Chapter 4 Architecture as Titulature: al-Madīnat al-Zāhira

In: Articulating the Ḥijāba: Cultural Patronage and Political Legitimacy in al-Andalus
Author:
Mariam Rosser-Owen
Search for other papers by Mariam Rosser-Owen in
Current site
Google Scholar
PubMed
Close
Open Access

When Ibn Abī ʿĀmir reached the apogee of his power, when his light shone in all its splendour and his autocratic power was fully manifest … he conceived the grand design, prerogative of kings, of constructing a palace where he would reside.

Ibn Khāqān, apud al-Ḥimyarī 1938, 80–811

One of the most visible ways in which al-Manṣūr articulated his role as ḥājib was through monumental construction – more specifically, his decision to assume the ruler’s social responsibility to undertake public works, the most obvious of these being his extension to the Great Mosque of Cordoba, the subject of Chapter 5. In contrast to those works commissioned (albeit ostentatiously) out of piety or for the public good, al-Manṣūr’s construction of a palace-city appears to be motivated less by reasons of legitimation than of demonstrating the power of his ḥijāba. More so than his mosque extension, al-Manṣūr’s foundation of a palace-city has prompted emotive responses in the primary and secondary literature, and has become inextricably implicated in the discourse of his usurpation of the Andalusi Umayyad caliphate. This viewpoint has arisen from the perceived appropriation by al-Manṣūr of an architectural form which was the ‘prerogative of kings’, and from the historiographers’ subsequent need to define and locate al-Madīnat al-Zāhira within the wider historical framework of Islamic palace-building. It raises questions about his ‘right to build’, especially his ‘right’ to adopt architectural forms that were explicitly associated with the first two Umayyad caliphs after the many decades it had taken to construct Madīnat al-Zahrāʾ. Since all traces of this palace-city have disappeared from the archaeological record, and no serious attempts have been made to find and excavate it, as we will see in the first part of this chapter, to understand what it looked like and how it functioned, we must turn instead to the evidence provided by textual and poetic descriptions of the foundation, construction, function and destruction of al-Manṣūr’s palace-city.2 The texts which preserve this information tell the same story, and it is likely that the common source for all of them was Ibn Ḥayyān’s lost history of the ʿĀmirids. A scattered number of surviving fragments of architectural decoration, which have been attributed to the ʿĀmirid period for various reasons, give a material aspect to this reconstruction. The fragmentary and sometimes circumstantial nature of the evidence available to us means that this discussion of al-Manṣūr’s palace-city and what it looked like can only remain speculative.

1 Looking for al-Zāhira

“The ruin of al-Zāhira was so complete that there remained no echo of its name in the local tradition, nor memory of its location, much debated in modern times …3

It is probable that a fortuitous find will one day allow its discovery and its unearthed remains will reveal the last and little-known phase of caliphal art.”4

Thus Torres Balbás expressed his lack of conviction about the much-hailed occasional ‘identifications’ of the site of al-Manṣūr’s palace-city, which began with Velázquez Bosco’s investigations at the turn of the twentieth century.5 Attempts to locate al-Madīnat al-Zāhira in the archaeological record continued throughout the twentieth century, but none of them – not even the apparently definitive identification detailed by Arjona Castro et al. in 1994 – has so far presented an argument convincing enough to lead to excavation.6

The debate over the location of al-Zāhira began with the discovery and, later, excavation of Madīnat al-Zahrāʾ. During a delay in his excavations of al-Zahrāʾ, which began in 1910, Velázquez Bosco explored a nearby site which is now identified as the Munyat al-Rummāniyya, but which he initially believed to be al-Madīnat al-Zāhira, or ‘Alamiriya’ as he called it, from the munya al-ʿĀmiriyya mentioned in the historical sources.7 Velázquez Bosco gave no reason for his initial identification of this site as al-Zāhira, but thus began a spate of attempts to locate al-Manṣūr’s palace-city archaeologically. Over the following decades, every newly discovered caliphal site was hailed as al-Zāhira. Ramírez de Arellano in 1918 believed it to be located in the grounds of the sanctuary of La Fuensanta, to the north-east of Cordoba.8 Its identification to the west, by Rafael Castejón, was reported in 1926 in the Anales of the Comisión Provincial de Monumentos Históricos y Artísticos de Cordoba,9 and was accepted as late as 1951 by Gómez-Moreno.10 Ocaña, in 1952, identified al-Zāhira with the Cortijo del Arenal in the eastern Ramla suburb of Cordoba.11 And in 1963, a celebration of the ninth centenary of Ibn Ḥazm’s death gave rise to an investigation into the topography of the ‘Cordoba of Ibn Ḥazm’, which located al-Zāhira in the area bounded by the two streams of Pedroche and Rabanales, to the north-east of Cordoba.12

Evaluating these varying ‘identifications’ of al-Zāhira is made all the more difficult by the use of local toponyms, which do not appear on modern maps, the lack of grid-references or low-scale regional maps, and the absence from these publications of plans or pictures. Furthermore, it is curious that the search for al-Madīnat al-Zāhira followed an opposite pattern to that of al-Zahrāʾ, in that it occurred in spite of the textual evidence. As we will see below, all the texts which mention al-Zāhira unequivocally state that it was located to the east of Cordoba; however, due to the large number of archaeological finds which issued from the west of the city compared with the relative poverty of finds from the east, the texts were discounted. This trend began with Simonet, in his novel-esque history of al-Manṣūr, who located the city in the Eras de la Salud, to the west of Cordoba.13 This neglect of the textual evidence and insistence on a western location characterises the first phase of the search for al-Zāhira, so enthusiastic that it overlooked the development of a scientific methodology. This state of affairs continued until 1952 when Ocaña redirected the debate towards the east of the city, soberly demonstrating through copious references that there was no doubting the Arab sources.14 He was followed by Torres Balbás, whose accounts of al-Madīnat al-Zāhira are entirely text-based.15

The truth of the eastern location is now universally accepted, and has been most recently applied in Arjona Castro’s 1994 identification of al-Zāhira in ‘the Cortijo de Las Quemadas, in an alluvial plain which forms a valley whose lower part is bathed by the waters of the Guadalquivir which surround it forming a meander’ (Figure 17).16 This Cortijo is located at the heart of an industrial estate, in the eastern zone of the second meander of the Guadalquivir to the east of Cordoba, on a quarternary terrace which forms a plateau overlooking a flat, lower area which borders the river. According to Arjona Castro, the toponym Las Quemadas is suggestive of an area harbouring ruins which were ‘burnt’, as al-Zāhira was at the start of the Fitna. In the westernmost part of this area, ‘important archaeological remains’ have been found, including the standing remains of 10 metres of wall built from blocks of sandstone measuring 1.20–1.30m × 0.45–0.60m, and conforming to tenth-century construction methods: ‘undoubtedly in [this] place there was a great palace with an important walled enclosure’.17 Arjona Castro further reports that coin hoards and construction remains have been found throughout this zone; most significantly, when the foundations of the Centro de Adaptación de Incapacitados (CAIPO) were laid here in 1974, abundant blocks of masonry and marble carved with arabesque decoration were uncovered, as well as an Arab cistern and gateway which was, according to witnesses, ‘of great beauty [and] very similar to those of the entrance gates to the Mezquita Catedral in Cordoba’.18 Apparently these finds were not reported to the proper authorities, and the foundations were filled in with concrete.

Figure 17
Figure 17

Topographical reconstruction showing the proposed location of al-Madīnat al-Zāhira in relation to the medina of Cordoba, the River Guadalquivir, the main arterial roads, and Faḥṣ al-Surādiq

© Matilde Grimaldi

Apart from Velázquez Bosco’s investigations at what turned out to be al-Rummāniyya, none of these attempts to identify the location of al-Zāhira has resulted in excavation. Even the most recent identification of al-Zāhira, which is most convincing in regard to the textual descriptions of the site, has not prompted local archaeologists to begin further survey work or trial excavations. While Arjona Castro’s field survey seems to have had interesting results, his reliance on aerial photography has not supported his argument, since none of the constructions which he identifies in the photograph can clearly be discerned,19 and in the absence of other photographs – of the terrain, for example, or the coin finds and decorated construction fragments – we are still not yet ready to firmly identify this site as al-Zāhira’s location.

Since the excavation in 1944 of the Salón of ʿAbd al-Raḥmān III and the ongoing project to excavate and understand the palatine terraces of Madīnat al-Zahrāʾ, the caliphal city has overshadowed that of al-Manṣūr. At the same time, twentieth-century political developments have, as discussed in this book’s Introduction, perpetuated an unfounded belief that the ʿĀmirid period was marked by artistic decline. The disinterest in locating al-Madīnat al-Zāhira seems due, therefore, to a belief that it will have nothing new to offer after the riches of al-Zahrāʾ. We are, then, almost back where we started, and cannot do more than trust in Torres Balbás’ belief that ‘a fortuitous find will one day allow its discovery’.20

2 Reconstructing the Palace

“In 368/978–9 al-Manṣūr ibn Abī ʿĀmir ordered that his palace (qaṣr), known by the name al-Zāhira, be constructed … In 370/980–81 he transferred his residence there and installed himself with his accompaniment of aristocracy and people, taking possession and placing his stores of weapons, treasuries and personal belongings.”21

According to the sources, al-Madīnat al-Zāhira was constructed extremely quickly, which has led some scholars to remark that its fabric must have been ‘fragile, hastily made, its walls of earth and rubble, since it is explicable in no other way that it should pass so quickly without leaving any trace’.22 This picturesque view cannot hold, given that we know the congregational mosque of Madīnat al-Zahrāʾ was built in only forty-eight days, and far from disappearing due to hasty construction and fragile materials, it has been carefully excavated and found to consist of stone, in imitation of the Cordoba mosque.23 The short two-year construction period mentioned in all the sources may refer to the residence (qaṣr) into which al-Manṣūr first moved, and it can be presumed that the construction of other zones, which ultimately formed the madina, continued around it for some time.

The site was located on a rise overlooking the Guadalquivir, to the east of Cordoba (Figure 17).24 The sources give various reasons for the choice of location: according to some, it was already ‘remarkable for its splendid palaces’;25 to others – clearly written with historical hindsight – a prophecy was associated with that site, which told that one day a palace would be built there that was destined to supplant the Umayyad palace.26 The legend, related by an ‘old crone seer’,27 mentions wells, and might preserve the topographical memory of a water source in this location, a much more practical reason for al-Manṣūr’s choice of site. However, the sources also indicate that the site was chosen strategically – it was high, dominating the area around it, and protected on various sides by the bending of the river Guadalquivir.28 It may also have been deliberately sited close to the mustering point of the Umayyad armies. As we will discuss below, its location contributed to the defensibility not only of al-Madīnat al-Zāhira but also of Cordoba, and the caliph within.

The city’s location along the river, at a point where it bends, is indicated in the following anecdote reported by Ibn Ḥazm, presumably on the eye-witness testimony of his father, which also speaks of the intimate gatherings in which al-Manṣūr indulged with his close courtiers, as discussed in the previous chapter:

“On a dazzling day, we were out in a boat with al-Manṣūr and a group of viziers, on the river which flows past al-Zāhira, [profiting] from a charming view in front and behind. As the river started to bend, we plunged into intimacy, the mouth refreshed by chewing on fresh aromatic herbs from here and there, the pleasure of the world and its games gathered in [that moment]. [Al-Manṣūr] found the location extraordinary and contemplated the brightness and ornament [of his town]. He approved of the view and raised [his eyes] to the shining palaces, their pleasant constructions which beautifully attract the eyes and renew the hopes of life. Then al-Manṣūr said, ‘O al-Zāhira, resplendent in beauty! How beautiful is your allure! You exhale your opulence! To contemplate you, as well as to live in you, is an experience surpassing all. Your earth is of good quality and your water of a sweet taste’.”29

At this point the ḥājib began to bemoan the destruction that would inevitably befall his beautiful constructions:

“‘I know that rebellion will destroy you … Misfortune to him who is not impressed by your beauty! How is it possible that you will disappear?’ We were astonished by such a speech and disapproved of his words. We thought that the wine had got to him and placed such thoughts in his head. All that he said exceeded common sense … But he explained what would happen and the reasons why: ‘By God, it is as if you know nothing! Yes! Our enemy will prevail quickly, demolishing everything and reducing it to nothing! It will be as if it had never been made!’ We attempted to calm and relax him, but were later astonished that his account was an unmistakable prophecy …”

Though obviously inflected by the historical hindsight of the Fitna years, which saw the looting and destruction of al-Zāhira, this anecdote also implies al-Manṣūr’s sagacity, his awareness that palaces constructed so identifiably by a figure holding a position that would inevitably be coveted by others were unlikely to survive his passing, especially given the precarious situation of the caliphate under Hishām. It also indicates the sheer delight in which al-Manṣūr apparently indulged at al-Zāhira, and the physical setting in which he chose to relax with his closest courtiers.

Al-Zāhira’s height is clearly implied in another of Ibn Ḥazm’s anecdotes, of a party held at his family’s house in the Rabaḍ al-Zāhira: ‘The women stayed indoors for the first part of the day, then moved to a pavilion (qaṣaba) belonging to our house, on an elevated site overlooking the gardens and providing a view from there onto the whole of Cordoba’.30 It seems to have been terraced in a similar way to Madīnat al-Zahrāʾ: during his account of the sack of al-Zāhira in 1009, Ibn ʿIdhārī talks of al-Mahdī’s mob pillaging the ‘lower zones’, while presumably al-Mahdī kept the upper, palatial zones for himself.31 Other descriptions of the site contain historiographical connections to the foundation of al-Zahrāʾ: ‘the construction was spacious and [al-Manṣūr] took much care to extend it over flat ground for a great distance … He dedicated himself to levelling the slopes and excavating the foundations’.32 He gathered artists and labourers to work on the construction, commissioned the use of gold, lapis lazuli and other precious materials, and ‘decorated [his palace] with such magnificence that the sight wearied the eyes’.33

Al-Manṣūr granted the neighbouring zones to the upper classes and officials of the state, such as the Banū Ḥazm and Banū Shuhayd,34 who built their own mansions there, while other members of the Banū Abī ʿĀmir constructed their own palaces in the environs of al-Zāhira. Al-Muẓaffar lived in the Qaṣr al-Ḥājibiyya, ‘which was near to (jānib) al-Zāhira, outside its walls’; this was where his mother, Asmāʾ/al-Ḍalfāʾ, and son, Abū ʿĀmir Muḥammad, were found at the time of al-Mahdī’s siege in 1009.35 Al-Manṣūr also built himself smaller munyas ‘on the road to al-Zāhira’, and names of several are preserved. One is the Dhāt al-Wādiyayn, ‘Having Two Streams’, which Laura Bariani suggests might be the majlis li’l-mulūk (‘hall of the kings’) located between two waterwheels with a view of a marble pool, described in a poem by Ibn Hudhayl; Lévi-Provençal suggested that the name might suggest the place where the Fuensanta tributary debouches into the Guadalquivir.36 Another was the Munyat al-Urṭāniyya, a name which probably derived from the Romance word ‘Huertanilla’, meaning ‘a small orchard-garden’, as many munyas in fact were.37 The Munyat al-Surūr, ‘Villa of Happiness’,38 had a garden and a large pool, and according to Bariani is the ʿĀmirid munya most frequently cited in the sources; while the Munyat al-Luʾluʾa, ‘Villa of the Pearl’, was located in a high part which allowed al-Manṣūr to look down on the city at his feet – the Qubbat al-Luʾluʾa, one of the towers of al-Zāhira mentioned in the textual descriptions, was probably part of the walled enclosure of this munya.39

Another of the ʿĀmirid munyas, which seems to have had a more important role than the others, was the Munyat al-ʿĀmiriyya, where we know that al-Manṣūr liked to relax on his return from campaign (Chapter 3), and where in spring 978 ʿAbd al-Malik celebrated his marriage to his niece, Ḥabība.40 Al-ʿĀmiriyya’s location is closely related to al-Zāhira by Ibn ʿIdhārī, who mentions it in the midst of his account of al-Manṣūr’s palace-city and calls it ‘one of his palaces’.41 Al-Ḍabbī (d. 1203) presents it as one of the palaces integrated within the al-Zāhira complex by calling it ‘one of his castles (quṣūri-hi) in the “Zāhiriyya”’.42 However, perhaps the most vivid indication of their intimate topographical relationship is given by Ibn Shuhayd in his poem about the lost glories of Cordoba, who talks about both palaces in the same breath (note the clever puns in verse 15 on the names of the two palaces):

(v.6) For the weeping of one who weeps with an eye the tears of which flow endlessly is not enough [to lament the loss of] such as Cordoba …
(v.13) … O for their pleasant circumstances in its palaces and curtained apartments when its full moons were concealed in its palaces!
And the palace of the sons of Umayya abounded in all things, while the caliphate was even more abundant!
(v.15) And al-Zāhiriyya shone brightly with pleasure boats, and al-ʿĀmiriyya was rendered flourishing by the stars (wa al-zāhiriyya bi’l-marākib tazharu, wa al-ʿāmiriyya bi’l- kawākib tuʿmaru).
And the Great Mosque (al-jāmiʿ al-aʿlā) was packed by all those who recited and studied whatsoever they wished, as well as those who looked on.
And the alleys of the markets bore witness that because of those who crossed them, doomsday’s assembly would hold not a few.
O Paradise such that the wind of separation has blasted it and its people so that both have been destroyed!43

Bariani speculates whether the Munyat al- ʿĀmiriyya in fact formed the first nucleus of al-Madīnat al-Zāhira, because of the otherwise early date at which al-Zāhira was founded.44 We will return below to a discussion of the motivations for al-Manṣūr’s foundation of al-Zāhira, but whatever the original relationship of al-ʿĀmiriyya to the wider palace-city, it seems that al-Zāhira’s private residential zone was supplemented by a constellation of smaller palaces, which were munyas in the sense of small, luxurious estate-houses, surrounded by gardens. From the various mentions of these munyas and the activities conducted in them, it seems that they were very much part of the ʿĀmirid ceremonial landscape, and fit with the picture that Glaire Anderson has reconstructed for how these suburban villas functioned during the reigns of ʿAbd al-Raḥmān III and al-Ḥakam II.45

The beauty of al-Zāhira’s buildings inspired poets to sing its praises – Ibn Saʿīd (d. 1286), for example, dedicated a whole book to describing the ‘splendid marvels and the beauty’ of al-Zāhira.46 From these poetic descriptions, corroborated with those of the historical sources, certain details can be inferred about its layout and decoration.

