these presuppose some midrashic reuse, suggests that the term kiv. marks the presumptive scriptural basis (‘as it were’) of the theological assertions of God's participation in Israelite history.84 The only exception is Jer. 49: 38, where God explicitly puts His throne in Elam.85

The mythic theology found in these teachings presumably preceded the proofs, which now support the teachings with texts requiring midrashic explication. If the term kiv. had been primarily employed to qualify the theological statements, their rhetorical impact and proclamatory assertions would be completely undermined, and the function of the scriptural proofs that follow would be inexplicable.86

2. Egyptian Bondage and Redemption

In other cases, interpretations of Exod. 24: 10 are presented as part of a series of related midrashic traditions. Three distinct interpretations are found. The first is traced to R. Levi b. Sisi (T5), as reported by R. Berekhia (A5) among others; the second is given by R. Berekhia alone; and the third is by Bar Qappara (T5). See the full explication of this passage in terms of God's participation in Israel's enslavement and redemption above (pp. 136–8).

a. Enslavement and redemption as two poles of divine activity (i) LevR 23.8 Aḥarei Mot (MM, 537 f.). The first and third cases both use the formula ‘before they (the Israelites) were redeemed’, but only in the first case is it preceded by the lemma from Exod. 24: 10a; they then both give the formula ‘but after they were redeemed’, and follow it with citations from Exod. 24: 10b. The midrashic details are somewhat apocopated or elided in the first case, but more fully articulated in the third one (see below). The use of the formulas in both instances (but not in the second) suggests that this was an original component of the tannaitic rhetoric. Both R. Levi and Bar Qappara are T5. The Shekhinah is not mentioned in any of the three cases, the subject being ‘the God of Israel’ mentioned in the lemma. Only the third case uses the term kiv., but does not cite the whole lemma (presumably because it was adduced earlier).

The third teaching reads as follows:

Bar Qappara said: Before Israel was redeemed from Egypt—kivyakhol87 it (the brick, lebeinah)88 was inscribed in the firmament; (but) after they were redeemed, it was no longer seen in the firmament. What is the proof?89—‘And like the purity of the heavens for splendour’ (Exod. 24: 10b); namely, when they (the heavens) were cleared of clouds.

(p.353) The term kiv. is linked to the first interpretation of the lemma in this cluster; the query ma͗i ṭa͑ama͑ (‘what is the proof?’) introduces the scriptural proof for the second half of the teaching. In both instances, the myth is based on an exegetical presumption. The interpretation here is softer than that found in MdRI (1(i) ). In that case, the word ke-ma͑aseh could be construed to mean that God's action was ‘like the labour’ of the Israelites; while here it suggests that God only put an image of the work (something ‘like’ it) in the firmament, out of sympathetic concern. (Weaker yet is the reference in SifNum Zuṭa, Beha͑alotekha 35 (H, 267), where the people saw a zekher lebeinim, ‘a memorial of the bricks’.) Note in this regard that 1(i) refers to the enslavement of the Shekhinah ‘with’ the Israelites; whereas 2(i) mentions neither, suggesting that God remains transcendent while attentive to Israel's plight.90

Notably, the first interpretation indicates that R. Levi delivered the teaching in a synagogue in Nehardea; it was still transmitted down to the end of the Amoraic period (A5).

The foregoing cluster of traditions recur (with minor variations) in JT Sukkah 4.3, 54c; and SongR 4.8, but the term kiv. does not occur in these instances.

A striking addition to LevR 23.8 is found in MSS Ox4 and JT. After the conclusion (‘cleared of clouds’) it states: ‘(He is) the God of Israel; but when they were not cleared of clouds—kivyakhol, He is not the God of Israel.’ The purport of the remark is to introduce the visibility of God after the redemption from bondage and thus expand the midrash. However, the comment paradoxically subverts the original teaching, whose very point was to state that God was active as Israel's God—even doing a manner of brickwork which obscures His appearance.91

b. Emphasis solely on the pole of divine redemption or restoration from Egypt. Three proof-texts are critical: (i) MdRI Bo͗ 14 (H–R, 51). The bold rereading of 2 Sam. 7: 23 by R. Akiba (T2) is quoted and discussed above, pp. 139–41. By construing the preposition + pronoun suffix lakh in the lemma in a self-reflexive sense (as ‘Yourself’B;; not ‘for Yourself’), and by taking the ensuing reference to ‘a nation and its God’ to indicate God's direct involvement, the sage gave scriptural proof of divine self-redemption with the people from Egypt. Normally the idiom ͗ilmalei sheha-dabar katub (‘if it were not written’), which is also found here, indicates a reinterpretation that does not require any change to the received scriptural spelling or vocalization.92 Just this is the case at hand. Hence, the term kiv. seems pleonastic in this instance; though it still functions in context to underscore the exegetical basis of the interpretation put forth. The force of the term here is to indicate that the theological assertion is found ‘as it were’ in the written Scriptures.

(ii) JT Sukkah 4.3, 54c. R. Akiba explicitly rebuts his interlocutor and stresses that the word ͗elohav in the lemma must designate Israel's God, not an idol—for otherwise one would make a ‘sacred’ designator secular. This brief exchange is a further indication that the adduced interpretation in the passage is the issue here.93

(p.354) Notably, the ͗ilmalei formula is not given in this version; and the term kiv. is itself qualified by the word ke͗ilu (‘as if’). This usage is a clear pleonism, and suggests that kiv. was not perceived by subsequent tradition as qualifying R. Akiba's theology per se.

(iii) ExodR 15.12. R. Meir (T3; a disciple of R. Akiba) makes the co-redemption explicit. He explained the phrase ‘This month is for you (lakhem)’ (Exod. 12: 1) to mean: ‘For me (li) and for you is the redemption (ge͗ullah)—kivyakhol I am redeemed with you (nifdeh ͑immakhem); as it says, “When You redeemed yourself (padita lakh) from Egypt, people and its God”’ (2 Sam. 7: 23).94 Notably, the phrase following kiv. is terminologically linked to 2 Sam. 7: 23, after R. Meir makes his theological point. It thus appears that kiv. precedes a supportive supplement to the bold reading of Exod. 12: 1 by R. Meir, and that this presumptive explication of the first teaching is itself reinforced by a proof-text.

It is possible that R. Meir had the full verse in mind when he conceived his midrash of Exod. 12: 1, for the subsequent phrase reads, ‘It shall be the first for you (ri͗shon hu͗ lakhem).’ One may wonder whether the sage perceived this clause as doubling the first teaching about God's redemption and as asserting that God, the ‘Primus shall be with you’ at the redemption. Notably, the term ri͗shon is a cognomen for God in Isa. 44: 6 and elsewhere. The pronoun hu͗ would reinforce the point; namely, that ‘The First One, He shall be with you.’

Precisely this cognomen is explicated in ExodR 15.1, where Exod. 12: 1 is cited and commented upon: ‘Kivyakhol, the Holy One, blessed be He, is called “the first (ri͗shon)”—as it says, “I am (the) first (ri͗shon) and I am (the) last” (Isa. 44: 6).’ In this case, the term kiv. marks the reading of Exod. 12: 1 in terms of God. The teaching is anonymous and now heads a cluster of comments marked by the term ri͗shon known from amoraic homilies on Lev. 23: 40 (but without this prologue; see GenR 63.8 (T–A, 687), LevR 30.15 (MM, 713), and PdRK 28.10 (BM, 416 f.) ).

c. A Godly Exodus (i) PesR 21 ͑Aseret Ha-Dibrot (A) (MIS, 110a). Interpreting the lemma, ‘(I am the Lord, your God) who took you out (hotz͗eitikha) of the land of Egypt, from the house of bondage’ (Exod. 20: 2),

Ḥananiah the nephew of R. Joshua says: (The scriptural word hotz͗eitikha) is written hwtz͗tykkivyakhol I and you went out of Egypt.

Relying on the orthography involved, Ḥananiah (T2) reinterprets the verb hotz͗eitikha (which indicates that God supremely took Israel out of their Egyptian bondage) and implicitly repoints it as hutz͗eitikha, ‘I was taken out with you’ (construing a 1st-pers. hoph͑al with pronoun suffix); possibly presuming a crasis of hutz͗eiti (‘I was taken out’ (hoph͑al) ) + ͗itkha (‘with you’; i.e. the preposition ͗et plus a pronoun). The term kiv. marks this remarkable exegetical presumption, and in that sense qualifies the theological point being construed.

JT Sukkah 4.3, 54c, cites this tradition up to but not including the term kiv. It is thus truncated and inaccessible, except for one who would already know this midrash. Presumably, an earlier version of the teaching with kiv. plus an explication was edited (p.355) out. Had kiv. been a true qualifier of the theology, it is unlikely that this opaque redaction would have been necessary.95

d. An anthology of redemptions (i) Tan Aḥarei Mot 12. In a long catena of texts and interpretations documenting divine participation in Israel's redemption (cited, pp. 151 f.), Exod. 14: 30 is adduced, presumably by R. Berekhia (A5) who gives several related explications based on scriptural orthography. In TanB Aḥarei Mot 18, the teacher is R. Meir (T3).

‘And YHWH saved…Israel.’ It is written wywš͑kivyakhol, when Israel was redeemed, He was redeemed (nig͗al).96

As above in c, the midrash is based on the received spelling of Scripture. In this case, wywš͑ is presumably read as vayivvasha͑, ‘He (YHWH) was (Himself) saved with (͗et) Israel’—not as vayosha͑, ‘He saved’ them. Once again, an act of divine deliverance is reconfigured as a conjoint event (the particle ͗et is read as the preposition ‘with’, not as a nota accusativa). Note also that the concise orthographic point is clarified by a theological paraphrase, as in PesR 21. Again, the term kiv. functions to mark the remarkable exegetical presumption—and thus characterize the theological assertion being made.97

A brief citation of this interpretation from the lost Midrash Abkir is preserved by R. Eliezer Ha-Darshan in his Liqquṭim Mi-Sefer Gemaṭriot 94a.98

wywsˇ͑ (namely, vayosha͑). Read it wywsˇ͑ (namely, vayivvasha͑)—kivyakhol, He was saved (nosha͑) with them.

The qere (read) tradition is emphasized, not the ketib (written) orthography. Kiv. marks this midrashic intervention. As elsewhere, an active verb depicting divine redemption is construed to denote God's inclusion in the process.

e. For God's sake (i) ExodR 30.24. A series of bold teachings on God's own redemption begins with a comment on Isa. 56: 1, ‘For My salvation is soon to come.’ In context, this divine proclamation refers to God's mighty act of salvation on behalf of Israel in exile. The language of the proclamation allows the rhetor to make a striking observation.

‘For My salvation is soon to come’ (Isa. 56: 1). Scripture does not say ‘your (pl.) salvation’, but ‘my salvation’. May His name be blessed! If it were not written, it would not be permitted to say it. The Holy One, blessed be He, said to Israel: If you do not have merit (for salvation), I shall do it for Myself (bishvili)—kivyakhol. All (p.356) the time you are there in trouble, I am with you; as it says, ‘I am with him in trouble’ (Ps. 91: 15)—and I redeem My self (go͗el le-͑atzmi), as it says, ‘He saw and there was no one (person) and was astonished…so His arm saved him (va-tosha͑ lo zero͑o)…’ (Isa. 59: 16). And it also says, ‘Rejoice greatly, O daughter of Zion; shout aloud, O daughter of Jerusalem—for behold your king is come for you: a righteous one and saved (nosha͑)’ (Zech. 9: 9). It is not written ‘who saves’, but ‘who is saved (nosha͑)’. Truly, even if you do not have good deeds (as merit), the Holy One, blessed be He, does it for Himself (bishvilo); as it says, ‘For My salvation is soon to come.’

In this remarkable (anonymous) homily, the specification of the personal pronoun suffix ‘my’ allows the preacher a bold turn: the salvation of the people is done by God, for Himself—even if Israel is without merit. The term kiv. hardly qualifies the overall force of the original proclamation, for then the absolute graciousness of this theological pronouncement would be diminished. It rather marks the subsequent exegetical presumption, which transforms the act of divine agency—into an act done by God for Himself. Indeed, the full force of this teaching is spelled out through a series of proof-texts in which the sovereign action of God includes a self-reflexive aspect. The specific phrases and terms used (note especially the term le-͑atzmi and the use of Isa. 59: 16 to refer to God, not Israel) make it clear that the action is an act of divine self-deliverance, and not solely a supreme act of beneficence. In ways unspecified, the act of God's participation in exilic suffering requires His own salvation. The mythic dimension of this teaching is bold in every respect, and presumes to explicate divine prophecies of salvation with respect to God Himself. Particularly notable in this regard are the epexegetical comments (bishvili, bishvilo, and le-͑atzmi). By this means, the preacher has God Himself concur and underscore his own exegetical presumptions.

D. Divine Participation in Babylonian Bondage, Exile, and Redemption

1. God in Exile

(i) ME Pet 16 (B 13; ed. prin. Pet 15). In a teaching of R. Simeon b. Lakish (A2) about the exile of Israel among the nations, it is recorded that God overheard these people deride His divinity and assert that He has now been punished for His defeat of ancient kings such as Pharaoh, Sisera, and Sennacherib.

Will He always be a youth?—kivyakhol, ‘things’ have aged; as it says, ‘And he went (va-yabo͗) to the nations among whom they (Israel) came (ba͗u) and they desecrated His holy Name’ (Ezek. 36: 20). Scripture should have said ‘they went’ (va-yabo͗u), and you say ‘he went’? Rather, kivyakhol (this verb indicates) He (God) Himself, as it says, ‘He went to the nations’…

In this homily the term kiv. occurs twice with respect to the putative senescence of YHWH, after the nations sent Israel into exile. In the first case, it comes after the charge that God is no longer a youth, and introduces the theme of debility with a supporting verse from Ezek. 36: 20. If kiv. is not simply used euphemistically in this instance, because of the assertion of divine ageing, it anticipates a verse referring to the desecration of God's Name, which could only happen if He were weak and (p.357) powerless.99 In any event, the application of this verse to God in exile (stated at the outset of the homily) needs reinforcement, and this is adduced by means of the odd use of the singular verb va-yabo͗ alongside the reference to Israel by the verb ba͗u.100 According to R. Simeon, had Scripture intended only to indicate Israelite exile, the verse would have said va-yabo͗u; hence we must conclude that God Himself came into exile, as a weakened and exiled God. The term kiv. precedes this exegetical application of the singular verb to God Himself; and, at the same time, this explication reinforces the earlier assertion of divine senescence. That is, God has become old and His Name has been desecrated by foreign peoples in the exile to which He now comes—a matter that stands in stark contrast to His vigorous, victorious past. Scripture proves the point, by an exegetical presumption.

R. Simeon's teaching is blunt and may refer to contemporary polemics. In other instances this sage also expresses a theology of divine transformation and weakness (above, p. 170).

(ii) TanB Behar, 2. This passage comments on Lev. 25: 35, ‘If your kinsman becomes destitute and comes under your authority (u-maṭah yado ͑immakh; lit. and his hand is bent toward you)’. The original rule deals with treatment of a compatriot in economic straits; it is reread as dealing with the Babylonian enslavement of Israel and God's presence with them in exile.

‘When your hand shall bend’ towards Nebuchadnezzar, kivyakhol the Shekhinah ‘is with you (͑immakh)’.

The legal lemma is cited and its clause reinterpreted: the first phrase is construed as Israel's subjugation to Nebuchadnezzar; the second understands the Shekhinah to be the subject of ‘is with you (͑immakh)’. Through this reformulation, a totally new sense is given to the biblical passage. The term kiv. here introduces the bold specification of divine involvement in Israel's travail—marking the exegetical presumption, by which God's hand is putatively subjugated along with Israel. The proof-text adduced thereafter is Isa. 43: 14, which was commonly also reread in terms of the Shekhinah and its exile (cf. MdRI Bo͗ 14; see above, p. 139).101

2. Israel's Degradation and God

(i) SongsR 4.2 (end). In the context of comments about Israelite exile and redemption there occurs the following midrash by R. Simeon b. R. Yannai.

‘Now I shall arise’ (Ps. 12: 6). Whenever she (Israel) is wallowing in ash (͗efer)—kivyakhol, He is also (ve-hu͗ khen),102 as Isaiah said, ‘Shake yourself from the dust (͑afar), arise, O captive of Israel’ (Isa. 52: 2). At that time, ‘All flesh is silent before the Lord’ (Zech. 2: 17). Why? ‘Because He is aroused from His holy dwelling’ (ibid.)

(p.358) The opening scriptural passage (Ps. 12: 6) is a divine assertion, announcing God's decision to requite the needy people. This is understood by R. Simeon as God's own act of self-aroused redemption on behalf of exilic Israel. The kiv. clause marks the ensuing theologoumenon of divine participation in Israel's suffering. It is reinforced by correlating Isa. 52: 2 and Zech 2: 17. Just as Israel is called upon to shake off (hitna͑ari) the dust of degradation (Isa. 52: 2), God's arousal (nei͑or) to redemptive action suggests that He, too, has shaken off His sympathetic suffering (Zech. 2: 17). This exegetical correlation works out the theological presumption indicated by the term kiv. Possibly the word ‘Now‘ in the lemma indicates for the teacher the transition from the earlier condition of Israelite suffering to the heralded moment of God's arousal.

(ii) GenR 75.1 (T–A, 878; Lo). The version of this midrash found here is garbled. After citing Ps. 12: 6, it says ‘Whenever she is wallowing in ash, kiv.’ This account depletes the force of a teaching about the divine involvement in national suffering and treats the metaphor of Israel's wallowing in ash as a kind of euphemism. The relationship between Israel's arousal in Isa. 52: 2 and God's arousal in Zech. 2: 17 is thus also without motivation or precedent. However, the versions in Pa and Ox4 have ‘kiv. even He also (͗af hu͗ khen)’.

The tradent in GenR is R. Simeon b. R. Jonah.

3. Israel's Bondage and the Arm of God

(i) PdRK 17.5 Va-To͗mar Tziyon (BM, 286 f.; Ox1). A tradition of R. Azariah (A5) and R. Abbahu (A3) ‘in the name of Resh Lakish’ (A2; R. Simeon b. Lakish, as in Ox3).103 The tradition presumably derives from Tiberias. R. Abbahu studied with R. Yoḥanan, head of that academy, and was a companion of his brother-in-law, R. Simeon.104

This teaching is translated and analysed above (pp. 147–9). In the face of the enemy's acts, God empathically binds His arm behind His back. Lam. 2: 3 is the proof-text, marked by kiv. The citation originally indicates divine aggression and the withdrawal of aid. Rereading the preposition mipnei (as ‘on account of’, instead of ‘before’) transforms the verse. The new event is God's concrete imitation of a ‘historical’ event befalling Judaean soldiers. The mythic realism is direct; but the proof-text needs qualification to mark its new sense (hence the term kiv.).

The projected release of God's arm at the end of days reinterprets Dan. 12: 13—but without the term kiv. The introductory lemma for the homily is Ps. 137: 5, ‘If I forget you, O Jerusalem, let my right hand wither’. This human oath is reread here (implicitly) as a divine asseveration, that indicates God's faithful remembrance of Zion: He will not forget Jerusalem just as He will not forget His arm. PesR 31 (MIS, 144b–145a) iterates this tradition and cites Lam. 2: 3, but without kiv. God does not explicitly imitate Israel's travail, or speak in the first person. Rather, the teacher says that God's withdrawn arm is ‘caused to dwell (memushkan)’ with the people in exile.105 However, in this version the interpretation of Ps. 137: 5 as dealing with (p.359) God's oath is explicitly spelled out (this version also suggests the shape or scope of the original homily). In MidPs 137.5 (B, 524), Lam 2: 3 is cited as God's response to the Levites' mutilation of their fingers, so that they will not play songs of Zion while in exile (Ps. 137: 5). The withdrawal of His arm is construed in terms of a divine oath, by which God assures the nation that He will ‘not forget’ Zion. The term kiv. does not occur here.

(ii) PesR 28 ͑Al Naharot Babel (MIS, 136a). This tradition also regards the withdrawal of the divine arm as an oath, as in MidPs 137.5, though now in reaction to the joyous martyrdom of the Levites. The act of asseveration is introduced by the query, ‘Is it possible (kivyakhol) that He enacted’ this oath? And following the citation of Lam. 2: 3, indicating the withdrawn arm, there is a second query, ‘Is it possible (kivyakhol) that the Holy One, blessed be He, did not restore His arm to its place?’ That question is now answered by God's explication of Ps. 137: 5, to prove His good intent (and also v. 6).

These two uses of kiv. are most unusual, since the term is employed somewhat like yakhol, which is regularly used to introduce hypothetical queries in halakhic discourse. They have thus been contaminated by other rhetorical forms, and indicate nothing of exegetical practice or euphemistic pronouncement. Their value lies solely in the semantic confusion and transformation of usage to which they attest.

4. Divine Bondage and the Exile

(i) PdRK 13.9 Dibrei Yirmiyahu (BM, 232)106 and ME (B, 38). This teaching has been translated and discussed above (p. 145). The lemma from Jer. 40: 1, deals with Jeremiah's incarceration in chains on his way into exile; in the reinterpretation, the phrase ‘and he (ve-hu͑) was bound (͗asur) in chains’ is applied to God (‘He’) Himself by R. Aḥa.107 The pronoun hu͑ is taken as a direct reference to the divine.108 The term kiv. introduces the exegetical presumption involved in this bold theological teaching.

(ii) TanB Tetzaveh, 2 (B, 96). The plain sense of Songs 7: 6 (melekh ͗asur birehaṭim, ‘A king held captive in the tresses’) is uncertain. It presumably indicates the erotic entanglement of the lover (king) with his beloved. In this passage, it is midrashically construed as the divine king saying that He is ‘bound (͗asur)’ or foresworn to Israel on account of two episodes in which Abraham ran to fulfil a commandment.109 The term kiv. precedes God's application of the verse to Himself. The context is unstated, and possibly indicates God's presence with Israel in exile. The verse is not directly used here as an oath; but this is how it is used in LevR 31.4 (MM, 718 f.), SongR 7.6, and Tan Tetzaveh, 6.110

In these last three cases, Songs 7: 6 introduces a statement by God saying that He is ‘bound in oath’ to Israel, on account of Abraham's deeds. In the first two instances, this oath binds the Shekhinah—either to remain in the Temple, and thus not leave it (p.360) on account of the people's sins, or to dwell with the people in exile.111 In the last case, kiv. marks a divine act of importance to the people.

