MUʿTAZILAH . A religious movement in early Islam, the Muʿtazilah turned into a theological school that become dominant in the third and fourth centuries ah (ninth to tenth century ce) and persisted in certain areas until the Mongol invasion at the beginning of the thirteenth century ce. The history of the movement is comprised of three different phases: (1) an incubation period that lasted roughly through the eighth century; (2) a short period of less than half a century (c. 815–850) when the Muʿtazilī school, after having defined its identity, developed an astonishing variety of individual, sometimes contradictory ideas and permeated the intellectual life at the Abbasid court; and finally (3) several centuries of scholastic systematization channeled into two branches or schools that were named after the towns of Basra and Baghdad respectively.
Each of these phases presents its own problems for the researcher: the first is badly documented and can only be reconstructed on the basis of later reports, which are frequently distorted and tendentious; the second is better attested but needs detailed monographical treatment; and the third has only recently begun to attract scholarly attention. On the whole, knowledge of this movement is still rudimentary. Detailed research is hampered by the lack of original texts. This is due to the fact that, after the middle of the ninth century, the Muʿtazilī movement was gradually driven into the position of a heresy; in the areas where it was considered "unorthodox" its books were no longer copied. Therefore we have to rely, at least up to the third phase, mainly on heresiographical reports. For the later centuries we possess a few texts, some of which reach the size of a summa theologica, but they belong to a rather restricted period; outside of this limit, many of the thinkers still remain mere names to us.
The Muʿtazilī movement is usually traced back to the end of the Umayyad period, the years between 740 and 750. But during the first century of its existence the movement was far from the most important factor in the development of Islamic theology.
The Muʿtazilah began in Iraq, but there the Shīʿah (in Kufa) and the Ibāḍīyah (in Basra and Kufa) initially had the better thinkers, while the school of Abū Hanīfah, which combined juridical competence with an "ecumenical" outlook in theology, enjoyed greater missionary success. We are not even sure whether we can assume—as all of our sources do—that a real continuity existed between the first and second phases: there seems to be almost no doubt that the great thinkers of the second phase did not have any precise knowledge about their spiritual ancestors. When they moved from the old intellectual centers, Basra and Kufa, to the Abbasid court in Baghdad they felt the urge to preserve the memory of their past, but they evidently could not rely on any established historical tradition of the "school." The gap was widened by the fact that they disagreed with certain opinions held in the preceding generation and therefore tried to keep their immediate predecessors out of the picture. Under these circumstances we must reckon with the possibility that they constructed a past which never belonged to them or was only partially true. We should also not forget that the historical reports that were transmitted at that time were not collected and written down until some time later, from the middle of the second half of the ninth century onward.
The lack of historical recollection is amply demonstrated by the discovery that no Muʿtazilī author had any precise information about the original meaning of the name Muʿtazilah. Our sources offer a number of explanations, but all of them are secondary guesses and beside the point; some are blatantly tendentious. Modern scholarship has contributed a few more suggestions, but the question still remains open. All that we can prove is that the movement bore this name when it became involved in an insurrection against the caliph al-Manṣūr in 762. But the name does not seem to have been invented at that time, for it does not fit the situation. It means "those who dissociate themselves" or "those who keep themselves apart" and thus calls for political neutrality rather than revolutionary activism. The name already existed in the first century ah (seventh century ce), with this connotation, as a term designating some renowned companions of the Prophet who abstained from any participation in the first civil war (the Battle of the Camel in 656 and the Battle of Ṣiffīn in 657). It was then probably applied to the first Muʿtazilī thinkers since they, too, did not side with any political party of their time.
This attitude was distinctive insofar as it was adopted in a period when almost everybody had to make personal alignments clear, namely during the last years of the Umayyad caliphate, which saw the breakdown of the political order in Iraq and elsewhere. The founder of the movement, a cloth merchant from Basra by the name of Wāṣil ibn ʿAṭāʾ, intended to create a missionary organization working inside Islam; he sent his disciples, as "propagandists," to the most remote regions of the Islamic empire—the Arabian Peninsula, Armenia, Iran, India (the Punjab), and the Maghreb—so that they could interpret the Muslim creed and win people over to his own cause. Unfortunately, we do not know what this "cause" really implied. We cannot exclude the possibility that it was originally political, for Wāṣil copied a model that clearly had a political character, that is, the network of agents built up by the Ibāḍīyah, with whom he lived closely in Basra. Most of our information, however, contradicts this hypothesis: Wāṣil wanted reform, not revolution. Islam, was, after all, still the religion of a minority; outside the great centers, the knowledge of what Islam really meant was rather limited, and its definition differed from area to area. It any case, Wāṣil did not live to see the fruit of his efforts; he died in 749, one year before the triumph of the Abbasids.
Wāṣil's "propagandists" were mostly merchants like himself, and when they traveled, they combined business with missionary zeal. This pattern explains how the movement financed itself but does not say much about its spiritual impetus. The inner motivation of Wāṣil's circle seems to have derived from a certain feeling of inferiority: all of the participants were non-Arabs, that is, they did not enjoy the natural prestige of the aristocracy but came, as "clients," from Iranian or Aramaean families who had been converted to Islam one or two generations before. They possessed considerable wealth and were, as a matter of fact, recognized in their society, but they had to rely on Islam as the basis of their identity. They knew something about Islamic law, for Wāṣil advised them to win the favor of their audience by delivering fatwā s ("legal opinions") to demonstrate their juridical expertise. But they also deliberately distinguished themselves from normal, "worldly" people: they clipped their mustaches and wore turbans (which, at that time, were characteristic of certain nomadic tribes but not of the urban population); they also wore special sandals and wide sleeves. This was the attire of ascetics; it gave them a certain "alternative" touch.