From the descriptions of ceremonies considered in Chapter 2, especially the passage detailing the reception of Sancho Abarca in 992, we know that a road led from the main entrance gate (bāb) of the qaṣr to that of the majlis, the inner sanctum of the palace where al-Manṣūr sat on his throne (sarīr) to receive his visitors. This calls to mind the arrangement at Madīnat al-Zahrāʾ, where a processional road led from the monumental entrance of the Bāb al-Sudda, beyond which only the privileged few passed, to the majālis, of which al-Zahrāʾ had three.47 Similarly, in addition to the majlis where Sancho Abarca was received, the sources mention various reception halls at al-Zāhira, some of which may have been located in munyas. Ibn Ḥayyān mentions the majlis al-kabīr, ‘Great Hall’, and notes that it had views towards the Guadalquivir.48 A majlis al-sāmī, ‘High’ or ‘Exalted Hall’, is also mentioned in an anecdote in which al-Manṣūr staged a particularly impressive display of wealth and power, on the occasion of the visit to his court of ‘an ambassador from the most powerful of the Christian kings of Rūm’.49 Al-Manṣūr ordered that ‘a vast lake, several miles in length’ be planted with as many water-lilies as possible, and that four qinṭārs (about four hundred pounds) of gold and silver pieces be broken into tiny fragments and placed inside the buds of every flower when they closed at night. At dawn, al-Manṣūr summoned the ambassador to present himself in the High Hall, ‘which had views over the lake’; they watched as one thousand Ṣaqāliba appeared, ‘dressed in silken robes embroidered with silver and gold’. When the flowers opened to the sunlight, the Ṣaqāliba plucked the silver and golden ‘pollen’ from the flowers, placing the silver pieces on gold trays and the gold pieces on silver trays, which they finally placed at al-Manṣūr’s feet, ‘thus raising a mountain … before his throne’.

This obvious metaphor of the natural bounty and riches of al-Manṣūr’s kingdom was recognised both by the Christian visitor, who immediately requested a truce, and the historian who cited this passage in order to demonstrate the ‘splendour and magnificence with which [al-Manṣūr] used to surround his person while residing in his palace of al-Zāhira’.50 While it resounds with historiographical hyperbole, this passage also preserves a clear sense of the manner in which al-Manṣūr manipulated al-Zāhira as the ceremonial stage for his ḥijāba and, like ʿAbd al-Raḥmān III, the conscious use of setting to overawe his visitors. This passage specifically mentions the ‘views over the lake’ from the High Hall, and the ensuing incident is carefully contrived to make the most of the view. These features, together with the indication in the sources that the private zones of the palace were in an elevated position, imply that the reception halls at al-Zāhira were miradors, windows which framed a ‘constructed view’ of the ruler’s domains, which Ruggles argues was such an important and innovative feature of the palaces at Madīnat al-Zahrāʾ.51 What the ruler sees is a landscape carefully and artificially constructed – by means of cultivated gardens or an aesthetically manipulated vista – to represent productivity and order, a pleasing metaphor of his ‘good stewardship’ of Islam and of his kingdom. This cultivation of the ‘wilderness’ – the creation of gardens where before was only unruly nature – signifies the ruler as the earthly counterpart of God, and he looks out as viewer/creator, via the mirador, over his view/creation (Figure 101). Is there a reference to this poetic imagery or these actual views in the decoration of al-Manṣūr’s basin (Figure 114), created in the palace-city itself, or the Braga pyxis (Figure 15), whose designs of arches framing luxurious vegetation evoke the idea of looking through a porticoed pavilion onto a flourishing garden?

Bariani believes that the incident of the waterlilies took place in the Munyat al-Surūr, which we know had a large pool,52 while as we saw above, the Munyat Dhāt al-Wādiyayn may have housed the majlis li’l-mulūk. Though this anecdote involved the reception of a Christian ambassador, it gives no indication of the ceremonial prelude as described for the reception of Sancho Abarca (Chapter 2) though, as we have seen, that unusual embassy involved the king himself. While Sancho was received formally and officially in the qaṣr, the Christian ambassador mentioned here was ostensibly subjected to a more intimate gathering in one of the ʿĀmirid munyas. This implies a functional difference between the two types of majlis, in which the qaṣr proper was reserved for more ceremonial activities, though the performance could be no less intimidating in a more ‘private’ space.

The hyperbolic mention of revetments of gold and lapis lazuli, and the ‘magnificence that … wearied the eyes’, conveys the impression that the rooms of al-Manṣūr’s palace were as splendidly decorated as those of the caliphal palace;53 but apart from these hints, very little is known about the buildings’ physical appearance. This contrasts with the wealth of information transmitted in the sources about al-Zahrāʾ, which in turn implies that the caliphal city ‘was recognised as an extraordinary creation’ in its own day.54 Poetic descriptions exist of al-Zāhira’s Munyat al-Surūr, ‘the place of Happiness’, such as the following verses by Ibn Darrāj:55

“The high Dār al-Surūr whose balconies are above the high and brilliant stars,

is raised on columns of marble as if it wished to imitate the arrangement of the stars in Gemini.

They seem two squadrons, on foot and on horseback, which faithfully guard the battle, all equal.”

Another poet, known simply as ʿAbd Allāh, describes another of al-Zāhira’s munyas (ṣarḥ):56

“[The halls] are [raised] on columns which seem in their beauty like the necks of youths and whose arches seem like waxing half-moons.
They are adorned with cupolas in which the echo seems like the cooing song of a dove.
When the singer lowers his voice, the roll of thunder comes to the listeners.
It is as if the red rooves between the patios were poppies fed by the damp dew.”

In another verse, the poet likens the munya to Solomon’s palace. These descriptions of arcades, formed of smooth, marble columns, moon-shaped horseshoe arches and red-tiled domed rooves evoke the appearance of the Hall of ʿAbd al-Raḥmān III at al-Zahrāʾ. The mention of ‘two squadrons, on foot and on horseback’, imitating the stars in Gemini, also suggests the doubling of columns (and therefore capitals), which becomes significant when we discuss below the unique double-capitals that are associated with al-Manṣūr’s architectural projects.57 Ibn Hudhayl also described the arches of one of al-Manṣūr’s palaces as ‘half moons’,58 and elsewhere they are likened to the curve of a bow.59 The marble columns seemed to Ibn Hudhayl to be formed from clear water which one could almost drink,60 and the smoothness of the marble also appeared wet with springwater to Ṣāʿid al-Baghdādī.61 It gleamed like perfumed camphor, and the columns were so slender they recalled the necks of beautiful youths.62 The palace doors must have been plated with a metal such as bronze, since ‘you can see in each one … such light that it seems the sun is silhouetted in them’.63 The mention of ‘balconies … above the high and brilliant stars’ again implies the presence of miradors, and the terracing of the landscape to construct the view.

The reference to ‘perfumed camphor’ interestingly evokes the multi-sensory experience of being in one of these palaces, in which expensive scents must have been all-pervading. There was also the sense of sound from pools which furnished palace interiors, adorned with decorative basins and fountain heads, even live animals. The presence of pools is indicated in the poem by al-Jazīrī and the description by Ibn Jabīr of the reception hall of al-Maʾmūn at Toledo, both discussed in the previous chapter. Al-Jazīrī’s poem, for example, describes ‘a large basin of green water’ standing ‘in the centre of the hall’, ‘in which the turtles continually make sounds’.64 Its lion-shaped fountain-head ‘is of scented aloeswood’ (nadd), perhaps a description born of poetic exaggeration, but intriguingly again evoking the notion of strong scents pervading the air of the palace interiors – aloeswood being one of the five important scents.65

Also located in the private zone of al-Manṣūr’s palace complex were the harem,66 and apparently the public and private treasuries, presumably because this location was particularly secure. The plural used in Ibn ʿIdhārī’s text (buyūt al-amwāl) may imply that the treasury was formed by various rooms, all containing different types of precious possession, in the same way that the Fatimid treasuries are known to have been thematically organised.67 Descriptions of the palace’s destruction mention al-Mahdī’s mob looting ‘money, jewels and precious objects’, which were presumably stored in these treasuries; over a period of two days, they ‘cleared out the greater part of the stores of clothes, carpets and furniture, perfumes and jewellery, treasures, arms and provisions’.68 The descriptions of the looting of al-Zāhira also mention carved marble dadoes,69 and tell us that the ‘strong gates, the sturdy wood and the rest of what was found in the palace’ were ripped out and ‘sold in all parts’.70 It is likely that some of the objects discussed in this book were looted from al-Zāhira at this time: the marble basin in the Alhambra (Figure 156, Chapter 7 4.2.1), for example, bears an inscription which indicates that the Taifa ruler of Granada, Bādīs ibn Ḥabbūs al-Ṣanhājī (r. 1038–73) brought the basin ‘to the palace (ila qaṣri) of his capital’, the obvious implication being, he brought it from somewhere else (see Appendix 4.20). Its provenance is unrecorded, though its iconographic relationship to the two basins produced in al-Zāhira for al-Manṣūr (Figures 113–118) and his son ʿAbd al-Malik (Figures 128–133) argues that it was from the ʿĀmirid palace-city, or one of its munyas. The enormous size and weight of the basin and the effort of transporting it 170 km from Cordoba to Granada, suggests that Bādīs’s acquisition of this object was not driven merely by a search for building materials. Rather, his ownership of this marble object was a statement of association with the ʿĀmirids. As to the whereabouts of the luxury arts ateliers themselves, we can only speculate, though the inscription on al-Manṣūr’s basin saying that it was made bi-qaṣri’l-zāhira (Appendix 4.7), indicates that they were well established here by 377/987–8, ten years after the foundation of the city (Chapter 6).

All these buildings were surrounded by luscious gardens which were planted to be aromatic as well as aesthetic,71 and which formed an essential element in the framed view from al-Manṣūr’s miradors. It is possible that the gardens included a menagerie, as al-Manṣūr would have needed somewhere to house the exotic animals that he received as a diplomatic gift from Zīrī ibn ʿAṭiyya (see Chapter 2). While the two hundred pure-blood horses and fifty mahara camels were probably housed in al-Zāhira’s stables,72 the bird ‘which spoke eloquently in a marvellous voice’ would perhaps have been housed in an aviary, if not a cage in one of the palaces, while more particular arrangements would have to be made for the musk ox, the ‘tiger of marvellous physique and enormous body’, civet cats, and ‘varieties of desert fauna’ including the lamt antilope; we can only imagine where al-Manṣūr housed the poor giraffe that had died en route and arrived stuffed, though ‘there was great amazement when people saw it’.73 In comparison, we know that at Madīnat al-Zahrāʾ there was a large zoological park (al-ḥāʾir), and an aviary protected by netting.74 Here ʿAbd al-Raḥmān III housed the animals he received as diplomatic gifts from his North African allies, which also included camels and racehorses, as well as sheep, gazelles and ostriches.75 For his lions, ʿAbd al-Raḥmān constructed a dedicated lion-house behind his palace at Cordoba, ‘on a bridge that was elevated over a gully … which still today bears the name of the Bridge of Lions’.76 The ownership of a reserve filled with exotic animals may have been one of the features that ʿAbd al-Raḥmān imitated from the Abbasid palace model that influenced his construction of al-Zahrāʾ.77 There are also textual references to collections of lions in the Abbasid cities of Iraq: al-Muhtadī ordered the lions of the Dār al-Khilāfa at Samarra to be killed in 869, while Byzantine envoys to Baghdad in 917 were shown ‘one hundred lions – fifty to the right and fifty to the left, each handled by a keeper, and collared and muzzled with chains of iron’.78 A reserve for wild or hunting animals seems to have been a ‘must-have’ feature of the new palace style and, given the number of other features the two extra-urban palace-cities had in common, as well as the practical need to house these exotic diplomatic gifts, it may well have been adopted at al-Zāhira as well.

As well as live animals, the palace gardens were no doubt also populated with inanimate creatures in the form of decoratively-carved fountain basins and heads, mirroring those inside the palace halls. Poetic descriptions refer to the lakes (ṣahārīj) constructed in front of the palaces and adorned with fountain-heads in the forms of lions.79

The lavishness of al-Zāhira’s gardens is implied in the anecdote of the waterlilies and the gold and silver coins, cited above. The ‘vast lake’ which the High Hall of the Munyat al-Surūr overlooked evokes the monumental lake which has been excavated at the site of the Munyat al-Rummāniyya, to the west of Madīnat al-Zahrāʾ. In the northwestern corner of the complex, on the uppermost terrace, was found a huge water basin, with a depth of more than 4 metres. It is the largest and deepest such basin yet found in Cordoba.80 It was surrounded by a walkway projecting on cantilevers over the water. Recent excavations have revealed the mechanism by which water from the basin was distributed to the various terraces of the munya’s gardens and orchards; the water may also have supplied the palace and stables with fresh water, and the basin may have contained fish to provision the estate owner’s table.81 Arnold notes that the volume of water stored here could also serve as a means of tempering the microclimate of the palace, cooling adjoining spaces in the summer.82 It was also used as a boating lake – such as those of the Dār al-Baḥr at the Qalʿa of the Banū Ḥammād,83 and the Qaṣr al-Baḥr at Aghlabid Raqqāda84 – and Arnold cites texts that speak of rafts floating across the water on which musicians sat and guests drank wine.85 The existence of boating lakes in the ʿĀmirid palaces is indicated by the sources: Ibn Shuhayd’s poem cited above mentioned that ‘al-Zāhiriyya shone brightly with pleasure boats’, and the description of ʿAbd al-Malik and Ḥabība’s wedding at al-ʿĀmiriyya mentions ‘boats on its clear lakes [resounding] with sweet music’.86 And as we saw from the Ibn Ḥazm anecdote quoted above, al-Manṣūr also liked to go boating with his courtiers on the River Guadalquivir.

This outline of the physical appearance and spatial organisation of al-Zāhira’s palace levels is obviously inconclusive, since the textual evidence can only take us so far, and as yet nothing can be verified archaeologically. Nevertheless, it is clear that there were many formal respects in which al-Zāhira resembled its caliphal model, Madīnat al-Zahrāʾ, and it goes some way to helping us reconstruct the physical setting of al-Manṣūr’s private life, in particular, the intimate majālis discussed in Chapter 3. However, perhaps the more important relationship between the two palace-cities was at the institutional level, and we shall now turn to consider which institutions were housed at al-Zāhira, and which were not.87

3 Reconstructing the City

Neither Madīnat al-Zahrāʾ or al-Madīnat al-Zāhira was just another munya writ large, but was also ‘a self-sufficient city, conceived on a monumental and urban scale’.88 In studying these royal foundations, it is impossible to disentangle this dual character. Both madinas housed all the appropriate institutions of a city, but they nevertheless remained dependent on Cordoba, so that the two seats of power together functioned as what Mazzoli-Guintard has termed ‘une capitale à double polarité’.89 After al-Manṣūr transferred all the important organs of the state bureaucracy to al-Madīnat al-Zāhira in 981, the palace-city became his centre of operations until his death.

3.1 The Mosque

After the palace, or dār al-sulṭān, the most significant formative entity of a Muslim city was the congregational mosque. The sources do not describe the foundation of al-Zāhira’s mosque in any detail, but a number of anecdotes attest to its existence, and its location next to the eastern (and possibly only) gate, the Bāb al-Fatḥ.90 It must have been constructed sometime between 980 and 990.91 The most detailed account is found in the Tartīb al-Madārik, detailing the controversy over whether al-Manṣūr was permitted to celebrate the Friday prayer in the al-Zāhira mosque:92 that is, whether his mosque was recognised as a jāmiʿ or congregational mosque, as was the al-Zahrāʾ mosque. The Cordoban fuqahāʾ were divided over this issue: the grand qāḍī of Cordoba, Ibn Zarb, was opposed, and the majority of jurists seconded the fatwa he issued against al-Manṣūr. However, others supported a favourable fatwa issued by Ibn al-ʿAṭṭār, which argued that the extensive urbanisation of Cordoba’s environs made it impossible for much of the population to attend the Great Mosque in the centre of Cordoba. These favourable jurists ruled that the distance between the Cordoba and al-Zāhira mosques, as measured by the longest route, was about a parasang (3–3.5 miles) and therefore great enough to justify the existence of a jāmiʿ mosque in both places. This accords with Islamic legal rulings on the circumstances in which the multiplicity of jāmiʿ mosques is permitted,93 though certain conditions were attached: the older mosque always took precedence and, if possible, the believer should pray there, even if he had already said the Friday prayer in the newer mosque. This would explain why the Tartīb al-Madārik tells us that a group of jurists who lived in the Rabaḍ al-Zāhira attended the prayer in the al-Zāhira mosque, so that ‘al-Manṣūr would not harbour rancour against them’, but afterwards repeated it in the Cordoba mosque.

