In this year [381/991–2] al-Manṣūr began Hishām’s extension (al-ziyyādat al-hishāmiyya) to the Great Mosque of Cordoba. He extended it almost by half again of what had been made by earlier sovereigns. The works began on the first of Rajab of the year 381 [13 September 991] and the prayer could be made in Rajab of the year 384 [August–September 994]; they lasted, therefore, three years. On the works there laboured Berbers, noble knights from Galicia and Frankia, chained together with the workmen until it was concluded.
Dhikr Bilād I:193–4 [II:182]
∵
Al-Manṣūr pursued a policy of ostentatious piety, carrying with him on campaign an autograph copy of the Qurʾān, and a shroud woven by his daughters in which he should be wound on the event of his death, collecting dust from every campaign to be scattered over his grave. As Hugh Kennedy wryly observes, al-Manṣūr would have ensured this piety was widely publicised: ‘We can be sure that news of these private austerities was not kept from the wider Cordoban public, any more than it has been kept from us’.1 We should not undervalue al-Manṣūr’s personal piety, but his more dramatic acts – in particular the purge of inappropriate and quasi-heretical works from al-Ḥakam II’s library – had a public significance, undertaken to secure the approbation of the strictest and most orthodox of the Maliki religious scholars in al-Andalus, whose support he needed to legitimise his adopted role as the State’s religious champion, given the incapacity of the caliph himself to fulfil this role. This same motivation caused al-Manṣūr to lead the Umayyad armies in jihad against those deemed enemies of the State, whether in Christian Iberia or the Fatimid Maghrib. However, his single greatest and most conspicuous act of public piety was his massive extension to the Great Mosque of Cordoba (Figures 42, 49).
Phased plan of the Cordoba Mosque, indicating the different construction phases
© Matilde GrimaldiWhile we possess textual and archaeological evidence for a handful of al-Manṣūr’s public architectural projects – especially road and bridge-building, to facilitate the movement of the Umayyad army, as discussed in Chapter 4 – none of these had the symbolic status of the Umayyad dynastic monument: the Great Mosque of Cordoba. The massive extension he built here, with its attendant infrastructure, is the only example of al-Manṣūr’s public works to remain largely in its original form and, unlike his palace-city, can be discussed from standing remains. However, it has not been subject to anything like the attention paid to the spectacular extension added by the caliph al-Ḥakam II between 961 and 965. As Pedro Marfil observed, ‘the ʿĀmirid extension has been unjustly undervalued, having been considered by many authors as a second rank achievement and not comparable to the previous constructions, in not providing the decorative profusion or artistic boasting of al-Ḥakam II’.2 Studies of the Cordoba mosque have concentrated on al-Ḥakam’s section, and have considered al-Manṣūr’s extension in its light, finding it artistically wanting. In publications on the Cordoba mosque, there is usually no more than a short paragraph on the ʿĀmirid extension, which is surprising given that it is by far the largest addition to the mosque and the only one to be built by a non-Umayyad patron.3 This lack of interest in al-Manṣūr’s extension stems directly from the verdict passed on it by both Gómez-Moreno and Torres Balbás in the 1950s, which has been upheld, without question and almost without exception, by those who have written on the mosque since then: ‘the new work copied exactly the arrangement of that of al-Ḥakam’,4 or was a ‘slavish copy’ of Cordoba III, ‘with hardly anything new artistically’; its decoration was ‘uniform and monotonous, … repeated without the richness of designs and carving of that of the second of the caliphs’.5 On the other hand, Marfil sees it as ‘the definitive crystallisation of the architectonic forms of the Cordoban mosque, contributing in a resounding way to the monumentality of the mosque’.6
Al-Ḥakam’s extension utterly changed the aesthetic of the Cordoba mosque, and at the time was highly controversial. He contemplated changing the position of the qibla ‘because it was oriented too far to the west, but he forsook this task, since the people (ʿāmmat al-nās) were scandalised by the idea of breaking with ancestral practice’.7 As Rius points out, the cohesion of the umma was considered more important than the accuracy of the qibla.8 Another controversy was generated by the expense of the project: the judge Mundhir ibn Saʿīd had to swear a legal oath with witnesses in front of the qibla in order to overcome the scruples of the people who refused to pray in al-Ḥakam II’s extension because they doubted where the money had come from which paid for it.9
Nevertheless, today this extension has come to epitomise the monument: the decoration, of the maqṣūra especially, has become a metaphor for the high level of artistic and cultural excellence attained during al-Ḥakam’s reign. One should, however, be careful never to judge a mosque by its maqṣūra: this section of any mosque will always be more elaborately decorated, because here religious, ceremonial, political and symbolic messages are at their most meaningful. While al-Ḥakam’s maqṣūra is unique, the rest of his extension follows exactly the arrangement of both Cordoba I and II, which were also the models for al-Manṣūr’s extension.
To understand the potency of the Cordoba Mosque as a monument, and the ways in which construction at this site has long been interpreted in terms of political symbolism, we need to introduce the building’s pre-ʿĀmirid architectural history. But its postscript is also enlightening. The Christian conquest of Cordoba in 1236 was immediately commemorated by the consecration inside the mosque of the first cathedral of Santa María.10 While this changed the fabric of the original building very little, the Baroque cathedral built between 1523 and 1607 was a huge, interventive construction, implanted mostly in al-Manṣūr’s extension; the cathedral’s expanse, high dome and clearly-defined nave ‘proclaim its Christian identity to all who view the mosque from without’.11 In addition, ʿAbd al-Raḥmān III’s minaret was encased in a new salient, and thus the most visible symbol of Muslim, indeed Umayyad, rule in al-Andalus, which had overlooked Cordoba for centuries, was ‘christened’ as a bell tower.12 The superimposition of the Baroque cathedral made a loud political statement of Christian superiority over Islam in the Iberian Peninsula, and demonstrated that the significance of construction at this site was still as potent three centuries after the Castilian conquest as it was when the first addition to Cordoba I was conceived. It is not far-fetched to consider that modern scholars of the mosque have projected some of the issues raised by the Christian construction – of appropriation and usurpation – backwards on to al-Manṣūr’s extension, whose motivation has consequently been understood in similar terms.
1 The Pre-ʿĀmirid Mosque
The congregational Mosque of Cordoba was the dynastic and ceremonial heart of Umayyad rule in al-Andalus. By the mid-tenth century, it had become the physical symbol of the Umayyad caliphs as the legitimate rulers of Islam, in opposition to the claims made by the Abbasids and Fatimids. From the beginning, the mosque’s architectonic forms and decoration were symbolic of Umayyad rule in al-Andalus, as emphasised by foundation myths which are associated with construction projects at the site by the Arabic sources, and which explicitly relate Cordoba to that Umayyad dynastic mosque par excellence, al-Walīd’s mosque in Damascus.13 As a consequence, its distinctive repeat unit, of double-tiers of horseshoe arches (Figure 43), was imitated and perpetuated by all later builders at the monument, no doubt out of ‘a respect for tradition and continuity’,14 but also because of their architectural (what Flood styles ‘architextual’15) allusions to the key Umayyad monuments of Bilād al-Shām, which showed ʿAbd al-Raḥmān I’s concern for ‘the authority provided him by virtue of his lineage’.16
Vista through Cordoba IV, showing painted stone voussoirs and different coloured marbles of the columns (section 2.1.3.3)
© Mariam Rosser-OwenIt is no coincidence that every Umayyad ruler of al-Andalus – with the single exception of al-Ḥakam I (r. 796–818), whose entire reign was occupied with quelling revolts and fighting the Christians – celebrated his rule by engaging in construction at this building.17 In its time, each addition would have been an important political statement as well as a visible symbol of rule and the aspirations of the ruler. However, their original meanings are now eclipsed by the architectural interventions of the first two caliphs. In 951, ʿAbd al-Raḥmān III constructed a huge new minaret, 73 cubits (34 metres) high, surmounted by a domed pavilion, and containing two independent staircases which spiralled round each other.18 This minaret would have been the tallest building in Cordoba, visible for miles around, indicating the mosque ‘like a beacon’;19 furthermore, in so dramatically embracing the minaret as a form, it had a clear polemical meaning, asserting the legitimacy of Umayyad rule vis-à-vis the Fatimids, who for doctrinal reasons had rejected the minaret in favour of the monumental bāb.20 The Cordoban minaret would thus have ‘[called] the faithful to [al-Nāṣir’s] caliphate rather than any other’.21 A few years later, in 956, he commissioned a matching pair to this minaret at the al-Qarawiyyīn Mosque in Fez, which still survives in its original form to this day (Figure 44); again the symbolism of this architectural gesture was potent, claiming Fez for the Umayyads in the face of Fatimid expansionism (on which see Chapter 1).22
Minaret at the al-Qarawiyyīn Mosque, Fez, 956
© Péter Tamás NagyMost spectacular, however, in terms of its decoration and Umayyad symbolism was the enormous extension (Cordoba III) which al-Ḥakam II began on the very day he acceded to the caliphate in 961 (Figures 42, 46).23 He extended the mosque to the south by 12 bays, necessitating the destruction of the mihrab zone built a century earlier by ʿAbd al-Raḥmān II, and elongating the sanctuary to a total area of 9,087 m2. The double qibla wall contained the bayt al-māl to the east (still today the site of the Cathedral treasury) and, to the west, the corridor leading to the sābāṭ and caliphal palace, preceded by a chamber where the caliph received homage. The mihrab at the centre took the form of a hexagonal chamber, though it appears to have imitated the form of ʿAbd al-Raḥmān II’s mihrab, and we know from Ibn ʿIdhārī that al-Ḥakam ordered the four columns which ‘supported the entrance arch’ of the earlier mihrab to be carefully removed and reused in his own.24 It functioned as a sound-box to carry the imam’s voice into the mosque behind, but also, metaphorically, as a ‘gate opening towards the invisible or the unknown’.25
Al-Ḥakam’s extension was clearly demarcated by more intensely-articulated decoration, in particular of the three central naves, which constituted the new prayer hall’s maqṣūra. This space was also physically enclosed by a wooden screen, described by al-Ḥimyarī as ‘a balustrade of wood artistically carved with different decorative motifs’, and which, according to Ibn ʿIdhārī, measured 75 cubits long and 42 cubits wide. Concepción Abad has plotted these dimensions on a groundplan of al-Ḥakam’s extension, showing that they perfectly encompass the three central naves from the qibla wall all the way down to and including the space now called the Capilla de Villaviciosa (Figure 42).26 The height of this wooden enclosure is likely to have reached the same level as the capitals, though it is not known whether it was solid or pierced. Perhaps, like the early eleventh-century wooden maqṣūra that still survives at the Great Mosque of Qayrawān, it was both, with a solid, elaborately carved framework, filled with smaller turned-wood windows.27 The Cordoba maqṣūra also appears to have had a door, presumably located at the northern end (communicating with the mosque), since the caliph and his court habitually came and went via the Bāb al-Sābāṭ, located to the right of the mihrab. However, after the Fitna, in 1010, ‘the people of Cordoba broke the door of the maqṣūra, took it and sold it to the Franks [probably the Catalans]’, as mentioned by Ibn ʿIdhārī.28
The entrance to this privileged zone was marked by a particularly elaborate twenty-one lobed arch (Figure 45), leading into what is now the Capilla de Villaviciosa, topped by one of the elaborate domes that were constructed in this new maqṣūra. The rectangular dome that crowns this space is frequently referred to in contemporary literature as the ‘lucernario’, because of the sixteen small windows that pierce its upper level, providing much-needed illumination to this deep zone of the mosque’s interior (Figure 47). Ahead and to the east of this space is an elaborate screen formed of interlacing five-lobed arches, supported from below by three further lobed arches (Figure 46). Originally there was probably another of these screens to the west, forming a sort of enclosed vestibule to the maqṣūra, but any trace of what might have existed here in the late tenth century was destroyed at the end of the fifteenth, when the Gothic cathedral was constructed in this zone.29 The bay to the east is now occupied by the Capilla Real, established in 1371. Ruiz Souza has argued that the dome above it is Umayyad, though refurbished and redecorated with muqarnas in the fourteenth century, to conform to the architectural fashions of the day. The western bay would have had a matching dome, so that this entrance zone recalled the tripartite portico of the courtly spaces at Madīnat al-Zahrāʾ, in particular the Hall of ʿAbd al-Raḥmān III. However, others have noted that the physicality of the architecture in this part of the mosque does not support this view: the size, height and number of the columns in this zone indicate support for one central dome, the so-called ‘lucernario’ (Figure 47).30 Abad and González have recently revisited the eastern bay in this zone, now occupied by the Capilla Real, and have interpreted several interventions here as dating from the ʿĀmirid period:31 this will be discussed in detail in section 4 below.
View of 21-lobed arch at entrance to maqṣūra (now Capilla de Villaviciosa), with 11-lobed arch flanking entrance on north side
© Mariam Rosser-OwenView of al-Ḥakam’s maqṣūra, from the screen of interlaced arches (Capilla de Villaviciosa) towards the mihrab
© Mariam Rosser-OwenView up into Capilla de Villaviciosa dome (the so-called ‘lucernario’)
© Glaire AndersonNevertheless, the organisation of the maqṣūra in three bays reinterprets the palatine space of Madīnat al-Zahrāʾ inside the mosque, so that ‘the space where al-Ḥakam and his court would arrange themselves in the Great Mosque deliberately recalled the great reception hall of the caliphal city founded by his father’ (Figure 42).32 The final structure marking the entrance to al-Ḥakam’s maqṣūra ‘does not just mark the transition from one constructive phase to another, but from the amiral mosque to the caliphal’.33 The windows above allowed a curtain of light to pour down amidst the gloom of the now extremely deep mosque, and from the naves constructed by ʿAbd al-Raḥmān I and II, it would have been astonishing to observe this ‘luminous façade’ which announced the beginning of the exuberant extension.
At the opposite end of the three central naves opened by this internal façade, three other elaborate domes demarcate another vestibule which projected in front of the mihrab. These domes are more complex than those at the northern end of the maqṣūra, as they transition from square plan to octagonal dome.34 The central dome, immediately in front of the mihrab entrance, is beautifully incrusted with gold mosaic, which al-Ḥakam II is said to have been gifted along with a mosaicist by the Byzantine Emperor. The unique – and, in al-Andalus, unprecedented35 – use of mosaic decoration in the immediate vicinity of the mihrab makes a clear visual allusion to the decoration of key Umayyad monuments: the Dome of the Rock and Al-Aqṣāʾ in Jerusalem, al-Walīd’s mosque in Damascus, the Prophet’s Mosque in Madīna – all of which are known to have had decoration in their key zones in the Byzantine technique of glass mosaic.36 In terms of meaning, it seems likely that the entirely vegetal iconography of the mosaic friezes in Cordoba’s mihrab chamber and ante-mihrab dome evoke Paradise, which is how the iconography of the mosaic decoration of other Umayyad mosques has been interpreted.37 However, the style and content of the mosaics in Cordoba are rather different, since the architects of Cordoba III adapted the characteristic Andalusi decorative form of carved marble to this new medium, one to which its sinuous forms were not best suited.38 The fact that such a decorative adaptation was contrived at all underlines the pan-Umayyad significance of the use of glass mosaic as a material, and gives emphasis to the messages of continuity and legitimacy which Cordoba III embodies.39
Above the mihrab is a fifth dome, in the form of a scallop-shell.40 Great care was taken over the hierarchisation and planning of this space, and the pink and black marble columns used throughout this extension alternate in a carefully-articulated and rhythmical pattern.41 This use of Late Antique spolia, the copying by Andalusi carvers of Roman capitals, and the employment of Byzantine mosaics creates a Classicising aesthetic in Andalusi art that seems to be a way in which the Umayyads distinguished themselves artistically from the Fatimids.42 These messages are reinforced in the inscription programme used throughout the mosque, both in key locations of the interior as well as prominent places on the exterior, which assert the Sunni and especially Maliki orthodoxy espoused by the Andalusi Umayyads, against any form of heterodoxy. We will return to the selection and meaning of the Qurʾānic passages which were employed in the maqṣūra and elsewhere in section 3 of this chapter.
When ʿAbd al-Raḥmān III began to stake the Umayyads’ claim to be the only legitimate caliphs of Islam, the mosque’s statements of dynastic continuity assumed new resonances: as Safran comments, ‘The monument came to symbolise the continuity of Umayyad rule … The expansion and elaboration of the mosque celebrated the success of Umayyad promotion of Islam on the Peninsula … The mosque’s architectural and decorative references to Syrian Umayyad monuments identified Cordoba as an Umayyad city and capital’.43
The issues raised here defined the Great Mosque of Cordoba as a monument loaded with subtle and not-so-subtle visual messages – located especially in what is now the ceremonial heart of the building, al-Hakam’s maqṣūra – which emphasise its Umayyad-ness by association with other Umayyad monuments, and make clear statements of the dominance and permanence of Umayyad rule in al-Andalus. The disinterest of modern scholars in al-Manṣūr’s extension seems to reflect their disappointment that he did not express his own political aspirations in visual messages as clear as those of the Umayyad and Christian phases. Hence Cordoba IV is criticised as a ‘slavish copy’, the implication being ‘if only al-Manṣūr had made his political statement more readable’. However, as we shall discuss below, his statement is readable, but modern scholars have been looking for the wrong message in the wrong place.
2 The ʿĀmirid Mosque (Figure 49)
Robert Hillenbrand describes the Cordoba mosque as a ‘palimpsest’ of its successive building campaigns,44 and nowhere is this more true than in al-Manṣūr’s extension (Figure 43). It adjoins every earlier section of the mosque, and it was primarily in this section that the Christian additions were imposed, likely because it provided the largest surface area of repeat units and was decoratively less varied than other parts of the building. Apart from Pedro Marfil’s brief chapter on the ʿĀmirid intervention,45 this part of the mosque has never been subject to a full and detailed discussion, until now.
The extension was begun in 377/987,46 and probably completed in 384/994, since the Dhikr Bilād al-Andalus tells us that prayer was inaugurated in the space in Rajab 384/August–September 994. The mosque extension thus took seven years to build. The reliability of this information is not certain, however, since the Dhikr Bilād gives the commencement date as Rajab 381/September 991, a date which is not provided by other, earlier sources.47 As Souto noted, this ‘chronological imprecision’ likely derives from the lack of a foundation inscription: al-Manṣūr ‘did not want to leave any inscription commemorating his intervention, he did not allow that his name should be written in stone in the [Mosque]. He was extremely conscious of the importance of such a document and of the insult that it would imply against the Umayyad legitimacy that he claimed to be defending and in support of which claim this construction was clear propaganda. However, neither is there any foundation inscription in the building in the name of the caliph, Hishām II’.48
Inscription naming Hishām II, late tenth century, marble; Instituto Valencia de Don Juan, inv. 5.266, H 40, W 38, D 7 cm
© Cuauhtli GutiérrezA few sources mention that al-Manṣūr’s mosque extension was carried out on Hishām’s behalf,49 and it seems certain that, had a construction inscription survived from this extension, it would have been in the caliph’s name in the same manner as those inscriptions commemorating the restoration of Lisbon’s city walls (Figures 19–20, Appendix 4.6), the construction of bridges at Écija (Figure 21, Appendix 4.8) and Toledo (Figure 22, Appendix 4.9) or the new backrest for the Andalusiyyīn minbar (Figures 5–7, Appendix 4.4). Though Hishām would have been in his 20s by this time, as we saw in Chapter 1 he was likely incapable of overseeing such a construction himself. There is only one surviving inscription which names Hishām in the role of active commissioner: this is the small and fragmentary foundation inscription now in the collections of the Instituto de Valencia de Don Juan in Madrid (Figure 48, Appendix 4.3).50 It only partially preserves the beginning of his laqab al-Muʾayyad bi-llāh, though it clearly reads the active verb amara (at the start of the penultimate line) and the phrase amīr al-muʾminīn. No information survives about the monument ordered or its date.51 The preserved wording does not fit closely with the standard formulae for constructional inscriptions under the first two Andalusi caliphs (see the semantic discussion in Chapter 8). According to Lévi-Provençal’s reconstruction, the third line contains part of a pious invocation, from which it could be argued that the foundation to which this fragment belonged was a pious work, such as a fountain or a mosque – could this be the missing foundation inscription from al-Manṣūr’s mosque extension? However, in building his extension to the Córdoba mosque al-Manṣūr would surely have proclaimed his own role in the undertaking, albeit while acting in the caliph’s name (as on the Andalusiyyīn minbar inscription), and there is no mention here of the ḥājib. Hishām’s role is expressed in the active not the passive tense. This epigraphic fragment thus offers a tantalising glimpse of Hishām’s engagement in the patronage of pious works, of the kinds of architectural foundations which a caliph would normally be expected to undertake. Perhaps he was not so closeted from the operations of the court and its artisans as has long been assumed. However, to date, no other inscriptions have been found in his name at Madīnat al-Zahrāʾ, which implies that unlike his energetic father, Hishām commissioned no additions or refurbishments at the palace throughout his long residence there.52
Plan of the Cordoba Mosque showing the ʿĀmirid interventions
© Matilde GrimaldiIf construction began in 987, this was only twelve years after the completion of al-Ḥakam’s extension. The primary motivation for the expansion given in the sources was the need for space due to the swelling population.53 This is something of a topos in the accounts of all the mosque’s extensions, but a significant level of demographic growth at this period is not unlikely,54 and may well have influenced the decisions to enlarge the Friday mosque. Immigration may also have been an important factor in population growth, especially under al-Manṣūr who, from the late 970s, actively encouraged Berbers displaced by the Umayyad war with the Fatimids to cross the Straits and join the Umayyad army (see Chapter 1).55 Berber soldiers continued to cross into al-Andalus throughout the ʿĀmirid period. Many of them were enrolled in al-Manṣūr’s army and became personally loyal to him, which probably contributed significantly to the increased efficiency of the Andalusi army under his command. Did these thousands of ʿĀmirid troops attend the mosque on Fridays, filling the new extension?
