Stations of the Ottoman Period in Iraqi Kurdistan: Preliminary Results from the Bazhera Archaeological Project

1.1 Stations and Outposts of the Ottoman Period: An Overview of the Archaeological Evidence

Archaeological evidence of later periods in the Kurdistan region of Iraq, especially from non-urban contexts, is rare and scarcely studied; among this, khans and other outposts for the protection of travellers, located in strategic areas for the control of the main routes, are certainly clear material evidence of the complex territorial management policy of this border region of the Ottoman empire.

Despite the not always stable situation of the Kurdistan region during the late Islamic period, on the border between these great empires, the area included between the Tigris in the west and the Zagros mountains in the east was traversed by several communication routes that branched off in a north-south and east-west direction. In this region, military actions intersected with commercial activities along major routes and merchants and pilgrims travelled along the same routes used by the two armies. The trade routes that departed from Baghdad in the direction of Mosul, then continuing to Diyarbakir and finally Aleppo, were marked by the presence of khans and forts that functioned as stopover stations (Peacock 2009, 2), as also recounted by Ottoman travellers, such as Matrakçi Nasuh (sixteenth century) and Eviliya Çelebi (seventeenth century) (Kale 2017, 132‑3).

In the more distant frontier territories, the Ottomans delegated the construction of bases that protected borders and principal roads to local chiefs, without any financial support and without providing any particular guidance on how to build them. For this reason, in these regions, it is difficult to find regularity and continuity of construction (building techniques and materials) of the forts and stations; a different situation appears instead in urban centres or those with greater strategic importance for the control of borders and movement of people and armies (Peacock 2009, 19‑20).

At the site of Bazhera (described in the following section, 2. The Bazhera Archaeological Project), which was the object of an initial archaeological investigation in 2023 with survey and excavation activities, the remains of a structure still visible today above ground may in fact provide a good example of a regional khan. This structure consists of an irregular square-shaped building with buttresses that may have served as a road station and point of control of the area between the Navkur plain (and the River Al-Khazir basin), to the south, and the access to the Zagros mountains, to the north [figs 1‑2].

While these kinds of shelter structures have been studied in detail in other chronological and geographical contexts, for example in works on medieval caravanserais in the Levant and Middle East region (Önge 2007; Cytryn-Silverman 2010) or on Ottoman-era stations on the hajj route (Petersen 2008), this region of northern Iraq has not been investigated so far. Except for study and valorisation projects concerning a few khans in urban settings,⁷ no archaeological study has so far been carried out to determine how they functioned, who they housed, and how they fit into the management framework of borders, major trade routes and land or water resources during the late Islamic period. Nevertheless, various evidence confirms the presence of stations and territorial protection or control buildings, built in the late Islamic period in the region.⁸ Written sources also confirm this presence.⁹ Some of these structures appear today in isolated contexts, far from urban centres or main villages, but always located in strategic areas, from the point of view of land management [fig. 2]. Only a few of them have been identified and partially studied. To the east of the Yazidi shrine of Lalish, located in Sheikhan district in Nineveh Governorate, about 30 km west of Bazhera, stands a station (19.5 m × 9.5 m), the foundation or dating of which is uncertain, but generally associated with the early Ottoman period. It is rectangular in shape, with a domed central room, considered to be associated with the accommodation of visitors to the shrine.¹⁰

Following the direction of the mountains, south of Koya – a city located about 50 km east of Erbil, along the road that connects this centre with Klkesimaq, Dukan and Sulaymaniya – there is an inn, outside of the urban area. It is rectangular in shape (16 m × 4.5 m), with domed rooms and a large space supported by columns, built of stone, mud and plaster. Within the city of Koya itself is located another large caravanserai (Mahmud Agha Khan); it is a khan with two floors and a large central courtyard, with rooms for storage, accommodation and stables, and decorative elements on the doors. Both structures are attributed to the mid-nineteenth century.¹¹

Moving westwards, on the River Tigris where today there is the Mosul Dam Lake, in the area of the so-called Saddam Dam Salvage Project,¹² several buildings, which have been interpreted as khans, were documented. In the southern region of the lake, a stone-built khan, consisting of a courtyard (22.5 m × 21 m) and rows of rooms on three sides, was identified in the village of Mishrifieh (site no. 17), probably founded during the Abbasid period but still in use in the Ilkhanid and likely Ottoman periods.¹³ Another khan with a similar plan has been identified in Tell Baqaq 3 (site no. 23), generally attributed to the Islamic period;¹⁴ caravanserai facilities have also been found in Hatara Kebir (Simpson 1997, 88).

The presence of these structures along the River Tigris during the Islamic period may be explained by the function of this watercourse as a transport and trade route that connected large urban centres, from Basra to Mosul, passing via Baghdad, and then directed towards western trade itineraries. Moreover, from Baghdad and Basra started the Iraqi hajj route towards Mecca and Medina (Petersen 1994, 47‑56).