Al-Manṣūr cautiously waited until after Ibn Zarb’s death in 992 before acting on the more favourable of the two fatwas, whereupon he appointed Ibn al-ʿAṭṭār as imam and mufti; the al-Zāhira mosque subsequently became that qāḍī’s office, where he issued fatwas and lectured on fiqh. However, the congregational status of the al-Zāhira mosque remained controversial and unpopular among the fuqahāʾ: al-Manṣūr’s first two choices for imam refused the appointment.94 Ibn Dhakwān (d. 1024), grand qāḍī of Cordoba (1001–1010) and a good friend and influential advisor of al-Manṣūr, who also had a residence within al-Zāhira’s walls, nevertheless refused to accompany him in prayers at the al-Zāhira mosque, out of deference to Ibn Zarb’s ruling.95 These anecdotes underline al-Manṣūr’s respect for Maliki juridical opinion, as discussed in Chapter 1; they are also interesting for the insistence they place on al-Zāhira’s dependence on Cordoba, and the higher authority of the Umayyad dynastic mosque in the historic capital. Al-Zāhira’s mosque was not intended to supplant that monument, but remained subordinate to it. Ibn al-Khaṭīb reports that, for a while, the Friday prayer was only celebrated at al-Zāhira – but this is likely to have been due to the works on al-Manṣūr’s extension to the Cordoba mosque rendering it unusable for a time (see Chapter 5).96

Other textual references to the al-Zāhira mosque are brief and sparse in detail, though they confirm that it remained a congregational mosque for the rest of its short life. As mentioned in Chapter 3, Ṣāʿid al-Baghdādī wrote his Kitāb al-Fuṣūṣ whilst sitting in the ṣaḥn of the ‘jāmiʿ mosque of the city of al-Zāhira’.97 The mosque is mentioned in the Almohad period as a ruinous and isolated Sufi refuge, indicating that parts of it were still standing almost two hundred years after the city was razed by al-Mahdī.98 While none of these texts indicates the appearance of the al-Zāhira mosque, it is probable that, like the al-Zahrāʾ mosque, it recreated the form and features of the Cordoba mosque on a reduced scale. The al-Zahrāʾ mosque was a small monument of five naves, a courtyard and a minaret, and despite its rapid construction was richly decorated in carved stone and plaster. According to al-Maqqarī, 1000 skilled artisans worked on it, and its maqṣūra was said to be ‘of wonderful construction and ornamented with costly magnificence’, containing a minbar of ‘extraordinary beauty and design’, installed on the day the mosque was completed. Friday prayers were held there for the first time in Shaʿbān 329/May 941, indicating it was a jāmiʿ mosque from the start.99 It seems likely that the al-Zāhira mosque would have had similar, if not smaller, dimensions. Did it have its own minaret and minbar as at al-Zahrāʾ? It is doubtful that al-Manṣūr would go so far as to appropriate such potent symbols of royal power. Chapter 5 discusses al-Manṣūr’s extension and interventions at the Cordoba Mosque, and highlights some innovative architectonic and decorative features that seem to have been used there for the first time. Since it is likely that the construction of the al-Zāhira mosque predated the ʿĀmirid extension at Cordoba, had al-Manṣūr already tried out some of these features in his own mosque? Indeed, at Cordoba he was constrained by the need to work within the mosque’s existing vocabulary and the weight of Umayyad precedent, while at al-Zāhira he could do as he liked. Its decoration may have been highly innovative and experimental, as perhaps suggested by some of the unusual capitals that have been attributed to this period, discussed below.

3.2 Organs of State Bureaucracy

Turning to the bureaucracy of al-Zāhira, the sources tell us that when al-Manṣūr took up residence in his palace-city, he established there the administrative departments of the Cordoban civil service, in which the high officials of the state took office (al-dawāwīn wa al-ʿummāl).100 The axis of the government administration thus shifted east, but which offices and officials were these? Significantly, it does not appear to have included the fundamental caliphal institutions of the Dār al-Sikka and Dār al-Ṭirāz, though the Dār al-Ṣināʿa or at least some of its craftsmen do seem to have relocated to al-Zāhira, because we know al-Manṣūr’s marble basin was made there (see Appendix 4.7). The operation of the luxury arts industry, including the Dār al-Ṭirāz, under the ʿĀmirids will be discussed in Chapter 6. Though the mint had been transferred to Madīnat al-Zahrāʾ in 947, it had followed al-Ḥakam back to Cordoba in 976, a shift reflected in the coinage in the change of the mint-name to ‘al-Andalus’. This is the mint named on all coins issued during al-Manṣūr’s ḥijāba (with the obvious exception of North African issues) and, though al-Manṣūr was clearly in control of the mint, as evidenced by the presence of his kunya on the coins (Figure 4) and his appointment of mint governors, there is no evidence for a mint at al-Zāhira. It must therefore have remained in Cordoba after 976.101

Nevertheless, al-Zāhira became the state’s fiscal centre. Ibn ʿIdhārī and Ibn Khāqān both recount that al-Manṣūr sent orders throughout the provinces of al-Andalus and the Maghrib that all tax revenues should be sent to al-Zāhira, and that local governors (ʿummal) should direct all petitions there.102 As mentioned above, the state treasuries (buyūt al-amwāl) were located in the inner sanctum of al-Manṣūr’s palace, presumably for increased security. As we saw in Chapter 1, after the domestic crisis (waḥsha) of 996–7, the caliphal purse was also moved to al-Zāhira. At the time of its sack by al-Mahdī, the al-Zāhira treasury contained the equivalent of 5,500,000 dinars in silver and 1,500,000 dinars in gold: it is said to have taken three days to transfer all of it to the caliphal palace in Cordoba.103

Mention in the sources of kuttāb (scribes) and ḥussāb (accountants) indicates that the Chancery, or Dīwān al-Inshāʾ, was moved to al-Zāhira.104 This makes sense not only in terms of the governance of the state, but also in relation to the private majālis discussed in Chapter 3; since many of al-Manṣūr’s nudamāʾ were kuttāb or held some other post in the Dīwān, their proximity to the ḥājib was desired as well as expected. We also know that the state viziers were based at al-Madīnat al-Zāhira:105 we saw above that al-Manṣūr granted the surrounding lands to his high state officials, and members of prominent families such as the Banū Ḥazm, Banū Ḥayyān and Banū Shuhayd moved their residences there. This made it all the easier to convene emergency council meetings, as al-Manṣūr did in 996–7, to discuss the response to Ṣubḥ’s theft from the Cordoban treasury.106 During the Fitna it was the ‘viziers of al-Zāhira’ who were in charge of the city’s defence, while Sanchuelo was absent on his fatal last campaign.107 Perhaps their council was held in the dār al-sulṭān, which is where the people addressed their petitions.108 Lastly, al-Zāhira had its own ṣāḥib al-madīna,109 and ṣāḥib al-shurṭa, whose headquarters were located next to the Bāb al-Fatḥ.110 There is no indication that al-Zāhira had its own qāḍī, as did al-Zahrāʾ,111 unless this role was fulfilled by Ibn al-ʿAṭṭār, once he was appointed imam of the al-Zāhira mosque, or Ibn Dhakwān, qāḍī of Cordoba, who had a residence in the city.112 Al-Zāhira had a prison and executions were held there, which implies access to the services of a qāḍī, though perhaps in this crucial legal office al-Zāhira again deferred to Cordoba.113

In addition to the treasuries, there were important stores of weapons (asliḥa) at al-Zāhira.114 Ibn al-Khaṭīb recounts that 15,000 suits of armour were deposited in this Khizānat al-Silāḥ, in addition to all the arms required.115 Small circular bronze discs that have been recovered from the Cordoba area offer a glimpse of the physicality of this military equipment, as ingeniously reconstructed by Ana Labarta.116 Only five are known, of which three can be dated to the ʿĀmirid period, two of them naming al-Manṣūr himself. These discs average 35 mm in diameter and 1 mm in thickness, and have two or three perforations, made before an inscription was incised into the metal. A near-complete example was found in the excavations at Cercadilla (Cordoba Archaeological Museum, inv. CER 93/S1/VE1/10-9-93; Appendix 4.22): it consists of two discs of equal size, each with two perforations, and preserves one of the rivets that connected them through an object – this object was not very thick because the rivet is only 3 mm long. Both external sides bear an inscription, which would need to be easily accessible for reading; the reverses of the discs are left plain. The inscriptions refer to weights, in a very precise manner. On the second (probably interior) of the discs from Cercadilla, the inscription reads ‘Its weight is 21 raṭls’, which has been struck through and rewritten above as ‘Its weight is 20 raṭls’. But what do the weights refer to? The exterior disc refers to ghilāla, the term used for a chainmail tunic, while another example in the Cordoba Museum (Figure 18, Appendix 4.21) mentions badan, the term for a short mail coat, or hauberk. These terms are qualified by adjectives: the tunic is called ḥarrāniyya, probably after the city of Ḥarrān in Mesopotamia, an important military centre in the Umayyad period. Another adjective used on these discs is kūfī, after the Iraqi city of Kufa. Labarta believes armour of these types was used for military campaigns, as opposed to the more decorative armour used for parades (burūz). An example of the latter is probably indicated on the disc in Figure 18, where the mail coats (al-abdān) are called al-ṣaqlabiyya al-muzayyana, ‘decorated in the Slav style’. The historical evidence does not allow us to reconstruct the appearance of these different types of armour, or what ‘the Slav style’ meant, but this parade armour could have been made in nielloed silver or silver-gilt, as the arms given in khilaʿ presentations frequently were.117

Figure 18
Figure 18

Bronze disc in al-Manṣūr’s name, found in Calle Cruz Conde, Cordoba, diam. 3.4 cm; Museo Arqueológico y Etnológico, Cordoba, inv. CE009509

© MAECO / Foto: Sil via Maroto Romero

The disc in Figure 18 is the 141st of the abdān decorated in this style, and the inscription further identifies it as the property of the ‘ḥājib Muḥammad ibn Abī ʿĀmir’. The discs from Cercadilla refer to the 11th of the Ḥarrānī mail tunics, also property of the ḥājib. Labarta has reconstructed that these numbers refer to inventories of arms and armour kept in the Khizānat al-Silāḥ, that must have been used to record to whom certain items were loaned when they were withdrawn for war or burūz, and from whom they were safely returned. These discs were attached through the rings of mail as tags to identify them. The reference to the ḥājib in both cases is because he would have paid for this armour to be made. The phrase fī sabīli-llah (‘for the cause of God’) on the Cercadilla disc further implies that this donation of armour to the arsenal was established as a waqf, or pious endowment, since this phrase is used in documents endowing arms or horses in this way.118 These unassuming little discs therefore contain a wealth of fascinating information about the operation of the state arsenals, as well as our only physical survival of the armour worn by the Umayyad army at this time.

Al-Manṣūr had granaries (al-ahrāʾ) constructed into the city walls, and mills (al-arḥāʾ) were erected along the banks of the Guadalquivir.119 Both these foundations suggest that al-Zāhira was intended to be self-sufficient in the supply of food to its inhabitants. They should also be considered as pious foundations, in the sense that they demonstrated the ruler providing for his people. Indeed, during a ‘terrible famine’ throughout al-Andalus, the Maghrib and Ifrīqiya, which began in 379/989–990 and lasted three years, ‘al-Manṣūr ordered that every day … should be made 22,000 loaves of bread, which were distributed amongst the poor’.120 Significantly, the description of al- Manṣūr’s generous behaviour during this time is comparable to that of al-Ḥakam during an earlier famine, in 353/964–5, and they are both credited with caring personally for the needy.121

Another indication of al-Zāhira’s intended self- sufficiency was the establishment of markets, another way in which al-Manṣūr’s city resembled Madīnat al-Zahrāʾ.122 These were ‘frequented by numerous caravans’ (wa qāmat bi-hā al-aswāq, wa kathurat fī-hā al-arfāq).123 A Yemeni jewel-seller who is mentioned anecdotally in the sources must have reached al-Zāhira’s markets in one of these caravans,124 though there is otherwise no indication of what commodities were traded in these souks. The people of Cordoba were attracted by these new markets and their associated job prospects, as well as the desire ‘to be near the head of state (ṣāḥib al-dawla)’,125 and so they also settled in al-Zāhira’s environs. The surrounding land that was not built on was divided into orchard gardens and fiefdoms, administered by the wuzarāʾ, kuttāb, quwwād and ḥujjāb to whom al-Manṣūr had granted it; this land was farmed by tenant-farmers (wa tanāfasa al-nās fī al-nuzūl bi-aknāfi-hā).126 Perhaps here was produced the grain that was stored, ground and made into the city’s bread, as well as other food products that were sold in the city’s markets. So many people settled in al-Zāhira’s environs that in a very short time her suburbs ‘joined those of Cordoba’;127 indeed, the way between al-Madīnat al-Zāhira, Cordoba and Madīnat al-Zahrāʾ was so intensely urbanised that it was said that, at night, one could walk more than ten miles under the uninterrupted light of the torches illuminating the streets of the three cities.128

3.3 Walls

A final and potent symbol of al-Zāhira’s self-sufficiency was its high and heavy walls.129 Contrary to Terrasse’s belief that al-Zāhira’s construction was rapid and flimsy, which is why the city disappeared so easily from the archaeological record, her walls appear to have been so well built that some of its towers were still standing in the twelfth century: on his way to campaign in Shawwal 567/June 1172, the Almohad caliph Abū Yaʿqūb Yūsuf (r. 1163–1184) made a stop ‘on the mountain of the Faḥṣ al-Surādiq, which dominated the towers of the site of al-Zāhira (abrāj arḍ al-Zāhira)’.130 This Almohad reference also allows us to deduce the significant topographical information that the Faḥṣ al-Surādiq, the ‘Field of the Tent’ or ‘Pavilion’ and the main mustering point for the Umayyad armies, not only lay to the east of Cordoba, but that al-Zāhira was sited close to it.

The first mention of the Faḥṣ al-Surādiq occurs in Ibn Ḥayyān’s history for the year 319/931–2, when he mentions that ʿAbd al-Raḥmān III began to plan a campaign against Toledo, ‘beginning the preparations and bringing out the pavilion and the tents to the encampment of Faḥṣ al-Surādiq, to the east of Cordoba’.131 Ibn ʿIdhārī tells us only that the Faḥṣ al-Surādiq was located on the northern bank of the Guadalquivir (bi-jawf al-nahr al-aʿẓam). In recent years, however, Manuel Acién Almansa and Antonio Vallejo Triano have been keen to locate the Faḥṣ al-Surādiq to the northwest of the city, closer to Madīnat al-Zahrāʾ: specifically, they want to identify it with the site of Turruñuelos.132 The Faḥṣ al-Surādiq is mentioned several times in al-Rāzī’s Annals, so was clearly well-established as the location of military staging by the 970s.133 Before then, the mustering point had been the Faḥṣ al-Rabad, on the left bank of the river in front of the Alcázar, next to the maṣalla where great public religious acts were celebrated, such as prayers for rain.134 Vallejo is keen to relate the transfer of the focus of military staging to the development of Madinat al-Zahrāʾ to the northwest of Cordoba.

