The Interlaced Arches and the So-Called sebka Decoration: Origin and Materialisation in al-Andalus and Its Reinterpretation in Medieval Castile

Abstract

In this article, I aim to address one of the most characteristic decorative elements of the Almohad period, the so-called sebka decoration. With this aim in mind and through the research carried out and the examples that have been preserved, I consider it appropriate to know the origin of this ornamental motif that is so recurrent in the Andalusian architectural panorama and to analyse not only its compositional scheme but also the different formal variants that arose around it. Furthermore, its use in other buildings in the Kingdom of Castile is a further indication that allows us to approach a scenario where cultural and artistic transfer between al-Andalus and the Christian territories was a reality.

1. Introduction

As a decorative element that constitutes one of the hallmarks of Almohad art, the sebka is characterised by its presence in a large number of Andalusian and Maghrebi constructions during this period. According to the surviving material evidence, this motif had a significant impact in later dynasties, such as the Nasrid and Marinid dynasties or during the Sa’did and Alawite periods, and was even used systematically in Christian works—mainly between the 13th and 15th centuries—and in many variants. All this leads me to corroborate the relevance of this ornamental resource, on which there are only a few research studies (Pavón Maldonado 1996, pp. 45–129; Villalba Sola 2013, pp. 141–42; Villalba Sola 2019, pp. 47–52; Giese and Varela Braga 2021, pp. 431–59) and numerous specific references that deal with its analysis when dealing with a particular period or work in question. For these reasons, I believe that a detailed study is needed to shed more light on this element as a whole, as well as to answer some of the questions that I will be raising.

In this sense and having been mentioned by Manuel Gómez-Moreno (1929, p. 237) with the term “sebca”, authors such as Leopoldo Torres Balbás (1949, 1955) saw in this decorative motif a rhomboidal pattern that was repeated over a large part of the architectural surface. Hence, because of its similarities to this geometric figure, it was called a “diamond pattern” or “rhombus pattern”, forming a kind of mesh or ornamental network. George Marçais (1954) echoed this decoration with the expression des réseaux losanges (“diamond latticework”) due to its similarities to the shape of the rhombus, as also referred to by Buresi and Ghouirgate (2013, p. 132) and Pierre Guichard (2013, p. 145), among other French specialists.

This conception has been maintained to the present day, and the term sebka is widely used in academic circles to refer to this ornamentation found in Islamic constructions of different types (religious, palatial, domestic, etc.) and whose use in Christian works exemplifies the artistic transfer that had such an impact at the time in Aragon—with special attention being paid to it by authors such as Cabañero Subiza and Sábada Lizanzu (1996, pp. 25–74)—and in Castile. The latter has not received the same attention as in the case of Aragon, and I will examine it in more detail below.

However, it is not my aim to carry out a detailed formal and stylistic analysis of this decorative motif in each of the works—Islamic, Christian, and Hebrew—in which it appears and in which we already have in the different works published, something that would be too lengthy for an article; rather, I aim to reflect on its origin and evolution for a greater understanding of the subject and to know to what extent it transcended into other cultural spheres. For this reason, I will take as a reference the most significant examples from the Iberian Peninsula, where allusions to North Africa will be constant as a consequence of the unification of both sides of the Strait and the cultural exchange that took place between al-Andalus and the Maghreb between the end of the 11th century and the middle of the 13th century.

2. About the Term and Origin of the sebka Decoration

Before focusing on this type of decoration, I would like to briefly consider some questions related to its origin and meaning. As I have already mentioned, the term sebka has been systematically used in historiography to refer to the rhomboidal architectural decoration that appears on surfaces, mainly from the Almohad period onwards. However, in view of the variety of designs used—among which Basilio Pavón Maldonado (1996, pp. 47–53) himself differentiated between up to seven different types—I would like to clarify this further, starting with its meaning.

According to the definition provided by the Thesaurus of the Cultural Heritage of Spain (Ministry of Culture), it is an “ornamental architectural element widespread in Almohad architecture. It is composed of a decorated rhomboidal network formed by a series of superimposed arches, each one riding on the keystones of the arches below it”1. As can be interpreted from the latter, the sebka is made up of different arches that are superimposed on top of each other in an ascending fashion, with the outline of a network of lozenges. The term is therefore directly attributed to a decoration defined by specific motifs and promoted by the Almohad dynasty. Dolores Villalba Sola (2019, p. 51) points out that this word comes from the Arabic šabak (to cross, to entangle), whose variant šabaka (plural: šabak, šabakāt, or šibāk) means “net” or “mesh”2.

The latter definition is, a priori, far removed from the limitations offered by the former and is also used by historians in a broader sense to refer to the rhomboidal decoration made up of different ornamental elements, such as arches, vegetal motifs, or simple lines. In one way or another, the rhombus design becomes the leit motif of this decoration and has been considered by some specialists as one of the characteristic features of Almohad art (Pavón Maldonado 1996, pp. 45–129) and, as a consequence, of the decorative pattern generated by its repetition (Villalba Sola 2018, pp. 90–91).

This double reading is what leads me to reflect not only on its etymological meaning but also on the origin of this decorative motif, aspects of which, in my opinion, are fundamental for understanding the subject under study. According to Dolores Villalba Sola (2019, p. 47), for Juan Clemente Rodríguez Estévez (1998, pp. 112–17), the birth of this decoration based on a grid of lozenges dates back to the Minoan culture, an ornamental element that was also used—according to the aforementioned specialist—by the Amazigh peoples (primitive Berbers of North Africa) and which, in her opinion, may possibly constitute the germ of the Almohad sebka.

To this last point, he adds that the earliest evidence of Islamic roots, known in the West as sebka, is found in the Umayyad period, specifically in the apodyterium of the bath of the palace complex of Quṣayr ’Amra (Jordan), built during the caliphate of Walid I (705–715). On its vault, we can see the fresco painting of a “network of lozenges” (Almagro Basch et al. [1975] 2002, p. 45) formed by the crossing of small leaves that follow one after another with a figural decoration inside (Figure 1), as also noted for the towers flanking the main entrance of Qaṣr al-Ḥayr al-Gharbī (Syria) (Villalba Sola 2013, p. 147; 2019, p. 48), built between 724 and 727 during the caliphate of Hisām b. ‘Abd al-Malik (724–743), and in whose interior the figures have been replaced by vegetal motifs. Moreover, this type of decoration bears some similarities to the stuccoes of Samarra in Iraq (9th century).

