The article below by Professor Eckhard Neubauer on the history of music in Iran (650-1370 AD) was originally posted in the Encyclopedia Iranica on February 20, 2009.


When in 31/651 Yazdgerd III, the last Sasanian king, left Iran fleeing from the Arab troops, he took with him “1,000 cooks and 1,000 musicians” (Ṯaʿālebi, Ḡorar,p. 742; Ḥamza Eṣfahāni, Taʾriḵ seni I, p. 63). This statement, along with other historical accounts, testifies to the great importance given to music at the Sasanian court. ʿAmr b. Baḥr Jāḥeẓ (d. 255/868) even used the fact that music was considered an art among the Persians and the Greeks as an argument in favor of music in Islam (Jāḥeẓ, Ketāb al-qiān, tr., Beeston, pars. 28-31). Under Arab-Islamic rule local musical traditions lived on in Iran, though on a provincial level. Abu Yusof Kendi (d. after 256/870) saw a difference between the languages and the musical styles of Persians, Turks, the people of Deylam, Ḵazar “and others” (Kendi, Moʾallafāt, p. 137; Idem, Resāla fil-loḥun, p. 26). From pre-Islamic times on the Arabs had introduced Iranian elements into their own music. They adopted the ‘Persian lute’ (ʿud fāresi) and made it the principal instrument of their urban and court music. After the predominance of the harp in pre-Islamic Iran and the lyre in the Greco-Byzantine culture, the lute came to personify a new period of music history in the Islamic world and beyond. Its Arabic name ʿud, whose etymology is not yet convincingly explained, may well have been derived from the Persian word rud, since other terms belonging to the instrument like zir (highest string), bam (lowest string) and dastān (fret) were also adopted from the Persian language, and the names of the second and third strings, maṯnā and maṯlaṯ, may well have been translated from the Persian terms dotār and setār.

The new art music of the Arab-Islamic world had emerged from older local traditions, including Persian ones, and we know of a number of musicians of Iranian origin who acted first in the early musical centers of Mecca and Medina, and later in Baghdad under the ʿAbbasids as singers, composers, and writers on music, such as Našiṭ (2nd half of the 1st/7th cent.), Yunos Kāteb (1st half of the 2nd/8th cent.), Dārā Fāresi (Ebn Ḵordāḏbeh, Moḵtār, p. 54), Salmak Rāzi, who served under Hārun al-Rašid (r. 809-13) and is said to have introduced the musical meter ramal into Persian music (Abu’l-Faraj Eṣbahāni, Aāni, I, p. 379) or Ebrāhim Mawṣeli (d. 188/804) and his son Esḥāq Mawṣeli (d. 235/850), who dominated the court music from the time of al-Mahdi (r. 775-85) down to al-Motawakkel (r. 847-61). The main activity of the Mawṣelis, however, was not directed towards Persian music but towards the heritage of the great masters of Mecca and Medina. Esḥāq defended an unadulterate performance of the traditional Arab music against his colleague and rival prince Ebrāhim b. Mahdi (d. 224/839), who followed a less rigid, “romanticist” style. Singing girls from Khorasan performing at the courts in Damascus and Baghdad may have sung and danced in Persian as well as in Arab fashion.

According to Arab authors, who are our principal source for Persian music in the early Islamic period, musicians such as Ebn Mesjaḥ (d. ca. 90/710) and his pupil Ebn Moḥrez traveled to Iran and Syria to learn Persian and Byzantine music respectively (Aḡāni³ I, p. 378, III, p. 276). Though being of Persian descent and living in the mixed Arabo-Persian society of Iraq, the young Ebrāhim Mawṣeli traveled to the cultural center of Ray to study Persian as well as Arab music (Aḡāni³ V, pp. 157-58). This is not astonishing when we consider that an Arab composer such as ʿAbd-Allāh b. Moʿāwia Bāheli, who had accompanied his tribesman Qotayba b. Moslem to Khorasan, lived himself in Ray (Ebn Ḵordāḏbeh, Moḵtār, p. 24). Arab and Persian music existed side by side and were handed down separately. Abu Naṣr ʿAbd-al-Raḥim Qošayri (d. after 542/1147), a son of ʿAbd-al-Karim b. Hawāzen Qušayri (d. 465/1072), the author of the famous Sufi Resāla, compiled in Nišāpur a Ketāb al-aḡāni containing Arabic song texts (ʿEmād-al-Din Eṣfahāni, Ḵaridat al-qasr… II, p. 104).