Although the site of Turruñuelos remains unexcavated, it was obviously an extraordinary construction since it can be detected clearly by aerial photography and is visible on Google Earth. Rectangular in form and of huge dimensions, measuring 530 by 380 metres, the aerial images suggest that much of the interior space was empty, possibly planted as a huge garden. A large palatial complex stood on slightly elevated ground along the northern side. In this area fragments of marble decoration and column shafts can be seen lying on the surface. Additional buildings stood outside the enclosure.135 Vallejo believes this structure ‘should undoubtedly be related’ to the Faḥṣ al-Surādiq, ‘or, more likely, that this enclosure formed part of the Faḥṣ, as the most important piece of infrastructure of the military cantonment of the State’. He goes on to note that the lack of excavation impedes the precise identification of its caliphal chronology, though ‘the perfect geometry of its outline, its identical orientation to that of Madīnat al-Zahrāʾ, and the regularity of the preparation of some of its foundations suggest a date around the 950s for the construction of its enclosing wall’.136

Murillo et al seem to accept the identification of Turruñuelos with the Faḥṣ al-Surādiq, although they admit that the ‘interpretation of this immense and enigmatic site remains open for the moment’.137 They consider that its association with the military necessities of the state – ‘such as arsenals, the fabrication of arms or whatever other complex of buildings related to the function of a military camp’ – could be related to the ʿĀmirid period, because of the ‘discovery of a direct connection between Turruñuelos, by means of the road which leaves from the gate in the centre of its southern wall, and the complex formed by large paved patios and narrow corridors, recently excavated in the Plan Parcial Huerta de Santa Isabel Oeste’. While this complex is still being studied, the archaeologists provisionally identify it as ‘an enormous stables, complemented by surrounding areas of pasture and various auxiliary buildings’. There seems no reason in what they describe to associate this structure with al-Manṣūr, though they note that ‘al-ʿĀmiriyya, al-Manṣūr’s residence before the construction of Madīnat al-Zāhira … contained workshops and arsenals for arms as well as storage and stables for the breeding of horses’. In citing al-ʿĀmiriyya, they may be following Arjona Castro’s association of Turruñuelos with the caliphal munya of al-Ruṣāfa, rebuilt by al-Manṣūr after the fall of al-Muṣḥafī.138 This identification has not been accepted (indeed Murillo et al discuss another large structure which they identify as al-Ruṣāfa), but Arjona’s association of Turruñuelos with al-Manṣūr was largely based on the discovery of a marble capital with a signature in its abacus, which he read as ʿamala Saʿāda ibn ʿĀmir ʿabduhu.139 While Juan Souto originally concurred with Arjona’s reading, in a later study he reread the inscription as ʿamala Saʿīd ibn ʿUmar, and discounted the ʿĀmirid association.140 León and Murillo also mention a residential quarter excavated to the west of the suburb of al-Ruṣāfa, constructed ‘ex novo on a strictly orthogonal street grid’, which they interpret as the housing for the Berber troops serving the ʿĀmirid ḥājib (Chapter 1).141

All these interpretations might be correct: Turruñuelos may represent a massive military installation established in the 950s at the same time as Madīnat al-Zahrāʾ, and this area may indeed have witnessed reinforcement and redevelopment during the ʿĀmirid period, to house Berber troops and cater to the breeding and stabling of the cavalry and as a factory and arsenal for arms and armour. This does not also mean that this site should be identified as the Faḥṣ al-Surādiq, which all the Arabic histories state was to the east of the city. Vallejo himself admits that his identification is ‘despite the insistence of the sources in situating its placement at the eastern tip of Cordoba (ṭaraf Qurṭuba al-sharqī)’.142 Indeed, there is no need to have a permanent building on the site where the Umayyad armies came together before a campaign. As its very name indicates, this was a city of tents, and as Ibn Ḥayyān’s text mentions, the most important preparation for campaign was erecting the caliphal pavilion. In 1009, when Muḥammad ‘al-Mahdī’ seized Cordoba, ejected Sanchuelo and forced Hishām to abdicate, he ‘camped in Faḥṣ al-Surādiq, ordering provincials from every corner of al-Andalus to settle around his tent’.143 All the sources indicate that it was a temporary city of textile structures that sprang into being when the army came together.

The location of the Faḥṣ al-Surādiq to the east of the city, and in proximity to al-Madīnat al-Zāhira, also makes sense of Ana Echevarría’s observation that al-Zāhira’s main gate was located on the ancient Via Augusta, which led directly to the Marches.144 Taking Arjona Castro’s suggestion of the area of Las Quemadas as the most likely site for al-Zāhira (see above), just to the north of the second bend in the river Guadalquivir, this indeed sits close to the main arterial road departing Cordoba to the east, as plotted by Murillo et al.145 It is logical that the Umayyad army setting off on campaign would have taken the main road out of Cordoba, and that the Faḥṣ al-Surādiq would have been located close to this road (Figure 17).

As Lévi-Provençal describes, once a campaign had been announced, the ceremonial of war began with the burūz, or grand parade, during which the ruler, surrounded by a luxurious cortège and followed by the Cordoban population, would process from the capital and install himself at the Faḥṣ al-Surādiq; here he would review and organise the troops, and distribute their payments, before departing for the front. This mustering period could last between 20 and 40 days, while the troops congregated here from all over the south of the peninsula.146 Ibn ʿIdhārī also tells us that a garrison of six hundred men was permanently camped outside al-Zāhira, where they could be called upon in case of need;147 and that ‘from [al-Zāhira] the standards marched direct to certain victory’.148 These passages may indicate the constant readiness of al-Manṣūr’s troops to head off on campaign, as well as implying the city’s proximity to the army’s usual mustering point. The vista presented by al-Zāhira may thus have been one in which banners of the different regiments were constantly fluttering in the army camp around the city, a distinctive vista that may be significant in interpreting some of the visual imagery of the artwork created for the ʿĀmirid ḥājibs, in ateliers located within the city’s walls, as we will discuss further in Chapter 8.

As the end of the city’s history demonstrates, al-Zāhira was built to be defensible. Its walls were said to be impregnable, and its gate(s) could be closed against attack, as the city’s viziers urged its ṣāḥib al-madīna to do when its fall to al-Mahdī seemed imminent.149 Bariani argues that al-Zāhira was, in fact, intentionally small, with most of the buildings located outside the city walls, which would facilitate its defensibility; further, it may only have had one gate, as common for defensive structures.150 While Ibn ʿIdhārī mentions ‘gates’ in the plural, the only gate to be mentioned by name is the ‘easternmost gate’, the Bāb al-Fatḥ, or Gate of Victory, where the heads of al-Manṣūr’s executed foes were displayed,151 and where the people congregated to see the ʿĀmirid ḥujjāb depart on campaign.152 If the Bāb al-Fatḥ was the only gate, its location on the eastern side of al-Madīnat al-Zāhira put it at the farthest possible remove from Cordoba: arriving from Cordoba, you would have to travel all the way to the far side of the city in order to enter, and your progress would be perfectly visible along the way. According to Bariani, this enhanced the impression that the city was conceived as a fortress. It also meant that al-Zāhira shielded Cordoba, in the same way that it was the ḥājib’s role to shield the caliph. During the absences of al-Manṣūr and his sons from al-Zāhira, they would seal the Bāb al-Shikāl, a gate in the eastern wall of Cordoba’s al-Sharqīyya suburb, to enhance the defensibility of the capital.153

That it was apparently impregnable does not imply that al-Manṣūr feared an internal uprising against his regime, nor that he attempted to seal himself in a hermetic world of which he alone was master, or to further isolate Hishām – all suggestions that surface in the historiography of al-Zāhira. The heaviness of the city’s walls may have been another feature that was imitated from al-Zahrāʾ, whose walls were so substantial that the city’s maximum extent can be determined solely by their appearance in aerial photographs.154 But given the extent of al-Manṣūr’s campaigning, and the very real military threats from Iberian and North African neighbours, the defensibility of al-Zāhira was also a sensible precaution against potential backlashes, especially if it was sited in close proximity to a standing garrison and the main mustering point of the Umayyad army. It is only in the aftermath of the outbreak of Fitna that the idea of defence against internal threats was written into the historiography of al-Zāhira’s foundation.

4 Why Did al-Manṣūr Build al-Madīnat al-Zāhira?

This discussion of al-Zāhira’s defensibility raises the question of al-Manṣūr’s motivation for building this palace-city. The reconstruction presented here of what al-Zāhira looked like and how it functioned shows that in many important respects the palace-city imitated Madīnat al-Zahrāʾ, both in terms of the formal similarities of their palatial buildings, and in the centralisation of all the necessary institutions of a city. Both al-Zahrāʾ and al-Zāhira ‘duplicated civil, commercial and religious functions already in place in Cordoba’;155 however, an examination of the textual evidence for the state’s political and religious calendar has found no substantive functional differences between Cordoba and al-Zahrāʾ. For example, ambassadors were just as likely to be received at al-Zahrāʾ as they were at the qaṣr in Cordoba, and proclamations of military victories or against heretics were as likely to be read in the Cordoba mosque as the al-Zahrāʾ mosque.156

This is the context in which al-Madīnat al-Zāhira should be understood, as a self-sufficient, autonomous capital, functionally no different from the twin capital of Cordoba/al-Zahrāʾ. The city’s deference to Cordoba in important respects created a second axis of government, that is Cordoba/al- Zāhira, which eventually displaced the Cordoba/ al-Zahrāʾ axis. Cordoba continued to be the traditional and principal capital of al-Andalus: despite the relocation of the treasury to al-Zāhira after the waḥsha, the fundamental caliphal institutions – the Dār al-Sikka and Dār al-Ṭirāz – were never transferred to al-Zāhira, and the city’s congregational mosque had subordinate status to the Great Mosque of Cordoba, which continued to function as the primary jāmiʿ mosque in the area. It is also likely that the majority of the population continued to live in Cordoba,157 so that the ancient Umayyad capital was still very much at the heart of the state.

However, the historiography surrounding the foundation of al-Madīnat al-Zāhira inextricably implicates it in the culmination of al-Manṣūr’s appropriation of the power and prerogatives of the Andalusi caliphate. Some passages are explicit in their assertion of usurpation, such as those which speak of Hishām’s destitution of all caliphal insignia, or his lonely imprisonment at al-Zahrāʾ;158 while this theme is implicit in other sources – the ‘usurpation’ of the site on which al-Zāhira was constructed, or of the Friday prayer in the city’s mosque. Again this bias derives from Ibn Ḥayyān who gives three primary motivations for al-Manṣūr’s construction of a palace-city: the desire to manifest his independence, having reached the height of his power; the need for better security from his many enemies, since he feared for his life every time he needed to visit the caliphal palace; and lastly but most significantly, the desire to isolate Hishām and strip him of all power and influence.159

This view of al-Manṣūr’s city has been perpetuated in the modern attempts to understand and define it: Ruggles, for example, states that al-Manṣūr ‘strategically developed Cordoba’s east side … to draw building attention and activity away from Madīnat al-Zahrāʾ which was too closely identified with the Umayyads’;160 and that by building such a palace, al-Manṣūr cunningly ‘elevated himself to the level of kings’ and ‘put himself in the place of the caliph’.161 She even interprets the archaeological evidence according to an utterly literal reading of Ibn Ḥayyān: a subterranean passage, running from the Hall of ʿAbd al-Raḥmān III’s west corridor under the upper garden to the lower south terrace, was later blocked, ‘presumably by al-Manṣūr as part of his modifications to Madīnat al-Zahrāʾ when he incarcerated Hishām’.162 Such readings imply that al-Madīnat al-Zāhira was constructed specifically to efface al-Zahrāʾ, which became no more than a luxurious jail-cell for the caliph. On the contrary, as Bariani’s study of the sources has shown, after the crisis of the waḥsha was resolved, Hishām ‘often stayed at al-Zāhira’.163

In order to understand the motivations for al-Zāhira’s construction more objectively, it is essential to properly contextualise it in light of the historical discussion presented in the previous chapters. As we saw, the years 976 to 981 were the crucial period in the development of al-Manṣūr’s career, during which he rose to a position in the regency government and became one of the most powerful political players on the peninsular stage. The beginning of the end for his rival, al-Muṣḥafī, was sealed when al-Manṣūr contracted his dynastic marriage to Ghālib’s daughter, Asmāʾ, which was celebrated in Muḥarram 368/August 978. As a wedding present, the caliph Hishām promoted him to ḥājib. A few months later, al-Muṣḥafī fell from grace, and al-Manṣūr assumed full control of the regency government (Chapter 1). This was the situation in 368/978, when the sources tell us that al-Manṣūr ‘ordered his palace to be constructed’. It is likely that, ultimately, it was founded both to triumphalise his new role at the centre of government, as well as to begin to rationalise the organs of government under his administration.164 After the defeat of Ghālib in 981 and the culmination of al-Manṣūr’s triumph, al-Zāhira took on an extra, symbolic role. Just as Madīnat al-Zahrāʾ had been ‘the architectural equivalent of … caliphal titulature’,165 so al-Zāhira was as important a symbol of victory as his adoption of a laqab: both emblematised the beginning of al-Manṣūr’s rule as sole regent. When he transferred the organs of government to al-Zāhira that same year, the palace truly became ‘a new seat for the rule of al-Andalus’.166

Furthermore, just as al-Manṣūr was not the first ḥājib to adopt a laqab (see Chapter 1), so he was not the first to build a palace: al-Muṣḥafī’s home in elegant al-Ruṣāfa was ‘one of the most magnificent residences in Cordoba’ (Chapter 6).167 But just as al-Manṣūr’s laqab was undeniably more ‘caliphal’ than those of his predecessors, we can identify a similar difference in scale in his construction of al-Zāhira: in graduating from munya to madīna, al-Manṣūr was adopting a caliphal architectural form that clearly articulated his ḥijāba as if it were the caliphate. After 981, al-Zāhira was undoubtedly perceived by al-Manṣūr, his court and the population of Cordoba, as a ‘celebration of triumph and a testament of power’.168 It became a physical and metaphysical symbol of the new centre of power in al-Andalus.

While this symbolic aspect cannot be denied, there were also very practical reasons for al-Zāhira’s foundation. In establishing a new form of government, al-Manṣūr would have surrounded himself with administrators who were personally loyal and answerable to him, just as he reformed the organisation of the regiments in the Andalusi army. Inheriting the administrators of the old regime, who might harbour loyalty to al-Muṣḥafī, was not a shrewd move for someone who wanted to stamp his own identity on government or seek to organise it in a more efficient way, just as happens when governments change today. At the same time, while Madīnat al-Zahrāʾ nestled cosily up against the foothills of the Sierra Morena, al-Zāhira was outward facing and strategically located: the city sat in a wide plain which opened onto the Guadalquivir, and was clearly visible from the river. Of course this also implies that anyone approaching from the east, by boat or road, was clearly visible to watchmen on the city’s walls. The main approach to the Umayyad capital was from the east, from the major ports of Algeciras and Almería, which provided al-Andalus’s access to North Africa, the Mediterranean, and the wider world;169 and this was the direction in which the Andalusi troops would head out on campaign.

Figure 19
Figure 19

Roman stele reused to carve the inscription commemorating the restoration of the city of Lisbon by al-Manṣūr in 985; Museu da Cidade, Lisbon, inv. MC.ARQ.CSJ.40.EP.0009

© José d’Encarnação
Figure 20
Figure 20

Drawing of inscription in Figure 19

© Carmen Barceló

Facilitating the mustering and movement of troops for the army’s biannual campaigns was a central concern for al-Manṣūr, and much of the evidence we have for his other architectural construction – especially bridge- and road-building and the fortification of garrison towns – is connected with this. Indeed, the booty from al-Manṣūr’s successful campaigns probably contributed funding towards these architectural projects.170 As we saw in Chapter 1, he ordered the construction of a defensive wall at Ceuta, and the historical sources tell us that he created a new base for the Andalusi fleet at Alcáçer do Sal, on the Portuguese coast south of Lisbon, which became a mustering point for the combined land-sea attack on Santiago de Compostela in 997.171 An inscription found in excavations at Castillo Sao Jorge in Lisbon provides the information that al-Manṣūr also undertook reforms at this strategic coastal site, which are not otherwise known from the historical record. Eleven lines of Arabic are carved in relief below the Latin text of a Roman tombstone made from a local stone and datable to the first century AD (Figures 19–20; the full text is given in Appendix 4.6).172 The inscription refers to a restoration of the city – tajdīd Madīnat al-Ashbūna – in 374/985, which Barceló interprets as the repair of the city’s walls. This would connect the fortification of Lisbon with al-Manṣūr’s wider policy of reinforcing the frontier zones of al-Andalus in the wake of the campaign against Barcelona (also 985) and leading up to his attack on Santiago.173 The text is in the name of the caliph Hishām, though the works mentioned were orchestrated by ‘his servant (ʿabdihi), his ḥājib and the sword of his State (sayf dawlatihi)’, Muḥammad ibn Abī ʿĀmir. Significantly, the title ‘al-Manṣūr’ is omitted from this inscription, though he is referred to by what appears to have been the standard ḥājibal title of Sayf al-Dawla.

Ibn ʿIdhārī and al-Maqqarī both mention bridge-construction as among al-Manṣūr’s pious works: in 378–9/988–9 he built a bridge over the Guadalquivir which cost 140,000 dinars and was ‘appreciated by the people’. He spent even greater sums on a bridge over the river Genil at Écija, which ‘smoothed out the difficult roads and the steep ravines’. The choice of the location of this second bridge is interesting, since Écija was the site of two fountains commissioned by ʿAbd al-Raḥmān III and Ṣubḥ, and as such had an important association with the caliphal family.174 A fragmentary marble inscription preserved in Fuentes de Andalucía (prov. Seville, some 35km from Écija) has been identified by Carmen Barceló as part of the foundation inscription commemorating this bridge construction (Figure 21, Appendix 4.8). Though only two lines of Kufic script survive – of excellent quality carving, in the style of the late 970s or 980s – they name [Abū ʿĀ]mir Muḥammad ibn Abī ʿĀmir and allude to damage caused by a flood in the winter of the year [98]3 (ẓalama-hu sayl shatwa sanat thalath …). Barceló notes that the historical sources speak of the Genil flooding Écija and the damage caused to the bridge and mills along its banks; if similar damage occurred in 373/983, it is likely that this inscription celebrated the completion of the bridge’s repair five years later, in 378/988.175 The fragment from Fuentes may thus be a physical survival of an example of ʿĀmirid bridge construction.

