On the south bank of the Yeşilırmak, in the Şamice quarter, on a narrow street close to the Gökmedrese, stands an octagonal tomb. It is topped by a pointed pyramidal cone, and on its west side rise the remains of a half-ruined madrasa wall. The name that locals pronounce as "Hilfet Gazi" is properly Halîfet Gazi. The man who bore it, Mübârizüddin Halîfet Gazi in his full name, was an emir of the Danishmendid principality; when the principality came to an end he entered Seljuk service, served as governor of Amasya, and, when he fell as a martyr in the campaign against the Georgians in 1232, was buried in this tomb. The complex is one of the few monuments in Amasya that show how the Danishmendid and Seljuk legacies were intertwined, and it is also a building whose date specialists still do not fully agree on.

Location and Present Condition

The complex lies in the city centre, in the Şamice quarter, on Torumtay Street. Some publications also write the quarter's name as "Şamlar", but the form established in the encyclopedia entries is Şamice. The tomb sits right next to the Gökmedrese and within the walkable core of Amasya; it can easily be seen on the same route as the Yeşilırmak embankment and the Gökmedrese.

The stairs leading to the upper floor of the tomb have collapsed, so the doorway now sits high above the ground. The badly ruined lower floor cannot be entered. The building can always be viewed from outside, but its interior is generally not open to visitors. Because access conditions can change over the years, anyone who wishes to see the inside would do well to obtain current information from the Amasya Provincial Directorate of Culture and Tourism. The tomb has undergone repairs and is in sound condition today; the madrasa survives only as ruins.

Architecture

The tomb is a two-storey structure with a square base, an octagonal body and a pyramidal cone. The base is built of rubble stone, while the facades and the cone are clad in regular cut stone; rubble stone is used in the interior. The octagonal body is covered inside by a shallow dome resting directly on the walls, and the dome's inner surface and the walls are lime-plastered. The entrance is a depressed-arch opening on the axis of the south facade. This facade was made wider than the others to form a monumental front, and nearly all of the building's decoration is concentrated here while the remaining facades are left plain. It is easy to see from this difference that the street-facing side was conceived as the ceremonial face.

The doorway on the south facade sits within a stepped, pointed portal niche with a three-tiered muqarnas hood, and above the niche is a window with twin arches. Thin zigzag mouldings surround the arch, and the whole arrangement is framed by a geometric outer border. In the spandrels above the arch and along the outer edges of the facade there are rosettes filled with geometric interlacing and stars. The rosettes vary in size and are placed freely so as to fill the gaps; the ornament is described as showing an extremely decorative style with asymmetrical features. The carving of the portal combines braiding, zigzags, hexagonal interlacing, six- and eight-pointed stars and half stars, together with curling branch, rûmî and palmette compositions. Geometric motifs predominate and floral elements are fewer.

The other faces of the octagonal body also have windows. On the west facade is a depressed-arched window with a tracery grille, on the east a round-arched one; the east window is topped by three tiers of muqarnas. In the upper parts of the southeast and north facades there is one twin-arched window each. The east and south window frames draw attention with a row of dentils, stars, rûmî and palmettes, inscriptions and rosettes. On the west facade of the tomb there is also a rectangular door opening into the madrasa; it is the clearest sign that the two structures were used as a connected whole.

In some publications the lower floor is called the burial chamber. According to M. Oluş Arık it has an oval plan and gives the impression of a space covered with a shallow vault supported by a large rectangular pier in the middle. Because it cannot be entered today, this description cannot be checked on site.

Its Place in the Kümbet Tradition

Funerary monuments in Anatolia are called by two different names. In a türbe the dead are buried directly in the ground and the body is usually square or polygonal; in a kümbet, by contrast, a cylindrical or polygonal body rises above a basement level known as the burial chamber, mummy chamber or crypt, and is closed by a dome on the inside and a conical or pyramidal roof on the outside. The Halifet Gazi Tomb, with its square base, octagonal body, pyramidal cone and two storeys, explains why the sources call it both "türbe" and "kümbet": its outward form belongs to the kümbet type, and so do the burial level beneath and the visiting chamber above. Sema Doğan's TDV entry stresses that in structures of this kind the custom of mummification continued in thirteenth- and fourteenth-century Anatolia, which shows how deep-rooted burial customs were among the Turks. The same entry also reports a view going back to Josef Strzygowski: kümbets are tent art transferred into architecture, and the conical, fluted bodies preserve the memory of the tent.