The organization was taken over by Wāṣil's colleague ʿAmr ibn ʿUbayd (d. 761), a prominent disciple of Ḥasan al-Baṣrī, the great figure of religious life in Basra during the preceding generation. ʿAmr had to cope with the new situation created by the Abbasid seizure of power. He seems to have given up all relations with the cells outside Iraq and in Basra, where he controlled a considerable number of adherents (who possibly formed youth groups), he kept quiet. This position became increasingly precarious at the end of his life as discontent with the Abbasid government mounted in Iraq. After his death the activists among the Muʿtazilah followed the call of the Shīʿī pretender Muḥammad ibn ʿAbd Allāh al-Nafs al-Zakīyah (or rather, that of his brother Ibrāhīm) and took part in the revolt of 762. When the attempt failed, the Muʿtazilah were persecuted and went into hiding; those who had compromised themselves mostly fled to Morocco.
This event seems to be a decisive turning point. We hear that afterward the Muʿtazilah still possessed a mosque of their own in Basra, but we do not know of any leading personality for at least thirty years. Above all, there is no hint of any specific theological activity. Then, toward the end of the eighth century, two figures emerged: al-Aṣamm in Basra and Ḍirār ibn ʿAmr in Kufa. But neither of them was a typical Muʿtazilī; as a matter of fact, the later school kept a certain distance from them. Al-Aṣamm was obviously an Ibāḍī, whereas Ḍirār, a judge by profession and one of the most original thinkers of this period, differed from the communis opinio of the following generation in his ideas concerning free will and therefore fell victim to a damnatio memoriae. The original concept of the Muʿtazilah as a popular missionary movement seems to have survived best in Baghdad where, during the same period, Bishr ibn al-Muʿtamir, a slave merchant by profession, exhorted the masses by expressing his theological ideas in simple poetry. That was appropriate for the social climate in the newly founded capital; the town had attracted many people who came to make their fortune and ended up by being uprooted.
The period of success
The Muʿtazilah were propelled from provinciality to prime importance by the theological interest which emerged at the Abbasid court. The change came in two shifts, first through the influence of the Barmakids, the viziers of Hārūn al-Rashīd (r. 786–802), and then, after a short setback caused by their downfall, thanks to the initiative of the caliph al-Maʾmūn (813–833). Both the Barmakids and al-Maʾmūn were not so much interested in theology itself as in listening to disputations: they liked to have representatives of different religions and confessions argue against each other. This predilection may have been stimulated by a non-Iraqi environment: the Barmakids originally came from Balkh, and al-Maʾmūn first resided in Merv; in Transoxiana, where both towns were situated, Islam co-existed with Zoroastrianism, Buddhism, Christianity, and Judaism.
However, the main stimulus came from the intellectual atmosphere of the capital itself. Islam was no longer the religion of a minority, as in the time of Wāṣil, but a creed which had rapidly expanded at the expense of other religions. The conversions to Islam had been prompted mostly by social considerations, but theology had to furnish an a posteriori justification: thus the outlook of the new theology was strongly apologetic and its style predominantly dialectical. The Muslims were not entirely unprepared; they had experienced enough internal strife between different "sects" in order to know what methods to use in disputes. Therefore, the Muʿtazilah were not the only ones to sharpen these weapons for the fight with their pagan adversaries. But besides being skillful dialecticians (mutakallimūn ) they offered a concept of Islam which, by its rationality, transcended the divisions among the old theologico-political factions (Shīʿah, Murjiʿah, and others) and therefore had broad appeal, at least among the intellectuals. The Muʿtazilah thus became the first overall, "orthodox" school of theology.
Their path to success can still be traced. Ḍirār ibn ʿAmr took part in the sessions arranged by the Barmakids, but there he was only one among many. In Merv, one generation later, the situation was different; the Muʿtazilī Thumāmah ibn Ashras acted as a kind of counselor to al-Maʾmūn, and Bishr ibn al-Muʿtamir was among those who put their signature as witnesses to the document in which the caliph nominated the Shīʿī imam ʿAlī al-Riḍā as his successor. The real breakthrough came when, in 820, al-Maʾmūn moved the court back to Baghdad. Two figures dominated the scene there: the Basran theologian Abū al-Hudhayl al-ʿAllāf, who was already about seventy (he died a centenarian in about 840), and his nephew al-Naẓẓām. The latter showed all the features of a courtier: he mocked at asceticism and excelled in light and imaginative poems which celebrated wine and the beauty of youths. He therefore acquired the reputation of being a drunkard and a homosexual. He was not necessarily either, since poetry is not reality, but these characterizations demonstrate that, with their success, the Muʿtazilah also came under scrutiny. Al-Naẓẓām's open identification with the ideals of high society did not tally with Bishr ibn al-Muʿtamir's earlier attempt to convert the masses. The split deepened when those who still understood the Muʿtazilah as a popular movement and kept to the old ascetic tradition started adopting Ṣūfī tendencies: they dressed in wool and asserted that the Muslim community should abstain from electing a caliph (a merely symbolic viewpoint, for the community did not have any influence in this respect anyway). In their view, however, court life was a scandal and the entire Muslim world corrupt, full of injustice and violence.
