The results of the present study allow us to reaffirm the evaluation of Ibn al-Rawandi as a heresiarch, without disregarding Maturidi's important evidence. A close analysis of Maturidi's text in the context of the other sources showed that there is no contradiction between Maturidi's account and those of Muayyad and Abd al-Jabbar. Rather, when the accounts of these three

INTRODUCTION

A CENTRAL ISSUE IN EARLY ISLAMIC THEOLOGICAL DISCUSSIONS is the question of who is to be counted as a believer (mumin). In asking this question medieval Muslim thinkers sought to define the essence of Islam and to delineate the borders of the Muslim community. Their views were sometimes formulated in positive terms, but quite often they addressed the issue in a roundabout way, seeking to define who is not a believer. In this context, Muslim writers pay relatively little attention to non-Muslims, who are seen as outright infidels, and concentrate on the study of those who are seen as heretics, that is to say, people who are supposedly Muslims, but who adhere to non-orthodox beliefs. The study of accusations of heresy can thus serve as an important tool in the attempt to understand the development of Muslim self-perception.

Not all heresies were seen as equally deviant. While some opinions, though fiercely disputed, were treated as wrong interpretations of Islam, others were viewed as definitely outside Islam.(2) The present study will focus on the case of one of the most notorious heretics, Ibn al-Rawandi. In Muslim consciousness Ibn al-Rawandi has become a byword for denial of the phenomenon of prophecy. But in the last few decades scholars have given attention to certain Muslim texts that seem not to share this harsh view. Josef van Ess, in particular, has offered a radical reinterpretation of this thinker, arguing that in Ibn al-Rawandi's eyes, as well as in the eyes of many of his contemporaries, his views concerning prophecy were still acceptable for aMuslim.(3)

Our knowledge of Islamic theology in the third century is admittedly still very deficient. It is mostly dependent on later sources, and as new sources become available, we must be prepared to rethink traditional views. Some ideas, about the legitimacy of which later Muslim orthodoxy had a definite opinion, were probably still open for debate in the third Islamic century.(4) Views that could be defined as heretical in the fifth century may well have been treated with more leniency in the third and early fourth. But prophecy in general, and the prophethood of Muhammad in particular, was not one of those open issues.(5) By the third century, belief in prophecy was not open for discussion; a person who denied that Muhammad was a prophet stood outside the pale of Islam. Ibn al-Rawandi's case, when properly understood, vividly demonstrates this fact.

The present study will deal with what is perhaps Ibn al-Rawandi's most notorious book, "The Book of the Emerald" (Kitab al-Zumurrud, henceforth, Zumurrud). Part I of this paper presents in brief the available data concerning Ibn al-Rawandi and the Zumurrud, and reviews the relevant scholarly opinions. Part II is a detailed analysis of one passage from the Zumurrud. Part III reviews the Muslim sources that are supposedly favorable to Ibn al-Rawandi, and discusses the connection between Ibn al-Rawandi's heresy and the form of the Zumurrud; and part IV is an attempt to summarize what we know about the Zumurrud and its contents.

I. IBN AL-RAWANDI AND THE KITAB AL-ZUMURRUD

1. IBN AL-RAWANDI

If one were to write a biography of Ibn al-Rawandi that contained only the information on which there is consensus, it would be quite short.(6) Abu al-Husayn Ahmad b. Yahya b. Ishaq al-Rawandi was born in Marwarrudh about the year 815 A.D. He joined the Mutazila of Baghdad, and gained prominence among them. But when he approached the age of forty he became estranged from his fellow Mutazilites, and formed close alliances with non-Mutazilites, both Muslims (Shiis) and non- Muslims (Manichaeans, Jews and perhaps also Christians). He wrote against the Mutazila, and they reciprocated in kind.

Our sources contain much material beyond this short biography. But for every detail other than those mentioned above there are at least two contradictory versions. Some sources suggest that Ibn al-Rawandi may have died eround 860, others that he lived to the year 910.(7) And while most of the sources describe him as an outspoken and dangerous heretic, some appear to present him in a neutral or even positive light.

These differing perceptions of Ibn al-Rawandi are reflected in Ibn al-Nadim's Fihrist (compiled ca. A.D. 988), which offers two lists of Ibn al-Rawandi's books. One contains respectable scholarly publications, the other an impressive collection of heretical works: in support of the eternity of the world, against the idea that God is wise, against the Quran, against the prophet Muhammad, against all prophets. This dichotomy is often explained in the sources as the result of the development of Ibn al-Rawandi's thought. Thus al-Balkhi (d. 931), cited by Ibn al-Nadim, can speak of the books written by Ibn al-Rawandi in his righteous period (ayyam salahihi), as opposed to those written after he had gone astray (ayyam fasadihi).(8)

Modern scholars are as divided as the medieval authorities. It is generally agreed that Ibn al-Rawandi was indeed a heretic, but there is no agreement as to the nature of his heresy. Some look for the roots of his heresy in his connections with Shiism,(9) and depict him as a Mutazilite gone wild;(10) some regard him as an Aristotelian philosopher,(11) while others see him as a radical atheist,(12) and some stress the political challenge he presented to the Islamic polity.(13)

At the same time, scholars try to account for the more positive view of Ibn al-Rawandi in some Muslim sources.(14) Josef van Ess in particular has suggested an original interpretation that aims at accommodating all the contradictory information. Van Ess notes that the sources which portray Ibn al-Rawandi as a heretic are predominantly Mutazilite and stem from Iraq, whereas in eastern texts he appears in a more positive light. As an explanation for this difference, van Ess suggests "a collision of two different intellectual traditions," i.e., those in Iran and in Iraq. He further suggests that Ibn al-Rawandi's notoriety was the result of the fact that after Ibn al-Rawandi left Baghdad, "his colleagues in Baghdad ... profiting from his absence ... could create a black legend."(15) In other words, van Ess believes that Ibn al-Rawandi, although admittedly eccentric and disputatious, was not a heretic at all.

2.THE KITAB AL-ZUMURRUD

A. Facts and Probabilities

Although the contradictory information about Ibn al-Rawandi touches on various aspects of his personality, the greatest difficulties concern his book against prophets, the Kitab al-Zumurrud, regarding which the sources appear to offer two diametrically opposed accounts. I shall begin by setting out some facts that we know for certain about the Zumurrud, and some inferences that can be drawn with a high degree of probability from such facts. This will be done on the basis of only those sources that explicitly mention the Zumurrud.

i. The character of the book. There is unanimous agreement that the Zumurrud was a book that was directed against prophecy. Al-Khayyat, for instance, describes the book in the following words:

The book known as Kitab al-Zumurrud, in which he [i.e., Ibn al-Rawandi] mentioned the miracles of the prophets, peace upon them, such as the miracles of Abraham, Moses, Jesus and Muhammad, God's blessing on them. He disputed the reality of these miracles and claimed that they were fraudulent tricks (makhariq) and that the people who performed them were magicians and liars; that the Quran is the speech of an unwise being, and that it contains contradictions, errors and absurdities. He included in it a chapter entitled: "Against the Muhammadans in particular," meaning the community of Muhammad, God's blessing on him.(16)

ii. The authorship of the Zumurrud. Almost all the sources present Ibn al-Rawandi as the author of the Zumurrud. Only two sources deviate slightly from the consensus. In a passage of the Tathbit dalail al- nubuwwa of Abd al-Jabbar al-Hamadhani (d. 1025), the Zumurrud appears twice: once as a book written by Abu Isa al-Warraq,(17) and two lines later as written by Ibn al-Rawandi.(18) And Ibn al-Jawzi (d. 1201) says that Ibn al-Rawandi and al-Warraq each charged the other with writing this book, although they agreed on the criticism of the Quran that the book contained.(19)

iii. The form of the Zumurrud. The book contained arguments both for and against the existence of prophets. This is stated explicitly in the Majalis Muayyadiyya of the Ismaili al-Muayyad fi'l-Din al-Shirazi (d. 1077). The Majalis Muayyadiyya is the only source that contains relatively long citations of arguments against prophecy which are identified explicitly as belonging to the Zumurrud.(20) Al-Muayyad does not claim to have read the Zumurrud itself; he says that he relies on a refutation of the book written by an Ismaili missionary (dai), and he claims to quote from this refutation. In the quotation the dai says:

We have come across an epistle composed by Ibn al-Rawandi. He called it The Emerald and attributed it to the Barahima.(21) The epistle is about the rejection of the existence of prophecy. In it he [i.e., Ibn al-Rawandi] listed arguments advanced by those who support the existence of prophecy, and others presented by those who deny the existence of prophecy.(22)

Books dealing with kalam often take the form of a dialogue. There were two standard ways of presenting such a dialogue. One was for the author to respond to a hypothetical interlocutor, so that the form would be in qala ... qulna ("if he were to say ... we would reply" ) or in qala ... yuqal lahu ("if he were to say ... he should be told" ). The other was to present the dialogue as having occurred between two real people.(23) The Majalis Muayyadiyya shows that the dialogue in the Zumurrud was of the second kind. Discussing a certain passage from the Zumurrud, Muayyad cites "the heretic" Ibn al-Rawandi and "his opponent" (khasmuhu) and reports their exchange of arguments. This passage, whose evidential value on this point has not been noted before, shows that the Zumurrud was constructed as a dialogue between two participants, the one arguing for the existence of prophecy, the other against it.(24)

iv. The participants in the dialogue. In such dialogues one of the participants was often the author himself.(25) It is therefore likely that one of the participants in the Zumurrud was Ibn al-Rawandi.(26) Some of the sources indeed use the phrase: "Ibn al-Rawandi said in the Zumurrud."(27) But such phrases do not, in themselves, prove that Ibn al-Rawandi was represented in the Zumurrud as one of the participants in the dialogue. They could also mean that, as the author of the book, Ibn al-Rawandi was held responsible for its content, regardless of whether or not he presented himself as one of the participants.

As to the other participant, there is reason to suspect that he was Abu Isa al-Warraq. Al-Warraq is often mentioned as Ibn al-Rawandi' s colleague and mentor,(28) and both Abd al-Jabbar and Ibn al-Jawzi associate al-Warraq specifically with the Zumurrud.(29)

v. The roles of the participants. Which role was assigned to which speaker? Given that the Zumurrud was a book whose purpose was to disprove the existence of prophecy, and that Ibn al-Rawandi was its author, one would expect Ibn al-Rawandi to take the role of the one who denies the existence of prophecy. But this could hardly be the case if his interlocutor was indeed al-Warraq. Al-Warraq is consistently portrayed in the sources as denying the existence of prophecy, whereas Ibn al- Rawandi is portrayed as sometimes arguing against the existence of prophecy and sometimes arguing for it. The balance of probability is therefore that in the Zumurrud it was al-Warraq who had the role of the one who attacks prophecy. This would leave the role of the defender of prophecy to Ibn al-Rawandi.

While the first two points are fairly certain, the third and fourth points rely mainly on the Majalis Muayyadiyya, and the fifth point is based wholly on inference. I now want to show that these probabilities can be strengthened by looking at some sources which do not refer to the Zumurrud by name.

B. Maturidi's Kitab al-Tawhid

The most problematic of these sources is the K. al-Tawhid of Maturidi (d. 942).(30) Whereas the Majalis Muayyadiyya presents Ibn al-Rawandi as someone who denies the existence of prophecy, Ibn al-Rawandi appears in Maturidi as one who argues in favor of prophecy. For this reason, Joseph Schacht, Wilferd Madelung and (in the past) I myself, assumed that the relevant passages in Maturidi reflect the views of Ibn al- Rawandi before he wrote the Zumurrud, i.e., before his break with Islam.(31)

In sharp contrast, van Ess suggests that the passages cited by Maturidi are derived from the Zumurrud;(32) and he further implies that the Zumurrud was written in the form of a dialogue in which al-Warraq presented arguments against the existence of prophecy, while Ibn al- Rawandi presented arguments in its favor. Van Ess advances his ideas with some caution, saying that "the final decision [concerning the form and nature of the Zumurrud] must be left to a detailed philological analysis."(33) The main reason for van Ess' caution is his assumption that the evidence of Maturidi is inconsistent with the evidence of the Majalis Muayyadiyya. Van Ess thinks that the dialogue form of the Zumurrud can be seen only in the passages cited by Maturidi,(34) whereas the Majalis Muayyadiyya offers a different picture of the Zumurrud, a picture which van Ess believes to be distorted.(35)

In fact, however, as we have just seen, van Ess' points can be inferred with reasonable probability from the Majalis Muayyadiyya and other sources, without any reference to al-Maturidi. That these points fit well with the relevant passages cited by Maturidi constitutes in fact an independent argument for the view that the passages cited by Maturidi are derived from the Zumurrud. In the following section I shall offer another argument to the same effect, establishing it on the kind of detailed analysis that van Ess calls for. Through an analysis of one passage from the Zumurrud I will show that Maturidi, like Muayyad, draws from the Zumurrud itself, and that the information that each one of them offers complements rather than contradicts what the other has to say.

II. THE MUBAHALA VERSE IN THE KITAB AL-ZUMURRUD

1. THE BACKGROUND

According to Muslim tradition, Q. 2:94 was revealed when the Jews of Medina claimed arrogantly that the hereafter belonged to them alone. In Muhammad's presence, the Jews reigned confidence in the bliss awaiting them in the world to come. But God revealed to Muhammad that the Jews were in fact afraid of him, and that if he challenged them to wish for death, they would not dare to do so. Muhammad then presented the Jews with this challenge (tahaddi) and, indeed, the Jews never expressed a wish for death, neither then nor on a second occasion, when Q. 62:6- 7 was revealed. The commentators note the fact that the Jews were eager to prove that Muhammad was not a real prophet. They fought Muhammad with all their might, sacrificing their property and their dear ones; yet they were not willing to add such a simple, easy thing as saying that they wished for death. The fact that Muhammad predicted this strange behavior is taken by the commentators as proof that he was a prophet.(36)

Like the Jews, the Christians of Najran did not accept Muhammad as a true prophet. When they came to sign an agreement with Muhammad, they had not planned to accept his terms. Muhammad challenged them to let God be their judge by calling down His curse on the party which lied, a challenge which is alluded to in Q. 3:61 (the mubahala verse).(37) After a consultation with the Jews, the Christians decided not to risk the ordeal and accepted Muhammad's terms. In Muslim tradition the two confrontations are frequently connected, and both are regarded as proving that Muhammad was indeed a prophet.(38) The term mubahala is used by the commentators in reference to both these verses, and we can thus call them "the mubahala cluster."

The commentators mention other verses which they vegard as proofs that Muhammad was a prophet who possessed secret knowledge. Among such verses are Q. 29:48 and Q. 48:27. Muslim tradition regards Q. 29:48 as referring to Muhammad's illiteracy, and as implying that Muhammad's knowledge of the content of the Jewish and Christian scriptures must have come to him through revelation.(39) Q. 48:27 is seen as referring to Muhammad's prediction of his victory over the Meccans.(40) It should be noted, however, that unlike the verses of the mubahala and the tahaddi, which the commentators constantly mention together, Q. 29:48 and Q. 48:27 are not, as a matter of course, connected with each other or with the first two.

Muhammad's confrontations with the Jews and Christians are discussed in the Zumurrud, and fragments of the discussion are preserved in several sources.