In the Great Seljuk period, kümbets in eastern Iran were generally cylindrical with conical roofs, while octagonal bodies appear from the late eleventh century in the brick kümbets of Harakan and at Damavand. In Anatolia kümbets are mostly two-storeyed, of stone and polygonal; square-plan kümbets appear only after the thirteenth century. The octagonal body with a pyramidal cone recurs across a wide geography in Anatolia. The Yusuf b. Küseyr Kümbet at Nakhchivan, dated 1162, is eight-sided with a pyramidal cone. At Divriği the Kamereddin and Sitti Melik kümbets of 1196, from the Mengujekid period, are octagonal with pyramidal roofs, and the Sitti Melik Kümbet stands out for its showy entrance facade and geometric ornament. At Erzurum the Emir Saltuk Kümbet is an octagonal structure of two-coloured cut stone with triangular pediments on its facades. At Kayseri the kümbet of 1238 in the Huand Hatun complex has an octagonal cut-stone body and a pyramidal cone on a marble base with muqarnas. The Halifet Gazi Tomb fits easily into this series in its cone and plan; what sets it apart is that it gives all the weight of its ornament to a single facade.

In terms of dating, the Halifet Gazi Tomb has a special place in the kümbet literature. Sema Doğan notes that no independent funerary monument of the Artuqid period is known in Anatolia, and that the status of some of the monuments accepted as Danishmendid is disputed; she counts the Halifet Gazi Kümbet at Amasya among the disputed ones. The same entry dates the simple octagonal Kulak Kümbet at Niksar to the late twelfth century. As the beginning of the Seljuk line it cites the Kılıçarslan II Kümbet at Konya (before 1192), the Gevher Nesibe Kümbet in the Double Madrasa at Kayseri (1205-06) and the İzzeddin Keykâvus Kümbet at Sivas (1217). Whether the Halifet Gazi Tomb is dated to 1145-46, as Gabriel and Aslanapa propose, or to 1225, as the common opinion indicates, places it at the head or in the middle of this series; the date of the building therefore also affects how the early phase of Anatolian kümbets is to be constructed.

The Roman Sarcophagus

The most striking object inside the tomb is a marble sarcophagus placed in an east-west direction. It was not made in the time of Halifet Gazi; it is a Roman-period work reused here a second time. It is decorated with erotes and winged medusas leaning on garlands tied at the corners with ram's heads. On the south-facing side, reliefs of ram heads with curled horns can be distinguished. The lid shows a sleeping woman in a garment with pronounced folds, and each corner of the lid has an acroterion. From these features the sarcophagus is understood to be a Roman-period sarcophagus of a woman.

The sources differ over which floor of the tomb the sarcophagus stands on. The text of the Amasya Provincial Directorate of Culture and Tourism places it in the lower floor, the burial chamber, while the entry in the TDV Encyclopedia of Islam places it in the interior of the upper floor, and the photograph illustrating that entry shows the south face of the sarcophagus from inside the upper floor. Since both floors are closed to visitors, this difference cannot readily be settled today.

The notes that Tanju Cantay published in the Sanat Tarihi Yıllığı (Art History Yearbook) treat the sarcophagus separately. Cantay states that a mummified body was found in the sarcophagus, a Roman one. In his view a section with the function of a mummy chamber is unnecessary in a kümbet; this can be explained by the existence beneath the building of a vaulted memorium attached to a church structure, or of a burial chapel. The interpretation offers an answer to the unusual plan of the tomb's lower floor, but it is not a conclusion confirmed by excavation. The reuse of a Roman-period sarcophagus in a Seljuk-era tomb is a common occurrence in Anatolia, where material from ancient times lived on by changing hands over the centuries.

