In the Middle Ages the madrasa (lit., “place to study” Ar. darasa “to study”; for discussion of darasa as a technical term meaning “to study jurisprudence” and darrasa meaning “to teach jurisprudence,” see Makdisi, 1961, pp. 10-11) was a college for the professional study of the Islamic sciences, particularly jurisprudence (feqh) but also the Koran, Hadith, and such ancillary fields as Arabic grammar and philology, knowledge of which helped in understanding sacred and legal texts. The so-called “foreign sciences,” like philosophy and medicine, which also formed part of a learned education, were most often studied in the teachers’ homes, as was literature when conceived as a field apart from the Islamic sciences.

The first stage of Islamic religious education was memorization of the Koran. Information on this stage is scanty (see iii, above). Many Islamic scholars boasted of having learned the Koran by the age of eight or nine years, though presumably most students took somewhat longer. Boys in elementary school (kottāb) were also taught reading, writing, and arithmetic. Next came the memorization of Hadith. In biographical notices the age at which a scholar began to “hear” Hadith is often mentioned; as young as six years of age was not uncommon. A traditionist might receive payment for teaching the Hadith he knew, and there was greater demand for traditionists who had heard from authorities that few others could cite. There was thus an incentive for starting to collect Hadith very early (Bulliet, 1983). Other scholars are said, however, to have begun collecting Hadith in their teens and twenties; there are occasional reports of traditionists who would not relate Hadith collected at an early age, for fear of making mistakes. As serious collecting of Hadith entailed travel to other cities, it can be surmised that it normally occupied the teen and later years. For example, the celebrated traditionist Moḥammad b. Esmāʿīl Boḵārī (d. 256/870) left elementary school at the age of eleven years, devoted himself to memorizing important collections of Hadith until he was sixteen years old, then traveled with his mother and brother to Mecca (Taʾrīḵ Baḡdād II, p. 7). Until the 10th century, in Hanbalite schools, the collection of Hadith and the study of jurisprudence could scarcely be separated; both meant sitting before various masters and memorizing whatever they had to teach, whether the words and deeds of the Prophet and early Muslims (Hadith) or the legal opinions of later teachers (jurisprudence).

Travel in pursuit of Hadith was crucial to the development of the madrasa as a separate institution, inasmuch as it stimulated the foundation of residences specifically for students. In the first Islamic centuries the Islamic sciences were always studied in mosques or private houses. Itinerant students needed places to stay. Often they slept in the very mosques where they heard lectures in the daytime. An ascetic might make the mosque his home for thirty years or more. Yet more private quarters were usually preferred, especially for students who remained in one place for long.

During the 10th century the number of long-term students increased, especially as jurisprudence developed into a field of study distinct from Hadith. The prerogative of issuing juridical opinions came to depend less on having collected the opinions of various judges and legal scholars as transmitted by teachers than on having studied under a particular master and received a license from him. Students of jurisprudence were therefore less given to continual travel than students of Hadith.

In the early centuries long-term students who did not sleep in the mosque usually stayed in the merchants’ ḵān. Often the students themselves were long-distance traders; in biographical dictionaries religious scholars are more often associated with the occupation “trader” than with any other, and there are many stories of traveling students who supported themselves by selling goods they had brought with them.

Religious scholars were recipients of alms in early Islam, so it is not surprising that eventually ḵāns were endowed specifically for them, usually next to mosques, and that sometimes teaching took place in them. For example, the celebrated traditionist Moslem b. Ḥajjāj Qošayrī Nīsābūrī (d. 261/875), though he lived off agricultural rents, lectured at Ḵān-e Maḥmeš in Nīšāpūr (Ḥākem Nīsābūrī, cited in Ḏahabī, p. 570). Ḵān-e Saʿd, named for a trader and traditionist, stood next to the mosque of Dīnār, the most prominent in 10th-century Gorgān (Sahmī, p. 109). Badr b. Ḥasanūya (d. 405/1014), governor of the Barzīkānī Kurds under the Buyids, was said to have founded two or three thousand mosque-ḵān complexes in Persia (Ebn al-Jawzī, s.a. 405; Ebn Kaṯīr, XI, p. 354).

The first notices of buildings called madrasas also appeared in the 10th century (Halm, pp. 438-39). The earliest examples appear to have been founded by the men who taught in them. From the later 11th century madrasas normally included living quarters for students, as in the old mosque-ḵān complexes. Customarily each was devoted to the study of jurisprudence according to one school (Makdisi, 1961, pp. 10-11).