The construction was likely paid for by ‘part of the riches which al-Manṣūr had garnered through his campaigns against the Christians and in the Maghrib’.56 Direction of the works was entrusted to ʿAbd Allāh ibn Saʿīd ibn Batri (d. 1010), who was ṣāḥib al-shurṭa of Cordoba, but about whom very little else is known.57 Something more can be said about the labour force who produced the new stone supports in the mosque extension, thanks to the more than seven hundred signatures and masons’ marks have been recorded on columns, bases and capitals in Cordoba III and IV (Figure 50).58 Their presence was interpreted by Ocaña as representing the presence of Christian workers in the mosque, basing this idea on a passage in al-Maqqarī in which he mentions the employment of chain-gangs of prisoners-of-war in the construction of al-Manṣūr’s extension.59 Ocaña’s ‘extremely suggestive theory’ was that the most common Arabic names found in the monument (Masʿūd, Mubārak, Naṣr) have ‘exact Latin equivalents in Felix, Benedict and Victor … which leads one to presume that [the names] belonged to Christian craftsmen’. His argument was compounded by ‘certain undoubtedly Christian signs’ among the masons’ marks: the Greek letter Thau, the Anchor, the Ship of Fishermen, the Morning Star and the Mustard Seed. Were these secret declarations of Christian faith and identity? Moreover, on one column in Cordoba IV, the name Naṣr occurs on the same column as a Thau symbol.60
Selection of signatures in the ʿĀmirid extension to the Great Mosque of Cordoba: Aflaḥ, Fatḥ, Masʿūd, Naṣr, ʿĀmir, Khalaf al-ʿĀmirī
© Mariam Rosser-OwenThis theory was debunked by Juan Souto in one of his last publications.61 What he called ‘marcas de identidad’ had a practical function, to identify the author of a particular piece of stone carving (p. 36); ‘they had no symbolic character, they had no intention or implication beyond identifying their authors, with a practical purpose within the process of constructing a building’ (p. 58); ‘I firmly believe that before pronouncing on the religious confession of an individual who is dispossessed of any references that are unequivocally religious, it is necessary to investigate the documentation and refer it to the context to which it belongs’ (p. 46); ‘I am convinced that if the authors of these marks had wanted to represent crosses, they would have carved them’ (p. 49). Finally, ‘it is obvious that the signatures in the mosque could have belonged to some of those individuals of a certain rank who acted at the time as ‘supervisor’, ‘chief’ or some other office linked to the carving of the pieces’.62
In fact, this large number of signatures comprises only twenty-five names,63 of which thirteen occur as signatures in contexts outside the mosque, including on objects.64 The repetition of such a small number of names on such a large number of architectonic elements argues for the production of those elements within a regulated workshop which was part of the Dār al-Ṣināʿa, as we will discuss in Chapter 6.
Cordoba IV was by far the largest of the three extensions to ʿAbd al-Raḥmān I’s mosque. Al-Manṣūr extended the mosque to the east along its entire length, including the courtyard, in which he installed a massive new cistern. He added to the prayer hall eight aisles, which are on average narrower than the original aisles,65 and widened the building by nearly 50 m.66 It increased the surface area of the prayer hall by more than 5,300 m2, or 8,167 m2 including the courtyard. Through this work the mosque acquired its definitive dimensions: a total surface area of 22,400 m2, and the capacity of some 40,000 Muslims at prayer. It thus became the third largest mosque in the medieval Islamic world, only surpassed by the two Samarra mosques.67 As Ana Echevarría points out, the extended mosque thus also surpassed any of the Fatimid mosques.68
The sources tell us that east was the only direction in which al-Manṣūr could build, since further south lay the river, to the west was the caliphal palace, and to the north was ʿAbd al-Raḥmān III’s minaret. To the east were only houses, and Ibn ʿIdhārī reports that al-Manṣūr amply compensated the owners in order to acquire their land.69 However, in building towards the east, al-Manṣūr was also establishing a geographical and physical relationship between his mosque and the palace-city he had established a decade earlier. His ‘colonisation’ of this side of the city – as opposed to the west, which contained both the Cordoban qaṣr and Madīnat al-Zahrāʾ – further indicated that this was an ʿĀmirid, rather than an Umayyad, mosque.
As Marfil notes, al-Manṣūr’s builders faced the challenge of setting out eight new naves over a built-up space, whose ground level was descending towards the river: in the northern zone, this was approximately 2 m below the mosque’s interior level, dropping to more than 3 m below in the southern zone. To maintain a single level in the new interior, the builders laid a solid foundation of masonry reused from the demolition of parts of the old eastern façade (which had to be penetrated in order to connect the new extension with the old mosque – this will be discussed further in Section 2.1.3.5) and the lateral platform and street: excavations in this zone have revealed the original street that ran alongside al-Ḥakam’s mosque extension.70 The space was infilled by river pebbles and earth, on top of which courses of masonry blocks were laid until they reached ground level. His builders thus constructed an enormous artificial platform on which to erect the mosque extension, and a wide terrace flanked the new eastern façade, allowing access to the mosque from the new street that was also constructed alongside.
The covered prayer hall thus became almost a square (114.6 m N-S × 128.41 m E-W), though this had the effect of decentralising the mosque’s old axis, along which lay the minaret, the main entrance gate (now the ‘Puerta de las Palmas’) adorned with the eighth-century foundation inscription,71 and al-Ḥakam’s maqṣūra – all key elements in the mosque’s Umayyad identity. One of the main criticisms expressed by modern historians is that ‘al-Manṣūr destroyed [the mosque’s] symmetry’,72 though to maintain it would have meant the destruction or subordination of these focal elements, in order to create a new axis. Echevarría considers that the ʿĀmirid extension displaced the mihrab and maqṣūra, ‘leaving it semi-hidden in a corner’. She notes that whoever leads the community in prayer would need to place himself in a central position along the qibla wall, thus ‘leaving the caliph in a lateral position, as corresponded with Almanzor’s vision’.73 However, there is no indication of any feature that could be defined as a subsidiary mihrab, either in Cordoba IV or along what would have been the mosque’s new central axis, so we must assume that the main axis of the mosque remained as it had been, and that it was the congregation assembled in Cordoba IV who instead found themselves in ‘a lateral position’. The construction of an even number of aisles in Cordoba IV precluded any one of them being privileged as a central nave, and none of the aisles is articulated through greater width or more elaborate decorative features as more important than the other seven. This contrasts with the articulation of the central nave of the western part of the mosque through its greater width (7.85 m) in relation to the others (an average width of 7.2 m), and the concentration of decorative elements in the maqṣūra area of Cordoba III. The lack of such features indicates that Cordoba IV was not meant to displace the original axis of the mosque. Indeed, as Pedro Marfil has pointed out, the extension along the eastern flank allowed the very elongated mosque to regain proportion, and improved problems with illumination, as we will discuss below.74 Nevertheless, there are indications of possible ʿĀmirid interventions in al-Ḥakam’s maqṣūra, which indicate the continued importance of this zone during the ʿĀmirid period, as will be discussed below.
Extending the mosque to the east required the destruction of the mosque’s original eastern façade and its gates, constructed only a few years earlier by al-Ḥakam. Al-Manṣūr pierced eleven large arched holes in this wall, to communicate between Cordoba IV and the older mosque (Figure 51). The traces of the pre-existing façade were blocked up and covered over, though Pedro Marfil believes that the Treasury gate (popularly known today as the Puerta del Chocolate) remained visible (Figure 52).75 What was left of the original decoration of al-Ḥakam’s gates remained hidden until the restoration work by Velázquez Bosco in the early twentieth century, and these remains are now visible in the mosque today. Their traces, though sometimes ghostly, can be related to the seven new gates of al-Manṣūr’s eastern façade with interesting conclusions, as will be discussed below.
View along the communicating wall between Cordoba III and IV, showing al-Ḥakam’s original eastern façade
© Mariam Rosser-OwenThe mosque’s north and south walls were also extended, and the courtyard was proportionally enlarged and refurbished. Certain other ʿĀmirid additions to the mosque complex were entirely practical, and reformed the mosque’s infrastructure. The first of these was the construction of a large subterranean cistern, abutting the foundations of the original eastern façade of Cordoba I.76 This cistern presumably fed the new ablution pavilions (on which see below), and any new fountains which might have decorated the courtyard, as well as irrigating the trees.77 It is a square enclosure, measuring 14.5m a side and 5m deep, divided into nine equal compartments by cruciform pillars, supporting round arches below a vaulted ceiling, in which are pierced three openings. All the surfaces are covered by a hydraulic coating of ochre.78
It is curious to note that the cistern was built on a nine-bay plan, a plan sometimes used for the construction of mosques and churches across a widely dispersed geographical area. One of the most famous is the Mosque of Bāb al-Mardūm in Toledo, built only a few years after the ʿĀmirid extension, in 390/999–1000.79 It is still enigmatic in the sense that it is not clear whether there was architectural precedent for this mosque within al-Andalus, though the scholars who have studied it tend to concur that there must already have been a mosque of this form in Cordoba. The nine-bay concept was certainly already present in Cordoba as it was used in the ʿĀmirid cistern. Could this be one of the earliest uses of this form in al-Andalus? Gómez-Moreno noted that the vaulted arrangement of the ceiling of the ʿĀmirid cistern was the same as in a cistern found at Madīnat al-Zahrāʾ.80 But according to Vallejo, ‘the palace did not have cisterns or subterranean tanks for storing rainwater’;81 rather, the pattern at Madīnat al-Zahrāʾ was one of open reservoirs which evacuated directly into a network of drains, qanats and aqueducts.82 Monumental, vaulted cisterns are not in themselves unusual: one of the best-preserved in Iberia is in the city of Cáceres, on the western frontier of al-Andalus. This huge cistern measures 14 by 10 metres, and had a capacity of 700m3 when full. It is arranged in five naves separated by twelve columns supporting horseshoe arches, forming 20 bays covered by barrel vaults. The date of the cistern is debated, but is likely to be in the twelfth century, when Cáceres was heavily fortified under Almohad rule.83 It is thus unlikely to be a precedent for the ʿĀmirid cistern at Cordoba. Much earlier precedents existed in the Islamic East, such as the cistern constructed in Ramla in 789, by order of the Abbasid caliph Hārūn al-Rashīd. This had 24 bays, and according to Creswell manifests the ‘earliest known systematic and exclusive employment of the free-standing pointed arch’.84 With the limited state of knowledge – pending the excavation and publication of more medieval cisterns – it is difficult to say whether al-Manṣūr’s architects were innovating in their construction of this cistern, or working within a model that had already been established in al-Andalus.85
Archaeological investigations undertaken in 1996 indicated that part of the courtyard had been paved with slabs of purple-coloured limestone, and also located the remains of a large drain that diverted the waters towards the new eastern street. Marfil also notes that new gates were opened in the courtyard’s extended perimeter wall: one of these is the gate known today as the Puerta de Santa Catalina, now a large Baroque gate that opens into the patio close to the entrance to the mosque’s new wing (Figure 67: Gate 8 on the East side). Another small gate was opened on the northern wall, abutting the precinct that had been constructed by ʿAbd al-Raḥmān III (Figure 67: Gate 2 on the North side).86
Another important aspect of al-Manṣūr’s refurbished infrastructure for the mosque was only discovered during rescue excavations in 1998, in the street that runs parallel to the mosque’s eastern façade.87 The remains of an ablutions pavilion, or mīḍaʾa, were found, orientated at a slightly oblique angle to the mosque’s façade but undoubtedly related to Cordoba IV.88 The mosque’s first mīḍaʾa had been built to the east of Cordoba I by Hishām I, but was destroyed during al-Ḥakam’s construction of a new eastern façade.89 Al-Ḥakam himself constructed four new ablution halls, ‘two large ones for the men and two smaller ones for the women’, one of each on the east and west sides of the mosque.90 Al-Manṣūr’s extension would have necessitated the destruction of the two eastern ablution halls, and he built at least one huge replacement (plotted on Figure 49). However, the Dhikr Bilād al-Andalus records his construction of ‘three hydraulic complexes dedicated to the ritual ablutions (wuḍūʾ)’,91 to the east, west and north of the Mosque: ‘each one of these complexes (sāqiyāt) had 20 cubicles (buyūt) for private use (taṣarruf) and a patio (ṣaḥn) for ablutions with a pond (ṣahrīj), at the centre of which was a fountain (fawārra) of clear water’. The text dates their completion to 390/999–1000.92 Though there would have been no practical need to reconstruct the western and northern ablution halls, which would have been unaffected by the construction of Cordoba IV, al-Manṣūr may have refurbished them for the sake of symmetry with his new eastern pavilion.93
The excavated foundations of this mīḍaʾa reveal a rectangular plan,94 of dimensions c. 16 m wide × 28.23 m long (until the foundations disappear beneath the Hotel Conquistador). It was a freestanding structure, of a ‘monumental quality’, with roads on all four sides. Its foundation walls are extraordinarily wide, varying between 1.5 and 2.25 m. The most outstanding feature is its network of water channels measuring 50–60 cm wide × 90 cm high, indicating a high volume of water flow. These channels run around the perimeter of the structure’s interior, and are supplemented by smaller transverse channels. The walls were rendered impermeable by the usual hydraulic coating. The upper structure seems, from the remains and the textual description, to have comprised 20 latrine cubicles around the perimeter, with a central paved area in which there was a pond and stone fountain, presumably intended for the ablutions.95 Though we have no idea what it would have looked like, Montejo comments that the decoration would probably have had ‘the same characteristics of solidity and permanence observable in [al-Manṣūr’s] extension to the mosque’.96
This mīḍaʾa was obviously a massive and expensive structure, all the more so if al-Manṣūr built two others like it. Together with the huge new cistern under the courtyard, we can understand the building campaign as ‘not an isolated monument but … integrated into an architectural and urban complex’97 – a complex whose Umayyad origins were respected and enhanced under ʿĀmirid patronage. Though the construction of such features in al-Manṣūr’s extension stem from a practical need to reform the mosque’s associated infrastructure, they were also significant pious works, and the kind of necessary institution which rulers were expected to provide for their subjects. Indeed, the supply of life-giving water is a topos associated with the good ruler, as is shown by the following qaṣīda which Ibn Shukhayṣ composed in celebration of al-Ḥakam’s reform of the mosque’s water infrastructure in 967, when he refurbished a Roman aqueduct to bring water to the mosque from the Sierra Morena:98
“You have torn the sides off the mountain in search of water springs of great purity, which you have brought to the house of worship in order to clean bodies of their impurities and to ease the thirst of men. At the same time you have done a rare thing, a glorious act and a work of piety for the people, whose pastor and protector you are.”99
We will return in Chapter 8 to the poetic image of the ruler’s role in bringing life-giving water to his subjects, however it should be noted here that panegyrics composed in honour of al-Manṣūr likewise praised him in terms of water metaphors. His reforms of the water infrastructure of Cordoba IV, indeed the ʿĀmirid mosque extension as a whole, were meritorious public works, and in commissioning them al-Manṣūr was, at a fundamental level, fulfilling his duty as de facto ruler.100
2.1 Analysis of the Interior
Turning to the interior of Cordoba IV, its internal decoration has been called ‘uniform and monotonous’, ‘a slavish copy’ of Cordoba III with ‘hardly anything new artistically’; in the prayer hall, ‘there isn’t a single innovation’.101 The critical neglect of al-Manṣūr’s extension has perpetuated such views, which suppose that there were no artistically innovative features in Cordoba IV. On closer inspection, however, such views are unfounded, since there are a number of decorative differences and technical innovations that distinguish this campaign from earlier phases. For the most part, these are ingenious practical solutions to problems posed by or lessons learned from the pre-existing structure, though in a few cases these features are purely decorative, possibly representing the evolution of certain forms. In combination, these features indicate a more general theme, which is that no expense was spared in making Cordoba IV look just like the rest of the mosque. As Christian Ewert conceded, a distinctive element of al-Manṣūr’s extension was its adaptation to the pre-existing construction, even the subordination of it in order to highlight even more the extraordinary quality of the caliphal works.102
In the following sections we will discuss:
-
those differences in the decoration of Cordoba IV that can be explained through the need for practical solutions to existing construction issues;
-
architectural differences that seem purely decorative, and can be read as ostentatiously costly;
-
architectural features that were imitated from the earlier mosque;
-
evidence for ʿĀmirid intervention into al-Ḥakam’s maqṣūra.
In section 3 of this chapter, we discuss the Qurʾānic inscriptions that are employed in the mosque, especially in Cordoba III and IV, and on the new eastern façade of al-Manṣūr’s extension.
2.1.1 ‘Decorative Differences’ That Can be Explained through the Need for Practical Solutions to Existing Construction Issues (See plan in Figure 49)
2.1.1.1 Double Qibla Wall Not Continued
The double qibla wall of Cordoba III was not continued into Cordoba IV, probably because the Bayt al-Māl stayed where it was. This would explain why its gate – the so-called Puerta del Chocolate, originally an external gate of al-Ḥakam’s eastern façade – has been so well preserved (Figure 52). As noted above, Pedro Marfil believes that this gate remained visible in Cordoba IV,103 presumably providing a secondary point of access to the Bayt al-Māl. The naves of al-Manṣūr’s extension are therefore two bays deeper than in the western portion of the mosque.
