Krikor (Grigor) Balakian’s Ruins of Ani A Surprising Source for Armenian Architecture

Krikor (Grigor) Balakian’s 1910 work, The Ruins of Ani (Ngarakrut‘iwn Anii Aweragnerun Badgerazart), documents the visit of the Armenian Catholicos Matt‘ēos Izmirlean (1845-1910) to Ani in 1909. Largely neglected by historians of architecture, The Ruins of Ani nevertheless offers an extraordinary account of the city and its monuments. After considering Balakian’s sources and scholarly perspectives, this paper explores his report on the buildings and the archaeological museum of Ani, highlighting discrepancies from the known record. Balakian’s often surprising remarks require careful scrutiny and cross-checking; at the same time, they highlight the value of any eyewitness source on Ani composed during the period of Russian control. Summary


Introduction
Situated on the modern closed border between the Turkish and Armenian Republics, in the Akhurean (Turk. Arpaçay) river valley, Ani is a place of astonishing natural and architectural beauty. While access to the site was restricted for much of the twentieth century, Ani has long been known as a rare intact, uninhabited medieval city. In 2016, UNESCO entered Ani onto its World Heritage List, but that was just a few weeks before the attempted coup d'état of July 15. As of this writing, future plans for the preservation of Ani are unclear.
With its rich array of medieval monuments, many dating from the tenth to thirteenth centuries, Ani forms a central subject in the history and historiography of Armenian architecture. Two recent bibliographies on the city include thousands of titles devoted to the site, including travel accounts, critical studies of architecture and history, corpora of epigraphy, archaeological reports, and exhibition catalogues (cf. Gechyan 2006 andYazıcı 2017b). Many conferences and workshops have focused on Ani; the virtualani. org website, moreover, offers a comprehensive sense of the city and posts periodic condition reports on its monuments. 1 Recent scholarship on Ani has explored issues of cultural heritage, as well as the period of Russian control of the city , when the site was excavated (cf. Watenpaugh 2014, Pravilova 2016. Such close and sustained attention to Ani makes the relative neglect of Krikor Balakian's 1910 work, The Ruins of Ani, all the more surprising. Originally published in Western Armenian in Constantinople by the Y. Mattʿēosean Press as Ngarakrutʿiwn Anii Aweragnerun Badgerazart (Description of the Ruins of Ani, Illustrated), it is a 90-page account of the two-day visit of the Armenian Catholicos Mattʿēos Izmirlean  to Ani in 1909. Balakian (1875Balakian ( -1934 was Venezia Arti, 27, 2018, 67-80 e-ISSN 2385-2720 at the time a 34-year-old priest; he later became known as a church leader and author of Armenian Golgotha, a memoir of the Armenian Genocide. He is the granduncle of the poet Peter Balakian, whose forthcoming translation of Ruins of Ani is eagerly awaited. 2 Ruins of Ani was not, to my knowledge, reissued after its initial publication, and judging from the scarcity of copies available, its printrun was modest. Nevertheless, it has earned increasing attention in recent years. It has appeared in Turkish translation (Usta, Hazaryan 2015) and was featured in a major essay in the Journal of the Society of Architectural Historians (Watenpaugh 2014). Tracing the history of Ani from the Middle Ages through the periods of Ottoman, Russian, Armenian, and modern Turkish rule, Watenpaugh situated Balakian's work, and the pilgrimage of the Catholicos, within the period of the city's rediscovery at the turn of the nineteenth century. Along with the European travellers who went to Ani, Watenpaugh notes, so too did Armenians from the Ottoman Empire, for whom Ani was a painful sign of prior (and lost) glory. These travellers, and the excavations of Nikolai Marr (1865Marr ( -1934, brought the dead city to life again, as processions wound their way through the city, open-air cauldrons bubbled for communal feasts, and museum visitors feasted their eyes on unearthed antiquities. As Ekaterina Pravilova has shown, this narrative offers only one view of the Russian period of Ani, which also characterized by conflicts between Marr and the Armenian institutions that supported him (Pravilova 2016). Nevertheless, it is wrenching to contemplate in light of the Genocide of the Ottoman Armenians only a few years later, and the annexation of the Kars region by the Republic of Turkey.