2. THE MAJALlS MUAYYADIYYA

Let us first examine quotations from the Zumurrud which are explicitly ascribed to it, those found in the Majalis Muayyadiyya. As mentioned above, Muayyad claims that his source was a refutation of the Zumurrud by an Ismaili dai. This dai's refutation contained arguments against the existence of prophecy, derived from the Zumurrud, and responses to those arguments, written by the dai himself. Although the Zumurrud had already presented arguments for the existence of prophecy, the dai preferred to replace them with pro-prophecy arguments of his own.(41) The voice of the proponent of prophecy in the Zumurrud is thus usually silenced by that of the dai. But in one rather obscure passage this voice emerges unannounced. The difficulty as well as the importance of this passage were already noticed by Paul Kraus. What has remained unnoticed is the fact that this short passage provides the key to a correct understanding of the original structure of the Zumurrud.(42)

The exchange of arguments in the dialogue is not quoted in full. Consequently, both the arguments and the identities of those presenting them remain obscure. In what follows, I divide the passage into numbered paragraphs. The following abbreviations are used in identifying the persons referred to: IR is Ibn al-Rawandi; D is the speaker who denies the existence of prophecy; and P is the speaker who supports the existence of prophecy.

1. The heretic [IR] refuted the mubahala verse [Q. 3:61] and what is said concerning the circumstances under which it was revealed; he rejected [what is known to be] the [true] meaning of God's saying "Then wish for death, if you are speaking the truth" (Q. 2:94, 62:6); and he also mentioned various other verses of the same kind [and refuted them].

2. The dai said in response to this that if he [D] and his opponent [P] agree about the meaning of these verses, then the way is open to him [D] to refute them and to reject what they imply. But if his opponent [P] says, "The meaning of these verses is not what the assumptions on which you [D] base your reckoning demand," then his [D's] refutation is invalid and his exertions have been in vain.

Exactly the same applies also to his [D's] refutation of God's words, "You have not recited a book before ..." [Q. 29:48] and his commentary on God's words, "You shall enter the holy mosque, if God so wills" [Q. 48:27] where he [D] says that these were mere guesses, and not authoritative announcements of what he [Muhammad] desired should occur.

3.The opponent [D] in his eagerness to refute, lumped all these verses together in his interpretation, without pausing to reflect; the impropriety of this is obvious.(43)

As mentioned previously, much of this passage is unclear. For example, the text of the Majalis Muayyadiyya gives no explanation of "the assumptions on which he bases his reckoning."

The passage preserves some remnants of the dialogue form of the Zumurrud, but the exchange of arguments in the dialogue is not quoted in full. We can see the presence of two speakers in the Zumurrud. The form of introduction which al-Muayyad uses here ("If his opponent says ...") results from the fact that Muayyad summarizes rather than quotes, and does not indicate in this case that the dialogue is hypothetical. We may assume that in the Zumurrud itself the discussion had the form of an actual debate ("He said ... Then he answered ...").

As to the identity of the speakers, it had been assumed by scholars that the Majalis Muayyadiyya shows that Ibn al-Rawandi presented in the Zumurrud the arguments against the existence of prophecy. But a close reading of this passage--the only passage in the Majalis Muayyadiyya where the two speakers are actually heard--calls for a different interpretation. Let us summarize this passage:

1. The dai says that the heretic disputed the mubahala verse (Q. 3:61) and Q. 2:94 and the relevant Muslim traditions concerning the circumstances of their revelation.

2. Rather than present the arguments of D and refuting them, as he usually does, in this passage the dai summarizes the arguments of both D and P in their discussion of these verses. D attempts to interpret these verses by introducing an assumption that will enable him to criticize them, and his opponent P (khasmuhu) rejects the assumption. D mentions Q. 29:48 and Q. 48:27.

3. The dai says that the opponent (al-khasm), in his eagerness to refute the Quran, grossly misunderstood those verses.

The "heretic" in the first paragraph, as everywhere in Muayyad's account, is Ibn al-Rawandi, the author of the Zumurrud. We then have three possibilities, all problematic.

One's first inclination would be to identify Ibn al-Rawandi with the speaker who criticizes the Quran in the second paragraph (D) and with the "opponent" in the third paragraph. This would mean that Muayyad applies the word "opponent" (khasm) in the same short passage to two people who hold opposite views. For "the opponent" is harshly criticized by Muayyad, whereas "his opponent" is clearly a proponent of the existence of prophecy.

A second possibility is the one suggested by Kraus, who identified Ibn al-Rawandi with the speaker who criticizes the Quran in the second paragraph (D). Kraus did not realize that the passage reflects the dialogue in the Zumurrud. He suggested that the words "the opponent" and "his opponent" in the second and third paragraphs refer to the same person, and that this person is not Ibn al-Rawandi, but rather "a Muslim writer who had attempted to refute the Zumurrud before the author" (i.e., before the dai, whom Kraus believed to have been Muayyad himself).(44) Kraus's suggestion seems rather arbitrary. Nowhere in the Majalis Muayyadiyya is there evidence for the existence of such a "previous Muslim writer." Nor is there any other case where al-Muayyad abruptly abandons his attack on Ibn al-Rawandi and turns to chide someone else; and the language Muayyad uses to berate "the impropriety" of "the opponent" seems much too strong for criticizing the wrong methods of a Muslim who defends the Quran. And though Kraus's suggestion avoids a proliferation of "opponents," it does not address the other difficulties of this passage.

A third possibility is to accept Kraus's idea that the term "opponent" must refer at both occurrences to the same person, but to assume that throughout the passage Muayyad continues to target the Zumurrud. "The heretic" of the first paragraph, "his opponent" of the second paragraph and "the opponent" of the third would then be one and the same person, namely, Ibn al-Rawandi. This interpretation implies that the pronoun "he" changes its reference in the middle of the passage. Such a change is quite plausible, given the common tendency to use pronouns in an unclear manner and the obscure wording of this particular passage. This interpretation also means that Ibn al-Rawandi, who is described in the beginning of our passage as seeking to refute the Quran, appears in the second paragraph as the one who seeks to defend it. This would be possible only if we assume that Ibn al-Rawandi, the author of the heretical Zumurrud, is not the one who presents the attack on the Quran (D) but rather "his opponent" (P).

None of these three possibilities is entirely without problems, and the Majalis Muayyadiyya does not provide us with enough data to decide between them. I believe, however, that the third possibility is the least problematic, and I shall presently try to show that other sources support this view.

In another passage Muayyad attacks Ibn al-Rawandi for things that he said, "citing one of those who reject the existence of prophecy." (45) This could be taken to mean that Ibn al-Rawandi presented the attack on prophecy as he had heard it from someone else (the Brahmans, for example). But if our last suggestion is correct, then this passage may indicate that the author of the Zumurrud left the role of presenting the arguments against the existence of prophecy to someone else, and that this someone was not only a vague ethnic group of Brahmans,(46) but a specific person, someone known for his disbelief in prophecy. It would follow that Ibn al-Rawandi took upon himself the role of the proponent of prophecy (P).

The mubahala passage in the Majalis Muayyadiyya is too cryptic to allow us to go any further.(47) In order to identify the speaker who attacks the Quran and denies prophecy in the Zumurrud and to understand his relation to Ibn al-Rawandi, we must now turn to Maturidi's K. al-Tawhid.

3. MATURIDI

As mentioned above, Maturidi does not refer to the Zumurrud by name. But his book contains a dialogue on prophecy in which the arguments against the existence of prophecy bear a striking resemblance to those cited by Muayyad and by Ibn al-Jawzi as part of Ibn al-Rawandi's Zumurrud. Maturidi identifies the participants in the dialogue clearly: Abu Isa al-Warraq, who attacks the prophets, and Ibn al-Rawandi, who responds, defending them. It is the passage on the mubahala verse in Maturidi' s work that proves that Maturidi was using the Zumurrud. The relevant passage in Maturidi reads:(48)

We have already set out some of the arguments that show that Muhammad was a prophet. In addition to this, he [i.e., IR] cites Muhammad's saying to the Jews, "Then wish for death."(49) He [IR] argues from it in two ways [to show that Muhammad was a prophet]. First, there is [God's] pledge that if they wish for death, they will die. Second, that though nothing would be easier for them than to wish for death, they will never do so. He [IR] also used the mubahala with the Christians, and [Muhammad's] announcement that [a divine] curse would befall them to prove that Muhammad had already been described in their scriptures.