Archaeological evidence of the presence of other Ottoman period khans also comes from more southerly contexts, mainly located along the banks of the River Tigris. In the nineteenth century, khans were built in the area south-east of Samarra and Balad, to provide shelter stations for pilgrims (visiting the Shiʿa shrine of Samarra) and travellers on their way to Baghdad.¹⁵

While these structures fulfilled the function of caravanserais, located on main routes and with rooms that were evidently used for housing travellers, goods and animals, it cannot be ruled out that this function could have also been provided by other types of buildings with a more clearly protective and defensive aspect, such as castles or forts. These presumably served to protect territories, most of them associated with Kurdish emirates, but may also have been used to accommodate travellers and merchants. Several forts, castles and military outposts are attested in the region throughout the late Islamic period, which fulfilled various functions. A series of relatively small forts seem to have been built around the sixteenth century and restored or renovated in the nineteenth century; they are located on main roads that connect Erbil with northern and eastern territories, usually in elevated positions (Khanzad, Dere and Dwin castles are located on the main road from Erbil to Ruwanduz).¹⁶ They have been seen as part of the construction activities for the protection of the Soran Emirate (sixteenth-nineteenth century), which controlled a wide region east of the Tigris, including the cities of Erbil and Ruwanduz. Khanzad is a two-floor rectangular building with large circular corner towers, made of small sized stones bound with mortar; Dere is a hexagonal structure with rounded-shaped towers at each corner. The Dwin walled citadel, on the top of the Pirman mountains, was also once the capital of the emirate and was occupied by residential areas, a mosque, places for keeping animals and a water collection system.

The site of Qalat Lokhan is a fortress located on the north edge of the modern village of Ruwanduz, which was investigated by the Rowanduz Archaeological Program and seems to have been part of a network of towers and forts that protected the Soran Emirate (Danti 2014) – which likely included also the construction of the Sartka castle, situated further south near the modern village of Dukan, built with stones bound with mortar, used to guard the borders with the Baban Emirate¹⁷ and to protect commercial caravans.¹⁸

In the late Ottoman period, in the eighteenth and nineteenth centuries, several outposts (qishle) of the Ottoman empire were established in Kurdish cities for the administration of territories.¹⁹

The region shows traces of various structures that served to protect borders and main routes, especially in the area of the mountains and along the rivers. These sites were part of a broader framework of infrastructures in support of land and water management, such as bridges to cross the main watercourses and mills.

A better understanding of the archaeological evidence associated with stations and khans of the Ottoman period, which also includes buildings similar in plan to the Bazhera structure, comes from the hajj route in the Arabian Peninsula, where the forts erected to protect trade and pilgrims’ roads were constructions of small size (circa 20 m across), employing simple construction techniques with the use of local materials, whose function was not so much to maintain military control of a region as to protect trade and pilgrims (Peacock 2009, 19). In his study of the hajj road during the Ottoman period, Petersen identifies two main phases of building typologies for these structures. A first phase, linked to the early years of the Ottoman conquest in the region of the Middle East and the Arabian Peninsula (early sixteenth century), saw the construction of several buildings, mainly square in plan (about 20 m per side) and with a rectangular central courtyard surrounded by vaulted rooms; these generally have a ground floor, a first floor and an additional floor with a parapet. Generally, these buildings are constructed with locally available materials and have few decorative elements, usually located on the main door (Petersen 2008, 32‑3). Another important phase, which took place in the eighteenth century, involved the construction of small, square-shaped stations with corner towers with gun slits (2008, 32‑4). These buildings have no decorative elements, not even at the main entrance. They are organised around a central court and have one ground floor, usually with vaulted rooms used for sheltering animals or for storage, and a first floor (Shqairat 2020, 108).

These structures were all supplied by cisterns or water reservoirs, an extremely important feature for their function but also for the reception of pilgrims; other infrastructures along the roads, such as bridges, ensured their accessibility (2020, 104‑5).

4.1 The Enclosure

One of the main goals of the first season of fieldwork at Bahzera was to study and document the structures preserved above ground, in order to derive an initial set of data on the construction history of the structure and to formulate specific questions for further research.

The structures currently preserved above ground belong to an enclosure of which three sides – east, south, and west – are visible today [fig. 6]. The standing structures are preserved to a maximum height of about 4 m, measured on the outside; the interior is now covered by layers of rubble.

The enclosure has four three-quarter circle corner buttresses, plus two semicircular side-buttresses on the south and west sides. The north side of the enclosure has not survived, but its original course is indicated by the two corner towers that once connected it to the west and east sides. The south and west sides appear to be preserved for their entire lengths (the western side is, however, partially collapsed in the northern part). The south side measures about 34 m (measured on the inside) and has a semicircular outer buttress at the centre. The west side measures about 27 m (measured on the inside) and, like the south side, has a semicircular buttress in the centre. Middle buttresses on the other sides are not currently visible. The east side has a peculiar shape: the south-east buttress is not in line with the north-east buttress, and therefore, to connect, the south-eastern portion bends westwards to meet the portion coming from the north-east, creating a recess. The wall fragments visible today do not allow us to understand the exact configuration of this side, or to hypothesise the presence of specific defensive elements that may have determined this specific layout, such as a bent entrance.

