The Sacred Direction and City Structure: a Preliminary Analysis of the Islamic Cities of Morocco Author(S): Michael E

The Sacred Direction and City Structure: A Preliminary Analysis of the Islamic Cities of Morocco Author(s): Michael E. Bonine Source: Muqarnas , 1990, Vol. 7 (1990), pp. 50-72 Published by: Brill Stable URL: http://www.jstor.com/stable/1523121

REFERENCES Linked references are available on JSTOR for this article: http://www.jstor.com/stable/1523121?seq=1&cid=pdf- reference#references_tab_contents You may need to log in to JSTOR to access the linked references.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at https://about.jstor.org/terms

Brill is collaborating with JSTOR to digitize, preserve and extend access to Muqarnas

This content downloaded from 196.12.203.5 on Mon, 14 Sep 2020 14:01:20 UTC All use subject to https://about.jstor.org/terms MICHAEL E. BONINE

THE SACRED DIRECTION AND CITY STRUCTURE: A PRELIMINARY ANALYSIS OF THE ISLAMIC CITIES OF MOROCCO

Turn your face toward the Sacred Mosque [the Ka'ba in Mecca]; wherever you may be, turn your faces toward it. Qur'an 2:144.

City planning in Islam appears to have been rather King,2 for instance, based on an analysis of Arabic rudimentary compared to planning in the cities of an- historical, legal, and astronomical texts, has shown evi- cient Greece and Rome; public space and buildings dence that the qibla was actually determined in several were less common, and the strict grid pattern of the ways and consequently the qibla in a given location classical city was unknown. Yet in most pre-modern could end up being a number of possible different direc- cities of the Islamic Middle East a clearly orthogonal tions. Acceptable ways of determining the qibla in- structure of housing and streets is manifest. In an earli- cluded some based on astronomical sighting, such as er analysis of Iranian cities I demonstrated that their the rising or setting of the sun or the rising of the star orthogonal morphology resulted principally from the Canopus, and on the prevailing winds. The method of rectangular irrigation and field systems into which determining the qibla could also change over time at housing spread and that slope characteristics played the any particularly location, leading to rather different predominant role in the orientation of those systems.' directions for the qibla even within the same city. One factor which was investigated as a possible influ- Both approximate and exact mathematical methods ence on the orientation of Iranian cities was the (based on latitude and estimates of longitudes) for find- mosque. In Islam, Muslims pray toward the Ka ing 'ba the in qibla were widely known in the medieval Islam- Mecca, the sacred direction, or qibla. The mosque ic world had from the ninth century onward. Astronomical then to have a qibla wall where the mihrab or niche handbooks was (zijes) usually included chapters on the de- located that faced toward Mecca, and this almost termination al- of the qibla by such mathematical means, ways rectangular building also ought to have hadtables rec- for latitude and longitude, and sometimes qiblas tangular streets around it, which would also be orthogo-for important cities.3 Yet even using these medieval nal to the qibla. If the mosque were built first, as mathematical it often calculations, many, if not most, medieval was in newly established Islamic cities, then the mosquesstreet are not properly aligned toward Mecca. As pattern could be expected to evolve around and King from has noted: the mosque. In Iran, however, no such correlation between the qibla direction and the street pattern could The be earliest qibla determinations were, in fact, associated established; in those few cities that were oriented to with the the risings and settings of the sun and fixed stars, and mosque orientations in the seventh and eighth centu- qibla the relation appeared to be a coincidence of ries,slope and even thereafter, were made by astronomical and direction to Mecca. Possibly this was because alignments.most Thus, for example, some of the earliest Iranian cities existed before Islam, and so the basic mosquescity in Egypt and Andalusia faced the rising sun at structure and orientation were established before the midwinter, and some of the earliest mosques in Iraq, Islamic period. Iran, and Transoxania (Central Asia) faced the setting sun at midwinter. Directions perpendicular to the solsti- Some relatively recent scholarship provides new in- tial directions were also used. Cardinal orientations were sight into the orientation problem. Work by David popular as well, even where they were really quite inap-

This content downloaded from 196.12.203.5 on Mon, 14 Sep 2020 14:01:20 UTC All use subject to https://about.jstor.org/terms SACRED DIRECTION AND CITY STRUCTURE 51

propriate. Occasionally, of course, The existence mosques of the weremedinas built(actually onor archaeolog- the sites of churches and pagan ically) temples, means that thewithout structure mod-and orientations of ification of the orientation of thethese earliercities can edifices.4be fairly well determined, although Hence even though the mathematically dating the foundation "correct" of particular sectionsqibla of a city might have been available always in remainsthe medievala problem. Doing so,Islamic however, assumes world, many other qiblas were that evidently the present orientation found is basically accept- the same as the able in the Middle East at the same time. original layout, but that assumption is reasonably What is the significance of all this for the structure sound. of Archaeological excavations of cities in the Mid- the city? King in fact makes the statement that "entire dle East tend to confirm that the basic street patterns cities with more-or-less orthogonal street plans were and building orientations have not changed in most sometimes laid out facing either an accepted astrono- instances, particularly with the cultural continuity such mical orientation for the qibla or a mathematically com-as existed during Islamic times. Mosques (or their qibla puted qibla. Religious architecture in these cities could walls) in some instances have been reoriented in later thus be aligned with the street patterns."5 He does note, periods during reconstruction or renovation, a situation however, that "more commonly, mosques would which be could present major problems for their study. oriented to the qibla regardless of the street patterns, However, or this practice seems to have been rather rare in would be aligned externally with the street patterns andMorocco. internally with the qibla."6 In another publication onIn the summer of 1985 a survey of a number of cities medieval Cairo, King states that three sections of Mam- of Morocco and Tunisia was conducted.'" The qibla luk Cairo are oriented in three different directions cor- direction of mosques and madrasas was determined by responding to three distinct ways of determining the compass readings; information was gathered about the qibla.7 history of particular buildings and cities; compass read- It remains to be established, then, what the correla- ings and slope determinations of streets were made; and tion between mosque orientation, street patterns, and maps and aerial photographs were obtained for many of city structure really is. Even in the Cairo study, King the cities." In this paper only some of the results of this does not provide any evidence of actual mosque orien- work in Morocco will be discussed. Although a number tations, or of their relationship to street orientations, of Moroccan cities were surveyed, the paper will con- and neither does he consider the influence of slope on centrate on six of the major medinas: Rabat, Sale, the direction and orientation of the sections of the city. Meknes, Marrakesh, Fez, and Taza. All six of these If Islamic cities, or even some of them, were indeed settlements were founded in Islamic times, providing laid out in relation to the qibla it is most significant. It excellent case studies to test the theses previously men- would mean that Islam played a major role in the actual tioned. layout of cities; that a sacred direction and orientation, at least in some instances, was important in city plan- MOROCCAN QIBLAS IN TIME AND SPACE ning. Yet, there still is no real evidence to show that Islamic city structure was influenced by the sacred In Morocco the true direction to Mecca (the direction of direction to Mecca. a great circle route [= shortest distance] to Mecca) The Maghreb (northwest Africa) and specifically varies from an azimuth of 970 in the north of Morocco Morocco provide an area in which these ideas can (e.g., be Tangiers and Tetouan) to 910 in the southern city tested. This region is especially noted for a great num- of Marrakesh. Yet, considering only the principal ber of cities newly established in early and medieval mosques and shrines of each city (generally thejami or Islam, and unlike the eastern Islamic region of South- Great Mosque), one can see the great variety of read- west Asia with its long pre-Islamic urban heritage, ings (and hence orientations) (table 1, fig. 1).12 It is many of these cities of the Maghreb were entirely readilynew apparent that the qiblas vary considerably from settlements. Hence, as Abu-Lughod8 has pointed theout, "correct" azimuth to Mecca.'3 Although the pos- the Maghreb is a prime area for investigating Islamic sible reasons for the various orientations are beyond the urbanism. Equally as important, the medinas (old scope ci- or purpose of this paper, it can be speculated at ties) of many of these towns still exist, partly a result this of point that the great number of orientations in the the "serendipity of conservation" that stemmed frommid-to-late 1500's (adjusted to true north) may be due French colonial policies.9 to the similar alignment of the Ka'ba and the 120's and

This content downloaded from 196.12.203.5 on Mon, 14 Sep 2020 14:01:20 UTC All use subject to https://about.jstor.org/terms 52 MICHAEL E. BONINE

Table 1. Qiblas of Major Moroccan Mosques and Shrines

City Name of Building Date of Founding Period of Founding Qibla Direction Correct Direction (Azimuth*) to Mecca**