The perpendicular arcades thus run all the way to the exterior wall at the south, and large windows pierced high up in the walls allow this zone farthest from the courtyard to be illuminated (Figure 53).104 This solved one of the problems created by al-Ḥakam’s extension, which had made the prayer hall excessively long. Not only was the new maqṣūra area far from the courtyard arches, which illuminated as well as ventilated the mosque, but the double qibla wall with ancillary spaces behind (the sābāṭ, chamber for the minbar, mihrab, and the Bayt al-Māl) had no communication with the exterior wall and thus no illumination, apart from the small windows which surrounded the maqṣūra’s domes. As discussed above, Juan Carlos Ruiz Souza has argued that the windows piercing the dome above the entrance space to al-Ḥakam’s maqṣūra (Figures 46–47) created a symbolic curtain of light that penetrated the shade of the mosque’s interior, thus drawing attention to the monument’s most potent space.105
Window grille in situ in Cordoba IV
© Mariam Rosser-OwenHowever, the tabula rasa of the ʿĀmirid wing allowed a chance to resolve these illumination problems. These new windows were also filled with large marble window grilles, carved with elegant geometric designs, each of them different (Figure 54). Similar grilles are also seen flanking the gates of al-Manṣūr’s eastern façade (for example, Figure 71). While no windows were necessary on the southern façade of al-Ḥakam’s extension, window grilles were used in the lights that punctuated the domes of the maqṣūra: in 1998, one extant original grille was found in the dome over the Capilla de Villaviciosa, which is pierced by sixteen windows. In the window situated most to the west of the dome’s northern side, the grille consists of a piece of marble with a design of intersecting lobed arches, imitating the elaborate arcaded ‘screens’ inside the Capilla de Villaviciosa space beneath, its painstaking carving even going so far as to include miniature capitals and column shafts.106 The use of marble window grilles was thus not an ʿĀmirid innovation in the mosque, but the skill with which they are carved, their huge size and the sophistication of their designs speaks to the artistic skill of artisans working under ʿĀmirid patronage, as we will see in Chapters 6 and 7, in relation to other marble objects carved for the ʿĀmirids. It also speaks to the availability of large rectangular sections of marble which could be used for carving these new grilles, and as we know from the ʿĀmirid basins (discussed in Chapter 7) this material was indeed readily available.
Window grille from Cordoba IV, 987–994, marble; Museo Arqueológico y Etnológico, Cordoba, inv. 3.448, 153 × 98 cm
© Mariam Rosser-Owen2.1.1.2 Five-Lobed Arches to Resolve Disparity along the New Courtyard Façade
The north wall, which opens onto the courtyard, is thicker in the western part of the mosque than in Cordoba IV, since when ʿAbd al-Raḥmān III refurbished the courtyard in 346/957–8, he added an outer layer of piers to those which survived from ʿAbd al-Raḥmān I’s original mosque, to provide greater stability to this otherwise weak façade.107 This resulted in a double thickness of piers supporting the northern façade. When al-Manṣūr extended the courtyard façade to the east, he only continued the outermost layer of piers, to maintain a consistent appearance along this façade. This results in a disparity between the depth of the first few bays of Cordoba IV’s north wall, and those of the western part of the mosque. This is compensated for in al-Manṣūr’s extension by the uncharacteristic replacement of the usual horseshoe arch by five-lobed arches in the first two bays of each of the seven naves which begin at the north wall (Figure 55). Only the lower arches are lobed, while the upper arches are semicircular as in the rest of the mosque. After this ‘compensatory zone’, the arches in Cordoba IV maintain parity with arches in Cordoba I, II and III, and the style reverts to the superimposed arrangement of semicircle-above-horseshoe-arch that is the mosque’s standard repeat unit, except in one other location which we will discuss in the next section (2.1.1.3).108
The interior of the courtyard façade of al- Manṣūr’s extension seems originally to have had painted decoration, now hidden behind the lateral chapels of the northern wall. An archaeological survey in this zone revealed the partial survival of these painted schemes, which can be accessed in the elevation of the wall above the low ceiling vaults of the lateral chapels. According to Pedro Marfil, large Arabic inscriptions painted in a red pigment were discovered here, on the interior of the courtyard façade facing into the mosque. These seem to have been framed within long rectangular friezes above the arches that communicated with the courtyard, of the same type that frame the carved inscriptions on the mosque’s eastern façade (see below). While the content of these painted inscriptions is not yet known, it is likely that, like the inscriptions on al-Manṣūr’s eastern façade, they were citations from the Qurʾān that affirmed the ideological messages of the Cordoba Mosque’s pre-existing epigraphic programme.109
2.1.1.3 Five-Lobed and Pointed Arches to Resolve Disparity within the Northern Zone
The use of the five-lobed arch occurs again in al-Manṣūr’s extension, in an area where it is easy to miss.110 When ʿAbd al-Raḥmān II built the first extension to Cordoba I, he incorporated the earlier building’s external buttresses into his extension by using them as the springers for his new arcades.111 In Cordoba IV, the thickness of these buttresses creates a new problem of symmetry. The plan published by Christian Ewert suggests that this problem was solved by the imitation of these buttresses, however the buttresses in Cordoba IV only occur in the westernmost four aisles, in the area which is now occupied by Hernán Ruiz’s cathedral. It is likely that these buttresses are architectural traces of Hishām I’s mīḍaʾa, which is known through excavations to have been located in this area, or that constructed by al-Ḥakam II.112 For the three arcades to the east, al-Manṣūr’s architects resorted to the same solution as at the north wall, which was to compensate for the disparity in width created by these buttresses by inserting three bays with narrower intercolumniations, capped by five-lobed arches (Figure 56).
Cordoba IV: double bays of lobed arches in the compensating zone along the interior of courtyard façade (section 2.1.1.2)
© Mariam Rosser-OwenCordoba IV: five-lobed and pointed arches in the compensating zone around where Hishām I’s mīḍaʾa was located (section 2.1.1.3)
© Mariam Rosser-OwenSome of the arches in this zone are pointed. Herrero Romero suggests that these were added by Hernán Ruiz;113 but these could also date from the tenth century, as there are other examples of the pointed arch in the mosque. Ruiz Cabrero accepted these as dating to the ʿĀmirid period, attributing the narrower intercolumniations here to an error in planning on the part of al-Manṣūr’s architects.114 In discussing the pointed arches used in the transverse arcade that runs parallel to the qibla wall in al-Ḥakam’s maqṣūra (Figure 75, discussed below), Antonio Momplet observes that these might be the earliest known examples of the pointed arch in Andalusi architecture. Following Torres Balbás, he cites a number of examples from the Islamic East, including the palace at Mshatta, the mosques of Samarra, and early Egyptian examples such as the Nilometer and Mosque of Ibn Ṭūlūn; however, in accordance with his own argument about Byzantine structural influences on the Cordoba Mosque, he opts for several Byzantine models, including from Armenia.115 Since the transverse arcade is possibly an ʿĀmirid intervention in the maqṣūra (see 2.1.4.1 below), the introduction of the pointed arch in the Cordoba Mosque may be an innovation of al-Manṣūr’s architects; the idea might have been suggested by the overlapping arches in the frieze above the Puerta del Chocolate, which creates a series of pointed profiles (Figure 52). As we will see, this is not the only location in which unusual arch profiles can be associated with the ʿĀmirid extension.
2.1.1.4 Introduction of Lead Discs
Another innovation in the ʿĀmirid extension was to introduce the use of lead discs between base and column, column and capital, and capital and cyma. This acts as a bond between these rigid elements, as well as serving as a cushion between them and allowing for better damping of the elements in relation to each other. At the same time the column shafts are anchored into the building’s foundation by resting on top of large blocks of marble or limestone, which are in turn supported on a strong bed of lime mortar.116
2.1.2 Architectural Differences in Cordoba IV That Seem Purely Decorative, and Can Be Read as Ostentatiously Costly
2.1.2.1 Stone Construction
Most obviously, the whole extension is built from stone, rather than stone combined with brick as in the western mosque. The arcades are plastered and the voussoirs painted in imitation of the rest of the mosque. There seems no practical need for this, and since the use of stone throughout such a large surface area would undoubtedly have been extremely costly, the most likely explanation is that the expense was intended to be conspicuous. Pedro Marfil calls it a sign of ‘the maturity of caliphal architecture at the end of the tenth century’.117
2.1.2.2 Increasing Complexity of Roll Corbels
Roll corbels (modillions) made from plaster are employed throughout the Cordoba mosque, including along the courtyard façade, above the abacus of every column, in the position where they effectively serve as springers for the upper of the two superimposed arches (Figure 57).118 In the earlier phases of the mosque, these have a simple form: between five and seven horizontal rolls are intersected at the centre by a vertical band, which is usually decorated simply with incised lines. This element of the corbels, however, becomes increasingly decorative in Cordoba IV, and comes to feature quite complex vegetal or geometric designs. In some cases, traces of paint survive (especially red and blue), though these are likely to be from later Christian interventions in the space. In three cases in the ʿĀmirid extension there are also inscribed examples, containing religious phrases carved in relief in simple Kufic script (Figures 58, 67, Appendix 3B).119 A fourth is now blank, its Arabic inscription having been carved away at some point; another bears the shahada, the Muslim declaration of faith; the third contains the verse 65:3, ‘Whosoever puts his trust in God, He is sufficient for him’, which does not occur elsewhere in the Mosque. This may be an example of the phenomenon noted by Calvo, of the use in the mosque’s inscriptions of pious formulae that may have been common in everyday life.120
Cordoba IV: increasingly decorative roll corbels, some with traces of colour (section 2.1.2.2), and the standard capital used throughout the ʿĀmirid extension (section 2.1.3.4)
© Mariam Rosser-OwenCordoba IV: roll corbels with inscriptions the ʿĀmirid extension (section 2.1.2.2, Appendix 3B)
© Mariam Rosser-OwenThese subtly more decorative modillions may be no more than an example of stylistic evolution, though these elements in Cordoba III remain conservatively in the style of the previous two building campaigns. Alternatively, it may be a subtle indicator of the great expense on this extension.
2.1.2.3 New Decorative Elements on the External Gates
We will discuss the decoration of the exterior façades in more detail below, however we should note here what Torres Balbás calls the ‘most important variation’ of the façade of Cordoba IV:121 on the western façade, each gate is flanked by a carved marble window grille, hooded by a blind five-lobed arch, the whole group set in a rectangular panel filled with decoration in carved stone and inlaid brick. The gates on al-Manṣūr’s eastern façade preserve this arrangement but shift it upwards, introducing below it an additional element: a pair of small, blind horseshoe arches, filled with panels of carved vegetal designs (Figure 82). This new element has the effect of enlarging the decorative surface area of each gate and, again, would have been more expensive to produce than the decoration of the western façade gates.
While it is important for a more balanced appreciation of Cordoba IV to enumerate the ways in which it differs from the previous building campaigns, these features contribute to two simple conclusions: firstly, that its architects were highly skilled in their invention of practical solutions to the problems of leftover architectural features from earlier construction phases; and secondly, that the use of increasingly decorative features for their own sake implies that no expense was spared on this extension, and that its patron wished this to be made obvious (albeit subtly). It is also important to examine those features that were imitated from the earlier mosque, since it is in these elements that there resides al-Manṣūr’s own conception of the relationship of his extension to the building’s Umayyad phases.
2.1.3 Architectural Features That Were Imitated from the Earlier Mosque
2.1.3.1 Continuation of Transverse Arcade Running Parallel to Qibla Wall
The feature of continuity that is most obvious when one studies a plan of the mosque is the continuation of the transverse arcade running parallel to the qibla wall twelve (fourteen in Cordoba IV) bays from the south (Figure 49). In the western part of the mosque, this arcade, composed of large double semicircular arches each spanning a nave and supported by two pairs of columns and capitals, marked the junction between Cordoba II and III at the time of al-Ḥakam’s extension. Today this junction is most visible at the entrance to the Gothic cathedral. The same style of arcade is found at the junction between Cordoba I and II. While the appearance of some of this transverse arcade, in the vicinity of the Baroque cathedral, has been altered in later interventions, it is clearly visible in the four easternmost bays of Cordoba IV, especially at the points where it intersects with the perpendicular arcades. This intersection is mediated by means of a thick pier, with the two pairs of columns carrying the transverse arcade and the single columns carrying the perpendicular arcade embedded into it, resulting in a cluster of six columns (Figure 59). These column clusters are still visible at the perimeter of the Baroque cathedral, despite the fact that the piers have been refurbished in a more sixteenth-century style.
View of the transverse arcade running from junction of Cordoba II+III into Cordoba IV (section 2.1.3.1)
© Mariam Rosser-OwenThere was no earlier building at this point in Cordoba IV which necessitated such a transverse arcade, though it does add structural support to the perimeter walls of the new extension. However, Ruiz Cabrero observes that the channel which drained rainwater from the mosque’s roof ran along this section: it was constructed when al-Ḥakam extended the mosque, and had to be extended by al-Manṣūr’s architects, in order to drain the roof. The greater weight of the water channel was thus supported by these much larger arches.122 Nevertheless, the imitation in Cordoba IV of the appearance of these arches created symmetry and physically identified the new extension with the earlier phases of the mosque.
2.1.3.2 Arcade of Large Horseshoes along Longitudinal Wall
The same style of arcade – large double horseshoes supported on pairs of columns and capitals – is found running longitudinally north-south at the junction between Cordoba IV and the mosque’s original eastern exterior wall (Figure 51). As noted above, adding an extension along the mosque’s entire eastern flank necessitated the destruction of its original exterior façade – those entrance gates constructed by ʿAbd al-Raḥmān I, ʿAbd al-Raḥmān II, and, only a few years earlier, by al-Ḥakam II (discussed in 2.1.3.5). Eleven large arched holes were pierced into this wall to communicate between Cordoba IV and the older mosque. Again, at the points at which this longitudinal arcade intersects with the transverse arcade, the six columns meet in clusters set into thick piers.
Where this longitudinal arcade joins the courtyard façade, at its northernmost end, it culminates in a single large eleven-lobed double arch supported on paired columns and capitals, at the junction between Cordoba IV and Cordoba I (Figure 60). Unlike in the rest of Cordoba IV, the marble capitals used here appear to be earlier spolia, perhaps reused from a previous phase of the building (such as an external gate built by ʿAbd al-Raḥmān I or II?) that was demolished as a result of the ʿĀmirid extension.
This is not the only location in the mosque where a large eleven-lobed double arch is found: two such arches are found at the northern entrance to al-Ḥakam’s maqṣūra, flanking the even more elaborate central arch of twenty-one lobes that leads into the space now occupied by the Capilla de Villaviciosa (Figure 45). The easternmost of the two spaces designated by these eleven-lobed arches is now occupied by the Capilla Real (Figure 78); the southern side of this space is marked by a further eleven-lobed arch (though not double), supported on pairs of columns which are topped by double-capitals, carved from a single block of marble (Figures 39–40, 79). As discussed in Chapter 4 (5), double-capitals like these may have been an ʿĀmirid innovation, and in fact this southern eleven-lobed arch seems to have been part of an intervention by al-Manṣūr in al-Ḥakam’s prayer hall, as will be discussed below (2.1.4.2). If so, perhaps this inspired his architects to try the feature again, in another part of the ʿĀmirid extension.
The fact that the precedent for this decorative arch at the northernmost end of the longitudinal arcade was the symbolic entrance to al-Ḥakam’s maqṣūra indicates a special significance to this point of junction between Cordoba IV and Cordoba I. Was it because here the ʿĀmirid mosque was connecting with the oldest and most venerated part of the Umayyad dynastic monument? This significance might be underlined by the reuse of spoliated capitals, in a manner similar to the respectful reuse of the columns and capitals from ʿAbd al-Raḥmān II’s mihrab in the new mihrab erected by al-Ḥakam II.
There is no balancing eleven-lobed arch at the southern-most end of the longitudinal arcade because this is where the caliphal treasury and its now-internal entrance, the ‘Puerta del Chocolate’, was located.
2.1.3.3 Alternating Pattern of Coloured Columns
Ewert complains that al-Manṣūr ‘brusquely abandoned the simple rhythmical order of the columns’ which he so admires in al-Ḥakam’s extension, and instead ‘imitated the chaos of reused pieces in the mosque’s first two phases’.123 He is referring to the unbroken pattern in Cordoba III in which red marble columns alternate with black marble. However, in general the columns in Cordoba IV do maintain an alternating pattern, although this extension employs three coloured marbles (red, black and grey, colours which Ewert calls ‘sober’) (Figure 43).124 More than twice the number of columns were required for Cordoba IV than for al-Ḥakam’s extension, and the breaks in the pattern may be due to marble of one colour being less available than another. However, since al-Manṣūr’s builders imitated a pattern established by al-Ḥakam, and were also careful to maintain the original pattern at the junction with Cordoba III, they were clearly aware of it and used it as another subtle means to identify this extension with its predecessor.
2.1.3.4 Newly-Carved Stone Capitals
The capitals in Cordoba IV were all custom-made from stone, and imitated the simplified Corinthian capitals used throughout al-Ḥakam’s extension. On the basis of one capital per column, some 350 new capitals were specially produced for this extension, which would have been a further huge expense.125 In Cordoba I and II the majority of the capitals were reused spolia, from ruined Roman and Late Antique buildings in Cordoba itself or brought by river from Seville and Mérida (Figure 25). By the time of the construction of Cordoba III, there cannot have been as many standing ruins available to despoil, hence a standardised design was introduced for the mosque’s capitals. Ewert suggests that this simplified Corinthian design was deliberately used in Cordoba III as a ‘means of contrast with [its] decorative accents’,126 and seems disappointed by the fact that only in one small section of Cordoba IV does al-Manṣūr introduce some variation in the design of his capitals.127 However, this imitation of the design of the capitals from Cordoba III provides another example of the deliberate relation of Cordoba IV to the immediately preceding extension.
11-lobed arch at the junction of Cordoba I and Cordoba IV, terminating the longitudinal arcade (section 2.1.3.2)
© Mariam Rosser-Owen2.1.3.5 Imitation of al-Ḥakam’s Original Eastern Façade
The last but most significant way in which al-Manṣūr’s extension imitated Cordoba III is now difficult to appreciate, but would have been obvious to contemporaries who knew both façades. Evidence from the remains of al-Ḥakam’s original eastern façade, which now forms the transitional wall between Cordoba III and IV (Figures 51, 61–65), shows that al-Manṣūr imitated the scheme of its gates on his eastern façade (Figures 66, 68–74).
Reconstruction of the elevation of al-Ḥakam’s original eastern gates
© Matilde GrimaldiDetail of the arch of al-Ḥakam’s original Gate 2
© Mariam Rosser-OwenDetail of the arch of al-Ḥakam’s original Gate 3
© Mariam Rosser-OwenTraces of blind arcade above al-Ḥakam’s original Gate 3
© Mariam Rosser-OwenSurviving section of al-Ḥakam’s original Gate 4
© Mariam Rosser-OwenVista along the ʿĀmirid eastern façade
© Mariam Rosser-OwenPlan showing the names and locations of the mosque’s gates and key internal inscriptions
© Matilde GrimaldiʿĀmirid eastern façade: Gate 1: view and detail of inscription in archivolt
© Mariam Rosser-OwenʿĀmirid eastern façade: Gate 2: view and detail of inscription in archivolt
© Mariam Rosser-OwenʿĀmirid eastern façade: Gate 3: view and detail of inscription in archivolt
© Mariam Rosser-OwenʿĀmirid eastern façade: Gate 4: view, detail of inscription in archivolt, detail of inscription in frieze above arch
© Mariam Rosser-OwenʿĀmirid eastern façade: Gate 5: view, detail of inscription in archivolt which retains traces of red pigment on the surface of the letters, detail of inscription in frieze above arch
© Mariam Rosser-OwenʿĀmirid eastern façade: Gate 6: view and detail of inscription in archivolt
© Mariam Rosser-OwenʿĀmirid eastern façade: Gate 7: view and detail of inscription in archivolt
© Mariam Rosser-OwenAll the mosque’s external gates follow the same basic scheme as introduced by the Bāb al-Wuzarāʾ, built by Muḥammad I in 855–6.128 All are surmounted by a frieze of five blind arches. On al-Manṣūr’s eastern façade the types of arches in these friezes alternate with each gate (Figure 82): Gate 3 bears five small horseshoe arches (Figure 70), Gate 4 bears five trilobed arches (Figure 71), Gate 5 horseshoes (Figure 72), and so on. Accordingly, we can extrapolate that the decoration of the friezes that have not been reconstructed on Gates 1 and 2 consisted of horseshoe and trilobed arches, respectively (Figures 68, 69).
