Figure 1. Mosque of al-Ḥākim, main entrance, frieze on northeast side.
Table of Contents
When the Fāṭimids arrived in Egypt they had never built in fine stonework and they found no one doing so. Yet in their newly founded city of Cairo they set about building a caliphal palace in stone (if only on the exterior), and they began using carefully carved decoration in fine stone masonry. The first preserved example of Fāṭimid architecture featuring fine masonry is the Mosque of al-Ḥākim, and its decoration is a window into early Fāṭimid design and its sources.
In this study I describe the carved stone ornament of the Mosque of al-Ḥākim, emphasizing how its designs were constructed rather than their motifs, and I follow out certain related aspects of these compositions. A fuller consideration of the carved stonework would take in the profiles and use of its moldings, but as I believe such a study would require new work on scores of monuments I leave it for someone else. I am not concerned here with the overall plan of the mosque, its other decoration, its epigraphy, or its iconography. My cut-off date for comparative material is ca. 1400 A.D.
It is not necessary to read all of my detailed descriptions of the ornamental patterns to follow all of my argument, and these descriptions, however necessary, are likely to be as boring to most readers to read as they were to develop. I have marked sections of detailed description as such and provided overall summaries for the various types of ornament, but it has not been possible to avoid all conclusions in the detailed descriptions; they are not simply supporting material.
I thank Dr. Sheila Canby, Dr. Jens Kröger, Prof. Bernard O'Kane, and Prof. Lorenz Korn for help with and, or, comments on this work. All errors not the fault of my written sources are mine.
Images of negatives from the Creswell Archive, Dept. of Eastern Art, Ashmolean Museum, are referred to in this work by means of links. The copyright notice for this material is at http://creswell.ashmolean.org/Copyright.html. (I have found that some of the images in the Creswell Archive have been cropped when compared to the corresponding plates in Creswell's works.) I have also linked to an image from the Thesaurus Islamicus Foundation's Islamic Art Network, for which the copyright notice is at http://www.islamic-art.org/Legal/legal.asp. Images from Samuel Flury's Die Ornamente der Hakim- und Ashar-Moschee, Materialen zur Geschichte der älteren Kunst des Islam, published just a hundred years ago, are from a digital copy obtained from archive.org.
|
Figure 1. Mosque of al-Ḥākim, main entrance, frieze on northeast side. |
The Mosque of al-Ḥākim has been well known to scholarship for over a century. Its form, planning, and urban context have been examined in detail. Its decoration has been remarked as notable and documented to a fairly satisfactory degree. The building was sadly dilapidated before successive partial restorations and it has been very heavily rebuilt,1 but the published record, which I supplement with my photographs of the mosque, taken in 1983, and negatives from the Creswell Archive, will support a closer study of its decoration. (I describe the building as it existed before reconstruction.)
The definitive, though by no means complete, description of the Mosque of al-Ḥākim is that by K.A.C. Creswell.2 The essential details are that the mosque was built outside what was at the time the northern wall of Cairo, to the east of the north-south axis of the city. It is approximately 120 m. wide by 113 m. deep, built mostly of rough masonry faced at the exterior of the entrances with finely finished stonework. The two minarets at the corners of the northwest facade appear to have been built entirely of fine masonry; they are encased in later masonry and have lost their original tops in an earthquake. In Creswell's words,
The northern minaret consists of a cylindrical tower diminishing slightly in diameter from 7.43 to 6.50 m., and resting on a richly profiled socle about 3.65 m. high and 7.68 m. square. … The Western minaret … consists of a square shaft, 13.90 m. in height, surmounted by five receding octagonal storeys. … The total height preserved is, therefore, 24.26 m., against 23.70 for the other minaret.3
The main, central entrance in the northwest facade is a projecting block. Inside the building, the arcades of the structure are set on mostly brick piers braced by wooden tie-beams. The prayer hall is articulated by a wider and higher central aisle, a dome before the mihrab, and domes in the two end bays next to the qiblah wall. The interior decoration was apparently mostly in stucco, with some wooden and stone elements (the lost original columns of the mihrab), and probably paint.