For the moment, the court party had the upper hand. But enjoying the favor of the caliph also meant supporting his policies. When al-Maʾmūn, in a decree sent throughout the empire in 833, asked his governors to enforce the doctrine of the createdness of the Qurʾān as a kind of state dogma, the Muʿtazilah were immediately identified with this measure. This evaluation was only partly justified. The caliph had certainly made the decision by himself, as a demonstration of his spiritual leadership of the community, and his main adviser had been a theologian by the name of Bishr al-Marīsī, who, through his belief in determinism, stood apart from the Muʿtazilah. But the Muʿtazilah subsequently had to lend their intellectual support to the measure. When the policy of the caliphs led to a persecution (the so-called miḥnah, or "inquisition") the chief judge was a Muʿtazilī: Ibn Abī Duwād. The miḥnah lasted for fifteen years, and the government succeeded in purging the ranks of the state officials of any opposition. Resistance remained strong, however, among the population of Baghdad who, with a clear anti-intellectual bias, rejected rational theology in favor of the prophetic tradition (ḥadīth ). Therefore the caliph al-Mutawakkil, the third successor of al-Maʾmūn, decided to steer another course. In 848 he ordered some traditionists to preach about (spurious) sayings in which the Prophet allegedly condemned the Muʿtazilah and similar groupings; a few years later, any occupation with dialectical theology was prohibited. The Muʿtazilah were removed from the court.
But the movement was still very strong. Measures taken in Baghdad did not always have consequences outside the capital, and the Muʿtazilah had established themselves in almost all parts of the Islamic world: in Upper Mesopotamia and in the Syrian Desert (among the Kalb); in several suburbs of Damascus and in Lebanon (for instance in Baalbek); in Bahrein and even in the Maghreb (again among certain tribes in what is today Morocco and Algeria); in Armenia; above all in western Iran, in the provinces of Kerman, Fārs (for instance in Arradjān and in Sirāf), and Khuzistan (for instance in Shūshtar, Susa, ʿAskar Mukram, and Gundē-shāpūr, at that time the seat of a famous medical academy directed by the Nestorians); and finally in India, in the area along the shore of the Indian Ocean to the west of the Indus Delta. In these centers the trend toward individualistic thinking and dialectical pyrotechnics had certainly not been as predominant as in Baghdad. Many of the Iranian towns mentioned are situated on the main trade routes: it seems that the common theological outlook created an atmosphere of confidence essential for better business.
This extended geographical base helped the Muʿtazilah to survive. However, it also fostered misunderstandings and tensions which came about through separate regional developments. In Baghdad, the caliph al-Mutawakkil had not only acted against the Muʿtazilah but also—and even more violently—against the Shīʿah. Consequently al-Jāḥiẓ (d. 869), a prominent Muʿtazilī author who had maintained close relations with Ibn Abī Duwād and other high state officials, wrote a book in which he praised the Muʿtazilah and at the same time attacked the Shīʿah. This opportunistic turn irritated a colleague of his by the name of Ibn al-Rāwandī (d. about 910), who had just come from the East, where he had acquired a sound reputation within and outside of his school. Iran was governed by the dynasty of the Tahirids, who did not follow al-Mutawakkil's anti-Shiʿah policy. Ibn al-Rāwandī therefore joined the Shīʿah in Baghdad and refuted al-Jāḥiẓ's book. But he did more than that: in a series of treatises he showed that some of the axioms accepted by the Muʿtazilah, such as the createdness of the world or the justice of God, that is, theodicy, were not based on solid premises and that the Qurʾān was full of contradictions. His enemies within the school called him a freethinker for those views, but he apparently wanted only to point out that, with respect to certain positions, the dialectical method allowed for arguments pro and con which simply neutralized each other. Thus the Muʿtazilah, shortly after being deprived of political power, were also faced with the inadequacy of their intellectual instruments.
The scholastic phase
Ibn al-Rāwandī's pinpricks produced a shock. His books were refuted by several authors, in Baghdad as well as in Basra. Although these two towns were far from the only strongholds of Muʿtazilī theology, their names served as labels for the two different schools that took up Ibn al-Rāwandī's challenge and, going beyond mere refutation, began systematizing the material accumulated in the past. At the beginning of the tenth century their main representatives were Abū al-Qāsim al-Kaʿbī (d. 931), who was identified with Baghdad although he only studied there and then taught in his hometown of Balkh in eastern Iran, and for Basra, al-Jubbāʾī (d. 915) with his son Abū Hāshim (d. 933). In their efforts to build up a coherent theological framework they had to care more than their predecessors about epistemological and terminological problems; this accounts for their growing interest in precise definitions and questions of logic.
The range of the Basran school, which, like the school of Baghdad, gradually shifted to Iran, is well attested by the work of the qāḍī ʿAbd al-Jabbār (d. 1024/5), chief judge at Rayy (near modern Tehran). His Mughnī ([The Book] That Makes [Other Books] Superfluous ), a twenty-volume summa theologiae, has recently been edited as far as it is preserved and also subjected to some research. Besides this valuable source, further texts written by his pupils and other theologians who followed his views are also available. For Muʿtazilī hermeneutics our best source is Abū al-Ḥusayn al-Baṣrī (d. 1044), with his Kitāb al-muʿtamad (The Reliable Book), although this book is concerned with the criteria of jurisprudence, not those of theology.