Al-Warraq offered two retorts to this. One was that if the Jews had indeed expressed verbally the wish to die, then it would have been said [by Muhammad] that what he [Muhammad] meant was that they should wish in their hearts to die. The other was that they [the Christians] believed in Moses and Jesus, and that these two had predicted to them [the coming of the prophet Muhammad] in the same way that an astrologer predicts things.

The answer to the first argument is that the mubahala will not bear this interpretation.(50) Furthermore, the Jews are clever people, so that if they had been given such a reply, they would have retorted that they also wish in their hearts to die.

As to the second argument, the answer is that had it been so [i.e., that upon learning of the prophecies of Jesus and Moses, the Christians surrendered to Muhammad], then they would not have refused to accept Muhammad when God said, "You shall indeed enter the holy mosque,"(51) and His saying "That He may uplift it above every [other] religion." (52)

. . . And he [i.e., IR] criticized him [i.e., al-Warraq, saying] that if this [i.e., what Moses and Jesus prophesied] had been a prediction of the same kind as those made by astrologers, then it would not have been repeated by them [so frequently] . . .(53)

. . . And he [i.e., al-Warraq] attacked the verse "You have not recited a book before,"(54) saying that memory can take the place of a book.(55)

In this passage Maturidi refers to all the verses mentioned by Muayyad, and almost in similar order.

Muayyad: Q. 3:61, 2:94, 29:48, 48:27

Maturidi: Q. 2:94, 3:61, 48:27, 29:48

The fact that neither source offers exact quotations may account for the slight difference in the order of the verses referred to, as well as the different choice of words. Notwithstanding these different elaborations, the similarities of the two texts are remarkable. It is not only that, as had been noted in the past, they contain similar arguments. Rather, the two texts present the same combination of arguments, attached to the same Quranic verses, treated in practically the same order. Such a similarity is too great to be explained as the result of quotations from two books on the same subject. In other words, Maturidi, like Muayyad, is probably relying on the Zumurrud itself.

This conclusion is strengthened by the fact that taking the two texts together one gets a better grasp of the original argument in the Zumurrud than from either one alone. Maturidi's account leaves much to be desired, but he is at least clearer than Muayyad. Only with Maturidi's help can we make sense of Muayyad's allusion to "the assumption." From Maturidi's presentation it appears that what is meant by "assumption" is the notion of the heart's intention which al-Warraq tries to introduce into the discussion of the "wish for death." The information given by Maturidi and Muayyad is in fact complementary rather than contradictory.

The analysis of the mubahala passage in the Majalis Muayyadiyya indicated that in the Zumurrud a person whose name we are not told argued against the existence of prophecy and that Ibn al-Rawandi acted as "his opponent." Maturidi's version of this paragraph of the Zumurrud corroborates this analysis and allows us to identify the first speaker as al-Warraq. Both the Majalis Muayyadiyya and Maturidi thus show that Ibn al-Rawandi, the author of the Zumurrud, presented in it the views of the proponents of prophecy, whereas the role of the speaker who denies the existence of prophecy was taken by al-Warraq.

4. THE TATHBIT DALAIL AL-NUBUWWA

The mubahala cluster appears in still another source, the Tathbit dalail al-nubuwwa of Abd al-Jabbar.(56) Abd al-Jabbar cites arguments against the existence of prophecy which closely resemble those of the Zumurrud, but he does not identify the source of the arguments, and his formulation of these arguments differs slightly from the formulation in the Majalis Muayyadiyya. Scholars are therefore hesitant in identifying these fragments as belonging to the Zumurrud.(57) In fact, the mubahala cluster allows us to see not only that the Tathbit cites the Zumurrud, but that the information it gives is entirely consistent with what we learn from the two other sources.

In the Tathbit, Abd al-Jabbar deals at length with the mubahala cluster and its relevance to prophecy. He adduces the arguments for and against the existence of prophecy that we have already seen cited by Muayyad and Maturidi. And, most interesting, Abd al-Jabbar seems unable--or unwilling--to make up his mind as to who presented the arguments against prophecy. Was it one person (to be addressed in the singular) or more than one (to be addressed in the plural)? Was it Ibn al-Rawandi, or Ibn al-Rawandi in the name of al-Warraq, or perhaps both of them together? The relevant passages read:

1. . . . The heretics were confounded by these verses [i.e., Q. 2:94- 95, and Q. 62:6-7]. . . . This is why they curse [Muhammad's] Arab [opponents, saying], "Why did they not produce an equivalent to the Quran?" And [they curse] the Jews, [saying], "Why did they not wish for death? [Had they done so] they would have proved that he [i.e., Muhammad] was a liar; they would thus have been rid of him, and so would we!" This is what people like al-Haddad and his friend Abu Isa say, may God disgrace them.(58)

2. Ibn al-Rawandi said that al-Warraq used to say: "The only reason that they did not wish for death is that the Jews and the Christians believed in Moses and in others who claimed to be prophets. In their books these [prophets] had announced the coming of Muhammad as a prophet, therefore they [i.e., the Jews] did not dare to wish this [i.e., to wish for death] . . .

He was told: This, then, proves the prophethood of those prophets as well as the prophethood of Muhammad, so you [pl.] are obliged [lazimakum] to believe in all their books, whereas you [pl., antum] deny all of it!

He [sic] said, "The ability of these people [i.e., Moses and Jesus] to predict the coming of Muhammad is similar to the ability of astrologers to predict the future." He was told, "And since when can astrologers predict something like this, that is to say, something like the advent of Muhammad, the date of his appearance, the content of his message, his country of provenance, the generation to which he belongs and his ancestry, in the same detailed manner that this was announced? This kind of prediction is not to be found in the announcements of the most proficient astrologers. They cannot even come close to it or do anything of a similar nature. Only by chance they [i.e., the astrologers] may happen to hit on a general, marginal thing, after a thousand cases of errors and mistakes.(59)

. . . Then these zindiqs say: The only reason why they did not wish for death is that, had they expressed such a wish, he [i.e., Muhammad] would have answered that he had meant them to wish it in their hearts. Had they said to him, "We did wish it in our hearts," he would have answered "Gabriel told me that you did not."(60)

3. Abu Isa al-Warraq and Ibn al-Rawandi criticized the story of the mubahala, [saying] that it is a mutual curse, and that these people [i.e., the Christians] withdrew from it of their own accord. He [sic] said: "Your claim that he [i.e., Muhammad] said to them [i.e., the Christians]: 'If you accept to have a mubahala with me, a God-sent revenge will hit you'--that is not in the Quran, it is only one of your hadiths."(61)

4. Ibn al-Rawandi claimed also that when he [i.e., Muhammad] had challenged$the Christians to the mubahala, or [when he challenged] the Jews to wish [for death], this was not meant to prove that he was a prophet; had this been his intention, they would have been quick to respond.(62)

Many of the central elements in these passages are also to be found in the citations of Muayyad and Maturidi. Abd al-Jabbar mentions the claim that the Jews and the Christians had foreknowledge of the coming of Muhammad, due to the prophecies of Moses and Jesus. He also mentions the comparison of these two prophets to astrologers, and retorts that astrologers can predict the future only in general terms, and that they hit on the truth only rarely. He dwells on the question of whether the wish for death meant an oral expression alone or had to include also the intention of the speakers. And he mentions the connection of these arguments to Q. 3:61 and 2:94.(63) It appears, therefore, that although Abd al-Jabbar does not name the Zumurrud, he is in fact drawing from it. And Abd al-Jabbar, who is aware of the fact that Ibn al-Rawandi was citing al-Warraq, holds both of these heretics accountable for the arguments against the Quran and against the existence of prophecy.