The term kiv. occurs only in the foregoing two Tanḥuma versions, and may therefore be a later editorial element.112 These traditions are anonymous. In LevR 31.4 and SongsR 7.6, the merit feature is attributed to R. Abba b. Kahana and R. Levi in the first case; R. Levi alone in the second. They are both A3.

E. Divine Suffering, Sorrow, and Mourning

1. Universal

(i) GenR de-Rabba. From an Oxford MS published by A. Epstein.113 Herein a passage from Jer. 30: 6 (‘Why then do I see every man (kol geber) with his hands on his loins, like a woman in labour? Why have all faces turned pale?’) is radically reinterpreted.

kol geberkivyakhol, this (refers to) the Holy One, blessed be He, for all Power (kol ha-geburot) is His, yet He suffers for His world.

The scriptural term kol geber (every man) is applied here to God (not as ‘Man’ but as the supreme ‘Omnipotent’ one; lit. ‘all the powers are His’). So, too, is the subsequent image of pain. Such a reading of the verse confirms the explication adduced by Raimundo Martini (‘kol geber refers to the Holy One, blessed be He’), though without the term kiv.; this version occurs in a citation from MidPs 20.114 However, the exegetical application to God is missing from the text of MidPs 20 known to moderns (B, 174), and from all MSS.115 It cannot be determined from Martini's citation whether kiv. occurred in the full version known to him. The term geber is also applied to God in BT Sanh. 98b, with the better reading, ‘He who has all Power (geburah)’;116 but no reference is made there to divine suffering.

2. For Individuals

(i) ExodR Shemot 2.5 (1) (S, 110 f.). An anonymous teaching that interprets the phrase ‘I am with him in sorrow (tzarah)’ (Ps. 91: 15) as God's sharing in Israel's suffering on the basis of His appearance to Moses in the thornbush—‘kiv. I share (shutaf) with them (Israel) in their pain (tza͑ar)’.117 The teaching is elliptical and without exegetical symmetry. It is taught elsewhere that God's participation in Israel's tzarah (sorrow) is by appearing in a ‘contracted’ or ‘narrow’ (tzar) place (i.e. the thornbush). Cf. Tan Shemot 14 and TanB Shemot 12; but both are without kiv.

The biblical verse refers to God as being ‘with’ Israel as a helper in their sorrow; whereas in the interpretation, pain and constriction are assigned to Him as well. (p.361) Presumably the term kiv. serves to mark the reinterpretation of ͑immo (be with him) in terms of the verb shutaf, and the construal of tzarah (sorrow) in terms of the ‘narrow place’ of the thornbush. In MidPs 20.3 (B, 173), tzarah (based on Ps. 20: 1) is also applied to God's participation in human suffering (mishtatfin kebodi, ‘My Glory shares’); but there is no reference to the Egyptian locale and the term kiv. does not occur. In view of the link between the lemma, ‘He (God) will answer you (ya͑ankha) on the day of sorrow (tzarah)’ and the reference to divine suffering, it is possible that the lemma was construed as if it said: ‘He will be pained with you (ye͑unnekha) on the day of sorrow.’

(ii) TanB Aḥarei Mot 13. The twofold occurrence of the phrase lifnei YHWH (‘before the Lord’) in Num. 3: 4, in connection with the death of Nadab and Abihu, is noted and its meaning pondered. The answer given is this:

The Holy One, blessed be He, said, ‘Remove the dead from before Me’; as it is written, ‘Draw near and carry out your brethren from before (penei) the Sanctuary (ha-qodesh)’ (Lev. 10: 4). Kivyakhol, when Israel is in sorrow, He is also with them; as it is written, ‘He is with them in all their sorrows’ (Isa. 63: 9).

The logic of the homily is opaque. Presumably, the double occurrence of the phrase lifnei YHWH marks the occurrence of the sin and its punishment before God Himself; and the presence of the corpses before God is reinforced by the divine command to remove them. This non-scriptural command functions here as the focus for Moses' otherwise autonomous order to the priests to remove their brethren from the sanctuary. However, the phrase ‘from before ha-qodesh’ was apparently taken in this midrash as an indication of the presence of the corpses before God, the ‘Holy One’, who was like a mourner whose dead are placed before him. This point is underscored by the kiv. clause. The midrash thus introduces a teaching of God's compassion into a scriptural scene of divine wrath and punishment.

(iii) M. Sanh. 6.5 (Vilna edn., 1908–9; Vienna edn., 1801).118 At the beginning of the mishnah we read:

R. Meir said: When a man is sorely troubled (mitzṭa͑er), what does the Shekhinah say?—kivyakhol, ‘My head is ill at ease (qlny), My arm is ill at ease.’119

The teaching of R. Meir (T3) follows mishnah 4, which concludes with a reference to the need to remove a hung criminal ‘at once’, and the citation of the scriptural proof: ‘His body shall not remain all night upon the tree, but you shall surely bury him the same day; for (ki) he that is hanged is a curse against God (ki qillelat ͗elohim taluy)’ (Deut. 21: 23). The final motive clause (beginning with ki) is then clarified, ‘As if to say, why is he hanged?—Because he cursed (beireikh) the (divine) Name, such that the Name of Heaven (God) is profaned.’ Following R. Meir's comment, there is a homiletical expansion of this point: ‘If God (the Omnipresent, ha-Maqom) is sore troubled (mitzṭa͑er) at the blood of the wicked that is spilt, how much the more so for the death of the righteous?’120 The mishnah then returns to the subject of keeping (p.362) corpses overnight. It is thus clear that R. Meir's remark is a distinct homiletic unit within the legal discourse.

However, R. Meir's dictum raises questions. The first of these bears on the origin of the formulation itself, since not all MSS and editions refer to the Shekhinah or use the term kivyakhol. Indeed, the reference to God first appears in the mishnah cited in a medieval MS of the Jerusalem Talmud (Leiden Scal. 3; 1289), and the term kiv. is first attested in the citation in Tosefot Yom Ṭov (16th c.), ad loc. In addition, there is a problem of meaning, since the verb qlny is not immediately understood. The earliest discussion of the term is found in BT Sanh. 46a, by Abbaye (BA 4) and Rabba (BA 4). According to the first, the sense is, ‘I am not qal’ (i.e. qalleini);121 according to the second, ‘I am qal’ (i.e. qallani). But the exact sense of the element qal- is questionable. Following the variant of Abbaye's position in BT Ḥag. 15b, it appears that the speaker in the mishnah takes the verb qalal to mean ‘lightness’ in the sense of ‘debasement’ or even ‘contempt’;122 whereas Rabba's position is that the term bespeaks a sense of ‘worthlessness’. Variants of these positions recur in the Middle Ages, as found respectively in Rashi (ad BT Sanh. 46b) and the Geonim (ad BT Ḥag. 15a).123

But who is speaking? If the words ‘Shekhinah’ and kivyakhol are not original, we must wonder at the motivation of R. Meir or those who transmitted his teaching; for it would appear to introduce the cry of despair for a criminal into the mishnah. But even if this is the case, the sense of the phrase ‘what does it say?’ is puzzling; for to what idiom and why are we so directed? If it is directed to the phrase qillelat ͗elohim (‘curse of God’) in Scripture, then R. Meir would have us think that the cry of despair qlny is the criminal's (or hanged one's) curse of God. But if this is so, we are hard pressed to explain the sequel in the mishnah, which speaks of divine pain for the wicked and even more so for the righteous.

It would thus appear from the mishnaic addendum to R. Meir's comment that God Himself is the subject of the sentence; and this is also the import of the variant of Abbaye's remark in BT Ḥag. 15a (which says that qal is used as a euphemism with respect to divinity).124 Starting from this point, and noting the verbal link between qlny and qillelat ͗elohim, we may suggest that R. Meir proposed a midrash on the motive-clause in Deut. 21: 23, construing the phrase as if it meant ‘the hanged person is a qillelat ͗elohim’. He taught that when a person suffers such a death (i.e. by hanging), God Himself experiences a loss (or depletion) or debasement in His own being, since a person created in His image has endured pain to his limbs. He thus cries qellani (meaning ‘depleted’ or ‘debased’).125

(p.363) Alternatively, God's cry (qellani me-) is over a depletion or debasement of (me-) the structure (or limbs) of His own being (namely, ‘I am weakened in My head’, etc.). Such an understanding of the cry would bring R. Meir's midrashic teaching into alignment with a remarkable statement (ultimately also a midrash) of his teacher, R. Akiba. That sage taught: ‘Whosever sheds blood (shofekh dam), Scripture accounts him as one who has diminished (mi͑eṭ) the (divine) image. What is the reason? (It says:) Whoever sheds the blood of ha-͗adam (a man), by man will his blood be shed (Gen. 9: 6)’ (GenR 34.6; T–A 326). R. Akiba was apparently struck by the definite article in the verse, and understood it to be referring to the Primordial Adam—namely, the divine Anthropos itself!126

The mishnaic redactor, as noted, also understood R. Meir's midrash on Deut. 21: 23 as a divine cry; and he shows his secondariness to R. Meir's comment by his use of the term ha-Maqom for God (not Shekhinah), by his rhetorical expansion of the scope of divine pain for the death of the righteous, and by his reference to spilt blood (damamshe-nishpakh) here (where hanging is the subject). If the latter expression is merely used idiomatically, it may also tally with R. Akiba's theological dictum. In any event, three layers of composition are evident in mishnahs 4 and 5: (1) the law of hanged criminals, which concludes with a midrashic explication; (2) R. Meir's piece of midrashic theology; and (3) a mishnaic supplement that uses a fortiori argumentation.

We may thus conclude that the reference to the Shekhinah is original to R. Meir's teaching, and that the term kiv. marks the exegetical reading of the words qillelat ͗elohim as God's cry ‘qlny.’ But even a euphemistic reading of this midrash was apparently too much for some later tradents, who deleted the words ‘Shekhinah’ and ‘kiv.’ entirely—preferring an incomprehensible solecism to such a bold piece of mythic theology.

R. Meir's theology of a God who ‘is sorely troubled’ (mitzṭa͑er) for members of His people may be related to another midrash by him. In Sifre Deut. 319 (F, 365), the phrase ‘You forgot the God who bore you (meḥollalekha)’ (Deut. 32: 18) is explained by R. Meir as referring to ‘The God who writhed (he-ḥal) for you, who was sorely troubled (she-nitzṭa͑er) for you; as it says, “writhing (ḥil) like a woman in labour” (Ps. 48: 7).’ Presumably the pollel form ḥll in Deut. 32: 18 (‘writhe’) was associated with the qal stem ḥwl and had the double entendre of ‘writhe’ and ‘be ill’. Cf. Jer. 4: 31, where the verbal noun ḥolah (a person ‘in travail’) puns on the word ḥolah (meaning an ‘ill person’), and is also linked to the adverb tzarah, ‘trouble’.

(iv) SifNum 82 Beha͑alotkha (H, 93). At the end of a collection of teachings dealing with the kabod (‘honour’; ‘glory’) which God gives the elders of Israel (H, 92 f.), it is taught that

The Omnipresent is as sorely troubled (mitzṭa͑er) for a singular elder (zaqen) as He is for the whole of Israel; as it says ‘I was angry at My people, I defiled My heritage (ḥillalti naḥalati)’—kivyakhol, troubled (meḥolalim) for all; but ‘upon the aged (zaqen) you made your yoke exceedingly heavy (hakhbadet)’ (Isa. 47: 6).

(p.364) This midrash is a striking reinterpretation of a divine condemnation of Babylon, in which God says that He defiled His people and inheritance in His anger, but they showed no mercy even upon the aged. Read in the light of the initial theological assertion that God is as sorely troubled for the affliction of each elder as for the whole nation, the negative phrase ḥillalti naḥalati in the lemma is radically transformed. Guided by the explication ‘kiv. troubled (meḥolalim)’, the expression of destruction in the lemma is retrospectively understood as expressing God's affliction on behalf of His people (something like ḥolalti naḥalati, ‘I was troubled for My inheritance’). The final clause is now set off in counterpoint—by God's emphasis on the travail put on ‘the zaqen’. In context, this undoubtedly refers to the oppression of the elderly in general. But the midrash picks up on this specification of the elderly by a collective noun and reads it as if only one elder is indicated—one who stands out for special emphasis within the larger context of national suffering.127

The term kiv. bridges the lemma and its explication, and marks the exegetical presumption read into the biblical text. Given the opening reference to divine trouble, there can be no doubt that the word meḥolalim must be read in this light. Other versions of this teaching read meḥullalim (in the sense of ‘defiled’) or even maḥul lakh—as if God has ‘forgiven’ Babylon for its actions against all except ‘the elder’. But this reading weakens the thematic structure of the teaching and undercuts the exegetical transformation.128

3. God Suffers with Israel—General

(i) BT Giṭṭin 58a. A teaching of Resh Lakish (A1) on the sorrows of the destruction. An episode of a renowned woman is recited, and this leads to an interpretation of Jer. 6: 26. According to the first clause, the nation (called here ‘daughter of My people’) is enjoined to mourn in the dust; but the second phrase is interpreted to indicate God's participation in the suffering of individuals. Thus, following the citation ‘mourn as for an individual (child),129 wail bitterly, for suddenly the destroyer is come upon us (͑aleinu)’, the sage notes:

It does not say ‘upon you’ but ‘upon us’—kivyakhol upon Me and you has the destroyer come.

The term kiv. marks the exegetical presumption whereby the word ͑aleinu is interpreted to indicate God's shared suffering with Israel—not Jeremiah's call to the nation to begin mourning in anticipation of the doom to befall it, himself included. Resh Lakish uses the same exegetical emphasis on ͑aleinu in PdRK 15.4 to teach about God's request that a dirge be recited for His suffering along with that of Israel. See below, 7b(i).

(ii) SongsR 2.3. After the qualifying comment ‘If it were not written, etc.’ R. Reuben goes on to explicate Isa. 66: 16 (‘For YHWH is judged through fire’) as applying to (p.365) God Himself. The term kiv. occurs between the qualifying phrase and proof-text—though not in all versions. A full citation and a consideration of parallels is given above (p. 169). Following the qualification, the term kiv. sharpens the presumptive aspect of the exegetical teaching.

4. God Suffers with Israel—as His Twin

(i) SongsR 5.2. This teaching is translated and discussed above (p. 165 and n. 26). The female is addressed by her beloved as tamati (‘my pure one’) in Songs 5: 2. In the midrash, this designation is reinterpreted as te͗omati (‘My twin’), and understood as God's characterization of His relationship to Israel. The teaching first emphasizes the sympathetic pain felt between twins, and then boldly applies this observation to God's suffering ‘with’ Israel (Ps. 91: 15) in her times of travail. The term kiv. precedes this theological point, which is dependent upon the aforementioned exegetical presumption concerning the noun tamati.

The teaching is attributed to R. Joshua of Sikhnin in the name of R. Levi (T5). See the variant in ExodR 2.5 (ShR, 111).130 The teaching of shared suffering also occurs in PdRK Haḥodesh, 5.6 (BM, 87 f.) and PesR Haḥodesh 15 (MIS, 70b), but without kiv.131

Kiv. is an unstable element in these traditions, and probably a secondary element in SongsR (and the later ExodR). If it were primary in PdRK, it is unlikely to have been removed.

5. God Suffers with Israel—as His Eye

(i) ME Lamentations 1.16 (B, 88). This remarkable teaching is translated and discussed above (pp. 165 f.). God mourns for His damaged eye, which is midrashically interpreted as the destroyed people Israel. The term kiv. marks the application of the reinterpreted lemma (‘My eye, My eye’) to God, after a bold reuse of Zech. 9: 1.

The teaching is attributed to R. Levi (T5). Cf. his teaching adduced in 4a (above).

(ii) MdRI Beshalaḥ 6 (H–R, 135). In a comment on Exod. 15: 7, praising God for defeating enemies, the commentator stresses that Scripture says ‘Your opponents (qamekha)’, not ‘our opponents’. This observation leads to the exegetical proposal that ‘Whoever rises (qam) against Israel is as if he rises against the Holy One, blessed be He.’ The reinterpretation turns on a rereading of qamekha as those ‘who rise against You’ (i.e. God is now the object of the opposition, not the Lord of battles). Among a series of proofs is a list of eleven passages that are said to refer to God, though they contradict the plain sense. The first of these passages, and its interpretation, is based on Zech. 2: 12.

‘For whoever harms (noge͑a) you harms the pupil of his eye’ (MT Zech. 2: 12).132 R. Judah says that it (Scripture) does not state ‘pupil of (p.366) an eye’, but rather ‘pupil of His eye’ is written—kivyakhol the reference is to God, but Scripture speaks euphemistically (kinnah ha-katub).133

According to R. Judah (b. Ila͗͑i?, T 3), the phrase ‘his eye’ refers to God's own eye, which is harmed when Israel is harmed. The sage presumes that writing ‘My eye’ would have been too bold, so a softer formulation is used (‘His eye’). He asserts this point on the basis of the written formulation of Scripture; and while the proof itself is weak, the mythic assumption behind it is strong. The term kiv. does not, therefore, qualify the reference to God's eye per se, since this is precisely the locus of midrashic concern—which would otherwise be left rhetorically inconsequential. It rather serves to indicate the presumptive reading of the third-person pronoun in the text with respect to God. R. Judah's teaching is adduced again in SifNum 84 (H, 81 f.), though here only eight of the examples of such euphemisms are found. Following it, a teaching by R. Yose is given, which takes ‘his eye’ to be none other than a human eye.134

The list of euphemisms in these tannaitic sources are noted in some later lists as ‘scribal corrections’ or ‘emendations’.135 This latter idiom first appears in amoraic traditions, where it notably occurs with regard to our passage—according to a teaching by R. Joshua ben Levi (A1): ‘“Whoever harms you, harms the pupil of his eye” (Zech. 2: 12). R. Joshua says that this is a scribal correction, (that) was (originally) written “My eye”’ (ExodR Bo͗ 13.1).136 There are thus two distinct traditions: one that speaks only of original euphemistic formulations; another that refers to secondary corrections due to divine honour.137 Notably, the term kiv. occurs distinctively in the former cases, thus marking a midrashic construal of the pronoun with respect to God. Alternatively, if the verb kinnah (marking the euphemism) actually indicates a change of the pronominal attribution, it would mark a secondary feature, and the term kiv. would then relate to the final formulation.138 Significantly, references to the scribal formulation ͑eini (‘My eye’) not only occur in traditions about corrections, but also in those referring to euphemistic renderings. The lemma quoted in MdRI Beshalaḥ 6 (H–R, 135) reads ͑eini;139 and the word ͑eini is used with the phrase kinnah ha-katub in Tan Beshalaḥ 16, according to Col3, where the presumption (p.367) is that it was the original reading.140 Here, also, kiv. marks the mythic interpretation based on the final scriptural formulation.

6. Divine Sorrow

a. God is shattered (i) MZ Lamentations 18 (B, 58). The lament in Jer. 8: 21, ‘Because my people are shattered (͑al sheber bat ͑ammi) I am shattered (hoshbarti)’, is connected with that in Lam. 3: 41, ‘My eyes shed streaming water because my people are shattered (͑al sheber bat ͑ammi).’ This latter lament is explicitly ascribed to God in the midrash; and the one in Jer. 8: 21 is presumed to teach the same thing. The midrash adds:

If Scripture did not say (this), the tongue that speaks such would have to be cut in pieces; but earlier (generations) have preceded (us in astonishment at this clause); as it says, ‘Previous ones are seized in terror’ (Job 18: 20).

This powerful teaching of divine shattering is stated concisely, without any qualification. A parable follows about a king who, hearing that his son was shattered by a stone said, ‘I am shattered.’ The application to God then follows.

Similarly, the Holy One, blessed be He, said—kivyakhol ‘because My people are shattered, I am shattered’ (Jer. 8: 21).

This version seems to qualify the previous one; but it arguably has been adduced to explicate the full theological force of the initial assertion. Since the application of the parable has God speaking, to regard the term kiv. simply as a theological qualifier undermines the rhetorical force of the passage. Alternatively, it marks the midrashic presumption that God is to be deemed the speaker in Jer. 8: 21—with all the ensuing mythic implications. For a similar use of kiv. in v. 23, see below, 6b (iii).

b. Sorrow and tears (i) PesR 29 Bakhoh Tibkeh (MIS, 136b).141 A series of passages are adduced to show divine tears at the destruction of the Temple. After a comment about Jeremiah's lament (Lam. 3: 20), the teaching continues:

Kivyakhol (Jeremiah) saw what is written about the glory of the Holy One, blessed be He, ‘On that day, the Lord God of Hosts called for crying and lamentation’ (Isa. 22: 12).142 And it also says, ‘Therefore I say, “Let Me be, I shall weep bitterly”’ (ibid. 4). Kivyakhol, the Holy One, blessed be He, said: Just as it is below, so also there is crying above, before Me, concerning what has occurred. Kivyakhol there is crying before Me, since I left (heinaḥti)143 My Shekhinah (below), as it says, ‘This is My Place of habitation (menuḥati) forever’ (Ps. 132: 14). And she cries, insofar as I have left the glory of My Place of habitation and left her. Therefore, she cries. And (p.368) the proof?—That which is said in the passage, ‘Cries, she cries (bakhoh tibkeh) in the night’ (Lam. 1: 2).144

The three uses of kiv. in this homily appear to be euphemistic; and since they disturb the syntax and sense of the passage, may be secondary additions. This notwithstanding, the primary teaching evokes great mythic pathos. It appears that the word menuḥati is used here as a substantive term for God's indwelling presence, the Shekhinah (presumably called here ‘My Indwelling’). After the destruction, the Holy One leaves this earthly modality of the divine presence and withdraws to heaven. The teaching stresses that there is crying below and on high—and this is taken to be the crying of the Shekhinah on earth and the Holy One in heaven. The scriptural proof for this event is the phrase bakhoh tibkeh (Lam. 1: 2)—understood here not as a stylistic construction that marks a verbal intensification (‘She shall surely cry’), but as a twofold clause: ‘He (The Holy One) cries, when She (the Shekhinah) cries.’

What is theologically remarkable and unique here is the presentation of God in terms of male (transcendent) and female (immanent) modalities which separated after the destruction—the assumption being that the two were conjoined when the Temple stood. On the formal level, this division integrates different midrashim which alternatively depict the Shekhinah remaining on earth after the destruction, and God as ascending to the heavenly world (see above, pp. 157–9). But the result is most striking, and anticipates features found in a variety of later Kabbalistic sources (see pp. 296–9)

(ii) ME 1 (B, 59). This is a more condensed version of the preceding teaching. Hereby, the phrase, ‘Cries, she cries’, is explained and applied.

