To the south of Amasya, at the point where the Yeşilırmak leaves the Kazova plain and turns north, Turhal Fortress rises today on a steep mass of rock. The town that clings to its slope is the continuation of the settlement known in antiquity as Gaziura. Gaziura was a fortress town in the core lands of the Kingdom of Pontus along the Yeşilırmak basin, tied into the same road network as Amaseia, Zela and Komana Pontika. The geographer Strabo, a native of Amaseia, calls it "an ancient royal residence" but adds that in his own day the town lay deserted. That single sentence makes up most of what the written sources say about Gaziura; the rest is supplied by the fortress itself, by coins, by rock inscriptions and by the tombs in the surrounding countryside.

Its Name and Early History

The earliest traces of settlement around Turhal reach back to the Chalcolithic Age. Pottery of the Chalcolithic and Early Bronze Age has been found on Ali Tepesi in the township of Şenyurt, and finds from a rescue excavation carried out by the Tokat Museum in 1976 at Gevenlik Tepesi, near the village of Ulutepe (Kelit), have been dated to the Late Chalcolithic and Early Bronze Age. The oldest known name of Turhal, however, comes from a written document. A court record among the Kültepe texts, dated to the Old Assyrian Trade Colony period, around 1860 BC, mentions a place called Gazabura. The tablet concerns a copper transaction between a man named Kuzallum and another person whose name cannot be read because of a break in the tablet. Gazabura is considered to be the same place as the Gazzimara and Gazziura of the Hittite texts, and hence to point to the Hellenistic Gaziura and to modern Turhal. The Gaşipura of the Maşat Höyük letters is also generally accepted as the same town. According to Forlanini, the spelling Gaşipura appears in letters from the years before and just after the accession of Šuppiluliuma I, while the spelling Gazziura appears in texts of the thirteenth century BC. In the intervening period the region had been under Kaska occupation, and when the Hittites retook it under Hattušili III they went on using the old name in the form distorted by Kaska pronunciation.

Iron Age pottery has also been recovered on the fortress hill, and studies of Iron Age pottery in the provinces of Tokat and Sivas report examples from Turhal Fortress. The hill therefore appears to have been settled without interruption from the Early Bronze Age to the Ottoman period, through the Assyrian Trade Colony, Hittite, Iron Age, Persian, Pontic and Roman periods. This continuity also explains why the settlement was always founded at the same spot. Turhal stands where the roads linking Amasya, Tokat and Zile cross, on the bank of a river and in the middle of a fertile plain.

From Persian Satrapy to the Kingdom of Pontus

Under Persian rule these lands belonged to the satrapy of Cappadocia. In the reign of Artaxerxes III (358-338 BC) the satrapy was divided into a southern and a northern part. According to Strabo, when the Macedonians took Cappadocia the country had already been divided by the Persians into two satrapies, one called "Cappadocia proper" and the other Cappadocia Pontica, which the natives called Pontus. Turhal's lands lay within the latter. When Alexander, after the Battle of the Granicus (334 BC), appointed a new man to replace the satrap of Greater Cappadocia, the northern satrapy was in the hands of Ariarathes I, son of the Persian nobleman Ariaramnes. Ariarathes had remained outside the route Alexander took to the south and in 332 BC declared himself independent. After Alexander's death he resisted Perdiccas, to whom Cappadocia had fallen as his share, but he was defeated in two successive battles in 322 BC, taken prisoner and put to death.

Gaziura was the centre of this satrapy, and Ariarathes resided there as royal seat, at least for a time. The best evidence for the town's importance in this period is its coinage. Ariarathes had silver coins struck at Gaziura on the Persian weight standard. On the obverse the god of Gaziura, Baal, sits enthroned; he leans with his left hand on a long staff, holds an ear of wheat and a bunch of grapes in his right hand, and an eagle perches on his wrist. On the reverse a griffin attacks a stag, and the name of Ariarathes appears in Aramaic. The name of Baal, too, is read in Aramaic as "BAL-GZOUR". Wheat and grapes accord with the abundance Strabo describes for the plains of the Yeşilırmak.