Another is preserved on the Puente de Alcántara in Toledo (Figure 22, Appendix 4.9). This tells an interesting epigraphic tale that was reconstructed by María José Rodríguez and Juan Souto.176 The bridge was reconstructed by Felipe II (r. 1556–1598) and bears several sixteenth-century commemorative inscriptions arranged around another inscription, which dates from an earlier rebuilding of the bridge by Alfonso X in 1259. Preserved in the text of this inscription are elements of a third, tenth-century inscription, since it summarises in Castilian an Arabic inscription from 387/997–8 commemorating the first restoration of this bridge by al-Manṣūr. The text of the original Arabic inscription can be reconstructed as reading: ‘Al-Manṣūr Muḥammad ibn Abī ʿĀmir, the wazīr of the amīr al-muʾminīn, al-Muʾayyad bi-llāh Hishām – may Allāh prolong his life! – ordered the reconstruction of this bridge. And it was completed, with the help of Allāh, under the direction of the qāʾid of Toledo, Khalaf ibn Muḥammad al-ʿĀmirī, in the year 387’.177

Figure 21
Figure 21

Fragment of an inscription commemorating the restoration of a bridge at Écija in 988; Museo Fuentes de Andalucía, H 17 cm, W 34 cm

© Carmen B arceló
Figure 22
Figure 22

Inscription on the Puente de Alcántara, Toledo

© Kent Rawlinson

Since no foundation inscription survives from al-Manṣūr’s mosque extension (Chapter 5), these fortuitous epigraphic survivals preserve the only known construction inscriptions in his name. In the Lisbon inscription, he omits his laqab ‘al-Manṣūr’ but is identified as ḥājib and by the title Sayf al-Dawla; in the Toledo inscription, it is interesting that he should have chosen to give himself the relatively inferior title of wazīr at a date that marks the apogee of his career. Rodríguez and Souto attribute this to the fact that al-Manṣūr had designated his son ʿAbd al-Malik with the title ḥājib in 991–2 and had adopted the title al-malik al-karīm in 996–7, while retaining the title wazīr which can be held by more than one person; they do not attempt to explain why he does not appear in the inscription as al-malik al-karīm (a title of which Bariani is rightly dubious, see Chapter 1).178 The Castilian word used in the thirteenth-century inscription is ‘alguacil’, deriving from the Arabic but still today a widely-used Spanish word for a governor or minister or someone with a legal role. The thirteenth-century inscription may have substituted the word ‘ḥājib’ with an already well-understood Castilian equivalent. In both cases, however, the inscriptions underline that Ibn Abī ʿĀmir’s authority derives from the caliph Hishām.

Returning to al-Madīnat al-Zāhira, Ruggles may be right to note that al-Manṣūr ‘colonised’ the zone to the east of Cordoba, but his motivations were not to do with the effacement of the Umayyad zone to the west. Ultimately, the reasons were practical, associated with the efficient centralisation of government, the strategic dominance of the main approach to the Umayyad capital, and with the mustering and movement of troops on campaign. All that al-Manṣūr might stand accused of was aggrandising his palace-city to visualise and triumphalise his position as sole regent and de facto ruler of the state.

One final point relates to the deliberate and ideological use of onomastics by al-Manṣūr in the choice of the name ‘al-Zāhira’ for his palace-city, and the ‘Bāb al-Fatḥ’ for the city’s main – or only – gate. Such naming strategies mark a new departure for Andalusi city gates, since those of both Cordoba and al-Zahrāʾ bore names that simply described them or the destination to which they led, such as the Bāb al-Jinān (Gate of the Gardens) in Madīnat al-Zahrāʾ, or the Bāb al-Qanṭara (Gate of the Bridge) or Bāb Ishbīliya (the gate leading to Seville) in Cordoba.179 Hitherto, there seems to have been no deeper level of meaning to the nomenclature of city gates in al-Andalus, though the names themselves were significant enough that the twinning of the double capital, Cordoba/al-Zahrāʾ, was underscored by the recapitulation in the latter of the names and relative placements of the former’s Bāb al-Ṣūra and Bāb al-Sudda – what Safran calls ‘commemorative referencing’.180 Rather, in naming the main gate of his city the Bāb al-Fatḥ, al-Manṣūr approximated the Fatimid practice of naming gates or cities in accordance with their ideology and propaganda: all the Fatimid cities had gates that evoked the concept of victory and the rise of Ismaʿilism, such as the Bāb al-Naṣr and Bāb al-Futūḥ in Ṣabra-Manṣūriyya and Cairo. Furthermore, mutual referencing often occurred between the names of the Fatimid caliph and his city, for example, al-Mahdī and al-Mahdīyya, built circa 916, soon after ʿUbayd Allāh had adopted the caliphal title; or al-Manṣūr and Ṣabra-Manṣūriyya, both names adopted in 947 to commemorate the victory of the third Fatimid caliph over the Kharijite rebellion led by Abū Yazīd.181 In this respect, the use of fatḥ for the name of the ʿĀmirid gate might deliberately pun on the meaning of the founder’s laqab, and imply a similar element of referencing between the names of the ruler and his city.

Was al-Manṣūr consciously evoking a Fatimid practice in thus naming his gate? It has been suggested that the Andalusi Umayyads engaged in a ‘war of words’ with the Fatimids, specifically that in giving his city the name ‘al-Zahrāʾ’ (‘Shining’, ‘Resplendent’, bright with its white marble buildings), ʿAbd al-Raḥmān III consciously evoked the epithet of the Fatimids’ ancestress, Fāṭima, also ‘al-Zahrāʾ’, and thereby appropriated their ‘symbols and language’ in his attempt to represent his caliphate as the only legitimate one.182 Given the anti-heterodox messages that can be read in the epigraphic programme of al-Manṣūr’s eastern façade of the Cordoba mosque, it is not unlikely that the emergence of unorthodox religious trends in al-Andalus prompted the ḥājib himself to adopt this onomastic game in his own articulation of power (Chapter 5, section 3).

It should be remembered that the Bāb al-Fatḥ was also the gate from which the ʿĀmirids marched their armies to battle – as Ibn ʿIdhārī said, ‘from [al-Zāhira] the standards marched direct to certain victory’ – and the adoption of this name may have had a more auspicious and ceremonial character. It may also have deliberately evoked the Surat al-Fatḥ, the Qurʾānic sura of ‘victory’ that was recited in the Great Mosque of Cordoba during the ritual of knotting the banners, one of the culminating moments in the ceremonial of preparing for jihad as discussed in Chapter 5.

This would still indicate an awareness of onomastic and ideological resonances between the city’s gate and the ritual of preparing for war. Al-Manṣūr clearly used onomastics to express his ideology of rule, and it can be argued that he followed a pattern of name-selection similar to that of the Fatimids. Most significantly, his choice of name for his city in the same z-h-r root as al-Zahrāʾ was surely as deliberate as his choice of a laqab in the same n-ṣ-r root as ‘al-Nāṣir’ (the title of ʿAbd al-Raḥmān III) and ‘al-Mustanṣir’ (the title of al-Ḥakam II). In both cases, al-Manṣūr was making a statement of alignment with the Umayyads, and in choosing the name ‘al-Zāhira’ he was again attempting to define himself as the practical (rather than actual) successor to the Umayyads. We will see this meaningful use of words again in the deliberate semantic punning employed in the inscriptions on ʿĀmirid luxury arts, discussed in Chapter 8. Interestingly, there was a contemporary Fatimid instance of ‘al-Zāhira’ as the name of a palace: upon the completion of the Eastern and Western Palaces at al-Qāhira, al-ʿAzīz (r. 975–996), who himself bore the epithet ‘al-azhar’, renamed the entire complex ‘al-Quṣūr al-Zāhira’.183 Perhaps al-Manṣūr again consciously adopted a central element of ʿAbd al-Raḥmān’s model of caliphal legitimacy – namely, the appropriation of Fatimid ‘symbols and language’.

5 What Did al-Madīnat al-Zāhira Look Like?

A small and scattered collection of architectonic elements – mainly capitals – can be associated with the ʿĀmirids through the deployment of the same motifs and vocabulary as seen on their portable objects (see Chapters 6–8). This allows us to propose that at least some of these elements came from ʿĀmirid buildings, whether in al-Madīnat al-Zāhira itself or in one of the family’s munyas. One group of capitals has stylistic elements that relate it closely to the ʿĀmirid marble basins, and if they came from the same spaces this would imply a decorative coherence and interplay between the architecture of these spaces and the works of art which furnished them. As with most capitals that were made in caliphal al-Andalus, the order employed is composite, in that they combine the volutes of the Ionic order with the acanthus leaves of the Corinthian order.184 These works are characterised by a simplification of the Corinthian elements, a greater ‘flatness’ of style, and the use of what Marinetto Sánchez has called ‘less noble materials’, such as sandstone (though it is important to remember that Madīnat al-Zahrāʾ was by no means all marble – most of the wall decoration was carved from limestone).185 But the picture is far from one of artistic decline: we see innovations such as double-capitals (Figures 37–40), as well as pieces which play with the capital form, ‘melting’ the rigid marble material and the standard elements of Andalusi capitals. There is a greater frequency of zoomorphic motifs, and the addition of a figural scene in the echinus (the space between the volutes) – as, for example on Figure 23 – is a stylistic and technical innovation of the ʿĀmirid period which has never been noted and deserves more detailed investigation. The capitals discussed below all have different shapes, sizes and decoration, and as such probably come from different buildings and building types. None of them has a clear provenance.

Two capitals of unknown provenance now in the Museo de la Alhambra seem to come from the same original location, since their material, size and carving style are basically identical (Figure 23), though only one of them bears figurative decoration in its echinus; this capital also has one row of acanthus leaves where the other has the usual two.186 Both are carved from sandstone, which is not the usual material for architectonic supports in the caliphal period – columns, capitals and bases made for Madīnat al-Zahrāʾ and other government buildings were usually formed from marble. However, if Arjona Castro was correct in suggesting that the location of al-Madīnat al-Zāhira should be plotted around the area of Las Quemadas, this communicates with the alluvial zone of the Arenal, or ‘al-Ramla’ as it was called in the Arabic sources, both toponyms deriving from words for ‘sand’. As discussed at the start of this chapter, ‘important archaeological remains’ were found in this area, including large blocks of sandstone.187 It would be logical to use the stone quarried out for the levelling and foundations of the site to construct the buildings and carve architectonic elements – especially if it were being built quickly – and it is therefore possible that these two sandstone capitals came from buildings in al-Madīnat al-Zāhira. Being a softer material, this contributes to the more worn appearance of the carving, as well as to their red colour.

Figure 23
Figure 23

Capital with ducks and a worm in the echinus, late tenth century, sandstone; Museo de la Alhambra, inv. R.4489

© Mariam Rosser-Owen

Furthermore, stylistically they relate to other objects whose manufacture can be attributed with certainty to ʿĀmirid patronage. One of the two capitals unusually features a figurative scene in the echinus at the front of the capital; on the back, the band is filled with vegetal scrolls.188 The scene features two ducks both pecking at a worm or water-snake, which wriggles in the centre. These ducks are of exactly the same type as those seen in the ʿĀmirid marble basins (Figures 118A, 129, 157), in some of which the ducks also peck at worms. They are so close stylistically that the capitals must have emanated from the hands of the same craftsmen as the basins, which argues that the same artisans worked on the architectonic decoration of the palaces as on the luxury objects produced in the Dār al-Ṣināʿa (discussed further in Chapter 6). The use of sandstone relates these capitals to the small basin also in the Alhambra Museum, while that in Seville is made of limestone, another ‘less noble material’ (Figures 164–168). The fact that there is only figurative decoration on one side of the capital suggests that this side faced forwards: it may have supported an arch in a garden pavilion, with one of the stylistically-related basins located close by, so that viewers could draw the visual connection between the two elements.

Its pair is a composite capital with two superimposed rows of eight ‘plain acanthus’ leaves around the basket (body), while its volutes project less than on the previous example.189 Its decoration is entirely vegetal, in a low relief style with minimal undercutting of the decoration, very similar to that of the previous capital; and on both sides the echinus contains floral scrolls identical to those on the back of that capital. These similarities and the fact that this capital is also carved from sandstone suggest that it shares a provenance with the previous capital.

A third capital can be related to these two, through its flat carving style and the presence of a wriggling worm or snake in the space above the front right volute (Figure 24), though this example is marble.190 Its decoration is again entirely vegetal, formed by two rows of acanthus leaves, which rather than being plain are decorated quite naturalistically. It has the same characteristic flat carving style as the previous two examples, which can be most clearly seen in the large palmettes that fill the volutes. The word ʿamal appears in the gusset at the front of the capital, the beginning of a craftsman’s signature the rest of which is now lost. To the right of this is a small worm with a smiling face. Its proximity to the signature suggests this might be a visualisation of the craftsman’s name or perhaps a workshop marker. It is not uncommon to encounter rather playful motifs in ʿĀmirid art, especially on the basins: the nature of the figural carving sometimes conveys a real sense of humour, as well as a reference to the natural setting in which these objects were located. While the ducks, turtles and fishes seen on some of the basins allude to their function as water reservoirs, the presence of a worm out of context on this capital seems to have no significance apart from playfulness on the part of the carver. However, it is interesting to observe that some of the Roman capitals reused in the Great Mosque of Cordoba – in particular, late first- or early second-century AD examples which probably came from the theatre at Mérida (Figure 25) – have a fleuron at the centre of their abaci containing a small serpentine motif, which looks a lot like a worm.191 Did the stone masons working on the ʿĀmirid mosque extension – who likely saw employment on other ʿĀmirid constructions, as we will see below – take inspiration from the Roman capitals they saw inside the mosque?192

Figure 24
Figure 24

Capital with worm or snake above the volute, late tenth century, marble; Instituto Valencia de Don Juan, inv. 5.234, H 22.5, Max W 32, W (base) 22 cm

© Cuauhtli Gutiérrez
Figure 25
Figure 25

Capital from Mérida with fleurons in its abacus, reused in Great Mosque of Cordoba, late first- or early second-century AD, marble

© Rose Walker / Photo: JohnBattenPhotography

The echinus decoration of these ʿĀmirid capitals invites comparison with a number of other, rather more elaborate examples of this phenomenon, and leads to the consideration that these could also be examples of ʿĀmirid art. The first is a marble capital coming from the Casa del Gran Capitán (Gonzalo Fernández de Córdoba, 1453–1515) in Cordoba and now in the Museo Arqueológico Nacional (Figures 26–7), whose figurative elements relate iconographically to the large ʿĀmirid basins.193 Apart from the basin in León (Figure 159, discussed in Chapter 7), this capital is the only other object to make use of ‘stilts’ below the legs of both lions and griffins, which might serve to confirm its ʿĀmirid chronology. Another composite capital, the wide space between the volutes is occupied by scenes of two lions devouring a deer, and affronted pairs of winged griffins, which alternate around the circumference of the capital. Gómez-Moreno clearly believed this capital to date from the ʿĀmirid period, since he introduced it by mentioning that ‘the appearance of marbles with animated themes’ was ‘a novelty in the closing stages of the caliphate …, under al-Manṣūr and his sons’. The lion-gazelle combat on this capital recalls the prominent use of this motif on the large basins, as does the presence of the winged griffin, which we see on the narrow sides of the ʿĀmirid basins, though on the capital they are represented with an unusual loop in their hind quarters, which may recall the representation of seamonsters on Late Antique sarcophagi.194 Both motifs of the lion-gazelle combat and the griffin are emblematic of the ʿĀmirids’ projection of their power and legitimacy as ḥujjāb, as argued in Chapter 8. This capital shows that these emblems were carried over into the architectonic decoration of ʿĀmirid spaces, marking these in the same way that the caliphs’ names and titles inscribed on earlier capitals physically marked those as ‘Umayyad’. It also indicates that a coherent programme was enacted in the decoration of al-Madīnat al-Zāhira – if that is where this capital ultimately came from – which served to underline the ʿĀmirids’ messages of power. Compared to the capital with the ducks and worm, the carving of the figures here is much more three-dimensional and sculptural. Though the undercutting is deeper, the flat-relief style of the background carving nevertheless relates it to the capitals just discussed, while the prevalent vertical motif of the basket is a version of the ‘chainlink’ seen on the double-capital, discussed below, which may be a chronological indicator of ʿĀmirid capitals.

Figure 26
Figure 26

Capital with lions and griffins in the echinus: side with griffins, late tenth century, marble; Museo Arqueológico Nacional, inv. 52117, diam. 28, H max 36 cm

© Mariam Rosser-Owen
Figure 27
Figure 27

Detail of the capital in Figure 26

© Mariam Rosser-Owen

A very worn marble capital now in Girona shifts the figurative decoration from the abacus to the volutes, which are formed by the heads and trunks of four elephants, though only two corners remain (Figure 28).195 Its lower part is composed of two superimposed rows of ‘plain acanthus’, and the elephants’ trunks curve downwards to combine with the topmost row. The whole capital is very worn from water erosion, but between the heads runs a band of decoration that seems to be entirely vegetal, executed in the undercut technique familiar from Umayyad capitals. Castejón was unsure about dating this piece to the tenth century, partly because its proportions are slenderer than those of caliphal capitals (it is just over 14 cm in diameter at its widest point); he suggested instead a twelfth- or thirteenth-century date. The closest comparanda appear to be a set of capitals in situ in the early eleventh-century crypt of Vic Cathedral, which are carved from a local stone with very roughly-hewn decoration. Nevertheless, as Rose Walker discusses, the volutes are ‘enlarged and, more inventively, turned into pairs of lions or birds whose hindquarters form the curl of the volutes, whilst their heads almost meet where the fleuron would otherwise be’.196 The capitals at Vic might be another example of the legacy of ʿĀmirid art on the Peninsula in the early eleventh century, as discussed in this book’s Conclusion. Nothing comparable to the Girona capital exists to provide a strong ʿĀmirid connection, besides that fact that its closest parallels are the capitals we have been discussing, with figurative decoration on their abaci. The volute from al-Rummāniyya with a lion-head and birds (Figure 29), which probably dates from the 960s, indicates a precedent for the use of figurative carving in the volutes of capitals; while a number of extraordinary examples from Carthage, dating to the fifth–sixth centuries, have projecting sculptural volutes in the form of eagles and rams.197 Could the carvers of this capital have looked again to Late Antique precedents? It is tempting to associate the depiction of elephant heads on the capital with al-Manṣūr’s control of the trans-Saharan trade routes, through his victories in North Africa and alliances with Zīrī ibn ʿAṭiyya, which gave him access to elephant tusks to be carved into ivory objects (Chapter 2). Elephants with similarly long, thin trunks with a curl at the end appear on the Pamplona casket (Figures 122, 124B), but otherwise the carving on the capital is too worn to discern any other comparisons between these two objects.

