The Role and Meaning of Religious Architecture in the Umayyad State: Secondary Mosques

Historiography and archaeological research have traditionally defined mosques mainly as religious spaces or places to pray, without further specifications. This simplification has usually dominated the analyses of mosques, while other uses or functional aspects of these buildings were put aside. The scarcity of material information available for years to approach these buildings, together with the dominance of the more monumental examples—such as the great mosque of Córdoba—provoked that analyses about other more modest mosques were scarce or almost inexistent. However, in recent decades, the proliferation of real estate building activities has led to the recovery of many new and fresh archaeological data related to other mosques different from the Friday ones. Specifically, in Córdoba, the volume of information recovered has been enormous, and concerns not only mosques as isolated buildings, but also their urban environments, construction processes, and evolution along the centuries. Therefore, in this paper, we offer a summarized overview of the state of the arts about research on mosques in al-Andalus, presenting the main problems and limitations of the topic until now, and also the case of Córdoba and the main results achieved there as a reference for further actions to be undertaken in the rest of the territory.


Introduction: Mosques, Beyond Religion
Religion is one of the main modelers of Islamic civilization and Muslim lifestyles.The need to perform daily prayers explains why mosques, as hearts of the quotidian religious activity, are usually preeminent elements in the arrangement of settlements and cities.Not only were they the place for the congregation of believers, but also means of monumental display for the benefit of ruling authorities.Together with this, the non-confessional use that they received from the inhabitants also constituted an important part for their raison d'être.
Such a concurrence of factors has been frequently derived in the affirmation that there is no Islamic city without mosques.However, whereas the notion of an 'Islamic city' has been rethought and reformulated in recent decades (see, for instance, Abu Lughod 1987;Behrens-Abouseif 2000;Neglia 2008;or Brogiolo 2011), the term 'mosque' has remained unreviewed.Mudun are no longer conceived as non-planned sites, produced by a chaotic society, but as organic living realities that change according to the needs and ways of life of those who inhabit and build them.In other words, the Muslim city is not defined through its morphological characteristics any more, but through the functions it performs along time.
In stark contrast, mosques have barely been explained beyond their formal aspects.This is quite contradictory, given that the Quran refers to a 'mosque' as the place for prayer for the Islamic community, without stipulating specific architectural characteristics apart from the orientation towards Mecca (The Quran, II, 144-150;Grabar 1979, p. 119).In fact, from the very beginning of the Islamic Arts 2018, 7, 63 2 of 17 expansion, the architectural forms of these buildings were multiplied according to diverse building traditions and to the personalities of the different regions.Because of this formal variety, there have been several intents to classify existing mosques in certain typological groups.
One classification is suggested in The Encyclopedia of the Quran, where mosques are arranged according to their morphological characteristics and geographical locations. 1The Encyclopedia of Islam also systematizes mosques, but considers the reasons that promoted their foundation. 2All of these criteria allow a primary organization of mosques despite their strong architectural plurality, but this current overview is insufficient to understand all their functions in urban scenarios or in Muslim ideology. 3The panorama becomes more complex when it comes to the explanation of what a mosque is beyond its material form.The Encyclopedia of Islam (Bosworth 2010) defines "mosque" as the place where Muslims prostrate during the religious service.This comes from the etymological origin of the word mosque-in Arabic masŷid or masgid, plural masāŷid-which derives from the verb sadjada, that is to say, to prostrate.Thus, 'mosque' alludes to the ritual followed by the faithful when they perform the prayer: standing in parallel rows, believers execute a series of positions that culminate in the total prostration as a symbol of submission to Allah.This explanation 4 is mainly focusing on the religious facet of mosques and defining them as a place for worship, but disregards many others aspects.It has been also stated that at least at the beginning of the Islamic conquest mosques were built as territorial landmarks as well, located in strategic spots where they could show the presence of the Islam in the conquered lands (Calvo 2014, p. 28).Mosques were designed to represent and maintain the memory of certain historical episodes or people too, and they were sometimes linked to the policies of Arabization and Islamization of the new territories (Calvo 2014, p. 41).
Together with this, at least during the medieval centuries, rulers used masāŷid as propagandistic scenarios to launch certain messages.These large architectural works served to establish signs of political legitimacy too, through carefully designed architectural and decorative programs.First in the Syrian area, the appearance of mosques was progressively regularized through a "direct imperial patronage" (Ettinghausen and Grabar 1987, p. 54).Later, similar processes will be undertaken in al-Andalus (Almagro 2001;Juez 1999, pp. 86-ff.).In this region, the great mosque of Córdoba is considered the best example of monumental religious architecture to the service of political legitimization and propaganda.
Apart from these official intentions, mosques were also modeled by the normal and daily use that the inhabitants of the cities gave to them.Regarding this, it is well documented that, at least in al-Andalus, the attendance of Muslims to mosques was not only to perform the daily prayers there, but also to solve other issues, such as those related to justice or neighborhood conflicts (see Juez 1999, pp. 116-17;Peláez 2000;Al-Jusani 2005).This social interaction turned mosques into a stage in which the pretensions of the power and the habits of the daily life concurred.