However, other specialists (Torres Balbás 1949, p. 51; Caillé 1954, p. 123; Marçais 1954, p. 257; Ewert 2006, p. 228; Guichard 2013, p. 145) affirm that the origin of this ornamentation can be found in the interlacing of arches in the extension of the old mosque of Córdoba3, built under the Caliph al-Ḥakam II (961–976). This formula will acquire an important development in the Almohad period when a network of lozenges is generated in an ascending direction along the surface to be decorated (Gómez-Moreno 1929, p. 237), the evolution of which Basilio Pavón Maldonado (2016) examines, corroborating its genesis in the aljama of Córdoba.

In view of this approach and taking into account the etymological meaning of the term šabaka (plural: šabak, šabakāt, or šibāk), any type of decoration with these characteristics would correspond to this concept and could, in my opinion, fall into a certain ambiguity. Hence, I wonder whether we should really speak of sebka when referring—in addition to the examples cited from Quṣayr ’Amra and Qaṣr al-Ḥayr al-Gharbī—to the rhomboidal pattern that appears to ornament one of the interior tympana of the Great Mosque of Susa (Tunisia, 9th century) or the southwestern façade of the ancient mosque of Bāb al-Mardūm (Toledo, late 10th century) (Figure 2), to cite a few cases from the Islamic West. The presence of a grid of lozenges is evident; so, in principle, there is no doubt about its occurrence. However, what is the relationship between this composition and that which appears in the great Almohad works on both sides of the Straits of Gibraltar?

In relation to this last point, the concept we have of the Almohad sebka is far removed from the simple weave of lozenges that we have seen previously, and its configuration derives, rather, from the ascending interlacing of arches that, progressively, will acquire a certain autonomy and varied forms, on which I will comment below. This could be one of the reasons that leads us to think about the lack of chronological and formal continuity in terms of this decoration between the aforementioned examples from the 8th century and the works from the second half of the 12th century, to whose problem Dolores Villalba Sola (2013, p. 142) has already referred. This is why, in my opinion, we must turn our gaze to the ancient mosque of Córdoba to find out the origin and evolution of the sebka decoration, where the arch is the basis of this ornamental resource.

3. From the Structural to the Decorative: The Systems of Interlaced Arches in al-Andalus

Bearing in mind that the horizontal interlacing of semicircular arches can already be found in early Islamic works, as in the main façade of the Qaṣr al-Ḥayr al-Gharbī as a decorative frieze, in the extension of the mosque of Córdoba by al-Ḥakam II, the predominant typology in this system of interlacing arches is the lobed arch, showing a tendency towards verticality. This arrangement, of which no precedents have been found and on which Christian Ewert made a detailed analysis (Ewert 1968, pp. 15–25), can be found in the most important spaces of the aljama of Córdoba, which were intended to be exalted from an architectural and decorative point of view. I refer to the entrance of the longitudinal maqṣūra—which included the three central naves of the Umayyad caliph’s intervention—and the transversal maqṣūra located in front of the miḥrāb (Abad Castro 2009, pp. 9–30), each section being covered by a vault of interlaced arches (Figure 3), which constitutes another of the contributions of the 10th-century extension4.

In addition to the presence of these vaults, these spaces also have a particular configuration of openwork façades made up of different types of arches in various bodies and planes. The structural function of the horseshoe arches predominates, reinforced by pentalobulated arches, in order to support the aforementioned roof system. This means greater stability and distribution of the loads of the vaults, being fundamental to the grouping of two or four columns in some of the angles of the sections and having one or two columns on the sides (Gómez-Moreno 1929, p. 236; Torres Balbás 1952, p. 46). However, by focusing on the subject that concerns us, in these systems, we can see different architectural solutions carried out in a short space of time.

The Chapel of Villaviciosa or the Lucernario—the former entrance to the longitudinal maqṣūra of al-Ḥakam II and whose space was occupied from 1236 onwards by the presbytery of the primitive Christian church—was delimited on three of its sides (east, south, and west) by this system of façades, while on the northern front, a double horseshoe arch was opened up with a lobed arch behind it, as can be seen today. Of these three façades, and without forgetting the later Christian interventions and the restoration work carried out by Ricardo Velázquez Bosco in the aforementioned chapel (Ruiz Cabrero 1985, p. 54; García Cuetos 2015, pp. 119–20), only the eastern façade, built in the 13th century, and the one located to the south of the Lucernario have been preserved, the one corresponding to the western side having been lost in the 15th century when the first Gothic temple was remodelled.

The eastern façade facing the interior of this space5 is made up of three bodies of high arches—pentalobulated and horseshoe arches—on different levels of depth (Figure 4). In the second body, we can see a simple horizontal interlacing in three planes, configured, on the one hand, by the arrangement of the pentalobulated arches with respect to the horseshoe arches over which they are superimposed (displaced half a span and partially covering them) and, on the other hand, in relation to the start of the pentalobulated arches that occupy the third body. We could therefore speak here of a first step towards the development of interlaced arches in an ascending direction that would progressively materialise in the Islamic West, where the location of the pentalobulated arches, supported by their corresponding crests, create a framework between them along the entire façade.

The southern façade of the Lucernario (Figure 5) is structured in two main sections separated by a cornice that acts as an additional reinforcement. In the upper section, and also superimposed on different levels, there is a new interlacing of pentalobulated and horseshoe arches in the same form as in the eastern façade but framed by a thirteen-lobed unloading arch. In this interlacing, we can already see how the thread of the pentalobulated arches is prolonged in an ascending direction with three more lobes, constituting, in my opinion, an advance in this respect. However, it is not known whether these new interlacing and unloading arches were originally planned (Abad Castro 2009, p. 16).

With respect to the arcades of the maqṣūra that delimit the section that precedes the miḥrāb, the horseshoe arches located in the second body of the southern façade (Figure 6) are not intersected by other pentalobulated arches in the same way as we have seen in the Lucernario, that is, using different planes for them. The only thing we find, sheltered by these horseshoe arches, are the starts of some stabilising lobed arches whose piece—as a central voussoir—breaks into the axis of the lower pentalobulated arches (Ewert 1968, p. 22) and dies in the soffit of the first ones. This has even led to the suggestion that the arcade as a whole was not developed correctly (Camps Cazorla 1953, p. 51) and that the stabilising arches were the result of a later addition (Abad Castro 2009, p. 18)6. However, the central piece referred to above is replaced by a voussoir with a curved profile—which simulates the circumference of the lobe—on the interior faces of the arcades perpendicular to the qibla wall of the section preceding the miḥrāb (Figure 7), thus better resolving its layout. In this way, the lower pentalobulated arches are given a sense of continuity in an upward direction, which can be clearly distinguished on their opposite faces as the lines of the extrados intertwine between them (Figure 8).

