Alba Regia. Annales Musei Stephani Regis. – Alba Regia. Az István Király Múzeum Évkönyve. 10. 1969 – Szent István Király Múzeum közleményei: C sorozat (1969)
Tanulmányok – Abhandlungen - Éry Kinga, K.: Investigations on the Demographic Source Value of Tomb Stones Originating from the Roman Period. – Vizsgálatok a római kori sírfeliratok demográfiai forrásértékéről. X, 1969. p. 51–67.
no doubt, as I have already referred to above, that according to S z i 1 á g y i' s data there are many inscriptions in Carthago relating to slaves or liberated slaves, but the total is no more than 30 per cent of the entire inscription material published by him from the city. Even this possiblity is precluded in the case of Caesarea, because the inscriptions of merely a few slaves or liberated ones are known, while the average age at death in this town is lower even than in Carthago. Since, in both cases, other than North African customs regarding commemoration by inscriptions prevailed in these two cities, it seems highly probable that their deviating distributional pattern results from their different ethnic composition. As was already mentioned, some of the authors explain the extremely low average ages at death in Rome by the highly unsanitary conditions prevailing in the capital. In accordance with this assumption, the unfavourable conditions should then theoretically have afflicted the entire urban population. The inscription material apparently substantiates this conjecture since, for instance, the average age at death of the slaves in Rome had been no more than 18 years against the 32 of those in Carthago. Or, for example, the average age at death of the merchants and artisans in Rome had been merely 31 years while, in other parts of the Empire, the same stratum attained an average 39 years of age 2,i . If, however, the data are analysed according to linguistic or ethnic differences, a very different picture is obtained. Among the 9980 tomb inscriptions published from Rome, there are 822 written in Greek 24 . If the age group distributions of those having spoken Latin or Greek are separately plotted onto graphs, we receive a striking result of completely constrasting distributions (Fig. 6). The Latin inscriptions show a characteristically Italian distributional pattern, while the Greek ones correspond to the African pattern. Accordingly, the average age at death calculated from the Latin inscriptions is merely 21 years, whereas that of the Greeks is 30 years more, that is, 51 years! Obviously therefore the two different linguistic groups inhabiting Fig. 6: Age group frequencies of the Greek and Latin tomb inscriptions in Rome the city also followed different customs in commemoration. That the differing age group patterns cover divers ethnic substrates can be illustrated by another example. Szilag y i publishes the average age at death values of persons deriving from divers geographical areas but deceased in foreign localities 25 . Among these, it is again the Africans who show the highest average age at death and the Italians the lowest ones, proving that these persons had retained, even abroad, the commemorating custom prevailing in their place of origin. With regard to the ethnic differences presumably hidden behind the distributional patterns, the separate studying of the patterns deriving from cities and settlements, shown on Figures 3,4,5, may be most instructive. The problem, for instance, of the distribution in the Italian Tarquinii, differing characteristically from the majority of the other peninsular settlements and rather approaching the African pattern. Or the question whether the appearance of Italian distribution or, let us say, the distributional pattern of the capital in non-Italian cities might imply the appearance of a population of Italian origin, that is, whether the social effects of Romanization had been greater in such places than in other localities. The solution of these problems are, however, the task of historians of antiquity. As a next step in our investigations, let us examine sex differences per age group as revealed by the inscriptions in the combined material of the provinces. Figure 7 displays the data plotted on the graph by the usual method. According to Fig. 7, the sex distribution of the deceased persons shows slighter to greater age group differences per province. For a correct interpretation of the apparent differences in rate and trend, the model life-tables of the UNO again submit ample information 26 . From the data of the model life-tables giving life expectancies of 25, 30, 35 years, I calculated which sex is represented by the greater frequency among the dead of the different age groups, as well as the per cent order of magnitude of the sex differences per age group. These data are contained in Table 4. According to the model distributions, the mortality frequencies of sex patterns per age group are as follows: 1.5—2.5 per cent more boys than girls die in the age group 0—9 years, 4—6 per cent more women between the years 10—39, 2—5 per cent more men between 40—69 years, and it is only in the age group 50—59 years that their ratio increases above 8 per cent. Finally, above 70 years of age there are again more dead women, their ratio above 80 years exceeding by 11 — 15 per cent that of the men. For a similar evaluability of the material from the provinces, I have compiled the per cent differences of the sex patterns per age group in Table 5. Since the average age at birth was surely about 25—35 years in the Empire, the values of Table 4 can be aligned as a natural pattern with those obtained from the provinces. In Table 5, the cases where the higher frequency per sex corresponds to that of the model are given in cursive, and in bracket those where the rate of difference, beyond this, more or less agrees with the model distribution. 23 J. SZILÁGYI, A. Arch. Hung., 14, 1962, p. 347, table B. 2i J. SZILÁGYI, AntTan, 6, 1959, pp. 45-47, 25 J. SZILÁGYI, A. Arch. Hung., 15,1963, p. 181, table C. 26 cf. Note 10. 60
