An Arab room with a view: resource extraction and transcultural exchange

Cardiff Castle is a site with multiple collaborations. The 3rd Marquess of Bute commissioned Burges to develop architectural and interior designs to renovate the castle, and this project led to the co-development of many properties across Bute’s portfolio. Prior to the building work at the castle in the 1870s, Burges prepared a report of the castle’s existing elevations and historical contexts for the trustees of Bute’s estate in 1866.¹⁷ Bute was still under the age of 21 and his estate was entrusted to a board upon his father’s early death. The report contained imaginings of the castle’s Roman, medieval, Celtic, and Norman past lives and anticipation for its future life embodying these pasts. Of particular interest is the frontispiece that Burges commissioned Edward Sherratt Cole to produce to visualize Cardiff Castle’s past and future (not pictured here).

Throughout Burges’s designs and report, he articulated the medieval castle and its society to be located within a cosmopolitan Cardiff society engaged in transcultural exchange.¹⁸ The frontispiece is a watercolour of the exterior of the castle from street level. In the illustration a man in armour walks two chain-mail covered horses, another man leans on a jousting lance as he speaks to a monk with a tonsure, and another stands wearing a blue turban. These figures signal the courtly and knightly culture of medieval Catholic Wales, with pilgrims and militaristic crusaders who travelled to, and warred in, Palestine. The presence of militarism and/or police is never neutral, but the seeming antagonistic divide between East and West articulated in later medieval and crusading sources is challenged in this watercolour.¹⁹ The knight with the lance leans on it casually as he speaks to both the monk and the man wearing a turban, suggesting that tensions do not, at least, prevent amicable relationships, including trade. The men stand beside barrels, possibly of alcohol or other goods, indicating local or global trade.²⁰

Burges contextualized Cardiff Castle through this visualization of the past with what would become his additions — including the Clock and Herbert towers, animal wall, and gardens — and transtemporal, cosmopolitan viewers of medieval and nineteenth-century Wales. The actualization of Burges’s designs culminated in exterior polychrome, particularly on those features facing street-level viewers, including the animal wall and the Clock Tower. Perhaps, like a crack in the seam which may come to reveal gold deposits (or, in Bute’s case, coal leading to gold bank deposits), these polychrome elements hint at the luscious interior decorations. The rest of the castle’s exterior, however — behind its grey stones and, at times, boxy towers — conceals the interior delights, enticing and tantalizing street-level viewers to wonder what lay within the exterior mix of polychromy and stone.

The Arab Room is located in the Herbert Tower, tucked behind the banqueting hall on the first floor. This means that unlike the Clock Tower, where Bute’s male-centred smoking rooms looked out into the public world of the city’s streets, the Arab Room faced the fenced-off gardens of Bute’s private estate. Designed by Burges for the 3rd Marquess’s wife, Lady Gwendolen Mary Anne Stuart, the Arab Room was, functionally, a tea room and occasional overnight guest room. There is a complex relationship in Burges’s designs at the castle regarding both looking in and looking out, and the power and directionality of the gazes from viewers with varied and intersecting identities. Described throughout the literature as a harem space because of its ‘Arab’ aesthetic and the fact that it was designed for Bute’s wife, the room, at first glance, seems to be solely about looking within and concealing external gazes.²¹

But like a geode, cracked open for its glittering crystal and mineral centre, the room dazzles and asks to be looked at. And, if I may lean on the geode-mineral simile for a moment more, it was the vast mineral wealth in coal on the Bute estates and its global transiting from the Bute docks that footed the bill for such a room. Significantly, similar to the trade figured in the frontispiece to Burges’s report, nineteenth-century Cardiff was a significant site of trade, immigration, and temporary workers from throughout the globe. The 2nd Marquess of Bute established the West Bute Dock which opened on 9 October 1839 and, under the board of trustees, the East Bute Dock opened in 1855. The two docks enabled the shipment of Welsh coal to France, ports in the Mediterranean, Baltic, and North seas, as well as to South America, coastal colonial Africa, North America, South and East Asia, and Australia.²²

Similar to Cole’s and Burges’s envisioning of medieval Wales, in the nineteenth century, shipment to, and from, this constellation of global ports facilitated the movement of peoples between any and all of these ports, especially Cardiff (i.e., 2<n<infinity). As I describe above, Europeans, Arabs, Africans, and East Asians temporarily and permanently settled in Cardiff around the port at Tiger Bay. In terms of Edwards’s theory of the cosmopolitan, then, Cardiff Castle and the surrounding area was a site of ‘multi-valuedness’ as multi-ethnic and polycultural communities traversed past, looked at, and worked at the castle. These transcultural networks of exchange are embedded in Burges’s designs, including the cosmopolitanism of the surrounding community and their viewership(s). Moreover, as I demonstrate below, the Cappella Palatina’s designs reflect an ideal of transcultural exchange to display the patron’s claims to imperial wealth and authority (within its specific contexts). In Bute’s case, it neatly and aesthetically avoided the culpability of empire’s crimes.²³