For the school of Baghdad, on the contrary, we lack extensive original documentation. But its ideas are relatively clear to us thanks to a development that deeply influenced the later history of the Muʿtazilah: the winning over of the Shīʿah. This process took place within two of the three main branches of Shīʿī theology; the Zaydīyah and the Twelver Shīʿah. Only the Ismāʿīlīyah preferred to seek support in Neoplatonic philosophy instead. Among the Zaydīyah the door had already been opened to Muʿtazilī thought by the imam al-Qāsim ibn Ibrāhīm (d. 860), though he did not intend to be a Muʿtazilī himself. The final decision was taken somewhat later and in two different regions: first by his grandson al-Hādī ilā al-Ḥaqq (d. 911), who founded a Zaydī principality, an imamate, in Yemen; second, though with certain setbacks, by the Zaydī pretenders in northern Iran, in the area near the Caspian Sea. Among the Twelvers, Muʿtazilī theology was introduced by two members of the Banū Nawbakht, a famous family of state officials and scholars: Abū Sahl Ismāʿīl al-Nawbakhtī (d. 924) and his nephew Ḥasan ibn Mūsā (d. between 912 and 922). The trend originally met strong resistance there, especially from the traditionist theologian Ibn Bābawayhī (d. 991), but then it prevailed because of the influence of Shaykh al-Mufīd (d. 1022).
The motives for these dogmatic shifts are not altogether clear. The Muʿtazilah still retained, among their theological principles, the obligation to spread the right belief; this may explain their usefulness for the Zaydī pretenders. The Twelvers had lost their spiritual leader through the disappearance (ghaybah ) of the twelfth imam in 874 and may have clung to rational theology for new reliable guidance. In Baghdad, Muʿtazilī theologians had always had moderate Shīʿī leanings, although they did not normally side with the Rawāfiḍ (i.e., the Twelvers). At the end of the ninth century, even a Basran like al-Jubbāʾī recognized ʿAlī 's son Ḥusayn as a righteous ruler. Abū al-Qāsim al-Kaʿbī was the secretary of an early Zaydī pretender in Iran. Later on, the crisis of the caliphate revived the hopes for political support. The Muʿtazilah in Baghdad entertained relations with Sayf al-Dawlah (r. 944–967), the Hamdanid ruler of Aleppo, and then with the Buyids; both dynasties had Shīʿī leanings. A Buyid vizier, al-Ṣāḥib ibn ʿAbbaād (d. 995), promoted the qāḍī ʿAbd al-Jabbār in Rayy.
The Baghdad school exerted its strongest influence among the Zaydīyah, though only in its Yemeni wing. The Caspian imams were under Basran influence, but their principality did not survive. However, theological works of the Iranian wing were taken over into the Yemen, for instance the great Qurʾān commentary written by al-Ḥākim al-Jushamī (d. 1101). Among the Twelvers, al-Mufīd followed the Baghdad school, but in the generation after him a pupil of the qāḍī ʿAbd al-Jabbār, Sharīf al-Murtaḍā (d. 1044), turned to Basran ideas and determined the Twelver outlook for the centuries to come. On the whole, the Zaydīyah adopted Muʿtazilī doctrine more fully than the Twelvers.
Outside of the Shīʿī areas, the Muʿtazilah were in retreat. But their impact persisted in at least two places. I have mentioned the Berber tribes in the Maghreb; some of them remained Muʿtazilī even after the Fatimid invasion, at least up to the second half of the eleventh century. They called themselves Wāṣilīyah, with reference to Wāṣil ibn ʿAṭāʾ. In fact, they seem to have lost contact with the Iraqi Muʿtazilah very early and did not participate in the move toward intellectualism instigated by the Abbasid court. We find scarcely any traces of theological activity; the claim to be Wasiliyah merely constituted a symbol of identity, the implications of which we do not know. In addition, Muʿtazilī theology continued to exert a certain attraction in Baghdad and in Iran, among jurists belonging to the school of Abū Hanīfah. In Baghdad, this combination became precarious after the weakening of Buyid power. In 1017, the caliph al-Qadīr forced the Ḥanafī judges and witnesses to make a public disavowal; pressure from traditionist, especially Hanbali, circles increased steadily. Eastern Iran, with its fragmented political landscape, offered better conditions. Muʿtazilī thought flourished under the Khwārizm-shāhs up to the beginning of the thirteenth century. We even know about a Muʿtazilī in the environment of Timur Lenk (Tamerlane, d. 1405). He, however, seems to have been an exception; generally speaking, the end of Muʿtazilī influence on Sunnī circles in Iran came earlier, with the Mongol invasion in the first half of the thirteenth century.
In the contemporary Muslim world, Muʿtazilī ideas are evaluated in different ways. In Iran, they still permeate theological thinking, especially after the revival of Shiism. In Yemen, they belong to the Zaydī heritage, but have lost all reproductive vigor. In certain Sunnī countries undergoing the impact of modernist movements, they have been thought of as giving witness to the essentially rational character of Islam; this has led, especially in Egypt during the last two generations, to a certain scholarly interest which was sometimes hailed as a "renaissance." Modern fundamentalism, however, has proved that view premature. Muʿtazilī ideas are again pushed back into the corner of heresy.
That Muʿtazilī doctrine changed over the centuries should go without saying. What is perhaps more important is the fact that its function also changed. During the incubation phase doctrine was less important than group solidarity; at the Abbasid court, Muʿtazilī theology represented the first attempt at a rational and universal description of Sunnī Islam; and in the later centuries certain basic positions, especially the doctrine of free will, served as a label which indicated that a person belonged to the Muʿtazilī "school."
The "five principles"
The decisive step toward creating a reliable "dogmatic" framework was apparently taken rather late, by Abū al-Hudhayl, who defined five principles that he considered indispensable to Muʿtazilī identity. These have determined the structure of Muʿtazilī theological works for centuries, in spite of the fact that two or even three of these principles did not retain much importance in later discussion. One of them was already dated when Abū al-Hudhayl took it up: the princple of "enjoining what is good and forbidding what is evil" (al-amr bi-al-maʿrūf wa-al-nahy ʿan al-munkar ), that is, active admonition to follow the right path and resist impiety. This had been the device of the revolution against al-Manṣūr in 762 and was probably the justification for Wāṣil's missionary projects, but it was rather out of place at the Abassid court.
