Chapter 6 The Dār al-Ṣināʿa: ʿĀmirid Patronage of the Luxury Arts

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

It was said that the cause of his triumph was service to the sayyida Ṣubḥ al-baskunsiyya, mother of ʿAbd al-Raḥmān and Hishām, and that she was the main reason by which the government and authority passed to him in a short time. He gained this woman’s trust by the quality of his service, by his success in gaining her satisfaction, and by his generosity in the offer of gifts and presents, until he captured her soul and dominated her heart, [which in turn] dominated that of her lord, [al-Ḥakam al-Mustanṣir]. Ibn Abī ʿĀmir strove to do good for her, to make ever more frequent his attentions to her, and he created [new means] to do so, by bringing her things which had no equal [and no-one had ever seen before]. [He did this] until he conceived for her a palace of silver, which he did when he had in his hands the control of the mint. He worked on [the palace] for a time and spent on it an immense quantity of money, to create a novelty [the like of which] nothing more marvellous had ever been seen. They transported it from the house of Ibn Abī ʿĀmir so that the people could see it and they spoke about it for a long time [afterwards]. Ibn Abī ʿĀmir occupied the highest place in her heart: [she] strove to help him, she gave him her support and spoke for him, until [the point] where the people gossiped about her passion for him. One day al-Ḥakam said to some of his confidants: ‘Who is he who [has used] this youth to bring my women to his side, until their hearts are seized? Who is he, who even though they have the pleasures of the world at their disposal, they do no more than describe his gifts and are not satisfied with anything unless he has given it to them? Either [I have] a mage full of wisdom or I have a diligent servant. I am nervous of what might come from his hands …’

Ibn ʿIdhārī, Bayān II:268–9 [translation 416]1

The ability to afford beautiful products executed in expensive, precious and preferably exotic materials has been a signifier of status or aspiration in all cultures,2 and as such patronage of the luxury arts has always been an important facet of rulership. This is epitomised by the bestowal of luxury objects, especially textiles and garments woven with the ruler’s name and titles and intended to be worn by the recipient as a mark of loyalty (khilʿa); as well as by the assembly of precious collections in royal treasuries (as is best-documented for the Fatimid period, thanks to the Book of Gifts and Rarities),3 and the formation of imitative collections by members of the court elite as symbols of their social position, close to the ruler. In the adab literature which developed at the increasingly ‘ritualised’ Islamic courts,4 much space was devoted to ‘describing, naming, and enumerating … substances and articles, objects and victuals’ which members of the court circle considered appropriate to their status;5 and it is telling that the word for ‘luxury objects’ (mustaẓrafa) shares a root with the concept of courtly ‘elegance’ (ẓarāfa) so prized by that circle.6

The passage, quoted above, describing how al-Manṣūr inveigled his way into Ṣubḥ’s affections and secured her support for his rise to power highlights the unprecedented manufacture of a miniature building made of silver, so large that it called people’s attention as it was transported through the streets of Cordoba. No further detail is provided of the palace’s appearance, nor what Ṣubḥ did with this amazing gift once she had received it, though it was so astonishing that it caused the people of Cordoba and the women of the caliphal harem to talk about it for a long time afterwards. The anecdote goes onto to discuss the shadow of fraud that fell over Ibn Abī ʿĀmir because of the huge sums that he had ostentatiously spent on this magnificent creation: his position at the time (around 967) as ṣāḥib al-sikka led to the accusation that he had embezzled from the mint to fund or to find sufficient quantities of precious metal to have the palace made.7 Ibn Abī ʿĀmir was obliged to borrow money from his friend and relative, Aḥmad ibn Muḥammad ibn Ḥudayr, in order to demonstrate that no resources were missing from the mint (Chapter 1). As a consequence of this incident, the young Ibn Abī ʿĀmir came directly to al-Ḥakam’s attention – surely the intention all along. Al-Ḥakam dismissed Ibn Abī ʿĀmir’s accusers as liars, ‘and his admiration for [him] grew and he maintained and confirmed him in his position. Ibn Abī ʿĀmir eventually returned the money to Ibn Ḥudayr, he grew close to al-Ḥakam and came to be one of those among whom al-Ḥakam could not do without in the affairs of government’.8

The framing of this anecdote reflects the teliological tendency in the historiography of al-Manṣūr’s rise to power, where anecdotes and details of his early career are imbued by later historians with a significance implying his destiny to become ḥājib. The gift of the silver palace was, according to Ibn ʿIdhārī, a defining moment in al-Manṣūr’s career, marking a turning point at which his cultivation of Ṣubḥ brought him to the caliph’s attention, after which his astuteness, efficiency and competence led to him only rising higher within the caliphal administration. While Echevarría is no doubt right to say that the anecdote offers a ‘picturesque metaphor’ of the riches at al-Manṣūr’s disposal during his governorship of the mint,9 it nevertheless contains further interesting implications. First, it implies a connection – physical and material – between the mint and precious metal workshops; secondly, though the silver palace was ‘a novelty the like of which nothing more marvellous had ever been seen’ (a clear topos), its manufacture was not beyond the abilities of the Cordoban precious metalworkers; thirdly, it implies that it was possible to access those workshops to order a private commission. Perhaps al-Manṣūr’s already significant role in the state administration, as ṣāḥib al-sikka, allowed him this privileged access, especially if the intended recipient of the commission was a member of the royal household. Finally, this anecdote speaks to the way in which al-Manṣūr was highly strategic about gift-giving and the use of objects, implying that from an early stage in his career he was commissioning objects from the luxury arts infrastructure that were calculated to solidify personal relationships and loyalty. If the anecdote can be dated around 356/967, when Ibn Abī ʿĀmir was appointed governor of the mint for the first time, Ṣubḥ was already someone of importance at court: the birth of her first son, ʿAbd al-Raḥmān, around 965 meant she was umm walad, mother of the heir to the throne. A canny politician, Ibn Abī ʿĀmir may already have been looking to align himself with the interests of the caliphal succession as a way of advancing his career at the heart of the court.

Though no architectural sculpture in silver survives from al-Andalus, an example in silver, silver-gilt and niello made in the Byzantine world at an exactly-contemporary moment does survive. This is an artophorion (container for the Eucharist), held since the twelfth century in the treasury of the cathedral at Aachen where it was repurposed as a reliquary for the head of Saint Anastasios the Persian (Figure 83). It is made in the shape of a small but elaborate Byzantine church, with a central dome rising above an arcaded drum, a small apse pierced by long windows, and large portals decorated with applied crosses. A dedicatory inscription framing the apse indicates that it was commissioned by Eustathios, who served as Byzantine military commander of Antioch in 969/70.10 This precious survival of a miniature building crafted from silver may now be unique, but it illustrates the potential of the luxury metalworking ateliers in al-Andalus to have produced something similar for al-Manṣūr’s gift to Ṣubḥ. As we know from the casket made for Hishām (Figures 1–2), sophisticated objects in precious metals were being made in the Cordoban ateliers at this time; and they were not just producing containers – sculptures were produced in bronze in the form of animals for the caliphal palaces (Figure 10). Perhaps a Byzantine example similar to the artophorion had been received in Cordoba during one of the exchanges with Constantinople, and inspired the silver palace for Ṣubḥ. The anecdote about al-Manṣūr’s ingratiation may thus preserve testimony of a type of object that was produced in Cordoba but which has not survived.

Figure 83
Figure 83

Reliquary of Saint Anastasios the Persian, in the shape of a miniature church, Byzantine, dated 969–970, silver, silver-gilt and niello, dims 39 × 19.6 × 20 cm; Cathedral Treasury, Aachen

© Mariam Rosser-Owen

Alongside their grand architectural projects, the second element in the ʿĀmirids’ use of cultural and artistic patronage to articulate the legitimacy of their rule was their sponsorship of the luxury arts industry. The remaining chapters of this book will thus be dedicated to the discussion of objects. This chapter will explore the mechanisms by which objects were commissioned, arguing that the ʿĀmirids adopted direct control of the caliphal Dār al-Ṣināʿa, perpetuating the organisational structures that had developed during the reigns of the first two caliphs, though with the artistic commissions now ‘in the hands of’ – ʿala yaday – Ṣaqāliba who belonged to, and thus were personally loyal to, the ʿĀmirids. This allowed the ʿĀmirid patrons to dictate how their objects were decorated, and to create a deliberate programme of ʿĀmirid self-expression, so that the messages we read in them are not accidental. Chapter 7 will identify for the first time a corpus of ʿĀmirid objects, by a close examination of the surviving objects that are associated with the ʿĀmirids through epigraphy, or through close stylistic connections with the inscribed objects. Many of these are well-known in Andalusi art history, because of their spectacular proportions and decoration – the Pamplona casket (Figures 120–127) is a case in point – but they have usually been discussed in broad terms as products of Umayyad al-Andalus, and have rarely been properly located within a discussion of the specific context in which they were created.11 Lastly, in Chapter 8, we will discuss the meanings of ʿĀmirid art by examining the messages embedded in their iconography and inscriptions, with particular reference to the poetic imagery composed for these same patrons. This will elucidate the ways in which the ʿĀmirid dynasty understood and visualised their role. While my main focus in these chapters will be objects, I take the term ‘luxury arts’ to encompass both the architectural decoration commissioned as adornment for palaces and mosques (the specific examples of which were discussed in detail in Chapters 4 and 5), and the precious objects commissioned to furnish them. As we will see, there was a close infrastructural relationship between the production of both, which shared materials, skills and craftsmen.

The objects that survive from al-Andalus – and in particular the formulae used in their inscriptions – allow us to reconstruct quite a clear picture of the structure of the luxury arts industry, and to see that it was tightly regulated within workshops operating at the heart of the court. Whether these functioned solely as a royal monopoly or might have been more widely accessible to members of the court elite is a question that will be considered below. What is clear is that, more so than with other medieval courts in the Mediterranean – possibly in the entire medieval world, both Christian and Muslim – we have evidence for the production of objects within workshops. Where doubt exists about the existence of workshops in other ivory carving centres, such as Constantinople or southern Italy under the Normans,12 we can be confident that Andalusi ivory carving was the product of an organised team operating according to a strict hierarchy. That these ivory objects were produced by the same craftsmen who worked on architectural decoration, responding to the availability of the raw material – which might have been more exceptional than we have come to think – will also be elaborated below. The point that this discussion hopes to make is that understanding this workshop structure allows us to better understand the processes of patronage. This aspect of Andalusi luxury arts production has (surprisingly) never been fully considered, and it has therefore never been securely established that a patron was able to directly control an object’s message. While the historical evidence for this remains largely absent, the names and signatures on surviving objects and architectural adornments, and especially the deployment of fityān in the service of the patron, allows us to reconstruct, to an extent, how these objects were commissioned and to suggest how they might have responded to the direct wishes of the patron.

1 The Origins of the Dār al-Ṣināʿa

The Dār al-Ṣināʿa was one part of a full-scale industry, which was an important facet in the development of the caliphal state in al-Andalus and which, perhaps most importantly, provided a vehicle for the rulers’ self-expression. The foundation of the Dār al-Ṣināʿa is credited to the Umayyad amir ʿAbd al-Raḥmān II, at the same time as he established the Dār al-Ṭirāz and the Dār al-Sikka.13 The products of these state workshops are widely recognised as royal prerogatives, but it is also important to include the production of luxury objects beyond textiles in this picture, given their importance in the distribution of khilʿa by the ruler to members of his court and allies, and in diplomatic gift exchange (Chapter 2). As we know from the sources, patronage of the luxury arts industry played a fundamental role in the foundation of ʿAbd al-Raḥmān III’s caliphate. He is said to have spent equal quantities of the state revenue on funding his army and his architectural patronage,14 and caliphal expenditure on building projects was thus ‘the most important aspect of appropriation of territory and power by the Umayyad caliphs, next to the military’.15 Surviving objects datable to the ninth century might be early products of this institution, and textual evidence exists for highly sophisticated scientific instruments being produced, probably in brass, in the mid-ninth-century by Ibn Firnās, an intellectual at ʿAbd al-Raḥmān II’s court.16

We know that Dār al-Ṣināʿa was located alongside the Cordoban qaṣr, at the south-west of the city. It must have been close to a gate at the north of the palatine complex which bore the name Bāb al-Ṣināʿa. It was also close to the Bāb Ishbīliya (the Seville Gate), also known as the Bāb al-ʿAṭṭārīn since beyond this gate was located the quarter of the perfumiers/pharmacists, next to which was that of the parchment makers.17 Presumably the latter were physically located close to those who produced books, including calligraphers, illuminators and leather binders. This concentration of workshops and artisanal skills near to the palace is telling of the close relationship of the luxury arts to the heart of the state, from an early period. They were also close to the Great Souk – which, following a fire in 936, was reconstructed under the direct patronage of ʿAbd al-Raḥmān III18 – suggesting that the same workshops served the court as well as a wider market, an idea that will be elaborated further below. The mint was in this location as well, likely to have been in its own building, until it was moved to Madīnat al-Zahrāʾ in 947.19 This would mean the precious metals being stored and processed here were physically close to the luxury arts ateliers, in order to produce mounts and other fixtures for objects, and perhaps the mint workers also processed these materials into precious metalwork objects, as implied by the anecdote of the silver palace quoted above. It is also interesting that in the ninth century there was already a connection between craft production and perfumes and cosmetics, which remains the case in the tenth century, as we will see.

The exception to this centralisation of the state workshops in the area to the west of the palace was the Dār al-Ṭirāz, which at some point in the early tenth century was moved to the north of city, beyond the walls, near to the cemetery of Bāb al-Yahūd and the Mozarabic church of San Zoilo. The Dār al-Ṭirāz was already well-established in this area by 961, since this suburb is called vicus tiraceorum several times in the Calendar of Cordoba, written in that year.20 This move – presumably initiated by ʿAbd al-Raḥmān III, perhaps in the 940s when the Dār al-Sikka, Dār al-Ṣināʿa and other ateliers moved to Madīnat al-Zahrāʾ – displaced the previous model of administrative centrality, but the Dār al-Ṭirāz remained in this location thereafter. It is not known why the textile factory was moved, though perhaps it was for practical reasons, to be closer to the raw materials, in particular the tanneries, which being an industrial production requiring lots of space and generating unpleasant odours are usually located beyond the city walls.21

Ceramics were also normally produced beyond the city walls, since again they required space – for the storage and processing of raw materials, the making, drying, storage of half-finished or finished objects, decorating, kilns – and the fuel burnt off during the firing was smokey and full of impurities, so located away from residential quarters.22 Excavations in recent years have uncovered more than 140 kilns in Cordoba’s periurban zones, more than have been found in any other Andalusi city.23 These are concentrated in two broad zones: the first is to the north of Cordoba’s eastern suburb, al-Sharqiyya, in an area now known as Las Ollerías, reflecting its historical association with ceramics. Pottery production seems to have taken place here since Roman times, and continued to the end of the Islamic period. This zone seems to have been where domestic pottery was primarily produced. Arabic sources mention a quarter known as al-Fakharīn located in al-Sharqiyya, very close to the city walls. The guild of potters was located here, which also indicates a level of centralised and standardised production.24 As Salinas points out, the siting of the pottery quarter here makes sense in terms of the geomorphology of the city, as it was located on the second fluvial terrace of the Guadalquivir, and was a region rich in clay. There was an ample water supply from the river’s tributaries – in particular, the Fuensanta stream passes through this zone – and was close to the foothills of the Sierra Morena which supplied wood to fuel the kilns. It was extramuros but close to the city walls and to one of the city’s main arterial roads. In the caliphal period, during the massive urban expansion to the west of the city caused by the foundation of Madīnat al-Zahrāʾ, a new zone for pottery production was created to the west of the madina of Cordoba, also located on the second fluvial terrace. The kilns that have been excavated here show that this zone primarily produced building materials, such as tiles, for the new construction, and very large storage jars, to store construction materials or to service the agricultural and fishing activities in this part of the city. This zone appears to have been largely abandoned during the Fitna and did not continue in use into the later Islamic period.

The putative site of the Dār al-Ṭirāz was excavated in 1991 as a salvage operation during the construction of the high-speed train station and railway lines, and this area and any further information it might offer is now effectively buried under the train tracks.25 The building that was uncovered during these excavations was a very large, monumental edifice constructed around 11 courtyards, according to construction methods which date it to the third quarter of the tenth century. Its configuration with numerous courtyards and halls but no basilical structures implies that its function was not public or official. Initially identified as a luxurious private residence, no decorative elements were found, although the structure had been heavily robbed out. Given the topographical information from the sources that clearly indicates the Dār al-Ṭirāz was located in this zone, Arjona Castro and Marfil have identified this structure as the Dār al-Ṭirāz. It might have previously been a munya, repurposed by ʿAbd al-Raḥmān III to house caliphal textile production when he reorganised and relocated the other ateliers. Arjona Castro and Marfil suggest that it was later expanded by al-Ḥakam II, and the completion of this expansion or refurbishment might have been the reason for his famous inspection visit in 361/972.26 In this year, al-Rāzī’s Annals tell us that ʿAbd Allāh ibn Aḥmad al-Iflīlī was appointed amīn (superintendent) of the Dār al-Ṭirāz, and Muḥammad ibn al-Walīd was elected by his colleagues to the post of kātib al-ṭirāz; he was apparently ‘one of the most outstanding and practical of the kuttāb, and one of the most capable and experienced of the men in that office’. The caliph was received at the factory ‘by the administrative directors and by the directors of the workshops, who paid him the necessary tribute. [Al-Ḥakam] asked them for details of their work and favoured them with his instructions’.27

It is clear from this anecdote that the Dār al-Ṭirāz was regulated by a structured hierarchy of administrators, headed by a superintendent (amīn), a kātib al-ṭirāz, other kuttāb below him, and other directors and overseers of individual workshops within the factory – perhaps each one producing textiles according to different techniques or designs. All of this operated under the direct control of the caliph, who possibly paid to have the building enlarged or refurbished, but certainly took enough interest in its operation to pay this special visit. This speaks to the importance of ṭirāz as a caliphal product. We can also extrapolate from this that the Dār al-Ṣināʿa, and no doubt other ateliers, needed a physical location from which to operate, as well as a structured administrative hierarchy, probably formulated along the same lines: an amīn, chief kātib, other kuttāb, heads of workshops and master craftsmen. These must be the roles indicated by the different names – of fityān and craftsmen – inscribed on some of the products of these industries. According to the few historical texts that mention Andalusi ṭirāz, it would appear that the Cordoban Dār al-Ṭirāz functioned along similar lines to the Fatimid industry, in which some workshops apparently worked exclusively for a court clientele (known as ṭirāz al-khāṣṣa) while others worked more openly for the market (ṭirāz al-ʿāmma).28 In the account of a gift sent to the Zanāta chieftain, Muḥammad ibn Khazar, soon after ʿAbd al-Raḥmān declared his caliphate (discussed in Section 2, below), the accompanying letter is at pains to stress that the clothing being sent was made ‘fi ṭirāz-hu al-khāṣṣ’, ‘in his – the caliph’s – private ṭirāz’. Compared to the huge numbers of ṭirāz textiles that have survived in Egypt’s dry climatic conditions, only two Andalusi textiles survive which are inscribed with names of the caliph: one possibly made for ʿAbd al-Raḥmān III, now in the Cleveland Museum of Art;29 and the so-called ‘Veil of Hishām’ in the name of Hishām II (Figure 9, Appendix 4.2, discussed below).

It is therefore possible that, as he reformed the coinage and other offices of the state bureaucracy, ʿAbd al-Raḥmān also reorganised the internal structure of the Dār al-Ṣināʿa and Dār al-Ṭirāz, so that they functioned more efficiently under the strain of the increased output expected from them by the caliphs. All these ateliers were moved to Madīnat al-Zahrāʾ, during the period when ʿAbd al-Raḥmān was aggrandising it to be a more potent expression of his power as caliph. Al-Maqqarī records that in 335/946, the Dār al-Ṣināʿa was relocated; the mint was moved in 947, as indicated by the surviving coin issue. Both were well established there by 949, when – interestingly – they were included on a tour of the palace-city by a delegation of Byzantine envoys: ‘they visited the Dār al-Ṣināʿa and the Dār al-ʿUdda (the house of military equipment), on the outskirts of al-Zahrāʾ, and the Dār al-Sikka (the mint)’.30 However, the first extant dated objects which identify Madīnat al-Zahrāʾ as their place of production come from twenty years later, in 355/966 (the ivories made for al-Ḥakam’s sister, Wallāda).31 Other services that apparently remained in Cordoba were the Dār al-Zawāmil, the stables for the beasts of burden (sing. zāmila) and the Barīd, post office. In 971, these were moved to the west of the city, in the expansion zone that had sprung up between the old city of Cordoba and Madīnat al-Zahrāʾ.32 The ḥājib Jaʿfar al-Muṣḥafī was charged with masterminding this move, and this could be why his title is given as ṣāḥib al-khayl wa-l-burud on the marble pipe-cover in the David Collection (Figures 104–107).

The Dār al-ʿUdda mentioned during the visit of the Byzantine envoys might be the same as the Khizānat al-Silāḥ at Madīnat al-Zahrāʾ.33 This government arsenal also manufactured weapons in enormous quantities: 3000 tents, 13,000 shields, 12,000 bows a year, and 20,000 arrows a month. As we saw in Chapter 4, the sources tell us that there were weapons stores (asliḥa) at al-Madīnat al-Zāhira:34 of course, al-Manṣūr and his enormous army needed to be appropriately apparelled for their regular campaigns, and weapons also played a major part in khilʿa.35 Could the production of weapons have taken place here too?

Another Khizānah mentioned in the historical sources seems to have been similar to the Fatimid treasury, in that it was simultaneously a storage facility that was also intended to display precious objects. This treasury is mentioned in the Kitāb al-Mafākhir al-Barbar in the anecdote relating to Ḥasan ibn Qānūn’s marvellous piece of ambergris, which al-Ḥakam would stop at nothing to acquire:

“A fragment of ambergris of unusual form and extraordinary dimensions … obtained from a certain place along his coasts. Ḥasan had ordered it to be worked until it had the form of a leather cushion, and he used it in the period of his splendour. Word of this came to al-Ḥakam, who requested that Ḥasan send it to him for his treasury (khizānati-hi), on the condition that he would be recompensed in justice for the fragment of amber.”36

Ḥasan commissioned an extraordinary object to be crafted from this precious material, and showed it off ‘in the period of his splendour’. Stories of this amazing piece of ambergris provoked al-Ḥakam’s jealousy, and he became obsessed with acquiring it. What was the point of going to such lengths if the amazing piece of ambergris was going to disappear into a store where it would remain invisible? This anedcote surely implies a treasury that the caliph could take visitors around to boast proudly of his art collection, as collectors do today. It is worth keeping this in mind when we consider the visibility of such luxury objects as the Cordoban ivories. This Khizānah even had its curator, al-khāzin, a role played by al-Ḥakam’s favourite Durrī al-Saghīr in the 960s (see below). The lengths to which al-Ḥakam would go to obtain important works of art is seen in other cases, not just his desire to have Byzantine glass mosaics for his maqṣūra at the Cordoba mosque (Chapter 5): he paid high prices to obtain a copy of Dioscorides’ Materia Medica, which then initiated a whole translation enterprise, as well as copies of the Kalīla wa Dimna and the Kitāb al-Aghānī, all of which no doubt were then housed in his famous library (which the texts call al-khizānat al-ʿilmiyya, or Treasury of Knowledge).37 As discussed in Chapter 1, this Library functioned as a scriptorium for the copying and production of manuscripts, as well as providing a repository of books which could be consulted by Cordoba’s elite scholars. Could we imagine that the Khizānah functioned, or was at least accessible, in a similar way? It is perhaps significant that the inscription on the famous mantle woven in Palermo in 528/1133–4 for the Norman king Roger II (r. 1130–54) states that it was woven in the Khizānah, which appears in that case to be equivalent to the caliphal institution of the Ṭirāz.38 Could this imply that as well as being an open access storehouse for precious possessions, the fabrication of luxury objects also took place here? Perhaps Khizānah was an equivalent term to Dār al-Ṣināʿa.

While we can speculate that the library and Khizānah were located within the palace precincts itself, where were the workshops located after their transfer to Madīnat al-Zahrāʾ? Al-Maqqarī’s text, cited above, about the visit of the Byzantine ambassadors to the workshops states that that the Dār al-Ṣināʿa and Dār al-ʿUdda were bi aknāf, ‘on the outskirts of’, al-Zahrāʾ. The topography of Madīnat al-Zahrāʾ replicated that of the old city of Cordoba; gates were placed in the same relationship to each other as in Cordoba, and were even given the same names.39 This would suggest that the workshops would be located in the far south-west of the site, replicating their old placement near the Souk and Bāb Ishbīliya/al-ʿAṭṭārīn in Cordoba. The equivalent zone at Madīnat al-Zahrāʾ remains unexcavated but aerial photography has identified ‘six large structures, identical, contiguous and aligned north-south, which constitute an extraordinary group of constructions of a state character’.40 These are extremely large buildings with a large open space in front of them. Vallejo has proffered as possible interpretations of these structures that they could be the barracks for the city’s guard, or the royal workshops. Vallejo admits that this identification is hypothetical until geophysical survey can shed clearer light on the function of this zone.41 If these are the buildings housing the workshops, are they too monumental? Ivory carving, or the working of precious metals, would not require such extensive space for equipment, storage or working, but the ateliers for architectural decoration certainly would – to store the stone and marble and other large-scale materials required for making columns, capitals and bases, and then to store the finished pieces until such time as they could be installed in their intended buildings. Ibn Ghālib states that ‘the houses of the fityān … and the functionaries were outside the palace, on the western side’ of Madīnat al-Zahrāʾ.42 As we will discuss below, the fityān were the cohort who regulated the luxury arts industry, as well as other branches of the state infrastructure, and as such, if these buildings can be identified as the workshops, they themselves would be physically located nearby.