The reuse of a Roman-period sarcophagus in a Seljuk-era structure is not unexpected. Amasya was an ancient city that grew up as a fortress-town and where, in antiquity, Strabo was raised. The TDV entry notes that spolia were used in the rubble fill of the madrasa wall, and it is natural to think that the sarcophagus belongs to the same tradition. Whether the sarcophagus was found on the site or brought from elsewhere is not stated in the sources; Cantay's interpretation of a church-related memorium comes closer to the first possibility, but there is no firm document. If one considers that the kümbet tradition maintained the custom of mummification, a mummified body lying in an ancient marble sarcophagus suggests that the reuse may also have been connected with burial custom.

The Madrasa

The madrasa to the west of the tomb was built, according to its inscription, in AH 606, that is, in 1209-10. Today its south wall and east section stand in part; in an account on the Turkish Culture Portal, on the other hand, only one arch column and stone remains on the west wall of the tomb are said to be left. The difference between the two statements depends on how the remains are read.

The building is made of stone and brick. In the south section one can make out the springing of an arch resting on a pier that is stone below and brick above. Old photographs, however, show a round double arch resting on three piers. The rubble fill of the east wall adjoining the tomb largely survives, and from this fill it is understood that spolia were used in the construction.

Whether the madrasa had two storeys is also disputed. Cantay says there was a second floor above the vaulted space extending towards the kümbet, and that this section is important because it made the madrasa a two-storey building.

Two repairs stand out in the madrasa's history. After the earthquake of 1647, Müderris Hasan Efendi had the collapsed parts rebuilt in timber. In 1888 the mufti of Amasya, Hacı Osman Hilmi Efendi, had the building thoroughly repaired. The madrasa's endowment deed (vakfiye), dated 622 (1225), has survived; Refet Yinanç, in his article published in the Vakıflar Dergisi, treated the madrasa together with its endowments.

Among the Madrasas of Anatolia

The first examples of madrasa buildings in Anatolia date from the Danishmendid period. The Niksar and Tokat madrasas built by the Danishmendid ruler Nizameddin Yağıbasan, from the mid-twelfth century, survive in part, and both use a closed-courtyard scheme. When the Seljuks took a city, the first thing they did was to found a mosque, a madrasa and a zawiya, thereby settling merchants, men of religion and Turkish population there. Madrasa architecture is classified within this framework by whether the courtyard is open or closed. The closed-courtyard type, in which the courtyard is covered by a lantern dome, matured in the Ertokuş (1224) at Isparta, the Karatay (1251) and İnce Minareli (1265) at Konya and the Caca Bey (1273) at Kırşehir. In the open-courtyard type, the Sırçalı Madrasa (1243) at Konya, the Taşmedrese (1250) at Akşehir and the Burûciye Madrasa (1272) at Sivas made the plan progressively more symmetrical, and this line reached its fullest form in the twin-minaret madrasas of Sivas and Erzurum in the last quarter of the thirteenth century.

The date of the Halifet Gazi Madrasa, 1209-10, falls in the early phase of this development, just after the Gevher Nesibe Double Madrasa at Kayseri (1205-06). The surviving remains are not enough to read the madrasa's plan with certainty; the standing south wall, the east section and the arch springings suggest a building arranged around a courtyard, but whether it was open or closed is not stated. Even so, two details tie the building to the madrasa concept of its time. The first is that the madrasa and the tomb were built side by side with a door between them. The oldest tomb to survive from the Anatolian Seljuks was set up in a cell of the madrasa that İzzeddin Keykâvus had built at Sivas, and it is counted among the early examples of the tomb-madrasa relationship. The second is that the founder's name appears in the madrasa's endowment deed; in the Seljuk period, madrasas were charitable institutions kept alive by the founder through endowment income.

The Gökmedrese, built at Amasya in the mid-thirteenth century, shows a different solution from the Halifet Gazi Madrasa. Referred to as Gökmedrese in Seyfeddin Torumtay's endowment deed, the building was used as both a mosque and a madrasa and forms a complex with the tomb adjoining it on the northeast. This comparison shows that at Amasya the functions of madrasa, mosque and tomb were combined in a single complex in different proportions. That Perrot, in 1861, called Amasya "the Oxford of Anatolia" and wrote that nearly 2,000 students studied in its madrasas shows that the tradition continued in later centuries.