The second principle, concerned with the intermediate state of the Muslim sinner, was still valid at that time, but derived its importance from an earlier debate that had its roots in the period before Wāṣil. It was a compromise between two attitudes that had arisen in the seventh century, on the one hand the rigorist belief that every Muslim who commits a grave sin excludes himself or herself from the community, and on the other, an "ecumenical," communitarian position which understood the revelation of the kerygma as the decisive event by which every Muslim, whether sinner or not, was ultimately saved. Wāṣil was rigorist enough to abhor any laxity or minimalism, but he lived late enough to realize that the exclusiveness practiced by the rigorists could not serve as the basis of the world religion that Islam had meanwhile become. He wanted the grave sinner to remain a member of the Muslim community, with all the rights that involved (safety of life and property, inheritance from other Muslims, etc.), but he insisted that the sinner would be condemned to eternal punishment in Hell like the pagans if that person did not repent. Thus, the grave sinner is to be treated neither as an unbeliever nor as a true believer. This doctrinal position was apparently not far from that taken by Wāṣil's teacher, Ḥasan al-Baṣrī, but it was Wāṣil who brought it into focus and sharpened it by changing the terminology. Ḥasan had called the Muslim sinner a munāfiq ("hypocrite"), using a word taken from the Qurʾān, where it referred to enemies of Muḥammad who had not openly sided with the unbelievers. This led to exegetical problems; the Qurʾanic context was not always appropriate for the theological definition wanted. Therefore Wāṣil used the term fāsiq ("transgressor") instead. This term was equally Qurʾanic but had merely moral and no historical connotations.
The eternal punishment of the "transgressor" was Abū al-Hudhayl's third principle. It turned out to be somewhat flexible. The school sometimes tolerated members who only believed in a kind of prolonged purgatory; the dogma was also mitigated among the Twelvers. This tolerance derived from the fact that the discussion shifted to a related issue that had been Abū al-Hudhayl's fourth (or, according to his own counting, second) principle: God's justice. God does not do wrong: he punishes the sinner and rewards the good. He has the right to do so because he has put human beings under an obligation; he has revealed the law, and men and women have the choice to obey or to disobey. This choice presupposes two things: freedom of decision and an instrument to grasp this possibility of decision, that is, reason. However, the fact that man is an intelligent being then implies that the obligation already existed before revelation; the prophets are sent to confirm what reason knows, or should have known, beforehand. To a certain extent, revelation is merely a sign of God's mercy or favoring help which makes insight easier for man. It may, however, add new commandments; there are rational laws (e.g., the interdiction of lying) and revealed laws (e.g., the prohibition of the eating of pork).
Abū al-Hudhayl's last principle was the unity of God. This had always been an indispensable postulate for Islam, as opposed to the Christian trinitarian beliefs and the dualism of Iranian religions such as Zoroastrianism or Manichaeism. But this unity had been understood in different ways. Early anthropomorphists such as the Qurʾān commentator Muqātil ibn Sulaymān (d. 767 or later) had found it sufficient to assert that God is one and not two or three, which did not exclude his having human shape. They merely held that God's body is compact, not hollow, for he neither eats nor drinks. He is thus one in number and consistence but he is not one in form, for he has limbs like man. The Muʿtazilah, on the contrary, understood unity as incorporeity, unity in essence: God is beyond time and place, he is unchangeable. They did not even agree when some Shīʿī anthropomorphists in Kufa, Hishām ibn al-Ḥakam among others, refined their position so far as to conceive of God as a body of light that radiates like an ingot of silver. For the Muʿtazilah, God is not a body at all but only "something," a being that cannot be perceived by the senses but is exclusively known to us through revelation or through reason, by the effect of his creative will in nature. God will therefore not even be seen in Paradise: he is unlike all other beings.
Unity then also means that the attributes ascribed to God in the Qurʾān are identical with him and not different entities or hypostases. This is true at least for "essential attributes" such as knowledge, power, or life; they are eternal and unchangeable like God himself and merely tell us something about certain aspects of his nature. "When one states that the Creator is knowing," said Abū al-Hudhayl, "he has asserted the reality of an act of knowing that is God and has denied ignorance in God and has indicated [that there is] some object known [to God] that has been or will be." However, in addition to "essential attributes" there are also "attributes of action" such as willing, hearing, seeing, or speaking, which describe God's temporal relationship with his creation; they are other than God and subject to change, for they come into being when God acts and cease when his action ceases. They do not subsist in him. This is why the caliph al-Maʾmūn declared the Qurʾān to be created: the Qurʾān is God's "speech" or "discourse" and came into being at a certain historical moment.
This theory presented certain problems. The attributes were inferred from Qurʾanic statements, where they normally took the form of predicates or "names," in sentences like "God is knowing," "God is (all) mighty," and so on. These sentences then had to be reformulated as, for example, "God has knowledge" and "God has power." How could this be done without reifying the attributes and transforming them into independent entities? And if knowledge really was God, as Abū al-Hudhayl seemed to say, how could it be differentiated from his power and other attributes? Finally, the question of whether there are attributes beyond revelation that have to be deduced rationally must be asked; predicates like "eternal" or "existent" were absent from the Qurʾān.
