Saadia Gaon's kavod nivra
and its place in his philosophy of Judaism

Jay Michaelson
January 20, 1994

Outline of Argument

I. Introduction: Saadia as philosopher

II. Saadia's theology
     A. Unknowability of Gcd
          1. Total abstraction and incorporeality of Gcd
          2. Can thus not be known
          3. If revelations cannot give knowledge of Gcd,
               they cannot be of Gcd
     B. Who is Saadia arguing against?
          1. "Those who reject the sciences"
          2. Those for whom his ideas are "too subtle"
          3. Those who think that G-d and the First Cause
               are different things
     C. The price of Saadia's theology: Revelation
          1. Why Saadia must invoke created kavod
               a. He didn't: alternatives
               b. Anti-anthropomorphic interpretation of "facts"
               c. Interpretative technique fails in whole theophany
          2. Revelation was of a created object

III. Saadia's doctrine of the kavod nivra & its discontents
     A. Its createdness
          1. Contradicts scripture
               a. Theory of Biblical ellipsis
               b. Interpretive method
          2. Changes nature of revealed and revelation
               a. No noetic value to theophany
               b. Divergence from traditional view of prophecy
          3. Distinction between Mosaic and other revelations
               a. Redefines the special status of Moses
               b. Leads to the same "can of worms"
                    i.  Saadia's speech-as-apprehension idea
                    ii. Kavod seen, Gcd 'heard'
          4. Possible alternatives to Saadia's theory
               a. Appeal to Miracle
               b. Appeal to limits of reason
     B. Its kavod-ness
          1. Departure from the Talmudic View
               a. Midrashic/Talmudic definitions of kavod
               b. The exception that tests the rule
          2. Problems 
               a. Departure from Talmud: Karaism
               b. Praying to a created object: Idolatry
               c. Some of Saadia's critics
     C. Consequences of Saadia's position in subsequent thought
          1. Philosophy: Maimonides
               a. Kavod as a description?
               b. Kavod as created being?
               c. Kavod as equivocal term
          2. The development of Kabbalah
               a. Explanation of anthropomorphism
               b. 'Inner' and 'Outer' kavod
               c. Pleroma in the Bahir and elsewhere
          3. Implications

IV. Conclusion: Saadia's philosophical endeavour 

Saadia Gaon's kavod nivra
and its place in his philosophy of Judaism

Although several passages in the appear to suggest that its author, Saadia Gaon, is interested in the independent use of reason only insofar as it can affirm the truth of the Jewish revelation, such a conclusion cannot account for those occasions on which Saadia changes the very meaning of that revelation by the application of rational theory. Perhaps the most dramatic of these is Saadia's (re)interpretation of the Biblical revelations to (among others) Isaiah, Ezekiel, and Daniel in his treatment of the Divine in Treatise II of Beliefs. Not only does Saadia deduce that, because of their anthropomorphic characteristics, the revelations were of the kavod (Divine Glory) and not of the godhead itself, but Saadia states--in an innovation entirely his own, and in a break from the Talmudic tradition--that the kavod is a created object, similar to the angels (Beliefs 121). Saadia's theory radically alters the character of these prophetic revelations, as they are no longer revelations of Gcd, and in so doing problematizes the doctrines of revelation as inherited from Saadia's predecessors and as interpreted by his successers. It will be the objective of this paper to analyze Saadia's treatment of the kavod, its radical novelty, and some of its consequences, as a means to suggest that Saadia is not at all willing to sacrifice reason at the altar of revelation and instead changes the idea of revelation itself to make it more philosophically tenable.

The first step in such an examination is to try to understand why such an innovation was necessary--and whether that necessity springs from rationalist or fundamentalist criteria. Treatise II of Beliefs, of which chapter ten is Saadia's treatment of the kavod, is intended to defend Saadia's conception of the Jewish Gcd against potential adversaries, and to this end begins with a lengthy disclaimer for speaking in the abstract about Gcd. Saadia himself recognizes its unusual size and explains it in fairly strong language: "the reason for my verbosity and my dwelling at such length on this preliminary observation [i.e. that Gcd is incorporeal because He is the most subtle and developed ideas] in contrast to my customary brevity is that it constitutes the foundation and the axle of the entire book." (Beliefs 93) Saadia's insistence on the absolute incorporeality of Gcd, which prefigures Maimonides' obsessive attention to the same topic three hundred years later, will be a primary reason for his radical redefinition of the kavod: Saadia cannot tolerate either outright or implied anthropomorphism. "Those who do not seek to affirm that Gcd is a body but yet insist on arrogating for Him motion or rest or anger... act like the individual who says, 'I do not demand of Reuben 100 drachmas, but I demand of him the square root of 10000.'" (Ibid.) As Saadia introduces it, Treatise II can be seen as a proto-Guide to the Perplexed, which will assuage any discomfort in the minds of believing rationalists and convince non-believing rationalists that the Jewish idea of Gcd is a rational one.

