Arab music.

1. Introduction.

Viewed historically, Arab art music needs to be related to a wider set of traditions in the Middle East embracing also the art musics of Persia/Iran and Turkey. Although these various traditions are now quite clearly distinct, their development has been marked to an unusual extent by periods of reciprocal influence and convergence, and they still have much in common. Furthermore, they have all suffered (if in varying degrees) from the ambiguities and tensions resulting from the generally hostile attitude of Islam to entertainment music and its practitioners (see Islamic religious music).

Within the Arab world, two major complexes may be distinguished, each with internal subdivisions: that of the eastern Arab world (principally Egypt, Lebanon, Syria and Iraq), and that of the western (principally Morocco, Algeria and Tunisia). Both may derive ultimately from the court music that evolved during the Umayyad and early Abbasid periods (7th to 9th centuries), but they probably began gradually to diverge not long thereafter. The western tradition, which is often considered by its modern practitioners to represent a direct descendant – or even a survival – of the Andalusian music of Moorish Spain, was certainly perceived to be distinct in certain respects from the eastern by the early 13th century, and it remained largely insulated from the effects of the process of interaction between Arab and Persian (and, subsequently, Turkish) elements that characterized developments in the eastern Arab world.

The precise contribution of each tradition is impossible to assess, but the eventual result was a musical lingua franca, tolerant of local variation but evolving in a fairly uniform fashion. From the 13th century (if not before) to the 17th it was propagated and appreciated at cultural centres stretching across a vast area from Egypt through the Fertile Crescent to Anatolia, Persia and parts of Central Asia. At different stages in its development this composite idiom was patronized at, for example, the courts of the last Abbasids in 13th-century Baghdad, the Jalairids in 14th-century Tabriz, the Timurids in 15th-century Samarkand and Herat, the Ottomans in 16th-century Istanbul and the Moghuls in 16th-century Delhi. It is only from the 17th century that we encounter increasing evidence of regional differentiation.

For the most part patrons and practitioners of Arab art music have been concentrated in major cities, and there are large areas within the Arab world where it has made little impact. Beyond the boundaries of effective influence lie Mauritania, with its own separate griot tradition; Sudan, whose local traditions include strong Nilotic African elements; and Arabia, which has been peripheral to the development of art music during the last millennium, but nevertheless serves as the inevitable starting-point for any survey of its historical development.

(i) The pre-Islamic period (before 622).

2. The early period (to 900 ce).

(ii) The early Islamic period and the Umayyad caliphate (622–750).

(iii) The early Abbasids and Baghdad (750–900).

(iv) Early theory.

Arab music, §I, 2: Art music: The early period (to 900)

(i) The pre-Islamic period (before 622).

However fragmentary, the earliest available sources for West Asia and the eastern Mediterranean offer a tantalizing variety of materials: textual references (e.g. Old Testament), iconography (e.g. ancient Egyptian representations of instruments, musicians and dancers), archaeological evidence (e.g. surviving Mesopotamian instruments) and fragments of notation (e.g. Ugaritic and Greek). But Arabia has yielded no such material remains, and even if Arabic contains a few instrument names for which there are cognates in other Semitic languages, such evidence indicates only that certain generic instrument types were widely diffused at an early date.

The fertile and relatively densely populated south-west of Arabia yields an extensive body of inscriptions that reveals aspects of the society and culture that flourished there from the 1st millennium bce to the 6th century ce. But music, unfortunately, is not among them; and no comparable record can be found in the arid and inhospitable centre and north, thinly populated by nomadic pastoralists and settled oasis-dwelling agriculturalists. Yet the Arabs of this area were not culturally isolated: links both north and south were reinforced by population movements, kinship and by the maintenance of important trade routes. Contacts were cultivated not only with the south but also with the northern buffer kingdoms of al-Hīra and Ghassān through which, respectively, Persian and Byzantine influences could percolate.

Nevertheless, it is not until the 6th century ce that we can discern anything about Arab musical practices, and even then there is no contemporary evidence. The earliest testimony is exclusively textual and consists of a few incidental references in poetry, recorded later but reflecting pre-Islamic Bedouin values and mores, and of more substantial prose writings of no earlier than the 9th century which, given their date, need to be treated with considerable caution as repositories of information for the pre-Islamic period. Reflected in them is a general concern with origins: the origin of song is credited to the Old Testament ancestor Jubal, while the inventor of the most important of all art music instruments, the ‘ūd (short-necked lute), is identified as Lamak (Lamech), who took as his model body parts of the dead son he mourned.

Early textual sources also supply brief and not very revealing references to what are supposed to be the earliest genres of music. The most ancient, for which is proposed the rather more prosaic origin of a cry of pain, is claimed to be the hudā’ (caravan or cameleer’s song), and together with the nasb, of which no description is given, it was associated with male performers. The nasb is subsumed under the wider heading of ghinā’ (‘song’, later sometimes used generically for musical practice), which is clearly differentiated from the hudā’ (although, according to one account, derived from it).

Of the two other subcategories of ghinā’ mentioned, one is characterized as ‘heavy’ and ornate, the other as ‘light’ and gay. These two types were probably normally performed by the Qayna (‘singing slave-girl’), with whom may be associated the emergence of entertainment music in the nascent urban centres of the Hijaz, possibly under the influence of the court fashions of al-Hīra and Ghassān. In early poetry the qayna is sometimes depicted playing an instrument, and instrumental accompaniment to song may have provided a further contrast with the performing practice norms of the hudā’. Women also performed the nawh (funeral lament), although, again, there is no description of this genre. Its apparently separate status possibly reflected its special function and ritual context. The contrast between hudā’ and ghinā’ might likewise encode a distinction between the categories of work-song and entertainment music. On the other hand, the ‘heavy’–‘light’ opposition within the latter may be a projection back of a later stylistic discrimination: it can hardly be expected that the spasmodic transmission of information about an area of cultural expression that had undergone dramatic changes during the intervening centuries would have ensured an accurate record of pre-Islamic distinctions.

Unfortunately, very little can be added from other sources to supplement the meagre textual record. A degree of familiarity with certain forms of liturgical chant may be inferred from contact with scattered Jewish and Christian communities. But our knowledge of pagan pre-Islamic religious practices is scanty, and in the absence of iconographical evidence only the most obvious conclusions can be drawn, for example that chant and possibly also dance were probably integral to ritual occasions and worship at certain pagan shrines. The extent of overlap between chant and the other genres is impossible to assess, but the range of instrument names mentioned in relation to entertainment music is at least compatible with the equally obvious supposition that it was in this context that the highest levels of technical proficiency were required. The instruments played by the qayna included the frame drum (duff), the lute (kirān, muwattar; presumably types with a skin table) and reed instruments (mizmār). Other terms found in poetry – not all of which can be indentified with confidence – include the mi‘zafa (?psaltery), mizhar (lute or frame drum), qussāba (end-blown flute), jalājil (bells), sanj (cymbals) and jank (harp).

Arab music, §I, 2: Art music: The early period (to 900)

(ii) The early Islamic period and the Umayyad caliphate (622–750).

By 650, less than 20 years after the death of the Prophet Muhammad, Muslim armies had swept aside Byzantine resistance in Syria and Palestine, conquered Egypt and destroyed Sassanian power, seizing Iraq and pressing on into the heartlands of Persia itself. The dramatic transformation of the political map soon instigated (or accelerated) musical developments.

These represent another major theme in 9th-century writings, which acknowledge change and, where the source was thought to be external, identify it. In fact, innovation was stressed, new techniques or practices being associated with particular named individuals; and whether or not the particular ascriptions are well founded the general emphasis is surely justified, both in relation to the introduction of novel elements and to changes in the social role and identity of the musician.

Early Persian influence is associated with Nashīt, who is said to have gained popularity in Medina with his Persian songs, but in order to maintain his reputation was obliged to add to them a repertory in Arabic. The same conjunction of change and cultural symbiosis is developed further and more subtly in the portrayal of two of the most famous singers of the next generation, Ibn Misjah and Ibn Muhriz (both d c715). They are described, albeit in suspiciously similar terms, as having travelled through former Byzantine and Persian territories, adopting the more readily acceptable features of the music they encountered. Such accounts ring true, at least in so far as they point to a period of transition during which the emerging art music idiom was characterized by the capacity to assimilate new elements without compromising what were perceived to be the essentials of the Arab tradition that lay at its core. At the same time, they imply that the differences between the Byzantine, Persian and Arab traditions were not so radical as to impede borrowings. (Indeed, a degree of familiarity with aspects of Byzantine and Persian practice may already have developed in the pre-Islamic buffer states of al-Hīra and Ghassān.)

With the establishment of the Umayyad dynasty in 661, the political centre of power shifted to Syria. Innovation in music, however, continued to be associated with the Hijaz, and especially the cities of Mecca and Medina. Here, patterns of social behaviour were affected by the new level of affluence, which had as one consequence an increased demand for (and appreciation of) entertainment music. Previously the preserve of the female entertainer (qayna), this gradually came to be dominated by male performers, being associated in the first stages of transfer with the effeminate mukhannaths, some of whom were as renowned for their immoral behaviour as for their artistic accomplishments.

For the pious, entertainment music could hardly fail to be tainted by such links. Reports identifying certain qaynas as conveyors in song of opposition to the Prophet articulate a negative image that would have been reinforced (if authentic) or provoked (if not) by later associations of the qayna and the mukhannath with increasingly unacceptable patterns of behaviour involving frivolity, licentiousness and wine-drinking. Ultimately, however, it was a social rather than a moral code that made music an unsuitable profession for a free-born Arab, and the important male singers who dominated the scene towards the end of the 7th century, supplanting the mukhannaths, were still generally non-Arabs of low social status. There is, however, another side to the coin: even if, again, the possibility of historiographical refraction cannot be excluded, it should be noted that certain female performers of the 7th century, notably ‘Azza al-Maylā’ and Jamīla, were celebrated not only for their musical achievements but also for the propriety that characterized both their performances and the behaviour expected of their audiences.

Little is known about the technical qualities of the musicians. Among the innovations mentioned are cases of musical diffusion, as with the introduction of a Persian type of short-necked lute with a wooden table (‘ūd fārisī); aspects of performing practice, as in the shift towards a norm of singers accompanying themselves on a melody instrument (usually a lute); and technical features, such as the introduction of a particular rhythmic cycle.

The major source for this period, the monumental Kitāb al-aghānī (‘Book of songs’) by al-Isfahānī (897–967), provides evidence of the evolution of an increasingly precise theoretical vocabulary. Song texts are accompanied by an identification of the rhythmic cycle and sometimes also the melodic mode. But no definitions are offered, and despite the evidence of a keen interest in assessing the characteristics and relative merits of various singers, the judgements made are generally lapidary.

The most extensive catalogue of discriminations comes in the form of a succinct list of the qualities needed in a good singer, which is ascribed to another major Umayyad figure, Ibn Surayj (c637–726). It begins, instructively, with accuracy of diction and grammar and then proceeds to binary oppositions that suggest a precise conceptualization both of vocal control and of a distinction between fundamental rhythmic and melodic structures and ways in which they can be embellished. Important to the singer-composer was not only rhythmic and formal fidelity to the text but also the ability to select from a wide range of poetry. Contemporary verse was frequently drawn upon, and several Umayyad singers had close contacts with poets: they may, indeed, have contributed to the creation of the new sensibility that Hijazi poetry reveals. There is no convincing evidence, however, to suggest that musicians influenced poets in the specific sense that certain rhythmic cycles or patterns inspired innovations in poetic metre.

Arab music, §I, 2: Art music: The early period (to 900)

(iii) The early Abbasids and Baghdad (750–900).

Many of the anecdotes making up the biographical notices of great Umayyad musicians revolve around journeys, not just within the Hijaz but far beyond. Several were invited to Syria to perform before caliphs, but they do not appear to have been permanently attached to their courts, so that even if we can reasonably speak of an increasingly distinct form of art music (recognized in the phrase ghinā’ mutqan: ‘perfected song’) it would be only partially true to call it a court tradition. But with the defeat of the Umayyads by the Abbasids in 750 and the subsequent founding of a new capital at Baghdad the pattern changed: prominent musicians were closely connected to the caliphal court, and the sources – again principally al-Isfahānī’s Kitāb al-aghānī – concentrate overwhelmingly upon persons and events in the imperial capital. This perspective reflects, in addition to the crucial role of the caliph and other notables as patrons, the wider importance of Baghdad as the preeminent centre of an increasingly mature and self-confident culture. The early Abbasid period was one of immense activity, a coming of age when, as a result of rapid development involving both the acquisition of new knowledge through translation (e.g. in science and philosophy) and the elaboration of indigenous systems (e.g. in law and grammar), Islam established itself fully as a major intellectual as well as religious and political enterprise.

Something of this vitality and variety is evident in the surviving 9th-century textual corpus dealing with music. It contains historical and biographical material characterized by wide-ranging curiosity, psychological analysis and the elaboration of theory, and the earliest surviving treatise debating the legal position of music. This last, by Ibn Abī al-Dunyā (823–94), is unremittingly hostile, classing music not only with idle pastimes such as chess but also with unacceptable sexual practices. The link with sexuality was explored further, and far more subtly, by al-Jāhiz (d 868–9) in his Risālat al-qiyān (‘Epistle on singing slave girls’), which also illustrates the cultural sophistication of the qayna and the profits to be made from producing such a highly desirable commodity. A similar sophistication was expected from her teachers, prominent among whom was the eminent musician Ibrāhīm al-Mawsilī (742–804). But it was his son, Ishāq al-Mawsilī (767–850), who was the supreme example of the cultivated musician-courtier: the foremost composer and performer of his day, he was also an accomplished poet and an expert in philology and jurisprudence.

In addition to information about the careers of such musicians, the events portrayed in Al-Isfahānī’s Kitāb al-aghānī give us some insight into the relationship between patron and singer and the norms of performing practice. Often performers would be summoned at short notice to appear at court and to take part, usually with others, in a majlis (‘assembly’). More prominent musicians may also have been assigned their nawba (‘turn’), a particular day of the week on which to perform. But this term does not imply the existence of the cyclical form it later came to designate: at this stage there is no hint of songs being organized into sequences governed by modal, rhythmic or other structural criteria. In any case, a patron would frequently desire to hear a song repeated several times over, to the exclusion of others. Songs needed to fit a particular mood or occasion, but appropriateness was primarily textual; identity of mode or rhythmic cycle does not appear to have been a determining factor. The patron would also often demand a particular song or an extempore setting of a specific text, so the best performers needed to possess both an extensive repertory and the ability to improvise. Their reactions to such tests of expertise are the subject of several anecdotes, especially where competition was involved, with rivalry sharpened by the considerable material rewards that might be lavished on the successful. These stories thus offer comparative evaluations of style and technical competence.

Despite one or two intriguing references to the use of some form of notation, transmission was essentially oral, a song being repeated until mastered. However, the effectiveness of some musicians as teachers was reduced by their tendency to vary each repetition, while others deliberately practised variation as a means of preventing piracy.

But the extent to which variation might be either cultivated or avoided was also coloured by attitudes to tradition, and in parallel with the literary debate on the respective merits of the ancients and moderns, we find advocates of faithful musical transmission opposed to innovators. Chief among the latter was Ishāq al-Mawsilī’s great rival, the princely amateur Ibrāhīm ibn al-Mahdī (779–839). Renowned for the quality and reputed four-octave range of his voice, he was portrayed as a champion of greater freedom of expression. The innovations espoused appear to have involved a further injection of Persian elements, but exactly what these might have been is by no means clear, for again we encounter curt indications of stylistic contrast rather than analysis. When used in relation to Umayyad musicians, the distinction between ‘heavy’ (thaqīl) and ‘light’ (khafīf) appears to have implied a contrast between a more complex and serious style and a simpler, gayer one, the former commanding more prestige, the latter greater popularity. In its Abbasid manifestation, however, it appears that the lighter, more Persianate style involved an association of freedom of interpretation with greater melodic elaboration, in contrast to the sobriety of the traditionalists. But whatever its nature further Persian influence was only to be expected: the Abbasid revolt had begun in Khorasan, and the shift of the capital eastwards from Syria to Iraq symbolized a new cultural as well as political balance, in which the Persian contribution would became more prominent. No trace, however, seems to have remained of the two different types of ensemble portrayed on the Sassanian reliefs (c600 ce) at Tāq-i Bustān, one an outdoor wind and percussion band (including, as a striking early example of the diffusion of instruments across the breadth of Asia, sheng-type mouth organs), the other a court ensemble dominated by harps (see Iran, §I).

Although occasional references are made to a number of musicians singing or playing together (in unison), the norm for Abbasid court performances remained the solo singer, accompanying him- or herself, or with another accompanist. Despite the often-emphasized importance of rhythmic accuracy, accompaniment on percussion instruments appears to have been the exception rather than the rule. With the decline of the earlier use by some singers of a qadīb (percussion stick), it is the ubiquitous duff (frame drum) that tends to be mentioned, while the sporadic references to the tabl possibly refer in this context to a double-headed waisted drum. There are also occasional references to aerophones, nāy and mizmār, although it is not certain that at this period these terms are always to be identified with, respectively, the end-blown flute and the shawm: they may refer to single-reed instruments as well. But among melodic instruments plucked chordophones generally predominated, even if the open-stringed harp and psaltery remained marginal. By far the most common instrument was the ‘ūd (which, in the absence of percussion, may have played a rhythmic as well as melodic role in accompaniment). The only serious rival to its supremacy was the increasingly popular tunbūr, a long-necked lute, which was indentified particularly with certain Persian provinces and on which a number of qaynas excelled.

Arab music, §I, 2: Art music: The early period (to 900)

(iv) Early theory.

While Persian influences seem to have had a significant impact on performance context and aspects of practice, references by Arab writers to the seven ‘royal’ Persian modes may merely reflect an antiquarian interest in Sassanian culture (see Iran, §I, 5). The organization, or at least the classification, of the modal system of Umayyad and early Abbasid music seems, rather, to have been influenced by the recently formulated Byzantine Oktōēchos. Some features of the Arab system, which likewise consisted of eight modes, are described by al-Munajjim (856–912), who discussed them in terms of the diatonic fretting, to which their names relate, on the two upper strings of the ‘ūd. Assuming a tuning in perfect 4ths, the fretting yields a series of intervals consisting of the Pythagorean whole tone (T) of 204 cents, the limma (L) of 90 cents and (by subtraction) the apotome (A) of 114 cents (Table 1).

table 1
Strings	Frets
	G	c	f	b	mutlaq	(open string)
T
T				g	c'		sabbāa	(first finger)
L							sabbāa
L				a	d'	wusịā	(second finger)
A							(second finger)
A				a		d'		binsir	(third finger)
L									(third finger)
L					b	e'	khinsir	(fourth finger)

The eight modes are divided into two sets (termed majrā: ‘course’) of four, distinguished by the occurrence in one set of the note produced by the second-finger fret, and in the other of that produced by the third-finger fret, these two being mutually exclusive. Each string thus supplies two pitch sets: 1 2 3 4 and 1 2 3 4. The names of the modes (e.g. mutlaq fī majrā al-binsir: ‘open string in the course of the third finger’) are a descriptive shorthand, specification of the set being preceded by a term that presumably refers to a note of particular importance.

Although applied to some of the Umayyad repertory, the final formulation of this nomenclature may be rather later. It has been attributed to Ishāq al-Mawsilī, who was al-Munajjim’s principal authority; it is certainly the case that where Ishāq is not the source for the technical description of a song only the classification according to set may be given. During the Umayyad period, therefore, the internal differentiation between the members of each set was possibly not very clear-cut. Even when the boundaries became more securely established, the fact that only eight modes were recognized, as against the far higher numbers recorded later, suggests that at this period the concept of mode was probably located nearer to the pitch set rather than the specific melody end of the spectrum. Modal distinctions could have served initially as a classificatory tool for segmenting a pre-existing repertory and may not yet have evolved into complexes of guidelines enabling and governing the production of new songs. On the other hand, accounts of virtually extempore composition suggest the extensive use of melodic formulae; and whereas seven diatonic modes can be distinguished on the basis of intervallic structure alone, eight cannot. It is plausible, therefore, to conjecture the presence of criteria of differentiation (e.g. authentic versus plagal, position of prominent notes, identity of finalis) or a contrastive emphasis on ascending versus descending melodic movement.

Al-Munajjim’s neat 2 x 4 scheme probably also tidies up a more complex reality. One evident anomaly is that it takes no account of the neutral 3rd fret said to have been introduced by Zalzal (d after 842), the ‘ūd teacher of Ishāq al-Mawsilī himself, and named after him (wustā zalzal) (see Table 4 below). Even if it was irrelevant to the codification of the Umayyad repertory one would not expect a precise description of 9th-century songs to ignore it. The adoption of Ishāq al-Mawsilī’s terminology throughout the Kitāb al-aghānī also serves to conceal the existence of another set of mode names – one that later resurfaced and, in fact, proved more durable. But of greater significance is the fact that as a means of identification (or at least labelling), specifying the melodic mode of a song seems to have been less crucial than indicating the rhythmic cycle. In addition, there is some evidence that for musicians a melodic mode was not readily conceived of in abstraction from the particular rhythmic structures through which it was manifested. Nevertheless, as early as the Umayyad period analytical categories were clearly being developed, whether separately or together, for both melodic mode and rhythmic cycle, and on the modal side the formulations of Ishāq al-Mawsilī represent the culmination of this process.

Contemporary with his final translations, however, was a new development: the emergence, during the 19th century, of a theoretical literature incorporating ideas disseminated through translation and derived in the main from Greek sources. Of particular importance are several short treatises of the philosopher al-Kindī (c801–c866), the first major theorist whose works are extant. Eclectic in approach, they adumbrate two of the most significant areas of concern to the later theoretical literature: cosmological speculation and the analysis of intervals and scales.

Al-Kindī’s cosmological interest focussed upon numerical correlations. He ordered chordophones according to the number of their strings, giving geographical and ethnic as well as cosmological associations for each instrument. He also initiated what became a major strand in later speculation by relating the (normal) four strings of the ‘ūd to, amongst others, the elements, seasons, points of the compass and the humours. He introduced, further, the frequently recurring notion of appropriate correspondences (e.g. relating a certain structural feature of music to the time of day supposedly best suited to it), although there is no evidence that this was of any relevance to contemporary practice.

Al-Kindī also provided a model that was followed by the later theoretical tradition in discussing intervals and scales, for even when Greek-derived concepts are used his form of presentation involved projecting scale structures onto the fingerboard of the ‘ūd. To the basic diatonic tuning, in terms of which the pitch sets of the early modal system were defined (Table 1), he added a notional 5th string, which extended the ‘ūd’s theoretical range to the two octaves of the ancient Greek Greater Perfect System (see Greece, §§I, 6(iii)(a) and (d)), and a hypothetical fret supplying certain octave equivalences (and other notes lacking them; see Table 2).

It has been argued that the diatonic scale structures al-Kindī presented are relatable to the contemporary system of eight modes, even if he himself did not make the connection. He certainly referred to actual practice when noting possible variations in the tuning of the lowest string, an aspect ignored by many later theorists, who also frequently failed to follow his example by disregarding the physical reality of instruments, thus making his description of the dimensions and construction of the ‘ūd all the more valuable.

Al-Kindī was also the first theorist to illustrate (even if only fitfully) rhythm or, more precisely, the internal articulation of the various rhythmic cycles (īqā‘āt), which constituted the primary system of classification for the Umayyad repertory. The inclusion of the terms thaqīl (‘heavy’) and khafīf (‘light’) in the names of certain cycles suggests not only that they may have been divided, like the melodic modes, into two sets, but also that the identity of the cycle was possibly important in coding stylistic differences. The pioneering definitions supplied by al-Kindī, which presumably relate to early Abbasid practice, exhibit none of the later terminological indebtedness to prosody and are somewhat approximate, so that the version given in Table 3, showing relative durations only, must be considered tentative.

(i) Political fragmentation.