Unfortunately, in the case of al-Manṣūr’s palace- city, we have no physical evidence from which to extrapolate the location of the ʿĀmirid Dār al-Ṣināʿa, but one precious piece of information informs us that a branch of the industry indeed existed there. Among the earliest dated of the extant ʿĀmirid objects is the marble basin made for al-Manṣūr in 377/987–8 (Figures 113–118, Appendix 4.7), which tells us through its inscription that it was made bi-qaṣr al-zāhira. This makes the basin the only securely identifiable survival of the ʿĀmirid palace-city. It was made under the direction of Kh-, probably Khalaf or Khayrah, the fatā al-kabīr al-ʿāmirī, indicating the physical presence of that fatā at the ʿĀmirid palace-city. The extant ʿĀmirid objects date from the mid 980s onwards, coinciding with the construction of al-Madīnat al-Zāhira, the adoption of a laqab, and the culmination of al-Manṣūr’s rise to power. The fact that the marble basin was made at the palace-city underlines the significance of the relationship between al-Manṣūr’s projection of power and his utilisation of the luxury arts, as we will see below.

1.1 Structure of the Industry

It is possible to reconstruct to some extent the organisation of the Dār al-Ṣināʿa, and to identify some of its employees. The caliph’s direct patronage of the luxury arts industry is implied by the inscriptions on surviving objects, especially carved ivories and marbles, which give the name of the caliph or a member of his family as the commissioner or recipient of that object. Of the seventeen ivories with an historical inscription, nine are datable to the period of the first two caliphs, of which eight were made for members of the highest level of the ruling family; two of them even state that they were made bi-madīnat al-zahrāʾ.43 This implies a close association between the ivory-carving industry and the person of the caliph – as Makariou puts it, ‘in caliphal Spain, the art of ivory-carving was essentially an art of power, just like the ṭirāz’.44 In fact, the formal relationship between the iconography of ivories and textiles argues for a close executive relationship between the Dār al-Ṣināʿa and the Dār al-Ṭirāz – compare, for example, the scenes of horsemen inside lobed medallions on the al-Mughīra and Ziyād ibn Aflaḥ pyxides (Figures 3, 12, 171) and textiles such as the Oña embroidery (Figures 13, 137) or the Suaire de Saint Lazare (Figures 84, 136). We know from inscriptions and textual sources that Ṣaqāliba were sometimes running two ateliers at the same time: around 350/961–2, Jaʿfar al-Ṣiqlābī was acting as ṣāḥib al-khayl (Head of the Cavalry) wa al-ṭirāz,45 and in 361/972–3, Fāʾiq was made ṣāḥib al-burud (Head of the Postal Service) wa al-ṭirāz.46 We will discuss below the possibility that the craftsmen themselves moved between different media.

Figure 84
Figure 84

Suaire de Saint Lazare, late tenth century, embroidered silk and gold thread; general view of the largest fragment

© Musée Rolin, Autun
Figure 85
Figure 85

Marble basin with heads of lion and deer from al-Rummāniyya, 960s; Museo Arqueológico y Etnológico, Cordoba, inv. CE006418, H 28.5 W 97 D 71 cm

© Mariam Rosser-Owen

The patronage structure of the luxury arts industry is most easily reconstructed at the stage at which an object was commissioned.47 Inscriptions on objects not only provide information but also mark them as royal products of the Dār al-Ṣināʿa, as does the way that the inscriptions are carved in relief. The standard formula commonly used in al-Andalus to express the commission was mimma amara bi-ʿamali-hi, ‘what he [patron already specified] ordered to be made’; when the object was commissioned as a gift for someone else, this phrase is followed by the preposition li- and the recipient’s name. The Girona casket inscription provides a clear illustration of this: it begins by invoking blessings on the caliph al-Ḥakam, and continues with the standard formula, as above, followed by, li-Abī al-Walīd Hishām walī ʿahd al-muslimīn (Figures 1–2, Appendix 4.1). This implies that this object was commissioned as a gift for the caliph’s son, and even though the object is not dated, it has been convincingly argued that it was created to celebrate Hishām’s official designation as walī al-ʿahd in 976.48 The commissioning formula may also be accompanied by the name of a court official (fatā) through whom the commission was carried out (designated in the inscriptions by the phrase ʿalā yaday), and these names are often qualified by information indicating the personal relationship of that official to his patron: for example, the suffixes ‘ibn ʿAbd al-Raḥmān’,49 ‘ibn Hishām’,50 ‘ibn Muḥammad al-ʿĀmirī’,51 or just ‘al-ʿĀmirī’.52 All of these phrases indicate that these fityān were slaves or freedmen of the patron. These names may also be qualified by the short phrases ʿabdu-hu, ‘his servant’, which should perhaps be translated in the more general sense of ‘at his [the patron’s] service’; or mawlā-hu, ‘his client, freedman’; or, more unusually on the Pamplona casket, mamlūku-hu, ‘his slave’.53 Though the name is partially lost on al-Manṣūr’s marble basin, the designation given is al-fatā al-kabīr al-ʿāmirī, i.e. the ʿĀmirid fatā, presumably to make a distinction from the caliphal fatā. This is perhaps logical when one considers that there were now two palace households, at Madīnat al-Zahrāʾ and al-Madīnat al-Zāhira, but also indicates that the Dār al-Ṣināʿa at al-Zāhira replicated the organisational structure of that at al-Zahrāʾ, possibly on a smaller scale, unless the whole infrastructure had in fact been transferred to al-Zāhira and brought under direct ʿĀmirid control.

In several cases, the named fatā can be identified with an individual known from the historical record, or from the signatures on other objects. These were men of high social standing, who held important offices in the state bureaucracy, and were often very wealthy. Though no inscription explicitly describes them as aṣḥāb dār al-ṣināʿa, it must be assumed that this is what the mention of their names on the commissions implies. This is confirmed by the Girona casket (Figures 1–2), where the Jawdhar named in the inscription can be identified with the ‘superintendent of the gold- and silver-smiths’ (ṣāḥib al-ṣāgha) who plotted in 976 to displace Hishām from the succession (Chapter 1).54 At the same time, Jawdhar also held the post of ‘Grand Falconer’ (ṣāḥib al-bayāzira), which implies his high standing in al-Ḥakam’s entourage, the ‘bayzara’ (sport of falconry) being an important caliphal pastime.55

The social and political importance of the men in charge of the luxury arts industry is illustrated by the career of Jaʿfar ibn ʿAbd al-Raḥmān al-Ṣiqlābī. His name occurs in inscriptions between the 950s and 970s, and Ocaña used this epigraphic record to trace his career, corroborating this information where possible from the historical sources. In inscriptions from the Hall of ʿAbd al-Raḥmān III dated 345/956–7, Jaʿfar is styled simply fatā-hu; by 348/959–60 he seems to have been manumitted, since his name is suffixed fatā-hu wa mawlā-hu.56 Within a couple of years he had received an important promotion, since a series of inscriptions from the year 350/961–2 name him as ṣāḥib al-khayl wa al-ṭirāz fatā amīr al-muʾminīn … wa mawlā-hu.57 This evidence is corroborated by al-Maqqarī, who links the promotion to al-Ḥakam’s accession, and also gives important information about Jaʿfar’s wealth.58 By the mid 960s he had been promoted again, to ḥājib, and given the laqab ‘Sayf al-Dawla’.59 It is under this title that he occurs on the remaining inscriptions, which come mainly from al-Ḥakam’s extension to the Cordoba mosque, a project which he supervised.60 We will return to Jaʿfar below, in the context of his own luxury arts patronage while ḥājib, as a possible precedent to al-Manṣūr.

A similar career path is followed by two fityān charged with the production of some important ivories, though there is less external information about their careers. Durrī al-Saghir, whose name appears on the pair of pyxides made circa 964 for al-Ḥakam (Figure 86) and his concubine Ṣubḥ, is known from the contemporary historical accounts to have been a favourite at al-Ḥakam’s court, and to have accumulated great riches in his service: in 973, Durrī made a gift to the caliph of the Munyat al-Rummāniyya, which he had constructed for himself, and which became a favoured caliphal retreat from nearby Madīnat al-Zahrāʾ.61 While no official title is given in the inscriptions, we encounter him in al-Rāzī’s Annals (§9) as ‘al-khāzin’ (treasurer or curator – equivalent to being in charge of the Dār al-Ṣināʿa?), and ‘khalīfa’ of Madīnat al-Zahrāʾ.62 By 976, he was serving as qāḍī of Baeza, but he fell victim to al-Muṣḥafī’s and Ibn Abī ʿĀmir’s attempts to curtail the wealth and prestige of the most powerful Ṣaqāliba. An incident in which Durrī insulted Ibn Abī ʿĀmir by pulling his beard – at the very gate of his residence, in front of members of his personal guard – indicates the social faultlines that were beginning to divide state and society: on the one hand, the increasing reliance on Berber support by al-Muṣḥafī and Ibn Abī ʿĀmir, and on the other the more ‘traditional’ reliance of the Andalusi elite on the slave faction of the Ṣaqāliba.63

The second fatā who was involved in the production of ivories was Zuhayr ibn Muḥammad al-ʿĀmirī, named in the inscription of the Pamplona casket as the fatā al-kabīr who directed this commission, as well as al-Manṣūr’s manumitted slave (mamlūkihi). This tour de force of ivory carving, which will be discussed in more detail in Chapter 7 (2.1), appears to have been such an important work – that needed to be completed as soon as possible to commemorate ʿAbd al-Malik ibn al-Manṣūr’s victory over Sancho García of León, for which he received the new title ‘Sayf al-Dawla’ – that a whole team of carvers worked on it, led by Faraj who perhaps also led the team that carved the unprecedented double-capital, and no doubt others, to adorn the palaces of al-Madīnat al-Zāhira (Figures 37–38, Chapter 4). But over Faraj was Zuhayr, the fatā who at that time held a position of responsibility in the (ʿĀmirid) Dār al-Ṣināʿa. It is probable that Zuhayr was also named as the fatā al-kabīr in the inscription on the Braga pyxis, carved shortly afterwards, where the name of the director of the commission is missing because of a loss to the ivory (Figures 11, 15, Appendix 4.12). It is probable that this Zuhayr ibn Muḥammad al-ʿĀmirī was the same man who was governor of Jaén, Baeza and Calatrava, and who later became a Taifa ruler, like many ʿĀmirid officials, in this case of Almería (Conclusion).64

It can thus be argued that wherever a fatā is designated as the director of works by the phrase ʿala yaday, this equates to the official role of ṣāḥib within the luxury arts manufactory, the Dār al-Ṣināʿa. These aṣḥāb were important men in society and in the state bureaucracy, and exercised a ‘function of control’ over the skilled workers who engaged in the actual making.65 They may have been the men who worked with the patron to conceive and develop a design for an object. It is likely that different ateliers, working in different media, functioned under the umbrella of the Dār al-Ṣināʿa, which is why the same names are found on architectural decoration, ivories, marbles, even ceramics and metalwork. While textiles were produced in a separately-organised industry, as we have seen, men of the status of Durrī al-Saghir, Jaʿfar al-Ṣiqlābī or Fāʾiq held roles across different branches of this industry, linking in one person the luxury arts with the Dār al-Ṭirāz. This suggests the means by which iconographic relationships between textiles and luxury objects might have come about. It has often been observed that the medallion-based designs of the figurative ivories were influenced by the design of textiles such as the Suaire de Saint Lazare (Figures 84, 136), whose processions of hunters or warriors inside lobed medallions are stylistically very close to the mounted falconers that recur on the carved ivories. Thinking about the ʿĀmirid marble basins, for example (Chapter 7: 1.2, 2.3.1, 4.2.1), the fact of a centralised crafts industry operating under direct royal patronage could facilitate design choices between textiles and carved marbles. If the ʿĀmirid marble basins copied textile models, this would go some way to explaining both the anomalous depiction of ‘frontal-splayed’ eagles in al-Andalus, and why they are stylistically so similar to contemporary textiles, especially those produced in Byzantium (Figure 87).

Figure 86
Figure 86

Pyxis made for al-Ḥakam II, c 964, ivory and silver-and-niello mounts; Victoria and Albert Museum, 217–1865

© Victoria and Albert Museum, London
Figure 87
Figure 87

Heraldic eagles on the Suaire de Saint Germain, Byzantine, about 1000, woven silk

© Musée d’Auxerre

The names of the skilled workers at the next level down within the Dār al-Ṣināʿa are also represented epigraphically, in the form of the large number of signatures inscribed on luxury arts of all media, usually preceded by the word ʿamala, ‘made [by]’. Many of the names were common among slaves and freedmen, being auspicious or aesthetic abstract nouns such as ‘Shining’ (Durrī), ‘Consolation’ or ‘Joy’ (Faraj), ‘Victory’ (Fatḥ and Naṣr), ‘Elegant’ (Rashīq), ‘Happiness’ (Saʿāda), or ‘Full Moon’ (Badr).66 We find many of these names carved onto architectonic elements in the ʿĀmirid extension to the Cordoba Mosque, as we saw in Chapter 5. The recurrence of the same name in different artistic contexts may be pure coincidence, and need not necessarily indicate the involvement of the same craftsman. Juan Souto was rightly wary of stating that the same name on different objects indicated their makers were one and the same, since such names were common among the slave class in al-Andalus, which numbered in the thousands – though a much more restricted pool would have been employed in the Dār al-Ṣināʿa, especially with the level of skill that would merit signing their name.67 Since men at this level of society are not mentioned in the historical sources, it is difficult without further evidence to say for certain that these were the same craftsmen, however the circumstantial evidence, of the same name occurring in contemporary and related contexts, is highly suggestive. As we will discuss below, it is likely that the same men worked across different media and different projects. It is significant that the seven hundred signatures documented on architectonic elements in the ʿĀmirid mosque extension comprise only twenty-five names, of which thirteen occur as signatures in contexts outside the mosque, including on objects (Chapter 5). This would seem to indicate that this relatively small number of craftsmen were working together within one branch of the Dār al-Ṣināʿa, on projects that ranged from ivory carving to the production of columns and capitals for the mosque.

The Pamplona casket is unusual in bearing five signatures, incised in inconspicuous places throughout its decoration (Figure 126). This indicates that five craftsmen worked simultaneously on its production, though the style of its iconographical programme is surprisingly unified for a shared project. This might imply a designer, perhaps the fatā, Zuhayr, or Faraj, whose role as master craftsman is indicated by an invisible inscription, located on the underside of the topmost plaque of the lid; this extremely important signature was only discovered when the casket was dismantled for cleaning and conservation in the 1960s.68 It reads ʿamal Faraj maʿa talāmiḍihi, ‘Made by Faraj with his apprentices’ (Figure 126A), and indicates that a craftsman whose name occurs in two contexts associated with al-Manṣūr’s mosque extension was, a decade later, a master craftsman in the luxury objects atelier.69 Perhaps because of this role, Faraj is known from other epigraphic evidence to have worked in teams: as discussed in Chapter 4, his name appears alongside Bāshir and Mubārak on the unprecedented double-capital probably from al-Madīnat al-Zāhira (Figures 37–38); and on a (single) capital, reused in the Alcázar at Seville, alongside Mubārak again, and Khayr who, significantly, also signs the Pamplona casket.70

Faraj’s signature also appears in the carving on the casket itself, incised on the right lion-slayer on the right side of the lid (Figure 126B). The other names given are Miṣbāḥ (‘Light’), incised beneath the lion-borne platform in the right medallion on the front, beneath the bearded figure; Khayr (‘Goodness’), incised in the centre of the shield carried by the lion-fighting man on the back central medallion; Rashīq (‘Graceful’), incised on the hindquarters of the right deer in the left medallion on the left side; and Saʿāda, incised on the hindquarters of the left deer in the right medallion on the right side.71 Clearly a whole team was needed to carve such a large and elaborate creation as the Pamplona casket, and perhaps to finish it quickly. As Sheila Blair points out, the location of the signatures is significant, and helps us to infer the division of labour across the workshop. The lid is signed (visibly and invisibly) by the master craftsman Faraj, then his four apprentices each sign one of the four remaining plaques: Miṣbāḥ the front, Khayr the back, Rashīq the left side, Saʿāda the right side. The positioning of Miṣbāḥ’s name – literally ‘under ʿAbd al-Malik’s foot’ – creates a visual pun of the craftsman’s lowliness vis-à-vis the patron; it also implies that these ivory carvers were by this date controlled not by the caliph but by the ʿĀmirid ḥājib.72 This connection to the patron is underlined by the presence of the word ʿāmir incised on the hindquarters of both horses in the left medallion on the back of the casket (Figure 126E). They have in the past also been read as signatures, but Blair suggests that instead they are brands, ‘identifying the horses as belonging to the family herd owned by the [casket’s] patron, ʿAbd al-Malik ibn al-Manṣūr’.73 Like the wording of the inscription, and the visual imagery that we will discuss further in Chapters 7 and 8, everything connects the carving of this casket with the ʿĀmirids as patrons.

Three of Faraj’s ‘apprentices’ are also known from other contexts,74 of whom the third, Khayr or Khayra(h?), can be found among the signatures in Cordoba IV, and may appear on al-Manṣūr’s marble basin. Only the first letter of the name (kh-) survives on the basin, and it may be more likely to have originally read Khalaf: this individual occurs throughout Cordoba IV, most frequently as Khalaf al-ʿĀmiri, but is also known to have been qāʾid of Toledo in 387/997–8, when he oversaw the construction of the bridge over the Tagus in al-Manṣūr’s name (Chapter 4).75 Since the person named on al-Manṣūr’s basin is both the director of the commission and fatā al-kabīr al-ʿāmirī, indicating a man of high status, Khalaf seems more likely than Khayra(h), whose signature on three columns in the ʿĀmirid extension to the Cordoba Mosque and on capitals now in Seville and Marrakesh identifies him, rather, as a skilled workman.76

Furthermore, the phrase al-fatā al-kabīr al-ʿāmirī also proclaims a direct connection between the ḥājib and the overseer of the marble basin. This direct relationship is likewise proclaimed in the designation of Zuhayr as an ʿĀmirid freedman (‘ibn muḥammad al-ʿāmirī’). His name and status, positioned prominently in the Pamplona casket’s main inscription, designated Zuhayr as ṣāḥib dār al-ṣināʿa, a post he still occupied when the Braga pyxis was created a short while later. While Faraj was certainly a talented master craftsman, in the case of the Pamplona casket’s manufacture, he reported to Zuhayr. Men like Zuhayr and Khayr/Khalaf were no doubt the executors of ʿĀmirid will regarding the designs and inscriptions on the objects made under their patronage. This cohort of elite slaves formed the backbone of the Andalusi court adminstration, and were personally loyal to their masters.77 This direct and perhaps physically close relationship – if the model of Jaʿfar’s residence near to the caliph’s at Madīnat al-Zahrāʾ is anything to go by – meant that the ḥujjāb themselves could keep a close eye on the planning and execution of the decorative schemes on objects and buildings being made for them. The fityān provided the mechanism by which the ʿĀmirid regents themselves could control the luxury arts industry and invest its products with certain messages conveying the construction and external projection of their image of power.

Could the Khalaf al-ʿĀmirī who may have been the overseer of al-Manṣūr’s marble basin and worked within the Cordoba mosque extension, and who was qāʾid of Toledo by the late 990s, be the same person who was carving (and signing) ivories in the 960s?78 Would a fatā designate himself as Khalaf al-ʿĀmirī to distinguish him from another well-known Khalaf, or because he was now operating under ʿĀmirid as opposed to caliphal patronage? Is it possible for someone who was a skilled worker carving ivories in the 960s to have worked his way up through the ranks to become fatā al-kabīr? And, having distinguished himself in that role, to be appointed qāʾid? We may never know the answer to these questions but it does raise the issue of the lifespan of craftsmen in the Dār al-Ṣināʿa. Would the Khalaf of the 960s still even be alive in the late 990s?

Oliver Watson has reflected on this issue in his study of the Doha casket (Figures 145–146).79 Observing that only 30 complete or near complete objects survive from a hundred-year period – taking the first textual reference to ivories (in the year 934) as the starting point, and the last dated Cuenca piece (from 1049) as the end point – he calculates that this equates to ‘a survival rate of some two to three pieces per decade’, though the spread across the decades is not even. In his Fig. 87, he charts the distribution of the surviving ivories datable from the 960s to the 1040s, against the working life of a trained craftsman, which he proposes was roughly three decades. He includes overlaps to account for the need to train apprentices by those already skilled.

This purely statistical exercise suggests that a craftsman practising in the 960s could still be working in the 980s. It does not factor in Khalaf’s age when he became skilled enough to sign his name on ivories, or the life expectancy at that level of society. If Khalaf had been apprenticed at age 15 (around 960), by 966 when he signed the Fitero casket he would have been in his early 20s; by 990, when he ‘retired’ – possibly having worked his way up to the position of fatā al-kabīr – he would have been about 50. Could he then have enjoyed another seven to eight years in a more honorific role, as qāʾid of Toledo, where he oversaw the bridge construction, perhaps as one of his last acts? It is perhaps unlikely, but not inconceivable. Such a principle of longevity in the craftsmen’s careers allows for the distinct possibility that craftsmen who had worked in the caliphal Dār al-Ṣināʿa later operated in the ʿĀmirid industry. This would explain the stylistic relationships, for example, between the tall deer with elaborate antlers or the mustachioed lions that are such a distinctive feature of the ʿĀmirid marble basins (Figure 157), and the lions and deer that appear on the surviving fragments of architectonic decoration from the Munyat al-Rummāniyya (Figure 85). The craftsman who carved the basins found at al-Rummāniyya, or the elaborate capital decorated with lions and birds of which only one volute survives (Figure 29), could conceivably have been employed on the ʿĀmirid marble basins some twenty to thirty years later, if he had been at the start of his career or still an apprentice around 970.

This discussion leads to thoughts on hands – the distinctive approach to carving that creates a ‘tell’ which allows us to tentatively attribute different objects to the same maker. This might be the recurrence of a particular motif or the way of rendering fur on animals’ bodies. Other examples of this will be highlighted in the discussion of particular objects in Chapter 7, however, to establish the principle, it is worth looking at the two caliphal ivories we know were carved by the same craftsman, the Fitero casket and the pyxis in the Hispanic Society.80 Both ivories feature a distinctive floral motif with four petals with tiny drill holes around its centre and incised lines along the petals,81 and the innovative use of uncarved areas of ivory to create a contrasting texture to the decoration. We could go further and start to define other characteristics of Khalaf’s carving style, such as the comparatively large amount of empty background he leaves around the vegetal designs on these two objects. ‘Highlights’ of uncarved ivory are seen to a more limited extent on the larger of two caskets in the V&A datable circa 961 (A.580-1910), whose bolder, blockier carving style is close to the Fitero casket, and might suggest this earlier casket was also made by Khalaf. It is quite different from the more delicate ‘miniature’ style seen on the smaller V&A casket (301-1866), which in turn exhibits a carving style very close to that of the cylindrical cosmetics box in Burgos. This is most evident in the inscriptions of these two ivories, but they also share a motif in the long thin leaf with two circles where it joins the stem.

These stylistic connections start to indicate where two or more objects might have been carved by the same craftsman, and again imply information about how the ivory workshops operated: in the case of the circa 961 group, at least two different craftsmen – Khalaf and another whose name is unknown – were working at the same time to produce objects for a daughter of ʿAbd al-Raḥmān, perhaps as part of a bridal trousseau.82

These stylistic connections also operated across media. There are many examples that could be cited. The decoration of the earliest ivories made for a daughter of the first caliph evokes the vegetal motifs of the Hall of ʿAbd al-Raḥmān III; while a marble panel from Madīnat al-Zahrāʾ features identical palmettes to those which adorn the Girona casket, signed by Badr and Ṭarīf (Figure 2). Similar palmettes are also seen in the wooden ceiling beams from al-Ḥakam’s maqṣūra at the Cordoba mosque,83 where the marble carving around the mihrab is also signed by Badr. The extremely elegant and well-executed inscription on the Pamplona casket is unusual when compared to other inscriptions in the ivory group: as Watson noted, ‘its broad, beaded letters, monumental scale, and sure calligraphy proclaims that it bears an important message. It is in an authoritative style hardly matched by any other Spanish ivory’.84 However, its distinctive beaded letters relate to the ‘pearl beading’ on carving at Madīnat al-Zahrāʾ, in particular on panels that come from the royal baths constructed at the heart of the palace zone in the 950s (Figure 88). The braided guilloche borders enclosing the medallion decoration of the Pamplona casket is also seen, for example, in the carved plaster designs added to the double-capitals in the possibly ʿĀmirid tribune in the Cordoba Mosque;85 while the same distinctive curlicue capitals are seen on the fragments of a marble basin now in the Alhambra Museum (Figure 134) and on the Braga pyxis (Figure 15). We can also note the recurrence of the heraldic eagle on ʿĀmirid objects. Other instances of such interconnections will be highlighted in the discussions of particular objects in the next chapter.

Figure 88
Figure 88

Pearl beading on a marble carving from the rooms adjacent to the Hall of ʿAbd al-Raḥmān III at Madīnat al-Zahrāʾ, 956–61

© Mariam Rosser-Owen

The pearl beading in the Hall of ʿAbd al-Raḥmān III and on the Pamplona casket was carved fifty years apart, so there is no question of the same craftsmen being involved in both. It is a technique that must have continued in the Dār al-Ṣināʿa, perhaps among the particularly skilled carvers (perhaps suggesting that on, the Pamplona casket, the master craftsman Faraj – who signed the lid – also carved the more tricky inscription). However, this shared approach across these different materials likely indicates the presence of a central overarching design authority, at a higher level of state control. This may indeed have been the case, especially where the inscriptions were concerned, which were probably conceived in the state Chancery. The phrasing, formulae and also physical appearance of the inscriptions on the objects – which are carved in relief in the same manner as official inscriptions – underline the ‘state’ level of the workshops that produced them. This notion also leads to the consideration of an issue that demands much fuller discussion than it has been granted hitherto: the likelihood that craftsmen worked across different, perhaps multiple, media.