Although apparently within the Aristotelian tradition, Saadia gives his own legitimation for why Gcd must be incorporeal. He states in the introduction to Treatise II that scientific knowledge refines the object of knowledge and thus moves gradually and steadily towards the abstract. "The idea of the Creator," in turn, "must of necessity be subtler than the subtlest and more recondite than the most recondite and more abstract than the most abstract... so that it would be impossible to fathom its character at all." (Beliefs 92) This defense prompts the first of many occasions on which we might ask to whom Saadia is writing. If he is--as some have assumed--couching a theological, revelation-based argument in philosophical terms, why use a non-scriptural argument for the abstraction of the Divine? Surely the Aristotelian Deity is more abstract than the Biblical one, so why defend ineffability and abstraction? It would seem that the implied opponent of Saadia is not the believer in an abstract-philosophical Gcd but the believer in a personal-fundamentalist One.

We will return to this point below, but for the time being it is worth noting that Saadia's insistence that Gcd must be totally unknowable plays an important part in his interpretation of revelation. Essentially, if a revelation gives information about the revealed, the revealed cannot be Gcd. So if Daniel sees the 'Ancient of Days' with a flowing beard, not only can that vision not be of Gcd, but it cannot be of an element of Gcd (i.e. Divine Glory as non-created), because the revelation would have some noetic content with respect to the Divine. Further, Saadia here seems to be giving preference to reason over revelation, because there are bounds to what can be revealed. He does not say that Gcd cannot reveal himself, but he does imply that were he to do so, humans would still not have learned anything about him.

Having shown the need for Gcd to be unknowable, Saadia now turns to the question discussed above, that of whom he is arguing against. He states on page 71 (Beliefs 91) that he is directing his argument against, first, those philosophers who reject "the idea of the Creator... on the ground that they had never seen Him... on account of the profundity of the matter and its extreme subtlety... [or because they] think that beyond the idea of Gcd there is another idea," and second, those theologians who "believe that they can picture Gcd in their imagination as a body." The latter group is clearly described, and Saadia makes clear that a fundamentalist reading of the Jewish tradition will not be acceptable. The former group is more complex, and in fact divides into three sub-groups, each of which will be discussed in turn.

Contrary to Rosenblatt's note in the Yale edition, the first group--"those who assert, 'We believe only what our eyes can see'"--are not rationalists but fundamentalists; the sentence continues "and who thus reject the sciences." (ibid.) The error here is not empiricism but anti-rationalism. Yet Saadia finishes the sentence as follows: "they have been refuted by me sufficiently in my exposition of the theory of the advocates of eternity, as well as that of the Sophists and the Skeptics." (Ibid.) So are they fundamentalists or rationalists? I do not believe this confusion to be incidental. Rather, Saadia appears to be deliberately disguising an argument against fundamentalism as an argument against rationalism, similar to the tactics of Al-Kindi in the introduction to his Metaphysics. Philosophers were the advocates of science, fundamentalists its opponents. So the sentence under analysis effectively states that fundamentalists will find their views refuted in the argument against philosophers ("Sophists and Skeptics"). Saadia is playing the familiar medieval-philosophical shell-game.

From those who reject Gcd because they have not seen them, Saadia attacks those who find his description (or anti-description) of Gcd too 'subtle and profound.' Once more, this would not be a complaint we would expect from philosophers, rather from fundamentalists clinging to a literal reading of the Bible. Next he attacks the much more ambiguous third sub-group: "those who seek something beyond the idea of Gcd," on page 78 (Beliefs 92). Once again, Saadia's refutation is complex; in fact, it does not seem to be a refutation so much as an identification of goals. Saadia states that our knowledge of Gcd is limited by our body and faculties, and it follows that since we cannot know what Gcd is, we cannot form a conception of anything beyond Gcd. Secondly, Gcd is infinite and endless; humans finite ("what is infinite and endless cannot be embraced by the [human] mind," [Beliefs 93]). So Saadia seems to be denying somehow the power of reason to know Gcd in Himself. But he does not deny reason. His third 'refutation' works "from the standpoint of the [material] basis upon which all the sciences are founded." (Beliefs 93) This was discussed above; that Gcd is the First Cause, and since he is utterly refined and abstract, he is unknowable, so it makes no sense to speak of what is beyond the unknowable. What Saadia is here doing is identifying the Gcd of the Hebrews with the Causeless Cause. If those who seek something beyond the idea of Gcd are, perhaps, philosophers, and Saadia answers their concerns by saying that Gcd is the 'last stop' on the deductive train line. Again he masks this third point, which appeals (in again proto-Maimonidean fashion) to an earlier passage in the text, with the first two apparently-anti-rationalist "refutations," which actually do not say anything about Gcd. The earlier passage in the text identifies Gcd with the First Cause. The 'refutation' of the third sub-group of Saadia's opponents is effectively a sign-post to careful readers which uses rationalist concepts as proof for Saadia's ideas, and vice versa. So, while it may seem on an initial reading of some introductory passages in Beliefs that Saadia is hostile to the project of philosophy, here it seems he is adopting it as his own. Still, it is more useful to examine what Saadia does with philosophy--not what he says about it--and it is to this topic that we now turn: Saadia's stance on the otherness of Gcd (the indescribability of Which leads to its hyphenated non-spelling, by Saadia and others, as a refusal to claim that Gcd can be ultimately named or referred to as a subject or object) does not come without its price.