3. The later Abbasids (900–1258).

(ii) Theory: al-Fārābī to Ibn Sīnā.

(iii) Music and religion.

(iv) The Arabs and Europe.

Arab music, §I, 3: Art music: The later Abbasids (900–1258)

(i) Political fragmentation.

Towards the end of the 9th century the power of the Abbasid caliphs was much reduced. Unable to prevent the establishment of virtually independent dynasties in various provinces, they increasingly became puppets in the hands of their Turkish generals. Political fragmentation and cultural efflorescence, however, often go hand in hand. The break-up of the empire probably accelerated the diffusion of artistic norms established in Baghdad to the regional centres vying with it. Many of these centres lay outside the Arabic-speaking heartlands, and although those writers who had produced the early texts on music were mostly Arabs, the greatest theorists from the 10th century to the 13th were of Persian and Turkish origin. During this period, therefore, and especially with regard to the practitioners and theorists of the eastern tradition, the label ‘Arab’ becomes increasingly restrictive and misleading, even if Arabic remained the dominant language of religion, culture and science and was the preferred vehicle for articulating aspects of what was evolving into a more generalized Near and Middle Eastern court music tradition.

Abbasid control in the far western Arab world was equally tenuous, and Spain, which had been conquered under the Umayyads, served as a refuge for a surviving member of the Umayyad house who succeeded in founding a dynasty (756–1031) there. Although Córdoba, its capital, soon rivalled Baghdad in prosperity, the eastern Arab world was still regarded as culturally superior, and it is against this background that the mostly late and largely fragmentary materials on the early history of music in Muslim Spain need to be judged. They provide a narrative of beginnings according to which the legacy of the first important singers – qaynas coming or returning from the Hijaz – was eclipsed by that of Ziryāb (d 850), who arrived at the Umayyad court in Córdoba in 822.

Ziryāb is portrayed as a supremely dominant cultural figure, not only as a performer and teacher but also as an arbiter of taste, setting standards in, for example, fashion and cuisine. But to dismiss some of the material surrounding him as apocryphal is not to deny his historical importance; rather, it is to ask where his significance truly lies, and as a cultural icon he symbolized, ultimately, parity between Córdoba and its erstwhile model Baghdad. On a more mundane level, even if the details of his reported changes to the ‘ūd are unreliable, he was possibly associated with improvements in both the construction and playing technique of the instrument. It is also tempting to ascribe his fame as a teacher not just to novel methods of vocal training, but also to the wider access he is reported as promoting.

In the century following Ziryāb's period of influence at Córdoba, North Africa became the testing ground for the Fatimid cause, a sectarian movement that rejected the legitimacy of the Abbasid caliphs, and was to pose one of the most serious threats to their position. Advancing eastwards from Tunisia, the Fatimids seized Egypt in 969, creating a new capital city, Cairo. For the greater part of the next two centuries they also held much of Palestine/Syria, until they were finally extinguished by Saladin in 1171.

Information on musical patronage in Fatimid Egypt is scanty, despite the substantial evidence of mercantile activity, a flourishing economy, and the large-scale production of luxury goods. Although the personal attitudes of rulers could have a significant effect on the standing of music at court, we should not conclude from this paucity of evidence that there was any marked diminution of patronage. The phrase ‘golden age’, used in relation to music in the early Abbasid era, reflects not so much a pinnacle of status or quality of performance and invention but rather a comparative wealth of documentation: thereafter the picture simply becomes much sketchier. The continuing importance of art music as integral to the opulent court life during the Fatimid period is certainly attested, but historical sources refer more frequently to spectacular public ceremonies and the wind and percussion ensembles that accompanied them.

Only one text that is specifically about music survives from the Fatimid period. By Ibn al-Tahhān (d after 1057), a court musician and teacher, it is rather unrevealing as a historical source: it has more to say about Ishāq al-Mawsilī than the author's own contemporaries, and the lists of musicians active in Fatimid Egypt and Syria, while useful, provide nothing beyond their names.

Arab music, §I, 3: Art music: The later Abbasids (900–1258)

(ii) Theory: al-Fārābī to Ibn Sīnā.

Despite the fascination of al-Kindī’s wide-ranging and exploratory treatises, it is the works of the great philosophers who succeeded him that contain the most sustained and elaborate theoretical analyses. In fact, nothing in the preceding literature anticipates the scope and intellectual rigour of the Kitāb al-mūsīqī al-kabīr (‘Great book on music’) of al-Fārābī (d 950).

The introduction, which contains speculations on the origins of music and the nature of musical talent, sets an Aristotelian tone and is important for its general methodology. But despite its recognition of the priority of practice, the main body of the work is determinedly theoretical and much of its material, whether on intervals and their combinations or the elaboration of rhythmic and melodic structures, is at some remove from contemporary realities. Thus al-Fārābī went beyond al-Kindī’s adoption of the framework of the Greater Perfect System to codify in detail the various tetrachord types that can be combined within it, providing a numerical analysis of their constituent intervals. Since many of these tetrachords were not in current use, the amount of attention paid to them points to a new concentration on theory for its own sake – a development of the purely speculative side of music viewed as one of the mathematical sciences (the Western medieval Quadrivium).

Compared with al-Kindī’s diatonic fretting, al-Fārābī’s seems extraordinarily complex (Table 4). But the complexity is not gratuitous: it results from the superimposition of different analytical strands. Thus to the diatonic values inherited from al-Kindī are added (at 98 and 303 cents) variant definitions of the semitone and minor 3rd, arrived at not by ratios but by an empirical technique of halving the distance between other frets. Of the remainder, 142 and 168 are again alternative approximations to a value one whole tone below the neutral 3rd at 354. The introduction of this last, the wustā zalzal (‘Zalzal’s second-finger fret’), is a clear reflection of current practice, and on a single string (e.g. tuned to c), the fretting as a whole corresponds to the following set of functional discriminations (the symbol b representing a value roughly halfway between and ): c (d, db), d, e, eb, e, f. The hesitation over the values to be assigned to d and db suggests that at this period performers were only just beginning to explore the area between open string and first-finger fret.

Reference is again made to the modal distinction provided by the mutual exclusivity of the notes produced by the second and third fingers. However, aside from the suggestion that on any given string the pitch set 1 2 3b 4 now enjoyed equal status with 1 2 3 4 and 1 2 3 4, little insight is afforded into the development of the modal system since its codification by Ishāq al-Mawsilī a century earlier.

Matters are further complicated by the fact that the other main melody instruments are each described as having their own distinctive scales, even if, in most cases, these can be reconciled with structures derived from the ‘ūd scale norm. Those that differ most markedly are associated with two varieties of long-necked lute, the tunbūr baghdādī and tunbūr khurāsānī (of Baghdad and Khorasan). The former is said originally to have had a scale proceeding by approximate quarter-tone steps over a range of little more than a minor 3rd, while the scale of the latter, which may in reality have included neutral intervals, is analysed in terms of limma and comma (comma being the 24 cents difference between apotome and limma). Whereas al-Fārābī provided a composite fretting for the ‘ūd, he was content simply to juxtapose the varied scalar resources of the other instruments; and from this we might infer the coexistence of various regional traditions. But al-Fārābī is silent on this subject, as on practical organological matters. Indeed, his initial classification of instruments identifies only chordophones and aerophones. However, his recognition of the category of bowed strings is of particular historical significance, his account of the rabāb being the earliest attestation of a bowed instrument.

Also highly theoretical is the treatment of rhythm in the Kitāb al-mūsīqī al-kabīr. This makes an initial distinction between conjunct and disjunct cycles, the former equating essentially with a pulse set at various tempi. The latter, on the other hand, incorporate both variables and the concept of a disjunction relating to the analysis of a cycle as ending with the last marked time unit rather than with the one preceding the onset of the next cycle. The result is a generative framework, analogous in some respects to the variable schema of South Indian theory, from which a vast number of possible cycles can be derived (see India, §III, 4(iii)).

Al-Fārābī turned more directly to existing cycles in two later works. His treatment suggests that variability was a marked feature of contemporary practice, involving not only the predictable addition of supplementary and timbrally differentiated percussions within a standard grid of marked time units, or the omission from certain time units of the percussions normally associated with them, but also, and far more radically, the repetition or suppression of segments of the cycle, thereby altering the total number of time units. This suggests that the early categorization of rhythm could represent a sophisticated segmentation of spectra perceived as having common core elements, despite the markedly different surface structures produced by various transformational processes. However, although later writers may also have acknowledged the existence of variants, most of them reverted to the more straightforward type of definition offered by al-Kindī, having no apparent difficulty in deciding on the most representative form of any given cycle.

Al-Fārābī also treated melodic typology schematically, but not in relation to the general binary conceptualization of fundamental structure plus embellishment. Rather, he considered two aspects, on the one hand distinguishing between melismatic and syllabic styles of text setting, and on the other establishing various abstract series of pitch changes that offer an analytical parallel to the elaboration of possible rhythmic structures.

Somewhat closer to practice is his treatment of vocal quality. This aspect was further developed in treatises by al-Hasan ibn Ahmad (late 10th/early 11th century) and Ibn al-Tahhān. The former, even if unoriginal and somewhat confused in his discussion of abstract topics, had much to say about such practical issues as appropriate behaviour. He also touched upon the interaction of performer and audience, and marshalled an extensive range of descriptive and evaluative terms relating to features of both vocal and instrumental technique. His treatise, therefore, points to a considerable subtlety of aesthetic discrimination among discerning listeners. His treatment of mode confirms the independence (on any given string) of the pitch sets 1 2 3 4, 1 2 3b 4 and 1 2 3 4, but it also suggests that other neutral intervals functioned as variants of diatonic ones and did not yet serve to identify discrete structures.

Ibn al-Tahhān covered much of the same ground, concentrating particularly on the character, technique and care of the voice (including diet). He followed al-Kindī in providing quite a full account of the dimensions and construction of the ‘ūd and in listing various other instruments, several of which are stated to be specific to the Byzantine Greeks (rūm). For the rhythmic cycles, he gave a symmetrical set of four ‘heavy’ and ‘light’ pairs, adding that each has two fourfold subdivisions. The vocabulary for these, however, is modal, which suggests that for practising musicians the parameters of pitch (mode) and rhythm (cycle) were still conceptually wedded. That the perception of a modal outline was habitually projected onto a particular rhythmic organization is also suggested by a Judaeo-Arabic fragment of this period (from the Cairo Geniza, a store in the synagogue), which contains an elementary ‘ūd exercise. Durations are not given in this fragment, but a particular rhythmic cycle is identified and the text spells out the physical pattern of plectrum strokes, as in ex.1, a representation of the thānī thaqīl cycle in its mutlaq modal configuration.

The mode names in the Geniza fragment are the same as those of Ibn al-Tahhān and al-Hasan ibn Ahmad. They provide evidence for the survival in the eastern Arab world, at least until the early 11th century, of the terminology that continued in use thereafter in Spain. In the light of contemporary evidence for the emergence of a new modal nomenclature in Persia, it would thus appear that at this very early stage in the differentiation between the western and eastern art music traditions, the dividing line could well have corresponded to the geographical, ethnic and linguistic demarcations between Arabs and Persians.

The new Persian terminology first appeared in the chapter on music within the section on the mathematical sciences in the Kitāb al-shifā’, an encyclopedia by another great philosopher, Ibn Sīnā (d 1037) – known in Europe as Avicenna. Inevitably more compressed than that of al-Fārābī, his treatment of the subject is nevertheless comparable in scope, and its organization is in some respects more logical. The introduction considers musical sound a means of aesthetic expression ultimately derived from a signalling device helping to preserve the species. It is followed by an exhaustive analysis of intervals and tetrachordal combinations thereof, then by a section on rhythm that models itself on the work of al-Fārābī in its application to a given base of variables that generate cycles of differing lengths. When dealing with instruments, Ibn Sīnā refined al-Fārābī’s criteria in classifying both chordophones and aerophones, noting, for example, the presence or absence of a reed, but he followed his predecessor in ignoring membranophones.

Ibn Sīnā's chapter on music concludes with an ‘ūd fretting followed by outline sketches of a number of common modes, and it is here that the new terminology appears. Noteworthy is the first mention of one of the principal Middle Eastern modes, rāst (‘straight’; mustaqīm in Ibn Sīnā’s Arabic translation), here defined as a series of conjunct 1 2 3b 4 tetrachords. Despite the inclusion of some basic diatonic structures that presumably survive from the earlier system, the prevailing impression conveyed by Ibn Sīnā is of a somewhat heterogeneous stage of transition which is still some distance from the more complex system codified by theorists from the 13th century onwards.

The Neo-Platonic tradition was largely ignored by al-Fārābī and expressly rejected by Ibn Sīnā. Nevertheless it proved just as important as the Aristotelian, and cosmology and numerology reappear in another encyclopedic work, the Rasā’il (‘Epistles’) of the 10th-century Ikhwān al-Safā’ (‘Brethren of Purity’), a group of scholars based in Basra. This covers much the same ground as al-Kindī's writings, touching on various aspects of practice and also propounding the theory of the spherical propagation of sound. Its most striking feature, however, is an overarching concept of cosmic harmony and balance. The earlier sets of associations for the four strings of the ‘ūd are further elaborated; Platonic numerical relationships invoke the music of the spheres; and a claim is made for the medical and moral efficacy of music. The cosmological views of the Ikhwān al-Safā’ remained significant in the culture for many centuries and were echoed in several subsequent works; and there is also a later literature on the therapeutic value of music.

Arab music, §I, 3: Art music: The later Abbasids (900–1258)

(iii) Music and religion.

Here we seem to have reached a position diametrically opposed to that of the 9th-century tract condemning music. Yet any notion of an ideological reversal would be illusory: for those who dominated religious discourse, such views were esoteric and, at best, marginal. As before, performance of art music normally involved patterns of behaviour that ran counter to orthodoxy. Such music could not fail to be seen as an adjunct or incitement to immorality, being intimately linked with hedonism and wine-drinking, and intensifying the charge of the erotic verse it generally set (especially when performed by a qayna, who was expected to be physically attractive as well as accomplished). Consequently, entertainment music provoked continuing condemnation among Muslim legists.

Much of the literature on the permissability of music – still a live issue today – rehearses essentially the same arguments and often uses instruments as symbolic representations. The grudging acceptance normally afforded the music associated with important ceremonies integral to social life was articulated by declaring the frame drum licit, since it is central to, for example, marriage celebrations. Lutes and flutes, however, have particular associations with entertainment music, and the legists universally condemn them.

Not surprisingly, specifically religious usages provoked a conceptual rupture. The core phenomenon of the cantillation of the Qur’an was excluded from the domain of music and accorded a separate terminology. While the technical literature prescribes in great detail how the sacred text should be realized phonetically, there is no mention of pitch organization, despite the implications for intonation in the rules for pause and juncture. Yet the very fact that several authorities warn against the use of secular melodies for Qur’anic cantillation indicates that, in practice, such separation was not always rigorously maintained, and it later became quite common for male performers to excel in both secular and religious repertories.

The elaboration of a fundamentally negative ideology nevertheless had an important consequence: music theory became excluded from the citadel of religious scholarship. As the tradition of philosophical and scientific enquiry began to weaken, the intellectual territory of music shrank. Among its professional practitioners, who remained for the most part of low social status, with a significant number later coming from marginal or non-Muslim minority communities (e.g. see Jewish music, §V, 1), theory increasingly became practical rather than speculative, and was transmitted orally. For its cultured patrons, the horizons of musical discussion gradually dwindled to a core set of literary topoi, ranging from entertaining anecdotes to philosophical dicta and cosmological lore, the materials in either case being fundamentally derivative.

That mainstream Islamic thought eschewed further engagement with the rationalist Aristotelian philosophical tradition is largely attributable to Abū Hāmid Muhammad al-Ghazālī (1058–1111), whose major contribution was to integrate the spiritual and emotional core of Sufism within an orthodoxy tending towards the drily legalistic. His Ihyā’ ‘ulūm al-dīn (‘Revival of religious sciences’) contains an extended defence of music which articulates a Sufi standpoint, essentially arguing that it is possible to use music in a morally positive way, despite its potential for erotic arousal. Al-Ghazālī drew a distinction with regard to the listener’s reactions, stressing the possibility of a spiritual apprehension legitimizing music as an aid to progress on the Sufi path. (A parallel may be drawn with Islamic mystical literature, where the language of profane love became a recognized vehicle for the expression of religious and yearnings.)

Arab music, §I, 3: Art music: The later Abbasids (900–1258)

(iv) The Arabs and Europe.

The philosophical riposte to al-Ghazālī came from the Córdoban Ibn Rushd (also known by the Europeanized name Averroes; 1126–98), who influenced the development of medieval scholasticism in Europe. But it was his predecessor Ibn Bājja (Avempace; d c1139) who provided an explicitly musical link with Europe. The western conterpart of al-Fārābī as a philosopher-musician, Ibn Bājja is credited as a composer with the creation of an influential new style representing a synthesis of ‘Christian’ and ‘Eastern’ song. Both he and Ibn Rushd illustrate facets of the relationship between Christian and Islamic (to which should also be added Jewish) culture during the medieval period. Though by no means exclusive to Spain, it was there that contact was most intense and prolonged, involving both the intellectual world of translators and, more importantly, social interchange over several centuries between the ethnically mixed populations of Christians, Muslims and Jews (with the Muslim population consisting more of converts and Berbers than Arabs).

Scholarship has tended to concentrate on the nature and extent of Arab musical influences on Europe, an area where paucity of evidence allows conflicting interpretations. But one thing is clear: European interest in the Arab intellectual heritage did not extend to music theory, and none of the major texts was translated.

Turning to practice, however, a very different picture emerges. There is abundant lexical and iconographic evidence for the European acquisition of a wide range of instruments, the lute (‘ūd), rebec (rabāb) and nakers (naqqāra) being only the most obvious. Depictions of these and others are provided by the miniatures of the Cantigas de Santa Maria, which represent Christian and Moorish musicians at the court of Alfonso el Sabio (1252–84; see Cantiga, fig.1). In parallel to the synthesis achieved by Ibn Bājja, it seems that their melodic, rhythmic and formal conventions were sufficiently similar to allow mutual comprehensibility and transfer. But despite the likelihood of a considerable degree of cross-fertilization it is impossible to tell what specific melodic, rhythmic and/or formal structures might have been involved: we have the Cantigas de Santa Maria and a number of surviving troubadour melodies, but no comparable record of a medieval Arab repertory. Many of the Cantigas are in the zajal form, and it is likely that the strophic Arabic Mūwashshah and Zajal were also normally sung, but the parallel may be variously interpreted. Within Arabic poetry, these forms are Spanish innovations: there is no evidence for the emergence of strophic forms in Umayyad and early Abbasid court music, and although other Arabic literary antecedents have been mooted it is legitimate to speculate that they may have originated in contrafactum songs based on Romance prototypes. In short, although the music of the Arab courts must have provided a cultural model to be emulated, musical influences were probably not unidirectional.

(i) Eastern tradition: theory of the Systematist school.

4. Mongols and Mamluks (1258–1517).

Within the Mongol empire, which stretched from Russia to China, the part that included Persia and Iraq was ruled for a century by Hulegu and his descendents, the Il-Khans (1256–1353). Hulegu sacked Baghdad in 1258, executed the last Abbasid caliph, and then proceeded to invade Syria, but the westward advance of the Mongols was halted in 1260 by the Mamluks of Egypt, who dominated the eastern Arab world until the early 16th century. The Mamluk rulers were something of a caste apart, a military dynasty constantly replenished by slave recruits of Turkish (and later Circassian) origin. Little is known of their tastes in music, despite the occasional references of chroniclers to musicians at court.

Confrontation with the Mongols did not entail lack of contact: the Mamluks maintained diplomatic and artistic ties with the Il-Khans and musicians were invited from Iraq to Cairo. It is thus possible that rather than maintaining the tradition that had been fostered by the Fatimids, the Mamluks encouraged the further westward expansion of the eastern tradition, already dominant in Iraq by the beginning of the 13th century if not before.

Wherever the boundary between eastern and western art music traditions was situated, by the late 12th century they were clearly divergent. A major witness to the differences between them is al-Tīfāshī (1184–1253), a North African who lived for some years in Egypt and Syria. Echoing earlier contrastive encapsulations, he characterized the eastern tradition as more innovative and more affected by Persian taste, and the western as more conservative and, by implication, more complex. He underlined this distinction by identifying an intermediate style in Tunisia that was lighter than the Spanish but more florid than the eastern. He also confirmed differences in modal nomenclature and norms of formal organization, but supplied no definitions – for these we have to turn to the more specifically theoretical literature.

(ii) Eastern tradition: practice.

(iii) Western tradition.

Arab music, §I, 4: Art music: The Mongols and Mamluks

(i) Eastern tradition: theory of the Systematist school.

With the loss of the treatise by Ibn Bājja, virtually nothing of theoretical interest survives from the 200 year period between the works of Ibn Sīnā and his pupil Ibn Zayla (d 1048) and the appearance in 1235–6 of the Kitāb al-adwār (‘Book of cycles’) by Safī al-Dīn (d 1294). If ultimately indebted to the earlier theoretical tradition, the immediate antecedents of this work are unknown, although the particular combination of rigour and clarity of presentation that characterizes it is probably original to the author. This treatise inaugurated the school of ‘Systematist’ theorists (as conventionally termed by Western scholars), providing a model that profoundly influenced analytical writing until the end of the 15th century (and even some 20th-century ideas; see Iran, §II, 2).

The Kitāb al-adwār concentrates initially on the analysis of scale, and derives from al-Fārābī’s first tetrachord division on the tunbūr khurāsānī (Khorasan long-necked lute) an octave division of 17 pitches arranged in three symmetrical layers of two identical larger elements followed by one smaller one:octave = tetrachord + tetrachord + whole tone
tetrachord = whole tone + whole tone + limma
whole tone = limma + limma + comma.

Accordingly, the neutral intervals of practice, which were difficult to reconcile with the traditional analytical stress on the primacy of simple ratios, were now treated virtually as just intonation intervals. Safī al-Dīn then elaborated combinatorial consonance rules generating in the first instance all and only those tetrachords occurring in practice. From these and a largely overlapping set of pentachords he produced a series of 84 possible octave scales, and it is largely in terms of these that the principal melodic modes are presented.

Safī al-Dīn’s analysis of rhythm was also innovative, at least in its clear and unambiguous use of a traditional prosodic terminology. It also set the fashion for displaying the structure of rhythmic cycles in circular format. Other areas are, however, ignored: there is nothing on melodic structure, form, vocal quality or technique, and instruments. The Kitāb al-adwār ends, nevertheless, with yet another innovation – the inclusion of a few fragments of notation.

After Safī al-Dīn, theoretical writing is less original and tends to rework an existing body of material. But several later Systematist treatises are important in their own right, whether for the critical elaboration of ideas or for the additional information they include. Like those of Safī al-Dīn, all such treatises relate specifically to the eastern art music tradition. Most are written in Persian, and in several the factual content corresponds in the first instance to musical practice at the 15th-century Timurid courts of Samarkand and, later, Herat, while the use in others of Arabic reflects its continuing importance as a language of scientific discourse rather than a geographical locus within the Arabic-speaking world. Of the major texts in Arabic, theRisāla al-fathiyya of al-Lādhiqī (fl late 15th century) was dedicated to an Ottoman sultan; the Majalla of al-Shirwānī (d c1453) was probably written in Anatolia; and the Sharh mawlānā mubārak shāh was dedicated, in 1375, to Shāh Shujā‘ of Kirman, in west Persia. (This last is an extensive commentary of the Kitāb al-adwār, and in its general coverage of the thoretical domain, to which it adds the anatomy of the larynx, perhaps stands second only to al-Fārābi’s Kitāb al-mūsīqī al-kabīr.)

