• 1 Usually 1896 is given as the end-date, but there were two massacres in 1897 (in the Sivas vilayet), (...)

1The mass violence of the years 1894 to 1897 that started with the bloody repression of the so-called Sasun “revolt” in 1894, climaxed with attacks on Armenians all over Asia Minor in the autumn of 1895 and continued until 1897.¹ These comprised a traumatic series of events of unprecedented proportions. Often called the “Hamidian massacres” nowadays (referring to Sultan Abdülhamid II), Armenians before 1915 knew them as the “Great Massacres” (“medz charte” or also “medz yeghern”). Despite their enormous scope and significance, however, and an overwhelming number of primary sources, the Hamidian massacres are under-researched.

• 2 For example, R.G. Hovannisian, 1984; V. Dadrian, 1995; D. Bloxham, 2003; D. Bloxham, 2005; A.D. Mos (...)
• 3 S. Duguid, 1973; J. Verheij, 1998; J. Verheij, 1999; S. Deringil, 2009, on the forced conversions t (...)
• 4 There are two contemporary inventories: 1) an 1896 list by the Embassies of the Powers in Constanti (...)
• 5 To my knowledge, C. Mouradian, 2006 and my own articles (J. Verheij, 2012a and 2012 b) are the only (...)

2Scholars approach the events in question mostly within the framework of general histories of the Armenian Question or as “pre-history” to the First World War genocide.² Indeed, the number of studies exclusively devoted to the 1894-1897 massacres remains strikingly small. The existing literature consists of some research articles, while monographs are entirely absent.³ As a result, many key questions have not only gone unanswered but have barely been addressed. In particular, no satisfactory inventory of the incidents exists.⁴ There are only a couple of regional studies on specific cities or regions,⁵ and there is little analysis of the perpetrators and their motives.

• 6 Some examples include V.M. Kurkjian, 1964, p. 296; L. Nalbandian, 1963, p. 102; H. Pasdermadjian, 1 (...)
• 7 E.g. K. Gürün, 1985, p. 147-150; Belgelerle Ermeni Sorunu, 1992, p. 108; Y. Halaçoğlu, 2002, p. 28; (...)
• 8 For a more extensive treatment of the historiography on the Hamidian massacres see J. Verheij, 2012 (...)
• 9 R. Melson (1982) considers possible organization by the Ottoman state, but not in any depth. L. Nal (...)

3In addition to the paucity of research, there exists a strong dichotomy between an Armenian/Western perspective and an Ottoman/Turkish one. In the mainstream Armenian/Western perspective the central theme is that the violence against the Armenians was organized or at least triggered by the Ottoman authorities.⁶ The (official) Turkish historiography describes the 1894-1896 events as a series of revolts by Armenian revolutionaries.⁷ While in the Turkish tradition the violence against Armenians is either downplayed (to the point that the term “massacres” is avoided) or described as “provoked” by Armenian action, the Armenian/Western tradition has tended to minimize the role of the Armenian revolutionary organizations.⁸ Few have tried to explain and reconcile these almost diametrically opposed positions. Of course, the lack of thorough research is one of the factors at the heart of this impasse; neither the role of the Ottoman authorities nor that of the Armenian revolutionaries has been well studied.⁹

• 10 J. Verheij, 2012a and 2012c.
• 11 English: Hizan/Khizan, Shirvan/Sherwan. Turkish spelling is generally preferred for contemporary pl (...)

4The research reported here takes a micro-historical approach. It builds on the assumption that the wider issues can be better understood through a close study within a limited geographical frame, one that enables an identification of the various actors, local and external. For the 1895 crisis in Diyarbekir, my previous work focused on a single city and its surroundings and yielded results that questioned both the Armenian/Western and Ottoman/Turkish accounts.¹⁰ This study demonstrated that in Diyarbekir, Armenian unrest and protest were a more important factor than shown in the existing accounts, and that, interestingly, this had not drawn the attention of the defenders of the Turkish provocation theory. Secondly, my research indicated that the anti-Armenian violence, as far as could be ascertained, emanated primarily from a local level, actively supported by the local Ottoman authorities, with the central state mainly occupying a mitigating role. In fact, the actions of the local Muslims bore the characteristics of a protest against the central state. The 1895 events in Diyarbekir, thus seemed to be more of a Muslim revolt than an Armenian one, challenging the authority of the Sultan in the city and region. The results of this research on the Diyarbekir region both showed the value of a micro-historical approach and, given the great differences in the character of the various regions in the eastern provinces, also indicated that much more local research of the same kind was needed. Consequently, this article is devoted to another geographical area, and one that is even more limited: two rural districts in the vilayet (province) of Bitlis, namely, Hizan and Şirvan.¹¹

• 12 In a British report of March 1896 the total number of killed in the sub-province (sancak) of Siirt (...)
• 13 British National Archives [BNA], Foreign Office [FO] 881-6820 710 (Currie to Salisbury, 12 Dec. 189 (...)

5These districts were selected for this study for three main reasons. First, researchers have neglected the events in the countryside even more than those in the cities. Second, there is a notion in the British sources that the massacres in Hizan and Şirvan were particularly serious.¹² In fact, the British government thought that the extent of the forced conversion to Islam there was so serious that it employed it as an argument in trying to induce the other Powers into some kind of concerted action against the Ottomans (which was not forthcoming, as the Powers remained remarkably inactive in 1895).¹³ Finally, I came upon detailed British investigations of the events in these two districts, located in files compiled two years after the massacres.

• 14 In recent years, this notion has been frequently appearing on social media. For an example of the u (...)
• 15 Ottoman documents published by E.Z. Ökte (1989) suggest that a Hamidiye regiment from the Hayderanl (...)
• 16 J. Verheij, 2012a, p. 134-135. Other than the terminological confusion, the view that the Hamidiye (...)

6This article thus sets out to reconstruct the events in a geographical area, aiming at an inventory of incidents and estimate of the scale of the violence and number of victims. This approach also offers an opportunity to explore several other neglected issues, including the identity of the perpetrators, their motives and their relations with the authorities. In the process, I further try to find clues to the important question of government involvement. Since Hizan and Şirvan were rural areas where the Muslims were all Kurds, light is also shed on the role of the Kurdish irregular regiments, the Hamidiye, as well as of the Kurds in general. There is a popular notion that the Hamidiye regiments were among the prime perpetrators in 1894-1897, prompting some even to call the events the “Hamidiye massacres.”14 However, it seems that the Hamidiye played little or no role in the Sasun massacre of 1894.15 In Diyarbekir, similarly, my research did not reveal prominent Hamidiye activity.16 Finally, particularly because of the strong emphasis in Turkish “official” historiography on the role of the Armenian revolutionaries, I also seek evidence for any Armenian protest movement and its role.