At this point, we could say that different solutions were produced in these systems of horseshoe and pentalobulated arches during the Caliphate of al-Ḥakam II that allow us to corroborate the possible origin of the so-called sebka panels in the aljama of Córdoba. Consequently, there would have been a change from an interlacing and superimposition of arches in height—whose upper arrangement supports the load of the vaults with interlacing ribs and the lower one acts as a reinforcement (Ewert 1968, p. 15)—to an ascending interlacing with a clear ornamental intention and that only allows, as Gómez-Moreno (1929, p. 236) pointed out, the lobed arch. This sequence could respond to the different construction phases to which the mosque was subjected during the expansion of the second caliph of Córdoba (Abad Castro 2009, pp. 10–22), in the same way as happened with the vaults of interlaced ribs to which I have just referred (Marfil Ruiz 2004, pp. 91–107), with the section occupied by the Lucernario being configured prior to that which precedes the miḥrāb.

It was in the 11th century that this interlacing that we saw in the internal arcades of the maqṣūra of the mosque of Córdoba survived in some palatial constructions, even acquiring a significant development in others. On the one hand, I refer to the Alcazaba of Málaga, a work assigned to the ḥammūdí dynasty during the first half of that century7 and in whose Patio de los Surtidores a small pavilion can be seen, to the west of the access to the southern hall, closed on its four sides by openwork arcades (Figure 9) found in 1934 (Ordóñez Vergara 2000, pp. 251–52). As we can see, the pentalobulated arches intertwine with each other in an ascending manner until they form, at the top, a knot as a finial, now lacking an upper body of horseshoe arches due to the lower height of its façades. Moreover, according to Cabañero Subiza and Lasa Gracia (2004, p. 43), this solution would no longer make much sense in a pavilion of such reduced dimensions. Despite this, the layout of these façades allows us to look back to the aljama of Córdoba, whose continuity at this time is evident. This space has been linked to the existence of a possible palatine oratory (Íñiguez Sánchez 2018, p. 338), where the presence of these intertwining pentalobulated arches with a certain distinctive character could reinforce such an approach.

On the other hand, this compositional scheme of the Alcazaba of Málaga will evolve towards an interlacing of mixtilinear arches in the Aljafería of Zaragoza (Qaṣr al- Ŷaʿfariyya)8—built during the reign of Abū Ŷa’far Aḥmad al-Muqtadir bi-Llāh (1046–1081), a member of the Taifa dynasty of the Banū Hūd—(Cabañero Subiza and Lasa Gracia 2004, pp. 40–41), where some authors already saw the mosque of Córdoba as its most direct precedent (Ewert 2006, pp. 227–28; Guichard 2013, p. 155). In this sense, Bernabé Cabañero Subiza (2018, p. 351) states that the southern front of the access arcade to the Throne Hall or Golden Hall (Maŷlis al-Ḏahab) of the Aljafería in Zaragoza (Figure 10) is a reinterpretation (even taking into account the restoration criteria carried out at the time9) of the southern façade of the transverse maqṣūra of Córdoba, with the second body of arches disappearing in the Taifa palace to make way for continuous interlacing mouldings.

As I have already mentioned, the latter was already evident in the internal arcades that delimit the space preceding the miḥrāb of the Aljama Mosque of Córdoba and in the pavilion of the Patio de los Surtidores in the Alcazaba of Málaga, thus losing the independent character of the lobed arches and the superimposition of arcades that we saw in the Chapel of the Lucernario. However, in the case of the Aljafería of Zaragoza, this vertical interlacing of mixtilinear arches in the access to the Golden Hall gives its façade, as in the case of the Alcazaba of Málaga, a greater decorative charge (Gómez-Moreno 1929, pp. 231–37; Torres Balbás 1949, pp. 31 and 51; Pavón Maldonado 1996, p. 45; 2016, p. 13), the raison d’être of the Almohad sebka.

Hence, the hūdí palace has been considered by traditional historiography as a key work in this formal and stylistic evolution I have been talking about, and it is significant that the restoration of the northern façade of this hall referred to above was carried out “as a true 12th-century sebka cloth, when its original appearance was that of a series of mixtilinear arches (with a large loop at the top) that were linked together to create a pendular line” (Cabañero Subiza and Lasa Gracia 2004, p. 37). Although in the words of the aforementioned specialists, this intervention was not the most successful, it is a clear indication of the origin of this ornamental motif and of the fact that the Aljafería is a reference point in this context.

This was the opinion of Pierre Guichard (2013, p. 145) when referring to the rhomboidal network of the Almohad minaret (ṣawma’a) of the Hassan Mosque in Rabat, whose decorative motif derived from the lobed and criss-crossed arches of the mosque of Córdoba was greatly developed in this 11th-century palace. Even with regard to the Aragonese Christian panorama, Gonzalo M. Borrás Gualis (1985, I. 177–78 and 181) already pointed out the influence that the Aljafería had on the configuration of the sebka panels present in its constructions and the appearance of the mixtilinear arch in the Taifa art of Zaragoza10, without forgetting the repercussions of its promotion and development by the Almohads from the mid-12th century onwards (Ewert 2006, pp. 227–28).

However, although no material evidence has been preserved to corroborate this, some authors point out that in the Almoravid period, an ornamentation based on this interlacing of lobed and mixtilinear arches must have been used (Torres Balbás 1949, pp. 51 and 53; Gómez-Moreno 1951, p. 283; Cabañero Subiza and Sábada Lizanzu 1996, pp. 28–29; Pavón Maldonado 1996, p. 45), which would be the first step towards the rhomboidal forms that we will see with the Almohads11. This could be evidenced by the decoration that appears on one of the interior wooden panels of the minbar of the Kutubiyya Mosque in Marrakech (Figure 11), which the Almoravid Amir ‘Alī b. Yūsuf (1106–1143) ordered Andalusian craftsmen to make (Bloom et al. 1998). It features arabesque motifs framed by two vertically interlacing mixtilinear arches and two others superimposed in a second body that seem to simulate horseshoe arches12, reminding us of the composition of the Aljafería, as already suggested by Manuel Gómez-Moreno (1951, pp. 294–95). The decoration on the outer façade behind the miḥrāb of al-Qarawiyyīn Mosque (Fez) is also significant in this regard as a material testimony between the Taifa and Almohad periods.