Mas ā ŷid in Al-Andalus: Friday Mosques and Other Examples
The Encyclopedia of Islam points out that mosques which housed the noon Friday prayer could receive the rank of major or "Friday mosques" (Bosworth 2010, pp. 325-26).Conversely to daily prayers, which could be practiced almost anywhere (Hillenbrand 1994, p. 31), Friday prayer in the Arts 2018, 7, 63 3 of 17 mosque was of compulsory assistance for all adult male belonging to highest classes.It was not only a religious duty, but an official instrument through which inhabitants expressed their community membership and their adhesion and support to certain political regimes.This is the reason why many main mosques became a clear sign of the authority and position of the ruling dynasty.In the case of al-Andalus, the Umayyads were the first encouragers of the Islamic faith.Since they also had the need to legitimate their commanding position and to manifest their building capacity, they became the highest promoters of these masāŷid al-Jami' (Juez 1999, p. 86). 5 In these buildings, they could arrange official and political ceremonies, as well as to check adhesion of the umma to the regime (see Juez 1999;Calvo 2014, pp. 62-ff.).As a consequence, these facilities showed great monumentality and artistic quality which were interpreted by the believers as the material expression of the unity of the community, and by the non-believers as a manifest sign of the hegemony of Islam.
Their eye-catching and attractive formal characteristics turned mosques, in the opinion of P. Wheatley and especially between the 7th and 10th centuries, into the most striking structures on the Islamic cities6 (Wheatley 2001, p. 231).Because of this, Friday mosques in urban areas have usually received the most of the attention in research, which has tended to leave other minor prayer spaces aside.However, those monumental features are not perceptible in all andalusí cases nowadays, often due to the frequent transformation of mosques into later churches.The case of Jaén may be very illustrative.Only after a hard architectural study in the current Iglesia de la Magdalena it was possible to recognize the Islamic mosque that remains underneath (first Berges (2007), recently reviewed by Rütenik (2017, pp. 249-57).In some other cities, basic data such as the location and fundamental structure of the main mosque are still being the subject of scientific discussion (in Granada, for example, (Torres Balbás 1945;Fernández Puertas 2004); or Toledo, where only the location of its main mosque is barely known).
In the cases mentioned above, readjustments to the Christian cult and the continued use produced strong modifications in the original structures of these buildings, which are nowadays difficult to perceive.Nevertheless, in some other cases several fundamental traces can be still identified in spite of the Christian modifications.At this respect, the Friday mosque of Zaragoza, currently called Seo de San Salvador, is a highlighted example.For a long time, the mosque was only approachable through written information, and barely recognized in some reused structures, such as its minaret (Souto 1989;Almagro 1993).However, the use of techniques related to the archaeology of architecture has led to the documentation of other archaeological traces, allowing a more complete formal and typological knowledge of the old mosque (Souto 1993a(Souto , 1993b;;Hernández Vera 2004).
Similar situations can be observed in Huesca (Carrero 2005); Tudela (Gómez Moreno 1945), Málaga (González Sánchez 1996), Mértola (Torres Balbás 1955;Ewert 1973;Macías 2006;Macías et al. 2011), Carmona (Anglada et al. 2017;Jiménez Hernández 2018) or Sevilla see (Valor 1993) or (Cómez 1994) for the Ibn Adabbas mosque, (Jiménez Sancho 2016) for the Almohad mosque), where Friday mosques are progressively being discovered or better documented.In addition, in this last city the Almohad mosque has been addressed from new viewpoints, in regard to its urban insertion and its impact in the nearest urban fabric (Almagro 2007;Arévalo 2011;Jiménez Sancho 2016). 7he variability in the better or worse knowledge of andalusí Friday mosques becomes clearer when it comes to the great mosque of Córdoba.This complex is considered by many scholars the most emblematic building of those erected in al-Andalus due to its forms, functions and meanings (Souto 2009, p. 9).Its significance, together with its state of conservation, has produced great fascination among researchers and other authors, who have widely written about it.As a consequence, Córdoba's mosque has received plenty of attention, materialized in a wide range of existing bibliography of different nature. 8 While research has been traditionally concentrating more on these principal mosques, for which there is a considerable amount of literature available, less monumental ones have barely appeared in the scientific discussion.Recent historiography has popularized the names 'neighborhood', 'secondary', or 'minor' mosques for these oratories, which were built all along the cities and which do not seem to fit properly in the typologies provided by the Encyclopedia of Islam (see above).
Written sources transmit first order information about the juridical and legal details of these foundations-internal maintenance, methods of financing, and economic sustainability, among other aspects-can be studied through the documents related to awqāf.The works by García Sanjuán (2002) and Carballeira (2002) have shed new light on this pious institution, and on how these foundations were engaged with the social life of cities' inhabitants.In this respect, texts reveal how these smaller mosques were places for justice, social gathering, teaching, for hosting different social ceremonies and, even in certain cases, scenarios for funerary activities as well (see some examples in González Gutiérrez 2016a, pp.419-ff.).