Other than Watenpaugh's essay, Balakian's Ruins of Ani is virtually unstudied among specialists of Ani. Yet Ruins should be studied both for what it reveals about the early historiography of Armenian art and architecture, and, equally important, for what it adds to, and challenges 2 For Peter Balakian's own engagement with Ani, see for example Balakian 2013. 3 Obviously, any claim of 'surprising' information depends on the knowledge level of the writer. I have sought out as many sources as possible -textual, visual, and oral -in order to verify Balakian's claims, from early travel accounts, to the archaeological reports and catalogues, to the most recent explorations of the city by Sezai Yazıcı and Vedat Akçayöz. The main sources used are listed in the bibliography. 4 For transcriptions of Ani's epigraphy see Orbeli 1966. in, the known archaeological record of Ani. The specialist will be surprised, for example, to learn of Latin inscriptions in the Ani museum, masons' marks at the church of Tigran Honentsʿ, and the existence of an undamaged, complete model of the church of Gagkashēn. Whether or not we are able to refute or confirm these remarks, they highlight the importance of pursuing every known source on Ani from before the destructive events of the twentieth century. They also suggest that even after centuries of interest in Ani, surprises still await the researcher. 3

Balakian's Sources and Historiographical Context
Balakian's text provides the reader with a general account of Ani, first considering its history, then its topography and urban plan followed by his own eyewitness observations of the site, concluding with an account of the scholarship on Armenian architecture (and on Ani's monuments in particular  (1842), as well as the pictorial albums of Garabed Basmadjian (1904) and Arshag Fetvajian (1866-1947. Balakian presented his account of Ani as an update to these works in light of the discoveries made during the exca- Balakian's commentary on the monuments of Ani, and on art more generally, follows contemporary scholarly trends in the European literature. For Balakian, the monuments of Ani are works of Armenian genius հանճար, thus reflecting the perceptions of monuments as expressions of nation (Balakian 1910) (ԺԱ). Along with Lynch and Texier, Balakian viewed Armenian architecture as originative and creative, departing from Karl Schaase's view that it derived from Byzantine, European, or Persian tradition (Schnaase 1844, 248-76; see also Maranci 2001 andAzatyan 2012).
Like many of his contemporaries, Balakian was also interested in the relationship between medieval Armenian and Gothic architecture, drawing heavily on the available literature. He grouped Ani Cathedral among the great expressions of Gothic architecture: San Marco in Venice; Notre Dame in Paris; and Westminster Abbey in London. Indeed, for Balakian, Ani Cathedral (c. 989-1001) bore a "preliminary imprint" (նախնական դրոշմը) of Gothic architecture, exhibiting the Gothic style as a kind of primordial impulse, rather than a historically conditioned product (Balakian 1910, 75). 6 Balakian's work thus demonstrates the engagement of Ottoman Armenian writers with European and Anglophone scholarship, challenging any illusion of neat borders between an 'Armenian' and 'European' history of Armenian architecture. 7 His Armenian translation of Lynch, moreover, actively reworks and edits the original English text, a project that deserves historiographical study in its own right. 8 Noteworthy, too, is Balakian's repeated reference to the role of human figures in Armenian 5 A complete bibliography of either Marr or Tʿoramanyan exceeds the limitations of space; nevertheless for the former, see principally Marr 1934; for the latter, Tʿoramanyan 1942-47. 6 As I have discussed elsewhere, this perception would take an explicitly anti-Semitic and pan-German turn eight years later with Josef Strzygowski's vision of an Aryan 'North' as the common explanation for Armenian and Gothic. See Strzygowski 1918 andMaranci 2001. 7 On the dynamic relations between German-and Armenian-speaking academic circles with regard to the study of Armenian medieval art, see Azatyan 2012. 8 See Balakian 1910, 76 compared to Lynch 1901 9 միշտ զգոյշ մնացած է մարդկային էակի ընդօրինակումէ... See also his comments in relation to the monastery of Hoṙomos, when he writes that figural carvings were "something which our ancestors always avoided" (…որոնցմէ միշտ խոյս տուած են մեր նախնիք. Balakian 1910, 86). 10 Although later in this text, he praises this statue's quality; see below.
