^{page 21 note 1} I wish to thank the directors and staffs of the Başbakanlik Arşivi in Istanbul and the Kuyud-i Kadime Arşivi of the Tapu ye Kadastro Dairesi in Ankara for their generous hospitality. The research for this paper was done with grants from the American Research Institute in Turkey and from the Center for Asian Studies, University of Illinois at Urbana-Champaign. I wish to thank Professor Andreas Tietze for helpful suggestions.Google Scholar

^{page 21 note 2} Tables most recently published in ‘Research in the Ottoman Fiscal Surveys’, in Cook, M. A. (ed.), Studies of the Economic History of the Middle East (London, 1970), pp. 163–71, at pp. 168, 169.Google Scholar

^{page 21 note 3} See Braudel, F., The Mediterranean and the Mediterranean World in the Age of Philip II, trans. Reynolds, S. (London, 1972), vol. 1, 402 f., 326 f.Google Scholar

^{page 22 note 1} Reaching puberty is the sign of adulthood according to the Hanefi school of Islamic law. Cf. Schacht, J., An Introduction to Islamic Law (Oxford, 1971), p. 124.Google Scholar ‘Majority is determined by physical indications, by the declaration of the youth in question, or, failing this, by reaching the age of fifteen lunar years.’ M. A. Cook has shown that among 751 boys from 51 Anatolian villages in the fifteenth/sixteenth centuries many were considered legally adults even before age 12 and all were considered so by age 15 (Population Pressure in Rural Anatolia, 1450–1600 [London, 1972], p. 64).Google Scholar

^{page 22 note 2} See Barkan, Ö. L., ‘Daftar-i Khakani’, Encyclopedia of Islam, 2nd ed.;Google Scholaridem, ‘Research in Ottoman Fiscal Surveys’, pp. 163–71, in Cook, Studies, pp. 163–71; idem, ‘Essai sur les données statistiques des registres de recensement dans l'Empire ottoman au XVième et XVIième siècles’, Journal of the Economic and Social History of the Orient, vol. 1 (1957), pp. 9–36.Google Scholar

^{page 23 note 1} Cook, Population Pressure, pp. 26 f., esp. p. 27 n. 1.Google Scholar

^{page 23 note 2} Ibid. pp. 26 f.

^{page 23 note 3} ‘These marked swings in average household size clearly indicate that the medieval household was not a rock of social stability but was very sensitive to what French historians call the conjuncture, the long-term economic trend’ (Herlihy, David, ‘Mapping Households in Medieval Italy’, The Catholic Historical Review, 8 ], 1–24).Google Scholar

^{page 23 note 4} Ibid. pp. 5 f.

^{page 23 note 5} Ibid.

^{page 23 note 6} Ibid. p. 5.

^{page 24 note 1} Ibid. p. 13.

^{page 24 note 2} The proportion of mucerred (unmarried adult males) among the population of the city varies from survey to survey, from city to city, and from mahalle to mahalle. Fertility rates, infant mortality rates, and other mortality rates can vary with time and place. Some cities offered greater economic attraction to immigrants, and the economic circumstances in some agricultural regions may have encouraged migration to cities while conditions elsewhere may have discouraged it. There was immigration from outside to at least some of the cities, and some of the immigrants helped to swell the proportion of mucerred. The defters do not provide information that enables one to judge for certain whether or not a mucerred was an immigrant. However, a mucerred who was the son or brother of another adult male living in the city was listed after the relative; in such cases the name of the mucerred's father was not usually repeated, but the relationship was usually noted. Some among the other mucerred may have had no living adult male relatives, but others were immigrants. It does not seem possible to distinguish more precisely, and it is not possible to determine whether married adult males were immigrants.Google Scholar

^{page 25 note 1} What are apparently the last defters for all the provinces, preserved in the Tapu ve Kadastro Dairesi in Ankara, come mostly from the period 1575–90. In 1609 Ayn-i Ali set forth his complaints about the government's failure to maintain the timar defters during the previous twenty or thirty years. The tapu defters must also have fallen into neglect about the same time, although it is hard to imagine how any successful government enterprise could have been carried out without up-to-date defters.Google Scholar

^{page 27 note 1} For this idea, see Akdağ, Mustafa, Türkiye'nin İktisadî ve İçtimaî Tarihi (Ankara, 1971), vol. 11, pp. 22 f.Google Scholar