Unfortunately, these textual data are too undetailed to allow approaches to the formal aspect and the functions performed by these buildings.Written sources rarely transmit enough information about the appearance of these buildings, but only bare mentions, without descriptions, related to further narration or other events.In the best of the cases, some inaccurate counts have been made. 9Texts' inaccuracy turns archaeological information into an indispensable source to reach a better knowledge about mosques.Nevertheless, archaeological data may also be hard to process frequently due to insufficient registration, the poor state of conservation of the remains found, or the difficulties to identify them as mosques.Because of this, in the former urban territory of al-Andalus there are still many mosques to be detected.
Regarding Seville, for instance, it has been conventionally assumed that medieval mosques played a very important role in the arrangement of the later Christian parish districts in the city, even though there is very little archaeological news of other Islamic religious buildings rather than the principal mosques (Valor 2008).Difficulties when it comes to the identification of mosques-which is not exclusive for the Sevillian area-are too often related to their almost total remodeling after the Castilian conquest, and to the scarcity of adequate urban archaeological interventions to document them.The frequent lack of publications derived from the few interventions made should be added to this problem (Valor 2008, pp. 142-45).
Jerez de la Frontera may be taken as another example.Christian sources written after the Conquest register the existence of 18 mosques in the city-some of them turned into parish churches-not registered by archaeology until now (Calvo 2014, pp. 308-9)  10 .Some minor mosques have been located in former cities, currently abandoned, such as in Madīnat al-Zahrā' (Vallejo 2009) or in Vascos, Toledo (Izquierdo andPrieto 1993-1994).There is also evidence of other religious spaces different from Friday mosques located in citadels or palaces, instead of in neighborhoods.These are the cases of the small mosques excavated in the citadels of Vascos (Izquierdo andPrieto 1993-1994;Izquierdo 1990;Izquierdo and Ares 2004), Murcia (Sánchez and García 2007), Badajoz (Valdés 1999), the Aljafería 8 Publications on the great mosque of Córdoba are countless, and offer different approaches in terms of focus, depth and quality.Without the will to be exhaustive, the works of Terrasse (1932); Hernández (1975); Stern (1976); Creswell (1979); Papadopoulo (1977); Grabar (1979) alone and with Ettinghausen and Grabar (1987); Ewert (1987Ewert ( , 1995)); Cressier (1984Cressier ( , 1985) ) or Hillenbrand (2010), among many others, should be always regarded.9 There are several counts for Córdoba (see González Gutiérrez 2012, pp.51-ff.)which reflect an excessive number of mosques for the city.These exaggerations have been interpreted as the intent to show the great monumentality of the city made by chroniclers, as well as to reflect the importance that mosques had for its imperial image. 10The remains of a former mosque were identified at the Alcazaba, but it was not corresponding to a secondary or neighborhood mosque, open to the majority of the population (Calvo 2014, pp. 546-51).
Arts 2018, 7, 63 5 of 17 palace in Zaragoza (Borrás and Cabañero 2012;Cabañero 2007), and probably in the citadel of Merida as well (Féijoo and Alba 2002),11 among others (Calvo 2014, pp. 445-ff.).These locations point towards a private or restricted use of these spaces by specific collectives rather than by the majority of the population, and it is also confirming the existence of places for prayer different from Friday mosques or from neighborhood mosques.The clearest example may be the renowned Cristo de la Luz mosque at Toledo (Calvo 1999;Jurado Jiménez 2000;or Romero 2006).Even though it was not strictly located at the Alcázar enclosure, its constructive characteristics point towards a possible private use as well.This infrequent discovery of other smaller mosques in al-Andalus urban areas has, fortunately, two very important counterpoints.The first of them is Toledo, where many smaller places of prayer have been identified so far, exhaustively compiled by Romero (2006).In addition, the application of innovative techniques when analyzing these buildings' elevations is leading to suggest diachronic readings of their evolution and, in many cases, subsequent conversion into churches.The main example in this regard is the work of Rütenik (2008Rütenik ( , 2017)), who has raised the number of mosques documented in the city to almost a dozen (see Figure 1).12 the cases of the small mosques excavated in the citadels of Vascos (Izquierdo andPrieto 1993-1994;Izquierdo 1990; Izquierdo and Ares 2004), Murcia (Sánchez and García 2007), Badajoz (Valdés 1999), the Aljafería palace in Zaragoza (Borrás and Cabañero 2012;Cabañero 2007), and probably in the citadel of Merida as well (Féijoo and Alba 2002), 11 among others (Calvo 2014, pp. 445-ff.).These locations point towards a private or restricted use of these spaces by specific collectives rather than by the majority of the population, and it is also confirming the existence of places for prayer different from Friday mosques or from neighborhood mosques.The clearest example may be the renowned Cristo de la Luz mosque at Toledo (Calvo 1999;Jurado Jiménez 2000;or Romero 2006).Even though it was not strictly located at the Alcázar enclosure, its constructive characteristics point towards a possible private use as well.
This infrequent discovery of other smaller mosques in al-Andalus urban areas has, fortunately, two very important counterpoints.The first of them is Toledo, where many smaller places of prayer have been identified so far, exhaustively compiled by A. Romero (2006).In addition, the application of innovative techniques when analyzing these buildings' elevations is leading to suggest diachronic readings of their evolution and, in many cases, subsequent conversion into churches.The main example in this regard is the work of T. Rütenik (2008Rütenik ( , 2017)), who has raised the number of mosques documented in the city to almost a dozen (see Figure 1). 12The second one is Córdoba, where, from the end of the 20th century, urban archeology is revealing numerous archaeological remains of former secondary mosques, inserted into very wealthy urban contexts.

