Introduction

Within the present context of Mediterranean architectural history, a fundamental historiographical issue is the extent to which Anatolia has been defined as a landmass, and hence separated from Mediterranean studies. The land-based way of defining the region has its roots in the early years and decades of the Republic of Turkey, founded in 1923 (Blessing 2023). From the beginning, the Anatolian landmass played a central role in the foundation narrative, from Mustafa Kemal (later Atatürk)’s movements through Anatolia, with congresses held in cities across the region, to the expulsion of Greek forces that culminated in the siege and fire of Izmir in September 1922 (Amygdalou 2020; Özkan 2012). Even the decision to establish a capital in Ankara, at the center of the Anatolian plateau, and to build government buildings from scratch rather than adapting what was there in Istanbul, the political and cultural center of the Ottoman Empire, tied into that land-based narrative (Cengizkan and Cengizkan ed. 2022). And yet, the coastline of Turkey today measures over 8,000 km (5,000 miles), divided among the Mediterranean, the Aegean, the Black Sea, and the Sea of Marmara.¹ Hence, just as much as Anatolia does not equal Turkey in critical reevaluations of recent years, Anatolia should also not equal landmass. In this article, I will first explore the land-based historiographies, before turning to examples of architecture and historical relationships to Anatolia’s seas.

Anatolia’s Seas

Where does the sea (or the many seas) bordering Anatolia appear? In the Republican foundation narrative, it appears in the account of Mustafa Kemal’s victory over Greek forces, which is summarized in the textbook phrase ‘Yunan’ı denize dökmek’, driving the Greeks into the sea. The tragic side of this push is epitomized in the siege of Izmir, where civilians became trapped between the fire and the harbor’s water until evacuation began ten days after the fire had started (Amygdalou 2020: 315–17). The cosmopolitan city of Izmir in the 19th century, and up until the events just described, is also one of the central focal points for the study of Mediterranean narratives on the shores of Anatolia (Zandi-Sayek 2012). Such studies connect to a wide range of studies on Mediterranean port cities such as Marseilles, Beirut, or Alexandria. Once Anatolia is included into narratives of Mediterranean trade, connections to other regions on this shore come into focus, such as ties to Venice and Genoa, both of which had trading posts on the Aegean and Black Sea shores of Anatolia, and to Egypt. Beyond these direct trade connections, Anatolia also emerges as a region that is connected not only overland to Iran, Central Asia, and China through the silkroad(s), but also to the broader Mediterranean basin, and hence finds a place within Mediterranean studies.

For the medieval and early modern periods, the seas of Anatolia appear especially in connection to trade, with extensive work on Genoese merchants in the Black Sea, and the presence of Venetians and Genoese in the Aegean (Fleet 1999; Yenişehirlioğlu 2020; Quirini-Popławski 2023). Other studies have examined the Saljuq presence in the Black Sea with a focus on Crimea and the port cities of Sinop (Peacock 2007; Redford 2014), and Antalya on the southern coast (Preisler-Kappeler 2015: 121–27). Historians have also studied the seafaring role of some of the beyliks of western Anatolia (İnalcık 1985), and the Ottomans’ wharfs, navy (Uzunçarşılı 1984; Isom-Verhaaren 2022), and fortifications along the straits of Gallipoli and the Bosporus (Thys-şenocak 2006; Holmes 2012). In terms of architectural history, however, much remains to be done, especially to understand the relationship between monuments built along the coasts and those located further inland.

As noted above, Anatolia is defined first and foremost as a landmass in much of the writing about it (Blessing and Goshgarian 2017: 4–5). This is true for the trend known as Anadoluculuk (Anatolianism), established in the 1920s and 1930s to emphasize Anatolia as a cultural center reaching back far in time, at least to the Bronze Age; in this narrative, Islam was not a defining feature (Redford 2007). The related Türk Tarih Tezi connected the Turks living in Anatolia to (semi-mythical) Central Asian ancestors, and equated Anatolia to a Turkish homeland defined by ethnicity (Turkishness) rather than religion (Islam) (Akboy-İlk 2023: 4–6; Redford 2007: 243). During the same period, historian Mehmed Fuad Köprülü developed a narrative in which the early Ottoman Empire, and the world of medieval Anatolia from which it emerged, is a center of Turkish and Islamic culture and literature (Köprülü 1935 [1978] and 1919 [2006]). In large part, Köprülü developed this framework to counter Western accounts of the Ottoman Empire as derivative, copying the Byzantine Empire without making any original contributions. Increased emphasis on the Seljuqs beginning in the 1950s developed into a focus on a specific Turkish-Islamic identity (Türk-İslam sentezi), most prominently in the work of historians Osman Turan and İbrahim Kafesoğlu (Strohmeier 1984: 135–60). This too, is a land-based narrative that began with the mythologization of the battle of Manzikert (today Malazgirt) in 1071, a crucial victory of the Seljuqs over the Byzantine army, and culminated in anniversary volumes issued in 1971 (Hillenbrand 2007). Perhaps in an attempt to simplify the complex political and cultural realities of medieval Anatolia, the text-book-ready sequence of Seljuq-beylik-Ottoman was developed (Pancaroğlu 2007). In architectural history, this sequence was picked up, for instance, when scholars such as Oktay Aslanapa and Aptullah Kuran argued that the Green Mosque of İznik, a late 14th-century Ottoman monument, marked the transition from Seljuq to Ottoman architecture (Akboy-İlk 2023: 7; Kuran 1968: 61–63). The building itself, however, offers a much more complex reality in that it also includes stylistic references to Mamluk architecture, similar to those discussed below for beylik architecture in western Anatolia.