5. THE MUBAHALA CLUSTER IN THE ZUMURRUD

Maturidi, Abd al-Jabbar and Muayyad belong to three different schools of thought. It is not likely that Abd al-Jabbar or Muayyad would have copied from Maturidi, and it is only remotely possible that Muayyad would have copied from Abd al-Jabbar. Moreover, notwithstanding the similarities in their treatment of the mubahala cluster, each of the three has some information which is not found in the other two. It is therefore evident that they do not draw their extracts of the Zumurrud from each other.(64)

We are now in a position to summarize our analysis of the mubahala cluster. Ibn al-Rawandi wrote the Zumurrud in the form of a dialogue with al-Warraq, in which the latter presented the arguments against the existence of prophecy and Ibn al-Rawandi responded, defending prophecy and presenting the arguments in favor of its existence. The discussion does not seem to have been very systematic. Al-Warraq presented an argument, and when Ibn al-Rawandi responded, al-Warraq either answered again to the same point, or moved to attack the same verse from another angle, or to criticize another verse.(65) Although much of the discussion is still obscure, the three texts taken together allow us to reconstruct one part of it, and to assign each participant his role, as follows:

Ibn al-Rawandi: In Q. 2:94 Muhammad challenged the Jews to wish for death. Muhammad was confident that if the Jews would wish for death, they would immediately die. [The Jews knew that too, and therefore they did not wish for death. This shows that, despite their refusal to accept Muhammad, the Jews knew that he was a true prophet.]

Muhammad also knew that, despite the certainty they feigned, the Jews would not dare to express the wish for their own death(66) [and the verse says explicitly that they will not do so. This proves that Muhammad was a prophet, because without God's help, he would not have been able to guess what the Jews would do].

Al-Warraq: [If the Jews did not accept Muhammad's challenge to wish for death, this was not because they believed that he was a prophet, but rather because they did not take his challenge seriously.]

When Muhammad challenged the Christians to take part in the mubahala and the Jews to wish for death, the words he used did not imply that he was doing so in order to prove that he was a prophet. Had the Jews and Christians realized that this was his intention, they would have been glad to accept the challenge.(67)

Ibn al-Rawandi: Q. 3:61 refers to the mubahala with the Christians. Muhammad warned the Christians that [if they did not submit to him] they would be cursed. [This proves that he was a prophet, because he knew what the outcome of the mubahala would be, should it take place. The Christians preferred to accept Muhammad's terms rather than perform the mubahala and be cursed. This proves that] the Christians realized that he was a true prophet. They knew it because their scriptures foretold the coming of a prophet whose description is like Muhammad' s.(68)

Al-Warraq: People who claimed to be prophets, like Moses and Jesus, had indeed foretold the coming of Muhammad as a prophet. The Christians and Jews believed in prophets; therefore, they did not dare to respond to Muhammad's challenge.(69)

Ibn al-Rawandi: But if you agree that Muhammad is described in the scriptures of the Jews and Christians, then you must admit that prophecy exists, and that Moses and Jesus, as well as Muhammad, were prophets.(70)

Al-Warraq: Moses and Jesus did indeed predict the coming of Muhammad [but this does not imply the existence of prophecy or that these people were prophets]; any astrologer can make correct predictions.(71) In the same way, the fact that Muhammad could predict certain events does not prove that he is a prophet; he may have been able to guess successfully, but this does not mean that he had real knowledge of the future. And certainly the fact that he was able to recount events from the past does not prove that he was a prophet. [He could have read about those events in the Bible] and, if he was illiterate, he could still have had the Bible read to him.(72)

Ibn al-Rawandi: The Jews and Christians had access to a very detailed description of the Prophet and the circumstances of his future arrival. No astrologer could predict the future in such a precise manner. Astrologers rarely succeed in predicting the future and then only by chance.(73)

Al-Warraq: The Jews knew that if they had accepted Muhammad's challenge [and declared that they were so confident of prospering in the hereafter that they wished for death], then Muhammad would have said that they did not really wish for death, but only said so.(74)

Ibn al-Rawandi: Muhammad's challenge did not include the condition that they have to mean what they say, they only had to say it.(75) Also, the Jews could have replied that they did wish it in their hearts.(76)

Al-Warraq: If the Jews were to say that they did wish it in their hearts, Muhammad would have answered that Gabriel had revealed to him that they were lying, and that they did not wish it in their hearts.(77)

III. THE HERETIC IBN AL-RAWANDI AND HIS ZUMURRUD

1. THE SOURCES SUPPOSEDLY FAVORABLE TO IBN AL-RAWANDI

Despite the fragmentary state of our sources, we were able to derive from them a consistent account of one paragraph of the Zumurrud. This information confirms van Ess' suggestion as to the form of the Zumurrud, namely that it was a dialogue between al-Warraq, whose role was to present arguments against the existence of prophecy, and Ibn al-Rawandi, who presented the arguments in its favor.

Another part of van Ess' suggestion must now be examined. He believes that Ibn al-Rawandi was not a heretic, and that "the extreme and shocking statements which have been filtered out of his books cannot be entirely representative of his own ideas."(78) He suggests that in the Zumurrud "Ibn al-Rawandi defended Islam, or perhaps his special interpretation of Islam."(79) He believes that Ibn al-Rawandi "wanted to waken the spirit of doubt and hesitation in the hearts of the people [of the Baghdad school], without being a skeptic himself."(80) And he further suggests that even al-Warraq, who presents the arguments against the existence of prophecy in the Zumurrud, was not really a heretic, and that "the dialogue between Ibn ar-Rewandi and Abu Isa was . . . apparently a dialogue between Muslims both of whom were still right."(81)

This contention is based on the assumption that Muayyad and Maturidi give inconsistent accounts, which I have argued is not the case. It is further based on preferring the evidence of Maturidi to the evidence of the many other sources that unambiguously condemn Ibn al-Rawandi. Van Ess believes that Muayyad and Abd al-Jabbar, al-Khayyat. and Ibn al-Jawzi are all biased, and that they draw a distorted picture of Ibn al-Rawandi. Although van Ess' mainstay is Maturidi, he supports his claim by referring to other sources that are supposedly favorable to Ibn al-Rawandi.

A closer look at these sources shows that the information in most of them is not inconsistent with his being a heretic. Many of them discuss Ibn al-Rawandi's scholarship, and are thus irrelevant to the subject of his heresy. Tawhidi, for example, describes Ibn al-Rawandi as a competent theologian and praises his mastery of Arabic style.(82) If this were our only evidence of what Tawhidi thought of Ibn al-Rawandi, we might have inferred that Tawhidi regarded Ibn al-Rawandi as a respected Muslim. But fortunately we have passages in which Tawhidi' s discussion leads him to the question of prophecy, and there he unequivocally condemns Ibn al-Rawandi's heretical views.(83) Tawhidi saw no inconsistency between, on the one hand, giving Ibn al-Rawandi the praise due to his erudition and, on the other, censuring his heretical ideas.(84)

The same is true of al-Ashari, who several times mentions Ibn al-Rawandi' s opinions in neutral or even favorable tones.(85) This attitude of al-Ashari may account for Abd al-Jabbar's sour remark that the Asharites are impressed with Ibn al-Rawandi.(86) But on another occasion, al- Ashari refutes Ibn al-Rawandi's theology, adding to his name the unmistakable epithet "the cursed one" (al-la in).(87) Al-Ashari also wrote refutations of Ibn al-Rawandi's heretical books.(88)

According to Abd al-Amir al-Asam, who painstakingly collected all the medieval references to Ibn al-Rawandi, only two authors attempted to defend Ibn al-Rawandi: al-Sharif al-Murtada (d. 1044) and Ibn Khallikan (d. 1283).(89) But a closer look at what these two authors have to say shows that they too are well aware of Ibn al-Rawandi's heresy. Al-Murtada had an account to settle with the Mutazila, and his book al-Shafi fi'l-imama was written as a retort to Abd al-Jabbar. It is therefore understandable that in this book he blames the Mutazila for Ibn al-Rawandi's heretical writings and plays down Ibn al-Rawandi' s faults. He mentions the claim that Ibn al-Rawandi's main object in his heretical books was to spite the Mutazila. But even al-Murtada sums up this discussion by saying, "There can be no doubt that he sinned in writing these books, regardless of whether he believed in it (i.e., their contents) or not."(90) And when, as an ilzam, or counter- attack, against the Mutazila, al-Murtada compares Ibn al-Rawandi to al-Jahiz, his clear intention is to incriminate the latter, not to exonerate the former.(91)