She cries and causes others to cry (bokhah u-mebakkah); (meaning,) she cries and kivyakhol causes the Holy One, blessed be He, to cry.

In this teaching, various ‘others’ are presumed to have cried.145 Arguably, this version does not go as far as b(i) (above), since here it is Zion that cries, not the Shekhinah. The use of the term kiv. may be secondary here (as also above), and was added to temper the notion that Zion influences God's emotions. Notably, kiv. does not occur in the version found in LamR 1.23 (D).

(iii) MZ Lamentations 18 (B, 59). In the context of divine mourning and sorrow, the angels instruct God in the proper praxis—one feature of which is tears. God responds:

I too shall cry day and night, as it is said: ‘And I shall cry day and night for the slain of my people’ (Jer. 8: 23). Who cries?—kivyakhol, the Holy One, blessed be He.

The citation adduced by God is arguably the voice of Jeremiah. Hence, the point is made that God is, ‘as it were’, the speaker. The term kiv. marks this exegetical presumption, whose striking depiction is the very purpose of the teaching. If kiv. functioned solely as a euphemistic qualfier, the rhetoric effect would be entirely eviscerated.

(iv) MZ Lamentations (version 2) 1.20 (B, 142). After the destruction, God determines to remove His presence from Zion. The homily then adds: (p.369)

kivyakhol the Holy One, blessed be He, cried and said: What have I done that I have caused My Shekhinah to dwell below for Israel?

God goes on to lament that He will now withdraw and fears being a laughing stock before the nations. In this case, the term kiv. expresses a pious qualification and has no apparent exegetical function.

(v) MZ Lamentations (version 2) 1:3 (B, 74). In conjunction with divine sorrow and the rejected angelic attempts at consolation,146 R. Joshua b. Levi states:

From the day the Temple was destroyed and Jerusalem laid waste, the Holy One, blessed be He, has no joy—kivyakhol until the Temple and Jerusalem shall be rebuilt and Israel returned to its midst; as it is said, ‘I shall rejoice over Jerusalem and be happy for My people, and the sound of weeping and wailing shall not be heard any more’ (Isa. 65: 19).

The use of kiv. does not qualify the present divine sorrow, but marks its future cessation on the basis of Isa. 65: 19. In that verse, God's joy is stated explicitly; whereas the end of sorrow, applied in the prophecy to the nation alone, is understood in the midrash as pertaining to God.

R. Joshua b. Levi is A1.

The expression that God ‘had no joy’ recurs elsewhere with kiv., but only as pious insertions and without any exegetical function. Various reasons are given: (1) the death of the wicked, noted in MdRI Beshallaḥ 1 (H–R, 118), MdRS (E–M, 72), and Tan Beshallaḥ 10.1;147 and (2) because of various considerations bearing on the state of the world, noted in LevR 20.2 ͗Aḥarei Mot (MM, 450), Tan ͗Aḥarei Mot 21 (also Ca, Tan Shemini 3, and MidPs 148.4 (B, 538) ).

(vi) SER 28 (MIS, 154). In the context of dealing with divine sorrow and tears, the special character of God's feelings is underscored:

‘My eye must stream and flow with tears’ (Jer. 13: 17)—kivyakhol there is no being more compassionate for Israel than the Holy One, blessed be He, alone.

The term kiv. appears to qualify this concrete assertion of divine compassion. Alternatively, it may mark the presumptive attribution of Jeremiah's human speech to God.

c. God's cry of ‘Woe’ (i) PdRK Eicha 15.4 (BM, 251 f.). R. Yoḥanan (A2) likened the successive exiles of the nation to a king who punished and banished his sons. Just as the king lamented in woe on these occasions, so also when God exiled the ten northern tribes He recited the verse ‘Woe (͗oy) to them for they have wandered from Me’ (Hos. 7: 13). And subsequently,

When the tribes of Judah and Benjamin were exiled—kivyakhol the Holy One said ‘Woe (͗oy) is Me for My shattering [shibri]’ (Jer. 10: 19).

In this teaching, the term kiv. occurs before the second expression of ‘woe’ to mark the presumptive transfer of the pronoun (from Jeremiah) to God. There is no reason to employ the term it in the first instance (Hos. 7: 13) since God is clearly the speaker. The same teaching recurs in ME Pet 3 (B, 4).

(p.370) R. Yoḥanan's teaching is linked to another by his brother-in-law, R. Simeon b. Lakish. See 7b, below.

(ii) NumR Naso͗ 12.7. An interpretation of the opening word of Num. 7: 1, ‘And it was (va-yehi) on the day that Moses completed’ building the Tabernacle. The meaning of va-yehi is midrashically construed as a contraction of vay hayah, ‘There was (a cry of) woe.’ As to who said this,

R. Abin said, kivyakhol the Holy One, blessed be He, said vay.

By way of explication, a parable is given of a king who had a contentious wife and had her make him a royal garment—during which time she was preoccupied and did not complain.148 When she finished, he cried ‘Woe’, out of concern that she resume her contentious behaviour. In a like manner, we are told, Israel frequently complained in the desert, but ceased to do so when preoccupied with the building of the tabernacle. When they finished, God Himself cried vay—recalling Israel's past behaviour and hopeful that she would not resume her complaints. The term kiv. thus marks the injection of a divine dimension into the archival notice found in Scripture, underscoring the presumptive nature of the interpretation proposed.

R. Abin is A4.

A related version of the teaching occurs in PesR 5 Vayehi Beyom (MIS, 20b), though it is more opaquely rendered. The attribution here is to a R. Abba.

7. Divine Mourning

a. God observes bereavement (i) LamR 1.1 (D, Vilna). A teaching of R. Naḥman and R. Samuel (A 3) ‘in the name of R. Joshua b. Levi’ (A1). God asks the ministering angels to teach Him how a human king mourns, so that He could also grieve for His people and Temple. The rites include putting on sackcloth, darkening lamps, going barefooted, overturning beds, and ripping one's robe. The proof-texts given are, respectively, Isa. 50: 3, Joel 4: 15, Dan. 7: 9, Nahum 1: 3, and Lam. 3: 17. After the citation of Dan. 7: 9, regarding overturning the bed (‘As I looked on, the thrones were set up (remiv)’), the text adds: ‘kivyakhol, which were overturned’.

The term kiv. does not occur elsewhere in this teaching, nor in the versions found in PdRK 15.3 (BM, 250 f.) and ME 1.1 (B, 42 f.).149 Possibly, kiv. marks the fact that the verb remiv is used (in the midrash) in its explicit sense (‘thrown down’), and not in the applied or figurative sense of setting up furniture (found in Dan. 7: 23). The term is most likely a secondary addition, underscoring the act of divine mourning.

(ii) Tan Shemini 1. In this passage we learn that God observed

seven days of mourning before He brought the flood— kivyakhol. And from what passage (can we deduce) that He mourned? (From the fact that it says), ‘And the Lord…was saddened (va-yit͑atzeb)’ (Gen. 6: 7), and the meaning of sadness (͑atzibah) is mourning (͗eibel).

The use of kiv. here is loose and disconnected from the exegetical application. As it stands, it appears to be a secondary or euphemistic element. The term does not occur in Ca.

(p.371) b. God needs lamentation (i) PdRK Eicha 15.4 (BM, 252). This teaching of Resh Lakish (A 1) follows a related one about divine suffering by R. Yoḥanan, discussed above, 6c (i). The present instruction takes a more radical position, stating that ‘after the tribes of Judah and Benjamin were exiled—kivyakhol the Holy One, blessed be He, said: “I do not have the strength to mourn them.”’ God then requests professional mourners, invoking Jer. 9: 16. A bold interpretation of v. 17 follows. It emphasizes the inclusive pronoun ‘us’ or ‘our’ used in Scripture in order to mark the fact that the keeners mourn for God as well as for the people.

This teaching is translated and discussed above (pp. 170–1). The term kiv. appears to be euphemistic here; since it does not occur with the explanation of v. 17, where such words as ͑aleinu (us) are interpreted in an inclusive sense. Withal, it proleptically sets up the ensuing exegesis. A similar interpretation of ͑aleinu occurs in BT Giṭṭin 58a, in another teaching by Resh Lakish. In that case, however, the term kiv. does occur and marks the transition from the specific scriptural term ͑aleinu to its explication with reference to both God and the people.

ME Pet 3 (B, 4) gives the same teaching.

8. Divine ‘Sleep’

In striking contrast to the dominant motif of divine participation in Israel's historical suffering, a number of sources portray an apparent divine disengagement from protective care—compared to sleep—during which time Israel hopes in and evokes God's arousal. This activation of divine power will result in a positive new era for Israel, and an end of their enemies and oppressors.

a. During Israel's suffering (i) Tos. Soṭah 13.9 (TL, 234; MS Vienna 46). According to M. Ma͑aser Sheni 5.15, Yoḥanan the High Priest abolished three things—one of which were the ‘Awakeners’ (me͑orerim). The list in Tos. Soṭah 13.10 is different, and the reference in 13.9 simply says: ‘These are the Levites, who recited on the (Temple) dais, “Arise, Why do You sleep, O YHWH?” (Ps. 44: 24).’ The specific time and purpose is not indicated.150 Instead, an exegetical teaching is given by R. Yoḥanan b. Zakkai.

And does He sleep? Is it not already stated, ‘Surely He shall neither slumber nor sleep’ (Ps. 121: 4). Rather (understand it): Whenever Israel is in sorrow and the nations of the world are at ease—kivyakhol, ‘Arise, why do You sleep?’

R. Yoḥanan resolves the contradiction between Pss. 44: 24 and 121: 4 by means of a metaphorical interpretation. The term kiv. marks the shift to the more figurative undertanding of God's withdrawal and arousal. Indeed, its present function as an epexegetical ligature is clear from the formulation in the Erfurt MS (TL, 233), where the term kiv. is absent and the verb ‘they recite’ occurs. This version gives the recitation an exhortative tone that is much stronger and more concrete. Moreover, following the initial condition (‘When Israel is in sorrow’), the Levitical recitation assumes a somewhat time-bound character.151

(p.372) (ii) JT Soṭah 9.11, 24a offers a version that is set within the context of activities of Yoḥanan the High Priest, but does not indicate a reform against the Awakeners; and the teaching on Ps. 44: 24 is given anonymously. The first part is basically like the toseftan passage, where a textual contradiction is raised. However, in this case, another verse is added:

Why then does Scripture say, ‘The Lord awoke like one who sleeps’ (Ps. 78: 65)? Rather (understand it): kivyakhol, as if He sleeps when Israel is in sorrow, etc.

Hereby a different rhetorical strategy was followed. By introducing Ps. 78: 65 with its simile (‘like one who sleeps’) the redactor has prepared for a qualified reading of the divine sleep mentioned in Ps. 44: 24. Now the sleep is deemed entirely figurative. The added qualifier ke͗ilu (‘as if’) shows just how a later scribe wanted the term kiv. to be taken.

Use of the sleep motif in connection with the topos ‘When Israel is in straits and the nations at ease’ provides a sharp contrast with those teachings where this topos is used to introduce God's participation in Israel's travail (e.g. PdRK 17.5 (D.3(i) ).152

R. Yoḥanan b. Zakkai (T1) is explicitly the teacher of the midrash only in the toseftan version, MS Vienna 46. He is mentioned in JT as reporting the deeds of Yoḥanan the High Priest; hence, he could be construed as speaking the midrash there, as well.

b. Before the Redemption (i) MidPs 59.5 (B, 303). The lemma discussed is Ps. 59: 6, ‘O You, YHWH, God of Hosts, God of Israel, bestir Yourself (haqitzah) to bring all the nations to account; have no mercy on any treacherous evildoer.’ This exhortation is explicated as follows:

(It is) because in this world the Holy One, blessed be He, makes Himself as one who sleeps. And why?—Since it is not time that they (i.e. Israel) be redeemed; as it is said, ‘And He awoke (va-yiqatz) as one who sleeps’ (Ps. 78: 65). But in the future, when the end (ha-qetz) shall come—kivyakhol, the Holy One, blessed be He, is aroused (nei͑or); as is said, ‘Bestir Yourself to bring all the nations to account.’ At that time, ‘Have no mercy on any treacherous evildoer.’

The term haqitzah (in the lemma), va-yiqatz in the citation, and ha-qetz in the reference to future time establish the temporal coordinates of this midrash (present plea; present reality; future time). The terms are used to indicate that God in the present makes Himself ‘as’ one who is asleep. He is thereupon evoked by the imperative haqitzah—meaning here both, ‘bestir Yourself’ and ‘bring about the end’. The term kiv. mediates this double sense for the verb. The semblance of divine sleep serves as an indication of divine absence in this world; and it provides the metaphorical basis for hope in God's immediate engagement with the enemy.

This teaching also occurs in YalShim §777; but it is > in various MSS, and in the ed. prin. (Venice, 1546).

(ii) MidPs 78.18 (B, 356). A condensed paraphrase of b(i), though God's future arousal is as if from a wine-stupor.

(p.373) (iii) MidPs 121.3 (B, 505 f.). An expansion of earlier traditions, with proofs why there can be ‘neither sitting nor sleeping on High’. Having established this point, the query is then posed:

But what (is the meaning) of ‘Let Your guardian not slumber’? Just that: kivyakhol, it is from the afflictions that befall Israel in the world that the Holy One, blessed be He, sleeps.

In this version, the lemma (Ps. 121: 3) is taken to mean that although God ‘sleeps’ and is disengaged from the troubles of the world, lo͗ yanum ve-lo͗ yishan—He will neither slumber nor sleep in the world to come. The force of the term kiv. is thus apparently to mark this exegesis. This construal of the verb in the future sense (rather than as a present conditional) adds a further exegetical dimension to the meaning of Ps. 121: 3.


A. Gracious Kindness and Forgiveness

1. God ‘Forgets’ Sin

(i) PdRK Ki Tissa͗ 2.7 (BM, 26 f.). The homily links the verse that immediately precedes this lection, dealing with atonement (Exod. 30: 10), with the opening injunction of this unit: ‘When (ki) you raise up (or, count, tissa͗)…the Israelites’ (ibid., v. 12). Exod. 30: 10 thus serves to establish the overall thematic focus of the homily; whereas v. 12 deals with the specific issue of concern. The scriptural citation adduced (v. 12) does not, however, follow the MT spelling ts͗, but has tsh—a consonantal cluster that may be read taššeh, ‘lend’. This latter reading is then midrashically tallied with the same verb in Deut. 24: 10 (dealing with loans). On the basis of this hermeneutical revision of Exod. 30: 12, Moses is presumed to have addressed God as follows:

‘When Israel has merits (for atonement), let them alone’; but when they do not have such merit—kivyakhol lend them credit (hinnasheh) once a year, so that the Day of Atonement will come and provide atonement, ‘For by means of (lit. “on”; ba-) this day atonement shall be made for you’ (Lev. 16: 30).

The homily boldly transforms a divine rule dealing with counting into an appeal of intercession by Moses. The term kiv. marks the midrashic reading of the written orthography of Scripture known to the homilist and its application to an appeal for divine grace. It is unlikely that the term kiv. functions here to qualify God's ability to grant merit, for this would undermine the entire force of Moses' request. It would also contradict a feature of normative rabbinic theology; indeed, elsewhere God is asked to forgive unilaterally, if necessary, without the precondition of human repentence (cf. BT Yoma 86a).

The MT spelling tissa͗ is found in MhG Ki Tissa͗ (M, 641), but the verb taššeh is implied since the teaching invokes God to ‘forgive them—kivyakhol, once a year’. In TanB Ki Tissa͗ and YalShim Ki Tissa͗ (§ 386), neither tissa͑ nor taššeh occurs. None the less, a trace of the older sense may still be preserved insofar as God tells Moses to ‘go do it’ (zekof͗oto; i.e. the act of atonement) immediately. Significantly, the verb zakaf is often used by the ancient rabbis to refer to the establishment or erection of a loan (cf. (p.374) SifreDeut. 34; BT Metzi͑a͗ 72a). However, in these two cases Moses does not act as an intercessor on behalf of Israel; and the term kiv. does not occur.

(ii) PdRK 25.2 Seliḥot (BM, 381). In the context of extended explications of the divine attributes in Exod. 34: 6–7, the term nose͗ ͑avon (‘forgives sins’) is taken up. Among the interpretations, we read:

R. ḥuna (taught) in the name of R. Abbahu—kivyakhol, there is no forgetting (shikheḥah) before Him; yet for Israel He becomes forgetful. What is the proof? ‘Who is a God like You, nose͗ ͑avon and passing by transgression’ (Micah 7: 19).

In this terse teaching, nose͗ is midrashically construed as nošeh, ‘forgetting’ (cf. Gen. 41: 51); and the second phrase of the proof-text from Micah is assumed to reinforce the point. The use of the term kiv. here is somewhat opaque, since it qualifies the fact that God does not forget, but does not qualify the more radical point. One must therefore prefer the version in JT Sanh. 10.1, 27c, where kiv. immediately precedes the words ‘yet for Israel’ (and also JT Sheb. 1.6, 33c). While the term may be taken as offering a qualification here, as well, it may more likely mark the midrashic presumption attributed to God (i.e. that He forgives Israel). The midrash is spelled out in the version in JT Sanh. There after the query concerning the proof, we have the clause: ‘nosheh ͑avon (“He forgets sin”)—but the text writes nose͗.’ But as this explication interrupts the query from the proof itself, it may be secondary or misplaced. By comparison, PdRK is remarkably (and perhaps intentionally) laconic. Another variant occurs in MidPs 32.2 (B, 242).

R. ḥuna here is presumably R. ḥuna b. Abin (A4). He received the teaching from R. Abbahu (A3). The same tradents appear in JT Sanh (R. Abbahu) and MidPs (R. ḥuna in the name of R. Abbahu).

(iii) Midrash ḥaser Ve-Yater.153

Every (occurrence of) nose͗ (forgiving creatures) is written with (the letter) shin. Why? Whereas forgetting occurs among humans, with regard to the Holy One, blessed be He, (it says) ‘Who is a God like You, who forgives (nose͗) transgression?’ (Micah 7: 19). Why (is it written) with a shin? (To teach) that even though there is no forgetting with Him—kivyakhol, He causes the transgressions of Israel to be forgotten.

This formulation specifies the orthographic issue, and agrees with the theological content noted in (i)–(ii). The term kiv. marks the exegetical presumption.

2. God Forgives the Repentant

(i) MdRI Beshalaḥ 3 (H–R, 166). In connection with a discussion about the meaning of the manna, described ‘as hoar frost’ (ka-kefor),

R. Ṭarfon says, It only came down upon the hands of those who gathered it (ha-͗osfim).154ka-kefor upon the earth’—kivyakhol God stretched forth His hand and received the prayers of our ancestors who lay in the dust of the earth; and He (p.375) thereupon brought down the manna for Israel. As it says, ‘Then He has mercy upon him, and decrees: “Redeem him from descending into the Pit; for I have obtained a ransom (kofer)”’ (Job 33: 24).

R. Ṭarfon (T2) teaches that the manna that descended upon Israel was given in response to the prayers of the ancestors. He apparently construes the scriptural phrase ka-kefor as ke-kippur (‘as a ransom’ or ‘as an atonement’).155 That is, the prayers resulted in divine forgiveness and atonement and culminated in the physical gift of the manna. Alternatively, God receives the kofer (ransom) of intercessory or supplicatory prayers of the dead ancestors—and the complaining people receive kippur (atonement) and the gift of manna in return. On either reading, the term kiv. would seem to mark the bold theological meaning given to ka-kefor rather than serving to qualify God's reception of intercessory prayer.

MdRS (E–M, 110) has a similar reading.156

3. God Overlooks Sin

(i) PdRK 24.13 Shubah (BM, 370). A condensed exposition of Job 11: 11 is preserved.

R. Issaḥar of Kefar Mandi interpreted ‘For He knows false men, He sees iniquity but will not consider it’ (Job 11: 11). It is man's way in the world to heap up piles and piles of transgressions; but if he repents—kivyakhol, ‘He sees iniquity but will not consider it’

In this teaching the two phrases of the verse are exegetically transformed. According to the plain sense, the opening assertion of divine providence is reinforced by a rhetorical query: ‘And if He sees iniquity, will He not consider it?’ R. Issaḥar transforms the two phrases into a dramatic contrast: in the first part, God knows human iniquity; but the second part boldly indicates God's merciful overlooking of transgressions, if a person has repented. The term kiv. marks this exegetical presumption in which standard theology is taught through a hermeneutical rereading of the verse. It is unlikely that the term kiv. simply qualifies standard rabbinic theology about divine mercy.

The version in MidPs 5.8 (B, 54) is without kiv. and provides a fuller explication of the dynamic of human repentence and divine mercy. However, as only the second phrase of Job 11: 11 is cited, the radical rereading is obscured, and the quoted unit fits in with standard theology. The absence of the full exposition would explain the absence of the term kiv. in this case.157

B. Prayer and Intercession

1. God to Himself

(i) MidPs 76.3 (B, 341). On the verse ‘And His booth was in Shalem (va-yehi be-shalem sukko), and His dwelling in Zion’ (Ps. 76: 3).


R. Berekhia said: From the beginning of the creation of the world, the Holy One, blessed be He, made a booth (Temple) in Jerusalem—kivyakhol He would pray in it (thus): May it be (yehi ratzon) that My children do My will (retzoni) so that I do not destroy My House and Shrine…And when it was destroyed He (would) pray in it, May it be favourable before Me (yehi ratzon milefanai) that My children repent so that I may hasten the building of My House and Shrine. Indeed, yehi be-shalem sukko.

In this exposition, the Temple is projected back to the creation and God prays for the obedience of His children—so that His Shrine should not be destroyed. The term kiv. marks the transition betwen the prologue of the primordial Temple and God's ongoing prayer. Presumably, then, va-yehi be-shalem sukko was triply instructive. In the first instance, the lemma was taken to mean ‘His booth (Temple) was in Shalem (i.e. Jerusalem)’—from the beginning of creation. However, on the basis of the two cases of God's prayerful petition (yehi), we may understand that the lemma was understood as stating God's hope that His Temple ‘be’ whole (shalem) in Zion—namely, that it remain complete, and not be destroyed; and (after the disaster) that it become complete once again. In these two cases, va-yehi would have been construed as the optative: viyhi, ‘may it (the Temple) be’. In the first case the appeal invokes obedience, lest the Temple be destroyed; in the second it invokes repentance, that the Temple be rebuilt.

