In the present chapter on demography, we have chosen to distinguish the pre-1895 and post- 1895 periods for an obvious reason: such a distinction makes it easier to assess the direct consequences of the 1894–96 massacres, which particularly affected the population of the Armenian high plateau.
Shortly before these massacres began, in 1894, an Ottoman census arrived at figures similar to those given for 18818/2–83 (with differences well under one hundred individuals). By and large, the statistics utilized for 1894 were the same as those cited from 1881 to 189418 in both official publications and the documents transmitted to foreign governments or specialists such as Cuinet.
The demographic evolution of the Armenian high plateau is only perceptible in an official document said to have been produced in 1897, but is in fact based on results obtained in 1895, on the eve of the biggest massacres.
Interestingly, this document reveals radical changes that had supposedly come about in a single year: in the vilayet of Erzerum, the number of Muslims jumped from 445,648 to 509,948; in Van, the figure of 60,448 Armenians established by both the 1894 census and earlier censuses fell to 59,433, while the number of non-Christians soared from 59,412 to 80,773.
Nothing about the events that the populations of the area lived through in 1895 allows us to explain this 36 per cent increase in the number of Muslims over a 12-month period. Thus, it is quite obvious that no serious census had been carried out for decades. Rather, the old figures were regularly recycled; the number of Muslims was methodically inflated and the number of Christians just as methodically reduced.
We must, however, distinguish the way the Ottoman authorities handled statistics for European Turkey and western Anatolia from the methods it employed in the case of the Armenian high plateau: there were, it seems, fewer distortions in its figures for the former regions, whereas the census figures for the Armenian plateau were subject to careful manipulation.
The Ottoman census of 1906/07 is hardly more explicit. It indicates that the vilayets of Erzerum, Bitlis, Van, and Mamuret ul-Aziz had an Armenian population of 354,577 (352,035 in 1895) and a Muslim population of 1,194,778 (1,139,041 in 1895).
According to these figures, then, practically no demographic change had occurred in 11 or 12 years: this was tantamount to affirming that the massacres of 1895 had never taken place except in the Armenians’ imagination, and in the Western newspapers that claimed to have described a slaughter organized by Sultan Abdülhamid.
In fact, the Census Bureau of the Ministry of the Interior resolved the problem by reissuing the figures for previous years, which had already been reduced, on our estimate, by 200 per cent. Given the importance of the demographic weight of a given group in the resolution of a territorial conflict – consider the example of Bulgaria, Romania, and Greece in 1878 – it was simply not possible to obtain figures accurately reflecting the real situation from the states concerned.
It was, no doubt, because the Armenian Patriarchate was aware of this fact that it carried out censuses of its own. In the preceding pages, we have already discussed the patriarchal censuses for 1878/82, which, it should be added here, generally assessed only the overall Armenian population, without giving precise information as to how it broke down by kaza. Moreover, it is clear that the Armenians adopted a low profile under Abdülhamid. There could be no question of their conducting a census of any kind in the period.
Thus, it was only with the Young Turks’ assumption of power in 1908 that it became possible to engage in an undertaking of this sort. In 1912, the Patriarchate produced a preliminary assessment, hardly more precise than that of 1878/82, of the number of Armenians living in the vilayets of Van, Bitlis, Mamuret ul-Aziz, Dyarbekir, and Erzerum, arriving at a figure of 804,500.
Probably because it was dissatisfied by the imprecision of this document, the Patriarchate methodically carried out a second census, necessitated by the plan to reform the eastern provinces. According to Vahan Papazian, who was, as we have seen, executive director of the Security Commission directly responsible for the question of the Armenian reforms, the Commission took the initiative of organizing the 1913 census.
Of course, the government and the Patriarchate were in this period fighting a fierce battle around population statistics, since the future of the reform plan was inseparable from the number of Armenians living in the eastern provinces.
The main argument advanced by the Young Turk cabinet to justify its rejection of the principle of reforms in “Armenia” was that the Armenians represented a tiny “minority” there and that there was therefore little reason to make changes in the local administration, and even less to give the Armenians responsibilities in the management of local affairs. Papazian says that, while he was waiting for the results of the census then in progress, one of the best sources of information at his disposal was a census carried out under the sultan’s authorization with a view of collecting “one per cent more” from each Armenian family in order to pay off the colossal debt of the Jerusalem Patriarchate.
The records of this census, with lists of the sums collected, were preserved in the archives of the Patriarchate in Istanbul. On 20 February 1913, the offices of the Patriarchate sent a circular and the required forms to all the dioceses in the empire: the dioceses were charged with distributing these documents to the parish councils, collecting and synthesizing the results, and forwarding them to Constantinople.
The census was supposed to be completed by May. In summer 1914, forms countersigned by the primates and the members of the diocesan councils were still arriving in Constantinople. Most of the work, however, had already been done by then. Based for the most part on community structures, and especially the thousands of parish councils in the empire, the census was carried out more carefully in some regions than in others. Despite imperfections and lacunae, however, the document that resulted is of crucial importance, since no other one like it was produced in the period that interests us. It is the only source to reveal the real weight of the Armenian population, particularly that of the Armenian plateau.
Furthermore, this census was taken precisely in a period when it was assumed that the reforms in the eastern vilayets would be supervised by two European inspectors. The Patriarchate accordingly had no interest in doctoring its figures, which it knew would be promptly verified by the two officials in question.
In light of the figures we have just given, it appears very plainly that the Armenian population of the high plateau, far from growing between 1874 and 1914, sharply diminished. Indeed, without adjustment for probable population growth under normal conditions, the Armenian statistics show a drop of more than a million individuals in the 37 years between these two dates. This drop cannot be explained by the massacres of 1894–96 alone. The 300,000 “émigrés” registered in this period also have to be taken into account, together with the many villages forcibly converted to Islam.
It must further be noted that the representatives of the dioceses who conducted the censuses did not always have access to all the areas inhabited by Armenians, especially when Kurdish tribes controlled them. Consequently, not all Armenians were counted. This was particularly the case for the vilayet of Dyarbekir, from which the Turk authorities deported, to their surprise, 120,000 Armenians in spring 1915,31 although there were only 106,867 in the vilayet according to the Patriarchate’s 1914/1914 statistics and only 73,226 according to the 1914 Ottoman census.
The figures on tax revenues in 1914–16, published by the empire’s Ministry of Finance, are no less significant. They indicate that, overall, the five eastern vilayets contributed diverse fees and taxes totaling more than 110 million piasters to the Ottoman budget: 64,683,935 piasters in 1915 and nothing in 1916.33 Even if the war obstructed the collection of these taxes in the combat zones, it is hard to understand why the regions to the south, although spared by the fighting, contributed nothing to the war effort – unless we agree that the massacres and deportations of the Armenian populations of these vilayets, which was significantly underestimated in the statistics, brought on the financial ruin of these regions and their Muslim inhabitants.
Note- this chapter is from Raymond Kévorkian’s book ARMENIAN GENOCIDE: A Complete History, pp. 270-272.