In Mordaunt Crook’s sweeping study of Burges’s designs under the mantle of High Victorian Gothic, he notes that the Arab Room’s designs are linked to the ‘shimmering vaults of Siculo-Norman Sicily — the Cappella Palatina; the Zisa, or House of Delights’, but, beyond this, he does not specify the design citations from these buildings in Burges’s designs at Cardiff.²⁴ Burges travelled to Palermo in 1845.²⁵ To solidify the links between Burges’s designs and twelfth-century Sicily, I located watercolour drawings that Burges made of the mosaic designs from La Zisa, built between 1164 and c. 1189 during the reigns of William I and II (Burges’s drawings not pictured here).²⁶ Burges used watercolour to recreate the mosaic pattern of the peacock as well as some of the border decorations of the fountain at La Zisa (Fig. 2). These tracings demonstrate his awareness of the design elements of Norman Sicily in their multiple scales — from mosaic tiles at La Zisa to architectural details of the Cappella. Additionally, in Burges’s notebook, I located schematic drawings of the Cappella Palatina which outline the architectural qualities of the nave and its unique use of muqarnas, an Islamic architectural style, in a Christian space of worship. I attribute these drawings to Axel Herman Hägg because, in the bottom right-hand corner, Hägg signed authorship with ‘A.H.H., 1875’ (Fig. 3).

Fig. 2: Once-known artists, Mosaic under muqarnas at La Zisa, Palermo (c. 1165–89). Photograph by Magnus Manske.

Fig. 3: Axel Herman Hägg, Schematics of the Cappella Palatina in Palazzo dei Normanni, Palermo, Sicily (1875), in William Burges, ‘Prints, Drawings & Paintings Collection’. Victoria and Albert Museum, London.

My attribution of these schematics to Hägg stems from his and Burges’s long-time collaboration and partnership. Born in Sweden, Hägg studied shipbuilding in the government dockyard at Karlskrona, before relocating in 1856 to Scotland to work for Lawrence Hill & Co.²⁷ From the draughting skills he learnt from shipbuilding, he was commissioned to design a house for his employer, which led to a connection via Hill to the architectural office of Ewan Christian, the architect of the Ecclesiastical Commissioners, through which Hägg met Burges.²⁸ During their many years of collaboration and friendship, Hägg visualized Burges’s designs for patrons or jury panels at design competitions. In particular, Hägg made architectural drawings of Burges’s designs at Cardiff Castle which were displayed at the Royal Academy in 1870.²⁹ Around this time, Burges suggested that Hägg travel to Palermo to see and sketch the architecture and environs of Monreale and the Cappella Palatina, the royal chapel contained within the complex of the Palazzo dei Normanni, or the Norman Palace.³⁰

Hägg’s schematic records of the Cappella Palatina are a significant example of nineteenth-century architectural and design interest in the Cappella. In the 1840s Charles Joseph Hullmandel produced chromolithographs based on George Belton Moore’s drawings which were printed in Henry Gally Knight’s Saracenic and Norman Remains (1840) (Fig. 4). Although Hullmandel’s chromolithograph of the Cappella nave represents the space and its imagined dynamic use in the twelfth century, the image distorts the nave’s architectural reality in order to survey the composite designs. The muqarnas ceiling of the nave is recorded, but not so that it may be replicated architecturally or even at the level of design — some of the formal shapes are recorded, especially the eight-pointed stars containing eight-lobed flowers, but replicating the precise geometry and intricate design schemes are not the point of the image.

Fig. 4: Charles Joseph Hullmandel after George Moore, Cappella Palatina in Palazzo dei Normanni, in Henry Gally Knight, Saracenic and Norman Remains: To Illustrate the Normans in Sicily (Murray, 1840). Via Getty Research Institute.

During the 1870s and 1880s in Palermo, extensive research was put into producing Amari and others’ La cappella di S. Pietro nella reggia di Palermo with chromolithographs by Andrea Terzi.³¹ These chromolithographs are meticulous recreations of the architectural plans and design elements. The publication includes plans of the muqarnas in the nave ceiling as well as more detailed design information, such as examples of the Kufic script which decorates the eight-pointed stars of the muqarnas. It is likely, given the dating, that Hägg met, worked alongside, and even benefited from Amari and Terzi at the Cappella in 1875 as he recorded the schematic details and design facets. (The 1889 edition of Amari’s La cappella, however, was published after the completion of the Arab Room at Cardiff Castle and Burges’s death.) Notably, the replication of the Cappella’s employment of Islamic-world designs at Cardiff demonstrates that Burges’s designs were part of contemporary, cutting-edge research into Norman Sicily and its historical cosmopolitan, polycultural art production.