1.2 Craftsmen Working across Media

Implicit in the foregoing discussion of signatures – especially where the same names are seen on the Pamplona casket and on architectonic elements from the ʿĀmirid extension to the Cordoba Mosque – is the idea that the same craftsmen worked on carving ivory caskets and these various elements in stone and marble. This idea has not received much support in ivory studies of recent years, or in medieval art history more generally, and appears to reflect a post-Renaissance perception of the specialisation of artists and craftsmen.86 The likelihood that medieval craftsmen did not specialise, but worked wherever they could apply their artistic skill, is currently the subject of some reconsideration. Recent studies have reminded us how Abbot Desiderius rebuilt the church of Montecassino (in southern Italy) before he became Pope Victor III in 1086, and because of the lack of craftsmen, he had the young monks trained in the full range of arts: gold, silver, bronze, iron, glass, ivory, wood, plaster and stone (ex auro vel argento, aere, ferro, vitro, ebore, ligno, gipso, vel lapide).87 Admittedly, it is not specified that the monks worked in all, or more than one, of these materials. For a later period, Sarah Guérin has pointed out that ivory carvers in Gothic Paris operated in guilds according to the types of images they produced, and were not specialised by material. Seven different guilds worked in ivory: the guild of painters and carvers of images were authorised to work in ‘all manner of wood, stone, bone, ivory and all types of paint good and true’; carvers of images and crucifixes specialised in bone, ivory and wood, though ‘all other materials can be used as long as the carver is familiar with the trade’.88 Guérin concludes, ‘In the various carving guilds, therefore, the same artisans worked in wood, stone, ivory, bone and any other suitable material. There were no “ivory specialists” in thirteenth-century Paris, but rather the discerning patron sought a highly skilled carver capable of working in a number of different media’. This documentary evidence has recently been reinforced by archaeological discoveries, which as Guérin observes, ‘encourage scholars to relax such formerly conceived categories as enmancheur versus coutelier, and to celebrate the dexterity and diversity of the artisans’.89 Though these formal categories relate to the products carved in Gothic Paris, they might easily be substituted with the words ‘ivory’ versus ‘architectural decoration’, and this reconsideration of non-specialisation in craft production has implications for the specific context of tenth-century Cordoba.

As mentioned above, in discussing the tiny percentage of ivory objects that have survived, at a rate he calculates as ‘two to three pieces per decade’, Watson observes, ‘This is surely nowhere near the required output needed to sustain an industry of such high quality and provide for the practice and training of craftsmen’.90 Such practice and training presupposes a constant and consistent supply of ivory material, which may not in fact have been the case, as we will discuss below (‘Materials’). At the end of his caption to Fig. 87, charting the potential working life of craftsmen, Watson comments, ‘It is important to note the very long periods for which nothing survives but during which time skills must have been preserved and objects produced’. In his note 38, he asks, ‘What other work might they have undertaken? I doubt that they were engaged in masonry or stucco work, as the fine motor skills needed for small-scale carving are totally different from those needed in working stone or on a large architectural scale’. Watson suggests woodwork as one possibility. The problem we immediately face when discussing this further in the Andalusi context is that very little carved wood survives from the caliphal period, unlike in Egypt where the conditions are dry and there are few forests, meaning that wood was a more valuable commodity in the first place, encouraging its reuse and preservation.91 The Cordoba Mosque was roofed with flat wooden beams, carved with abstract vegetal designs and painted, in a similar fashion to Umayyad monuments in Bilād al-Shām.92 But the carving was done in one plane only, is flat on the surface, and forms a comparatively simple pattern based on a framework of interlaced lozenges containing a palmette, alternating with split palmettes in the interstices. The ceiling beams which roof al-Ḥakam II’s maqṣūra display a rather more complex pattern, with some motifs that relate quite closely to the ivories – for example, the leaf bud with two circles where it joins the stem, or the distinctive palmette form seen on the Girona casket. But here again, the carving is flat and in one plane. In all these cases, the decoration was applied in paint on top of a layer of plaster, so there is no real comparison between the production of these ceiling beams and the sculptural carving of the ivories.

Another possible parallel, unfortunately now lost, is the world-famous minbar commissioned by al-Ḥakam II to adorn his extension to the Cordoba Mosque. According to al-Idrīsī’s eye-witness account, written around the middle of the twelfth century, this minbar was ‘unequalled for craftsmanship in the populated world’.93 The frequent comparisons made in its own day between the minbar now in the Kutubiyya mosque and the lost Cordoba minbar lead us to imagine its original appearance as resembling the minbar now in Marrakesh. The Kutubiyya minbar was even made in Cordoba, in 1137, so conceivably represents an updated copy of the Cordoba minbar. The finely-carved panels of this impressive piece of mosque furniture, carved from woods of various sorts with borders and details inlaid in bone and ivory, feature decoration that is stylistically unified across the different materials, and whose carving style is technically very close to that of the ivory caskets. On this basis Bloom argued that perhaps the best ivory carvers did not leave Cordoba after the Fitna,94 implying that they turned to working in wood after ivory material became less available. If we can indeed judge by the carving style of an object dating from 200 years after the Cordoba minbar, this demonstrates that the same craftsmen could work equally well in both wood and ivory.

The Cordoba minbar was made from ebony (though Bloom comments that this is more likely to have been African blackwood, a wood native to West Africa that was used on the Kutubiyya minbar), box and aloeswood. Ibn ʿIdhārī’s description probably draws on the account of the near-contemporary Ibn Ḥayyān; he writes that it was ‘inlaid’ (mudkhāl) with red and yellow sandalwood, ebony, ivory, and Indian aloeswood, and cost al-Ḥakam 35,705 dinars.95 After completing his new mihrab in 355/965, al-Ḥakam ordered the pre-existing minbar to be moved there – presumably a ninth-century minbar dating from ʿAbd al-Raḥmān II’s mosque extension, just as al-Ḥakam had the columns and capitals from the old mihrab to be relocated to the new. Al-Ḥakam then commissioned the new minbar, which was finished in five years according to Ibn ʿIdhārī, seven according to al-Idrīsī.96 Its ‘carpentry and inlay’ were carried out by six craftsmen, ‘apart from those apprentices serving them’. This gives us another insight into the workshop structure required to produce such a large and impressive royal commission: assuming a minimum of one apprentice per master craftsman, at least twelve craftsmen were employed on al-Ḥakam’s minbar, working simultaneously. By way of comparison, there would originally have been more than a thousand panels on the Kutubiyya minbar.97

The materials employed for the Cordoba minbar were extremely expensive, as appropriate for a royal object. Sandalwood and aloeswood are among the principal aromatic substances imported at high cost from great distances away, and highly sought after in caliphal Cordoba’s sophisticated culture.98 It is also interesting to consider the implications of carved ivory panels on the minbar. All the Andalusi ivories we know of – both extant objects and through textual references – have a secular function, but given the abundance of ivory available at the exact time the minbar was being made, it seems logical to suppose that some ivory panels were carved to decorate it.99 This raises a tangential question of what other ivories might have been produced for religious purposes. Maribel Fierro has noted that, ‘in an effort to overcome his deficient legitimacy’, Hishām II devoted himself to collecting relics associated with pre-Islamic prophets. She asks, ‘Could some of the ivory caskets preserved from this period in al-Andalus have been used as containers for these relics?’100 It is highly unlikely that any of the extant ivories served this purpose, being so blatantly secular in their iconography and epigraphy, but it is certainly plausible that other ivory caskets, which have not survived, were made for this purpose.

Ivory and wood have an integral relationship to each other in the construction of furniture, where the wood provides the framework into which carved ivory panels are inset. Though this technique is predominantly used for the ivory objects that survive from Fatimid and Mamluk Egypt, only one potential example of this is so far known from al-Andalus: this is the rectangular panel now in the Metropolitan Museum in New York (Figure 32).101 Unusually, this panel has no space left in the carving for mounts, which suggests that rather than being one of the walls of a casket, it was instead inlaid into a larger piece of furniture. The decoration would originally have continued at the right and left sides of the panel, where today is only preserved half of the original pair of birds or animals. Perhaps it originally formed part of a frieze on a door or cupboard. If so, it would be a unique example of this kind of inlaid furniture from al-Andalus at this period – but if ivory was inlaid into the Cordoba minbar, why not into other large pieces of furniture, for secular contexts? Another aspect that might suggest that this panel was part of the furnishing in a palace interior is the fact that this is the only example among the Andalusi ivories where tiny pieces of glass have been found set into the beaded bands which surround the decoration.102 If the object into which this panel was inset were to move, as doors were opened or closed for example, or if light were to move in the room in which it stood, these tiny pieces of glass would catch the light and enhance the apparent movement of the dancers.

Wood was also integrally related to the production of luxury objects, since it was used as a core for the Girona casket and several of the large ivory caskets. There can be no doubt that wood was available to be carved, since it was also needed for architectural construction. However, we cannot rule out the possibility that ivory carvers were also working with other materials besides wood.

Significantly, it has gone unnoticed until recently that the hardness of ivory is equivalent to that of limestone on the Mohs scale.103 In fact, at 2.5–3 it sits between gypsum (2.0) and calcite/limestone (3.0), and close to marble (recrystallised limestone, at 2.5–5), with gold (2.2) and silver (2.4) on a similar scale. As Jeheskel Shoshani explains, ‘Being a biological product, a tusk … is relatively soft, equal in hardness to calcite mineral (number 3 on the Moh’s field scale of hardness used by mineralogists). Hardness and therefore carvability of ivory differs according to origin, habitat, and sexual dimorphism [of the elephant]. [Thus] the ivory from western and central Africa (where the forest subspecies, Loxodonta africana cyclotis, is more prevalent) is said to be the best ivory because it is harder than all other ivories’.104

The ivory being carved in al-Andalus, which was sourced from West Africa as discussed below, was at the harder end of the scale of ivory hardness, and as hard to carve as limestone and some marble. This implies that a carver skilled in one material could equally turn his skill to the carving of the other. When we consider that limestones of various kinds were the primary material used for the support elements and decoration throughout Madīnat al-Zahrāʾ (see below), this opens up the distinct likelihood that craftsmen in the Andalusi Dār al-Ṣināʿa worked across these three materials.105 This would explain the stylistic connections across media which imply the same ‘hands’ at work, as mentioned above.

As a twentieth-century manual of ivory-carving explains, the craft generates little waste, and so can effectively be conducted anywhere.106 The main requisites would be storage space for the tusks – which, though potentially huge at 2 to 3 m long, may never have been numerous at any one time – and for the finished objects, which, if made for specific commissions as they probably were, may not have spent much time in the workshop after they were finished; and one or more turning tables to power the bow drill, though this may also have been done by a craftsman sitting on the ground and powering it with his feet, as still happens today in Middle Eastern wood-turning workshops. At the height of the industry – at the time of the production of the Pamplona casket, say – space would be needed for around six craftsmen working simultaneously, and their equipment. Consequently, the Andalusi ivory workshop did not need to occupy much space, and this carving could well have taken place in the same or a neighbouring space as the workshops producing architectural decoration, since they probably employed the same makers. The stone carvers, or the most skilled among them, may have been redeployed to ivory work as and when the raw material became available. Similarly, Lawrence Nees made the suggestion that the Carolingian ivory carving industry came into being because of the sudden availability of the two elephant tusks following the death of the war elephant, Abū’l-ʿAbbās, which had been a gift to Charlemagne from the Abbasid caliph Hārūn al-Rashīd: artisans familiar with other media applied themselves to this new material.107 A scenario of dependence on availability could explain how, as early as the 930s, when, according to the sources, the ivory industry begins, apparently founded ex nihilo by ʿAbd al-Raḥmān III (see below), there were craftsmen available to carve it, as they would previously have worked on caliphal architectural projects. And why, by the first extant objects of the 960s, the craft had been so utterly mastered.

There is, admittedly, a difference in scale to account for: that is, ivory is worked ‘in miniature’ while marble and limestone are worked on a more monumental level. However, not all surviving architectonic elements from tenth-century al-Andalus are on the same scale, and there is great variety in the sizes of marble capitals. By way of an example, a small marble volute from a capital, now in the V&A (Figure 89), is carved on a miniature scale that has close affinities, technically as well as stylistically, with the ivories. Not all stone carving is at the macro level, but it is also important to recognise that carving is carving. An Andalusi workman who ‘did carving’ may have had to use slightly different tools and materials, but he would adapt, or he would not advance within the Dār al-Ṣināʿa. Let us not forget that the crafts and craftsmen we are discussing were not responding to the whims of the open market, but functioning within a rigidly organised industry directed towards serving the ruler’s needs and desires.

Figure 89
Figure 89

‘Miniature’ carving in marble on a volute from Madīnat al-Zahrāʾ, 950–970; Victoria and Albert Museum, A.104-1919

© Victoria and Albert Museum, London

This argument implies the mutuality of the different crafts produced under caliphal patronage in tenth-century Cordoba. Under the umbrella organisation of the Dār al-Ṣināʿa there may have been physically separate ateliers – though likely in close proximity to one another – for carved marble and limestone (encompassing both architectural decoration and objects, such as fountain basins), ivory, wood and metalwork (gold and silver, at least), while textiles and ceramics were produced in specific locations in an industrial quarter to the north of the old city of Cordoba; perhaps foundries for bronze production were located there too. Though each separate atelier may have had its own ṣāḥib, some of whom controlled two or more ateliers, at the lower level the artisans moved between them, transferring their technical skills across media, responding to demand and the availability of raw materials. Such a transfer of craftsmen facilitates the sharing of styles and decorative devices across media, through the possibility that the same ‘hands’ were employed on different types of object. It also implies a lack of craft specialisation within the Cordoban Dār al-Ṣināʿa. It is also intriguing to observe that, as far as we can tell from the extant pieces, the dates of important objects do not tend to overlap with those of large-scale architectural projects (because they both employed the same master craftsmen?), perhaps indicating that the artisans only worked on objects in the slower periods between royal construction projects. This needs a fuller comparative study and the picture might change as more inscribed architectural elements and objects are published. An interesting exception to this rule seems to be al-Manṣūr’s basin, to which we will return below.

The recurrence of the same names in signatures on luxury objects and architectural decoration is thus largely due to the same craftsmen working on both. The Girona casket, for example, is signed under the lockplate by two men (ʿamal Badr wa Ṭarīf ʿubaydayhi, Figure 2, Appendix 4.1) whose names also occur in the Hall of ʿAbd al-Raḥmān III,108 and as observed above, there is close artistic similarity between the two contexts, as well as the ceiling beams in al-Ḥakam’s maqṣūra. Given the relative closeness of silver to marble and limestone on the Mohs scale, as well as the carving relationship between wood and ivory (which implies a carving relationship between wood and marble), we might postulate that Badr and Ṭarīf worked in precious metals and woodwork as well as marble. It is perhaps an accident of survival that we have no ivory caskets featuring their signatures.

But craftsmen might have been shared between other media as well. We have seen that the craftsmen Bāshir and Mubārak put their names to the innovative double-capital probably from al-Madīnat al-Zāhira, alongside that of the master craftsman Faraj (Figures 37–38). Bāshir and Mubārak are also names that appear ‘in the manner of signatures’ on some of the Andalusi green-and-brown ceramics.109 Other green-and-brown ceramics are signed by Rashīq and Naṣr,110 and Manuel Acién suggested the same Rashīq might be the craftsman who signed bronze oil lamps found at Liétor.111 The relationship between these crafts might seem obscure, but the ability to model clay would have helped in the manufacture of metal objects using the lost wax technique. Of course the signatures on the ceramics may indicate the author of the painted design rather than the potter, though the principle of the lack of specialisation still holds. Certainly a facility with drawing would make it easier to transfer (or conceive) designs on whatever medium was about to be carved. That designs were drawn on the surface first is indicated by a fragment of wall decoration from Madīnat al-Zahrāʾ, now in the V&A collection, where the top of the block preserves the etched outline of a floral scroll which was not, for whatever reason, actually executed (Figure 90).

Figure 90
Figure 90

Incised design along the top of a section of wall decoration from Madīnat al-Zahrāʾ, indicating how the carving was planned, 950–970, limestone; Victoria and Albert Museum, A.118-1919

© Victoria and Albert Museum, London

1.3 Materials

The last but in fact the most important aspect of the luxury arts industry was the supply of raw materials. The sourcing and gathering of these in sufficient quantities was also controlled by the Umayyad state according to well-organised and centralised structures, as we can see most explicitly in the Calendar of Cordoba, written c. 961 for the new caliph al-Ḥakam II, by his kātib ʿArīb ibn Saʿd (d. 980) and the bishop of Elvira, Recemundo. This stipulates the specific times of the year at which raw materials destined for the state industries – including the army, the Dār al-Ṣināʿa and especially the Dār al-Ṭīrāz – should be planted and harvested. The local officers of the state bureaucracy were charged with gathering them by letters sent out from the chancery in Cordoba: in February, ‘they sent out a missive with instructions for the recruitment of men for that year’s summer campaign; in March, the orders consisted of purchasing horses for the government; in May, they should collect cochinille, silk and bentonite (Fuller’s earth) for the caliphal textile workshop (ṭirāz); in June they estimated the extent of the year’s harvest and assigned guardians for the granaries that would receive the tithes (ʿushūr); in this same month they should also send the antlers of deer for the fabrication of bows; in August they requested them to send silk and blue dyes for the caliphal textile workshop; and in September they were requested to send madder’.112 The process for securing other necessary materials for building projects and the luxury arts industry was surely no less carefully structured and regulated, though the state probably had less control over the sources of these materials. In important objective of Andalusi foreign policy in the late tenth century was surely to secure the requisite natural resources that this now highly complex state needed to operate.

1.3.1 Stone and Marble

Though we do not have such explicit written evidence for the gathering of larger scale material such as stone and marble, anecdotal evidence exists for the sourcing of marble spolia from sites around the Mediterranean, especially North Africa, to supply sufficient numbers of columns, capitals and bases to begin construction at Madīnat al-Zahrāʾ. The initiative for this is credited to ʿAbd al-Raḥmān III himself, and the harvesting of heavy items of spolia on such a large scale could only have been undertaken at the level of the ruler’s fiat. Recent excavations at Utica in Tunisia, for example, are suggesting the systematic despoliation of the classical ruins in the early Islamic period to provide building materials for the new urban centres.113 Another source of raw materials was extraction. The use of marble in the Visigothic period to carve elements such as basins and bas-reliefs may indicate a continuous but low-level working of Iberia’s abundant Roman quarry sites between the third and seventh centuries, though perhaps a more likely source for the material are Roman objects or inscriptions that were recarved.114 However, the scale of the caliphal architectural projects in the mid-tenth century meant that it was no longer sufficient to resort to spolia for the supply of architectonic elements; instead, marble needed to be freshly quarried. As Antonio Vallejo’s work has shown, marble was introduced to Madīnat al-Zahrāʾ in its second phase of construction, in the 950s, implying that it took some time for ʿAbd al-Raḥmān to mobilise the resources sufficiently.115

In general, the stone quarries were located close to Madīnat al-Zahrāʾ: sources of the limestones used for construction, architectonic supports and paving were located within kilometres of the site.116 Other limestones, particularly sought-after for their aesthetic effect – such as the reddish ‘false breccia’ used for columns, and the limestones used for capitals, bases and wall decoration – were extracted from the Sierra Subbética, some 50 km from Cordoba.117 Marbles, such as the coloured marbles used for columns in the Cordoba Mosque, either came from Roman or Late Antique spolia (see below), or were extracted from the region of Almadén de la Plata (Seville) or Estremoz (Portugal).118 The ability to maintain the harmonious pattern of alternating varicoloured marble columns in al-Ḥakam’s mosque extension demonstrates the caliph’s tight control over the marble industry. There was also an element of conspicuous display: the maintenance of consistent designs and colours throughout a monument, and the unified decorative programme of the Hall of ʿAbd al-Raḥmān III, was a clear statement of the caliph’s power over the state infrastructure, and this was proudly signalled in the inscriptions that adorned nearly every purpose-carved marble element.

By the time of al-Manṣūr’s ḥijāba, the infrastructure for quarrying marble on a massive scale was obviously well-enough established to supply the huge numbers of newly-carved columns and capitals that furnish his mosque extension, which required more than twice the number of elements than al-Ḥakam’s extension, and to maintain the pattern of alternating coloured marble in the columns (Chapter 5). I will focus my discussion here on the stones employed in the production of objects for the ʿĀmirids, in particular the water basins. Some of these stones relate to those used to create architectural adornment for al-Manṣūr’s palaces, discussed in Chapter 4.

The most likely source of huge trough-shaped blocks of marble from which the impressive ʿĀmirid fountain basins were made are Late Antique sarcophagi, which were imported into Hispania from Rome in the late third and fourth centuries AD (Figures 91, 92, 97). By the tenth century, these were clearly still available, as more than twenty figural sarcophagi have been recovered from Madīnat al-Zahrāʾ, some of them reused as fountain basins in the centre of courtyards.119 Of course not all Roman sarcophagi were originally elaborately decorated – there is an uncarved sarcophagus among the ruins at Madīnat al-Zahrāʾ, for example (Figure 91);120 and Ambrosio de Morales described another located within the principal cloister at the Monastery of San Jerónimo, on the slopes above Madīnat al-Zahrāʾ, with which the two bronze fountain heads in the form of deer were associated (Figure 10).121 The backs of Roman sarcophagi were often uncarved, where they were placed against a wall.

Figure 91
Figure 91
Uncarved Roman sarcophagus excavated at Madīnat al-Zahrāʾ
© Conjunto Arqueológico Madinat al-Zahra
Figure 92
Figure 92
Roman sarcophagus with decoration of arches, fourth-fifth century AD, found at Avenida de la Cruz de Juárez, Cordoba; Museo Arqueológico y Etnológico, Cordoba, inv. CE021992
© Rose Walker / Photo: JohnBattenPhotography
Figure 93
Figure 93
Marble panel with decoration of three arches, said to be from the Alcázar at Cordoba, 950s–60s; Museo Arqueológico Nacional, Madrid, inv. 50369
© Museo Arqueológico Nacional, Madrid / Foto: Ángel Martínez Levas

The ʿĀmirids may have made use of uncarved sarcophagi for their basins – taking the reuse one step further, and recarving the sarcophagi with their own designs. Such reuse would clearly go some way to explaining the size and shape of the ʿĀmirid basins. Other ʿĀmirid objects, such as the enormous marble window grilles (measuring 1 metre wide by nearly 2 metres high) that were newly carved for the mosque extension, could also have been carved from the sides of sarcophagi (Figure 54).

That Roman marble objects were reused by the ʿĀmirids is shown by the foundation inscription in Lisbon discussed in Chapter 4, carved at the base of a reused Roman tombstone (Figure 19, Appendix 4.6). That they were also converted into basins is shown by an object that tells an intriguing story of multiple reuse. A funerary stele now in the Archaeological Museum in Rabat commemorates Abū Yaʿqūb Yūsuf, the Marinid sultan of Morocco who died in 1307 (Figure 95).122 His stele is carved on a thick slab of marble whose other face preserves part of a Latin inscription in the name of Aulus Caecina Tacitus, likely to have been the governor of Baetica – the fertile Roman province in the south of Hispania, of which Corduba was the capital – and probably dating from the third or fourth century AD (Figure 94). However, the narrow sides of the stele also bear traces of carving: though one side is badly damaged, the other shows that this slab was clearly once part of a fountain basin, stylistically related to the group of ʿĀmirid basins, as discussed further in Chapter 7 (Figure 96). The stylistic significance of these surviving traces of carving on the sides of the stele has not been sufficiently noted, and where it has been mentioned it has been called (oddly) ‘Byzantine’. The object can thus be reconstructed as an official Roman inscription, carved on a large thick block of good-quality white marble, which was reused in the tenth century to create another fountain basin for ʿĀmirid palatial gardens. This basin was later brought to Morocco, perhaps at the same time as ʿAbd al-Malik’s basin (Chapter 7: 2.3.1), though in this case its eventual destination may have been Rabat rather than Marrakesh. In the fourteenth century, a beautiful piece of marble, which perhaps retained a nostalgic association with al-Andalus or, by extension, with the Umayyads, was chosen to create a fittingly royal memorial for one of the Marinid sultans, and installed in their necropolis at Shāla.

Figure 94
Figure 94
Funerary stele of Abū Yaʿqūb Yūsuf: side with Roman inscription, third–fourth century AD
© Archaeological Museum, Rabat
Figure 95
Figure 95
Funerary stele of Abū Yaʿqūb Yūsuf: side with Marinid tombstone, 1307; Archaeological Museum, Rabat
© Péter Tamás Nagy
Figure 96
Figure 96
Funerary stele of Abū Yaʿqūb Yūsuf: side with basin decoration, late tenth century; Archaeological Museum, Rabat
© Mariam Rosser-Owen

The form and function of the ʿĀmirid basins may have been inspired by the sarcophagi redeployed as fountain basins in the royal zones of Madīnat al-Zahrāʾ or still visible in other parts of Cordoba, and this relationship might also explain some aspects of the style of the ʿĀmirid basins. It may, for example, have suggested an origin for the arrangement of the decoration within an arcade, as seen on the back of al-Manṣūr’s basin. On sarcophagi it is not uncommon to see figures, such as philosophers or characters from the New Testament, discoursing under arches whose architectonic elements are very clearly defined, just as they are on al-Manṣūr’s basin. A particularly monumental example from the fourth to fifth centuries, with carving almost in the round, was excavated in Cordoba and is now in the Archaeological Museum there (Figure 92). The triple arch motif was already in use under the Umayyads, as evidenced by the thick marble panel that probably comes from the Alcázar at Cordoba (Figure 93).123 But other features of the basins as well as artistic motifs that seem to enter Andalusi art at this time – such as the griffin and the heraldic eagle – could reference the decoration of surviving Roman or Late Antique carvings. Many of the sarcophagi feature on their short ends a monumental griffin, depicted with ears, sitting in profile (Figure 97); and, of course, the heraldic eagle was the pre-eminent Roman Imperial symbol, available as models on such objects as the late second-century marble altar dedicated to Venus Victrix from Mérida and now in the Museo Arqueológico Nacional (Figure 98).