^{page 27 note 2} On Kayseri, see Darkot, B., ‘Kayseri’, Islam Ansikiopedisi; R. Jennings, ‘Kayseri’, El’. For a map of the city and a restored plan of the city walls and inner citadel,Google Scholar see Gabriel, A., Monuments Turcs d'Anatolie, vol. 1, (Paris, 1931);Google ScholarGökbilgin, M. T., ‘XVI. asir başlarinda Kayseri şehri ve livasi’, in Zeki Velidi Toğan'a Armagan (Istanbul, 1950–1955), pp. 93–108.Google Scholar

^{page 27 note 3} Başbkanlik Arşivi, Istanbul, Tapu defter no. 33, Kayseri, Maliye; dated 906 (1500). Recorded in the defter at the end of the mahalle by mahalle enumeration of the adult male populace are the totals 2,293 nefer and 1,425 hane-i 'avariz. The number of nefer is virtually the same as my count but the number of hane differs by 423 hane. My figure of 1, 848 hane is derived by subtracting the 439 mucerred (unmarried, noted by ‘m’) from the 2,287 nefer (the total number of names listed). Perhaps the names of a number of tax exempt men (exempt from the 'avariz tax) were omitted.Google Scholar

^{page 27 note 4} Ermeni - Armenian Gregorian millet; Rum - Greek Orthodox millet. Muslims are identified only as such, not as Turks, Arabs, Kurds, Circassians, etc.; but it is unlikely that there were many non-Turkish Muslims in the cities studied here.Google Scholar

^{page 28 note 1} An excellent recent survey of the Karamanli problem, with good bibliographical references, is found in Vryonis, S., The Decline of Medieval Hellenism in Asia Minor (Berkeley, 1971), pp. 448–462. I use ‘Turkish’ names to designate names used by Turks, not just names of Turkish origin.Google Scholar

^{page 29 note 1} As shown below, zimmis were often identified with community (cema'et or taife) rather than mahalle. Perhaps the distinction is just an affectation, although in such cases it cannot be considered certain that they lived in one single community.Google Scholar

^{page 29 note 2} Başbakanlik Arşivi, Istanbul, Tapu defter no. 387, Karaman, Rum, and Erzurum vilayets, Maliye, evkaf (icmal), undated, reign of Kanuni. In his study of Trabzon, M. T. Gökbilgin uses the date 1523 (929) for this defter (‘XVI. Yüzyil Başlarinda Trabzon Livasi ye Doğu Karadeniz Bölgesi’, Belleten, vol. 26 [1962], p. 337).Google ScholarKonyali, I. H., in his study Karaman Tarihi, uses the date 1522 (929) ([Istanbul, 1967], p. 106). For convenience the date 1523 is used in this article. Written at the end of the icmal listing for Kayseri are the ‘totals’: 2,139 nefer, 1,344 hane-i Muslim, 429 hane-i gebran (zimmi). These figures are plainly not in accord with the sums of the total number of nefer and hane listed by mahalles; perhaps the names of some tax-exempt Muslims were omitted. The 133 muhafizan (soldiers of the castle) are included in the hane as well as nefer by my count, as are the ulema of Kayseri.Google Scholar

^{page 30 note 1} Başbakanik Arşivi, Istanbul, Tapu defter no. 976, Kayseri, Maliye, evkaf, undated. This defter was compiled after the one of 1523 and before that of 1583. Tax surveys sometimes were carried out at thirty-year intervals - in Trabzon, for example. Moreover, a date of about 1550 is consistent with the rate of change going on in Kayseri during the sixteenth century.Google Scholar

^{page 31 note 1} Tapu ve Kadastro, Ankara, Tapu defter no. 136, Kayseri liva, dated 1583 (992).Google Scholar

^{page 31 note 2} The twentieth-century increase in the population of urban Kayseri is at a faster rate than the sixteenth-century growth, but the modern growth has been accompanied by a large-scale migration from villages to city. Also, it comes at a time when widespread health improvements were introduced, greatly reducing the mortality rate without any contingent effect on the fertility rate.Google Scholar

^{page 34 note 1} For Chalcedonian Armenians of Kayseri (Haikhrum), see Vryonis, Decline, pp. 65 f., esp. n. 323;Google ScholarMordtmann, A. D., Anatolien, ed. Babinger, F. (Hannover, 1925), p. 492.Google Scholar