Given their origins and destinations, such texts do not relate directly to the norms of court music in Cairo and the major cultural centres of the Fertile Crescent. But in the eastern tradition a broad uniformity of idiom was maintained from the 13th century to the 16th, and much of the information in works such as those of the first great Timurid theorist ‘Abd al-qādir al-marāghī (d 1435) is as valid for the eastern Arab world as that found in the treatises of Safī al-Dīn, who had performed before the last Abbasid caliphs in Baghdad.

Arab music, §I, 4: Art music: The Mongols and Mamluks

(ii) Eastern tradition: practice.

The material contained in Systematist treatises, while selective, provides a more detailed picture than hitherto of the structural bases of the various modal, rhythmic and formal parameters of the eastern art music tradition in the course of their evolution from the beginning of the 13th century to the end of the 15th.

(a) Mode.

Safī al-Dīn cited 20 modes in all. Among these, two canonic groups predominate: the 12 shudūd (sing. shadd) and the six (later seven) āwāzāt (sing. āwāz). (The obvious association of the number 12 with the zodiac was, however, generally disregarded by Systematist writers, few of whom touched upon cosmology.) The names of the 12 shudūd are (in alphabetical order): abūsalīk, buzurk, hijāzī, husaynī, ‘irāq, isfahān, nawā, rahāwī, rāst, ‘ushshāq, zankūla and zīrafkand. The six āwāzāt are kardāniya, kawāsht, māya, nawrūz, salmak and shahnāz. Wherever possible (distortions notwithstanding), Safī al-Dīn described modes in terms of octave scales, that is, as fixed combinations of tetrachord and pentachord species. From the chapter on music in an encyclopaedia by the next major Systematist theorist, the polymath Qutb al-Dīn (1236–1311), it is also possible to discern the importance of certain non-tetrachordal species and non-octave structures.

The basic units from which more extended modal structures were formed in practice are shown in ex.2. (In both this example and those below the appropriate neutral values have been recuperated from the virtual just intonation equivalents of the theorists.) These structures were not combined at random: (Pythagorean) diatonic units, for example, were normally segregated from those containing neutral intervals. But the most important controlling factor was consonance: nearly all the recognized scalar combinations of these units are characterized by complete or partial parallelism (normally at the 4th, less frequently the 5th). Similarly, Qutb al-Dīn described analogous combinations at the same pitch level, which are likewise characterized by minimum differentiation.

Within this general framework, more specific features contributed to modal identity. A few modes were characterized by particular melodic movements, while a common convention concerned the identity of the initial and/or finalis, so that in some cases the order of occurrence of the constituent units was fixed. They might also be variably weighted, and the note linking the two units would frequently be prominent, as with the g in rāst (ex.3).

The pivotal importance of this linking note was such that it led to the creation of a new mode, and kardāniya (ex.4) may well have evolved from rāst in this way.

A different form of derivation is represented by a set of modes that probably emerged during the 13th century. This was undoubtedly derived from rāst (now with an optional additional major 7th). The new modes were differentiated from rāst by a successive raising of the locus of prominence to the 2nd, 3rd, 4th and 5th degrees above the originally prominent 4th (ex.5). It is from these degrees that they take their names: dugāh (‘second place’), segāh (‘third place’), and so on.

During the 14th and 15th centuries the system was further enlarged, partly because changes in its structural basis allowed slightly different patterns to develop. Two new sets, shu‘ab (sing. shu‘ba: ‘branch’) and tarkībāt (sing. tarkīb: ‘combination’) supplemented the canonical shudūd and āwāzāt sets. Some modes were identified with particular units, and melodies were probably articulated in terms of unit-based segments. One segment would be explored, at least to the extent of establishing its identity, before passing on to the next in a sequence that might also modulate through units not intrinsic to the mode in question.

(b) Notation.

Some faint idea of such processes of composition can be gleaned from the surviving examples of notation, even if their primary (and essentially didactic) purpose was no more than to demonstrate how notation could be done. As in the analysis of scale, the examples provided by Safī al-Dīn define pitch by single or paired letters of the alphabet arranged in a conventional sequence that assigns them numerical values. Relative duration is indicated by adding numerals below. The examples begin with a deliberately elementary descending sequence repeated in slightly varied form as a text-setting and continue with a further instrumental and vocal pair. They are followed by two instrumental fragments which unexpectedly have ancient mode designations otherwise preserved only in the western traditions. Recalling the Geniza outline (ex.1), they appear to preserve archaic and otherwise abandoned features in the context of a set of elementary ‘ūd exercises. Ex.6 is the second vocal piece, a song about love and the pain of separation. It shows a modal structure in which the melody is confined to a 5th and is dominated by the prominent central c, which is both initial and finalis.

For a more accurate reflection of 13th-century practice we must turn to the one example that transcends such limitations: the extraordinarily detailed, wholly unprecedented, and sadly unique example of notation – of a song ascribed to Safī al-Dīn (yā malīkan bihī yatību zamānī, ‘O sovereign, through whom fortune smiles on me’) – which concludes Qutb al-Dīn's chapter on music. He used a grid with vertical divisions for time units and five discrete horizontal layers for, respectively and in descending order: pitch (in Safī al-Dīn’s alphabetic-numerical represention); a percussion part differentiating two layers (presumably representing timbral contrast); changes of mode; indications of prolongations, pauses, expression and dynamics; and the song text.

Ex.7 gives one block of Qutb al-Dīn's original notation (the third rhythmic cycle) and transcribes the setting of the first hemistich, which displays a compositional technique that is presumably typical of the period. Thus the setting of the first seven syllables of the text, containing just the one initial pitch change, can be contrasted with the extended melismatic treatment of the next syllable. One may also note the segmentation signalled by breath pauses, the first phrase outlining a g–d tetrachord unit, the second filling it in but suggesting a shift of emphasis to f–c and the third beginning with a further shift to eb, which the descent to Bb then confirms as a modulation, only to end by restoring the original mode with a prominent g. The percussion part closely follows the two dimensions of syllabic versus melismatic and short versus long duration in the vocal part. The piece also contains an extended modulatory nonsense-syllable section, and the evidence of 15th- and 16th-century song text collections suggests that this was a typical feature.

(c) Rhythm.

Following Safī al-Dīn, Systematist theorists normally defined rhythmic cycles in two ways. One was to divide a circle into the same number of segments as there are time units and add symbols indicating those normally sounded. The other, parallel to the standard presentation of poetic metres in prosody, was to employ the syllables ta, na (each equivalent to one time unit) and tan, nan (each equivalent to two time units) and to divide the rhythmic cycle into feet of two to four time units, ta and tan always being initial in a foot, na medial and nan final. A representation such as tanan tanan tananan tan tananan suggests, accordingly, an internal 3 + 3 + 4 + 2 + 4 division and an associated distribution of percussions, with the initial time unit in a foot always sounded, the final one almost always not and the sounding of any medial ones generally optional. Not indicated, however, is how such schema would be varied in performance, and it is only from Qutb al-Dīn's notation that some insight can be gained into the ways in which percussions could be added (or suppressed) and contrasts of timbre employed. Table 5 gives Safī al-Dīn’s definitions of the eight rhythmic cycles listed in the Kitāb al-adwār. Several have variants, and the versions in his other treatise, the Risāla al-sharafiyya, are sometimes different again so that rather than definitive versions the list gives representative examples taken from sets of related forms.

The relationship between members of a set is sometimes obvious: another form of ramal, for example, is 2 + 2 + 2 + 2 + 4, and a variant of thaqīl thānī is 3 + 3 + 2. Other cases, however, recall transformational processes adumbrated by al-Fārābī, for it appears that there might have been considerable flexibility in the arrangement and even the identity of feet: there are cases of reordering (2 + 4 + 2 + 4 and 4 + 2 + 4 + 2 are listed as variants of ramal); substitution (2 + 4 + 2 + 4 is also listed as a variant of khafīf al-ramal, given in Table 5 as 2 + 3 + 2 + 3); elision (a variant of hazaj is 4 + 2); and extension (fākhitī has a reordered variant with two extra feet: 2 + 4 + 4 + 4 + 2 + 4 + 4 + 4).

After Safī al-Dīn, writers generally presented a greater number of rhythmic cycles than his eight, but with fewer or no variations. Qutb al-Dīn added a further five cycles, while towards the end of the 15th century al-Lādhiqī asserted that 18 cycles were in common use, and listed a further nine rarer ones. Several are extremely complex, involving an exceptionally large number of time units: thaqīl, 48; darb al-fath, 88; and chahār darb, 96. The longer cycles are mostly duple (combinations of feet of 2, 4 and 8 time units), and aksak combinations of duple and triple occur mainly among the shorter cycles (e.g. rawān 2 + 3 + 4 and samā’ī 3 + 3 + 4). The incidence of the various cycles in the song-text collections reinforces al-Lādhiqī’s perception that several were peripheral, but also suggests that the core cycles were not primarily the shorter ones: thaqīl, darb al-fath and chahār darb, for example, all occur frequently.

Several modes (e.g. isfahān, hijāzī and ‘irāq) are named after cities and regions, but there is no suggestion that they were specific to them. The great majority was widely known throughout the domain of the eastern tradition. In rhythmic cycles, on the other hand, there were clear regional preferences, at least during the 13th century. Safī al-Dīn stated explicitly that the Arabs used all the cycles here presented in Table 5, except the final one (fākhitī), while the Persians preferred the thaqīl al-ramal and also used fākhitī (if less frequently). Later, Qutb al-Dīn introduced a cycle called turkī (which may indicate use by Turks). However, such regional distinctions probably became less marked: the cycle darb al-fath, created by ‘Abd al-Qādir al-Marāghī in Tabriz, was known both by Al-Lādhiqī and his contemporaries in Herat, and survived into the later Ottoman system.

(d) Form.

The fullest account of forms provided by a Systematist theorist is that of ‘Abd al-Qādir al-Marāghī. His definitions single out as significant whether the form or section in question is set to words or not and whether the language of the verse is Arabic or Persian. But they also refer to structural features: differences between various sections, their presence or absence in a given form, and whether or not a fixed rhythmic cycle is used. The initial analysis of section types is accompanied by simple notated examples that contrast declamatory with slightly more florid styles of text-settings in a manner reminiscent of the much fuller notation of Qutb al-Dīn (ex.7).

The most common structural pattern seems to have been AABA, with B providing contrast through a shift to a higher register and/or modulation. As the song-text collections confirm, each of sections A and B might be characterized by internal contrast, both between styles of setting for different segments of the verse and between levels of textual specificity. Verse-setting subdivisions often ended with extraneous conventional phrases and contrasted with nonsense-syllable subdivisions, which were sometimes extensive.

‘Abd al-Qādir also described forms designed to display technical skill. Presumably marginal, these could have been appreciated only by an audience of some sophistication and may have appealed more to the musicians themselves. In one form two rhythmic cycles were played simultaneously, and in another several rhythmic cycles were introduced successively. The latter form could be combined with a modal counterpart of two possible types, one introducing a large number of modes successively, the other introducing modes at different pitch levels, thereby including all the notes of the theoretical octave.

Among the more basic forms, he listed the basīt, a setting of Arabic verse with an instrumental prelude, and the nashīd, a setting of verse with sections in free rhythm alternating with others in a fixed rhythmic cycle. These terms had already been used in the 10th century, when the term nashīd designated an initial section setting one or two lines of verse in free rhythm, in contrast to the remainder of the song, called basīt. Lack of documentation, however, makes it impossible to trace the intermediate stages of development.

A similar obscurity surrounds the emergence and evolution of the most extended form, the nawba. In the 13th century, five constituent parts were reported, and in the 14th, three. However, it is clear that for most of the 14th and 15th centuries the eastern nawba consisted of a cycle of four songs, all in the same mode, and using a restricted range of rhythmic cycles. They were distinguished more by linguistic than structural features: the first song set Arabic verse, the second Persian, the third either Arabic or Persian, and the fourth again Arabic. For ‘Abd al-Qādir, the nawba was clearly the most important form (and he attempted, unsuccessfully, to enlarge it by adding a complex fifth element). By the late 15th century, however, it was evidently in decline, and soon afterwards it disappeared.

Arab music, §I, 4: Art music: The Mongols and Mamluks

(iii) Western tradition.

Lack of documentation makes it impossible to give a comparable outline of the modal, rhythmic and formal structures underpinning the western tradition. No more than four mode names are mentioned by al-Tīfāshī; he makes no attempt at description and the rhythmic cycles are passed over in silence. The few later texts are equally uninformative, and although the philosopher-historian Ibn Khaldūn (1332–1406) integrated art music into his analysis of the cyclical rise and fall of dynasties, regarding it as the most delicate and vulnerable flower of luxurious court life, his interests concerned general laws of human society rather than specific musical phenomena.

However, al-Tīfāshī does provide information about form, repertory and performance. His characterization of the Spanish style as heavy and complex probably related to material reserved for professional soloists, who might improvise at great length: he cited one singing girl who spent two hours over a single line of verse. He certainly viewed the serious solo forms as more prestigious than the muwashshah and zajal.

Both types were included in the western nawba, which was quite distinct from the eastern. The introduction of a conventional sequence of song types is ascribed, along with so much else, to Ziryāb. But whatever its origin, it seems to have involved, much more than in the eastern Arab world, the notion of contrast: the first two forms, termed, interestingly, nashīd and basīt, were associated with slow (or ‘heavy’) rhythms. They were followed by muharrakāt and ahzāj, songs in faster (or ‘lighter’) rhythms. But by the beginning of the 13th century, as al-Tīfāshī reveals, the western nawba consisted of nashīd and sawt followed by muwashshah and zajal. As the performance of the latter probably involved a choral element, the change points to a further development, soloistic performance of the more ornate slower pieces being now counterbalanced by group participation in the faster and more repetitive strophic forms. This contrasts with the eastern nawba, which was the preserve of professional musicians. The implied integration of popular and art music forms points forward to a strand of communal music-making that later proved particularly important in North Africa for the preservation of the traditional repertory.

Such communal performances frequently occurred as an informal appendage to Sufi gatherings. Many of the major Sufi orders were founded, diffused and integrated within society at large from the 12th century to the 15th. The importance attached to music as an integral part of the Dhikr ceremony varied between the orders, but membership often allowed exposure to a religious repertory that shared structural features with the secular. The institutionalization of the orders facilitated diffusion and probably narrowed the gap between popular and art music, and between sacred and secular, at least to the extent that some performers of secular art music were partially trained through participation in Sufi activities.

5. The Ottoman age (1517–1918).

Already long established in the Balkans and western Anatolia, the Ottoman Turks captured Constantinople in 1453 and in the following century expanded southwards. By 1517 they had defeated the Mamluks and taken control of Syria and Egypt. During the rest of the 16th century they made advances into North Africa, but in the east it was not until 1638 that Baghdad was definitively incorporated into the Ottoman domains. After this, their empire shrank, first in Europe, but the connection with Egypt and the Fertile Crescent was not severed until the end of World War I. Thus for four centuries most of the Arabic-speaking world formed part of the Ottoman empire, with Istanbul as its principal cultural centre.

Frequently dismissed by Arab scholars as a time of political and cultural decline, the Ottoman period from the 16th century to the 18th has been insufficiently studied, and the texture of musical life in the major cities remains largely unknown, as does the nature and extent of Ottoman influence. It is, no doubt, legitimate to relate the gradual development of a quite separate Iraqi art music tradition to its peripheral position within the empire, but from this it does not follow that Ottoman models were imitated in, for example, Aleppo, Damascus and Cairo. Musicians from the provinces may have been attracted to Istanbul, but it is not known how many travelled in the opposite direction, although some were probably enticed by provincial Ottoman governors, the most striking example of such patronage being provided by Muhammad al-Rashīd, the mid-18th-century ruler (bey) of Tunis to whom the introduction of Ottoman instrumental forms is credited. In general, however, the western tradition continued along a separate path, and in North Africa more than elsewhere there remains an ideological stress on the conservative nature of the art music repertory. But although certain song texts are demonstrably ancient, the songs themselves may not be: the local differences of repertory, style and structure between the various urban traditions reinforce the obvious assumption of continuing, if gradual, development and innovation.

Similar local variations no doubt occurred in the eastern Arab world, particularly from the 17th century onwards, when the consolidation of a distinctly Ottoman repertory based on Istanbul practice must have created or reinforced differences between it and the other regional inheritors of the earlier pan-eastern tradition. Ex.8 gives the beginning of an instrumental composition, notated in Istanbul c1650 by ‘Alī ufkī and ascribed to a certain Seyf el-Mısrī (‘Sayf the Egyptian’). But this appears to be a typically Ottoman composition, and the extent to which it might also be representative of contemporary Egyptian style is unclear.

That practice in the Arab provinces remained largely independent is demonstrated by later Ottoman and Arabic song-text collections, which, in addition to the separation of repertories dictated by differences of language, are only partially congruent in modal and rhythmic nomenclature. Indeed, it may be that local styles retained with some tenacity elements that would be considered archaic by comparison with new Ottoman norms that had evolved by the early 17th century. Despite the possibly levelling effects resulting from the diffusion of the religious repertories of the more widespread Sufi orders, there is no need to assume that, say, either secular song or Qur’anic cantillation in Cairo imitated that in Istanbul. That there was still a clear separation at the beginning of the 19th century is suggested by the fact that although the tunbūr, a long-necked lute associated particularly with Turkish art music, had been introduced into Egypt, neither it nor its repertory had been assimilated. In effect, although Ottoman elements in 20th-century Arab art music are quite noticeable, particularly in the instrumental repertory and the mode stock, Ottoman influence appears to have been rather patchy and intermittent before the latter part of the 19th century, when visits to Istanbul by, for example, major Egyptian singers and composers point to a potential but also paradoxical fusion of Arab and Turkish traditions. Centrifugal strivings for cultural (as well as political) autonomy entailed centripetal competition, so that Ottoman musical influences became stronger as the power of the state ebbed away.

See also Ottoman music.

(i) Theory.

Theoretical treatises of the 16th century to the 19th differ markedly from earlier Systematist models, containing none of their more rigorously analytical and mathematical elements. But rather than a new departure, they represent a continuation of other types of text produced alongside the Systematist writings. They consist of didactic expositions in verse usually categorizing the two canonic sets of melodic modes according to their cosmological affiliations; and prose works that either explore cosmological and affective ramifications or present the conceptualizations of practising musicians with no access to, or interest in, the more scientific theoretical tradition.

The first major representative of the latter tendency was Ibn Kurr (d 1341), who predated the Ottoman period. His nomenclature of scale degrees repeats the same set of terms on each string of the ‘ūd, and he provided a type of modal definition found in several later works. Thus with the exception of al-Saydāwī (d 1506), who gave a graphic, almost stave-like representation of modal outlines, later writers followed Ibn Kurr in eschewing notation. Instead, their works spell out the various pitch steps forming a modal nucleus which often takes the form of a rudimentary melodic contour akin to some of those provided by the 15th-century Systematist al-Lādhiqī. Ex.9 compares versions of three modes provided by Ibn Kurr, al-Lādhiqī and an anonymous work probably dating from the 17th century, the Shajara dhāt al-akmām (‘Tree with calyxes’).

The Shajara is one of a number of similar treatises that are difficult to date, but although its terminology is different, the basic conceptualization of scale in the Shajara is identical to that found in the Ottoman treatise of c1700 by Dimitrie Cantemir (1673–1723): interval sizes are not defined, but two sets of pitches are distinguished, one principal (the notes of the rāst scale), the other secondary (those intermediate between them). However, no evidently Ottoman features can be detected in the Shajara, and the work is somewhat conservative in comparison with Cantemir's. It also, like several other texts, differs sharply from Cantemir's in its strong interest in cosmological affiliations, relating the modes to the humours, elements and the zodiac. (Similar conceptualizations appeared in the western Arab world, where, at least since the time of Ibn al-Khatīb (1313–75), the modes have been called generically tubū‘: ‘natures’; sing. tab‘.)

In its account of rhythmic cycles, the Shajara draws directly on Safī al-Dīn. It accords primacy to his set of eight (see Table 5 above) – even if some rather different definitions are offered – and considers more recent cycles to be derived from them. But if the names it mentions are not among those listed by, for example, al-Lādhiqī in the late 15th century, they are equally absent from Cantemir’s account of late 17th-century Ottoman practice, and its explicit reference to the existence of many further cycles among non-Arabs is substantiated by other texts that distinguish between Arab and Turkish cycles. Their rich and varied nomenclature suggests that the gradual diffusion of Ottoman norms did not seriously erode the independence and creative vitality of local traditions.

The Risāla al-shihābiyya by the Lebanese mīkhā’il Mushāqa (1800–1880) points to a rather similar situation on the modal side, at least in Syria. Here modes are described in the traditional form of brief melodic matrices, but there are also innovations: the reintroduction of the mathematical definition of interval sizes and a new theoretical concept, the quarter-tone scale. But this innovation did not lead in an unbroken line to the arguments about intonation that later proved to be one of the dominant strands of 20th-century debate: when the quarter-tone concept surfaced in two 19th-century Egyptian works (by Shihāb al-Dīn al-Hijāzī (written c1840) and ‘Uthmān ibn Muhammad al-Jundī (1874 or later)), it did so in a distorted form indicating that it had not yet been naturalized.

(ii) Practice.

The two large-scale 19th-century song text collections accompanying the works of Shihāb al-Dīn al-Hijāzī and ‘Uthmān ibn Muhammad al-Jundī are devoid of Ottoman material, and thus indicate a continued independence of repertory in Egypt. But they also suggest a shift in taste: they are rather colloquial in tone and show a predilection for strophic muwashshah poetry.

The formal labels they exhibit are still in use today, but post-Systematist texts generally ignore any discussion of form, so the relationship between the major 19th- and 20th-century forms and earlier ones is obscure. The 15th-century vocal form ‘amal resurfaced as the Ottoman kâr (both words meaning ‘work’). But in contrast to the instrumental forms bashraf (Turkish peşrev) and samā‘ī (semai), the major Ottoman vocal forms have apparently had no impact on the Arab world. With differing degrees of complexity, they preserve the AABA form of earlier song types, and thus contrast with the ritornello muwashshah type.

Both traditions recognized improvisation as a separate genre, whether instrumental (taqsīm, or vocal (layālī), but within composed forms there was greater fluidity on the Arab side, allowing the interjection of much improvised material. Sometimes improvisation predominated as, for example, in the qasīda and mawwāl (the difference in terminology relating to the nature of the text, classical in the qasīda and colloquial in the mawwāl). Even within the most extended vocal form, the dawr (possibly a 19th-century innovation), the composed frame expanded to allow the insertion of an increasing number of improvisation sections termed hank.

These Egyptian song text collections are organized into blocks containing material in the same or related modes (maqāmāt; sing. maqām). These are headed by the term wasla, which also designates a large-scale cyclical form. How and when this evolved is not known. In the west it is tempting to assume a direct development from the 13th-century nawba as described by al-Tīfāshī to the more complex modern form. But the latter is, in evolutionary terms, a new species: an extended cyclical framework within which selections from the traditional repertory are performed; whereas its medieval forebear, to judge from the number of nawbas an expert performer needed to memorize, had to be, as in the east, a sequence of particular pieces. In the eastern tradition there is a similar contrast with, in addition, clear evidence of temporal discontinuity: the nawba disappeared long before the earliest references to the Ottoman fasıl – the first of the modern cyclical forms to be attested. The fasıl was made up of a particular sequence of improvisations and composed genres into which modally relevant selections from the repertory would be slotted.

The eastern Arab cyclical forms are similarly complex, but it would be facile to assume they simply derive from the fasıl: the sequence of events in the wasla is rather different, and although lexically identical the Iraqi fasl is at an even further remove. Less a framework for selection than the articulation of a repertory, it may be considered as intermediate between the Turkish structural model and that of the Persian dastgāh (see Iran, §II, 3). However, it could also be said that in its greater emphasis on integrating a degree of vocal improvisation within a relatively fixed corpus of pre-composed material, the Iraqi maqām presents a type of modal practice that is closer to the melody end of the spectrum when compared with the early 19th-century modal definitions offered by Mushāqa, whose adherence to the earlier matrix type has affinities with Ottoman practice as represented by Cantemir (see Mode, §V, 2).