11 See for example Balakian 1910, 25, 38. art. Armenian artists, he writes, "were always cautious about the representation of human beings" (Balakian 1910, 78). 9 Their presence in Armenian art, for Balakian, arose alongside cultural contact with Byzantium and Europe; when Persian and Arabic contacts were stronger, on the other hand, ornamental and vegetal forms become dominant (75, 78). Balakian regarded the lavish fresco program of the Tigran Honentsʿ church at Ani, and the freestanding, larger-thanlife statue of the Bagratid King Gagik (discussed below) as exceptional: the former the result of Byzantine and European influence, and the latter lacking refinement (78). 10 As is well known, however, figural representation is commonplace in medieval Armenian architecture, whether in two or three dimensions. Within Ani itself, there is almost no church standing which does not preserve some kind of interior figural painting -with more 'discovered' every year. At Hoṙomos (his Ghōshavankʿ) wall paintings survive in the church interiors, while the central vault of the gavitʿ of the upper monastery features a striking figural representation of Christ with church patriarchs (Vardanyan 2015). That Balakian mentions the dark interiors of the churches suggests that rather than overlooking these images, he simply could not see them. 11 Nevertheless, his perception of Armenian aniconism also reflects the complex historiography of the role of images in Armenian art, also expressed for example in the works of Josef Strzygowski (1891, 77-9;1918). The subject of Armenian aniconism, including its possible prehistory in medieval Armenian treatises and church councils, and its relations to the historiography of Islamic art, awaits closer scrutiny (cf. Der Nersessian 1973;Eastmond 2017, 77-122;Rapti 2009, 72-4).

Balakian versus the Known Archaeological Record of Ani
If the Balakian's general perceptions of Armenian architecture as expressed in Ruins of Ani find echoes in contemporary scholarship, his specific remarks regarding Ani sometimes challenge the archaeological record. Ruins of Ani will thus surprise the specialist on Ani, who might wonder about Balakian's viewing and recording habits.
It is important to note that Balakian's trip was very brief -really one afternoon and one morning -and it was made difficult by the "scorching summer sun" (ամառնային բոցակէղ արեւը) of late June (Balakian 1910, 92). Balakian also mentions the difficulty of sleeping during the night, due both to the merry-making of pilgrims which continued into the early morning, and his own excitement and "haste" (աճապարանք) to see Ani (94, 99-100).
One might therefore regard the anomalies in Balakian's report as a casualty of the rushed and fraught conditions of the trip, and simply discard it as an archaeological document. Yet entirely to disregard Balakian would be unwise, both in light of contemporary and subsequent looting of the site (which Balakian himself records), and of course the almost total disappearance of the contents of the Ani museums. Further, Balakian was a trained engineer and architect; he was later involved in the construction of Armenian churches in Marseilles and Nice (a subject, once more, deserving of separate study). Balakian and his group, moreover, were offered expert guidance on site by the archaeologist and architectural specialist Tʿoros Tʿoramanyan (84-9 and 99).
