In the Ottoman city, the mahalle (quarter) was the basic unit of settlement and administration. In Amasya, too, life was organized less around broad avenues and squares than around quarters clustered about a small mosque or a fountain. Tax registers recorded households quarter by quarter, endowment deeds named mosques, schools and fountains together with their quarters, and the Muslim and non-Muslim communities were entered in separate quarters or in separate parts of the same one. Thanks to these records, Amasya's population, its number of quarters and the distribution of its communities in the Ottoman period can be followed in figures, quite apart from the accounts that describe the historic fabric of its quarters. Some of the figures come from a foreign traveler, some from the sixteenth-century tahrir (survey) registers, and some from the population and temettuat (income) censuses of the nineteenth century; in places they confirm one another, and in places they diverge markedly.

How a Quarter Was Founded and Named

In Ottoman custom a quarter usually formed around a small mosque (mescit). A well-to-do person would build the mosque, add a school and a fountain beside it and tie their upkeep to an endowment; in time households settled around the nucleus and the area came to be known by the benefactor's name. Hüseyin Hüsâmeddîn, the historian of Amasya, lists the quarters alphabetically in the first volume of his Amasya Tarihi and explains to whose memory each name is owed. The examples he gives show how widespread the rule was. The Uzun Mustafa Quarter bears the name of the merchant Hoca Hacı Mustafa Çelebi, whom sultans sent several times as an envoy to Egypt; the Hacı Hamza Quarter that of İsa Beyzâde Hacı Hamza Bey, tutor and adviser to Bayezid II during his governorship of Amasya; the Pirinççi Quarter that of Pirinççi Sinan Bey, one of Bayezid II's officers; and the Hızır Paşa Quarter that of Hızır Paşa, who built a mosque, a madrasa, a school, a dervish lodge and two bathhouses. Each of these men founded at least one mosque and one school in his own quarter and endowed his property to them. When the Hijri dates Hüsâmeddîn gives are read, many of these namings turn out to fall in the second half of the fifteenth century and the early sixteenth, the period in which Amasya shone as a princely governorship.

A name did not always derive from a person. What gave the Bozahâne Quarter its name was a large boza house built there in the time of Prince Ahmed, when the drink had become widespread. Devehâne took its name from being the resting place of the state camels that carried munitions and provisions on Ottoman campaigns, and the Kağnı Pazarı (Ox-cart Market) in the middle of the quarter arose when a deputy governor founded an endowment and a market there. Sarachâne is attributed to a saddlery that Çelebi Mehmed had built there during the years he ruled in Amasya, and the Kazancı Quarter appears in older records as the Kaya Paşa Quarter. The Tatar Quarter, by contrast, grew up around Tatar immigrants settled there from H. 1295 (1878); the official registers entered it as the İhsâniye Quarter, but among the people the name Tatar prevailed. A quarter could be founded around an old building, and equally around a newly arrived community.

The spelling of quarter names also shifted over time with local pronunciation, and Hüsâmeddîn notes examples one by one. The Eslem Quarter, named for Eslem Hatun, became the İslâm Quarter in everyday speech; the Şeyh Kırık Quarter turned into Çıkrık, and some scribes even wrote it Şeyh Garîk. The Gökceğiz Quarter, named for the Seljuk emir Bahâeddîn Gökceğiz, gradually became Küpceğiz and appears in the nineteenth-century register as Köyceğiz. The name Savakça derives from a small wooden water basin (savak) that Abdurrahman Halvetî of Erzincan had made in front of his dervish lodge; after H. 980 (1572) it began to be entered in records as Sevâdiye, which is the origin of the Suvadiye quarters of the nineteenth-century registers. Such variations mean that a researcher who reads sixteenth-century surveys and nineteenth-century population registers side by side will meet the same quarter under different names, and comparative work calls for care.