According to the Arabic sources, mostly preserved by Aḥmad b. ʿAlī al-Maqrīzī,4 the mosque was founded by al-ʿAzīz, the father of the notably unstable Fāṭimid caliph al-Ḥākim, in Ramaḍān 380/990 and was finally furnished by al-Ḥākim in 403/1012–13, including its matting. I assume that at this date it was complete and ready for regular use. In between these dates certain other events are noted, and it is worth reëxamining them. In Ramaḍān 381/991, a year after work commenced, al-ʿAzīz said the khuṭbah there, and (the implication is soon after) courses were given there. Al-ʿAzīz died in 386/996 and al-Ḥākim acceded to the caliphate. In 393/1003 al-Ḥākim ordered the completion of the mosque (the execution of some of the carved stone ornament is dateable ca. 393/1003 by inscriptions on the minarets and the main entrance). In 401/1010 large cubical masonry buttresses (one of them later enlarged when a new city wall was built) were constructed, enclosing the lower parts of the minarets.5 In 403/1012–13 al-Ḥākim ordered furnishings, as already noted. And in 403/1013 the mats were laid down, the other furnishings were installed, and al-Ḥākim “made the first prayer there after its completion”.
Creswell interpreted this information to mean that the “main work” on the mosque was completed in a year, and it seems that he has been followed by all later authors. I find it remarkable not that so large a building could be built in a year but that subsequently it required ten years of additional “completion”, during which time it lacked even mats. More likely, al-ʿAzīz delivered the khuṭbah in an unfinished and possibly unfurnished building; how unfinished and unfurnished we do not know. The detail about courses being given in the mosque suggests it was useable, but perhaps actually describes a temporary arrangement or later use.
The question of the state of the mosque in 381/991 is connected to the problem of understanding whether the minarets were part of the original design—and vice versa. Creswell attributed them to al-Ḥākim because they bear inscriptions in his name.6 He did not express an opinion on whether they were added to the original design. Doris Behrens-Abouseif entertained the notion that they were built by al-ʿAzīz and that al-Ḥākim disliked them and hid them, although she was aware of the inscriptions in al-Ḥākim's name.7 Jonathan Bloom attributed the main entrance to al-ʿAzīz's original plan despite its inscriptions naming al-Ḥākim (of which one could have been an addition to the original structure) but concluded that al-Ḥākim added the minarets.8 Bloom argued that the minarets are shown to have been added to an earlier design because the main entrance contains stairways, which were envisioned at the outset as the access from within the mosque to its roof. That the ground-level entrances to these stairways were later blocked (the walls in which they stood were rebuilt9), Bloom contended, indicates that they had become “redundant, for the tower staircases [i.e. the minarets] would have provided ample access to the roof”. In other words, the main entrance stairways would not have been required had the minarets already been planned. I do not see how the minarets would have provided any convenient access to the roof (see below for the question of access to the western minaret from the roof), and so far as Creswell's observations go the entrances to the stairways may not have been blocked until the nineteenth century, possibly after the roof on the northwest side disappeared (cf. Creswell's fig. 28, from the Description de l'Égypte, in which the northeast half of the aisles on the northwest side of the mosque still stands but the southwest half seems not to).
The minarets can certainly be seen as additions to the plan of a building that was previously planned to have simple corners instead. But, as I understand Creswell,10 the exterior walls were bonded with the masonry of the minarets, so they were built together. As the inscription in the eighth register of the western minaret, which is well below the top of the exterior wall of the mosque and appears to be part of the original structure, names al-Ḥākim (see below), it cannot be the case that al-ʿAzīz started the minarets and al-Ḥākim finished them unless al-ʿAzīz did not complete the northwestern wall to its full later height. Otherwise, if the minarets were additions to al-ʿAzīz's building, that building had no northwestern side at all. (A simple solution that occurs to me is that al-Maqrīzī and, or, his sources are confused, and that al-ʿAzīz did not found a mosque on the site of the Mosque of al-Ḥākim at all.)