Further reason to take seriously Balakian's account is the amount of verifiable documentation within it. Part Three of Ruins of Ani collates the author's detailed historical and epigraphical knowledge with eyewitness observations at the site (21-71). This section contains descriptions of the fortifications, the Cathedral, the church of Tigran Honentsʿ, the church of the Holy Apostles, the church of Abughamrentsʿ, Gagkashēn, the 'Georgian' church, the Palace, the so-called Mosque of Minuchir (therein referred to as the Residence of the Catholicos), the monuments of the citadel, the Monastery of the Virgins (Բեխենց Վանք), the Virgin's Castle (Kiz Kalesi, Աղջկայ Բերդ), the bath, and the nearby Monastery of Hoṙomos. Balakian's comments on individual monuments demonstrate informed and close observation. For example, he knows the early eleventh-century account of Stepʿanos Tarōnetsʿi, which mentions that the church of Gagkashēn was based on the seventh-century church of Zvartʿnotsʿ (79). Balakian also notes that the crenellations of the fortifications have largely lost their "comb-toothed points" (սանտրագլուխ ցցուածքներ), also barely visible today (22). Regarding Ani Cathedral, Balakian rightfully notes the many cavities in the vaults and arches of the structure, invisible to the naked eye, but verifiable by intrepid climbers (29-30). Finally, he pays attention to interior decoration: at the church of Tigran Honentsʿ, he writes, the depictions of the martyr Hṙipʿsimē and her companions are depicted with "such vivid postures 12 այնքան կենդանի դիրքերով, որ ականատեսին մարմինը կը փշաքաղուի. that the hair on the body of an eyewitness will stand on end" (36). 12 Alongside this close and verifiable reporting, however, are several remarks which are either incorrect or cannot be verified. For example, Balakian reports that the eleventh-century church of Abughamrentsʿ ( fig. 1) is entered by three doors, and that it could hold "40 and 60 people" -a surprise to anyone who knows this petite monument of roughly twelve meters' diameter, entered by a single door at the southwest (46). It may be that Balakian confused in his notes this church with the much larger Gagkashēn, also dedicated Saint Gregory, which measures around 37 meters in diameter and is entered by three doors. The mistake, further, could also be a printer's error, because 40 to 60 appear as numerals rather than words in the published text, and because the sentence reads "This small church, which can only hold 40 to 60 people, has three beautiful doors" (emphasis added). 13 Another passage in Ruins is not so easily attributable to accident. In describing the interior of the Cathedral, Balakian counts twelve niches within the curvature of the apse, likening them to the apostles (31). In fact, there are only ten ( fig. 2). Balakian's miscount may the result of hastiness and a perhaps an enthusiasm for number symbolism, rather than a printer's error. 14 Elsewhere in Ruins we find statements that are entirely new, and that either have not yet been verified, or are unattested in the archaeological record and now cannot be verified because 13 Այս փոքրիկ եկեղեցին որ հազիւ 40-60 հաւատացեալ կրնայ պարունակել` 3 սիրուն դուռ ունի. On this count, it would be ideal to see Balakian's handwritten notes, should they survive.
14 The number ten however also contains symbolic value, however, and Ani's ten niches may be related to the number of canon tables prefacing Armenian gospel books, which were and are referred to in Armenian as khoran (lit. tent, canopy, but also used for the church sanctuary). the evidence is lost. An example of the former is found in Balakian's comments on the church of Tigran Honentsʿ ( fig. 3), in which he noted that the exterior walls bear masons' marks: each carved stone of this church, built of polished and uniform stones, bears the letter, Ա, Բ, Գ, Դ, Ե, and the succeeding letters of the Armenian alphabet. Consequently we can assume that the sculptures of each of these stones were separately carved, and in order not to create confusion for the stonecutters, they added the characters before they placed the stones in their present positions. Otherwise, at the height of the capitals, it would have been difficult to carve in such a delicate Masons' marks are quite commonly found on seventh-century Armenian monuments, and appear, with less regularity, on those of the tenth and eleventh centuries (including the Cathedral 16 For masons' marks, see Tʿoramanyan 1984, 52-7;Barkhudaryan 1963, 212;Kazaryan 2012, 2: 23-7;Maranci 2015a. 17 I thank Yavuz Özkaya and Armen Kazaryan for discussing this problem with me. from inspecting my own detailed photographs taken at the church over multiple years. Could Balakian's report therefore constitute an error or a mistaken memory -a product of the 'haste' which possessed the 1909 pilgrims to Ani? Balakian's specific observation regarding the forms of the marks, and his rational explanation for their role in the construction process, would suggest otherwise. Further, masons' marks on Armenian churches are typically only shallow scratches, rather than the deeply-carved incisions of the formal epigraphy, and so it is perfectly possible that they either weathered or are just imperceptible unless one hunts for them in raking light. If Balakian's marks do indeed exist, though, they would be striking examples of the convention: typically, Armenian masons' marks do not appear on Armenian churches in alphabetic sequence, nor do they draw exclusively from the Armenian alphabet, but assume various geometric and diagrammatic, as well as Greek-alphabetic forms. Should Balakian's masons' marks exist (and obviously further investigation is necessary) they might shed important light on building practices in thirteenth-century Armenia. Equally striking, although impossible now to verify, are Balakian's reports regarding the churches of the Holy Apostles (Surb Aṙakelotsʿ) and Gagkashēn. The former church, whose earliest inscription dates to 1031, is a now-ruined inscribed tetraconch; on its south façade stands a gavitʿ dated before 1215. The church has long earned the attention of scholars, both for its eleventh-century structure and potentially fivedomed plan, and for its astonishing gavitʿ with muqarnas vaults, polychrome, and Islamicizing façades. Yet nowhere, to my knowledge, is mention of what Balakian claims to have seen there: The visitor is amazed at what care and skill these massive monolithic stones were raised on these high walls; as in the Gagkashen church of Grigor, here too there were beautifully carved capitals in the form of eagles (Balakian 1910, 41). 18 No eagle capitals survive today at the site of the church of the Holy Apostles, whether within the eleventh-century structure or in the gavitʿ, nor did they appear in any of the early photographs and drawings I have surveyed.