Hüsâmeddîn's List of Quarters

In his book Hüsâmeddîn treats fifty-two quarters entry by entry (in his general introduction he says the city had fifty-three) and adds, separately, the farm-born settlements of Karasenir and Kayabaşı outside the town. The list includes names such as İslâm, Eski Kethüdâ, Üçler, Uzun Mustafa, Bâyezîd Paşa, Bozahâne, Pirinççi, Pervâne Bey, Tatar, Temennâ, Câmi-i Enderûn, Çırakçı, Çeribaşı, Çıkrık, Hacı İlyâs, Hacı Hamza, Hekîm Çelebi, Hâtuniye, Hızır Paşa, Hoca Süleyman, Dârü's-Selâm, Dere, Devehâne, Receb, Ziyâre, Sâbıkeddîn, Sarachâne, Sarây, Sa'deddîn, Şamlılar, Şâmîce, Şehre Küsdü, Savakça, Sofîzâde, Sofîler, Acem Ali, Fethiye, Kuba, Karatay, Kazancı, Kamereddîn, Kocacık, Kurşunlu, Kılıççı, Küpceğiz, Köprü Başı, Gök Medrese, Gümüşlüzâde, Mehmed Paşa, Helkis and Yâkûtiye. Each entry describes the quarter's boundaries through its neighbors (for Eski Kethüdâ, for instance, "Köprübaşı to the east, the Bayezid Tabhane Row to the west, Bozahâne to the south, the Yeşilırmak to the north"), then gives the source of the name, the mosques, schools, fountains and bathhouses in the quarter, and sometimes the notable people who came from it. In this respect the book records the quarters as a geography of endowments as much as administrative units.

The headings also show that quarters replaced one another, split, or took new names. The Bâyezîd Paşa Quarter was founded south of the Kuş Köprüsü (Bird Bridge) where the Foka Quarter had once stood; the Mehmed Paşa Quarter received that name from H. 890 (1485), and its earlier name, Dugrâkiye, survived only as the name of the madrasa. Çırakçı and Câmi-i Enderûn were later separated from other quarters to become independent units, while Karatay and Kurşunlu are names with roots in the Danishmendid and Seljuk periods that remained in use under the Ottomans. Hüsâmeddîn bases these older names on what he calls kuyûd-ı atîka, the old records, and on endowment deeds. The Hekîm Çelebi Quarter is a melancholy example in the list. Lying east of the Hızır Paşa Quarter at the foot of the Kırklar Mountain and bearing the name of the renowned physician Bedreddin Mahmud Çelebi (active around H. 890), it was destroyed entirely in the war of 1293 (1877-78), and by the time Hüsâmeddîn wrote, only ruins were left.

Quarter Administration and the Communities

The daily order of a quarter was maintained by the imam and the headmen (muhtar) in Muslim quarters and by the priest and the kocabaşı (lay community leaders) in non-Muslim ones. The role of these officials is clearest in the conduct of the nineteenth-century population and temettuat censuses. In the temettuat survey the first and second muhtar and the imam took part in Muslim quarters, and the priest and kocabaşıs in non-Muslim ones; when the recording was finished they stamped their seals at the end of the register to confirm its accuracy. In most of the Amasya registers the seals of both muhtars and the imam are present; in the registers of four quarters only the two muhtars' seals appear and the imam's is missing. In mixed quarters the seals of the priests and kocabaşıs responsible for the reaya are not found. Muslims and non-Muslims living in the same quarter were entered in the same register, first the Muslim households and then, after a blank page, the reaya households. The title of the registers says as much: the first section contains the property, land and income of the Muslim inhabitants of the quarter, the second the records of "the reaya population dwelling in the said quarter."

The picture that emerges from the records is that in Amasya the communities did not live entirely apart, yet quarter boundaries were still clear. Alongside the Muslim quarters there were quarters of the Armenian, Greek and Jewish communities. In 1523 there is no quarter in which Muslims and dhimmis are recorded together, in 1576 six mixed units appear, and by the middle of the nineteenth century more than half of the city's quarters were ones in which both communities lived together. As will be seen in detail below, the structure of the quarters did not remain fixed over the centuries.

Tahrir Registers and Reading the Population

The Ottoman state had the lands it conquered or took under its administration recorded in terms of land, taxes and population. This process was called tahrir, and the resulting books were called tapu tahrir defterleri (land and tax survey registers). In the classical period surveys were generally made at intervals of twenty-five to thirty-five years, and causes such as the accession of a new sultan, a conquest or a disturbance in revenue balances could bring a new survey. The main purpose of the registers was to determine tax revenue and to regulate the timar system, so population information is obtained only indirectly. The entries show, for each quarter and village, the number of households, the male taxpayers (nefer), unmarried adults (mücerred) and revenues one by one. Researchers multiply the number of nefer by a coefficient such as 3 or 3.5 to arrive at a total population; the method gives a rough estimate, and results vary with the coefficient.

