Frontiers in Armenia: Hellenistic, Late Antique, Medieval, and Islamic Art (History) in the Caucasus

May 17, 2016 by Patricia Blessing, 2014 H. Allen Brooks Travelling Fellow
In this blog about medieval architecture in Armenia, I will return to two connected questions that I have discussed before. First, I will address the question how classical heritage is integrated in later buildings and, second, I consider once more at the ways in which we define medieval art. Both questions have multiple dimensions that involve ancient and recent history, as well as the historiography of various sub-fields of art history. I certainly do not claim to solve any of these complex questions, but my encounter with late antique and medieval architecture in Armenia has pushed me to reflect again on some of the answers that I have previously given. 

Rather than offering an overview of all the monuments that I visited, I will concentrate on three sites: Zvartnots, Geghard, and Noravank. While the first is a ruined seventh-century basilica, the latter two are monasteries with standing buildings dating to the thirteenth to fourteenth centuries. As I discuss these case studies, it will become clear how the issues above, as well as questions of site, scale, geography and history are crucial elements for the understanding of these buildings. 

Figure 1: Zvarnots with Mt. Ararat in the background (P. Blessing) 

Geography and history immediately come to mind at the first sight of the cathedral Zvartnots (Figure 1). Granted, it is hard to take a photograph to good effect when the sky is not entirely clear, but Mount Ararat (Masis in Armenian, Ağrı Dağı in Turkish) emerges in the background. The symbolic mountain seems near, and is yet far: the linear distance between Zvartnots and Ararat is about 60 km, but the mountain is also located 40 km inside Turkey, beyond the closed border between that country and Armenia. Hence, directions from Google Maps end up looking rather dramatic, with options of crossing through Iran or Georgia (Figure 2). Beyond this flippant observation based on playing with maps lies of course the serious reality of the ongoing tension between Armenia and Turkey on the subject of the Armenian genocide in 1915. The Turkish government still refuses to recognize the genocide as such, and this has not changed with the 100-year anniversary in 2015. 

Figure 2: Screenshot for Google Maps directions from Zvartnots to Mt. Ararat (

Figure 3: Zvartnots cathedral, partial view with portal (P. Blessing)

Thus, the site of Zvartnots is within reach of geopolitical issues; nevertheless, the architecture even in its ruined state is such that it quickly raises other questions (Figure 3). The cathedral, built in 643–652 under katholikos Nerses was destroyed in the tenth century. Hence, the parts that remain standing today are only a small section of the monument, in fact the inner row of columns that would have formed an ambulatory (Figure 4). The elevation of the church has been the subject of several suggested reconstructions, as Christina Maranci discusses in detail in her recent book about three major seventh-century Armenian churches. Fragments of the upper sections, with striking carving of pomegranates and wine leaves, are found on the site and are in part arranged to form a possible reconstruction (Figure 5). 

Figure 4: Zvartnots, columns of ambulatory (P. Blessing)

Figure 5: Zvartnots, fragments of elevation (P. Blessing)

Particularly intriguing are the large eagle capitals (Figure 6 and Figure 7): The wings appear wrapped around the capital, and the bird ready to take flight. The scale of these capitals, and of many of the fragments remaining on site or shown in the History Museum of Armenia in Yerevan, is a point that brought me to think, and that emerges also in other monuments. 


Figures 6 and 7: Zvartnots, eagle capital (P. Blessing)

Figure 8: Geghard, view (P. Blessing)

Geography is joined by topography in the elements that are striking while reflecting on medieval Armenia. The site of Geghard (Figure 8) is at the end of a valley 40 km east of Yerevan, yet this seemingly short distance takes a good hour even by car. So what of the period when the monastery was built? Of course, this originally being a hermitage, remoteness was the point as monks wanted to live in a community far from worldly concerns. Those, however, reached the monastery when princely burials were established at the site in the twelfth and thirteenth century. The addition of princely burials led to the construction of the gavit, a large fore-hall much more elaborate than a narthex, to the Katoghike church, in 1215-25 (Figures 9 and 10).

Figure 9: gavit, Katoghike church, Geghard, exterior 




Figures 10, 11, 12, and 13: gavit, Katoghike church, Geghard, interior and details of muqarnas dome (P. Blessing) 

Entering the gavit, the visitor is struck at first by the large scale of the structure (Figure 10), of unexpected height after the low entrance. Similarly unexpected may be the muqarnas dome (Figures 11 and 12) as well as capitals (Figure 13). While these forms are associated—in the mind of the art historical survey literature—with Islamic architecture, they are common in thirteenth- and fourteenth century Armenian architecture. The number of labels used in just one sentence makes me nearly cringe, yet I am struggling to find a better way to describe the discrepancy between historiography and architecture right here. 