Azemmour Great Mosque ? ? 1280 920 53' Casablanca Great Mosque ? ? 1350 93041' Chechauen Great Mosque ? ? 1480 960 52' El Jadida Great Mosque ? ? 1310 920 45' Essaouira Great Mosque 1764 + Alawite 1040 900 17' Fez Qarawiyyin Mosque 9th c. (859) Idrisid 1630 950 47' Fez Andalousian Mosque 9th c. Idrisid 1510 950 47' Fez (Jadid) Great Mosque 13th c. (1289) Merinid 1570 950 47' Marrakesh Ya(qub al-Mansur Mosque (Qasba Mosque) late 12th c. Almohad 1590 910 21' Marrakesh Kutubiya Mosque 1) 1147 1) Almohad 1) 1540 91021' 2) 1162 2) Almohad 2) 1590 91021' Meknes Great Mosque mid-14th c.? Merinid 1510 950 16' Mulay Idris Mausoleum of Idris I 1660/1690 Alawite 1240 950 28' Qsar es Seghir Great Mosque late 13th c. Merinid 1450 970 26' Quazzane Great Mosque 1727 or later Alawite 1090 960 38' Rabat Great Mosque ? ? 1390 940 36' Rabat Hasan Mosque (Tour Hasan) 1195 Almohad 1550 940 36' Rabat Qasba Mosque 12th c.? Almohad 1530 940 36' Sale Great Mosque late 12th c. Almohad 1240 94039' Sefrou Great Mosque ? ? 1510 950 39' Tangier Great Mosque late 17th c. Alawite 1370 970 12' Taza Great Mosque mid-12th c. (1142- 1146) Almohad 1540 960 38' Tetouan Great Mosque ? ? 1260 970 15' Tinmal Great Mosque 12th c. Almohad 1570 900 33'

130?'s based perhaps on the Greatmathematical Mosque incalculations Sal6. The Merinid period (mid-thir- and tables of the astronomers, teenth-early-fifteenth or on the winter sunrise, century), on the other hand, or were simply a mid-point betweenshows a greatereast and variety, south. Itwith is some concentrations in the apparent that in medieval Islamic1200's asMorocco well as thethe prob-1500's and with one building, the lems of measuring direction Madrasa on a spherical 'Attarin surface in Fez, as being almost directly south well as the problem of the correct (1820) distance- a direction were eitherfound for a few other buildings not properly understood or, moresurveyed likely, (whose mathematical- dates of origin are unknown), and ly calculated qiblas were being which ignored also inwas preference one of the to accepted qiblas in parts of the other ways of determining it. Islamic In any Middlecase, it Eastshould and be Spain. stressed that the modern ("correct") Only the qiblaAlawite values period are (mid-seventeenth century- basically irrelevant to any discussion present) ofshows medieval a rather archi- significant deviation from the tecture and city planning. pattern, and this late period's qiblas approach the But what are the qibla directions? known Are"correct" there directionvarious of Mecca more closely. It is orientations by specific dynastic important, periods? in Unfortunate- fact, to note that the cities of Essaouira ly, the paucity of available (and (formerly sometimes named conflicting) Mogador) and Ouazzane were found- information from the literature ed onlyon the in date the ofeighteenth origin of century, and Moulay Idris buildings means that only some began structures to grow canas a be town com- only after the building of the pared (table 2).'" The Almohad mausoleum period (mid-twelfth-of Idris I in the mid-seventeenth century thirteenth century) is the most (even consistent; though Idris all the I had qibla died at the end of the eighth readings are in the mid-to-late century). 150"'s Oneexcept pattern for the which is discernible over the

This content downloaded from 196.12.203.5 on Mon, 14 Sep 2020 14:01:20 UTC All use subject to https://about.jstor.org/terms SACRED DIRECTION AND CITY STRUCTURE 53

"Correct" Oibla QSAR ES SEGIR (Great Circle Rout.) - 970

TE TOUAN

CHECHAUEN

OUAZZANE 10os9.

TAZA

SALE MULAY IDRISS FEZ RABA 1' R BMEKNES0 SEFROU

CASABLANCA

EL JADIDA AZEMMOUR r r o

E AMARRAKESHCorrecto Qibla (Great Circle Route) - 91*

TINMAL

1. Qiblas of major Moroccan mosques and shrines. From field data as tabulated in table 1.

This content downloaded from 196.12.203.5 on Mon, 14 Sep 2020 14:01:20 UTC All use subject to https://about.jstor.org/terms 54 MICHAEL E. BONINE

Table 2. Qiblas of Moroccan Religious Architecture by Dynastic Period

City Building Date of Qibla Direction Correct Direction Founding (Azimuth*) to Mecca**

Idrisids (late 8th-early 10th c.) Fez Qarawiyyin Mosque 9th c. (859) 1630 950 47' Fez Andalousian Mosque 9th c. 1510 950 47' Almoravids (late 11th-mid-12th c.) (No buildings confirmed) Almohads (mid-12th c.-13th c.) Marrakesh Ya(qub al-Mansur Mosque (Qasba Mosque) late 12th c. 1590 91021' Marrakesh Kutubiya (1 & 2) 1) 1147 1) 1540 91021' 2) 1162 2) 1590 91021' Rabat Qasba Mosque 12th c. 1530 940 36' Rabat Hasan Mosque (Tour Hasan) late 12th c. (1195) 1550 94036' Sale Great Mosque late 12th c. 1240 940 39' Taza Great Mosque mid-12th c. (1142-46) 1540 960"38' Taza Sidi Azouz Mosque 12th c. 1550 960 38' Tinmal Great Mosque 12th c. 1570 900 33' Merinids (mid-13th-early 15th c.) Fez (Jadid) Great Mosque 1289 1570 950 47' Fez (Jadid) El Hamra Mosque 14th c. 1500 950 47' Fez Madrasa Sahrij 1321-23 1370 950 47' Fez Madrasa Bu (Inaniya 1350-55 1360 950 47' Fez Madrasa (Attarin 1325 1820 950 47' Fez Madrasa Saffarin 1271 1200 950 47' Fez Sherabliyyin Mosque 14th c. 1640 950 47' Fez Zawiya Mulay Idris early 15th c. 1720 950 47' Fez Bab Guissa Mosque 14th c 1460 950 47' Fez Abu al-Hasan Mosque 1341 1600 950 47' Meknes Great Mosque mid-14th c.? 1510 950 16' Meknes Zaituna Mosque 14th c. 1450 950 16' Meknes Madrasa Bu (Inaniya mid-14th c. 1500 950 16' Marrakesh Mu (assin Mosque ? 1400 910 21' Marrakesh Ben Salih Mosque ? 1470 91o 21' Rabat Great Mosque ? 1390 940 36' Rabat (Chellah) Abu Yusuf Mosque early 14th c. 1290 940 36' Rabat (Chellah) Zawiya Abu'l Hasan early 14th c. 1270 940 36' Sale Sidi Ben Ashir 14th c. 1270 94039' Sale Madrasa Abu'l-Hasan 1335/1342 1290 Taza Madrasa Abu'l-Hasan 1323? 1540 960 38' Qasr es Seghir Great Mosque late 13th c. 1450 970 26' Sa (dians (early 16th c.-mid-17th c.) Marrakesh Madrasa Ben Yusuf 1562 1460 910 21' Marrakesh Bab Dukkala Mosque 1557 1250 910 21' Marrakesh Sa (dian Tombs late 16th c.- early 17th c. 1600 91021' Marrakesh (Badi( Palace) late 16th c. 1650 91021' Marrakesh Zawiya Sidi Bel- (Abbas 16th c. 1650 91021' Alawite (mid-17th c.-present) Fez Madrasa Sharratin 1670 1380 950 47' Fez Er-Rsif Mosque 18th c. 940 950 47' Meknes Mausoleum of Mulay Isma(il 1727 + 1050 950 16'

This content downloaded from 196.12.203.5 on Mon, 14 Sep 2020 14:01:20 UTC All use subject to https://about.jstor.org/terms SACRED DIRECTION AND CITY STRUCTURE 55

Table 2. (Continued)

City Building Date of Qibla Direction Correct Direction Founding (Azimuth*) to Mecca**