In studies over the past two decades, scholars of the history, archaeology, art, and architecture of Anatolia have critically evaluated how such narratives influenced scholarship in their fields beginning in the 1920s. In a special issue of Muqarnas, with the poignant title ‘History and Ideology: Architectural Heritage of the “Lands of Rūm”’, Sibel Bozdoğan and Gülru Necipoğlu gathered articles that examine these questions from a variety of viewpoints (Bozdoğan and Necipoğlu 2007). The volume’s title carries a core concept for such reevaluations: the ‘Lands of Rūm’. The term, much used since, is the English translation of Bilād al-Rūm, the term used for Anatolia in Arabic geographies written when the region was still under the rule of the Byzantine Empire. Rūm means Rome or Roman, mirroring the Byzantine emperors’ use of the term basileus ton Romaion, ruler of the Romans, as an extension of the Roman empire.² The term’s main advantage is that it provides a somewhat more neutral ground for discussion than ‘Anatolia’, with its historiographical ballast and the inherent danger of equating it with the borders of the present-day Republic of Turkey (which differ from those of 1923, notably with the addition of the south-eastern province of Hatay in 1939). For Necipoğlu and Bozdoğan, the term ‘Lands of Rūm’ allows for a critical reevaluation of historiography and cultural geography in an area spanning from the Balkans to Anatolia, without relying on dynastic or national(ist) categories (Bozdoğan and Necipoğlu 2007: 2). In his article in the same volume, Cemal Kafadar examines at great length meanings of the terms Bilād al-Rūm and Rūmī and coins the term Rūmī-ness to define Ottoman self-definitions of identity that were not based on ethnic categories and found their literary expression in Ottoman Turkish (Kafadar 2007: 12–15). Well into the 18th century, the Lands of Rūm in this imaginary extended into areas of the Balkans (Rūm-ili or Rumeli in Ottoman Turkish) under Ottoman Rule (Kafadar 2007: 17–18).

In my own work, I have examined these terms and categories in detail, and analyzed especially how they affected the ways in which the architectural history of central and eastern Anatolia from the 12th to the early 14th century was long framed as exclusively Seljuq (Blessing 2014). But I, too, have so far focused on land and overland connections, with little attention paid to coastal areas. Recently, I extended my studies to the beyliks of western Anatolia as a conduit through which workers from Mamluk Egypt reached the region and their impact on Ottoman architecture in the early 15th century (Blessing 2022: 119–40). Below, I will tie into these examples to develop pathways for a Mediterranean and Aegean framework for the architectural history of medieval and early modern Anatolia.

Coastal Beyliks: From Trade and War to Architecture

Thinking of the Aegean today, two opposite narratives come to mind: the carefree days of beach vacations, and the tragedy of migrants and refugees drowning in attempts at crossing to what they hope will be a better future in Europe. In the Middle Ages, the Aegean coast of Anatolia was a dynamic zone of both war and trade, with Christian and Muslim rulers vying for domination over ports, islands, and straights. The same situation continued in the Ottoman period, with trade and rivalry alternating in the region, as the Ottoman Empire and Venice both worked to expand their possessions. To complicate matters, the Knights Hospitallers, or Knights of Saint John, were also present in the region, with a major citadel in today’s Bodrum (Halikarnassos), in addition to their stronghold on Rhodes (Vatin 1994).