The two most important sixteenth-century records for Amasya are the register compiled in 1523-1525, known in the Ottoman Archive of the Presidency as Tapu Tahrir 387, and the detailed register of 1576, recorded as no. 26 in the Kuyûd-ı Kadîme Archive of the General Directorate of Land Registry and Cadastre. According to the encyclopedia entry, Amasya in 1525 had fifty-five quarters, five of them non-Muslim, and a population of about eight thousand, of whom some two thousand were non-Muslims. Ronald Jennings, who examined the same register, draws a more detailed picture for 1523. According to him the city had 1,990 nefer and 1,403 households recorded, 78 percent of the population was Muslim and the remainder Armenian, Greek and Jewish. Jews made up 17 percent of the non-Muslims, that is, 73 of 435 nefer. Jennings notes that among the cities he studied, Amasya was the only one with a separate Jewish community of over two hundred. The same author recommends multiplying nefer by 3 to 3.5, and by this calculation Amasya's population in 1523 comes to roughly 5,970 to 6,965. The eight thousand given by the encyclopedia lies somewhat above this range, and the two calculations evidently rest on different multipliers.

Jennings writes that in 1523 the city had sixty quarters, five of which were units close to non-Muslim villages. Dhimmis lived in twelve units, and in none of these settlements were Muslims and dhimmis mixed. Of the twelve units where dhimmis lived, five were villages on the city's edge, three were called communities and one was recorded as a taife. The quarters were small: on average thirty-three nefer per quarter, only two quarters had more than eighty nefer, fourteen quarters had no more than twenty and thirty-nine no more than forty. The most populous quarters were Dere Mescid (94 nefer), Sûfîler (91), Şamlular (72) and Yahuda (73); an unreadable quarter thought to be Savadiye was also in this group with seventy-three. The smallest were Alaca Mescid (9 nefer), Du'a (10), Mevlana Sûfî (10) and the Hızır Bey quarters. Four quarters bearing names tied to trade, Deve Hane (34 nefer), Şamlular, Aşağı Bazar (24) and Serraçlar (25), were recorded with an almost entirely Muslim population.

The Growing City of 1576

By the 1576 survey the city had grown markedly more crowded. According to Jennings's figures the population rose 67 percent to 3,326 nefer; Muslims rose 64 percent to 2,551 and dhimmis 78 percent to 775. The Christian population grew 97 percent, from 362 to 712, while the Jewish population fell 14 percent to 63 nefer. Muslims made up 77 percent of the total. Converted with the nefer multiplier, these numbers describe a city of about ten thousand people; the encyclopedia entry likewise puts the population at the end of the sixteenth century at ten thousand and the non-Muslim population at around 1,500, noting a decline from 1525. That the two calculations conflict on the non-Muslims depends on which units the registers counted as part of the city; it appears that Jennings counted the endowment villages as well as the quarters within the city.

Jennings gives the number of quarters as sixty for 1523 (villages included) and, for 1576, in one place 57 and in another 54; the difference evidently arises from which units were counted. The average number of nefer per quarter rose from 33 to 58, to 102 in mixed quarters, 52 in Muslim quarters and 60 in dhimmi quarters. A new development in the structure of the quarters is also noteworthy: the thirty-eight Muslim quarters of 1523 continued in 1576 and nine new Muslim quarters were added, bringing the total to forty-seven, while the ten dhimmi quarters were kept without new units. Among the six mixed units are Dede Kasım (26 Muslims, 2 dhimmis), Bayezid Paşa (58 Muslims, 1 dhimmi), the Mescid of Çeribaşı Mahmud Bey (42 Muslims, 2 dhimmis) and Eski Kethüda (25 Muslims, 9 dhimmis). Besides these there was a village of 108 nefer and the village of the Kamer Hatun Imaret with 339 nefer; in the latter 101 Muslims and 238 dhimmis were recorded, whereas in 1523 there had been only 24 dhimmis. Amasya was the only city in whose registers villages were also entered as part of the urban population, and the revenues of three of these villages were endowed to the Tomb and Imaret of Kamer Hatun. One of the city's new quarters was Bezaristan, meaning the covered bazaar, recorded with 13 nefer.