On the basis of comparison with the surviving gates of the western façade, it would appear that the decoration of the gates on al-Manṣūr’s eastern façade was very different, even that al-Manṣūr introduced a new decorative system for the façade gates. This would be surprising, given the ways in which al-Manṣūr otherwise physically identified his extension with Cordoba III through the architectural imitations and allusions just discussed. However, the true relationship between the two eastern façades has not been fully appreciated. As mentioned, the best-preserved gate of al-Ḥakam’s original eastern façade is that popularly known as the ‘Puerta del Chocolate’ (Figure 52), which is decorated with an upper blind arcade of intersecting horseshoe arches. It has been assumed that all the remaining gates on both eastern and western façades originally had the same intersecting horseshoe design in their upper frieze.129 Two gates on the western façade were in fact restored by Velázquez Bosco in the late-nineteenth century as exact copies of the Puerta del Chocolate.130 One of these restored gates is the Bāb al-Sābāṭ, whose façade would not originally have been visible, since this was the entrance that opened into the covered passageway connecting the mosque with the palace. However, there is no evidence to suggest the existence of such a design anywhere but on the Puerta del Chocolate, a special entrance (from the street into the Bayt al-Māl) that could be expected to be more elaborately decorated.131 Just as al-Ḥakam’s maqṣūra was marked as a privileged space through decoration different from the rest of his extension, it is likely that the original exit from the mosque’s treasury was also marked as special by means of a different decorative theme, one which evoked the interlacing arches of the maqṣūra itself.
The ghostly evidence of the surviving gates of al-Ḥakam’s eastern façade tells a different story (Figures 61–65): that they originally bore alternating friezes of horseshoe and trilobed arches in exactly the arrangement as seen today on al-Manṣūr’s eastern façade; furthermore, that the location of al-Ḥakam’s gates corresponds exactly to those of al-Manṣūr’s gates; finally, that the alternating horseshoe/trilobed arrangement occurs in the same place on al-Manṣūr’s gates as it did on al-Ḥakam’s gates. In sum, far from being a stylistic innovation, the decorative system employed on al-Manṣūr’s eastern façade seems to have been a copy of al-Ḥakam’s eastern façade.
It must be acknowledged that Gates 3 to 7 of the ʿĀmirid façade were restored by Velázquez Bosco between 1908 and 1914.132 In line with his restoration approach throughout the Mosque, he conserved as many original elements as had survived, albeit in a deteriorated state, and only executed anew those parts which had totally disappeared. On the eastern façade, the decoration above the gates was in a particularly bad state of preservation, having suffered from interventions in the eighteenth century. Velázquez’s original proposal was to substitute copies of the interlacing arcades above the Puerta del Chocolate, as he had done on the western façade. However, this was not the scheme that was ultimately executed. Though the documentation does not survive to explain this change of direction, given what we know of Velázquez’s approach to preserving the original integrity of the building, it is probable that it was due to discoveries of aspects of the original decoration of al-Manṣūr’s eastern façade.133 While the overall scheme clearly relates to that of al-Ḥakam’s original eastern façade, there are significant differences in the details, in the surviving decoration and in the content of the inscriptions, for example, so it is clear that Velázquez’s interventions did not merely produce a reconstructed copy of al-Ḥakam’s façade.
Looking more closely at the vestiges of al-Ḥakam’s original eastern façade that survive inside the mosque, we can reconstruct three gates, in addition to the Puerta del Chocolate: their locations are marked on Figure 67 and the elevation of the façade is reconstructed in Figure 61. The tympanum of the first gate from the south (Gate 2) is well preserved, though its left edge is cut off by the first of the large arches which al-Manṣūr inserted into this wall to connect his extension with the rest of the mosque (Figures 51, 62). Above and to the left of this tympanum, a lone, blind horseshoe arch decorated with a pattern of inlaid brick is the only trace of the original upper frieze of five horseshoe arches which once decorated this gate. The remains of the next gate (Gate 3) occur just to the left of where the third large arch of the transitional arcade opens, and though the voussoirs of its arch retain much of their original carved decoration, this tympanum is otherwise quite damaged (Figure 63). Immediately above it, this section of the wall retains the ghost of this gate’s original blind arcade of five trilobed arches (Figure 64). A fourth gate (Gate 4) retains only a small portion of the archivolt of its arch (Figure 65), which contains a Qurʾānic inscription (see section 3, and Appendix 3B). To its left are fragments of decoration in inlaid brick, including a small trilobed arch, whose function it is difficult to determine in comparison with the gates on the current eastern façade: it is possible these fragments represent an aspect of the original decoration of all the gates, which has not been preserved by the other extant examples.
A last element occurs between Gates 2 and 3: located in the spandrels of the second arch of the transitional arcade are fragments of two blind, lobed arches. From their lower position relative to the blind horseshoe arcade of Gate 2, and also from the size of the lobes, it is possible to identify these as a gate’s lateral decoration, in which a blind arch of five (sometimes more) lobes framed the marble window grilles that flanked the gate. However, it does not seem likely that the two lobed arches seen here indicate the two flanks of one gate, since this would presuppose the existence of a third gate set very closely between Gates 2 and 3: it would be located where the transitional arch is now, and would therefore not correspond to an opening on the current eastern façade. It is more likely that the left-hand arch represents the right-hand flank of Gate 2, and the right-hand arch the left-hand flank of Gate 3, as reconstructed in Figure 61.
This analysis of the surviving traces of three gates of al-Ḥakam’s original eastern façade allows us to conclude that the arrangement employed by al-Manṣūr on his façade was copied from that of his predecessor – if we allow that Velázquez Bosco based his restoration on the identification and incorporation of original elements. This decorative imitation would have been obvious to contemporary viewers of the mosque, accustomed to walking past al-Ḥakam’s eastern façade until a few years before. It would have been read by them as a clear (and clearly visible) statement by al-Manṣūr of the continuity of his extension with that of al-Ḥakam. What we cannot judge with certainty, due to the lack of extant decoration on Gates 1 and 2 of al-Manṣūr’s façade, and the extent of the twentieth-century restorations, is whether the decoration itself – the patterning of the brick inlay or the carved stone – was also imitated. However, as will be examined in section 3, the Qurʾānic inscriptions used on al-Manṣūr’s façade did not imitate those on al-Ḥakam’s, although they conveyed the same overall messages.
The only gate which al-Manṣūr did not copy was the Puerta del Chocolate (Figure 52), and the reason for this is clear: there is no gate in this location on the current eastern façade. The mosque’s treasury (and contents) stayed where it was. Nor was there a need to replicate the Sābāṭ exit on the western façade for the sake of symmetry, since this was not originally an exterior gate.
It is also clear from the survival of ornamental elements of al-Ḥakam’s original eastern façade inside the mosque that, though al-Manṣūr’s builders stripped them back, they did not reuse their decorative features as building material, and this is another testament to the amount of money lavished on the ʿĀmirid extension. Nevertheless, as mentioned above, excavations under the pavement of the mosque have shown that some of the masonry from the demolished eastern façade was reused in levelling the ground and laying the foundations for the new construction.134
2.1.4 ʿĀmirid Interventions into al-Ḥakam’s Prayer Hall
There is intriguing evidence that al-Manṣūr sponsored interventions in two places in al-Ḥakam’s maqṣūra, one of which (2.1.4.1) seems to have been born from technical needs to add greater strengthening to the structures of the qibla zone; the other (2.1.4.2) is a more mysterious space and may have had a more ceremonial function.
2.1.4.1 Addition of Transverse Arcade along the Qibla Wall
The first possible intervention is the transverse arcade that runs parallel to the qibla wall two bays from the south (Figures 49, 75), to east and west of the screens of polylobed arches that frame the three central bays of the maqṣūra. At its westernmost end, the arcade abuts the external wall of al-Ḥakam’s prayer hall right in the middle of the interior lintel of Gate 2 (Puerta de San Ildefonso, see Figure 67), indicating that the transverse arcade was constructed when the perimeter of the prayer hall was already built. Antonio Momplet first pointed out that this feature was a late addition to al-Ḥakam’s extension, but he understood it as implying a change of project halfway through the construction of the prayer hall, according with his theory that the three ante-mihrab domes were not part of the original architectural concept of al-Ḥakam’s maqṣūra. Once it was decided to add them, the exterior walls had already been constructed, and the transverse arcade was therefore retrofitted to provide greater strengthening for the heavy weight of the domes.135
Elevation of the transverse arcade along the qibla wall
© Concepción Abad and Ignacio González CaveroHowever, the qibla wall of the maqṣūra was probably always intended to be crowned by three domes, mirroring the tripartite palatine structures of Madīnat al-Zahrāʾ (Figure 42). Moreover, Momplet did not take into account how the transverse arcade abutted the eastern end of the prayer hall, an issue which Concepción Abad has since resolved.136 Here the arcade is supported by a column, located in what would have been the centre of Gate 2 of al-Ḥakam’s original eastern façade (Figure 76). These gates were blocked up when the ʿĀmirid extension was constructed, in order to create the regular openings of the longitudinal arcade which runs south to north at the zone of transition between Cordoba IV and the rest of the mosque (2.1.3.2). As Abad points out, this column has to have been placed here after the original gate had been blocked up, thus indicating that the column can be associated with the ʿĀmirid extension.137 Indeed the diameter of the shaft relates it to the other columns of Cordoba IV, which are slightly slimmer than those of Cordoba III. The column also bears the name Masʿūd, one of the masons’ marks which, as discussed above (Figure 50), are particularly characteristic of the mosque’s ʿĀmirid extension.
Transverse arcade, where it abuts the eastern end of prayer hall, supported by an ʿĀmirid column
© Concepción Abad and Ignacio González CaveroLogically, then, the transverse arcade came to rest on this column after it had been added to the building during the construction of Cordoba IV, indicating that the transverse arcade is itself an ʿĀmirid addition to the building. Since Gate 2 on the western façade remained in use as an entrance to the building, the architects had no option but to brace the arcade’s western end against the gate’s interior, with additional support coming from the existing perpendicular arcade of Cordoba III’s westernmost nave.
Why build this transverse arcade? As Abad notes, when the original eastern façade of al-Ḥakam’s prayer hall was opened up to construct al-Manṣūr’s extension, the domes of the maqṣūra lost the bracing support of the external wall.138 This was, therefore, another ingenious solution by al-Manṣūr’s architects (Figure 75). Ruiz Cabrero also notes the presence of another water channel running along the roof above this location.139 He admits that the post-Islamic works on the roof make it difficult to determine if this channel had been constructed by al-Ḥakam or later, or even if it continued into the ʿĀmirid extension; but if this arcade is an ʿĀmirid addition, so too is the water channel. Supporting the channel is thus the reason why the line of this arcade continues west to east across Cordoba IV (see Figure 49), as well as giving extra bracing support.
This arcade incorporates the ‘screens’ of the three bays of the ante-mihrab domes, which previously would have projected as a kind of portico within the mosque, the arcade is almost invisible – it blends in to the extent that it is not included on several published plans of the mosque.140 The outer three arches on either end are simple open horseshoes, high, wide and lightened by the unusual solution that their spandrels are pierced. As Momplet notes, the structural support is provided by the arch itself, so the spandrels here are not doing anything.141 The spandrel designs are simple: at the far eastern end they consist of rows of eight-lobed rosettes, resembling sliced oranges (Figure 76). The idea for piercing the decoration may have come from the ‘screens’ in front of the Bāb al-Sābāṭ and Bāb Bayt al-Māl, where the pair of five-lobed arches at the lower level each have a single polylobed shape pierced in their spandrels. The arcade that continues on either side of these screens picks up on this idea but elaborates it. At the same time, al-Manṣūr’s builders were carving window grilles in openwork, so this innovative notion could have transferred from there.
The arches that directly flank the three central naves of the maqṣūra are more decorative, and again of an unusual type (Figure 77). Only the western arch survives, though part of the eastern arch is visible, embedded into the later projecting wall of the Capilla del Cardenal Salazar. They are pointed, and thus among the earliest examples of the use of the pointed arch in al-Andalus.142 As mentioned above (2.1.1.3), pointed arches may have been used by al-Manṣūr’s architects in Cordoba IV, to adapt the intercolumniations of the new extension to those of the existing structure (Figure 56). Perhaps this gave them the idea to monumentalise this new feature, to develop a special architectonic device that was appropriate to their privileged location, flanking the royal maqṣūra in front of the qibla wall. These two arches also feature more complex decoration in their openwork spandrels than the flanking arches, and they are rendered yet more decorative by another unusual feature, which also seems to have been used here for the first time:143 their intradoses are filled with rolls, twenty-two on either side of the arch, a purely decorative device added in plaster. At the outer surface of the arch, the profile of these rolls are painted in red pigment; a central vein is also painted along the inner length of each intrados, evoking the shallow recess between double arches that we see elsewhere in the Mosque (Figure 78, for example). This painted decoration is not original but probably follows the original paintwork. These rolls seem to be monumental versions of the more decorative modillions added above the capitals throughout Cordoba IV (2.1.2.2). As such, this monumentalised combination of two experimental elements from Cordoba IV – the pointed arch and increasingly elaborate modillions – draws a visible connection between the two spaces: the new ʿĀmirid prayer hall, and al-Manṣūr’s intervention in the maqṣūra. As the sources tell us, the three central naves of the maqṣūra were enclosed by a wooden screen (indicated on Figure 49), and as a result the arches with rolls in their intradoses would have been particularly visible, flanking the most ‘privileged’ space on either side of the maqṣūra.
Transverse arcade along the qibla wall: arch to the west of maqṣūra naves, with rolls in intradoses
© Concepción Abad and Ignacio González CaveroFinally, there is evidence of further decoration in this area, or at least the implication of the transverse arcade into the maqṣūra’s epigraphic programme (discussed in detail below; see also Appendices 3A and 3B). As Calvo notes, the far western end of this arcade bears a painted inscription, visible on entering the mosque through Gate 2 of the western façade. Containing the Qurʾānic passage 10:106–7, this verse reaffirms the theme of predestination which had been introduced in the inscriptions outside the Mosque.144 If the transverse arcade is an ʿĀmirid addition, this painted inscription is also, and introduces the way in which al-Manṣūr’s choice of Qurʾānic inscriptions to adorn his additions to the mosque recapitulated and emphasised the existing epigraphic programme devised for al-Ḥakam’s extension. The medium of painting also calls to mind the inscriptions that were painted on the interior wall of al-Manṣūr’s new courtyard façade (2.1.1.2), whose content would also no doubt have underscored the building’s existing messages. These will be discussed in detail in section 3 below.
2.1.4.2 An ʿĀmirid Tribune in the Maqṣūra?
In their detailed study of the Capilla Real, located in the bay to the east of the Capilla de Villaviciosa, Concepción Abad Castro and Ignacio González Cavero have noted an interesting series of features which appear to antedate any Christian interventions in this space (Figures 49, 80–81).145 As discussed above, the bays now occupied by the Capilla Real, Capilla de Villaviciosa and the bay to the west, formed a tripartite space that functioned as an internal façade and entrance vestibule into the maqṣūra constructed by al-Ḥakam; the wooden maqṣūra screen would have run along the front of these bays, and its door was probably located under the elaborate central arch. The eastern and western bays were introduced by eleven-lobed double arches that rested on plain stone capitals like those used throughout al-Ḥakam’s extension. Ruiz Souza has posited that the bays were also enclosed on their southern sides, by screens of interlaced arches matching those opposite them at the far south of the naves, in front of the Bāb al-Sābāṭ and Bāb Bayt al-Māl.146 However, according to Abad, it is unlikely there were originally closing arches on the southern side of these bays, as the column of the perpendicular arcade delimiting the eastern side of the maqṣūra gives no indication that it supported an additional arch at this point.147 She considers that both eastern and western bays would originally have been open to the maqṣūra, leaving the central domed space projecting into the central nave.148
All agree, however, that the eleven-lobed arch that now closes the southern end of the eastern bay (Figure 79) was added later, in imitation of the arch on its northern side. But while Ruiz Souza attributes this addition to Christian interventions in the fourteenth century, Abad and González associate this and other features with an ʿĀmirid transformation of this space.149 This southern arch has no central vein, as we normally see in the larger arches inside the mosque (including that other eleven-lobed arch added by al-Manṣūr, at the intersection of Cordoba IV and Cordoba I: 2.1.3.2, Figure 60). This is a single arch of double-width, and presumably originally its voussoirs were painted to match those on the northern side, just as the arcades of Cordoba IV have painted voussoirs. The capitals that support this arch are carved, composite capitals, in the style known from Andalusi palatine structures, and thus call particular attention to this space because elsewhere in Cordoba III and IV the capitals employed are plain. Ruiz Souza attributes this to spolia from Madīnat al-Zahrāʾ introduced into the Capilla Real in the fourteenth century, as are the miniature capitals and columns supporting the arches of the chapel’s crypt at its northern and southern elevations (Figure 78).
ʿĀmirid tribune: 11-lobed arch at northern end, showing crypt of fourteenth-century Capilla Real beneath
© Mariam Rosser-OwenʿĀmirid tribune: 11-lobed arch at southern end. The western double-capital is just visible at the bottom of the photo.
© Concepción Abad and Ignacio González CaveroHowever, the capitals employed here are carved as perfectly integrated double capitals, which, as discussed in Chapter 4, were probably an ʿĀmirid innovation (Figures 39–40). The only other example of such double-capitals is that which bears the signatures of its three carvers, now in the Cordoba Archaeological Museum and discussed by Juan Souto (Figures 37–38).150 As Abad has noted, the perpendicular arcades of the maqṣūra’s three naves have twinned columns at the point where they intersect with the façades of the mihrab, Bāb al-Sābāṭ and Bāb Bayt al-Māl (Figure 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 the other capitals of al-Hakam’s extension.151 Stylistically the two capitals here match each other, and seem to have been carved as a matching pair. While the interlacing guilloche bands along the abacus between the volutes are plaster additions, possibly from the fourteenth-century refurbishment of the chapel,152 both capitals make much use of the vertical ‘chain link’ motif that is also employed on the double-capital in the Cordoba Museum. That capital is too fragmentary to see if other motifs are shared with the mosque capitals. No egg-and-dart motif is used in the mosque capitals, but there is a similar treatment to the spikiness of the leaves on the bodies of all three. It thus seems that the two double-capitals used in this space were created specifically for it by al-Manṣūr’s architects, or brought as an existing pair from another ʿĀmirid monument (presumably at al-Madīnat al-Zāhira). The double-width arch that rises above was thus adapted to the double-capitals beneath.
Most significant, however, is Abad and González’s observation that the bases of the columns do not reach the same groundlevel as the pavement of al-Ḥakam’s extension, but fly in the air 70 cm above it (Figure 80).153 This confirms the existence of another groundlevel at this elevated height, approximately three steps up from the groundlevel of the maqṣūra. This indicates, then, a space with a slightly higher elevation in relation to the rest of the mosque, monumentalised by the polylobed arches that delimit it at north and south sides; to its west was the screen of intersecting polylobed arches of the domed central bay, and to the east perhaps it was open to the mosque, though this bay was also enclosed at north and east by the wooden screen of the maqṣūra (see reconstruction in Figure 49). Perhaps the steps up to the elevated platform opened into the mosque at the bay’s southern side, facing towards the qibla and connecting this space with the maqṣūra. The fact that the double capitals are decorated, in contrast to the plain capitals used elsewhere, gives a sense of greater privilege to this space. There may also be evidence of another painted inscription at high level: during restoration work on the exterior of the Capilla de Villaviciosa dome in the 1990s, a painted inscription was found around the embrasure of the northern-most window on the western side of the Capilla Real, ‘written in floriated Kufic, in letters of a monumental character, which correspond to the repetition of the word “al-mulk” (“sovereignty”)’.154 Because the transformations of the Capilla Real space have traditionally been considered to date from the fourteenth century at the earliest, Pedro Marfil interpreted this inscription as preserving the traces of an external decorative programme around the central dome, dating from the period of al-Ḥakam II: they represented the ‘few preserved remains of what must have been a frieze that crowned the exterior of the caliphal lantern’.155 However, the exterior decoration of the Capilla de Villaviciosa dome support is also the internal decoration of the space now occupied by the Capilla Real. Could this painted frieze, repeating the phrase al-mulk in floriated Kufic around the upper walls, have related to the transformation of this space by al-Manṣūr? Such a phrase could equally exalt God and highlight the royal associations of the space beneath.