Since these questions referred to statements contained in the holy text, they were answered by linguistic analysis. The doctrine of attributes is based not on metaphysics but on grammar. Our understanding must therefore proceed from the Arabic language and medieval grammatical theory. The first differentiation can already be found in the translation of Abū al-Hudhayl's statement mentioned above: instead of "God has knowledge" we have to say "There is an act of knowing belonging to God"; Arabic linguistic feeling and Muʿtazilī theology insisted that "knowledge" is "knowing," that is, an infinitive rather than a noun, and as such not necessarily permanent or independent. This was not, however, a definitive solution. One could still, as did certain opponents of the Muʿtazilah such as Ibn Kullāb (d. 855?), conclude that this formulation referred to a separate entity that subsists in God. Al-Jubbāʾī therefore rephrased Abū al-Hudhayl's doctrine in the following way: "The meaning of one's describing God as knowing is (1) the assertion of [God's] reality, (2) that [God] is contrary to whatever cannot know, (3) that he who says that [God] is ignorant states a false proposition, and (4) an indication that there are things that [God] knows." Here, the act of knowing is completely excluded: what is asserted is only God's reality. This was enough to protect God against any plurality, but there was some reason for doubting whether it was enough to affirm his knowledge.
Al-Jubbāʾī's son Abū Hāshim then established a compromise by going back to the original Qurʾanic statements and inserting a copula into them (which is normally absent from nontemporal statements in Arabic): Allahu ʿālimun thus became kāna Allāhu ʿāliman, "God is knowing." The copula was then understood as a "complete verb," that is, it gained existential meaning: "God is"; the assertion of God's reality had been made explicit. The participle for "knowing," however, now put into the accusative instead of the nominative, was no longer interpreted as a predicate but as a ḥāl, a "state" of the subject instead of an attribute. In the words of Abū Hāshim himself: "Since it is true that [God] has a state in his being knowing, the knowledge that he is knowing is a knowledge of the thing itself [that is, the subject as] in this state rather than a knowledge of the act of knowing or of the thing itself." This theory allowed the above statements to be understood univocally of all knowers; a theological problem had been put into the general frame of grammatical analysis.
However, since the purpose of this description was ontological we find the later authors unfolding a whole range of different ways of predication and attribution. There are, for example, besides the "essential attributes" also "attributes of the essence," that is, attributes which not only make some statement about the essence but denote the essence itself and thus are not shared by anything else (in the case of God, a word like eternal). Then there are attributes grounded in the presence of an accident. They were subdivided by the qāḍī ʿAbd al-Jabbār into several categories: for example, those that specifically qualify the substrate (or object) and necessitate a state in it as a whole, like motion or rest; and those which specifically qualify the substrate but do not necessitate a state of it as a whole, as, for example, colors (which may inhere in a few single atoms of the substrate, but not the whole of it). From them we may distinguish attributes determined by the agent who causes the existence of the thing, like "well ordered" or "skillfully wrought" (with respect to the creation, for instance); attributes directly derived from his action, like "speaking," "commanding," and so on; and finally attributes that are grounded neither in the essence nor in an accident, like "self-sufficient" said with respect to God (since this word which, in Arabic, originally means "rich" is positive only in its form of expression whereas "its strict sense is the denial of need on the part of one who is specifically characterized by a state having which need and self-sufficiency are actually possible").
The Muʿtazilī worldview
The way in which the theory of the divine attributes developed shows that the achievement of Muʿtazilī theology consisted not only in defining basic tenets by which it distinguished itself from other schools but also in the conceptual and systematic framework through which these tenets were expressed. Another case in point is the atomistic model by which the relationship between God and his creation was explained. This model had been conceived as early as the eighth century by Ḍirār ibn ʿAmr, but the theologian who gave it its final form was again apparently Abū al-Hudhayl. He thought of atoms as mathematical points that do not have any spatial extension until they touch each other. Atoms therefore are not three-dimensional; only bodies are. Normally it is only in spatial existence, that is, in combination with each other, that atoms take on accidents: color, for instance, and consequently visibility. Only a few accidents are connected with atoms when they are still isolated, namely those that, in the form of an alternative, make combination possible: composition versus separation and, as the medium through which composition and separation take place, motion versus rest. Motion then means that the atoms join each other and receive extension, or that they part and thus lose their corporeity again; rest means that they retain their status of being isolated and therefore cannot be perceived, or of being composite and therefore remain spatial.
What is decisive, however, is that the atoms do not join by themselves, for their composition is an "accident," and accidents, by definition, are created by God. A body exists therefore only as long as God allows this accident of composition to endure. This is Greek atomism turned upside down. Democritus and Epicurus intended to explain nature by the principle of chance (therefore the church fathers abhorred their philosophy). For the Muʿtazilah, on the contrary, atomism served as an accomplice of divine omnipotence. As mere conglomerates, things do not possess any essence of their own; this demonstrates their dependence on God. However, although objects are reduced to purely material composites, this was not a materialistic theory. The Greek atomists were searching for the monad: they wanted to explain nature. The Muʿtazilī atomists were searching for the pivot of God's will and power: they wanted to explain creation.
That is why, in contrast to the Greeks, the Muʿtazilah extended atomism to the realm of human action. The freedom of choice that is necessary for human responsibility is based upon a capacity which the individual receives from God. This capacity is not a permanent quality, an inborn power of acting, but a momentary capability to do one specific act. The Muʿtazilah shared this basic assumption with their determinist or predestinarian opponents. What made the difference was merely the fact that they did not think that this capacity was given simultaneously with the act, as the determinists did, but one moment before, leaving an interval of one time atom so that people have the chance to make their decisions, to choose whether to perform the action, to leave it, or to do something else. There were Muʿtazilah, especially in the Baghdad school, who believed that the capacity for action was not merely momentary since it was identical with health and the intact functioning of the body, but even they understood this continuum as a mere accumulation of single isolated moments.