Placing the component parts from Hägg’s schematic drawings side by side with photographs of the Arab Room and the Cappella Palatina (Figs. 5, 6; see also Figs. 1, 3), it is evident that Burges’s designs are based on the schematics and designs of the Cappella, as well as the polycultural style of twelfth-century Palermo and Sicily. Looking directly up towards the pinnacle of the ceiling of the Arab Room reveals that Burges recreated one unit of the muqarnas — a central octagonal dome surrounded by four smaller domes within a squared niche — from the Cappella. Figure 5 is a photograph from the Arab Room at Cardiff Castle that shows a cross-section of the muqarnas, and Figure 6 is an image of a cross-section from the Cappella. At Cardiff, the cells are minimally modified from the original source material, and the component parts directly cite the Cappella muqarnas. At Cardiff the cells are streamlined and the pointed shape of the arches is emphasized through the use of a gold-band border. Burges eschewed the use of human figures and Kufic script as decoration within the cells, and he substituted these with painted foliage designs, multifoil pointed arches, and animals. Where Burges used one unit of the muqarnas design, the Cappella contains twenty repeated muqarnas along the nave in two rows of ten. This pattern marked a transition in the mid-twelfth century to horizontal muqarnas from the pine-cone shaped muqarnas in which the vault of a single dome is composed of niche-like cells and stalactites.

Fig. 5: William Burges, Cross-section of muqarnas in the Arab Room, Cardiff Castle (c. 1866–81). Photograph by author.

Fig. 6: Once-known artists, Central nave detail of muqarnas, Cappella Palatina in Palazzo dei Normanni, Palermo, Sicily (c. 1132–40). Photograph by Ghio Roli. Bridgeman Images.

In Owen Jones and Jules Goury’s monumental study of the Alhambra Palace complex in Granada, Spain (1842–45), they noted that muqarnas are an architectural feature born from, and which transversed, dār al-islām, or the Islamic world.³² Muqarnas are a three-dimensional vault structure which, like a squinch and spandrel, enables a structure to move from a rectangular shape to a dome. The complex patterns are based on two-dimensional geometric designs that are tessellated and stacked in space to produce a three-dimensional honeycombed vaulting effect that creates negative space that is similar in shape to a pine cone.³³ At the Cappella, the nave ceiling used the vertical principle of the pine-cone muqarnas and translated it to a horizontal barrel vault. Contemporary examples of this are rare, with the closest example found in Almoravid Fez at the al-Qarawiyyin Mosque (c. 1132 and 1142–43).³⁴

Vicenza Garofalo writes that the muqarnas in Norman Palermo are based on orthogonal geometry, which means that ‘through the rotation of squares they create all sorts of different shapes such as polygons, stars, and flowers […] [and are] composed of figures obtained by the rotation of polygons within circles’ (p. 372). For instance, Hägg’s schematics subdivide the layers of the muqarnas at the Cappella in a cross section of one dome which shows the vertical surfaces, with tiers labelled B, C, D, and E (see top right, Fig. 3). The schematics also show the horizontal surfaces of the respective layers and the layers’ orthogonal geometry. These culminate in the shapes which form the horizontal structural elements and some design features — Kufic script, eight-pointed stars, eight-lobed flower petals, and painted designs — of the dome seen from directly below (see bottom half of Fig. 3).

At Cardiff Castle, within the space allowed by the Herbert Tower, Burges translated the horizontal muqarnas system at the Cappella to a vertical pine-cone style. Otherwise, the structural elements are replicated from the Cappella to the Arab Room. The vertical and horizontal surfaces (as we see in Fig. 3, B, C, D, and E) are replicated based on the subdivisions outlined in Hägg’s drawings of the Cappella, culminating in the geometrical design of the dome. At Cardiff Castle the rounded dome shape is contained within a square base structure; these designs are replicated from Hägg’s schematics. In the Arab Room there are a total of five domes, one large central dome surrounded by four smaller ones (one quarter of the larger dome), each of which contains an eight-pointed star. The stars of the external, smaller domes are decorative and painted; whereas, in the central dome, the eight-pointed star is part of the structural design of the muqarnas itself. The angled spandrel-shape of the tier (Fig. 3, E) that meets the cupola of the dome creates the first outline of the star. There are then two arabesque and braided-band borders of the star, which culminate in a final star that contains an eight-lobed flower. The exterior arabesque border replaces the Kufic script that is present in the Cappella muqarnas; the interior braided border is replicated exactly, as we can see in the painted ceiling at the Cappella (Fig. 7).