Why would al-Manṣūr have presided over the transformation of the maqṣūra’s northeastern bay in this highly interventive manner? There is no evidence that the bay to the west was monumentalised in a similar manner, because all traces of the mosque’s tenth-century structure were erased by the construction of the Gothic cathedral. Abad and González consider the possible function of this space (Figures 80–81). One possibility they consider is an early example of a dikka, the elevated ‘respondents’ platform’, used by a muballigh for Qurʾānic recitations, chants and calls to prayer inside the mosque (iqāma).156 Given the vast size of the mosque after the completion of Cordoba IV, such a feature might have become necessary to relay the words of the imam or the stages of the prayer to the assembled congregation. However, if this were a dikka, it seems to be placed too deep in the mosque, in a location where the muballigh’s calls would not be effectively heard by congregants closer to the courtyard end of the prayer hall. Extant dikkas tend to be located midway between the mihrab and the mosque’s courtyard, on the same axis as the qibla; they are also generally raised on columns two to three metres high, and are thus much more elevated than the platform at Cordoba, allowing the voice of the muballigh to carry further. Though George Michell notes that ‘the dikka is a very early innovation and was already in widespread use by the eighth century, increase in congregation size having decreed its invention’,157 all extant dikkas are much later in date. They are seen especially in Mamluk and Ottoman mosques, and in Cairo the earliest dated examples were erected in the fifteenth century; the dikka at the al-Aqṣā Mosque in Jerusalem was built with reused Crusader capitals, and thus appears to date from the twelfth century at the earliest.158 In the absence of apparent precedents for such a feature, it is potentially anachronistic to associate this intervention with the architectural feature of the dikkat al-muballigh. In the same way, this only slightly elevated platform located so deep in the mosque and not on the qibla axis would not have been particularly effective for relaying the iqāma or stages of the prayer.
ʿĀmirid tribune: reconstructed elevation
© Concepción Abad and Ignacio González CaveroʿĀmirid tribune: reconstructed interior view
© Concepción Abad and Ignacio González CaveroHowever, there are many other activities for which such a space might have been used, indeed it may well have been multipurpose, for special activities and ceremonies. Recitations of the Qurʾān may well have happened here, as well as discussions of Qurʾānic interpretation (tafsīr) and hadith. An Arabic inscription found in the Capilla de Villaviciosa in 1766 was translated as indicating that ‘this is the room where our learned doctors dispute our law’.159 It is entirely plausible that this liminal zone, connecting the maqṣūra with the public areas of the mosque, was used to dispense justice, one of the regular activities that took place inside the mosque. As Lévi-Provençal has discussed, there was no special building in Cordoba that the qāḍī used for his audiences, which took place in an annex (‘une dépendance’) of the Great Mosque or in the prayer hall itself, if not in the judge’s own home.160 The paraphernalia of justice did not require much space: the qāḍī sat, legs folded or leaning on cushions, surrounded by councillors and his clerk (kātib). The two parties stood before him, while the other litigants waited their turn to be called by an usher (ḥājib). They could be assisted by advocates (khasm or wakīl), who presented their defence. The judge dealt with complaints of all sorts – marriages, repudiations, alms donations, certificates of bankruptcy; Calvo notes that acts of conversion to Islam or repentance for or condemnation of apostasy also took place before him.161 The chief qāḍī could also summon the majlis al-shūra, or council of fuqahāʾ. Perhaps this is what the inscription found in 1766 was referring to. Abad and González also suggest it could have been used as a space for teaching fiqh, and that by creating a dedicated space for this in such a privileged location in the mosque, al-Manṣūr was courting the favour of the fuqahāʾ, whose support he needed to legitimise his regime.162
Other public acts took place in the mosque, including the swearing in of public officials and all types of legal oaths, which Maliki jurisprudence recommended to be taken on foot and next to the minbar.163 However, with the construction of al-Ḥakam’s maqṣūra, the minbar was largely inaccessible to the Cordoban populace, not just because during the week it was stored in a chamber next to the mihrab and only brought forward for the khuṭba on Fridays. The public bayʿa (bayʿat al-ʿāmma) to Hishām as al-Ḥakam’s heir apparent took place next to the (new) minbar – which, as Fierro argues, may have been commissioned specifically to be used on this occasion – and this may well have been one of the few instances on which members of the public were allowed into the new maqṣūra; this would have underlined the specialness of the occasion.164 Could al-Manṣūr have conceived of a subsidiary, quasi-public space, near enough to the minbar, where such activities could have taken place? Could the rituals associated with the Qurʾān of ʿUthmān and the weekly viewing of the relics have taken place in this new space?165
Decrees, such as those condemning the work of Ibn Masarra or other heterodoxies, could have been proclaimed from here as well as from the mosque’s external gates, as could important declarations, such as the regular reports of al-Manṣūr’s victories in campaigns against the Christians and the Fatimids. Indeed, this elevated area could have provided a more public space for the rituals surrounding the declaration of jihad, which required religious authority and was thus undertaken within the physical framework of the congregational mosque. In particular, the new space could have provided a stage for the ritual of knotting the military banners to their lances (ʿaqd al-alwiya) before the Umayyad army set off on campaign.166 This ‘religious ceremony’ was an important part of initiating the ceremonial of war: it took place in the Cordoba Mosque on the Friday before the date of the army’s departure, presided over by a high member of the caliphal administration – as Bariani notes, it was ʿAbd al-Malik ibn al-Manṣūr himself during his ḥijāba, and it is not unlikely that al-Manṣūr would likewise have presided over this ritual in person. While the banners were being tied to their lances, the mosque’s imams recited the sura of Victory (Sūrat al-Fatḥ, Qurʾān 48), which contains numerous passages relating to the conduct of jihad.167 The final verse was recited at the same moment that knotting the banners was concluded. These banners were the symbols of power which the army carried with them into battle, and this ceremony placed the banners and by extension the entire army and its leaders under divine protection. The sources do not tell us but perhaps the banners were stored in the mosque’s treasury when not in use, to invest them with yet more baraka.168 Three banners are named in al-Rāzī’s Annals: al-ʿuqda, al-ʿalam and al-shaṭranj; the last apparently bore a chequered design, and seems to have been the most important ensign during al-Ḥakam’s reign.169 These symbols of the Umayyad army were also significant for the ʿĀmirid ḥājibs’ role as leaders of jihad on the caliph’s behalf: we will return in Chapter 8 to the importance of the military banners and their designs, in particular relation to the visual imagery of ʿĀmirid art.
This elevated space, or tribune, at the heart of the mosque could thus have been constructed as a site of ceremonial, a stage for activities relating to the prosecution of justice, religious orthodoxy and jihad. Its location was offset in relation to the mosque’s axial nave, so it did not interrupt the visual axis of the Umayyad mosque or the possible processional route in and out of the maqṣūra. The western side of the maqṣūra was orientated towards the Umayyad qaṣr, and during the reign of al-Ḥakam, following earlier tradition, the caliph was accustomed to sit in the bay in front of the Bāb al-Sābāṭ, or in the room just inside the door, to hold audiences after the congregational prayer.170 In contrast, the tribune was orientated towards the east, in the direction of al-Manṣūr’s mosque extension and, beyond, his own palatine city, al-Madīnat al-Zāhira. This was also the direction in which the Umayyad armies departed. As we have noted, architectural features, such as the painted voussoirs and the double-capitals, relate this space with al-Manṣūr’s other architectural projects. Was this space used by al-Manṣūr for his own audiences after Friday prayers? It is unlikely that he would have sat here during the prayers, which would have been an explicit and visible statement that he was placing himself at an elevated level vis-à-vis the caliph (whether or not Hishām was physically present in the mosque) and the chief qāḍī, who led the prayers.171 What we have seen of al-Manṣūr’s delicate balancing act in terms of his relations with the ʿulamāʾ makes such a statement highly improbable. Another or additional possibility is that this space was occupied by Hishām II on the rare occasions when he visited the mosque, such as at the end of the waḥsha.172 Given the pains taken to hide Hishām and his condition from the Umayyad court and Cordoban people, the space could have been configured with textile hangings to enclose the person of the caliph and hide him from the congregation.
This section has discussed in detail for the first time the ʿĀmirid extension to the Cordoba Mosque. It has surveyed what we know of the labour force used in building the mosque and creating its physical elements; the new infrastructure that was created to support the expanded mosque, in the form of a new cistern and ablutions pavilion, and the possible restoration of the two remaining caliphal ablutions pavilions; the ingenious engineering solutions found to level the terrain on which the mosque was constructed, and the architectural solutions used within the mosque to solve disparities caused by building over or connecting to existing structures; the subtle but increased level of decorative devices within Cordoba IV; the features that were imitated from Cordoba III, in particular in the arrangement of the gates on the eastern façade, that underlined the relationship of continuity from the caliphal into the ʿĀmirid mosque; and the probability of ʿĀmirid interventions in al-Ḥakam II’s maqṣūra, for reasons of structural necessity as well as to provide a stage for a range of ceremonial activities that took place inside the mosque, including the prosecution of justice, the teaching of religious law, and the preparation for jihād. All these elements combine into an extremely significant building which has not been recognised as such. We now turn to discuss another significant and understudied aspect of the Cordoba Mosque, which is its epigraphic programme.
3 Qurʾānic Inscriptions at the Great Mosque of Cordoba
See the plan indicating the location of the gates in Figure 67. The Qurʾānic inscriptions in the Cordoba Mosque and on its external gates are detailed in Appendices 3A and 3B.
As discussed, no foundation or commemorative inscription survives from al-Manṣūr’s extension to the Cordoba mosque. However, an important set of inscriptions do survive, whose potential significance has only recently begun to be considered.173 These are the Qurʾānic inscriptions that form an essential element of the exterior decoration of the seven gates on the mosque’s eastern façade. Since these inscriptions contain no ‘historical’ information, they have suffered the same fate as many purely religious monumental inscriptions all over the Islamic world, which is to be ‘dismissed as mere banalities’.174 In general, Qurʾānic verses have only been identified if they form part of an historical inscription; otherwise the use of religious writing on buildings has been considered no more than decoration, what Grabar styles ‘the faithful handmaiden of architecture’, or Ettinghausen ‘symbolic affirmation before God of the faith’.175 However, more recent studies have begun to appreciate that ‘any text presupposes a readership’, that the selection of certain Qurʾānic passages for a decorative programme is never mere accident, but instead part of ‘the discourse of architecture’, in which iconographic and epigraphic imagery illuminate each other, and ‘constitute … an address to multiple audiences’.176 Thus, Oleg Grabar’s reading of the anti-Christian statements in the mosaic inscriptions of the Dome of the Rock,177 Irene Bierman’s study of the ‘Fatimid public text’,178 and Finbarr B. Flood’s discussion of the lost epigraphy of the Damascus Mosque as a complement to its iconography,179 have led the way in suggesting how political or religious statements were encoded in the choice of Qurʾānic inscriptions by the patrons of such monuments, how these statements may have been read by the contemporary Muslim and non-Muslim populations, and how reading the inscriptions can lead to a better understanding of a building’s decorative programme.
This consideration of the significance of religious epigraphy was first applied to the Cordoba Mosque by Nuha Khoury, and more recently by Susana Calvo Capilla. Both scholars largely focus on the inscriptions of al-Ḥakam’s maqṣūra and mihrab, though Calvo has also considered the external façade inscriptions of both al-Ḥakam’s and al-Manṣūr’s extensions.180 As with the Damascus Mosque, the emphases in the Cordoba inscriptions are on the spiritual obligations of all Muslims that will lead to the rewards of Paradise, such as prayer and faith in God, while rejecting Christian trinitarianism and appealing to other ‘Peoples of the Book’ to turn to Islam. Khoury’s main conclusions are political: the inscriptions on the two imposts of the mihrab’s arch stress the truth (ḥaqq) and divine guidance (hudā) of Islam, and by extension, of the Andalusi caliphate.181 In the mihrab itself, below the scallop-shaped dome, is the exhortation ‘Believers, … cling one and all to the faith of God and let nothing divide you’ (3:103): this can be said to ‘issue a call for a unified caliphate’.182 Janina Safran also reads a political meaning in the epigraphy – ‘The inscriptional programme in its entirety … explicitly asserts the caliph’s guidance and leadership of the community’ – but she adds that an essential aspect of this role was as ‘spiritual guide’.183
These interpretations of the Cordoba inscriptions emphasise the caliph’s position as God’s earthly representative and thereby the divinely-ordained head of the Muslim umma in al-Andalus. These statements have potentially global implications, seeming to lay claim to the Andalusi Umayyad caliphate’s superiority over its competitors, the Abbasids and especially the Fatimids. The message that many of the verses convey is that the population – including Jews and Christians – should unite in the worship of God (islām), and thereby in obedience to their ruler. This concern for political and religious unity hints at something deeper, and Calvo has convincingly identified this as the Maliki response to a local debate on heterodoxy that was of great significance in tenth-century al-Andalus.184 The state-sponsored proselytising on the part of the Shiʿi Fatimids at this period created an atmosphere of anxiety in which anyone professing unorthodox religious views could fall under suspicion.
One heterodox trend in al-Andalus was the rational theology of Muʿtazilism, in which the main ideas were free will, human responsibility for their acts (encompassing both human agency and the denial of intercession), an insistence on the created, non-eternal nature of the Qurʾān, man’s ability to interpret the Qurʾān for himself, and the denial of God’s corporeality.185 As Stroumsa discusses, while some Muʿtazili ideas certainly infiltrated al-Andalus, mainly through travellers to the Islamic East, ‘enough to influence some people and make others quite anxious’, the number of adherents to these ideas remained very small, and it seems clear that there was no Muʿtazili school as such in al-Andalus.186 The term ‘Muʿtazili’ seems merely to have been used as a term of opprobrium, ‘designed to black-smear an unorthodox opponent’.187 Ibn Ḥazm, for example, accused the 979 plot to assassinate Hishām II and replace him with another grandson of ʿAbd al-Raḥmān III as having a Muʿtazili ideology behind it, since one of the ringleaders – ʿAbd al-Malik ibn Mundhir, son of the former chief qāḍī, who was the only one punished for this conspiracy – professed Muʿtazili beliefs.188 Nevertheless, as Fierro notes, it is evident that he was not crucified for being a Muʿtazili, as his brothers were unharmed, but for his role in the conspiracy against the caliph.189 There is no indication that other Muʿtazilis recorded by the sources were persecuted or charged with heresy.190
A more radical heterodox movement that gained some adherents in al-Andalus in the mid- tenth century followed the mystical and Neoplatonic philosophical tenets developed by Ibn Masarra (883–931). His philosophy was poorly understood until 1972, when two of his writings were rediscovered in a manuscript in the Chester Beatty Library in Dublin. These have been used to reconstruct elements of his thought. As Stroumsa and Sviri discuss, Muʿtazili and Sufi elements undoubtedly fed into this, but the greatest influence was Neoplatonic trends as preserved in pseudo-Aristotle. There are also unmistakable elements of Ismaili/Shiʿi teachings: during Ibn Masarra’s travels in the Islamic East, he spent time in Qayrawan just as the Fatimids were coming to power. Stroumsa and Sviri also suggest Ibn Masarra’s association with ‘an intellectual-mystical milieu close to that which, later on, produced the Epistles of the Pure Brethren [Ikhwān al-Ṣafāʾ]’,191 and indeed Godefroid de Callataÿ has argued since that Ibn Masarra was influenced by the Ikhwān al-Ṣafāʾ, whose writings were probably circulating much earlier than has previously been thought. He argues that these had been introduced to the Iberian Peninsula at the end of the 930s, by Maslama al-Qurṭubī, following his long travels in the Middle East, but that Ibn Masarra could have had direct access to them during his own travels in the East and North Africa.192
These connections with Ismailism and with the Fatimids’ philosophy of bāṭinism, or esoteric interpretation of sacred texts, made Ibn Masarra’s views extremely dangerous in the eyes of the Andalusi Umayyad caliphs – though the persecution of his followers did not start until two decades after his death. As Fierro notes, the caliphs seized an opportunity to assert Malikism as the state orthodoxy by clamping down on Masarrism.193 An edict condemning this ‘sect’ (firqa) was read in the congregational mosques of Cordoba and Madīnat al-Zahrāʾ in 952, and a second and third were read from the entrances of the Cordoba mosque in 956 and 957.194 In 350/961, some of their books and writings were burnt by Ibn Zarb, then one of the jurists of the shūra, before the eastern gates.195 It is not clear if this act took place before the death of ʿAbd al-Raḥmān III, or if this was an indication of al-Ḥakam II’s reiteration of his father’s policy against Masarrism and a new statement of Maliki orthodoxy.196 The main accusations made in the edicts of the 950s were that the Masarris believed that the Qurʾān was created; that they renounced and caused others to renounce the spirit of God; that they made a wrong interpretation of Prophetic tradition; they denied repentance and intercession; they insulted the previous generations of pious Muslims; they did not respond to the Muslim greeting; and isolated themselves from the common people, living as ascetics.197 In contrast, ʿAbd al-Raḥmān III ‘represented himself as the champion of the faith against deviation and as guardian of the community against fitna’; the edicts also included a selection of Qurʾānic verses which portrayed ʿAbd al-Raḥmān as the scourge of God delivering his punishment, and emphasised the importance of preserving the sunna of the Prophet.198 Of the fifteen verses cited in the edict of 952,199 eleven can be associated with verses used in inscriptions in the Cordoba Mosque.200 This has led Calvo to read the epigraphic programme of al-Ḥakam’s extension, in particular as concentrated in the maqṣūra, as the ‘definition of orthodox theology’,201 and she demonstrates that the Qurʾānic verses used inside the mosque even respond to specific points of Masarrī heterodoxy.
First, the great emphasis placed in the inscriptions on the rewards of Paradise (Qurʾān 39:46, 40:7–9, 41:30–32: see Appendix 3A) responds to the rejection by the Masarris of the ‘eternal recompense for the believers’; they believed instead that the soul purified by following an ascetic life would receive neither reward nor penalty in the afterlife, but once it had been released from the body would join with the divinity.202 Secondly, the Masarris rejected the existence of the ‘Throne of God’, from which God ‘rules the world’, since ‘God is too great for there to be attributed to Him the act of actually doing something’.203 Whether or not the mosaic-covered dome in front of the Cordoba mihrab symbolically represents the ‘celestial sphere’,204 God’s ‘rule’ over his Creation is also stressed in the Cordoba inscriptions, especially in 32:6. A third emphasis of the mosque’s interior inscriptions is on the ‘central miracle of the faith’,205 the revelation of the Qurʾān (3:4, 3:19, 22:77–78, 39:46), which the Masarris rejected, stating that it was a man-made creation: see, for example, 3:98, ‘“People of the Book, why do you deny the revelations of God?”’.206
Perhaps the most significant aspect of the maqṣūra’s internal epigraphic programme is the orthodox response to the contemporary debate concerning predestination versus free will. Whereas heterodox philosophy taught that man created and was responsible for his actions, according to a power that was created in him by God, orthodox Malikis believed that, though man was free in the choice of his acts and through them could influence his salvation or damnation, the final decision lay with God, and took place on the Day of Judgement.207 In the Cordoba mosque, this translated into a selection of certain Qurʾānic texts which emphasise predestination (3:191–193, 33:70–71).