^{page 35 note 1} Konyali, Karaman, pp. 100 ff., 106 ff;Google ScholarBeldiceanu, N. and Beldiceanu-Steinherr, I., Recherches sur la Province de Qaraman au XVIe siècle (Leiden, 1968).Google Scholar

^{page 35 note 2} Başbakanlik Arşivi, Istanbul, Tapu defter no. 455, Konya, Maliye, evkaf, undated, reign of Kanuni. Also, Tapu defter no. 387, Kararnan, Rum, and Erzurum vilayets, Maliye, evkaf (icmal), undated, reign of Kanuni. The icmal, as mentioned, is dated 929 (1523) by Gökbilgin and 929 (1522) by Konyali. For convenience, 1523 is used here. Defter no. 455 must be dated shortly before 1523, for its contents are summarized in defter no. 387.Google Scholar

^{page 36 note 1} Of course it is possible that a mahalle named for a commercial function at some time in the past might cease to be the center of that function.Google Scholar

^{page 36 note 2} Tapu defter no. 104 in the Tapu ye Kadastro Arşivi, Ankara. Most of the information can be found in Konyali, Karaman, pp. 121 f., 103 if.Google Scholar

^{page 36 note 3} Thirteen of these mahalles survive to the present. Karaman has been remarkably stable and conservative (Konyali, Karaman, p. 135).Google Scholar

^{page 36 note 4} For some of Konyali's readings, see Karaman, p. 122.Google Scholar

^{page 37 note 1} For Amasya see B. Darkot and A. Gabriel/M. H. Yinanç, ‘Amasya’, IA; Fr. Taeschner, ‘Amasya’, EI²; Hüsameddin, Hüseyin, Amasya Tarihi, 5 vols. (Istanbul, 1330–2/1927–1935);Google ScholarGabriel, , Monuments, vol. 2.Google Scholar For a map of modern Amasya and restorations of the old city see ibid., vol. 11, pp. 13, 9, 10.

^{page 37 note 2} Başbakanlik Arşivi, Istanbul, Tapu defter no. 387, Karaman, Rum, and Erzurum, Maliye, evkaf (icmal), undated, but identified above as 1523 (929).Google Scholar

^{page 38 note 1} Tapu ve Kadastro, Ankara, Tapu defter no. 26, Amasya liva, dated 1576 (984).Google Scholar

^{page 39 note 1} See Akdağ, M., Celâlî İsyanlari 1550–1603) (Ankara, 1963), pp. 250–257.Google Scholar

^{page 39 note 2} Başbakanlik Arşivi, Istanbul, Tapu defter no. 776, Amasya, Tapu, dated 1642 (1052). The defter is badly damaged. Pages survive regarding only one-fourth of the population (459 nefer). It is not possible to calculate the proportion of Muslims and zimmis, although 80 per cent (365 to 90) of the surviving names were Muslim - more or less in accord with the earlier ratio. The approximate figure of 1,736 nefer can be derived in two ways: (1) the ulema and military class (i.e. tax-exempt) are given as 353 and the reaya (tax-bound) as 1,383 = 1,736 total; (2) elsewhere, the figure 5,767 nefer is given for the population of the province of Amasya and 4,031 nefer for that of the province exclusive of the city = 1,746.Google Scholar

^{page 40 note 1} Evliya Çelebi visited Amasya in c. 1640. He describes the city as large and flourishing: 5,000 inhabited buildings of all sizes, 1,060 shops, tradesmen of 160 crafts, rich orchards, gardens, and vineyards (Seyahatnamesi [Istanbul, 1314/1896], vol. 11, pp. 186, 189). The population he reports is far in excess of the figures in the defters, and the number of shops seems equally exaggerated. Yet his description of the site of the city and its buildings is so accurate in details that anyone who has visited Amasya must feel this is a first-hand report. Numerical exaggeration is a fault common among travelers, but the mere fact that Evliya describes Amasya as flourishing is disquieting - given the evidence of the defters. Evliya himself mentions how wealthy people (‘a‘yan ve eşraf’) of the city had hidden valuables in caves in the castle during the insurrections of the Celali, Kara Yazici, and Kara Said (vol. 11, pp. 184 f.).Google Scholar

^{page 40 note 2} Barkan, ‘Research’, p. 170. There were 28 Jewish hane at Ankara, 1520–35. Tokat was another exception.Google Scholar

^{page 42 note 1} For the development of this idea, see Vryonis, S., ‘The Conditions and Cultural Significance of the Ottoman Conquest in the Balkans’, p. 9.Google Scholar Paper presented at 2nd International Congress of South-east European Studies, 7–13 May 1970 (Athens, 1970), repr. in idem, Byzantium: Its Internal History and Relations with the Muslim World (London, 1971), chap. X.