• 17 BNA FO 195-2021, p. 158v, Report on a Journey in the Cazas Sherwan, Sairt and Aroh, May and June 18 (...)
• 18 R. Kévorkian and P. Paboudjian, 1992 (hereafter abbreviated KP), p. 506. I encountered the same phe (...)

8Hizan and Şirvan were adjacent districts (kaza) in different sancaks (sub-provinces) in the vilayet of Bitlis; Hizan was situated in the merkez sancak (central sub-province, also named Bitlis), and Şirvan in the sancak of Siirt. In terms of ethnic composition, Şirvan was a predominantly Muslim Kurdish district. Only  28 of the 200-odd district settlements were Christian, although the Christian villages were generally larger, so some 20% of the district population seems to have been Christian.¹⁷ A large proportion of the Christian villages in Şirvan were Assyrian (Suriyani, Syrian Orthodox), which was not always recognized in Armenian sources of the time.¹⁸ Hizan, however, was much more Armenian in character.

• 19 Ottoman census of 1914, in K.H. Karpat, 1985, p. 174-175.
• 20 Report of Boghos Natanyan, in A. Yarman, 2010, vol. 2, p. 221-247, passim.
• 21 MO, p. 158v-159 (Şirvan), V.T. Mayewsky, 2007, p. 389-395. According to an article in an 1881 issue (...)

9According to the Ottoman government in 1914, Hizan had a population of about 70% Muslims and 30% Armenians.¹⁹ Given the high number of Armenian settlements in Hizan, this looks like an undercount, and in view of the sizeable emigration before 1895,²⁰ the actual proportion may well have been 50% or higher. Mixed Christian-Muslim villages were rare in Şirvan, but more numerous in Hizan.²¹ While a large part of the Kurdish Muslim population was sedentary and the villages in the area inhabited yearlong, Şirvan and Hizan were also both “home” to a number of nomadic Kurdish tribes, which used the high mountain meadows in the summer and wintered in lower regions outside the districts.

• 22 For the centralization, see M. Van Bruinessen, 1992, p. 175-184; H. Özoğlu 2004, p. 59-68; S. Aydın (...)
• 23 A. Safrastian, 1948, p. 52 claims that the defeat of Bedirhan Bey was accelerated because the beys (...)

10Both Hizan and Şirvan suffered a lot from the Ottoman centralization in the 1830s and 1840s, when local Kurdish overlords (bey-s or mir-s) were driven out and replaced by direct Ottoman administration.²² In the Hizan area, there were two dynasties of beys, those of Hizan and those of Sparkert/İspayirt, both probably already weak and largely defunct at the beginning of the 19^th century. The beys of Şirvan, described as vassals to the famous Bedirhan Bey of Bohtan, were more important than the Hizan and Sparkert beys.²³ The new Ottoman administration was largely nominal, however, and in Hizan and Şirvan, even late into the 19^th century, there were just a few officials headed by a kaymakam (district governor), with few troops or police at their disposal and barely able to enforce the law.

• 24 MO, p. 159, 172.
• 25 M. Van Bruinessen, 1992, p. 296, 337-338; B. Dickson, 1910, p. 370.
• 26 MO, p. 160v, 174, 187, 194v; B. Dickson, 1910.

11Although ousted, the old Kurdish notables were not completely diminished, at least in Şirvan, and their descendants still commanded respect in the area at the end of the 19^th century.²⁴ First among the groups that gained power were the sheikhs (şeyh-s) of the influential Nakşibendi order. In Hizan, these sheikhs had their center in Gayda, close to the district’s administrative center, Karasu, and strongly influenced politics even in the provincial capital Bitlis until well after WWI.²⁵ In Şirvan, meanwhile, there were influential sheikhs in the village of Verkanis (Kasımlı) and in Tillo, outside and directly to the southwest of the district.²⁶ By marriage and in other ways, the sheikhs were highly connected to members of the order in other areas, notably with those in Norşin (Güroymak), north of Bitlis, Arvas in Müküs/Moks (nowadays Bahçesaray) and Şemdinan (Şemdinli) in Hakkari.

• 27 Notably the Miran, “the most formidable of all the nomad tribes” (MO, p. 189v); on the smaller trib (...)

12After the sheikhs, the second group to gain ascendancy following the centralization were the Kurdish (semi-)nomadic tribes. These were the nightmare of the sedentary population, due to their constant demands and exactions. The tribes in Hizan and Şirvan were typically small, although some of the larger tribes from elsewhere passed through the area seasonally.²⁷

• 28 S.H. Astourian, 2011, p. 59.

13Thus, the Ottoman centralization essentially created a power vacuum in the region, one that was partly filled by the Nakşibendi sheikhs and rendered it vulnerable to the Kurdish tribes. This may be characterized as a period of turmoil, therefore, only exacerbated by Tanzimat legislation, like the Land Code of 1858, which opened the way for a land grab.²⁸ All formal and informal power remained in the hands of Muslims, however. The Christian population, both Armenian and Syriac-Orthodox/Assyrian, had no administrative role, except for a quite meaningless representation (some notables placed in the local administrative councils created by the Ottoman bureaucratic reform). So, it was against this background of uncertain authority and shifting power that the Ottoman-Russian wars of 1853-1856 and 1877-1878 were fought, which fueled a generalized hatred of Christians among the Muslim population(s).

• 29 MO, p. 162, 165v, 178v.
• 30 MO, p. 162. Similar observation for Hizan: Report on a Journey through the Caza of Khizan in Septem (...)

14Both Hizan and Şirvan were mountainous districts, with numerous valleys and some plains separated by mountain ranges, forming isolated micro-worlds that differed in terms of culture and the character of their peoples. They were not arid districts, and some parts were quite wooded. Due to its lower altitude and hotter climate throughout the year, the south of Şirvan, in particular, was known for its horticulture and high-quality fruit. Sedentary peasants practiced agriculture, horticulture and husbandry. Christians (Armenians and Assyrian) also engaged in cloth production, and in some villages, every household had a loom.²⁹ Their customers were primarily their Kurdish neighbors, who in this respect were thus dependent on the Christians. Indeed, an informed British visitor noted that “the Kurds, with the exception of some of their women, are as a rule incapable of making anything.”³⁰

• 31 J.-M. Thierry, 1970, p. 164-186; Id., 1973-1974, p. 202-232. Three monasteries still operated in th (...)
• 32 KP, p. 475-477, list eight village schools in Hizan; MO has no Armenian schools in Şirvan, while Ké (...)
• 33 Notably those of Tağ in İspayirt, Hizan, and in Tillo, southwest of Şirvan.
• 34 MO, p. 159.