The fragmentary remnants that we have of the work of certain authors sometimes include only a single short reference to Ibn al-Rawandi. If such a reference happens to be neutral or even positive, we would be mistaken to deduce from it that the author condoned Ibn al-Rawandi' s views on prophecy. This principle is particularly to be observed when dealing with relatively late sources, which draw selectively on earlier ones. Such is the case of Ibn Khallikan, whose entry on Ibn al-Rawandi describes him as "the well-known scholar." Ibn Khallikan goes on to say that Ibn al-Rawandi wrote a treatise on kalam, that he was among the distinguished men (fudala) of his time, and that he wrote some one-hundred-and-fourteen books. Of these books, Ibn Khallikan names only five, the Zumurrud and four other heretical books. Ibn Khallikan is aware of the fact that Ibn al-Rawandi "had peculiar doctrines, which the theologians transmitted in their books The theologians from whom Ibn Khallikan learned of the five heretical books were surely quite explicit in evaluating these peculiar ideas. Ibn Khallikan's reluctance to repeat the theologian's words bespeaks his own prejudices and preferences rather than these of his sources. And it would certainly be a mistake to rely on Ibn Khallikan's information in determining the tenor of Ibn al-Rawandi's ideas.(92)

H. S. Nyberg argued that the philosophers and physicians "took Ibn al-Rawandi seriously and sometimes defended him," and in support of his claim he cited the titles of books written by al-Kindi and Ibn al-Haytham (d. 1039).(93) But Nyberg's evidence does not prove his claim. Al-Kindi wrote "a discussion with Ibn al-Rawandi on God's unity." (94) We have no way of knowing what the contents of this book were, but it is likely to have reflected some disagreements with Ibn al- Rawandi. As to Ibn al-Haytham, he was not at all supportive of Ibn al-Rawandi; a treatise in which he criticized Abu Ali's refutation of Ibn al-Rawandi also included his suggestions for a better refutation.(95) Although it is probable that Ibn al-Rawandi's intellectualist ideas were indeed attractive to some philosophers, the philosophers were careful to conceal such dubious influences and they avoided mentioning his name, as the case of Abu Nasr al-Farabi shows.(96)

The only early source in which Ibn al-Rawandi appears as a good Muslim in the context of the discussion of prophecy is Maturidi's K. al-Tawhid, and we have seen how this account is to be understood. Maturidi's book is also the source which most clearly shows the role of Abu Isa al-Warraq in the Zumurrud as well as his Manichaean beliefs. For, contrary to van Ess' view, there can be no doubt as to al-Warraq's Manichaeism.(97) Van Ess is probably right in assuming that the peculiarity of Maturidi's evidence (namely, the fact that he does not attack Ibn al-Rawandi) was related to his eastern provenance. I doubt, however, that the correct interpretation of the geographically based difference is that in the east people were less dogmatic than in Baghdad, and therefore more understanding toward Ibn al-Rawandi.(98) Maturidi lived in Samarqand, where in the tenth century Manichaeism was still considered a real threat to Islam.(99) The interest of the Zumurrud for Maturidi lay in the fact that in this book, the Manichaean al-Warraq appeared as a person who denies the existence of prophecy. For Maturidi, the urgent task was to attack the Manichaean al-Warraq, not to unmask the real face of Ibn al-Rawandi.

2. IBN AL-RAWANDI AS A FREETHINKER

If Ibn al-Rawandi denied the existence of prophecy, why did he choose to present the case of its proponents? In order to understand the motives of the author of the Zumurrud, we would need to have a fuller view of his person and his books. As new material comes to light or old material is analyzed more closely, we may hope to confirm our views or to correct them. Until then, we can offer only a tentative analysis, on the basis of very scanty information.

One possible answer would follow the approach developed by Leo Strauss in his Persecution and the Art of Writing. According to this approach, Ibn al-Rawandi was afraid to publish his ideas in his own name, and therefore attributed them to al-Warraq.(100) This explanation, however, does not account for the fact that Ibn al-Rawandi's other books, though no less offensive to Muslims, were published under his own name. It also does not account for the apparently close and unbroken friendship between Ibn al-Rawandi and al-Warraq.(101) It is unlikely that al-Warraq volunteered to present Ibn al-Rawandi's views, or that he would have appreciated finding himself in the position of having to take the blame for these views, yet there is no indication that there was any animosity between the two after the publication of the Zumurrud.(102)

I think it is more likely that the structure of the book reflects the real relationship between al-Warraq and Ibn al-Rawandi. Al-Warraq is usually described in the sources as the one who started Ibn al- Rawandi on the path of heresy, introducing the young Mutazilite to criticism of the Quran and the prophets.(103) While debating with him, Ibn al-Rawandi used the method of ilzam (i.e., counter-attacking by turning the opponent's own arguments against him). He used al-Warraq' s arguments against revealed religions to attack al-Warraq's own religion, Manichaeism. When Ibn al-Rawandi was finally won over to al-Warraq' s heretical views, he went even further than his teacher. Whereas al-Warraq was a Manichaean, Ibn al-Rawandi did not consider the Manichaean tradition to be any better than the Muslim one. The Zumurrud may well be a record of the dialogues between the two, as edited by Ibn al- Rawandi. The dialogues reflect their collegiality as well as their disagreements. They record the process of discussion by which al-Warraq persuaded Ibn al-Rawandi to reject prophetic religions, as well as Ibn al-Rawandi's refusal to see Manichaeism as an exception.(104)

One should also note the connection between Ibn al-Rawandi's character, his style and his ideas. I know of no other book where the author, as one of the participants in a dialogue, ascribes to himself an idea in which he no longer believes. Yet if anyone were to do it, Ibn al-Rawandi would be the man. I also know of no other author who systematically wrote refutations (naqd, not just corrections or retractions) of his own works. Ibn al-Rawandi did. He is known to have systematically written first a book and then its refutation.(105) Muslim authors considered this baffling habit a proof of Ibn al-Rawandi's frivolous, mercenary tendencies.(106) Van Ess regards it as a peche mignon, "a certain dose of intellectual coquetry."(107) In contradistinction to both these views, I think that this practice of writing self-refutations reflects, in Ibn al-Rawandi's provocative way, his deep convictions.

As the quotations from Ibn al-Rawandi's books show, he was a true freethinker, in the sense that he rejected the authority of any scriptural or revealed religion.(108) This is borne out by citations from his other writings, besides the Zumurrud: the K. al-Damigh against the Quran, and "The Futility of (Divine) Wisdom" (Abath al-hikma). Ibn al-Rawandi's complete break with Islam is reflected not only in the contents of these books, but also in their tone. Some of the questions presented by Ibn al-Rawandi in the Damigh could perhaps be asked by a Muslim commentator with an inquisitive mind who was trying to reconcile two apparently contradictory verses. But no Muslim would introduce such a question with a declaration that God does not know how to add two to four and make six.(109) A tormented Muslim might inquire into theodicy, but no Muslim would conclude the inquiry with the statement that God behaves like a wrathful, murderous enemy.(110) And a pious Muslim might question the inherent value of the rituals in the pilgrimage. But for a Muslim, this question highlights the notion of submission to the superior knowledge of God and His Prophet, whereas for Ibn al-Rawandi it implies the primitive nature of the Muslim rites.(111) A person who presents Ibn al-Rawandi's questions in the tone that he adopts, could not remain a Muslim in any meaningful way.(112) Indeed, it was the tone of Ibn al-Rawandi's arguments as much as their content that infuriated Muslims. This is reflected in Ibn al-Jawzi's comment that Ibn al-Rawandi was worse than Iblis, because although Iblis disobeyed God, at least he addressed Him with respect.(113)