The local Persian traditions were supported when, in the 3rd/9th century, Iranian dynasties regained power and music became once again “one of the signs of rule” (yak-i az amārāt-e pādešāhi, see Naršaḵi, Tāriḵ-e Boḵārā, 1892, p. 258). The poet Rudaki, a boon companion of the Samanid Naṣr II (r. 301-31/914-43), composed songs to his own verses and accompanied himself on the lute or the harp (Naršaḵi, 1892, p. 251; Neẓāmi ʿArużi, pp. 49-54). The same is reported of his poet colleague Farroḵi Sistāni (d. 429/1038). He served at the court of Sultan Maḥmud Ḡaznavi (r. 388-421/998-1030) along with a singer called ʿAndalib and a tanbur player called Buqi. Maḥmud’s son and successor, Sultan Masʿud (r. 421-32/1031-40), was entertained by the lute player Moḥammad Barbaṭi and the songstress Setti Zarrin-kamar, also called Setti Zarrin Moṭreba (Mašḥun, I, p. 163).

Several Saljuk rulers were fond of music. A famous lute player from Khorasan, called Kamāl-e Zamān (Perfection of the age), performed at the court of Sultan Sanjar (r. 511-52/1118-57) in Marv (Juzjāni, Ṭabaqāt I, 1949, p. 308). ʿAlāʾ-al-Din Kayqobād (r. 616-34/1219-37), the Saljuq ruler of Anatolia, presented the Ayyubid ruler in Damascus, Malek Ašraf Musā (r. 626-35/1228-37), who came to visit him, a very capable and beautiful female harp player (Duda, p. 148). The female poet and musician Ferdows Moṭreba from Samarqand was favored by the Ḵᵛārazmšāh ʿAlāʾ-al-Din Moḥammad (r. 596-617/1200-20; Jovayni, II, p. 56). When, in 617/1220, Bukhara and Samarqand were captured by Čengiz Khan, Ferdows was taken over by the Mongol ruler, who is said to have saved the artists of the towns he seized (Abu’l-Ḡāzi Bahādor Khan, II, 119).

During the rule of the Il-khans the main musical activities shifted west, and the rulers became accustomed to Irano-Arab urban art and court music. After capturing Baghdad in 656/1258, the Il-khan Hülegü (Hulāgu) Khan (r. 654-63/1256-65) saved the life of the eminent musician and writer on music Ṣafi-al-Din Ormavi (d. 693/1294), who had served al-Mostaʿṣem (r. 640-56/1242-58), the last ʿAbbasid caliph (Neubauer, “Ṣafi-al-Din,” p. 806). Šams-al-Din Jovayni (d. 683/1284), Hülegü’s ṣāḥeb-e divān, made his own house a center of musical activities. He not only supported Ormavi, who dedicated to him his second book, al-Resāla al-šarafiya, but also three of his pupils, Ḥasan Nāʾi, ʿAli Setāʾi, and Ḵᵛāja Zaytun, as well as other musicians such as Abu Bakr Tawrizi and Yaḥyā Ḡarib Wāseṭi (b. in 661/1263; Neubauer, 1969, pp. 251-60). With the exception of Ḡāzān Khan (r. 694-703/1295-1304), most of the Il-khanid rulers were fond of music. Abu Saʿid Bahādor Khan (r. 716-36/1316-35) even took lessons from his favored musician Kamāl Tawrizi; he also played the lute and composed songs. Another musician at his court, who still served under Musā Khan (r. 737/1336-7), was Neẓām-al-Din b. Ḥakim (d. ca. 760/1360), a pupil of Ṣafi-al-Din Ormavi in the second generation (Neubauer, 1969, pp. 257-8). Several of Ormavi’s students had emigrated to places like Mardin, Ḥamāt, Damascus, and Cairo.