Figure 28
Figure 28

Capital with elephant-headed volutes, late tenth century, alabaster; Museu d’Arqueologia de Catalunya-Girona, inv. 550, H: 22, diam 11.5, Max W: 14.3 cm

© Arxiu d ’imatges. Museu d’Arqueologia de Catalunya-Girona
Figure 29
Figure 29

Volute from al-Rummāniyya, c. 960s, marble; Museo Arqueológico y Etnológico, Cordoba. A. side with birds; B. front view with lion head © John Patterson/DAI Madrid; C. reconstruction drawing of the volute and associated fragments © Felix Arnold/DAI Madrid

Another capital without clear parallels also came from the Casa del Gran Capitán in Córdoba and is now in the Museo Arqueológico Nacional (Figure 30); presumably both this and the capital with lions and griffins (Figures 26–27) had an original provenance in the Cordoba area.198 This unique capital is completely cylindrical, with an abacus of a wider diameter projecting above it, and four lions’ heads as ‘volutes’ which link the abacus with the body of the capital. Other than this, the body is not divided into zones, and instead is filled with a free and naturalistic decoration of vegetation which contains a single bird among its foliage. The lions’ manes spread up above their heads and they have rather startled expressions on their faces. From their mouths descends a vegetal roll which spreads out into the body of the capital like water, and becomes the basis for the covering of vegetal decoration: thus, the relationship between volutes and body decoration functions ‘metaphorically’ like four lion-shaped fountain heads around a circular pond. The flat-relief surface of the capital again relates it to other possibly-ʿĀmirid capitals, though much of the marble has been excavated from around the foliage, leaving a plain ground, which is technically sophisticated. The low relief further invokes the flat surface of water, through which the viewer looks to the plants waving below. Though the style of foliage does not really have close parallels, it can be related to the waving aquatic plants on the small ʿĀmirid basins (Figures 164–168), though here it is more naturalistic. If this capital is not actually much later, it seems to represent a clever visual pun, and the naturalism of its decoration is a sophisticated technical accomplishment.

Figure 30
Figure 30

Capital with lion-headed volutes, late tenth century, marble; Museo Arqueológico Nacional, inv. 52118, diam. 25, H 37 cm

© Mariam Rosser-Owen

Finally, among the figurative capitals, is another unicum – the well-known but still enigmatic capital with musicians (Figure 31).199 This large capital of the ‘Madīnat al-Zahrāʾ type’ has a single lower zone of eight ‘plain acanthus’ leaves on which four musicians rest their feet; each musician occupies the centre of each side of the capital and forms its main decoration. The figures have all been decapitated, presumably victims of aniconism at a later period. Each musician carries a slightly different instrument, which seems to suggest a troupe: the figures at two ‘sides’ of the capital both hold an ʿūd in their left hand and strum it with their right (Figures 31A, B), while the figure at the ‘back’ holds a smaller stringed instrument that Dwight Reynolds has identified as one of the earliest representations of a three-string fiddle (Figure 31C).200 The figure at the ‘front’ holds no instrument at all and has been identified as a singer (Figure 31D). Each musician wears an outfit of baggy trousers, whose folds are represented as prominent stripes. This way of depicting clothing is most closely related to the figures on the Xàtiva basin (Figures 146–147), though there the style is rather over-evolved. Musicians depicted on the ivories do not wear this style of short, baggy pants, though the comparison is not direct since few of them are depicted in standing poses. We see a standing ʿūd-player on the al-Mughīra pyxis (Figure 3), but this figure wears a long robe that reaches down to his feet; seated musicians are seen on the lidless pyxis in the Louvre and on the Pamplona casket, however in both cases their robes again appear to reach to their feet. Short tunics – though not gathered in the same way as the musicians’ trousers on the capital – are seen on the dancing figures on the rectangular ivory panel in the Metropolitan Museum, which has been attributed to the ʿĀmirid period (Figure 32). Though not conclusive, this might imply that the capital is also of this period. Perhaps short, gathered clothes were the usual garb worn by musical performers.201

Figure 31
Figure 31

Capital with musicians: A. ʿūd player; B. ʿūd player; C. bowed instrument player; D. singer; late tenth century, marble; Museo Arqueológico y Etnológico, Cordoba, inv. DOCC133, diam (base) 26, H 43 cm

© Mariam Rosser-Owen
Figure 32
Figure 32

Panel with dancing figures, mid to late tenth century, ivory, inv 13.141

Metropolitan Museum of Art, New York

The background to the musicians is formed by deeply-undercut but simply-represented vegetation, which is almost identical to capitals produced under al-Ḥakam II; since capital-carvers working for al-Ḥakam no doubt continued to work for al-Manṣūr, this may not in itself be significant for dating purposes. If it were a caliphal-period capital it would be highly unusual in bearing figurative ornament, though perhaps it came from the private context of a munya, rather than the official context of Madīnat al-Zahrāʾ, where no figurative decoration has been found among the thousands of architectonic fragments recovered during excavation. In its original context, the ‘musicians’ capital’ presumably would not have been alone: at least two capitals are needed to support the standard three-arched porticoes of Umayyad buildings, and if it came from a garden pavilion in a munya, there might have been capitals like this supporting arcades on all sides, conveying an impression of being surrounded by music and song. These hypothetical other capitals may have provided the prototype for the concept of projecting three-dimensional decoration from the abacus and body, which was subsequently developed under the ʿĀmirids; but since the only other examples of this technique can be associated with ʿĀmirid art, it seems most logical to group this capital with them. It also fits well with the privileging of poetry and music at the ʿĀmirid court (Chapter 3) and, in the same manner as other ʿĀmirid examples, its physical location in a space where musical entertainments occurred would create a cycle of referentiality between the setting and the works of art which adorned it.202

The addition of decorative elements, especially figurative ones, to the echini of these capitals seems to be a development of the ʿĀmirid period. It is a playful invention, ‘melting’ the capital’s rigid marble form and subverting its classical orders. It is not seen on any of the many surviving capitals from Madīnat al-Zahrāʾ, however there are other indications that experiments were made with adding figures to architectonic elements in the 960s. The site of al-Rummāniyya has yielded a significant quantity of material with figurative decoration. The marble volute found there decorated at its apex with a lion’s head and on its sides with a procession of birds, indicates that the capital from which it came had, at the very least, three other such volutes, and may have been completely covered with figurative decoration (Figure 29C).203 Another fragmentary element recovered from al-Rummāniyya – from a basin or possibly an impost block – depicts the snouts of two sharp-toothed animals snarling at each other from either side of a floral motif.204 Two marble basins have also been found here, depicting the heads of moustachioed lions and horned gazelles peaking out between spirally curling leaves.205 Another significant architectonic object with figurative decoration is the cyma – the impost block that sat above the abacus of a capital, and from which sprung the arch above – inscribed with the date 353/964–5 and apparently bearing the name ‘Jaʿfar al-ḥājib’ (Figures 33–35).206 This could refer either to Jaʿfar al-Ṣiqlābī, favoured eunuch and courtier of ʿAbd al-Raḥmān III and al-Ḥakam II, or Jaʿfar al-Muṣḥafī, vizier at the end of al-Ḥakam’s life and the first regent of Hishām II’s reign. (This object will be discussed in detail in Chapter 6, in the context of the architectural patronage of both Jaʿfars.)

Figure 33
Figure 33

Cyma in the name of Jaʿfar al-ḥājib, front view; 960–980, marble; private collection, Cordoba

© Mariam Rosser-Owen
Figure 34
Figure 34

Left side of the cyma in Figure 33

Figure 35
Figure 35

Right side of the cyma in Figure 33

All these architectonic elements came from private villas (munyas) built by high level courtiers of al-Ḥakam II: al-Rummāniyya was constructed by Durrī al-Saghīr for his own enjoyment some time before 973 when he gifted it to al-Ḥakam; while we know that al-Ṣiqlābī built a munya in the Cordoban hinterland around 964, and al-Muṣḥafī owned ‘one of the most magnificent residences in Cordoba’. If architectonic experimentation with introducing figurative elements into capitals and impost blocks occurred during al-Ḥakam’s reign, it is perhaps significant that it took place within the less rigid environment of the elite slave patronage that flourished at this time.207 The construction in a relatively short time of so many private spaces at al-Madīnat al-Zāhira – in which al-Manṣūr relaxed with his close courtiers without being bound by the rigid protocol that governed access to the person of the caliph (Chapter 3)– perhaps provided the crucible in which the craftsmen of the Dār al-Ṣināʿa could indulge their ideas and experimentation, allowing the figurative capital to truly develop at this moment.

This also seems to have been part of a wider trend in the decoration of capitals in the late tenth century, though the common inspiration may have derived from Cordoba. Walker has discussed the figurative capitals which start to be carved for Mozarabic churches in Christian Spain at this same moment, highlighting the narrative examples located in the crossing at San Pedro de la Nave, just outside Zamora, which could date to the 970s based on comparisons with illustrations in Beatus manuscripts. She also dates to the tenth century the large limestone capital from a Cordoban church, with simplified volutes and large single figures of the evangelists occupying the entire body of the capital between the volutes (Figure 36). The drapery of these figures’ clothes is treated in a very similar way to that on the musicians’ capital, as well as on the Xàtiva basin (Figures 142–143, see Chapter 7 3.2). It seems coincidental that carvers of capitals start to play with figurative ornament and to subvert the capital form in similar ways at similar times, and Walker speculates that the carvers of the Christian examples were trained in al-Andalus.208

Figure 36
Figure 36

Evangelists capital, late tenth century? limestone; Museo Arqueológico y Etnológico, Cordoba, inv. CE007931, W (abacus) 42, H 36 cm

© Mariam Rosser-Owen

The development of figurativism in architectonic elements may not have been the only example of the ingenuity of ʿĀmirid craftsmen. Juan Souto has associated an unusual double-capital in the Cordoba Archaeological Museum with ʿĀmirid patronage (Figures 37–38).209 Though badly damaged, this was clearly a virtuoso piece of carving, as indicated by the fact that it is signed by three craftsmen – Bāshir (front right), Faraj (front left) and Mubārak (side right) – all of whom have signatures in the ʿĀmirid extension to the Cordoba Mosque (Figure 50). It is probably no coincidence that a Faraj was the head of the ivory workshop that produced the Pamplona casket (Figure 126A, B): he was clearly a master carver leading a team, and the presence of three names on this double-capital probably indicates that it was also produced by a team of carvers. The undertaking to produce an unprecedented double-capital may have necessitated a pooling of skills in the workshop. According to Meouak, there was a Bāshir among the twenty-six al-fityān al-kubarāʾ or kibār of al-Manṣūr’s son and successor, ʿAbd al-Malik, who had the privilege of exclusively serving the ḥājib; Souto wondered if this could be the same person who signed architectonic pieces as well as objects, perhaps in the capacity of ‘“supervisor”, “director” or some other position of authority’.210 Another (single) capital, reused in the Alcázar at Seville, also bears three signatures, of Faraj, Mubārak, and Khayr (rather than Bāshir this time); significantly, Khayr was another signatory to the Pamplona casket.211 Mubārak’s name occurs again in the ʿĀmirid mosque extension as ‘Mubārak ibn Hishām’.212 This links the double-capital to the ʿĀmirid period without a shadow of doubt, and the connections with the Pamplona casket probably also account for the shared motifs between casket and capital. The pointed palmette leaf with round buds at its base that fills the vegetal scroll on the capital’s echinus is also seen on the lid of the Pamplona casket (Figure 125B); but, more particularly, the cable or chainlink motif which runs vertically between the lower range of acanthus leaves on the capital is also seen as an axial element on the back of the Pamplona casket, on either side of the central medallion (Figure 125D). This ‘chainlink’ motif may be a simplification of a criss-cross device used on the stems of projecting acanthus leaves on caliphal capitals,213 but in its simplified version it seems to become a characteristic device indicating an ʿĀmirid chronology. We can see it on the capital with lion-gazelle combat and griffins (Figures 26–27). It is also seen on the only other known examples of double-capitals: an identical pair in the Cordoba Mosque, which were probably installed during an ʿĀmirid intervention in al-Ḥakam II’s maqṣūra (Figures 39–40, discussed in Chapter 5).214 Additionally, the echinus decoration of these two capitals features an identical guilloche design to that which fills the bands surrounding the medallions all over the Pamplona casket (Figures 120–127).

Figure 37
Figure 37

Double capital, front view, signed by Bāshir (right) and Faraj (left), 980–990, marble; Museo Arqueológico y Etnológico, Cordoba, inv. CE028529

© Mariam Rosser-Owen
Figure 38
Figure 38

Double capital, back view, 980–990, marble; Museo Arqueológico y Etnológico, Cordoba, inv. CE028529

© Mariam Rosser-Owen
Figure 39
Figure 39

Double capital in ʿĀmirid tribune in Cordoba Mosque (east)

© Concepción Abad Castro and Ignacio González Cavero
Figure 40
Figure 40

Double capital in ʿĀmirid tribune in Cordoba Mosque (west)

© Antonio Vallejo Triano

All three capitals were carved as a perfectly integrated double-capitals, which introduces a new type of capital into Andalusi architecture. Again the idea for it may have derived from caliphal-period capitals: as Concepción Abad has noted, the perpendicular arcades of the Cordoba mosque’s three maqṣūra naves have twinned columns at the point where they intersect with the façades of the mihrab, Bāb al-Sābāṭ and Bāb Bayt al-Māl (compare the plans in Figures 42 and 49), providing extra structural support for the domes above. These twinned columns are crowned by twinned capitals, which have been made ‘double’ by the insertion of a vertical element; but the capitals are plain, like all the other capitals of al-Ḥakam’s extension.215 The ʿĀmirid period is the earliest one for which we have preserved examples of this innovative double-capital, though this prefigures the use of double capitals at the Taifa palace of the Aljafería (Figure 41), which may be another example of the legacy of ʿĀmirid artistic developments into the eleventh century. The surviving elements thus seem to verify the poetic description of arches supported by double columns in the Munyat al-Surūr at al-Zāhira, discussed above.

Figure 41
Figure 41

Twinned capitals at the Aljafería, Zaragoza, late eleventh century

© Kent Rawlinson

It is difficult to say what the other elements of the palaces’ architectural decoration could have looked like, as so many elements were decontextualized by the looting of the Fitna years. The walls of al-Madīnat al-Zāhira may have been lined by carved stone panels such as those now in Denia and Toledo (Figure 177), which present vertical patterns of floral scrolls featuring the distinctive pointed, star-like flowers that decorate the vegetal sides of al-Manṣūr’s and ʿAbd al-Malik’s large basins. These may have been removed from al-Zāhira’s palaces and taken to Taifa states by former ʿĀmirid lieutenants (Denia was a post-ʿĀmirid Taifa state, as discussed in the Conclusion), or carved in the Taifa period but reflecting ʿĀmirid style. In Chapters 7 and 8, we will examine the interplay between the decoration of objects which probably furnished palace gardens and the iconography of the poetry composed in and for these spaces, which can give further clues as to the appearance and setting of the ʿĀmirid palaces.

Fragmentary survivals such as these can start to build a picture of what ʿĀmirid architecture looked like. The capitals discussed here begin to indicate some of the innovative artistic developments that took place under al-Manṣūr’s architectural patronage. These may have been highly experimental in their architectonic forms, and highly figurative in their decoration, with deliberately playful and punning references between architectural spaces and the objects that furnished them. Grand capitals without any figurative decoration probably adorned the official spaces of al-Madīnat al-Zāhira’s palaces, where al-Manṣūr gathered his court for ceremonial occasions. In the more private reaches of al-Zāhira, in garden retreats or pavilions, or in the ʿĀmirid family’s munyas, a more figurative style prevailed, enforcing the ʿĀmirid dynastic emblems to underscore their message to those admitted to these spaces, but at the same time characterised by a playful inventiveness – breaking out of the rigidity of traditional forms, and setting up visual puns and metaphors between the artefact being viewed and the surroundings in which it was set. It is highly likely that the craftsmen who worked on al-Zāhira’s decoration also worked in other media within the ʿĀmirid Dār al-Ṣināʿa, as will be discussed further in Chapter 6, as well as on al-Manṣūr’s most audacious architectural project, the enormous extension he constructed at the congregational mosque of the Umayyad capital, Cordoba, as we will discuss next.

1

See also Rubiera Mata 1981, 132–133; Arjona Castro 1982, 212–213.

2

I undertook this exercise while researching and writing my DPhil thesis. Shortly after this was submitted, Bariani 2002a was published, which also looked at the Arabic sources to see what could be reconstructed of al-Zāhira’s original appearance. This chapter incorporates her discussion into my own.

3

Torres Balbás 1956, 356.