In any event, there does not seem to be any evidence, physical or otherwise, that the minarets were built up against an existing building.
Two possibilities occur to me: that al-ʿAzīz delivered the khuṭbah in 381/991 in a largely unbuilt mosque that was not in fact used until 403/1013, and, or, that al-Ḥākim demolished and rebuilt at least the northwest side of the building al-ʿAzīz had begun. In any event, while there is no way of knowing whether the minarets were part of the original plan, as they exist today they were built by al-Ḥākim.
The carved stone ornament was first published in detail by Samuel Flury, whom Creswell followed closely.11 Creswell himself described its individual elements only briefly.12 It occurs on and within the main entrance of the Mosque of al-Ḥākim and on and within the minarets.
In the main entrance the ornament consists of squares rotated 45° within the shallow niches, and above them a cornice, a frieze (as Flury called it), another cornice, niche heads, their borders, and a border that may have run around the zone above the springing of the niche heads and may have been coupled with an epigraphic border. The minor entries of the northwestern facade had door frames distinguished by moldings but not carving. The ornament of the minarets includes borders, moldings, round medallions, horizontal bands (both applied and as what Creswell called a cresting), window frames, the fillings of spandrels adjacent to niche heads, and an inset panel set above a doorway.
I consider the frieze of the main entrance and comparative material first, then the rest of the ornament, sorted by type.
In this study I use certain existing terminology in restricted ways and introduce a handful of useful neologisms. Below is a list of new terms, with links to the text in the order of appearance of the terms. First, a few geometrical terms.
I use “intertwining” to denote a small group of topologically simple and regular combinations of generally linear elements, or strands, such as twists, including guilloches, and braids. In these patterns the individual strands do not branch or combine. If seen as having a direction of movement (for example, right-to-left or left-to-right; it does not matter), each of the strands of these patterns moves in one direction only, without looping back in the other direction.13 Braids can be more complex, branching and combining to surround several rectangular panels, but such patterns do not occur in the ornament of the Mosque of al-Ḥākim.
I use “interlacing” for more complex regular combinations of generally linear elements, which may branch and combine, and may loop back on themselves.
As regards star-shaped elements, I adhere to what I understand to be convention and describe as n-grams (for example, pentagrams) only those figures with internal construction lines. So far as I am aware all these are regular interlacings. Similar figures without internal construction lines properly might be called “star-shaped polygons with x points”, but that is cumbersome, so I call regularly formed (that is, having equal angles and sides) star-shaped polygons with (for example) five points simply “five-pointed stars”.
Finally, I have adapted Creswell's terminology for the levels of the two minarets; see below.
Note that Creswell's elevation of the northern minaret (fig. 36) shows its northeast side (for the vertical lines on the left side, outside the shaft of the minaret proper, see Muslim Architecture of Egypt, v. 1, 90–91, discussion of a “flat face”); his elevation of the western minaret (fig. 40) shows its southwest side. For further discussion of the second elevation see “The Niched Room in the Western Minaret”.
Both Samuel Flury and K.A.C. Creswell observed color on the window frames of the northern minaret of the Mosque of al-Ḥākim.
I believe that Flury's pl. 25, 1, shows the window facing east (in Creswell's terminology) in the third register (my terminology). Flury described it as facing north and observed that the interlacing bands of the outer frame were colored blue on a red ground.14 Elsewhere he also noted red and blue paint, alternating in some fashion, in connection with an inscription frieze and an ornamental band.15
In the foliate border of the lintel of the entrance to the staircase of the northern minaret Creswell noted “blue paint … in the hollows of the ornament”.16 Flury mentioned traces of red and blue on the same ornament, but did not describe their disposition.17 Creswell also described a window above the entrance, which is “set in a bevelled frame with its lower edge 6.32 m. above the pavement. It is decorated with a beautiful undulating stalk, with a palmette in the centre, like that on the door lintel. The sides of each deeply cut motif are coloured blue and the background red”.18
In the fifth register there are three small windows with grilles, bevelled frames, and “arabesque” borders, according to Creswell, who observed blue in the background of one border. He did not publish photographs of these window frames and remarks that Flury ignores them.19
All this is very valuable evidence. This colored paint must have been present already in 401/1010, when the painted areas were concealed behind the new buttresses, because no one would have bothered to apply it later. Unlike some other exterior paint on stone buildings, this must be more or less original, and probably part of the original decorative intent. I do not fully understand how these zones were colored; I imagine that much more evidence of the sort could be obtained by careful archaeological work.