Even more bewilderingly, Balakian compares these eagle capitals with those at the church of Gagkashēn. Eagle capitals at Gagkashēn would be a surprise to anyone who specializes in Armenian architecture; yet Balakian mentions these not once, but again in his more detailed discussion of the church, when he writes that 20 For detailed discussion of the archaeology and reconstruction theories of Zvartʿnotsʿ, see Maranci 2015b. massive stones whose four carved capitals are eagles with spread wings. (Balakian 1910, 48) 19 Like those of the church of the Holy Apostles, these birds are also unattested in the literature, and cannot be found at the site or in archaeological records of it, as far as I know. Tʿoramanyan's published field notes and drawings of Gagkashēn do not preserve mention or images of eagle capitals. They appear elsewhere at Ani, certainly: the twin eagles on the south façade of Ani Cathedral; an eagle and hare carved in bas-relief on a carved spandrel (preserved, at the time of Lynch's visit to Ani, in the interior of the Cathedral); and the eagle spandrels of the church of Tigran Honentsʿ (see Lynch 1901, 1: 372-3). Yet none of these forms a compelling parallel for what Balakian describes: namely, eagles, with wings outspread, positioned on columns.
Where we do indeed find such capitals, of course, is in the ruins of Gagkashēn's famous prototype, the aforementioned seventh-century church of Zvartʿnotsʿ, near the Mother See of Holy Etchmiadzin (mod. Republic of Armenia). There, magnificent carved birds once roosted, precisely as Balakian describes, on the four capitals under the dome, their wings outspread ( fig. 4). 20 Perhaps, then, Balakian confused the churches; Zvartʿnotsʿ had already been excavated and recorded by Tʿoramanyan, Balakian tells us that Tʿoramanyan showed him his drawings during his Ani visit, and Balakian mentions that he visited the excavations of Zvartʿnotsʿ. Therefore, Balakian might simply have projected the eagle capitals of Zvartʿnotsʿ onto Gagkashēn.
Yet ought one to dismiss definitively Balakian's eagle capitals as an accidental misremembering? If so, could he have misremembered eagles both at Gagkashēn and the church of the Apostles? This is unlikely; moreover, given the much better state of preservation of the archaeology at Ani in 1909, the difficult circumstances of the excavations and the upheaval of the following decades, it would be rash to dismiss this account out of hand. Gagkashēn, furthermore, imitated its prototype Zvartʿnotsʿ not only in ar-Venezia Arti, 27, 2018, 67-80 e-ISSN 2385-2720 chitecture but also in its sculptural program of Ionic basket capitals, spandrels featuring human figures and vinescrolls, and even its sundial, further discouraging one from discarding Balakian's report. Equally surprising is Balakian's description of the famous donor portrait of Gagkashēn, which was lost during the looting of Ani in the First World War and the Genocide. 21 Excavations of the church unearthed an over-life-size portrait of the patron, King Gagik Bagratuni I, together with a stone model of the church. After the statue and model were unearthed, they were brought the so-called Mosque of Minuchir, a large threebayed vaulted structure located on a cliff of the Akhurean River, which had been transformed into one of the two on-site museums holding finds from the excavations. According to Nikolai Marr, the model was already broken upon discovery, with only its lower part intact. This is seemingly confirmed by excavation photographs, which show the careful and elaborate design of the model, with its exterior arcade and projecting capitals, incised spandrel decoration, profiled oculi, roof ribs, and its portal with jambs and a denticulated cornice (fig. 5).