In 301 BC Mithradates I Ktistes founded the Kingdom of Pontus in the basins of the Iris (Yeşilırmak) and the Lykos (Kelkit). Gaziura remained within the borders of the new kingdom. Its position at the point where the roads joining Zela, Amaseia and Komana Pontika crossed gave the town the advantage of controlling both the Dazimonitis and the Zelitis districts. Under Mithradates VI Eupator Gaziura is also reported to have been one of the Pontic fortresses that housed a mint. This information shows that the fortress was not merely a refuge but worked as one of the administrative nodes of the kingdom, where coin was struck.

Strabo's Gaziura and the Geography of the Plain

The name Gaziura appears for the first time in the ancient sources in Strabo. Describing the course of the Iris, he writes that the river flows through the middle of Komana in Pontus and westward through Dazimonitis, a fertile plain, and then bends north "toward Gaziura, an ancient royal residence, though now deserted". It then turns east again after receiving the waters of the Scylax and other rivers, flows past the very walls of his native Amaseia, and joins the Lykos at Phanaroea. In the same passage he says that the plain is always moist and grassy, that it feeds herds of cattle and horses and allows millet and sorghum to be sown, and that the mountain slopes yield wild grapes, pears, apples and nuts in great quantity.

Turhal fits this description exactly. The Yeşilırmak comes down from Kazova in an east-west direction, passes through the Katmerkaya Gorge into the Turhal plain, and there begins to flow from south to north. The first person to identify Turhal with Gaziura was the British geologist William Hamilton, who travelled in the region in the nineteenth century. Hamilton noticed that the river turns from west to north at Turhal and observed that this matched Strabo's account. The French archaeologist Charles Texier reached the same conclusion, and the American missionary H. J. van Lennep, the German orientalist A. D. Mordtmann and the traveller Heinrich Barth likewise accepted the fortress as Gaziura.

The identification is not beyond dispute. Some scholars equate Gaziura with Talaura, others with Ibora. Writers who follow Ramsay, such as Munro, Anderson and Grégoire, accept that Turhal is Gaziura and think that the town's name in late antiquity may have been Ibora; G. de Jerphanion's article "Ibora-Gazioura", published in 1911, is one of the basic texts of this debate. Today most reference maps place Gaziura at Turhal, and recent archaeological assessments of Turhal Fortress proceed on the assumption that its name in the Pontic period was Gaziura.

The Roman Wars and the Dazimonitis Plain

The best-known event in Gaziura's history took place during the Third Mithridatic War between Mithradates VI and Rome. Cassius Dio relates that in the consulship of Manius Acilius and Gaius Piso (67 BC) Mithradates pitched camp near Gaziura to draw the Roman commander Triarius into battle, and that Triarius left Gaziura reluctantly. At the end of the campaign the Roman army suffered a heavy defeat near Zela, and Gaziura Fortress was cleared of Roman soldiers. The victory did not last. When Mithradates learned that Lucullus had set out against him, he withdrew eastward from Gaziura towards Dazimon, and as he went he had part of the crops in the fields of the plain harvested and the rest of the grain destroyed. Lucullus, following him, was unable to feed his army as he crossed this fertile plain. The episode shows plainly that Dazimonitis, that is Kazova, was rich enough to feed an army.

Before it passed into the hands of the Pontic kings, Dazimonitis made up the large temple estate of Komana. The king governed the plain through a high priest whom he appointed, and the revenues were gathered at the temple. This arrangement appears to have continued in the Roman period, except that the high priest was now appointed by Rome. When Dyteutos, the last high priest of Komana, died in AD 34-35, these lands were annexed by Rome and attached to the province of Galatia. In the Byzantine period the plain was an imperial estate, as the boundary stone of the Emperor Maurice (AD 582-602) proves, and it served as an assembly point on the military road that ran through Sebasteia down to the Armeniac theme.