Musical instruments. While pointing to the sound of instruments as an indicator of differences in musical taste, Abu Yusof Kendi mentions that “the Iranians are not moved by the organ as the Indians or the Greeks are not moved by the pandore from Khorasan” (ṭonbur ḵorāsāni, see Kendi, Moʾallafāt, p. 72). Long necked lutes (ṭanābir) were in favor “with the people from Ray, Ṭabarestān, and Deylam. The Persians prefer the ṭonbur to most other instruments” (Ebn Ḵordāḏbeh, Moḵtār, p. 16; Masʿudi, ed. Pellat, V, p. 128). The Arabs, who spelled the Persian word tanbur as ṭonbur, were ravished by the sound of its two strings whose seven frets were said to match the number of days and planets (Kendi, Moʾallafāt, p. 74). Abu Naṣr Fārābi (d. 339/950) described the different tunings of the instrument (see FĀRĀBI v. MUSIC). The poet Rāʿi (early 2nd/8th cent.) called it “harsh-sounding,” while in a verse by the Arab poet Ḏu’l-Romma (d. 117/735) the instrument “raises its voice in intoxication, its melody containing what is foreign to the dialects of the Arabs” (Mofażżal b. Salama, Ketāb al-malāhi, tr., p. 15). The “agility” (ḵeffa) and “velocity’ (sorʿa) of ṭonbur players (Kendi, Moʾallafāt, p. 137) formed their notion of Persian music more than anything else. Biographical and artistic data of male and female ṭonbur virtuosi were collected in particular books (e.g., Ketāb al-ṭonburiyin by Jaḥẓa and Ebn Ṭarḵān’s Aḵbār al-moḡannin wa’l-ṭonburiyin; Ebn al-Nadim, ed. Tajaddod, pp. 163, 173, tr. Dodge, pp. 319, 342). Quotations from some of these titles have survived in Abu’l-Faraj Eṣfahāni’s (d. 356/967) Ketāb al-aḡāni, and names of pandore players were listed by the 5th/11th-century Egyptian court musician Ebn al-Ṭaḥḥān (Ḥāwi al-fonun,p. 119).

The upper chested angular harp (čang) was one of the instruments favored in Sasanian Iran and later. It formed, together with the lute (ʿud), the main body of urban and courtly chamber music (Ebn Ḵordāḏbeh, Moḵtār, p. 15). Both instruments are well documented in literary and iconographic sources. The couple was completed by the instrument called nāy(-e siāh), which in those days was not the rim-blown flute (nāy-e safid) but a chalumeau (mezmār) played with a reed (Ebn Zayla, Kāfi, p. 78; Kāšāni, Kanz al-toḥaf, p. 114). A Persian ʿud player and a nāy player are depicted in a fresco in Qaṣr al-Ḥayr al- Ḡarbi in Syria, which dates back to ca. 90/710 (Farmer, Islam:Musikgeschichte, p. 35). Other pairs of instruments used in Iran in early Islam were the reed instrument zonāmi and the tanbur, sornāy and ṭabl, mouth ogan and čang (Ebn Ḵordāḏbeh, Moḵtār, p. 15). The rim-blown variety of the nāy was used in folk music. It was emancipated in Sufi circles and entered art music not much earlier than in the 8th/14th century. In the early Islamic period, the barbaṭ was not identical with the ʿud. It had a longer neck and a smaller body. It was compared with the Roman balance (qarasṭun, see Mofażżal b. Salama, Ketāb al-malāhi, tr., p. 6-7), and may be depicted in Qoṣayr ʿAmra (see Farmer, Islam:Musikgeschichte, p. 32). Only later, when the instrument had fallen into oblivion, did the name barbaṭ come to refer to the short necked lute (e.g., Ebn Sinā, Šefāʾ, p. 144). The best barbaṭ players were said to come from Marv (Jāḥez, al-Tabaṣṣor p. 37).

Among the instruments of chamber music mentioned or described in later sources we find those with “open strings” like qānun (its tuning is first given in Ormavi, Adwār, ms. Fatih 3662, fols. 22r-23r), santur (first mentioned by Manučehri), the rectangular psaltery nozha “invented” by Ormavi and described by Ḥasan Kāšāni (Kanz al-toḥaf, pp 116-17), the lute type robāb, a “lute psaltery” called moḡni (pp. 113-14, 118-19), which was a predecessor of the Indian sārangi, and the four-octave hybrid šāhrud (invented near Samarqand in the early 4th/10th century, and described by Abu Naṣr Fārābi), the bowed instruments kamānča (first? mentioned in a verse by the poet Sanāʾi, d. 525/1131) and ḡešak (Kāšāni, Kanz al-toḥaf, pp. 112-13), the rim-blown flutes šabbāba (ʿAlišāh Boḵāri, Ašjār wa-aṯmār, unpubl. transcription by Amir Ḥosayn Pourjāvadi, p. 3) and biša (Kāšāni, Kanz al-toḥaf, pp. 114-55), and the daf that “takes care of the meter and rhythm of all the other instruments” (ʿAlišāh Boḵāri, Ašjār wa-aṯmār, transcription by Pourjāvadi, p. 3). Descriptions of these and more instruments were later given by ʿAbd-al-Qāder b. Ḡaybi Marāḡi (d. 838/1435). Pathbreaking studies on the development of musical instruments in Iran were published by Henry George Farmer in The Encyclopaedia of Islam and elsewhere (in particular Idem, Islam:Musikgeschichte; idem, Studies in Oriental Music; see also Mallāḥ, Farhang, ssv.).