15Although there were several medieval Armenian monasteries in Hizan, testifying to a richer Armenian past,³¹ by the 19^th century, Hizan and Şirvan were relatively underdeveloped districts. Thus, while Armenian Church and societies dedicated to the development of education elsewhere had a large network of village schools, in Hizan, only around 10% of the Armenian villages were reported as having schools, and in Şirvan they were virtually non-existent.³² With just a few medresses³³ and small Koran schools offering classical religious and linguistic education, education for the Muslim community was just as poor. British Vice-consul Monahan wrote of Şirvan that “the people in general have remained more ignorant and backward than those of any part of the Vilayet outside the Sairt Sanjak.”³⁴

• 35 MO, p. 174v, 175 (on Syriacs in Şirvan), 199v (on Kurds in Şirvan); CR, p. 209 (on Armenians in Hiz (...)
• 36 Apparently, there was never “any seditious or political agitation or movement among the Christians (...)

16The people of Hizan and Şirvan did have a lot of contact with the outside world, however, since many Kurdish and Christian men worked seasonally in Constantinople.³⁵ This may have been one of the reasons for the support among peasants of the Armenian revolutionary movement – in which respect, the proximity of Van, an important center of Armenian political activism, may have been crucial. Support for the revolutionary movement generally seems to have been more widespread in Hizan, particularly in Spargert/İspayirt, than in Şirvan.³⁶

• 37 Notably, there are reports in the Armenian Patriarchate archives (some of them used by R. Kévorkian (...)
• 38 E.L. Cutts, 1878, p. 116-121; H. Rassam, 1897, p. 102-103; F.R. Maunsell, 1904, vol. 3, p. 489-493 (...)
• 39 T. Kotschy, 1860; J. Wünsch, 1890.
• 40 MO, p. 159v.
• 41 See CR. Before 1914 more British representatives visited the region (B. Dickson, 1910, p. 368; F.R. (...)

17Şirvan and Hizan are largely terra incognita. Reflecting the weak influence of the state in the area, Ottoman sources are scarce, and while for other regions there exist rich repositories of Armenian regional histories, the so-called houshamadyan genre, for Şirvan and Hizan these seem virtually non-existent, only compensated for by some contemporary Armenian Church sources.³⁷ There are also few travel reports from foreign travelers. Only one of the horse-tracks from Bitlis to Siirt ran through the region, passing by the west of Şirvan and its administrative center of Küfre. Some travelers described that route.³⁸ Before 1895, only two foreign travelers seem to have explored the rest of Şirvan and Hizan: the Austrian botanist Theodor Kotschy, and the Czech geographer Josef Wünsch.³⁹ When Monahan travelled through Şirvan in 1898, he was under the impression that “most of Sherwan seems never to have been previously visited by any European.”⁴⁰ The Vice-consul’s report on Hizan from the previous year was a similarly pioneering work.⁴¹

• 42 CR, p. 210.

18The chaotic situation of the 19^th century took a heavy toll on the sedentary population in the area. State authorities collected revenues through an oppressive tax-farming system, while the nomads frequented villages with their own incessant demands, including for unofficial “taxes.”⁴² The sheikhs were active as tax-farmers and money-lenders, not hesitating to demand excessive interest rates or forced labor. In Şirvan a system called miribalik was implemented, in which some peasants paid half of their harvest as interest, inevitably leading over time to the loss of their land to the creditors.

• 43 Temettü was a tax introduced in the Tanzimat period, on (estimated) business profits. Monahan showe (...)
• 44 A. Yarman, 2010, p. 225; CR, p. 208; M. Van Bruinessen, 1992. The most famous of the sheikhs, Celal (...)
• 45 E.L. Cutts, 1878, p. 188; T. Kotschy, 1860, p. 75; MO, p. 161v.
• 46 A. Yarman, 2010, p. 241; MO, passim.
• 47 For cases of emigration of Kurds from Şirvan see MO, p. 167-168. Unfortunately, we do not have good (...)
• 48 Armenian school inspector Mirakhorian, who extensively travelled in many of the eastern provinces, (...)

19Although the peasantry in general suffered, Christians were subject to further exactions. To the state they paid an extra tax for exemption of military service, those engaged in textile production also paid the special temettü tax.⁴³ Moreover the sheikhs seem to have specifically targeted non-Muslims. The Hizan sheikhs, who were known for their religious fanaticism and hatred of Christians, exerted a localized reign of terror.⁴⁴ As a consequence of oppression and economic hardship, many peasants sought employment elsewhere, by seasonally migrating, or fleeing the land forever. By all accounts, migration was already extensive before the 1880s. Locals told travelers stories of a glorious past when the villages were much larger and prosperous.⁴⁵ In the two decades before 1878, the Armenian rural population of Hizan was said to have declined by 80%.⁴⁶ Since none of the regional capitals – Diyarbekir, Bitlis or Van – showed any remarkable population growth during the 19^th century, presumably the final destinations of the migrants were faraway places where many other Armenian emigrants went too, like Russian Armenia, other territories of the Russian Empire or Constantinople. Although it is impossible to ascertain the extent to which Kurdish peasants from Hizan and Şirvan also left their land, we may assume that the majority of the emigrants were Christian.⁴⁷ Hence the Kurdish population percentage increased after the centralization. A gradual process of Kurdification was also observed in nearby districts in the province of Van.⁴⁸

• 49 Letter from Bitlis, 15/27 November, 1895, in A. Tchobanian, 1896, p. 73-79 (hereafter abbreviated M (...)
• 50 Notably “Lettre de Sa Béatitude le Catholicos Khatchadour d’Akhtamar,” 19/31 Déc. 1895, in A. Tchob (...)
• 51 Regularly detailed reports of the Patriarchate were included in the British Confidential Papers (pr (...)

20Sources for the study of the 1895 events in Hizan and Şirvan are far from abundant, although the situation is not much worse than for other rural districts. Ottoman sources are fragmented, but offer some interesting information that largely corroborate other sources. The only Armenian source for Şirvan, in 1895 that I was able to locate was a letter from an anonymous correspondent in Bitlis, probably a clergyman.⁴⁹ For Hizan there are more Armenian accounts, because this district was under the jurisdiction of the Catholicos of Akhtamar. His staff produced some detailed reports.⁵⁰ An examination of the unexplored archives of the Istanbul patriarchate is required.⁵¹

• 52 Ottoman administrators accused Crow of contacting Armenian revolutionaries (BOA DH.TMIK.M. 71-45).
• 53 As recorded in MO and CR. It is unlikely that other foreign powers produced such detailed reports, (...)