Waqf and the Endowment Deed

The 1225 endowment deed of the Halifet Gazi Madrasa, which is in hand, is a narrow slice of the Seljuk waqf system. A vakfiye is the official document showing for what charitable purposes an endowed property is to be used and how the waqf is to be managed. The number of endowment deeds identifiable for the Anatolian Seljuk period exceeds eighty; the earliest is dated 440 (1048), and for most only later copies, containing copying errors, have survived. Apart from one Persian deed, all are in Arabic. Seljuk deeds come in scrolls; there are examples such as Karatay's caravanserai deed, more than 0.27 metres wide and 10.64 metres long. In original Seljuk deeds, after the basmala, "es-sultan" is written in red ink in place of the tughra, and the sultan's name is mentioned in the opening section of almost all of them. The deed is registered by the qadi, whose decision on the validity and binding force of the waqf is recorded in the document.

The deed of Halifet Gazi itself has not been read in full text; for its content one should consult Refet Yinanç's article and Sadi S. Kucur's doctoral thesis on the Seljuk and principality-period waqfs of Sivas, Tokat and Amasya. The Gökmedrese offers a useful comparison: the deed drawn up by Seyfeddin Torumtay himself is dated 665 (1266-67), and because the building's inscription could not be found, its construction is considered to have been finished by that year at the latest, on the basis of this document alone. At the Halifet Gazi Madrasa the situation is reversed: the inscription gives 1209-10 and the deed 1225, and the reason for the fifteen-year gap is not explained in the sources.

Who Was Halifet Gazi

His full name was Mübârizüddin Halîfet Gazi. His place and date of birth are unknown; according to Sefer Solmaz's article on his life, his father's name was Tulf and his grandfather's name was Türkân-şah. He was one of the Danishmendid emirs, and after the Danishmendid principality came to an end he entered the service of the Seljuks. According to the TDV Encyclopedia of Islam, Izz al-Din Kaykaus I, who conquered Sinop in 1215, appointed him commander of the Black Sea coastal guard. In the time of Alaeddin Kayqubad I, in 619 (1222), he was appointed governor of Amasya. He later took part in Kayqubad's campaign against the Mengujekids and, once victory was won, was appointed governor of Erzurum. In 1232 he was given command of the army in the campaign against the Georgians, fell as a martyr during it, and his body was brought to Amasya and buried in the tomb here. Solmaz notes that he served the Seljuk state in administrative, military and cultural fields.

The sources part ways at two links of this chain. The TDV entry places his governorship of Amasya in the time of Kayqubad, in 1222; more general accounts tie the post to the later years of Izz al-Din Kaykaus I, the early thirteenth century. The second divergence concerns Sinop. Some publications write that the city was taken in 1215 by Halifet Gazi, yet it was Sultan Izz al-Din Kaykaus I who took it; what fell to Halifet Gazi was the defence of the Black Sea coast after the conquest. That an emir of Danishmendid origin followed a line of posts from the Black Sea coast to Amasya and from there to Erzurum shows that the new state took administrators from the former principality's circle into its service.

Halifet Gazi's name is not remembered only for military and administrative posts. A medical book was translated from Persian into Turkish in his name; this is Hekim Bereket's Lübâbü'n-nühab, a translation whose Persian original is the Tuhfe-i Mübârizî, and the epithet in the title of the original comes from the same root as the emir's own Mübârizüddin. A book on the sports practised in Anatolia at the beginning of the thirteenth century, the Kitâb-ı Hulâsa der İlm-i Tıb, was also written in his name. These two works indicate that the emir was known as a patron of learned circles as well.