Figure 97
Figure 97
Griffin on the side of a Roman sarcophagus, 230–270 AD; Museo Arqueológico Nacional, Madrid, inv. 1999/99/185
© Mariam Rosser-Owen
Figure 98
Figure 98
Roman eagle, on an altar dedicated to Venus Victrix, Mérida, late second century AD; Museo Arqueológico Nacional, Madrid, inv. 20220
© Mariam Rosser-Owen

Could the form of the Roman and Late Antique objects that provided the models, if not actually the raw material, for the ʿĀmirid basins also have inspired their decoration, with designs of arches as well as apotropaic animals on the short ends? Might the carvers of the Dār al-Ṣināʿa have looked to pre-Islamic models for inspiration for certain motifs, especially those with which they were unfamiliar? As will be discussed in Chapter 8, it was ʿAbd al-Raḥmān III who decided to introduce eagles to his military banners, and al-Ḥakam who may have had them first carved onto objects (Figure 86). If the Andalusi craftsmen had no other precedents for how to carve a sufficiently regal-looking heraldic eagle, they would reasonably have looked to the available Roman and Late Antique models. In the same way, as discussed in Chapter 4, the worms or water-snakes on ʿĀmirid capitals might imply a misreading of fleurons on Roman capitals from Mérida that had been reused in the Cordoba Mosque (Figure 25), or the looped body of the griffin on one of the capitals probably from al-Madīnat al-Zāhira (Figure 26) may have been inspired by the representation of sea-monsters in Roman carvings.124

1.3.2 Ivory

Ivory was another luxury material for which a reliable supply was needed once its use to make spectacular objects had been established in the Andalusi luxury arts industry. There is no evidence for ivory-carving in al-Andalus before the caliphate of ʿAbd al-Raḥmān III, and it seems that the industry was created ex nihilo under the direct auspices of this ruler. Though no extant objects date before the 960s, the industry was clearly established by 934 – only five years after the declaration of ʿAbd al-Raḥmān’s caliphate – when ivory and silver perfume containers were listed among the extensive and luxurious diplomatic gift that the caliph sent to his main ally among the Berber lords of the Maghrib, Mūsā ibn Abī’l-ʿAfiyya:125

“… nine [containers] ranging from pyxides to caskets, all of which were filled with different kinds of perfume; among them a round silver pyxis containing incense mixed with ambergris; a pyxis of white ivory, containing sticks of frankincense seasoned with ambergris; another ivory pyxis, also with silver hinges, that had an Iraqi vase inside filled with an excellent ghāliya;126 a third pyxis of ivory with silver hinges and a flat lid, containing royal frankincense; a glass casket with a silver lid and silver chain, containing pungent musk powder; a fourth ivory pyxis with silver hinges, too, containing the powder used by kings against sweat in the summer; a gilded Iraqi flask with rosewater of the Iraqi caliphs …”127

Though only four ivory containers are mentioned here, the gift included another ivory object: ‘a brocaded silk (dibaj) sheath containing a sultan’s large ivory comb to comb the beard’.128 This precious historical passage thus indicates that the ivory industry was clearly established by this date, and that its products were already good enough to send out as diplomatic gifts worthy of a newly-minted caliph.

As discussed in Chapter 1, raw ivory was one of the commodities of trans-Saharan trade and was sourced via the Umayyads’ Berber clients, who commanded the entrepôts where these materials were gathered and traded with the nomadic tribes who trafficked the trade routes along the western flank of Africa. The supply – and, I would argue, the meaning – of ivory is intimately connected with the supply of gold. Both came from the sub-Saharan region (along with black slaves, salt and other commodities) and it is interesting to consider the two materials in parallel, to see what this suggests about the abundance or regularity of ivory availability in al-Andalus. Though the extant Andalusi ivories were made over a broad period of a hundred years, from the mid-tenth to the mid- eleventh century, their dates of production coincide with three major phases of patronage: the Umayyad caliphate, with all the extant objects attributable to the 960s; the ʿĀmirid period, with the extant objects dating from the 990s and 1000s; and the industry founded by the Taifa kings of Toledo, in an effort to legitimise their rule by reviving this splendid facet of the caliphal court. The extant Taifa objects date between 1026/7 and 1049/50, and it is likely that they were made with stocks of raw ivory appropriated from the caliphal Khizānah during the Fitna years, rather than by sourcing new reserves for themselves. While it is inadvisable to base firm arguments on the absence of evidence, these clusters in production – in particular between 960–70 and 990–1010 – may also reflect the patchy availability of the raw material.

As we have seen (Chapter 1), Umayyad desire for gold was stimulated by competition with their caliphal rivals, the Fatimids. ʿAbd al-Raḥmān III struck his first dinars in 929, as soon as he declared his caliphate, probably using gold from other coins (Aghlabid, Midrārid and possibly Fatimid) that were already circulating in al-Andalus. The growing interventionism of the Umayyads in the Maghrib – heralded by the conquest of Ceuta in 931 – marked a major policy objective, which was to source sufficient gold to mint dinars, such a potent symbol of being a caliph. Under ʿAbd al-Raḥmān, control of the trans-Saharan trade routes was far from total, as indicated by the fact that dinars were not minted annually, and there was great fluctuation in quality, especially in the last decade of his reign when there appears to have been a gold shortage and almost no dinars are minted.129 This apparent gold shortage continues into al-Ḥakam’s reign, so that there is a gap in gold emissions between 958 and 967 – curiously, this is exactly the period when the first extant ivory objects were produced, so either mapping ivory supply against gold supply is not totally reliable, or the craftsmen of the Dār al-Ṣināʿa were able to make use of a stockpile of ivory that had allowed them to produce the perfume containers for Mūsā ibn Abī’l-ʿAfiyya’s gift thirty years before. As the caliphate matured and the Umayyads’ control in North Africa was consolidated – thanks largely to al-Manṣūr’s policies and carefully maintained relationships with his Berber clients – the gold supply becomes more regular and the high quality of the dinars is very stable with only minor variations. From 967, there is an annual series of gold emissions until the end of al-Ḥakam’s reign. During Hishām’s caliphate, there is a growth in dinar emissions, but a gap in minting from 368/978 to 377/987, and the year 382/992 when no gold coins are minted. There are particularly high outputs of dinars between 390/999 and 394/1003, a period in which several ʿĀmirid ivory objects are also produced. There is then a gradual decline in gold emissions between 395/1004 and 399/1008, after al-Manṣūr’s death.

Mapping the dates of the extant ivories against the gold coin emissions does not provide a perfect match. The Pamplona casket and the Braga pyxis (Chapter 7: 2.1, 2.2) were produced in the period when gold coin emissions were declining. This is perhaps because the ivory had entered al-Andalus at an earlier period, and had been stockpiled. The point to keep in mind is that there may not have been a constant and consistent supply of ivory during the second half of the tenth century, in the same way that interruptions in the gold supply are reflected as gaps in coin emissions. This inconsistent supply may reflect the way that elephants were hunted and ivory was sourced. This is an area that still requires further research, as more information emerges from African archaeology, but the picture that is emerging does suggest the deliberate hunting of large mammals for their ivory as well as the probable harvesting of tusks from already-dead elephants. Ivory obtained in this way by farming communities in southern Africa was brought to trade entrepôts south of the Sahara and gathered there until traders carried it north into the territories dominated by the Umayyads’ Berber allies.130

Ivory’s possible exceptionality of course makes it a more expensive material. Though we do not have written evidence about the value of ivory in al-Andalus at this period, its apparent near-monopoly as a material employed at the heart of the Umayyad court is telling in this regard. Sheila Blair reflects on the expense of ivory, by drawing a comparison with Hishām II’s casket, executed in wood and silver-gilt (Figure 1). She argues that this imitates an ivory casket, since the mounts are cast integrally to the surface decoration thus replicating the mounts that give external support to the ivories, which are not necessary on metalwork.131 Intriguingly, the form and dimensions of the Girona casket are very similar to those of the Pamplona casket.

Ivory may also have arrived at irregular intervals in the form of tribute or diplomatic gifts from the Umayyads’ Berber allies. As we saw in Chapter 2, the information apparently transmitted by al-Maqqarī, that Zīrī ibn ʿAṭiyya gave ‘8,000 pounds weight of elephant tusks’ to al-Manṣūr in his diplomatic gift of 994, can no longer be relied upon. Nevertheless, this type of tribute or gift is not unlikely as a mode of transmitting raw ivory to the Umayyad overlords in al-Andalus. Given the great victory Zīrī had just won over the Fatimids, extending his own power from the Atlantic coast into the mountains of modern-day Algeria, if ivory had been available he would surely have sent it to al-Manṣūr. Only a few years after Zīrī’s embassy, in 389/999, the ivory pyxis for al-Manṣūr’s younger son, Sanchuelo, was created, the first of the extant ʿĀmirid ivories (Chapter 7: 3.1). Following the dates of the extant objects – and admittedly this relies on the accident of survival – this object implies the revival of the ivory industry after a thirty-year gap, since no dated ivories survive between this and Ziyād ibn Aflaḥ’s pyxis, made in 969 (Figure 12). After Sanchuelo’s pyxis there follow in close succession the Doha box (4.1.1), the Pamplona casket (2.1), the Braga pyxis (2.2), probably the uninscribed casket now in the V&A (Figures 150–155) and the small box in the Bargello (4.1.2) – the reasons for assigning these ‘anonymous’ objects to the ʿĀmirid period are discussed in more detail in Chapter 7.

This flourishing in the 990s no doubt reflects the stimulus that the industry received if not specifically from Zīrī ibn ʿAṭiyya’s gift, then from the Umayyad allies’ conquests in North Africa, which put the trans-Saharan trade routes under direct Umayyad control. Is it even possible that al-Manṣūr specifically sought raw ivory in order to reinvigorate the craft of ivory carving, which had developed such a close caliphal association through ʿAbd al-Raḥmān’s sponsorship in the 930s and al-Ḥakam’s in the 960s?

Given the enormous size of African elephants and the great lengths to which their tusks can grow – up to 2 or, exceptionally, 3 metres132 – we may reasonably posit that several objects could be carved from a single tusk. No precious material would have been wasted, and the blocks excavated from the centres of pyxides and caskets were probably used for smaller items such as chess pieces, so that a single tusk could produce a whole ensemble of objects. Given the relatively small size of the objects made during the 960s, it is conceivable that the entire group of extant objects made during al-Ḥakam’s reign might have been carved from a single pair of enormous tusks.133 If the ivory supply mirrored the gold supply in becoming more abundant and reliable during the late tenth century, this makes it probable that many more ivories were produced by the ʿĀmirid Dār al-Ṣināʿa than have survived. Of course it is also possible that the raw ivory was not all consumed within Cordoba. After the Fitna, some of this leftover ivory may have been taken to Cuenca, where the Banū Dhū’l-Nūn established an ivory-carving workshop (Figure 174).134 The production techniques employed for the objects made there – thin openwork plaques attached to a wooden core to produce caskets, or tiny pyxides such as that in the David Collection – indicate that ivory as a raw material was much scarcer, and the craftsmen were trying to make the most out of the material available to them. This is likely to be because they no longer had direct access to the trans-Saharan trading network. As discussed in this book’s Conclusion, carved objects in ivory also began to be made in the Christian Iberian kingdoms in the late tenth century, though Glaire Anderson has recently argued that one of the key milestones in the supposed development of that industry – the processional cross from the Monastery of San Millán de la Cogolla – was in fact made in Cordoba in the early 960s.135

The ivory supply in al-Andalus might have been irregular, but the material was clearly abundant enough to allow a production technique in which objects were carved from solid blocks of ivory. Andalusi ivory carvers of the caliphal and ʿĀmirid periods, at least, show no sign of needing to adapt their construction methods to maximise the amount of available material. As Anthony Cutler has pointed out in reference to the Hispanic Society pyxis, the craftsmen even seemed to luxuriate in the material.136 In analysing their carving techniques, Cutler highlights the way these ivories show a ‘display of skill for its own sake’: for example, the fact that the background walls, behind the carving, are usually only millimetres thin, resulting in translucency when viewed against the light. The caskets were not meant to be viewed this way, this feature has nothing to do with their function – it is purely incidental, a ‘demonstration of artistic virtuosity’.137 This virtuosity is also seen in the carving of their decoration, in a rich sculptural style that Cutler calls ‘hypertrophic abundance’. Such construction methods imply a plentiful supply of raw material.

Even the largest surviving object – the Pamplona casket (Chapter 7: 2.1), whose walls are nearly 40 cm long – is structured from individual, thick ivory plaques, pinned together at a 45 degree angle, without a wooden core. Additional support to the interiors of such caskets would have been provided by their external mounts in precious metals but also by textile linings, providing a material and visual connection between the textiles manufactured in the Dār al-Ṭirāz, and the ivory caskets made for the caliph’s family or powerful political allies. As we saw above, there was likely also a personnel connection across these industries, facilitating the transfer of decorative modes between media. The association of most of the inscribed ivories with members of the caliphal family, and the close iconographical relationship between the ivories and surviving examples of Andalusi textiles, indicates that ivory-carving in al-Andalus was perceived as an ‘art of power’.138 For this reason, these objects played an important role in diplomacy and gift exchange, as we can uniquely reconstruct for the Braga pyxis (Chapter 2).

The creation of these spectacular objects was part of the regents’ construction and external projection of their image of power, carefully modelled on that created by ʿAbd al-Raḥmān some fifty years earlier. As ʿAbd al-Raḥmān had done, the ʿĀmirids took raw materials from their Berber clients, and ‘converted’ them into finely-worked precious objects which they then ‘returned’ in the guise of diplomatic gifts, sending a message of political superiority through their ability to excel at the creation of beautiful artworks. The gifting of objects carved from this precious material – together with expensive perfumed contents, likewise sourced from wide-ranging international networks – to clients or allies in North Africa, Christian Iberia, Byzantium, even Ottonian Germany (Chapter 2), showed the world that Cordoba could attract the best craftsmen to work in a material that was sourced via trans-Saharan trade – trade that the Umayyads controlled at the specific expense of the Fatimids and their North African clients. Some objects seem to have been deliberately designed to flaunt this association, by displaying as much of the ivory material as possible. The cosmetics box in Burgos, for example, when viewed closed, presents an entirely plain surface of pure white ivory; remarkably straight-sided for its stunning length of 46.5 cm, the Burgos box boasts of the huge size of the elephant tusk from which it was created (Figure 99).

The creation of these spectacular ivory objects and their presentation as gifts was intricately connected to the projection of the caliphal message, that Umayyad hegemony extended over the western Maghrib and its peoples, and thereby controlled access to the West African trade routes and the curious and precious goods that these routes supplied. Control of this trade symbolised victory over the Fatimids. The reality of this message only held more true as the Umayyads’ hold on North Africa strengthened during the reigns of al-Ḥakam and Hishām, in both cases due to the clever politics played by al-Manṣūr (Chapter 2).

1.3.3 Perfumes and Perfume Containers

As the list of gifts from ʿAbd al-Raḥmān to Ibn Abī’l-ʿAfiyya makes clear, the ivories were containers for luxurious perfumed contents, which may actually have been the more expensive and sought-after gift, with the ivory casket its precious ‘wrapping paper’.139 Perfumes and cosmetics were an important element in the sophisticated culture of Umayyad al-Andalus, and the raw materials to make them were themselves sourced from great distances away. Musk and camphor, for example, both mentioned in the poetic inscription on the Hispanic Society pyxis, were imported from South and East Asia. Camphor is a white granular substance distilled together with camphor oil from the sap of the tree Cinnamomum camphora, a large evergreen native to East Asia, from India to Japan;140 while musk derives from the gland secretion of the male musk deer, and hails from Tibet and China.141 This speaks eloquently to al-Andalus’s position within far-flung networks of international trade – the Iberian Peninsula was by no means isolated on the periphery. The other important substance for perfume production was ambergris, the secretion of the gall-bladder of the sperm whale, which washes ashore along the eastern and western coasts of Africa.142 This is another commodity that might have been sourced by the Umayyads from their connections with North Africa. Ḥasan ibn Qānūn’s marvellous piece of ambergris, which al-Ḥakam desired to possess at all costs, was mentioned above.143 This odd anecdote, related in the Kitāb al-Mafākhir al-Barbar, speaks of a ‘piece of ambergris of an unusual form and enormous dimensions, obtained on one of the Maghribi coasts’.

Given the discussion at the start of this chapter about the location of the luxury arts industry vis-à-vis the other caliphal manufactories, and in particular its original proximity in old Cordoba to the Rabaḍ al-ʿAṭṭārīn – the perfumiers’ quarter – it is likely that the perfume makers of Madīnat al-Zahrāʾ (and possibly of al-Madīnat al-Zāhira, if some of them moved there) were physically close to the workshops of the craftsmen who made the receptacles to hold their perfumes. There was even a pharmacy at Madīnat al-Zahrāʾ, established by one Aḥmad ibn Yūnis in a room at the caliphal palace, by order of al-Ḥakam II.144 According to contemporary recipes – including that written by the most famous Andalusi pharmacologist of the caliphal period, Abū’l-Qāsim al-Zahrāwī (936–1013), known in the west as Abulcasis, but whose kunya derives from Madīnat al-Zahrāʾ, where he was born and lived – liquids such as ghāliya, rosewater, and other essential oils should be contained in clean bottles made from gold, silver or glass.145 Of course at Madīnat al-Zahrāʾ, these workshops were also physically close to the mint (until it moved back to Cordoba), and – as noted above, in reference to the silver palace which al-Manṣūr commissioned for Ṣubḥ – precious metals are likely to have been worked in or alongside the mint, where the materials were probably stored under lock and key. The glass receptacles may well have been made locally, though examples of mould-blown and cut glass imported from the Islamic East have also been found at Cordoba and Madīnat al-Zahrāʾ.146

As the gift to Ibn Abī’l-ʿAfiyya indicates, receptacles for liquid – or even solid – perfumes, whether they were made from gold, silver or glass, were placed inside the ivory containers. This fact calls to mind the few perfume bottles that have survived from tenth-century Cordoba, including two (one silver, one bronze) with spherical bellies and narrow tapering necks, and a hemispherical lid-cum-stopper which is attached by a chain to the body of the vessel (Figure 100). These two objects have abstract repoussé decoration, which evokes carved details on the extant ivories or their metal mounts, and it is intriguing that their dimensions seem to match the interior dimensions of the cylindrical ivories, indicating that they could have fitted snugly inside.147 We have to assume that objects such as the Braga pyxis originally contained perfume gifts contained in bottles, but obviously there is no way of knowing which material these bottles were made from. Such precious contents might have inspired the repurposing of the Braga pyxis by its Christian owner, as a container for the vessels of the Eucharist.148 Might it be going too far to speculate that, once the perfumed contents had been finished, the precious metal was melted down and reused to make the chalice and paten now stored inside? (Figure 11)

Not only would the metalwork ateliers have made perfume containers like these, but they also made mounts to adorn the ivories and other objects emerging from the nearby Dār al-Ṣināʿa. While the silver and niello mounts on several of the ivory pieces made during al-Ḥakam’s caliphate appear to be the original mounts, none of the ʿĀmirid ivories retains its original mounts: Sanchuelo’s pyxis lid and ʿAbd al-Malik’s casket have lost all trace of their mounts, while the Braga pyxis has no space allowed in the ivory for mounts, so originally did not have anything more, perhaps, than textile cords attaching the lid to the body.149 The mounts on the caskets in the Bargello (Figures 148–149), V&A (Figures 150–155) and Doha collections (Figures 145–146) have been added much later. We do not know, therefore, what the original ʿĀmirid mounts would have looked like, but they probably followed the silver-and-niello arrangement seen on surviving mounts from the 960s.

The same decorative technique is used on a number of silver or silver-gilt containers that may have been made in al-Andalus, among them the small heart-shaped case that now holds relics of San Pelayo, in the treasury of San Isidoro de León.150 This tiny box does not have much space for decoration, but the form of the split palmettes that decorate its front and back closely resembles the mounts on a small ivory casket made for one of al-Ḥakam’s sisters around 961, now in the V&A (301–1866). San Pelayo’s relics were translated to León in 967, which might allow us to date these heart-shaped boxes to the 960s as well. It seems likely they were also made in the caliphal metal workshop. Were the relics and accompanying luxury containers part of a royal gift, from al-Ḥakam to the young Ramiro III (r. 966–84), who had just come to the throne of León, to express a wish for good relations?

1.3.4 Other Object Types

There are no surviving metal objects that can be directly associated with the ʿĀmirids, but historical evidence and the survival of artefacts datable to this period attest to the continuation of metalworking practices of different kinds. The two repoussé perfume bottles mentioned above – one silver, one bronze (Figure 100) – were found in two hoards datable to the ʿĀmirid period, according to coins present in the hoards that were minted between 393/1003 and 399/1009. According to Azuar Ruiz, this indicates the existence of a workshop producing objects in precious metals at the turn of the tenth and eleventh centuries – though of course the bottles could have been survivals from a slightly earlier period.151 Because of the use of the repoussé technique, Azuar Ruiz associates with them a cylindrical lidded casket made in tinned brass, from the church of Santa María de Lladó, near Girona, and now in that city’s cathedral treasury.152 This casket is decorated on its body with lobed roundels containing floral motifs, and on its lid with interlinked roundels containing animals: a peacock, lion, harpy, gazelle, lion and hare. Around the lid and the upper frieze of the body is an inscription in Kufic epigraphy, in the style of the late tenth/early eleventh-century: this appears to read the same string of blessings, mirroring each other on lid and body. Two bronze or brass pitchers were found in excavations in the crypt of San Vicente in Valencia, together with a cache of green-and-brown ceramics and a dinar dated 398/1007–8.153 The continuity of bronze casting into the Taifa period – when it seems that the magnificent sculpture known as the Pisa Griffin was made154 – indicates that it must have received patronage through the ʿĀmirid period, not least to make the great gates for the Great Mosque of Cordoba and for al-Manṣūr’s new palaces, and objects such as the fountain heads that adorned the basins in ʿĀmirid palace gardens. Finally, a word may be said about jewellery at this period: this consists of bracelets or the links from bracelets and necklaces, pendants, rings including those with cabuchons, earrings, and loose beads, made from gold, silver, but especially silver gilt. These have principally been recovered from hoards and, as Labarta points out, most of the datable hoards – ie those that contain coins, although these have typically been perforated for attachment to clothing and thus might already have been old when the hoard was buried – date from the time of the Fitna. The Loja hoard, from Granada and now in the Instituto de Valencia de Don Juan, is datable 1010, as is the hoard from Ermita Nueva, in Alcalá la Real; that from Cortijo de la Mora, Lucena, is datable 1013; the Lorca hoard, from Murcia and now in the V&A, contains ten dirhams minted during the reigns of al-Ḥakam II and Hishām II.155 A systematic study of the objects found in these hoards is necessary to improve our understanding of the techniques and aesthetics of Andalusi jewellery and how these might compare with better-documented finds from the Islamic East.

Figure 99
Figure 99
Ivory cosmetics box in Burgos, reverse; Museo de Bellas Artes, Burgos
© Archivo Oronoz
Figure 100
Figure 100
Perfume bottle, repoussé bronze, found in the hoard from Olivos Borrachos, Cordoba, buried c. 1010; Museo Arqueológico y Etnológico, Cordoba, inv. CE003772
© Mariam Rosser-Owen

One type of metalwork that al-Manṣūr required in great quantity was arms and armour, and as we saw in Chapter 4, the sources tell us that there were important weapons stores (asliḥa) at al-Madīnat al-Zāhira. Ibn al-Khaṭīb informs us that, in the campaign against Barcelona in 985, the ʿĀmirid soldiers wore armour of Indian steel (qarāmid al-Hind).156 This was made in al-Andalus – there was an important steel factory in Seville, whose products were exported around the world. At this period, two arms factories operated in Cordoba and Madīnat al-Zahrāʾ: each year, they provided 13,000 shields, 12,000 Arab and Turkish bows, and each month they supplied 20,000 arrows.157 In the ʿĀmirid period, the Cordoba factory was directed by Abū’l ʿAbbās al-Baghdādī – Vallejo observes that an ‘eastern craftsman’ in this role might have inspired the introduction of oriental techniques and object types into Andalusi metalwork.158 The factory at Madīnat al-Zahrāʾ was directed by the Slav, Ṭalḥa. Their products were stored in the Khizānat al-Silāḥ, the state arsenal, from where they were signed out when needed for military parades or campaigns.159 The bronze disc in Figure 18 and that excavated at Cercadilla bear physical testament to this organised system of signing arms and armour in and out of the ʿĀmirid arsenal. They also indicate the huge amount of money that must have been spent on making these suits of armour: the tag in Figure 18 was attached to the 141st mail coat ‘decorated in the Slav style’, which means that al-Manṣūr had paid for at least another 140 such coats, potentially made from or decorated in precious metals. Maybe it was Ṭalḥa, the Slav, who made this particular type of armour, hence its designation. Arms and the apparel for horses, which would have been made in these state arms factories, were also important components in khilʿa presentations to the Umayyads’ Berber allies; as the descriptions of these show, these were frequently made of precious metals, often silver-gilt or nielloed silver, and highly decorated.160

Finally, the sources tell us of the huge number of lamps that illuminated the Great Mosque of Cordoba during al-Manṣūr’s period. The mosque contained 280 lamps, mostly of bronze, not counting those which illuminated the external gates – this is equivalent to a lamp hanging in every other bay of the mosque. Four of these lamps were made from silver, each of which consumed more than 200 kg of oil a night.161 These hung in the mosque’s central nave, the largest of them hanging in the great dome in front of the maqṣūra, in order to illuminate the sacred Qurʾāns.162 This could mean the mosaic-incrusted dome in front of the mihrab, next to the Bayt al-Māl where the Qurʾāns were stored, or the dome now known as the ‘lucernario’, just to the west of the possibly-ʿĀmirid tribune (Figures 47, 49). This lamp was hung with 1020 lights, which were covered with gold (presumably silver-gilt), and it had a circumference of fifty palms. The description calls to mind the early thirteenth-century ‘Grand Lustre’, which hangs in the main dome of the axial nave of the al-Qarawiyyīn mosque in Fez, made to replace an earlier, deteriorated lamp, and possibly preserving the form of older Andalusi lamps.163 All these lights blazed throughout the month of Ramadan and especially on Laylat al-Qadr, the 27th night, when 35 hundredweights of oil were burned, together with three hundredweights of wax and a quantity of ambergris and other perfumes. Finally, Vallejo speculates that the bells brought to Cordoba after the campaign against Santiago de Compostela, which were turned into lamps for the mosque (Chapter 1), might have necessitated the introduction of a new type of metalworking tradition.164

We can assume that those metalworking techniques that required more industrial processes and made more noise and mess were located further afield, perhaps close to the potters’ quarter or the Dār al-Ṭirāz. Their production cannot be localised with certainty, because no archaeological evidence of metal workshops or foundries has yet been excavated in or around Cordoba.165

Likewise, though no examples of ceramic objects can be associated directly with al-Manṣūr and his court, we can assume that the production of tin-glazed ceramics decorated in green and manganese continued and presumably developed under the ʿĀmirids, because it flourished so widely across al-Andalus during the eleventh century. The terminus ante quem for the invention of this distinctively Andalusi ceramic technique is provided by the tiles used in the coving of the mosaic-incrusted ante-mihrab dome in the Great Mosque of Cordoba, completed around 965. The huge number of so-called ‘green and brown’ ceramics that have been recovered in excavations at Madīnat al-Zahrāʾ, in the western suburbs of the medieval city of Cordoba, and in the excavations of the street that runs beneath the ʿĀmirid mosque extension,166 indicates that the production of this ware was well established by the tenth century. Apart from the early association with the court, through their employment in the mihrab dome, the fact that this type is found so abundantly probably implies it was not an elite ware, or used by the caliph himself. It is more likely that the upper levels of the court hierarchy used more expensive and exotic imported ceramics: Abbasid and Fatimid lustrewares have been excavated at Madīnat al-Zahrāʾ, and there are also examples of Chinese imports found in possibly tenth-century contexts.167 These are the types of wares the caliph and his courtiers are likely to have used, perhaps later imitated in this by his ḥājib. It may be that the exotic imports from China and Iraq – which travelled via Fatimid Egypt despite political rivalries – inspired the caliph or his potters to develop their own fine tablewares in imitation, just as Iraqi potters in the eighth century had been stimulated by the arrival of Chinese whitewares in the Middle East to invent tin-glaze technology.168 It speaks to a heightened atmosphere of artistic innovation and experimentation because of a strong and wealthy patronage power.