Among the largest quarters, Dere Mescid stood out with 152 nefer, Mehmed Paşa with 105 and a dhimmi quarter with 129. Jennings also observes that the Bayezid Paşa Quarter lay at the city's far end, south of the Kuş Köprüsü, and that Gök Medrese likewise lay along the river, and that the densest areas stretched along the river. Hüsâmeddîn's placing of the Bâyezîd Paşa Quarter south of the Kuş Köprüsü matches this observation.

Evliya Çelebi and Hüsâmeddîn's Figures

Evliya Çelebi, who visited Amasya in the seventeenth century, counted fifty-three quarters, forty-eight of them Muslim and five non-Muslim, and put the population at five thousand households, that is, an estimated twenty thousand people. The same traveler also writes of some 1,060 shops. This figure is far above the populations derived from the tahrir registers and the record of 1642; rough estimates by travelers and the tax-based counts of the registers often differ in this way, and both numbers should be approached with caution. The number of quarters, however, agrees with the registers: the figure moving between fifty-three and fifty-five stayed about the same from the sixteenth century to the seventeenth.

Hüsâmeddîn writes that in his own time the town of Amasya had fifty-three quarters, some eight thousand households and about thirty-seven thousand inhabitants. He says that in the period when the princes resided there as the seat of the sultanate it had about seventy quarters and some sixty thousand inhabitants, which he infers from old registers; he adds that when the princes later stayed in Istanbul and Amasya was attached to the province of Sivas, its former prestige was lost and its quarters and population declined. The tahrir registers, however, paint a far more modest picture of the princely age. Jennings's calculation for 1576, between ten and eleven thousand six hundred, falls well below Hüsâmeddîn's sixty thousand. The gap is hard to explain; Hüsâmeddîn presents his figures as inferences from old registers and does not detail his sources, so these numbers should be read as round estimates. In the same book he mentions three separate Defter-i Nüfûs ve Mahallât records for Amasya: one covers H. 816-918, one H. 1000-1057 and one H. 1112-1163, and they contain the population, the quarters, the notables and, for each family, about three generations of lineage. The second is said to be complete and the other two incomplete; it is also noted that each was certified by the Ottoman administration.

The Celâlî Devastation and the Register of 1642

The Celâlî rebellions that began in Anatolia at the end of the sixteenth century, and the attacks of the levends and bandit bands, affected Amasya and its surroundings severely. Because almost no tahrir records were kept after 1590, the effect of this destruction on the population can be measured only from a register of 1642. According to Jennings, the condition of this register is very poor; only a quarter of its pages, covering 459 nefer, survives, and from that fragment the city's population has been calculated at 1,736 nefer. The figure can be derived in two ways: 353 nefer of ulema and military class exempt from taxes, and 1,383 reaya. This corresponds to about half of the 1576 figure and to 13 percent below that of 1523. Since 80 percent of the readable names are Muslim, it can be concluded that the non-Muslim population fell proportionally more, but Jennings adds that this proportion cannot be regarded as certain. It is also recorded that in this period Tokat won the commercial rivalry with Amasya that had lasted for centuries and became the main depot for goods from eastern Anatolia, while Amasya remained dependent on the income of the main road to Istanbul.

Although the loss of population appears so great, Jennings leaves it as a matter for further research whether 1642 was a low point for Amasya, whether conditions improved under Murad IV, and what influence the Köprülü family, who governed Amasya and Merzifon, had over the city. For this reason it is not possible to give a reliable number of quarters or population for the seventeenth century. The encyclopedia entry says that the city's population grew in the seventeenth century but that it did not develop physically, and that its position in a narrow valley probably played a part in this. The encyclopedia's estimate for the nineteenth century puts the population between twenty-five and thirty thousand and the number of quarters at about thirty-six; this number of quarters is very different from the records of the 1840 census, and it is not stated which date or counting criterion it rests on.