The beyliks were localized principalities ruled by lords (bey in Turkish, hence the term beylik — something belonging to a lord) that emerged in the aftermath of the decline of Saljuq power beginning in the second half of the 13th century. The beyliks emerged and thrived in the frontier lands of western Anatolia — on the edge of the Saljuq zone of influence, and increasingly independent as the Saljuqs became vassals of the Mongol Empire (Peacock 2019: 1–17). Lands were conquered as the Byzantine Empire, newly restored after the end of Latin rule in Constantinople in 1261, shrank. These conquests also happened in the name of gaza, a holy war to expand the lands under Muslim rule (dār al-islām), but, as Cemal Kafadar showed in his work on the emergence of the Ottoman Empire, adventure and profit also played major roles (Kafadar 1995; for a discussion of understandings of dār al-islām and dār al-ḥarb: Özkan 2012: 12–18). In the period under discussion here, the Ottomans ruled just another beylik, centered in northwestern Anatolia, and their first massive and short-lived expansion did not happen until the reign of sultan Bayezid I (r. 1389–1402), stopped short by Timur’s defeat of this ruler in 1402 (Kastritsis 2007). The ruler of Aydın in this period, Umur (d. 1348), was known for his naval exploits, which were later wrapped into an epic by the 15th-century Ottoman poet Enveri (İnalcık 1985; Isom-Verhaaren 2022: 9–28; for the source: Mélikoff 1954)

During the late 14th and early 15th centuries, trade in the Aegean boomed. The increase in trade was closely connected to the lifting of a papal ban with Islamic lands in 1344 (Howard 2000: 16). Even after the ban was lifted, notions of crusading complicated trade relations, as did the presence of Christian military orders such as the Knights Hospitallers (İnalcık 1985). This change allowed the trading powers of Venice and Genoa to establish a presence once more in the Eastern Mediterranean, the Black Sea, and Egypt. Trade connections to Egypt were also strong before the papal ban, especially through the southern Anatolian ports of Antalya, Alanya, and Ayas (Preisler-Kappeler 2015: 124–26).

All this was part of what Johannes Preisler-Kappeler refers to as ‘maritime Asia Minor’ (Preisler-Kappeler 2015). In western Anatolia, ports were revived, and new ports were created in cities, such as Balat-Miletus and Ayasuluk-Ephesus in particular, because those used in antiquity had long been silted up (Caner Yüksel 2019). Thus, the geological and topographical conditions of Anatolia play a central role in this story: major ports were located on the estuaries of rivers, causing issues with harbors that silted up over the centuries (Brückner et al., 2017).

Miletus, for instance, is now an archaeological site located about 10 km from the shore in the estuary of the Büyük Menderes River (Figure 1). In antiquity, however, the city was located on the shore of the Gulf of Miletus, which silted up over time and had nearly disappeared by 1500 CE (Figure 2).³ The location of the İlyas Bey complex, the major beylik monument in the area, was all water as late as the Roman period, when the water line was just west of the Theater (Brückner 2017: 878). A similar situation was in place at the site of Ephesus, where new harbors were created as old ones silted up, and the medieval settlement of Ayasuluk was created at some distance from the antique city (Brückner 2017: 888–89).

Figure 1

Büyük Menderes River Delta near Balat-Miletus. Photograph by Patricia Blessing, 2019.

Figure 2

Büyük Menderes River Delta, geological changes 1500 BCE to 1500 CE, after figure 1 in Brückner et al. (2017: 878).

Further, estuaries and their silting up brought further challenges with the presence of mosquitoes and the diseases they would bear, especially malaria (Brückner 2017: 878; Foss 1979: 175).⁴ The location of the medieval port of Ayasuluk, and its identification with the port named as Scalanova in medieval and early modern sources, is disputed (Pfeiffer-Taş 2014: 1093).⁵ Based on archaeological evidence and historical sources, şule Pfeiffer-Taş has convincingly argued that it could have been located at the site of South Pamucak, about six miles from the historical center of Ephesus (Pfeiffer-Taş 2014: 1096–1102). Clearly, the Venetian and Genoese presence, supported by treaties allowing trade, settlement, and the construction of churches, concluded with the bey of Aydın in 1337 and was renewed by the Ottoman sultan Bayezid I, after the area first came under Ottoman rule in 1389–90 left substantial physical traces (Pfeiffer-Taş 2014: 1089–90). These connections were enhanced by the movements of scholars, such as Hacı Pasha (d. ca 1425), who moved from Cairo to the court of the beys of Aydın in 1370 (Yıldız 2014), and scholars from the Ottoman realm who studied in Cairo and Damascus before returning home, as part of larger intellectual networks that spanned from Egypt to Central Asia (Binbaş 2016).