11
This study is also suggesting a possible use of this mosque as a signal tower.12 However, a study which tries to analyze the role of these mosques in the urban landscape still remains to be conducted.The second one is Córdoba, where, from the end of the 20th century, urban archeology is revealing numerous archaeological remains of former secondary mosques, inserted into very wealthy urban contexts.

The Case Study of Córdoba: Secondary Mosques for Al-Andalus
Major archaeological evidence about neighborhood mosques in Córdoba did not arise until the end of the 1990s.However, and despite the vast shadow casted by the Friday mosque, from Arts 2018, 7, 63 6 of 17 the beginning of that same century some researchers began to raise questions about the possible presence of other simpler mosques in the landscape of the city.We are referring to authors such as Castejón (1929) or Lévi-Provençal (1932, etcetera), who aimed to recreate the image of the caliphal city in the 10th century.To do so, they combined the information contained in written sources with the observation of the heritage remains preserved in the present city. 13Approaches to the surviving Islamic remains were scarce but of great value.Among them, we should highlight the efforts made by Escribano (Escribano 1964(Escribano -1965)), Pavón (1976), Golvin (1979) and, most of all, Hernández (1975), who intended to approach the basic structure and organization of former mosques.Hernández also tried to distinguish different minaret typologies, which led him to suggest a possible Cordobesian building school and style tendencies depending on chronological moments.However, further conclusions should wait, according to Hernández, pending further archaeological evidence.
This state of the arts had to wait until the second half of the 20th century and the first years of the 2000, when Córdoba experienced a rapid increase of real estate activities which provoked the progressive and quick exhumation of enormous areas of the Islamic city.These excavations, mainly developed in areas located outside the medieval walled enclosure, meant the discovery of archaeological evidence of great quantity and quality.Among others, multiple vestiges interpreted as the remains of mosques were uncovered, which were rapidly compiled into a specific article (López Guerrero and Valdivieso 2001) 14 and progressively incorporated into the urban-diachronic and landscape studies of the Islamic city.Acién and Vallejo (1998), Murillo et al. (2004), Casal et al. (2006) or Vaquerizo and Murillo (2010) published works which rose awareness about these elements, often disregarded when speaking about the urbanism and landscape of Madinat Qurtuba.
The historiographical precedents and the availability of this new archaeological information motivated our study about secondary mosques in Islamic Córdoba.Our main aim was to try to give answer to a whole series of questions related to the appearance, function, and continuous evolution of the city: how were these mosques structured and organized, and why?What were the typologies?How did these mosques act in the Islamizing work promoted by the Umayyads?What urban and social transformations aroused in the neighborhoods in which they were founded?etcetera. 15 For achieving this, the study of these buildings per se was required first.This involved the identification of all the former mosques and the analysis of their structure and general arrangements, typologies, and layouts, and their chronological identification as well.Despite the number of mosques identified in Córdoba at present is lesser than the figures transmitted by written sources (see González Gutiérrez 2014, pp.298-99), up to date scarcely a dozen of complexes have been identified as former secondary mosques (see Figure 2).Their states of documentation and conservation are, besides, quite uneven. 13Among them, they included some minarets reused as belfries after the Christian conquest, which still are a popular part of Córdoba's heritage and cultural landscape (González Gutiérrez 2017). 14This compilation article has been, for many years, the only publication available on this specific subject.However, even though discoveries continued to emerge from 2001 onwards, no author considered necessary to review or enlarge this contribution.Despite that, it continued to raise awareness not only about how these buildings were, but about their role in the city and their importance in people's lives 15 First through our master's thesis (González Gutiérrez 2012), and finally with our doctoral work (González Gutiérrez 2016a).There are still standing remains of some of these mosques, specifically several minarets which were transformed into bell towers.Even though most of them have been recently restored and, therefore, the observation of the medieval Islamic techniques becomes hard, there are still some examples in which the traditional wall-reading techniques can be applied.It is the case of the minaret reused as bell tower in the church of San Lorenzo, whose chronological evolution has been reinterpreted through the architectural analysis of its visible faces (González Gutiérrez 2016a, pp.293-96).Its possible foundational plaque states that the minaret was erected or refurbished in caliphal times, during the rule of al-Hakam II.Nevertheless, recent research has suggested a possible emiral origin of the tower (Marfil 2010, p. 54;De Vicente 2014, p. 27).Our revision on these readings, put in relation with other andalusí examples, has brought different results16 that are summarized in Figure 3.
These analyses could be only applied in very specific cases, because usually mosques are documented just partially.There is also a very frequent lack of other architectural or ornamental material, as well as a shortage of accurate stratigraphic information in many cases.Given these conditions, an approach to their structure and architectural design should be made through new methodological techniques.We proposed a metrological and modulation study tested in four mosques documented in Córdoba (further methodology and results at (González Gutiérrez 2017)).Results obtained suggest the possible existence of a building scheme for caliphal mosques, which consisted of a squared prayer hall and a quadrangular patio on the west side, equal or slightly bigger than the haram's square (see Figure 4).This type of masŷid roughly reproduces the scheme established by the Great Mosque during the Caliphate, which consists of two equal squares as well.