This 'price' is paid largely in chapter 10 of Treatise II, in Saadia's invocation of the created-kavod for when all other attempts at reinterpretation fail. It is useful to note that he did not have to take this route. Maimonides, possibly because he was a better manipulator of language than Saadia, never "threw up his hands" (to use a metaphor both would interpret) the way Saadia did. He interpreted, for example, Isaiah's vision piece by piece using his analogical hermeneutic until there was no 'vision' left (that word being one of the terms interpreted). While a comparison with Maimonides' doctrines of the kavod and of revelation will be discussed later, I mention them now to note that, even with Saadia's rationalistic program, it would be possible to preserve the Divine aspect of Biblical theophany. Saadia, however, does not.

Saadia first sets out in Chapter I of Treatise II five "facts" about Gcd learned from the prophets and verified by reason. (There is a discrepancy between the Arabic and the Hebrew texts, only minimally relevant to the present topic; the Arabic reads 'six,' probably because in the first paragraph of Chapter 1, Saadia lists "there is nothing that resembles Him," and "He does not resemble any of his works," seperately. Ibn Tibbon translated 'five' apparently because the latter is not dealt with in the subsequent paragraphs.) He then applies reason to prove Scripture's doctrines of uniqueness, vitality, omnipotence, omniscience, and transcendence of Gcd, and discusses in Chapters two and three the incoherence of dualism and Judaism's freedom from it. He clarifies the five 'facts' in chapter four, and argues against the trinity in chapters five, six, and seven. In chapter eight, he begins the Guidesque project of purging Scripture of anthropomorphism, beginning with emotional terminology and moving to figurative language of resemblance in chapter nine. This poses more problems for Saadia than it would later for Maimonides, because Saadia's doctrine of transcendence renders the project of figurative language itself problematic. Saadia also does not perform the same sleight-of-hand as Maimonides regarding attributes; Maimonides has them actually describing nature, which, were it created by a human, that human could be said to be compassionate, gracious, etc., while Saadia tries to have the attributes describe Gcd directly, albeit figuratively. With one notable exception at the end of chapter nine ("feet=And under the throne of His glory", which will be returned to later), Saadia carries out a reinterpretation of biblical anthropomorphisms that will not look unfamiliar to any reader of the Guide. It is useful to bear in mind that Saadia preceded Maimonides by 300 years, but the problem for Saadia arises when his deft interpretive machinery runs up against a passage it cannot fix.

Peradventure, however, someone, attacking our view, will ask: "But
how is it possible to put such constructions on these
anthropomorphic expressions and on what is related to them, when
Scripture itself explicitly mentions a form like that of human
beings that was seen by the prophets and spoke to them and to which
they imputed Gcd's words, let alone the description by it of Gcd's
being borne by the angels on top of a firmament...                     
                               (Beliefs 120)

Figurative language is a suitable defense for such metaphors as "finger of Gcd" and "the Lord said in His heart," but the theophanies of Ezekiel and Michaiah (I Kings 22) cannot be so interpreted. (The use of the term 'imputed' is itself interesting, as it implies that the prophets interpreted their visions as being from Gcd and the words therefrom being Gcd's words, and that there might be some checking-procedure for verifying whether a revelation is genuine.) Saadia does not go to extremes to defend the claim of the prophets that their vision and words are from Gcd by continuing his interpretive method. Instead, he decides by deduction that such visions cannot be of Gcd. His answer to the objection:

Our answer to this objection is that this form was something
created.                      (Beliefs 121)

Saadia further specifies what created form it was: the kavod, or glory of Gcd, which Saadia--here following the Talmudic conception--also identifies with the shechinah, or 'Presence' of Gcd. Since the passage under discussion in this essay is relatively short, I have decided to quote it in its entirety (underlines are Rosenblatt's). Saadia continues his answer as follows:

Similarly the throne and the firmament, as well as its bearers,
were all of them produced for the first time by the Creator out of
fire for the purpose of assuring His prophet that it was He that
had revealed His word to him, as we shall explain in the third
treatise of this book.  It is a form nobler even than [that of] the
angels, magnificent in character, resplendent with light, which is
called the glory of the Lord.  It is this form, too,
that one of the prophets described as follows: "I beheld till
thrones were placed, and one that was ancient of days did sit." 
(Daniel 7:9) and that the sages characterized as shechina. 
Sometimes, however, this specially created being consists of light
without the form of a person.  It was, therefore, an honor that Gcd
had conferred on His prophet by allowing him to hear the oracle
from the mouth of a majestic form created out of fire that was
called the glory of the Lord, as we have explained.
                                   (ibid.)   

There are several important ideas in this passage. Chiefly, there are two innovations: that what the prophets saw was a created object, and that that object was the kavod/ shechinah. The second of these innovations will be discussed on pages 14ff. below, but there are several important consequences of the first alone. Most obvious of these is that Saadia's statement that Ezekiel, Daniel, and Micaiah, (and presumably, although not mentioned here, Isaiah) saw a created being contradicts what they themselves say. The passage Saadia himself quotes from Kings explicitly states: "I saw the Lord sitting on His throne." (I Kings 22:19) Saadia takes a fairly radical (pre-Maimonidean) stance: the prophets left out words.