No traces of the original entrance are preserved; the original limits of the opening seen today on the eastern side are not visible, and the northern side is completely collapsed, therefore without excavation it is not possible to establish its contemporaneity with the original layout. The masonry visible today has no openings of any kind, or specific defensive elements, such as crenelations [fig. 7].

The collapse in the interior means that it is not possible to understand whether multiple storeys were present. However, the thickness of the rubble deposits, as well as a greater width of the lower courses of the internal walls seem to suggest the presence of a roofing system [fig. 8]. The thicker masonry, in fact, could have supported vaulting.

The extreme homogeneity of the construction that can be observed today suggests the absence of major interventions on the structure, and thus would seem to indicate a single construction phase. Only by means of the excavation of specific portions of the deposit will it be possible to acquire information useful for reconstructing the layout of the enclosure and understanding the details of its occupation phases.

4.2 Building Technique and Materials

The foundation of the structure is not visible today. The standing walls are very homogeneous in terms of building techniques and materials, including the buttresses. The masonry is built in regular banks that can be identified all along the internal and the external faces. The height of the bank varies from 27 cm to 58 cm; the choice of a specific height for a given course does not seem to relate to its position in structure: thicker banks are not necessarily used in the lower parts [fig. 9]. Unworked cobbles of medium (abound 25 × 13 cm) and small size (abound 12 × 9 cm) are bound together with abundant mortar [fig. 10].

These cobbles were likely been collected from the banks of the neighbouring wadis, where their presence can also be observed today. The masonry is built with two faces and a core, with a thickness of about 70 cm. The building material is carefully laid in series of rows in the faces, with elements of larger-sized pieces at the bottom of the bank and smaller ones at the top; blocks are laid lengthways. In the core small sized cobbles are generally used, with a higher percentage of mortar.

The mortar is tough, made of abundant sand of variable grain size and lime;²⁷ it is abundant and extrudes between stones. The original finishing can be observed east of the south-west buttress [fig. 11]: the mortar is not smoothed but extrudes against the formwork; no clear traces have survived on the surface of the mortar to indicate the nature of the formwork (wooden planks?). The upper layer of each bank shows a regular surface obtained by a rough smoothening of the mortar [fig. 12]. The construction seems to have proceeded homogeneously by horizontal levels, as a single bank can be followed along the whole structure visible today. The passage from one portion of formwork to the adjacent one has left no traces in the faces of the walls. An interruption can only be seen on the external faces in proximity to the circular buttresses [fig. 13], where a different formwork was presumably required; no interruption is perceived on the corresponding portion of the internal face.

The internal face of the walls presents a peculiarity: in several cases the blocks are cut as to obtain a more regular surface; this operation seems to have been carried out once the wall was already completed. It never occurs on the exterior, where irregularities were presumably not inconvenient.

4.3 Interpretation and Dating

On the basis of the preliminary analysis carried out so far, no clear indication as to the function of the structure seems to emerge. The absence of elements clearly related to defence is certainly relevant, although the fragmented condition of the building clearly limits the interpretation of negative evidence. However, the building of a robust structure in such a prominent position, overlooking the valley and dominating the surroundings, is compatible with the function of a khan.

As to its dating, only hypotheses may be put forward at this stage of the research.

Other khans are present in the region. They have never been studied in detail, but they are usually ascribed to the last phase of the Ottoman period.

The masonry typology may provide some chronological indications, but for the area this is still in the process of being established.²⁸

The masonry in banks seems to have appeared in the region in the late Ottoman period. So far, this type of masonry has been studied in detail only in relation to milling installations, and therefore any similarities should be treated with caution. In the group of mills excavated in the Wadi Bandaway (site no. 124 of the LoNAP survey), this banking technique is employed for the building of the millhouse. In Period 4 (late nineteenth-early twentieth centuries, Tonghini, Usta 2021, 109‑13), natural cobbles are used together with split stone (usually limestone), the latter in much larger quantities. In Period 3, the use of cobbles prevails over that of other types of material, and the masonry shows similarities to that studied at Bazhera (2021, 105‑9). Period 3 at Wadi Bandawai pre-dates Period 4; as regards absolute chronology, it has been tentatively ascribed to the early Ottoman period, i.e. to the sixteenth century (2021, 115). Given certain technical similarities, and taking into account the historical context, a similar chronological horizon may fit the building of the khan at Bazhera. The continuation of fieldwork, especially excavation, may in the future provide more conclusive evidence in this respect.