21Particularly for the late 19^th century, British sources are among the most detailed for this area. For strategic reasons (mainly a possible Russian advance in the direction of the Mediterranean) Britain had a tight network of consulates in the eastern provinces. In its self-assumed role of protector of the Christians in the area Britain had long since closely followed the situation of the Armenians. In the autumn of 1895, there were British representatives in Van and Muş. Occupied with the serious disturbances in the cities, and unable to travel, they could not visit Şirvan or Hizan. Therefore, British sources on the districts appear disappointing at first sight; however, valuable reports exist for later years. Eyed with great suspicion by the Ottoman authorities,⁵² British Vice-consuls Monahan and Crow toured both districts in 1897 and 1898 and collected important data on the massacres and their consequences.⁵³

• 54 MO p. 159.
• 55 BNA FO 881-6775 91/1.
• 56 Ibid. 65/3, Substance of a Letter from the Catholicos of Akhtamar.
• 57 Ottoman report BOA Y..A….HUS 320-18, 17 Febr. 1895.
• 58 Some reports: BNA FO 881-6695 167/4, Evidence of Karapet Garzariantz, 167/7, Evidence of Arakya Har (...)

22Looking back in 1898, British Vice-consul Monahan characterized the situation in Şirvan on the eve of the autumn 1895 massacres as relatively quiet, saying that “the people, Kurdish and Christian, were living together in tolerable peace and comfort.”⁵⁴ Contemporary sources do not confirm that impression, however. A report on Şirvan made by the Armenian Patriarchate in July 1895 mentions various sorts of oppression of Christians in the preceding years, highlighting cases of forced conversion to Islam, murders and emigration.⁵⁵ Similarly Hizan was already in a state of relative disorder. A report by the Catholicos of Akhtamar contains dozens of cases of plunder.⁵⁶ In some European newspaper reports, it was even rumored that Hizan might be the next Sasun.⁵⁷ Actually, the unrest was probably triggered by events in Sasun, which prompted violent Kurdish attacks on Armenian villages, perpetrated by Hamidiye, which occurred in a much wider area of the provinces of Bitlis and Van (particularly in the Muş plain, and the Ahlat, Adilcevaz and Erciş districts) in the Autumn and Winter of 1894, incidents that largely escaped the attention of external observers.⁵⁸ Şirvan and Hizan, however, seemingly were not affected by these series of attacks to the same degree.

• 59 KP, p. 506.
• 60 This register has been published (I. Beheiry, 2009), but the original lists are used here, kindly p (...)
• 61 V.T. Mayewsky, 1904. The Turkish edition of B. Bayraktar (2007) was used. This edition, in turn bas (...)
• 62 Neither the Ottoman state nor the ecclesiastic authorities ever used a standard spelling of place n (...)

24For an inventory of incidents (below), a two-stage process was followed. First, as complete as possible a list of all Christian villages was made, using the above-mentioned reports of Monahan and Crow and the work of Kévorkian and Paboudjian,⁵⁹ a register of Syrian-Orthodox in Şirvan by the Syriac priest ‘Abd-Allah Sattuf of Sadad⁶⁰ and data provided by Mayewsky.⁶¹ No attempt is made at any standardization of names, the contemporary official name is added in italics, where possible).⁶² Second, all the data located on the 1895 massacres were categorized per sub-district and village. For each sub-district, a characterization of the events, and if possible, an estimate of the number of victims and the identity of the perpetrators is given.

• 63 Natanyan describes this sub-district as belonging to Hizan (A. Yarman, 2010, vol. 2, p. 241).
• 64 The village of Madar (Sarıtepe), listed by KP (p. 506) under Şirvan and said to have lost its entir (...)

25There were four clusters of Christian habitation: 1) the district center of Küfre; 2) a north-western group, in an area presently in the district of Baykan; 3) a southern group, on the north side of the Bohtan Valley, bordering on Pervari; 4) the valley and sub-district (nahiye) of İrun (now Cevizli sub-district),⁶³ to which is added the isolated village of Maden, to the south.⁶⁴

• 65 Kefra/Koufra (ibid., p. 506), Kufra (MO, p. 169), Kafrah (SA).
• 66 MO, p. 169-169v. The tribe names are the (anglicized) forms as they appear in the text.
• 67 KP, p. 506; MO, p. 168v-169v.

26Although it was the administrative center of the kaza, Küfre was nothing more than a village. It had a mixed population of Kurds, Syrian Orthodox and some Armenian families. It was attacked by nomads of the “Mahometan” and “Strugan” tribes,”⁶⁶ reportedly allowed by Fatha Bey, a local notable and acting kaymakam. All the Christian houses were reported as plundered, with 25 Armenians and Syriacs killed.⁶⁷

• 68 Gounde-Déghan (KP, p. 506), Gundediesan (MO, p. 195), Kourtétizan (MT p. 75), Gunde-Dizal (SA), the (...)
• 69 Syrian-Orthodox according to MO, p. 195 and SA, but Armenian-speaking Nestorian according to KP, p. (...)
• 70 MO, p. 195.
• 71 MT, p. 75.
• 72 MO, p. 195-195v.

271. Kondudizan (Suludere)⁶⁸: a Syrian-orthodox village⁶⁹ attacked by Kurds of the “Demli” tribe in November 1895, with local Kurds partaking in the plunder.⁷⁰ According to an Armenian report, half of the population was massacred, and the headman burned alive;⁷¹ Monahan however, says 12 were killed, including the village priest, and three families died from hunger or disease after the massacre; the rest of the population converted to Islam, but reconverted in the summer of 1896.⁷²

• 73 Derzin (KP, p. 506), Dersen (MO, p. 161).
• 74 MO, p. 161v-162.
• 75 KP, p. 506.

282. Derzin (Adakale)⁷³: a collective name for three settlements, only one Armenian populated. The place was probably attacked by Motki and “Atmangli” Kurds; all villagers converted to Islam, but after their migration to Bitlis, reconverted in 1897. There were no casualties reported.⁷⁴ According to Raymond Kévorkian, the Armenian population disappeared in 1895, which would mean they never returned from Bitlis.⁷⁵

• 76 Mnar (KP, p. 506.), Minar (MO, p. 166v, SA), Menar (MT, p. 76).
• 77 KP, 1992, p. 506.
• 78 MO, p. 166v, SA.
• 79 MT, p. 76.
• 80 KP, p. 506
• 81 MO, p. 166v-167

293. Minar (Dilektepe)⁷⁶: another village named Armenian in Armenian sources⁷⁷, but Syrian-orthodox by others.⁷⁸ According to one source it was completely ruined, and no inhabitants were left because of massacre, plundering, and kidnappings.⁷⁹ Kévorkian and Paboudjian assert that the village lost its population in 1895.⁸⁰ Monahan heard that the inhabitants were not molested, but he could not speak with them when passing the village.⁸¹

• 82 Kikan/Garvé (KP, p. 506), Gigan (MO, p. 164v), Guiguan (MT, p. 75).
• 83 MT, p. 75.
• 84 MO, p. 165-165v.