4

Torres Balbás 1957 [1996], 598.

5

Velázquez Bosco 1912, and below.

6

I am not aware of any more recent attempts than this to locate al-Zāhira or attempt excavation.

7

Velázquez Bosco 1912, 18–19, and passim. For more information on the Munyat al-Rummāniyya, see Anales de la Comisión Provincial 1926, 17–21, 34, 55; Arjona Castro et al 1994a, 252–253; Castejón 1945, 1949, 1954; Gómez-Moreno 1951, 166–171; Ocaña 1984a; Terrasse 1932, 166–67; Torres Balbás 1957 [1996], 594–597; Vallejo Triano 1987; Anderson 2013; Arnold, Canto García and Vallejo 2008, 2015 and 2019.

8

Ramírez de Arellano 1918, 132, 328–9, 337.

9

Anales de la Comisión Provincial 1926, 55.

10

Gómez-Moreno 1951, 165–166.

11

Ocaña 1952, reprinted 1964–5.

12

Ocaña 1964–5.

13

Simonet 1858 [1986], 86–87, n. 121.

14

Ocaña 1952, reprinted 1964–5.

15

Torres Balbás 1956; 1957 [1996], 597–600.

16

Arjona Castro et al. 1994b, 262. This argument is recapitulated in Arjona Castro et al. 1995.

17

Arjona Castro et al. 1994b, 264; Arjona Castro et al. 1995, 186–7.

18

Arjona Castro et al. 1994b, 266. The witness he spoke to ‘recuerda perfectamente que tenía un arco con dovelas de color azul brillante’: Arjona Castro et al. 1995, 187, n. 86.

19

Antonio Vallejo, personal communication.

20

Torres Balbás 1957 [1996], 598.

21

Bayān II:294 [translation, 457]; Ibn Khāqān, apud al-Ḥimyarī, 82; al-Maqqarī, 242; Dhikr Bilād I:181 [II:192].

22

The words of Henri Terrasse, delivered in an unpublished paper at the Ibn Ḥazm conference in 1963, and quoted in Castejón 1964–5, 62. Elsewhere in this article, Castejón speaks of the ‘fortaleza pétrea’ of Madīnat al-Zahrāʾ in comparison to the ‘fugaz delicadeza’ of al-Zāhira.

23

On the ‘mezquita aljama’ at Madīnat al-Zahrāʾ, see Vallejo 2010, 197–219. Though the texts indicate that the mosque was constructed in 941, fragments of the foundation inscription give the date 333/944–5. The rapidity of its construction could be reflected in the distinctive flat carving style employed for many of its decorative elements: see Escudero 2015, 73–5 (capitals); 24, 78–9 (fragments of an epigraphic frieze) – all these elements are carved in limestone.

24

All the sources concur in locating al-Zāhira to the east of Cordoba: Ibn Ḥazm, for example, who lived there, clearly locates it in the east when he relates how his family moved from ‘their new houses in the eastern part of Cordoba, in the suburb of al-Zāhira’ to the ‘old [houses] in the western part’ (in the suburb of Balāṭ Mughīth), in Jumādā II 399/January–February 1009, after the success of al-Mahdī’s rebellion: cf. Ibn Ḥazm 1949, 286; 1953, 212. Cf. also see Bayān II:313 [translation, 485]; al-Maqqarī, 209–211 (Analectes, I:261, 268). According to Ibn Bashkuwāl (1101–1183), the Rabaḍ al-Zāhira was still in his day one of the seven suburbs comprising al-Sharqiyya, the populated zone to the east of Cordoba: see Zanón 1989, 29, §2.1.4.

25

Bayān II:294 [translation, 457]; al-Ḥimyarī 1938, 82.

26

Bayān II:275 [translation, 427–428]; al-Ḥimyarī 1938, 82. This prophecy, which derives from Ibn Ḥayyān, was said to have been known to al-Ḥakam, who tried to preempt it by being the first to build at that site. By a neat conceit, al-Manṣūr ‘still young, hard-working and unknown’, worked on this project and thus knew of the ‘favourable prognostications’ associated with the site, which is why he later chose to raise his palace there: see Torres Balbás 1956, 358–359.

27

The ‘old crone seer’ is a topos of other city foundation legends cited in O’Meara 2007.

28

Bariani 2002a, 333.

29

This translation comes from Franssen 2008, Appendix, xxix–xxx. I am very grateful to Élise Franssen for sharing with me her unpublished MA thesis on the ʿĀmirids and their ‘material traces’.

30

Ibn Ḥazm 1949, 282; 1953, 209–210.

31

Bayān III:63 [translation, 66].

32

Al-Ḥimyarī 1938, 82; Bayān II:295 [translation, 457].

33

Al-Ḥimyarī 1938, 82; Bayān II:295 [translation, 457]. It is an age-old topos of royal patronage that the builder is influential enough to summon the best artisans and materials from the farthest places: ʿAbd al-Raḥmān III had done so at al-Zahrāʾ, importing marble columns from North Africa and Byzantium; and al-Ḥakam II brought a mosaicist and tesserae all the way from Constantinople for his extension to the Cordoba mosque (Chapter 5).

34

Ibn Shuhayd’s family residence was the Munyat al-Nuʿman: see Dickie 1964, 251. ʿAbd al-Raḥmān III had also granted the lands surrounding al-Zahrāʾ to the nobility. In addition, he issued financial incentives to the people so that they would settle there and ‘urbanise’ his city: see Ruggles 2000, 53–85.

35

Bayān III:62–64 [translation, 65–66]. Arjona Castro et al 1994b, 256–258 discusses the location of the Qaṣr al-Ḥājibiyya, which they believe was located in the area of Rabanales, to the north-east of Cordoba.

36

Bariani 2002a, 331, citing in Ibn al-Kattānī 1966, 80–81; Lévi-Provençal 1957 [1996], 247, n. 125.

37

HEM III:381, n. 4; Anderson 2013, 117–9.

38

Bayān II (1951):398–99, cf. HEM III:381, n. 4.

39

Bariani 2002a, 330.

40

Ruggles 2000, 113.

41

Bayān II:297 [translation, 461].

42

Ocaña 1984a, 371–372.

43

Ibn Shuhayd 1963, 64–67 (metre: kāmil); cf. Monroe 1974, 160–163; Pérès 1990, 128–129.

44

Bariani 2002a, 332–3.

45

Anderson 2013, Chapter 5, ‘The Landscape of Sovereignty’.

46

Bariani 2002a, 328.

47

These were the Majālis al-Gharbī (West Hall), al-Sharqī (East Hall) and al-Dhahab (Golden Hall): see Ruggles 2000, 66–67. She believes the Hall of ʿAbd al-Raḥmān III (Salón Rico) can be identified with the Majlis al-Dhahab, which opened onto gardens with a pavilion, pools and water basins with figural sculpture.

48

Bariani 2002a, 331: ‘al-mushrif ʿalā al-nahr’.

49

Al-Maqqarī, 243–244 (Analectes, I:349). See also Bariani 2002a, 330–331. The date and purpose of this embassy is unspecified, apart from ‘the king of Rūm’ wanting to inform himself of the ‘circumstances and force of the Muslims’ (Bariani, ibid).

50

Al-Maqqarī, 243: ‘The legate was stunned by the vision of servants so richly dressed … This spectacle provoked stupor and admiration on the part of the Christian, who requested a truce with the Muslims. After returning quickly to his country, the ambassador said to his king: “Avoid enmity with this people! [With my own eyes] I have seen the land give up its treasures to them!”’ (Bariani’s translation: 2002a, 330–331).

51

See Ruggles 2000, esp. 107–108.

52

Bariani 2002a, 330.

53

Compare, for example, the description of the Qaṣr al-Khilāfa at Madīnat al-Zahrāʾ, in Ruggles 2000, 67.

54

Ruggles 2000, 53.

55

Ibn al-Kattānī 1966, 67–68, cited in Rubiera Mata 1981, 133–134. See also Pérès 1990, 137 and n. 73, citing al-Maqqarī, Analectes, I:385, 406. It is interesting to note that there was a Qaṣr al-Surūr in the caliphal palace at Cordoba: cf. Pérès 1990, 129, n. 31.

56

Ibn al-Kattānī 1966, 69–70, cited in Rubiera Mata 1981, 134.

57

Puerta Vílchez 2013a, 62, interprets this poetic description as ‘columnas de mármol dobles’.

58

Ibn al-Kattānī 1966, 68, cited in Rubiera Mata 1981, 133.

59

In a poem by al-Jazīrī, cited in Continente 1969, 133.

60

Ibn al-Kattānī 1966, 69–70, cited in Rubiera Mata 1981, 133.

61

Bayān II:297 [translation, 460].

62

Bariani 2002a, 331, citing Ibn Hudhayl, apud Ibn al-Kattānī 1966, 75–6; and Rubiera Mata 1981, 92–3.

63

Ibn Hudhayl, quoted in Rubiera Mata 1981, 133, citing Ibn al-Kattānī 1966, 69–70.

64

Continente 1969, 131–132; Ruggles 2000, 124.

65

The Abbasid court physician, Yuhanna ibn Māsawayh (d. 857), defined the five most important aromatic substances (uṣūl al-ṭīb) used in the medieval Islamic world as musk, camphor, ambergris, aloeswood, and saffron: see King 2008, 175. For the use of luxury fragrances in tenth-century Cordoba, especially in connection with artistic production, see Anderson and Rosser-Owen 2015, esp. 40–45.

66

Bayān III:63 [translation, 66]: during the sack of al-Zāhira, al-Mahdī embargoed his mob from sacking this private zone, which he claimed as his own. When the harem of the Banū Abī ʿĀmir fell into his hands, he released the free women and kept the female slaves for himself. On the women of al-Manṣūr’s harem, see Echevarría 2011, 104–111.

67

Bayān III:63 [translation, 66]. On the structure of the Fatimid Treasury, see al-Ḥijjāwī al-Qaddūmī 1996.

68

Bayān III:63 [translation, 66].

69

Bayān III:64 [translation, 67].

70

Bayān III:61 [translation, 64].

71

Anderson 2013, 105–35 (chapter on ‘Gardens’).

72

Mentioned by Ibn Khāqān, apud al-Ḥimyarī 1938, 81–82. See also Arjona Castro 1982, 214.

73

Ballestín 2004a, 185–6.

74

Ruggles 2000, 65; Torres Balbás 1957 [1996], 435, n. 70.

75

Ruggles 2000, 65, and Ibn Ḥayyān 1981, passim.

76

Ibn Ḥayyān 1981, 40–42 (§§23–25).

77

At Samarra, the Jawsaq al-Khāqānī contained a zoo (‘Ḥāʾir’, ‘al-Ḳali’, Encyclopaedia of Islam, 2nd Edition; Ruggles 2000, 90), while Rogers 1970, 137, describes the Musharrahāt as a zoological palace and garden on the opposite bank of the Tigris. Northedge 2005 disputes these identifications and interprets these instead as game reserves for princely hunting, see esp. 150–165, chapter on ‘al-Ḥayr’.

78

Northedge 2005, 151.

79

Bariani 2002a, 331, citing Ibn Hudhayl, apud Ibn al-Kattānī 1966, 75–6; and Rubiera Mata 1981, 92–3.

80

Velázquez Bosco 1912, 24; Arnold 2017, 107.

81

Ruggles 2000, 116; Anderson 2013, 113, who notes that the word birka used in the texts to refer to the munya’s pools also means ‘fishponds’.

82

Arnold 2017, 107.

83

Ruggles 1994; Ruggles 2000, 116, n. 18.

84

Mazot 2000, 137. This lake appears to have been many times bigger than the Rummāniyya lake: it was trapezoidal, measuring 89 and 130 metres on its short sides, and 171 and 182 metres on its long sides. It was used for aquatic jousting and boating parties.

85

Arnold 2017, 108.

86

Ruggles 2000, 113.

87

There is a very brief list of al-Zāhira’s institutions in Vallvé 1986, 260–262.

88

Ruggles 2000, 53.

89

Mazzoli-Guintard 1997.

90

Bariani 2002a, 333, citing al-Maqqarī (Nafḥ al-Ṭīb, I, 156) who explicitly states that the mosque was located near to the only gate: wa masjid [al-Zāhira] fi’l-bāb al-munfarid [=single] bi-hā.

91

Bariani 2002a, 336, because of the involvement of Ibn Zarb, chief qāḍī from 977 to 992, in the controversy about the congregational prayer. She also comments (p. 337) that this controversy ‘must have influenced [al-Manṣūr’s] decision to start constructing his extension to the Cordoba Mosque in 987’, the implication being that he was trying to pacify the fuqahāʾ and regain their support and approval.

92

ʿIyāḍ ibn Mūsā 1968, II:657–658; cited in Ávila 1980, 107–109.

93

See Calero 2000, esp. 128–130. According to her examination of the surviving fatwas, one of the circumstances in which the multiplicity of jāmiʿ mosques is permitted is when the extent of the city and suburbs is so great that the distance between the two mosques exceeds two miles.

94

As Bariani 2002a, 337, n. 61, notes, Aṣbag al-Ṭāʾī was sent into exile, while Abū Bakr ibn Wāqid was stripped of membership of the shūra and placed under house arrest. On the ʿulamāʾ publicly rejecting political office, which was quite common practice, see Marín 1994.

95

HEM II:295, 303. On the location of Ibn Dhakwān’s residence, see Bariani 2002a, 333, n. 32.

96

Bariani 2002a, 338.

97

Al-Maqqarī, 202.

98

Zanón 1989, 104–105, §9.23, citing Ibn al-Abbār 1955, 304, who relates that ‘the Sufi Abū al-Qāsim Khalaf ibn Yaḥyā ibn Khaṭṭāb al-Zāhid (d. 576/1187) had in it a place where he used to sit in order to preach’. For other twelfth-century references to the al-Zāhira mosque, see Ibn Bashkuwāl 1966, 233, 460, 598, 646.

99

Al-Maqqarī, 237–8. See also HEM II:137, n. 1; Torres Balbás 1957; Pavón Maldonado 1966.

100

Bayān II:295 [translation, 458]; Ibn Khāqān, apud al-Ḥimyarī 1938, 81–82; al-Maqqarī, 187; Dhikr Bilād I:181 [II:192]. Bariani 2002a, 336, comments that Ibn Ḥawqal, writing c. 378/998 and thus contemporary with al-Manṣūr, does not mention al-Zāhira as the centre of government. Perhaps the transfer of power to al-Zāhira was slow, and never total?

101

On al-Manṣūr’s role as ṣāḥib al-sikka, see the discussion in Chapter 1.

102

Bayān II:296 [translation, 459]; al-Ḥimyarī 1938, 82; cf. also Ibn al-Kardabūs 1986, 84 (§62).

103

Bayān III:61 [translation, 65]. Ibn ʿIdhārī also notes that after the sack of al-Zāhira, coin hoards were found buried in jars, which amounted to a value of 200,000 dinars. Presumably these hoards were buried by wealthy residents fleeing al-Zāhira, who for whatever reason did not return to reclaim their assets. This calls to mind the ‘varios tesoros de monedas’ which Arjona Castro et al 1994b, 266, notes are ‘frecuentísimos’ in the area of the Cortijo de Las Quemadas, where they believe al-Zāhira to have been located.

104

Bayān II:295 [translation, 458]; al-Ḥimyarī 1938, 82; Ibn al-Kardabūs 1986, 84 (§62). Dhikr Bilād I:181 [II:192] also mentions kuttāb.

105

Bayān II:295 [translation, 458]; al-Ḥimyarī 1938, 82.

106

Bariani 1996a, 45.

107

Bayān III:57–8 [translation, 62–3], which relates how the bad news of Hishām’s surrender to al-Mahdī reaches the ‘viziers of al-Zāhira’.

108

Bayān III:28 [translation, 34], which relates that ʿĪsa ibn Saʿīd, al-Muẓaffar’s ḥājib, was so important that no-one dared talk to him when he was en route from his house to al-Zāhira, and the only time anyone could catch him was during office hours at the dār al-sulṭān.

109

Vallvé 1986, 261, mentions that al-Manṣūr’s great-uncle, ʿAmr ibn ʿAbd Allāh ibn ʿAsqalāja, was ṣāḥib al-madīnatayn. Perhaps this role was shared with that of al-Zahrāʾ from the beginning, since according to Ibn Ḥayyān, as late as 1023–4 there was a ṣāḥib al-madīna of al-Zāhira and al-Zahrāʾ: see Bariani 2002a, 339–40.

110

Dhikr Bilād I:181 [II:192] says: ‘[Al-Manṣūr] placed a ṣāḥib al-shurṭa at [al-Zāhira’s] gate’; Bariani 2002a, 329. The poet al-Jazīrī held the post of ṣāḥib al-shurṭa at some point before his incarceration in al-Zāhira in 1003: see Continente 1969, 136 and n. 2.

111

Ruggles 2000, 53.

112

Bariani 2002a, 333, n. 32.

113

Various prominent figures were imprisoned or executed at al-Zāhira: al-Jazīrī was jailed there because of his involvement in an attempted coup against ʿĪsa ibn Saʿīd, and was reputedly assassinated in his cell: see Continente 1969, 123, 127, who says the dungeon was subterranean.

114

Bayān II:295 [translation, 458]; Bayān III:61, 63 [translation, 64, 66], which mention ‘arms and munitions’ as a subdivision of the treasury. On the weapons stores at al-Zahrāʾ, see Anales, §9.