In the Ottoman, Syrian and Egyptian traditions, modal structure provided a flexible set of directives for progression through a given pitch set (or sequence of pitch sets), emphasizing certain nodal points and providing conventional ways of approaching and prolonging them, but allowing considerable freedom with regard to reordering, repeating or omitting them before reaching the invariable finalis. Equally, the 18th- and 19th-century Egyptian, Syrian and Turkish corpora of modes resemble each other in the effacement of previous internal hierarchies. Where they differ is that the potential for the creation of new modes was more fully exploited on the Turkish side than in Syria, and hardly at all in Egypt, where the late 19th-century additions to the mode stock are generally Turkish loans.

As already noted, there was evidently a higher degree of regional variation among rhythmic cycles than among modes. Continuities are difficult to trace, especially when later writers fail to define the total number of time units in the cycle (although some do specify the dum and tak strokes that outline the fundamental pattern of timbre distinctions (see Turkey, §IV, 3). Some new cycle names may distinguish those entities previously lumped together as variations of a single cycle, while others point to innovation. This probably involved reshuffling conventional stroke sequences, but for all their variety and complexity, it is difficult to think of the cycles as a systematic exploitation of combinatorial possibilities.

The 19th century was a period of increasing exposure to (and subjugation by) the West. In reaction, there were efforts at political, educational and cultural renewal and reform. As far as music was concerned, conscious attempts at Westernization, the military band apart, still lay in the future. Despite the obvious symbolism of the Cairo Opera House, which opened in 1869 with a performance of Rigoletto, there was no adverse effect on indigenous art music. On the contrary, its status seems to have risen, and it may have benefitted from an increasing interest in the arts among the newly educated classes. In Egypt, the Khedives patronized major composers and singers whose social respectability was reinforced by their initial training in, and association with, the religious repertory.

If the first half of the 19th century remains poorly documented, the history of the modern Arab music in the east does stretch back to include the major figures of the second half. Among them may be cited Ahmad Zaydān (d 1912) in Iraq, and in Egypt ‘Abduh al-Hāmūlī (1843–1901), his wife Almaz (1860–96) (whom al-Hāmūlī prevented from performing after their marriage), and Yūsuf al-Manyalāwī (1847–1911). The Syrian Ahmad Abū Khalīl al-Qabbānī (1841–1902) and the Egyptian Salāma Hijāzī (1852–1917) were pioneers of musical theatre; through their recordings something has been preserved of late 19th-century Egyptian art music and the abundant creativity of its outstanding singers.

(iii) Ensembles and instrumental music.

6. Developments since 1918.

(i) Introduction.

(ii) Theory.

(iv) Notation.

(i) Introduction.

Two main lines of thinking defined the aesthetic of Arab music throughout the 20th century: the terms applied to them are modernization or development (tatwīr or tatawwur) and tradition as conveyed by the word ‘patrimony’ (turāth), used instead of the term for ‘tradition’ itself (taqlīd), which arouses only a lukewarm response. While the first concept (tatwīr) was dominant in the first part of the 20th century, the second (turāth) was more usual in the second half, although the defenders of tatwīr were not entirely defeated.

After World War II, when a number of former colonies won independence, national forms of music as distinct from Arab music in general emerged. In these forms dialect song takes precedence over the language of the classical repertory (klāsikī). They employ local instruments, usually from country areas, and display local colour (lawn), which distinguishes them from one another: in Tunisia, for instance, the prominence of the mizwid bagpipe has made it almost a national symbol. In Syria and Lebanon, the long-necked lute known as buzuq is prominent, and in Morocco the ginabri lute has become popular. In Sudan, on the other hand, the tanbūra or rabāba lyre has given ground before the increasing popularity of the ‘ūd. These national musical styles often draw on the general form of song, ughniya. Lebanon also has a form of sung drama, the ‘sketch’.

The clear reversion to tradition in the second half of the 20th century was due partly to Western influences and the musical dominance of Egypt, all leading to a style condemned by many critics who would like to see a return to a ‘purer’ form of song, with poetic texts of high quality. Among other factors involved is the production of poor-quality, mass-market music for radio and television. The movement towards a return to tradition does not aim merely to revive whatever is most authentic. It is usually envisaged in terms of modern music, conveyed through the medium of the firqa orchestra or other technical means. However, many marginal forms of music are still not well known, such as the huge paraliturgical repertory. Another source of nourishment for the revival of traditional Arab music comes from the West, where much archival material is recorded on disc and recitals of traditional music are given. At the end of the 20th century the takht (a traditional chamber orchestra consisting of three to five instrumentalists and a singer) and certain vocal techniques were being revived, thanks to their preservation on old 78 r.p.m. records, which are valuable archives and sources of reference. In 1990, after the war in Lebanon, there was a new musical development with the rediscovery of the art of the qasīda, long set aside and almost forgotten while national Lebanese song attained prominence.

Arab music, §I, 6(ii)(c): Art music after 1918: Form

(ii) Theory.

In 1905–06 the Kitāb al-mūsīqā al-sharqī (‘The book of eastern music’) by Kāmil al-Khulā‘ī (1879–1938) established the equidistance of quarter-tones in the octave. This scale of 24 quarter-tones was the subject of fierce discussion at the Congress of Cairo in 1932, where the participants divided into two opposing camps: the Egyptians supported division of the octave into 24 equal quarters, while the Turks (represented by Yekta Bey) and the Syro-Lebanese (Sabra and Tawfīq al-Sabbāgh) rejected the system of equal division.

In 1959 and 1964 the Egyptians organized two symposia to settle the differences of opinion arising from the controversy at the 1932 Congress over the equidistance of quarter-tones. The aim of these symposia was to establish the principle of equal temperament on the basis of the quarter-tone and give official sanction to its teaching. However, theoretical thinking between 1949 and 1974 took a different direction in Syria, especially in the works of Mikhā‘īl Allawīrdī (1904–81) and in volume four of al-Samā‘ ‘inda al-‘Arab (‘Music among the Arabs’) by Majdī al-‘Aqīlī (1917–83), published in Damascus (1969–79). Both writers employed abstruse mathematics to reach conclusions contrary to those of their Egyptian colleagues, theoretically reaffirming the existence of the natural scale in opposition to the artificial quarter-tone scale recommended in Egypt. In his theoretical research, Allawīrdī also proposed introducing the notion of the cent. In Egypt Youssef Shawki (1925–87), one of the last theorists of the second half of the 20th century, suggested in his vast retrospective and theoretical survey of the state of musical research, Qiyās al-sullam al-mūsīqī al-‘arabī (‘Measuring the Arab musical scale’; Cairo, 1969), that it was time to turn to applications deriving from electronics. At this point interest in theoretical studies began to wane, and in the second half of the 20th century they were no longer considered a priority. Practical theory became more appealing than pure speculation, which gave way to a great many publications of all kinds on the rudiments of music and solfeggio, as well as teaching manuals offering a combination of theory and practice. These educational publications were innumerable. Two of them became bestsellers, the work by Mahmoud Ahmad El Hefny (1896–1973), entitled al-Mūsīqā al-nazariyya (‘Theoretical music’; Cairo, 1938, 6/1972), and the later method by the Lebanese Salīm al-Hilū (1893–1980), al-Mūsīqā al-nazariyya (Beirut, 1964).

The change from purely theoretical studies to a more practical approach also affected the nature of the quarter-tone itself. The quarter-tone, nīm (derived from dividing a tone into four equal parts), was later largely replaced by the concept of the three-quarter tone tīk. Introduced in the last quarter of the 20th century, this has been predominant ever since. Musical practice shows that intervals of three-quarters of a tone occur much more frequently than quarter-tone intervals, which determine the structure of only two rarely performed modes: ‘awj and sāzkār whose first intervals are quarter-tones. Contemporary discussions concerning the quarter-tone imply a return to purely theoretical considerations; practical usage conforms to the three-quarter-tone system.

(a) Mode and modality.

(b) Rhythm.

(a) Mode and modality.

The study of modes became the chief preoccupation of the 20th century among both Arab and Western scholars. The latter, following Baron d'Erlanger, applied their minds to the subject in meticulous detail while attributing great complexity to it. In Arab research, on the other hand, the concept of the mode has been simplified and tends to be reduced to an arrangement of intervals defining an abstract scale. Today, teaching of the subject is confined to the definition of a series of ascending or descending intervals, which the pupil must learn, and which no longer corresponds to the old modal concept, taught by means of melodic formulae. Western studies, however, see the mode (for which it retains the term maqām) as a developmental form and the supreme expression of musical thinking, in line with the use of the term in Iraq, where it also has the quite different connotations of a musical form par excellence in the name of the al-maqām al-‘irāqī (the Iraqi maqām).

The term for ‘mode’ used by practising musicians in the Middle East is naghm (pl. anghām). However, the term maqām, retained by the Turks among others, was also used by theorists at the beginning of the 20th century. It asserted itself as standard terminology through the agency of Western musicology, which sanctioned it and devoted much attention to it. As a result, the Congress of Cairo in 1932 decided to use the term maqām rather than naghm, while retaining the term tab‘ (‘nature’) in North Africa, where it is commonly used instead of the other two terms, which are absent from the local musical vocabulary. Tab‘ also has semantic connotations: it expresses a close link between the mode and the medieval humours, being the four elements or the four natures (in Morocco), or the mode and the hours of the day (in Algeria).

The modal system was based on the seven degrees of the scale, each degree generating one or more of the principal modes and their derivatives. These ideas are clearly illustrated in two theoretical works written at the same time, one by Ahmad al-Safarjalānī of Damascus (1818–93), who arrived at a total of 107 derivative modes (furū‘) of the seven fundamental modes (anghām) in his book al-Safīna al-adabiyya fī al-mūsīqā al-‘arabiyya (‘The vessel of belles-lettres in Arab music’; Damascus, 1889), the other by his Egyptian contemporary Muhammad Dhākir Bey (1836–1906), an essay entitled Kitāb hayāt al-insān fī tardīd al-alhān (‘The book of the life of man in the repetition of melodies’; Cairo, 1895). Its author adopts the same classification, also calls the mode the maqām, and relates it to the scale (of eight rather than seven degrees, called barda, singular of bardāt). This approach was retained in the 20th century as a whole, the modes being classified in terms of the seven degrees of the scale, with the difference that while the basic mode was rāst notated in the key of G until the Congress of Cairo in 1932; after that date rāst was transposed into the key of C and has become accepted throughout the Arab world in that form, although it should be added that musicians often tune their instruments in D. Another major difference, and one adopted by the Congress of Cairo on the basis of the work of the Lebanese Egyptian theorist Alexandre Chalfoun (1881–1934), is the idea that the mode goes further than the octave; defining a mode by relating it to either two octaves or the octave plus a 5th, the disposition of the degrees not always being the same in the lower and upper octaves.

The Congress of Cairo retained modes going beyond the octave in the Syro-Egyptian area. They employ a system of tetrachords (jins, pl. ajnās), suggesting that Arab music is essentially tetrachordal in origin. The mode is defined by the combination of two, three or four tetrachords, each pair of which may sometimes (but not systematically) be separated by a major degree (fāsil tanīnī). These tetrachords may bear the same name or various other names. Thus rāst consists of four similar and successive tetrachords, each called rāst and extending over two octaves, while bayyātī consists of two similar tetrachords repeating in the octave, called bayyātī and būsalīk. Hijāz consists of the following tetrachords: the hijāz, bayyātī, husaynī, hijāz and būsalīk. The modal signature, which enables the mode to be clearly recognized, does not stipulate development over two octaves; on the contrary, the nature of the mode is made evident by a trichord (in the case of sabā), a tetrachord (in the case of hijāz) or a pentachord (in the case of nawā athar). This system of the definition of the maqām by the addition of tetrachords is a feature of art music and is not found in popular or rural music, which remains within a narrow ambitus.

Modes within the octave, regarded as heptatonic, form the basis of the repertory of Arabo-Andalusian music and of the nawba genre, particularly in Morocco. In the resolutions of the Congress of Fez in 1969, they were definitively established and limited to a total of 26 named modes. They differ from the other modes of Arab music in not using the three-quarter tone and depending on diatonic scales. They are not envisaged as a system of tetrachords but constitute melodic units within a narrow ambitus and can be identified by their signature.

Pentatonic scales had already been recorded by collectors in the Yemen, Sudan, Egypt, Morocco and Tunisia; the genre employing pentatonic scales is known as rāst ‘abīdī in Tunisia Its existence was not officially recognized in Arab music, under the name of sullam al-khumāsī (pentatonic scale), until the Congress of Khartoum in 1984.

Certain countries of the Arabian peninsula, such as the Yemen and the Gulf States, never had a technical vocabulary defining modality, but sīkā (or segāh), the mode par excellence of Arab music, is the basis of a great deal of Saudi music. A proposition was put forward in Yemen at the Congress of Sana‘a in 1997 suggesting that the art music repertory of Sana‘a, known as ghinā san ‘ānī (song of Sana‘a), be provided with modal terminology borrowed from elsewhere. But support for the proposal was not unanimous, since it would have meant the introduction of foreign terms for such local traditional forms as husaynī ‘ushayrān (Sana‘a’s most widespread modal structure).

Between 1960 and 1970 political ideals envisaging the founding of a single Arab nation from the Atlantic to the Gulf prompted several attempts to identify similarities between associate modes with similar intervallic structures: ‘The rāst known in Morocco by the name of istihlāl … corresponds in Tunisia to the second kind of rāst al-dhīl mode' (S. el Mahdi, La musique arabe, Paris, 1972, p.38). This approach encountered opposition in the late 20th century, on the grounds that if two modes set out from the same intervals, under different names, their commas and modal signatures will still vary, and they can be distinguished on that basis. Furthermore, the same degree may be felt as higher or lower in different modes making use of it. Practising musicians preferred modal degrees that were felt to be higher or lower than theoretical equalities. It has now been proved that the rast of the Middle East does not have anything to do with the Moroccan rasd, the latter being pentatonic in concept.

Another idea ran through 20th-century theory: an obsession with transposition (taswīr). Once equal temperament has been introduced, nothing stands in the way of the transposition of a mode to different degrees of the scale. This has been a question of theory rather than practice. Although the idea of transposition may have been adopted in theory, and for some musicians transposition to various degrees of the scale of 24 quarter-tones represents a pure exercise in style and a technical tour de force, it has not won approval in the traditional repertory, and some transpositions have been condemned as contrary to the aesthetic of Arab music, although others are tolerated in practice.

Arab music, §I, 6(ii)(c): Art music after 1918: Form

(b) Rhythm.

20th-century theory was much concerned with rhythm, retaining the term awzān (sing. of wazn) or more usually īqā‘āt (sing. īqā‘) and imposing these terms on contemporary practice. Lists of rhythms have been drawn up. Ahmad Safarjalānī gives rhythm precedence over discussion of the intervals. However, this was reversed in the 20th century, whereby theoretical works open with the discussion of modes and follow it with the discussion of rhythm.

Muhammad Dhākir Bey's Kitāb rawdat al-bahiyya fī al-awzān al-alhān al-mūsīqiyya (‘The book of the garden of delights concerning the rhythms of musical melodies’, Cairo, c1895) was the first book entirely devoted to rhythm as distinct from modality. The author distinguished three kinds in practical use in his time: Arab rhythms, Turkish rhythms and the Western metrics known as afrank. Western influence on rhythm in Arab music was already present when he was writing, and it increased throughout the 20th century.

The work of the Syrian ‘Alī al-Darwīsh, a great expert on rhythm, enabled the Congress of Cairo to draw up an impressive, if not exhaustive list of the various rhythms known in the Arab world. They are not defined numerically, as in European metre, but by means of a combination of long and short values either with or without rests, the whole being articulated as a series of strong or weak beats determining the accentuation.

The concept of rhythmic cycles (usūl, adwār) in the art music of the Middle East has been increasingly abandoned. These cycles could be very extensive, for instance the cycle of 120 beats (known as the zanjīr or ‘chain’), as recorded in Tawfīq al-Sabbāgh's al-Dalīl al-mūsīqī al-‘āmm (‘General guide to music’, Aleppo, 1950), and a cycle of 176 beats (known as fath) recorded at the Congress of Cairo. During the 20th century the structural concept of long cyclical time was relinquished in favour of short, metrical time, with a decline in standards of memorizing. This led to a need for written notation as an aid, although that need has not yet been generally felt in the area of rhythm, where the majority of percussion players still rely on oral tradition. In their interpretations, complex rhythmic cycles are reduced and resolved into small units, as in contemporary interpretations of the Syro-Egyptian muwashshah verse form (where singers refuse to observe the rhythmic cycles literally). Thus a cycle in 24 or 16 beats is often reduced by performing musicians to simple cycles of three, four, six, eight beats and so on.

In the 1960s and 70s scholars discovered the polyrhythms practised in the Gulf States, particularly in the repertory of the pearl fishers on the seaboard of Bahrain, Kuwait and Qatar. This discovery has left its mark on Arab art music, which always used to be monorhythmically articulated. Polyrhythms have been exploited by composers such as the Egyptian Baligh Hamdī (1934–93), who introduced the principle into mass-market music.

There was a general tendency in the late 20th century to look back and use simple metres, a four-beat metre being increasingly popular, a phenomenon illustrated by the evolution of mass-market and ‘light’ music.

Arab music, §I, 6(ii)(c): Art music after 1918: Form

(c) Form.

In previous centuries, the idea of form was linked to the idea of composition (ta'līf). This term, used in the past, was replaced at the end of the 20th century by the term ibdā’, a concept differing from its predecessor in indicating a concern with creativity and innovation, whereas the older term suggested conservation and deference to the past. In the second half of the 20th century, the concepts of qālib and shakl were adopted to define musical form, which was also determined by the emergence of the term ughniya (and its plural ughniyāt or aghānī), ‘song’. It eclipsed all other terms and came into general use after World War II. Any discussion of form in the 20th or 21st century is implicitly understood to involve ughniya, a word probably coined in Egypt. It was preceded by the terms al-ghinā' al-jadīd (‘new song’; Qastandī Rizq, 1946), or al-aghānī al-hadītha (‘modern songs’; Mansī 1949), which occur in critical writings of their period and in themselves are evidence of a sense of conflict between the old and the new. In Egypt, the term ughniya very quickly replaced its predecessor, taqtūqa, which fell into disuse between 1930 and 1935. Taqtūqa, a simple couplet and refrain structure, served as the springboard for the successful emergence of the light popular song; it had been the province of the dancing women known as almeh in Egypt at the end of the 19th century, but on becoming fashionable it was adopted in the musical theatre (al-masrah al-ghinā'ī, hence maghnā, also designating the repertory of the musical theatre). Musical theatre became important in the revival of musical creativity in Egypt, the Middle East and North Africa (where the café-chamta of Tunisia is a distortion of the French café-chantant). The term ughniya also superseded the terms mūnūlūj (derived from French monologue), meaning a song for stage performance by a solo singer, and alhān (melodies), which were still found in written Egyptian texts of the first quarter of the 20th century.

The ughniya is a useful and easy musical form based on the alternation of couplets and a refrain, performed by a solo voice to which a chorus usually responds, and accompanied by an instrumental ensemble that has grown steadily larger. Since 1960 the term ughniya has designated ‘song’ in a very wide sense, whatever its structure and organization: with or without a chorus responding to the soloist, with or without an instrumental prelude, and with or without free parts alternating with metrical sections. Ughniya is now a generic term embracing all contemporary composition appealing to a wide public, and sometimes, as in certain long songs performed by Umm Kulthum, has been substituted for the old term qasīda.

After 1960 various qualifying terms were introduced to define the concept of ‘song’ more precisely. They include ughniya ‘ātifiyya (sentimental song, the most widespread genre in the Arab world), ughniya siyāsiyya (political song), ughniya rīfiyya (local song, country song), ughniya sha‘biyya (popular or traditional song), ughniya khafīfa (light song), ughniya badāwiyya (Bedouin song), and finally, the latest coinage, ughniya shabābiyya (song of youth), the designation officially adopted at the Third Congress and Festival of Arab Music held in Cairo in 1994 for a genre known in Europe as jil or gil. The ughniya shabābiyya is distinguished from other kinds of song in reintroducing an element of dance, absent from the other genres. Around 1970 a number of other terms, linked to geography, came into use: including ughniya khalījiyya (song of the Gulf), ughniya yamaniyya (Yemeni song) and ughniya sūdāniyya (Sudanese song), which thanks to Sayyid Khalīfa made a considerable mark on the repertory of contemporary Arab song through the huge success of the mambū al-sūdānī (Sudanese mambo).

In the last decade of the 20th century there was a new move, principally on the part of Arab musicologists, to replace the term ughniya by uhzūja, another word for ‘song’, and to denote a type more closely attached to its roots.

Although the general term ‘song’ does tend to infiltrate North Africa, North Africans have remained resistant, and faithful to the idea of separate genres rather than a generalized formal concept. Algerian terminology, for instance, includes the genres of sha‘bī (a popular urban song), hawzī (traditional urban song), and Raï, while in Morocco the term malhūn denotes the traditional urban responsorial song structure of couplets alternating with the refrain. The term ughniya is not used for these specific genres.

The word ughniya is also seen as opposed to the classical repertory, which was revived by common consent, particularly in the second half of the 20th century, and retains many specific terms inherited from past centuries. There are two aspects to this traditional repertory. The first perpetuates a series of forms handed down from previous centuries and now in current practice. They are known to us from accounts in written texts and treatises. In the Middle East, the dominant forms are bashraf (instrumental overture), samā‘ī (instrumental prelude or interlude), lūngha (instrumental piece with a bright, lively conclusion, a term probably derived from the Jewish klezmer repertory as hongah), mūwashshah (composition for chorus, usually based on a rhythmic cycle in a semi-classical manner), mawwāl (sung and improvised dialect poem), taqsīm (instrumental improvisation) and wasla (or suite). The wasla survives in Aleppo, Syria; it has now disappeared from Egypt (see Syria, §2(iii)(a)). The wasla unites several of the forms mentioned above in consecutive performance, with modal unity as the common denominator, and featuring a small ensemble and solo singer in whose part improvisation figures prominently. The other kind of suite is the nūba of North Africa, for chorus and instrumental ensemble, except in Algeria, where a solo voice is dominant and a chorus of members of the instrumental ensemble responds to the soloist (at Algiers and Constantine) The term nūba antedates wasla, which does not appear in texts until the 19th century. A form with a different name is the dawr, a vocal piece for soloist and small chorus originating in 19th-century Egypt; it fell into decline and died out in Egypt in 1962 with the death of its last performer Sālih ‘Abd al-Hayy. The dawr was also maintained in Syria and was revived in the latter part of the 20th century, through the efforts of a movement outside the Middle East (particularly strong in Paris) promoting a return to the music of the earlier part of the century.

The second aspect consists of orally transmitted local forms within their original geographical regions. They comprise the Iraqi maqām, a suite of free and metrical pieces; the sawt of the Gulf States, a solo song performed to the accompaniment of the ‘ūd and the small double-headed drums, mirwās; and the ghinā' san‘ānī (song of Sanaa), another suite in several movements.

(iii) Ensembles and instrumental music.

The takht still existed in the Middle East at the beginning of the 20th century, although the instruments of which it consisted were seldom the same: the qānūn and violin featured most prominently, sometimes with the ‘ūd. Although the takht did not die out, it retreated into the background in the second quarter of the 20th century as the firqa orchestra rose to prominence. The firqa consists of ten or more members, in essence a group of violins with a qānūn and other traditional instruments (nay, ‘ūd), together with cello, double bass and percussion; bongo drums have become standard. This kind of orchestra of variable composition, first found in Egypt and characterized by its increasing borrowing of Western instruments has conquered the Arab world, largely through the medium of radio. The firqa was countered in the 1960s by the creation in Lebanon of an orchestra with the same number of players but different instruments, in which the accordion and European piano feature prominently. The Egyptian type of firqa, however, predominates and has spread to the Arabian Gulf since 1970.