304. Kiğan (İkizler):⁸² an Armenian village where it is claimed that a massacre took place, with half of the population killed.⁸³ According to Monahan, who visited the village, it was attacked and plundered by Kurds from Motki (possibly the Atmangli tribe, since they continued to harass the village afterwards), but no lives were lost and no-one converted. Four men from the village, who worked as muleteers, were killed elsewhere, however, and 10 of 30 households left the village after 1895.⁸⁴

*

• 85 Guéli/Kélin/ Kialoum (KP, p. 506), Gheli (MO, p. 181).
• 86 MO, p. 182-182v. Two “Nestorians” families mentioned by KP evidently not found by MO.

321. Gili (Durankaya)⁸⁵: a small Armenian village of six houses that was protected by Kurds. No plunder or conversion. Monahan found the community relatively prosperous during his visit in 1898.⁸⁶

• 87 Birké/Perg (KP, p. 506), Birkeh (MO, p. 183).
• 88 MO, p. 183-184v. Five “Nestorian” families according to KP, p. 506, unmentioned by MO.

332. Birki (Yatağan, a hamlet [mezra] linked to Taşlı village)⁸⁷: Armenian village of 14 households, pillaged by the followers of a sheikh from Siirt. All Armenians converted to Islam but reconverted a year later. Eight families left the village.⁸⁸

• 89 Dérik/Térèg/Dirik (KP, p. 506), Direk (MO, p. 171).
• 90 The Armenian Patriarchate (FO 881-6775 91/1-3) claimed that the village counted formerly 50 househo (...)
• 91 MO, p. 171-172v.

343. Dirik (Ayram, mezra of Taşlı village)⁸⁹: a small Armenian village of four households, but reportedly formerly much larger,⁹⁰ attacked, pillaged, and destroyed, with two people killed. The survivors converted to Islam, and in 1898 were still nominally Muslim. The attackers are not mentioned.⁹¹

• 92 Khandak (KP, p. 506), Hindik (MO, p. 171v).
• 93 MO, p. 171v.

354. Hendek (İncecik, mezra of Belençay village)⁹²: an Armenian village, attacked, but no details available. The inhabitants converted to Islam and were still Muslim in 1898.⁹³

• 94 Smkhor (KP, p. 506), Simkhor (ibid.; MO, p. 180; SA).
• 95 Syriac village (MO, p. 180; SA); mixed Armenian/“Nestorian” (KP, p. 506).
• 96 MO, p. 180-181.

365. Simhor (Sarıdana)⁹⁴: described as the largest Christian village in the region, with 50 houses.⁹⁵ Not attacked in 1895, due to a hired Kurdish guard.⁹⁶

• 97 Guérian (KP, p. 506) Gurena (MO, p. 181) Kourina (MT, p. 75).
• 98 MT, p. 75; MO, p. 181v-182.

376. Kevijan (Yolbaşı)⁹⁷: formerly, a larger Armenian village of 20 houses, included in a Bitlis correspondent list of villages where half of the population was killed. According to Monahan there were no victims, only the church was pillaged and the population survived by converting to Islam but reconverted a year later.⁹⁸

• 99 Nibin, Nebayn (KP, p. 506), Nub’ain (MO, p. 196v), Napaine (MT, p. 74).
• 100 MO, p. 197-197v; MT, p. 74.

387. Nibin (Turgutlu, mezra of Nallıkaya village)⁹⁹: a relatively large Armenian village of 40 households. Scene of a terrible massacre, killing 33 people. Many of the survivors (38 households) fled to Bitlis, two remaining families converted to Islam. The Ottoman authorities sent the refugees back, but evidently they fled again, since in 1898 the village only counted seven households. The attackers were Kurds, without specification.¹⁰⁰

• 101 Kosakh (MO, p. 177v), Kosikh (SA).
• 102 MO, p. 177v-179v.

398. Kosih (İncesırt)¹⁰¹: a Syrian-Orthodox village. The villagers lost all their cattle, but were “protected” by an agha of a neighboring village, who demanded that the population convert to Islam. They reconverted in 1896.¹⁰²

• 103 Merj (MO, p. 196v) al-Marj (SA).
• 104 MO, p. 196v.

409. Merç (Suluyazı)¹⁰³: Syrian-Orthodox village. Two people killed, all 9 families converted to Islam, 7 families fled, 2 remained and were still Muslim in 1898. Some Kurdish families settled here.¹⁰⁴

• 105 MO, p. 197v; BOA DH.TMIK.M 43-15, 13 Nov. 1897.

41With the exception of the massacre in Nibin, relatively little bloodshed occurred here. Total number of victims: 39 (including four killed in 1897).¹⁰⁵

• 106 MO, p. 197v. According to some Armenian sources they were Armenian, but according to other sources, (...)
• 107 MO, p. 197v-199; MT, p. 74-75.

42A secluded valley in the north-eastern corner of Şirvan, on the border with Hizan, this district had a majority Christian population. According to Monahan, 11 of the 16-17 villages here were Syrian-Orthodox, two Armenian and three or four Kurdish. Armenian sources mention at least four Armenian villages.¹⁰⁶ One night in November, all the villages suffered a combined attack from three Kurdish tribes, comprising the Atmangli together with the “Modeki” and “Demli.” Reportedly, the local Kurdish inhabitants did not take part in the plunder and tried in many cases to protect their Christian neighbors. According to Monahan a total of 73 people were killed. The anonymous correspondent from Bitlis states that massacres took place, without supplying victim numbers, but adding that after the massacres, 23 people died from hunger in the village of Zınzık (Oya). It is unclear whether Monahan took note of that. Reportedly nearly all the survivors fled, but 18 families who stayed behind converted to Islam. They reconverted in April 1896, with the exception of the inhabitants of one small village, Direk (not the village of the same name in the south of Şirvan listed above).¹⁰⁷

• 108 Maden (MO, p. 174v, MT, p. 75), Ma’dan (SA).
• 109 MT calls the village Armenian, but all other sources have it as Syriac-Orthodox (MO; SA; T. Kotschy (...)
• 110 MO, p. 175; MT, p. 75
• 111 MO, p. 177.

43Maden (Madenköy):¹⁰⁸ a Syrian-Orthodox village¹⁰⁹ south of the İrun sub-district; it was attacked by “Demli” Kurds, plundered and burned, and 20 inhabitants were killed.¹¹⁰ Ten of the twenty families migrated to Bitlis, the rest stayed in houses rebuilt by the authorities.¹¹¹

• 112 In the late 19^th century, the upper part of this valley was placed in the neighboring (now defunct) (...)