Fig. 7: Once-known artists, Cappella Palatina in Palazzo dei Normanni, Palermo, Sicily (c. 1132–40). Photograph by Ghio Roli. Bridgeman Images.

Moreover, elements of the designs in the ceiling at the Cappella are carried through to the floor in the Arab Room, such as the replication of octagons and thick pointed bands (in yellow and white) in the interior pattern of the floor design (compare bottom left of Fig. 3 and Fig. 7). The geometry of the design elements at the Arab Room is meticulous, as are design principles from dār al-islām. The floor designs are based on an understanding of muqarnas principles which use a two-dimensional geometric pattern and stack it into three-dimensional space as described by Jones and Goury.³⁵ For instance, the floor design recreates and scales down the three-dimensional octagonal elements of Plan E in two dimensions and rotates them at a 22.5° angle (see bottom right-hand corner of Fig. 3 and Fig. 1). The rotation, interestingly, enables the eight-point star design in the middle to transition from a star with 90° angles to one with points that have 45° angles.

Visiting the Cappella through polycultural Norman and Victorian eyes

The Cappella Palatina is a marvel in design; it contains Byzantine mosaics, opus sectile designs on the floors and walls, and the painted muqarnas ceilings. Much scholarly attention has been paid to the Cappella, as well as to determining source material, architectural and design elements, and how the chapel signified to Roger II or Sicilian audiences more broadly.³⁶ Rosa Di Liberto describes the chapel, situated within the palace complex, as the manifesto of Roger II’s court and ‘one of the most complete artistic expressions of power’ by the Hauteville Normans (p. 141). Similarly, Jeremy Johns suggests that the use of various styles at the Cappella evinced the ‘range and universality’ of Norman sovereignty under Roger II and signalled the expansion of his empire.³⁷ Annliese Nef further characterizes Roger II’s rule as a period in which ‘different elements were intermingled’ in court life.³⁸

The Norman brothers Robert Guiscard and Count Roger I, Roger II’s father, worked together to conquer southern Italy and Sicily. Both brothers cooperated with Muslim allies in the area and, in 1060, Count Roger made a significant alliance with Muhammad b. Ibrahim Ibn al-Thumna, the Emir of Catania on the east coast of Sicily. Ibn al-Thumna was assassinated in 1062, which interrupted cooperation between Muslims loyal to him and Count Roger. Around this time, Zirid forces from Ifriqiya arrived in Sicily to challenge Roger’s incursion.³⁹ After 1062 Roger changed strategy and began incorporating Muslims into his own armies, with evidence in 1076 that Muslims fought for Roger in the siege of Salerno. By the 1090s Roger had thousands of Muslim soldiers in his army.⁴⁰

The wider dār al-islām of Sicily and the Mediterranean was composed of competing interests as well as diverse political and religious beliefs. Some Muslims were willing to cooperate with Roger. When he conquered the island, a peaceful treaty was negotiated with the Muslim population conducted by Roger and two qa’ids and elders from the city of Palermo.⁴¹ The treaty, as recounted by the contemporary monk and historian Goffredo Malaterra, outlined that the Normans would not interfere with existing Muslim legal systems. The treaty made use of the established jizya, a tax on non-Muslims on the island, and translated it into a tax on all populations except Christians.⁴² In addition to Muslim populations, a significant Greek Christian population had lived on the island continuously from before the 827 invasion by the Aghlabids of Ifriqiya.⁴³ Latin Christian populations grew, especially after the Norman rule over the island.

Jewish Sicilians were a smaller but notable presence on the island, and were recorded by the Jewish-Spanish traveller Benjamin of Tudela, who travelled to Sicily c. 1164.⁴⁴ From Johns’s important contributions to the field, we know that under Roger II the administration, diwan, of Sicily was conducted in three languages: Greek, Arabic, and Latin.⁴⁵ That said, these were far from the only languages on the island. In addition to Hebrew in religious ceremonies, Jewish Sicilians used Judeo-Arabic, which used Hebrew letters to convey the Arabic language.⁴⁶ These four scripts were memorialized on a quadrilingual tomb of a woman named Anne/Anna (d.1148). She was the mother of Grisandus, a cleric at the Cappella Palatina during Roger II’s lifetime.⁴⁷ The tomb is in a chapel dedicated to St Anna in the Church of St Michael in Palermo. Moreover, knowledge about this quadrilingual tomb would have been available to Burges and, especially, Hägg; a reproduction of the tomb was published by Amari in 1879.⁴⁸ Herbert Houben writes that this 1149 monument addressed multiple communities and viewers: Muslim, Jewish, Greek Christian, and Latin Christian Sicilians. It is possible that Anna was Jewish, or that the proximity of the tomb to a large Palmeritan Jewish community led to this design choice.⁴⁹ The act of communicating and addressing multiple viewers transculturally is significant. It suggests that Norman Palermo and its building campaign were engaged in polyvalent and, thus, ‘many-valued’ iterations of iconographies and meaning-making both at the court and towards the local community — similar to the ways in which the polyvalent designs at Cardiff Castle addressed multiple cosmopolitan viewers, as I discuss below.⁵⁰