It is impossible to know the extent to which the Mosque of al-Ḥākim and the comparative material I cite in this study involved effects created by color that extended those seen now or added to them. Felix Hernandez Gimenez indicated that some low-relief work at Madīnah al-Zahrā' had a red background but does not give examples; some elements of arches were also colored, and this is how the Salon Rico has been restored.20
One coloristic effect that was achieved, to judge from Creswell's description of “blue paint … in the hollows of the ornament”, was that of leaving the relief in the light-colored base material (here, stone) unpainted on a colored ground. This effect was used both earlier and later in Islamic architecture, at least in interior decoration.
Ernst Herzfeld, excavator of Samarra, wrote of the stucco decoration of Samarra houses in his First Style (Creswell's Style C, the “bevelled style”):
The finished dado was subsequently provided with a coating of chalk paste and originally appeared in a blinding white, which has taken an ivory or golden tone through age and weathering [interpolated remarks by Creswell omitted]. Painting was only employed in the very rarest cases, and was confined to the emphasizing of a few hollows in red and light blue.21
Of course the whiteness of the stucco would not have been blinding if it were not seen in full sun, as Herzfeld saw it.
Herzfeld's estimate of how common color was may have been inaccurate. Thanks to the kindness of Dr. Sheila Canby I was able to examine fragments of Samarra stucco at the British Museum in February 2009. I found no sign of any coating of any sort, and observed red and blue pigment, mostly on the background, in several pieces.22
The effect of painting the background of relief stucco while leaving the relief its natural white would have been generically similar to that of the restored stucco decoration in the Madrasah of Granada (which is of course much later).
 Figure 2. Granada, Madrasah. |
However, Flury's and Creswell's descriptions of blue on a red ground suggest a less refined effect.
The used of paint applied to flat exterior surfaces is more difficult to envision, but it was used. During the very caliphate of al-Ḥākim inscriptions were painted on the exteriors of both religious and secular buildings. Irene Bierman cited a passage from al-Maqrīzī23:
In the year 395 slander and monstrosity occurred concerning Abu Bakr and `Umar. … It was written in Safar of this year [395] on all of the mosques and on the old mosque of Fustat, outside and inside, and on all of its sides (walls) and on the gates of shops, and rooms and on tombs, insulting the ancestors and cursing them. It was variegated and colored with colors and gold. And that was done on the doors of houses and bazaars. And people were forced to [do] that.
Those were inscriptions, not ornament, and al-Ḥākim was an unusual person, so one should not conclude from this evidence alone that painted exteriors were common. On the other hand, they were mentioned by al-Maqrīzī only because of their content. It is possible that other painted exterior inscriptions—and even more likely ornament, which is seldom mentioned anyway—went unremarked. In any event, the Mosque of al-Ḥākim, and perhaps even the Fāṭimid palace in Cairo, may have featured painted exterior decoration.
1. Jonathan M. Bloom, Arts of the City Victorious, New Haven, 2007, pp. 72–81.
2. K.A.C. Creswell, The Muslim Architecture of Egypt, 2 v., Oxford, 1952, v. 1, pp. 65–106, with plans and elevations. Another plan, in two different resolutions, can be found at the site of the Museum With No Frontiers, Brussels: http://www.museumwnf.org/images/lo_res/monuments/isl/eg/1/4/plans/1.jpg and http://www.museumwnf.org/images/hi_res/monuments/isl/eg/1/4/plans/1.jpg; it appears to have been derived from the plan in Ministry of Waqfs, The Mosques of Egypt from 21 H. (641) to 1365 (1946), 2 v., Giza, 1949, p. 24; the same publication contains multiple sections of the two minarets and cross-sectional elevations of the minarets with their later surrounding masonry, p. 23.