16
The detailed observation of the face of the minaret in which the window is still conserved has been put in relation with the study of other pieces.Ashlars used in Córdoba can be compared with others used to build several emiral minarets in al-Andalus, such as San Juan or Santiago, both in Córdoba, or even the minaret of Ibn 'Adabbas in Seville or El Salvador in Toledo.Other elements taken into account are related with the typologies and dimensions of the windows.There are still standing remains of some of these mosques, specifically several minarets which were transformed into bell towers.Even though most of them have been recently restored and, therefore, the observation of the medieval Islamic techniques becomes hard, there are still some examples in which the traditional wall-reading techniques can be applied.It is the case of the minaret reused as bell tower in the church of San Lorenzo, whose chronological evolution has been reinterpreted through the architectural analysis of its visible faces (González Gutiérrez 2016a, pp.293-96).Its possible foundational plaque states that the minaret was erected or refurbished in caliphal times, during the rule of al-Hakam II.Nevertheless, recent research has suggested a possible emiral origin of the tower (Marfil 2010, p. 54;De Vicente 2014, p. 27).Our revision on these readings, put in relation with other andalusí examples, has brought different results 16 that are summarized in Figure 3.
These analyses could be only applied in very specific cases, because usually mosques are documented just partially.There is also a very frequent lack of other architectural or ornamental material, as well as a shortage of accurate stratigraphic information in many cases.Given these conditions, an approach to their structure and architectural design should be made through new methodological techniques.We proposed a metrological and modulation study tested in four mosques documented in Córdoba (further methodology and results at (González Gutiérrez 2017)).Results obtained suggest the possible existence of a building scheme for caliphal mosques, which consisted of a squared prayer hall and a quadrangular patio on the west side, equal or slightly bigger than the haram's square (see Figure 4).This type of masŷid roughly reproduces the scheme established by the Great Mosque during the Caliphate, which consists of two equal squares as well.We have not managed to calculate overall proportions that determined the general dimensions chosen for each building.It is probable that the size of these mosques depended more on the conditions of the plots where they were erected.Given that three of the four mosques analyzed were caliphal, it has not been possible to suggest a wider chrono-typology beyond the limits of the caliphate.Thus, new archaeological evidence is required to complete our vision and clarify our hypotheses.We have not managed to calculate overall proportions that determined the general dimensions chosen for each building.It is probable that the size of these mosques depended more on the conditions of the plots where they were erected.Given that three of the four mosques analyzed were caliphal, it has not been possible to suggest a wider chrono-typology beyond the limits of the caliphate.Thus, new archaeological evidence is required to complete our vision and clarify our hypotheses.We have not managed to calculate overall proportions that determined the general dimensions chosen for each building.It is probable that the size of these mosques depended more on the conditions of the plots where they were erected.Given that three of the four mosques analyzed were caliphal, it has not been possible to suggest a wider chrono-typology beyond the limits of the caliphate.Thus, new archaeological evidence is required to complete our vision and clarify our hypotheses.Although the Qu'ran, as we said at the introduction, does not dictate rules that determine the morphological appearance of masāŷid beyond their orientation towards Mecca, Cordobesian mosques soon defined their basic structure around two main parts.Both, prayer hall and courtyard, had apparently simple features, but they were also very easy to combine and model.They were, in short, highly elastic and plastic spaces that could be built almost anywhere.The type of caliphal oratory identified for Córdoba is flexible and its design could be changed according to the plots or spaces where the mosques would be built.This plasticity is not confined to Córdoba.In al-Andalus, many 'transgressions' of this common scheme are detected, probably favored by the absence of quranic guidelines.Thus, we have mosques whose mihrabs are not inserted in the center of the qiblas, as in the mosque of Córdoba, or the citadel of Badajoz; apparently random ground plans as the one drawn by the mosque in citadel of Vascos, or the generalized differential position of minarets.This adaptation ability also means that there were no strict buildings rules that can be extrapolated for all al-Andalus, and it is of paramount importance because it turned mosques into one of the most versatile and multifunctional buildings in the Islamic city.
Concerning their orientations, and despite being this the only characteristic imposed by the Quran, it has been revealed to be imprecise in all cases.Its accuracy, which is being subject of intense debate in historiography, did not seem to have capital relevance in Córdoba.Directions displayed by qiblas in Madinat Qurtuba were unrelated to their chronologies, situation repeated in the rest of al-Andalus; nor was a predominating trend over others.Our recent analyses, summarized in Table 1, suggest that the degree of accuracy of Cordobesian mosques' orientation was more influenced by topography and by the inherited road networks (González Gutiérrez 2016a, pp.454-ff.), as is the case in certain Moroccan cities (Bonine 1990).The deepening in the structural knowledge of Madīnat Qurt .uba's masāŷid led to approach the forms and lifestyles of the society that used them, as well as to a better understanding of the associated urban phenomena.The interpretation of information contained in written sources in the light of archaeological data confirms the functional ambivalence of mosques.Far from being places exclusively dedicated to prayer and worship, several texts speak of other activities such as the administration of justice, the education in specific disciplines or the development of some commercial operations.All this turned them into true centers for social meeting.Their material relevance has been confirmed through their spatial analysis from a diachronic perspective.As we will see, mosques were generally located in close relation with historic paths, major roads and the gates of the city.Sometimes, their topographic prominence has also been verified.Documentary sources register numerous mosques' foundations during the Emirate.Promoted by the sovereign and his closest social circle, they played an important role in the emergence and development of the earliest neighborhoods.17Archaeology has not been able to find all these examples, but has noted the existence of some whose location does not seem casual, but the result of a pre-planned project.We are referring to mosques close to the walled city's gates, or placed on entity roads leading to it.They may respond to the Umayyad government's desire to provide Madīnat Qurtuba with the needed infrastructure to develop very specific ways of life, and to create a determined landscape (González Gutiérrez 2016b, pp.280-84).Perhaps the clearest example of the Islamization and proselytism policy developed by the Umayyad dynasty is the mosque of Santiago (see Figure 5).Probably built at the request of the emir Hišām I along a road with Roman origins, near the Bāb al-H .adīd, it was located in a neighborhood already populated during the Late Antiquity, very close to a dhimmi cemetery.Thus, its erection would have been conceived to encourage urban dynamics and Islamization in a vernacular area (González Gutiérrez 2016b, pp. 280-84).
