The south-eastern outposts of the Armenian Dispersion were left to the last, although their immediate neighbours in the Cilician highlands had been taken at the very beginning. The villagers of Djibal Mousa were not summoned till the 13th July; Aintab was not touched till the 1st August, and then only cleared gradually during the course of the month. The summons to Ourfa, which was answered, as at Djibal Mousa, by defiance, was not delivered till the last week in September.
Glancing back over this survey, we can discern the Central Government’s general plan. The months of April and May were assigned to the clearance of Cilicia; June and July were reserved for the east; the western centres along the Railway were given their turn in August and September; and at the same time the process was extended, for completeness’ sake, to the outlying Armenian communities in the extreme south-east. It was a deliberate, systematic attempt to eradicate the Armenian population throughout the Ottoman Empire, and it has certainly met with a very large measure of success; but it is not easy to present the results, even approximately, in a statistical form. The only people in a position to keep an accurate account of the numbers affected were the Ottoman authorities themselves; but it is unlikely that they have done so, and still more unlikely that they would ever divulge such figures to the civilised world. We are compelled to base our estimates on the statements of private persons, who were excluded from detailed investigation by the jealous suspicion of the Government officials and were seldom able to observe events in more than a limited section of the field. We must make our computations by piecing together these isolated data from private sources, and since Oriental arithmetic is notoriously inexact (and this is scarcely less true of the Nearer than of the Further East), we shall only make use of testimony from foreign witnesses of neutral nationality. Such witnesses may be assumed to be comparatively free from unconscious exaggeration and completely innocent of purposeful misrepresentation, and we can accept their statements with considerable assurance.
The first step is to establish the number of Armenians living within the Ottoman frontiers at the moment the deportations began. All the other figures ultimately depend upon this, but it is harder than any to obtain, for there are no independent foreign estimates of this on record, and the discrepancy between the native estimates is extreme[[256]]. The Armenian Patriarchate, after an enquiry conducted in 1912, placed the number as high as 2,100,000[[257]]; the Ottoman Government, in its latest official returns, puts it at 1,100,000 and no more. Both parties have an equal political interest in forcing their figures, but the Armenians are likely to have had a greater respect for exactitude, or at any rate a stronger sense of the futility of falsification. The most “neutral” course under the circumstances is to halve the difference, and to take the number provisionally as being 1,600,000, with the qualification that the true figure certainly lies between this and 2,000,000, and probably approaches more closely to the latter. The rest of the necessary figures can fortunately be drawn from foreign neutral testimony, in which such baffling discrepancies are rarer.
The second step is to estimate the number of those who have escaped deportation. There are the refugees who have escaped it by crossing the frontier—182,000 into the Russian Caucasus and 4,200 into Egypt, according to detailed and trustworthy returns[[258]]. There are also two important Armenian communities in Turkey where practically all but the leaders have been left unmolested—those of Smyrna and Constantinople. At Constantinople about 150,000 Armenians must still remain. Then there are the Catholic and Protestant Millets, which were nominally exempted from deportation, and the exempted converts to Islam. It is impossible to estimate the numbers in these categories with any plausibility, for the conduct of the authorities in respect of them was quite erratic. Many of the converts to Islam[[259]], as well as Armenians of the other denominations, were given the same treatment as the Gregorians, and the actual percentage of conversions is unascertainable, for they were encouraged in some places and discouraged in others. We must also allow for those who managed to elude the Government’s net. As a general rule, this category is more numerous in reality than it appears to be, and this is especially so in the Near East. But in the present case the Young Turks seem to have put a Prussian thoroughness into the execution of their scheme, and the margin of ineffectiveness was evidently narrow. In the towns, such as Zeitoun, Hadjin, Sivas, X., and Erzeroum, where we have sufficient testimony to cross-check the estimates presented, the clearance, by deportation or massacre, seems to have been practically complete. At Erzeroum, for instance, there were 20,000 Armenians before the clearance began, and when it was over there were not more than 100 left[[260]]. Concealment on any considerable scale can only have been practised in the villages, yet the number of those who have emerged from hiding since the Russian occupation is extraordinarily small. According to the investigations of the Patriarchate, there were 580,000 Armenians in 1912 in the Vilayets of Erzeroum, Bitlis and Van, which are now within the Russian lines[[261]]. The American Relief Committee has recently been informed by its agents on the spot that there are now only 12,100 left alive there[[262]]. Whatever arbitrary margin of reduction the absence of confirmatory statistics may make it necessary to subtract from the former figure, the proportion borne to it by the 12,100 survivors remains infinitesimal. Putting the communities at Constantinople and Smyrna and the refugees together at about 350,000, we shall certainly not be reckoning too low if we allow a quarter of a million for the Protestants, Catholics, converts and others who were spared, and estimate the total number of Armenians in Turkeywho escaped deportation at not more than 600,000.
This leaves at least 1,000,000 to be accounted for by deportation and massacre, and probably 1,200.000 or more.