Meknes Lalla Auda Mosque late 17th c.- early 18th c. 1360 950 16' Meknes Berrima Mosque 18th c. 1300 950 16' Marrakesh Madrasa Ben Salah 18th c. 1480 91o 21' Marrakesh Ben Yusuf Mosque 19th c. 880 910 21' Marrakesh Berrima Mosque ? 1090 91o 21' Rabat Ahl al-Fahs Mosque 18th c. 1070 or 1100 940 36' Rabat Mulay al-Mekki Mosque 18th c. 1580 940 36' Essaouira Great Mosque 18th c. 1040 900 17' Mulay Idris Mausoleum of Idris I 1660/1690 1240 950 28' Ouazzane Great Mosque (Mulay Abdallah Sharif) 18th c. 1090 960 38' Tangier Great Mosque end of 17th c. 1370 970 12'

various periods (in some instances) Mosque (est. 1289) is and a royal consistency palace were founded of in the qiblas for specific cities. the new Now city. Succeeding we shoulddynasties, such turn as the Sa(dians to analyzing qiblas and the orientation and Alawites, also of added these a few more cities. mosques, shrines, and madrasas, as well as renovating a number of build- THE QIBLA AND MOROCCAN CITY ings. STRUCTURE The qiblas of the surveyed buildings indicate a wide and rather bewildering variety of orientations, ranging Fez. Fez was originally a small Berber town from constructed 940 to 1820 in Fez el-Bali and from 900 to 1570 in at the end of the eighth century by Idris FezI. However,Jadid (table 3, figs.the 2 and 4). The two major ninth- real foundation of the city dates from a settlement century Great estab-Mosques are generally in the approxi- lished by Idris II at the beginning of the mateninth direction century of the 1500's found for the early mosques on the left bank of the Oued Fez, a tributary of Morocco -of Qarawiyyin the at 1630 and Andalusia at Sebou River. Refugees from Qayrawan 1510, (Kairouan) and the qibla of in the thirteenth-century Merinid Tunisia constituted a major proportion Great of Mosque the in pop- FezJadid is also at 1570. The Merinid- ulation, and hence the major mosque, which period originally mosques and madrasas provide the greatest va- was founded in the mid-ninth century, riety was of orientation,called theconfirming that, at least for this one Qarawiyyin Mosque. Meanwhile, at about period, the either same there was no one accepted qibla (at least time on the right bank a separate town developed one which was with translated a in the siting of the monu- population ofAndalusians who had been banished mental architecture), from or there was no attempt to adhere Cordoba in Spain; they called their congregational to a particular direction (other than generally the south- mosque the Andalusian Mosque. Each of west these quadrant). towns developed separately around their Great Mosque,The correspondence until with the street and housing pat- they were enclosed within a single wall by tern the also showsAlmora- a rather confusing pattern (cf. figs. 2 vids (1060-1145). This later became known and 3). Theas irregularFez el- street orientations in Fez el-Bali Bali. are determined basically by the irregular, hilly topo- Fez was the capital of Morocco under the Merinids, graphy. The orientation of the houses, particularly evi- from 1248 to 1428. They not only built many mosques dent where field patterns and houses can be observed at and madrasas, but also founded a new town (FezJadid) the edges of the settlement, confirm this interpretation. in the thirteenth century to the southwest of the old The mosques and madrasas fit into this network in a medina (but not to be confused with the French Protec- way that is difficult to interpret. The buildings some- torate modern city, the Ville Nouvelle). A new Great times are oriented in the direction of the streets, which is

This content downloaded from 196.12.203.5 on Mon, 14 Sep 2020 14:01:20 UTC All use subject to https://about.jstor.org/terms 56 MICHAEL E. BONINE

Table 3. Qiblas of Religious Architecture in Fez

Identity Code Building Date Period Qibla Comments (Azimuth*)

("correct" qibla) 950 47' Great Circle route Q Qarawiyyin Mosque 9th c. (859) Idrisid 1630 Exact reading on mihrab (through door), verified in air photo A Andalousian Mosque 9th c. Idrisid 1510 Reading on side of inner courtyard, verified in air photo sf Madrasa Saffarin 1271 Merinid 1200 Exact reading on mihrab GM Great Mosque 1289 Merinid 1570 Reading to inner courtyard, verified on air (Fez Jadid) photo Sj Madrasa Sahrij 1321-23 Merinid 1370 Reading on sides MA Madrasa (Attarin 1325 Merinid 1820 Exact reading on mihrab AH Abu al-Hasan Mosque 1341 Merinid 1600 Reading on inner courtyard. Cannot locate on air photo BI Madrasa Bu (Inaniya 1350-55 Merinid 1360 Exact reading on mihrab, ID on air photo not positive H (Fez Jadid) El Hamra Mosque 14th c. Merinid 1500 Reading on inner courtyard, verified on air photo Sh Sherabliyyin Mosque 14th c. Merinid 1640 Reading on sides BG Bab Guissa Mosque 14th c. Merinid 1460 Reading on sides, verified on air photo MI Zawiya Mulay Idris early 15th c. Merinid 1720 Reading on sides, verified on air photo MS Madrasa Sharratin 1670 Alawite 1380 Exact reading on mihrab R Er-Rsif Mosque 18th c. Alawite 940 Reading on sides, verified on air photo SA Sidi Ahmed Tijani Mosque ? ? 1590 Reading on sides O Madrasa al-Oued ? ? 1280 Reading on sides SL Sidi Lezzaz Mosque ? ? 1660 Reading on sides. Cannot locate on air photo T Tijania Mosque ? ? 1300 Reading on sides. Cannot locate on air photo M El Menia Mosque ? ? 1570 Reading on sides Ab Mulay Abdallah Mosque ? ? 900 Reading on inside courtyard, verified on air (Fez Jadid) photo BJ (Bab Bu Jelud Mosque) ? ? 1390 Reading on sides. Name not known

* Magnetic declination in Morocco in summer 1985 was approximately 60 west of true north, and the true north readings have been adjusted by subtracting 60 from the compass reading. Directions are expressed in azimuths from true north. also emphasized by the buildings along cating Rue areasTalaa whichKe- normally would be built for sym- bira, the qibla being generally at a right metry. angle Similar of this examples of irregular mosque shapes major street, even as it arcs through much and differentof the city. orientations of the qibla to the street pat- But in a number of instances the building tern haveorientation been noted by Kessler for Cairo.'5 appears to have little correspondence with It theappears, street at or least in the case of Fez, that the qibla, housing pattern. Even in Fez Jadid with and its hencemore regu-the orientation of mosques and madrasas, lar street pattern (and less steep and has less little irregular influence on the street pattern or the morphol- topography), the Great Mosque has onlyogy ofa minimal the settlement. Topography is the overriding area of similarly oriented streets and houses, determinant the Hamra of the orientation of streets and housing. In Mosque is slightly out of kilter with the fact, street in Fezpattern, it appears that the great variation in direc- and the Mulay Abdallah Mosque is at tionan oblique of the angleqibla and the orientation of religious struc- to the street orientation (cf. figs. 4 and tures 5). is also a response to variation in slope, although The way in which some of the religious this structures relationship are does not always hold. Even though oriented differently, but have fit into themosques street andpattern madrasas are often oriented in the same nevertheless, can be seen in the plans of direction the Qarawiyyin as immediate, adjacent streets, the variation and Andalusian mosques and the Madrasa in qibla orientations,Bu 'In- the changing orientation, and the aniya (fig. 6). The buildings themselves greatare very irregularity often of the streets all seem to confirm that irregular, filling out the space between streets,the orientation or trun- of the streets and the morphology of Fez

This content downloaded from 196.12.203.5 on Mon, 14 Sep 2020 14:01:20 UTC All use subject to https://about.jstor.org/terms SACRED DIRECTION AND CITY STRUCTURE 57

...... - .

orr~ectew

Oibla - U#96o

tar <.<5. hSh

L 0 HMA

.... Y VE BAByy Y y.y

L A HALPO ,Y y Ab EL y y ~ 1 NO;Z M4 RA 13y Y yy. yY %A 6 AIAF Qt 9 EL AROA T y

2. Qiblas in Fez el-Bali. From field data as tabulated in table 3. Base map from Plan-Guide de Fes, 1:10,000. (See table 3 for building identifications.) evolved independently for the Lalla Fidila Mosque. A major exception, how- of the direction of the qibla. ever, is the mausoleum of Mulay Isma>il (d. 1727), Meknes. Meknes, where the qiblas of two separate mihrabsfounded were mea- in the tenth century by a Zenata Berber sured, one at 1120 (Magnetictribe, N) and another at 1100was fortified by the Almoravids and provided (Magnetic N) andwith the true north direction numerous of 1050 was mosques and madrasas by the Almohads and Merinids. Yet it remained a rather taken as an adjusted average of the two. insignificant town until becoming the Alawite capital Meknes,in like Fez, is built on rather hilly land. The the seventeenth century. As one of the four imperial streets and housing orientation have a most irregular cities, its glory is confined particularly to Sultan Mulay pattern (fig. 8). Residential districts of the medina Isma'il (1672-1727), a rather cruel, capricious ruler, curve consistently and even swirl about; the religious but one who wanted to make Meknes one of the most structures often incise and interrupt this pattern. An glamorous cities of the world. Toward that end, exceptionthe to this irregularity are the newer housing royal gardens and royal palace south of the old city were districts immediately south of the old medina. For in- established by the ambitious sultan. stance, the eighteenth-century Berrima Mosque has a Religious monuments in the southern part of qiblathe of 1300, and the streets are exactly orthogonal to medina were surveyed (table 4, figs. 7 and 8). thisThe direction. This axis of 2200 is at a right angle to the orientation of the qibla is generally in a similar direc- contours of a rather gentle slope. The oldJewish Mellah tion, ranging from 1360 for the Najarine Mosque (reput- (quarter) south of Berrima changes orientation, and in ed to be the oldest surviving building in the city) to 1540fact the slope has changed, and the Mellah streets are at