Music theory. The educated musicians who had access to Arabic books could deepen their knowledge of music theory by reading the treatises of a great number of authors active from the 2nd/8th century onward. Traditional Arabic theory was distinguished from that inspired by Greek music theory. The Greek and Byzantine traditions were present in the work of Abu Yusof Kendi (d. after 256/870), Abu Naṣr Fārābi (d. 339/950), Abu ʿAbd-Allāh Ḵᵛārazmi (wrote ca. 375/985), the Eḵwān al-Ṣafāʾ (4th/10th cent.), Ebn Sinā (d. 428/1037), Ṣafi-al-Din Ormavi (d. 693/1294), and others. An abridged Persian translation of the Rasāʾel-e Eḵwān al-Ṣafāʾ was made in the early 7th/13th century under the title Mojmal al-ḥekma. Its chapter on music found a wide circulation. Both titles by Ṣafi-al-Din Ormavi, Ketāb al-adwār and al-Resāla al-šarafiya, were several times translated into Persian and commented on in both languages.

As in other fields, the Arabian element was predominant even when the books were written in Persian like the Dāneš-nāma-ye ʿalāʾi, composed in 428/1037 by Ebn Sinā in Isfahan, or Faḵr-al-Din Rāzi’s (d. 606/1209) Ḥadāʾeq al-anwār fi ḥaqāʾeq al-asrār, which was composed at the command of the Ḵᵛārazmšāh ʿAlāʾ-al-Din Tekeš and completed in 575/1179. A certain exception to the rule can be seen in the Qābus-nāma (written in Persian, 475/1082) of ʿOnṣor-al-Maʿāli Kaykāvus b. Eskandar and in Ebn Zayla’s Arabic Ketāb al-kāfi, in which the Iranian element is recognizably represented. Some Persian versions of the ʿAjāʾeb al-maḵluqāt wa-ḡarāʾeb al-mawjudāt by Zakariyāʾ b. Moḥammad Qazvini (d. 682/1283) contain additional sections on music. An important Persian text on the subject was written by Qoṭb-al-Din Maḥmud b. Masʿud Širāzi (d. 710/1311), who was a pupil of Naṣir-al-Din Ṭusi and knew Ṣafi-al-Din Ormavi, in his encyclopedia Dorrat al-tāj le-ḡorrat al-Dobāj, composed on behalf of the prince of Gilān, Dobāj b. Filšāh.

Less original is the chapter on music in the Persian encyclopedia Nafāʾes al-fonun fi ʿarāʾes al-ʿoyun, written in about 740/1340 by Moḥammad b. Maḥmud Āmoli for the Inju ruler Abu Esḥāq b. Maḥmudšāh (d. 758/1357) in Shiraz. Another book on music was written under the title Laṭāʾef al-asrār le-maqāṣed al-adwār for Jalāl-al-Din Turānšāh, a vizier of Abu Esḥāq. The author may have been Yaḥyā b. Aḥmad Kāši, who translated Ormavi’s books into Persian and dedicated them to Abu Esḥāq. The Laṭāʾef al-asrār, which is closely related to Ormavi’s work, has survived in only one copy (see Raʿnā Ḥosayni, pp. 748-56). In about 750/1350, a certain Ḥasan Kāšāni wrote an impressive book on the theory and practice of music entitled Kanz al-toḥaf. He was the first in Iran to deal in greater detail with musical instruments. Additional treatises and anonymous titles are mentioned in the catalogues by Aḥmad Monzavi (1970), Moḥammad-Taqi Dānešpažuh (1976) and Moḥammad-Taqi Massoudieh (1996).