If, however, it (Scripture) mentions the name 'Lord' but does not
attach to it the word 'glory' or 'angel' but only such expressions
as 'vision' or 'throne' or some human attribute, there can be no
doubt but that there is something suppressed in the utterance, the
full form of which should be 'glory of the Lord' or 'angel of the
Lord', in accordance with the practice of the language of Scripture
to leave out words by ellipsis."   
                                        (Beliefs 122)

Saadia, however, does not make reference to such a "practice" earlier in Treatise II, and only once more in the ensuing chapters, which deal with attributes of quality, relation, place, time, possession, action, and the apprehension of Gcd in the abstract, does Saadia interpret Scripture as missing words: again, in the context of theophanic (although the term is thus rendered incorrect) revelation of the kavod. (Beliefs 130; II:12). He does with some rigor reinterpret the language of attributive passages in the Bible, but does not give any other examples of Biblical ellipsis. Further, while Talmudic sources frequently cite instances where the Bible has missing details (Cain's wife, for example), nowhere(to my knowledge) do they claim that the descriptions that are given are missing fundamental terms.

Saadia's interpretation does more than contradict what the prophets themselves say; it changes the nature of what they saw as well. Saadia states that the visions of the kavod are signs and wonders designed to convince the prophets of the authenticity of their visions. While it seems somewhat ironic that Saadia himself has undermined this authenticity by claiming that the vision could not actually be Divine, it is interesting that Saadia has Gcd presenting prophets with a majestic revelation in order to convince them that the messages they are receiving are not from the revealed (lower case Saadia's?). As he mentions inII.10, Saadia makes a similar argument in Treatise III, chapter five, where he wonders "how the prophet could be sure that the message that he heard came from Gcd.... and I arrived at the conclusion that this was effected by means of some sign that would appear to him at the beginning of the colloquy and remain until its termination. This sign could take the form of a pillar of fire or a pillar of cloud or a light that did not emanate from the ordinary luminaries. When, then, the prophet saw such a sign, he was certain that the message came from Gcd." (Beliefs 151) Effectively, Saadia purges the visions of the prophets--and the children of Israel at Sinai, which will be discussed below--of any intrinsic theological value. The elaborate description of the chariot in Ezekiel, which was at the time Saadia was writing the central text for entire schools of thought within Judaism, is no more relevant to theology than the pillar of fire, because both are mere signs to verify that it is Gcd who is at work. And of course, Saadia's interpretation renders the prophets no more in possession of actual knowledge about the Divine than average people. They are privileged to receive Gcd's word, but know nothing of Him. This conclusion fits well within Saadia's schema of Gcd as being utterly unknowable becaues He is utterly refined, but fits less well with the Talmudic and even Jewish-philosophic traditions to which Saadia was an heir. In other words, his conclusion works much better with rationalism than with traditional theology.

Returning to Treatise II, chapter ten, it is important to note that Saadia distinguishes between the aforedescribed revelations to prophets and 'The' revelation of Moses. He offers as proof for his argument that the prophets saw a created object the primacy of Moses in the Jewish tradition as stated in Deuteronomy: "And there has not arisen a prophet since in Israel like unto Moses, whom the Lord knew face to face." (Deut. 34:10, quoted in Beliefs 122) While Saadia may have included this statement as evidence for his position, it is quite significant in its own right, because it reproblematizes the Mosaic revelation: perhaps Moses did have knowledge of Gcd after all. It is clear that the Children of Israel did not--and this is relevant to Saadia's discussion of how they could accept the revelation as true, as they only saw a convincing-looking object--because Saadia states in Treatise II, chapter nine that the in the account of the revelation of the sapphire throne that reads "And under His feet," the word 'feet' is shorthand for "the throne of glory." (Exodus 24:10, in Beliefs 116) The large-scale theophany at Sinai, witnessed by the seventy elders, was of the created kavod, and the vision was of the same order as the pillar of fire and the cloud: a "sign" as described in Book III, Chapter V.

Yet Saadia implies by his statement that the theophany experienced by Moses was of Gcd Himself. This is problematic on two counts: first, the entire philosophical can of worms regarding knowledge of Gcd, which caused Saadia to postulate that prophetic revelations are of something else, is here reopened, and second, Moses himself actually asks to see Gcd's Kavod in Exodus 33:18: "And he said, I beseech you, show me your glory." Gcd's response is that, if Moses will hew two new stone tablets to replace the recently-broken first set, He will "make all my goodness pass before you [ani a`avir kol tuvi `al panecha]." These are the "back parts," shown to Moses in conjunction with the enunciation of the thirteen attributes. Saadia interprets this term in Treatise II,Chapter XII of Beliefs as being the second portion of the "special light which He creates and makes manifest to His prophets," i.e. the Kavod. (Beliefs 130) Wherein lay Moses' special revelation? Since it cannot be visual, we must return to Saadia's initial statement on the topic, which reads as follows:

Now it is inconceivable that this interlocutor [in Ezekiel 2:1] was
the master of the universe, because the Toray says that the Creater
has never spoken to anyone without an intermediary except to Moshe
Rabbeinu alone.  That is the import of the statement, "And there
has not arisen a prophet since in Israel like Moses, whom the Lord
knew face to face."
                                        (Beliefs 121-122)