4. The Decoration of sebka in Almohad Architecture and Its Formal Evolution

With this exclusively decorative function, the Almohads brought the interlacing of arches to a moment of great splendour in their constructions with the development of a new compositional scheme which, as I have already mentioned, may have already existed in the Almoravid period to a greater or lesser extent. This composition is based on the rhombus and derives from this “architectural interlacing” (l’entrelacs architectural), as Basset and Terrasse ([1932] 2001, pp. 118–19) called it, becoming the hallmark of this North African dynasty.

Nevertheless, and taking into account its origin, we believe that this ornamental resource was one of its many vehicles for legitimising its position in al-Andalus, as has already been mentioned in some publications (González Cavero 2018), where the interlacing of lobed arches that we saw in the Umayyad Caliphate of Córdoba was maintained until now with different formulas, being interpreted and adapted by the Almohads to the point of taking it even to very advanced models. This can be seen in the minaret of the Kutubiyya Mosque in Marrakech (Figure 12). Although some scholars have dated the beginning of its construction to the caliphate of ‘Abd al-Mu’mīn (1130–1163) (Torres Balbás 1949, p. 27; 1955, p. 14; Pavón Maldonado 1996, p. 46; Rodríguez Estévez 2004, pp. 190–91; Villalba Sola 2015, p. 133; Almagro Gorbea and Jiménez Martín 2022, pp. 255–89), most of them agree that it was completed by the third Almohad caliph, Abū Yu’sūf Ya’qūb al-Manṣūr (1184–1199)13, with its second body acquiring a different imprint from the first.

Thus, as far as the first body of the minaret of the Kutubiyya Mosque, which is made of masonry, is concerned, it is surmounted by four tumid arches, each with seven-lobed arches starting from their corresponding columns, the threads of which intertwine to form a second architectural interlacing of ascending form, which are knotted at the top (Figure 13a,b). Its layout not only corresponds to the internal arcades of the maqṣūra of the mosque of Córdoba, to the western pavilion of the southern portico of the Patio de los Surtidores in the Alcazaba of Málaga and, although with a different type of arch, to the access to the Golden Hall of the Aljafería in Zaragoza, which I have commented on, but also bears a marked similarity to the interlacing of arches that we observe in the upper part of the oratory of the palace in Zaragoza (Ewert 2012, pp. 103–4). This demonstrates, once again, the importance of the Aljafería as an artistic reference point in this context, as Basilio Pavón Maldonado (2016, pp. 2 and 8) has already pointed out.

We are therefore faced with a formula inherited from the Andalusian tradition, which, in my opinion, will reach an important stage of development at this time with the so-called sebka decoration. Even the interlacing of the arches of the minaret of the Kutubiyya marks the beginning of a network of meshes, as Basset and Terrasse ([1932] 2001, p. 120) stated, now independent of the structural function for which this composition was originally conceived. This influence should not come as a surprise since we have evidence of how the Almohad caliphs surrounded themselves with numerous Almohad architects (ʿurafāʾ) and Andalusian architects who were at the head of their works (Torres Balbás 1946, pp. 214–24).

The second section of the minaret of the Kutubiyya Mosque has been considered the earliest example of Almohad sebka (Marçais 1954, p. 244). Dolores Villalba Sola (2015, p. 89; 2018, pp. 90–91) finds the intrados of the lateral arches of the section preceding the miḥrāb to be the earliest Almohad example of sebka, to which Basset and Terrasse ([1932] 2001, p. 55) referred with the expression “vegetal interlacing” due to the decorative motifs of digitate palms. However, and in accordance with the discourse I am following, in which the arch is the basis of this decorative model and which is based on its interlacing, I think that it would more likely be a composition based on a superimposition of palms and with a parallel arrangement between them, as Henri Basset and Henri Terrasse pointed out, not generating any type of interlacing as such and also lacking the arches that could produce it. In a more recent publication (Villalba Sola 2019, p. 49), the aforementioned specialist even states that there is an evolution from the time of the caliph ‘Abd al-Mu’mīn—when the rhombuses of the decoration originate from a simple palmette pattern—to the caliphate of Abū Yu’sūf Ya’qūb al-Manṣūr, when the sebka tends to develop towards more mixtilinear profiles. On the other hand, Georges Marçais (1954, p. 258) saw this vegetal decoration of Tinmāl in the evolution of the interlacing of arches, an aspect to which we will return later.

However, returning again to the second body of the minaret of the Kutubiyya Mosque, its formal and stylistic character has been related to the vertical interlacing of arches of the minarets of the old Almohad aljama mosque in Seville, the Hassan Mosque in Rabat, and the Qaṣba Mosque in Marrakech (Pavón Maldonado 1996, p. 48; Villalba Sola 2015, pp. 159–61), built during the caliphate of Abū Yu’sūf Ya’qūb al-Manṣūr, which I will describe below. However, due to the recentness of their workmanship compared to the contemporary models mentioned above, I believe that the decoration of the lantern of the minaret of the Kutubiyya Mosque should be the subject of a detailed study.

In addition, according to the Kitāb al-istibṣār, its author attributes the construction of the foundation phase and the extension phase of the Kutubiyya mosques to the caliph ‘Abd al-Mu’mīn (Almagro Gorbea and Jiménez Martín 2022, pp. 255–88), in the translation of which Dolores Villalba Sola (2015, p. 133) states that the minaret was completed by the caliph Abū Ya’qūb Yu’sūf (1163–1184). As the aforementioned specialist points out, and thus responding to the information offered by Ibn Abī Zar’ in his Rawḍ al-qirṭās (Ibn Abī Zar’ [1918] 1964, II. 447)—in which he attributes the construction of the minaret to Abū Yu’sūf Ya’qūb al-Manṣūr—it is possible that the latter was simply the reform of its lantern and that, in my opinion, there may have been some other intervention at a later date. This could therefore be the reason for the way it looks.