In the second half of the 20th century three types of firqa developed: the strictly instrumental ensemble, whether accompanying a solo singer or not; the ensemble based on a large chorus accompanied by instruments; and finally the type known as the firqat al-funūn al-sha‘biyya (ensemble of popular arts), combining instruments of art music with rural instruments such as the zūrnā. This last-named type of firqa has also reclaimed the genre of dance and given it prominence. The larger such ensembles became, the more usual has it been to find a conductor (qā'id al-firqa), who has now become the key element of the firqa, regulating the relationships between the various instruments which once used to break into free and spontaneous improvisation. Unison now reigns supreme, and the heterophony of the past has tended to disappear. Around 1986 a general reaction against the firqa in Baghdad led to the emergence of a specific kind of ensemble known as the firqat al-bayāriq, consisting solely of traditional instruments, but with the ‘ūd, qānūn, santūr and nay grouped into instrumental desks, with support from a percussion quartet. The notable feature of the ensemble was its rejection of all Western instruments. A parallel development has been the emergence of permanent symphonic ensembles: the Cairo Radio Symphony Orchestra (1934), which in 1959 became the Cairo Symphony Orchestra; the Baghdad National Symphony Orchestra (1959); the Tunis Symphony Orchestra (1969); and the National Symphony Orchestra of Damascus (1992).

The importance of new and modernized instruments led in turn to the rise of a new phenomenon, the tawzī‘, literally ‘arrangement’ or ‘orchestration’. It is found principally in ‘light’ music, where the composer known as mulahhin (melodist) is now associated with the instrumental arranger (muwazzi‘).

Such developments generated the production of a good deal of instrumental music in the second half of the 20th century, often pieces that were originally songs and had been transposed for orchestral ensemble. The first transpositions of the sung nūba for orchestra were made in Algeria by Merzak Boudjemia (1933–85). The concept of descriptive music (mūsīqā taswīriyya) appeared in the 1930s with the emergence of the Egyptian film industry and is still evolving. The instrumental overture (iftitāh) became increasingly common; its purpose was to introduce the song (ughniya). In Egypt the philosopher Fu'ād Zakariyyā, in his al-Ta‘bīr al-mūsīqī (Cairo, 1956), denounced vocal music, traditionally the basis of Arab music, for its lack of expressive potential. He argued for the development of a strictly instrumental and symphonic repertory which would allow new expressivity to develop.

(iv) Notation.

Another development throughout the 20th century has been the adoption of European notation. This phenomenon was greatly encouraged by the proliferation of European military and brass bands in the Arab world, many with Turkish bandmasters with Western training. Musicians now had to accustom themselves to written notation, something that had also been introduced by European missionaries to the Middle East who transcribed eastern hymnology for use in worship by Arab Christians. 1867 saw the publication in Beirut of Mazāmīr wa-tasābīh wa-aghānī rūhiyya (‘Psalmody, chant and religious hymns’) by the American missionary Edwin Louis. His work was followed in Lebanon by that of the Maronite monk Būlus al-Ashkar (Paul Achkar, 1881–1962), who transcribed the music of the Maronite religious office. The first publication uniting notated sacred and secular music was the work of the priest Jirjī Ibrāhīm al-Rāhiba al-Dimashqī (1875–1920), in a collection of 24 pages entitled al-Rawd al-mustafīd (‘The garden of profits’), containing transcriptions of adwār.

In the first school of music in Cairo (founded by Mansūr ‘Awwād and Sāmī al-Shawwā in 1907) teaching concentrated mainly on Western notation (‘ilm al-nūta, the ‘science of notes’). Musical notation was soon faced with the problems of adaptation inherent in the quarter-tone and a symbol for an accidental to represent it. No one symbol has ever been generally adopted, but the symbols b, , and + have come into use to designate a half flat (nīm) and half sharp (tīk) (see ex.10). Similarly, the question of whether music should be notated for reading from right to left in the Arab way or left to right in the Western way was considered in the first quarter of the 20th century, and the Western way was generally adopted.

While the use of notation was an obvious sign of emancipation and progress in the first half of the 20th century, its significance was to develop during the second half of the century, when an increasingly large body of works recording the heritage of Arab music was assembled and published. It included several encyclopedic works transcribing muwashshahāt (Aleppo, 1955), or muwashshahāt and adwār (Cairo, 1959–63); the Tunisian nūba genre appeared under the title of al-Turāth al-mūsīqī al-tūnisī/Patrimoine musical tunisien in nine instalments from 1963 and various attempts to notate the 11 Moroccan examples of the nūba have been made between 1931 and the present day. In general these transcriptions do not take account of any variants, but offer a standard version of the works concerned and expect performers to follow it. However, interesting variations have appeared in transcriptions of the same piece by different hands; at this level, the authority of whoever is transcribing the piece is not questioned.

At the same time, notation has extended to song, and almost all the members of the firqa ensembles read printed music. These notations represent a ‘skeleton’ to which the interpreter can add ornamentation and creative touches. In 1983 a new system of notation that addressed ornamentation was put forward by the Egyptian Al-Hītamī (see ex.11). The proliferation of notated versions for performance of both the traditional and the modern repertory still continues to present problems in the field of interpretation, which is no longer determined by transmission of the style of a certain school or branch of a school.

7. Musical instruments.

(i) Pre-1918.

Early acoustic recordings from the beginning of the 20th century, such as those of Yūsuf al-Manyalawī and Salāma Hijāzī (see §5(ii)) give some idea of the role of the accompanying instrumental ensemble, providing rhythmic support, ostinato and drone accompaniment to vocal improvisations, gap-filling echo repetitions or variations, pre-composed but heterophonically realized concertante passages with or without the vocalist and, on occasion, a solo taqsīm improvisation. The late 19th-century Egyptian takht ensemble would typically consist of a small frame drum (Riqq) and two or three melody instruments chosen from the ‘ūd, violin, qānūn and nāy. Ensembles of similar composition (with spike fiddle for violin) have supported the singer for several centuries, but the various traditions have tended to make different choices between pairs of functionally equivalent instruments. In Persia and Iraq we find the santūr; in Turkey and the rest of the Arab world the qānūn. In Persia and Turkey long-necked lutes (Setār, Tār and tunbūr) predominate; in the Arab world the short-necked ‘ūd (and its North African variant, the kwītra). The traditional Iraqi ensemble includes a spike fiddle (jōza) to which correspond in Egypt the violin and in North Africa variously the violin, viola and boat-shaped fiddle (rabāb).

Unfortunately, hardly any specimens survive from before the 19th century, and the evolution and pre-modern geographical distribution of such instruments cannot be established with precision. In the absence of detailed written descriptions, it is only with the onset of an extensive iconography (the principal media being painting, ceramics, metalwork and ivory) that variants can begin to be distinguished.

The most frequently represented instrument is that most often described, the short-necked lute (‘ūd; fig.1). By the early Abbasid period its gut or silk strings, plucked with a plectrum, were arranged in four courses normally tuned in 4ths. Al-Kindī (c801–c866) provides details on materials, dimensions and construction, as does Ibn al-Tahhān (d after 1057). Although later writers generally withhold such information and do not discuss instrumental technique, it is clear that constant attempts were made to refine the instrument and improve its sonority. Its range was enlarged by the addition of a fifth and then later a sixth and seventh course, and thinner wood was introduced. In Timurid and Ottoman miniature paintings we encounter a variant (shahrūd) with an extremely large belly, but we do not know whether this was widely used in the Arab world.

With regard to other instruments, the rich variety of names indicates little more than the range available at various times and in various places. There is little description, and less to help pinpoint any technological changes that may have been introduced. Even al-Fārābī (d 950) was less concerned with the structure of instruments than with the scales that could be produced on them. Thus although membranophones and idiophones are mentioned in the introduction to his Kitāb al-mūsīqī al-kabīr, none is described in the section on instruments. The categories exemplified there are plucked chordophones (e.g. ‘ūd), aerophones (generically called mizmār, pl. mazāmīr) and bowed chordophones (Rabāb being the only example). A further classificatory distinction is maintained between instruments on which each note is produced by a separate open string (e.g. the harp, jank), and those on which the strings are stopped. Ibn Sīnā (Avicenna, d 1037) refined al-Fārābī's criteria in classifying both chordophones and aerophones (e.g. he notes the presence or absence of a reed). However, he too ignored membranophones. Safī al-Dīn (d 1294) pays even less attention to instruments; it is symptomatic that his comments in an early draft of the Kitāb al-adwār on the tuning of the jank and Qānūn (plucked zither) are omitted from the final version.

The most informative later texts are Persian: the anonymous 14th-century Kanz al-tuhaf and the Jāmi‘ al-alhān (‘Compendium of melodies’) by ‘Abd al-Qādir (d 1435). The Kanz al-tuhaf lists only nine instruments but describes them in more than usual detail, mentioning materials as well as proportions and construction. The Jāmi‘ al-alhān cites a large number of chordophones, but concentrates on defining the number of strings and the nature of the accordatura, aspects closely related to standard theoretical concerns. But ‘Abd al-Qādir’s list of instruments reflects the wide range of cultures he had encountered in Samarkand, and it is unclear how many of the entries in it were at all well-known in the Arab world. The summary that now follows is therefore tentative (and chronologically capricious).

Alongside the ‘ūd, plucked chordophones included the long-necked tunbūr, the popularity of which seems to have faded rapidly after its Abbasid heyday. It was first associated with Persia, and then with Anatolia, but its prominence in a later manifestation as a major instrument of Ottoman art music did not lead to its reintroduction in the Arab world (see §5 above): the long-necked buzuq of Syria/Lebanon is typologically different. Barbed lutes (Rabāb) seem to have flourished mainly in Persia and Central Asia, but might also have enjoyed brief favour in Iraq at least, as one is depicted on an Abbasid ceramic.

The jank is one of the earliest instruments to be mentioned and depicted, and in Timurid and Ottoman miniature paintings it frequently appears as an upper-chested harp with a compass of at least three octaves. By the 18th century, however, it was extinct, leaving the field among instruments with unstopped strings to the Santūr and qānūn. Both of these are trapezium box zithers, but of different shapes, the qānūn having one rectangular side. They also differ in playing technique, the santūr being hammered, the qānūn plucked with plectra inserted into a ring on each forefinger. By the 19th century the qānūn had some 24 triple courses of gut strings, but the flap bridges, which can be raised or lowered to effect changes of mode, are a more recent innovation. The santūr has metal strings, and its bridges rest directly on the wooden table, whereas the qānūn's single bridge has five feet resting on skin glued over an aperture in the table.

Bowed chordophones, previously termed rabāb, are of two quite distinct types. The more common was a spike fiddle called kamanja (see Kamāncheh) (from the Persian for ‘bow’), except in Iraq where it is currently called jōza (‘coconut’). It has horse-hair strings and a skin table glued over the resonator. It is held vertically and bowed like a viol, movement from string to string being effected by turning the cylindrical neck on its axis (in Morocco this technique is currently employed on the vertically held viola). The European viola and violin entered the Arab world during the 19th century, and the latter rapidly displaced the indigenous spike fiddle, except in Iraq. In North Africa, however, the result was coexistence with the other form of bowed chordophone, a boat-shaped fiddle with two strings termed rabāb, which in Morocco retains the vital role of ensemble leader projecting the core melody line.

From early references to aerophones it is not always clear whether the instrument had a reed, and if so, whether the reed was single or double. Reed aerophones (generically called mizmār) are sometimes mentioned in court contexts, possibly in connection with dancing, but they were generally used in outdoor performances, the shawm in particular being central to the military and ceremonial band. The Būq, a form of horn pipe in three sections (comprising mouthpiece, central section with finger-holes, and bell), was for a time an essential adjunct to court music in Spain, possibly serving to link it to a more popular repertory. The eventual exclusion of reed instruments from art music ensembles has left as the sole possible aerophone the rim-blown obliquely held flute (nāy, from Persian Ney: ‘bamboo’). It consists of a section of bamboo with three or four nodes into which are cut six finger-holes and one thumb-hole at the back. The nāy is important, both symbolically and in practice, for many Sufis, and became particularly associated with the Mevlevi order founded by Jalāl al-Dīn Rūmī, but it is still only sporadically present in the art music domain.

Idiophones are rarely mentioned in writings. They essentially occur only in the form of castanets and finger cymbals used by dancers. Among membranophones the frame drum (duff) is the instrument most frequently cited and depicted. In Timurid and Ottoman miniature paintings of the 15th century to the 18th it is normally held vertically, face high, and is depicted with five sets of jingles inserted into the frame. The single-headed goblet drum Darabukka is a relative newcomer in art music contexts.

The distribution of instruments by gender has varied considerably over the centuries. During the Abbasid period we read of many female entertainers (see Qayna) who were celebrated performers, especially on long- and short-necked lutes. In Timurid and Ottoman miniature paintings of the 15th to 18th centuries the drummer and the harpist are frequently female. But the harp became obsolete, and it is conceivable that the development of a slightly smaller form of frame drum with a new name, riqq (and often with rich mother-of-pearl inlay), was associated with its increasing use in the all-male 19th-century takht ensemble. Whereas women in the hareem might have been familiar with most of the main art music instruments, as public performers female instrumentalists have until recently had only a restricted range of percussion instruments available to them.

(ii) Post-1918.

Classical instruments have undergone technical modernization. The replacement of gut strings by metal or nylon strings is becoming standard, and among other innovations are the adoption from the West of certain materials that are more durable than their predecessors, for instance the skins for certain drums: the darbukka, a goblet drum, and frame drums such as the riqq and bandīr, now have stronger heads, but as they are often in man-made materials the traditional sound is not the same. The Qānūn had a system of brass levers added, which liberated the left hand and allowed playing in octave, transforming the aesthetic of the instrument.

The nay reed flute of the Mevlevi dervishes is now in the public domain. The first musician unconnected with the dervishes to introduce this change was the Egyptian flute-player Amīn al-Būzarī (c1855–1935). The nay has also been introduced into countries where it was not previously established, such as Morocco and Tunisia, where it has been very successful since 1933, thanks to the Syrian player ‘Alī Darwīsh who taught the most famous Tunisian exponent, of the instrument, Salāh al-Mahdī (b 1925), as well as his pupil Muhammad Saada (b 1938). There have been attempts to modernize the reed flute by adding keys, but they were abandoned.

The last decades of the 20th century have seen the disappearance of some of the traditional instruments of art music, such as the ‘ūd ‘arbī or Arab ‘ūd with four double strings, previously known in Morocco as ‘ūd ramāl Elsewhere it has been replaced by the Egyptian ‘ūd, which has five double strings and is considered a more flexible and sophisticated instrument in performance. Finally, the electronic amplification of instruments is becoming increasingly common, and musicians are now often incapable of playing without amplified sound, producing multiple echoes that provide a kind of ornamentation.

The 20th century saw the invention of new instruments such as the eastern piano, on which work began in Egypt in 1914, spurred by moves to introduce harmonies into what had always been monodic or heterophonic music. Their construction aimed to solve the problems involved in the quarter-tone controversy of the time. However, they were abandoned because the instrument failed to resolve the technical problem of the transposition of Arab modes, and its tone colour did not really seem to suit the spirit of Arab music. An eastern trumpet (with an extra valve) was made by the Lebanese Nasīm Maalūf (b 1941) in 1982, but it was not widely adopted.

While the end of the 19th century saw the successful incursion of the violin into Arab music, replacing the tentative use of the viola d'amore in previous centuries, there was large-scale introduction of European instruments, including the cello and double bass, from the second quarter of the 20th century onwards. As important as the introduction of these was that of the accordion, a key instrument in Lebanese and Sudanese music since 1960. Other instruments such as bongo drums were introduced when Afro-Cuban music became fashionable. In the second half of the 20th century the saxophone was introduced into Sudan. Less frequently used is the oboe, for instance in the ensemble of the Lebanese Marcel Khalīfé (b 1951). The electric guitar and synthesizer are also played in light music ensembles. This introduction of Western instruments and the concept of the instrumental desk, also borrowed from the West, has led to the formation of the modern Arab firqa orchestra. Such instruments also increase the tessitura of the music, which was formerly confined to three octaves but can now extend to four or five. The idea of intensity and nuances has also been introduced, as well as that of recitals for solo instruments (the ‘ūd, buzuq, nay and qānūn). In North Africa the European piano, the saxophone and less commonly the clarinet have become part of Arabo-Andalusian orchestras together with the violin and viola, that had already been adopted at the end of the 19th century. Under the name of snitra the mandolin has been introduced into the Andalusian-Algerian orchestra.

New instrumental tutors have been written, not simply in response to the perceived need for musical notation, but also as systematic teaching manuals based on exercises, a novel concept for Arab music. The first of these tutors, by Muhammad Dhākir Bey, was for the ‘ūd lute and was published in Cairo in 1903. Next came the first tutors for the qānūn board zither by Mustafā Ridā and Mahmūd Ahmad al-Hifnī (Mahmoud Ahmad El Hefny; Cairo, c1935), for violin by ‘Abdul Halīm ‘Alī (Cairo, 1939) and for nay by Mahmūd ‘Iffat (Cairo, 1968). Several accordion tutors have come on the Arab market since 1980. Tutors have also been published for harp (Cairo), piano (Cairo), saxophone (Casablanca) and darbukka (Aix-en-Provence, France). Some of these tutors promote a clear distinction between teacher and pupil, whereas their close relationship was the basis of traditional teaching.

b: sources in arabic, pre-1000 ce

BIBLIOGRAPHY

a: bibliographies

c: sources in arabic, 1000–1900

d: general works

e: history

f: the arabs and europe

g: theory

h: modes, modality and rhythm

m: notations and transcriptions

n: modernism and tradition

a: bibliographies

H.G Farmer: ‘Some Musical Manuscripts Identified’, Journal of the Royal Asiatic Society (1926), 91–3

H.G. Farmer: The Sources of Arabian Music (Bearsden, 1940, 2/1965)

H.G. Farmer: ‘“Ghosts”: an Excursus on Arabic Musical Bibliographies’, Isis, xxxvi (1945–6), 123–30

‘A. al-‘Alwajī: Rā’id al-mūsīqā al-‘arabiyya [Guide to Arab music] (Baghdad, 1964)

F. Sayyid: Fihris al-makhtūtāt al-musawwara, juz’ 4: al-ma‘ārif al-‘āmma wa-’l-funūn al-mutanawwi‘a (Baghdad, 1964)

Z. Yūsuf: Makhtūtāt al-mūsīqā al-‘arabiyya fī ’l-‘ālam (Baghdad, 1966–7)

‘A. al-‘Alwajī: Al-murshid ilā al-nitāj al-musīqī (Baghdad, 1975)

U.N. Al-Naqshabandī: Makhtūtāt al-mūsīqā wa-’l-ghinā’ wa-’l-samā‘ fī maktabat al-mathaf al-‘irāqī (Baghdad, 1979)

A. Shiloah: The Theory of Music in Arabic Writings (c. 900–1900) (Munich, 1979)

W.J. Krüger-Wust: Arabische Musik in europäischen Sprachen (Wiesbaden, 1983)

U.N. Al-Naqshabandī: ‘Makhtūtāt al-mūsīqā wa-’l-ghinā’ al-musawwara fī qism al-makhtūtāt bi-’l-mu’assasa al-‘āmma li-l-āthār wa-’l-turāth-Baġdād’, al-Mawrid, xiii/4 (1984), 117–26

E. Neubauer: ‘Henry George Farmer on Oriental Music: an Annotated Bibliography’, Zeitschrift für Geschichte der Arabisch-Islamischen Wissenschaften, iv (1987–8), 219–66

E. Neubauer: ‘Manuscrits de musique arabe: enregistrement et catalogue depuis le congrès du Caire’, Musique arabe: le congrès du Caire de 1932 (Cairo, 1992)

b: sources in arabic, pre-1000 ce

al-Jāhiz [d 869]: Risālat al-qiyān; ed. and trans. A.F.L. Beeston: The Epistle on Singing-Girls by Jāhiz (Warminster, 1980)

Abū Zayd Hunayn Ibn Ishāq [d 873]: Kitāb ādāb al-falāsifa [Sayings of the philosophers] (MS, GB-Lbl Or.8681, ff.47b–70b)

al-Kindī [d c866]: Mukhtasar al-mūsīqī fī ta’līf al-nagham wa-san‘at al-‘ūd [Summary on music with reference to the composition of melodies and lute making]; Fr. trans. A. Shiloah: ‘Un ancien traité sur le ‘ūd d'Abū Yūsuf al-Kindī’, Israel Oriental Studies, iv (1974), 179–205

al-Kindī: Risāla fī ajzā’ khabariyya fī ’l-mūsīqī [Treatise on informative sections on the theory of music] (MS, D-BSb 5503); trans. H.G. Farmer: Transactions of the Glasgow University Oriental Society, xvi (1957); ed. Z. Yūsuf: Mu’allafāt al-Kindī al-mūsīqiyya [The musical works of al-Kindī], (Baghdad, 1962)

al-Kindī: Risāla fī khubr sinā‘at al-ta’līf [Treatise concerning the inner knowledge of the art of composition] (MS, GB-Lbl Or.2361, ff.165–8); Eng. trans. C. Cowl in The Consort, xxiii (1966), 129–66; ed., with Eng. trans., Y. Shawqī (Cairo, 1969)

Abu Tālib al-Mufaddal ibn Salama [d c902]: Kitāb al-malāhī [Book on instruments]; ed. ‘A. al-‘Azzāwī in Al-mūsīqā al-‘irāqiyya [Iraqi music] (Baghdad, 1951), 73–89; trans. J. Robson in Journal of the Royal Asiatic Society (1938), 231–49

Ibn Khurradādhbih [d c912]: Mukhtār min kitāb al-lahw wa-’l-malāhī (Beirut, 1961); ed. ‘A. al-‘Azzāwī in Al-mūsīqā al-‘irāqiyya [Iraqi music] (Baghdad, 1951), 91–101

al-Munajjim [d 912]: Risāla fī ’l-mūsīqī [Treatise on music] (MS, GB-Lbl Or.2361); ed. Z. Yūsuf (Baghdad, 1964); Ger. trans. in E. Neubauer: ‘Die acht “Wege” der Musiklehre und der Oktoechos’, Zeitschrift für Geschichte der Arabisch-Islamischen Wissenschaften, ix (1994), 373–414

Ibn ‘Abd Rabbihī [d 940]: Al-‘iqd al-farīd; partial Eng. trans. in H.G. Farmer: ‘Music: the Priceless Jewel’, Journal of the Royal Asiatic Society (1941), 22–30, 127–44

al-Fārābī [d 950]: Kitāb al-mūsīqī al-kabīr [Great book on music] (MS, NL-Lu 651); ed. G.A.M. Khashaba (Cairo, 1967); Fr. trans. in R. d'Erlanger: La musique arabe, i (Paris, 1930), 1–306; ii (1935), 1–101

al-Fārābī: Kitāb ihsā’ al-‘ulūm [Book of the classification of the sciences] (MS, E-E 646); ed., with Sp. trans., A. González Palencia: Catálogo de las ciencias (Madrid, 1932, 2/1953)

al-Fārābī: Kitāb al-īqā‘āt [Book of rhythms]; Ger. trans. in E. Neubauer: ‘Die Theorie vom Īqā‘, i: Übersetzung des Kitāb al-īqā‘āt von Abū Nasr al-Fārābī’, Oriens, xxi–xxii (1968–9), 196–232

al-Fārābī: Kitāb ihsā’ al-īqā‘āt [Book of the comprehension of rhythms] (Manisa MS 1705, ff.59a–81b, 88a–89b); Ger. trans. in E. Neubauer: ‘Die Theorie vom Īqā‘, ii: Übersetzung des Kitāb Ihsā’ al-īqā‘āt von Abū Nasr al-Fārābī’, Oriens, xxxiv (1994), 103–73

al-Isfahānī [d 967]: Kitāb al-aghānī [Book of songs]; pubd (Cairo, 1868–9, suppl., Leiden, 1888, 2/1905–6, 3/1927–74); partial Fr. trans. in E.M. Quatremère: ‘Mémoire sur l'ouvrage intitulé Kitab al-agâni, c'est-à-dire “Recueil de chansons”’, Nouveau journal asiatique, 2nd ser., xvi (1835), 385–419, 497–545; Journal asiatique, 3rd ser., vi (1838), 465–526

al-Khwārazmī [d 997]: Mafātīh al-‘ulūm [Keys of the sciences]; ed. van Vloten (Leiden, 1895)