In 64 BC Rome joined Pontus to Bithynia to form the province of Pontus and Bithynia, and Gaziura came within its borders. The town then passed successively under Byzantine, Danishmendid, Seljuk, Ilkhanid, Eretna, Kadı Burhaneddin and Ottoman rule; the Ottomans are accepted as having taken it for good in the early fifteenth century. Strabo's remark suggests that the town had lost its character as a royal residence as it entered the Roman age. Texier writes that Pompey had the Pontic fortresses demolished and that Gaziura was already a ruin in Strabo's day; this interpretation rests on Strabo's word "deserted" and does not contradict the fact that the fortress was later reused with Byzantine repairs.

Turhal Fortress

Turhal Fortress stands in the town centre on a steep mass of rock about 180 metres above the floor of the plain. Local people also call it Dengiboz Castle. The Yeşilırmak flows past its west side from south to north and the Tokat-Amasya highway passes on its east; the town cemetery lies on the eastern and south-eastern foot of the rock, and houses on the western slope. The east and west sides are precipices and the south side is also partly steep, so the fortress is most easily approached from the north, where steps cut by levelling the bedrock survive, now largely worn away. The fortress extends north to south. Of the two surviving bastions, one in the south-east is square in plan and the other in the south is semicircular; it is reported that until 1933 the walls had three bastions. Most of the surviving wall remains are built of rubble limestone and date to the Byzantine and Ottoman periods. In the seventeenth century Evliya Çelebi described the fortress as a small pentagonal castle built of stone on an artificial mound in the Turhal plain, and added that it was not very strong.

The stepped tunnel cut into the rock is the most remarkable part of the fortress. Its entrance faces north and the tunnel descends towards the south; it is horseshoe-shaped, about 2.8 metres wide and 3.2 metres high at the entrance. The first steps of the tunnel are partly preserved, and the rest has been badly damaged. Hamilton writes that he went about fifty paces into the tunnel and turned back, heeding in the end his guide's warning, because the surface sloped at 45 degrees and was broken. Texier notes that the end of the tunnel was filled with stones and gravel and that one could go only about fifty paces. The German researcher Belck in 1901 found the tunnel skilfully worked, 2.80 metres wide and about 4 metres high, and wrote that after 230 steps it ended blindly in the rock. H. H. von der Osten, on 13 August 1926, was able to descend to the very end; he reported that the tunnel, which slopes at 45 degrees, was 250 metres long and that at its end two horizontal tunnels began, blocked with stone and earth. De Jerphanion recorded a second tunnel in the fortress, which begins at its highest point and winds down beneath the building remains. The same scholar noted an open round silo about 5 metres in diameter, and beside it a second silo or cistern of half that diameter. Today four cisterns cut into the rock survive on the hill; two are filled with rubble and earth, and two lie under vegetation.

Stepped tunnels are a feature of fortress architecture peculiar to the Pontic region. They consist of a stepped corridor cut into the rock at a slope of between 25 and 45 degrees. According to surveys, thirty-nine stepped tunnels are known in thirty separate places in the provinces of Tokat, Amasya, Çorum, Sivas, Yozgat, Ordu and Samsun. In the province of Tokat there are thirteen tunnels at eleven points, ten of them connected with a fortress; the fortresses of Turhal, Tokat, Zile, Niksar, Çördük, Geyraz, Karagöz and Kevgir are among them. Three views have been advanced about their function. Most researchers think the tunnels were made to supply the fortress with water, even during a siege; the fact that the tunnels at fortresses such as Turhal, Zile and Niksar descend towards the river flowing below supports this view. Excavations at Tokat Fortress showed that a spring was reached at the 138th metre of its tunnel. Von der Osten, judging from the extraordinary size of the Turhal and Çördük tunnels, argued that they also served to let part of the garrison descend to the valley out of the enemy's sight, and de Jerphanion thought they might have been secret sally routes. A cultic use is suggested on the basis of niches at the entrances of some tunnels in other regions; there is no such example in Tokat.