Figures 14 and 15: interior of Katoghike church, Geghard (P. Blessing) 

The gavit connects to several structures, first and foremost the Katoghike church, built in 1215 (Figures 14 and 15), where a tall dome is the substructure of the lantern seen in Figure 9. Branching off from the gavit to the North is a second gavit, belonging to the Proshian chapel; built in 1215 and enlarged in 1283, this structure and the rest of the monastery was carved into the rock (Figure 16). Here, striking stone reliefs of lions (Figure 17) and crosses adorn the walls under a small dome (Figure 18).



Figures 16, 17, and 18: Proshian chapel, Geghard, interior (P. Blessing)

Next to this structure is the mid-thirteenth-century Avazan cave church, another chapel with a muqarnas dome, but here cut into the rock; light only comes through the small opening at the top, and from candles lit within the chapel (Figures 19 and 20). Overall, candles are a central source of light in these structures, be it in the gavit of the Kathogike (Figure 21) or in the upper gavit (completed in 1288) on the level above the Avazan cave church (Figure 22). 


Figures 19 and 20: Avazan cave church, Geghard (P. Blessing) 

Figure 21: votive candles in gavit Katoghike church, Geghard (P. Blessing) 

Figure 22: Upper gavit, Geghard (P. Blessing) 

While the question of the connections to Islamicate vocabularies abounds in the interior, it is also present on the exterior of the gavit of the Katholike (Figures 23 and 24). The scroll and leaves seen here are familiar from contemporary monuments in central Anatolia, but also particularly from manuscripts produced in Cilician Armenia. Many examples, some on view, are preserved at the Matenadaran in Yerevan, Armenia’s major manuscript library and research institute that holds manuscripts in Armenian, Georgian, Arabic, and Persian among others. 


Figures 23 and 24: gavit, Katoghike church, Geghard, details of stonework over entrance (P. Blessing)

Figure 25: Landscape around Noravank (P. Blessing)

Figure 26: Church of the Mother of God, Noravank (P. Blessing)

Similar elements are present in Noravank, the second monastic site I will discuss. Even more remote than Geghard, Noravank also lies in a mountain valley, surrounded by forbidding rock formations (Figure 25). The main structures remaining standing are the Church of the Mother of God, built around 1330 (Figure 26) and the Church of St. John the Baptist with its gavit, built in 1261 (Figure 27).

Figure 27: Church of St. John the Baptist, Noravank, portal of gavit (P. Blessing) 


Figures 28 and 29: Church of St. John the Baptist, Noravank, interior of gavit (P. Blessing) 

Upon entering the gavit, once more a muqarnas dome emerges above sculpted walls, although on a smaller scale than at Geghard (Figures 28 and 29). On the floor, stone slabs mark burials, pointing to the funerary function of this structure (Figure 30). The carvings on the walls and in niches, many of them crosses (khatchkars) further enhance the memorial aspect of the monument. In addition to the muqarnas dome, the portal of the Church of the Mother of God in particular shows elements that play into the convergence between manuscript illumination, Islamic and Christian architecture evoked above (Figure 31). This renews the question of attribution, stylistic divisions, and the separate narratives created by art historical fields that are only slowly being broken up in recent studies on this and related material. 

Figure 30: Church of St. John the Baptist, Noravank, floor of gavit (P. Blessing) 

Figure 31: Church of the Mother of God, Noravank, detail of stonework on portal (P. Blessing) 

Similar observations are of course true not only for medieval monuments, but also other periods. In this vein, I would like to leave the reader with two images of the Temple of Garni (Figures 32 and 33), built in the 1st century CE by Armenian king Tiridates, who had been rather inspired by a visit to Rome, it appears. Now, is this structure Roman? Should it be studied with the Roman ‘canon’? As part of the frontiers of the Roman Empire? Or as part of Armenian architecture? The Caucasian landscape behind the site does not help to solve the question.


Figures 32 and 33: Temple of Garni (P. Blessing)  

Recommended Readings

Armenia Sacra - Mémoire chrétienne des Arméniens (IVe–XVIIIe siècles), exhibition catalog (Paris: Musée du Louvre, 2007).

Maranci, Christina. Vigilant Powers: Three Churches of Early Medieval Armenia (Turnhout, Belgium: Brepols, 2015).

Maranci, Christina. “Building Churches in Armenia: Art at the Borders of Empire and the Edge of the Canon,” The Art Bulletin 88, no. 4 (Dec., 2006), pp. 656-675 .

Load more comments
Thank you for the comment! Your comment must be approved first
New code
Comment by from