Whether the current decoration of the second body of the Kutubiyya Mosque is a more recent intervention or not, it is certain that the third Almohad caliph played an important role in the development of this ornamental element, among other constructive undertakings, to which we should devote special attention. Returning to Ibn Abī Zar’’s Rawḍ al-qirṭās, the Maghrebi historian says:

[Abū Yu’sūf Ya’qūb al-Manṣūr] When he passed on to al-Andalus to undertake the Alarcos expedition, he had the citadel of Marrākush built and the mosque adjoining it with its minaret, the minaret of the mosque of al-Qutubiyīn, the city of Rabat al-fath in the land of Salé and the mosque of al-Hasan. When he had finished the mosque of Seville and prayed in it, he had the fortress of Aznalfarache built on the banks of the Guadalquivir, and returned to al-Magrib, arriving at Marrākush in Sha’bān 594 (8 June to 6 July 1198). He found that everything he had ordered to be built had been completed, the citadel, the towers, the mosque and the minarets, all built with the fifth of the booty taken from the Christians.

(Ibn Abī Zar’ [1918] 1964, II. 447–48)

As we can read in this fragment, it is to Abū Yu’sūf Ya’qūb al-Manṣūr that we owe the construction and decoration of the minarets, which I will examine below. According to his account, everything suggests that the Seville Mosque was the first to be completed. Moreover, we know from the Almohad court chronicler Ibn Ṣāḥib al-Salā (1969, p. 195) that work on the mosque began in 1172 with the expropriation and demolition of the dwellings on the same site. As for its minaret, the author goes on to say that its construction was ordered by his father, the caliph Abū Ya’qūb Yu’sūf, in 1184, who was not even able to see the work begin because of his untimely death that same year in the Santarem campaign (Ibn Ṣāḥib al-Salā 1969, pp. 200–1).

Hence, it was during his son’s caliphate that its construction should be attributed, as is also corroborated by ‘Abd al-Wāḥid al-Marrākušī (1955, pp. 220–21) or by the anonymous author of the Ḏikr bilād al-Andalus (1983, II. 67); its construction began in the same year, and the foundations and the first courses of stone ashlars were laid by the alarife (al-’arīf) Aḥmad b. Baso and the almojarife (al-Mušrif) Muḥammad b. Sa’īd. A few months later, the latter was dismissed and the work came to a standstill until 1188, when a new almojarife—Abū Bakr b. Zuhr—was appointed to take charge of the minaret, this time with ‘Alī l-Gumārī remaining in charge and the work continuing in brick. It was not until 1198 that the minaret was completed with the crowning of the ŷāmūr (Ibn Ṣāḥib al-Salā 1969, p. 202; Ibn ‘Iḏārī 1953–1954, I. 202; Ibn Abī Zar’ [1918] 1964, II. 447).

It was precisely in this context, between 1184 and 1198, that work was carried out on the minarets of the Seville Mosque, the Qaṣba Mosque in Marrakech, and the Hassan Mosque in Rabat, the latter remaining unfinished after the death of Abū Yu’sūf Ya’qūb al-Manṣūr. The similarities in their decoration support their coetaneity, despite the fact that some authors claim that the minarets of the Qaṣba and Hassan Mosques are older in conception (Cabañero Subiza and Sábada Lizanzu 1996, p. 31), also taking into account the important role played by the Sevillian almojarife Abū Bakr b. Zuhr between Marrakech and Seville as an inspector of works, a vizier, and a poet, among other positions around the Almohad court (González Cavero 2013, pp. 278–82). It should even be noted that he was assassinated on 21 ḏū-l-ḥiŷŷa of the year 595H (14 October 1199), the same year in which the caliph died (22 Rabī’ first of the year 595H/22 January 1199), which, as has been generally accepted, meant that the Hassan Mosque was not completed and was subjected to its despoilment during the following years.

However, it is very significant that, as I have already mentioned, the change in almojarife conditioned the continuity of the work on the minaret of the Seville Mosque. This leads us to reflect on the leading role that this figure must have played in the realisation of this entire construction programme and to ask ourselves, on the one hand, whether the death of Abū Bakr b. Zuhr had something to do with the fact that the work on the Hassan Mosque came to a halt14 and, on the other hand, whether his figure was so decisive as to participate in the design and decoration of these three minarets.

If the latter is the case, this will help to further understand the similarities in their ornamentation and how, at this time, the system of interlacing arches that we saw in the minaret of the Kutubiyya Mosque is developing in an ascending direction that allows for the covering of large vertical surfaces, the essence of which is to be found in the interlacing of arches that start from columns. In this respect, there is no doubt that the Sevillian inspector and almojarife Abū Bakr b. Zuhr was well acquainted with the Andalusian artistic tradition, an aspect that could reinforce this approach.

As for the minaret of the Almohad mosque in Seville (Figure 14), made of brick, its four faces are structured in three lanes, the lateral ones being divided horizontally into two vertical panels of arches criss-crossed in an ascending direction and which we can already call, as referred to in historiography, “sebka panels”. According to Jacques Caillé (1954, p. 123), this ornamental element—the purpose of which is to decorate vertical surfaces with a possible meaning linked to the Unity of God15—is known in Morocco by the expression “derj ou ktef” (step or shoulder), with Alfonso Jiménez Martín finding its possible correspondence in Castilian with the term “adaraja”16 and to which he attributes not only the decoration of the minaret (Jiménez Martín 2007, p. 145) but also that which appears on the southern side of the belvedere of the Puerta del Perdón (Figure 15) in the aforementioned mosque (Jiménez Martín 2017, p. 312).

As in the case of the minaret of the Kutubiyya Mosque, the first section is topped by a succession of blind interlaced lobed arches. Here, however, there is a slight formal variation in that the arches of the second register—the result of this interlacing—have a series of hooks or curls, such as those that decorate the interior of the Baños de la Reina Mora (Seville) (Figure 16)17. This transformation in height from one type of arch to another could come from the Aljafería in Zaragoza, where the access to the southern hall is formed by an interlacing of lobed arches that change, in a second order, to mixtilinear arches.

It is precisely the twinned mixtilinear arches (on the lower side panels) and twinned lobed arches with hooks or curls (on the upper side panels), all supported on columns (from which an interlacing of their threads is generated on the four faces of the minaret), that are repeated vertically over the entire surface, giving the whole complex a carved brick decoration and making this interlacing of arches the hallmark of Almohad art. However, it should also be noted that all these vertical panels, with the exception of the lower panels on the northern and western sides (which have a single line of mixtilinear arches (Figure 17)), have a second ornamental mesh that is recessed in the background (Figure 18), whose intention, according to Juan Clemente Rodríguez Estévez (2004, pp. 208–9), is to make the upper part of the whole complex more fragile and thus create a sense of ascent. This second mesh is made up of smooth palm leaves that are chained together and superimposed one on top of the other, simulating rhomboidal shapes and arches, without interlacing with each other, which leads us to think of the decoration of the intrados of the lateral arches of the section preceding the miḥrāb of the Tinmāl mosque, which may have had a certain influence here.