Hence the power of the exposition is not only the divine prayer itself, but the reinterpretation of the lemma as a divine prayer. The term kiv. marks this exegetical presumption.

The version of R. Berekhia's exposition of Ps. 76: 3 in GenR 56.10 (T–A, 608) is abbreviated and adapted to other purposes. In it, R. Berekhia (A5) taught in the name of R. ḥelbo (A4) that God built a booth in Shalem in the time of Abraham, and prayed therein that this patriarch might see the building of the Temple. The daring thrust of the MidPs version does not occur in this case, and the lemma is given an unremarkable application. The term kiv. is not used.

2. God and Moses

(i) DeutR 3.15. The subject is Moses' entreaty of divine favour after the sin of the Golden Calf (Exod. 32: 11). Moses is portrayed as telling God that he knows of His love for Israel and desire for someone to speak on their behalf. R. Simeon (ben Pazzi? A3) offers a parable of a king who argued with his son and cried out, ‘haniḥah li (leave me be) that I may kill my son’—but really wanted someone to intercede on the son's behalf. This is applied to Israel's sin and God's wrath.

Thus the Holy One, blessed be He, said to Moses: ‘haniḥah li that I become angry against them and destroy them’ (Exod. 32: 10). Whereupon Moses thought, Am I holding (tofes) onto the hand of the Holy One, blessed be He, that He said ‘leave Me be’?—kivyakhol He is requesting someone to speak on their (Israel's) behalf. Hence, immediately thereafter, ‘Moses entreated YHWH’ (ibid., v. 11).

This striking homily tries to contextualize the sequence of actions in vv. 10–11. The divine divine cry haniḥah li is deemed puzzling to R. Simeon. Why would God request Moses to let Him destroy Israel before Moses did anything to prevent this? (In the homily, this point is thematized by Moses' deliberation when he construes the phrase (p.377) haniḥah li as read concretely—as if God asked Moses to leave Him alone!) Moses thus draws the conclusion that haniḥah li is, in fact, a divine request whereby God says: ‘Assuage Me’ or ‘Calm Me down (by speaking on the people's behalf) lest I get angry, etc.’ The term kiv. marks this presumptive reinterpretation of God's command as a bold request. Moses' supplication follows.

DtRL ͑Eqeb 1.15. (SL, p. 90) has the same tradition.

In BT Ber. 32a, R. Abbahu (A3) interprets haniḥah li to be God's command to Moses to ‘let go’ of him, since Moses was ‘holding’ (tofes) onto Him as one holds onto a garment. (The term kiv. is not used here, since the contextual meaning of the term is not transformed, but only read more concretely.) In DeutR 3.15 R. Simeon uses this motif, but rejects it in favour of a request by God for Moses to calm His wrath.

C. Human Action and Divine Effect

1. Doing God's Will (General)

a. Affects divine protection (i) MdRI Beshalaḥ 5 (H–R, 134). An anonymous interpretation of Exod. 15: 6, where the phrase ‘Your right hand (yeminkha), O YHWH’ is followed by a second reference to God's ‘right hand’ in the next bi-colon. This is explained:

When Israel does (͑osin) the will of the Omnipresent,158 they make (͑osin) the left (hand) a right (hand); as is stated, ‘yeminkha YHWH, yeminkha YHWH’ twice. But when Israel does not do the will of the Omnipresent, kivyakhol they make the right (hand) a left (hand); as it says, ‘He put His right hand (yemino) behind Him’ (Lam. 2: 3).

Doing the commands of God affects His protection of Israel. This is part of a varied theology of the powerful effects of human action.159 The term kiv. only appears in this teaching to mark the negative pole, as is common in midrashic teachings of this type.160 From a hermeneutical standpoint, the second proof-text is more presumptive than the first, and this may explain the occurrence of kiv. here. However, this term is not well attested in texts and MSS (> MH, Ox 2, YalShim, and MdRS (E–M, 84) ). It is also frequently absent in the other examples of this type (see below). Tan Beshalaḥ 15 is like MdRI.

(ii) ibid. (H–R, 134). A teaching of the preceding type. It states that when Israel obeys God, He does not sleep (proved by Ps. 121: 10); but when they do not,

kivyakhol He does sleep, as it says, ‘YHWH was aroused like one who had been sleeping’ (Ps. 78: 65).

The term kiv. occurs only in the negative position, which is also the one with the bolder and more presumptive exegesis (the positive position simply reaffirms Scripture). However, the term is missing in many MSS, and is thus an unstable element. It may therefore be used here as a euphemistic qualification of God's inactivity.

(p.378) (iii) Midrash Panim ͗Aḥeirim (B) (B, 74).161 The reference to the ‘sleep of the king’ in Est. 6: 1 is identified allegorically with ‘the sleep of the Supreme King of kings, the Holy One, blessed be He’.162 And on the substantive issue, the query is posed:

And is there sleep on High? Rather (understand it thus): When Israel sins, kivyakhol He makes Himself like one who sleeps; as it says, ‘Arise, why do You sleep?’ (Ps. 44: 24). But when Israel does the will of the Omnipresent, ‘Behold, the Guardian of Israel neither slumbers nor sleeps’ (Ps. 121: 4).

Like the preceding case, this teaching sees the ‘sleep’ topos in conjunction with Israel's observance of the commandments. As in other cases of this type, the term kiv. is in the negative position163 though in this instance it occurs first, not second. However, the fit between the topos and the proof-text in the first part is weak; a citation from Ps. 78: 65 (with its simile, ‘like one who has been sleeping’ would have been more fitting, given the statement that God makes Himself ‘like one who sleeps’ (see (ii) above).164

(iv) SifNum 157, Maṭṭot (H, 211). Another teaching of the previous type. It states that when Israel obeys, God fights for them;165 but when they do not,

kivyakhol He fights against them; as it says, ‘He turned into their enemy, He fought against them’ (Isa. 63: 10).

The term kiv. is found only in the second position;166 and it may be a scribal addition here, since it is hermeneutically extraneous. Scripture makes the case explicitly—even stating that the change is due to disobedience. The term kiv. does not occur in MdRI Beshallaḥ 5 and Tan Beshallaḥ 15

The idea that whenever Israel does God's will, He aids them, but when they do not, He is their enemy, is presented in SifNum 157 as a theological principle (kelal) and adduced in connection with the lemma, ‘Yet they (the females) are the very ones who, at the bidding of balaam’, led the people astray (Num. 31: 16). According to the homily, balaam counselled the Midianites to lead the Israelites into perversity as the only way to destroy them—for he said that if Israel sinned their God would be angry and punish them Himself. After the principle, the midrash repeats the point with another proof-text: ‘Moreover (if Israel disobeys God), they make the Merciful One cruel; as it says, “YHWH became an enemy and destroyed (billa͑) Israel” (Lam. 2: 5). This verse alludes to the opening lemma from Num. 31: 16, and suggests that the name balaam (Hebrew, bil͑am) was interpreted as an allusion to the fact that it was God Himself who ‘destroyed them’ (bille͑am; i.e. the Israelites) for their sins.

This second proof (and the summary principle) was carried over to the list in MdRI Beshalaḥ 5, where it seems extraneous.167 Hence the entire unit about God becoming (p.379) an enemy is probably secondary there, and its primary example concerning God's right hand in MdRI (C.1a(i) ) was subsequently supplemented by an anthology of similar types.

(v) MdRI Beshalaḥ 4 (H–R, 130). On the lemma ‘YHWH is a man of war’ (Exod. 15: 3), the midrash compares a human warrior to God.

With respect to a (human) warrior (gibbor) in the province—when an arrow departs from his hand, he cannot retrieve it; but the Holy One, blessed be He, is otherwise. When Israel does not do His will—kivyakhol a decree departs from Him; as is said, ‘When I whet168 (shannoti) My flashing sword’ (Deut. 32: 41). (But if) they repent, I immediately retrieve it; as is said, ‘My hand lays hold of judgement’ (ibid.).

This teaching utilizes an analogy in order to laud God, who is both just and merciful—in response to human behaviour. As is common in this midrashic type, observance and non-observance are juxtaposed, with the term kiv. in the negative position (here given first).169 The midrash utilizes the theme of the sword of justice as a metaphor for a decree (gezeirah; something ‘cut’); and most probably construes shannoti as meaning God's ‘change’ from mercy to judgement.170 The second phrase is correspondingly also given a new sense. According to the plain sense of the passage, the two stichoi comprise a complementary parallelism. In the midrash, the second one is interpreted as denoting God's return to mercy and restraint.171 Altogether, the biblical image of God as a divine warrior is thoroughly transformed.

The term kiv. marks the allegorical reading of Deut. 32: 41. It does not occur in MdRS (E–M, 82).

b. Affects divine protection or withdrawal (i) SifDeut 355 Ve-Zo͗t (F, 422). An anonymous tannaitic midrash deals with Deut. 33: 26, ‘There is none like the God of Jeshurun,172 Who rides through the heavens to help you, through the skies in His majesty.’ In the teaching, the last two clauses are not construed as a synonymous parallelism, lauding the mighty warrior Lord in heaven, but as two aspects of divine providence which respond to Israelite behaviour.

When Israel is upright173 and does the will of the Omnipresent, ‘He rides through the heavens to help you’; but when they do not do His will, kivyakhol ‘and in His majesty, (the) skies’.

The term kiv. occurs only in the second part of this rhetorical figure, which regularly indicates a negative consequence.174 The first part uses Scripture in a straightforward manner, whereas the second part is the product of exegesis. In it, the reference to God's ‘majesty’ (ga͗avato) is boldly construed as Israel's rebellious ‘pride’, which results in God's withdrawal to the heights of heaven as an act of (p.380) punishment and rejection.175 The term kiv. marks this hermeneutical presumption; otherwise, its function is opaque.176

c. Affects divine strength or weakness (i) LamR i.6, 33 (D, Vilna). Two teachings are given for Lam. 1: 6, ‘And they went without strength (belo͗ koaḥ) before the pursuer’.

(1) R. Azariah said in the name of R. Judah b. R. Simon: When Israel does the will of the Omnipresent (Maqom), they increase (mosifin) strength (koaḥ) in the dynamis (geburah) on high;177 as it is said, ‘We shall make strength (ḥayil) in God’ (Ps. 60: 14). But when Israel does not do the will of the Omnipresent, kivyakhol they weaken (metishin) the great strength (koaḥ gadol) on High; as it is said, ‘You have forgotten (teshi) the Rock who bore you’ (Deut. 32: 18).178

(2) R. Judah b. R. Simon said in the name of R. Levi b. R. Ṭarfon: When Israel does the will of the Holy One, blessed be He, they increase strength in the dynamis on High; as it is said, ‘And now, let the strength (koaḥ) of YHWH be made great’ (Num. 14: 17). But when Israel does not do the will of the Holy One, blessed be He, kivyakhol they weaken the great strength on High, and they also go ‘without strength before the pursuer’.

In these two similar traditions about increase or decrease of the divine geburah, the term kiv. occurs only in connection with the second, negative consequence179—even though the positive positions are equally bold. It should be noted that the proof-texts in positive positions presume no midrashic changes. By contrast, the negative consequences in (1) presume the reading teshi (from the stem nšy, ‘forget’) as if it were from twš (‘weaken’, as in metishin);180 hence in this instance kiv. marks an exegetical presumption. Oddly, the term kiv. in (2) appears separated from the proof-text, which is now adduced as a plain sense reference to the flight of Israel (Lam. 1: 6). This doubling of the negative consequences (for God and for Israel), and the appearance of the term kiv. only in connection with the divine effect (note that the second consequence is set off by a ‘they also’ clause), gives the teaching about divine weakening an apparent euphemistic dimension.

These two traditions are also found in PdRK 25.1 Seliḥot (BM, 380) with various differences. One will note that the teaching in PdRK 25.1 (1) and LamR 1.6 (1) have the same tradents as LamR (1), but that the PdRK unit focuses on ‘the righteous’ (not Israel), and uses Num. 14: 17 as the proof-text.181 In addition, teaching (2) in PdRK gives a variant version of the father of R. Levi (referred to here as Parṭa͗),182 and uses (p.381) Ps. 60: 14 as the proof-text in the positive first part. These are not substantial differences. By contrast, it is notable that PdRK (2) cites Lam. 1: 6 solely in connection with a decrease in divine strength. This results in a different conclusion from the one noted earlier in LamR. Rather than suggesting different consequences for God and Israel, one is led to the conclusion that ‘Israel went before the pursuer’ into exile because there was ‘no strength’ in the divine dynamis to protect her. If this is correct, the term kiv. in PdRK (2) marks this strong midrashic presumption about God's impotence. By contrast, LamR (2) shows a weakened use of the proof-text—linking it solely to Israel's exile.183

The overall relationship between LamR and PdRK is fairly complex, and leads to ambivalent and complicated conclusions.184 The present instance is a case in point, and it is arguably best to regard the two pair of traditions as two variations of a common type—with each one evincing primary exegetical features along with secondary scribal adjustments. For example, among its primary features LamR 1.6 has two symmetrical traditions about Israel's actions; and preserves the older DN Maqom (Omnipresent) in (1).185 On the other hand, the formula koaḥ geburah (strength of the dynamis) has been expanded by the word ‘on High’ in (1) and (2); and the conclusion to (2) refers to both God and Israel, thereby breaking the symmetry of the rhetorical structure (dealing with the effects of human behaviour upon God).

For its part, PdRK 25.1 has a symmetrical structure in the second portion; and does not have the word ‘on High’ in either teaching. These are its primary features. On the other hand, teachings (1) and (2) evince asymmetrical elements, insofar as (1) speaks only about the deeds of the righteous and (2) only about those of Israel. This variation is undoubtedly due to the impact of the opening rhetorical sequence of the pericope, which refers to the triad Moses—the Righteous—Israel. Were PdRK the primary version overall, the reference to the righteous would presumably have been carried over to LamR; moreover, the conclusion to PdRK speaks about exile and redemption—a topic that is irrelevant to its purposes, but is entirely befitting LamR. One may therefore conclude that a version like that preserved in LamR influenced the incorporation of the exilic features now found in PdRK, but without either its stylistic expansions or its theological softening.

R. Azariah is A5; R. Judah b. R. Simeon (ben Pazzi) is A4. No R. Levi with either patronymic is known. If the name Ṭarfon is original, it is first attested by the Ṭanna R. Ṭarfon (T2) and possibly related to the Greek name Tryphon.

d. Affects divine realm in heaven (i) SifDeut 346, Zo͗t (F, 403). A comment on the phrase ‘together (yaḥad) the tribes of Israel’ (Deut. 33: 5) evokes a comment on the benefit of national unity, which by implication establishes God as King in Jeshurum (the beginning of the lemma). The comment uses the phrase ͗aguddah ͗aḥat (‘one band’), and this language is correlated with Amos 9: 6, ‘Who (God) builds His (p.382) chambers in heaven and establishes His vaults (͗aguddato) on earth’186—presumably to make the same point. Rabbi Simeon bar Yochai (T3) then explicates the theological import of this verse (and its reiteration of the effect of national unity upon God) by a parable about how separate boats can provide the basis for a residence to be built upon them, if they are lashed together. This parable is then applied it to Israel:

When they do the will of the Omnipresent, He builds His upper chambers in heaven; but if they do not do His will, kivyakhol ‘He establishes His vaults on earth.’

This application is puzzling, since the second half of Amos 9: 6 should make a negative point—as is common in this rhetorical type, where the term kiv. appears in the second position.187 R. David Pardo tellingly observed that the language here is ‘garbled’, and went on to reconstruct the meaning on the basis of the opening (anonymous) comment on Deut. 33: 5. There the ideal of national unity is referred to by the noun ͗aguddah (‘group, band’) and divisiveness by its plural form ͗aguddot.188 The teaching would thus be that when Israel collectively does God's will, God's realm is exalted above,189 but not when the nation is religiously disobedient (and fragmented).

Following Pardo's lead, we may say that the scriptural word ͗aguddato (in Amos 9: 6) was understood to allude to this negative effect. On this basis, the term kiv. marks the midrashic presumption in which ͗aguddato was construed to refer to disobedient ‘divisions’ or ‘groups’ separate from God on earth. Negative or divisive behaviour by Israel was thus presumed to impede the establishment of God's realm (kingship) on high.190

In both Tan Beha͑alotkha 11.2 and MidSam 5.15 (B, 62) only the positive first part is given, regarding Israel's obedience. The term kiv. is found in both phrases, but now functions merely as a euphemism regarding the effects of Israelite action upon God's upper realm.

(ii) Ibid. (F, 404). At the end of a series of proofs connected to R. Simeon's teaching (above, d(i) ), the establishment of the divine throne in heaven is also linked to human action (unspecified, but contextually joined to doing the will of God). The teaching states:

‘I have raised my eyes to You, who dwell in the heavens (ha-yoshbi ba-shamayim)’ (Ps. 123: 1). Were it not for me, kivyakhol You would not dwell (yosheb) in the heavens.

The relationship between the plain sense and its midrashic transformation is unclear. Presumably, the shift from a human acknowledgement of God's heavenly transcendence to a constitutive act of divine enthronement is based on construing the old genitive ending -i in ha-yoshbi (who dwell) as a self-referential pronoun suffix (p.383) uttered by the speaker, such that it means ‘who dwell on my account’ (namely, God ‘is enthroned’ in heaven ‘through my agency’). The term kiv. marks the exegetical explication of the genitive ending in the lemma as the pronoun ‘me’ (͗ani), and with it the great presumption that affects divine sovereignty.191

(iii) MZ Songs 1.1 (B, 9). In a discussion of songs recited by the angelic host, it is said that the angels spoke of God's glory (in Isa. 6: 8) in order to indicate (inter alia) that:

If His nation does not enthrone Him on earth, kivyakhol He has no kingdom in the highest heavens.

The passage goes on to indicate that when Israel recites the Shema͑-prayer below, proclaiming divine sovereignty there (Deut. 6: 4), they give the angels permission to proclaim His sovereignty above (Isa. 6: 2).192 This is a lapidary use of the term kiv., transparently euphemistic, and does not mark any exegetical feature.

MZ Songs 3.11 (B, 30) gives a more expansive version of this topos (also with the term kiv.). Just prior to that reference, Songs 3: 11 is used to introduce a lapidary formulation of crowning God with the crown of kingship (also with kiv.). See similarly in AgSongs 3.1 (SS, 34).

e. Affects the divine Name (i) MdRI, Beshalaḥ 3 (H–R, 128). Commenting on Exod. 15: 3, R. Simeon b. Eleazar (T4) taught:

When Israel does the will of the Omnipresent, His Name becomes great (mitgaddel) in the world; as it says…(Josh. 5: 1 and 2: 10–11 are cited, respectively); but when they do not do His will—kivyakhol His Name is desecrated (mitḥallel) in the world; as it is said, ‘But when they came to these nations,193 they caused My holy Name to be desecrated (vayiḥallelu)…‘ (Ezek. 36: 20–3).

This is a tannaitic teaching of the ‘When Israel does the will’ type, with positive-negative valences; and as is common to this form, only the negative implication has kiv.194 However, in this instance the proof-text from Ezek. 36: 20–3 states the point of desecration explicitly and no midrashic interpretation is involved. Hence, the term kiv. qualifies neither the content nor the exegesis; it was presumably added to conform to the general rhetorical type.195 Notably, kiv. is not found in Ox2, M2, and YalShim; nor in MdRS (E–M, 80). The topic of enhancement and desecration of the divine Name focuses here on God's reputation. This is different from the substantive effect upon the divine Name found elsewhere (see 2a below).

2. Doing God's Will (Specific Acts)

a. Affects the divine Name (i) LevR 23.12 Aḥarei Mot (MM, 548). At the conclusion of a homily dealing with the effect of adultery on the changed characteristics of (p.384) an embryo, a supplement by R. Isaac considers the effect on God.196 In one version, that effect is not on God's image in the embryo, but on the divine Name itself, and the minuscule letter yod in the verb teshi (tšy) in the scriptural lemma (Deut. 32: 18) marks this diminution. See the analysis above (pp. 188–9).197 The term kiv. introduces the reference to this divine effect. For a different midrash on the incompleteness of the divine Name because of evil, see 3a(i) below.

b. Affects divine power (i) Tan, Naso͗ 4. In this version of the impact of adultery on the embryo, the teaching is given in the name of R. Isaac;198 and the effect is on the image of God in the unborn child (whose face is changed from that of the woman's husband to that of her paramour). The teaching opens with the forceful assertion that the adulterer weakens ‘the power (koaḥ) of the Shekhinah’. This point is then developed. The proof-text from Deut. 32: 18 is introduced by kiv., so that the meaning of the phrase tzur yeladekha teshi is not, literally, ‘You have forgotten (teshi) the Rock that bore you’, but:

A minuscule yod (occurs, indicating that) you have weakened (tashash) the hands (yedei) of the Artisan.

In this case, God (the Rock, tzur) is interpreted as the Artisan (tzayyar) and the act of Israel's forgetting is midrashically transformed into a weakening of the hands of God. The term kiv. marks this midrashic presumption, which includes understanding the minuscule letter yod as signalling a weakening of divine power (‘the hands (yedei) of God’) because of adultery.199

Another version of the act of divine transformation occurs at the end, in the name of R. Abbahu (A3). Hereby there is a pointed reference to the changing of the ‘signs’ (sammanim) of the embryo's appearance:

What does the Holy One, blessed be He, do?—kivyakhol He goes back and changes the original form of the husband's image into the image of the adulterer. Indeed, ‘He hides (seiter) his face’ (Job 24: 15).

The term kiv. marks the general theme of divine transformation together with its proof-text. Specifically, a verse stating that an adulterer has masked or hidden his face by an act of nightly stealth is midrashically construed to suggest that ‘He (God) has changed (satar) his face (i.e. of the embryo).’ Alternatively, He has changed the image of the embryo into the face of the adulterer. Either way, kiv. underscores the bold exegetical presumption. The first half of Job 24: 15 (about the adulterer at twilight) also occurs at the beginning of LevR 23.12, but in this version the second half of the verse is used to refer to God who dwells in ‘the hiddenness’ of the world and sees all acts. Hence the passage is thus construed to mean that ‘He who dwells in hiddenness puts His face’ (namely, providential regard) to observe the acts of humans. The term kiv. does not occur with this more routine theology of divine omniscience.200

(p.385) c. Affects the covenantal relationship (i) LevR 23.9 ͗Aḥarei Mot (MM, 538). Basing himself on a digest of Lev. 18: 2–3,

R. Ishmael taught, ‘Do not follow the practices of the land of Egypt or land of Canaan’—‘I (͗ani) am YHWH, your God’; otherwise, kivyakhol I am not (͗aini) YHWH, your God.