The tonal system. Music theory proper may be divided into three disciplines: the tonal, modal, and metrical systems. As to the first mentioned, later Persian sources have preserved traces of an archaic tonal system. From our limited information we can conclude that it comprised eighteen pitches or bāngs (lit. voices, sounds, tones), divided into whole-tones (bāng) and half-tones (nim-bāng). It seems to have been a method of music instruction inherited from pre-Islamic times, and it remained unaffected by Arab influence. Abu Aḥmad b. Monajjem (d. 300/913) confirms that eighteen was also the number of notes of the two-octave system of the “ancients.” As a consequence, the microtones used in urban and court music since the 3rd/9th century were neglected. Still, in the 7th/13th century and later the method served to define the melodic modes by a standardized number of bāngs. In doing so, the number eighteen stood for the total number of bāngs of the twelve main modes. The system was related to the two-stringed tanbur that was used in Persian music theory comparably to the use of the four-stringed lute (ʿud) in the Arab world, and it was based on the scale of the mode rāst. The method recalls that of Esḥāq Mawṣeli (d. 235/850) in Baghdad, who defined the modes according to the first two or three notes of their course (majrā) on the second string (maṯnā) of the lute. He also disregarded the microtones, but in contrast to the bāng system he gave the notes precise positions on the finger board. The term bāng lived on in Persian theoretical writings until the 11th/17th century. In some later texts the bāng figure was further specified by indicating the more precise fret (parda) position.

In contrast to the bāng method used in musical practice, the pitches were exactly measured on the fingerboard of the lute and recorded by scholars such as Abu Naṣr Fārābi, Ebn Sinā, Ṣafi-al-Din Ormavi, and Qoṭb-al-Din Širāzi in their respective books. As a result of physical and mathematical endeavors the tetrachord was divided into eleven steps by Fārābi, and seven steps by Ebn Sinā and Ormavi. Qoṭb-al-Din adopted the figures of Fārābi and Ormavi and added some data of common (ʿorfi) intervals from the practice of the time. The final division of the octave into seventeen steps was recorded by Ormavi. It is a Pythagorean scale, and it became the accepted norm of later Persian music theory (Manik, Das arabische Tonsystem, pp. 63 ff.).

The modal system. The notion of bāng was closely related to the local modal systems, the oldest of which was inherited from Sasanian times. It was a set of seven modes, whose names are listed in Arabic sources but are distorted by copyists and only few of them can tentatively be identified, such as bahār, nayruzi (nowruzi), and mehrajāni (Farmer, “The Old Persian Musical Modes”). Ebn Ḵordāḏbeh (d. 300/913), who speaks of eight traditional Persian modes but gives only seven names, describes some of their characteristic features in detail (Moḵtār, p. 15).

The notion of seven modes survived in Iran even after the character and names of the modes, now called parda, had changed. Four of the new names were first mentioned by Ebn Sinā in the chapter on music of his Ketāb al-šefāʾ, including the main mode rāst in its Arabic form mostaqim. Nine additional names (māda, ʿerāq, ʿoššāq, zirāfgand, busalik, sepāhān (= eṣfahān), navā, goḏāšta and rāhavi) are listed by Kaykāvus b. Eskandar (Qābus-nāma, p. 237). Additional names are found in poems by Manučehri Dāmḡāni (d. ca. 432/1041), and Anwari Abivardi (d. 565/1170), and in Neẓāmi Ganjavi’s Ḵosrow wa Širin, written ca. 581/1180. An enlarged system of two times seven modes is listed in the music chapter of Ašjār wa-aṯmār by ʿAlišāh Boḵāri (Storey, 1972, pp. 61-62), an astrological work written in 686/1287. Here the seven fundamental (aṣl) modes and their seven branches (farʿ) are both related to the seven planets. The author also lists a fundamental (aṣl) group of āvāz modes along with their higher (tiz) counterparts, and he mentions that tarkib modes are composed of the āvāz modes. He describes this system as if it were the leading one in his day while, at the same time, he highly praises the name of Ṣafi-al-Din Ormavi, who represented a quite different tradition.

In stages of development unknown to us, the older system of seven modes developed into a system of twelve main modes (parda), six secondary modes (āvāz), and additional šoʿbe or tarkib modes in sexagesimal order. It was also connected with astronomy. Several sources confirm that the previous system of seven modes, which had been a “planetary” one, changed into the more sophisticated “zodiacal” system of twelve modes (Anon., al-Moḵtaṣar al-mofid, p. 56). By the early 7th/13th century at the latest the new system had been completely developed. It was described by the North African writer Aḥmad b. Yusof Tifāši (d. 651/1253) as a recent Persian system adopted by the Arabs and containing twelve bardas (i.e., parda) and six āvāz modes whose names are mostly Persian (see Ṭanji, “al-Ṭarāʾeq,” pp. 96-97) but only partly identical with those known from Ormavi and his school.