3. Creswell, op. cit., v. 1, p. 66 for the earthquake, pp. 90 and 94 for the measurements.
4. Ibid., pp. 65–66.
5. Ibid., pp. 65–66, 71, 93, and 96. Although the buttresses could have been added for earthquake protection, there was no earthquake of note during the few years before they were built. Irene Bierman, Writing Signs, The Fatimid Public Text, Berkeley, 1998, p. 93, suggested that the minarets may have “proved structurally unsound”, which is plausible. Bloom, “The Mosque of al-Ḥākim in Cairo”, Muqarnas, v. 1, 1983, pp. 15–36, pp. 20 and 28, asserted that the minarets were originally designed to make a visual reference to Makkah and that the buttresses were built to cover up this reference; the minarets were not torn down, he says, because they were useful as “beacons or watchtowers”. In Minaret, Symbol of Islam (Oxford Studies in Islamic Art, v. 7), Oxford, 1989, pp. 134–36, Bloom abandoned the idea of the minarets as watchtowers and claimed that “had [al-Ḥākim] been able to destroy the towers, he probably would have done so, but even he was forbidden by law from tearing down parts of a mosque already given in waqf and recorded in a foundation inscription”, without citing any legal authority or parallel case whatsoever. Waqf buildings were of course routinely modified; in any event, al-Ḥākim could have done as he liked, as he frequently did.
6. Creswell, op. cit., v. 1, p. 86.
7. Islamic Architecture in Cairo, Leiden, 1989, pp. 63–65.
8. Minaret, p. 134.
9. Creswell, op. cit., v. 1, p. 72.
10. Ibid., pp. 88–89.
11. Samuel Flury, Die Ornamente der Hakim- und Ashar-Moschee, Materialen zur Geschichte der älteren Kunst des Islam, Heidelberg, 1912.
12. See Creswell, op. cit., v. 1, pp. 65–106.
13. The borders of the ambulatory vault mosaics in the Mausoleum of Sta. Costanza, Rome, for example, are composed of four intertwined strands, helpfully distinguished by color (illustrations may be found online at http://sacred-destinations.com/italy/rome-santa-costanza-photos/slides/eosa_089p and http://sacred-destinations.com/italy/rome-santa-costanza-photos/slides/eosa_092p).
14. Flury, op. cit., p. 44.
15. Ibid., pp. 45–46 and p. 48.
16. Creswell, op. cit., v. 1, p. 91.
17. Flury, op. cit., p. 44, n. 119.
18. Creswell, op. cit., v. 1, p. 91.
19. Ibid., v. 1, p. 93.
20. Hernandez Gimenez, Madīnat al-Zahrā': arquitectura y decoracion, Granada, 1985, pp. 125–26; for the Salon Rico, see Marianne Barrucand et al., Moorish Architecture in Andalusia, Cologne, 1992, double-page plate pp. 52–53.
21. Der Wandschmuck der Bauten von Samarra und seine Ornamentik, Berlin, 1923, p. 10, trans. Creswell, Early Muslim Architecture, 2nd ed., 2 v., Oxford, 1969, v. 2, pp. 287–88). In the original:
Die fertigen Wandsockel wurden schließlich mit einem Überzug von Schlemmkreide versehen und zeigten sich ursprünglich in blendendem Weiß, das durch Alter und Verwitterung einen elfenbeinernen oder goldigen Ton angenommen hat.
Bemalung war nur in den allerseltensten Fällen angewandt, und beschränkt sich auf die Betonung einiger Tiefen durch Rot und Hellblau.
22. OA+10974, OA+11018, OA+11024, and OA+11016.
23. Al-mawāʿiz wa'l-iʿtibār bi-dhikr al-khitāt wa'l-āthār, 2 v., Bulaq (Cairo), 1294/1877, v. 2, p. 341; trans. Bierman, op. cit., p. 175, n. 55 to p. 76.