In this formulation of Scripture, R. Ishmael (T2) condenses v. 3 (‘Do not follow’, etc.) and shifts the divine formula (‘I am YHWH, your God’) from before the exhortation in v. 2 to its conclusion. This transposition serves a rhetorical end, and allows R. Ishmael to contrast the positive ‘I am’ (͗ani) with the counterpoint, ‘I am not’ (͗aini). The midrashically revised formula is preceded by the term kiv., which marks both this exegetical innovation and its bold theological implication.201

(ii) SifDeut 346, Ve-Zot (F, 403 f.). In a series of teachings linked to an opening midrash attributed to R. Simeon b. Yochai (T3) on Amos 9: 6,202 and dealing with the overall theme of the effect of Israel's behaviour and convenantal commitment upon God, we read:

‘You are My witnesses—declares YHWH—and I am (ve-͗ani) God’ (Isa. 43: 12). When you are My witnesses I am (͗ani) God; but when you are not My witnesses, kivyakhol I am not (͗aini) God.

This teaching has rhetorical and exegetical features similar to (i) (above). Here also the term kiv. marks the homiletical point. This extraordinary midrash is also found in PdRK 12.6, Ba-ḥodesh (BM, 208), where it is explicitly linked to the Tanna R. Simeon b. Yochai. However, the final phrase, ‘kiv. I am not (͗ain ͗ani) God’ obscures the exegetical point and daring evident in SifDeut 346. The result is that kiv. appears to serve a more euphemistic function. The term is used in the negative position.203 Theologically speaking, this teaching marks a limit case for God's dependence upon humans.

(iii) SongsR 1.4. In a homily on Torah and its effects, Hos. 4: 6 is explained in a striking way.

‘Because you have forgotten (va-tishkaḥ) the Torah of your God, I too (͗af ͗ani) shall forget (͗eshkaḥ) your children’ (Hos. 4: 6). Said R. Aḥa: ‘I, too’—kivyakhol I too am forgotten (be-shikheḥah).

In R. Aḥa's transformative exegesis, the result clause is not God's action against Israel, but His being forgotten by them. He puts his emphasis on the finals words ͗af (p.386) ͗ani, and indicates his hermeneutical intent by using be-shikheḥah to explain ͗eshkhaḥ. This latter was thus presumably construed as a niph͑al reflexive (i.e. ͗ishakaḥ, be forgotten). In the version of this exegesis found in MidPs 8.4 according to the ed. prin., (Venice, 1546), and Ro1, the teaching is attributed to R. ḥiyya and the exegetical point is clarified: ‘Kivyakhol, I also am forgotten (mishtakeaḥ), for the children used to bless Me.’ In the Buber edition of MidPs 8.4 (B, 77), this same phrase occurs; but instead of a named sage, it says, ‘the Holy One, blessed be He, said’. The comments in MidPs appear to be epexegetical and secondary. In all cases, the term kiv. marks the bold transfer of the effect from Israel to God, mediating between the proof-text and the comment.204

The names R. Aḥa and R. ḥiyya are undoubtedly scribal variations but it is difficult to assert which is original, or which Aḥa or ḥiyya is meant.

(iv) Tos ͑Abodah Zarah 5.5 (Z, 466). Quoting Lev. 25: 38, ‘To give you the land of Canaan, to be for you as a God’, the following rhetorical assertion is made:

Whenever you are in the land of Canaan, I am a God for you; (but whenever you are) not in the land of Canaan, kivyakhol I am not (͗aini) a God for you.

This unit exploits the two phrases of the lemma, construed in conditional terms. Indeed, the idiomatic phrase lihyot lakhem lei͗lohim (‘to be your God’) is now read in concrete terms which stress the resultative nature of the phrase. The word kiv. marks the inversion and qualifies the negative implication. No special exegetical aspect is adduced at this point; and kiv. evidently serves only a euphemistic function.205

d. Affects the divine attributes (i) NumR Shelaḥ 17.3. In a homily Deut 1: 45 is cited and explicated.

‘Again you wept before YHWH, but YHWH would not heed your cry’ (Deut. 1: 45)—kivyakhol, you have made the Attribute of Judgement as if it were (ke͗ilu) cruel. R. Samuel b. Naḥman said that they made the Attribute of Judgement as if it were cruel.

This teaching, given twice, reacts to the failure of supplication to move God. The remark is apparently based on the fact that the divine name YHWH is used here, and although it normally denotes God's merciful presence, no mercy occurs in this proclamation. This example is one of several cases where divine judgement occurs with the Tetragram and is explicated by R. Samuel b. Naḥman as denoting a change in divine providence. See discussion above (pp. 183–4). The term kiv. seems to qualify the theological assertion, and is likely a secondary addition.

e. Affects divine judgement (i) PdRK 5.13 Ha-ḥodesh Ha-Zeh (BM, 102 f.). In a teaching of R. Hosha͑yah, reference is made to the determination of a human court regarding the dates of Rosh Hashanah—a decision is accepted by God Himself, who then convenes His court. A reason is proposed on the basis of Scripture:

‘For it is a law for Israel, a judgement (mishpaṭ) for the God of Jacob’ (Ps. 81: 5). If it is not a law for Israel—kivyakhol there is no judgement for the God of Jacob.

(p.387) The striking exegetical point of R. Hosha͑yah turns on transforming the positive parallel clauses of the lemma into a conditional sequence. The sage does not do this directly, but by rereading the clauses as a negative condition followed by a negative result. The term kiv. marks the bold theological implications of his reinterpretation—calling attention to the implications if Israel does not determine the date of the holiday. Indeed, the sage's ‘reason’ goes so far as to suggest that if Israel does not actually establish the date, God does not bring His judgement upon the world. The implications clearly exceed the point that human calendration is decisive for the Jewish festival year.206

This teaching also occurs in JT Rosh Hashanah 1.3. R. Hosha͑yah is the teacher here as well. He is presumably R. Hosha͑yah Rabba (A1, even though the teaching in PdRK opens with teni, a verb that regularly marks a tannaitic tradition). The version in MidPs 4.4 (B, 43 f.) is an elaborate rendition; and although Ps. 81: 5 is adduced, the negative implication is not, and thus kiv. is also absent. In this form, the theology is quite unremarkable.

f. Divine augmentation (i) BT Baba Batra 10a. In a succinct teaching, R. Yoḥanan (A2) asks, ‘What is the meaning of malveh YHWH ḥonen dal (Prov. 19: 17)?’ He answers thus:

If it were not written in Scripture, one could not say it—kivyakhol ‘And the borrower is slave to the lender (͗ish malveh)’ (Prov. 22: 7).

The teaching seems opaque. It answers the meaning of one verse with another, which it presumes to be clear. But this is hardly the case. The relationship between the passages is all the more puzzling if Prov. 19: 17 simply means that ‘YHWH lends to one who is generous to the poor’, as one might assume from the parallel phrase (namely, ‘and his good deed He will repay him’).

It would appear that R. Yoḥanan intends a bolder point, and even takes Prov. 19: 17 as saying that, ‘One who is generous to the poor person makes a loan to YHWH’. For if he merely meant the plain sense, what is the purpose of the rhetorical question? This suspicion is strengthened insofar as the sage adduced a bold qualifier (‘If it were not written’) before the seemingly innocuous citation Prov. 22: 7—which is is offered to explain the first passage! On this logic, we may assume that R. Yoḥanan rereads the second verse to mean, paradoxically, that ‘a slave makes a loan to a lender (i.e. God)’. Only such a theological presumption would allow one to bring the two passages into alignment, which is R. Yoḥanan's exegetical purpose. The term kiv. marks the bold exegetical presumption offered here about divine augmentation.

In LevR 34.2 (MM, 774 f.), R. Yoḥanan's brother-in-law, R. Simeon b. Lakish (A2) offered a similar interpretation of Prov. 19: 17. However, in the tradition of his teaching presented there, he simply says, ‘it is the (normal) way for the borrower to become a slave of the lender’, and then adduces Prov. 22: 7. While R. Simeon's explication is both without the clause ‘If it were not written, etc.’ and the term kiv., it is nevertheless clear from his comment that he is inverting the sense of Prov. 22: 7—and thus also rereading Prov. 19: 17 in a radical manner. Hence R. Simeon's laconic version confirms the foregoing explanation of R. Yoḥanan's exegesis, although it completely conceals the bold reinterpretation. It is not hard to assume that this (p.388) condensation is the work of an editor, who fully perceived the radical force of R. Simeon's remark.

Both Tan Mishpaṭim 15 and TanB Mishpaṭim 6 (B, 85) are later reflexes of the teachings just noted. In the first case, R. Tanḥuma cites Prov. 19: 17 and clarifies it with the remark: ‘Kivyakhol, he makes a loan to YHWH’; in the second, a fuller comment is given: ‘Whosoever lends to a poor person, kivyakhol it is as if (ke͑ilu) he makes a loan to the Holy One, blessed be He.’ In both instances, the term kiv. marks an exegetical reuse of Prov. 19: 17. But given the absence of Prov. 22: 17 in these cases, the teachings (especially the second) tend to have a more rhetorical character.

(ii) MZ Songs 1.15 (B, 21). Applying the opening words of the verse to acts of charity, the word ra͑ayati (‘my darling’) is identified with the ‘community leaders’ (parnasim). This leads to a general point:

When you sustain (mefarnesim) the poor, I deem it favourable on your behalf, kivyakhol (as if it were possible) you are sustaining Me.

The term kiv. is used in a lapidary way here, as part of the formulation itself. The strong exhortation is thus qualified with regard to the divine effect; see similarly in AgSongs 1.15 (SS, 27 f.).

There is no exegetical justification for this bold teaching. The one that follows is instructive in this regard, and shows how the point could be developed. The term kiv. does not occur, but the teaching is itself qualified at the end.

‘Behold you are fair, my darling’ (Songs 1: 15). The Holy One, blessed be He, says to Israel: You are sustaining Me; as it said, ‘My sacrifice and My food’ (Num. 28: 2). Is it possible (yakhol) that He eats and drinks? Scripture therefore says le-͗ishai (lit., ‘for My fire-offerings’), (to teach that) you (should) give (gifts) to persons (le-͗ishim). If so, why is ‘My food’ stated?—(To teach that) though you give to persons, I regard (or: credit) you as sons who sustain their father.207

3. Existence of Amalek

a. Affects the divine presence (i) PdRK 3.16, Zakhor (BM, 53). In connection with Exod. 17: 16, regarding the existence of Amalek,

R. Berekhia (said) in the name of R. Abba bar Kahana: Whenever the seed of Amalek exists in the world—kivyakhol as if (ke͗ilu), a wing (kenaf) covers the divine Face; (but when) the seed of Amalek is destroyed from the world, ‘then your Teacher will no longer be concealed (yikkanef), and your eyes shall see your Teacher’ (Isa. 30: 20).

The evil deeds of Amalek result in the occultation of God's presence, presumably by a wing of the seraphic host (cf. Isa. 6: 1–20); but in the future, when the evil nation will be destroyed, this situation will change. Then all will see God—called here ‘your Teacher’ or Guide. The assertion of covering, as with a wing, anticipates the proof-text that refers to the end of God's concealment (indicated by a verb that alludes to the covering of a wing). The term kiv. marks this exegetical instruction overall; though, in (p.389) point of fact, it is not directly conjoined to the proof-text from Isa. 30: 20, which marks the positive pole of the teaching. Indeed, such a placement gives the term a euphemistic tenor. In consequence, other tradents felt that further euphemistic augmentation was necessary; hence the term ke͗ilu. Both kiv. and ke͗ilu recur in the versions found in PesR 12 Zakhor (MIS, 51a) and TanB Ki Tetze͗ 18.M

R. Abba b. Kahana is A3; R. Berekhia is A5.208

b. Affects the divine throne and Name For this teaching and the interpretation of scriptural orthography, see above (pp. 189–90). The term kiv. occurs only in PesR 12 Zakhor (MIS, 51a),209 but not in parallel passages (see a(i) ). It is apparently secondary and added for symmetry with the previous instruction (a(i) ).

4. Existence of the Wicked

a. Affects the divine throne (i) YalShim §754. Using the standard trope, the following teaching is given in the name of R. Mesharshi͗a:

Whenever the wicked exist in the world—kivyakhol, the Holy One, blessed be He, does not sit upon His throne.

This point is proved by proof-texts from Hag. 2: 25 and Dan. 7: 9. But these accounts of an eschatological upheaval of the thrones of the nations do not fit the point. This appears to be a weak use of the topos.210

5. Curse Against Israel

a. Affects God (i) AgBer 66 (B, 131). In a striking teaching about the close relationship between God and Israel, it is reported in the name of R. Judah b. Pazzai that

When Balak sent for Balaam to curse Israel, Balaam said: What are you doing? If you want to injure (ligga͑) the Holy One, blessed be He, who is ‘with them’ (consider the analogy of)211 two persons who are bound (medubbaqim) one to another. If someone should smite one of them, is it not as if (ke͗ilu) both are smitten? (So, similarly) kivyakhol, the Holy One, blessed be He, is bound (medubbaq) to Israel. (Hence,) if I curse them, kivyakhol I injure (noge͑a) Him; as it is said, ‘How can I curse (mah ͗eqob) whom God has not cursed (lo͗ qabah)?’ (Num. 23: 8).

In this teaching, Balaam uses an analogy to suggest the intimate bond between God and Israel. The first occurrence of the term kiv. simply adduces a pious qualification to the bold notion. The second instance seems to function likewise. However, it will be noted that the plain sense of the proof-text does not seem to justify the rhetorical point made by Balaam; for according to the plain sense, one cannot curse Israel, the beloved of God. Hence one must assume that the verse was construed to mean: How could I curse (Israel) without (also) cursing God? This remarkable reading is signalled by the term kiv., which marks the theologoumenon and hints that the proof-text must be understood in accord with its point.

(p.390) This is a unique tradition. R. Judah b. Pazzi is R. Judah bar Simon (A 4), the son of Simeon b. Pazzi.

D. Human Denial of Divine Power

1. At the Conquest

(i) JT Ta͑anit 4.5, 68d. In the contexts of negative statements by the spies,

R. Simeon b. Lakish said, They spoke words against (God) on High: ‘For he (the enemy) is stronger mimmennu’ (Num. 13: 31). They said, kivyakhol He is not able (to defeat) them.

The formulation is compact. R. Simeon construes the phrase in Num. 13: 31 to indicate the spies' denial of divine power. He thus takes the pronoun suffix in mimmennu to refer to God (‘than Him’), not the spies themselves (‘than us’). The term kiv. marks this presumptive deviation from the plain sense (‘than us’), and the application of mimmennu to God. Indeed, this is its signal function; for were kiv. to be a euphemistic qualifier, the teaching of R. Simeon would make no sense.

In the two variants of this bold teaching found in BT Soṭah 35a and BT ͑Arakhin 15b, the exegetical device attributed here to R. ḥaninah b. Pappa is more clearly specified. After quoting from Num. 13: 31, it is stated: ‘Do not read mimmennu but mimmennukivyakhol, the master of the house is not able to remove his utensils from it.’ The effect of making the ͗al tiqre (do not read x) construction helps the reader see that nothing is changed in the text itself, but only that the pronoun suffix is construed differently.212 The statement about the impotence of the owner of the house (God) goes further than the laconic statement attributed to R. Simeon. The version of the teaching in NumR Masa͑ei 23 offers a conflation of the versions of R. Simeon and R. ḥaninah. It is anon.

R. Simeon b. Lakish is A2; R. ḥaninah b. Pappa is A3.

2. In Acts of Theft

(i) BT Baba Qama 79b. Discussed below (VI.C.2(i) ).

3. Near the Redemption

(i) BT Sanh. 97a. Among various comments offered concerning when the Messiah, son of David, would come, there is an interpretation of Deut. 32: 36 (‘YHWH will judge His people…When He sees that their might is gone, and neither bond nor freeman is left (ve-͗efes ͑atzur ve-͑azuv)’). This passage is construed to mean that the divine judgement will come when the people lose all hope in salvation. To reinforce the point the final phrase is invoked, but with a surprising twist:

Kivyakhol—there is neither a supporter nor helper for Israel.

Since the opening plain sense of the passage indicates that God will bring the salvation when hope is lost among persons, one must wonder what is indicated by this comment. Surely it cannot be that Israel has no warriors to help, since that would (p.391) make the comment redundant. Rather, the term kiv. alerts us to a more radical theological reading; namely, that the redemption will not come until the people themselves despair of God's salvation. This is the bold and presumptive interpretation of Scripture marked by the term kiv.213


A. Divine Rewards and Justice

1. Positive (Reward)

(i) ExodR 25.8 Beshalaḥ. Among a series of teachings on heavenly bounty, Isa. 33: 16 is variously adduced. The following is included:

‘He dwells on high (hu͗ meromim yishkon)’ (Isa. 33: 16). As it is written, ‘For YHWH, your God, brings you to the good and spacious land (͗aretz)’ (Deut. 8: 7),214 to see the table prepared in the Garden of Eden; as it says, ‘I shall walk before YHWH in the lands of the living’ (Ps. 116: 9)—kivyakhol, He sits above the partriarchs, and the patriarchs and all the righteous are in it (be-tokho); as it says, ‘and they followed at Your feet (tukku le-ragleykha)’ (Deut. 33: 3).215 And He distributes to them portions (from the Edenic table on High)…

In this teaching, Isa. 33: 16 operates on two levels. It first indicates the human being who shall dwell on high, in ‘Eden’. Such a reading extends the plain sense of the passage (vv. 15–16), in which the righteous are promised divine security—not destruction. Now the deserving ones may anticipate a heavenly dwelling—a point reinforced by an allegorical transposition of Deut. 8: 7 to refer to an entrance into a divine realm.216 However, the teaching abruptly shifts to a second level. The term kiv. marks a rereading of the initial lemma in terms of God's own supernal dwelling. The shift in subject is clearly indicated, since the first-mentioned righteous are now (along with the patriarchs) said to be ‘in it’—this being Eden—and below the one seated ‘on High’. However, the proof-text introduced to reinforce the ‘indwelling’ of the patriarchs and the righteous in Eden is surprising; for the citation from Deut. 33: 3 does not simply support the fact that these persons are ‘in’ Eden, but that they are at the feet of God. Presumably, the phrase ve-heim tukku le-ragleykha was construed to mean ‘they are in it at Your feet’ (reading tukku as tokho).217 In one stroke, therefore, both theological (p.392) assertions at the end (God dwells above; the patriarchs and righteous are below Him at His feet) are supported by the proof-text.218 At the same time, the image of God's dwelling in heaven also receives a new emphasis, since He is exalted high above those partaking of a heavenly reward.

The term kiv. marks the presumptive rereading of Isa. 33: 16 in terms of God's exalted dwelling in a heavenly paradise (Eden), where His faithful assemble at His feet for their reward.

(ii) ExodR 21.3 Beshalaḥ. The lemma from Exod. 14: 15 in which God asks Moses ‘Why do you cry out to Me?’ serves as the basis for teachings about God's response to humans in this world and the next. With respect to one ‘who does the will of the Omnipresent and directs his heart in prayer’, we learn that God answers them in this world and similarly rewards them in the next.

And (with respect to) in the world to come, ‘they speak (medabberim) and I shall hear’ (Isa. 65: 24). And what do they speak? Each stands and causes his study (talmudo) to be heard (mashmi͑a)—kivyakhol He sits and listens among them;219 as it is said, ‘Then (͗az) those who revere YHWH were talking with one another (nidberu); YHWH has paid attention and heard (va-yashma͑)’ (Mal. 3: 16)220

In this scenario God attends to the study of the righteous in heaven. This application of the scene of speaking to the world to come presumably derives from a construal of the word ͑od in the phrase ͑od heim medabberim (Isa. 65: 24) to mean ‘they shall yet (or again, in the future) speak’ (not: ‘while they are still speaking’).221 The term kiv. marks the shift from a prognosis of human speech in heaven to God's attention to heavenly study by humans, and marks the exegetical presumptions involved. Just as the particle ͑od is presumably the key to the use of Isa. 65: 24 as prophesying a future hearing by God of human speech, it appears that the particle ͗az triggers the application of Mal. 3: 16 to the world to come. This usage of the particle is a standard rabbinic trope, asserted most famously in the principle ‘In some instances ͗az indicates the past; in others, the future to come’ (MdRI Beshalaḥ 1).

The term kiv. does not occur with Isa. 65: 24, and its reinterpretation as a prophecy of future divine hearing; it is only found in connection with the more dramatic scenario of divine activity putatively supported by Mal. 3: 16.

2. Negative (punishment)

(i) MidPs 52.2 (B, 283). In a teaching against lashon ha-ra͑ (‘slander’), the principle is given that such speech leads to heresy. According to R. Yose, this point is proved by (the sequence of phrases in) Ps. 12: 15 (‘They say, “By our tongues we shall be mighty…who can be our lord?”’). The divine response follows: (p.393)

Kivyakhol, the Holy One, blessed be He, cries out against ‘the masters of the tongue’, ‘Who will arise with Me against the evildoers (merei͑im)?’ (Ps. 94: 15). Who is able to stand against them, and who will?—Gehinom. But Gehinom says, ‘I am not able to stand (against them).’ The Holy One, blessed be He, says to it: ‘I am above and you are below.’ And (Scripture) says, ‘The sharpened arrows of a warrior, with hot coals of broom-wood’ (Ps. 120: 4).

The passage is compact and dense. At the outset, God cries out in judgement against the slanderers. The term kiv. marks this transition and the presumptive application of Ps. 94: 15 to a divine response against those engaged in evil speech. God answers His own (rhetorical) query by noting Gehinom; but this embodiment of hell-fire claims insufficient power. The divine response and the subsequent proof-text leave the result opaque. The issue is clarified in BT ͑Arakhin 15b, where we are told that ‘R. ḥisda said in the name of Mar ͑Uqba, Whosever speaks slander, the Holy One, blessed be He, says (about him) to the Prince of Gehinom: I shall be against him above and you shall be against him below, (and together) we shall judge him.’222 Thereupon, Ps. 120: 4 is cited and allegorically applied to this topic: the arrows are slanders, the hero is God, and the hot coals are Gehinom. Presumably the word shenunim (sharpened) was midrashically construed to mean a ‘doubled (judgement)’. This would explain the formulation of the conclusion (we shall judge him); but nothing is stated in the received tradition.