45Contemporaries usually divided Hizan thus: (1) Hizan center, in the northwest of the district, with its administrative center, Karasu; (2) the lower part of the Gargar Valley in the extreme northeast of the district;¹¹² (3) Shenadzor, a valley to the south of Gargar; (4) Sparkert/İspayirt, the dominions of the ancient principality with that name, bordering on the district of Müküs/Moks (Bahçesaray); and (5) Mamerdank (Mamotank/Mamertunk) in the southeast of Hizan, bordering on Pervari. Some distinguish a separate district, Nzar, in the thinly populated highland of the southwest of Hizan; here, it is included in İspayirt. Hizan counted at least 75 Christian villages, all Armenian, although some of them, particularly in the northern and north-western part of the district, also had a Kurdish population.

• 113 CR, p. 212. MTAgh lists 30 villages, but including the villages of Shenadzor, discussed here separa (...)
• 114 CR, p. 212-213.

46Vice-consul Crow counted six Armenian villages here.¹¹³ They were attacked and plundered by Kurds from the area, and, according to Crow, 15 Armenians killed. This was the heartland of the sheikhs of Hizan. Under their pressure the Armenians converted to Islam, but with few exceptions reconverted a year later. Forty families fled the district. Crow highlights the unbearable oppression of the remaining Armenian villagers by the sheikhs after 1895.¹¹⁴

• 115 Sdorin Gargar (KP, p. 476), Yeghehis (CR).
• 116 MTAgh, p. 110-111, specified per village; CR, p. 209.

47In this district of 10-11 Armenian villages, according to the report of the Catholicos of Akhtamar, some 65 people were killed. Crow did not mention any victims. The attackers seem to have been Kurds from the Aliganli tribe. All surviving Armenians converted to Islam on the instigation of a local agha, a supporter of the sheikhs.¹¹⁶

• 117 Şenatsor (NT) Chentsor (KP, p. 476) Shinetsor (CR).
• 118 MTAgh, p. 109, CR, p. 209.

48According to the Catholicos, some 400 Armenians were killed here and in the central district of Hizan. In contrast to his data on other districts, no specification per village was offered, suggesting this was an estimate. Crow mentions 17 dead in this district. There was considerable plundering. The main perpetrators were again the Aliganli, who also attacked local sedentary Kurds. Followers of the sheikhs prevented further bloodshed on condition that the Armenians converted to Islam. When Crow visited the villages in 1897, he found local Kurdish guards there, and that the Armenians spoke “well of their Kurdish neighbours.”¹¹⁸

• 119 Sparkert (KP, p. 476), Spaguerd (MTAgh), Spiert (CR), Sıyapgerd/İspaygerd (NT); and Nzar (KP, p. 47 (...)
• 120 MTAgh, p. 106-109; CR, p. 209-210.

49The Catholicos mentions 300 dead in the 28 villages of the district, based on a count per village. Curiously, following his visit to the district, Crow concluded that it “seems to have suffered less during the disturbances of 1895 than any other part of the Vilayet of Bitlis.” He claims that 11 of the 25 villages converted to Islam, “while the rest were unmolested, and the proportion plundered [did] not exceed two or three.” Crow thought this was due to fatigue of the attackers. He noted however that the villages that were attacked suffered considerably, citing two instances, where a total of 25 people were killed.¹²⁰

• 121 Mamedank (KP, p. 476), Mamerdan/Nimran (CR), Mamırdunk (NT).
• 122 MTAgh, p. 110, CR, p. 210-212. We were shown this village by local people during a visit to Hizan i (...)

50Crow counted 12 Armenian villages in this district, the Catholicos 17. According to the latter, 160 Armenians were killed, while Crow gives a total number of 30, which seems too low since he mentions 26 victims in two villages only. More than 60% of the Armenians (100 of 163 families) fled the district, and at least one village, on the plain of Ov, was deserted forever.¹²²

51Crow described in detail the fate of the famous monastery of the Holy Cross of Aparank in Mamedank, close to the border of Hizan with Pervari. Due to the unsafe situation, the abbot had already left six years before, and a caretaker looked after the monastery. In 1895, Dahar Agha of the Aliganli, “described by the Kurds themselves as being the highest brigand of the neighbourhood,” took up his quarters in the monastery, compelled the caretaker and his family to convert to Islam and occupied both the monastery and the lands. When Crow visited the monastery two years later, he found it to be occupied by “Kurdish families” and the converted caretaker:

• 124 CR, p. 212; MTAgh, p. 108, 110-112; J.-M. Thierry, 1970, p. 164, 167; Id., 1973-1974, p. 228. For c (...)

52Aparank was not the only monastery in the area permanently abandoned in 1895. The year was the death blow to a number of monastic establishments that had historically been at the center of Armenian cultural and religious life.¹²⁴

• 125 MO, p. 200v.
• 126 MT.

53A count of the number of victims in Şirvan is relatively easy because of the limited number of Christian settlements and the rich detail of Monahan’s report. His own total figure of 179 dead (151 men and 18 women)125 corresponds with his count per village. Monahan’s findings are at odds with the descriptions of the only available Armenian source,126 which alludes to massacres costing the lives of half of the population of some villages. I am inclined to believe that Monahan’s figures are closer to the truth, since he interviewed survivors on the spot and aimed to collect evidence. The Armenian source was written shortly after the incidents, when general panic reigned and the numbers of victims were naturally rough estimations. Still, Monahan’s figures could be marginally low.

• 127 See my analysis of the number of victims of the urban massacres (J. Verheij, 1999, p. 126).

55The numbers of fatalities in both Şirvan and Hizan were lower than one would expect on the basis of the initial reports. This is in line with a general pattern in 1894-1897, with initial figures elsewhere being too high and later replaced by lower figures.127

• 128 Hizan: MTAgh, p. 108-109, 113-114; Şirvan: MO, p. 177v, 183, 195v.
• 129 MO, p. 160v, 164v, 182-183, 199. Some villages paid for protection: MO, p. 178, 180-180v, 192v.
• 130 Hizan: CR, p. 208, 209, 210, 211-212; Şirvan: MO, passim.
• 131 For the Kurdish tribes that took part in the attack on the Armenians of Sasun, see BNA FO 881-6583 (...)