Sets of twelve parda and six šoʿbe modes appear in the 7th/13th century in slightly different versions. There was an eastern or Khorasanian tradition transmitted by Moḥammad b. Maḥmud b. Moḥammad Neysāburi, the master of Khorasan (ostād-e Ḵorāsān, prior to 656/1258), who added the numbers of bāng to the main modes. There was another system of twelve modes called adwār (sing. dawr) or šodud (sing. šadd) plus six āvāz modes in a western or Irano-Arab version recorded by Ṣafi-al-Din Ormavi in the 13th century and, to judge from its terminology, intended to represent the predominant Persian and Arab local traditions, namely, ʿoššāq, navā, busalik, rāst, ʿerāq, eṣfahān, zirāfgand, bozorg, rāhavi, zangula, ḥosayni, ḥejāzi. In contrast, the beginning of Neysāburi’s series of pardas and bāngs (rāst, moḵālef-e rāst, māda, ʿerāq) resembles more closely that given by Kaykāvus b. Eskandar in his Qābus-nāma. Neysāburi emphasizes the importance of the mode rāst as being the “šāh of all the pardas” (ed. Purjawādi, p. 63). In other sources rāst is compared with a tree of which the other modes are the branches, or a town in which the other modes are the streets (Anon., al-Moḵtaṣar al-mofid, p. 56).

Comparison of different tables of modes from the 7th/13th and 8th/14th centuries reveals a relative similarity of the twelve main modes and a greater difference between the āvāz and šoʿbe modes which, however, was a frequent phenomenon in Persian and Arabic sources (cf. ms. Istanbul, Köprülü 1613, fol. 70b; Ebn Fażl-Allāh ʿOmari, Masālek al-abṣār X, p. 3, l. 10). Most unusual are the āvāz modes listed in the anonymous al-Moḵtaṣar al-mofid (ca. 755/1354). Their number is eight instead of six and includes some old-fashioned terms.

One generation after Ormavi an important account of the modal system was given by the versatile scientist Qoṭb-al-Din Širāzi (d. 711/1311) in the music chapter of his Dorrat al-tāj. In contrast to the other authors, he did not follow a strict system. He only once names the twelve parda modes as ʿoššāq, navā, busalik, rāst, nowruz, ʿerāq, eṣfahān, bozorg, zirāfgand, rāhavi, zangula, ḥosayni (Dorrat al-tāj, p. 124), and he mentions nine šoʿbe modes “used by the musicians” (dogāh, segāh, jahārgāh, banjgāh, zāvoli, ruy-e ʿerāq, mobarqaʿ, māya, šahnāz), the traditional system being of minor importance for his own distinction between and classification of the modes. He is the first to use the word maqām as a general term in the sense of mode (Dorrat al-tāj, pp. 122, 124), and he is the first to describe the melodic development (sayr) of modes, singling out characteristic notes such as the beginning (ebtedāʾ, mabdaʾ), a central pitch (wasaṭ), and the ending (entehāʾ, maḥaṭṭ, see Wright, The Modal System, pp. 143-292). At the end of the chapter on music, he has written down a song by Ormavi in the most sophisticated musical notation known from Islamic lands (see below).

Among his sources Qoṭb-al-Din quotes the Ketāb al-adwār by a certain Salmak (Dorrat al-tāj, p. 121). This was one of the books of the 6th/12th and 7th/13th centuries that have not survived but could have helped to detect the unknown predecessors and colleagues of Ormavi. Another book of this kind was written by a certain Šaraf-al-Din b. ʿAlāʾ, a distinguished contemporary of Ormavi in music theory, who is quoted in the anonymous Ketāb al-mizān fi ʿelm al-adwār wa-’l-awzān, written by a pupil of Ormavi.

Extra-musical phenomena. A specific aspect of the Persian modal system was its association to a number of extra-musical phenomena such as the stars, the seasons, or the hours of day and night. The musician was reminded to choose a song or a mode in correspondence with the nature and condition of his listeners, their age, complexion, status, and origin. The effect of music was treated in books by Abu Yusof Kendi, by his pupil Abu Zayd Balḵi (d. 322/934), by Ebn Hendu Nišāburi (d. ca. 420/1029), who was a student of Abu Zayd in the second generation, and by later writers. In the Persian language, this topic was first taken up in the Qābus-nāma (“the greatest art of the musician is to meet the nature of the listener”; Qābus-nāma, p. 237), and continued to be regarded as an essential aspect of musical practice down to the end of the Safavids. In the 7th/13th century at the latest, a pseudo-scientific relation between the twelve zodiacal signs, the twelve main modes, the twelve parts of the body, and the hours of day and night was made into a system of musical dietetics and music therapy that existed until the 8th/16th century (see Neubauer, “Arabische Anleitungen”).