Given this striking and expansive version of the tradition, it is clear that MidPs 52.2 means to indicate that God responds to Gehinom's refusal by indicating that the two of them would conjointly punish the slander. The citation from Ps. 120: 4 bears on this point, though not in any way that could be construed form this version alone.

MidPs 12.2 (B, 106) is an even more condensed and opaque version of this teaching (hence MidPs 52.2 was explicated first).223 Nevertheless, the citation from Ps. 2: 5 shows that this is the primary locale for the peculiar conflation of traditions kept separate in BT ͑Arakhin 15b (namely, the distinct teachings about the link between slander and heresy and the scenario of divine judgement). Moreover, after God's response to Gehinom (‘I am above’, etc.), the citation from Ps. 120: 14 given here is apocopated—ḥitzei gibbor shenunim. But this brief citation is sufficient to show that the verse was adduced to indicate a conjoint punishment. Read in the light of the Talmudic explication, these three words refer to the slanders, to God, and to the two powers that will bring punishment, respectively.

The teaching in BT ͑‰Arakhin 15b is not introduced by the query in Ps. 94: 15, and the term kiv. does not occur there. In the two versions in MidPs, the term marks the shift to a divine response to slanders, and thereby provides a bridge between distinct traditions.224

R. Yose (b. Zimra) is A1 (his tradent is R. Yoḥanan; A2); Mar ͑Uqba is BA2 (his tradent is ḥisda; BA3). MidPs thus combines EI and Babylonian traditions, preserved on the same page of the Talmud.

(p.394) (ii) MidPs 12.3 (B, 107). Commenting on Ps. 12: 5 (‘Because of the plundered poor and the groan of the needy, I shall now arise, says YHWH’),

R. Yudan said, The punishment for robbing the poor is greater than the sin of the generation of the Flood. Regarding the generation of the Flood, it is written, ‘YHWH sat (enthroned) at the Flood’ (Ps. 29: 10)—kivyakhol, He exacted punishment against them while seated; and hereby (it is written), ‘I shall now arise’, (meaning that) He does not exact punishment against them except while standing.

R. Yudan interprets Ps. 29: 10 to refer to the physical position of God during acts of judgement. He does this to emphasize God's arousal against those who harm the poor. The term kiv. marks the presumptive interpretation of Ps. 29: 10, and thus does not qualify the act of divine punishment per se.225

B. God and the Courts

1. Divine Warning to the Judges

(i) Tan Shofeṭim 7. In the context of the warning in Deut. 16: 19 (‘Do not bend judgement’), advice is given to the judges. Specifically, they are told to act with great trepidation, because

Kivyakhol, they judge on behalf of the Holy One, blessed be He; for just this did Jehoshephat say to the judges: ‘(Consider what you do,) for you judge not on behalf of man (le-͗adam), but le-YHWH (lit. on behalf of the Lord)’ (2 Chron. 19: 6).

At first glance this teaching seems to be a weak, lapidary use of the term kiv., which piously qualifies what is already stated explicitly in Scripture. Similarly, in TanB Shofeṭim 6. However, this concise formulation may conceal the more radical implication that the phrases le-͗adam and le-YHWH in the proof-text were construed to mean that the judges judge man and God Himself (not simply on His behalf). That this reading is not out of the question is indicated by the explicit teaching of R. ḥama b. R. ḥaninah (A2) that follows. Having 2 Chron. 19: 6 in mind, he says, ‘If it were not written in Scripture, one could not say it: one of flesh and blood (a human judge) judges his Creator. Hence the Holy One, blessed be He, said to the judges, exact trepidation upon yourselves as if you are judging Me.’

On this basis, the difference between the first anon. teaching and the second by R. ḥama is not one of content but of form, the second one being more epexegetical and specific. Non liquet.

2. Divine Presence and Absence

(i) ExodR 30.24 Mishpaṭim. In connection with Exod. 21: 1 (‘And these (ve-͗eilleh) are the laws’), the following is taught:

‘These also (gam ͗eilleh) are by the wise: to be partial in judgement is not good (bal ṭov)’ (Prov. 24: 23…What is bal ṭov?—When the judge sits and gives true judgement, kivyakhol, the Holy One, blessed be He, leaves the highest heaven and causes His Shekhinah to dwell near Him; as it says, ‘When YHWH raised up judges for them, YHWH was with the judge’ (Judg. 2: 18). And when He sees him being (p.395) partial, kivyakhol, He causes His Shekhinah to depart and ascend to heaven…as it says, ‘Because of the plundered poor and the groaning of the needy, I shall now arise, says YHWH’ (Ps. 12: 5).

This teaching opens with a verbal tally between Exod. 21: 1 and Prov. 24: 23 (the word ͗eilleh),226 and this serves as the basis for introducing the ensuing teaching against partiality. As to the meaning of bal ṭov, we first are told that God descends to be with the judge who administers truthfully. The term kiv. precedes this idea and the presumptive transformation of Judg. 2: 18. Correspondingly, we further learn that when there is partiality in judgement God leaves the judge and ascends. As in the first case, the term kiv. precedes this notion and the presumptive application of Ps. 12: 5.

But what is bal ṭov—in light of this teaching? It can hardly simply mean that partiality is ‘not good’, for this would be a banal instruction and not explain the central motif of divine presence and absence. One might therefore conclude that when God ascends to heaven in the face of injustice, this is ‘not good’ for people.227 But this view also moralizes the point, and is a weak explanation of the divine movement. Hence, it would seem that the phrase bal ṭov means that when there is partiality in human courts, God is no longer there; i.e. bal ṭov means that God, ‘the Good One’,228 is ‘not’ on earth. This would be the bolder reading, that takes into account the divine consequences of human injustice.

In both instances here, the term kiv. marks bold theological assertions and the striking uses of scriptural proof-texts brought to support them.229

C. Deceit of God as a ‘Theft’

1. At Sinai

(i) Tos Bab Qama 7.9 (TL, 31). In the context of types of theft (beginning 7.8), it is said that

When Israel stood at Mt. Sinai, they sought to deceive God;230 as is said, ‘All that YHWH has said, we shall do and we shall hear’ (Exod. 24: 7)—kivyakhol, He was deceived by them.

This use of the term kiv. merely serves as a pious comment, qualifying the assertion that God was deceived. It is probably a scribal addition.

2. During Acts of Theft

(i) BT Baba Qama 79b. In the context of a query as to why the penalty for theft is greater than robbery, the answer of R. Yoḥanan b. Zakkai is that it is an act of affrontery against the Creator:

Kivyakhol, he (i.e. the thief) makes the eye below as if it cannot see…

(p.396) This text is problematic. Given the opening comment about the thief's offence against God, and the proof-texts (from Isa. 29: 15; Ps. 94: 7; Ezek. 9: 9), that deal with assumptions and statements that God cannot see human acts, it is likely that ‘the eye below (maṭah)’ is a euphemistic substitute for ‘the eye above (ma͑alah)’ (of God).231 The term kiv. serves a euphemistic function here, qualifying the rabbinic assertion. It is therefore probably a scribal addition.


In this section are collected all other examples of the term kivyakhol that have neither been studied nor mentioned in Sections I–VI above. In different ways, they show usages at variance with the patterns and functions considered earlier. For example, (1) in many cases the term kiv. is used in a purely lapidary sense, piously added to references concerning God or His activities. These instances simply qualify the theologoumenon involved and render it a theological possibility. Such usages are notably extraneous when they occur in the second part of a parable, at the point when the figure is applied to God (e.g. ‘Thus kivyakhol the Holy One’; or ‘such (is the case) kivyakhol (regarding) the Holy One’). Proof-texts are either absent or irrelevant in these cases. By contrast, (2) another cluster of examples adduces the term kiv. solely in connection with the proof-text. In effect, such uses simply qualify an explicit statement made about God in Scripture—without further explication. Such usages are manifestly extraneous and secondary. Finally, (3) a number of examples actually apply the term kiv. to the people Israel. Such uses diverge totally from the preponderant occurrence of the term with respect to God's personality or actions in the Midrash.

The ensuing examples are merely listed, without evaluation. The categorization is designed to facilitate correlation with the materials presented above.

  1. I. Primeval Events

    • A. Creation: MdRI Yitro 7 (H–R, 230); ExodR 23.1; TanB Bereshit 11; PesR 42 (MIS, 175a); ibid. 46 (MIS, 187b); OM i. 67 (col. 2)

    • B. Eden: AgBer 22–3 (B, 48)

    • C. Flood: TanB Bereishit 36; Tan Noaḥ 4

  2. II. God and History

    • A. Events

      • 1. Egypt: During Bondage: ExodR 15.15; 18.7; NumR 2.6; PesR Add 2 (MIS, 196b); at Deliverance: MidPs 114.6 (B, 473)

      • 2. At the Sea: DtL 1.8 (SL, 87)

      • 3. At Sinai: ExodR 28.1

      • 4. Temple: MdRI Beshalaḥ (H–R, 150); SER 28 (MIS, 50); OM iv. 281 (col. 2)

    • B. Personalities

      • 1. Adam: Tan Shemini 2; SER 8 (MIS, 185)

      • 2. Patriarchs and Matriarchs: Adam, AgBer 28 (B, 57 f.); Jacob, ibid. (B, 108); Rachel, ibid. (B, 104).

      • (p.397)
      • 3. Moses: ExodR 18.1; TanB Shemot 18; DeutR 1.2; DtRL Ve͗etḥanan (SL, 49)

      • 4. Righteous: AgBer 50 (B, 100); PesR 2 (MIS, 5b); SER 18 (MIS, 97)

    • C. God and Israel (Helper, Guide, Protector, etc.)

      • 1. Special Relationship: GenR 18.5; DeutR 2.15; LamR 3.20; EstR 7.10; Tan Vayishlaḥ 28; Tan Vayiqra͗ 4; DtRL Re͗eh (SL, 95); DtRL ͑Eqev 12 (SL, 113); AgSongs (SS, 40)

      • 2. God Does Not Abandon Israel: TanB Beshalaḥ 15; Seeks Intercession: PesR 33 (MIS, 150b)

      • 3. Loves the Land of Israel: SifDeut 40 (F, 80)

      • 4. Redemption: OM i. 226 (col. 1) ⩵ BhM vi. 87 (Mid. Yelammedenu)

  3. III. God and Human Behaviour

    • A. Divine Providence

      • 1. Inaction: MidPs 10.5 (B, 94 f.); Action (Victory): PesR 9 (MIS, 32b)

      • 2. Mercy: MidPs 86.7 (B, 374 f.); Compassion: AgBer 47 (B, 94); SER 18 (MIS, 89); Attends Needy: ibid. 19 (B, 37)

      • 3. Judgement: SifDeut 326 (F, 377); MidPs 10.5 (B, 94 f.); ibid. 30.3 (B, 234); ibid. 75.2 (B, 338)

      • 4. Immanence: Tan Ki Tissa͗ 27; Tan Be-Midbar 13; Tan Naso͗ 12

      • 5. God as Source of Life: AgBer 35 (B, 69); Desires Life/Repentence: PesR 44 (MIS, 182b)

    • B. God Responds to Prayer: ExodR 23.8; 29.9; 42.5; 43.1

    • C. Sinful Behaviour: Tan Va-yeitzei͗ 5; RuthR Pet 7; Idolatry: MidPs 97.2 (B, 422); Improper Use of DN: Tan Va͗eira͗ 1; God's Response: Tan Balak 11; Mistreatment of Parents: SER 24 (MIS, 134)

    • D. God is Misled: Tan Qedoshim 8

  4. IV. God and Parables

    • A. Term Kiv. Qualifies Application to God

      • 1. With Formula ‘Thus (kakh) kiv. said the Holy One’: NumR 21.2 (Pinḥas); NumR 23.11 (Masa͑ei); Tan Be-Midbar 20 (TanB 23); Tan Pinḥas 11 (TanB 2)

      • 2. Without Formula: ExodR 20.11; NumR 11.26 (Shelaḥ); TanB Shelaḥ (Addendum 7)

    • B. Term Kiv. Qualifies Parable Itself: NumR 16.27 (Shelaḥ)

  5. V. Term Kivyakhol and Humans

    • A. Applied to Israel: Tos Soṭah 14.5; LevR 23.6; SongsR 2.9; 4.12; Tan ͗Emor 16; MidPs 20.3 (B, 173); ibid. 18.20 (B, 147); ibid. 81.5 (B, 367); SER 18 (MIS, 107); ibid. 28 (MIS, 148)

  6. VI. Other

    • A. Kiv. and Allegory: Divine Body: AgBer 59 (B, 118); Attributes: OM ii. 484 (col. 2) ⩵ BhM iii. 188 (Mid. Tadshe, 20)232

    • B. Desecration of DN: Mid. Yelammedenu (ad Gen. 1: 1).233

(p.398) Conclusions

I. 1. The hundreds of occurrences of the term kivyakhol in rabbinic sources (classical and early medieval), and the diversity and inconsistency of these textual versions (in MSS and in printed editions) make claims of authenticity hazardous and final determinations of limited value.234 Moreover, due to the impact of scribal factors, there is no clear progression over time of increased or decreased use of the term in the evidence at hand. Early sources (like the Tosefta, among tannaitic ones) evince examples of routinized or lapidary usages characteristic of later midrashic compilations (like AgBer, PesR, and MidPs)—though certainly not in the same degree or manner. Even the best manuscript evidence for certain works does not yield consistent results. Certainly in most comparative cases, one may presume that the absence of the term kiv. indicates the more authentic version of the teaching at hand (though not necessarily the more authentic version overall). But this does not mean that the term is always or only a secondary qualification, for certain stylistic conventions and hermeneutical patterns suggest otherwise (see II. 1–2, below).

Moreover, there is no way to get completely behind the manuscript evidence in which the term kiv. is a consistent feature. One is therefore obliged to understand the usages as they occur, and cautioned to eschew putative reconstructions in most instances. This was the procedure adopted above—with the main weight placed on hermeneutical, thematic, and rhetorical factors. Nevertheless, the evidence at hand is nettlesome, insofar as the occurrences of the term kiv. appear to undermine or qualify the very theological assertions being made. This factor cannot be excluded entirely or in specific instances, and the term would then mark the exegete's own hesitation in making bold theological assertions. This being so, one must wonder at the status of the teaching itself; the role of scriptural proofs adduced to justify it; and the implication of the use of kiv. before a biblical citation. These factors must be included in any comprehensive estimation of the phenomenon.

I. 2. Analysis of the data reveals three characteristic features (in different combinations): (a) a theologoumenon (usually highlighting bold anthropomorphic and anthropopathic elements); (b) the term kivyakhol; (c) a scriptural proof-text. Two patterns are most typical: (1) the term kiv. is incorporated into the theological assertion, and followed by a citation formula (like ‘as is said’) and a verse from Scripture (unqualified); and (2) the theological assertion (itself unqualified) is followed by the term kiv., after which occur the citation formula and a proof-text. The first of these patterns is the more frequent; the second sometimes includes an exegetical observation before the term kiv. (see below).

The combined occurrence of the three above-noted features is significant for understanding the meaning and function of the term kiv. in its various contexts. If there were only the theologoumenon plus kiv., there would be little doubt that the term (p.399) serves to qualify the content of the assertion (as especially in late collections); whereas if there were only kiv. plus proof-text, one might suppose that the term seeks to qualify the boldness of Scripture—thus implying a more figurative apprehension of the biblical phrase. However, the evidence shows that all three elements are standard.235 This raises the following questions. In what sense does the term kiv. qualify the theologoumenon, when these are followed (and justified) by an unqualified proof-text? And in what sense does the term kiv. qualify the proof-text, when these are preceded by an unqualified theologoumenon making a bold claim? And finally, what is the bearing on both queries of the fact that there is a marked disjunction between the assertion of the theologoumenon and the plain sense of the proof-text?

I. 3. Taking the foregoing factors into account, the following proposal is offered. The term kivyakhol marks a presumptive possibility of meaning (‘as in/by the possibility’): when incorporated into the theologoumenon, it qualifies that assertion insofar as it is based upon an exegetical possibility presumed of the cited Scripture; and when it precedes the proof-text, the term indicates that the preceding theological assertion is based upon an exegetical presumption concerning the cited Scripture's plain sense. Hence the role of the proof-text is crucial. It both justifies and qualifies the exegetical theology proposed.

On this understanding, kiv. is taken as having distinct rhetorical and hermeneutical functions in the contexts where it occurs. This does not obviate the possibility that with regard to pattern 1 the word kiv. introduces a tone of pious qualification to the theologoumenon; but the present argument is that this occurs within a broader hermeneutical awareness of the presumptive nature of the exegetical theology involved. With regard to pattern 2, the term kiv. pointedly marks a hermeneutical presumption. Over time, the different uses melded, and the term was perceived to be a pious qualifier and adduced by scribes in more indiscriminate ways. Increased usage of the term was undoubtedly affected by rationalistic suspicions of midrash ͗aggadah in Geonic times, and as part of internal self-censorship in the face of Karaite critiques of rabbinic midrash and its bold anthropomorphic theology. This notwithstanding, some occurrences undoubtedly reflect older and possibly authorial conventions (see below).

II. 1. In addition to the aforementioned three-part hermenutical structure, several interrelated stylistic patterns recur. (a) A pattern used in connection with divine participation in Israelite history characteristically makes the general assertion that ‘wherever’ or ‘whenever’ Israel is in exile/straits, God is with them. The term kiv. occurs before the divine reference. Proof-texts are adduced to support this positive situation, but their exegetical import is only implied and must be deduced. (b) Another pattern is used in connection with Israelite observance and its effect upon God is binary (‘If Israel does/does not do God's will’). The term kiv. occurs only in the negative part, after the introit (‘If Israel’) and before the proof-text regarding divine diminishment whose exegetical import is only implied. (c) A similar binary pattern (if/if not) is used in connection with certain deeds or proclamations by Israel and their effect upon God. Here, too, kiv. only occurs in the negative part before a statement (p.400) which exegetically inverts the positive assertion of the proof-text. In these cases, the exegetical procedure is made somewhat transparent. (d) A pattern is used regarding evildoers or evil nations, stating that ‘whenever’ (or, as long as) they exist, a divine feature or aspect is not complete. The term kiv. occurs in these negative cases, before the statement of divine consequences which explicates the cited proof-text. Finally, (e) the formula ‘whenever x is said, (this means) kiv. y’. This hermeneutical formula is clearly explicative, and the term kiv. marks the point.

It will be noted that the term kiv. occurs in the first four patterns ((a)–(d) ) in conjuction with a negative situation. In all, this is deemed a positive feature of God's sympathetic involvement; in the other cases, negative behaviour has a diminishing effect upon God. One may therefore claim that in patterns (b) and (c) the term kiv. is used to qualify the negative consequence—even though in (b) (especially) the positive consequence is also bold. However, it will also be observed that the proof-texts used in the positive parts either intensify the plain sense of Scripture (in (b) ) or simply reiterate it (in (c) ). Only in the negative part is a bold exegetical rereading of the proof-text required for an understanding of the relation between the proclamation and the citation. This suggests that a hermeneutical function for the term kiv. is also involved in such instances.

The regularity of the pattern (despite much thematic diversity) suggests that the placement of the term kiv. in the negative position is an old and fixed convention. Latter scribes never tamper with it.

II. 2. In structural pattern two (theologoumenon followed by kiv.+proof-text), an exegetical condition or procedure is sometimes noted (before the term kiv.), thereby setting up the basis for the presumptive rereading of Scripture that underpins the theological assertion. Among the conditions mentioned is an emphasis on the written text (‘x is written’; or ‘not x is written but y’). In this way, either the form of the written text is stressed in order to support a reconstrual of it; or to highlight the written form (ketib) over its traditional recitation (qere); or even to underscore an orthographic or stylistic oddity in received Scripture. Among the procedures used is the exegetical injunction: ‘Do not read x but y.’ Here a new recitation is the key to the new theology, although this does not necessarily mean that a different articulation is involved; for sometimes all that is involved is a different emphasis (e.g. of an ambiguous pronoun suffix). Occurrences of kiv. in these instances suggest secondariness; but in some cases the reference to a technique may itself be a later clarification incorporated into the text. Each instance must be assessed with its variants. At all events, the term kiv. now signals to the reader that a bold rereading of Scripture is required in order to understand the use of that verse as a proof for the preceding theologoumenon.

III. 1. Many teachings with kiv. are transmitted anonymously. The problems of attributions notwithstanding, a review of the evidence indicates several striking clusters of named traditions. One is a group of tannaitic traditions attributed to R. Meir and R. Simeon b. Yochai (both T3; both students of R. Akiba); a second cluster is amoraic traditions attributed to R. Yoḥanan and R. Simeon b. Lakish (both A2); a third group is attributed to R. Abbahu and R. Samuel b. Naḥman (both A3); (p.401) and a fourth bundle of traditions is linked to R. Berekhia (A5), who transmitted or reformulated earlier teachings and delivered newer o nes. All the foregoing traditions overlap but also exceed the mythic teachings of these masters discussed in Part II.

III. 2. The above evidence also highlights the occurrence of bold anthropomorphic theologies in certain schools and chains of disciples. Thus one master–disciple chain is centred around R. Akiba. A second chain begins with R. Yoḥanan (b. Nappaḥa), who taught first in Sepphoris and later at Tiberias. His close disciple and associate there was R. Simeon b. Lakish. R. Abbahu was also a student of R. Yoḥanan. R. Samuel b. Naḥman (a student of R. Jonathan b. Eleazar) studied in Tiberias and his main student and tradent was R. ḥelbo (A4), who taught R. Berekhia.236 Midrashic traditions linked to the term kiv. were thus associated with EI masters,237 and especially those with a connection to Tiberias. Further generalizations are more allusive, and the evidence dictates caution. Nevertheless, the great number of traditions associated with A3–4 masters is a matter of particular note.


Special Abbreviations for Appendix 2 Special Abbreviations for Appendix 2

  • AgBer

    Aggadat Berei͗shit, ed. S. Buber (Vilna: Romm, 1891)

  • AgSongs

    Aggadath Shir Hashirim, ed. S. Schechter (Cambridge: Dell, 1896)

  • anon.


  • AnOr

    Analecta Orientalia.

  • ArCom

    Arukh Completum, ed. A. Kohut (Vienna: Menorah, 1926)

  • B

    Solomon Buber.

  • b.

    ben; bar (son of)

  • BhM

    Bet ha-Midrasch, ed. A. Jellinek (3rd edn.; Jerusalem: Wahrmann Books, 1967)

  • BM

    see PdRK

  • Ca

    MS Cambridge 1212

  • Col1

    MS Columbia 64

  • Col2

    MS Columbia 81

  • Col3

    MS Columbia 205

  • Col4

    MS Columbia 262

  • D

    Defus (printed edn.)