115

Ibn al-Khaṭīb 1956, 119–20; Labarta 2016, 264.

116

Labarta 2016; Labarta 2019a.

117

On which see Ballestín 2006.

118

Labarta 2016, 274.

119

Bayān II:295 [translation, 458]; al-Ḥimyarī 1938, 81.

120

Dhikr Bilād I:153–154 [II:193]; see also Ibn Abī Zarʿ 1964, 115. The length of three years seems like a topos: the Dhikr Bilād mentions a ‘great plague of locusts which affected the whole country and caused great damage’, which began in 381/991–92, and again lasted three years: cf. Dhikr Bilād I:154 [II:193]. Ibn Abī Zarʿ 1964, 115, details al-Manṣūr’s response on this occasion.

121

Bayān II:251 [translation, 389]: ‘Al-Ḥakam took care of the sick and needy, whether in Cordoba, its suburbs or al-Zahrāʾ. He gave them nourishment and he thus saved their lives’. In comparison, ‘during this famine al-Manṣūr behaved as no king before him had ever acted, and made kind gestures: he helped the Muslims, fed the weak, waived the tithes, buried the dead and succoured the living’: Dhikr Bilād I:153–154 [II:193].

122

Al-Maqqarī, 238, lists markets among al-Zahrāʾ’s many ‘public and private establishments’, which also included baths: he gives two, while Bayān II:247 [translation, 383] gives 300, which seems like an exaggeration, since Bariani 2002a emphasises the small size of the city. Cf. also Torres Balbás 1957, 435.

123

Bayān II:295 [translation, 458]; al-Ḥimyarī 1938, 82.

124

Bayān II:313 [translation, 485]; al-Maqqarī, 209–211 (Analectes, I:261, 268).

125

Bayān II:295 [translation, 458]; al-Ḥimyarī 1938, 82.

126

Bayān II:295 [translation, 458]; al-Ḥimyarī 1938, 81.

127

Bayān II:295 [translation, 457]; al-Ḥimyarī 1938, 82; al-Maqqarī, 242–243. The topos of the suburbs joining together is also mentioned in relation to al-Zahrāʾ: see Ruggles 2000, 53–85.

128

Ibn ʿIdhārī, cited in Bariani 2002a, 330.

129

Bayān II:295 [translation, 458]; al-Ḥimyarī 1938, 81.

130

Zanón 1989, 79–80, §7.3.

131

Ibn Ḥayyān 1981, 216 (§190).

132

Lévi-Provençal 1932, 141, 225, citing Bayān II: 219/338; Acién and Vallejo 1998, 126.

133

García Gómez 1965, 359–60; García Gómez 1967, 170; Vallejo 2010, 125.

134

Vallejo 2010, 125.

135

Arnold 2017, 116.

136

Vallejo 2010, 126.

137

Murillo et al 2010b, 612.

138

Arjona 2000, 157 ff.

139

Arjona 2000, fig. 4; Arjona 2001, 384.

140

Souto 2002b was dedicated to this inscription; he revisited the inscription in Souto 2010a, 257 (no. 7.103). Arnold 2017, 117, returns to the ʿĀmirid association of Turruñuelos, saying ‘The possibility should not be discounted that Turruñuelos was a palace built by al-Manṣūr, either his country estate al-ʿAmirīya or the famed Madīnat az-Zāhira itself’ – but there is no evidence to justify this.

141

León and Murillo 2014, 25.

142

Vallejo 2010, 126.

143

Amabe 2016, 96.

144

Echevarría 2011, 182, citing Bayān III, trans. 12–13. She notes that its placement was therefore ideal for communicating quickly with Armilāṭ (Guadamellato), the Mozarabic monastery that served as a meeting and departure place for the armies. Armilāṭ is described in HEM II:282 and n. 1, as the first stopping station after Cordoba on the way to Toledo. It was located about 15 km north of Alcolea (prov. Cordoba), thus about a day’s march to the east from the capital. This was where ʿAbd al-Malik ibn al-Manṣūr died in 399/1008, on the road back to the city.

145

In Arjona Castro et al. 1995, 188–192, they explore the location of the Faḥṣ al-Surādiq in the neighbourhood of the Cortijo de Rabanales, to the northeast of the city, now one of the campuses of the University of Cordoba.

146

Lévi-Provençal 1932, 141. Bariani 2003, 210–11, notes that troops located in more northerly or frontier locations would probably join the army along the route.

147

Bayān III:58 [translation, 62].

148

Bariani 2002a, 328, citing Bayān II:277.

149

Bayān III:58 [translation, 62].

150

Bariani 2002a, 333–4.

151

ʿĪsa ibn Saʿīd was executed at al-Zāhira on 10 Rabīʿ I 397/4 December 1006 for his attempted coup against the ʿĀmirids, and his head was displayed ‘above the gate of al-Madīnat al-Zāhira’, where it remained until the fall of the ʿĀmirids: see Bayān III:33–34 [translation, 39–40]. His co-conspirator, Hishām ibn ʿAbd al-Jabbār, was thrown into al-Zāhira’s dungeons and never seen again: Bayān III:35 [translation, 41].

152

Bayān III:5 [translation, 13]. The gate is named in the context of al-Muẓaffar’s departure in 393/1003 on his first campaign as ḥājib. Bariani 2002a, 334, notes that postern gates are sometimes opened in the walls of defensive structures, which might explain the plural in Ibn ʿIdhārī’s text.

153

Lévi-Provençal 1957 [1996], 241, n. 101, citing Bayān III:56, though he notes that this name might also have been given to one of the gates in the Alcázar (Bayān III:89). See also Bariani 2002a, 334.

154

Vallejo 1992, 27; Vallejo 2010, 165, pl. 89.

155

On the institutions which al-Zahrāʾ housed, see Ruggles 2000, 53, 62–65; Vallvé 1986, 257–259.

156

Mazzoli-Guintard 1997, 50–52 for the enumeration of religious events; pp. 52–57 for political events; pp. 57–64 for her incisive discussion and conclusion.

157

Mazzoli-Guintard 1997, 63.

158

Bayān II:296 [translation, 459]; Ibn Khāqān, apud al-Ḥimyarī 1938, 82; Dhikr Bilād I:152–153 [II:190–192].

159

See, for example, Bayān II:294, 296 [translation, 457, 459–460].

160

Ruggles 2000, 113.

161

Ruggles 2000, 124, quoting Ibn Khāqān, apud al-Ḥimyarī 1938, 80.

162

Ruggles 2000, 85 and fig. 33. This assumption – that any later rebuilding at the site must provide physical evidence of the ʿĀmirids’ isolation of the caliph – is problematic, especially since we know from al-Idrīsī 1975, 579–580, that a small population continued to occupy the site of al-Zahrāʾ as late as the mid-twelfth century: see also Mazzoli-Guintard 1997, 62, n. 113. Such a modification could have been made any time, and Antonio Vallejo admits that it is incredibly difficult to date such examples of repair and reuse at al-Zahrāʾ (personal communication). Vallejo 2010, 162–3, says ‘Nuestro conocimiento sobre las reformas que cabe atribuir al prolongado gobierno del tercer califa de al-Andalus, Hishām II, es mucho más limitado y, hasta el momento, no tiene constatación epigráfica’.

163

Bariani 2002a, 339.

164

See Mazzoli-Guintard 1996a, 170–172, on ‘la ville triomphante’. Ballestín 2004a, 198 cites Ibn ʿIdhārī’s reasons for the construction of al-Zāhira as the centralisation of the administration of power, and to regulate closely access to Hishām.

165

Ruggles 2000, 92.

166

Safran 2000, 101.

167

Al-Maqqarī, 183.

168

Safran 2000, 102.

169

On the importance of Algeciras as a port, especially for access to North Africa, see Fierro 2008, esp. 588–592. On the port of Almería, which was refortified by ʿAbd al-Raḥmān III after a Fatimid attack in 954 as a home for the Umayyad fleet, see Lévi-Provençal 1932, 152–6; Lirola Delgado 1993, 198–212.

170

The important socio-economic impact of these campaigns, in terms of the availability of money and manpower from prisoners-of-war, is stressed by Echevarría 2000, esp. 109–110.

171

Echevarría 2011, 134.

172

Now in the Museu da Cidade (Inv MC.ARQ.CSJ.40.EP.0009), H: 111.2 cm × W: 40 cm × Depth: 29 cm. This was one of nineteen Roman inscriptions found during the excavations but, while these have been studied and published, the presence of an Arabic text was noted but assumed to be illegible. It was finally read and published by Barceló 2013.

173

Barceló 2013, 185. She says (p. 173) that this would also push the foundation date of the castle – previously thought to have been founded in the mid-eleventh century – back to the Umayyad period.

174

Bayān II:309 [translation, 479], and al-Maqqarī, 219. On the fountains at Écija, see Lévi-Provençal 1931, 36–37 (#29), 37–38 (#30), and Anderson 2012, 661–4.

175

Barceló 2013, 183–184; the full inscription is reconstructed in her Figure 6.

176

Rodríguez and Souto 2000b.

177

Rodríguez and Souto 2000b, 202–5, discuss the identity of Khalaf ibn Muḥammad al-ʿĀmirī. He will be discussed here in Chapters 6 and 7, in relation to the structure of the luxury arts industry and his possible role in the creation of al-Manṣūr’s fountain basin (Appendix 4.7).

178

Rodríguez and Souto 2000b, 201.

179

For the names of the five city gates known for Madīnat al-Zahrāʾ, see Ruggles 2000, 64; for Cordoba’s seven city gates, see Lévi-Provençal 1957 [1996], 236–238; on the eight gates of the caliphal palace at Cordoba, see Torres Balbás 1957 [1996], 592–593.

180

Safran 2000, 68: ‘Textual references make clear that the most important city gate (Bāb al-Ṣūra) and the most important gate into the qaṣr (Bāb al-Sudda) in Madīnat al-Zahrāʾ were named after their counterparts in Cordoba’.

181

See ‘Al-Mahdiyya’, and ‘Al-Mansur bi’llāh, Ismāʿīl’, Encyclopaedia of Islam, 2nd Edition.

182

Safran 2000, 58–59, 214 n. 28; Fierro 2004b, 319.

183

Rabbat 1996, 53.

184

The terms used here for the different parts of the capital follow Vallejo 2013; in particular see the diagram on p. 106.

185

Marinetto Sánchez 1987c, 185–186.

186

Marinetto Sánchez 1987c, 199–200.

187

Measuring 1.20–1.30 m × 0.45–0.60 m: Arjona Castro et al. 1994b, 264.

188

Inv. R4489; dims: 19 (D) × 27 (W) × 18.5 (H) cm; Marinetto Sánchez 1987c, 185–186; Arte Islámico en Granada, 264 (cat. 62).

189

Inv. 3749; dims: 20 (D) × 23 (H); Marinetto Sánchez 1987c, 184–185, 199.

190

Instituto de Valencia de Don Juan, Madrid, inv. 5.234. Bibliography: unpublished.

191

I am grateful to Rose Walker for discussing these Mérida capitals with me.

192

This inspiration might have happened at an earlier date. Vallejo 2013, fig. 113b (right) illustrates a capital from ʿAbd al-Raḥmān II’s mihrab (carved c. 848), reused in al-Ḥakam II’s mihrab: at the centre of its abacus is a fleuron with an abstract circular element at its centre that could be interpreted as a worm. ʿĀmirid craftsmen may equally have been looking to earlier Umayyad capitals for their models.

193

Gómez-Moreno 1951, 181, fig. 245a. Though he briefly discussed and illustrated this capital, Gómez-Moreno gave no indication of its dimensions or location, so its whereabouts were not obvious until the capital was displayed in the Islamic gallery of the Museo Arqueológico Nacional in Madrid following its refurbishment in 2014. It entered the collection in 1912 and bears inv. no. 52.117; dimensions: H: 36 cm; diam of calathos: 28 cm; max W: 36 cm. Information and images are available on the museum’s online catalogue. I would like to thank Isabel Arias for discussing this object with me. It is also mentioned in Martínez Enamorado 2006, 52, n. 123, who references Revilla Vielva 1932, 63, no. 152, plate 16.

194

I discuss this motif and its Late Antique origins in Rosser-Owen 2015b, 41–2. In Chapter 6, I argue that Late Antique sarcophagi were a likely source for the huge blocks of marble from which the ʿĀmirids carved their fountain basins. It would be highly probable for their craftsmen to have been influenced by the designs on such objects.

195

Found in the Castillo de Peratallada, Girona; now in the Museo Arqueológico, Girona (acquired 1934). See Dubler 1945; Castejón 1924, 163–164, plate 11.

196

Walker 2016, 262, fig. 107.

197

These capitals come from the rotunda at the site of Damous el-Karita. Made from marble, one group features rams as volutes while the other group features eagles. See Landes and Ben Hassen 2001, 151–4, cat. 53. My thanks to Glaire Anderson for bringing these extraordinary capitals to my attention.

198

Inv. no. 52.118. It entered the collection in 1912, together with 52.117 (see n. 193). Information and images are available on the museum’s website, which dates it broadly to the tenth century. See Castejón 1945, 206.

199

See Dodds in Al-Andalus, 248 (cat 40); Esplendor: Catálogo, 135–136; Fernández Manzano 1995.

200

Cited in Anderson 2018, 242–4.

201

Anderson 2018, 246–7, suggests that the singer on this capital is a woman: ‘This empty-handed figure wears a garment with an undisturbed diagonal drape … while the other three figures, who hold instruments, wear garments that appear divided up the middle to form what might be some type of pantaloons. Perhaps, then, the Musicians Capital gives us a representation of a female singer and three male instrumentalists’.

202

Anderson 2018 discusses the musical performances that took place in the semiprivate and private residences of the ruler and court notables during the caliphal period, and comments (p. 245) on the striking ‘reflexive qualities’ of the physical settings, in which the architectural decor, the functional use of the space, and the metaphors within the lyric performance, all mirror each other.

203

Velázquez Bosco 1912, 31–32, plate 10 (1); Anales de la Comisión Provincial 1926, 19; Castejón 1945, 206–208, 204; Gómez-Moreno 1951, 181, fig. 245b, 245d; Castejón 1954, 155, 153; Mann, Glick and Dodds 1992, 214, cat. 53; O’Neill 1993, 77; Anderson 2013, 73; Arnold, Canto García and Vallejo Triano 2015, 138–40 (no. 10).

204

Velázquez Bosco 1912, 32, plate 35; Castejón 1945, 208–209, 205; Gómez-Moreno 1951, 181, fig. 245c; Castejón 1954, 155, 154; Anderson 2013, 73–4. Arnold, Canto García and Vallejo Triano 2015, 132–3 (no. 1), believe this to be the fragment of a basin, as reconstructed in their Fig. 63b; Martínez Núñez 2015, 66 n. 33, has suggested that it might have come from a figurative cyma.

205

The larger of the two basins, which is housed in the Cordoba Archaeological Museum was published in: Anales de la Comisión Provincial 1926, 19–21; De los Santos 1926; Terrasse 1932, 166 n; Castejón 1945; Castejón 1949; Gómez-Moreno 1951, 191, fig. 252a; Castejón 1954. The second, smaller basin was published in Castejón 1945, 200–201, 203–206; Castejón 1949, 235–236; Gómez-Moreno 1951, fig. 252b; Castejón 1954, 155, 151; Anderson 2013, 73–5; Arnold, Canto García and Vallejo Triano 2015, 133–6 (nos. 2–3).

206

The main publication on this fascinating object (which is in the private collection of Fernando López Segura in Cordoba) is Martínez Enamorado 2006, though it was included in the small exhibition, Madīnat Qurṭuba, held at Casa Árabe, Cordoba: see Kedier 2013, 50–1, 83. It is discussed in Anderson 2013, 77–80, where it is described as a fountain basin.

207

Anderson 2012; 2013.

208

Walker 2015. On the capital with evangelists, see also O’Neill 1993, 50, cat. 8. The label in the Cordoba Archaeological Museum calls this capital ‘Visigothic’ and dates it to the sixth/seventh centuries.

209

Inv. 28.529. Souto 2007, #2.8; inscription no. 8; plates VII, VIII; Souto 2010a, 214 (2.18), 260 (7.108). Dims: H: 38 cm; total width of abacus: 68 cm; diam of each capital base: 25 cm.

210

Meouak 2004, 193, cited by Souto 2010a, 214 (2.18).

211

Souto 2010a, 221 (2.41), 260 (7.112), citing Martínez Núñez et al. 2007, n. 582.

212

Souto 2010a, 224 (2.55 and 2.56), 259 (7.107, fig. 26). Another capital signed by Mubārak is now in Cádiz, see Souto 2007, #2.9.

213

It is seen, for example, on a large composite capital from the Hall of ʿAbd al-Raḥmān III at Madīnat al-Zahrāʾ, dating 953–7, now in the Madīnat al-Zahrāʾ museum: Vallejo 2013, 106, fig. 95 (inv. 42.24077).

214

Abad Castro and González Cavero 2018, 235–9. I am grateful to the authors for bringing these capitals to my attention when they first began their study of the possible ʿĀmirid intervention into what is now the Capilla Real at the Cordoba Mosque (discussed Chapter 5, section 2.1.4.2), and for discussing them with me as their research developed.

215

Concha Abad, personal communication (22 June 2015); Abad Castro and González Cavero 2018.

  • Collapse
  • Expand