^{page 42 note 2} Janssens, E., Trebizonde en Coichide (Brussels, 1969), pp. 170 ff.Google Scholar For Trabzon see also Şevket, Şakir, Trabzon Tarihi (Istanbul, 1294);Google ScholarKaradenizli, Kemal, Trabzon Tarihi (Ankara, 1954).Google Scholar See also Kortepeter, C. M., ‘Ottoman Imperial Policy and the Economy of the Black Sea Region in the 16th Century’, Journal of the American Oriental Society, vol. 86 (1966), pp. 86–113.Google ScholarBryer, A. A., ‘The Tourkokratia in the Pontos: Some Problems and Some Preliminary Conclusions’, Neo-Hellenika, vol. 1 (1970), pp. 30–54;Google ScholarBijişkyan, P. Minas, Karadeniz Ktytlari Tarih ve Coğrafyast, trans. Andreasyan, H. (Istanbul, 1969).Google Scholar Most important is an article by Gökbilgin, M. T., ‘XVI. yüzyil başlarinda Trabzon livasi ye Dogu Karadeniz bölgesi’, Belleten, vol. 26 (1962), pp. 293–338.Google Scholar

^{page 43 note 1} Başbakanlik Arşivi, Istanbul, Tapu defter no. 387, Karaman, Rum, and Erzurum (icmal), undated, reign of Kanuni. Dated above to 1523.Google Scholar

^{page 44 note 1} Başbakanlik Arşivi, Istanbul, Tapu defter no. 288, Trabzon, Maliye, evkaf, dated 1554 (961).Google Scholar

^{page 45 note 1} Tapu ve Kadastro, Ankara, Tapu defter no. 29, Trabzon liva, 1583 (991).Google Scholar

^{page 47 note 1} For Erzurum, see: Inalcik, H., ‘Erzurum’, EI ²; B. Darkot, M. Yinan¸, and H. Inalcik, ‘Erzurum’, IA,Google ScholarKonyali, I. H., Erzurum Tarihi (Istanbul, 1960);Google ScholarBeygu, Şerif, Erzurum Tarihi (Istanbul, 1936).Google Scholar

^{page 47 note 2} Başkanlik Arşivi, Istanbul, Tapu defter no. 387, Karaman, Rum, and Erzurum (icmal), undated, reign of Kanuni. Dated to 1523.Google Scholar

^{page 47 note 3} Başkanlik Arşivi, Istanbul, Tapu defter no. 205, Erzurum, Maliye, evkaf, dated 1540 (947).Google Scholar

^{page 48 note 1} Cf. the 1540 (947) Erzurum kanunname: because the province is on the frontier, most of the reaya are scattered and dispersed; when spahis collect taxes from them, they should not be registered in other places (Barkan, Ö. L.. Osmanli Imparatorlugunda Ziraî Ekonominin Hukukî ve Malî Esaslart,, vol. I, Kanunlar [Istanbul, 1943], p. 67, no. 29).Google Scholar

^{page 48 note 2} Tapu ve Kadastro, Ankara, Tapu defter no. 41, Erzurum liva, dated 1591 (1000).Google Scholar

^{page 50 note 1} He describes Erzurum as a rich and flourishing trade center with a huge military garrison. Although he does not estimate the population directly, he asserts that the city had 70 Muslim quarters and 7 infIdel ones, 70 mihrabs, 77 mescids, and 110 schools for boys, convents, and houses, 70 hanes, and 800 shops. He mentions 13 Armenian churches in the suburb of the Erzincan gate and states that Erzurum was the third busiest customs post in the Ottoman Empire, after Istanbul and Izmir (çelebi, Evliya, Seyahatnamesi, vol. 2, pp. 202–16).Google Scholar

^{page 55 note 1} Akdag, Celâli Isyanlart: (1550–1603), pp. 68–77.Google Scholar