The disrespectful tone was an integral part of Ibn al-Rawandi's attack on established religion. In the same way, the form of his writings was closely connected to their skeptical content. Writing both a book and its refutation is a way of speaking with two voices at once. It may have been used by Ibn al-Rawandi to convey his mistrust of written authority and his complete commitment to a continuous effort to sharpen one's mind. It is possible that in the Zumurrud he tried to incorporate the same dual voice into the book itself, through dialogue with a colleague who shared some of his ideas. The fact that in the dialogue he took the role of the proponent of prophecy may have been a device to further enhance the critical approach. If this was his intention, there was probably more than a shade of mockery in it. But there was also the genuine drive to doubt--to doubt the authority of prophets, and then double-check the conclusion by doubting the doubt.(114)

Despite the peculiarity of the book's structure, Muslira authors were generally able to infer correctly the real beliefs of its author. We may conjecture that some section of the Zumurrud--the introduction or the conclusion--recorded Ibn al-Rawandi's authorship of the book. It is probable that, like kalam debates, this one too ended with one of the participants admitting defeat (inqita). We can therefore further conjecture that the conclusion of the book recorded al-Warraq's victory and Ibn al-Rawandi's ultimate rejection of the existence of prophecy.

Some of the Muslim sources which mention the Zumurrud reveal their awareness of the fact that the book was in some way the result of a collaboration between al-Warraq and Ibn al-Rawandi. Thus Ibn al- Jawzi remarks that Ibn al-Rawandi and al-Warraq charged each other with the authorship of this book, although they agreed on its criticism of the Quran,(115) and Abd al-Jabbar names both of them as authors of a K. al-Zurnurrud. In another passage of the Tathbit, Abd al-Jabbar mentions the Zumurrud's argument concerning the presence of angels in Muhammad's battles,(116) and then adds, "This question was raised by Ibn al-Rawandi after he had conspired with Abu Isa al-Warraq and Ibn Lawi the Jew."(117) This remark then develops to the statement that Ibn al-Rawandi "wrote most of his heretical books for the Jew Aba Isa [sic] of Ahwaz, in whose home he perished."(118)

This confused report(119) testifies to the notion retained in the Muslim sources of a certain collaboration between Abu Isa al-Warraq and Ibn al-Rawandi in the production of heretical books. This collaboration did not mean that they wrote books together, but that the books Ibn al-Rawandi wrote were the fruit of his discussions with al-Warraq.(120)

In the final analysis, however, people identified the author of the Zumurrud as Ibn al-Rawandi. The fact that in the Zumurrud Ibn al-Rawandi allotted to himself the role of the defender of prophecy did not mislead those who knew the work at first hand. For them, the Zumurrud was a heretical book written by a heretic.

IV. THE ZUMURRUD AND ITS CONTENTS

1. THE TITLE

The Hanbalite author Abu al-Wafa ibn Aqil (d. 1119) reports that when Ibn al-Rawandi chose to name his notorious book against prophecy The Book of the Emerald (Kitab al-Zumurrud), it was not, as Ibn al-Rawandi' s contemporary Abu Ali al-Jubbai had thought, in order to compare his science to this precious stone. His intention was rather to hint at a comparison between the power of the emerald to blind snakes, and the supposed power of the sophistries which the book contained to silence all arguments.(121) But he wished only to hint at this: this explanation of the book's name did not appear in the Zumurrud itself,(122) for Ibn Aqil claims to have seen it "in some other book of Ibn al-Rawandi."

After Ibn Aqil, this explanation was widely and generally accepted, and scholars, both Muslims and others, cite it with no apparent second thoughts.(123) In view of the late appearance of this explanation and of its flimsy basis, such general credulity is remarkable. Ibn Aqil (as cited by Ibn al-Jawzi) remains our only source for this explanation, and he is a source of questionable value. He probably did not read the Zumurrud itself, but only Abu Ali's refutation of it.(124) In fact, notwithstanding Ibn Aqil's claim, we have no reason to believe that he had direct access to any of Ibn al-Rawandi's books. Had there been "some other book" of Ibn al-Rawandi's that was unknown to Abu Ali and which contained this explanation, we would have expected other people with heresiographical interests to be aware of it. We would also expect that, in the writings of these other people, the mention of the emerald's special power would trigger a reference to Ibn al- Rawandi's book. This, however, does not seem to be the case. Nasir- i Khusraw, the Ismaili author who wrote half a century after Ibn Aqil, was generally sensitive to anti-prophetic sayings.(125) In his Jami al-Hikmatayn he refers to the supposed special quality (khassa) of the emerald, and he also mentions some of Ibn al-Rawandi's heretical views.(126) But he does not associate this special quality of the emerald with any aspect of Ibn al-Rawandi's work. It therefore seems legitimate to maintain a skeptical reserve toward Ibn Aqil's explanation.

It is apparently true that Ibn al-Rawandi had a keen interest in the special qualities of minerals. His gemological interest may partly account for the fact that he named several of his books after precious stones (although the practice is quite common among other people, too). It is also true that the existence of remarkable natural phenomena, such as the special powers of some minerals, was used by Ibn al-Rawandi in his refutation of revealed religions. Ibn al-Rawandi argued that those who claim to be prophets perform their supposed miracles by exploiting their knowledge of such natural phenomena. But we have no proof that Ibn al-Rawandi actually explained his choice of the title K. al-Zumurrud as connected with the special power of the emerald and to his own attempt to confound his opponents. There is a remote possibility that in his K. al-Khawass, along with the discussion of the emerald's power, Ibn al-Rawandi included such an explanation. But it is more likely that he did not, and that Ibn Aqil himself read this intention into Ibn al-Rawandi's title, in order to emphasize Ibn al-Rawandi's malevolence.(127)

2. THE STRUCTURE OF THE BOOK

Kraus believed that Muayyad faithfully followed his source, and that the order of the discussion in Muayyad's source followed the original order of the Zumurrud.(128) Kraus also assumed that the discussion in the Zumurrud moved in a systematic way from one subject to the next. This reconstruction is highly questionable. As Muayyad himself tells us, the dai discarded from the Zumurrud most of the arguments for prophecy that it originally contained. Occasionally, the dai's wording marks the places where he omitted a passage from the Zumurrud.(129) As to the material he does quote, it is not at all certain that the Majalis Muayyadiyya consistently followed the order of the discussion in the Zumurrud. Moreover, to the extent that the order of discussion in the Majalis Muayyadiyya (as well as in other sources) does preserve the original structure of the Zumurrud, it reflects a rather loose and unsystematic structure.

The book was apparently divided into chapters (abwab), for we find two authors referring to a bab of this book. Muayyad mentions a chapter in which Ibn al-Rawandi relied on the Barihima,(130) and al-Khayyat mentions a chapter "against the Muhammadans in particular."(131) Muslim authors would no doubt react particularly to this last chapter, and it is possible indeed that Muayyad and Ibn al-Jawzi drew mostly from it. This chapter apparently took the general arguments listed in the earlier parts of the book and showed how they apply to the particular case of Islam.

Nevertheless, there is no reason to assume that the chapter against Islam repeated the arguments in exactly the order in which they appeared in earlier chapters of the book. The discussion of some of the topics discussed in it (theodicy and criticism of the Quran) apparently repeated what Ibn al-Rawandi had said in other books. Ibn al-Rawandi reiterated his favorite arguments, and I suspect that even when an argument did not fit the plan of the book, he nevertheless used it. Rather than following a rigorous pattern, the Zumurrud was, in all probability, a potpourri of arguments. Our reconstruction can at best list the subjects dealt with in the book, but not the order in which they were presented.