However, the upper panels of the southern face are striking, where the second sunken mesh is the result of the interlacing of leaf arches (Figure 19), a typology characteristic of Almohad art. This novelty in terms of the composition of a double mesh, which Basilio Pavón Maldonado (1996, p. 47) had already echoed by including it among the seven modalities he established in his study of the sebka, will be greatly developed in the portico of house no. 10 at Siyāsa (Cieza, Murcia) (Navarro Palazón and Jiménez Castillo 1995, pp. 120–27; 2005, pp. 282–84; 2006, I. 249–303). In its side panels, the latter has a main (or “dominant”) pattern formed by lobed arches that intertwine and three superimposed patterns of plain palms, although, as we shall see, its configuration is of a different design (Figure 20a,b). However, as the abovementioned specialists point out, the use of plasterwork meant that even domestic architecture could be decorated with this type of decoration.

It has been interpreted that the sebka panels of the old minaret in Seville, together with the openings in the central section, form a horizontal tripartite façade and have been related to Almohad palace porticoes (Manzano Martos [1995] 2022, pp. 112–14; Rodríguez Estévez 2004, p. 208; Blasco López et al. 2009, p. 203; Márquez Bueno et al. 2024, p. 192). This is the case of the southern portico of the Patio del Yeso and the northern portico of the courtyard of the former Casa de Contratación in the Reales Alcázares in Seville. The Patio del Yeso (Figure 21a,b), discovered by Francisco María Tubino (1886, pp. 231–54) at the end of the 19th century, recovered by the Marqués de la Vega Inclán in the first decades of the following century, and restored by Rafael Manzano Martos between 1969 and 1971, has a structure that has an important impact on Nasrid palace architecture. Formed by a central arch of lambrequins, the lateral sections are made up of three leaf arches on which are juxtaposed a series of motifs that attempt to imitate an interlaced fretwork in an ascending direction with the same type of arches.

As we can see, there is no real interlacing of the quality we have seen in the minaret of the Almohad mosque of Seville. On the contrary, there is an initial intentionality between the leaf arches with the start—from what could be a small pillar on axis with the columns—of other new arches that did not develop correctly vertically, as was achieved in the lower panels of the northern and western faces of the minaret of the mosque (see Figure 17). Although it is a horizontal interlacing, the frieze fragment from calle Cristo del Buen Viaje nº 2 (Seville), kept in the archaeological museum of the same city, has been dated—albeit with doubts—to the first half of the 13th century (Lintz et al. 2014, p. 382). Rafael Manzano Martos ([1995] 2022, p. 114) has already pointed out that the decoration of the spandrels of the central arch shows the hand of an inexperienced plasterer, which leads me to think that the same was true of the lateral modules of the aforementioned portico.

However, it is strange that being a palatial area and coeval with the minaret, where the same master builder may have been involved, it has this appearance. In addition to this, it is important to note the deterioration and alteration to which it was subjected after passing into Christian hands in 1248, as well as the numerous alterations that took place between the 16th and 17th centuries (Marín Fidalgo 1990, I. 147–50 and 250–56) and the restoration criteria of the time influenced by the theories of John Ruskin (García Cuetos 2016, pp. 151–60). What if it was the initial test of a new formula that materialised and developed at that time in the Almohad minarets? If so, the model of the portico of the Patio del Yeso would not have been influenced by the latter, as traditionally accepted and as proposed by Francisco María Tubino; rather, the opposite would be true.

This approach could respond to Dolores Villalba Sola’s reading of this issue (Villalba Sola 2015, p. 214), in which she suggests that this irregularity, already highlighted by Rafael Manzano Martos for the decoration of the spandrels of the central arch, could possibly be due to the fact that we are dealing with an early model. Hence, in my opinion, and extending it to the side arcades as I have pointed out, the intention was to achieve a thicker sebka like the one that appears in the lower panels of the northern and western elevations of the minaret of the Seville Mosque mentioned above, and it is therefore one more sebka model among the varied compositional schemes that the Almohads used, even in the same work. To this, I should add that Leopoldo Torres Balbás (1949, p. 31) already stated that in the Patio del Yeso, we can appreciate some of the forms that would later be developed in the Nasrid period, finding, in my view, other intermediate formulas that would mark this evolution.

This can be seen in the minaret of the Qaṣba Mosque in Marrakech (Figure 22). Its elevation, made of brick, has a single decorative panel on each of its faces that occupies almost the entire surface. This ornamentation is based on a triple-lobed arcade on the northern and southern sides and lobed arches with pointed arches on the eastern and western sides, the threads of which are prolonged and intertwined in an ascending manner. Between these lower arches are small columns on axis with the columns supporting the latter, from which a series of tri-lobed arches (northern and southern canvases) and mixtilinear arches (eastern and western canvases) are placed in successive rows in the same way as Pierre Guichard (2013, p. 145) pointed out for the Hassan Tower in Rabat, helping to create a denser sebka as their arches also intertwine. It is worth noting that these arches are backed by a ribbon that tends to emphasise this interlacing and whose course is marked by turquoise-green tiles, like the interior of the tri-lobed arches, a decorative material that can also be seen finishing off the first body of the minaret (as well as its lantern), whose study has been carried out by Claire Déléry (2014, pp. 329–32).

It is likely that this minaret served as a model for the Hassan Tower in Rabat, where its decoration—different on each of its faces and made of stone—is reduced to the arches on which the sebka is developed (Figure 23) and where the reminiscence of Andalusian art is still present. In the second body of the western and eastern elevations, there is, framed by arches of a different typology18, an ascending interlacing of lobed and mixtilinear arches, respectively, which evokes the mosque of Córdoba and the Aljafería of Zaragoza, as I have commented on earlier. Its ornamental purpose is what makes this interlacing of arches develop vertically, forming a sebka, something that the system of interlacing arches of the mosque of Córdoba did not permit due to its structural function (Gómez-Moreno 1929, p. 237).

As in the minaret of the Seville Mosque, the third section of the façades is more fragile, with a very dense sebka of lobed arches and lobed arches with interspersed lancet arches on columns, between which small columns start on axis with the latter and support the tri-lobed arches (western and eastern elevations) and the mixtilinear arches (northern and southern elevations) that are developed and intertwined vertically. In addition, these arches also have a double ribbon separated by a sunken line and are arranged in a row on the model of the Qaṣba of Marrakech. As we can see, the existence of these small columns is essential for the proper development of this type of decoration, recalling those found in the mosque of Córdoba in the upper bodies of the spaces already studied (Caillé 1954, p. 123).