Among the Seljuk Buildings of Amasya

In the Anatolian Seljuk period Amasya was known by the title "Dâr al-Izz", and after Kılıçarslan II annexed the Danishmendid lands it became a city of that state. According to the TDV entry on Amasya, the works built in the period up to the passage of the city to Mongol governors after the Battle of Kösedağ in 1243 generally lie south of the Yeşilırmak. The oldest Turkish building in the city is known to be the mosque that Danishmendoğlu Melik Gazi had built just east of the citadel, on the site of today's Şamlar Mosque. Among the Seljuk-period works that have reached the present are the Burmalı Minare Mosque (1237-1247), the Gökmedrese Mosque (1266-1267), the Halifet Gazi Madrasa and Tomb (1209-1210) and the Torumtay Tomb. The Halifet Gazi complex is the oldest of these, and what distinguishes it from the others is that it is a foundation built by an emir of Danishmendid origin in Seljuk service.

According to its inscription, the Burmalı Minare Mosque was built in the reign of Gıyâseddin Keyhusrev II (1237-1246) by his emir Ferruh b. Selçuk. In this rectangular mosque, extending along the qibla axis, three domes are lined up one after the other on the central axis; the entrance facade is of regular cut stone and the other facades of rubble. To the left of the mosque's entrance stands a tomb that adjoins it and was built together with it. The body of this tomb, rising on a square base with triangular chamfered corners, is octagonal and is completed by a pyramidal cone; a burial chamber lies beneath it, and the upper floor is reached from inside the mosque. This arrangement belongs to the same family of buildings as the square base, octagonal body, pyramidal cone and burial level of the Halifet Gazi Tomb. In both tombs the window ornament is of muqarnas; in the Burmalı Minare Tomb the window facing the minaret, with its muqarnas niche, colonnettes and rosettes, has the character of an outer doorway. The difference lies in the density of ornament: the Burmalı Minare Mosque has no significant architectural decoration.

The Gökmedrese Mosque takes its name from the octagonal drum of the adjoining tomb, decorated with purple and turquoise tile and glazed brick. A brick drum rises above a cube-shaped cut-stone body on a square base; in the upper part of the drum a Kufic inscription frieze in turquoise glazed bricks runs around. The inside of the arch through which the tomb opens onto the mosque is covered with turquoise and purple bricks, and its pediment with green tile plaques. The Halifet Gazi Tomb lacks this wealth of tile; only stone is used on its inner and outer surfaces, and its ornament rests on the relief and carving of the stone. The mosque of the Gökmedrese, with its closed, thick and massive cut-stone walls, carries the characteristic features of Seljuk architecture; its facade is bounded on both sides by cylindrical corner towers, and in the middle a door is set within a pointed-vaulted iwan. On the borders at the entrance, geometric ornament appears in low relief, and the windows are arranged like prayer-niche compositions with muqarnas hoods. There is a resemblance between this facade arrangement and the south facade of the Halifet Gazi Tomb: in both the ornament is gathered on one face, geometric interlacing and star patterns are used, and the other faces remain plain.

When these three buildings are considered together, it can be seen that at Amasya the tomb type with octagonal body and pyramidal cone continued through the thirteenth century, but the method of decoration changed according to the period and the founder. The Halifet Gazi Tomb relies on stone relief, the Burmalı Minare Tomb on plainness and the Gökmedrese Tomb on tile and glazed brick. Because the decorative methods of the three buildings differ, dating the Halifet Gazi Tomb by its ornament alone does not appear reliable.

The Dating Debate

The most contested aspect of the building is the age of the tomb. The tomb has no inscription. The general opinion places it in the first quarter of the thirteenth century, most often in 1225. However, since the west facade of the tomb abuts the madrasa, it has been argued that it must have been built at the same time as the madrasa or earlier. As the madrasa's inscription gives 1209-10, the tomb is accepted to have been built then or before. The entry in the Online Turkey Tourism Encyclopedia likewise dates the building to the early thirteenth century on the basis of the construction date of the adjoining madrasa. In shorter accounts the tomb is placed around 1210 or in the range 1210-1225, a span that covers the fifteen years between the madrasa inscription and the endowment deed.