The dating of the tunnels is also disputed. No written document exists and the archaeological evidence is slight. Georges Perrot compared the barrel-vaulted entrances of the tunnels at Amasya Fortress with the workmanship of the royal rock tombs and dated them to the Hellenistic period. Bronze coins of the time of Mithradates VI found in the tunnels of Geyraz Fortress in Tokat and Kurul Fortress in Ordu also show that these structures were in use in the Pontic period. Although there is no direct dating evidence for the Turhal tunnel, the structure is considered within this tradition. Dereköy Fortress in the village of Dereköy in the district of Pazar, identified as a fortress of the Kingdom of Pontus by a surface survey in 2019, is a link in the same network. This small fortress, set at 1,102 metres above sea level, has a perimeter of 137 metres and an area of 838 square metres, and about 300 metres to its east and 70 metres lower down there is a stepped tunnel. It is reported that the Hellenistic settlements around Kazova were built on hilltops, that the farmland was left empty, and that along the ridge of mountains surrounding the plain stood a chain of fortresses that watched the road and the settlements. Gaziura is the largest knot of this chain within the plain.

The Fortress Inscriptions

The most important documents linking the fortress with antiquity are the Greek inscriptions cut into the rock. The British archaeologists Munro (1893) and Hogarth had reported ancient Greek inscriptions at the fortress. In 1899 J. G. C. Anderson saw, beside the path leading to the highest terrace, an inscription on a panel of the sheer rock face that had been deliberately defaced. According to Anderson's restoration the inscription looked like a military proclamation forbidding strangers to enter the fortress. In 1907 Henri Grégoire read it on a rectangular field 36 centimetres high and 65 centimetres wide and translated it as follows: no stranger may enter the fortress without the permission of the garrison commander; a slave who breaks the ban will be flogged and sold, and a free man will be exiled. The inscription has been dated to the late second or early first century BC, that is, to the time of Mithradates VI.

Grégoire had also recorded a second inscription on the upper terrace, of which he could decipher only a few letters. G. E. Bean found this second inscription in 1950, took a squeeze of it and read it: the son of Antiphilides, victor in the torch race of the youths at Byzantion, made this dedication to Hermes. From the letter forms Bean dated the inscription to the early second century BC, but in the same survey he could not find the military inscription that Grégoire had translated. Today neither can be read. The most recent surface survey identified on the bedrock of the south terrace a frame measuring 55 by 30 centimetres and noted that it coincides with the position of the inscription reported by Bean, and that the writing may have worn away over time. The researchers also record that the Hellenistic inscriptions reported earlier have disappeared from the fortress, destroyed knowingly or unknowingly.

The accounts left by those who worked at the fortress also record its history of wear. The parts of the fortress where Hamilton wrote that he had seen gates with lintels, jambs and thresholds of single large blocks were not found by Anderson in 1899, and Mordtmann said that only the stepped tunnel survived from antiquity and that everything else was modern stonework. Barth, while accepting that most of the fortress was new construction, wrote that the stairway cut into the rock could not belong to a period after Mithradates and that this was very probably a "place where the treasure was hidden". Strabo relates that Mithradates built seventy-five strongholds in the mountainous region of Lesser Armenia and deposited most of his wealth in them, but Gaziura is not among the principal fortresses he names; the treasure interpretation of Texier and Barth is therefore an inference rather than evidence.