Musical meters. A characteristic feature of both Arab and Persian musical practice in early Islam was the way (ṭariqa in Arabic, rāh in Persian) of a melody or a song. It was defined by a musical meter to which the melodic mode was subordinated. The musical meters (iqāʿ, oṣul al-iqāʿ)were described, on the Arab side, as being either heavy (ṯaqil) or light (ḵafif) versions of the three basic forms (oṣul) called ramal, al-ṯaqil al-awwal and al-ṯaqil al-ṯāni, and a separate group of light meters called hazaj. Fārābi brought these three layers into a system of basic beats in the relation of 1 : 2 : 4 or (in modern terms) eighth note, quarter note and half note. Ramal was the first of the heavy meters. It was described (in modern terms) as a sequence of two half notes and a half-note rest. The two remaining heavy meters consisted of three and four half notes respectively with again a half-note rest at the end. The three light meters had the same structure as their heavy counterparts, but double speed. The hazaj was described as a sequence of five quavers and a quaver rest. The final rest could be filled up by notes, and the notes of the basic patterns could be subdivided. Fārābi’s system was abstracted, with slight differences, from the patterns of the musicians as written down by Esḥāq Mawṣeli and others. According to Ebn Zayla (d. 440/1048), who was a pupil of Ebn Sinā, these were “all the meters used by musicians in Arabic, Persian, and Khorasanian

[songs]” (Kāfi, p. 62).

Abu Yusof Kendi had already described the Persian meters as “well-defined ways (ṭoroq maʿluma), similar to the [Arab] oṣuls” (šabiha be’l-oṣul; Kendi, Moʾallafāt, p. 137; Idem, Resāla fi ‘l-loḥun, p. 26). On their part, Persian writers saw their meters in a similar manner. In the Qābus-nāma (p. 232-33),ʿOnṣor-al-Maʿāli Kaykāvus first distinguishes between the two basic levels of heavy (gerān) and light (sabok), then proceeds to a threefold division. The first layer contains the “heavy” meters of the dastān-e ḵosravāni that is performed in the presence of kings, people of serious nature (ḵodāvandān-e jedd), and for old people. The second layer contains the “light meters” that correspond to light prosodic meters in songs (sorud) performed for younger people. The third layer is designated for the “delicate temper” (ṭabʿ-e laṭif) of women and children and contains the songs called tarāna whose meter (wazn) is lighter (laṭiftar) than those of others. Compared with Fārābi’s metrical system, the first layer can be assigned to the “heavy” meters, the second to the “light” versions of the “heavy” meters, and the third to the meter hazaj. In later Persian sources this threefold metrical system was reduced to cryptic formulations such as: “There are three kinds of beat (żarb), one beat (yak żarb) for old people, two beats (do żarb) for young people, and three beats (seh żarb) for women and children” (ms. Tehran, Majles-e Sanā 13682; ms. Istanbul, Köprülü 1613).

The metric system underwent a significant change between the 5th/11th and the 7th/13th centuries, which was comparable to the development of the melodic modes. At the end of this constitutive period, the meters had increased in length following the general principle of augmentation that governed the development of musical meters in Iran and, later, in Ottoman Turkey. Most of the meters of the 7th/13th century retained the older names while their form had considerably changed. In his Ketāb al-adwār, Ṣafi-al-Din Ormavi distinguishes between four different lengths of notes: letter alef or syllable ta is called sariʿ and represents the time unit of hazaj (=1), letter bāʾ or syllable tan is called sabab or ḵafif (= 2), letter jim or syllables tanan are called wated or ḵafif al-ṯaqil (= 3), letter dāl or syllables tananan are called fāṣela or ṯaqil (= 4). He also uses the mnemotechnic patterns of the root faʿala known from prosody. Accordingly, the basic form of al-ṯaqil al-awwal, described in Ketāb al-adwār as tanan (= 3), tanan (= 3), tananan (= 4), tan (= 2), tananan (= 4), is later recorded in al-Resāla al-šarafiya as mafāʿelon (3 + 3), faʿelon (4), moftaʿelon (2 + 4). One period of al-ṯaqil al-awwal (3 + 3 + 4 + 2 + 4 = 16) equals two periods of al-ṯaqil al-ṯāni (3 + 3 + 2 = 8) and four periods of ḵafif al-ṯaqil (2 + 1 + 1, or 1 + 1 + 2 = 4). A Persian meter was called fāḵeti. Ormavi gave it the pattern 4 + 2 + 4 + 4 + 2 + 4 = 20, Qoṭb-al-Din listed the form 2 + 4 + 4 + 2 + 4 + 4 = 20 and, under the name of fāḵetʾi-e zāyed, an extended form 2 + 4 + 4 + 4 + 2 + 4 + 4 + 4 = 28. Ormavi mentions that “very many melodies of the Iranians” were composed in the meter możāʿaf al-ramal (= ṯaqil al-ramal) in the version of al-Resāla al-šarafiya (= 4 + 2 + 4 + 2 + 4 + 2 + 4 + 2 = 24). Qoṭb-al-Din modifies this statement by saying that the meter had been favored by the Iranians in older days (dar qadim) and that the version recorded in Ketāb al-adwār (= 4 + 4 + 2 + 2 + 2 + 2 + 2 + 2 + 4 = 24) was more common than that of al-Resāla al-šarafiya (Ormavi, al-Resāla al-šarafiya, p. 208; Qoṭb-a-Din, Dorrat al-tāj, p. 137).