The Fire of 1555 and the Renewal of the Quarters

A city squeezed into a narrow valley and built mostly of two-story mud-brick houses was vulnerable to fire and earthquake. Hans Dernschwam, who came to Amasya on 7 April 1555 with the Austrian embassy, writes that the houses seemed almost piled on top of one another. According to the same traveler a great fire broke out in the city on 14 April 1555, and the burned houses were quickly rebuilt. At that time the city had eleven mosques and one wooden and three stone bridges over the river. Hüsâmeddîn states that in H. 1099 (1687-88) a fire that started in the Eski Kethüdâ Quarter, and that at dawn of the same night also began in the Hâtuniye Quarter in the inner city, burned two thirds of the city; the fire grew larger because of looters, and goods carried out to the countryside to save them were stolen. The same source records earthquakes in H. 999 (1590-91), H. 1057 (1647-48), H. 1079 (1668-69) and H. 1240 (1824-25). From Hüsâmeddîn's quarter entries it is learned that the earthquake of 1057 damaged the Fethiye Mosque and the mosque in the Hacı Hamza Quarter, and that the fire of 1011 burned the mosques of the Bozahâne, Eski Kethüdâ, Receb and other quarters. After such disasters the quarters were usually rebuilt, and the person who rebuilt a mosque was recorded as a second founder of the quarter. Local accounts also speak of a fire at a later date, in 1915, that destroyed the bazaar and many buildings; I have not been able to confirm this date in an independent source.

The Nineteenth Century: Census and Temettuat

The nineteenth century changed the nature of quarter records. After the abolition of the Janissary corps (1826), Mahmud II began a population census in 1828-1829 in order to know the military strength and taxable population of the army to be formed; the process, interrupted by the Russian war, was completed in 1830-1831. Its difference from earlier surveys was that it was carried out independently of land registration and counted all Muslim and non-Muslim males from one year of age to one hundred. The males in each household were classified by age. Muslims were divided into "matluba muvafık" and "matluba gayr-i muvafık" (suitable or unsuitable for the state's requirements) for military and tax purposes; non-Muslims were divided into those liable for the jizya and those not, and those liable were placed in the ala, evsat and edna classes according to the amount they would pay. The registers also recorded the physical appearance of the men, with descriptions such as "tall, black-bearded, broad-browed, cross-eyed," and for men aged fourteen to forty fit for service the mark "mim" was added. To follow up the records after the census, the Ceride Nezareti (Registry Office) was established in 1831, and defter nazırlıkları (registry offices) were opened in provinces and sanjaks. These offices recorded births, deaths and settlers and sent the registers to the center every three months at first and later every six. Subsequent general or regional censuses were carried out in 1844, 1852, 1856, 1882, 1885 and 1903-1904.

Amasya's population register of H. 1256 (1840) is kept in the Ottoman Archive of the Presidency as NFS.d 2155. Soon afterwards, in H. 1260-1261 (1844-1845), a temettuat census based on household heads' incomes was made. The figures that two censuses only four years apart give for the same city are astonishingly different. According to Selim Özcan's comparison, the 1840 population register records fifty-five quarters in the city; twenty-three are quarters of Muslims alone and thirty-two quarters where both communities lived together. In twenty-one of the mixed quarters Muslims are the majority and in eleven non-Muslims. The temettuat census, on the other hand, has forty-eight quarters; seventeen are Muslim, thirty-one mixed, with Muslims in the majority in twenty-one and non-Muslims in ten. According to both censuses there is no quarter in which only non-Muslims lived. The seven vanished quarters comprised 368 households and 784 men in the population register, about a tenth of the city. Özcan attributes this to registers being kept quarter by quarter and to some quarter registers perhaps having been mixed into other classifications.

The differences in household numbers are also large. In 1840, 3,180 households were recorded, 2,236 Muslim and 944 non-Muslim; in 1844-45, 2,709 households, 1,857 Muslim and 852 non-Muslim. Of the 471-household difference, 368 come from the seven vanished quarters, leaving 103 households. The decline is about 17 percent in Muslim households and about 9 percent in non-Muslim households. Özcan attributes the fall in Muslim households to the losses suffered in long wars and to military service lasting over twenty years, which delayed marriage, and the fall in non-Muslim households to migration to large cities; he also adds the plague and cholera epidemics. Marked movements are also seen at the quarter level. The Bayezid Paşa Quarter fell from 112 Muslim and 104 non-Muslim households to 92 and 87; in the Üçler Quarter Muslim households increased by 12, while in Sıraböcekli 21 households were lost. In Acemali, where the population register had no non-Muslims, 39 non-Muslim households appear four years later, in Saraçhane 32 and in Pirinççi 2; Özcan explains this by migration from the countryside and from other quarters.