57To begin with, no references exist of Ottoman police or army involvement in Şirvan or Hizan, which is perhaps simply a reflection of their absence there. In Hizan, quite a number of local actors operated, including relatives and affiliates of the sheikhs and local aghas, many of them named by the Catholicos but difficult to identify. In Şirvan also, various sorts of local people were involved, but to a lesser extent than in Hizan.¹²⁸ The sedentary Kurds seem to have largely abstained from participation or protected their neighbors there.¹²⁹ In both districts the majority of attacks were carried out by members of Kurdish semi-nomadic or nomadic tribes. Seven tribes are mentioned in the sources (in their original spelling): In Hizan the Aliganli and also the “Saidanli,” in Şirvan the Atmangli, Strugan, Demli and Modeki, together with the Mahmedan.¹³⁰ None of these were among the tribes that attacked the Armenians of Sasun in 1894.¹³¹

• 132 The Atmangli stayed in Baykan in the summer (northwest of Şirvan) (F.R. Maunsell, 1904, vol. 4, p.  (...)
• 133 BNA FO 881-6775 91/1 (report of Armenian Patriarchate); MO, passim.
• 134 CR, p. 208 says that the Alikanli were enrolled in the Hamidiye in 1895; this could not be corrobor (...)
• 135 Ibid. A document from 1873 shows that the Ottomans already aiming to settle the Alikanli (BOA A:}KT (...)
• 136 The Ottoman authorities held the Dimili responsible for the killing of 14 Armenians in the villages (...)
• 137 H. Rassam, 1897, p. 87, 102, 106.
• 138 F.R. Maunsell, 1894, p. 82.
• 139 BNA FO 881-6958 352, Currie to Salisbury, 25 May 1897, 386/1, Crow to Currie, 25 May 1897.

58With the exception of the Modeki (more of a collective name for all tribal Kurds from Motki than a tribe), all these tribes were relatively small and seem to have had their summer pastures in or close by Hizan and Şirvan.¹³² Almost all already had a reputation for harassment and exactions.¹³³ None served in Sultan Abdülhamid II’s Hamidiye regiments or otherwise had close relations with the government.¹³⁴ On the contrary, most of the tribes listed here had a strained relationship with the Ottoman administration. The Aliganli and Saidanli had previously been forcibly removed from the area upon complaints from the local population;¹³⁵ the Demli also found themselves the subject of government measures in 1895, because of raids in southern Bitlis, close to Hizan.¹³⁶ The Motki Kurds had had a reputation for lawlessness for decades.¹³⁷ Motki was one of the tribal mountain areas that were famous for their lack of government control – like neighboring Sasun, Dersim and parts of Hakkari – places where administration was nominal and ineffective, and no Hamidiye was recruited.¹³⁸ In fact, just two years after the massacres, in 1897, a serious confrontation between the government and the Motki Kurds erupted, in which the district governor, the gendarmerie commander of Bitlis and several other Ottoman officers were killed.¹³⁹ As a preliminary conclusion, we may say that in Hizan and Şirvan, 1895, a multitude of perpetrators were active, but that (semi-)nomadic tribes, many of whom were even in normal times not or hardly under government control, were the most conspicuous perpetrators.

• 140 Some evidence from primary sources: BNA FO 881-6820, 370/1 (Longworth to Currie, 21 Oct. 1895), 690 (...)
• 141 MO, p. 180v. Vice-Consul Monahan himself evidently thought the story was true (MO, p. 200).
• 142 MO, p. 180-180v.

59Among Ottoman Muslims and Christians alike there was a widespread notion that the violence against the Armenians in 1895 was desired, and even organized and ordered by the Ottoman government.¹⁴⁰ Şirvan offers an interesting illustration of this perception. Monahan reported three years afterwards that “the year of the massacre is universally spoken of by Mussulmans and Christians in Sherwan as the year of the Firman.” This firman, or order, was said to have been given by the acting kaymakam of the district, one Fatha Bey, in October 1895. The most extraordinary element in this story, actually rendering it quite improbable, is the idea that a relatively unimportant official, like a temporary (acting) kaymakam, could issue such an important order.¹⁴¹ Moreover, Monahan describes how a Kurd hired for protection by the Assyrian inhabitants of the village of Simhor (Sarıdana), went to the office of the sub-province governor (mütessarıf) in Siirt, thus the superior of the kaymakam, to obtain a document that annulled the firman. Thus, so the story goes, the village was not attacked.¹⁴² The idea of an Ottoman ordered extermination was so strong that even during the Motki revolt of two years later Kurds believed it was going to be “their turn:”

• 144 BNA FO 881-6820 738/1, Fontana to Currie, 3 Dec. 1895; idem 854/1, Longworth to Currie, 17 Dec. 189 (...)
• 145 In the Fatha Bey case, I searched the Ottoman archive catalogue in vain. For a refutation by a high (...)
• 146 A local Muslim notable in Muş asserted that “after the disorders in Constantinople [those on Septem (...)

60The (hypothetical?) firman of Fatha Bey is one of a rather limited number of cases in 1895 in which (written) orders are said to have existed for massacres.¹⁴⁴ No Ottoman sources to corroborate such claims have surfaced (including for the supposed Fatha Bey firman).¹⁴⁵ Possibly the perception of an order for massacre was more of a popular belief than something based on reality. Under Sultan Abdülhamid II’s autocracy, it was commonplace to assume that nothing could happen without an order of the Sovereign.¹⁴⁶ Of course, the fact that the Ottomans did not do much to counter the violence (in Hizan and Şirvan they could not even do so had they wished), and that few perpetrators were punished, only reinforced the belief that the violence was intended.

• 147 Most of the other incidents known in the Armenian/Western literature as anti-Armenian violence were (...)

62Reports on other areas contain abundant information on the motivation of the Muslims. Both the Ottoman government and Muslims saw the Sasun massacre of 1894 as an Armenian revolt. Particularly after the Huntchak demonstration of September 30, 1895, many Muslims believed that Armenians in general were ready to rise against Ottoman rule. This belief was further reinforced by the Zeytun revolt. The content of the administrative reforms proposed by the Powers following the Sasun massacre was unknown or misinterpreted and widely seen as a project for the establishment of Armenian rule.¹⁴⁷ Religious principles were invoked that allowed for massacre of the (male) infidels, plundering of their goods and enslavement of their women. In fact, it seems, many perpetrator groups were moved much more by a desire to plunder than by political motives, though the two are difficult to distinguish. This was particularly true for the Kurdish tribes, who even under ordinary circumstances plundered when they could. For them, the general anarchy in 1895 and the lack of government response was an occasion not to be missed. It has often been observed that they mostly only killed, when people resisted. Sadly, the information on Hizan and Şirvan is too scanty to allow for much interpretation. While the high incidence of conversion to Islam points to relatively strong political motivation, the situation also bears the hallmarks of chaos-induced tribal action to plunder and enrich.