3. THE ARGUMENTS AGAINST PROPHECY

The subjects discussed in the Zumurrud can be divided into several main groups:

A. Arguments relating to the primacy of the intellect

God has bestowed upon human beings the gift of intellect, by which they can judge right and wrong. If what the prophets announce corresponds to what the intellect decrees, then prophets are superfluous. If it contradicts what the intellect decrees, then one should not listen to them.(132) The discussion with the Barahima, the issue of the abrogation of the law, and the question of the possibility of substituting one law for another are also part of this argument.(133) The argument is then applied to Islam in particular.(134)

Connected with the claim of the sufficiency of human intellect is the discussion of various expressions of this intellect.(135) Human children are taught to speak by their parents, from one generation to another, and this has always been the case.(136) Ibn al-Rawandi is here probably addressing the question of whether human speech is natural or conventional. He seems to favor the solution of ilham (i.e., natural, innate knowledge), although the term itself does not appear. From the dais answer we can see that Ibn al-Rawandi gave various examples of innate knowledge (the ability of birds to Communicate with each other, the ability of ducks to swim, the ability of infants to suck milk), and that these were mentioned by him as being analogous to speech and understanding.

The sciences are also mentioned by Ibn al-Rawandi as proof for the sufficiency of the intellect. According to him, people developed the science of astronomy by watching the skies. They did not need a prophet to teach them how to watch. Nor did they need prophets in order to teach them how to build lutes. It is absurd to assume that without prophetic revelation people would not have learned that the intestines of a sheep, when dried and stretched upon a piece of wood, can produce pleasant tones. All these skills are acquired by the assiduous application of the inborn human intellect, discernment and power of observation.

Kraus thought that this part of the book opened with a paragraph praising the intellect in rhymed prose, one sentence of which is to be found in the dai's refutation.(137) Kraus noted that neither the dai nor Ibn al-Rawandi were given to writing rhymes. He argued that the only place one could expect either one of them to use such a sentence would be in an opening chapter of a conventional nature, where the praise of the intellect is sung before the real discussion begins. He therefore suggested that the Zumurrud had a poetic opening in which Ibn al-Rawandi glorified the intellect, and that the dai opened his response with a poetic paraphrase of Ibn al-Rawandi's verse.(138)

It is indeed possible that this, sentence is taken from an introduction written in flowery style. I do not think, however, that it could come from Ibn al-Rawandi's pen. Had this been the case, the dai would probably have said so explicitly, as he always does when he wants to attack something said by Ibn al-Rawandi. It is more likely that this sentence was written by the dai.(139) Furthermore, in the debate with the Kitab al-Zumurrud the proper estimate of the role of the intellect was not a side-issue, but stood at the core of the discussion. It is therefore likely that, rather than being a conventional opening, the reference to "the person who claims to cover the horizons of science with the wings of the intellect" is the dai's direct assault on Ibn al-Rawandi' s intellectualist pretensions.

B. Arguments relating to the Quran

The Zumurrud attacks from various angles the doctrine of the unrivaled beauty of the Quran (ijaz).(140) It suggests several natural explanations for the fact that the Arabs did not produce anything similar to the Quran. It argues that this may be due to the fact that the people of Quraysh were more eloquent than members of other tribes, and that Muhammad was a particularly gifted individual; or to the fact that the Arabs were too busy fighting with Muhammad, and had no time to invest in poetic competitions; or to the fact that the Arabs were uneducated people. The Zumurrud also argues that the Quran is not really that beautiful; in the sayings of Aktham b. Sayf there are more beautiful sayings than "We gave you in abundance" (Q. 108:1). The Quran is written in a faulty Arabic (lahn) and its language can be corrected by human beings.(141) The Quran is full of contradictory, absurd sayings, and it cannot possibly be the speech of the Wise One.(142) And, at any rate, while the language of the Quran may seem to the Arabs to be proof that Muhammad was a prophet, it cannot be expected to impress non-Arabs.(143)

The Zumurrud also discussed specific Quranic verses, attempting to point out their logical flaws and to call into question the Muslim tradition about the circumstances of their revelation. As the example of the mubahala cluster shows, such discussions also touched on other issues, such as the validity of Muslim tradition and the Prophet's alleged miracles.

C. Arguments relating to Muslim traditions

According to the Zumurrud, traditions concerning miracles are inevitably problematic. At the time of the performance of a supposed miracle only a small number of people could be close enough to the Prophet to observe his deeds. Reports given by such a small number of people cannot be trusted, for such a small group can easily have conspired to lie.(144) The Muslim tradition thus falls into the category of flimsy traditions, those based on a single authority (khabar al-ahad) rather than on multiple authorities (khabar mutawatir).(145) These religious traditions are lies endorsed by conspiracies.

The Zumurrud points out that Muhammad's own presuppositions (wad) and system (qanun)(146) show that religious traditions are not trustworthy. The Jews and Christians say that Jesus really died, but the Quran contradicts them.(147) If statements made by so many people cannot be trusted, all the more so the testimony of a handful of people like Muhammad's followers.

Ibn al-Rawandi also points out specific Muslim traditions, and tries to show that they are laughable. The tradition that the angels rallied round to help Muhammad is not logical, because it implies that the angels of Badr were weaklings, able to kill only seventy of the Prophet' s enemies. And if the angels were willing to help Muhammad at Badr, where were they at Uhud, when their help was so badly needed?(148)

D. Arguments relating to miracles

An important part of the Zumurrud is devoted to arguing that the miracles of the prophets are products of legerdemain. Like magicians, prophets exploit unusual natural phenomena, similar to the magnet but not as well known.(149) A number of Muhammad's miracles are specifically mentioned: the ablution basin, Umm Mabad's sheep, Suraqa, the wolf who talked, the intoxicated sheep who talked, and the isra.(150) According to the Zumurrud, the distance between Mecca and Jerusalem is not very great, and it is conceivable that a person could go from one of these cities to the other and back in one night, so the Prophet's presentation of his ability to describe Jerusalem as a miracle is a fraudulent trick (makhraq).(151) Even the Prophet's ability to predict the future (as in the case of the slaying of Ammar b. Yasir) is not regarded as a miracle, since it is claimed that any astrologer can do that.(152)

E. Arguments relating to the Muslim rituals

The Zumurrud criticizes prayer, preoccupation with ritual purity, and the ceremonies of the hajj: throwing stones, circumambulating a house that cannot respond to prayers, running between stones that can neither help nor harm. It goes on to ask why Safa and Marwa are venerated, and what difference there is between them and any other hill in the vicinity of Mecca, for example the hill of Abu Qubays, and why the Kaba is any better than any other house.(153)

This sketch of the arguments contained in the Zumurrud confirms several of the conclusions presented in previous sections of the present study. It shows the close similarity between the Majalis Muayyadiyya, Maturidi and the Tathbit, and thus corroborates the claim that they derive from the same source, the Zumurrud. It also strengthens the impression that all three sources summarize rather than quote the Zumurrud. And it shows that none of the three sources relied on either of the other two for its information, since in each one of them we find elements that are lacking in the other two. A comparison of the sources allows us to see that the same arguments are attributed at times to Ibn al- Rawandi, at times to al-Warraq and at times to both of them. A correct understanding of the Zumurrud must allow for the active participation of both Ibn al-Rawandi and al-Warraq in the dialogue, and take into account the heretical convictions of both of them.

CONCLUSION

Even if Ibn Aqil cannot be trusted concerning the intended meaning of the title K. al-Zumurrud, we must admit that his explanation grasps the effect that this book had on its audience. Its "sophistries" cannot be said to have silenced all argument--on the contrary, they drew furious tirades from many Muslims who read or heard them. But the book did seem to puzzle and confuse its readers. In modern scholarship, in particular, the discrepancy between Ibn al-Rawandi's position and the role that he took in this book combined with the fragmentary state of our data to produce a generally inaccurate picture of the Zumurrud.

Among the Muslim authors who mention Ibn al-Rawandi without attacking him, Maturidi is the only one who presents him in positive light in the context of