The third step is to estimate what proportion of these million Armenians has perished and what proportion survived, and here again our material is scanty and generalisation unsafe, the procedure of the authorities being erratic in this respect also. In certain vilayets, like Van and Bitlis, there was no deportation at all, but massacre outright; in others, like Erzeroum and Trebizond, and again at Angora, deportation and massacre were equivalent, the convoys being butchered systematically at an early stage on the road. In Cilicia, on the other hand, the men as well as the women seem to have been genuinely deported, and the convoys seem only to have been reduced by sickness and exhaustion. Yet even where there was no wholesale massacre on the journey, a convoy might practically be exterminated by degrees. A large combined convoy, for instance, of exiles from Mamouret-ul-Aziz and Sivas, set out from Malatia 18,000 strong and numbered 301 at Viran Shehr, 150 at Aleppo[[263]]. In this case, however, the wastage appears to have been exceptional. We have one similar instance of a convoy from Harpout which was reduced on the way to Aleppo from 5,000 to 213, a loss of 96 per cent.[[264]]; but in general the wastage seems to fluctuate, with a wide oscillation, on either side of 50 per cent.; 600 out of 2,500 (24 per cent.) reached Aleppo from a village in the Harpout district[[265]]; 60 per cent. arrived there out of the first convoy from the village of E. (near H.), and 46 per cent. out of the second; 25 per cent. arrived out of a convoy from the village of D. in the same neighbourhood[[266]]. We shall certainly be well within the mark if we estimate that at least half those condemned to massacre or deportation have actually perished.
We can check this estimate to some extent by the record of arrivals at certain important centres of traffic on the exile routes, or at the final destinations of the convoys. On the 16th August, 1915, for instance, an exceedingly competent neutral resident at Constantinople stated that, to his knowledge, there were then 50,000 exiles scattered along the route from Bozanti (the first break in the Baghdad line) to Aleppo; on the 5th November, another witness[[267]], who had just traversed this route, wrote back from Aleppo that he had passed 150,000 exiles between there and Konia. Again, 13,155 exiles had reached or passed through Aleppo by the 30th July, 1915, and 20,000 more arrived there between that date and the 19th August[[268]]. By the 3rd August 15,000 of these had been transmitted alive to Der-el-Zor, and this was only the beginning of the arrivals in the Zor district. No exiles reached Damascus before the 12th August, but between that date and the 3rd October, 1915, 22,000 of them had come through[[269]]. These are isolated data, and prove little in themselves, but in its Bulletin of the 5th April, 1916, the American Relief Committee has published a cable recently received in the United States from a competent source, in which the total number of Armenian exiles alive at that time in the regions of Der-el-Zor, Damascus and Aleppo is estimated roughly at 500,000[[270]]. This figure is possibly an exaggeration, but it is not incompatible with our two previous conclusions, that the total number of Armenians affected by the Young Turks’ scheme was at least a million, and that at least 50 per cent. of these have perished. To the alleged 500,000 survivors in the three regions mentioned we must add an uncertain but inconsiderable margin for the exiles who may have been planted at Mosul or who may still, in March, 1916, have been held up on the road; and this will raise the original number affected to something approaching 1,200,000, which we considered, on other grounds, to be nearer the real figure than the bare million which we accepted.
We can sum up this statistical enquiry by saying that, as far as our defective information carries us, about an equal number of Armenians in Turkey seem to have escaped, to have perished, and to have survived deportation in 1915; and we shall not be far wrong if, in round numbers, we estimate each of these categories at 600,000.
The exact quantitative scale of the crime thus remains uncertain[[271]], but there is no uncertainty as to the responsibility for its perpetration. This immense infliction of suffering and destruction of life was not the work of religious fanaticism. Fanaticism played no more part here than it has played in the fighting at Gallipoli or Kut, and the “Holy War” which the Young Turks caused to be proclaimed in October, 1914, was merely a political move to embarrass the Moslem subjects of the Entente Powers. There was no fanaticism, for instance, in the conduct of the Kurds and chettis, who committed some of the most horrible acts of all, nor can the responsibility be fixed upon them. They were simply marauders and criminals who did after their kind, and the Government, which not only condoned, but instigated, their actions, must bear the guilt. The peasantry, again (own brothers though they were to the Ottoman soldiery whose apparent humanity at Gallipoli and Kut has won their opponents’ respect), behaved with astonishing brutality to the Armenians who were delivered into their hands; yet the responsibility does not lie with the Turkish peasantry. They are sluggish, docile people, unready to take violent action on their own initiative, but capable of perpetrating any enormity on the suggestion of those they are accustomed to obey. The peasantry would never have attacked the Armenians if their superiors had not given them the word. Nor are the Moslem townspeople primarily to blame; their record is not invariably black, and the evidence in this volume throws here and there a favourable light upon their character. Where Moslem and Christian lived together in the same town or village, led the same life, pursued the same vocation, there seems often to have been a strong human bond between them. The respectable Moslem townspeople seldom desired the extermination of their Armenian neighbours, sometimes openly deplored it, and in several instances even set themselves to hinder it from taking effect. We have evidence of this from various places—Adana[[272]], for instance, and AF.[[273]] in Cilicia, the villages of AJ. and AK.[[273]] in the AF. district, and the city of Angora. The authorities had indeed to decree severe penalties against any Moslem as well as any alien or Greek who might be convicted of sheltering their Armenian victims. The rabble naturally looted Armenian property when the police connived, as the rabble in European towns might do; the respectable majority of the Moslem townspeople can be accused of apathy at worst; the responsibility cannot rest with these.
The guilt must, therefore, fall upon the officials of the Ottoman Government, but it will not weigh equally upon all members of the official hierarchy. The behaviour of the gendarmerie, for example, was utterly atrocious; the subordinates were demoralised by the power for evil that was placed in their hands; they were egged on by their chiefs, who gave vent to a malevolence against the Armenians which they must have been harbouring for years; a very large proportion of the total misery inflicted was the gendarmerie’s work; and yet the gendarmerie were not, or ought not to have been, independent agents. The responsibility for their misconduct must be referred to the local civil administrators, or to the Central Government, or to both.