  • DE

    Massekhet Derekh ͗Eretz Zutta ve-Pereq ha-Shalom, ed. D. Sperber (3rd edn.; Jerusalem: Tsur-Ot, 1994)

  • DH

    Midrasch Tannaim zum Deuteronomium, ed. D. Z. Hoffmann (Itzkowski: 1908–9; repr. Jerusalem, 1984) ⩵ Midrash Tannaim

  • DN

    Divine Name.

  • DtRL

    Midrash Debarim Rabba, ed. S. Lieberman (Jerusalem: Wahrman Books, 1964)

  • ed. prin.

    editio princeps

  • EI

    Eretz Israel (land of Israel)

  • E–M

    see MdRS

  • F

    see SifDeut.

  • H

    see SifNum.

  • H–R

    Mechilta D͗-Rabbi Ismael, ed. H. S. Horovitz and I. A. Rabin (Jerusalem: Bamberger & Wahrman, 1960) ⩵ Mekhilta de-Rabbi Ishmael

  • Jer

    MS Jerusalem 245

  • JTSA

    Jewish Theological Seminary of America.

  • kiv.


  • Lo

    MS London (British Museum Add. 27, 169)

  • Lo1

    MS London (British Museum Add. 16, 406)

  • LXX


  • M

    Midrash ha-Gadol (on Exodus), ed. M. Margulies (Jerusalem: Mossad HaRav Kook, 1967)

  • M


  • M1

    MS Munich 97

  • (p.403)
  • M2

    MS Munich 117

  • MdRI

    Mekhilta de-Rabbi Ishmael, ed. H–R

  • MdRS

    Mekhilta d͗-Rabbi Šim͑on b. Jochai, ed. J. N. Epstein and E. Z. Melamed (Jerusalem: Hillel Press, 1955) ⩵ Mekhilta de Rabbi Simeon bar Yochai

  • ME

    Midrash ͗Eicha (Lamentations), ed. S. Buber (Vilna: Romm, 1899)

  • MH

    Midrash ḥakhamim (MS of V. Aptowitzer; cited in F, H; H–R)

  • MH1

    Midrash ḥakhamim, abbreviated version (MS of A. Epstein; cited in T–A)

  • MhG

    Midrash ha-Gadol

  • MidPs

    Midrash Psalms (Shoḥer Tov), ed. S. Buber (Vilna: Romm, 1891)

  • MidSam

    Midrash Samuel, ed. S. Buber (Cracow: J. Fischer Press, 1893)

  • MidTan

    see DH

  • MinY

    Minḥat Yehudah (commentary of J. Theodor in T–A)

  • MIS

    Meir Ish Shalom.

  • MM

    Midrash Wayyikra Rabbah, ed. M. Margulies (Jerusalem: Wahrmann Books, 1966) ⩵ Vayiqra͗ Rabbah (Leviticus Rabbar)

  • MZ

    Midrash Zuṭa, ed. S. Buber (Berlin: Itzkowski, 1894)

  • OM

    ͗Otzar Midrashim, ed. J. D. Eisenstein (New York: Eisenstein, 1915)

  • Ox

    MS Oxford 147 (Neubauer Cat.)

  • Ox1

    MS Oxford 151 (Neubauer Cat.)

  • Ox2

    MS Marshall Or 24 (Oxford Bod.)

  • Ox3

    MS Oxford 2334-11 (Neubauer Cat.)

  • Ox4

    MS Oxford 2335 (Neubauer Cat.)

  • Pa

    MS Paris 149

  • Par

    MS Parma (de Rossi 261)

  • PdRK

    Pesikta͗ de-Rav Kahana, ed. B. Mandelbaum (New York: JTSA, 1962)

  • PesR

    Pesikta͗ Rabbati, ed. M. Ish Shalom (Vilna: Kaiser, 1880)

  • Pet

    Petiḥta͗ (proem)

  • PH

    Peirush Ha-͗Aggadot le-R. ͗Azriel, ed. I. Tishby (Jerusalem: Magnes Press, 1982)

  • PN

    Personal name.

  • PRE

    Pirkei de Rabbi Eliezer, ed. M. Higger (in ḥoreb, 8 (1944), 82–119; 9 (1946–7), 94–166; 10 (1948), 185–294)

  • R


  • rd.


  • RH

    Rabbeinu Hillel, annotated commentary on the Sifre

  • Ro

    MS Rome (Vatican 32)

  • Ro1

    MS Rome (Vatican 76)

  • S

    see ShR

  • SER

    Seder Eliahu Rabba, ed. M. Ish Shalom (Vienna: Achiasaf, 1904; 3rd printing, Jerusalem: Wahrmann Books, 1969)

  • ShQ

    The Shi͑ur Qomah: Texts and Recensions, ed. M. S. Cohen (Tübingen: J. C. B. Mohr, 1985)

  • ShR

    Midrash Shemot Rabba, Chapters I–XIV, ed. A. Shinan (Jerusalem–Tel Aviv: Dvir Publishing House, 1984)

  • (p.404)
  • SifDeut

    Sifre on Deuteronomy, ed. L. Finkelstein (2nd edn.; New York: JTSA, 1969)

  • SifNum

    Siphre D͗Be Rab, ed. H. S. Horovitz (Jerusalem: Wahrmann Books, 1966)

  • SL

    see DtRL

  • SS

    see AgSongs

  • St

    MS Stuttgart (Orient 32)

  • T

    Tanna (followed by generation number)

  • T–A

    Midrash Bereshit Rabba, ed. J. Theodor and C. Albeck (2nd printing; Jerusalem: Wahrmann Books, 1965) ⩵ Theodor–Albeck.

  • Tan


  • Tg. Jon.

    Targum Jonathan (to the Prophets)

  • TL

    The Tosefta, ed. S. Lieberman (New York, JTSA: 1955–88)

  • Tz

    Tzefat-Safed (MS of Bibliothèque de l'Alliance Israelite)

  • var.


  • Y

    MS Yemen (MS of E. Adler; cited in T–A)

  • YalShim

    Yalkut Shim͑oni

  • Yeb.


  • Z

    Tosephta, ed. M. S. Zuckermandel (Trier, 1882; ne, Jerusalem: Wahrmann Books, 1970)

  • >

    missing; absent.


(1) Special abbreviations listed at end, pp. 402–4.

(2) The more conventional spelling kivyakhol is used throughout; variant vocalizations are given below.

(3) See Sefer ha-Kuzari, trans. into Hebrew and annotated by Judah Even Shmuel (Tel Aviv: Dvir, 1963), 149. Isaak Heinemann has rendered Halevi's gloss on the term by ‘if it could be so, it would be so’; see in Three Jewish Philosophers, edd. H. Lewy, A. Altmann, and I. Heinemann (New York and Philadelphia: Meridian Books and Jewish Publication Society, 1960), 105.

(4) The example he gives is, ‘The sages said: When the Lord of the world descended to Egypt’. As Halevi notes, this teaching is not found in the Talmud, but in ‘certain Passover prayerbooks (siddurin)’. The full version (of which this is the incipit) is preserved in the Maḥzor Vitry (see Machsor Vitry, ed. S. Hurwitz (Nuremberg: I. Bulka, 1923), i. 293). Recitation of this midrash was customary in Provence, according to R. Simḥah of Vitri.

(5) ḥiddushei ha-Riṭba, ed. E. Lichtenstein (Jerusalem: Mosad Ha-Rav Kook, 1976), ad loc.

(6) Halikhot ͑Olam was written in 1467 and first published in Venice, 1634. I have used the third printing with additions and corrections (repr.; Jerusalem: Talpiyot, 1960–1), 9b.

(7) For ‘notarikon’, see simply H. L. Strack and G. Stemberger, Introduction to the Talmud and Midrash (London: T. & T. Clark, 1991), 33.

(8) In his super-commentary to Halikhot ͑Olam, R. Joseph Karo (Kelalei Ha-Gemara) adverts to Rashi's comment on BT Yoma 3b, thus construing the letters kiv- as a prefix marking ‘as if’ such a reading were possible.

(9) See in the reprint edition (Bene Brak: A. Kaufman & Sons, 1976), s.v. kavyakhol (so, as justified by the author), 38 f. The Tosafist comment found in Halikhot ͑Olam is cited with a minor variant ad loc. Ha-Tishby was first published in Isny, 1541. The opinion of Ha-Tishbi is also adduced by R. Yom Ṭov Lipmann Heller (1579–1654), in his glosses ‘Tosafot Yom Tov’ ad M. Sanh. 5.4. In his edition of Pesikta de-Rav Kahana (Mekize Nirdamim: Lyck, 1868), 120 a–b, n. 24 (s.v. kbykl), S. Buber adduces these foregoing opinions, without evalution. For the formula veha-kol mimmekha lo͗mar, ‘you are able to say all this’, see the comment of R. Jacob of Kfar ḥanin in PesR 33, and the comment of MIS in the note ad loc. Kivyakhol is there understood as indicating a teaching which one is able to give.

(10) See overall ‘kvykl’ in Jeschurun (ed. Kobak), 7 (1871), 1–6. For his vocalization, see below, n. 18.

(11) See his ‘Recensionen’, Wissenschaftliche Zeitschrift für Jüdische Theologie, 5 (1844), 271 n. 3.

(12) See s.v. kivyakhol, in his ͑Erkhei Midrash, trans. A. Z. Rabinovitz (Tel Aviv, 1923), 50 (in Hebrew) (orig. Die exegetische Terminologie der jüdischen Traditionsliteratur (Leipzig, 1899), i. 72).

(13) Ibid.

(14) In Aruch Completum (Vienna: Hebraïscher Verlag Menorah, 1926), iv. 130 (col. 2); and cf. the gloss s.v. ͗ikuniyon, ibid. i. 257 (MS Halberstamm).

(15) The Sages (Cambridge, Mass.: Harvard University Press, 1987), 709 n. 1.

(16) The Old Rabbinic Doctrine of God, ii: Essays in Anthropomorphism (1937; repr. New York: Ktav Publishing House, 1968), 131.

(17) Ibid.

(18) The element yakhol in the compound thus has a substantive quality, as already noted by Brüll (above, n. 10); but I do not think it necessary to follow his vocalization kivayakhol, ‘as by the possibility’.

(19) Note that passages employing the term kivyakhol are cited by italicizing the text and the subsection; but in cases without the term, only the text is emphasized (note the citation GenR 12.10, later in the paragraph).

(20) See T–A, 108, n. 3, and the reference to the gloss memshalah (dominion) in MS Liqquṭim. W. Braude, The Midrash on Psalms (New Haven: Yale University Press, 1959), ii. 520 (n. 7) renders ‘power’; he adduces the observation of S. Lieberman that Greek bia is equivalent to Latin defensor civitatis or defensor loci. For other official titles in GenR 12.10, see my comments in The Exegetical Imagination: On Jewish Thought and Theology (Cambridge, Mass.: Harvard University Press, 1998), 14; and p. 188 nn. 14–15.

(21) For this exegetical form, cf. A. Rosensweig, ‘Die Al-tikri-Deutung’, in M. Brann and I. Elbogen (edd.), Festschrift zu I. Lewy's siebzigsten Geburtstag (Berlin, 1911), 204–58.

(22) Lo; in Pa, Ox, M, and ArCom (s.v. cs, b) .

(23) rd. niymanim with EccR 2.12; so also T–A, MinY, ad loc. In Lo, memunim.

(24) For aspects of this topic, see above pp. 232 f.

(25) Cf. GenR 8.9. From this perspective, the deletion of the exegetical hypothetical in GenR 12.1, and the retention of Deut. 32: 6 with the singular verb could be a tendentious redaction, based on this principle. A different proof occurs in MdRI Beshalaḥ, 4 (H–R, 130).

(26) This is one reason to regard the version in EccR 2.12 as secondary to and derivative of GenR 12.1. More significant is the fact that the creation context is valid for GenR, but not so for EccR, which truncates the entire proemial introduction. Moreover, EccR carries over all the subsequent interpretations of creation found in the GenR pericope—matters irrelevant to the EccR context. J. Theodor, ‘Recensionen’, MGWJ 29 (1880), 186, noted that the present instance exemplifies the fact that EccR draws and assembles materials from proems in Midrash Rabba. He does not elaborate. The present analysis confirms his insight (at least in this case). By contrast, the form tiqqunkha in EccR may be original, as the noun tiqqun occurs earlier in the proem found in GenR. If so, the formulation tikkunkha in the latter was ‘corrected’ to conform to the proof-text from Deut. 32: 6.

(27) binyan ͗ab indicates a scriptural genus by which a species of topically related passages may be adduced. The exegetical norm here is like Hillel's third hermeneutical principle, ‘A binyan ͗ab from one scriptural text’ (see Tos. Sanh. 7.11).

(28) See the stylistic argument put forward by H, ad loc. This passage > YalShim, MH.

(29) Job 25: 2b is adduced pleonastically, either because it was used in an immediately prior teaching, or also because v. 2a refers to ‘Dominion and Dread’, and these may have been understood as indicating (just) two angels. Cf. PdRK 1.2 (BM, 5), where these powers are named Michael and Gabriel by R. Jacob of Kfar ḥanan.

(30) An even bolder mythic version of this type occurs in ShQ 46, ll. 73–6 (Siddur Rabba), where Song 5: 11 is interpreted to speak of the awesome size of the divine brow, and indicates that ‘after the Temple was destroyed—kivyakhol, the expanse of this brow was reduced (nitme͑u; rd. nitma͑eṭ)’.

(31) The sequence of citations is inverted.

(32) This resolution implies the sequence as in SifNum.

(33) It is introduced as a beraitha; this designation may belong earlier.

(34) Lit. ‘From where to us?’; namely, ‘From where in Scripture may we derive this?’ (Or, colloquially, ‘How do we know this?’)

(35) The textual relation is stated more strongly in MM 723 n. (‘influenced by’), and even more so in MIS 103a, n. (‘copied from’).

(36) But with qilluso, not shibḥo, and ‘diminished from His retinue’ .

(37) Rd. qillusi with Ox, Pa.; also in PesR 21.

(38) The verb used here is mi͑eṭ.

(39) See the explanation of this verse below.

(40) Jer. 25: 37 was understood as the Temple in BT Ber. 3b. Cf. Ps. 79: 7.

(41) ye͗shamu; the stem ͗sm is a by-form of šmm (cf. Ezek. 6: 6).

(42) Cf. Tan Tzav 7.

(43) The account is preceded by a midrashic play on Isa. 41: 2, which is applied to Abraham. The use of kiv. appears to be out of place.

(44) See ͑Arugat ha-Bosem, ed. E. E. Urbach (Jerusalem: Mekize Nirdamim, 1963), iii. 154.

(45) See in the edition of I. Davidson (New York: JTSA, 1934), 75.

(46) Cf. SifNum 61 (H, 58 ff.).

(47) Piyyuṭei Yannay (Berlin: Schocken, 1939), 33; and cf. S. Lieberman in ‘ḥazanut Yannai’, Sinai, 4 (1939), 238.

(48) PRE 7 understands ‘came’ as ‘was born’. Instead of ‘with …cruelty’, there is also the reading ‘as a cruel person (ke-͗akhzari)’; cf. see D (Venice, 1545), and Ox1. The reading is also cited by R. David Pardo in his commentary Peirush Sifre de-Bei Rav (Jerusalem: Lev Sameaḥ, 1990), iv. 234.

(49) Cited ibid. (n. 47).

(50) Kiv. > MidTan (DH, 189 f.).

(51) Attributed to R. Pinḥas (A5).

(52) GenR 59.8 may also have originally been linked to Deut. 32: 10, otherwise it is difficult to explain the theme ‘I made Him known’ (hoda͑tiv) in a verse dealing with swearing. Presumably, this verb interprets yeboneneihu (‘He (Abraham) made Him known’).

(53) The passage concludes with the midrashic lemma, Ps. 18: 36, hoy ͑anvatekha tarbeni, which is presumably understood as Abraham's response: ‘Surely, Your humility makes me great.’ In his disputation, Abrabanel had accused Hieronymus de Santa Fide of forgeries with respect to various ͗aggadot; however, with respect to a mythical reading of Ps. 110: 5, confirmation of authenticity may be found in Oxford MS Opp. 22 fo. 66 (Neubauer Cat. 167), as adduced by A. Neubauer in The Book of Tobit (Oxford: Clarendon Press, 1883), xxii–xxiv.

(54) In MidPs 110: 1 (B, 465), Ps. 110: 1 is midrashically applied to Abraham; and cf. the comments in AgBer 18 (B, 37).

(55) The versions in NumR 4.1 and Yelammedenu Bemidbar (cited in ArCom, i. 257; s.v. ͗kwnyn) read ‘on’ and ‘under’ respectively. The vocalization may be ͗ikoniyn. It renders Greek eikonion.

(56) See Rabbeinu Baḥye, Bei͗ur ͑al ha-Torah; ed. ḥ. Chavel (Jerusalem: Mosad Ha-Rav Kook, 1982), i. 288 f. And similarly, Recanati, ad loc.

(57) Tan B Vayishlaḥ 10, p. 168 note ad loc.

(58) In the firse case, ͗im (if) is read as an oath clause; in the second, ve-shabti (and (if) I return) is cut off from the subsequent apodosis, and possibly construed as an imperative (and (You) return me!).

(59) This is also the version preserved in the Geniza, published by L. Ginzberg in Seridei Ha-Yerushalmi (New York: JTSA, 1909), i. 210 (ll. 1–2), though the word da͑atah (‘her mind’) was accidentally dropped.

(60) See in his Tzeror Ha-Mor, ed. J. Alnaqawa (Tzeror Ha-Mor Institute; Jerusalem: Tzur ͗Ot, 1985), 141.

(61) Also in ArCom iii. 433a (s.v. ḥamor), and the epexegetical comment reported.

(62) A midrashic compilation of R. Samuel b. Nissim Masnut, ed. M. Hacohen (Jerusalem: Mosad Ha-Rav Kook, 1962), 239.

(63) Buber, Shoḥer Ṭov, 362 n. 14, indicates the absence of the clauses marking Joseph's and God's favour in various MSS and the ed. prin.; such readings weaken the homily.

(64) Reading with lishrot, ‘to dwell’, as in the MSS; however, if one reads without lishrot, as in the ed. prin. (Venice, 1545), one must point mnḥ as maniaḥ (‘I shall not give the Shekhinah rest’, or ‘have it settle’).

(65) The verb is used hermeneutically, to introduce a scriptural conclusion.

(66) Only in Ro and RH.

(67) Here, Moses says, ‘I shall not (͗eini) permit the Shekhinah to descend (tered)’. The variant ‘he shall not (͗eino)’ softens the statement but garbles the syntax; see in S. Horovitz, Der Sifre Sutta nach dem Jalkut und andere Quellen (Breslau: Alkalay & Son, 1910), 79.

(68) Following ed. prin. (Mantua, 1563), where par. 16–18 are addenda.

(69) On this power of the tzaddiq to influence God, see R. Mach, Der Zaddik in Talmud und Midrasch (Leiden: E. J. Brill, 1957), 112 f. See further below, p. 348.

(70) Similarly in Coll.

(71) For ͗ish as the divine Husband of Israel, see Hos. 2: 18.

(72) On this reading, the nota accusativa (͗et) must be disregarded.

(73) According to some MSS.

(74) The homily is given anonymously and very tersely in the Tan Terumah 3 and TanB Terumah 2 (p. 89), in both cases without kiv.

(75) For the phrase dabar qasheh, see above, p. 153.

(76) This is also the structure in an important MS not adduced by H–R; namely British Museum Or 5559 fo. 18. It was published by Z. M. Rabinovitz, Ginze Midrash (Tel Aviv: Tel Aviv Univ. Press, 1977), 4–14. For the value of this MS unit, see M. Kahana, Tarbiz, 55 (1986), 500, 508.

(77) For this verse, see above, pp. 132–6, and in C, below.

(78) > Tan Beshalaḥ 28; and thus also Midrash Hizhir, ed. J. Freiman (private printing, 1873), 43.

(79) For numerous Geniza variants, see M. Kahana, Ha-Mekhilt͗ot Le-Parashat ͑Amaleq, 101. His point (n. 196) that Hannah is speaking about God's salvation tallies with R. Abbahu's teaching in MidSam 4.4 (B 5); but it does not explain the radical midrashic point here.

(80) The term kiv. >Ox2 and M2; it occurs in PH 59.

(81) Kiv. >D, M2, and NumR 7.10.

(82) Lo, meshu͑abedet.

(83) Only Egypt and Babylon are mentioned as exiles; only Deut. 30: 3 is given as proof of divine participation in restoration—with the term melammed used to narrate the inference drawn from it.

(84) Isa. 43: 14 is cited in MdRI and SifNum with an orthography indicating the midrashic sense (šwlḥtyshullaḥti, ‘I sent Myself’), not the MT version (šlḥtyshillaḥti, ‘I sent’). This is a single scribal adaptation to the new ‘sense’ (H, 83 n. is unnecessarily perturbed).

(85) The language of restoration in v. 39 uses a phrase similar to Deut. 30: 3, but was never used in the midrashic attestations.

(86) In MZ Songs 2.9 (B, 26), ad Songs 2.9 (p. 26), Isa. 41: 25 is uniquely adduced to indicate divine arousal from the north. The passage continues: ‘Why does He come from the north?—because when Israel were exiled to the north, kivyakhol the Glory was exiled with them to the north.’ But here kiv. merely suggests the possibility of God's exile to the north from His later arousal from that place. No proof-text is provided, and kiv. is used to mark a hypothetical event.

(87) kiv > D, M2, and Pa.

(88) The opaque reference refers to the specification of the brick in the first case; the second case deals with both the brick and the work implements. See above, p. 187.

(89) Lit. ma͗i ta͑ama͗.

(90) For such theological differences, see above, pp. 136–8.

(91) The secondariness of the remark is also indicated by its use of Aramaic. For a weak version of this topos with kiv., see AgBer 72 (B, 140).

(92) See p. 140. Kiv. occurs in PH, 122.

(93) The interlocutor structure seems original here, and is neutralized in MdRI Bo͗ 14. In this setting Akiba takes the more radical position, contra his debates with Pappius (see ibid., Va-Yehi Beshalaḥ 6 (H–R, 112 f.) ). See similarly in EccR 7.3.

(94) MT goyyim is cited; as noted (p. 000), the original midrash had the version goy (‘a people/nation’).