Furthermore, certain verses were ‘manipulated’,208 in order to turn the Masarris’ argument against them, by means of ‘the “partial” citations of the two āyas which were considered essential to the theory of free will’:209 7:43 (inscribed around the entrance to the mihrab chamber) and 2:286 (on the Bāb al-Sābāṭ). Calvo points out that in both instances those sections that might support the Masarris’ argument in favour of free will are omitted: that is, the final section of 7:43, ‘And a voice will cry out to them, saying: “This is the Paradise which you have earned with your labours”’; and the first section of 2:286, ‘God does not charge a soul with more than it can bear. It shall be requited for whatever good and whatever evil it has done’. Instead, the latter passage was replaced with the words of verse 3:8, which ‘proclaims the mercy of God as the only source of salvation and omits all allusion to the value of actions’.210
The arguments considered here combine to read the maqṣūra’s inscriptions as embodying a reactive statement of the official Maliki line against the contemporary threat of religious heterodoxies; these could lead to the infiltration of Ismaili bāṭinism and endanger the unity of the Andalusi community and the absolute authority of the Umayyad caliph, since the Masarris’ rejection of the existence of the Throne of God undermined the position of his earthly representative. As such it can be said to embody a clear epigraphic programme. The creators of this epigraphic programme sought to reaffirm the caliph’s position as the legitimate, divinely-appointed, spiritual and political guide of the Andalusi umma, and to ‘issue a call for a unified caliphate’.211 The crucial importance of the meaning of these Qurʾānic quotations is underlined by their placement in the inner sanctum of al-Ḥakam’s extension, in the carved marble inscriptions of the maqṣūra and the mosaics of the mihrab vestibule and chamber, which make precise references to the placement of Qurʾānic inscriptions in the Dome of the Rock and the Damascus Mosque. The fact that these inscriptions were located in places where they were not especially visible or legible is not problematic. Ettinghausen argued that the presence of Qurʾānic verses on a building could be said to be symbolic, even talismanic;212 but what counted was that the people building the mosque chose particular verses to inscribe on it, to express a statement of policy. In the case of Cordoba III, the location of these significant Qurʾānic quotations within the maqṣūra, meaning that they could only be accessed by the caliph and his close courtiers, further underlines the association between these inscriptions and state policy. Calvo suggests that the qāḍī of Cordoba, Mundhir ibn Saʿīd, whom the sources indicate assisted in the development of the works at the mosque, may have been involved in or responsible for choosing the particular verses to inscribe on al-Ḥakam’s extension. He is said to have been a man who always had the mot juste, together with the appropriate hadith, which he employed in his sermons and official letters.213 The involvement of such men in the daily running of the mosque would also have ensured that the messages contained in these inscriptions were preached to the worshipping congregation, through the appointment of imams who followed the same branch of strict Maliki Islam. No doubt al-Manṣūr would have done the same, perhaps relying on the same group of jurists with whom he worked on the purge of al-Ḥakam’s library to develop an epigraphic programme for his addition to the mosque.
3.1 Reading al-Manṣūr’s Extension: the ʿĀmirid Epigraphic Programme on the Eastern Façade (Figure 67)
All the façade inscriptions are compiled in Appendix 3B for ease of reference.
In contrast to the inscriptions in al-Ḥakam’s maqṣūra, which could only be accessed by the caliph and his inner circle, the Qurʾānic inscriptions on the outside of the Cordoba Mosque were highly visible and highly legible. One of the surviving gates of al-Ḥakam’s eastern façade (Gate 2: Figures 61, 62) preserves an inscription whose background is painted blue while the letters themselves stand out in plain white stone, recalling the strongly contrasting colour combinations in the mosaic inscriptions around the mihrab and flanking doorways. Similarly, the letters of the archivolt inscription of ʿĀmirid Gate 5 have remains of red pigment on the surface (Figure 72). As noted in section 2, red pigment was also used for painted inscriptions on the interior wall of the courtyard façade, and at the western end of the transverse arcade along the qibla wall (2.1.4.1). If all the gate inscriptions were enhanced by coloured pigment in the same way, the building’s external inscriptions would have been clearly legible.
As discussed above (2.1.3.5), very little survives of the original external decoration of al-Ḥakam’s mosque: the only gates to retain traces of their original Qurʾānic inscriptions are three on the western façade, and four on the caliph’s original eastern façade (Figures 51–52, 61), now part of the longitudinal arcade at the transitional zone of al-Manṣūr’s extension.214 While these form self-contained and coherent statements in themselves, the external inscriptions both encapsulate and refer to the messages of the mosque’s internal epigraphic programme (Appendix 3A), though they may also have had direct reference to the building’s urban context and the activities that happened in front of these gates. The only two surviving inscriptions on al-Ḥakam’s western façade both quote from Sura 40: verse 3 on Gate 4 summarises the concepts of God’s mercy and judgement, and the punishments or rewards awaiting man in the hereafter, while verse 13, on Gate 3, speaks unequivocally of the revelation of the Qurʾān: ‘It is He who reveals His signs to you, and sends down sustenance from the sky for you. Yet none takes heed except the repentant’. These verses refer to others inside the mosque – 40:7, 8 and 9 are inscribed around the base of the dome of the maqṣūra’s western bay (in front of the Bāb al-Sābāṭ) – and thus sustain the themes of the mosque’s internal epigraphic programme.
However, as Calvo points out, Sura 40 is not frequently employed in mosque inscriptions: ‘called Ghāfir, “he who pardons”, or al-Mūʾmin, “the Believer”, these verses transmit the idea that all men will be submitted to Divine Justice, and insist on the value of faith as a means of salvation, and on the mercy and equity of God for those who deserve it, by following the recommendations of the Holy Book or by repenting. The verses recall the omnipotent power of God, in whom resides the final decision to save or condemn on the Day of Judgement, independent of human actions. On that day only He will judge – there are no intercessors’.215
Calvo believes that the use of these verses on the western gates had a specific reference to the institutions that lined the eastern façade of the Umayyad qaṣr opposite, which were related to the administration of justice and meting out of punishments and executions. On this façade of the palace was a gate known as the Bāb al-ʿAdl, or Gate of Justice, located opposite the mosque gate now known as Puerta de San Miguel, but known in Umayyad times as the Gate of the Mosque since it was used by the amirs to enter the prayer hall until the sābāṭ was built. The amir ʿAbd Allāh sat in this gate on Fridays, after the communal prayer, to listen to the people and impart justice. At the southern end of the palace walls was located the Bāb al-Sudda, which is where the death penalty was carried out, where the crosses and spikes for the public display of the bodies and heads of criminals, heretics and rebels. The heads of the apostates Ibn Ḥafṣūn and his two sons were displayed here after their rebellion was suppressed with the taking of Bobastro in 316/928, an event which had a great repercussion for the Andalusi state – a few months afterwards ʿAbd al-Raḥmān III assumed the title caliph. Also found on the western side of the mosque was the Dār al-Ṣadaqa, or Alms House, over whose gate was a chamber, used on at least one occasion (in 360/971) for the public derision of a malefactor afterwards led to prison.216
Apart from the Bāb al-Wuzarāʾ (now known as the Puerta de San Esteban), built in 785–6 and the only surviving gate from Cordoba I, whose epigraphic content is entirely historical, none of the other gates on the western façade preserves its original inscription or appearance.217
Though now very fragmentary, three of the four gates on al-Ḥakam’s original eastern façade retain traces of their inscriptions (Gates 1, 2 and 4). These again highlight themes of Divine Justice and the need to follow the precepts of the revealed truth (the Qurʾān), or suffer the penalties after death; the phrase ‘He forbids all shameful deeds and injustice and rebellion’ (16:90 on Gate 2) again seems to allude to the perils of heterodoxy. Calvo also associates this verse with the exhortation to practise charity and do good, in particular the phrase ‘God commands justice, the doing of good, and liberality to kith and kin’. Again this seems to relate to actual activities at the mosque. She notes that in March 974, the chief qāḍī of Cordoba had a proclamation read from the gates of the mosque, reminding the population of Cordoba of the needs of the city’s poor and to give zakāt, threatening them with the judgement of the Last Days.218
Calvo also discusses the anti-Christian content of 19:35 on Gate 1, which denies the divinity of Christ: ‘It is not befitting to (the majesty of) God that He should beget a son. Glory be to Him! When he determines a matter, He only says to it ‘Be’, and it is’. This very same verse forms the centrepoint of the anti-trinitarian message of the Dome of the Rock in Jerusalem, as interpreted by Oleg Grabar.219 Key to Calvo’s argument is the fact that three Qurʾānic quotations with anti-trinitarian import are used in different locations along the qibla wall at Cordoba: 6:101 (‘How should he have a son when He had no consort?’), in the mosaic inscriptions around the window grille above the Bāb al-Sābāṭ; 19:35 on the Bāb Bayt al-Māl; and 112 (Sūrat al-Ikhlāṣ) in combination with 33:56 on Gate 1 of the ʿĀmirid eastern façade (Figure 68). Indeed, the only other place where Calvo has found verses 19:35, 112 and 33:56 in association with each other is at the Dome of the Rock, which she implies is a deliberate reference.220 She goes onto argue that Christianity posed enough of a threat in the late-tenth century to have inspired a strand of the mosque’s epigraphic programme, focusing on apostasy as another branch of heresy.221 There is no evidence that Sūrat al-Ikhlāṣ was used in any of the inscriptions of al-Ḥakam’s maqṣūra, so if this combination of verses was intended as a deliberate reference to the Dome of the Rock or to an earlier Umayyad policy to deny the divinity of Christ, then this was an association conceived by al-Manṣūr, or the fuqahāʾ with whom he developed this epigraphic programme, since the verse is employed on one of the ʿĀmirid gates (Gate 1). Calvo gives no evidence of anti-Christian sentiment or policy relating specifically to the ʿĀmirid period, but such statements may have added weight and legitimacy to the increasingly intense campaigning against Iberia’s Christian kingdoms that al-Manṣūr had undertaken in the period preceding the mosque construction. Such clear statements as that given by Sūrat al-Ikhlāṣ – ‘Say: He is God, the One and Only … He begetteth not, Nor is he begotten; and there is none like unto Him’ – could offer further arguments to the justification of jihad.
The specific interrelationship between the mosque’s internal and external inscriptions allowed readers on the outside to understand the meaning of the internal epigraphic programme ‘in a nutshell’. Again this evokes the organisation of the mosaic inscriptions at the Dome of the Rock, where the outer face of the octagon is divided into six ‘miniature composition[s] encapsulating the major themes of the inscription on the inner face’.222 Several of the inscriptions on the ʿĀmirid façade – those words which al-Manṣūr chose to write on his building – are similarly intertextual, since five out of the seven ʿĀmirid gates bear Qurʾānic passages that can be directly related to those adorning the key ceremonial foci of al-Ḥakam’s extension. While verses 3:18–19 on Gate 2 (Figure 69) could be read as another borrowing from the Dome of the Rock,223 it is more likely that its direct source was the inscription on Gate 4 of al-Ḥakam’s original eastern façade (Figure 65), or that which runs along the top of the central nave of the maqṣūra.
The first attempt to record the mosque’s epigraphic programme was made by Amador de los Ríos in 1879, and while his renderings of the historical inscriptions have since been supplemented by Lévi-Provençal and Ocaña,224 it is only recently that the surviving Qurʾānic inscriptions on the eastern façade have been reread or republished.225 Furthermore, Amador de los Ríos’ work was only partial, and it is often difficult to understand which inscription he is discussing: of the seven gates constructed by al-Manṣūr, he published the inscriptions of only five, referring to them by the names they bore in 1879, and without mentioning whether his ordering began at the north or south end of the eastern wall. Fortunately, an Arabic transcription of the inscription is always included, and it is therefore possible to identify which of the five ʿĀmirid gates he published, and to verify his readings. The Qurʾānic citations on the two remaining gates (Gates 1, Figure 68; and 6, Figure 73) have now been identified (Appendix 3B).
Though the decoration around these gates has been extensively restored, the inscriptions were left untouched. They seem to have been longer than the inscriptions on the western façade, and each inscription was originally organised in four zones (Figure 82):
-
1) The main passage began with the basmala and followed the archivolt of the horseshoe arch above the door; in most cases this inscription occupies only the top half of the tympanum, but on Gates 4, 6 and 7 (Figures 71, 73, 74) the inscription runs around the whole circumference of the archivolt.
-
2) The passage continued in a short horizontal frieze running across the middle of the tympanum, and frequently ended with the taṣliyya (wa ṣallā allāh ʿalā muḥammad) – an exaltation of the Prophet Muḥammad – or other blessing, such as ḥusn allāh on Gate 5 (Figure 72). As Calvo discusses, these may reflect pious expressions in use in daily life at this period.226
-
3) Another horizontal frieze, above the arch, generally contained a different verse, beginning again with the basmala (Figures 71, 72).
-
4) Originally there would have been a fourth section, above the frieze of five blind arches (Figures 70–73), but no inscriptions survive on the eastern façade in this location.
Safran considers that the verses used on these gates ‘sacralised the building and marked it territorially’, and ‘were probably not meant to ‘missionise’ so much as to encourage the Muslims … in their faith’.227 While this is no doubt true, other levels of meaning can also be read in the ʿĀmirid inscriptions on the Cordoba mosque. There are strong statements of the primacy of Sunni Islam: Gates 2 and 6 both assert the shahāda, the Muslim credo, and 3:19 states ‘The only true faith in God’s sight is Islam’. Verse 33:56 (Gate 1) highlights the importance of the Prophet Muḥammad as God’s messenger (‘God and His Angels send blessings on the Prophet: O ye that believe! Send ye blessings on him, and salute him with all respect’), and as Calvo observes, the taṣliyya is particularly frequent on the ʿĀmirid gates (occurring on Gates 4 (twice), 5, and possibly 7). This prayer exalts the Prophet Muḥammad and was linked to the belief of Sunni Muslims in Muḥammad’s mediation between God and his community; it was also recommended to recite this phrase on entering and exiting a mosque.228 Another theme here is the unity of God, which may carry an anti-Christian message: Gates 2, 3, 5 and 6 repeatedly state this theme, for example 59:22–23 (Gate 6), ‘He is God, besides whom there is no other deity’. Another theme is the revelation of the Qurʾān, including 3:3 (Gate 3), ‘He has revealed to you the Book with the Truth, confirming the scriptures which preceded it; for He has already revealed the Torah and the Gospel for the guidance of mankind’; and the vivid verses of 59:21–23 (Gate 6), ‘[Had We sent down] this Qurʾān on a mountain, verily thou wouldst have seen it humble itself and cleave asunder for fear of God …’.
Standardised arrangement of the decoration and placement of the inscriptions on the gates of the ʿĀmirid façade
© Matilde GrimaldiThe main message insisted on here may be the punishments due to those who reject this message, for example, 3:4 (Gate 3): ‘Then those who reject Faith in the Signs of God will suffer the severest Penalty’; or 3:192 (Gate 7): ‘Our Lord! Any whom Thou dost admit to the Fire, truly Thou coverest with shame and never will wrong-doers find any helpers!’ Similar warnings are found on Gates 2 and 5, while other verses emphasise the corollary, that is, the rewards to (Sunni) Muslims on the Day of Judgement. As well as verses on Gates 3 and 7, most significant are two long inscriptions on Gate 4 (neither of which was used in al-Ḥakam’s extension, as far as we can reconstruct): 43:68–71 (‘My devotees! No fear shall be on you that Day, nor shall ye grieve, (being those) who have believed in Our Signs and bowed (their wills to Ours) in Islam. Enter ye the Garden, etc’); and 36:78–79 (‘“Who can give life to (dry) bones and decomposed ones?” Say: “He will give them life Who created them for the first time!”’). The eschatological emphases here have been interpreted by Calvo as a possible reference to the practice of saying the prayer for the dead (ṣalāt al-janāʾiz) in front of these gates as, in the Maliki rite, dead bodies were not brought inside the mosque for the funeral oration.229 The particular emphasis on the rewards of the afterlife expressed on Gate 4 might suggest that it was in front of this gate that the funeral cortège stopped for the recital of the funeral prayers.
Finally, there is the unique use of Sūrat al-Ikhlāṣ (112) on Gate 1, which may have the anti-Christian message that Calvo reads into it; on the other hand, as she also notes, Sura 112 is very frequently used in mosques and in epigraphy of a public character. Most significantly, however, it was used as the Umayyad motto on the first coinages, both gold and silver, issued by ʿAbd al-Raḥmān III after his assumption of the caliphal title, in 316 and 317/929; this evoked the Syrian Umayyad use of this motto after the seventh-century coin reforms of ʿAbd al-Malik.230 Even after new formulae were developed for Andalusi coins, did this Sura retain an oral association with the Umayyad dynasty, which al-Manṣūr evoked through its employment on the most important gate of his eastern façade – that giving access to the qibla zone of his new prayer hall?
Thus the inscriptions on these gates do ‘missionise’. But perhaps the most important level on which they should be understood is their reiteration of the mosque’s internal messages. It is important to note that only one of al-Manṣūr’s gates repeats the surviving inscriptions of al-Ḥakam’s original eastern façade: Gate 2 (Figure 69), which repeats 3:19 from Gate 4 of al-Ḥakam’s original eastern façade. The words, ‘Those to whom [the scriptures] were given disagreed among themselves’, is no doubt a warning against the divisiveness of religious heterodoxies. So while the appearance of al-Ḥakam’s gates was recreated on the eastern façade, their inscription programme was conceived anew. Another case (Gate 1) reflects similar themes to those expressed on the parallel gate from the old eastern façade (the Puerta del Chocolate), though these are the anti-Christian messages which Calvo argues for, and perhaps the placement of both at the qibla wall might suggest a special case. Otherwise, the gates directly quote passages from the central nave of al-Ḥakam’s maqṣūra. Gates 2, 3 and 7 between them repeat about half of the carved inscription which runs along the top of the two arcades of the central nave of the maqṣūra, and the archivolt inscription of Gate 5 also evokes this area by continuing a citation from Sura 39. Gate 6 repeats 59:23, with its extremely powerful statement of revelation, from the mosaic inscription above the entrance to the mihrab chamber itself. This method of ‘quotation’, by the exterior of the interior, naturally served to reprise the anti-heterodox statements made by al-Ḥakam in his mosque construction twelve years before. As such, al-Manṣūr made a clear demonstration of political and religious alignment with the Umayyad dynasty’s stance on heterodoxy, evoking architecturally the political action he had taken in purging al-Ḥakam’s library. The employment of Surat al-Ikhlāṣ, with its direct association with the Umayyad dynasty, underlines this statement of alignment and continuity.
Perhaps it is more significant to consider how those passages which are not quotations from al-Ḥakam’s maqṣūra, and which may therefore represent al-Manṣūr’s own statements on the subject, expand on the sentiments of the caliph’s message, and contribute new strategies to the debate. It is important to bear in mind, however, that the mosque’s extant epigraphic programme has not been fully recorded, and future research may reveal that these passages also reprise the monument’s internal messages. As noted above, in the introductory discussion to the ʿĀmirid mosque, al-Manṣūr also opened two new gates in the expanded courtyard wall, and evidence of painted inscriptions has been found on the interior of the courtyard façade of his extension, but as yet we have no information about what these inscriptions said.
Two of the eastern gates contain verses that cannot so far be identified with any extant Qurʾānic passages used inside the mosque. Verses 43:68–71 (Gate 4, Figure 71) enumerate the luxuries and rewards of Paradise (‘Abiding there for ever, you shall find all that your souls desire and all that your eyes rejoice in’), but make the clear statement that this is only available to Muslims (‘those who have believed in Our revelations and surrendered yourselves’, wa kānū muslimīn, that is, become Muslims). Verse 39:53 (Gate 5, Figure 72) evokes the inscription of the mihrab vestibule (where 39:47 is used), and adds new ammunition to the debate over predestination, by emphasising God’s ultimate mercy: ‘Say: “Servants of God, you have sinned against your souls, do not despair of God’s mercy, for God forgives all. It is He who is the Forgiving One, the Merciful”’. Of perhaps greatest interest are the unequivocal warnings, on Gates 4 and 5, which can be read together as a direct address to adherents of heterodoxies: for example, verses 36:78–79 – ‘He argues back with arguments, and forgets His own Creation. He asks: “Who will give life to rotten bones?” / Say: “He who 71, 73, 74 first brought them into being will give them life again”’231 – warn against the perils of philosophical reasoning, on which Muʿtazilism and Masarrism were based, which leads the believer from the straight (and narrow) path of orthodox Islam. The immediately preceding verse, not quoted inside the mosque but which would have been evoked in the minds of all Muslims familiar with the Qurʾān, is also pertinent: ‘Does man not see We created him from a little germ? Yet he is flagrantly contentious’. A final warning is found on Gate 5, which might be said to encapsulate the mosque’s entire epigraphic programme: ‘This is a warning to mankind. Let them take heed and know that He is but one God. Let the wise bear this in mind’ (14:52).
The painted inscription on the transverse arcade running parallel to the qibla wall – which is probably an ʿĀmirid intervention, as discussed above (2.1.4.1) – also contains verses (10:106–7) that are not used elsewhere in the mosque, but again they reaffirm messages about predestination which are found in al-Ḥakam’s epigraphic programme (‘If God do touch thee with hurt, there is none can remove it but He …’).232 This arch is supported on the interior of Gate 2 on the western façade and its painted inscription would have been legible as soon as a worshipper crossed the threshold into the mosque. If the transverse arcade was an ʿĀmirid intervention, were these painted verses directed for Hishām’s attention, or at detractors who might not have been happy with ʿĀmirid de facto rule? It is also likely that the painted inscriptions on the internal courtyard façade of al-Manṣūr’s extension likewise recapitulated the same themes as expressed in al-Ḥakam’s maqṣūra (Figure 67).
