Licata tells us that hungry Danakil go to their chief and say: “I am hungry, give me something to eat”[97].
Among the Larbas free labourers “work for one more fortunate than themselves, but not for a superior; for notwithstanding the relation of employer and employed, equality prevails”[98].
It is easy to understand that slaves are preferred to such servants. Only in one case is this preference mentioned by an ethnographer. Munzinger states that the slaves bought by the rich Beduan for household work are generally more trusted [[286]]than ordinary servants, as they are riveted to their position[99]. But we may safely suppose that in other cases also this circumstance has furthered the growth of slavery.
We have explained why pastoral tribes have no great use for slave labour. We have also mentioned some motives that may induce such tribes to keep slaves. But the fact has not yet been accounted for, that some pastoral tribes keep slaves and others do not. Whence this difference? It has been shown that slavery does not only exist among pastoral tribes that till the soil to a limited extent. Among all pastoral tribes subsistence is dependent on capital. Wealth, too, exists among all these tribes[100]; and we cannot see why slaves, as a luxury, would be wanted by one such tribe more than by another. As slaves are sometimes employed as warriors, we might be inclined to suppose that slavery exists among all warlike tribes, and among these only. But there are several pastoral tribes which, though very warlike, do not keep slaves: Kazak Kirghiz, Turkomans, Massai, and some pastoral nomads of South Africa.
That the subjection of tribes as such in stead of individual slaves, of which we have spoken in the last paragraph, cannot account for all cases in which slavery does not exist, becomes evident, if we take into consideration that most of the pastoral tribes of North-East Africa, which keep other tribes in subjection, practise slavery, whereas in Central Asia and Siberia we find neither subjected tribes nor slaves.
Therefore there must be other causes.
In chapter II we have spoken of external causes: it may be that slaves would be of great use, and yet cannot be kept, because the coercive power of the tribe is not strong enough. We have also seen that this coercive power is most strongly developed where men have fixed habitations, live in rather large groups and preserve food for the time of scarcity, and where there is a group of somewhat homogeneous tribes maintaining, constant relations with each other. Pastoral tribes are nomadic, [[287]]do not live together in very large groups, and do not want to preserve food, for they have their supply of food always at hand. Yet the fact that several pastoral tribes keep slaves proves that at least among these the coercive power is strong enough. We shall try to find a cause peculiar to these tribes, that enables them to keep slaves. Now it is remarkable that our positive cases are nearly all of them found in a few definite parts of the globe: North-East Africa, the Caucasus, and Arabia; whereas the pastoral nomads of Siberia, Central Asia, India, and South Africa, with one exception (the Ovaherero), do not keep slaves. And the parts where slavery exists are exactly those where the slave-trade has for a long time been carried on on a large scale. Accordingly, the slaves these tribes keep are often purchased from slave-traders and in several cases belong to inferior races.
The slaves of the Aeneze Bedouins are Negroes[101].
The slaves kept by the Larbas are Negroes purchased from slave-trading caravans[102].