Another type of artistic production for which very little physical evidence survives is books. However, we know from the historical sources – not least those that highlight al-Ḥakam’s erudition and famous library – and the discussion in Chapter 3, of books and learning under the ʿĀmirids, that this was an art that would have been sponsored by the court. It has been suggested that the intense atmosphere of book collecting and copying in late-tenth-century Cordoba led to the codification, if not the invention, of the distinctive Maghribi script, though the only surviving manuscript that can be associated with al-Ḥakam’s library – a copy of the Mukhtaṣar Abī Muṣʿab, dated 359/970, now in the library of the al-Qarawiyyīn Mosque in Fez – is not written in Maghribi script.169 Bongianino argues that this scribal development was due instead to the activity of the many private libraries and book collectors in caliphal Cordoba: ‘the largest and most famous of these private institutions was no doubt the library of the affluent scholar Ibn Fuṭays (d. 402/1012), where six scribes were employed constantly to increase their master’s collection of books, hired at a fixed salary rather than at piece rates, lest they be tempted to rush their handwriting’.170 Julián Ribera calculated that more than 60,000 manuscripts were being copied in the city’s private libraries every year.171

It is surely significant that of the eleven extant Andalusi manuscripts copied in Maghribi scripts, all of them date from the ʿĀmirid period, with the exception of a ninth-century manuscript that was owned by the faqīh ʿAbbās ibn al-Aṣbagh ibn ʿAbd al-ʿAzīz al-Hamdānī (306/918–386/996), who died during the ʿĀmirid period, and an anomalous tenth-century manuscript with an ownership note in Hebrew script.172 Again, these manuscripts have survived through accident, so it would be unwise to base any firm conclusions upon them, but this fact would seem to support an impression of a heightened – certainly, at least, continued – scribal and intellectual activity through the period of the ʿĀmirid ḥijāba, despite al-Manṣūr’s purge of the caliphal library. Al-Manṣūr even indulged in scribal activity himself: as mentioned in Chapter 5, one aspect of al-Manṣūr’s ostentatious piety, that has been memorialised by historians over the centuries, is the fact that he himself copied a volume of the Qurʾān, which he always carried with him on campaign. Ana Echevarría has speculated further that al-Manṣūr or other members of this family could have commissioned beautifully ornamented Qurʾān manuscripts to present to the mosques he patronised – at al-Madīnat al-Zāhira and his Cordoba Mosque extension – but again there is no evidence for this.173 However, we know that al-Manṣūr had his own librarian, and hence his own physical library and collection of books. ʿAbd al-Raḥmān ibn Muḥammad (or Muḥammad ibn ʿAbd al-Raḥmān) ibn Maʿmar al-Lughawī (d. 423/1032) is mentioned in biographical dictionaries as a ‘grammarian, an expert in letters and the sciences of language’. He was employed by al-Manṣūr and his son (presumably ʿAbd al-Malik) to be their ‘official chronicler and in charge of their library’, and was the author of a lost dynastic history of the ʿĀmirids, al-Taʾrīkh fī l-dawla al-ʿāmiriyya. After the Fitna, he found protection under one of the ʿĀmirid fityān who became a Taifa ruler, Mujāhid al-ʿĀmirī, and lived at his court in the Balearic islands, where he ‘presided over judgments’ until his death.174

Vallvé notes that Andalusi rulers took abundant supplies of parchment with them on their military expeditions, ‘in order to concede mercies and privileges to the governors and lords of the provinces and to confirm accords signed with the Christian counts and kings’. He cites an anecdote about Sanchuelo at the outbreak of the Fitna, returning rapidly to Cordoba and exhausting his supply of parchments on the road, ‘in a last effort to attract to himself the Berber chiefs of his troops by means of royal concessions’.175

The types of materials and techniques discussed in this last section formed the background to the material world in which the ʿĀmirids lived, and it is important to keep in mind that objects in these media existed and were used by the ḥujjāb. However, since no surviving examples can be directly associated with ʿĀmirid patronage, they will not feature prominently in the discussions to come.

2 Iṣtināʿ: The Strategic Use of Objects

The commissioning of precious objects was closely implicated in ʿAbd al-Raḥmān III’s attempts to set al-Andalus as the stage of a truly international state. The two explicit textual descriptions of his luxury arts patronage are both recorded in the context of diplomatic exchange. We saw in Chapter 2 how ʿAbd al-Raḥmān commissioned from the Dār al-Ṣināʿa twelve gold and jewel-incrusted fountain heads to adorn the green marble basin he had received as a gift from the Byzantine Emperor. While no extant objects can be securely associated with ʿAbd al-Raḥmān III (apart from the fragmentary ṭirāz in the Cleveland Museum), the scant historical texts that mention objects capture something of that caliph’s deployment of the products of this luxury arts workshop.

A gift sent to the Zanāta chieftain, Muḥammad ibn Khazar, soon after ʿAbd al-Raḥmān declared his caliphate, contains crucial information about the way in which ṭirāz texiles – that prerogative of being a caliph – were already being deployed as gifts.176 The text records ‘excellent garments and sublime suits, luxury jewels and strange (foreign?) wonders, distinguishing [the recipient] with the special gift of clothing which [the caliph] ordered, from his own ṭirāz, to be embroidered with the name “Muḥammad ibn Khazar”, a great distinction, which had never before been conceded by a king to one of his partisans, and about which al-Nāṣir himself boasted in the letter that accompanied the gift’. This letter stresses that the clothing was made ‘fi ṭirāz-hu al-khāṣṣ’, ‘in his private ṭirāz’, and refers to ‘ten pieces of various types, to be used for clothing, of pure ʿubaydī silk, embroidered with your name (muṭarraza bi-ismi-ka), the like of which were never made in the Abbasid factories nor in others’. ʿAbd al-Raḥmān seems to be gloating of his ownership of a ṭirāz workshop, which was already superior to those of his political rivals, in its unprecedented production of textiles specifically commissioned for a particular recipient, so that they had his name embroidered onto them. Alongside other textiles, the gift also included ‘a short sword in the Frankish style, adorned with silver, gilt and relief decoration’, and the text dwells on the description of its luxuriously decorated and bejewelled sheath and cords, and on the belts and spurs which accompanied it, likewise adorned with gold, precious stones and pearls.

The magnificent gift sent by ʿAbd al-Raḥmān in 934 to Mūsā ibn Abī’l-ʿAfiyya, mentioned above, did not just include ivories and perfume containers; these are embedded within a long list that included ‘ṭirāzī khāṣṣī’ from Iraq and other textiles, including turbans; a ‘large silver caliphal perfume chest, with gilded engraving in its plating and a white base, and whose interior was lined with purple fabric’, presumably a Byzantine imperial silk received in an earlier embassy and repurposed here to show the importance of ʿAbd al-Raḥmān’s international relations; banners with figurative emblems of an eagle, a lion and a white horse, as well as one decorated with epigraphy; and arms and armour, including two swords, ‘four braided pieces of rope for striking water buffaloes’, ‘six fully mechanised gilded drums’, and ‘twenty thousand Persian arrows’.177

The context of this magnificent gift to Mūsā was a reward for a victory over the Fatimids. The banners (to which we will return in Chapter 8) were intended to be displayed as a visual symbol of Mūsā’s Umayyad allegiance. The gift as a whole, and the significant presence of these early ivories within it, can be associated with the political projection of the Umayyad caliphate into the Maghrib. Both these gifts to Berber leaders should be understood within the concept of the Umayyads’ deliberate policy of khilʿa, the strategy executed by al-Manṣūr on their behalf to attract support in the Maghrib (Chapter 2). In this role – negotiating a delicate balancing act of hierarchies and sensitivities – al-Manṣūr learned early on in his career important lessons about the value of strategic relationships and the power of gift-giving. Ibn Bassām uses the term iṣtināʿ to describe relationships built through gifts and special favours, which rely on the gratitude that the beneficiary feels towards his benefactor; he attributes al-Muṣḥafī’s eventual downfall to his failure at building such relationships, in contrast with al-Manṣūr’s skill at doing so.178 Ibn Abī ʿĀmir dedicated himself to a policy that was the opposite of al-Muṣḥafī’s, ‘calming jealousies by giving gifts and satisfying needs, seeking out the company of men while al-Muṣḥafī held himself aloof, increasing their assets while al-Muṣḥafī diminished them’.

Al-Manṣūr was not the first ḥājib to make use of artistic patronage for political ends, and it is instructive to consider the patronage of the two men who had occupied this office before him (no artistic patronage has yet been attributed to either of the men who held the office of ḥājib under ʿAbd al-Raḥmān III, on whom see the Introduction). In accessing the processes for commissioning art and architecture, al-Manṣūr was not ‘usurping’ another caliphal prerogative, but following a precedent for the use of the luxury arts industry by high level members of the state administration. As Martínez Enamorado put it, ‘the case of Jaʿfar al-Ṣiqlābī serves to light the way for al-Manṣūr’.179 However, al-Manṣūr’s use of the industry manifests a massive increase in scale and intent, and a deliberate exploitation of the visual arts to express messages of power.

2.1 Precedents for Patronage: The Two Jaʿfars

The two ḥājibs who served under al-Ḥakam II were confusingly both named Jaʿfar: al-Ṣiqlābī (in office c. 353/964–5 to 360/971–2) and al-Muṣḥafī (in office ?360/971–2 to 368/979). As Martínez Enamorado notes, even the medieval historians confused them.180 Both these men were extremely important in the state administration, as indicated by the passage in al-Rāzī’s Annals describing the location of Jaʿfar al-Ṣiqlābī’s residence close to that of the caliph at the heart of Madīnat al-Zahrāʾ:

“At the beginning of Muḥarram of this year (361/972–3), the caliph al-Ḥakam ordered that his khalīfa and the first among his favourites, the fatā al-kabīr Fāʾiq, ṣāḥib al-burud wa’l-ṭirāz, transfer himself from the house which he occupied in the east wing of the Qaṣr al-Zahrāʾ to that which had belonged to the ḥājib Jaʿfar ibn ʿAbd al-Raḥmān al-Ṣiqlābī, [who] had died the previous year (360/971–2). This house was much more important and was situated in the west wing. [The caliph] decided to honour [Fāʾiq] in this way, because of the high esteem in which he held him, and as proof of his distinction and preference.”181

Manuel Ocaña carefully studied the surviving epigraphy which names Jaʿfar, as a means of charting the development of his career.182 It is remarkable that so many inscriptions survive that allowed him to do so, compared to other state officials. The list includes the many architectural commissions that Jaʿfar carried out on the caliph’s behalf (ʿala yaday-hi), and indicates how active Jaʿfar was in the service of the state, as well as the intensity of building projects under the first two caliphs. Indeed, it implies how much more the caliphs came to rely on state officials, especially their ḥājib, to carry out such commissions (as well as other tasks). As discussed above (‘Structure of the industry’), Jaʿfar’s name is linked with all the major architectural projects of al-Ḥakam’s caliphate, in particular the Cordoba mosque extension: his name and titles appear five times in the marble and mosaic work around the mihrab, along with the caliph’s own name. He also came to be ṣāḥib al-ṭirāz (from 347/957), thus heading up the two branches of the arts industry that were most important in the political projection of the Umayyad caliphs. In so doing, he amassed great wealth, and is said to have left a considerable fortune on his death. This was managed as a waqf (al-nāẓar bi’l-aḥbās) by the Cordoban faqīh Abū ʿAbd Allāh Muḥammad ibn Aḥmad ibn ʿUbayd Allāh al-Ruʿayni, known by the name Ibn al-Mashshāt (d. c. 1009), though Martínez Enamorado does not say whether it is known what pious works were funded with this money.183

Al-Ṣiqlābī’s official residence in the west wing at Madīnat al-Zahrāʾ has been excavated, reconstructed and well-studied by the site’s excavators, in particular Antonio Vallejo.184 New inscriptions have come to light since Ocaña’s study, including two on architectural supports that are probably associated with private residences. Indeed, it is likely that during al-Ṣiqlābī’s lifetime, some of his immense fortune would have been spent on building and embellishing a private, suburban villa in addition to his ‘official’ residence at Madīnat al-Zahrāʾ. One highly significant new piece is an element now in a private collection in Cordoba (Figures 33–35). This is a cyma or impost block, the element that sits above a capital and supports the springers of arches. Since there is always more than one springer for an arch, we can imagine that there were originally several matching impost blocks in the building from which this element came. Normal for this period would be an opening of three arches, each supported by impost blocks, of which the flanking two would have abutted against the wall: we see this arrangement all over Madīnat al-Zahrāʾ (Figure 101). We can therefore reasonably assume that there were originally four similarly decorated impost blocks. Only the front half of this cyma survives, the back half having apparently suffered damage by being torn from its original placement. The fact that the inscription starts at the front and wraps all the way around the cyma suggests that it was originally visible from all four sides, and thus supported the springer to the right or left of the central arch, rather than occupying one of the flanking positions against a wall. However, this example is unique. No decorated cymas have appeared at ‘in any of the significant spaces’ at Madīnat al-Zahrāʾ, though there are examples of non-figurative cymas; those with inscriptions mostly date from the period of al-Ḥakam II’s reign.185 Martínez Núñez has suggested that the fragment decorated with the heads of sharp-toothed beasts, found at al-Rummāniyya, might have come from a figurative cyma.186

Figure 101
Figure 101

Tripartite arcade in the Upper Basilical Hall, looking onto the garden, Madīnat al-Zahrāʾ

© Mariam Rosser-Owen

At the front, the cyma shows two affronted griffins, represented as winged animals combining the body of a lion and the upper torso of an eagle, amongst deeply drilled vegetation; this vegetation occupies all the upper corners of the cyma and is carved in the manner of a capital, perhaps evoking the style of the capitals that originally sat below these impost blocks. On its damaged sides are seen the rear ends of two hooved quadrupeds, no doubt gazelles. There is no suggestion of a combat scene on the sides, suggesting that paired gazelles and paired griffins alternated with each other around the four sides of the cyma. This arrangement recalls the ʿĀmirid capital reused in the Casa del Gran Capitán in Cordoba (Figures 26–7, Chapter 4). The figurative decoration of this cyma suggests that the residence from which it came was a private one, given the uniformly non-figurative nature of the official art of the Umayyad state, while munyas as sites of otium seem to have led to a more relaxed attitude towards figurative representation.187

The cyma preserves part of an inscription, which begins on the front, at top right, with a long basmala, and appears to name ‘Abū Aḥmad Jaʿfar al-ḥājib’; it ends with the date, which Martínez Enamorado reads as 353/964–5, while María Antonia Martínez Núñez suggests 356/966–7.188 The name Jaʿfar would appear at the far left of the front inscription band, which is oddly smoothe compared to the clear relief of the remainder of the inscription, and invites speculation that the name was later deliberately removed, by a rival or later occupant of the building from which this element came. Martínez Enamorado interprets Abū Aḥmad as Jaʿfar al-Siqlābī’s kunya, though this was never used in the many commissions he carried out in the caliph’s name. On those inscriptions, Jaʿfar habitually celebrated his relationship to the caliph and the Umayyad dynasty, by employing the nasab Ibn ʿAbd al-Raḥmān, and proclaiming himself the caliph’s mawlā-hu wa ḥājibu-hu.189 Nor does it appear in the inscriptions from the House of Jaʿfar at Madīnat al-Zahrāʾ. That was his official residence, which he occupied at the grace and favour of the caliph: as such, epigraphy from this residence underlines Jaʿfar’s relationship with his caliphal patrons. There is no reference to the caliph in the cyma’s admittedly incomplete inscription – the blessings in the opening eulogy on the cyma are for (li-) Jaʿfar himself. Furthermore, unusually, the name as given here follows an Arab-style onomastic system, rather than that usually employed for freedmen.

Two capitals survive which likewise bear the kunya Abū Aḥmad.190 One, of which only a volute remains, excavated in Valencia, clearly reads the phrase ‘li-Abī Aḥmad Jaʿfar’ (Figure 102), while the other is a capital now in the Romero de Torres collection in Cordoba (Figure 103), whose inscription has been reconstructed as reading a full basmala (as on the cyma) followed by: li-Abī Aḥmad Jaʿ[far ḥāj]ib amīr al-muʾminīn aṭāla a[llāh baqā-hu fa-tamma bi-ʿawn al]lāh wa-niʿmatu-hu fi [sana … thala]th miʾa.191 In both cases the complete date is missing, but assuming an identification with Jaʿfar al-Ṣiqlābī, Carmen Barceló (in her various publications of these elements) dates the capitals broadly to the period 350–360/960–971, coinciding with the apogee of his career. Given the coincidence with the use of the kunya on the cyma, as well as its probable date, it seems highly likely that the two capitals and impost block all come from the same building, possibly a munya, constructed in the mid-960s, perhaps by Jaʿfar al-Ṣiqlābī.

Figure 102
Figure 102

Volute inscribed li-Abī Aḥmad Jaʿfar, excavated in Valencia

© SIAM (Servicio de Investigación Arqueológica Municipal), Ayuntamiento de València, inv. 1/237
Figure 103
Figure 103

Capital inscribed in the name of Abū Aḥmad Jaʿfar; Museo de Romero de Torres, Cordoba

© Antonio Vallejo Triano

On the other hand, this may be an example of the confusion between the two ḥājibs named Jaʿfar. Martínez Núñez has instead identified the ‘Abū Aḥmad Jaʿfar [ḥāj]ib amīr al-muʾminīn’ on the two capitals as referring to Jaʿfar al-Muṣḥafī.192 Though he is known to have had the kunya Abū’l-Ḥasan, she notes that his first-born son was named Aḥmad, and that it was not unusual to use a double kunya. But al-Muṣḥafī was not, apparently, named ḥājib until the very end of al-Ḥakam’s life. However, the word ḥājib in the inscriptions is actually a reconstruction, again based on the identification with Jaʿfar al-Ṣiqlābī, and only the final -bāʾ of the word is legible. Martínez Núñez suggests that an alternative reading of this word would be kātib, which could then refer to al-Muṣḥafī, who occupied the post of personal kātib to al-Ḥakam before he acceded to the caliphate.193 These two capitals may thus have come, instead, from a construction sponsored by al-Muṣḥafī – but, until this was suggested, it was thought that nothing material had survived in his name. Now, perhaps, we can also associate with him a unique figurative impost block, and another inscribed object, discussed below.

As mentioned in Chapter 3, al-Muṣḥafī was celebrated for his poetry, and perhaps his patronage interests lay more in literature than in art, though as a member of the court elite, al-Muṣḥafī would surely have wanted to surround himself with a beautiful setting that made a physical statement of his lofty position, especially once he was promoted to ḥājib and became the chief political figure in the state. Indeed, there are incidental references to a munyat al-Muṣḥafiyya in the sources: al-Maqqarī tells us that once he had fallen from grace, al-Muṣḥafi was ‘so ruined and impoverished that he was compelled to sell [al-Manṣūr] his munya in al-Ruṣāfa, which was one of the most magnificent residences in Cordoba’.194 This loss was nostalgically evoked by his grandson, Muḥammad Abū Bakr ibn Aḥmad ibn Jaʿfar al-Muṣḥafī, who, on passing by one day, lamented in a poem: ‘Stop a moment before al-Muṣḥafiyya and weep for an eye without its pupil! Ask Jaʿfar about his power and generosity in times past!’.195 Ibn ʿIdhārī mentions a Munyat Jaʿfar, where Hishām II stayed briefly in 399/1008–9, and where he designated ʿAbd al-Raḥmān Sanchuelo to be his heir to the caliphate.196 Lévi-Provençal believed this to have been a residence of Jaʿfar al-Ṣiqlābī, while Barceló and Cantero actually rule out the possibility that it could be identified with al-Muṣḥafī’s munya, because elsewhere this is called ‘al-Muṣḥafiyya’.197 This is far from conclusive, however, as al-Muṣḥafiyya is an adjective, and could merely have been employed poetically in Muḥammad Abū Bakr’s lament, rather than indicating an actual name for the munya. In the same way, munyas associated with the ʿĀmirids were called al-ʿĀmiriyya or al-Ḥājibiyya (see Chapter 4).198 Martínez Núñez seems happy to associate the Munyat Jaʿfar with al-Muṣḥafī.199 The sources also contain hints of other levels of patronage: according to Ibn Juljul, al-Muṣḥafī had his own personal physician, Aḥmad ibn Ḥakam ibn Ḥafṣūn, who was also in charge (muqīm) of the ḥājib’s residence.200 Al-Muṣḥafī or another member of his family may also have founded a mosque in the centre of Cordoba (dākhil Qurṭuba), though the only evidence for this is a topographical memory in a later source.201 If the cyma did come from al-Muṣḥafī’s munya, later taken over by al-Manṣūr after the former’s fall from grace, this might explain the deliberate erasure of the name Jaʿfar from the inscription at the front.

Another object has come to light in recent years that can now be associated with al-Muṣḥafī’s artistic patronage because of its epigraphic relationship with the cyma and two capitals discussed above. This is a vertical element carved from white marble, now in the David Collection in Copenhagen, bought on the art market in 2009 (Figures 104–107).202 It is carved on three sides with a design of trilobed arches with pronounced voussoirs enclosing flourishing vegetal decoration; its fourth side has a much rougher surface and a concave depression running top to bottom, about the same height and diameter as a lead pipe; the top of the object is flat, with the pipe hole passing up into it. As such it has been identified, by María Antonia Martínez Núñez and Antonio Vallejo, as the decorative cover for the piping that would have fed a fountain head standing on the flat top of the object, pouring water into a fountain basin of the ʿĀmirid type (Figures 108–109).203 This also helps to explain what would have abutted and hidden the blank spaces seen on the short ends of the three complete ʿĀmirid basins. The decoration on the sides of the David Collection piece might have reflected or continued whatever decoration was carved on the sides of the fountain basin against which it was set; we can understand the relationship perfectly if we look at al-Manṣūr’s basin (Figure 113), where the arcade on the back is stylistically extremely close to the trilobed arches of this marble ‘pipe cover’.

Figure 104
Figure 104

Pipe cover for a fountain basin: side 1, 971–9, marble; The David Collection, Copenhagen, inv. 25/2009

© Pernille Klemp
Figure 105
Figure 105

Pipe cover for a fountain basin: side 2, 971–9, marble; The David Collection, Copenhagen, inv. 25/2009

© Pernille Klemp
Figure 106
Figure 106

Pipe cover for a fountain basin: side 3, 971–9, marble; The David Collection, Copenhagen, inv. 25/2009

© Pernille Klemp
Figure 107
Figure 107

Pipe cover for a fountain basin: interior, 971–9, marble; The David Collection, Copenhagen, inv. 25/2009

© Pernille Klemp

An inscription runs around the lowest level of the piece, where it is least visible, though this placement recalls the column bases from Madīnat al-Zahrāʾ that are inscribed with the names and titles of the ruler. In the object’s more recent history, it was (appropriately) used as the base for a fountain, so the original sharpness of many of its details has been eroded away; the ‘front’ of the object (side 2) is most worn, and here the inscription is most difficult to make out. However, according to Martínez Núñez’s reading, the ‘pipe cover’ again mentions Abū Aḥmad Jaʿfar: (1) baraka min allāh li-ab[ī] (2) [a]ḥmad jaʿfar al-hāj[ib] (3) [ṣāḥib a]l-khayl wa al-burud. The extremely close stylistic relationship of this piece to al-Manṣūr’s basin, made in 377/987–8, suggests a similar chronology, though obviously before 979 when al-Muṣḥafī fell from grace. The title ṣāḥib al-khayl wa … was initially taken as associating the pipe cover with al-Ṣiqlābī, who is named as ṣāḥib al-khayl wa’l-ṭirāz in an inscription on one of the small decorative arches from the baths in the House of the Pool (‘Vivienda de la Alberca’) at Madīnat al-Zahrāʾ, though this dates from the year 350/961, before he became ḥājib. However, as mentioned in Part 1, in 971 al-Muṣḥafī received the office of head of the postal service (barīd), when he was put in charge of moving its headquarters.204 This later chronology and the stylistic parallels point to al-Muṣḥafī as the patron, and in fact this title allows a more precise dating for the pipe cover, of 971–9.