Continuity was a factor that never came first in this model. With respect to the physical world, consequences had to be taken into account especially regarding the explanation of movement, for in the context of atomism movement meant only that the atoms of a moving body were, at different moments, opposite to the subsequent atoms of the surface on which this body was moving. Consequently, one had to discuss the problem of when movement takes place at all, whether during the first moment, when it begins, or during the second, when it has already ceased to exist. Movement, it could be said, is only a convention of language; in reality, bodies are always at rest, though at subsequently different places. Continuity is then only an illusion of our senses, as in a film.
With respect to human existence, this lack of continuity comes to the fore in the Muʿtazilī concept of person. For Abū al-Hudhayl the human being was first a mere complex of atoms and "accidents." It is true that he or she is alive whereas other bodies are not, but life is again only an "accident," a quality added to the conglomerate of atoms which form the body. There is something like a soul, but it is conceived of merely as a kind of breath that permeates the body as long as that body is alive. The soul may leave the body during sleep; this explains the phenomenon of dreaming. But the soul is not immortal: it guarantees life and it disappears together with it. What Abū al-Hudhayl and his colleagues wanted to explain was not the continuity or uniqueness of the human person, but God's power to create human beings anew in the hereafter. God adds the "accident" of life to the atoms that form them when God creates them; God withdraws this accident when he makes them die; and adds it again when he resurrects them.
This concept of a fundamentally disjointed world seems to exclude causality. But the Muʿtazilah usually did not go so far as to deny causality completely. To begin with, they gave it a different name: "production," which they defined as the dependence of one act on another. The examples they adduced show, however, that they thought in terms of human actions alone. They did not want to formulate a law of nature but to bind the human being to the consequences of his or her own behavior. They were mainly interested in man and his responsibility; their approach was juridical rather than metaphysical, for they started from man's obligation toward God. Later on, they recognized that they could not neglect the universe completely and would have to give an explanation for the phenomenon of a world that functions in an orderly and foreseeable way in spite of its being dependent on God's will at each moment. Abū al-Qāsim al-Kaʿbī spoke of God's "habit" in this respect, the coutume de Dieu that changes only in the case of a miracle.
Abū al-Hudhayl's worldview that I have described so far mainly influenced the Basran school, through al-Jubbāʾī. But during his lifetime he was already contradicted by his nephew al-Naẓẓām, who did not believe in atomism but proceeded from the infinite divisibility of bodies. Al-Naẓẓām also reduced God's immediate omnipotence. According to him, inanimate things now have a nature of their own that is independent in its activity, although ultimately created by God. They are still composites of different elements; however, these elements do not simply agglomerate like atoms but mix and grow into organic units. Nor do they depend on God's will at each moment of their existence; they are rather created all at once and then behave according to their own character. The human being, too, is no longer merely unique because of outward form but possesses a soul which, though still material, persists beyond death; al-Naẓẓām was the first Islamic theologian to take over the Platonic proofs for the immortality of the soul. This soul, a "subtle body," permeates all the limbs and keeps them alive. It can mix with other "bodies" that come to it from outside, such as sounds; this explains sense perception.
For some time, al-Naẓẓām's alternative approach had enormous success. But there were certain excesses that discredited it. Ultimately it did not supplant atomism. Nevertheless, it was not without influence on the Baghdad school. The Baghdadis continued to believe in atomism, but they held that the atoms have extension and endure by themselves, as do their basic accidents. Characteristically enough, it was a Baghdadi like Kaʿbī who relied on the concept of the coutume de Dieu ; like al-Naẓẓām, he believed in the existence of natural qualities that determine the functioning of bodies and guarantee the preservation of the species. However, his natural philosophy was closer to Greek concepts than was al-Naẓẓām's; he followed the classical doctrine of the elements. In most of these points he was attacked by the followers of al-Jubbāʾī, who stuck to the Basran system.
Arguments for the existence of God were used within the circle of the Muʿtazilah at least from the time of Abū al-Hudhayl. The cosmological proof was known from Christian sources. But the Muʿtazilah preferred by far the argument e novitate mundi ; deriving the existence of God from the "accidental" character of creation corresponded to their atomistic worldview. Originally, the notion of God had been considered as a priori, "necessary," as one said. Abū al-Hudhayl even believed that the "proof" for God's existence, namely the createdness of the world, was also immediately evident. This insight then implied, as he thought, the obligation to speculate further about God's nature and to look out for his commandments; it had juridical consequences. When later theologians, in an attempt to create an overall rational system, gave up the a priori character of the notion of God, they were confronted, just because of the juridical aspect mentioned, with a very typical problem: how can man be obliged to know God if he does not already know him, that is, is it not that the obligation to recognize God's existence presupposes that man already possesses a notion of God? Al-Jubbāʾī then answered this question by assuming that man necessarily feel the obligation to know God because when he reaches intellectual maturity he becomes aware of being constantly exposed to the merciful assistance of an unknown reality and then realizes that in order to be grateful for this anonymous help he should know where it comes from; otherwise his benefactor might become angry at being unduly ignored. Man, as it were, awakens with an existential feeling of fear that only ceases when he has recognized that there is a God who will be just and merciful if his commandments are fulfilled. This cognition then grants that "tranquillity of the soul" that always results from knowing the truth.
Influence and originality
Muʿtazilī theology certainly participated in the process of the hellenization of Arabo-Islamic thought that started in the eighth century. But we should not forget that Abū al-Hudhayl and even al-Naẓẓām developed their ideas when most Greek texts were not yet available in Arabic or were just being translated. Obviously neither Aristotle nor the Neoplatonists exerted any impact on their thought. Al-Naz-zam's system reminds us in some places of the Stoics (especially their theory of krāsis di holōn, "total mixture"), but that influence was filtered through Iranian intermediaries and reached him not in the form of a written translation but through his contacts with followers of Bardesanes or Manichaeans who lived in Iraq.