This content downloaded from 196.12.203.5 on Mon, 14 Sep 2020 14:01:20 UTC All use subject to https://about.jstor.org/terms 58 MICHAEL E. BONINE

3. Aerial photograph of Fez el-Bali. From Service G6ographique et Topographique, Rabat. (See table 3 for building identifications.) right angles to that ben Tashfin slope. as the imperial capitalThis of the Almoravids.slope continues to the southeast, becoming Conquering Muslimeven Spain toward more the end of thegentle. elev- It is this area where the royal enth-century,gardens this Berber dynasty(now was also aresponsible golf course) and the royal palace were for built bringing Andalusian by culture Mulay to embellish the capitalIsma)il, on an axis of approximately 1150 (and other cities).and The sonat of the afounder, right (Ali b. Yusuf, angle to the slope. The mausoleum ofbuilt the Mulayhuge walls to enclose Isma)ilthe city that still stand at the edge of the royal developments today; however, is theseoriented walls did not prevent theat Almo- 1050, approximately the same direction. hads from capturing the city in 1147. Destroying most of Conclusions concerning the main medina of Meknes the Almoravid monuments, these Berber tribesmen are similar to those of Fez; the orientation of the qibla soon built their own mosques and palaces (including does not affect the streets and housing orientation; slope the Kutubiya and Qasba mosques). The Almohads by is the determining factor. Yet, in the less steep areas the end of the twelfth century transferred their capital to south of the medina, streets and housing are the same Rabat, and Marrakesh was not a royal capital again direction, and the orientation is at right angles to untilthe the Sa'dians came to the city in 1521. main slope. In general, it is true that the main axis of the The qibla orientations of Marrakesh present a now medina is in the same basic direction as the qiblas familiar pattern - a wide variety of directions, ranging within the medina. The significance of these patterns from 880 for the Ben Yusuf Mosque to the 160?'s for a will be dealt with later. number of mosques, and even 1710 (table 5, fig. 9). The Ben Yusef Mosque, in fact, presents a most interesting Marrakesh. Marrakesh was founded in 1062 by Yusuf case. At this site the Great Mosque of the Almoravids

This content downloaded from 196.12.203.5 on Mon, 14 Sep 2020 14:01:20 UTC All use subject to https://about.jstor.org/terms SACRED DIRECTION AND CITY STRUCTURE 59

\ ?Z44R? YMAWd

Y CYY Y .V I u ~ Atq CINEVI1RE ? ' " In"e :r - s - " COrrecto Oi ila MI 960 y lt ...... -..

": " ":~ '.": ,II3jr= ...... ".-...? ..T OLMIT 4: BAN ? C .:' .

?e . '', .

MECHDUR0M, . QARAVIYYIN NOSQUE 7

am 4. Qiblas in Fez Jadid. From field data as tabulated in table 3. Base map from Plan-Guide de Fes, 1:10,000. (See table 3 for building identifications.)

-~~~ ~':~P~Td ' -L~\-.-I. ? N"RASA &V *'lNANIYA

ANOALOUS' POSQU .

6. Irregular building shapes and street patterns: examples from Fez. From Gaudio, 1982; Michelin, Maroc, 1975.

was built by 'Ali b. Yusuf. However, only a nearby small cupola, the Qubba Barrudiyyin, which was part of the Almoravid palace, remains from this period. The present Ben Yusuf Mosque dates from the nineteenth century, and hence its much more "accurate" qibla of 880. However, Deverdun, in a plan reconstructing the original orientation of the Great Mosque, indicates an axis of approximately 1090 (fig. 10).16 The Almohad buildings, the Kutubiya and Qasba mosques, are oriented in the familar (Almohad) direction of approximately 1500. The Kutubiya provides a most interesting example of a slight change in orientation arising from a different calculation of the qibla. The first Kutubiya was built in 1147 and the qibla 5. Aerial photograph orientation isof 1550. ItFez evidently Jadid. was destroyed afterFrom only Service Geographique et Topographique, Rabat. a few years, and(See at least bytable 1162 a second 3 Kutubiya for was building identifications.)

This content downloaded from 196.12.203.5 on Mon, 14 Sep 2020 14:01:20 UTC All use subject to https://about.jstor.org/terms 60 MICHAEL E. BONINE

Table 4. Qiblas of Religious Architecture in Meknes

Identity Code Building Date Period Qibla Comments (Azimuth*)

("correct" qibla) 950 16' Great Circle route GM Great Mosque mid-14th c.? Merinid 1510 Reading on sides, verified on air photo BI Madrasa Bu (Inaniya mid-14th c. Merinid 1500 Exact reading on mihrab Z Zaituna Mosque 14th c. Merinid 1450 Reading on sides, verified on air photo LA Lalla Auda Mosque late 17th-early 18th c. Alawite 1360 Measured only on air photo MI Mausoleum of Mulay Isma(il 18th c. (c. 1727) Alawite 1050 Exact readings on two mihrabs, one at 1060 and one at 1040 B Berrima Mosque 18th c. Alawite 1300 measured only on air photo N Najarine Mosque ? ? 1360 Reading on sides, verified on air photo. Reputed by populace to be the oldest mosque LF Lalla Fidila Mosque ? ? 1540 Reading on sides, not positive ID on air photo (Mosque of the Suq) ? ? 1420 Exact reading on mihrab, located at the entrance of the suq, name unknown

constructed adjacent to part of the older building. Yet, CIMETIEBE HE sib i? J 8~h~d "Correcto aijbia ,95 it has a qibla of 1600, and the axis of the building was 'SEN.- AISSA I y.?... changed accordingly, which is evident in the plan of the building as well (fig. 11). Obviously, in this case the qibla was deemed incorrect, and a change was institut- ed to fit a different calculation (which, of course, actual- Y C IN:? ? ly is even farther from the "correct" qibla of 910 for Marrakesh). The orientation and location of the Almohad ILAI mosques, both Kutubiyas and the Qasba Mosque, may Y Y y be significant in terms of the morphology of Marrakesh. In the south, the axis of the city is approximately 1660, RNY y U?M Fig~i~T MAR HAS- which is exactly at a right angle to the main, actual S3SSW upslope (fig. 12). The Qasba area, the Sa(dian Badi ~:~?::::~-~~L~:~~?~~FL Palace, the nineteenth-century Bahia Palace, the nine- i:6i`?;-=-js m teenth-century Agdal Gardens, and the sixteenth-century Jewish Mellah (quarter) are all oriented in this same basic direction. The slope in this case has deter- K.-- mined the orientation of most of the street pattern and y-~:. ~ j ?:-Z, the morphology of Marrakesh in the south - a very

MONA V MELMNA~nr regular, orthogonal pattern. The correspondence of the slope with the qibla of the Almohad mosques and their

7. Qiblas in Meknes. From field location data considerably as tabulated south of the in main table Almohad 4. medi-Base map from Plan-Guide de Meknes, na could 1:10,000. mean that (Seethe specific table sites 4 of for these buildingmosques identifications.) may have been selected for their slope characteristics.

This content downloaded from 196.12.203.5 on Mon, 14 Sep 2020 14:01:20 UTC All use subject to https://about.jstor.org/terms SACRED DIRECTION AND CITY STRUCTURE 61

N;~l 114iii'i 41r: "r "...... ":(.*:, ,4.

S0, 4- 7 4L

...."

... ..

? ..; r .7 F qo.

8. Aerial Photograph of Meknes. From Service Ge6graphique et Topographique, Rabat. (See table 4 for building identifications.)