(95) The verb in the JT version is quoted in plene-form also in the Minḥat Shai, ad loc. The ‘Qorban Ha-͑Edah’ gives a defectiva version; but if this were original, the midrash would lack any orthographic basis.

(96) R. Meir's point is given tersely and without kiv. in TanB Aḥarei 13. The exegesis is weakened in NumR 2.2 and TanB Bemidbar 10 (B, 9), where after ‘kiv. when Israel was redeemed’, the qualifier ‘as if (ke͗ilu)’ is interpolated. Cf. Tan Aḥarei 12 (R. Berekhia).

(97) According to Pesiqta ḥadita, as published in Bet ha-Midrasch, ed. A. Jellinek (3rd edn.; Jerusalem: Wahrmann Books, 1967), vi. 37 (Cod. Heb. xii, Leipzig Rathsbibliotek), kiv. is used with an assertion of God's own salvation. The punctuation vayoša͑ is explicated as an ͗al tiqre reading, given in the name of R. Abbahu.

(98) Munich 22710; cited by A. Epstein from his MS, in ‘Yalqut Shim͑oni’, Ha-͗Eshkol, 6 (1909), 207.

(99) This is the midrashic reading; the plain sense suggests that it is Israelite behaviour that has caused this degredation. See also S. Blank, ‘Isaiah 52.5 and the Profanation of the Name’, HUCA 25 (1954), 1–8.

(100) The LXX, Syriac, Vulgate, and Targum have a plural verb. R. Eliezer of Beaugency says that va-yabo͗ ‘is like va-yabo͗u’!—a deft emendation. Cf. his comment on Ezek. 14: 1, noted by S. Poznanski, R. Eliezer Mi-Belgantzi: Peirush Yeḥezkel (Warsaw: Mekize Nirdamim, 1909), i. 58 n. 5.

(101) The term kiv. seems attached to this proof-text in the version in TanNum Behar 1.1, and Ca.

(102) This reading is underscored by ‘Matnot Kehunah’; it was presumably lacking in his MS copy.

(103) In LamR 2.6 the tradents are R. Azariah (A5) in the name of R. Judah b. R. Simon (A4).

(104) Cf. SongsR 1.6

(105) Elsewhere, the arm is ‘in bondage (meshu͑abedet)’; see p. 148. The use of kiv. with the enslaved arm in MidPs 98.1 (B 1422 f.) is apparently extraneous—no proof-text is given there. Notably, the tradent is R. Aḥa (see 4(i) below).

(106) Minor variants with kiv.; cf. Pa.

(107) Presumably, R. Aḥa of Lod (A4), later of Tiberias.

(108) Hu͗ is considered to be a divine Name in the Tosafot to BT Sukkah 45a; s.v. ve-hu͗. But see the position of R. Judah in the Gemara. Also, cf. above, III.B.4(i).

(109) The word rehaṭim is understood as the plural noun ‘runnings’, and thus related to the Hebrew verb rutz. Philologically, rehaṭim has a double plural marking (internal h and final-im).

(110) On a divine oath and the exodus, cf. LevR 23.2 (MM 528).

(111) In the case of LevR, rehaṭim refers to the ‘Temple’; in SongsR it marks the ‘tents’ of Jacob. In SongsR 1.17 rehiṭeinu (‘our bower’) is also interpreted as the Temple.

(112) Kiv. also occurs in Tanḥuma MS Col4.

(113) Magazin für Wissenschaft des Judentums 15, p. 75.

(114) Pugio Fidei adversus Mauros et Judaeos, ed. B. Carpzov (Leipzig, 1687), 845. This passage is also adduced as authentic by Lieberman, Sheqi͑in, 64. Martini's text also gives the second clause a heavenly referent, since the ‘faces’ that have turned pale are interpreted as the ‘angelic retinue’ (mal͑akhei ha-sharet).

(115) The application of the last part to the angels does occur; kiv. does not occur.

(116) A teaching attributed to Rab (BA1), who often engaged in esoteric theology.

(117) In the Jerusalem MS used by Shinan (National and University Library, 24o 5977); kiv. and shutaf are missing in the printed editions.

(118) For other versions, see the discussion below.

(119) I simply cite the phrase ‘is ill at ease’ from H. Danby, The Mishnah (Oxford: Clarendon Press, 1933) 390 f., as a basis for the ensuing discussion of the difficult word qlny (possible vocalizations and interpretations follow).

(120) On the langauge of this clause, see below.

(121) As if a crasis of qal / ͗eini.

(122) Cf. Nah. 1: 14; 2 Sam. 6: 2. The reading of qlny as from qalon (‘contempt’) occurs in the Yemenite tradition in a copy of Maimonides' commentary on the Mishnah in MS Jerusalem 8o, 596. In this text, the terms Shekhinah and kiv. are absent, but they do occur in MS Jerusalem 4o 1336.

(123) ͗Otzar Ha-Ge͗onim, ed. B. M. Lewin, vol. iv, ḥagigah, pp. 30 f. For a full analysis of the term and its occurrence in the MSS, see G. Alon in ͗Alummah, 1 (1936), 124–8.

(124) Rabba's comment could also be spoken by God, if we assume that when it says ‘I am light’ or ‘weak’ it means ‘I (God) am heavy (burdened)’ by this death (cf. Alon, ͗Alummah, 1 (1936), 125).

(125) The vocalization follows G. Yalon in ḥ. Albeck, Shishah Sidrei Mishnah (Jerusalem: Mosad Bialik, 1953), iii. 188.

(126) The language is even stronger in Tos. Yeb. 8.7 (TL, Nashim, 26), where it says that one who sheds blood ‘annuls (mebaṭṭel) the image’.

(127) Presumably, the reference to the ‘exceedingly heavy’ yoke which Babylon imposed (hikhbadeti) upon ‘the elder’ encoded for the midrashist the special kabod (or honour) which the Lord would account him by suffering on his behalf.

(128) Cf. in the versions found in Lo1 and Sifre Num Zuṭa (H, 271), respectively. In the first case, kiv. is transformed into the hypothetical yakhol; in the second, the term kiv. occurs but the whole daring theology is neutralized by the reference to forgiveness.

(129) In the citation, ͗eibel yaḥid literally means an ‘only child’, not an individual sufferer.

(130) The formulation is condensed; the word tomim (‘twins’) is used. The personal pronoun suffix is thus dropped, leaving the exegesis somewhat obscured. The tradent here is R. Yannai (A1).

(131) Also a condensed formulation, referring to the twins as te͗omim. In these cases, the teaching attributed to R. Yannai interprets tamati as tomyati/teyomati (‘My twin’) to teach about Israel's equivalence with God at Sinai—neither one superior to the other.

(132) I give the MT formulation, which triggers the ensuing comment. Remarkably, the citation adduced in the midrash reads ‘My eye’, thus introducing the very reading into Scripture which the teaching claims was altered for euphemistic reasons. See further below.

(133) This passage is repeated as an editorial Wiederaufnahme at the end of the list (see ll. 13, 13), before a new subject is taken up.

(134) This source also repeats R. Judah's teaching in Wiederaufnahme (pp. 81, l. 16; 82, l. 6).

(135) See the lists and discussion in C. D. Ginsburg, Introduction to the Massoretic-Critical Edition of the Hebrew Bible (1897; Ktav Reprint, New York: Ktav Pub. House, 1965), 347–63. W. E. Barnes, ‘Ancient Corrections in the text of the Old Testament (Tiḳḳun Sopherim)’, JTS 1 (1900), 387–414, has argued that the tradition of changes is midrashic conceit and not a Massoretic feature. C. McCarthy, The Tiqqune Sopherim and other Theological Corrections of the Massoretic Text of the Old Testament (OBO, 36; Fribourg: Universitätsverlag, 1981), basically agrees but allows that Zech. 2: 12 (inter alia) is a correction.

(136) So ed. prin. (Constantinople, 1512). In Ox the reading is ‘it was written with a yod’ (i.e. ͑einiy, ‘My eye’).

(137) See Lieberman, Hellenism in Jewish Palestine (New York: JTSA, 1962), 26–37). The reading ‘My eye’ is adduced in Minḥat Shai ad Zech. 2: 12, but the author piously rebuts it. See also the ‘Matnot Kehunah’ commentary at ExodR Bo͗ 13.1.

(138) Ten of the eleven euphemisms in MdRI involve changes in the pronominal suffix.

(139) Cf. also the lemma edited from early texts, ArCom iv. 181.

(140) The texts states: ‘͑eini hu͗—it is (originally) “My eye.”’ Reference is also made to a ‘scribal correction’ here, but this seems secondary. The version in Tan Beshalaḥ 16, 1 is more hypothetical in nature (i.e. Scripture ‘should have said ͑eini’—but chose a euphemistic formulation).

(141) The beginning of this section is of the ‘Yelammedenu’-type. It only appears in the ed. prin. (Prague, 1656).

(142) Possibly construed as ‘The Lord God called the (angelic) host to crying’, etc. See above, pp. 228–9, and b (ii) below.

(143) Emended to ‘I took away’ by R. Ephraim Margoliot, in his commentary, ‘Zera͑ Ephraim’, printed in the Lvov edn. of PesR (1853).

(144) The ‘passage’ refers to the scroll of Lamentations, recited on Tisha͑ be-͗Ab.

(145) On the basis of Isa. 22: 12, first God and the angels were interpreted as crying. In addition, the tears of the heavenly bodies and earthly elements were adduced, with proofs brought from Joel 2: 10 and Jer. 4: 23–4 (B, 60).

(146) Citing Isa. 22: 4.

(147) kiv. > B. Meg. 10b; this is arguably the original formulation.

(148) So also, the ‘Yad Moshe’.

(149) In PdRK, in the name of Bar Qappara (T5); in ME in the name of R. Naḥman (A5).

(150) See the full discussion above, pp. 177–8.

(151) The version in BT Soṭah 48a has the same effect. It, too, is without the term kiv.; but instead of the ligature ‘they recite’ we have ‘for that reason it was said’. Presumably, this shift is due to the redactors' sense that since reference to the abrogation of the Awakeners is mentioned earlier, the recitation is a thing of the past.

(152) Discussed above, pp. 147–50. Cf. also M, 3.2, p. 45 (with kiv.); and similarly in Batei Midrashot, ed. A. Wertheimer (Jerusalem: Ketav Ve-Sefer, 1980), ii. 103.

(153) Text printed in OM, i. 195b; repr. from A. Berliner, Pletath Soferim (Breslau, 1872), 36 ff. (Hebrew section).

(154) With Ox2 and M2; H–R with D reads ha-͗ofsim (presumably a metathesis of the s/f). The emendation of I. Löw to ͗iskufim, ‘thresholds’, in ‘Lexicalische Mizellen’, Festschrift zum siebzigsten Geburtstage David Hoffmanns (Berlin, 1914), 199 f., is unjustified.

(155) This construal of ke-kofer is presumed by the gloss of the Gaon of Vilna: ‘This teaches that the sins of Israel were atoned for (nitkapperu)’. See ‘Efat Tzedeq’; and cf. ‘Birkat Havetzib’.

(156) However, it reads ha-͑ofsiyim—an even more incomprehensible version of D (see n. 154).

(157) The reference to this teaching in JT Rosh Hashanah 1.3, 57a is apocopated, and without the exposition.

(158) For this term, see Urbach, The Sages, 66–9.

(159) See above, pp. 173–90; and further on in this section. This theurgical theology is thus, presumably, at least tannaitic in age.

(160) In addition to 1.a(i), see below for other examples (1.a(ii–v); b(i); c(i); d(i–iii); e(i); 2.c(i–iv); e(i) ).

(161) In Sifrei de-͗Agaddeta ͑al Megillat ͗Ester, ed. S. Buber (Vilna: Romm, 1886).

(162) Cf. BT Meg. 15b, in the name of R. Tanḥum: ‘The sleep of the King of the world’.

(163) See above, n. 160.

(164) I have translated yashein differently, in these cases, to bring out the temporal modalities implied.

(165) Here, ‘does His will’, not ‘does the will of the Omnipresent’, as in (i) and (ii).

(166) See n. 160.

(167) It is the only use of this rhetorical type with a second proof.

(168) Cf. Saadia, Rashi, and Rashbam.

(169) See above, n. 160.

(170) Namely, ‘If I change’. The particle ͗im is thus construed as introducing a conditional clause (if), not a temporal one (when).

(171) In SifDeut 331 (F, 380 f.), the first phrase denotes the swiftness of punishment; the second, the fact that due process will be done.

(172) For the bold midrashic treatment of this phrase, see my Exegetical Imagination, 57–9.

(173) The adjective yesharim plays on the people's name Yeshurun (Jeshurun).

(174) See n. 160.

(175) My interpretation of this phrase here differs from that found in Exegetical Imagination, p. 60, where I suggested that God's Majesty ascended to the heights. For this latter view, see R. David Pardo, Peirush Sifre de-Bei Rab, ad loc; repr. with an edition of Sifre (Jerusalem: Lev Sameaḥ Institute, 1990), iv. 317.

(176) The term kiv. is a stable feature in the MSS.

(177) Regarding the dynamis as the divine Power or Might, see above, p. 180 and n. 82.

(178) Translating the lemma according to the plain sense; see further, below.

(179) As customary with this rhetorical form; cf. n. 160.

(180) A by-form of the term tšš.

(181) Also the terminology is more terse, stating only geburah—without the words ‘on High’. In LamR, ‘on high’ is written with slight variations (ma͑alah/ma͑alan).

(182) The two names (Ṭarfon/Parṭa͗) are apparently scribal metatheses of each other.

(183) The formulation in ME 1.6 (B, 70) is even weaker. In it, the second position states: ‘But when they anger Him, “they go without strength before the enemy”.’ This revision aligns the passage with more normative rabbinic theology about the sovereignty of God.

(184) See Jos. Abrahams, The Sources of the Midrash Echah Rabba (Dessau: H. Neubürgen, 1881 (doctoral diss., Leipzig) ), 45–59.

(185) On the antiquity of this epithet, and variants in the MSS, see Urbach, The Sages, 67 f., 77.

(186) ‘Vaults’ is a figurative rendition for something like the ‘bond’ of heaven; cf. the Akkadian term markas šamê (‘bond of heaven’) in CAD, M, i. 283, 4a. This cognate phrase was noted by S. Paul, Amos (Hermeneia; Minneapolis: Fortress Press, 1991), 280 n. 77.

(187) See above, n. 160.

(188) See his Peirush, iv. 288.

(189) It is not certain if the verb boneh in the midrash means that God or Israel ‘builds’ up the divine realm.

(190) NumR 15.18 uses kiv. and specifically refers to the establishment of the divine throne ‘on high’.

(191) The style here is different from the rhetorical form found elsewhere in this series, which sets forth both positive and negative consequences of human action. Here, too, the second position contains a negative valence (cf. n. 160).

(192) This issue is part of a larger liturgical and thematic topos; see overall A. Green, Keter: The Crown in Early Jewish Mysticism (Princeton: University Press, 1997), passim. Again, the negative valence is in the second position.

(193) Lit. vayyabo͗, ‘When he came’; this verse was understood elsewhere to refer to God's participation in exile. See above, p. 155.

(194) See n. 160 for other examples.

(195) The proof-texts of the first part have no relationship to the key word, and are out of order (they are similar in content).

(196) Presumably, R. Isaac Nappaḥa (A3)

(197) This addition is > in Tan Naso͗ 4.

(198) In LevR 23.12 the teaching is traced to R. Levi b. Parṭa, through R. Judah b. Simon.

(199) In NumR Naso͗ 9, R. Isaac's opening assertion about weakening the power of the Shekhinah is preceded by kiv.; this usage appears to be a euphemistic addition.

(200) Two other uses of kiv. occur at the beginning of Tan Naso͗ 4; but both seem to be euphemistic supplements, without any exegetical import or function.

(201) Variants of the positive first part of this teaching (also using Lev. 18: 2–3) occur in MdRI, Yitro 6 (H–R, 222 f.) and Sifra, ͗‰Aḥarei Mot 13.3—both times in the name of R. Simeon bar Yochai. The version from the school of Akiba lacks R. Ishmael's bold conclusions. Note the use of kiv. in the negative position; see n. 160.

(202) The beginning of this unit is given above, C.1.d(i). These teachings are incorporated into a comment on Deut. 33: 5, which provides a rhetorical inclusio. Amos 9: 6 comes first, as it has a verb tally with the comment on Deut. 33: 5.

(203) Two other teachings in this series (citing Exod. 15: 2 and Deut. 32: 3) follow this contractive rhetoric; here, too, kiv. marks only the negative application. Neither instance involves any apparent exegetical dimension, and kiv. serves a purely rhetorical and euphemistic function (i.e. to qualify the assertions that without human praise or proclamation, God or His Name would be somehow diminished). Note the use of kiv. in the negative position; see n. 160.

(204) See n. 160

(205) Z is based on MS Erfurt; kiv. also occurs in D and MS Vienna (Heb. 20; in list of A. Schwartz (1925) N 46).

(206) As usual, the term kiv. is in the negative position; see n. 160.

(207) These teachings are unique. In SongsR 1.15 the emphasis is on various human deeds and commandments, but nothing affecting God.

(208) In PesR 12 only R. Abba is mentioned.

(209) Before the phrase ‘whenever’.

(210) Weaker yet is the use of the trope in OM (i. 19b) regarding Esau, where the point is made that so long as they practise idolatry, God's kingdom is not complete. The formulation is entirely lapidary; there is no proof-text provided. Also regarding Edom, cf. AgBer 58 (B, 116).

(211) The text is lacunose; I have reconstructed it ad sensum.

(212) See Rashi, BT Soṭah 35a, s.v. mimmennu. In BT ͑Arakhin 15b, the Tosafists (s.v. ͗al) adduce another opinion that claimed that the letter nun had a dagesh forte only in the second (midrashic) one, but reject this as invalid.

(213) This point was perceived by Rashi, who comments on the term kiv. used here as indicating ‘as if one could say this thing against God (kelapei ma͑alah)’.

(214) The copyist has mistakenly conflated Deut. 8: 7, which only has ‘to the good land’, with the phrase ‘to the good and spacious land’ from Exod. 3: 8.

(215) The verb is unknown; this translation follows Saadia and Kimḥi, who construe the phrase contextually. Ibn Ezra takes ragleykha metaphorically (‘your ways’). For the use of the lamed to indicate a noun (namely, according to), see 1 Sam. 25: 22. In the midrash, it expresses locality.

(216) The divine bounty of sustenance in the midrash is undoubted influenced by the end of v. 16. Cf. also the comment in ‘ḥiddushi ha-Radal’, s.v. kivyakhol, where v. 17 is also adduced. In linking the vision of God to the eating below the divine feet, R. David Lurie (Radal) may have had Exod. 24: 10–11 unconsciously in mind.

(217) Reading the difficult verb tukku in terms of an entrance underlies Rashi's comment, though in his point of entering the protective ‘shade’ of divinity he seems to have been influenced by Targum Onkelos.

(218) In his commentary ‘Peirush Meharzu’, R. Zeev Wolf Einhorn makes a similar observation; the link between tukku and tokho is also noted.

(219) D (Vilna) reads mashmi͑a (‘makes Himself heard’); but I have emended to ve-shome͑a (‘and listens’) following ‘ḥiddushei ha-Radal’, who correctly notes the formulation ‘I shall hear’ (͗eshma͑) in Isa. 65: 24 (cited earlier in the homily) and the use of the qal-form in the proof-text itself. See similarly in ‘ḥiddushei ha-Rashash’. This is the commentary of R. S. Strashun.

(220) The last phrase is not cited in D, but it is obviously central to this teaching.

(221) The midrash does not explicitly cite ͑od; but its exegetical importance is clear and was pointed out by the Rashash (cf. n. 219, above).

(222) This teaching is separate from the one about the link between slander and heresy, which is given earlier as a teaching of R. Yoḥanan in the name of R. Yose b. Zimra, in a different form.

(223) The term kiv. occurs in the same position in the teaching.

(224) The formulation in the Talmud shows a common epigrammatic style; the formulations in MidPs evince a more dramatic dialogue.

(225) Use of Ps. 29: 10 to indicate God's seated judgement at the Flood also occurs in ExodR 17.4; but the term kiv. does not occur.

(226) See also the ‘Peirush Meharzu’.

(227) Cf. the ‘ḥiddushei ha-Radal’.

(228) For this designation, note the eulogy ha-ṭov veha-meiṭiv, ‘The Good One, who does good’ (M. Ber. 9.2; J. Ber. 9.2, 13d; and the discussion in GenR 13.15 (T–A, 124) ). Ṭov is one of the seventy names of God at the beginning of Midrash Lekaḥ Ṭov on Song of Songs, by R. Tuvia b. Eliezer, ed. A. Grünhut (London, 1909), 5.

(229) Notably, the term kiv. does not appear in MidPs 12.3 (B, 107), but this may be because the teaching is structured around the two parts of Ps. 12: 5, and thus the teaching seems more of a sharpened reiteration of the biblical verse.

(230) Lit. ‘they sought to steal, da͑at ha-͑elyonah’.

(231) The designation ma͑alah is itself a periphrastic reference to God in old rabbinic discourse. Cf. GenR 53.14 (T–A, 572).

(232) Cf. A. Berliner, Beiträge zur Jüdischen Alterthumskunde (Vienna: Lippe, 1887), vol. i, p. xxxix.

(233) See MS fragment published in Batei Midrashot, i. 141 (no. ‘b’). A related tradition occurs in GenR 1.12 (T–A, 10 f.), to teach divine humility; it is > kiv.

(234) Note, for example, the following: kiv. in MdRI Bo͗ 14 (H–R, 51, l. 10), and also in D, but > Ox1, M2, YalShim; kiv. in MdRI Bo͗ 14 (H–R, 51, 1. 18), but > D; however, in MdRI Beshalaḥ 5 (H–R, 134, l. 3), kiv. only in D. The inconsistency is also evident in parallel teachings. Thus, for example, kiv. in MdRI Bo͗ 14 (H–R, 54, ll. 5, 7), but > in SifNum 84, Beha͑alotekha (H, 83, ll. 5, 7); and the opposite, kiv. in SifNum 157 Maṭṭot (H, 211, l. 20), but > in MdRI Beshalaḥ 5 (H–R, 134, ll. 10–11).

(235) For the weakening of this structure, see Sect. VII above.

(236) R. Jeremiah (A4) is also affiliated with these late chains of tradition, and was an authority of the school of Tiberias.

(237) This exegetical theology was apparently not developed in Babylonian academies (although such theology was preached there by EI sages; cf. LevR 23.8 (MM, 573) ).