Within this framework, the architectural patronage of the beys of Aydın and Menteşe, the rulers of Ayasuluk and Balat-Miletus, respectively, was an important conduit for new stylistic trends to be introduced into Anatolia starting in the late fourteenth century. The harbors, full of traders from various Mediterranean port cities, were entry points for those working in the building crafts, from stone masons to designers, allowing for faster travel from, for instance, Egypt to western Anatolia. Eventually, these transfers also reached Ottoman architecture far inland, for instance in Bursa and Amasya in the second decade of the 15th century (Blessing 2022: 99–107 and 121–34).

In the İsa Bey of Ayasuluk-Selçuk (Figure 3), commissioned in 1375, the Mamluk impact is directly apparent in the signature of Ali ibn Mushaymish al-Dimashqi, a building professional whose nisba, ‘the Damascene’, suggests a connection, whether ancestral or direct, to lands under Mamluk rule (Meinecke 1992: vol. 1: 137–38; Blessing 2022: 121–22). The connection to the Mamluk context is apparent in the building itself, which shows strong ties to 14th-century Mamluk buildings, and in many ways would not have looked out of place in Aleppo or Damascus; indeed, the building’s style is thoroughly Mamluk, whereas its contemporaries in Anatolia just tend to exhibit occasional Mamluk elements (Tanman 2011; Tanman 2001). The strongest references to Mamluk architecture in this building appear in the interlocking stonework on the portal and on the mihrab (Figure 4). The stonework on the mihrab and portal was common especially in northern Syria since the Ayyubid period, with one of the most elaborate examples appearing in the Madrasa Firdaus in Aleppo, built in 1235 (Tabbaa 2000).⁶ The closest comparisons to much of the details — the inscription on the portal, especially the chrysanthemums marking its start and end — are found in mid-14th-century Mamluk architecture. Given the chronological gap, there was more than enough time for motif transfer through migration of workers or the exchange of designs (Blessing 2022: 123).

Figure 3

İsa Bey Mosque, 1375, Ayasuluk-Selçuk. Wikimedia Commons © José Luiz Bernardes Ribeiro/ CC BY-SA 3.0.

Figure 4

İsa Bey Mosque, 1375, Ayasuluk-Selçuk, mihrab. Photograph by Patricia Blessing, 2019.

Such Mamluk elements also appear in early Ottoman monuments, as Michael Meinecke notes, for instance in the Yeşil Cami (1378–91) in İznik, and Bayezid I’s mosque-zāviye (1394–95) in Bursa (Meinecke 1992: vol. 1: 141). The use of muqarnas to top windows and portals also appears in 14th-century Mamluk monuments, such as the Mosque of Altinbugha al-Maridani built between 1338 and 1340 in Cairo (Figures 5 and 6; on the building, see Meinecke 1992: vol. 2: 178). At the İsa Bey Mosque, this element is turned into a full window frame composed of muqarnas on all four sides (Figure 7). The same element will appear on the İlyas Bey Mosque (1404) in Balat-Miletus, and on the Mosque-zāviye of the Ottoman sultan Mehmed I (1419–21) in Bursa (Blessing 2022: figures 60, 64–66).

Figure 5

Portal of the Mosque of Altinbugha al-Maridani, 1340, Cairo. Photograph by Patricia Blessing, 2006.

Figure 6

Detail of portal, Mosque of Altinbugha al-Maridani, 1340, Cairo. Photograph by Patricia Blessing, 2006.

Figure 7

Window frame, İsa Bey Mosque, 1375, Ayasuluk-Selçuk. Photograph by Patricia Blessing, 2019.

In Balat-Miletus, the mosque complex of İlyas Bey was completed in 1404; it consisted of a madrasa and two hammams, all of which are in a ruined state, and have been covered with protective roofs in recent restorations (Batuman and Kayhan Elbirlik 2011: 199–264). At the mosque, the Mamluk stylistic impact is strong in the stonework of the portal, where intrinsic geometric patterns are created with accents in different types of grey and reddish stones (Figure 8) (Arel 2011; Tanman 2011). The insets of small pieces of turquoise tile, however, are a new touch. The long band of palmettes (Figure 9) on the same façade also evokes earlier Mamluk examples (see Figure 7). On the mihrab, the alternating elongated and round cartouches form a precedent to an identical pattern appearing on the exterior window frames of the mosque-zāviye of Mehmed I in Bursa, and the calligraphic lunettes over windows (in the interior in the İlyas Bey Mosque and on the exterior in the mosque-zāviye of Mehmed I) are also clearly related (Blessing 2022: 126–27). How these motifs moved is unknown, but clearly, expertise in stonework from Syria and Egypt was available in Anatolia in the early 15th century and spread across the patronage networks of the western Anatolian beyliks and the Ottomans.