Eight meters are solely listed by Qobṭ-al-Din and referred to as being well-known in his time. Among them is moḵammas (2 + 2 + 4 = 8), żarb-e rāst or żarbe aṣl (2 + 2 + 2 + 2 + 4 = 12), and čahārżarb (8 + 4 + 4 + 4 + 4 = 24). In Šarḥ-e adwār (pp. 266 f.) Ebn Ḡaybi Marāḡi (d. 838/1435) adds that čahārżarb was “invented” by a certain Moḥammadšāh Rabābi. Qoṭb-al-Din (p. 138) also gives a description of the meter torki “in theory” (3 + 3 + 4 + 4 + 3 + 3 = 20) and “in practice” (2 + 2 + 2 1/3 + 2 1/3 + 2 + 2 = 12 [2/3]). This is the earliest known reference to a limping (aksak) meter in music theory. It was not repeated by Qoṭb-al-Din’s successors Ebn Ḡaybi Marāḡi and his son, who content themselves with the common theoretical description of torki and its variants (12, 20, and 24 beats), nor by later Ottoman writers.

The above meters were still current, with some variants and additional forms, in the theoretical writings of the Timurid period. Ebn Korr (d. 763/1362), who had left Baghdad for Cairo, uses the same terms while most of his patterns are different. In his book a variant of ṯaqil al-ramal (6 + 6 + 6 + 6 = 24) is called ḵorāsāni, and he lists a meter called żarbḵosravāni with eighteen time units (Ḡāyat al-maṭlub, fols. 8r-9v, 19r-v).

Musical forms. The development of musical forms in the early Islamic period was closely related to that of the modes and meters. Some of them were inherited from Sasanian times and survived in a more or less altered form, others developed under the new cultural conditions. Here also the mutual stimulation between the Arab and the Persian ʿErāq was the origin and the center of the Irano-Arab symbiosis. The fundamental couple of an instrumental introduction in free meter and the following metrical song was expressed by Keykāvus by the terms rāh and navā. As mentioned above, the sequence of dastān, sorud, and tarāna represented the gradation of heavy, medium, and light meters in the respective types of song. The musical performance of epic texts (dastān) in which the syllabic structure of the verse determined the metric structure of the melody was taken up later by professional reciters of the Šāh-nāma (Šāh-nāmaḵᵛān) and practioners of similar professions.

Bukhara was known for a musical style of its own, including amazing songs (sorudhā-ye ʿajib) about the mythical Kayanid prince Siāvoš, and traditional dirges transformed into songs of art music (Naršaḵi, Tāriḵ-e Boḵārā, pp. 23-24, tr. p. 23). Ebn Zayla distinguished between a Khorasanian style of dastān singing and another one from Eṣfahān and lists some (distorted) names of ravāšin and dastānāt. A verse of a dastān and of a ṭariqa could be followed by a special formula called naḡma (Kāfi, pp. 66-67).

Sorud was a more general term, comparable to the Arabic ṣawt. The early tarāna was a light song. Later it became a part of the nawba (see below). The name of another musical form, ḡazal, was borrowed from the poetical form bearing the same name. The musical ḡazal was equated in the Qābus-nāma (p. 235) with a ‘song without meter’ (tarāna bi-wazn). Later it was also included in the nawba<