Balakian again makes mention of the undamaged model of Gagkashēn later in his text, in an account of the contents of the aforementioned Ani Museum. Much of this report can be verified by the 1906 and 1910 catalogues of Ani, by early photographs, and by existing objects today in the Historical Museum in Yerevan (see for example Marr 1906;Orbeli 1910a and1910b). Balakian describes the layout of the interior, its wooden drawers and glass vitrines; he mentions the skeletal remains, fragments of shirts made of leather and embroidered silk, bows and arrows, iron and stone axes, tools, censers, lances, iron ornaments, porcelain vessels, candelabra, glass, guns, and shields (61-2). In this context, he mentions again the model of Gagkashēn: The statue of Gagik I Bagratuni…is in a special glass showcase in this museum, and it remains a beautiful example of Armenian art of the tenth century. As we can see in the photograph, his Holiness the Catholicos was photographed next to it. Near this statue, there is a miniature model of the St. Gregory Church which was found undamaged and is now placed high up on the wall. It is a fine piece of work and thus appears to be the work of an accomplished artist. (Balakian 1910, 55) 23 Venezia Arti, 27, 2018, 67-80 e-ISSN 2385-2720 Photographs of the museum, including Balakian's photograph of Catholicos Mattʿēos confirms his description of the vitrine sheltering the image of Gagik ( fig. 7). Alas, the "undamaged" model, "placed high up on the wall" (որմին վրայ, բարձրը հաստատուած է), is not included in the frame. Another early photograph, showing Nikolai Marr in the museum, confirms that at least at one point, the statue of Gagik and the model were placed on adjacent walls; Gagik in a glass vitrine, and the model resting on a shelf with brackets. 24 What accounts for Balakian's anomalous reporting? It is tantalizing to imagine that the upper tiers of the model still survived upon excavation. It is not impossible that the model was completed with additionally excavated fragments, which somehow escaped mention or documentation. A more sobering possibility also exists: that Balakian was looking at a modern representation of Gagik and his model. The excavation artist, S. Poltoratski, composed a drawing showing Gagik as he is known from photographs, but holding a complete, three-tiered model of the church. A three-dimensional reproduction of the statue (now held at the Historical Museum in Yerevan) was also created, and is also attributed to Poltoratski (Kavtaradze 1999, 63). 25 Tʿoramanyan, too, composed many reproductions of Zvartʿnotsʿ, both as drawings and a three-dimensional model; perhaps one of these (or a similar model of Gagkashēn) was on view, and Balakian confused it with the original stone model. Yet would a modern copy be encased in a museum vitrine? And would Balakian, himself an engineer, be incapable of differentiating between a modern copy and medieval original? These speculations deserve careful attention because of the high stakes involved. Architectural models, in the medieval Armenian tradition, often followed closely the building with which they were associated, so the appearance of the model of Gagkashēn has real bearing on the original construction of the church, and by extension, its prototype, the seventh-century church of Zvartʿnotsʿ. I have elsewhere examined the complicated archaeology and reconstruction theories of Zvartʿnotsʿ; suffice it to state here that since both it, and the church of Gagkashēn, are in ruins, and that no Zvartʿnotsʿ 'copy' (there are others) survives intact, the model at Ani was a crucial piece of evidence (Maranci 2015b). If there were any chance that Balakian was accurate in his reporting, we would thus gain precious new insights about medieval Armenian architecture.