Arts 2018, 7, x FOR PEER REVIEW 10 of 17 development of the earliest neighborhoods. 17Archaeology has not been able to find all these examples, but has noted the existence of some whose location does not seem casual, but the result of a pre-planned project.We are referring to mosques close to the walled city's gates, or placed on entity roads leading to it.They may respond to the Umayyad government's desire to provide Madīnat Qurtuba with the needed infrastructure to develop very specific ways of life, and to create a determined landscape (González Gutiérrez 2016b, pp.280-84).Perhaps the clearest example of the Islamization and proselytism policy developed by the Umayyad dynasty is the mosque of Santiago (see Figure 5).Probably built at the request of the emir Hišām I along a road with Roman origins, near the Bāb al-Ḥadīd, it was located in a neighborhood already populated during the Late Antiquity, very close to a dhimmi cemetery.Thus, its erection would have been conceived to encourage urban dynamics and Islamization in a vernacular area (González Gutiérrez 2016b, pp.280-84).The cited patterns were respected during the Caliphate, in which mosques continued to occupy a prominent place in road networks and neighborhoods.The ultimate expression of the latter is documented in the western suburbs.Here, there were mosques built before the rest of the infrastructure, as it is the case of the mosque in Fontanar (see Figure 6a).Placed at the confluence of the neighborhood's main streets, this area probably experienced dense traffic of people and activities.Next to it, another masŷid was built in the most topographically prominent spot of the suburb (mosque in the Centro de Transfusión, see Figure 6d, (Sánchez Madrid 2005)).In addition, some sectors already emerged during the Emirate will be now provided with new infrastructure, among which mosques can be detected (mosque of Naves de Fontanar, see Figure 6c).Emiral mosques continued to be active, and even some of them were renovated or enlarged.The continuation of the Islamization policy initiated then will be continued now as well.It can be detected, for instance, in 17 On many occasions, they were erected as pious foundations.There are many written testimonies of relatives of the Emirs or the Caliphs, or by members belonging to the Court, defraying these constructions.This was made to encourage the social support to the Umayyads' regime and to increase the policies of Islamization.The cited patterns were respected during the Caliphate, in which mosques continued to occupy a prominent place in road networks and neighborhoods.The ultimate expression of the latter is documented in the western suburbs.Here, there were mosques built before the rest of the infrastructure, as it is the case of the mosque in Fontanar (see Figure 6a).Placed at the confluence of the neighborhood's main streets, this area probably experienced dense traffic of people and activities.Next to it, another masŷid was built in the most topographically prominent spot of the suburb (mosque in the Centro de Transfusión, see Figure 6d, (Sánchez Madrid 2005)).In addition, some sectors already emerged during the Emirate will be now provided with new infrastructure, among which mosques can be detected (mosque of Naves de Fontanar, see Figure 6c).Emiral mosques continued to be active, and even some of them were renovated or enlarged.The continuation of the Islamization policy initiated then will be continued now as well.It can be detected, for instance, in the Christian suburb located north of the walled perimeter, which was provided with a mosque in the 10th century.The introduction of this building brought the urban regression of the Christian areas in benefit of the new Islamic suburb, which experienced a notable growth from this moment on.
The fall of the Caliphate meant the destruction and abandonment of Western mosques, as well as few new foundations.These were more modest than the complexes projected during the Umayyad centuries (see Figure 7).At the same time, the ones that remained in the Medina and the Axerquía do not show significant changes in this period.Their functional continuity along the centuries is more related to the symbolic, emblematic, and territorial meaning that they acquired in each of the districts than to their architectural characteristics.These ideological implications in the urban fabric probably also favored the transformation of some of these mosques in churches after the Christian conquest in the 13th century.Some of them, in certain cases emerged in emiral times, reached an extraordinary longevity that might have depended on their condition of pious foundation.However, this statement does not mean that all the late-Islamic mosques were transformed into churches.Such a transformation was fruit of a different cultural group that transformed the city according to new needs and stimuli.Therefore they must be addressed and studied from other approaches and objectives henceforth (González Gutiérrez 2016b, pp. 285-87).
Arts 2018, 7, x FOR PEER REVIEW 11 of 17 the Christian suburb located north of the walled perimeter, which was provided with a mosque in the 10th century.The introduction of this building brought the urban regression of the Christian areas in benefit of the new Islamic suburb, which experienced a notable growth from this moment on.
The fall of the Caliphate meant the destruction and abandonment of Western mosques, as well as few new foundations.These were more modest than the complexes projected during the Umayyad centuries (see Figure 7).At the same time, the ones that remained in the Medina and the Axerquía do not show significant changes in this period.Their functional continuity along the centuries is more related to the symbolic, emblematic, and territorial meaning that they acquired in each of the districts than to their architectural characteristics.These ideological implications in the urban fabric probably also favored the transformation of some of these mosques in churches after the Christian conquest in the 13th century.Some of them, in certain cases emerged in emiral times, reached an extraordinary longevity that might have depended on their condition of pious foundation.However, this statement does not mean that all the late-Islamic mosques were transformed into churches.Such a transformation was fruit of a different cultural group that transformed the city according to new needs and stimuli.Therefore they must be addressed and studied from other approaches and objectives henceforth (González Gutiérrez 2016b, pp.285-87).