However, the decoration of the second section of the southern façade is striking (Figure 24). Here, twinned mixtilinear arches on columns are linked to another series of mixtilinear arches superimposed in height that do not start from the keystones of the lower arches like the others as they do not have these small intermediate columns, and there is no development of their main arches in an ascending direction that would allow for an interlacing, as we have seen in the upper sections. The forms generated are reminiscent of the portico of the Patio del Yeso, which leads me to think that they were completed in the Hassan Tower or, on the contrary, that in the case of the aforementioned portico, the alterations that this decoration underwent over the centuries caused it to lose its original appearance. In one way or another, with this minaret, we are witnessing the consolidation of the Almohad sebka, whose influence was decisive not only at this time but also in later dynasties (Rodríguez Estévez 2004, p. 213; Villalba Sola 2015, p. 268) and even in the Christian sphere.

This may have contributed to the evolution of the sebka decoration, which progressively acquired a certain autonomy and became independent of the columns and arches that were so important for its composition from its origins in the Aljama Mosque of Córdoba until its conceptualisation in the Almohad period. In fact, and in the words of some specialists, “it is the most advanced chronological and formal solution”, as in the bell towers of Aragon (Cabañero Subiza and Sábada Lizanzu 1996, p. 40). This can be seen in the northern portico of the courtyard of the old Casa de Contratación in Seville (Figure 25a,b) where, thanks to the recovery and restoration work carried out by Rafael Manzano Martos ([1995] 2022, p. 122), it was possible to restore it. Here, we can observe an important change since the portico’s frame, which is configured on leaf arches on both sides of the central lobed arch, is not the result of the ascending interlacing of the latter but rather, small pointed arches of leaves are mounted directly on its extrados, thus eliminating the small columns that we have seen at their junction in other previous examples.

Although the concept we now encounter is different, it would not be strange if it was derived from this decorative typology promoted by the Almohads. Moreover, this model of arches mounted directly on the extrados of the main arches is very common for covering surfaces where, as there is only one main arch, there can be no interlacing. Even on occasions, with two arches, these columns begin to disappear and start directly from the main arches, as in the case of the old minaret of the church of San Juan de los Reyes in Granada. This would therefore be a technical solution. I think that this is the case, to cite a few examples, at the junction between the first and second Kutubiyya Mosque (Marrakech, in the vicinity of the minaret), from the remains of which we can intuit that the arches above the single lower arch are of leaves19 (Figure 26), or in the case of the aforementioned portico of house no. 10 in Siyāsa, there is a clear predominance of vegetal elements and the lobed form of the arches that make up the layout of its mesh seems to be losing importance.

In this sense, we have seen in the Patio del Yeso and in the minaret of the Seville Mosque how the arch of leaves and the plain palm leaf are acquiring a certain prominence in these wefts, moving away, as Georges Marçais (1954, p. 258) pointed out, from their primitive character and enriching themselves with new forms. For his part, Christian Ewert pointed out that Almohad decoration tended towards a “vegetalisation of architectural structures, especially in arches” (Ewert 2006, I. 227), finding a possible precedent in an alabaster capital from the Aljafería in Zaragoza, which shows a frieze of intertwined arches whose lobes become digitating leaves. This can be seen not only in the examples mentioned above but also in the walls of the flowerbeds in the courtyard of the former Casa de Contratación (Figure 27a,b). In spite of their state of preservation, everything suggests that they are drawn with arches of leaves that criss-cross each other in an ascending manner.

5. Granada and Castile Between the 13th and 15th Centuries: Towards a New Conception of sebka?

As we know, the sebka decoration and its formal evolution promoted by the Almohad dynasty would survive in later dynasties (Nasrid, Marinid, Hafsid, etc.), materialising in some buildings in Castile between the 13th and 15th centuries. However, unlike in the case of Aragon, we believe that insufficient attention has been paid to the Castilian kingdom in this regard, with the exception of the occasional work (Pavón Maldonado 2011, pp. 1–21; Giese and Varela Braga 2021, pp. 431–35). This was a time when Castile and Granada participated in a common artistic language motivated not only by the taste of the period but also by the appropriation and reinterpretation of certain forms and even by the intervention of Castilian and Nasrid artists in both kingdoms.

In this context, Toledo evokes the past of Caliphate Cordoba on the façades of some of its buildings with the incorporation of a frieze of intertwined lobed arches which also reminds us of the top of the first body of the minarets of the Kutubiyya mosque or the old Almohad aljama mosque in Seville. This can be seen in the churches of Santa Leocadia (late 12th century), Santiago del Arrabal (second half of the 13th century), or Puerta del Sol (late 14th century), where we can perceive this intention to continue developing vertically to draw a perfect framework with an already widespread artistic language.

Therefore, referring to the Almohad sebka, the different variants that this decorative motif adopted can also be seen in the Christian sphere and even in this same period. I refer, for example, to the chapel of the Assumption in the monastery of Santa María la Real de las Huelgas (Burgos), founded at the end of the 12th century by Alfonso VIII of Castile and his wife Elena of Plantagenet. In this space, specifically on the side walls between the two arches, there is a plaster decoration which, according to Leopoldo Torres Balbás (1949, p. 41), consists of lobed arches that extend and intertwine above the keystone to form a “network of lozenges”.

It is therefore a single double-ribbed lobed arch with interspersed eyelets that has very similar formal features to the three arches on the northern elevation of the Hassan Tower, on which the corresponding sebka panel is developed. In addition, its spandrels are decorated with vegetal motifs framed by smooth palms that are joined together, a decoration that we have seen in some Almohad examples and that had a significant development in the Nasrid period, being reinterpreted in Christian constructions of the 14th century to the point of abstraction (Figure 28a,b) and without being, in my opinion, a sebka decoration as it is not generated from an interlacing of arches, as I have explained earlier. However, in the interlacing referred to by Leopoldo Torres Balbás in the Chapel of the Assumption, the arches that are generated are reminiscent of the small arches of leaves found in the northern portico of the courtyard of the old Casa de Contratación in Seville, which seem to force this interlacing by basically arranging them one on top of the other.

