Tatev Monastery

Where Tatev is

Tatev stands in Syunik, the southernmost province of Armenia, on a shelf of basalt at the edge of the Vorotan river gorge. The nearest settlement is the village of Tatev, and the nearest town of any size is Goris, roughly thirty-five kilometres away by a road that spends most of its length getting into and out of the gorge.

This is a long way from the other places in this section. Khor Virap, Etchmiadzin, Erebuni and the Matenadaran are all within an hour of Yerevan; Geghard and Garni share one valley east of the capital. Tatev is several hours south of all of them, and that distance is part of what the monastery was for. A bishop's seat in Syunik was not an outpost of somewhere else. It was the centre of its own region.

The setting is the first thing anyone notices and the last thing to understand properly. Orbelian, writing in the thirteenth century, described the site as standing like a lofty watchtower, level and healthful, ringed by mountain peaks as though drawn with a compass, with the river passing below through deep ravines with an awesome roar. He was describing a defensive position as much as a view.

The gorge, and what it is worth

The Vorotan gorge is the deepest in Armenia, and how deep is a question sources answer differently. Armenia's own submission to UNESCO calls it eight hundred and fifty metres and the biggest gorge in the country; other accounts give seven hundred to eight hundred metres at Tatev itself. The difference is partly about where the measurement is taken, and no single figure should be quoted as though it settled the matter.

What the gorge did for the monastery is easier to state. It made the plateau defensible on three sides without anyone building anything, which is why the site could hold a treasury, a library and several hundred people through centuries in which Syunik changed hands repeatedly. When the monastery's own valuables were sent away for safekeeping in the twelfth century, they went to a fortress; the point of Tatev was that for most purposes it was already one.

The same gorge is now the reason most visitors arrive the way they do. A cable car opened in 2010 crosses it from the village of Halidzor, and at 5,752 metres it holds the record for the longest non-stop double-track aerial tramway. It is worth being clear that this is modern access infrastructure and nothing more: it is not part of the monastery, and the monastery stood for eleven centuries without it.

What stood here before

There was a church on this rock before the one that dominates it now, and the best source for it is candid about how little he knows. Orbelian describes an obscure church of undressed stones set in lime, dating from very ancient times, from the period of Saints Nerses and Sahak — which would put it in the fourth or fifth century — with a few clerics living beside it in continual austerity.

He then says something that most later retellings drop. He was unable to discover who the first bishop here was, or when; as for the oldest of them, he writes, he knows nothing, since there are no memorials. The primary source for Tatev's early history states plainly that its early history is undocumented. That is the honest starting point, and it is stronger than any confident date.

The name has its own tradition. The thirteenth-century writer Vardan Arewelts'i records that Tatev held the see of the apostle Eustathius, a disciple of Thaddeus — so the derivation of the name from a first-century figure is at least a medieval belief and not a modern invention. It is still a tradition rather than a record. The folk etymology that turns the name into the Armenian for give wings, spoken by a builder throwing himself from the roof, has no medieval attestation at all.

Documented history begins in the ninth century, and begins with property. Under Bishop Dawit', the prince Philippe paid for adjacent land, by a deed dated to the Armenian year corresponding to 839. From this period Tatev was the episcopal seat of Syunik, and it was that status, rather than any relic or miracle, that made the building campaign of the following decades possible.

Building Saints Paul and Peter

The church that gives Tatev its silhouette was begun in 895, and we know because its founder said so on the wall. Orbelian transcribes the inscription: in the year 344 of the Armenian Era, at Easter, which fell on the fourth of Navasard, I, Lord Yovhannes, who succeeded Lord Saghomon as bishop of Syunik, began the construction of this church. Brosset, translating Orbelian in the nineteenth century, worked the date out to 20 April 895.

The patrons were the princes of Syunik. Prince Ashot, son of Philippe, was the chief donor, along with the bishop himself, who is said to have spent enormous sums and labours on the work; other contributions came from Gabur and Sahak, sons of Grigor Supan, and from Prince Dzagik. The consecration drew King Smbat of the Bagratid kingdom, the Catholicos Yovhannes, Gagik of Vaspurakan and the Catholicos of the Aghuans, and lasted eight days.

The often-quoted span of 895 to 906 needs one clarification, because it is regularly turned into something it is not. The 895 is documentary — it is on the wall. The 906 is arithmetic: Orbelian says the work was finished in eleven years, and Brosset added those to 895 in a footnote. There is no inscription reading 906, and Orbelian gives the consecration no year at all. Sources that describe 906 as the date of consecration have inferred it rather than read it.

Orbelian describes the finished building as having a dome resembling the sky, standing a hundred cubits high, forty-eight long and twenty-four wide, with four altars besides the principal one and relics of Peter and Paul laid beneath its columns. Medieval writers also call the church Surb Arakelots, the Holy Apostles, and inscriptions record gifts made to it under that name — a detail most modern accounts of the monastery leave out.