It will not be possible to judge the overall situation adequately until we have further studies of individual Muʿtazilī thinkers and the "dark period" between the great Hellenistic philosophers, up to the time of Proclus (410?–485 ce) and Iamblichus, and the arrival of Islam. What cannot be doubted, however, is the originality of the Muʿtazilī approach: Ḍirār's and Abū al-Hudhayl's atomistic theory is a case in point. They took Greek spolia but used them for an edifice that was entirely theirs, a theological system that was juridical in its outlook rather than metaphysical. They rarely quoted the Qurʾān because they wanted to rely on reason, but nevertheless they always took the Qurʾān as their guide. They offered a coherent worldview that was different from that of the later Muslim philosophers. These therefore reacted with reticence and, finally, contempt. Al-Kindī still tried to adjust his thinking to Muʿtazilī axioms in certain points, but al-Fārābi treated the Muʿtazilī theologians as "dialecticians" who used the wrong method. The Muʿtazilah, on the other hand, never changed their approach. A few of them, like Abū al-Ḥusayn al-Baṣrī, became interested in Aristotelian philosophy, but ultimately they did not adopt Aristotle's basic categories in either logic or metaphysics.
There is no satisfactory work on the Muʿtazilah as a whole. Albert Nasri Nader's Le système philosophique des Muʿtazila (Beirut, 1956) treats the development up to the beginning of the scholastic (third) phase, but this work is marred by philological misunderstandings and gives almost no biographical information. W. Montgomery Watt's The Formative Period of Islamic Thought (Edinburgh, 1973) is an introduction to early Muslim theology where the Muʿtazilah appear in the context of the other theological movements existing during the first three centuries of Islam (until al-Ashʿarī). Harry A. Wolfson's The Philosophy of the Kalam (Cambridge, Mass., 1976) follows the author's earlier works on Philo and the church fathers: it presents stimulating ideas but ignores some recent editions and secondary literature. My article "Une lecture à rebours de l'histoire du muʿtazilisme," in Revue des études islamiques 46 (1978): 163–240 and 47 (1979): 19–69, gives a biographical and systematic account of the period between Wāṣil and Ibn al-Rāwandī.
For the rest, one has to resort to monographs, of which there are not many. Abū al-Hudhayl's system has been analyzed in a perceptive way by R. M. Frank in his study The Metaphysics of Created Being according to abû l-Hudhayl al-ʿAllâf (Istanbul, 1966) and in an article in Le Muséon 82 (1969): 451–506, "The Divine Attributes according to the Teaching of Abû l-Hudhayl al-ʿAllâf." Hans Daiber's Das theologisch-philosophische System des Mu ʿammar ibn ʿAbbād as-Sulamī (Beirut, 1975) treats another early thinker but also presents much material about his contemporaries. The qāḍī ʿAbd al-Jabbār has been dealt with most extensively in J. R. T. M. Peters's God's Created Speech (Leiden, 1976). For the Muʿtazilī influence on the Zaydīyah, see Wilferd Madelung's Der Imām al-Qāsim ibn Ibrāhīm und die Glaubenslehre der Zaiditen (Berlin, 1965). The relationship with the Twelvers is analyzed in detail, at least as far as the beginnings are concerned, by Martin J. McDermott in The Theology of al-Shaikh al-Mufid (Beirut, 1978). Much information on individual thinkers is hidden in the relevant articles in The Encyclopaedia of Islam, new ed. (Leiden, 1960–), and in the Encyclopaedia Iranica (Leiden, 1982–), but it is not easy for the nonspecialist to find this.
Studies of specific problems of Muʿtazilī theology tend to compare the Muʿtazilah with other "schools," especially the Ashʿarīyah. On the doctrine of attributes we have Michel Allard's Le problème des attributs divins (Beirut, 1965), which contains a chapter on al-Jubbāʾī, and especially R. M. Frank's Beings and Their Attributes (Albany, N. Y., 1978), which gives a subtle analysis of the teaching of the Basran school. The arguments for and against the doctrine of free will are listed and treated in detail by Daniel Gimaret in Théories de l'acte humain en théologie musulmane (Paris, 1980). Gimaret's interpretation of the Muʿtazilī model is modified by R. M. Frank in his article "The Autonomy of the Human Agent in the Teaching of Abd-al-Djabbar," Le Muséon 95 (1982): 323–355. The consequences for Muʿtazilī ethics are described in George F. Hourani's Islamic Rationalism (Oxford, 1971). The corresponding physical ideas become clear in Judith Katz Hecker's "Reason and Responsibility: An Explanatory Translation of Kitab al-Tawlid from al-Mughni fi Abwab al-Tawhid wa-l-Adl by Qadi Abd-al-Jabbār al-Hamadhani" (Ph. D. diss., University of California, Berkeley, 1975). For atomism, Salomon Pines's Beiträge zur islamischen Atomenlehre (Berlin, 1936) still retains its value; it has now to be compared with Frank's book on Abū al-Hudhayl. Carmela Baffioni's Atomismo e antiatomismo nel pensiero islamico (Naples, 1982) attempts a new comparison with Greek atomism, but unfortunately ignores Frank's study. On epistemology, see Marie Bernand's Le problème de la connaissance d'après le Mugni du cadi ʿAbd al-Ǧabbār (Algiers, 1982). Muʿtazilī polemics against dualism have been treated by Guy Monnot in Penseurs musulmans et religions iraniennes (Paris, 1974).
Josef van Ess (1987)