Final Remarks
We have shown how research on religious urban spaces in al-Andalus has been concentrating mainly on Friday mosques.Among all of them, the Great Mosque of Córdoba has stood out.The sponsors of these mosques were commonly the rulers, who legitimized and displayed the power of their dynasties through them.However, there were other smaller mosques, usually disregarded by research, which had a multifunctional nature and were key to understanding the arrangement of cities and their meaning in the Umayyad building strategies.The great variety of these mosques in terms of formal appearance and, more specifically, in terms of functions performed, does not fit in the existing typologies.
Because of this, we suggest a new classification in which the andalusí examples can be arranged more accurately.This classification, which prioritizes functions over architectural appearance, is only composed of two main groups.The first one is for major or Friday mosques, that is to say, the ones that congregate the Muslim community for the noon Friday prayer.The second one includes all the so-called secondary mosques, in different categories: neighborhood mosques, which were at the general service of all the inhabitants; private masāŷid, only for the use of specific groups; and others, such as funerary mosques, for more delimited uses.This classification is flexible and, therefore, able to be extended and enriched from now on in the light of future discoveries.

Final Remarks
We have shown how research on religious urban spaces in al-Andalus has been concentrating mainly on Friday mosques.Among all of them, the Great Mosque of Córdoba has stood out.The sponsors of these mosques were commonly the rulers, who legitimized and displayed the power of their dynasties through them.However, there were other smaller mosques, usually disregarded by research, which had a multifunctional nature and were key to understanding the arrangement of cities and their meaning in the Umayyad building strategies.The great variety of these mosques in terms of formal appearance and, more specifically, in terms of functions performed, does not fit in the existing typologies.
Because of this, we suggest a new classification in which the andalusí examples can be arranged more accurately.This classification, which prioritizes functions over architectural appearance, is only composed of two main groups.The first one is for major or Friday mosques, that is to say, the ones that congregate the Muslim community for the noon Friday prayer.The second one includes all the so-called secondary mosques, in different categories: neighborhood mosques, which were at the general service of all the inhabitants; private masāŷid, only for the use of specific groups; and others, such as funerary mosques, for more delimited uses.This classification is flexible and, therefore, able to be extended and enriched from now on in the light of future discoveries.
Surely, the number of mosques that existed in al-Andalus was far higher than the figures documented by archaeology to date.However, the cases located in Córdoba and Toledo show that the general unawareness about mosques and their importance in the urban fabric is not due to researchers' apathy or to the lack of buildings to be analyzed, but to the need of new viewpoints to be applied.Remains of many former masāŷid are still to be uncovered, but this task requires new approaches that are able to go beyond architectural styles, and landscape analyses to help us understand the role performed by mosques in the cities. 18  The erection of mosques in the Umayyad capital was directly related to the development of Islamization policies.They also served to manifest the presence and hegemony of the Umayadd dynasty.Their locations were usually chosen to encourage the development and growth of new Islamic neighborhoods, especially in areas far from the walled perimeter.Understanding mosques in their urban context has been, indeed, key to comprehending the historical processes occurred in Córdoba during the Islamic stage.
Because all of these reasons, we believe that developing these analyses in other Islamic cities, even beyond the al-Andalus area, will be surely fruitful for a better knowledge of mosques, but also of cities themselves.Studying mosques addressing all their multi-functional aspects and their impact in the urban fabric is crucial to transcending the historicist and descriptive vision of the Islamic city and its elements, and also encourages understanding urbanism from an evolutionary and functional perspective.We are not able to determine whether the urban insertion of Qurt .uba's mosques constituted a pattern for other Muslim settlements, but this study case may be a reference for other cities, in which the diachronic study of urban processes, with mosques as the focal point, should be undertaken.
Consequently, we hope that our results help to stimulate research on mosques not as an isolated architectural element, but as key pieces to define Islamic cities and the urban dynamics associated with them.Since mosques are the unmistakable symbol of Islam, they should have a privileged position in urban, global, and diachronic analyses of Medieval Islamic urbanism.