Surveys and the Present Day

At Turhal Fortress in 1955 C. Burney recorded pottery of the second millennium BC, the Iron Age and the Classical period. In 1997 the Turhal District Governorate cleaned the stepped tunnel by decision of the conservation board. In 2011 the Tokat Museum carried out eleven days of cleaning and excavation to expose the fortress walls as the basis for a survey drawing for the fortress's repair. This work showed that the walls were on average one metre thick and largely destroyed, and it produced copper slag, pipe fragments, a bone ring and Iron Age pottery. Following the traces of the bastions and walls, most of the walls were later repaired. The most recent surface survey, carried out by a team headed by Şengül Dilek Ful of Tokat Gaziosmanpaşa University, confirmed that many of the remains in earlier researchers' accounts can no longer be seen, but that the hill on which the fortress stands was settled without interruption from the second millennium BC to the Ottoman period. The pottery found in quantity on the south and north sides is the most concrete proof of this continuity. On the south summit of the fortress there are the remains of demolished reinforced-concrete buildings.

Rock-Cut Tombs

The tradition of rock-cut tombs in the interior of Pontus can also be followed around Gaziura. The region is little researched archaeologically. Apart from the Hellenistic royal rock tombs at Amasya, the Late Hellenistic-Early Roman rock tombs at Ordu (Kotyora) and the few Hellenistic tombs in the provincial centre of Tokat, the rock-tomb tradition of Pontus has not been treated in detail. Surface surveys in the Turhal and Zile districts therefore try to fill this gap. Among the tombs examined, the Sivrikaya Rock Tomb in the township of Şenyurt, Turhal, and the Kapıkaya Rock Tomb at Evrenköy, Zile, stand out for their monumental appearance. At the locations called Göz and Ağılönü in the village of İğdir, in Zile, there are many tombs close to one another; since the other tombs in the survey area are scattered single examples, this concentration shows a distinct character.

No inscription or relief that could help with dating has been found on these tombs. The researchers therefore looked first at the architecture of the tombs and then compared them with rock tombs in Paphlagonia, Pontus's western neighbour, and in the village of Karabacak in the Sarıkaya district of Yozgat. The keywords of the study name the Hellenistic and Roman periods together; the reason no precise date can be given is that no written or relief evidence exists.

Chronology

Around 1860 BC: A place called Gazabura appears in the Kültepe texts.

Thirteenth century BC: The name Gazziura appears in Hittite texts; Hattušili III retakes the region from the Kaska.

332-322 BC: Ariarathes governs his independent satrapy from Gaziura; coins with the image of Baal are struck in the town.

301 BC: Mithradates I Ktistes founds the Kingdom of Pontus.

67 BC: Mithradates VI camps near Gaziura against the Roman commander Triarius.

64 BC: Pontus comes under Roman rule within the province of Pontus and Bithynia.

Early first century AD: Strabo mentions Gaziura as a deserted ancient royal residence.

1899-1926: Anderson, Belck, Grégoire and von der Osten examine the fortress and the stepped tunnel.

1950: G. E. Bean reads one of the fortress inscriptions.

2011: The Tokat Museum carries out wall clearance and excavation at the fortress.

Bibliography

Primary sources

  • Strabon, Geographika XII (Antik Anadolu Coğrafyası, trans. Adnan Pekman).

Academic studies

  • Uyanık, H. (2023). Tarihi ve Arkeolojik Araştırmalar Işığında Eskiçağ'da Turhal Kalesi. Akademik Tarih ve Düşünce Dergisi, 10 (5), 1529-1549.
  • Tekin, M., Ful, Ş. D., Uyanık, H. (2022). Kazova'da (Dazimonitis) Yeni Keşfedilen Basamaklı Tünelli Bir Pontos Krallığı Kalesi: Dereköy Kalesi. Belleten, 86 (306). belleten.gov.tr
  • Pontos Bölgesi'nin İç Kesimindeki Kaya Mezar Geleneği Üzerine Bir Değerlendirme (Gaziura ve Zela Örnekleri Üzerinden). TÜBA-AR (Journal of Archaeology of the Turkish Academy of Sciences). dergipark.org.tr
  • Bean, G. E. (1953). Pontos Yazıtları. Belleten, XVII (66), 151-178.

Other sources

  • Arslan, M. (2007). Mithradates VI Eupator: Roma'nın Büyük Düşmanı. Odin Yayıncılık.

Online sources