The differently oriented Berrima EvenMosque though (1090) theis Ala- variety of qibla directions cannot be wite, a much later period, when explainedthe calculation without of themore information on the dates of qibla had obviously changed considerably. origin for the buildings, the religious structures have not The core of the medina, however, influenced is as perplexing the basic as street patterns or housing orien- ever. Irrespective of the orientation tations of the in original most of Ben the medina. Although certainly less Yusuf Mosque, the houses and streets steep aroundthan Fez the or area Meknes, slight variations in slope are oriented in all sorts of directions. probably Throughout are accounting most for these patterns as well. In of the city the mosques and madrasas the northeastern often do partnot of the city the slope down to the correspond to the flow of the housing Oued (cf. Issid figs. is slightly9 and 13). irregular, with the area near Bab

This content downloaded from 196.12.203.5 on Mon, 14 Sep 2020 14:01:20 UTC All use subject to https://about.jstor.org/terms 62 MICHAEL E. BONINE

Table 5. Qiblas of Religious Architecture in Marrakesh Identity Code Building Date Period Qibla Comments (Azimuth*) ("correct" qibla) 91021' Great Circle route K Kutubiya 1 1147 Almohad 1540 Reading on old mihrab K Kutubiya 2 1162 Almohad 1590 Reading on sides, verified on Marrakesh, 1:6,500 Y Ya(qub al-Mansur late 12th c. Almohad 1590 Reading on sides, verified on (Qasba) Mosque Marrakesh 1:6,500 M Mu(assin Mosque ? Merinid 1400 Reading on sides, verified on air photo B Ben Salih Mosque ? Merinid 1470 Measured only on air photo and Marrakesh 1:6,500 BD Bab Dukkala Mosque 1557 Sa (dian 1250 Reading on sides, verified on Marrakesh 1:6,500 MBY Madrasa Ben Yusuf 1562 Sa (dian 1460 Exact reading on mihrab SB Zawiya Sidi Bel- (Abbas 16th c. Sa (dian 1650 Measured only on Marrakesh 1:6,500 BP (Badi( Palace) end of 16th c. Sa(dian 1650 Reading on wall (axis of palace), verified in Marrakesh 1:6,500 ST Sa(dian Tombs late 16th-early 17th c. Sa (dian 1600 Exact reading on mihrab in the tomb complex MB Madrasa Ben Salah 18th c. Alawite 1480 Measured only on air photo and Marrakesh 1:6,500 BY Ben Yusuf Mosque 19th c. Alawite 880 Reading on inner courtyard, verified on air photo. A reconstruction and reorientation on site of older mosque Br Berrima Mosque ? Alawite 1090 Reading on sides, verified in Marrakesh 1:6,500 SM Sidi Mimun Mosque ? ? 1470 Measured only on Marrakesh 1:6,500 BP [Bahia Palace Mosque] ? ? 1620 Measured only on Marrakesh 1:6,500. Exact name unknown RJ Riad ez-Zaitun Mosque ? ? 1350 Measured only on Marrakesh (El Jadid) 1:6,500. No positive ID on air photo RK Riad ez-Zaitun Mosque ? ? 1200 Measured only on air photo (El Kadim) and Marrakesh 1:6,500 Gs Guessabin Mosque ? ? 1210 Measured only on air photo and Marrakesh 1:6,500 SY Zawiya Sidi Yusuf Ben Ali ? ? 1410 Measured only on Marrakesh 1:6,500 SI Sidi Ishak Mosque ? ? 1120 Measured only on air photo and Marrakesh 1:6,500 H Hart Hart es Sura Mosque ? ? 1080 Measured only on air photo and Marrakesh 1:6,500 Ks [El Ksour]? Mosque ? ? 1710 Measured only on air photo and Marrakesh 1:6,500. No positive ID on air photo S [Mosque of the Suq] ? 1220 Measured only on Marrakesh 1:6,500. Exact name unknown. Cannot locate on air photo BA Sidi Bu Amur Mosque ? ? 1520 Measured only on Marrakesh 1:6,500 MA Mulay al-Sharif Mosque? ? 1270 Measured only on Marrakesh 1:6,500 Si Sidi Ben Sliman ? ? 1600 Measured only on Marrakesh 1:6,500

This content downloaded from 196.12.203.5 on Mon, 14 Sep 2020 14:01:20 UTC All use subject to https://about.jstor.org/terms SACRED DIRECTION AND CITY STRUCTURE 63

-a EWAN~?) rb~Y

yy "Correct" oibla - 91*.

I Yam" ~J sw

wtT V v

BD BYM

v W MA

v: M v ,Si WS

Ks . r ..

V v V v v y v T v y y v~CLDUIUL~tl li: ?UYli.Y v y

r' 1?5 ?c ,d 1

RJ 'MY" M i -~~~~,~ ~ -YA--~ C~~-a~b-U~~l'~~?r~iy y v??

AR VIE Y y y y Yy YY VY VY v VK VVYYY Y ' VYvY V V V Y! V 1 ~~Et~~PPi ~ nr in:~aa 1J n~pgIdr ~-~ec~;sHB~~g_~'b"rrul?GsULA000E v ? 'Y Y v y 'Y v Yv y vY yv T v yy y vv y Y v YT I Yy XY yv y v v11 y v y v I T y T T y y Vy vT Y

v v v cimmisty v v v Ado Amsm'rTY T v y yy 4.'

9. Qiblas in Marrakesh. From field data as tabulated in table 5. Base map from Plan-Guide de Marrakech, 1:6,500. (See table 5 for building identifications.)

This content downloaded from 196.12.203.5 on Mon, 14 Sep 2020 14:01:20 UTC All use subject to https://about.jstor.org/terms 64 MICHAEL E. BONINE

0I . A1T 'k- I ?r0.V SEt ZAOUIA AL-LAKHAR (UAZTIER

rvr It %t

l ARET

.7.

MW It:

-W. -AL?

ALMORAVID FONTANNE . 'f It-

S Io 3 o 5 c M.c Pss

A'Iy 11

10. Ben Yusuf Mosque in Marrakesh. From Deverdun, 1966.

X-ii A Ku Xp V. N oke.-: %ei Ill-4 A

SALMoLAD

( If ...... I

DAI MOuVL AiALI 12. Marrakesh axis and main slope. Contour interval = 10 meters. Base map from Carte du Maroc, 1:50,000 (Marrakech, Medina). o me Debbarah, for instance, being higher than the area south of it.'7 Detailed slope characteristics, however, 11. Kutubiya Mosque in Marrakesh. From Parker,would 1981. require an extensive field survey, and so the exact slope conditions cannot yet be properly evaluated.

Rabat and Sal. Astride each side of the mouth of the Bou Regreg, Rabat and Sale have been competing settlements for almost a millennium. In the tenth century a

This content downloaded from 196.12.203.5 on Mon, 14 Sep 2020 14:01:20 UTC All use subject to https://about.jstor.org/terms SACRED DIRECTION AND CITY STRUCTURE 65

.....:' ;; '? f' "@Corre to tbl

R A-1 A T e: "-r-:,... " ~YY YI 9 ......

OiblaCM 13

;: :: .....

...- .C=

14. Qiblas in Rabat and Sale. Contour Interval = 5 meters (taken from Rabat-Sale, 1:10,000). From field data as tabulated in table 6. Base map from Plan-Guide de Rabat-Sale, 1:12,500 (See table 6 for building identifications.)

ara. The present Great Mosque was founded in the mid-twelfth century by the Almohads, but the town 13. Aerial photograph of central Marrakesh. From Service Geogra- phique et Topographique, Rabat. (See never table became5 for building as prominent identifica- as Rabat, nor did it decline tions.) as drastically. The Merinids continued to show interest in the town in the thirteenth and fourteenth centuries small fortified monastery or ribat (= rabat) was estab- (and they built mosques and madrasas in Rabat as well, lished on the left bank of the Bou Regreg. A new (or such as at the Chellah, about two kilometers south of rebuilt) ribat was founded in the first half of the twelfth the medina), and as in Rabat, both Andalusians and century by the Almoravid Sultan Tashfin (Ali, al- piracy were important for the town's economy and though the real foundation of Rabat proper came in the culture. These two medinas had fewer major religious mid-twelfth century when the Almohad ruler (Abd al- structures than the previously examined cities, and are Mu)min rebuilt the ribat and constructed huge walls, a somewhat less difficult to analyze. In Rabat the qiblas palace, and a Great Mosque. By the end of the twelfth of the Almohad Qasba Mosque and Tour Hasan (the century, Rabat had become the Almohad capital, and uncompleted mosque) are 1530 and 1560 respectively, the ruler, Abu Yusuf Yaqub, began a most ambitious the Merinid Great Mosque is 1390, and a few other building program which, however, ended with his directions are represented as well (table 6, fig. 14). In death in 1199 and left the great Hasan Mosque (Tour Sale the Almohad Great Mosque's qibla and axis are Hasan) unfinished. The town declined considerably 1240, a rather unusual direction for an Almohad struc- after that until, in the seventeenth century, Andalusian ture (compared to the known Almohad qiblas). The refugees and then piracy helped revive it. orientation of the other religious buildings is basically Sale (Sala) was founded in the early eleventh century the same direction (table 6, fig. 14). on the right bank of the Bou Regreg by the Banu (Ash- What is significant, however, is the relatively regular