As an aside, Saadia's interpretation of Ezekiel implies that the lengthy preceding description of the Divine Chariot had as its rider not Gcd but the created kavod, which will have important ramifications for post-Merkavah Jewish mystics who must respond to Saadia. But the point here is the term "spoken." It is not Gcd that was seen by Moses, it was Gcd who spoke to Moses, and this is what is meant by 'knew face to face.' Consequently, only Moses had "the word of Gcd" direct from the source; all prophets got it from somewhere else. Saadia has here quite slyly done away with the most prevalent and least-acknowledged anthropomorphism in the Jewish Bible: that Gcd speaks. Of course, he has not done away with it entirely, because we are still left with the case of Moses. Saadia does not go to the Maimonidean extreme of stating that Gcd's "voice" is the "voice" of speculative reason, but he does indicate elsewhere that it is possible to have words without a mouth uttering them. Specifically, Saadia interprets "According to the mouth of the Lod (Exodus 17:1) = According to the word of the Lord." (Beliefs 116) Though this explanation remains somewhat unsatisfactory, it is outside the bounds of the present investigation. What is relevant here is that Saadia has stated (1) that all visual revelations, even Moses', are not of Gcd but of the kavod and (2) Moses is (still) unique insofar as only he is "spoken to" directly by Gcd, whereas the other prophets are "spoken to" by the kavod.

It is worthwhile to note that Saadia does not, as a fundamentalist might, claim that the contradiction of Gcd revealing Himself has only arisen because of the improper use of reason, or appeal to his own concept of miracle. There were other alternatives open to Saadia as well. He could have claimed a 'special status' for the unmediated vision of the prophets, in contrast to our own perceptions of the world, which may be faulty. Or, he could've claimed that Gcd need not be restrained by the laws of logic He himself created, and that at the very least Sinai constituted a once-in-eternity miracle (this Rosenzweigian interpretation, incidentally, averts the negation of the project of science because a logic-defying miracle happens once and only once). But Sinai is not equivalent to the giving of the manna, precisely because--setting the specific anthropomorphism of "feet" and so on aside--a revelation of Gcd Himself constitutes for Saadia an analytic impossibility, whereas the manna is logically permissible for an omnipotent being. Apparently, there are lines, dictated by reason, which Saadia will not cross.

He is similarly firm with respect to revelation and prophecy themselves. Instead of the complex artifice of a created messenger speaking the vast majority of Gcd's words in the Bible, he might've claimed that we cannot pretend to understand or limit Gcd by the use of our own reason. He certainly could've referred to his own explanation in Treatise I of how people are often in error because they apply their reason unguided by scripture. That Saadia instead claims, among other things, that scripture is itself missing words would seem to indicate that he takes his thesis that the Bible cannot contradict reason seriously, and is no mere apologist for the faith.

There is a second substantive innovation in the passage from Treatise I, chapter X, and that is that Saadia's revealed, created object is the kavod, with its associations and place in the history of Jewish theology. Saadia identifies the kavod with the Divine Presence, or Shechinah--"It is this form [the kavod]... that the sages characterized as shechinah--" (Beliefs 121, in passage cited above) following the Talmudic interpretation that Biblical verses such as Isaiah 6:3 (generally, though not always, translated "the whole world is filled with his glory [ HEBREW TEXT OMITTED]"), and the Mosaic request cited earlier, as referring to the immanent divine Presence. While the Shechinah went on to become a central concept in Kabbalah, it is perhaps more relevant here to note the large difference between the Talmudic conception of the Shechinah and Saadia's exposition in Beliefs I:10.

Onkelos' rendition of the Bible had used the term Shechinah whenever Gcd was said to dwell (Heb. shachan) in a place, such as Genesis 9:27, "And He shall dwell in the tents of Shem," rendered "And He will cause His Shechinah to dwell in the dwelling-place of Shem," and so forth. Midrashim often speak of the Shechinah dwelling among sages, with Israel in exile, and whenever "the manifestation of the Lord and His nearness to man are spoken of" (Uhrbach 43) Ephraim Uhrbach catalogs dozens of references to the Shechinah as referring to the immanence, even intimacy, of Gcd, as well as disagreements as to whether the Shechinah's indwelling was a 'natural' or 'intentional' phenomenon and finally concludes:

A survey of all the passages referring to the Shechinah leaves no
doubt that the Shechinah is no 'hypostasis' and has no separate
existence alongside the Deity.
                               (Uhrbach 63)

If the Shechinah is not a separate (created) entity, what is it? In the numerous passages Uhrbach cites, the Shechinah can best be described as representing the aspect of Gcd that is present and near in the world, different in quality but not distinct from the transcendance of Gcd. It might help to un-capitalize the terms 'glory' and 'presence' to understand how they were viewed, as properties of Gcd or descriptions of His degree of nearness with Israel (or the Kotel, as in R. Acha's statement that "The shechinah will never leave the Western Wall of the Temple, as it is said 'Behold, he stands behind our wall' (Lamentations 2:9)." (Exodus Rabba ii:2) And the Shechinah, as a decription of Gcd's nearness, is described by Rabbi Akiva as going into exile with Israel. (Mechilta d'R. Ishmael, Masechta de Pisha 12 in Uhrbach 54) also in a contemporary interpretation by philosopher Emil Fackenheim described as going into the gas chambers of the Holocaust) And in fact the Shechinah as immanent principle is enunciated by Rabban Gamaliel, who said that Gcd revealed Himself in the burning thorn-bush "to teach that no place on earth is void of the Shechinah." (Pesikta d'R. Kahana in Uhrbach 51) Urbach writes that the Shechinah is something of a theological tool: "The paradoxical concept that the specific presence of the Deity in a particular place not only does not contradict His presence throughout the world, but actually makes it possible, helped to solve a concrete, historical problem, namely the presence of Gcd among His people and the singling out of this people as His dwelling place." (Uhrbach 54) While the references to the Shechinah in Talmudic sources are somewhat diffuse, nowhere (with the one exception below) does it appear as (1) an entity distinct from Gcd, and nowhere is there any reference to it (2) being a created object. The former idea without the latter would seem to violate the concept of the unity and non-corporeality of Gcd (a price paid by many Kabbalists, who accept it in the guise of an 'emanated kavod', later, the Sefirot). To accept the latter without the former makes no sense, and yet to accept neither is to admit that Gcd can be known via revelation. Saadia's decision, to argue both, buys complete noncorporeality--a rationalist requirement--at the price of the theological preference for total veracity of the Written and Oral Torah. The position has its consequences (discussed below) but whatever they are, its origins are not in the Talmud.