The rest of the complex

Nothing at Tatev is all one century, and the buildings around the principal church are the proof. The church of Saint Gregory the Illuminator is the oldest structure with a date: Prince Philippe built it in 848, next to the old church on the south side, and was buried by its door. It has been destroyed and rebuilt at least three times since, most substantially in 1295 by Stepanos Orbelian himself, the historian, who was by then metropolitan of Syunik.

The gavit, the vaulted hall that Armenian monasteries place before a church, carries its own inscription: in the year corresponding to 1043, Lord Yovhannes recorded building these chambers. A small gate church of the Holy Mother of God went up on the north side in 1087. A west porch holds the tomb of Aruz-Khatun, wife of Prince Tarsayich, dated 1286. Each of these is a separate campaign, two centuries apart at the extremes.

The belfry is the structure most often described confidently and least often described accurately. A bell tower certainly existed by 1406, because a scribe working at Tatev names it. Whether the tower a nineteenth-century traveller described is that one is doubtful: an abbot demolished the medieval tower in 1890 and put up a new one, finished in 1897. What stands today is neither, but the lowest stage of a reconstruction begun in 1987 and abandoned in 1998.

The monastery was also a working settlement, and Orbelian lists what that meant: a wall of polished stone, underground storerooms, dining rooms, workrooms, shops, repositories for sacred objects, and libraries. An oil press survives, restored in 2010 and now a small museum. It pressed sesame, mustard and linseed rather than olives, which will not grow at this altitude — a small correction worth making because the mistranslation is common.

The Gavazan

Standing south of the church is an octagonal stone column about eight metres high, topped with a cross-stone, called the Gavazan — the staff. It is the most famous object at Tatev and the one most often described wrongly, so it is worth separating what is recorded, what is measured, and what has been claimed.

What Orbelian records is short. Bishop Yovhannes, after the church was finished, erected a marvellous pillar in the name of the Holy Trinity, close to the church and opposite it on the south side, built of small and large stones, reaching thirty cubits. That is the whole medieval account of its making. It says nothing about the column moving, and nothing about it measuring anything. Thirty cubits is also considerably more than the eight metres measured today, and nobody has explained the discrepancy.

That the column moves, however, is not in doubt and has been reported for well over a century. A traveller's account of 1858 names it the Gavazan and says that it shakes if touched with the hand, which led people to say its foundation had been artfully laid on a chest. Brosset, writing in 1864, noted that it still stood and still moved when leant upon, and that it had been examined by Hermann Abich — the founding geologist of the Caucasus — who could not explain the mechanism to him.

There is one medieval record of the column responding to an earthquake, and it is a good one. A scribe named Tovma Metsopetsi, working at Tatev, wrote in a colophon dated 29 November 1406 of an astonishing earthquake that shook the monastery, the bell tower and the Gavazan. That is an eyewitness in the building. It establishes that the column moved in a large earthquake; it does not establish that anyone built it in order to.

The modern explanation is that the column sits on a hinged joint at its base and rocks like a pendulum before returning upright, which is how a seismological paper of 2004 describes it. That paper also says the column could serve to record seismic oscillations, and credits the idea to a Soviet-Armenian architectural historian writing in 1962. This is the honest status of the famous claim: it is a twentieth-century interpretation of an observed property, offered tentatively, not a documented medieval purpose. The related stories — that it warned of approaching cavalry, or that it is an astronomical instrument aligned to the stars — have no support of any kind and should be set aside.

The school called a university

Tatev's second claim on Armenian memory is intellectual, and it comes with a label that needs handling. The institution is universally called the University of Tatev, and specialists who write about it in academic contexts generally call it the monastic school of Tatev instead. Both are describing the same thing; the second is more careful.

The difference matters because the modern word imports the wrong furniture. There was no charter, no faculty structure, no degree in the Western sense. What did exist was rigorous and institutional in its own way: the vardapet system, a ladder of fourteen ranks conferred on celibate priests after specialised study, examination before a panel and the presentation of a thesis. Grigor Tatevatsi codified those ranks himself, and the Armenian Church still uses them. That is a licence to teach rather than a degree, and it belongs to a church rather than to a university.

What was taught is reported consistently enough to summarise: scripture and its interpretation, the scriptural arts including manuscript illumination, and music. Reading extended to Plato, Aristotle, Philo and Porphyry, and — more striking — to Latin scholastic authors. In 1363 Grigor Tatevatsi copied a manuscript containing Bartholomew of Bologna and Peter of Aragon at his teacher's request. The Armenian scholars most opposed to union with Rome were studying Rome's philosophers closely.