Ikhwān al-Safā’ [Brethren of Purity; 10th century]: Rasā’il [The epistles] (MS, Gb-Ob Hunt 296); Ger. trans. in F. Dieterici: Die Propaedeutik der Araber im zehnten Jahrhundert (Berlin, 1865/R), 100–153; Fr. trans. with commentary by A. Shiloah: ‘L'épître sur la musique des Ikhwān al-Safa’’, Revue des études islamiques, xxxii (1964), 125–62; xxxiv (1966), 159–93

c: sources in arabic, 1000–1900

al-Hasan ibn Ahmad [fl late 10th–early 11th century]: Kamāl adab al-ghinā’ [The perfection of musical knowledge], ed. G. Khashaba (Cairo, 1975); ed. Z. Yūsuf in al-Mawrid, ii/2 (1973), 101–54; Fr. trans. A. Shiloah: La perfection des connaissances musicales (Paris, 1972)

Ibn Sīnā [Avicenna; d 1037]: Kitāb al-shifā’ [The book of healing] (MS, GB-Lbl Or.11190); Fr. trans. in R. d'Erlanger: La musique arabe, ii (Paris, 1935), 105–245; ed. Z. Yūsuf: Kitāb al-shifā’, al-riyādiyyāt 3: jawāmi‘ ‘ilm al-mūsīqī (Cairo, 1956)

Ibn Zayla [d 1048]: Kitāb al-kāfī fī ’l-mūsīqī [Book of sufficiency concerning music]; ed. Z. Yūsuf (Cairo, 1964)

Ibn al-Tahhān [fl 1st half 11th century]: Hāwī al-funūn wa-salwat al mahzūn [Compendium of the arts to comfort sad hearts] (MS, ET-Cn Dār al-kutub, funūn jamīla 539); facsimile edn., Publications of the Institute for the History of Arabic-Islamic Science, ser.C, no.52 (Frankfurt, 1990) [with introduction by E. Neubauer]

Muhammad ibn ‘Alī al-Hindī [fl 1135]: Jumal al-falsafa [On philosophy]; facs. edn. in Publications of the Institute for the History of Arabic-Islamic Science, ser.C, no.19 (Frankfurt, 1990) [music section: pp.114–23]

Ibn Bājja [Avenpace; d c1139]: Kitāb fī al-nafs [Book on the soul] (MS, D-Bsb 5061)

Ibn Rushd [Averroes; d 1198]: Sharh fī l-nafs li-Aristūtālis [Commentary on Aristotle's De anima]; ed. N. Morata: El compendio de anima (Madrid, 1934)

A. al-Tīfāshī [d 1253]: Mut‘at al-asmā ‘fī ‘ilm al-samā‘ [On the science of listening to music], in M. ibn T. al-Tanjī: ‘al-Tarā’iq wa-’l-alhān al-mūsīqiyya fī ifrīqiya wa-’l-andalus’, al-Abhāth: Quarterly Journal of the American University of Beirut, xxi (1968), 93–116

Nasīr al-Dīn al-Tūsī [d 1273–4]: Risāla fī ‘ilm al-mūsīqī [Treatise on the science of music]; ed. Z. Yūsuf (Cairo, 1964)

Safī al-Dīn al Urmawī [d 1294]: Kitāb al-adwār [Book of cycles]; ed. al-Rajab (Baghdad, 1980); ed. Khashaba and al-Hifnī (Cairo, 1986); facsimile in Publications of the Institute for the History of Arab-Islamic Science, ser.C, xxix (Frankfurt, 1986); Fr. trans. in R. d'Erlanger: La musique arabe, iii (Paris, 1938), 185–565

Safī al-Dīn al Urmawī: Risāla al-sharafiyya [The Sharafian treatise]; ed. al-Rajab (Baghdad, 1982); facsimile in Publications of the Institute for the History of Arabic Islamic Science, ser.C, xxix (Frankfurt, 1986); Fr. trans. in R. d'Erlanger: La musique arabe, iii (Paris, 1938), 3–182

Qutb al-Dīn [d 1311]: Durrat al-tāj [Pearl of the crown] (MS, GB-Lbl Add.7694) [in Persian]; ed. S.M. Mashkūt and N.A. Taqwā (Tehran, 1939–46)

Badr al-Dīn Muhammad al-Irbilī: Urjūzat al-anghām (1328); ed. in Al-mashriq, xvi (1913) 895–901; ed. ‘A. al-‘Azzāwī: Al-mūsīqā al-‘irāqiyya [Iraqi music] (Baghdad, 1951), 103–13

Ibn Kurr [d 1341]: Ghāyat al-matlūb fī ‘ilm al-anghām wa-'l-durūb (MS, GB-Lbl Or.9247)

Muhammad Ibn ‘Alī al-Bilbaysī [d 1348]: Al-mulah wa-l-turaf min munādamat arbāb al-hiraf; ed. (Cairo, 1866)

Shams al-Dīn al-Afkānī [d 1348]: Irshād al-qāsid ilā asnā al-maqāsid; ed. Sprenger: Biblioteca Indica (Calcutta, 1848), 92–4; (Cairo, 1900); Ger. trans. in Weidemann: Aufsätze, 589–95; Fr. trans. in A. Shiloah: ‘Deux textes arabes inédits sur la musique’, Yuval, no.1, (1968), 221–48

Shams al-Dīn al-Afkānī: Al-durr al-nazīm (fasc 57, Vienna 2, ff.42–4); ed., with Eng. trans., Gottheil: Jewish Quarterly Review, xxiii (1932), 164–80

Shihāb al-Dīn al-‘Umarī [d 1349]: Masālik al-absār fī mamālik al-amsār, vol.x, facsimile edn., Publications of the Institute for the History of Arabic-Islamic Science, ser.C, nos.46, 10, (Frankfurt, 1988)

Sharh mawlānā mubārak shāh bar adwār [The Mubārak Shāh commentary on the Kitāb al-adwār; written 1375] (MS, GB-Lbl Or.2361 ff.68b–153); Fr. trans. in R. d'Erlanger: La musique arabe, ii (Paris, 1938), 185–565

‘Abd al-Qādir [d 1435]: Jāmi‘ al-alhān [Compendium of melodies] (MS, GB-Ob Marsh 282; TR-Ino 3644) [in Persian]; ed. T. Bīnish (Tehran, 1987)

‘Abd al-Qādir: Maqāsid al-alhān [Purports of melodies] (MS, GB-Ob Ouseley 264, 385) [in Persian]; ed. T. Bīnish (Tehran, 1966, 2/1977)

Shams al-Dīn al-Saydāwi [fl 15th century]: Kitāb al-in‘ām bi-ma‘rifat al-anghām (MSS GB-Ob Marsh 82; Lbl Or.13019)

Fath Allāh al-Shirwānī [d c1453]: Majalla fī 'l mūsīqī [Codex on music] (first version MS, Istanbul Topkapi Ahmet III 3449); facsimile edn., Publications of the Institute for the History of Arabic-Islamic Science, ser.C, no.29 (Frankfurt, 1986) [with introduction by E. Neubauer]; (second version MS, GB-Lbl Or.2361, ff.168b–219b); Fr. trans. in R. d'Erlanger: La musique arabe, iv (Paris, 1939), 3–255

Muhammad al-Lādhiqī [fl late 15th century]: Al-risāla al-fathiyya fī ’l-mūsīqī [The victory treatise concerning the theory of music] (MS, GB-Lbl Or.6629; National Library, Cairo, f.j.7); Fr. trans. in R. d'Erlanger: La musique arabe, iv (Paris, 1939), 257–498

Muhammad al-Lādhiqī: Zayn al-alhān fī ‘ilm ta’līf al-awzān [The adorning of melodies in the composition of the measures] (MS, TR-Ino 3655)

Ahmad al-Mawsilī: Al-durr al-naqī fī ‘ilm al-mūsīqī [The science of music]; ed. J. al-Hifnī (Baghdad, 1964)

‘Abd al-Wāhid al-Wansharīsī [d 1549]: Fī 'l-tabā'i‘ wa-'l-tubū‘ wa-'l-usūl; Eng. trans. by H.G. Farmer: An Old Moorish Lute Tutor (Glasgow, 1933)

Shihāb al-Dīn al-‘Ajamī: Risāla fī ‘ilm al-anghām [Treatise on the science of melodies] (D-Bsb 5534; L-B (i) 220, 121–38; F-Pn 2865, ff.70–75v)

Abū Zayd al-Fāsī [d 1685]: Al-jumū‘ fī ‘ilm al-mūsīqī wa-'l-tubū‘; ed. H.G. Farmer: Collection of Oriental Writers in Music, i (Glasgow, 1933)

Al-shajara dhāt al akmām [Tree with calyxes]; ed. G.‘A. al-M. Khashaba and I. Fath Allāh (Cairo, 1983)

al-Hā’ik [c1800]: Majmū‘at al-nawbāt (MS, GB-Lbl Or.13235)

‘Askar al-Qādirī [fl 18th century]: Rāh al-jām fī shajarāt al-anghām (MS, D-GOl 1351)

Muhammad al-‘Attār [d 1828]: Rannat al-awtār fī jadāwil al-afkār fī fann al-mūsīqār (MS, US-PRu Yahuda 3233, ff.181v–187r)

Shihāb al-Dīn al-Hijāzī [d 1857]: Safīnat al-mulk wa nafīsat al-fulk; ed. (Cairo, 1892–3)

Mīkhā’īl Mushāqa [d 1888]: Al-risāla al-shihābiyya fī 'l-sinā‘a al-mūsīqiyya, Al-mashriq, ii, 1899; ed., with Fr. trans., P.L. Ronzevalle: ‘Un traité de musique arabe moderne’, Mélanges de la faculté orientale, Université Saint-Joseph (Beyrouth), vi, (1913), 1–120

‘Uthmān Ibn Muhammad al-Jundī [fl 19th century]: Rawd al-masarrāt fī ‘ilm al-naghamāt (MS, GB-LEbc 154)

d: general works

K. al-Khula‘ī: Kitāb al-mūsīqā al-sharqiyya [The book of eastern music] (Cairo, 1905–6/R)

H.H. Abdul-Wahab: ‘Le développement de la musique arabe en Orient, Espagne et Tunisie’, Revue tunisienne, xxv (1918), 106–17

H.G. Farmer: ‘Music’, The Legacy of Islam, ed. T. Arnold and A. Guillaume (London, 1931)

Q. Rizq: Al-mūsīqā al-sharqiyya wa-al-ghinā‘ al-‘arabī [Eastern music and Arab art singing] (Cairo, 1936–47/R)

H.G. Farmer: ‘The Music of Islam’, NOHM, i (1957), 421–77

‘A.al-K. ‘Allāf: Al-tarab ‘ind al-‘arab [Delighted pleasure among the Arabs] (Baghdad, 1963)

H.‘A. Mahfūz: Mu‘jam al-mūsīqā al-‘arabiyya [Dictionary of Arab music terms] (Baghdad, 1964)

S.M. Stern: ‘Andalusian muwashshahs in the Musical Repertoire of North Africa’, Actas del primer congreso de estudias Árabes e Islámicos: Córdoba 1962, ed. F.M. Pareja Maestre (Madrid, 1964), 319–27

H.G. Farmer: ‘Music’, A History of Muslim Philosophy, ii, ed. M.M. Sharif (Wiesbaden, 1966), 1124–78

A.S. Abū ‘Awf: Adwā’ ‘alā al-mūsīqā al-‘arabiyya [Light on Arab music] (Cairo, c1966)

M. ibn T. al-Tanjī: ‘Al-tarā’iq wa-’l-alhān al-mūsīqiyya fī ifrīqiya wa-’l-andalus’, Al-abhāth: Quarterly Journal of the American University of Beirut, xxi (1968), 93–116

M.A. al-Hifnī: Turāthunā al-mūsīqī (Cairo, 1969)

I. Sharrādī, ed.: Al-mu’tamar al-thānī lil-mūsīqā al-arabiyya [The 2nd conference on Arab music] (Tangiers, 1969)

H. Hickmann: ‘Die Musik des arabisch-islamischen Bereichs’, Orientalische Musik, H. Hickmann and W. Stauder (Leiden, 1970), 1–134

S. Jargy: La musique arabe (Paris, 1972, 3/1988)

E. Neubauer: ‘Tarannum und Terennüm in Poesie und Musik’, Mélanges de l'Université Saint-Joseph, xlviii (1973–4), 139–53

H.G. Farmer: Islam, Musikgeschichte in Bildern, iii/2 (Leipzig, 1976)

H.’A. Mahfūz: Qāmūs al-mūsīqā al-‘arabiyya (Baghdad, 1977)

S. al-Mahdī: Al mūsīqā al-‘arabiyya: tārīkhuhā wa-adabuhā [Arab music: history and literature] (Tunis, 1979)

M. Guettat: La musique classique du Maghreb (Paris, 1980, 2/1999)

L.I. al-Faruqi: An Annotated Glossary of Arabic Musical Terms (Westport, CT, 1981)

H.G. Farmer: Studies in Oriental Music (Frankfurt, 1986)

B.M. Liu and J.T. Monroe: Ten Hispano-Arabic Strophic Songs in the Modern Aural Tradition (Berkeley, 1989)

I. Eolian: Tradičionnaja muzyka arabskogo vostoka [Traditional Arab music of the Middle East] (Moscow, 1990)

S.Q. Hassan, ed.: Mūsīqā al-madīna [Urban music] (Beirut, 1991)

A. Shiloah: ‘La voix et les techniques vocales chez les arabes’, Cahiers de musiques traditionnelles, iv (1991), 85–101

J.S. Pacholczyk: ‘Towards a Comparative Study of a Suite Tradition in the Islamic Near East and Central Asia: Kashmir and Morocco’, Regionale maqam-Traditionen: Gosen, nr Berlin 1992, 429–63

A. and M. Hachlaf: Anthologie de la musique arabe (1906–1960) (Paris, 1993)

A. Shiloah: ‘Notions d'esthétique dans les traités arabes sur la musique’, Cahiers de musique traditionnelles, vii (1994), 51–74

A. Shiloah: Music in the World of Islam (Aldershot, 1995)

C. Poché: Musiques du monde arabe: écoute et découverte (Paris, 1996)

H.H. Touma: The Music of the Arabs (Portland, OR, 1996)

H.H. Touma: Die Nūbah Māyah: zur Phänomenologie des Melos in der arabisch-andalusi Musik Morokkos (Hildesheim, 1998)

e: history

A.P. Caussin de Perceval: ‘Notices anecdotiques sur les principaux musiciens arabes des trois premiers siècles de l'Islamisme’, Journal asiatique, 7th ser., ii (1873), 546–92

H.G. Farmer: A History of Arabian Music to the XIIIth Century (London, 1929/R)

H.G. Farmer: ‘Maimonides On Listening To Music’, Journal of the Royal Asiatic Society (1933), 867–84

H.G. Farmer: The Minstrelsy of ‘The Arabian Nights’: a Study of Music and Musicians in the Arabic ‘Alf Laila wa Laila’ (Bearsden, 1945)

M. Barkechli: L'art sassanide base de la musique arabe (Tehran, 1947)

‘A. al-‘Azzāwī: Al-mūsīqā al-‘irāqiyya fī ‘ahd al-mughūl wa-'l-turkumān [Iraqi music in the period of the Mongols and Turkmen] (Baghdad, 1951)

S.M. Imamuddin: ‘Music in Muslim Spain’, Islamic Culture, xxxiii (1959), 147–50

H.G. Farmer: ‘Tenth-Century Arabic Books on Music: as Contained in “Kitāb al- Fihrist” of Abu’l-Faraj Muhammad ibn al-Nadīm’, Annual of Leeds University Oriental Society, ii (1959–61), 37–47

A. Shiloah: ‘Réflexions sur la danse artistique musulmane au Moyen-Age’, Cahiers de civilisation médiévale, v (1962), 463–74

I. ‘Abbās: ‘Akhbār al-ghinā’ wa-’l-mughannīn fī ’l-andalus’ [Music and musicians in Andalusia], Al-abhāth: Quarterly Journal of the American University of Beirut, xvi (1963), 3–22

A. Shiloah: Caractéristiques de l'art vocal arabe au Moyen-Âge (Tel-Aviv, 1963)

E. Neubauer: ‘Eine musikalische Soirée am Hof von Hārūn ar-Rashīd’, Bustan, x/1 (1969), 27–33

‘A. ‘Alī al-Hajjī: Tārīkh al-mūsīqā al-andalusiyya [History of Andalusian music] (Beirut, 1969–70)

M.M.S Hāfiz: Tārīkh al-mūsīqā wa-’l-ghinā’ al-‘arabī [History of Arab music and song] (Cairo, 1971)

H.H. Touma: ‘Die Musik der Araber im 19. Jahrhundert’, Musikkulturen Asiens, Afrikas und Ozeaniens im 19. Jahrhundert, ed. R. Günther (Regensburg, 1973), 49–71

L. Manik: ‘Zwei Fassungen einer von Safī al-Dīn notierten Melodie’, Baessler-Archiv, new ser., xxiii (1975), 145–51

H.H. Touma: ‘History of Arabian Music: a Study’, World of Music, xxii/3 (1980), 66–75

A.J. Racy: ‘Music in Nineteenth-Century Egypt: an Historical Sketch’, Selected Reports in Ethnomusicology, iv (1983), 157–79

G. Sawa: ‘Musical Humour in the Kitāb al-Aghānī (Book of Songs)’, Logos islamikos: studia islamica in honorem Georgii Michaelis Wickens, ed. R.M. Savory and D. Agius (1984), 35–50

D. Granit: ‘The Music Paintings of the Capella Palatina in Palermo’, Imago musicae, ii (1985), 9–49

A. Shiloah: ‘Music in the Pre-Islamic Period as Reflected in Arabic Writings of the First Islamic Centuries’, Jerusalem Studies in Arabic and Islam, vii (1986), 109–20

G. Sawa: Music Performance Practice in the Early ‘Abbāsid Era (Toronto, 1989)

E. Neubauer: ‘Zur Bedeutung der Begriffe Komponist und Komposition in der Musikgeschichte der islamischen Welt’, Zeitschrift für Geschichte der Arabich-Islamischen Wissenschaften, xi (1997), 307–63

f: the arabs and europe

J. Ribera: La música de las Cantigas (Madrid, 1922; Eng. trans., abridged, 1929/R as Music in Ancient Arabia and Spain)

H.G. Farmer: ‘Clues for the Arabian Influence on European Musical Theory’, Journal of the Royal Asiatic Society (1925), 61ff; pub'd separately as The Arabian Influence on Musical Theory (London, 1925)

K. Schlesinger: Is European Musical Theory Indebted to the Arabs? (London, 1925)

J. Ribera y Tarragó: Historia de la música árabe medieval y su influencia en la española (Madrid, 1927/R; Eng. trans. abridged, 1929)

H.G. Farmer: Historical Facts for the Arabian Musical Influence (London, 1930/R)

H.G. Farmer: ‘The Influence of al-Fārībī's “Ihsa’ al-‘ulum” (De scientiis) on the Writers on Music in Western Europe’, Journal of the Royal Asiatic Society (1932), 561–92

H.G. Farmer: ‘A Further Arabic-Latin Writing on Music’, Journal of the Royal Asiatic Society (1933), 307–22

H.G. Farmer: Al-Fārābī's Arabic-Latin Writings on Music (Glasgow, 1934/R)

O. Ursprung: ‘Um die Frage nach dem arabischen Einfluss auf die abendländische Musik des Mittelalters’, Zeitschrift für Musikwissenschaft, xvi (1934), 129–41

M. Schneider: ‘A propósito del influjo árabe: ensayo de etnografía musical de la España medieval’, AnM, i (1946), 31–141

H. Anglès: La música de las Cantigas de Santa María del rey Alfonso el Sabio (Barcelona, 1958)

H.G. Farmer: ‘The Oriental Impingement on European Music’, Islamic Studies, ii (1963), 337–42

E.A. Bowles: ‘Eastern Influences on the Use of Trumpets and Drums in the Middle Ages’, AnM, xxvi (1971), 1–28

E.R. Perkuhn: ‘Die arabische Theorie und die Ursprungsfrage der Troubadourkunst’, SMH, xv (1973), 129–39

E. Neubauer: ‘Zur Rolle der Araber in der Musikgeschichte des europäischen Mittelalters’, Islam und Abendland, ed. A. Mercier (Berne, 1976), 243–66

E.R. Perkuhn: Die Theorien zum arabischen Einfluss auf die europäische Musik des Mittelalters (Walldorf-Hessen, 1976)

D.M. Randel: ‘Al-Farabi and the Role of Arabic Music Theory in the Latin Middle Ages’, JAMS, xxix (1976), 173–88

‘A. al-Jirārī: Athar al-andalus ‘alā ūrubbā fī majāl al-nagham wa-'l-īqā‘ (Rabat, 1982)

J.S. Pacholczyk: ‘The Relationship Between the Nawba of Morocco and the Music of the Troubadours and Trouvères’, World of Music, xxv/2 (1983), 5–16

I.J. Katz: ‘Higinio Angles and the Melodic Origins of the “Cantigas de Santa Maria”: a Critical View’, Alfonso X of Castile the Learned King: Cambridge, MA, 1984, 46–75

M. Haas: ‘Arabische und lateinische Musiklehre: ein Vergleich von Strukturen’, Miscellanea mediaevalia 17: orientalische Kultur und europäisches Mittelalter, ed. A. Zimmermann (Berlin, 1985), 358–75

C. Burnett: ‘The Use of Geometric Terms in Medieval Music: Elmuahim and Elmuarifa and the Anonymous IV’, Sudhoffs Archiv, lxx/2, (1986), 198–205

J.T. Monroe: ‘Poetic Quotation in the muwaššaha and its Implications: Andalusian Strophic Poetry as Song’, La corónica, xiv/2 (1986), 230–50

R. Alvarez: ‘Los instrumentos musicales en los codices Alfonsinos: su tipologia, su uso y su origen: algunos problemas iconograficos’, RdMc, x (1987), 1–38

J.T. Monroe: ‘The Tune or the Words? (Singing Hispano-Arabic Strophic Poetry)’, al-Qantara, viii (1987), 265–317

G. Braune: ‘Musik in Orient und Okzident’, Europa und der Orient, 800–1900: eine Ausstellung des 4. Festivals der Weltkulturen Horizonte, '89 (Berlin 1989), 210–30

C. Burnett: ‘Teoria e practica musicali arabe in Sicilia e nell'Italia meridionale in età normanna e sveva’, Nuone effemeridi, xi (1990), 79–89

A. Shiloah: ‘The Meeting of Christian, Jewish and Muslim Musical Cultures on the Iberian Peninsula (before 1492)’, AcM, lxiv (1991), 14–20

C. Burnett: ‘European Knowledge of Arabic Texts Referring to Music: Some New Material’, EMH, xii (1993), 1–17

D. Wulstan: ‘Boys, Women and Drunkards: Hispano-Mauresque Influences on European Song?’, The Arab Influence on Medieval Europe, ed. D. Agius and R. Hitchcock (Reading, 1993), 136–67

g: theory

M. Collangettes: ‘Etude sur la musique arabe’, Journal asiatique, 10th ser., iv and viii (1904), 365–422; (1906), 149–90

E. Arian: ‘Preuve irréfutable de la division de l'échelle musicale orientale en 24 quarts de ton’, Bulletin de l'Institut d'Egypte, vi (1924), 159–67