Figure 108
Figure 108

Exploded view of a reconstruction of al-Manṣūr’s basin showing pipe covers and lion fountain heads

© Matilde Grimaldi
Figure 109
Figure 109

Proposed reconstruction of al-Manṣūr’s basin with pipe covers and lion fountain heads

© Matilde Grimaldi

This reading would suggest that the reconstruction of the word ḥājib on the capital in Cordoba (Figure 103) is, in the end, correct, and that the two marble capitals should likewise date from the 970s rather than the 960s. These capitals were carved in the composite style, which is stylistically more consonant with the decoration on the pipe cover. The echinus of the capital in Cordoba is carved with a deeply drilled out vegetal scroll, which might mirror the border of drilled decoration that runs along the top of all three sides of the Copenhagen piece – but the carving of both is not well enough preserved to tell whether they were originally meant to match. The capitals both originally measured around (H) 26 cm × (W) 25 cm × 15 cm (diameter at the base), and had long inscriptions running all around the top of the echinus (which survive best on the volute in Valencia). They both represent considerable time on the part of the carver and investment on the part of the patron. Let us not forget that al-Maqqarī described the Munyat al-Muṣḥafiyya as ‘one of the most magnificent residences in Cordoba’.

A vast residential complex, known as the Plan Parcial de RENFE, uncovered during rescue excavations in 1997–8, has been tentatively identified as the physical remains of al-Muṣḥafī’s private residence. Its dimensions are huge – ten times the size of the largest domestic house yet excavated in Cordoba – arranged around seven courtyards, all configured around one large central courtyard containing a garden and pool.205 The archaeologists who excavated it consider that the property extended further to the north with a zone of gardens and orchards.206 All this indicates that this was the residence of someone of high social status: a minister or courtier of high rank. Felix Arnold lists the ten residences cited by the historical sources that can be associated with statesmen of this level, but only a few can be located in the neighbourhood where this structure was found:207 to the north of the city, between the suburbs of Cercadilla and the Bāb al-Yahūd, around 500 m from the northern wall of the madina of Cordoba, close to one of the arterial road of the city, in use since Roman times.208 This area was subject to surburban development in the third quarter of the tenth century (i.e. 950–975), precisely when al-Muṣḥafī became ḥājib and built ‘one of the most magnificent residences in Cordoba’. This chronology for the construction of the residence is supported by the archaeology, including the ceramic finds at the site.209 Of the munyas cited by Arnold, only one fits the location and chronology, and that is al-Muṣḥafī’s residence. The identification is speculative, but would appear to verify the information from the historical sources. The surviving fragments of architectural decoration that are discussed here may well have come from this site.

An interesting feature of the animals on the cyma is that they have a black substance inlaid into their eyes: this survives on the left-hand griffin but appears to have fallen out of the right-hand griffin, leaving a deep empty eye-socket. The only other example of this use of black inlay for the eyes, of which I am aware, is the volute from al-Rummāniyya (Figure 29): in the procession of birds on its sides, the birds’ eyes are likewise inlaid with a black substance.210 The accepted date for the construction of al-Rummāniyya by Durrī al-Saghīr is 965, after a fragmentary inscription that preserves part of a date.211 Might this shared technique indicate a date for the cyma in the 960s, even that it was produced in the same workshop as the architectonic elements for al-Rummāniyya? The eagles’ eyes on the two marble basins made for al-Manṣūr and ʿAbd al-Malik in later decades both have deep hollows for their eyes: could these have been intended to receive an inset substance (Figures 116, 132)? A ‘bright soft metal’, most probably silver, has recently been identified inlaid into the eyes of figures on the ivory casket in Doha, as discussed in Chapter 7 (4.1.1); and it was common for jet to be inlaid into the eyes of the ivory sculptures carved in the kingdom of León from the 1060s onwards.212 Could this have continued a technique that was already being practised in the caliphal sculpture workshops? If the musicians’ capital (Figure 31) had retained its heads, might they also have had black inlaid eyes?

The patronage of the cyma remains enigmatic: while it likely names Abū Aḥmad Jaʿfar, it clearly gives the title of ḥājib and a date in the 350s/ 960s – but al-Muṣḥafī was not ḥājib at this date, and al-Ṣiqlābī was no longer ṣāḥib al-khayl, if this is the title which follows ḥājib. There remains the possibility that al-Muṣḥafī was appointed ḥājib by al-Ḥakam II after the death of al-Ṣiqlābī in 360/970, and that this office was confirmed by Hishām II after his father’s death. (Is it significant that the inscription says ḥājib not al-ḥājib? As we saw in Chapter 1, various people could act in the role of ḥājib in court ceremonies.) The unusual figurative decoration of the cyma and the use of an inlaid substance in the eyes seems to belong to the 960s; if it did once adorn the Munyat al-Muṣḥafiyya, this suggests a range of styles, perhaps added at different times, over the course of al-Muṣḥafī’s long career.

We have noted the extremely close stylistic comparison between the Copenhagen ‘pipe cover’ and al-Manṣūr’s basin, though some motifs on the pipe cover are slightly different: the large flower at centre right of its best preserved side (side 3 following the inscription, Figure 106) does not appear on al-Manṣūr’s basin, while the Copenhagen piece does not feature the starlike flowers that become a characteristic motif on ʿĀmirid marble carvings (seen, for example, at top and bottom of the central arch section on al-Manṣūr’s basin). These variations may be explained by the slightly different chronologies. The Copenhagen ‘pipe cover’ is also very close stylistically to another decontextualised marble element, now in the Museo Arqueológico Nacional in Madrid (Figures 110–112). This has a large circular hole on its underside so was probably the socket for the post above a door, or was reused in this way at a later date. It does not seem to be another cover for a fountain pipe: though the socket might have supported a pipe, there is no indication that a pipe could have passed through the length of the element; and the decoration is orientated so that the flat end (on which a fountain head might have stood) is at the bottom. If it were originally a doorpost, this element would have been up against a wall. Only two (opposing) sides of this piece have carved decoration, and there is no suggestion of an inscription. The arrangement of the floral scrolls within trilobed arches with voussoirs is very close to the David Collection piece, though the leaf motifs within do not exactly match; while the border of drilled decoration that runs along the top of all three sides of the Copenhagen piece seems to find a match in a similar drilled border on the Madrid element. Could these stylistic consonances indicate that the piece in Madrid is another decontextualised architectonic element from al-Muṣḥafī’s munya? If so, the prevalence there of the designs based on trilobed arches – a motif that was employed in ʿĀmirid art from its earliest surviving example (al-Manṣūr’s basin) – might suggest that this formed part of an artistic vocabulary developed under al-Muṣḥafī, which was continued under al-Manṣūr.

The griffin motif depicted on the cyma appears to represent the earliest extant use of this motif in al-Andalus. Glaire Anderson has argued that the cyma’s griffins – prevalent as they are in Byzantine and Sasanian silks – were introduced to Cordoba through imported textiles, during Jaʿfar al-Ṣiqlābī’s tenure as ṣāḥib al-ṭirāz, and thus their employment on the cyma might specifically evoke that office (she is understanding al-Ṣiqlābī as the patron of this object).213 Alberto Montejo has argued for a deeper significance for the use of the griffin motif.214 Noting its prominent presence both here and on the basin made twenty years later for al-Manṣūr (Figures 116, 118B), and used subsequently on the other ʿĀmirid basins (Figure 132), as well as on the Pamplona casket (Figures 121, 123), Montejo argues that the griffin motif was a symbol of authority specifically associated with the Andalusi ḥujjāb; indeed that it was an emblem of the office of the ḥijāba, and it was for this reason that it was later widely used on works of art made for the ʿĀmirids. If this interpretation is correct, it seems that again al-Manṣūr was adopting and continuing an artistic motif that had already been established by his predecessors in the office of ḥājib.

Figure 110
Figure 110

Doorpost, 970–990, marble: right side; Museo Arqueológico Nacional, Madrid, inv. 1972/103/1

© Museo Arqueológico Nacional, Madrid / Foto: Ángel Martínez Levas
Figure 111
Figure 111

Doorpost, 970–990, marble: left side; Museo Arqueológico Nacional, Madrid, inv. 1972/103/1

© Museo Arqueológico Nacional, Madrid / Foto: Ángel Martínez Levas
Figure 112
Figure 112

Doorpost, 970–990, marble: underside; Museo Arqueológico Nacional, Madrid, inv. 1972/103/1

© Museo Arqueológico Nacional, Madrid / Foto: Ángel Martínez Levas

Where was the workshop that produced the architectural decoration for Durrī’s, al-Ṣiqlābī’s and al-Muṣḥafī’s private residences located? If the cyma was carved for Jaʿfar al-Ṣiqlābī, his access to the architectural workshops through all the commissions he carried out on behalf of the caliph put him in a position to order such works for himself. Certainly that would have been the case with the decoration of the House of Jaʿfar at Madīnat al-Zahrāʾ – but the implications of this are that the state workshops also undertook jobs for buildings outside the palace-city and outside the specific patronage of the Umayyad caliphs. Surely Durrī’s role in the luxury arts industry, as revealed by his involvement in the pair of ivories made in 964 for al-Ḥakam II (Figure 86) and Ṣubḥ, gave him similar access. This opens up the probability that the caliphal workshops also undertook work for members of the court elite, as Antonio Vallejo has recently argued.215 He notes, ‘The material evidence thus shows a more complex reality that is not restricted exclusively to Madīnat al-Zahrāʾ and official buildings. The proliferation of palaces among the members of the aristocracy linked to the state, which we know mainly from written sources, generated an increased demand for decorative materials made of marble, especially support elements and other elements about which we are little informed’. Furthermore, he notes that the anonymous expression baraka min allāh li-ṣāḥibihi, ‘blessing from God to its owner’, is inscribed on the gussets of some capitals, associated not only with palaces but also with mosques, whose construction was sponsored by this elite as pious works. He makes the intriguing suggestion that if these non-state commissions were indeed produced within the caliphal workshops, the state may have controlled this commercialisation in some way, perhaps by receiving ‘taxes or benefits of some kind from this activity’. This has wider implications for the discussion in the following section, that these workshops also produced portable objects under the wider patronage of the court elite.

Al-Manṣūr was thus not the first non-royal figure to make use of the luxury arts ateliers to produce decorative architectonic elements for a palatial construction that he was building for himself. By the time he commissioned the first of al-Madīnat al-Zāhira’s buildings, it may have been quite well-established practice that the caliphal workshops took on other work, especially if the state taxed it or benefitted in some other way. What is different about the ʿĀmirid patronage of these industries is the extent and scale to which the ḥājibs engaged them, and the deliberate messages that were encoded in their visual imagery and epigraphic programmes, as we will see in the following chapters.

2.2 Anonymous Objects

Though no portable objects survive in either Jaʿfar’s name, we can assume that ivories and other products of the luxury arts industry were made for such important court officials but have not survived. If an ivory gift was given by al-Ḥakam II to Ziyād ibn Aflaḥ, the ṣāḥib al-madīna and ṣāḥib al-shurṭa of Madīnat al-Zahrāʾ, in 968 (Figure 12), it seems inconceivable that an ivory gift would not have been given to Jaʿfar al-Ṣiqlābī, the caliph’s ḥājib and the person through whom the Cordoba mosque extension was carried out. Al-Ṣiqlābī was at the peak of his career in the 960s, precisely when the greatest number of ivories was made. The question remains whether the luxury arts ateliers functioned as a royal monopoly, or whether men such as al-Ṣiqlābī would have been able to commission their own objects; if Vallejo’s suggestion of the commercialisation of the state workshops is right, there may have been no obstacle to this, if the price were right and the materials were available.

A figure of the importance and stature of al-Ṣiqlābī, who did not hesitate to boast of his titles on the architectonic elements he commissioned on the caliph’s behalf, would not have hesitated to do the same with any ivory he commissioned for himself; and, on the model of the Ziyād ibn Aflaḥ pyxis, it is likely that a caliphal commission would have named him as well. But there is a possibility that, among the several extant objects (including ivories) with un-named patrons, we are looking at objects made not for the open market, but for members of the court elite or the state administration, beyond the restricted circle of the caliph, his family and closest officials. This is the scenario in which we can imagine al-Manṣūr’s commissioning of the silver palace for Ṣubḥ being possible. As we have seen, there were thousands of fityān of differing hierarchies within the Cordoban bureaucracy. These men would have wished to signify their high status through the symbolic ownership of objects associated with the ruling class. This elite market could afford the expensive materials employed in such commissions, and had access to the craftsmen of the Dār al-Ṣināʿa, who may have been allowed the flexibility of working for them. Protocol may have dictated that these non-royal/non-ḥājib patrons not name themselves. The standard commissioning formulae on objects produced in the court ateliers reaffirmed an institutional hierarchy at whose summit was the ruler (or regent), with the court bureaucracy arrayed below him. The inscriptions on luxury objects, as on architectonic elements, mark them as having been produced under caliphal (or ʿĀmirid) patronage. Official commissions may have had their inscriptions formulated within the court chancery, subjected to rigorous inspection to ensure that they conformed to court protocol. This access may not have been available to private patrons from outside the ruling circle. The craftsmen may thus have followed a text given to them by the commissioner, or resorted to a standard series of good wishes. These had the benefit of being politically neutral, for clients who did not enjoy high status at court, and may also have been easier to formulate if one’s Arabic was not so good.

In his study of the ʿĀmirid-period ivory casket now in Doha, whose anonymous inscription comprises a long list of blessings (Chapter 7: 4.1.1; Appendix 4.16), Oliver Watson discusses irregularities in the inscription, one reason why some scholars considered this object to be a forgery when it first appeared on the art market.216 These irregularities include the use of the full basmala and the presence of a month in the date, though, as we have seen, the architectonic elements made for al-Muṣḥafī included a full basmala; the inscription on al-Manṣūr’s basin may well have begun with the full basmala; and it will be seen later on the Palencia casket. The minbars from the al-Qarawiyyīn (Chapter 7: 2.5) and Andalusiyyīn mosques (1.1, Appendix 4.4) also specify the months in their inscriptions; the former (made in 395/1005, the year after this casket) also includes the phrase wa dhālik fī, as does the Doha box. The Doha box’s inscription features a number of spelling mistakes (yūmn, ṣābighah, li-ṣāḥibihā), though such mistakes are actually surprisingly common, even on the ivories made for members of the caliphal family. The word al-āʿāliyah is otherwise unknown in these inscriptions, though a related form, yad ʿāliyah, occurs on another anonymous casket, in the Musée des Arts Decoratifs in Paris (dated 355/966). As Watson points out, the misspellings on the Doha box seem to result from phonetic transcriptions of the way people actually spoke in contemporary society – for example, the lengthening or shortening of vowels (yūmn for yumn), the substitution of heavy for soft letters and vice versa (ṣābighah for sābighah). This kind of phoneticised spelling is characteristic of the inscriptions on tombstones, which also regularly employ the full basmala as well as more specific dates, including months: see, for example, the very specific date given on the epitaph of the son of an ʿĀmirid fatā (Jamʿah or Jumʿah ibn Fattūḥ/Futūḥ/Fatūḥ), where the date of death is given as ‘Wednesday eve at ni[ne (nights) from] Jumādā II 374/18 November 984’ (Figure 8).217 As Watson notes, tombstones were the one kind of formal inscription that every well-heeled ordinary person would have occasion to commission.218 The Doha box may thus have been, as Watson concludes, a ‘commercial piece ordered by a wealthy individual for his own pleasure’.

The lack of a named owner in the inscription on the Doha box associates it with other anonymous commissions, including the casket in the Musée des Arts Décoratifs, and another in the Museo Nazionale del Bargello in Florence, which is undated (Figures 148–149, Appendix 4.19). The inscriptions on both these caskets follow the same formula as the Doha box: a short string of blessings as single nouns, followed by a longer series of blessings pairing nouns and adjectives, before the phrase ‘to its owner’ and, on the Paris box, the date. For the blessings, words are used that are not part of the vocabulary of court inscriptions, where the blessings usually share roots with alqāb or other honorific titles, as we will discuss in more detail in Chapter 8. Words like sābighah are employed on these three anonymous boxes (misspelt as ṣābighah on both the Doha and Bargello boxes), while the phrase yad ʿāliyah on the Paris box has morphed into al-āʿāliyah on the Doha box, produced some forty years later.219

It is also highly likely that apprentices practised their carving skills on the objects made for non-royal patrons. On the lid of the Bargello box, for example, to the right of the central mount, something seems to have gone awry with the carving of the leaf motifs, which have got tangled and crowded; they are not clearly defined or well-executed, compared even with the left-hand side of the mount, as if the carver lost his way during the process and had to make the best of a bad job (Figure 149).

These anonymous commissions are an interesting phenomenon which require deeper study.220 But thinking of them as a commercial production for a totally open market, as Watson implies in his study of the Doha box, is missing the mark. This industry, physically based at the heart of the palace-city and principally functioning for the needs of the court, could only have been accessed by an elite associated with the court – that is, those who held offices (and, presumably, high offices) in the state bureaucracy. Indeed the fityān, the elite slaves who supervised the court commissions, would have had physical access to these workshops, and could sometimes have commissioned a piece for themselves, alongside the court commissions they were overseeing, hence the close stylistic and iconographic associations that can be noted between the Doha box and other contemporary ʿĀmirid ivories (Chapter 7).

A picture of less restricted patronage of the luxury arts implies a larger scale of production than we can now judge from the extant pieces. As Watson has noted, there are obvious reasons why court-commissioned objects have survived in greater numbers: ‘they were kept together in the court treasury, gathered centrally as individuals died, looked after carefully, given in groups as diplomatic gifts. They were ready to be looted en masse … and would subsequently be donated in groups to Christian churches, which have been the main if not the only guardians of Islamic ivories since medieval times’. The objects made in the more ‘commercial’ branch of the industry would not have been centrally stored, and ‘as individual items without a treasury structure to care for them [they were] much more vulnerable to the vicissitudes of history’.221 Just as the wealthy inhabitants of Cordoba had private libraries, it is likely that they also had private khizānahs, imitating those of the caliphs, in which to store their own precious objects; but the lower rank – and possibly lesser quality – of these objects may have made them less appealing to Christian kings and noblemen and thus less likely to be rededicated and preserved in the churches of northern Iberia.

2.3 How al-Manṣūr Used Objects

The technical and decorative similarities between ʿĀmirid objects, the recurrence of the same craftsmen’s names on objects and in ʿĀmirid architectural contexts, and the designation of the fityān through whom these works were carried out as ʿĀmirid freedmen through the nisbaal-ʿāmirī’, argues for a continuation of this centralised luxury arts industry under direct ʿĀmirid patronage. Furthermore, the inscription on al-Manṣūr’s basin states that it was manufactured bi-qaṣr al-zāhira, ‘in the palace of al-Zāhira’. Consequently it may be deduced that one branch of the Dār al-Ṣināʿa was also physically associated with the ʿĀmirids, located within the precincts of their palace-city. We do not know whether the core of the state atelier remained based at Madīnat al-Zahrāʾ, though the absence of any evidence for Hishām II’s own patronage of art and architecture combined with the shift of patronage power to the ʿĀmirid palace-city would argue against that. Was the Dār al-Ṣināʿa one of the state institutions transferred from Cordoba to al-Zāhira at the time of the city’s foundation? If so, it would presumably have been one of the first to go, since the ‘magnificent’ decoration of al-Manṣūr’s palaces required the involvement of craftsmen. Indeed, Ibn ʿIdhārī’s statement – that ‘he brought artists and labourers [to work on the construction]… and in this way revetted the palaces with a magnificence that wearied the eyes’ – may be interpreted not just as a topos or gesture of rulership, but as an indication of the physical transfer of a state institution that previously had an intimate patronage association with the caliphs.222 An obvious inference is that, in addition to objects and architectonic decoration in stone and marble, the extant ivories and objects in other media that can be associated with the ʿĀmirids (Chapter 7) were also manufactured at al-Zāhira.

As far as we can tell from the surviving dated inscriptions, the manufacture of luxury objects appears not to coincide with major architectural projects; the exception is al-Manṣūr’s basin, made in 987–8, the same year in which he began his construction of the Cordoba Mosque extension. It was possibly a special commission to commemorate that significant architectural intervention (see discussions in Chapters 7 and 8). In more general terms, the development of the luxury arts industry under the ʿĀmirids coincides with the increasingly elaborate court protocol that was discussed in Chapter 2: luxury objects and furnishings were an essential ingredient in this ‘ritualisation’ of the court, and in order to continue to cultivate his courtiers and allies, al-Manṣūr needed a ready supply of objects, textiles and money to present as khilʿa. Moreover, the increasing intensity of ʿĀmirid military campaigning and diplomatic relations also necessitated the production of luxury goods to fulfil the ritual of gift exchange, as may have been the context for the creation of the Braga pyxis (Chapter 2).

This notion of ‘strategic gift giving’ (iṣtināʿ) was introduced at the start of this chapter, with the anecdote about al-Manṣūr commissioning a model palace in silver as a gift for Ṣubḥ, which won her heart and support and in turn that of her consort, the caliph himself. Perhaps more important than the object was the staged presentation of the gift: as Ballestín comments, ‘[Ibn Abī ʿĀmir] did not make this gift to Ṣubḥ with discretion, but ostentatiously and with as much publicity as possible. He wanted the people to see this extraordinary gift … We can imagine that the fact that they talked about it for a long time afterwards, word running from mouth to mouth, was precisely the effect that Ibn Abī ʿĀmir wanted to achieve’.223 All the sources emphasise al-Manṣūr’s generosity to his close courtiers. Eduardo Manzano has called this a ‘profligate policy of gift-giving’, a canny calculation from the beginning of Ibn Abī ʿĀmir’s career of how to cultivate loyalty in the right places, recognising the ‘complicities and common interests, networks of power and mutual interest’ that according to Manzano characterised al-Andalus in the tenth century.224

Francisco Prado-Vilar has argued that al-Manṣūr’s employment of ‘the discourse of the gift’ had a darker side, that we can observe ‘a constant fluctuation in its field of meaning among reward, bribery and punishment’.225 Specifically, he argues that al-Manṣūr was the creative force behind the iconographic programme of the al-Mughīra pyxis (Figures 3, 171), made for the younger brother of al-Ḥakam II in 968, several years before the crisis in which al-Mughīra was murdered for being the main threat to Hishām’s legitimate succession (Chapter 1). Prado-Vilar’s ingenious reading of this pyxis, argued across two articles, is as a ‘hostile warning’ to al-Mughīra, to ‘enjoy his comfortable life as a member of the royal family … and leave to others the execution of political power’.226 Prado-Vilar’s perception of al-Manṣūr’s role in this commission is on the melodramatic side, referring to his ‘unbounded ambition’, even his ‘Machiavellian mind’.227 This belies a common misconception of the historical complexities of this period, which Prado-Vilar nuances somewhat in his later article. Whether or not al-Manṣūr was the commissioner of the al-Mughīra pyxis, technically his position at court by this date could have enabled his access to the luxury arts ateliers, as it allowed him to commission the silver palace for Ṣubḥ. Prado-Vilar’s reading prioritises a sophisticated mindset which conceived an iconographical programme replete with allusions to poetic language, as well as the visual language of mirrors for princes, in particular the international bestseller, the Kalīla wa Dimna. Such a programme relies on the equally sophisticated and educated mindset of the object’s recipient, to correctly read and interpret the allusions, and thus the warnings, embedded therein. This sophisticated interplay between poetic and visual language – what Prado-Vilar calls the ‘complex dialectic interface between the visual rhetoric of portable objects and courtly poetic discourse’ – certainly provides the context in which I understand the setting of ʿĀmirid works of art, and the appreciation and understanding of their iconographic and epigraphic programmes by al-Manṣūr’s close courtiers.

Al-Manṣūr also employed this tactic outside the internal world of the Cordoban court, in particular through the distribution of gifts to his soldiers and allies after his successful military campaigns. An especially elaborate gift was distributed in the aftermath of al-Manṣūr’s campaign against Santiago de Compostela in 997. Assembling his Christian allies ‘and others who had shown themselves the friends of the Muslims’, he ‘rewarded each man according to his rank, distributing robes of honour among them and their followers’. The list has been preserved as follows:

2285 pieces of various kinds of ṭirāzī silk (shiqqa); 21 pieces of sea-wool (ṣūf al-baḥr); two ʿanbarī [perfumed with ambergris] robes; 11 pieces of siqlatūn [scarlet coloured textiles]; 15 pieces of striped material (murayyash); 7 brocade carpets (namat); 2 robes of Byzantine (rūmī) brocade; and 2 marten (fanak) furs.228

This sumptuous list begs the question whether textile gifts to his allies were a feature of all al-Manṣūr’s campaigns, or whether participation in this exceptional campaign deserved exceptional recompense. The more than 2000 pieces of ṭirāz distributed on this occasion was not booty captured during the Santiago campaign. The techniques named in the list indicate that these were precious textiles produced or acquired in Cordoba, implying that al-Manṣūr had brought them with him on campaign, in order to distribute them later as rewards for good service in battle. This was likely something he did as standard: for example, in 978, Count García Fernández of Castile, while campaigning against al-Manṣūr, attacked the fortress of Gormaz, seizing it and a substantial booty, including many Andalusi textiles, which he immediately dedicated to the Monastery of San Simón in Covarrubias.229 As Xavier Ballestín has discussed, khilʿa did not consist solely of textiles and robes of honour, but often included coin and other gifts such as horses, accoutrements and arms.230 Perhaps the ivories that al-Manṣūr is said to have had with him in his campaign tent (see below) might likewise have been brought for later distribution as khilʿa. During the campaign, the display of these objects added to their value, and made them more appreciated at the moment of their distribution.