The exact details of the once-known artists who designed, fabricated, installed, and painted the muqarnas of the Cappella are lost to history. Thus the origins of the designs have garnered much scholarly interest,⁵¹ most notably with Johns’s and Slobodan Ćurčić’s respective theses that the designs came from Fatimid Egypt. They suggest that these artists then went to work on the now lost Mouchroutas ceiling in Constantinople; however, the extant examples from Fatimid Egypt are few and partial.⁵² Scholars also link the Cappella’s designs to the Almoravid dynasty, particularly to the al-Qarawiyyin Mosque in Fez because of its concurrent construction to the Cappella as well as the unique use of muqarnas horizontally across a barrel vault rather than in the pine-cone style.⁵³ Michel Écochard compares the geometry of the Norman muqarnas at the Cappella and the later Zisa with those at the hospital of Nur al-Din in Damascus, which were based on Iraqi models from Baghdad.⁵⁴ In contrast to other scholars, who looked elsewhere for the Cappella’s muqarnas-builders, Tom Klobe argues that local artists and carpenters were the source of the muqarnas, based on the use of local building materials. Elsewhere, muqarnas are made from plaster or stone, but at the Cappella they were made from pine wood native to the Sicilian mountains. Klobe contends that this innovation, especially for such a large-scale project, suggests that Muslim builders in Palermo had used this technique previously.⁵⁵ Unfortunately, there are no surviving examples to compare for reference.

By contrast, Johns argues that there was a period of decline of artistic patronage by Sicily’s Muslim population from the time of the Norman conquest of Sicily under Roger I to the patronage of the Cappella under Roger II. From this, he infers that a team of artisans must have come from Fatimid Egypt.⁵⁶ Klobe contends that, although we do not have records of buildings built nor even surviving examples, during this period there was a thriving art community in Sicily, especially in the carving of rock crystal, porphyry, ivory, and wood.⁵⁷ There may have been a decline in the intervening years between Roger I and II, but local artists still produced and sold art locally and to merchants for sale abroad. It is also unlikely that Norman rulers would do anything to harm the wealth of their kingdom by destroying the production and trade of art and local commodities.

The polycultural designs in Norman Sicily articulate cosmopolitan exchange throughout, and beyond, Muslim, Christian, and Jewish worlds, that is, 2<n<infinity. The island of Sicily and its local populations prior to, and under, the Normans were both diverse and wealthy. Its geographical location in the Mediterranean made it a prime location for merchants and traders to travel between the Iberian Peninsula, North Africa, southern Europe, and the Levant.⁵⁸ North African merchants established trade routes throughout West Africa, through which Muslim West Africans travelled on pilgrimages to Mecca. Merchants traded salt, gold, and copper as well as enslaved West Africans.⁵⁹ These were a significant boon to the economy, and Sicilian artists and merchants sold textiles and ceramics to Ifriqiya.⁶⁰ Within these trade networks, Sicily was a major exporter of goods, including pine wood, animal skins and hides, hemp, linen, and woollen cloth.⁶¹ During the Muslim dynasties, they introduced silk manufacture and tiraz workshops. Tiraz is a Persian word that describes a particular textile and embroidery practice. They are armbands embroidered with Kufic Arabic inscriptions on robes. Originally, these were items produced solely for the sovereign, who gifted tiraz to members of his court to signify the receiver’s courtly status. Tiraz could also be found on items like turbans and mantles. A key example made in Sicily is Roger II’s red silk mantle, now in the Kaiserliche Schatzkammer in Vienna. The gold and silk thread embroidery includes Kufic script that commemorates Roger’s coronation. The sumptuous and costly art production in Norman Sicily attests to its wealth and that it participated in cosmopolitan, ‘many-valued’, transcultural exchanges of art and material culture across the Mediterranean and its cosmopolitan cultures.