One final category of Balakian's anomalous reporting concerns the epigraphic corpus of Ani. Within the museum, Balakian writes, are small pieces of rock bearing Assyrian cuneiform inscriptions, as well as many large and small fragments of rock with Latin, Greek, Armenian, Arabic, and Georgian inscriptions. (Balakian 1910, 56-7) 26 Existing monuments and publications of Ani attest to Greek, Armenian, Arabic, and Georgian inscriptions (Jalaleantsʿ 1842, Orbeli 1966, Eastmond 2014. Cuneiform (presumably Urartian rather than Assyrian) is not attested, however, although circumstantial evidence certainly allows for the possibility. Marr reports the discovery of Urartian grave goods; additionally, there exists written correspondence, preserved in the Georgian National Museum Archives, concerning protests about the relocation of cuneiform inscriptions found at Ani (Pravilova 2016, 99). The presence of cuneiform inscriptions seems likely at least by the eighth century BCE, when the Kars basin was firmly under Urartian control. Finally, as I have discussed elsewhere, Urartian stelae were often reused in medieval Armenia (particularly in Van but also farther north), so it is possible that such objects were excavated from within the medieval strata (Maranci 2015a).
More surprising, however, is the mention of Latin inscriptions in the Ani Museum. Unfortunately, Balakian does not elaborate on them, so one wonders whether they were formal texts or graffiti, whether ancient, medieval, or modern. Ani's role as a world trade centre during the medieval period certainly allows for various occurrences of Latinity. Trade, embassies, and missionary activity all provide possible contexts, and European travellers to Ani are known from texts, including William of Rubruck (c. 1220c. 1293) and Ruy Gonzalez de Clavijo ( † 1412) (Yazıcı 2017, I). 27 One can also imagine Neo-Latin graffiti carved by early modern travellers to the city, such as Heinrich von Poser und Große Nedlitz (1599-1661), or Jean-Baptiste Tavernier (1605-1689). Without corroborating archaeological or textual evidence, however, the report remains a tantalizing possibility. If Balakian were right about the Latin inscriptions, the already impressive range of epigraphic languages known from Ani could be expanded yet further (Eastmond 2014).

Conclusion
Balakian's enigmatic comments open up a world of roiling doubt and tantalizing possibility, and, ultimately, confirm the tragedy that the majority of the excavated materials of Ani have simply vanished. The looting and destruction of Ani's museums, recently studied by Pravilova, coincided with the advance of Turkish troops towards Alexandropol (mod. Gyumri) in 1918. 28 By 1921, the museum was destroyed, its doors pried off the hinges and the roofs removed. There are reports of a train car packed with antiquities, headed for Tiflis, and then disappearing; there are other reports that the finds were reburied in the earth. In light of this uncertainty, and the continued destruction of the city in subsequent Venezia Arti, 27, 2018, 67-80 e-ISSN 2385-2720 decades, one is saddened to read of the missed opportunity to salvage the artefacts: Rumors reached us in Etchmiadzin that Saint Petersburg is considering moving Marr's Ani museum to Petersburg. I believe that the Petersburg Imperial Archaeological Academy will make a terrible mistake if it tries to move the museum there because the thousands of visitors from many nations who visit Ani will never have the opportunity nor the means to visit the museum in Petersburg. If the Academy is going to move the museum to Petersburg so it will be in the great Russian capital I would hope that both the patriarch in holy Etchmiadzin and Professor Marr himself will protest. It is appropriate and impressive for Ani's museum to be in Ani and furthermore the work of those conducting research is easer this way. Furthermore the Russian government should not take artifacts out of Ani -because the artifacts are both a source of great pride and even consolation. (Balakian 1910, 83) 29 Marr also shared Balakian's disapproval of the Saint Petersburg relocation, declaring in 1917 that artifacts from Ani should main in situ with 'only ideas' taken away by scholars (Pravilova 2016, 99). Had the museum been moved to Petersburg, however, many of the conjectures raised in the present essay about the archaeological record would probably be unnecessary. Nevertheless, given the lack of surviving evidence from Ani, we are obliged to acknowledge Balakian's documentation, whether or not it conforms to the previous scholarly consensus. Balakian's anomalous reports invite one to wonder what else might lurk in memoirs and personal diaries, as yet neglected or unpublished, of travellers to Ani during the Russian period.