All this implies that al-Manṣūr actively commissioned works of art in order to have a ready supply of textiles and other gifts to distribute after his successful campaigns or through his intense diplomatic activity. Given how many campaigns al-Manṣūr and ʿAbd al-Malik prosecuted, the Dār al-Ṣināʿa and Dār al-Ṭirāz under the ʿĀmirids may have been operating almost entirely to feed this need for khilʿa. It is significant, then, that there are no surviving textiles inscribed in the name of any of the ʿĀmirids. The hidden, incidental occurrences of ‘al-Muẓaffar’ on the Suaire de Saint Lazare (Figure 136A) do not make it a ṭirāz; indeed ʿAbd al-Malik’s laqab may have been kept deliberately small and imperceptible, so as not to risk appearing too much like a ṭirāz inscription.

On the other hand, the surviving silk textile inscribed with the name of Hishām II indicates that the official products of the Dār al-Ṭirāz, those intended to be given away as state gifts, continued to observe the royal protocol of naming the ruler and his titles. One of only two examples of ṭirāz to survive from al-Andalus, the so-called ‘Veil of Hishām’ is an almost full-width fragment (1.09 m is preserved, which is almost selvedge to selvedge) of a long headdress woven in silk taqueté and originally crimson in colour, which would have had tapestry-woven bands at each end (Figure 9, Appendix 4.2).231 One of these tapestried bands survives: it is divided into three equal zones, comprising two identical inscriptions in mirror-image on either side of a central band containing thirteen figurative medallions. The inscriptions invoke blessings in the name of ‘the khalīfa, the imām ʿAbd Allāh Hishām al-Muʾayyad bi-llāh, amīr al-muʾminīn’. The lettering is slightly floriated and is carefully plotted so that al-khalīfa appears at the centre. The figurative medallions are rounded octagons rather than circles, and are separated by stylised floral motifs. Eleven of the medallions feature alternating birds and quadrupeds, while the two medallions on the far right contain small human figures. Both have long hair, only vague facial features, and are seated on cushions; the right-hand figure extends a hand in the direction of his companion in the left-hand medallion, and holds in his other hand a round-bottomed bottle, which Fernández y González read as a sceptre. The left-hand figure has its hands folded in its lap.

In 1875, Fernández y González conjectured that the figure holding the sceptre/bottle was Hishām II himself, and that second figure was his mother, Ṣubḥ, even that she might have commissioned the textile, the blessings in the inscription representing her well-wishes for the future of the caliphate. This suggestion has lingered in the literature, though his interpretation of the symbolism of the animals in the other medallions has been quietly forgotten. However, as Pérez Higuera notes, the two human figures should be read as ‘symbolic manifestations of power and not concrete personifications’. In fact, they can be related to other works of art datable to al-Manṣūr’s ḥijāba. They evoke the gestures of the seated figures of the majlis scene on the Pamplona casket (Figure 120); as on the ivory, the figures are enclosed within medallions, separated by floral motifs. In other medallions on the textile, the birds that appear to hold worms or snakes in their beaks recall the ducks and worms on the ʿĀmirid capitals and basins (Figures 23–4, 118A, 161–170). Without taking these stylistic connections too far, given the different execution demanded by the different materials and their associated techniques, they again imply the close executive relationships between the different branches of the luxury arts industry, and that these continued during the ʿĀmirid period.

The presence of Hishām’s name on this textile in no way indicates that this was a piece commissioned by him or intended for his personal consumption.232 Since a major role of the industry was to produce textiles explicitly to be given as khilʿa, it cannot be assumed that Hishām ever even saw this textile, unless he presented it to someone. The textile was found in the mid-nineteenth century wrapped around relics in a casket under the altar of a church in San Esteban de Gormaz (prov. Soria), a region in which al-Manṣūr campaigned in 989 and 993–4 (see Timeline). It could have belonged to a member of al-Manṣūr’s army who received it as a reward for good service; if the Santiago campaign is anything to go by, this could have been a Christian ally, who subsequently rededicated the textile to his local church. Presumably the more than 2000 pieces of ‘ṭirāzi silk’ given away after the Santiago campaign would all have borne similar inscriptions in Hishām’s name. Thus, while al-Manṣūr adopted the royal role of bestowing ṭirāz as khilʿa, he did so explicitly in the caliph’s name.

In order to have such objects ready to distribute at the end of the campaign – before your armies and allies disperse to their regions – you need to carry an appropriate number of luxury objects with you into the field, so that the campaign tent becomes a sort of travelling Khizānah, where these objects are stored but also no doubt displayed, their presence spurring the army and allies on to heroic deeds in order to receive them as rewards. This may be why al-Manṣūr is said to have had ivories and other precious objects with him in his campaign tent, though this is related in verse in the mythologised account of one of the heroes of the so-called ‘Reconquista’: the Poema de Fernán González was written down c. 1260, ‘recording’ the heroic battlefield deeds of Fernán González, first Count of Castile (d. 970). Stanzas 276–280 sing of the looting of a number of ‘precious ivory caskets’, vessels made of gold, hangings and soft furnishings of silk, arms and armour made from gold and silver, from al-Manṣūr’s campaign tent at his frontier base in Medinaceli. The stanzas run:

The poem states that these ivories were rededicated by Fernán González at the church of San Pedro de Arlanza (prov. Burgos), where they were ‘placed on the altar’ as trophies; it is likely that the visible presence of ‘very precious ivory caskets’ in the treasury of that church led to the myth that they had been captured in battle from that supreme enemy of the Christians, ‘Almanzor’. The two Andalusi ivory caskets that survived in the Monastery of Santo Domingo de Silos also came to be associated with Fernán González (Figures 99, 174), even though the larger casket, which came to hold Santo Domingo’s remains, was not even produced until 1026, some fifty years after Fernán’s death.234 Obviously such mythologised accounts cannot be relied upon for their historical content, though they have been taken as factual accounts by several historians of ivory.235 Indeed Fernán González and al-Manṣūr would never have crossed paths: the ḥājib’s first campaign was not undertaken until 977 (see Timeline), seven years after Fernán’s death, although of course his son, García Fernández (d. 995), encountered al-Manṣūr on the battlefield. Nevertheless, what interests us here is the memory that this poetic account preserves of al-Manṣūr’s patronage of luxury objects or, at least, that luxury objects were a sine qua non of the status of Muslim potentate that al-Manṣūr held in the popular Castilian imagination in which context this poem was created.

Also interesting is the suggestion that these objects had been captured in battle, and had therefore been taken into the field. Though the notion that Islamic objects transferred across Christian borders purely as the result of booty in such campaigns has been challenged in recent years,236 the fact that stores of objects to be distributed as khilʿa were transported with the army is corroborated by what we know of the management of al-Manṣūr’s North African campaigns. This was the case from the time al-Ḥakam appointed the young Ibn Abī ʿĀmir to the post of qāḍī al-quḍāt and inspectorate (amāna) of the Maghrib in 973, when he was sent ‘with shipments of money, jewels, ornaments and presents of honour, so that he may distribute them abundantly’ among those Berber tribes who wanted to adhere to Umayyad authority.237 The functionaries that travelled with him on this occasion included a khāzin (treasurer), the ṣāḥib al-makhzūn, responsible for payments, and his friend Aḥmad ibn Muḥammad ibn Ḥudayr, ṣāḥib al-khizāna wa’l-sikka. Once al-Manṣūr was in a position to appoint his own officers, he entrusted his governor al-Sulamī with administering and distributing gifts of honour to the Berber chiefs. Al-Sulamī did not receive a fixed number of gifts to distribute as khilʿa, but had discretion to freely use all the resources necessary, implying that these resources were available on the ground and had been supplied from Cordoba for the purpose.238

It can be assumed that a military campaign provided the opportunity for a particularly potent form of display, and al-Manṣūr would surely have continued to hold court in his campaign tent, receiving not only his courtiers, generals and soldiers, but also his enemies and their envoys. A tent acted as temporary (‘soft’) architecture,239 and, in this respect, it might be suggested that the ceremonial described in Chapter 2 represented only a slightly more formalised version of that conducted in the campaign tent. Though the sources provide no information about ʿĀmirid campaign tents, an entry in al-Rāzī’s Annals mentions the large red tent (qubba) that al-Ḥakam gave as a gift to his general, Ghālib.240 This had a ‘marvellous appearance and was a beautiful sight’, and the caliph ordered him ‘with precise instructions and explanations, to erect it in the middle of his camp, to live and receive in it, and to use it to enhance his prestige and the desperation in the heart of his enemy’. Al-Manṣūr would surely have secured for himself just as marvellous a campaign tent as Ghālib, whose defeat marked the culminating moment of his rise to power, and used it in just this way; he could have enhanced its visual impact with luxury furnishings, including textiles and precious objects, to recreate the physical environment of his majlis at al-Madīnat al-Zāhira. ʿAbd al-Raḥmān III also filled his campaign tent with luxury contents, as we know from the account of what was looted from it after his defeat at Simancas in 939. Ramiro II of León (d. 951) and Fernán González of Castile made off with the tent itself, the clothing stored within it – presumably khilʿa ready to be distributed at the end of the campaign – and most notably ʿAbd al-Raḥmān’s personal copy of the Qurʾān.241 As we saw in Chapter 5, al-Manṣūr also took with him on campaign his autograph copy of the Qurʾān, as well as the winding sheet, woven by his daughters, in which he should be buried if he fell, and a casket in which to collect the earth from every battlefield. No doubt these almost talismanic objects had a prominent place in his campaign tent as well. The visibility of the portable objects he brought with him on campaign would provide an opportunity for al-Manṣūr to reassert their messages to those who visited him in his tent. While the Poema de Fernán González may be anachronistic, its verses nevertheless supply one possible context for how al-Manṣūr used the precious objects that he owned, one which also suits both the militaristic language and the aesthetic of the imagery on the extant objects’ epigraphy and iconography, as discussed in the following chapters.

In addition to al-Manṣūr’s use of objects in strategies of loyalty and legitimation and for the display of power, wealth and status, one final and important consideration is his apparent delight in engaging with the material thing. As we saw in Chapter 3, the poet Ṣāʿid al-Baghdādī was only admitted to al-Manṣūr’s dīwān of court poets after improvising verses in description of an elaborately laid tray. This tray contained objects executed in and decorated with precious metals and gems and scented woods, with hidden compartments and sculptural figures, perhaps even automata. Though Ṣāʿid’s initial verses were satisfactory, they were not spectacular, and in particular he overlooked a detail – ‘a boat, in which was a maiden rowing herself with oars of gold’. Al-Manṣūr pointed this out to him, and Ṣāʿid improvised new verses on this motif; finally ‘al-Manṣūr regarded the poem worthy of the object described’. As Prado-Vilar has pointed out, this anecdote, and especially the fact of al-Manṣūr’s initial unhappiness with the neglect of one aspect of its decoration, indicates an interested and deep engagement with the materiality and visuality of the object. It ‘reveals a model of viewing courtly portable objects that privileges close observation and attention to detail’; the viewer ‘finds pleasure in the discovery of nuances and variations on the commonplace’.242 Further, the appropriate language that brings the material objects fully to life is poetry. This tray and its sculptures would have been manufactured by artisans in the ʿĀmirid luxury arts industry, and given al-Manṣūr’s close engagement with each detail, as indicated in this anecdote, the craftsmen may have worked to his direct specification.

This chapter has aimed to outline the location, infrastructure and personnel of the caliphal luxury arts industry, the Dār al-Ṣināʿa, how it functioned and what material resources it consumed, to indicate the extent to which its caliphal, and later its regency, patrons could and did control the products that were made in it. We have established that already in the early years of Hishām’s reign, and certainly by the early 980s, the ʿĀmirids had taken control of this industry, which they relocated, in whole or in part, to al-Madīnat al-Zāhira. As far as we can reconstruct, the way in which the industry had functioned under the caliphs was continued under the ʿĀmirids; but since so few extant objects were made under direct caliphal patronage, compared to the surviving ʿĀmirid objects, it is difficult to speculate further as to whether the ḥājibs engaged more closely than the caliphs in the functioning of their luxury arts industry and in the creation of its products. We have begun to understand the ways in which the ʿĀmirids controlled and exploited this industry in their articulation of power; we have also outlined some of the functions and likely audiences of the objects they commissioned. It is now time to discover what these objects were. The next chapter introduces the works that can be associated directly or stylistically with ʿĀmirid patrons, identifying for the first time a corpus of ʿĀmirid art. In Chapter 8, we will look to the iconography of these works and the semantics of their epigraphy, to attempt to define more clearly the messages of self-expression that the ʿĀmirids hoped to convey through their patronage of the luxury arts.

1

This English translation is based on the Spanish translation in Ballestín 2004a, 63–4. See also al-Maqqarī, 179 (Analectes, II:61). On other gifts and favours presented to the women of the caliph’s harem, Bayān II:268 [translation 417]; al-Maqqarī, 179 (Analectes, II:62).

2

See Clark 1986.

3

On the Fatimid Treasuries, see especially Oleg Grabar 1969; Romberg 1985; Sanders 1994, 23–32; al-Ḥijjāwī al-Qaddūmī 1996.

4

Von Grunebaum 1955, 281.

5

Robinson 1995, 425, talking specifically of the Kitāb al- Muwashshāʾ (before 936), on which see eadem 2007, 19 (n. 39), 26.

6

Robinson 1995, 424.

7

Ballestín 2004a, 65, comments that the word used in the text is mawqūf, indicating (in his words) ‘dinero inmovilizado’, suggesting that the metal used came from reserves of stored precious metal that could not be minted into coin.

8

Bayān II:268–9, translated and discussed by Ballestín 2004a, 63–9.

9

Echevarría 2011, 50–51.

10

Bagnoli et al 2011, 118, cat. 55.

11

Exceptions to this are Robinson 2007 on the Pamplona casket and Makariou 2001 on the objects made for ʿAbd al-Malik more broadly.

12

A question that still besets many of the essays in the volume on the Salerno ivories, see Dell’Acqua et al 2016; on Byzantium, Cutler 1994, 66–78, ‘The Question of Workshops’.

13

HEM I:349.

14

Ibn Khaldūn (apud al-Maqqarī 1968, 524–5, 569).

15

Ecker 1992, 4.

16

On whom see Terés 1960, and Anderson 2020, on the earliest scientific instruments made in al-Andalus. The earliest extant examples all date from the Taifa period, but textual sources tell us that Ibn Firnās made an armillary sphere for ʿAbd al-Raḥmān II (r. 822–852), and a waterclock for Muḥammad I (r. 852–886).

17

Acién Almansa and Vallejo Triano 1998, 118–9, esp. n. 67: ‘No cabe duda que el conjunto de los talleres estatales que se conocen como Dār al-Ṣināʿa se emplazó en la parte occidental, en la proximidad de la Bāb al-ʿAṭṭārīn. La anécdota transmitida por al-Maqqarī, relativa al traslado del alfaquí Abū Ibrāhīm, desde la mezquita de Abū ʿUthmān al noroeste del Alcázar, a presencia del califa al-Ḥakam II, indica que la Bāb al-Ṣināʿa era la puerta más próxima al palacio desde dicha mezquita’, citing al-Maqqarī, 172–3.

18

Acién Almansa and Vallejo Triano 1998, 119.

19

Acién Almansa and Vallejo Triano 1998, 118, n. 67: ‘La ceca permaneció en ese lugar hasta su traslado a Madīnat al-Zahrāʾ, puesto que en los comienzos del s.X el geógrafo Ibn al-Faqīh al-Hamadānī confirmaba su situación “en el barrio de Bāb al-ʿAṭṭārīn”’, citing Lévi-Provençal 1932, 76, n. 2.

20

Acién Almansa and Vallejo Triano 1998, 119, and citing Pellat 1961, 73, 103, 163. See also Arjona Castro and Marfil Ruiz 2004.

21

As they still are today in the city of Fez (Morocco), for example. For a discussion of traditional processes of tanning and leather-working in late Ottoman Damascus, see Milwright 2018, esp. 251–4.

22

Ibn ʿAbdūn, in his ḥisba treatise written in twelfth-century Seville, recommends this: see García Gómez and Lévi-Provençal 1992, 113–4.

23

Salinas Pleguezuelo 2012, 695. The information given here is based on her thesis, especially Chapter 7, “Los hornos y la actividad alfarera”, pp. 579–698.

24

Salinas Pleguezuelo 2012, 583, citing Aziz Salem 1984– 5, 227.

25

Arjona and Marfil 2004, 142.

26

Arjona and Marfil 2004, 141.

27

Anales, §§77–78.

28

Sokoly 1997; Contadini 1998, 39–58; Sokoly 2017.

29

Now in the Cleveland Museum of Art (1977.188), this textile is woven in the name of the caliph ʿAbd al-Raḥmān III and dated 330/941–942. It also says that it was made under the supervision of ‘Durrī, ʿabdu-hu’: see Santiago-al-Andalus, 369–370; Mackie 2015, 172–3 (5.3, 5.4). It is length of linen over a metre long (and probably originally closer to 2 metres long); its inscription was embroidered with (originally crimson) silk and is largely now lost. The wording has been reconstructed from needle holes and the few remaining stitches. It is best seen in black and white photographs, as is almost impossible to read anything at all on the textile in its current state. I am grateful to Louise Mackie for allowing me to study this textile during a research visit to Cleveland in 2008.

30

Vallejo 2010, 184, citing Castejón (1961–2), 137.

31

Silva Santa-Cruz 1999; Bariani 2005.

32

Anales, §35; Ibn Ḥayyān 1965, 66; Acién Almansa and Vallejo Triano 1998, 134.

33

Nicolle 2001, 11.

34

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

35

Ballestín 2006.

36

Ballestín 2008, 146–7.

37

Puerta Vílchez 2013a, 76 (English version p. 63), relates that al-Ḥakam ‘sent 10,000 dinars of pure gold to buy the Kitāb al-Aghānī from Abū’l-Faraj al-Iṣfahānī, who sent him a copy of this magna literary encyclopaedia even before it saw the light in his native Iraq’.

38

Andaloro 2006, vol. 1, 44–49, cat. 1.1 ‘Manto di Ruggero II’; vol. 2, 171–181, ‘Manto di Ruggero II e le vesti regie’, by Rotraud Bauer; Mackie 2015, 162–5.

39

Lévi-Provençal 1957 [1996], 236–238 (gates at the Madīna of Cordoba); Torres Balbás 1957 [1996], 432 (gates at Madīnat al-Zahrāʾ); 592–593 (gates in the Alcázar at Cordoba); Safran 2000, 68.

40

Vallejo 2010, 133.

41

Antonio Vallejo, personal communication, 5 June 2014.

42

Vallejo 2010, 184.

43

The Fitero and Instituto de Valencia de Don Juan caskets, both made in 966, possibly for Wallāda, a daughter of ʿAbd al-Raḥmān III: see Silva Santa-Cruz 1999 and Bariani 2005.

44

Makariou 1999, 133.

45

Ocaña 1976, 219–220.

46

This is surely the same person as the Fāʾiq al-Nizāmī mentioned at Bayān II:277–279 [translation, 431–434], who together with Jawdhar led the campaign to displace Hishām from the caliphal succession.

47

Blair 2005.

48

Souto 2005, 255: ‘Though undated the work must be located between 5 February and 1 October 976, the period between the naming of Hishām as heir and the death of his father, the caliph al-Ḥakam II’.

49

Cf. the inscription commemorating the construction of a minaret and gallery at a mosque in Cordoba, by Sayyida Mushtaq, the mother of al-Mughīra: see Lévi-Provençal 1931, 24–5 (#18); Martínez Núñez 2006, §1; Anderson 2012, 656–661. This commission was carried out by Jaʿfar ibn ʿAbd al-Raḥmān al-Ṣiqlābī, on whom see below.

50

There is one instance on a column base in Cordoba IV of a signature Mubārak ibn Hishām (though Mubārak on its own occurs frequently). Ocaña 1986, 66, understood this man to be a mawla of Hishām II, which would provide the only evidence of this caliph’s personal involvement in the luxury arts industry. However, we cannot say for sure that this Hishām is the caliph. See Souto 2010, 2.56 (for Mubārak ibn Hishām), 2.55 (for Mubārak).

51

As on the Pamplona casket and Braga pyxis, where this kunya completes the name of the fatā al-kabīr, Zuhayr. See Appendix 4.11 and 12 for the inscriptions on these two objects.

52

As on the al-Manṣūr basin (Appendix 4.7).

53

Blair 2005, 85 says that Zuhayr’s designation as mamlūk identifies him as a ‘freed slave’.

54

Bayān II:277–279 [translation, 431–434].

55

‘Bayzara’, Encyclopaedia of Islam, 2nd Edition.

56

Ocaña 1976, 219, and Ocaña 1940, 441–2. These inscriptions and associated offices are updated and tabulated in Martínez Enamorado 2006, 66–7. As indicated above, ‘mawla’ did not necessarily imply Jaʿfar’s manumission, and may have had a general sense of ‘client’: however, as a ṣiqlāb, Jaʿfar would certainly have started his career as a slave, and it would not have been unusual if ʿAbd al-Raḥmān had manumitted him before his death: cf. Anales, §208, which tells us that as soon as al-Ḥakam fell ill, in Rabīʿ II 364/January 975, he freed over 100 of his slaves.

57

Ocaña 1976, 219–220.

58

Al-Maqqarī, Analectes, I:247–248; cf. Ocaña 1976, 220–221.

59

Bayān II:249 [translation, 385–6], confirmed by six inscriptions from the year 353/964–5, cf. Ocaña 1976, 221.

60

Ocaña 1976, 222: Jaʿfar is cited as ‘ḥājib’ in the inscriptions commemorating the placement of the columns which support the mihrab arch (dated 354/965–6), and those which commemorate the end of the construction of Cordoba III: cf. Lévi-Provençal 1931, 9–21 (#10–#14). Likewise, he is simply ‘ḥājib’ on an inscription (dated 358/969–70) now in the Capilla de Villaviciosa, and the inscription on al-Ḥakam’s fountain basin (dated 360/971–2).

61

See Anales, §104, and Ocaña 1984a, 376–377. On Durrī’s career, see Anderson 2012.

62

In the Anales, ‘khalīfa’ appears to be a salaried office within the caliphal administration, equivalent in status to ‘fatā al-kabīr’. In general, it seems to designate an ordinary freed slave, as opposed to a ‘mawlā’ who was a slave freed by the sovereign. See HEM II:126; Wasserstein 1985, 158, n. 8. Also Anales, §94, for the information on Durrī’s disgrace in Rajab 362/April 973 which required him to forfeit his salary as khalīfa. He only regained the caliph’s favour four months later, through the intercession of the young Hishām. Despite this debt to the walī al-ʿahd, Durrī nevertheless later became embroiled in the plot to install al-Mughīra on the throne: cf. Bayān II: 280–81 [translation, 436–7].

63

Bayān II: 280–1 [translation, 436–7], cited in Anderson 2003, 169.

64

Naváscues y de Palacio 1964a, 241–242; Wasserstein 1993a, 129–145.

65

Martínez Nuñez 1995, 144.

66

Souto 2001, 283; Souto 2010b, 39–45.

67

In his last article on this subject, Souto noted that the mutual identity of different artisans of the same name was ‘una de las varias preguntas sin respuesta que nos hacemos constantemente’ (Souto 2010a, 214). Sadly, as a result of his untimely death, we will never know the conclusion he might eventually have reached.

68

Naváscues y de Palacio 1964a and 1964b. It is interesting to speculate how many other ivories might have hidden inscriptions in this location or on the undersides of plaques hidden by a wooden core or textile lining.

69

Faraj occurs in signatures in Cordoba IV: cf. Rodríguez and Souto 2000a, and Souto 2001, 287–288, and on a capital probably from al-Manṣūr’s ablutions pavilion, Souto 2004, table 2. It should be remembered that the Dhikr Bilād I:40 [II:46] dates the completion of the ablutions pavilion to 390/999–1000, which is close in date to the manufacture of the Pamplona casket (395/1004–5).

70

Souto 2010a, 2.25 and 7.112, citing Martínez Núñez et al. 2007, n. 582.

71

Blair 2005, 84. Souto 2010a, 2.41 (Khayr) and 2.70 (Saʿāda).

72

Blair 2005, 84–5.

73

ʿĀmir appears on the coins minted throughout al-Manṣūr’s ḥijāba, which perhaps functions as a kind of brand. Nevertheless, ʿāmir on its own does also appear as a signature in the Cordoba Mosque (eg nos. 305 and 307): Souto 2010a, 7.107. See Figure 4.

74

Saʿāda’s name is inscribed on three capitals from Cordoba IV (Souto 2001, 295), a capital found at al-Ruṣāfa (Arjona Castro 2000, fig. 4, and 2001, 384), and could be one of the men mentioned in the foundation inscription of the Bāb al-Mardūm mosque in Toledo (999–1000), which reads ‘ʿala yaday Mūsā ibn ʿAlī al-bannaʾ wa Saʿāda’ (Souto 1998, 312). On the other instances of Rashīq, who because of the earlier chronology is probably not the same man as worked on the Pamplona casket, cf. Souto 2001, 294–295.

75

Rodríguez and Souto 2000a, and 2000b, esp. 202–205. On the bridge inscription, he is named as Khalaf ibn Muḥammad al-ʿĀmirī.

76

Souto 2007, 114–5, 126–7 (inscription no. 10) and n. 44: for the capital now in the Qaṣba Mosque in Marrakesh, see Souto 2010a, 7.110; on the capital in Seville, see above (n. 70).

77

Anderson 2012.

78

Souto 2001, 290–291, thought not, because of the time difference between these commissions.

79

Watson, 2005, 171–2.

80

Ferrandis 1935, cat. 8, plates XXI (Fitero); cat. 9, plates XIIXIII (Hispanic Society). This comparison is explored in Anderson and Rosser-Owen 2015, 34.