This content downloaded from 196.12.203.5 on Mon, 14 Sep 2020 14:01:20 UTC All use subject to https://about.jstor.org/terms 66 MICHAEL E. BONINE

Table 6. Qiblas of Religious Architecture in Rabat and Sale

Identity Code Building Date Period Qibla Comments (Azimuth*)

Rabat ("correct" qibla) 950 36' Great Circle route Q Qasba Mosque 12th c. Almohad 1530 Reading on northwest wall, verified on air photo H Hasan Mosque end of 12th c. Almohad 1550 Exact reading down center where mihrab (Tour Hasan) would be GM Great Mosque ? Merinid 1390 Reading through door, verified on air photo (but difficult to align exactly) (Chelleh) Abu Yusuf Mosque early 14th c. Merinid 1270 Exact reading on mihrab. Located c. 2 km s. of the medina (Chelleh) Zawiya Abu'l-Hasan early 14th c. Merinid 1270 Exact reading on mihrab. Located c. 2 km s. of the medina Ahl al-Fahs Mosque 18th c. Alawite 1070/1100 Reading on sides, but not square: one side 107, another 110. South of medina M Mulay al-Mekki Mosque 18th c. Alawite 1580 Measured only on air photo S Mulay Sliman Mosque early 19th c. Alawite 1270 Measured only on air photo Sali ("correct" qibla) 950 39' Great Circle route GM Great Mosque late 12th c. Almohad 1240 Reading on front door, verified on air photo A Madrasa Abu'l-Hasan 1335/1342 Merinid 1290 Exact reading on mihrab SB Sidi Ben Ashir 14th c. Merinid 1270 Reading on sides, verified on air photo

orthogonal axis of the streets and housing in both medinas. But it should also be noted that the direction of the two networks is slightly different, Rabat's being more southward than Sal's. The streets through the western part of Rabat's medina (Avenue Mohammed V and Rue Sidi Fatah) are oriented at 1580, but the main part of the medina, as shown by many of the small streets as well as the lower part of the Rue des Consuls and the Rue Oukassa, is oriented at 1400-1430. The Almohad mosques are oriented at 1530 and 1560 and the Great Mosque at 1390, which corresponds to the two principal directions. The eighteenth-century Mulay al-Mekki Mosque's axis is aligned exactly with the street, although the nineteenth-century Mulay Sliman Mosque at 1270 does not correspond with the pattern, and other mosques outside the medina are also oriented differently (cf. table 6). In Sal6 the streets are not as straight as in Rabat, but the main axis of the network is approximately 125Y-130' (with a few irregularities). As noted earlier, the qibla directions taken in Sal6 were 1240 and 1290, basically

15. Aerial photograph of Sale. From Service Geographique et Topo- graphique, Rabat. (See table 6 for building identifications.)

This content downloaded from 196.12.203.5 on Mon, 14 Sep 2020 14:01:20 UTC All use subject to https://about.jstor.org/terms SACRED DIRECTION AND CITY STRUCTURE 67

16. Aerial photograph of Rabat. From Service Geographique et Topographique, Rabat. (See table 6 for building identifications.)

the same orientation as the street city network, axis and and most somewhatof the qiblas are in the same different than the direction of direction. either the qiblas or the axis of Rabat (except for the nineteenth-century mosque). Taza. The Taza valley is strategically located in the The housing patterns correspond rather closely to the corridor between the Rif Mountains and the Middle streets of both medinas (figs. 15 and 16). Hence the Atlas and has been the location of various fortresses morphology of Rabat is very regular, and the court- since early times. The town of Taza proper dates from yards of the houses are in the same direction, along the the 1130's when the Almohad ruler, (Abd al-Mu'min, axis. In Sale the courtyards change direction slightly as built a Great Mosque and a wall around the city. Taza the streets curve, creating a wavy effect in the direction served briefly as a capital before the Almohads took of the main axis. Marrakesh from the Almoravids. In the mid-thirteenth Both the direction and degree of regularity of the century the Merinids took the town, and it again served morphology of Rabat and Sal6 are due to the slope as the capital for a short period at the end of the thir- characteristics. Each of the medinas is situated on a teenth century. Many mosques, madrasas, and other promontory, with rather steep drops to the Atlantic public and buildings were built or repaired by the various the Bou Regreg. But the medinas themselves slope dynasties, and the town continued its strategic impor- downward more gently to the southeast. The maintance even up to the twentieth century. The region of slopes are in fact in the same directions as the main Taza axes is quite hilly and irregular in topography, and the of the street networks, that is, at a right angle to medina the is located on a high steep hill, about a hundred main contours (cf. fig. 14). For both cities, the slope, meters the above the twentieth-century city to the north-

This content downloaded from 196.12.203.5 on Mon, 14 Sep 2020 14:01:20 UTC All use subject to https://about.jstor.org/terms 68 MICHAEL E. BONINE

Table 7. Qiblas of Religious Architecture in Taza

Identity Code Building Date Period Qibla Comments (Azimuth*)

("correct" qibla) 950 38' Great Circle route GM Great Mosque 1142-1146 Almohad 1540 Reading on sides and on inner courtyard, verified on air photo. Enlarged by Merinids, end of the 13th c. & by Alawites in the late 17th c. SA Sidi Azouz Mosque 12th c. Almohad 1490 Reading on sides, but may not be square AH Madrasa Abu'l Hasan 1323? Merinid 1540 Reading on sides, cannot locate exactly on map or air photo S [Mosque of the Suq] ? ? 1510 Reading on inner courtyard, verified by air photo A Andalusian Mosque ? ? 1480 Reading on inner courtyard, verified on air photo. 12th c. minaret (Parker, 1981, 155) P [Mosque of Place ? ? 1550 Reading on sides, verified on air photo Mulay el-Hasan]

east (although the top of the hill itself the isqibla quite based flat). upon The the shortest distance Great Circle Almohad Great Mosque, the Jama' route Lakbir, (from is970 located in northern Morocco to 910 in Marra- at the lower, northern end of the kesh). medina. Its qibla direction is 1540, a more common direction Correspondence for between Almo- the qibla and the street had buildings. The other mosques' qiblas patterns showare interesting, also in contrasting the patterns. In Fez 1500's range, except for the Andalusian and Marrakesh, Mosque for instance, which there are many different has a qibla of 1480 (table 7, fig. 17). qibla orientations, and the streets are quite irregular. In The street pattern and morphology hilly Fez of (el-Bali) the there city is an are accord of qiblas and most regular as well (fig. 18). The orthogonal immediately surrounding pattern streets in is many instances. As oriented at an axis of 1540, the exact the main direction street, the Rue Talaaof Kabira,the curves through Great Mosque's qibla and axis. The thecity medina, slopes the qiblas anddown adjacent streetsat are at right a gentle 2 percent grade to the northeast, angles to each part i.e., of the toward main thoroughfare. Topo- the Great Mosque. In the case of Taza, graphy appearsthen, to control we thehave basic streeta pattern in rather striking example of extremely Fez, andclose the orientation correspond- of mosques and madrasas is ence between the qiblas, the axis ofinfluenced the by city, both slope and and street the pattern. The qibla direction of the main slope. direction in this instance appears to have little effect in determining street patterns. SOME PRELIMINARY CONCLUSIONS Marrakesh has many different qibla directions and street directions, even though the city is not very hilly. This analysis of Moroccan medinas has In provided the central part dataof the medina, however, there is even to evaluate the relationship between the lessqibla agreement direction between building orientations and the and the axis of the city and its streets and immediate morphology. surrounding streets. The pattern of streets It has shown that qibla directions varied appears greatly, to have been formedal- quite independently of the though some patterns are discernable in qiblaparticular and vice versa. dy- Although it is probable that slight nastic periods. The Almohad period (mid-twelfth-thir- variations in slope produced the irregular city morphol- teenth century) is particularly uniform, ogy, with a more almost detailed study all of the topography needs to be of the qibla orientations in the 1500's range. conducted It is to only confirm in this speculation. Only in the the Alawite period (mid-seventeenth southerncentury part ofto the thecity is there a predominant corre- present) that the religious structures are spondence oriented of the fairly qiblas of the Almohad buildings with closely to the contemporary "correct" the calculations street patterns and of the axis of the city, which also

This content downloaded from 196.12.203.5 on Mon, 14 Sep 2020 14:01:20 UTC All use subject to https://about.jstor.org/terms SACRED DIRECTION AND CITY STRUCTURE 69

;~ a~~FB ~ d~~n~e~?~gg~ 1~tiill~~is- 4t~b s~i.:~~p~EEis d~~.;,'?lt k" ~ -*IAt ,, s ..., , /'itd'" "Correct*w O l,. _ . .~x r~Zi3PU44

...... ?rl~F~i*~llf~t~. ' ?I I,~ 7'OVA~ e" i~'P bi41

t,;m 9" bI'"

jf* a4

sp--

. .7:...... 4.0 r* t*~~ 5?~~r ~Op

i?-4!r ,.- .... ' ...... - on Aji'

16%1 ..... g5Salk . ....tor,- -' .Zi1401P

(L-8

N4..