The polycultural designs of the Latin Christian Cappella signified Roger II’s reign over polycultural Sicily and southern Italy, and evinced his access to, and control over, transcultural networks of exchange. This is emphasized by two further aspects of Roger’s patronage. First, Tronzo suggests that the paintings of the muqarnas ceiling contain Islamic natural world designs as well as people drinking, feasting, playing instruments, or dancing under what he characterizes as tents or tent-like designs indicating a kingdom in movement — despite the permanence of the chapel itself (see Fig. 6). The tent, he writes, was a dār al-islām symbol of the travelling ruler on military campaigns, wherein a ruler demonstrated control over his territories through his continual presence.⁶² Roger constructed his public image visually through his incorporation of Islamic world iconographies and designs to articulate the breadth of his rule. Movement and its depiction are tied to Roger’s kingdom building, as are the literal control of the multinode, or constellated, movement of art and commodities to and from the island. For instance, there are stylistic and iconographic similarities in the painted ceilings with contemporary Fatimid pottery, whose source material comes from Iran, once again suggesting another node on the cosmopolitan networks of exchange.⁶³ The very design of the muqarnas as multiple layers of interlocking geometric parts operates as a metaphor for the composition of Roger’s cosmopolitan (multiple and interlocking) kingdom and the iconographies his reign drew together — Latin Christian, Greek Christian, Jewish, and Muslim. There is, thus, literal and figural representations of movement (peoples and kingdoms) which signify across intersectional Sicilian identities and viewers as well as visiting dignitaries.

Second, Roger II signified his access to, and control over, transcultural networks of exchange when he commissioned the Arab geographer Ibn al-Idrīsi to produce the monolithic Kitāb nuzhat al-mushtāq fi ikhtirāq (The Book of Pleasure for Those Who Desire to Visit Far Away Places, c. 1100–66), also called Kitāb Rujār (the Book of Roger).⁶⁴ The atlas was recorded in Arabic, with an introduction and comprehensive descriptions by al-Idrīsi. In Katherine Jacka’s study of both the text and Roger’s commission, she characterizes his interest in the project as part of his reform of the Sicilian administration (pp. 60–61). In addition to this, I argue that it is a subsequent example of Roger’s interest in cosmopolitan networks of exchange throughout his and other kingdoms.

Records of the once-known artists that worked on the muqarnas at the Cappella are sadly lost. Their work enumerates a history of transcultural exchange — in the painted muqarnas ceilings as well as the Byzantine mosaics (completed at various stages) and the intricate opus sectile marble designs throughout. The elements that Burges recreated, muqarnas, opus sectile pavement, and stained glass have a shared materiality in that they are composed of interlocking parts in a network of designs. That is to say, differentiated composite parts work together to form a whole. For example, stained glass uses pieces of coloured glass in various shapes and sizes that are soldered together with strips of lead or joined with stone or plaster of Paris. These smaller differentiated pieces interlock to form a larger design. Mosaic and opus sectile operate in similar ways. Muqarnas use two-dimensional component parts of different shapes, sizes, and angles which stack and interlock to create a three-dimensional honeycombed dome. The very materiality of these elements is a metaphor that Roger II and Burges used to signify the differentiated and interlocking cosmopolitan communities and their aesthetics from which they both drew. The composite parts of the designs and the polycultural communities are part of a larger aesthetic network of art creation and transcultural exchange which occurs between cultures (n), where 2<n<infinity.

In this vein, Burges was well aware of the transcultural networks of exchange that occurred in Norman Palermo — which he also located in medieval Cardiff through his and Cole’s frontispiece. This understanding of transcultural exchange was likewise demonstrated in his 1865 lecture series Art Applied to Industry. ‘In the Middle Ages,’ he wrote, ‘the richer classes largely imported beautiful stuffs from the East, and afterwards from Sicily and Italy.’ Discussing embroidered clothing in particular, he argued that the designs originated from the Byzantine Empire and the ‘East generally’. The Norman conquest of Sicily meant that connections with Islamic world and Byzantine designs were strengthened. He suggested that Byzantine artists were imported to the island by Roger I and acknowledged that the island was still inhabited by thriving artistic Muslim communities. On this, he explained, ‘Of course many of the workers would be Mohammedans, and the old patterns, perhaps with the addition of sundry animals, would still continue in use; hence the frequency of Arabic inscriptions in the borders.’⁶⁵ The design choices, in Burges’s evaluation, were made by Muslim artists and makers, which demonstrates their agency over the art they produced in the Norman realm.