The Qurʾānic inscriptions on the eastern façade hold the key to understanding al-Manṣūr’s extension to the Great Mosque of Cordoba. The extant Qurʾānic passages that adorn these gates combine to make a clear statement of religious and hence political alignment with the Umayyad dynasty’s stance on heterodoxy, which in the tenth century presented a real threat to the unity of the umma and the caliph’s position at its head. The inscriptions demonstrate that al-Manṣūr had taken up the task, expressed by ʿAbd al-Raḥmān III through the edicts against Masarrism and by al-Ḥakam through the mosque’s internal epigraphic programme, of suppressing heresy and maintaining the role of the caliph as rightly-guided, Sunni ruler of the state. The new verses, those which are not found in al-Ḥakam’s extension, constitute a reaffirmation of this stance. Lastly, by proclaiming, in every aspect of his mosque extension, the continuity of his regime with that of the Umayyads, al-Manṣūr emphasised that it was in his hands, rather than theirs, that the ruler’s powers – especially for the protection of religion – now lay.
The Cordoba Mosque’s Qurʾānic inscriptions conformed to a deliberate and coherent programme with clear and targeted meanings that highlight the precepts of Maliki Sunni Islam. Anderson and Pruitt have seen in these targeted meanings a heightening of polemic and competition with the Fatimids at the end of the tenth century.233 Al-Manṣūr’s mosque extension coincided with the construction by al-Hākim (r. 996–1021) of his new congregational mosque in Cairo in 990, to which large-scale, exterior inscriptions were added in 1002–3 when he added two towers and a portal to the building’s main façade, though they were covered over again in 1010. If the Mosque of al-Azhar originally bore exterior inscriptions, these no longer exist, and it seems that monumental ‘public texts’ were really introduced during al-Hākim’s reign, after which this became a defining trait of Fatimid architecture and urbanism.234 That they were employed to express sectarian messages is shown by the large-scale anti-Sunni ‘cursing of the Companions’, inscribed throughout the city of Cairo-Fustat in colours and gold, on mosques, houses, and bazaars, sponsored by al-Hākim in 1005.235 However, as Anderson and Pruitt have interestingly realised, these Umayyad ‘public texts’ at the Great Mosque of Cordoba predate the first symbolic use of inscriptions by their Fatimid rivals. While the Cordoba inscriptions on al-Ḥakam’s and later al-Manṣūr’s façades primarily addressed local political and religious controversies, their rejection of non-Sunni heterodoxy also implied a rejection of Fatimid claims to religious and political authority. The direct competition between these two regimes’ approaches to ‘public texts’ became most clearly manifest through the vacillations of the Andalusiyyīn minbar and its recarved backrest (Chapter 7: 1.1). And as discussed at the end of Chapter 4, there may have been a Fatimid influence on the meaningful naming of gates and buildings at al-Madīnat al-Zāhira and in the significance of titulature adopted in the ʿĀmirid period. A more detailed comparative study of the specific Qurʾānic citations employed in contemporary Umayyad and non-Umayyad (or Sunni/non-Sunni) monuments would be interesting for the correspondences and differences this might reveal: for example, Sūrat al-Nūr (24:35), favoured by the Fatimids in both text and image as a metaphor for the hidden imam, does not seem to be much used on Umayyad monuments.
In sum, the statement that al-Manṣūr carefully constructed through his extension to the Great Mosque of Cordoba was that he was aiming not to usurp the rule of the Umayyad caliphs, but rather to take up their mantle as the protectors of orthodox Islam. This role was also expressed in his leadership of jihad, his censorship of suspect books in al-Ḥakam’s library, and his possible monumentalisation of a space for teaching fiqh within the heart of the caliphal mosque (2.1.4.2 above). As ʿAbd al-Raḥmān III had ‘refined the definition of his caliphal authority in the context of his campaign to suppress the heresy of Ibn Masarra’,236 so al-Manṣūr defined his de facto royal role as defender of the faith to include the Maliki sunna. He asserted this statement by consciously relating his extension symbolically and semantically to the mosque’s ceremonial and Umayyad heart – al-Ḥakam’s maqṣūra – by employing a variety of imitative devices. On the simplest level, his extension imitated the appearance of the earlier mosque by, for example, adopting the same alternating pattern of coloured columns, or by continuing the transverse arcade at the interface of Cordoba II and III into Cordoba IV, for the sake of symmetry. His architects were often highly ingenious in the solutions they invented to make Cordoba IV look exactly like the rest of the mosque. Most visibly, and perhaps therefore most importantly, al-Manṣūr imitated the form of al-Ḥakam’s original eastern façade, and clearly asserted his political and religious alignment with the Umayyads and the Maliki fuqahāʾ through the deliberate selection of Qurʾānic passages which restated the messages of the mosque’s internal epigraphic programme. In every respect, al-Manṣūr was claiming continuity with the Umayyads, while emphasising his control of the new political landscape.
At the same time, his architects were subtly innovating to correct problems with the older building, such as the lack of illumination, or the level of damping in the elevation of the columns and arcades. Interventions within the maqṣūra, such as the addition of the transverse arcade along the qibla wall, or the solutions adopted to compensate for disparities between Cordoba IV and the older mosque, make a decorative virtue out of practical necessity; but they also see the introduction of new and innovative features into Andalusi architecture, such as the early use of the pointed arch, the maturity of the multi-lobed arch, the use of public texts. The possible creation of a tribune at the heart of the mosque also speaks to the wide range of ceremonial activities that took place in this monument.
While there was no doubt a sense in which al-Manṣūr was seeking to ‘perpetuate his name and his government into the future’,237 his extension was surely a statement intended for a contemporary audience. The mosque took six or seven years to build, from 987/8–994, and over this long period the population of Cordoba was bound to be curious to see what would emerge when the scaffolding came down – perhaps even concerned, since we must not forget that al-Ḥakam’s extension had been highly controversial. We can speculate that this audience was relieved to discover that the new was essentially the same as the old, and that their new ḥājib was willing and able to fulfil his obligations as a ruler to undertake pious works for their benefit, and to defend the unity of the umma against heresy.
Al-Manṣūr’s deliberate use of epigraphic messages within and without the mosque shows his awareness of the impact such visual messages could have. Meaningful messages were also written and encoded on objects produced at a smaller scale; we will see this in the following chapters as we now turn to the discussion of the luxury arts of the ʿĀmirid period.
Kennedy 1996, 119.
Marfil 2003, 78.
For example, Torres Balbás 1957, 571–579; Gómez-Moreno 1951, 162–165; Creswell 1969, II: 144; Ewert 1981, 85–89; Dodds 1992; Singul 1997, 321–323; Pérez Higuera 2001; Puerta Vílchez 2013a, 52, who refers to al-Manṣūr throughout as ‘el caudillo’ and discusses the ʿĀmirid extension in just three sentences. Marfil 2003 is the exception to this rule.
Gómez-Moreno 1951, 162.
Torres Balbás 1956, 578.
Marfil 2003, 78.
Al-Maqqarī 1968, I:561–562, cited in Rius 2000, 174. The same passage tells us that, in wanting to change the qibla, al-Ḥakam was ‘in accordance with the astronomers (ahl al-ḥisāb), among whom were imams of great authority’.
Rius 2001, 426: ‘the fuqahāʾ counselled a diplomatic position: if it was possible to convince all the believers that a change in the qibla was necessary, then such a task could be undertaken, but if this question came to be a cause of disputes, then it had to be rejected and the original orientation maintained. The qibla was secondary to such a fundamental concern as the unity of the community of believers’.
Calvo 2010, 177: n. 83. This oath took place between the date of the placement of the imposts of the mihrab arch (dhū’l-ḥijja 355/Dec 965) and his death (dhū’l-qaʿda 355/17 Nov 966). On Mundhir ibn Saʿīd al-Ballūṭī, see Fierro 2010.
Dodds 1992, 24.
Dodds 1992, 24.
Hernández Jiménez made soundings through the 1.58m thick casing, and found infilled windows with marble capitals, shafts and bases, and other carved marble elements which match the colour and style of those at Madīnat al-Zahrāʾ, and the infilled passages of two helicoidal stairways: cf. Hernández Jiménez 1975. In recent years, these have been reopened and visitors to the monument can climb up the space between the minaret and its casing.
On which see Flood 2001. The Damascus and Cordoba mosques are specifically related through the ‘myths’ that both builders received gifts from the Byzantine Emperor of glass tesserae and a mosaicist to install them: on Cordoba, cf. Bayān II:253 [translation, 392]; on Damascus, see Flood 2001, 22–24; Gibb 1958; and George 2021. Both mosques started out life as a church which the Muslims shared with the local Christians: cf. Ocaña 1942; Dodds 1992, 11.
Safran 2000, 65.
Flood 2001, 194.
Dodds 1992, 15.
‘Al-Ḥakam I’, Encyclopaedia of Islam, 2nd Edition: ‘The reign of al-Ḥakam I was almost entirely devoted to suppressing the repeated rebellions which were ceaselessly breaking out on the three frontiers of Toledo, Saragossa and Mérida’.
Al-Ḥimyarī 1938, 155–156; al-Maqqarī, I:360; Torres Balbás 1957, 465–470. When its construction was finished, the caliph ceremonially visited it, inspected it by going up one flight of stairs and down the other side, and prayed two rakʿas in the mosque: Ecker 1992, 44.
Ecker 1992, 45.
On the minaret form in al-Andalus, see Bloom 1989, esp. 99–124; Bloom 2013, 137–188. See also Ecker 1992, 30–46, for a discussion of Bloom’s theory in specific relation to ʿAbd al-Raḥmān’s minaret.
Ecker 1992, 73.
Terrasse 1968, 12–14.
Bayān II:249 [translation, 385–386]. On this extension, see Torres Balbás 1957, 477–569.
Bayān II:254 [translation, 393]. Fernández-Puertas was convinced that, from the organisation and depth of the foundations at the interface between Cordoba II and III, uncovered in Hernández Jiménez’s excavations (viz. his 1961 publication), the earlier mihrab ‘was a chamber with a probable horseshoe-arched plan’, which projected from the building’s south wall as a salient, and had a horseshoe-shaped entrance arch: Fernández-Puertas 2000, 229–230 and 235–238.
Grabar 1988, 115.
Abad 2009, esp. 13–14, figs. 4 and 5; Abad 2013, figs. 1 and 2.
On the Qayrawān maqṣūra, see Golvin 1970, 233.
Abad 2009, 14.
Ruiz Souza 2001b, 436.
Abad 2009, 17–18. Nieto 2005, 95–118, has no doubt that the Capilla Real’s cupola was constructed in the fourteenth century when this space came to house the remains of Fernando IV of Castile (d. 1312) and later his son Alfonso XI (d. 1371). The high wall which surrounds the bay would have been constructed then to provide structural support for the dome. The style of the dome is Almohad, so either the fourteenth-century architects constructed it in an archaising manner, or there remains the possibility that it was one of the restorations undertaken by the Almohads in 1162, when Cordoba was briefly their capital (p. 109).
Abad and González 2018.
Ruiz Souza 2001b, 441.
Ruiz Souza 2001b, 440.
Abad 2009, 20.
Interestingly, Terrasse 1932, 101–102 observes, ‘Il était dans le palais des Califes des mosaïques byzantines, faites par des artistes venus de Constantinople … Toutefois un fragment d’enduit couvert de mosaïque d’émail, récemment retrouvé dans les fouilles, ne permet plus de douter de l’éxistence de ce renseignement’. Vallejo 2010, 364, pl. 307, mentions small fragments of glass mosaic found at various sites in the palace quarters at Madīnat al-Zahrāʾ, in similar colours to the tesserae used at the Cordoba mosque. He surmises that they might have formed part of the decoration of arches or domes.
The Prophet’s mosque was rebuilt by the Umayyads in 88–91/707–710, by the future ʿUmar II, under the orders of al-Walīd I: see Fierro 1991, 134. Huge sums were spent on it, and al-Walīd sent gold, marble and glass tesserae for its decoration. The evidence for a programme of mosaic decoration on the qibla wall, featuring ‘representations of gardens and palaces’ in the mode of the Damascus mosaics, is discussed by Khoury 1996, 90–94. For recent scientific analysis of the mosaic tesserae, see Gómez-Morón et al 2021.
See Flood 2001, 15–56. Calvo 2000, 25–26, interprets the mosaic-covered ante-mihrab dome in Cordoba III as representing the ‘celestial sphere, … the actual, physical representation of the baldaquin of the Throne of God’.
See Marçais 1965. Stern 1976, plate 38, features a pair of small birds, facing each other across the ‘rinceaux’, though he does not discuss this interesting figurative anomaly in his book.
The use of veined marble veneer panels in the mihrab of Cordoba III may have been another visual reference to Syrian Umayyad mosques, which extensively employ this form of decoration, in particular the Dome of the Rock and Al-Aqṣāʾ: see Milwright 2005.
On the motif of the scallop shell in niches and mihrabs in Umayyad monuments, see Flood 2001, 53–54.
Ewert 1987, 189.
I argue this in Rosser-Owen 2009.
Safran 2000, 67.
Hillenbrand 1986a, 182.
Marfil 2003.
Bayān II:307 [translation, 477].
Dhikr Bilād I:37 [2:43–44].
Souto 2007, 105. At p. 117, he talks of the ‘tiránico silencio constructivo’ imposed in this last extension.
Dhikr Bilād I:96 [II:124], I:182 [II:193] which says ‘al-Manṣūr began work on Hishām’s extension in the Great Mosque of Cordoba’; cf. al-Ḥimyarī 1938, 157, and al-Maqqarī, I:220, 227, who say al-Manṣūr ‘added to it by order of the caliph’.
Lévi-Provençal 1931, 27–28 (#20), plate 6d; Souto 2007, 115, 127–8 (inscription no. 11).
Lévi-Provençal 1931, 27, read the last visible word of the inscription as fataiy, i.e. the dual form of fatā, and wanted to associate this commission with the two fityān, Fāʾiq al-Nizāmī and Jawdhar, the second of whom was assassinated in 977, implying a date before that for this inscription. Juan Souto was more cautious: ‘No me atrevo yo a tanto’. Other pairs of fityān are known to have collaborated on commissions – Badr and Ṭarīf on the Girona casket, for example – and it is also possible that the final legible word reads the plural form (fityān) and not the dual.
Martínez Núñez and Acién Almansa 2004, 128: ‘Exista ningún indicio de intervenciones por parte del califa Hishām II. Frente a lo que sucede en el caso de al-Ḥakam, Hishām no aparece mencionado en ningún epígrafe, ni como heredero designado ni como califa’.
Bayān II:307–308 [translation, 477–478].
Various attempts have been made to estimate the size of Cordoba’s population: see Torres Balbás 1955; Bulliet 1979, 114–127, graphs 20 and 22; Glick 1979, 34–35, fig. 1; most recently, Chalmeta 1992.
Ibn Khaldūn 1934, 241; HEM II:266; Idris 1964, 80.
Echevarría 2011, 188. She comments further that the part used ‘corresponded to the ‘fifth’ allotted to the caliph’, without providing a reference.
HEM II:229, n. 2, citing the biography of al-Manṣūr in Ibn Bashkuwāl’s Kitāb al-Ṣila, no. 562.
Juan Souto wrote on this phenomenon several times throughout his career: Rodríguez and Souto 2000a, Souto 2001, 2002a, 2002c, 2004, 2010a, 2010b. There are some 150 columns in Cordoba III and 350 in Cordoba IV, and these numbers can be trebled to provide an approximate total number of 1500 architectonic supports (columns, capitals and bases). This means that on average there is a ‘lapidary sign’ on every second column/capital/base. In reality, however, more than one name or sign occurs on the same piece, or the same name occurs more than once. In 2010b, Souto’s last word on this was (p. 49) ‘son las [marcas] de otros tantos autores, o bien la del autor junto con la del aparejador, el maestro, el aprendiz, el jefe, el transportista, la cantera, repeticiones …’.
Al-Maqqarī, Analectes, I:359: ‘And the people considered that the most beautiful thing about the construction of this ʿĀmirid extension was the sight of captured Christians, from the land of Castile and elsewhere, bound with iron and employed on the work instead of Muslim labourers, out of scorn for polytheism and for the glory of Islam’. However, as Ocaña 1986, 66, himself points out, these signatures were carved by ‘a perfectly qualified personnel’, who were literate in Arabic, which seems unlikely for Christians just brought from the wars. These chain-gangs probably engaged in, literally, the dirty work.
Ocaña 1981; reprinted 1986.
Souto 2010b.
Souto 2010a, 221 (2.38: Khalaf).
See the index provided in Rodríguez and Souto 2000a, 384–387 (this has been checked against the index provided in Souto 2010a). The full 25 names are: Aflaḥ/Aflaḥ al-Farrāʾ; Badr; Badr ibn al-Ḥ[…] (?); Ibn Badr; Bāshir; Bishr; Bushrā; Ḥakam; Khalaf/Khalaf al-ʿĀmirī; Khayra; Durrī; Rashīq; Rizq; Saʿāda; Fatḥ; Faraj; Qāsim; Kāḥ; Mubārak/Mubārak ibn Hishām; Muḥammad; Masʿūd; Maysūr; Naṣr; Ibn Naṣr; Yaḥyā.
This information is collated and discussed in Souto 2001, Souto 2010a. The 13 names are: Aflaḥ/Aflaḥ al-Farrāʾ; Badr; Khalaf/Khalaf al-ʿĀmirī; Khayra; Durrī; Rashīq; Saʿāda; Fatḥ; Faraj; Muḥammad; Maysūr; Naṣr and Ibn Naṣr.
Grabar 1987, 106, gives the widths of all 19 aisles from west to east as follows (in metres): 5.35; 6.86 (× 4); 7.85 (the wider central nave); 6.26 (× 4); 5.35 (× 2, the second of which is the first aisle of Cordoba IV); and 6.86 (× 7). The average width of the aisles in Cordoba IV is thus 6.14 m, which is narrower than those in rest of mosque by about a metre (av. width 7.2 m).
Ewert 1968, 6 and figure 34. The width of the new extension was 49.12 m, making a total width of 128.41 m for the north and south walls. Fernández-Puertas 2000 theorises that the mosque’s architects consistently used the Pythagorean geometric system, which resulted in the building always doubling itself, a system which al-Manṣūr continued in his extension.
These calculations are from Marfil 2003, 78.
Echevarría 2011, 188.
Bayān II:308 [translation, 478–479].
Marfil 2003, 79–80. I am extremely grateful to Pedro Marfil for taking me under the mosque in November 2009, to see this excavated zone.
On which, see Lévi-Provençal 1931, 8–9 (#9).
Ewert 1968, 6.
Echevarría 2011, 188.
Marfil 2003, 78.
Marfil 2003, 79.
See the nineteenth-century drawing of a hypothesised elevation reproduced in Ruggles 2009, Figure 58. Ibn ʿIdhārī (Bayān II:308 [translation, 478]) says al-Manṣūr merely enlarged the cistern. He also says it was as large as the perimeter of the courtyard, which is not substantiated by the physical remains.
Ruggles 2009, 96: ‘A larger prayer hall and courtyard meant more worshippers and more trees, both of which demanded water, and so … a very large deep cistern was excavated in the courtyard floor’.
For details, see Torres Balbás 1957 [1996], 579–582; Gómez-Moreno 1951, 165; Marfil 2003, 83–4. Gómez- Moreno notes that, ‘in modern times’, the ʿĀmirid cistern was converted into an ossuary for the cathedral.
King 1972; Calvo 1999. I am also grateful to Asunción González Pérez for sharing with me her unpublished MA thesis on this subject: González Pérez 2009.
Gómez-Moreno 1951, 165.
Vallejo 2007, 10.