Two numbers should be treated with suspicion wherever they appear. Tatev is often said to have had five hundred or a thousand students. Both figures are Orbelian's, and neither describes the school: five hundred is his count of the clergy around 906, and the thousand is his description of the community in the eleventh century, some three hundred years before the school arrived. They are not student rolls, and attaching them to the university is a straightforward error of chronology.

Vorotnetsi and Tatevatsi

The two names attached to the school are Hovhan Vorotnetsi and his pupil Grigor Tatevatsi, and the teaching line behind them is well attested: Esayi Nchetsi at Gladzor taught Vorotnetsi, and Vorotnetsi taught Tatevatsi for twenty-eight years. What is often said about the first of them, however, does not survive a look at the dates.

Vorotnetsi is regularly called the founder of the University of Tatev. He died in 1386 or 1388, and the school is generally reckoned to have settled at Tatev about 1390. On the best-attested chronology he was dead before it arrived, and his own teaching was done at Gladzor, at Vorotnavank and at Aprakunis. The accurate formulation is that he led and reorganised the higher monastic school of Syunik in the generation after Gladzor, and that it moved to Tatev under his successor.

Grigor Tatevatsi headed the school at Tatev from about 1390 until his death in 1409, and he is the reason the monastery's name attached to it. His Book of Questions, finished in 1397, runs to ten volumes and has been called a Summa; his Book of Sermons, finished in 1407, contains three hundred and forty-four sermons. Scholars regularly compare him to Aquinas, and just as regularly qualify the comparison. He is venerated as a saint of the Armenian Church, with a feast in Lent, and he was buried at Tatev, where a chapel of 1787 now stands over his grave.

He was also a painter. He is credited with the miniatures in a Gospel manuscript of 1378, and official Armenian institutions describe him as having produced illustrated Gospels now held at Etchmiadzin and at the Matenadaran in Yerevan. That combination — the leading theologian of his church also illuminating books — is the clearest single picture of what this kind of monastic school actually was.

Damage, loss and rebuilding

Tatev has been badly damaged repeatedly, and the record of it is unusually specific. Orbelian reports that the dome of the great church collapsed onto the church of Saint Gregory and brought that down too, and he names the cause: an earthquake at Ganja, in the year corresponding to 1138. Modern seismology dates that earthquake to 30 September 1139. Ganja is about a hundred and forty-five kilometres away, which is far enough to make the report worth flagging rather than simply repeating.

The best-evidenced medieval earthquake here is the one of 1406, recorded by the scribe who felt it. The most destructive by far was modern: the Zangezur earthquake of 27 April 1931, with its epicentre less than thirty kilometres away, reached an intensity at Tatev sufficient to destroy its churches. The dome and drum of the principal church came down, the belfry was reduced to three pylons of its lowest stage, and the gavit, Saint Gregory and the gate church were all wrecked. The Gavazan tilted and returned, as it had in 1406.

One famous loss belongs to Tatev only indirectly, and is usually told wrongly. Orbelian describes the fall of the fortress of Baghaberd, where the monasteries of the region had sent their books and church vessels for safety, and says that more than ten thousand items were taken captive and scattered. This is commonly retold as ten thousand manuscripts burned at Tatev. They were not only manuscripts, they were not at Tatev, and the text says scattered rather than burned. Orbelian's own date is the Armenian year corresponding to 1160, though 1170 is the figure usually printed.

Restoration has been continuous and is incomplete. A Soviet campaign ran from 1974, and the dome of the principal church was designed in 1981 and finished in 1988 — not, as is often claimed, by the modern Tatev Revival programme, which began in 2008. The gate church was restored badly in 1979 and corrected between 2016 and 2018. The belfry has stood unfinished since 1998. Of eighteen restoration projects drawn up between 2013 and 2016, four were carried out.

What to understand before seeing it

The single most useful thing to know at Tatev is which surfaces are medieval. Broadly: walls are, and roofs are not. Every dome and every roof on the site is a twentieth- or twenty-first-century reconstruction, and the belfry is a stump. The medieval fabric survives largely at wall level, with seventeenth- and eighteenth-century work above and around it. This does not diminish the place; it simply means that a photograph of the skyline is a photograph of modern engineering standing where medieval engineering fell.

The exception is the Gavazan, which is the one substantially un-rebuilt monument on the site — which is part of why it draws the legends it does. Standing next to it is the closest thing at Tatev to standing in front of something the tenth century made and left alone.

It is also worth arriving without the wrong expectation about status. Tatev is not a World Heritage Site. It has been on Armenia's Tentative List since 1995, together with the hermitage of Tatevi Anapat and the Vorotan valley, and a nomination was announced in 2025 — which is a different thing from inscription. The rock-cut monastery at Geghard, some distance north, is inscribed; Tatev, so far, is not.