Figure 8

View of portal facade, İlyas Bey Mosque, 1404, Balat-Miletus. Photograph by the author, 2019.

Figure 9

Band of palmettes on portal façade, İlyas Bey Mosque, 1404, Balat-Miletus. Photograph by the author, 2019.

Beyond the Beyliks: Anatolian Architecture and the Sea

What of the architectural presence of the Venetians and Genoese, and the Knights Hospitallers, centered on Rhodes since 1308, until the island’s conquest by the Ottomans in 1522? Not everything can be discussed in detail here, but one monument that shows the coastal presence of the Knights Hospitallers is the citadel of Saint Peter they built in Bodrum (Figure 10). A monumental fortification directly on the harbor, the building contains numerous stone carvings of coats of arms related to the Knights Hospitallers, as well as a 15th-century Gothic chapel (Figure 11), later transformed into a mosque (Vaivre 2010; Berkant 2019).

Figure 10

View from the sea on the Citadel of Bodrum. Photograph by Patricia Blessing, 2023.

Figure 11

Chapel, 1498 with later additions, citadel of Bodrum. Photograph by Patricia Blessing, 2023.

While the building appears in accounts of the Knights of Rhodes’s presence in the Aegean, it is not included in the architectural history of medieval or early modern Anatolia, put instead into the context of late medieval crusading, and hence the fringes of European architectural history. However, much of the structure is contemporary to the monuments discussed above — so what if we considered them together, as part of Anatolia’s Mediterranean connection? The Knights Hospitallers built the citadel at Bodrum beginning in 1404 or 1406, in the aftermath of the Ottomans’ defeat against Timur, when the instability of the interregnum enabled a variety of actors to establish or reestablish (as in the case of the western Anatolian beyliks, which Timur revived) their power (Berkant 2019: 83; Kastritsis 2007). In the same campaign, the Knights Hospitallers lost their citadel in İzmir, and were left in need of a new stronghold on the Aegean cost (Vaivre 2010: 69). In the Bodrum citadel, a wide range of inscriptions and coats of arms, carved into stone, tell of the Knights Hospitallers presence at the site, recording the names of its commanders and the order’s grand masters (Berkant 2019: 85–95; Vaivre 2010: 79–111).⁷ The chapel in its current form was built in 1498, expanding or replacing an earlier structure; it was turned into a mosque in 1523, following the Ottoman takeover (Vaivre 2010: 114–15). The building, easily cast in terms of fortification architecture of Christian military orders, thus also holds elements tying it to the narrative of Ottoman architecture. In those latter phases, it thus also belongs into the context of 16th- and 17th-century Ottoman fortresses such as those of Çeşme, Seddülbahir, or Kumkale, as well as the 15th-century fortresses of Rumelihisarı and Anadoluhisarı, instrumental in the Ottoman conquest of Constantinople in 1453 (Holmes 2012; Thys-şenocak 2006: 107–86).

In introducing these monuments from a sea-based perspective, focusing on their location on or near shores, new challenges emerge. How do these buildings connect to the architectural histories of central Anatolia, land based as they are? How can we understand a bigger picture of a specific moment in time, in this case the late 14th and early 15th centuries, without falling back on a narrative that tries to present a false unity of the past in a geography that today largely is part of the Republic of Turkey? Would that not allow for a new perspective on Anatolia’s architectural heritage — one that exists alongside the land-based one, connecting Anatolia by way of land to the Caucasus, Iran, and Central Asia? Such an expanded perspective would certainly enrich our understanding of architectural history and integrate the study of medieval Anatolia further into the study of a global Middle Ages, valuing connections made overland as much as those forged across seas.

Landmass is also not as static as it may seem. As noted above not all the landmass we see today was there historically, especially on the western Anatolian coastline; sites that are nowadays landlocked had harbors and coastlines in antiquity or the medieval period. Think of the Troy of the Illiad, which is attacked from the sea; the archaeological site today is so far inland that one cannot see the shore. Clearly, the fluid nature of medieval and early modern Anatolia extends to land and sea, and both need to be considered in tandem as vectors of exchange, trade, and interchange between regions to reassess the region’s architecture in a Mediterranean context.

Author’s Note

I thank the editors of this special issue, Kıvanç Kılınç and Saygın Salgırlı, for their comments. I am also grateful to the peer reviewers for their comments and suggestions. For comments on earlier versions of this work, I thank Ali Yaycıoğlu. A first version of this paper was presented at the conference The Republic of Turkey at 100: Looking Back, Moving Forward, organized by the Abbasi Program in Islamic Studies and Center on Democracy, Development and the Rule of Law, Stanford University, 27 October 2023.

Competing Interests

The author has no competing interests to declare.