The same quarters are sometimes recorded under different names in the two censuses. Gümüşlüzade appears in the population register as Gümüşlüoğlu, Çevikçe as Çökçe, Bağıhelkıs as Helkisbağı, Cami-i Enderun as Enderun Camii and Pazar-ı Süfla as Aşağıpazar; Kayabaşı and Sıraböcekhane, which are entered as villages in the population register, are counted as quarters in the temettuat. In later volumes of Hüsâmeddîn's Amasya Tarihi, in the biography of İbrahim Ağa-Gümüşoğlu, there is also a detail concerning these registers: this former chief of the palace gardeners (bostancıbaşı) and coffee-server (kahvecibaşı) to Mahmud II is recorded in the population register of the H. 1254 survey as living in the Bağ-ı Helkis Quarter, aged forty-five. This record shows that the Bağıhelkıs Quarter of the 1840 register had also been entered in the previous census.

Özcan's most striking finding is the comparison of names and occupations. The names and occupations of household heads in the same quarters mostly do not match between the two censuses. When the population register of the largest quarter, Bayezid Paşa, with 112 households, is compared with the temettuat register with 92, only 30 household heads show a similarity that could make them the same person; of these, 18 pursue the same occupation, 6 changed occupation, two added a second one, two had an occupation in the earlier census and none in the later, and two had none earlier. Only the quarter's imam in the first household kept the same household number in both registers; the man written as "Mehmed Efendi, son of Yusuf" in the population register and "Mehmed Emin Efendi" in the temettuat register is most probably the same person. The remaining 62 household heads cannot be matched. Özcan seeks the reason in the way the censuses were made: the population registers may be copies prepared in the Ceride Nezareti at the center, whereas the temettuat registers are original records written on the spot and bearing the seals of the muhtars and the imam, and they would also be expected to be kept with greater care for the sake of tax calculation.

Later Censuses

After 1881 the salnames of the province of Sivas and the state archive records give Amasya's population at the level of the sanjak and the district. The population of the central district (the city together with its subdistricts and villages) appears as 50,432 in 1881, 63,640 in 1892 and 63,711 in 1915, about 82 percent of them Muslim. These figures cover the whole district rather than the city's quarters, and because women were not included in the 1881 figure, the author has doubled that year's value. The role of settlement of Muslim immigrants in the rise of the Muslim population is noted. The details of sanjak and district administration are a separate subject.

Chronology

1523-1525: According to Tapu Tahrir 387, Amasya has 1,990 nefer (1,403 households) and between fifty-five and sixty quarters; non-Muslims about one fifth.

1555: Dernschwam visits the city; a great fire breaks out on 14 April, and the city has eleven mosques and four bridges.

1576: The detailed register records 3,326 nefer and 54 to 57 quarters; Muslims 77 percent.

After 1590: Celâlî and levend pressure; no tahrir is kept.

1642: From a damaged tapu register, 1,736 nefer are calculated.

1687-1688 (H. 1099): A fire starting in the Eski Kethüdâ and Hâtuniye quarters burns two thirds of the city.

1831: The first general population census is completed; the Ceride Nezareti is established.

1840: Amasya population register: fifty-five quarters, 3,180 households.

1844-1845: Temettuat census: forty-eight quarters, 2,709 households.

1881-1915: The population of the central district rises from 50,432 to 63,711.

Sources

Primary sources

  • Hüseyin Hüsâmeddîn Yasar, Amasya Tarihi, vols. 1-4 (chapter on the quarters, introduction, disasters and monuments) and vols. 6-8 (biographies)
  • Republic of Türkiye, Ministry of the Interior, General Directorate of Population and Citizenship Affairs, "Osmanlı Dönemi" (tahrir and population censuses): https://www.nvi.gov.tr/osmanli-donemi

Academic studies

Online sources