Burges described Norman Sicily’s aesthetic links to the wider Islamic world and its networks of trade. In keeping with his contemporaries, he described sites in dār al-islām as the ‘East’, differentiating it from Byzantium and Sicily (p. 85). It may be argued that Burges’s word choice betrays a homogenizing of the Islamic world that participates in a binary view of ‘East’ and ‘West’. In some respects, Burges’s designs contributed to such a binarizing view; however, he offered a more nuanced view in his lectures. Despite his broad use of the word East to signify dār al-islām, he traced artistic practices that emerged in Europe to cosmopolitan networks of exchange across multiple sites. For example, he traced the history of Venetian glass arts to the Byzantine Empire and the ‘East’. In this context, his usage of the word East refers to the wider Islamic world because he was describing stained-glass window techniques found in, for example, Palestine, Oman, Iran, and Egypt. This style of stained glass uses small pieces of thin glass joined by plaster of Paris to create what he described as a ‘shawl pattern’. It is the same technique he recreated in the stained-glass windows at Cardiff Castle based on designs copied from A. C. T. E. Prisse d’Avennes’s L’Art arabe d’après les monuments du Kaire, 7^e–18^e (1869–77).⁶⁶ He contrasted this technique with those from Europe which used larger panels of thicker glass. Burges explained that the Venetians were able to refine their techniques of glassmaking through their trade with the Islamic world and the Byzantine Empire (p. 17). In this case, he did not differentiate between the specific locations and was speaking broadly about the Islamic world. However, in his discussion of bronzewares, Burges dismissed Dinanderie, or bronzeware from Belgium, and described bronze vessels in thirteenth-century Mosul, in present-day Iraq, as the golden age of bronzeware production, particularly vessels inlaid with gold and silver designs and Kufic script (p. 49). Burges made the term The East do a lot of heavy lifting, but he also demonstrated an awareness of distinct cultural styles and aesthetics as well as European art’s interconnectedness with the art of dār al-islām.

Through this lens, in Sicily Burges saw transcultural exchanges between constellated cultures and the port city of Norman Palermo. The use of animal and foliage designs from the Cappella and Norman Palmeritan sources in the Arab Room was based on his study of these. In his lectures he referenced a cap made in Sicily, now at the British Museum, that was found in the tomb of the Holy Roman Emperor and King of Sicily Henry VI. He described the pattern, which ‘consists of antelopes and parrots placed face to face, the ground being filled up with some Arabic-looking foliage’ (p. 85). These design elements, which Burges linked to networks of artistic exchange throughout Europe and the wider Mediterranean, are carried through in his designs of the Arab Room, and even his interior designs writ large. Among the painted foliage designs in the Arab Room, parrots fly or perch. Indeed, in the central dome, eight parrots in flight decorate the eight-lobed flower. At the border between the walls and the muqarnas, eight wooden parrots are perched on bell-cap posts. Rabbits and other mammals are tangled in foliage throughout the designs. At the Cappella, we saw that Roger II signalled his increasing and watchful eye over his domain by using the visual language of the tent to signify the sovereign on the move throughout his empire. By contrast, Burges’s use of the parrots in flight throughout the designs of the Arab Room signifies differently. It suggests that, rather than the sovereign on the move, it is the community which is in flight and on the move alongside the movement and transiting of coal excavated from Welsh land.

Polycultural Tiger Bay and Norman Sicily

With the opening of the Bute docks after the 1830s, demographic changes in the community felt rapid because technology allowed for faster movement of people, resources, and commodities. Burges contextualized these demographic changes in nineteenth-century Cardiff within his citations of cosmopolitan, polycultural Norman Sicily. In doing so, he was reckoning with contemporary anxieties about the fears of a changing Britain. These fears were fanned because non-white people were made more visible and their presence in Britain was politicized by white supremacy and developing European constructions of racial hierarchies.⁶⁷ I am by no means suggesting that Burges was tolerant of non-white people or that he had a utopic vision of equality; like many Victorians, he had prejudiced beliefs. Based on his study of medieval history, however, he paired nineteenth-century Cardiff with twelfth-century Norman Sicily to draw upon his understanding of their comparable contexts as cosmopolitan, polycultural communities. Additionally, he drew upon Norman Sicily’s cosmopolitan aesthetics and design elements to celebrate his patron’s imperial position of power over Cardiff’s polycultural communities and sumptuous networks of aesthetic and economic exchange.