H.G. Farmer: ‘The Influence of Music: from Arabic Sources’, PMA, lii (Madrid, 1925–6), 89–114

H.G. Farmer: ‘Greek Theorists of Music in Arabic Translation’, Isis, xiii (1929–30), 325–33

R. d'Erlanger: La musique arabe (Paris, 1930–59)

R. Lachmann and M. El Hefny: ‘Ja‘qūb Ibn Ishāq al-Kindī’, Veröffentlichungen der Gesellschaft zur Erforschung der Musik des Orients, i (Leipzig, 1931)

E.A. Beichert: ‘Die Wissenschaft der Musik bei al-Fārābī’, Kirchenmusikalisches Jahrbuch, xxvii (1932), 9–48

H.G. Farmer: ‘The ‘Ihsā’ al-‘ulūm’’, Journal of the Royal Asiatic Society (1933), 906–09

H.G. Farmer: ‘The Lute Scale of Avicenna’, Journal of the Royal Asiatic Society (1937), 245–57

E. Werner and I. Sonne: ‘The Philosophy and Theory of Music in Judeo-Arabic Literature’, Hebrew Union College Annual, xvi (1941), 251–319; xvii (1942–3), 511–72

H.G. Farmer: Sa‘adya Gaon on the Influence of Music (London, 1943)

H.G. Farmer: ‘An Anonymous English-Arabic Fragment on Music’, Islamic Culture, xviii (1944), 201–05

M.K. Allāh Wirdī: Falsafat al-mūsīqā al-sharqiyya fī asrār al-fann [Philosophy of Eastern music in the secrets of art] (Damascus, 1948)

A. Sautin: ‘La musique antique dans le monde oriental: la musique arabe’, Revue africaine, xciv (1950), 298–356

H. Avenary: ‘Abu ’l-Salt's Treatise on Music’, MD, vi (1952), 27–32

H.G. Farmer: ‘The Song-Captions in the Kitāb al-Aghānī al-Kabīr’, Transactions of the Glasgow University Oriental Society, xv (1953–4), 1–10

H.G. Farmer: ‘Al-Kindī on the “ēthos” of Rhythm, Colour and Perfume’, Transactions of the Glasgow University Oriental Society, xvi (1955–6), 29–38

H.G. Farmer: ‘The Science of Music in the Mafātīh al-‘Ulūm’, Transactions of the Glasgow University Oriental Society, xvii (1957–8), 1–9

H.G. Farmer: ‘Two Geniza Fragments on Music’, Transactions of the Glasgow University Oriental Society, ixx (1961–2), 52–62

Z. Yūsuf: Mūsīqā al-kindī [Al-Kindī's music] (Baghdad, 1962)

M.K. Allāh Wirdī: Jawla fī ‘ulūm al mūsīqā al-‘arabiyya [A voyage through the science of Arab music] (Baghdad, 1964)

E. Werner: ‘Greek Ideas in Music in Judeo-Arabic Literature’, The Commonwealth of Music: Essays in Honour of Curt Sachs, ed. G. Reese and R. Brandel (New York, 1965)

F. Rosenthal: ‘Two Graeco-Arabic Works on Music’, Proceedings of the American Philosophical Society, cx (1966), 261–8

H.M. al-Rajab: Hall rumūz kitāb al-aghānī li-l-mustalahāt al-mūsīqiyya al-‘arabiyya (Baghdad, 1967)

A. Shiloah: ‘Deux textes arabes inédits sur la musique’, Yuval, no.1 (1968), 221–48

Y. Shawqī: Qiyās sullam al-mūsīqā al-‘arabiyya [Measuring the Arab music scale] (Cairo, 1969)

E. Altwein: ‘Versuch über das arabische Komma’, Mf, xxiv/4 (1971), 432–37

A. Shiloah: ‘Un “problème musicale” inconnu de Thābit b. Qurra’, Orbis musicae, i (1971–2), 303–15

A. Shiloah: ‘Ibn Hindū le médecin et la musique’, Israel Oriental Studies, ii (1972), 447–62

H. Avenary: ‘The Hebrew Version of Abu’l-Salt's Treatise on Music’, Yuval, no.3 (1974), 7–82

Y. Shawqī: Risālat ibn al-munajjim fī ’l-mūsīqā [Al-Munajjim's treatise on music] (Cairo, 1976)

G.‘A. al-M. Khashaba: Kitāb al-mūjaz fī sharh mustalahāt al-aghānī [On the technical terms in the Kitāb al-aghānī] (1979)

B. Reinert: ‘Das Problem der pythagoräischen Kommas in der arabischen Musiktheorie’, Asiatische Studien, xxxiii/2 (1979), 199–217

A. Shiloah: ‘Les sept traités de musique dans le manuscrit 1705 de Manisa’, IOS, i (1979), 303–15

M. Cruz Hernández: ‘La teoría musical de Ibn Sīnā en el Kitāb al-shifā’, Milenario de Avicenna, ii (Madrid, 1981), 27–36

H. Avenary: ‘Paradigms of Arabic Musical Modes in the Geniza Fragment Cambridge, T.S.N.S. 90,4’, Yuval, no.4 (1982), 11–25

A. Shiloah and A. Berthier: ‘A propos d'un “petit livre arabe de musique”’, RdM, lxxi (1985), 164–77

E. Neubauer: ‘Das Musikkapitel der Ǧumal al-falsafa von Muhammad b. ‘Alī al-Hindī (1135 n. Chr.)’, Zeitschrift für Geschichte der Arabisch-Islamischen Wissenschaften, iv (1988), 51–9

S. Marcus: Arab Music Theory in the Modern Period (diss., UCLA, 1989)

A. Shiloah: ‘Techniques of Scholarship in Medieval Arabic Treatises’, Music Theory and its Sources: Antiquity and the Middle Ages, ed. A. Barbera (Notre Dame, IN, 1990), 85–99

E. Neubauer: ‘Arabische Anleitungen zur Musiktherapie’, Zeitschrift für Geschichte der Arabisch-Islamischen Wissenschaften, vi (1991), 227–72

E. Neubauer: ‘Die acht “Wege” der Musiklehre und der Oktoechos’, Zeitschrift für Geschichte der Arabisch-Islamischen Wissenschaften, ix (1994), 373–414

F. Shehadi: Philosophies of Music in Medieval Islam (Leiden, 1995)

h: modes, modality and rhythm

A.Z. Idelsohn: ‘Die Maqamen der arabischen Musik’, SIMG, xv (1913–14), 1–63

A.J Ellis: ‘On the Musical Scales of Various Nations’, Sammelbände für vergleichende Musikwissenschaft, i (1922)

T. Sabbāgh: Al-anghām al-sharqiyya [The Eastern modes] (Aleppo, 1954)

Carl Gregor, Duke of Mecklenburg: Ägyptische Rhythmik (Strasbourg, 1960)

H.G. Farmer: ‘The Old Arabian Melodic Modes’, Journal of the Royal Asiatic Society (1965), 99–102

L. Manik: Das arabischen Tonsystem im Mittelalter (Leiden, 1969)

J. Elsner: Der Begriff des Maqām in Ägypten in neuerer Zeit (Leipzig, 1973)

J. Jābir: Turāthunā wa-mafhūm al-sullam al-khumāsī [Our patrimony and the concepts of the pentatonic scale] (Khartoum, 1979)

A. Shiloah: ‘The Arabic Concept of Mode’, JAMS, xxxiv/1 (1981), 19–42 [repr. in A. Shiloah: The Dimension of Music in Islamic and Jewish Culture (Aldershot, 1993)]

S. al-Mahdī: Maqāmāt al-mūsīqā al-‘arabiyya [Modes of Arab music] (Tunis, 1982)

T. Kerbage: The Rhythms of Pearl Diver Music in Qatar (Doha, 1983)

G.D. Sawa: ‘Al-Fārābī's Theory of the Īqā‘: an Empirically Derived Model of Rhythmic Analysis’, Selected Reports in Ethnomusicology, i/9 (1983–4), 1–32

Arab Academy of Music: Nadwat al-sullam al-khumāsī [Round table on the pentatonic scale] (Khartoum, 1984) [Special issue of Majallat al-mūsīqā al-‘arabiyya]

S.H. al-Amīr: Dalīl salālim al-maqāmāt al-‘arabiyya [Guide to the scales of Arab maqāmāt] (Baghdad, 1986)

M.S. Firjānī: Maqāmāt al-mūsīqā al-‘arabiyya [The maqāmāt of Arab music] (Tripoli, 1986)

T. Ogger: Maqam Segah/Sikah Vergleich der Kunstmusik des Iran und des Irak anhand eines maqam-Modells (Hamburg, 1987)

J. Elsner: ‘Zum maqam-Prinzip, Tongruppenmelodik als Grundlage und Baustein musikalischer Produktion’, Maqām Raga Zeilenmelodik Konzeptionen und Prinzipien der Musikproduktion, ed. J. Elsner (Berlin, 1989), 7–39

S. Marcus: ‘The Periodization of Modern Arab Music Theory: Continuity and Change in the Definition of Maqāmāt’, Pacific Review of Ethnomusicology, v (1989), 33–48

S. al-Mahdī: Īqā‘āt al-mūsīqā al-‘arabiyya wa-ashkāluha [Rhythms and forms of Arab music] (Tunis, 1990)

‘U.‘A. Himsī: Usūl al-iqā‘āt al-sharqiyya [Eastern rhythms] (Damascus, 1991)

A. Shiloah: ‘Musical Modes and the Medical Dimension: the Arabic Sources (c.900–c.1600)’, Metaphor: a Musical Dimension, ed. J. Kessler (Sydney, 1991), 147–59

M. al-Ashhab: Ta‘līm al-maqāmāt al-‘arabiyya ‘alā al-ālāt al-mūsīqiyya [Teaching the Arab maqāmāt on musical instruments] (Casablanca, 1994)

i: traditional forms

H. al-Rajab: Al-maqām al-‘irāqī [The Iraqi maqām] (Baghdad, 1961, rev., enlarged, 1983)

al-Shaykh J. al-Hanafī: al-Mughannūn al-Baghdādiyyūn wa-’l-maqām al-‘irāqī [Baghdadi singers and the Iraqi maqām] (Baghdad, 1964)

S. El. Kholy: The Tradition of Improvisation in Arab Music (Cairo, c1978)

M. Guettat: La musique classique du Maghreb (Paris, 1980)

S. El Shawan: al-Musika al-‘Arabiyyah: a Category of Urban Music in Cairo, Egypt 1927–1977 (diss., New York U., 1981)

A.J. Racy: ‘The Waslah: a Compound-Form Principle in Egyptian Music’, Arab Studies Quarterly, v/4 (1983), 396–403

L.I. al Faruqi: ‘The Suite in Islamic History and Culture’, World of Music, xvii/3 (1985), 46–64

G. Braune: Die qasida in Gesang von Umm Kulthum: die arabische Poesie im Repertoire der grössten ägyptischen Sängerin unserer Zeit (Hamburg, 1987)

L. Plenckers: De Muziek van de Algerijnse mūwashshah (Alkmaar, 1989)

C. Poché: La musique arabo-andalouse (Arles, 1995; Sp. trans., 1997)

M. Garufi: Les formes instrumentales dans la musique classique de Tunisie (Tunis, 1996)

j: song forms

K. al-Najmī: Mutribūn wa-mustami‘ūn [Singers and listeners] (Cairo, 1970)

R. Abū Sariyya: al-Ughniya al-siyāsiyya al-jadīda [New political song] (n.p., c1972)

V. Sahhab: al-Sab‘a al-kibār fī ’l-mūsīqā al-‘arabiyya al-mu‘āsira [Seven giants in contemporary Arab music] (Beirut, 1987)

I. Sahhab: Difā‘an ‘an al-ughniya al-‘arabiyya [In defence of Arab song] (Beirut, 1980)

S. al-Sharīf: al-Ughniya al-‘arabiyya [Arab song] (Damascus, 1981)

M.F. Ibrahim and A. Pignol: L'extase et le transistor: la chanson égyptienne (Cairo, 1986)

N.M. Ghānim and K.M. al-Qāsimī: Asāla al-ughniya al-‘arabiyya bayna al-Yaman wa-al-Khalīj [Authenticity of Arab song from Yemen to the Gulf] (Damascus, 1991)

T. Fāri‘: al-Ughniya al-yamaniyya al-mu‘āsira [Contemporary Yemeni song] (Beirut and Sharjah, 1993)

M. Būdhina: al-Aghānī al-khālida: 100 ughniya tūnisiyyah mukhtāra [Immortal songs: an anthology of 100 Tunisian songs] (Hammamet, 1995–7)

k: instruments

Kanz al-tuhaf [Treasury of rarities] (MS, GB-Lbl Or.2361) [in Persian]

E. Wiedemann: ‘Über Musikautomaten bei den Arabern’, Centenario Amari, ii (1910), 164–85

B. Carra de Vaux: ‘Notes d'histoire des sciences, 1: Muristos, inventeur des orgues’, Journal asiatique, 11th ser., x (1917), 449–51

H.G. Farmer: ‘The Evolution of the tanbūr or pandore’, Transactions of the Glasgow University Oriental Society, v (1923–8), 26–8

H.G. Farmer: ‘Ibn Khurdādhbih on Musical Instruments’, Journal of the Royal Asiatic Society (1926), 509–18

K. Geiringer: ‘Vorgeschichte und Geschichte der europäischen Laute bis zum Beginn der Neuzeit’, ZMw, x (1927–8)

H.G. Farmer: ‘Meccan Musical Instruments’, Journal of the Royal Asiatic Society (1929), 489–505

H.G. Farmer: ‘The Origin of the Arabian Lute and Rebec’, Journal of the Royal Asiatic Society (1930), 767–83

H.G. Farmer: Studies in Oriental Musical Instruments, i (London, 1931/R); ii (Glasgow, 1939/R)

H.G. Farmer: An Old Moorish Lute Tutor (Glasgow, 1933) [four Arabic texts from unique MSS]

H.G. Farmer: ‘A Maghribī Work on Musical Instruments’, Journal of the Royal Asiatic Society (1935), 339–53

H.G. Farmer: ‘Was the Arabian and Persian Lute Fretted?’, Journal of the Royal Asiatic Society (1937), 453–60

H.G. Farmer: Ancient Arabian Musical Instruments as Described by al-Mufaddal ibn Salamah (Glasgow, 1938)

J. Robson: Ancient Arabian Instruments (Glasgow, 1938)

H.G. Farmer: ‘The Structure of the Arabian and Persian Lute in the Middle Ages’, Journal of the Royal Asiatic Society (1939), 41–51

H. Hickmann: Terminologie arabe des instruments de musique (Cairo, 1947)

J. Robson: ‘Some Arab Musical Instruments’, Islamic Culture, xxxii (1958), 171–85

M.A. El Hefny: ‘Ilm al-ālāt al-mūsīqiyya [The science of musical instruments] (Cairo, 1971, 2/1987)

T. Grame: ‘The Symbolism of the ‘ūd’, AsM, iii/1 (1972), 25–34

M. Kamel: ‘Al-takht al-sharqī’ [The takht al-sharqī traditional ensemble], Al-majalla al-mūsīqiyya, ii (1974), 23–4

L. Hage: Un piano occidental oriental: le piano Chahine (Kaslik, 1975)

A. Shiloah: ‘The ‘ūd and the Origin of Music’, Studia Orientalia, Memoriae D.H. Baneth dedicata, ed. J. Blau and others (Jerusalem, 1979), 395–407

P. Collaer and J. Elsner: Nordafrika, Musikgeschichte in Bildern (Leipzig, 1983)

C. Poché: ‘David and the Ambiguity of the mizmār according to Arab Sources’, World of Music, ii (1983), 58–73

A.J. Racy: ‘Sound and Society: the Takht Music of Early-Twentieth Century Cairo’, Selected Reports in Ethnomusicology, vii (1988), 139–70

J.F. Belleface: ‘Turáth, classicisme et variétés: les avatars de l'orchestre oriental au Caire au début du XXe siècle’, Bulletin d'études orientales de l'Institut français de Damas, xlii (1989), 39–65

E. Neubauer: ‘Der Bau der Laute und ihre Besaitung nach arabischen, persischen und türkischen Quellen des 9. bis 15. Jahrhunderts’, Zeitschrift für Geschichte der Arabisch-Islamischen Wissenschaften, viii (1993), 279–378

l: musicians

F. ‘Amrūshī: Al-jawārī al-mughanniyyāt [Singing slave girls] (Cairo, 1961)

E. Neubauer: Musiker am Hof der frühen ‘Abbāsiden (diss., U. of Frankfurt, 1965)

‘A.al-K. ‘Allāf: Qiyān baghdād fī ’l-‘asr al-‘abbāsī wa-’l-‘uthmānī al-akhīr [The qaynas of Baghdad in the Abbasid and late Ottoman periods] (Baghdad, 1969)

M.A. al-Hifnī: Ziryāb: abū ’l-hasan ‘alī b. nāfi‘ mūsīqār al-andalus [Ziryāb, the Andalusian musician] (Cairo, c1970)

A. Shiloah: ‘Le poète musicien et la création poético-musicale au Moyen Orient’, YIFMC, vi (1974), 52–63

J.E. Bencheikh: ‘Les musiciens et la poésie: les écoles d'Ishāq al-Mawsilī (m. 235 H.) et d'Ibrāhīm Ibn al-Mahdī (m. 224 H.)’, Arabica, xxii (1975), 114–52

M. Stigelbauer: Die Sängerinnen am Abbasidenhof um die Zeit des Kalifen al-Mutawakkil (Vienna, 1975)

‘A. al-Bakrī: Safī al-dīn al-urmawī, mujaddid al-mūsīqā al-‘abbāsiyya [Safī al-Din, reformer of Abbasid music] (Baghdad, 1978)

G. Sawa: ‘The Status and Roles of Secular Musicians in the Kitab al-Aghani’, AsM, xvii/1 (1985), 69–82

H. Engel: Die Stellung des Musikers im arabisch-islamischen Raum (Bonn, 1987)

m: notations and transcriptions

A. Chalfoun: Kitāb al-‘alāmāt al-mūsīqiyya [The book of musical signs] (Cairo, c1926)

A. Chottin: Corpus de musique marocaine, fascicule 1: Nouba de Ochchâk (Paris, 1931/R)

F. Rajā‘ī and N. al-Darwīsh: Min kunūzinā al-halqa al-ūlā fī al-muwashshahāt al-andalusiyya [From our treasures: the first collection of the Andalusian muwashshahāt] (Aleppo, 1955)

M.A. El Hefny and others: Turāthunā al-mūsīqī min al-adwār wa-’l muwashshahāt [Our musical patrimony from adwār and muwashshahāt] (Cairo, 1959–63)

el Hitami: Majmū‘at al-mūsīqā al-āliyya [Classical instrumental music of Egypt] (Cairo, 1983)

‘I. Bannani: Bughyāt wa-tawāshī: nawbat al-mūsīqā al-andalusiyya al-maghribiyya: majmū‘a kāmila bi-al-nūta al-mūsīqiyya [Bughyāt and tawāshī: the Andalusian Maghrebian Nouba: complete collection in staff notation] (Rabat, 1995)

M. Būdhīna: Mūsīqā al-ma’lūf [Corpus of the Tunisian ma’lūf] (Hammamet, 1995)

R. Aous, ed.: Les grands maîtres algériens du cha‘bi et du hawzi (Paris, 1996)

I. El-Mallah: Arabische Musik und Notenschrift (Tutzing, 1996)

2. History, theory and sources.

n: modernism and tradition

M.A.K. Hajjāj : al-Mūsīqā al-sharqiyya, mādīha, hādiruha, numuwwuha fī ’l-mustaqbal [Eastern music: its past, present and growth into the future] (Alexandria, 1924)

A.A. Mansī : al-Mūsīqā al-sharqiyya bayna al-qadīm wa-al-jadīd [Eastern music, between old and new] (Cairo, 1949, enlarged, 1965)

Z. Yūsuf : al-Takhtīt al-mūsīqī lil-bilād al-‘arabiyya [Musical planning for Arab countries] (Baghdad, 1965)

A.J. Racy : Musical Change and Commercial Recording in Egypt, 1904–1932 (diss., U. of Michigan, 1977)

G. Sawa: ‘The Survival of Some Aspects of Medieval Arabic Performance Practice’, EthM, xxv/1 (1981), 73–86

N. Bouzar-Kasbadji : L'émergence artistique algérienne au XXe siècle: contribution de la musique et du théâtre algérois à la renaissance culturelle et à la prise de conscience nationaliste (Algiers, 1988)

Z. Nassār : al-Mūsīqā al-misriyya al-mutatawwira [Egyptian modernized music] (Cairo, 1990, 2/1996)

S. El Kholy : al-Qawmiyya fī mūsīqā al-qarn al-‘ishrīn [Nationalism in 20th-century music] (Kuwait, 1992)

F. Lagrange : Musiques d'Egypte (Arles, 1996)

C. Poché: ‘Les archives de la musique arabe’, Revue d'études palestiniennes, viii (1996), 79–93

Arab music

II. Folk music

1. Introduction.

3. Vocal forms.

4. Music and performance.

5. Musical instruments.

1. Introduction.

Arab folk music presents a multiplicity of musical idioms and vocal and instrumental genres. They include several forms of religious chanting, work songs, narrative pieces, didactic and lyric songs, and songs and dances that provide entertainment at weddings and other special occasions or social events. Performed by amateurs and professionals or semi-professional specialists, these disparate genres reflect a wide range of human responses to varying conditions – from nomad encampments in desolate arid zones to small villages and urban centres.

Many factors contribute to this diversity. Arab folk music is an intricate mosaic involving peoples of different ethnic groups and sub-groups who speak and sing in numerous dialects of Arabic (and sometimes in languages other than Arabic). Although predominantly Muslim, peoples of the Arab world adhere to widely differing interpretations of religion, and their related concepts of music. Arab folk music and dance traditions also fulfil multiple functions on both the individual and community level.

These traditions exist alongside the urbanized art music which became established and widely accepted in the Abbasid era (750–1258; see §I above). The Islamic empire covered a vast expanse of territory stretching beyond the Arab world of today, from Central Asia to the Atlantic. Many genres which existed prior to the advent of Islam survived and preserved their own vitality, but the spread of Islam also had an important unifying impact. Through a long process of interaction between many diverse peoples, individual repertories within Arab music manifest an overall kinship. This is expressed in linguistic and musical affinities which cut across political boundaries.

Without over-generalizing, it is possible to single out salient common traits and distinctive particularities. The great importance of poetry and a predominantly vocal conception of music is characteristic of both nomadic and sedentary Arab peoples. A rich palette of styles of vocal production has developed. The vocal conception of music also applies to dance genres, which are usually accompanied by singing and clapping. Even dances accompanied by instruments frequently include sung verses.

The numerous genres of folksong surpass classical songs in both quantity and fecundity. Poetry connected with folk music and dance is primarily an accompaniment to ceremonial and other events, not so much a mode of literary expression as such. The poetry is based on consistent forms and methods of composition and rendition. Responsorial forms are frequently used, and these highlight the importance of audience participation. The poet-musician has a distinguished position as a narrator and mediator of events, and as a spokesman for fellow members of the community.

The use of instruments varies within the traditions and, although names vary, often the shape and playing techniques are similar.

2. History, theory and sources.

Arab art music has been systematically described, with its theory perpetuated in numerous scientific and literary works. Folk music, on the other hand, presents a different and problematic situation. A wealth of valuable information about folk music and dance exists, but it is scattered and generally unsystematic. Moreover, it is derived from various categories of sources which are often not directly concerned with folk music for its own sake.

The problem of sources relates to the history of basic concepts about music used by Islamic legists and by intellectuals. The Arabic word ghinā’ (‘singing’) has been used in theory and practice as a general term assigned exclusively to the art music which developed after the advent of Islam. Consequently, folk and religious music were not considered as music per se: their various forms received names denoting and emphasizing their verbal character, as vehicles for sung texts. Purely instrumental music was categorized in relation to musical instruments and classed with forbidden amusements.