Figure 1 .
Figure 1.Plan of Toledo with the location of mosques identified in current churches (Used by permission of T. Rütenik, with modifications.The original can be found at Rütenik 2008, p. 30).

Figure 1 .
Figure 1.Plan of Toledo with the location of mosques identified in current churches (Used by permission of T. Rütenik, with modifications.The original can be found at Rütenik 2008, p. 30).

Figure 2 .
Figure 2. Dispersion map of mosques documented in Córdoba through archaeological methodologies (Created by author).

Figure 2 .
Figure 2. Dispersion map of mosques documented in Córdoba through archaeological methodologies (Created by author).

Figure 3 .
Figure 3. Drawing of the southern face of San Lorenzo minaret (basis from Marfil 2010, p. 55, lám.5, used with his permission.Colors are added by author according to the different interpretations given to the minaret).To the left side, interpretations by De Vicente (2014, p. 145) and Marfil (2010, p. 55).To the right side, author's own interpretation.

Figure 4 .
Figure 4. Caliphal mosques in Córdoba.Their plant consists on a squared prayer room, followed by an also squared courtyard often provided with a minaret (Created by Author).

Figure 3 .
Figure 3. Drawing of the southern face of San Lorenzo minaret (basis from Marfil 2010, p. 55, lám.5, used with his permission.Colors are added by author according to the different interpretations given to the minaret).To the left side, interpretations by De Vicente (2014, p. 145) and Marfil (2010, p. 55).To the right side, author's own interpretation.

Figure 3 .
Figure 3. Drawing of the southern face of San Lorenzo minaret (basis from Marfil 2010, p. 55, lám.5, used with his permission.Colors are added by author according to the different interpretations given to the minaret).To the left side, interpretations by De Vicente (2014, p. 145) and Marfil (2010, p. 55).To the right side, author's own interpretation.

Figure 4 .
Figure 4. Caliphal mosques in Córdoba.Their plant consists on a squared prayer room, followed by an also squared courtyard often provided with a minaret (Created by Author).

Figure 4 .
Figure 4. Caliphal mosques in Córdoba.Their plant consists on a squared prayer room, followed by an also squared courtyard often provided with a minaret (Created by Author).

Figure 5 .
Figure 5. Location of the mosque currently visible in the church of Santiago (black star) in relation to the historical road (in black) and the Friday mosque (red star) (Created by author over the plan by Vaquerizo and Murillo 2010, p. 529, fig.249.Used by their permission).

Figure 5 .
Figure 5. Location of the mosque currently visible in the church of Santiago (black star) in relation to the historical road (in black) and the Friday mosque (red star) (Created by author over the plan by Vaquerizo and Murillo 2010, p. 529, fig.249.Used by their permission).

Figure 7 .
Figure 7. Mosques built after the fall of the Caliphate.(A) Mosque in Santa Rosa.(B) Mosque in Ollerías neighborhood (picture by author).Basis plan: Vaquerizo and Murillo 2010, p. 707, Figure 369.All images used by the permission of authors.

Figure 7 .
Figure 7. Mosques built after the fall of the Caliphate.(A) Mosque in Santa Rosa.(B) Mosque in Ollerías neighborhood (picture by author).Basis plan: Vaquerizo and Murillo 2010, p. 707, Figure 369.All images used by the permission of authors.

Table 1 .
Mosques documented in Córdoba: relation among their locations, chronologies and qibla orientations.