In some early references the terms for mosque and madrasa appear to have been used interchangeably, as teaching took place in both. Perhaps the distinction is clearest from the point of view of the founder. The mosque was among the earliest institutions of Islam and unalterably public in character, originally the site of government and justice, as well as of prayer and religious learning. Once a patron had endowed a mosque as a pious foundation (waqf), it passed out of his effective control; for example, though it might have been established for a particular teacher of jurisprudence, it was not in the patron’s power to appoint a successor to that teacher. By contrast, the patron who founded a madrasa might retain, for himself and his heirs, the right to appoint or even dismiss the faculty and staff. The importance to potential founders of such continuing control is illustrated negatively by the rarity of madrasas in lands where Malikite law prevailed, for Malikite law is peculiar in denying such control (Makdisi, 1981, pp. 37-38).

Pious foundations were better protected against confiscation and against division among heirs, who nevertheless might benefit from them, for example, as administrators. Teachers of the religious sciences often had important popular followings, so that patronage might also help secure the loyalty of these segments of the population. Max Weber suggested that ethnically alien regimes and especially personal rulers will always favor men of religion, in order to tame the subject populations (pp. 1160, 1176). Perhaps this suggestion helps to explain the growth of madrasas in the period when Turkish soldiers ruled over most Arabic- and Persian-speaking Muslims. Some patrons sponsored particular schools, as for example, the Saljuq vizier Neẓām-al-Molk (d. 485/1092), who founded madrasas in Baghdad, Balḵ, Nīšāpūr, Herat, Isfahan, Baṣra, Marv, Āmol, and Mosul for teachers of the Shafiʿite school of law (Makdisi, 1961, p. 44). Others sought to maintain a balance; for example, the chief of Bayhaq built separate madrasas for Hanafites, Shafiʿites, Karrāmīs, and evidently Zaydīs or Muʿtazilites (Halm, p. 443).

From the point of view of the teachers and students, however, there was no difference in curriculum or procedure between mosque and madrasa. Students preferred the madrasa if they received lodgings and small stipends; teachers often gained renown from association with particular madrasas. For example, Neẓām-al-Molk founded his madrasa in Baghdad for the Shafiʿite Abū Esḥāq Šīrāzī. Abū Esḥāq initially refused to transfer there from his mosque, and another Shafiʿite, Ebn Ṣabbāḡ, was named to teach there. Abū Esḥāq relented only when his students threatened to transfer to the new madrasa, even if it meant changing teachers (Makdisi, 1961, pp. 31-35). They must have been attracted by the prospect of lodging and stipends. Abū Esḥāq, in turn, must have been strongly influenced by the prospect of many students. Indeed, in the biographical dictionaries sixty-seven men who studied jurisprudence under Abū Esḥāq are named, compared to only eleven who studied under Ebn Ṣabbāḡ, even though contemporaries rated the two teachers equal or even preferred Ebn Ṣabbāḡ.

Students at each madrasa would gather about their teacher in a circle (ḥalqa). One’s position in the circle was an indicator of prestige, and there are stories of how a new student who impressed his teacher with some display of brilliance might be asked to move closer within the group. The class (majles) would begin with recitations from the Koran and prophetic Hadith, then proceed to the business at hand. Teachers of jurisprudence usually took about twenty students at a time, but audiences for simple dictation, especially of Hadith, might reach tens of thousands at the chief mosques.

The basic task of the practicing jurisconsult was to answer questions from laymen and, if he was a judge, to decide cases. He would derive his answers by combining relevant texts from the Koran and Hadith according to the principles of his legal school. The student had thus to memorize prodigious amounts of material, both Hadith and the opinions of previous jurists. The outstanding method of training students to use the material they had memorized was disputation (monāẓara), in which two scholars would argue different sides of a question, each seeking to reduce the other to silence or, failing that, to persuade the audience that he had overcome his opponent.