A noteworthy rare attempt to provide a scientific distinction between art music and folk and religious music was made by the eminent Arab historian and sociologist Ibn Khaldūn (d 1406). In his chapter on the ‘craft of music (ghinā’)’, he characterizes art music as based on well-established rules and conventions, concluding that long training is necessary to acquire skill. By contrast he sees folk music as simple and spontaneously created without special instruction: ‘Many people are gifted to achieve it by nature … for we find people who are by nature gifted in the metres of poetry, the rhythms of dance and similar things (Prolegomenes, V, 31)’. Interestingly, his first illustration of this natural aptitude relates to Qur’anic recitation, which he distinguishes sharply from art music. As an example of folk music, he refers to pre-Islamic Arab nomads’ chants and songs enhancing their social gatherings and pilgrimages to the shrine of the black stone, the Ka‘ba at Mecca.

The term samā‘ refers to both the act of hearing music and to music that is heard (in a spiritual context). A rich and varied literature exists on this topic, dealing with the admissibility of music from legal, theological and mystical points of view. An interminable debate about the use of music in ritual dates from the early Islamic period onwards (see above, §I and Islamic religious music, §I, 2 for further discussion). Quite aside from their central theme, these writings constitute a mine of information about different genres of singing, dances and typological details relating to musical instruments. Regardless of whether the writer wishes to defend or prohibit a particular instrument, the organological information is useful, and usually hard to find in theoretical treatises on music.

Jurists and theologians have used the term malāhī (‘diversions’) to designate musical instruments, assimilating them into the category of forbidden types of amusements. In the same spirit, another related term, la‘b (‘pastime, game’), has been used to designate the dance of the Sufis or dervishes to attain union with God. (The term still designates certain types of folk dances.) In their attack against the Sufi mystical dance, the Islamic jurists likened it to non-Islamic genres such as the ‘Golden Calf’ dance or Christian dances, thus rendering it heretical (bid‘a).

Religious prohibitions extended to other types of popular entertainment, with or without dance. The legists condemned pre-Islamic rituals which were enlivened with impressive music, such as the celebration of now-rūz (the ancient Persian New Year) and the different Nile festivals in Egypt. At now-rūz and other festivals in Egypt and Syria, dolls decorated with beautiful clothes and jewellery were displayed, serenaded with music from oboe-type instruments and drums. However, descriptions of travel writers attest that many non-Islamic customs persisted, with their attendant music. In his travel memoirs, the 11th-century Persian writer Nasīr-ī Khusraw describes a ceremony for opening a dyke on the river Nile: an ensemble of long trumpets and drums playing appropriate melodies was followed by a procession of 10,000 horsemen and 10,000 others.

These huge popular open-air ceremonial performances and parades were occasions for itinerant actors, jesters, shadow puppeteers, storytellers, dancers, singers and instrumentalists to display their talents. They were typical of urban and semi-urban areas. Late 18th-century European travellers described these perfoming practices, and in Egypt systematic surveys of folk music and dances were carried out by Villoteau (1809) and Lane (1836).

3. Vocal forms.

Text and music are indissolubly linked in Arab folk music. There are many types of sung poetry which differ from each other in prosodic form, in mode of performance and in their social or musical functions.

(i) The qasīda.

Sometimes translated as ‘ode’, this is one of the most ancient forms of sung poetry. It was originally created by Bedouin nomads in the Arabian desert, and its use now spans classical and popular traditions.

The classical qasīda consist of many lines, sometimes over 100. Each line is divided into two equal parts and subdivided into feet. Each qasīda has a single rhyme and uniform metre (any one of 16 traditional metres). In subject matter the qasīda follows a schematic development of themes and images. The basic compositional concept is that each line should be independent and contain a complete, self-sufficient idea. Hence the poet’s originality is measured by his ability to reformulate conventional material and to invent additional lines. Natural variants may occur through lapses of memory and the dependence on oral transmission, so the order of individual lines can seldom be established accurately. The poems are never sung in exactly the same way or in their entirety; original words and whole lines are often changed. Even when set to a standard tune, with the text only slightly changed or with lines rearranged, a qasīda is accepted as ‘new’ by singers and listeners.

In North Africa, the qasīda is considered a religious genre, usually performed by professional qassādīn at Muslim festivals. In Yemen, a poetic form akin to the qasīda, the nashīd, is performed by professional nashshādīn in similar contexts.

The folk form is known as qasīda. It differs from the classical qasīda both in prosodic structure and in language, which may contain colloquialisms. Musical structure takes precedence over textual metre. The classical and folk forms have similar performing practices and social functions, and are always sung with gravity.

(ii) Strophic forms.

Short strophic forms are prevalent in Arab folk music, particularly quatrains with or without refrains. Quatrains exist in differing forms, according to genre, rhyme scheme and number of syllables per line. Names vary regionally, e.g. ‘atābā, ubudhiyya, majruda, murabba‘, muthamman and ‘urūbi.

The ‘atābā, found in Lebanon, Palestine, Jordan, Syria and Iraq, has an AAAB rhyme scheme, the final line ending with the syllable -āb or -āba (the last syllable of the word ‘atābā). Here is an example of an ‘atābā stanza:

Bani-l-Dabbah bid-deyr hamūlī
Idha bindāq bi-l-hayja hamūlī
Bnifsi fī mhabbatkum humūlī
Il-shruqi wa-l-m‘anna wa-l-‘atāb.

(Sons of Dabbah, you form a clan in Dayr al-Asad.
When I am in trouble in battle, they defend me.
My load becomes lighter because of your love.
[Let us sing] the shruqi, the m‘anna and the ‘atāb.)

Akin to the ‘atābā is the Mesopotamian ubudhiyya (also ubuthiyya; see Iraq, §III, 3). As with the ‘atābā, the last line of each stanza identifies the form, ending with -iyya or -dhiyya. Both the ‘atābā and ubudhiyya are soloistic improvised song types, often performed antiphonally by two poet-musicians.

Another type of quatrain is known as murabba‘ in Lebanon, Syria, Jordan and Palestine; majruda in Libya. The lines usually have eight syllables; the rhyme scheme is AAAB, CCCB, etc. These songs are very rhythmic and always accompanied by hand-clapping. They include participation by members of the audience who repeat the final line of each stanza or sing a refrain. A related type, the muthamman, is found in Lebanon, Syria and Israel. It has eight seven-syllable lines to a stanza (rhyme scheme: ABABABAC, DEDEDEDC, etc.); the refrain does not rhyme with the stanzas (see above, ex.8).

The murabba‘ and muthamman are used without the refrain in disputation songs known as hiwār, mu‘ārada, ‘urūbi (North Africa) and mani (Turkey). In this largely improvisational genre two protagonists (men or women; sometimes a man and woman) compete, exchanging verses. In another quatrain form, the m’anna, the rhyme scheme is AABA. The audience sings a refrain consisting of the final line of each stanza. In contrast with the free and melismatic character of the soloist's material, it is set in a strict rhythm.

Arab folk music also has genres with stanzas of more than four lines. For instance, the Egyptian mawwāl has five lines (aaaba rhyme scheme); the short Zajal has six lines; the Baghdadi mawwāl has seven lines (AAABBBA). In these forms the first three lines end with homonyms.

Aspects of musical structure and performing practice also link these short poetic forms. These include the recurrence of certain vowel and consonant sounds, repetition of a line or part of a line, the addition of meaningless syllables and the use of refrains.

There are three main types of refrain. One reiterates the last line of each stanza (and consequently changes from one stanza to the next). Another type is the formulaic refrain, using a phrase (or phrases) unrelated to the song text, repeated exactly throughout the song, as in the muthamman. A third introductory type of refrain is recognizably derived from a pre-existing source (often an older non-strophic group song).

An example of formulaic refrain is Halālī ya mālī (‘the girl or the money’). Although this may have originally belonged to one particular song, it is now found in at least three genres, among them the muthamman. Some formulaic refrains use meaningless phrases, for example a genre known as dana dana from the Hadramawt region of Yemen. The soloist improvises stanzas and members of the audience sing the refrain Dana dana ya dani/Ya dani dana dana (sometimes preceded by a meaningful refrain line). Performances can last the whole night.

The introductory type of refrain consists of either an independent line or a formulaic opening line (or lines) taken from the first stanza. These lines identify a well-defined genre, such as ‘alā dal ‘ūnā, mijānā and rozānā. The genre that has been most studied is Barhum ya Barhum (diminutive of ‘Abraham, O Abraham’), well known in a wide area stretching from Egypt to Syria. In all versions the last two lines of the verse are repeated, preceded by exclamations such as eh wallah (‘O God!’), ya yumma (‘O mother!’) or ya weyli (‘woe is me!’). The song was created in the first two decades of the 20th century, and rapidly became popular. In its various versions it airs a great variety of topics and current social concerns, but the tune and consistent basic structure have remained relatively unchanged (ex.12). This pattern of stable tune and widely varying text seems characteristic of all the genres whose refrains come from pre-existing sources.

(iii) Non-strophic forms.

This category comprises a repertory of social group songs consisting of simple exhortatory or exclamatory formulae which accompany dances. It is a type commonly used for weddings, work, processions etc., and includes many texts and few tunes. An example is the Yemeni harvest song, mahjal. The far‘awiyya and mhorabe are also typical. They begin with a phrase known as the tal‘a, which is then repeated as a response after each solo verse. The solo verses are not related to each other or to the tal‘a. Some tal‘a phrases have religious connotations: ‘God is the greatest, O my land, glorify him’ (see below, ex.15). Others relate to a particular occasion or function: ‘The fair one stepped out of the bath’; or ‘Greet your guests, O Abū S’ud’.

These songs are usually accompanied by hand-clapping and sometimes by a kind of marching dance. In one such dance, the razīh of Sinai and Yemen, the shallāl (precentor) recites the first half-verse and the other participants sing one of the hundreds of traditional and very ancient refrains, such as ‘Praise to him who passeth not away/and whose power never ceaseth’; or ‘God give us good fortune’.

The same category includes the genres tansūrah, hanhunnāt and zaghārīd. These are sung by women during wedding festivals. They consist of improvised verses which start with the interjection ayha or iyha intoned at a high pitch, followed by verses sung in a fast parlando style and interrupted by the traditional yuyu (ululation). There are songs for the bridegroom (e.g. ‘O our bridegroom, we are your guests today/Prepare the bed for us and leave your house for us’; or ‘No one is like our bridegroom/The wedding is celebrated in his uncle’s house’) and others for the bride (e.g. ‘Your anklet, O beautiful one, resounds/Your skirt, this one of pepper colour, has in it life and death’; or ‘Walk gracefully, O beautiful one, O magnificent one/O rose which has bloomed in the garden’).

This category also includes the hōsa (an Iraqi Bedouin genre) and various religious processional songs, such as songs of pilgrimage to Mecca or to the tombs of saints, and songs for the mawlid (birthday of the Prophet). For instance, in the mawlid procession in Libya the formula ‘Pray, O worshipper, for the Magnificent (Prophet)’ is continually repeated.

Epics also belong to this category. They are popular among all classes throughout the Arab world. They are associated with the shā‘ir (bard or poet-musician), who sings them while accompanying himself on the rabāb (one-string fiddle).

In villages and towns there are also poet-musicians who function as entertainers, particularly during the evenings of religious festivals. They usually perform in coffee-houses, captivating the audience with their lively and dramatic manner of narration as well as with the story itself. Some poet-musicians specialize in particular epics: the ‘Abū Zaydiyya’, for example, perform the Abū Zayd romance about the life and adventures of the epic hero Abū Zayd al-Hilālī. This narrative is supposed to be based on events which took place in the 9th century. During a single evening of Ramadan the same epic, in various versions, can be heard in Libya, Yemen and Egypt. There are also epics about other famous heroes (such as Antar), which originated among the Bedouin and are performed in the same manner all over the region.

4. Music and performance.

(i) Music.

For the singer, text and melody form an integral unit. Such cohesion implies that they are also conceived together and a folksinger often finds it impossible to recite the text separately from the melody. The length of the lines is often determined by the melodic or rhythmic texture, while in some instances the text may determine the nature of the melodic and rhythmic events. It is impossible to understand the relationship between text and tune, especially the role of poetic metre, without a knowledge of the music used. As a further complication, both text and tune undergo constant changes from performance to performance.

The song genres utilize various musical structures. Some are solo; others are responsorial or mixed. Some are syllabic with rigid rhythm while others are melismatic and consequently performed with a certain rhythmic freedom. Some genres are accompanied by hand-clapping, dances, or percussion or melodic instruments, whereas others are for voice alone.

While ‘new’ texts are constantly added to the total song repertory, the number of ‘new’ melodies is very limited. In a given genre one tune can be used for several texts. There are various ways of fitting text to tune: contraction of syllables, elision of letters or addition of meaningless syllables. The links between simpler tunes and particular social functions are very flexible: a certain number of simple tunes are used without differentiation both in work songs and in processional, religious and wedding songs. This music is seen as a vehicle for the text; its function is not necessarily to provide a distinct setting for the text.

Within the various traditions there is generally a conscious differentiation between two types of song: syllabic and melismatic. The syllabic songs are narrow in tonal range, cruder and simpler than the melismatic songs, and may be older and less subject to variation. They employ many kinds of responsorial form and are usually performed communally at social gatherings or group activities. The melismatic songs are soloistic, lyrical and more complex in texture. To a large extent they fulfil individual and personal functions.

In syllabic songs the regular musical beats are adjusted to the scansion of the syllables, coordinating musical and textual expression. The end of each melodic phrase (which may occur within the framework of a line or a whole stanza) is distinctly marked by a cadential formula. In the melismatic songs stresses within the melody coordinate with textual stresses, and highly ornamented and melismatic passages are placed so as not to disturb the meaning and general flow of the text. Cadences do not always coincide with the ends of textual lines, particularly in songs which are musically more elaborate and perhaps influenced by art music. At the end of a line or stanza a singer may add meaningless syllables or vowel sounds before reaching the melodic cadential formula. Within the more soloistic genres musicians may use elements of the classical maqām system, but as yet the connections between folksong genres and art music are poorly understood.

(ii) Melodic structure.

We may generally observe that Arab vocal music ranges from simple melodies with short repeated formulae to more complex and soloistic melodies.

Songs closely associated with basic human physical and spiritual life tend to use only one or two scale-types and have narrow ranges, rarely more than a 4th (ex.13, ex.14 and ex.15). Some are performed on one note. This category of songs is usually performed antiphonally or responsorially. In some, formulae are exactly repeated throughout the song; in others, the repeated formulae use the principle of open and closed phrases. Open phrases frequently end on the 2nd below the finalis, or sometimes on the 2nd above. They are sung in a very plain manner, although they are often marked by various types of intonation and glissandos. Their melodic plainness contrasts with their varied and complicated rhythms and the scansion of the text often results in certain asymmetrical rhythms. The rhythmical richness is always emphasized and enhanced by hand-clapping, which is often doubled with drumbeats to create a polyrhythmic pattern with the melody (see ex.14). In the Egyptian song (ex.13) percussive accompaniment is relatively important, competing with the melody and creating independent cadences.

In addition to this complex rhythmic accompaniment, simple melodic songs are frequently interrupted or marked by exhortatory shouts, or by women's ululations (yuyu). An example is the zaghārīd, widely performed by women at wedding celebrations. Body movements give a spatial dimension to the rhythms and hand-clapping. The participants form a long row, gradually augmenting their tiny steps and movements as they sing and clap, thus effecting an unbroken transition into dance.

The dahhiyya dance-song of the Bedouin in Sinai and Iraq is a prime example, with its plain melodic style and emphasis on repetition and rhythm. It takes place in complete darkness, at a festival or evening party. At a given moment the men get up and form a row, repeating dahhiyya, dahhiyya until they feel inspired. The poet-musician then faces the line and starts improvising verses, or two soloists may sing antiphonally. Members of the group sing responses accompanied by uniform rhythmic clapping as they take small steps backwards and forwards, keeping their feet close to the ground. This dance reaches its climax at the appearance of an unmarried woman, clad entirely in black, brandishing a sword in her right hand. She takes large steps and jumps while waving her sword, and the dancers alternately move towards her and retreat, repulsed by the sword’s movement. The young woman is supposedly anonymous: all the participants know who she is, but they ignore her identity to protect her honour. The marbū‘a, a variation of the dahhiyya, is musically almost identical but sung at a higher pitch to attract young girls and women. Another form of dance-song, the radīh, occurs in Yemen and other parts of the Arabian peninsula. The form is physically as well as musically antiphonal, for the line of participants divides into two sections while performing the verses. There are many other forms of dance-song, particularly in Yemen where both men and women show a strong predilection for dance. In these responsorial songs melodic overlapping occurs when the soloist or chorus begin a phrase before the previous one ends. Certain scholars view this as a type of rudimentary folk polyphony.

Songs with frequent use of simple repeated formulae and the same rhythmic patterns are accompanied by hand-clapping and drum-beating. However, their melodies are more complex in form, organization, range and tonality. They tend to have a melodic range of a 4th or a 5th and the series of notes which forms the scales appears to be distantly related to the maqām tradition.

The Muwashshah (ex.16) illustrates some of these characteristics: the repeated melody is fairly melismatic, the range is a diminished 5th and the mode is possibly related to the maqām segāh. The refrain, musically identical with the soloist’s part, is sung twice by the chorus, both at the beginning of the song and after each stanza. The Egyptian Ramadan ‘lantern song’ (ex.17) shows another common pattern: the response, no longer identical with the soloist’s part, progresses within the tetrachord below the finalis while the soloist’s part undulates around the finalis. There is more complex organization in the muthamman (ex.18), which has a stanza of eight short lines followed by a refrain line. Each line is divided textually into seven syllables and musically into a duple metre of four beats. The fourth beat in each of the first six lines coincides with the finalis reached from the 3rd above, allowing the seven syllables to cover eight quaver beats. The melodic motif is exactly repeated six times. The seventh line and refrain constitute a developed cadence followed by the choral response, whose melody and text are unrelated to the main song.

More complex songs with a broader range and greater variety of pitches than songs of the other categories are marked by influences from art music and are essentially soloistic. The simplest have repeated melodic and rhythmic lines with slight variations, such as the Yemeni welcome song (ex.19). This has a systematic descent from the 6th, to the finalis characterized by an undulating progression and syncopated rhythm, ending on a cadence occurring after the finalis. The melody is also enriched by inflecting the 3rd (e).

Most of the free rhythmical songs have a more complicated form (e.g. the ‘atābā). They often begin with an introductory melismatic vocal improvisation (on meaningless words and syllables) to adjust everyone to the mode of the song. Short improvised passages of this kind sometimes precede each stanza. The musical lines are not repeated exactly; there are long musical periods which are highly melismatic and use frequently fluctuating intonations as an expressive device. The musical lines also have a wide range, often more than an octave. In this category improvisation and variability are much more evident in both music and text. Bartók described this phenomenon of variability as universal, but in Middle Eastern music it seems particularly marked.

Although most Arab tunes are sung or performed monophonically, rudimentary polyphony, both vocal and instrumental, is common. An interesting example is the nahami genre of Kuwait and Bahrain. The naham is a singer paid to accompany and entertain the pearl fishers during their long periods at sea. While he sings, the fishermen intone an extremely low note two octaves below the level of his melody. This may be a fixed drone, or the fishermen may attempt to follow the outline of the melody. This practice has similarities to the isōn in Byzantine chant and also occurs in other Arabian Gulf genres. Samaritans and Yemenite Jews still sing in various parallel intervals; heterophony and overlapping are also common in their music.

The striking style of singing at extremely low pitches also occurs in Egypt in the Laythi order’s dhikr (a type of religious ceremony): while the munshid (precentor) performs melismatic tunes, the worshippers intone at an extremely low pitch an ostinato with the repeated sentence ‘There is no other god but God’.

(iii) Song types and their social functions.

Many types of song fall largely within the orbit of professional and semi-professional specialists. These include sung narratives, didactic songs, songs connected with various Muslim festivals (e.g. Ramadan; ex.17), and many social songs marking major events in the human life-cycle.

Several types of lyric songs are performed by male and female non-professionals, singing as individuals or (sometimes) collectively. Nostalgic lyric songs are some of the oldest of the Bedouin repertory, and there is a great variety of lyric songs about love, including lovers’ complaint songs. The Moroccan ‘urūbī is a sung dialogue between a man and woman, and the Yemeni humaynī, a women’s genre, exalts genteel love. Caravan songs also come within this lyric category, variously known as hujaynī (Egypt), rakbānī (Iraq), hidā‘ (Yemen) and barakah (Libya).

Work songs present a variety of genres. Some relate to animals (hunting, herding, watering, etc.), some to sailing and fishing, and others to agricultural life, e.g. songs about fertilizing the date palm. In harvest songs such as the Yemeni mahjal, the rhythmic pattern stitmulates the work movements. Another Yemeni song, the women's hādī, is linked to the stripping of sorghum foliage. Work songs are among the most important within the Egyptian village repertory.

E.W. Lane, who lived in Egypt from 1825 to 1849, noted songs and chants of boatmen rowing, peasants raising water, porters carrying loads, sawyers, reapers and many other labourers. Most of these are still sung and also exist in other Middle Eastern countries. Some specific kinds of work have given rise to appropriate repertories (e.g. the flooding of fields and well-irrigation in Hadramawt, and the unique and beautiful repertory of the pearl fishers in the Arabian Gulf).

The zār, a song type of Egypt, Sudan and other places near the Red Sea, is connected with the practice of exorcism. It is part of a ceremony intended to expel evil spirits and mysterious diseases from people who are thought to be possessed. Specialized ensembles of singers, dancers and instrumentalists, particularly players of the simsimiyya (lyre), have an important role (ex.20).

(iv) The poet-musician.

Like all regional bards, the poet-musician (shā‘ir) has a uniquely important position as a spokesman who articulates the moods and aspirations of the community. In Bedouin tribal society he or she is a walking archive and interpreter of memorable events and nuances of custom. The poet-musician’s artistic contribution and very presence in the community is crucial for the creation, transmission and revitalization of many vocal genres.

As poet, composer and performer, multiple and inseparable talents are brought into play. The poet-musician usually has a good or pleasing voice and exceptional memory. Much creativity lies in the ability to adapt existing material from the traditional store of motifs and formulae. Drawing on memorized material, the poet-singer often repeats certain verses in different contexts and vocal genres, but always in a way that fits the particular form or situation.

Often a second poet-musician takes part in the performance and alternates with the first. The resulting exchange of improvised verses and stanzas introduces an element of competition which stimulates their imagination, gives them time to rest and think, and greatly amuses the audience. Sometimes, when one has difficulty continuing, the second immediately intones a vocal improvisation followed by a new stanza. For the poet-musician the presence of a large audience is also a stimulus – indeed, it is a necessary condition for improvisation.

Poet-musicians who are invited to sing at ceremonies are considered professional, but their payment does not amount to a living wage. Some specialize in a particular genre such as epics; dirges are the province of female singers who excel as keeners. There is no formal teaching: a gifted individual with a natural predisposition for music listens to the recognized poet-musicians of his group, memorizes some material, and starts occasionally creating songs based on these models. Having acquired a certain degree of confidence, the novice is invited to give an example of his ability. At a later stage he attends a small gathering to try to perform for a whole evening, and in return receives token payment. Finally, once he becomes recognized as an able singer, he is considered professional.

Although oral transmission has always been an essential feature of this tradition, educated poets now frequently write down their poems in notebooks and classify them according to genres, and some collections have been published. The poets use their notebooks as memory aids, and do not refer to them in performance.

The social status of the poet-musician is ambivalent: he is highly appreciated for his talent and as an indispensable entertainer, but in everyday life and human relationships he is treated with a certain distrust and even contempt. In Arab cultures music and other creative activities have always tended to be considered superfluous occupations and the lowest of professions.