Aynur Durukan's TDV entry goes further. Considering that the doorway lies under the rubble-stone fill of the madrasa, that some wall remains are noticeable east of the tomb, and the material and technical characteristics of the two buildings together, it can be thought that, contrary to the common opinion, the tomb is at the latest contemporary with the madrasa, and even older. Indeed Doğan Kuban holds that the building was commissioned in the mid-twelfth century, and Albert-Louis Gabriel and Oktay Aslanapa that it may have been commissioned in 540 (1145-46). There are also differing views on who commissioned it; even so, the general opinion is that Halifet Gazi had the tomb built.

Another dimension of the debate is the origin of the building. According to the Dânişmendnâme the building was converted from a church into a madrasa, and Gabriel shares this view. Metin Sözen suggests that the arched section may be the facade of an older Byzantine building. The Online Turkey Tourism Encyclopedia writes that the tomb was built on the remains of a Byzantine structure and abuts these remains on its west facade. These are the interpretations of researchers; no written document or excavation result that settles the question is known. When the unusual plan of the lower floor, the spolia and the Roman sarcophagus are read together, it appears plausible that traces of an older building are present here, but nothing can be said beyond that.

Notes for Visitors

The muqarnas and geometric interlacing of the portal are more legible in raking light, so morning or afternoon hours are better for studying the surface. The south portal, the twin-arched window, the window frames, the springing of the madrasa arch and the rubble-filled wall adjoining the tomb are the main things to look at. Because the interior is closed, the sarcophagus is usually seen only through written and visual sources. Framing the tomb together with the madrasa ruin makes it easier to grasp the true size of the complex.

Chronology

1145-46: The possible date proposed by Gabriel and Aslanapa for the tomb (conjecture).

1209-10: The madrasa is built, according to its inscription.

1215: Izz al-Din Kaykaus I conquers Sinop; Halifet Gazi is appointed commander of the Black Sea coastal guard.

1222: Halifet Gazi becomes governor of Amasya under Alaeddin Kayqubad I.

1225: The madrasa's endowment deed is drawn up; the date commonly cited for the tomb.

1232: Halifet Gazi falls as a martyr in the Georgian campaign, in which he had been appointed army commander, and is buried in Amasya.

1647: The madrasa sections collapsed in the earthquake are rebuilt in timber by Müderris Hasan Efendi.

1888: Mufti Hacı Osman Hilmi Efendi has the building thoroughly repaired.

Bibliography

Primary sources

  • TDV İslâm Ansiklopedisi, entries "Medrese", "Vakfiye", "Amasya", "Burmalı Minare Camii ve Türbesi" and "Gökmedrese Camii". islamansiklopedisi.org.tr

Academic studies

  • Sefer Solmaz, "Türkiye Selçukluları Devrinin Ünlü Gazisi Mübarizü'd-Din Halifet Gazi", Selçuk Üniversitesi Edebiyat Fakültesi Dergisi, no. 16, 2006, pp. 237-248.
  • Refet Yinanç, "Selçuklu Medreselerinden Amasya Halifet Gazi Medresesi ve Vakıfları", Vakıflar Dergisi, no. 15, 1982, pp. 5-14 (via the TDV bibliography).

Official sources

  • Türkiye Kültür Portalı, "Halifet Gazi Türbesi" (sourced from the Amasya Provincial Directorate of Culture and Tourism), kulturportali.gov.tr. kulturportali.gov.tr
  • Amasya Provincial Directorate of Culture and Tourism, "Halifet Gazi Kümbeti", amasya.ktb.gov.tr.

Other sources

  • Aynur Durukan, "Hilfet Gazi Medresesi ve Türbesi", TDV İslâm Ansiklopedisi, vol. 18, Istanbul 1998, pp. 30-31. islamansiklopedisi.org.tr
  • Göksu Demircan, "Halifet Gâzi Türbesi", Online Türkiye Turizm Ansiklopedisi, 2019. turkiyeturizmansiklopedisi.com
  • Tanju Cantay, "Bir Kuzey-Batı Anadolu Gezisinden Notlar", Sanat Tarihi Yıllığı, no. 7, 1977, pp. 21-40.
  • Sema Doğan, "Kümbet", TDV İslâm Ansiklopedisi, vol. 26, Ankara 2002, pp. 547-550.

Online sources