If the idea of the created kavod is not in Talmudic sources, where does it come from? Urbach dates the "change of conception" regarding the Shechinah, to "the philosophical exegesis of R. Saadia Gaon," i.e. the application of Saadia's rational deduction to a scriptural problem. (Uhrbach 64) More definitively, Joseph Dan has stated that the concept of the shechinah/kavod being created is "absolutely an innovation" of Saadia's, and he does not know of any previous texts which refer to it, with only one exception. (Dan *) This exception, noted by both Gershom Scholem and by Urbach, is exactly one midrash--again, amid about forty cited in the Urbach work--which interprets Proverbs 22:29 ("Seest thou a man diligent in his business?") in part as follows:

            
When the Sanhedrin wished to count him [King Solomon] with the
three kings and four commoners, the Shechinah stood before the Holy
One, Blessed be He, and said unto Him: "Sovereign of the universe! 
Seest thou a man diligent?--They wish to count him with mean men."
                    (Midrash Proverbs, XXX, cited in Uhrbach 63)

Scholem sees here the beginning of a new conception of the Shechinah, yet at the same time stresses that the Hechalot literature that follows does not continue to develop it. (Scholem 59) Nor does Dan find any evidence for the continuity of the idea. Still, although Saadia does not mention this source, it is possible to conjecture that he may have read it. For our purposes, however, this possibility is less pertinent than the comment by one of Saadia's critics, R. Moses b. R. Hasdai, on the matter. R. Moses blames a misinterpretation of the above midrash (of course, giving his own correct one) for Saadia's and others' misconceptions about the Shechinah being a created form and charges that the source is insignificant: "We need pay no heed to the reading in Midrash Proverbs, for our Talmud takes precedence." (Ktav Tamim, Oshar Nechmad III, cited in Uhrbach 63)

This observation is in fact what I believe to be most relevant in Saadia's redefinition of the kavod: that he goes against the apparent Talmudic opinion. Dan suggested, in fact, that--given the proto-Kabbalistic agreement with the Talmudic sages on the point--the only group from whom Saadia might've learned of the idea of a created kavod would be the Karaites, although even from them we have no texts supporting such an idea. (Dan *) A direct borrowing such as this would of course be shocking in Saadia's case, but even a divergence in principle from strict adherence to Talmudic authority is surprising. If Saadia diverges from the near-unanimity of Talmudic opinion, what is to stop the Karaites from doing so? One Karaite, R. Solomon b. Yeruhim, called Saadia on this point, using rationalist arguments to attack Rabbanism. (Here I am indebted to Joseph Dan's as-yet-unpublished examination of the original texts.) Using the extravagantly-anthropomorphic (at least on the surface) Shiur Komah, attributed to Rabbis Akiva and Ishmael, R. Solomon ridiculed those who stake their beliefs in such blatant anthropomorphists. Saadia's response, which we have in the writings of a late follower, R. Yehuda ben Barzilai, is to declare the book to be pseudepigraphic invention--a dangerous reply, given that it opens a wide body of Talmudic literature up to the same charge from the Karaites. Hedging his bets, Saadia states that if R. Akiva and R. Ishmael did actually write the Shiur Komah, then it is surely about the kavod.

Apparently, then, Saadia believed himself to be following the Talmudic opinion on the Shechinah--of course, given his thesis that the Torah and reason are non-contradictory, this would not be surprising--despite the evidence to the contrary. Needless to say, this is not the opinion of Saadia's many detractors, who point out that the consequences of accepting a created-kavod view are more troubling than the reasons to do so. Most obvious of these consqequences is that the kavod is itself (and Saadia claims that it makes sense to speak of the kavod 'itself') prayed to in Jewish tradition. Although the verse's grammatical ambiguity has been the source of wide-ranging interpretations, most often from adherents of the created-kavod idea, Ezekiel 3:12--[HEBREW TEXT OMITTED] (perhaps, Blessed is the kavod of Gcd from its/His place)--is the first such example. Reacting to Maimonides' exposition of the notion of the created kavod (discussed below), Nachmanides provides the most cogent objection:

If one should say that it is a created glory, in accordance with
the view of the Master [Maimonides] in regard to the verse "and the
glory of the Lord filled the tabernacle" and others, how can we
apply thereto 'baruch' [blessed is] and ha-mevorach [the blessed]? 
Moreover, one who prays to a created glory is, as it were, an
idolater.  
                (Nachmanides, Perush ha Torah, on Genesis 46:1) 

Nachmanides adds:

However, many statements by our Sages point to the fact that the
name Shechinah stands for Gcd, may he be blessed.
                                             (Ibid.)