In doing so, Burges visually communicated Bute’s aristocratic position and his immense property portfolio and coal-based wealth through the visual language of Norman Sicily. These citations also declare Cardiff’s polycultural community within the designs themselves, sometimes obscuring the population of Tiger Bay’s particularities through a romanticized lens of an idealized medieval past. That said, the designs at the Cappella purposefully integrated its local, polycultural communities into its designs to signal the ruler’s power. In recreating these designs for the Marquess of Bute, Cardiff Castle is likewise read as integrating its local and cosmopolitan communities to signal the patron’s wealth and power within his own context of Cardiff’s differently ‘glocal’ community of Tiger Bay. Thus, the composite building blocks of the Arab Room’s designs — the room’s very materiality — continuously remind viewers of the necessary differences of the interconnecting components that create the whole. Norman Sicily, as recreated by Burges at Cardiff Castle, is an excellent example of Christopher Pinney’s assertion of a ‘Creole’ Europe, which describes how transcultural exchanges of material cultures reconfigure the transacting cultures. That is to say, Pinney would view cosmopolitan twelfth-century Norman Sicily and nineteenth-century Cardiff — in which multiple cultures and their aesthetics interact — as the rule and not the exception (p. 128).

The Cappella was a palace chapel that was used as an audience hall, especially the area under the muqarnas, for visiting dignitaries, with some public access. The cosmopolitan, polycultural communities and artists, as well as wider networks of transcultural exchange, are evoked visually throughout the building campaigns of Norman Palermo. Translated to Wales, the cosmopolitan, multi-ethnic community of Tiger Bay was comparably, but differently, manifested in the designs of Burges’s Arab Room and the castle more broadly. Similar to Roger II, Bute’s commission of Burges’s designs enacted a building campaign to define and solidify his position as a marquess of considerable wealth and aesthetic sensibilities. Bute’s wealth was, after all, derived through resource exploitation. Coal was the fuel of the British Empire, and Welsh coal was the preferred coal of the empire’s navy and its ships. Welsh coal moved throughout the empire, and it also enabled the empire to move globally and colonize sovereign peoples and cultures. In this way Cardiff Castle is a site within the network of the British Empire and its designs speak to multiple audiences, similar to the Cappella, about the breadth of the empire and its cosmopolitan composition.

Although domestic interiors are read through the viewership of the patron and his circle, such a building campaign also spoke to the local community and was, thus, necessarily legible to them, too. The Arab Room at Cardiff Castle was restricted to those who gained access to the castle. It was used as a tea room by Gwendolen, but its location in the castle is within the public-facing sphere. It is tucked behind the grand dining room, and is highly visible within this space. Those who could gain access included members of Bute’s family, guests, musicians and entertainers hired for parties, and household staff. We should not assume that all of Bute’s guests and staff were exclusively white and/or British. It is likely that household staff included people hired from within the local community with strong ties to Tiger Bay, Cardiff, and Wales more broadly. Similar to the quadrilingual monument to Anna in Palermo, the Arab Room addressed multiple communities and cosmopolitan, polycultural viewers, whose gazes are anticipated and invited.

Many of these cosmopolitan viewers would have recognized their own aesthetic cultures cited throughout the design choices of the Arab Room and other spaces in the castle. Another example of the castle’s designs speaking to, and signifying, Cardiff’s populace may be seen in Burges’s rooftop Pompeian Roman Garden, a small rooftop space which has Hebrew script and depictions of the Old Testament story of the prophet Elijah painted on tiles which line the walls. These designs spoke to Bute’s piousness, his study of the Hebrew language, and linked him to the Roman Empire. Moreover, the designs also spoke to, and with, the significant and growing Jewish community in Cardiff during the eighteenth and nineteenth centuries. In these ways, Burges’s designs at Cardiff Castle, with its tantalizing hints of the interior designs from the polychrome exterior, invited and addressed multiple viewers across intersecting identities. Burges’s source material — Norman Sicily — was a culture that drew heavily on its character as a cosmopolitan society with broad and, at times, overlapping aesthetic tastes and practices. By referencing a culture that looked to its community to represent its sovereign, Burges’s design choices invite us to consider not only how they signify Bute’s wealth. More interestingly, we are also invited to see how his designs redirect our attention from Bute towards the eyes and multiple, complex viewerships of the cosmopolitan populations in Tiger Bay and Cardiff.

Acknowledgements

This article is part of my forthcoming project Embodied Beauty: Imperialism, Settler Colonialism, and Slavery, 1857–1900. Innumerable thank yous to Dr Jason Edwards for your feedback throughout this project. Thank you to Dr Maddie Hewitson for inviting me to share my research alongside so many thoughtful scholars. It was a pleasure to speak alongside Dr Charlotte Ribeyrol and her project on Burges’s Great Bookcase at the Ashmolean Museum. Thank you to both the Ashmolean and Barry and Deedee Wigmore for supporting this conference and for having me as a guest. Thank you, also, to all the speakers, commentators, and A.G. for the tour of Oxford chapels. As always, all mistakes are my own. Thank you to my loved ones, who support and guide me through praxes of community care towards a beautiful, embodied decolonial future.

Competing Interests

This project was kindly supported by a Paul Mellon Centre Research Grant.