17. Qiblas 18. Aerial in photographTaza. Fromof Taza. From field Service Geographiquedata as ettabulated Topo- in table 7. Base map from Taza: graphique, Plan Rabat. Urbain, (See table 7 for 1:5,000. building identifications.) (See table 7 for building identifications.) corresponds streetwith pattern, but the qibla the orientation as well.direction of a gentle slope. In Meknes, another Finally, a similar and almost exact correspondencerather of hilly town, the qiblas and street pattern streets and qiblas occursalso in Taza, where do the gentle slopenot agree closely, except in the newer sections south of the older medina. The main axis is also in the direction of the city axis. In this case, of the medina and its major streets are, however, in the however, the Almohad Great Mosque has a familiar same basic direction as the qiblas. In the western part of Almohad direction of 154'. Did the slope and therefore the city where the streets begin to curve, the qiblas the streets determine the qibla, or did the qibla deter- reflect this orientation as well. The topography also mine the streets? We are back to the original question appears to predominate in Meknes. posed at the beginning of the paper. Rabat and Sale indicate, once again, the preemin- When the evidence for all the cities is considered, ence of slope for the street direction, showing different however, it appears that the topography is the major orientations across the Bou Regreg. The axis of both determinant of both the street pattern and the city cities corresponds to their predominant qibla orien- structure. Streets and housing follow the slope, and if tations. The axis of Sale and its qibla direction are quite there is even a slight change in the gradient the mor- striking, and this is the only case where an Almohad phology of the city is altered as well. Yet exactly how building's qibla is not in the vicinity of the 150"'s. (Since changes in slope affect the street and housing pattern the Great Mosque's qibla is 1240, one wonders if it was requires much more detailed analyses of the slope char- indeed Almohad, or reoriented later.) It appears in the acteristics than we have, which would in turn require case of Sale that the slope not only has determined the more extensive field studies. There is also always the

This content downloaded from 196.12.203.5 on Mon, 14 Sep 2020 14:01:20 UTC All use subject to https://about.jstor.org/terms 70 MICHAEL E. BONINE

problem of centuries and hence the of axis of soilthe buildings. accumulation Smaller mosques, from dis- solving, crumbling madrasas, mud-brick and tombs perhaps tended structures to be controlled which might distort the original by theseslope conditions moreconditions. than some of the larger, more There is, however, important a buildings. very But the numbersignificant of instances in observation about city axes and which street the qibla directions orientations. of major buildings, principal With the exception of the extremely streets, and hilly the axis of theFez city and el-Bali, the main slope all the main axes of the medinas are orientation in are aaligned roughly leads to the conclusion thatsouthwest-northeast the orientation, instead site ofof a building being or a city was roughlyselected because it had a in cardinal directions. When the predominant slope which was in the correct direction, qibla that is, a slope directions of the specific cities are examined, that corresponded to the qibla. the close correspondence of the qiblas with the Thestreet qibla direction patternstherefore determined the (and street hence the main axes) exists in most pattern cases. and city axis Then only if the slope when conditions al- the slope characteristics are examined, lowed. Such athey slope must be arerather gentle often and oriented in the same direction. The examples in theof accepted the qibla direction, southern but the fact that qibla part of Marrakesh (as well as the general directions andaxis the city axisof can vary the slightly fromentire city to city), of Rabat, Sal6, and particularly city (or even withinTaza one medina) confirm indicates that the slope this correlation. Perhaps slopes in remainedthe paramountappropriate in determining the exact qibladirec- direction were selected for the location tion. The qibla influenced of the general many orientation ofof these medieval Islamic Moroccan cities. religious architecture, but the exact direction was more often controlled by the slope and topography, which In sum, the qibla directions used for the religious also archi- was the determinant of the city's orientation, street tecture of Morocco range from almost due east (900) pattern, to and morphology. due south (1800), but are concentrated in degree ranges David King's statement" that some Islamic cities in the 120's, 140's, and 150's. Although the calculation were laid out in the direction of the qibla is indeed true of the qibla changed over time, it is apparent for that Morocco. But the slope conditions determined consistency in any one direction (based on whatever whether or not that would take place. Only when an calculation) was never present, or at least was not appropriatecar- main slope oriented toward the qibla could ried out in the actual orientation and construction of be found would the streets and the religious architec- buildings, for even with a predominant direction, turethe also be aligned with the sacred direction. It may qibla and building axes can vary by a number of de-turn out that the various procedures outlined by King grees. as paramount for the qibla and the orientation of reli- The qibla direction and hence the orientation of religious buildings were in fact no more important, and gious architecture correspond to the adjacent street perhaps in a great number of cases even less important, pattern and direction when the local topography and than the slope conditions. slope are also oriented in that direction. When the street These conclusions must of necessity remain tentative pattern and the axis of a religious building are different, and preliminary. More detailed information on exact the qibla determines the building orientation and the slope conditions, more qibla readings (and in some slope determines the street pattern. cases confirmed readings), and more complete informa- The need to channel and direct water, both for distri- tion on the age and history of buildings are all needed bution and for proper drainage of the city, may well before definite conclusions can be made. In addition, have been the principal reason why the streets and any study of Morocco must be considered in the context houses follow the topography. Surrounding irrigated of the entire Maghreb as well as Andalusia and Libya. fields would also have been determined by the slope The subject deserves further research, for there are conditions, and these field patterns would have influen- indeed intriguing patterns which indicate that the ori- ced the morphology of the city as it expanded into the entation ofmosques, madrasas, and shrines did in many surrounding land.'"8 cases influence the orientation of a city - or the selec- The slight variations in qibla directions within a tion of a site for a city - or, significantly for architectu- single city and the correspondence of those directions ral historians, that the slope and street patterns deter- with a slightly changing street pattern indicate that mined the qiblas and the orientation of the buildings. local topography influenced the direction of the qibla How extensive these various patterns were remains to

This content downloaded from 196.12.203.5 on Mon, 14 Sep 2020 14:01:20 UTC All use subject to https://about.jstor.org/terms SACRED DIRECTION AND CITY STRUCTURE 71