81

Seen at the lower centre of the back of the Fitero casket, and to bottom right of the lockplate on the Hispanic Society pyxis.

82

On this group of three objects made as a ‘set’, see Anderson and Rosser-Owen 2015.

83

Nieto Cumplido 2005, 72.

84

Watson 2005, 168.

85

It must be noted, however, that Antonio Vallejo believes the plaster in the abacus and equinus zone of these capitals to be a much later addition (personal communication 6 September 2020).

86

As Sarah Guérin notes in the first chapter of her doctoral dissertation on Gothic ivory carving (Guérin 2009), ‘There has been a reticence to accept that ivory carvers and stone carvers could have been the same individuals, even though one of the guilds states this explicitly and the other implicitly … This reticence is based primarily on the similarity of the tools used by modern ivory carvers to those used by wood carvers’.

87

Walker 2015, 266.

88

Guérin 2015, 42. She cites Boileau 1837, Titre LXII.I: ‘Et puet ouvrer de toutes manieres de fust, de pierre, de os, de yvoire, et de toutes manieres de paintures bones et leaus’; and Titre LXI.I: ‘On face d’os, d’yvoire, de fust et de toute autre maniere d’estoffe, quele que ele soit, estre le puet franchement, pour tant que il sache le mestier’.

89

Guérin 2015, 44.

90

Watson 2005, 171.

91

For the importance of wood as an artistic material in the Islamic world, see ‘Islamic Art §VII. Woodwork’, Grove Art Online (consulted 24 January 2020). The author notes that, in Egypt, ‘wood was so scarce and expensive that it had to be imported’, mainly from Syria and the Sudan.

92

Torres Balbás 1957 [1996], 538–551.

93

Bloom 1998, 21–22 and n.33, citing al-Idrīsī 1866, 260; also Hernández Jiménez 1959, 381–99; Fierro 2007.

94

Bloom 1998, 23; Bloom 2005, 212.

95

Bloom 1998, 21–22 and n. 33, citing Ibn ʿIdhārī 1948–51, vol. 2, 238, 250.

96

Cited by Carboni in Bloom 1998, 50.

97

Bloom 1998, 7. He also notes (p. 21) that it took a modern Moroccan craftsman a week to carve a replica of one of the wooden hexagons. Extrapolating from this, he calculates that the carved panels alone represent a minimum of 1000 man-weeks of work, or four craftsmen working continuously for five years. This is only part of the decorative programme, so it would be reasonable to believe that the minbar represented the work of at least a dozen workers over some five years.

98

Anderson and Rosser-Owen 2015, 40–45.

99

As Watson 2005, n. 38, observes, the minbar was started in 966, ‘just as we begin to have the greatest number of dated extant [ivory] caskets’.

100

Fierro 2007, 162 and n. 53.

101

Metropolitan Museum of Art (13.141): Ferrandis 1935, cat. 6; Al-Andalus cat. 6, p. 203.

102

The eyes of some of the figures on the Met’s panel and some drill holes of the framing band are inlaid with a clear stone, perhaps quartz; and traces of green pigments can be seen clearly on many of the leaves, with red on some of the buds and blue on the flower at the left-hand edge of the frame. My thanks to Sheila Canby Curator Emerita of the Met’s Department of Islamic Art for sharing the conservation report on this object.

103

I owe this observation to Sarah M. Guérin, whom I sincerely thank for sharing with me the introduction to her doctoral dissertation (Guérin 2009), where this issue is discussed. I eagerly await the publication of her observations on this issue.

104

Shoshani 1992, 72. He goes on to say: ‘It is also elastic, and therefore more suitable for carving than those from eastern and southern Africa (where the bush or savannah subspecies, Loxodonta africana africana, is more prevalent) and those from Asia (where the Asian elephant, Elephas maximus, prevails). Some carvers also claim that a cow’s ivory is superior to that of a bull as it has a closer grain’.

105

Vallejo 2013, 113–4: ‘These workshops not only manufactured the so-called sumptuary objects used in the court … but also produced the different elements of the architectural decoration’.

106

See Ritchie 1969 and 1972.

107

Nees 2006.

108

Blair 2005, 83, n. 51, who corrects the reading given in the Al-Andalus catalogue; Martínez Nuñez 1995, 141; Labarta 2015. Badr’s name occurs in both the Cordoba Mosque and Hall of ʿAbd al-Raḥmān III, while Ṭarīf occurs in the Hall only. On Ṭarīf, see Souto 2005 and Souto 2010a, 2.83. Note that Labarta 2015, 15–16, believes the name to read Ẓarīf. On Badr, Souto 2010a, 2.15.

109

Souto 2007, 2.8, 112–3, citing Cano Piedra 1996, fig. 64: SA/456 (Bāshir), SA/361, 367, 402 (Mubārak). Cano Piedra attributes these to an earlier period than the ʿĀmirid, but the chronology of Andalusi green-and-brown ceramics is still under-studied, and these signatures may well be an argument for assigning these bowls to a later date.

110

Martínez Nuñez 1995, 143; Souto 2001, 293–294.

111

Acién Almansa 2001, 507.

112

Pellat 1961, 48–9, 62–3, 90–1, 102–5, 132–3, 144–5; discussed Manzano 2006, 444–5.

113

‘Quarrying the Roman City: an Islamic village at Utica (Tunisia)’, paper presented by Elizabeth Fentress and Corisande Fenwick, at the 5th annual Islamic Archaeology Day at the UCL Institute of Archaeology on 2 February 2019. They also noted that, at Carthage, no capitals, columns or bases have been found in situ, indicating wholesale spoliation at an early date.

114

Vallejo 2010, 116.

115

Vallejo 2010, 489. Marble, purple limestone and, to a lesser extent, alabaster, were introduced in the 950s especially for paving; Vallejo also notes the sumptuous decorative programme in marble that was introduced to the baths attached to the ‘Vivienda de la Alberca’ at this time.

116

Vallejo 2007, 4–5; Vallejo 2010, 103–111.

117

Vallejo 2010, 112–114.

118

Vallejo 2010, 115–117.

119

Vallejo 2010, 236-; Vallejo 2013, 119. This is in addition to various Roman portrait busts and other elements. He argues that it indicates cultural superiority of Islam over these civilisations. On the other hand, Calvo 2012 and 2014 has argued on the basis of the locations where these sarcophagi were found that these spaces were at the heart of the palace’s intellectual life.

120

For examples of plain-walled sarcophagi recovered at Madīnat al-Zahrāʾ, see Vallejo 2010, 238–40, pl. 189. He says that the basin described by Morales at San Jerónimo (see next note) ‘anticipates the ʿĀmirid basins’.

121

Vallejo 2010, 240; Morales 2012, 234 (fol. 116v): ‘Hanse hallado tambien en Cordova la vieja muchas antiguallas, de diversas maneras en diversas tiempos. Destas son la rica pila de marmol blanco de {quasi} dos varas en largo y mas de una en alto y otra en ancho, que sirve agora de Fuente en el {ins} monasterio de san Geronimo, en el claustro principal. Hallaronse dentro desta pila un ciervo y una cierva de laton {poco} ricamente labrados, poco menores que un cabrito. El ciervo echa el agua en la pila, y la cierva esta en el sumptuosissimo monesterio de n(uest)ra Señora de Guadalupe, en la fuente, que esta delante el refitorio’.

122

Boele 2005, 63; Rosser-Owen 2014, 193–7; Maroc Médiéval, 509, cat. 308. I am grateful to Péter Nagy for sharing his theory that the tombstone was likely commissioned by Abū Yaʿqūb Yūsuf’s successor, Abū Thābit (r. 1307–1308), when he had his predecessor’s corpse transported from al-Manṣūra (Algeria), where he died laying siege to the ʿAbd al-Wādids, and reburied in Shāla. Amid the turmoil of a civil war caused by his contested claim to the Marinid succession, Abū Thābit might have used whatever piece of marble he could get hold of (personal communication, 09/07/20).

123

Souto 2005. Montejo Córdoba 2006, 252–254, warns that the location of its find in the nineteenth century is not sufficiently well-recorded for scholars to attribute this panel to the Alcázar.

124

As depicted, for example, on the sarcophagus of the child Pomponia Agrippina, in the Museo Arqueológico Nacional, Madrid, inv. 1999/99/183.

125

Ibn Ḥayyān 1981, 264–265 (§§238–239). This translation from the Arabic by Stuart Sears is published in Anderson and Rosser-Owen 2015, 40.

126

Ghāliya was a luxury perfume made from musk and ambergris mixed with ben oil: see King 2008, esp. 181.

127

This may imply that Baghdad was the immediate source for this particular rosewater. While Iraq was clearly a major exporter of rosewater, it was not an important production centre; the best roses for distillation were considered to be either the red roses of Damascus or those from Fars in southern Iran, and in the tenth century the caliphs of Baghdad received 30,000 flasks of rosewater from Fars annually. See van Gelder 2006, and discussion in Anderson and Rosser-Owen 2015, 41.

128

Anderson and Rosser-Owen 2015, 40.

129

Canto García 2004.

130

See, for example, Coutu et al 2016; Dueppen and Gokee 2014; Guérin 2017.

131

Blair 2005, 89.

132

Two massive tusks in the Natural History Museum in London, on display in their Mammals gallery, are believed to be the heaviest ever recorded. They are said to have come from a bull elephant (Loxodonta africana) killed near Mount Kilimanjaro in East Africa, by an Arab hunter who had been trailing it for several weeks. They were sold in Zanzibar off the Tanzanian coast in 1898. The heavier of the two (on the left), 3.11 metres long and weighing 94 kg, was later purchased by the Museum for £350. The other tusk, 3.18 metres long and weighing 89 kg, was sold to the Museum in 1933 for an unspecified amount. Information taken from the gallery label.

133

As Sarah Guérin has argued was the case for the ensemble of the ‘Salerno ivories’, see Guérin 2016.

134

Shalem 1995; Makariou 1999.

135

Anderson 2014.

136

Cutler 2005.

137

Cutler 2005, 44–45.

138

Makariou 1999, 133.

139

A phrase I owe to Robert Hillenbrand. For a fuller discussion and contextualisation of luxury perfumes and luxury crafts, in specific reference to ivory, see Anderson and Rosser-Owen 2015, 40–45.

140

‘Kāfūr’, Encyclopaedia of Islam, 2nd edition; Grami 2013.

141

‘Misk’, Encyclopaedia of Islam, 2nd edition; King 2008; King 2011.

142

‘ʿAnbar’, Encyclopaedia of Islam, 2nd edition; Vallvé 1980, 219; Devisse 1988, 421–2; Sanagustin 1998, 189–202, n. 21.

143

Ballestín 2008, 146–7.

144

Hamarneh 1962, 62, n. 38.

145

For al-Zahrāwī’s description see Hamarneh 1965, 313.

146

Cressier 2000.

147

Anderson and Rosser-Owen 2015, 42–43.

148

Prado-Vilar 1997, 34.

149

Rosser-Owen 2012, 303. This means that its current mounts must have been added later, perhaps by Don Mendo when he commissioned the chalice and paten, or at a later date in the object’s life in the Braga Cathedral Treasury.

150

On this small group of metal objects, including these observations on the San Pelayo reliquary, see Rosser- Owen 2015a, esp. 53–55; Martin and Rosser-Owen forthcoming.

151

Azuar Ruiz 2018, 281.

152

Museu d’Art, Girona, inv. MD25, published in Giralt 1998, 207. My thanks to Therese Martin for sharing with me her photographs of this object.

153

Azuar Ruiz 2018, 282.

154

Contadini 2018.

155

Labarta 2019b.

156

Ibn al-Khaṭīb 1956, 74; Vallvé 1980, 213.

157

Ibn al-Khaṭīb 1956, 101–2; Vallvé 1980, 214; Labarta 2016, 266.

158

Vallejo 2018, 263.

159

Labarta 2016.

160

Ballestín 2006.

161

Vallvé 1980, 211. Ibn Ghālib says that there were only three silver lamps.

162

Vallvé 1980, 216, citing Ibn Ghālib, Ibn ʿIdhārī and al-Maqqarī.

163

Maroc Médiéval, cat. 190: the ‘Grand Lustre’ is datable 1202–13. As I have argued in Rosser-Owen 2014, the Maghribi dynasties who ruled al-Andalus consciously evoked the forms and objects of the Umayyad past to legitimise their role as successors to this territory.

164

Vallejo 2018, 259. This also evokes Moroccan examples that may continue this tradition, for example, the bell converted into a lamp in 1333–7, which now hangs in al-Qarawiyyīn in Fez: Maroc Médiéval, cat. 276.

165

Vallejo 2018, 263.

166

Pedro Marfil, personal communication, November 2009.

167

Heidenreich 2001; Heidenreich 2007.

168

On this well-known phenomenon, see for example Hallett 2012.

169

Bongianino 2017, 29–30. It is written in ‘composite variant of eastern Abbasid bookhands’.

170

Bongianino 2017, 37, citing Ribera y Tarragó 1928, 195.

171

Bongianino 2017, 37, citing Ribera y Tarragó 1928, 204.

172

Listed and discussed in Bongianino 2017, 41–50.

173

Echevarría 2011, 192.

174

Echevarría 2011, 16–17; María Luisa Ávila, Prosopografia de los Ulemas de al Andalus: http://www.eea.csic.es/pua/personaje/consulta_personaje.php?id=9354 (consulted 19/06/20); also Ávila’s “La sociedad hispanomusulmana al final del califato”, no. 665. With deep thanks to Umberto Bongianino for tracking down the information from the biographical dictionaries, and even sending me the Arabic text of Ibn al-Abbār’s al-Takmila li-Kitāb al-Ṣila, where Muḥammad al-Lughawī’s career is described at no. 417.

175

Vallvé 1980, 237.

176

Ibn Ḥayyān 1979, 177–179; 1981, 203–5.

177

Ibn Ḥayyān 1981, 264–265 (§§238–239), following the translation by Stuart Sears prepared for Anderson and Rosser-Owen 2015.

178

Discussed in Ballestín 2004a, 122–126. According to his definition (n. 77), iṣtināʿ, from /n/ʿ, is a term which means ‘to fabricate, create, construct, i.e. to “win the people contest” by means of gifts, goods, seeking partisans who will associate with someone because of their gratitude at the gifts received’.

179

Martínez Enamorado 2006, 20.

180

Martínez Enamorado 2006, 17. Even Lévi-Provençal confused these two Jaʿfars, a confusion which seems to have originated with the historian Ibn Khaldūn (n. 28).

181

Anales, §36.

182

Ocaña 1976, recently updated and summarised by Martínez Enamorado 2006, 16 ff. A full list of all the inscriptions naming Jaʿfar – though not including the David Collection piece, discussed below – is provided in tabulated form at pp. 66–7.

183

Martinez Enamorado 2006, 53; Meouak 1999, 215.

184

Vallejo 2007; Vallejo 2010; Vallejo and Montilla 2019.

185

Martínez Enamorado 2006, 51, citing Mora Vicente 2002, 180. Martínez Núñez 2015, 43 #33, 68, publishes a rare inscribed cyma from the reign of ʿAbd al-Raḥmān III (datable c. 345H). On p. 66 she observes that otherwise all the known cymas with inscriptions belong to al-Ḥakam’s reign.

186

Martínez Núñez 2015, 66 n. 33.

187

See the discussion on ‘decoration’ in Anderson 2013, 72–87.

188

The complete inscription is reconstructed by Martínez Enamorado 2006, 10, as follows: bismillāh al-raḥmān al-raḥīm barakah min allāh li-abī a[ḥmad jaʿfar] / [a]l-ḥājib ṣāḥib al-[khayl wa’l-ṭirāz ḥājib amīr al-muʾminin aṭāla allāh baqāʾahu bi-ʿawn allāh fa-tamma allāh wa-niʿmatu-hu fi sana tha]lātha wa khamsīn wa thalath miʾa (353/19 January 964–7 January 965). The missing section, including the reconstruction of the date, is extrapolated from the wording of the two surviving capitals also naming Abū Aḥmad Jaʿfar (discussed next): Barceló and Cantero 1995 reconstructed this as 353/964–5, though Barceló et al are more cautious in their 1998 publication, 29–31, dating the volute in Valencia broadly to the 960s–970s. Martínez Núñez commented (personal communication, 08/02/15) that the date ‘could be 356 H, since scarcely a fragment of the final grapheme of the unit [i.e. the ‘6’ in this case] has been preserved’. In a publication later that same year, she described it as ‘un ejemplar muy anómalo’, which she includes – in her list of known cymas with inscriptions – ‘con muchas reservas’, though without explaining what her reservations are (Martínez Núñez 2015, 66 n. 33). Carmen Barceló has also expressed doubts over the authenticity of this object: ‘Si comparamos los signos y nexos del cimacio con los de los capiteles que he atribuido al fatà Ŷaʿfar y otros cordobeses del mismo periodo y personaje, sus diferencias son bien notables y discordantes’ (Barceló 2015, 194–195). In a personal communication (29/07/20), she writes ‘Mi sospecha de que sea obra realizada ad hoc para el mercado de antigüedades se basa en el análisis de su inscripción que, además de anomalías en la redacción del texto, muestra muchas torpezas e incoherencias de carácter cronológico y estilístico en sus signos epigráficos’. If the cyma should be associated with al-Muṣḥafī rather than al-Ṣiqlābī, as discussed below, this might explain some of the stylistic differences she perceives from the capitals inscribed in al-Ṣiqlābī’s name.

189

See the table of his titles in Martínez Enamorado 2006, 66–7.

190

Martínez Enamorado 2006, 30, 48–52; Martínez Núñez 1999; Barceló, Cressier and Lerma 1988–90, 29–31, no. 2.1, fig. 5, plates III; Barceló, Cressier and Lerma 1990; Barceló and Cantero 1995; Barceló 1998, 225–6 (A1).

191

Barcelo and Cantero 1995, 424–5.

192

Martínez Núñez 1999, 85.

193

Meouak 1999, 187; de Felipe 1997, 179.

194

Al-Maqqarī, 183; Nafḥ, I, 471. On the Munyat al- Muṣḥafiyya, probably built between 961 and 979, see Ruggles 2000, 118.

195

Pérès 1990, 138; see also de Felipe 1997, 182.

196

Bayān III:42–43 [translation, 47–8].

197

Lévi-Provençal 1957 [1996], 247, n. 125: ‘una Munyat Chaʿfar (es decir, Chaʿfar el Eslavo), residencia de Hishām II’; Barcelo and Cantero 1995, 429. Martínez Enamorado 2006, 30, 48–52, follows them in this, though he notes ‘but this is far from being conclusive’.

198

Pérès 1990, 136–7, mentions that the late eleventh-century poet Ibn Zaydūn gives the names of various munyas in a poem (muwashshah) which he does not reproduce; these include al-Jaʿfariyya, but nothing further is known about this munya, such as whether it could be associated with one of the two ḥujjāb of that name, nor about most of the others named by Ibn Zaydūn. Pérès comments that apart from the first two (al-Ruṣāfa and al-ʿAqīq), Ibn Zaydūn is the only one to mention them: ‘Is this because he invented them? It seems more logical to suppose that all these places had existed in the period when Ibn Zaydūn went to them with Wallāda, but after the end of the eleventh century, they had ceased to attract the young desocupados of Cordoba’.

199

Martínez Núñez 1999, 94, n. 38.

200

Meouak 1999, 189.

201

Zanón 1989, 99, cited in de Felipe 1997, 185, 329: ‘La importancia adquirida por el linaje fue tal que su nombre pervivió en la topografía de Cordoba, donde hacia lugares ya denominación hacia referencia a ellos. Una muestra de ello lo constituye la mezquita llamada Masjid al-Mushafi situada dentro de la ciudad; aunque no existe ningun dato que prueba con certeza que esta denominación procede de un miembro de la familia, consideramos que queda dentro del ámbito de lo posible’.

202

David Collection, inv. 25.2009, H: 75 cm. I believe this object remains unpublished, although it is available on the David Collection’s website. My thanks to Kjeld von Folsach, Joachim Meyer and Will Kwiatkowski for sharing with me images and their reading of the inscription.

203

This identification was provided in a report written in January 2015 on the object’s inscription. I am extremely grateful to Prof. Martínez Núñez for taking the trouble to reread this inscription and share her thoughts with me. On examples of pipe covers found during excavations at Madīnat al-Zahrāʾ, which do not have carved decoration, see Vallejo 2010, 239–240, figs. 191–2 (for a reconstruction of how they would have worked to supply water to one of the plain sarcophagi also found at the site).

204

Meouak 1999, 186.

205

Arnold 2009–10; Arnold 2017, 101–2.

206

Ventura et al 2000, 340.

207

Arnold 2009–10, 270.

208

Ventura et al 2000, 341.

209

Salinas 2008, 248.

210

Castejón 1945, 208, noted ‘a most curious detail in these birds’, which is that each of their eyes contained what he considered to be a metallic incrustation, suggesting that they were originally inlaid with a piece of metal or jewel. Felix Arnold confirmed (personal communication 29/07/20) that they did not analyse this substance during their recent work on al-Rummāniyya.

211

Arnold, Canto García and Vallejo Triano 2015, 132–3, believe this to be the fragment of a basin, as reconstructed in their Fig. 63b; while Martínez Núñez 2015, 66 n. 33, intriguingly suggests that it might have come from another cyma. For the finds of al-Rummāniyya’s architectural decoration, see (in chronological order of publication) Velázquez Bosco 1912, 31–2, plates 10, 35; De los Santos Jener 1926; Anales de la Comisión Provincial de Monumentos Históricos y Artísticos de Cordoba 1926, 19–21; Terrasse 1932, 166 n.; Castejón 1945, 203–209; Castejón 1949, 235–236; Castejón 1954, 155; Gómez-Moreno 1951, 181, figs. 245, 252; Anderson 2013, 72–87. For publications of the new archaeology at the site, see Arnold, Canto García and Vallejo Triano 2008, 2015 and 2019.

212

Such as Art of Medieval Spain cats. 111, 114.

213

Anderson 2013, 87.

214

Montejo 2012.

215

Vallejo 2013, 116.

216

Watson 2005.

217

Barceló 2014, 129–30.

218

Watson, 2005, 169.

219

See the comparative discussion of these anonymous inscriptions in Watson 2005, 169–70.

220

Contadini 2017 revisits this issue, though only mentions the Andalusi ivories in passing (p. 438).

221

Watson 2005, 171.

222

Bayān II:295 [translation, 457].

223

Ballestín 2004a, 64.

224

Manzano 2006, 481.

225

Prado-Vilar 2005, 163, n. 88.

226

Prado-Vilar 2005, 153; 1997, 29.

227

Prado-Vilar 1997, 30.

228

Serjeant 1951, 33, after al-Maqqarī (Analectes, I: 271); see also Bayān II:319 [translation, 495]; Dozy 1913, 519–520; and Holod 1992, 44.

229

Ali-de-Unzaga 2012a, 568. Many Andalusi textiles remain in this ecclesiastical collection to this day.

230

Ballestín 2006.

231

Preserved in the Real Academia de la Historia, Madrid, inv. 298. For the bibliography on this textile, in chronological order of publication: Fernández y González 1875; Artiñano 1917, 10, cat. 43, plate 1; Gómez-Moreno 1919, 395; Lévi-Provençal 1931, 192, #211; Gómez-Moreno 1951, 345–347, fig. 404a; Serjeant 1951, 33; Bernis 1954, 198, plates 2, 4–6; May 1957, 14–17, figs. 3–5; García Gómez 1970, 44; Partearroyo 1982, 353; Al-Andalus 225–226, cat. 21; Partearroyo 1996, 62; Pérez Higuera 1994, 42, 86–87; Woolley 1995, 68–69, fig. 2; Santiago-Al-Andalus 370–1; Valdés Fernández 2001; Mackie 2015, 173–5 (5.5).

232

Valdés Fernández 2001, 386, believes this was an item of clothing owned by the caliph, i.e. part of his turban.

233

I have followed the spelling and formatting of the recent edition of The Poem of Fernán González, in Such and Rabone 2015, 174. The translation that follows is also taken from here (p. 175). Ruiz Souza 2001a, 31, also cites this poem.

(276) “With al-Mansur now a good distance away,
The field was left well peopled with Christians;
They gathered their possessions, granted them by God,
And found such great wealth as to be beyond tally.

(277) They found in the tents an abundant treasure:
Many cups and goblets made of fine gold;
Such riches as no Christian or Moor had ever seen
– it would have sufficed Alexander and Porus.

(278) Many cases they found there, along with many bags,
Filled with gold and silver – no sign of copper coins –,
Many silken pavilions and many tents of war,
Breastplates and swords and a great mass of armour.

(279) They found caskets of ivory very great in value,
With other noble objects impossible to count;
To San Pedro de Arlanza were most of these given,
Where to this day they are displayed upon its altar.

(280) From this great hoard of wealth, they took as they desired;
Two thirds and more remained they could not carry;
But the arms they had found they wished not to leave behind.
They came to San Pedro, bringing all of their gains.”

234

Ferrandis 1935, 51–2, cat. 1; 88–91 cat. 25.

235

Ferrandis 1928, 59; Ferrandis 1935, 21; Kühnel 1971, 5; Shalem 1995, 24 and n.16; Makariou 2001, 58–59.

236

Prado-Vilar 1997; Ruiz Souza 2001a; Anderson 2014; Rosser-Owen 2015a.

237

Ballestín 2004a, 54–8, citing Ibn Ḥayyān, Muqtabis, 123; Anales Palatinos, 156.

238

Ballestín 2004a, 159.

239

In his writings on the Fermo chasuble (Chapter 7: 2.4), in which he argues that the textile was originally a tent, Shalem uses the term ‘soft architecture’: see Shalem 2014a and 2017 (Introduction).

240

Anales, §119. On the Arabic terminology used by al-Rāzī for campaign tents, see García Gómez 1967, 169–170.

241

Ali-de-Unzaga 2012a, 567, citing Ibn Ḥayyān 1981, 257.

242

Prado-Vilar 2005, 156–7.

  • Collapse
  • Expand