• 148 Hizan: BNA FO 881-6820 881, Currie to Salisbury, 31 Dec. 1895; CR, p. 207; MTAgh, p. 106, 110-112, (...)
• 149 Hizan: CR, p. 208, 209.

63The events in Şirvan and Hizan attracted wide attention for their large scale of conversions to Islam. Indeed, all consulted sources confirm that almost all Armenians and Assyrians, some several thousand people, converted.148 The main reason for this was the fear of further violence and a hope for survival. Often it seems, conversion was an option not offered by the attackers but by Muslim neighbors, who either thought that the conversion of the Armenians was necessary for religious reasons or saw this as a practical method to avert further attacks by the tribes. In Hizan, the sheikhs were prime movers.149 Since they had previously been instrumental in cases of conversion, they seem to have had a vision of Islamization for their territory. The sheikhs look like the most important single factor in the mass conversions.

• 150 MO, p. 161v-162, 178.
• 151 CR, p. 212 mentions one village in Spagert/İspayirt, 3 families in Mamerdan (CR, p. 210) and 15 fam (...)
• 152 MO, p. 171v, 198, 200v.
• 153 MO, p. 171v-172.
• 154 MP, p. 183. Monahan also noticed a surprising low number of kidnappings of women in Şirvan, in cont (...)

64One of the most interesting finds of this research is that the large majority of the converts, away from public attention, reconverted within a few years. For some, this reconversion could only take place away from their oppressors, having fled their homes.150 In Hizan, only a small majority of the Armenians remained Muslim.151 In Şirvan, three converted small villages were still Muslim when Monahan visited the area.152 He left an interesting description of one of them. The church was used as a straw store. While the converted Armenians presented themselves as Muslims to the outside world, they continued to practice their religion in their houses and “use their old Christian names when there [were] no dangerous Mussulmans around.” To the apparent astonishment of the British visitor, “They talked as though they regarded the situation almost as a joke.” They had one wedding after 1895, secretly performed by an Armenian priest in a neighboring village.153 Also, they were not circumcised. Circumcision, reportedly more widespread during forced conversion in other regions, was rare in Şirvan.154

• 155 FO 881-6958 370/1, Memorandum by Mr. Monahan; MO, p. 178 mentions a case of reconversion of a villa (...)
• 156 FO 881-6958 370/1, Memorandum by Mr. Monahan, 2 Jun. 1897; CR p. 212; BOA DH.TMIK.H 30-24, 28 March (...)
• 157 FO 881-6820 684, Currie to Salisbury, 29 Nov. 1895.
• 158 FO 881-6958 370/1; S. Deringil, 2009.
• 159 MO, p. 172 relates how a descendant of the beys of Şirvan in Kormas did “not approve of the convers (...)

65The reconversion of so many people was, in principle, tricky, since in Muslim doctrine apostasy is punishable by death. Church authorities were of course against the conversions and, within their limited possibilities, struggled to regain their flock.155 An important factor was the position of Ottoman authorities, which did not accept the conversions.156 Diplomatic reports show that the Sultan himself was strongly opposed to conversion of non-Muslims in general, particularly if just for practical reasons.157 The same mentality existed everywhere in the bureaucracy. The anti-conversion stance that the Ottoman authorities had regarding Hizan and Şirvan fits in with the pattern identified by Selim Deringil in his research on the mass conversions of 1895. Interestingly, one of the arguments mostly used by the officials was that mass conversions would offer the foreign powers reason to criticize the Empire.158 Fear for loss of possible revenue (in particular the exemption tax for military service Christians made), seems to have played a role as well. Some local Muslims seem to have taken a similar position.159

• 160 Author’s personal encounters in Hizan and Şirvan between 2005 and 2015. In the most exhaustive list (...)

66Still, it is quite possible that some families who converted in 1895 never reconverted and stayed Muslim. And it is likely that they continued with their Christian beliefs and practices at home, thus falling in the category of “crypto Christians,” still extant in various regions of Turkey. In both Hizan and Şirvan, villagers may still occasionally be encountered who are identified as “Christians” by the local population, even though they are officially Muslims.160 Given the extreme restrictions on conversion to Islam during WWI, these people are more likely to be the descendants of converts from 1895, who thus survived 1915.

• 161 MO, passim.
• 162 MO, p. 199v (on Şirvan); CR, p. 209 (on Hizan).
• 163 CR, p. 207.

68Forced to lend money from rich Muslims, usually sheikhs and aghas, sometimes the very same people who had attacked them in 1895, the poor villagers accumulated debts that they could never pay back and started to lose their lands to their creditors.¹⁶¹ Security measures introduced by the authorities, such as a ban on seasonal migration (for Christians and Muslims alike),¹⁶² only worsened the situation. Moreover, all the oppression that was already apparent before 1895 seriously worsened. Those who reconverted were particularly at risk. It was reported from Hizan that “the more fanatical” Kurds despised those who reconverted.¹⁶³ Fear of violent reactions from neighbors kept many from reconversion.

• 164 My count on the basis of Monahan’s data.
• 165 Hizan in general: CR, p. 207; specific areas: CR, p. 210 (100 families that left the Mamedank distr (...)
• 166 CR, p. 210.

69The combined result of hunger, economic hardship and oppression was land flight. In and after 1895, high numbers of Christians in Hizan and Şirvan fled their lands. From 22 villages in Şirvan more than 50% of the population left.¹⁶⁴ A systematic count for Hizan is lacking, but contemporaneous sources confirm that a large emigration from here also took place.¹⁶⁵ That some 100 people from Hizan were said to have perished in the conflict in Van in 1896, gives some indication of the number of refugees.¹⁶⁶

• 168 FO 881-6820 186, Herbert to Salisbury, 28 Oct. 1895.

74There are few clues as to the motivation for such exceptional action. Probably it was a mix of political arguments and the fairly customary drive for plunder. Was there a movement among the Armenians that was seen as a provocation? Evidence from Şirvan and Hizan suggest no or minimal Armenian protest or revolutionary action, although in Hizan there was apparently some support for the Armenian revolutionary movement. Yet, perhaps, it was not so important whether there was real Armenian action or not, since the prevalent idea among Muslims after Sasun and the Huntchak demonstration in Constantinople, was that Armenians were in revolt. This perception among Muslims in Bitlis and its surroundings was probably reinforced by the government’s presentation of the massacre in Bitlis as prompted by an Armenian attack on a mosque.¹⁶⁸ As elsewhere, events in provincial capitals were an important impetus for subsequent violence in the surrounding countryside as news spread, and we may suspect that such a ripple effect occurred also in this case. For the unfortunate Christian peasants, the violence that suddenly engulfed Hizan and Şirvan was largely the import of an enmity that was rooted elsewhere and in which they had no stake whatsoever.