Similarly the Kabbalist R. Judah Hasid remarks in a commentary on the `alenu prayer that "if, as R. Saadia contends, that He is a created Glory, we would not be able to proclaim about a created Glory, that it is our Gcd." (cited in Dan, Iyyunim, 82) Although Saadia devotes many pages to prayer in Treatise V of Beliefs, primarily on why it may or may not be answered, he does not address this concern. (Maimonides does, below.) He of course condemns "worship of someone else than Gcd," in accord with the second of the ten commandments, but does not give the words of prayer the same rigorous interpretation as he did the words of scripture. (Beliefs 219) Nor does Saadia address the vaguely-Gnostic problem of an intermediary between Gcd and the world (cf. Philo and the logos). To him the kavod is a messenger/angel (the same word, of course, in the Hebrew) like any other. I do not mean to suggest that Saadia is ignorant of the consequences of his own innovation--let us recall that he only resorted to it to support "the axle of the book". (Beliefs 93) Rather, I see Saadia's deliberate choice of maintaining a rigorous philosophical conception of the Divine, at the expense of the traditional, revelation-based acceptance of It, reveals something of the nature of his program and priorities in The Book of Beliefs and Opinions.

Saadia goes to great lengths to make his theological point clear, and denies even incidental holiness of the kavod. He goes out of his way to state, in a refutation of trinitarianism, that the kavod in no way becomes holy: Christians are "compelled to assume that a physical being could become Gcd through the association with it of a divine element. They cite as an analogy the descent of the glory of Gcd on Mount Sinai and its appearance in the Burning Bush and the Tent of Meeting. Such a comparison would, however, compel them to acknowledge the Tabernacle and the Bush and the mountain also as deities, which would be going from bad to worse so far as they are concerned." (Beliefs 110) Although the phrasing of this passage is potentially misleading, Saadia is here implying that the kavod-as-object is no more divine by virtue of its association with Gcd than are the mountain, bush, and tabernacle in which it resides. Of course, once Saadia claims that the kavod is created, such terms as 'resides' are no longer problematic, and Saadia allows that the kavod takes several specific, physical shapes, sometimes also appearing as light. (Beliefs 121 and 177)

Saadia's innovation is by no means a minor one, and influences entire schools of both rationalistic and kabbalistic Jewish thought. As has been previously indicated, Maimonides adopts the idea, first giving the reader of the Guide to the Perplexed the choice of an interpretation of the term kavod ("The whole earth is full of His glory; the meaning of this verse being that the whole earth bears witness to his perfection") or Saadia's sense: "However, if you wish to consider that the 'glory of the Lord' is a created light that is designated 'glory' in every passage and that 'filled the tabernacle', there is no harm in it." (Maimonides 46, I:19) In typical fashion, Maimonides becomes more direct as the work progresses. The term 'shakhon,' to dwell, "is used in the sense of the permanence of his Indwelling--I mean his created light--in a place." (Maimonides 55, I:25, emphasis added) And again, soon after: Onkelos in Exodus 20 "referred the 'throne' to 'His glory', I mean to the Indwelling [shechinah], which is a created light." (Maimonides 60, I:28) It is interesting also to note that Maimonides follows both Onkelos' rereading of the Torah and Saadia's rereading of the concept of the shechinah (once more, lowercase his).

In chapter I:64 of the Guide, however, Maimonides characteristically confuses the issue, in an attempt to escape from the problem of praying to something created, anticipating Nachmanides criticism, by declaring 'Glory' to be an equivocal term.

The 'glory of Gcd' is sometimes intended to signify the created
light that Gcd causes to descend in a particular place in order to
confer honor upon it in a miraculous way...

However,

The expression is sometimes intended to signify 
His essence and true reality... as when he [Moses] says 'Show me,
I pray, Your Glory,' and was answered 'For man shall not see Me and
live.'  The answer indicates that the 'glory' spoken of here is His
essence...
'Glory' is sometimes intended to signify the honoring of Him, may
He be exalted, by all men... [even] those beings that have no
apprehension, as for instance the minerals, also as it were honor
Gcd through the fact that by their very nature they are indicative
of the power and wisdom of Him who brought them into existence... 
It is in this view of the notion being named 'glory' that it is
said "The whole earth is full of His glory", this being equivalent
to the dictum, "And the earth is full of His praise," [Habbakuk
3:3] for praise is called 'glory.'
                                      (Maimonides 157)

While I think this lengthy tripartite interpretation of the term Kavod serves mainly as an excellent example of the difference between Maimonides' method and Saadia's, it also makes clear the degree to which Saadia's notion was incorporated into rationalist thought and to which that stream of thought wrestled with it. While not all philosophers after Saadia agree with his interpretation of the kavod (a notable dissenter is Bahya ibn Pakuda, who writes in Hovot haLevavot that "it is impossible for us to concive of him from the perspective of the essence of His Glory, may He be blessed."), Maimonides' project indicates that he at least felt it to be a necessary solution to a philosophical problem that had severe consequences for traditional religious theory and practice.