Vallejo 2010, chapter on ‘La infraestructura hidraulica’. In a personal communication dated 6 September 2020, he wrote ‘En Madinat al-Zahra no hay cisternas, pero existen diversas construcciones lineales que cubren sus tramos cuadrados con bóvedas de arista. También alguna letrina, de planta cuadrangular, se cubre de la misma manera’.
See
Creswell and Allan 1989, 284–5.
Ruggles 2009, 96, says that the ʿĀmirid cistern ‘was a classic Andalusian type with nine vaulted bays arranged in a square’, but offers no comparanda.
Pedro 2003, 83.
The remains were found at 23 Calle Magistral González Francés, next to the Hotel Conquistador. See Montejo 1998, 253 and Montejo 1999, 209.
Montejo 1999, 211. The mīḍaʾa is oriented NW–SE.
For the details of the excavation of this mīḍaʾa, see Marfil 1999, 187–189, plates 5–7.
Torres Balbás 1957, 579.
Montejo 1998, 254.
Dhikr Bilād I:40 [II:46]; cf. Pinilla Melguizo 1998, who analyses this passage and produces a new translation which takes into account the archaeological remains (p. 231).
Montejo 1999, 219, says that documentary evidence from 1333 locates an Arab ‘lavatorio’ in the Corral de Cárdenas, on the western side of the mosque, to the north of the caliphal palace, where the Hospital Mayor de San Sebastián was built in the early sixteenth century. This has not been verified archaeologically. The northern ablution hall cannot be located with certainty either, and Montejo suggests a location between Calle Céspedes and Calle Velázquez Bosco.
These details are condensed from Montejo 1998 and 1999.
According to Montejo 1999, 216, the surviving fragments of the fountain bowl indicate that it had various holes, and it was probably designed, like the Fuente de los Leones in the Alhambra, to feed out of different spouts around the fountain.
Montejo 1998, 254. It is likely to have had a similar form to the ablutions hall of the Almohad mosque in Seville: Vera Reina 1995, 161–166.
Montejo 1998, 255, and 1999, 219.
Bayān II:256 [translation, 396–397]; Torres Balbás 1957 [1996], 579. According to Anderson 2013, 112, the refurbished aqueduct was the Fontis Aureae. This presumably now fed al-Manṣūr’s cistern.
Bayān II:256 [translation, 397], quoted in Torres Balbás 1957 [1996], 579–580.
Cf. Bayān II:307–309 [translation, 477–479].
Torres Balbás 1956, 578; Ewert 1968, 6.
Cited by Marfil 2003, 78.
Marfil 2003, 79.
Marfil 2003, 80.
Ruiz Souza 2001b.
Marfil 1998, 252–3. Unfortunately he includes no image, but Fernández-Puertas 2009, 195, illustrates an internal and external view in plates 97 and 98.
According to the inscription still affixed to the Puerta de las Palmas (Lévi-Provençal 1931, 8–9 (#9), and Répertoire, IV: 141–2), in 346/957–8 he enlarged the courtyard and reinforced the sanctuary façade, by adding a second, outer pier to the original piers of Cordoba I.
Torres Balbás 1957 [1996], 574. Marfil 2003, 80, notes that the distance between the first and second column from the courtyard is 2.26 m; between the second and third it is 2.29 m.
This information comes from Pedro Marfil (personal communication, November 2009); see also the photograph published (without commentary) in Marfil 2003, 83. While red pigment is clearly visible, it is not possible to make out any trace of inscriptions. According to Marfil, these spaces were fully documented and measured with laser scans, in order to facilitate the reading of the inscriptions. Unfortunately, I have not been able to obtain further information from him about the results of this survey, since he first kindly shared this information with me.
It is only mentioned in Fernández-Puertas 2000, 246.
Fernández-Puertas 2000, 226–231.
Marfil 1999, 187–189. Archaeologists excavated at exactly this transitional zone between Cordoba I and IV. Their trench covered the width of one nave, and its length was delimited by the Capilla de la Epifanía to the north and the Baroque cathedral to the south. The walls of Hishām’s mīḍaʾa were found to stretch 20 m N-S and 16 m E-W, which would bring its eastern façade to somewhere between the third and fourth buttress of Cordoba IV.
Herrero Romero 2017, 199. I am grateful to Antonio Vallejo for bringing this book to my attention.
Ruiz Cabrero 1990, 70–71.
Momplet 2012, 251–2.
Marfil 2003, 81.
Marfil 2003, 81.
Pavón Maldonado 1987.
Torres Balbás 1957 [1996], 575; Marfil 2003, 81–2.
Calvo 2010a, 180.
Torres Balbás 1957 [1996], 576–577.
Ruiz Cabrero 1990, 71-72. My thanks to Michele Lamprakos for bringing this article to my attention.
Ewert 1987, 189. For the discussion of al-Ḥakam’s extension, see Ewert and Wisshak 1981, 72–85, 89–94.
Marfil 2003, 81, says that the column shafts of violet colour were made from marmoreal limestone from the Sierra of Cordoba, while the breccia or red conglomerate was obtained from the Sierra de Cabra, in the mountains of the Subbéticas to the south of Cordoba.
According to Marfil 2003, 81, Cordoba IV employs a total of 342 columns: 44 in the courtyard wall, 32 in the transition zone with the old eastern façade, and 266 in the new prayer hall.
Ewert 1987, 189.
Ewert 1987, 189. He does not provide photographs of these few instances of more decorative capitals in Cordoba IV, however Souto 1990, 129, discusses them: this group is characterised by substituting the third row of plain acanthus leaves by an interlace of two small leaves with curly ends. I am grateful to Elise Franssen for bringing this to my attention: see Franssen 2008, 62–3.
Fernández-Puertas 1979–1981 and 1999; Marfil 2009. According to Fernández-Puertas’ drawn reconstructions, the upper frieze was formed by three independent horseshoe arches, between which were panels with carved decoration. Over time it seems these panels were also incorporated into arches, producing a five-arched arcade. The original pattern of the Bāb al-Wuzarāʾ continues in all later gates: two arches filled with carved decoration occur between three arches filled with decoration of inlaid brick.
Torres Balbás 1957 [1996], 552: ‘In the other three [gates] of the [eastern] façade, and in the corresponding gates of the western [façade] … there would have been a band of small decorative arches above the gate, as described for [the Puerta] del Chocolate’.
Herrero Romero 2017, 81–85, 87.
Torres Balbás 1957 [1996], 553: ‘The two [gates] which flank [a third gate which was redesigned in the sixteenth century] have been very much restored in modern times, copying the central part of the decoration of the [Puerta] del Chocolate, including the upper zone of small interlacing arches, of which there seem to remain no traces which could indicate their existence’.
Herrero Romero 2017, 85–92, 98.
Herrero Romero 2017, 89.
Marfil 2003, 78.
Momplet 2012, 240–1.
Abad 2013, esp. 15–20 on the transverse arcade. I am deeply grateful to Concha Abad for sharing her further thoughts and ideas about this architectonic intervention, which she attributes to the ʿĀmirid period, in a personal communication of 22 June 2015.
Abad 2013, 20. Marfil 2003, 79, noted the ‘existencia de una columnilla de la ampliación de Almanzor colocada en el oratorio de al-Hakam II’, though it was difficult without an illustration or a thorough knowledge of this part of the building to clearly understand his meaning. Abad’s study finally clarifies this.
Concha Abad, personal communication (22 June 2015).
Ruiz Cabrero 1990, 71–72.
As noted by Momplet 2012, 241–2. Compare, for example, Torres Balbás 1957 [1996], fig. 274: the pointed arches with rolls on their intradoses are marked, but not the arcade that continues to east and west.
Momplet 2012, 252.
Momplet 2012, 251: ‘No conocemos ejemplos anteriores en la arquitectura omeya de al-Andalus y son muy escasos posteriormente’. See also Torres Balbás 1957 [1996], 488: ‘El arco agudo hizo su aparación … en esta ampliación’. Torres Balbás gives a range of earlier Umayyad and Abbasid examples of the pointed arch which might have provided the model for Cordoba.
Torres Balbás 1957 [1996], 490: ‘No conocemos ejemplo anterior’. See also Momplet 2012, 252, who draws a connection with the ‘pillowed’ arch profile seen in Fatimid architecture, such as the Bāb al-Futūḥ in Cairo, constructed 1087 (see Bloom 2007a, 121–8), where it is also combined with a pointed arch.
Calvo 2000, 22–3; 2010, 161, citing Brisch 1959, esp. 184–5.
Abad and González 2018 (though published in 2020). I am extremely grateful to both authors for discussing with me their thoughts on this space, as well as sharing the unpublished text of their article at various stages of preparation. The ensuing discussion is heavily based on their findings, as I have not yet had the opportunity to visit this space in person.
Ruiz Souza 2001b, 436–7.
Concha Abad, personal communication (22 June 2015): ‘De hecho la columna de la fila longitudinal que discurría en el límite oriental del espacio se conserva y no presenta ningún elemento que haga pensar en la existencia de un arco transversal’.
Abad and González 2018, Illustrations 1 and 2.
Abad and González 2018, 233; Ruiz Souza 2001b, 437, n. 11; Ruiz Souza 2006, 19.
Souto 2007, §2.8, pp. 112–3, 125 (inscription no. 8).
Concha Abad, personal communication (22 June 2015); Abad and González 2018, 235.
Abad and González 2018, 239; personal communication from Antonio Vallejo (6 September 2020).
Abad and González 2018, 233–4 (Image 3, Illustration 3). This is only visible from inside the crypt of the Capilla Real.
Marfil 1998, 252–3. He provides no illustration of this inscription, but there is a very small image in Fernández-Puertas 2009, 196, plate 100.
Marfil 1998, 253.
Abad and González 2018, 244–5. There is little available literature on the architectural feature of the dikka, which has not been much studied. See: ‘Dikka’, Encyclopaedia of Islam, 2nd Edition; Behrens-Abouseif 1989, 54; Dickie 2011, 37.
Apud Dickie 2011, 37. He gives no reference but he is probably thinking of the same source as mentioned in ‘Masdjid’, D.2.e: ‘The platform or Dakka’, Encyclopaedia of Islam, 2nd Edition, where Maslama, Muʿāwiya’s governor in Egypt, is attributed with the invention of platforms (manābir) for the call to prayer. Pedersen (the author of this part of the article) notes that ‘this story … is not at all reliable’.
Nees 2015, 87, 141–2; Folda 1995, 441–456, pls. 10.13a– 10.13w.
Abad and González 2018, 245, referring to Eugenio de Llaguno y Amírola and Juan Agustín Ceán-Bermúdez, Noticias de los arquitectos y arquitectura de España desde su restauración (Madrid, 1829), vol. II, pp. 188–9: ‘Esta es la sala donde los santos doctores disputan nuestra ley. Alabado sea Dios todopoderoso’. As they note, it is possible that this is the space referred to by Jules Gailhabaud, Monuments anciens et modernes (Paris, 1850), vol. III, n.p., who mentions ‘a particular construction was found where questions relating to the Qurʾān were discussed’.
HEM III:126–136.
Calvo 2010a, 178.
Abad and González 2018, 248–9.
Calvo 2008, 99; Calvo 2010a, 177.
Calvo 2008, 99; Fierro 2007.
For al-Idrīsī’s description of the weekly ceremonial viewing of the Qurʾān of ʿUthmān, see Calvo 2008, 98. See also Bennison 2007b.
García Gomez 1967, esp. 168–9; HEM III:90; Lévi- Provençal 1932, 142; Bariani 2003, 211, following Ibn Ḥayyān 1965, 25–6/48–9; Bayān III:5 [translation, 13], and verses from Ibn Darrāj, cited in Bayān III:9 and Anales, §9 (pp. 48–50).
It is intriguing that apparently no verses from Sūrat al-Fatḥ were selected to be inscribed on the walls of the mosque as part of its epigraphic programme (see Appendices 3A and 3B).
HEM III:90, and Lévi-Provençal 1932, 142, says that at the end of the campaign, these ‘insignia of command were hung again on the walls of the sanctuary’.
García Gomez 1967, 168–9.
Abad 2009, 14: on one occasion, al-Ḥakam is said to have sat here with Ghālib discussing the Christian frontier.
HEM III:135.
‘After having conducted the prayer in the congregational mosque in Cordoba, contrary to his custom of not attending the Friday prayers in public, [al-Muʾayyad] directed his horse towards al-Zāhira with his mother Ṣubḥ’: Dhikr Bilād I:156–7 [II:196].
I undertook this study of the eastern façade inscriptions as part of my doctoral research, submitted in 2002, at which point I was indebted to Susana Calvo’s 2000 study of the epigraphic programme in al-Ḥakam’s maqṣūra. Since then she has published two further studies dedicated to this theme: Calvo 2008, which largely recapitulates her earlier discussion of the inscriptions, but makes suggestions about the symbolism of the maqṣūra’s decoration; and Calvo 2010a, which deals with the epigraphic programme on the mosque’s external gates, including the gates added to the east and west of the mosque by al-Ḥakam II, and the ʿĀmirid gates on the eastern façade. I have integrated the salient points of her more recent articles into my argument, and cross-referenced my own readings of the façade inscriptions with hers (see Appendices 3A and 3B).
Blair 1998, 19. Her article contains an interesting summary of the historiography on this issue. Van Berchem established the model for the compiling Islamic epigraphy, with the publication in 1894 of the first volume of the Matériaux pour un Corpus Inscriptionum Arabicarum, and since then the trend to record only those texts which ‘furnish some historical data’ has been persistent. There is no volume of the Corpus on al-Andalus; that gap was partly filled in 1931 by Lévi- Provençal’s Inscriptions, which nevertheless closely followed the Van Berchem model of a chronological ordering within geographic divisions. The work of the late Juan Souto aimed to compile the definitive corpus of Andalusi inscriptions, though again his work was concerned solely with historical inscriptions.
Grabar 1987, 126; Ettinghausen 1974, 307.
Flood 2001, 213. Moreover, the political or propagandistic use of Qurʾānic passages in monumental inscriptions of the Umayyad period has been taken as evidence for the early codification of the text of the Qurʾān. On the Dome of the Rock, for example, see Whelan 1998, 10, who says ‘there is abundant evidence from the Umayyad period that [the Qurʾān] was already sufficiently familiar to the community at large to provide easily recognisable claims to political legitimation and for religious propaganda’. Milwright 2016, 228–9, summarises the arguments for this.
Grabar 1959, 33–62; and idem. 1987, Chapter 3, ‘The Symbolic Appropriation of the Land’.
Bierman 1998.
Flood 2001, esp. 192–206, 222–225. The inscriptions, which once existed in a series of superimposed bands above a dado of marble revetments along the qibla wall, were destroyed in the post-Umayyad period, when the practice of inscribing Qurʾānic texts on the walls of mosques came to be criticised as a potential distraction from prayer, though it may also have been the result of Abbasid vilification of the Umayyads and appropriation of their key monuments: see Flood, 2001, 126, 242. Flood has recovered the Qurʾānic verses used from historical sources, and compiles them in his appendix (pp. 247–251).
Khoury 1996, 86–89; Calvo 2000, 2008, 2010a.
Khoury 1996, 88.
Khoury 1996, 86.
Safran 2000, 63.
Calvo 2000. The interpretations in her 2010a article are more focused on Umayyad responses to Christianity.
‘Muʿtazila’, Encyclopaedia of Islam, 2nd Edition; Fierro 1987, 1991; Stroumsa 2014.
Stroumsa 2014, 91.
Stroumsa 2014, 84.
Ibn Ḥazm 1953, 91–92; Ibn al-Khaṭīb 1956, 58; Dozy 1913, 488–490; HEM II:217–218; Ávila 1980, 92, 99.
Fierro 1987, 167.
Fierro 1992, 900.
Stroumsa and Sviri 2009, 214; Stroumsa 2006.
De Callataÿ 2014, esp. 262–3.
Fierro 1996c, 105.
See Ibn Hayyan 1981, 25–39 (§§11–22); Cruz Hernández 1981; Fierro 1987, 132–140; Safran 1998, 190–192l; Safran 2000, 32–37; Calvo 2000, 21.
Safran 2014, 148–151: Ibn Zarb had written a refutation of Ibn Masarra’s ideas, and ‘when a group of repenters came to him in 961, he carried their books and writings to the eastern wall of the congregational mosque of Cordoba and burned them while they, and others, looked on’. Safran mentions on p. 149 that the burning took place within the mosque’s courtyard, though Calvo 2010, 160, states that it was before the eastern gates of the mosque, where the edicts had been read, and which would have been a much more public act.
Fierro 1987, 139.
Fierro 1987, 135; Fierro 1996c, 105.
Safran 2000, 33; Safran 1998, 192.
Ibn Hayyan 1981, 31–35 (§§16–19); Safran 2000, 33, n. 55. The fifteen verses are: 2:85; 3:7, 3:10, 3:118; 4:86; 11:76; 12:87; 22:8–9; 39:18; 40:69–72; and 42:13.
‘Associated’ in the sense that, while they do not repeat the exact verses, the mosque inscriptions use verses from elsewhere in the same sūra, sometimes very nearby. These are 2:85; 3:7, 3:10, 3:118; 22:8–9; 39:18; and 40:69–72 (see Appendix 3A). Verses from sūras 3 and 39 are used on the ʿĀmirid façade (Appendix 3B).
Calvo 2000, 24.
Calvo 2000, 23–24.
‘Ibn Masarra’, Encyclopaedia of Islam, 2nd Edition.
Calvo 2000, 25–26. She returns to this interpretation in her 2008 article.
Blair 1998, 11.
Calvo 2000, 24.
Calvo 2000, 19–22.
Calvo 2000, 18.
Calvo 2000, 23.
Calvo 2000, 23.
Khoury 1996, 86.
Ettinghausen 1974.
Calvo 2010a, 180–1; Fierro 2010. According to Bayān II:236 [translation, 320], as administrator of the Treasury, he was responsible for deciding with the fuqahāʾ and ṣāḥib al-aḥbās how much they were going to pay for the works.
Only the inscription on the ‘Puerta del Chocolate’ was read by Amador de los Ríos 1879, 237–238.
Calvo 2010a, 155–6.
Calvo 2010a, 156–9.
See Fernández-Puertas 1979–81 and 1999, and Marfil 2009.
Calvo 2010a, 162.
Grabar 1959, 33–62; and idem. 1987, Chapter 3, ‘The Symbolic Appropriation of the Land’.
Calvo 2010a, 167.
Calvo 2010a, 165–178.
Whelan 1998, 5–6. Five of these sections contain Qurʾānic inscriptions, while the sixth contains ʿAbd al-Malik’s foundation inscription, later adulterated by the Abbasid caliph al-Maʾmūn. Milwright 2016, 62, fig. 2.18, plots the Qurʾānic citations on a plan of the octagonal arcade.
This passage comes at the end of the long mosaic inscription on the inner face of the octagon above the Rock, which contains the anti-trinitarian statement: see Whelan 1998, 5.
Amador de los Ríos 1879, Lévi-Provençal 1931, Ocaña 1976, Ocaña 1988–90.
Since I researched this issue for my thesis, Susana Calvo has published an article (Calvo 2010a) which reconsiders these inscriptions. We have come to similar conclusions, and while I have verified my readings against hers, my identification of the unread Qurʾānic passages do not rely on hers, with the exception of Gate 1 (the Puerta de Jerusalén), which, given its fragmentary nature, I was unable to identify.
Calvo 2010a, 180.
Safran 2000, 62.
Calvo 2010a, 169, 179. On the taṣliyya, see De la Puente 1999b.
Calvo 2010a, 164–5.
Calvo 2010a, 167 and n. 46.
Calvo 2010a, 163, calls these verses ‘chilling’ (escalofriantes), associating them with the ‘environment of intimidation’ (ambiente de intimidación) that she believes characterised al-Manṣūr’s de facto rule, and which she thinks is reflected in the inscriptions on the eastern gates.
Calvo 2010a, 161, 179; Brisch 1959.
Anderson and Pruitt 2017, 239–40. My thanks to both authors for sharing this essay with me when it was still unpublished.
Bierman 1998.
Anderson and Pruitt 2017, 240.
Safran 1998, 190.
De la Puente 1997, 396.