After about four years the master would examine a student and, if he found him qualified, give him a license (ejāza) to answer juridical questions. If a student wished not only to answer questions but also to teach jurisprudence, he became a fellow (ṣāḥeb) of the master for an indeterminate period. A fellow might assist as repetitor or tutor (molqī, moʿīd, mofīd). He might also produce a taʿlīqa, a compilation of lecture notes, often a commentary on some authoritative handbook like al-Moḵtaṣar of Abū Ebrāhīm Esmāʿīl Mozanī (d. 264/877) of the Shafiʿite school or al-Jāmeʿ al-kabīr of Abū ʿAbd-Allāh Moḥammad Šaybānī (d. 189/905) of the Hanafite. This work was virtually a doctoral dissertation. At last the master would license the qualified fellow to teach, perhaps when a position had opened up for him (see Makdisi, 1981; for arguments for a social basis of licensing, see Bulliet, 1972; cf. Mottahedeh).

In contrast to the professional jurist, a scholar was also expected to know a great deal about literature, philosophy, and other subjects. Although he might learn a little of these subjects in mosque and madrasa, such studies were mostly followed elsewhere. One reason was that the mosque and the madrasa, as Islamic religious institutions, had no place for the “foreign sciences.” A comprehensive literary education went well beyond what was necessary to understand the Koran and Hadith. In addition, it was neither practical nor desirable to require formal training and certification in literature and other subjects comparable to the training and certification of jurists; individuals learned through private reading, rather than by hearing lectures and undergoing examinations by certified masters (see Makdisi, 1990, pp. 68-70, 88-117, 217-27).

It must be stressed, however, that the study of jurisprudence in the madrasa, as well as studies outside, was informal and individual. Licenses to issue juridical opinions or to teach jurisprudence were always given by individual masters, not institutions or faculties. In biographical notices it is usually reported that someone learned jurisprudence from a particular master but never that he learned it at a particular madrasa. Moreover, there was never a uniform licensing procedure, even in specialized institutions like the madrasa. Madrasas were also never state institutions; politicians might endow them but from their personal fortunes and as Muslims individually promoting learning. Bureaucracies were staffed by scholars, and a knowledge of jurisprudence was part of a scholar’s education; it is anachronistic, however, to think of patrons as deliberately founding madrasas to train state functionaries (pace Tibawi, pp. 232-38).


Bibliography: (For cited works not found in this bibliography and abbreviations found here, see “Short References.”)

R. W. Bulliet, “The Age Structure of Medieval Islamic Education,” Stud. Isl. 57, 1983, pp. 105-17.

Idem, The Patricians of Nishapur, Cambridge, Mass., 1972.

Šams-al-Dīn Moḥammad Ḏahabī, Sīar aʿlām al-nobalāʾ XII,ed. Ṣ. Samr, Beirut, 1983.

ʿAbd-al-Raḥmān Ebn al-Jawzī, Montaẓam, ed. ʿA. Q. ʿAṭāʾ and M. ʿAṭāʾ, 18 vols., Beirut, 1992.

ʿEmād-al-Dīn Esmāʿīl Ebn Kaīr, Bedāya wa’l-nehāya, 14 vols., Cairo, 1932-39.

H. Halm, “Die Anfänge der Madrasa,” in ZDMG, Supplement 3/1, XIX. Deutscher Orien-talistentag, Wiesbaden, 1977, pp. 438-48.

G. Makdisi, “Muslim Institutions of Learning in Eleventh-Century Baghdad,” BSOAS 24, 1961, pp. 1-56.

Idem, The Rise of Colleges. Institutions of Learning in Islam and the West, Edinburgh, 1981.

Idem, The Rise of Humanism in Classical Islam and the Christian West, Edinburgh, 1990.

R. P. Mottahedeh, review of R. W. Bulliet, The Patricians of Nishapur, JAOS 95, 1975, pp. 491-95.

J. Pedersen and G. Makdisi, “Madrasa,” in EI2 V, pp. 1123-34.

Abu’l-Qāsem Ḥamza Sahmī, Taʾrīḵ Jorjān, Hyderabad, 1950.

A. Tibawi, “Origin and Character of al-madrasah,” BSOAS 25, 1962, pp. 225-38.

M. Weber, Wirtschaft und Gesellschaft, Tübingen, 1922; 4th ed., Tübingen, 1956, tr. E. Fischoff et al., ed. G. Roth and C. Wittich as Economy and Society. An Outline of Interpretive Sociology, Berkeley, Calif., 1978.

(Christopher Melchert)

Originally Published: December 15, 1997

Last Updated: December 9, 2011

This article is available in print.
Vol. VIII, Fasc. 2, pp. 182-184