Saadia's interpretation of the kavod had an even greater influence on the development of kabbalah. As has been convincingly argued in Dan, Torat Ha Sod, Saadia's kavod becomes the root of the Bahir's postulation of three such entities, which in turn multiply into keter (the highest sefirah, the ineffable link between the 'ein sof' and the sefirotic tree), the intermediate sefirot, and the Shechinah-as-tenth-sefirah. Scholem in Major Trends casts the history as follows:

The glory of Gcd, the Kavod, i.e. that aspect of Gcd which He
reveals to Man, is to the Hasidim [of medieval Germany] not the
Creator but the First Creation.  The idea is derived from Saadia
whose doctrine of divine glory was intended to serve as an
explanation of the Biblical anthropomorphisms and the appearance of
Gcd in the vision of the prophets....
  The importance of this conception for the religious thought of
[medieval] Hasidism is considerable.... The assertion that the
light of glory was created is, of course [!], a novelty introduced
by Saadia of which the ancient Merkabah conception of Kavod knows
nothing.  
                                   (Scholem 111-112)

Note that German Hasid R. Judah ha Hasid, in his "Book of the Glory" already distinguishes between an "inner glory" (kavod penimi) with no form but a voice, and a "visible glory" (kavod hachitzon) whose forms change with the will of Gcd, whose form convinces prophets that the voices they are hearing are not spurious. (cited in Scholem, 113) Scholem adds "The vision of the Kavod is expressly defined as the aim and the reward of Hasidic askesis." (Ibid.) So too does R. Nathan of Rome write in the year 1100 of "a Glory, which is above the Glory." (Cited in Verman, 22) Many interpret Proverbs 25:2--"It is the glory of Gcd to conceal a thing, but the honor of kings to search out a matter"--as suggesting that the kavod either hides the true essence of Gcd, or is that which is hidden; in either case, it is the kavod of Man to discover it.

The disjuncture between Saadia's kavod and the idea of the "Glory which is hidden from sight" (the 12th century Sefer Ha Iyyun, or Book of Contemplation, Verman trans., 38) form interesting rifts, outside the bounds of this paper, between those Kabbalists--primarily in Germany--who accept the created-ness of the Kavod, and those--primarily in Spain and Provence--who did not. Both views, in fact, are expressed in the Bahir, which uncharacteristically presents two conflicting parables regarding Ezekiel's "[HEBREW TEXT OMITTED]":one in which the kavod is the daughter of a far-off and invisible king, the other in which she is the invisible wife (i.e. noncontingent) of the king. In any case,Dan argues (this time contra-Scholem, although, as we have seen, not entirely) that it is in fact Saadia's doctrine of the Kavod, and not the influence of gnosticism, which sets off the chain of kabbalistic 'pleroma' in the twelfth and thirteenth centuries. Whatever the merits of this argument, and I think they are strong, the many positive and negative references Kabbalists make to Saadia's kavod themselves make a fairly strong case.

As revolutionary as Saadia's doctrine of the created kavod is, the lasting effect it had on Jewish thought would seem to indicate its usefulness in addressing the problem of the Divine acting in or penetrating the world, which can be seen as one of the fundamental contradictions between 'the Jewish Gcd' and 'the Aristotelian Gcd.' What I hope to have shown is that Saadia is not merely using Aristotelian methods to prove traditional ideas about the former, but that he takes seriously the task of proving that the two gods are the same. Because Saadia believes in the Aristotelian Gcd, as defined in the introduction to Treatise II, he cannot accept the revelation of that Gcd Himself, which is a basic claim of the Exodus from Egypt and revelation at Sinai, as traditionally understood. To make his rationalist program work, Saadia thus departs from this traditional understanding, radically reinterprets Biblical accounts of revelation, and introduces a new concept that was to take on tremendous significance in subsequent Jewish thought. While I believe that categorizing Jewish thinkers according to which method and tradition of thought they seem to prefer is to a priori deny the possibility of their projects, it would certainly seem that Saadia is being a rigorous philosopher in his interpretive endeavour. Rather than claiming that Saadia's Opinions about the rationality of Jewish Beliefs as bely a less-than-thorough philosophical worldview, I hope to have shown that the exposition of at least one of those opinions shows that Saadia takes his syncretism seriously, and, when reason demands it, changing not only the tradition received from revelation but the nature of that revelation itself, in order to make his integrative philosophical project legitimately work.

Works Cited

Dan, Joseph, Iyyunim be-Sifrut Hasidei Ashkenaz (Ramat Gan: 1975)

Dan, Joseph, Torat Ha-Sod, (Academon: 1977)

Dan * -- refers to conversations held with Professor Joseph Dan by the writer.

Fackenheim, Emil, To Mend the World (New York: Schocken, 1982)

Maimonides, Moses, The Guide of the Perplexed (Shlomo Pines trans., Chicago: University of Chicago, 1963)

Nachmanides, Moses, Commentary on the Torah (Volume I: Bereshit, Chavel trans., New York: Shilo, 1971)

Saadia ben Joseph haGaon, The Book of Beliefs and Opinions (Samuel Rosenblatt trans., New Haven: Yale, 1948)

Scholem, Gershom, Major Trends in Jewish Mysticism (New York: Schocken, 1941)

Urbach, Ephraim, The Sages: Their Concepts and Beliefs (Israel Abrahams trans., Jerusalem: Magnes/Hebrew University, 1979)

Verman, Mark, The Books of Contemplation (New York: State University of New York, 1992)

Back to library. Back to homepage. Email feedback to the author.