This leads us to ask, together with the scarce vertical development of the supposedly vertical weft in the Chapel of the Assumption, whether we can really speak of sebka or, rather, whether it is a different model derived from the latter, as in the Sevillian palace. Apart from this, its Almohad roots seem clear, as the aforementioned specialist (Torres Balbás 1949, p. 43) has already mentioned, suggesting the existence of an “Andalusian” workshop for its decoration and another “Toledan” one for its construction, without yet achieving the artistic syncretism that would reach great splendour in the 14th century and in which the sebka, as it was conceived in the Almohad period, would tend to be reinterpreted towards a certain geometrisation.

In view of the syncretism to which Leopoldo Torres Balbás alluded, it is in the palace of Tordesillas (ca. 1360), Valladolid, now the convent of Santa Clara, where we begin to appreciate this reality. As far as the subject of this study is concerned, the central street of the upper body of the façade is endowed with a system of interlacing arches that is developed in an exceptional manner and with forms very similar to the sebka of the Almohad period (Figure 29). It is decorated with two lobed arches with intercalated eyelets and small columns at their junction, as well as at the ends, making an axis with the columns that support this double arcade. Once again, we can see how, starting from these architectural elements, a series of mixtilinear arches are arranged in an ascending direction and intertwine to form a sebka. Some authors have related the latter to that of the Tower of Hassan in Rabat (Pavón Maldonado 1996, p. 57), except that in the Christian palace, there is a knot in the keystone of the arches and columns for purely ornamental purposes.

A similar compositional scheme—without being, for the most part, a replica of the façade of the palace of Tordesillas, as has been interpreted (Pavón Maldonado 2011, p. 5)—which will evolve towards this geometrisation, is found in the synagogue of Samuel ha-Leví (ca. 1360), Toledo (Figure 30), treasurer of Pedro I of Castile (1350–1369) and is now the Sephardic Museum. In the central panel of the hejal wall, the mixtilinear arches that start from the triple-lobed arcature with interspersed eyelets shelter inside them polychrome vegetal motifs in red, green, and black, forming a rhomboidal figure (Muñoz Garrido 2017, p. 129; Utrero Agudo et al. 2023, p. 84) that is getting closer and closer to this geometric form. However, there is no interlacing of the arches as we have seen so far, but they are arranged one on top of the other and the small columns of the spandrels disappear, with the exception of those in relief at the ends of the triple arcade, which act as support for the first row of mixtilinear arches.

This development was already present a few years earlier in the synagogue of Córdoba, built around 1315 under the direction of the master builder Ishaq Moheb. Despite the alterations it suffered from 1492 onwards, and thanks to the recovery work carried out since the end of the 19th century, the wall of the hejal preserves the remains of some mixtilinear arches (with their less pronounced outline) made in plasterwork, with vegetal decoration inside and a weft of smooth palms that some authors think could be a very subtle sebka (Muñoz Garrido 2017, p. 60). We do not know what the lower arches on which this decoration would have been arranged were like. Jesús Peláez del Rosal (1988, p. 147) has already suggested that there could be three arches supported by columns, as in the synagogue of Samuel ha-Leví. In the western wall, there is an opening through a lobed arch with intercalated eyelets that belonged to the tevah. On this arch, there are several mixtilinear arches with vegetal decoration on the inside, which cover the entire surface and start directly from the top surface, resembling a lozenge20.

At this point, we can see how the small columns arranged in the spandrels of the main arches, which allowed new arches to be started and interlaced, have gradually disappeared to make room for a network of arches that have practically lost their lobed or mixtilinear profile with a clear purpose, that of adopting straighter lines in accordance with the shape of the rhombus as a geometric figure. This can be seen, for example, in the southern portico of the Patio de los Arrayanes (first half of the 14th century) in the Alhambra in Granada (Figure 31), in the Patio de las Doncellas in the Palace of Pedro I (1364–1366), or in the Reales Alcázares in Seville (Figure 32), where I believe that the primitive sense of the Almohad sebka has been lost, giving way to a clear predominance of the rhomboidal shape as a decorative element. However, we should not forget that the Almohad portico of house no. 10 in Siyāsa or even in the Cuarto Real de Santo Domingo in Granada—an early example of Nasrid sebka—(Figure 33) already included this new form.

Consequently, at this time, ornamentation became widespread, acquiring a strong autonomy and becoming independent of the main arches, mounting directly on their extrados and even appearing as a completely individualised ornamental element. This is the case of the northern and southern walls of the Royal Chapel of Córdoba (second half of the 14th century), where several lobed lines are drawn to form lozenges that frame different plant and architectural motifs (Jordano Barbudo 2023, p. 177) (Figure 34). Could this ornamentation be considered sebka? Perhaps, and conceiving it as the result of a formal and stylistic evolution of what we saw in the Almohad period, there should be no confusion; however, further studies would be necessary to undertake a detailed analysis of this subject, as well as each individual case.

6. Conclusions

The different arguments that I have put forward regarding the origin and evolution of the sebka allow us to corroborate that it is an ornamental resource promoted during the years of the Almohad caliphate, reaching great splendour under the caliphate of Abū Yu’sūf Ya’qūb al-Manṣūr on the occasion, in my view, of the construction enterprise that he carried out on both sides of the Straits of Gibraltar. Its origin is to be found in the superimposition of interlacing arches in the mosque of Córdoba, a formula conceived from the outset with a structural function which, by its very essence, monumentalises and ornaments the space. This architectural system, in which the arch and the columns were to become the essential elements, gradually moved away from its original function to acquire a clear decorative connotation. Even its later development, derived from the interlacing of arches in an ascending direction, ended up covering large areas.

However, the importance that these arches (lobed, mixtilinear, or leaf arches) attained in the layout of the decoration itself meant that they became independent of the lower arches whose interlacing generated them, even eliminating the small intermediate columns and knots on which they started and acquiring a value in their own right and a certain autonomy. This scenario was to survive from the 13th to the 15th centuries, both in Muslim and Christian buildings, which is evidence, at the same time, of the cultural and artistic transfer that took place at that time. This was linked to the geometrisation of the network of arches, whose tendency towards the rhombus as a frame for the vegetal motifs that decorate their interiors was a constant feature.

All this allows us, on the one hand, to understand the origin of the sebka and, on the other, to highlight the arch and the column as architectural elements necessary for its initial configuration, reinterpreted at later times as part of the evolution to which this decorative motif was subjected. This also means that we can reflect on the ornamentation that, by the simple fact of configuring a rhomboidal lattice, has also been interpreted as sebka; future research is needed to shed more light on this.