be worked out, not only for Morocco, mausoleums, but and a fewfor mosques, other however, areas entry into these of the Middle East and the Islamic world as well. We Maliki religious structures is forbidden for non-Muslims, and lack of time prevented me from attempting to go through the red hope others will begin to conduct some of the impor- tape necessary to gain access to all the buildings. In most cases, I tant, primary field work which will enable us to begin discovered, the qibla could still be determined quite accurately. properly to assess Islamic city planning from this per- Often the mihrab could be seen through a doorway or window, spective. and in some cases it could be detected clearly on the outside back wall behind the building. Otherwise readings on outside walls, inner-courtyard walls, and roof alignments had to be used. If it Department of Geography were not obvious which wall the qibla wall was, the information Department of Near Eastern Studies was obtained from the local people. Double and triple checks by University of Arizona several readings and verification of the building alignments on Tucson, Arizona maps, aerial photographs, and, if available, architectural plans confirmed both the readings themselves and that the seemingly haphazard method did indeed provide reliable and accurate data. The margin of error is therefore quite small, although there NOTES is always the possibility that a few of the readings are inaccurate. Aerial photographs of Fez, Meknes, Marrakesh, Rabat, Sale, 1. Michael E. Bonine, "The Morphogenesis of Iranian and Cities," Taza from the 1960's and 1970's were available from the Annals of the Association ofAmerican Geographers 69 (1979): Moroccan 208-24. Service Geographique et Topographique in Rabat; 2. David A. King, "Astronomical Alignments in Medieval the Islamic photos are on such a small scale that their usefulness is Religious Architecture," Annals of the New York Academy of limited,Sciences however. Large-scale planning maps were not avail- 385 (1982): 303-12; Gerald S. Hawkins and David A. King, able, "Onalthough fairly accurate smaller-scale maps, ranging from the Orientation of the Ka(ba,"Journal for the History of Astronomy 1:5,000 to 1:50,000, were usable. These include: 1ditions Gauth- 13 (1982): 101-9; David A. King, "Architecture and Astronomy: ey, "Plan-Guide de F's," scale 1:10,000 (with inset "F's el-Bali- The Ventilators of Medieval Cairo and Their Secrets,"'Journal Fis Ja-did," of scale 1:5,000) (Casablanca, 1979); idem, "Plan- the American Oriental Society 104 (1984): 97-133; idem, Guide"The As- de Marrakech," scale 1:6,500 (Casablanca, 1979); idem, tronomy of the Mamluks: A Brief Overview," Muqarnas "Plan-Guide 2 de Meknes," scale 1:10,000 (with inset "Medina," (1984): 73-84; idem, "The Sacred Direction in Islam: A Study scale 1:5,000)of (Casablanca, 1979); idem, "Plan-Guide de Rabat- the Interaction of Religion and Science in the Middle Sal," Ages," scale 1:12,500 (Casablanca, 1979); Kingdom of Morocco, Interdisciplinary Science Reviews 10 (1985): 315-28; idem, "Makka:Division de la Cartographie, "Fis: Plan Urbain," scale 1:10,000 4. As the Centre of the World," Encyclopaedia of Islam, (Rabat, 2d ed. 1980); idem, "Meknes: Plan Urbain," scale 1:10,000 (Leiden, 1960-), vol. 6, pp. 180-87. (Rabat, 1982); idem, "Taza: Plan Urbain," scale 1:5,000 (Ra- 3. Idem, "Astronomical Alignments." bat, 1983); Kingdom of Morocco, Division du Cadastre et de la 4. Ibid., p. 304. Cartographie, "Rabat-Sale," scale 1:10,000, 2 sheets (with inset 5. Ibid. of Rabat Medina, scale 1:5,000) (Rabat, 1972); Kingdom of 6. Ibid. Morocco, Ministere de l'Agriculture et de la Reforme Agraire, 7. Idem, "Architecture and Astronomy." "Marrakech: Medina," scale 1:50,000, Carte du Maroc au 8. Janet Abu-Lughod, "The Legitimacy of Comparisons 50,000, in Com- Feville NH-29-XXIII-3a (Rabat, 1951); idem, "Fis- parative Urban Studies: A Theoretical Position and an Applica- est," scale 1:50,000, Feville NI-30-XIV-la (Rabat, 1973); idem, tion to North African Cities," Urban Affairs Quarterly 11 "Marrakech," (1975): scale 1:10,000, 2 sheets (Rabat, 1973); idem, 13-35; idem, Rabat: Urban Apartheid in Morocco (Princeton, "Rabat-Sale," 1980), scale 1:20,000 (Rabat, 1974). chaps. 1 and 2. 12. All of the compass readings (magnetic north) have been adjusted 9. Idem, "Moroccan Cities: Apartheid and the Serendipity of to true north for the analysis in this paper. Magnetic declination Conservation," in African Themes, Northwestern University Studies in in 1985 in Morocco was approximately 60 west of true north. Honor of Gwendolen M. Carter, ed. Ibrahim Abu-Lughod (Evan- Hence 60 has been subtracted from all the compass readings. ston, Ill., 1975), pp. 77-111. True north and grid north are less than 10 difference on any of the 10. Field research in Morocco and Tunisia in summer 1985 was maps used, and this slight difference was ignored. Considering partially funded by grants from the American Institute of Magh- the problems of obtaining exact (perfect) readings using a hand- ribi Studies and the University of Arizona, Social and Behavior- held compass (as well as the problem of access to the building), al Sciences Research Institute. Thanks are extended to both the difficulty of getting the exact magnetic declination for sum- these institutes for their support. A preliminary version of mer this 1985 for each of the Moroccan cities, and the slight differ- paper was given at the international symposium, "Traditional ences in grid north and true north, it is estimated that the margin Dwellings and Settlements in a Comparative Perspective," heldof error probably does not exceed 20-3' for the qibla directions at the University of California, Berkeley, April 1988. used in the analysis, and such "errors" may be canceling one 11. Slope was measured with a Suunto PM-5 clinometer. Unless another the out in any case. All readings are given in azimuths from slopes were gentle the topographic conditions were usually true too north, where east is 900 and south is 1800. complex to measure and record effectively in the pilot 13.survey. David King (personal communication) has noted that the me- Compass readings were made with a Suunto optical reading dieval "exact" or mathematically calculated qiblas also differ compass. The qibla was determined by taking a reading to slightlythe from the modern (correct) azimuths to Mecca because mihrab if it was accessible. Except for the madrasas, some longitude differences from Mecca were based on inaccurate

This content downloaded from 196.12.203.5 on Mon, 14 Sep 2020 14:01:20 UTC All use subject to https://about.jstor.org/terms 72 MICHAEL E. BONINE

distance values. For instance, man (Casablanca, 1949); Gaston in Deverdun, fourteenth-century Marrakech: Des ori- Marrakesh the qibla was calculated gines a' 1912,mathematically 2 vols. (Rabat, 1959, 1966); Janet Abu-Lughod, as 1030 (instead of the now known, correct value Rabat; Kenneth L. Brown,of People 910).of Sale: Tradition andYet, Change in a as the values in table 1 and fig. 1 indicate, the Moroccan majority City, 1830-1930 (Manchester, 1976);of and Mohamedthe Moroccan qiblas would not be close to the medieval Naciri, "Sal6: 1tude de geographie mathematically urbaine," Revue de geographie calculated qiblas either. du Maroc 3-4 (1963): 11-82. Information on specific cities as well 14. Data on the origins ofas general discussionsbuildings of Moroccan history can havebe found in, for relied not only on individu- al histories, but also instance,on Jamil standard M. Abun-Nasr, A History of the Maghrib guidebooks (London, of Morocco, including Der Grosse Polyglott, 1971); Jean Brignon et al., Histoire Marokko du Maroc (Casablanca, 1982); (Munich, 1987); Editions Marcus, Morocco: Pocket Charles-Andre Julien, Travel History of North Africa: Guide From the Arab Con- (Paris, 1984); Hachette, Morocco (Paris, 1966); questand to 1830 (New Michelin,York, 1970); and Abdallah Laroui, The HistoryMaroc (Paris, 1975). Gener- al surveys of Islamic/North of the Maghrib: An Interpretative African Essay (Princeton, 1977). architecture consulted include Lucien Golvin, Essai 15. Christel Kessler, sur "Mecca-oriented l'architecture Architecture within an Ur- religieuse musulmane, Archeologie Mediterraneenne, ban Context: On a Largely Unexplored vol. Building 5, Practice pt. of 4: L'art hispano-musul- man (Paris, 1979); Derek Mediaeval Hill Cairo," in Araband Architecture: Lucien Past and Present, ed. Anto- Golvin, Islamic Architec- ture: North Africa (Hamden, ny Hutt (Durham, 1984), pp. 13-20. Conn., 1976), with an excellent introduction by Robert 16. Deverdun, Hillenbrand; Marrakech, vol. 2, pl. 9. However, an orientationAntony of Hutt, Islamic Archi- tecture: North Africa 1090(London, is very unusual for an early Islamic1977); building, as this George re- Marqais, L'architecture musulmane d'Occident: search shows. I wouldTunisie, question the validity of Algerie,that reconstruc- Maroc, Espagne et Sicile (Paris, 1954); George Michell, tion for an early Moroccan ed., mosque (theArchitecture Almoravid cupola's axis ofthe Islamic World: Its History and Social Meaning is approximately 1380, for instance).(New York, 1978); Richard Parker, A Practical Guide to Islamic 17. See, for instance, KingdomMonuments of Morocco, Ministere de l'Agricul- in Morocco (Charlottesville, Va., 1981); and Charles ture, "Marrakech," Terrasse, 1:10,000. Mideras du Maroc (Paris, 1927). 18. Cf. Bonine, "Morphogenesis of Iranian Cities," for the case of Urban histories include Attilio Gaudio, FRs: Joyau de la civ- Iran. ilisation islamique (Paris, 1982); Roger Tourneau, FRs avant le 19. King, "Astronomical Alignments," p. 304. Protectorat: Etude economique et sociale d'une ville de l'Occident musul-

This content downloaded from 196.12.203.5 on Mon, 14 Sep 2020 14:01:20 UTC All use subject to https://about.jstor.org/terms