In Lifu, the word for ‘sister’ means ‘not to be touched,’ land this is a mere expression of customary law. A man ‘must not touch’ any one of the women of his generation whom the totem tabu and the rule of the exogamous ‘phratry’ (in origin, we suggest, totemic) forbid him to touch. All such women, in a particular grade, are his sisters. Many women, besides his actual sisters, stand to him in the degree thus prohibited. All bear the same name of status as a man’s actual sisters bear, but the name does not _mean_ ‘sisters’ at all, in our sense of that word: namely, daughters of the man’s real father and mother. It means tabued women of a generation. If the ‘classificatory’ terms which include our ‘fathers,’ ‘sisters,’ ‘wives,’ and the rest meant what our ‘fathers,’ ‘sisters,’ ‘wives,’ and so on mean, then the evidence from the terms, for communal or group marriage, would really be ‘extremely strong.’ But, as Messrs. Spencer and Gillen say, ‘unless all ideas of terms of relationship as counted among ourselves be abandoned, it is useless to try and (_sic_) understand the native terms.’[24] Yet the whole force of the argument for communal marriage derived from savage terms of relationship rests precisely on our not ‘abandoning’ (as we are warned to abandon) ‘all ideas of terms of relationship as counted among ourselves.’ The friends of group and communal marriage, it seems to me, keep forgetting that our ideas of sister, brother, father, mother, and so on, have nothing to do (as they tell us at certain points of their argument) with the native terms which include, indeed, but do not denote these relationships, as understood by us. An Urabunna calls a crowd of men of his father’s status by the same term as he calls his father. This need not point to an age when, by reason of promiscuity, no man knew his father. Were this so, a man of the generation prior to his father might be the actual parent of the speaker, and all men under eighty ought to be called ‘father’ by him--which they are not. The facts may merely mean that the Urabunna styles his father by the name denoting a status which his father shares with many other men; a status in seniority, ‘phratry,’ and totem. We really cannot first argue that our ideas have no relation to the terms employed by savages, and then, when we want to prove a past of communal marriage, turn round and reason as if our terms and the savage terms were practically identical. We cannot say ‘our word “son” must not be thought of when we try to understand the native term of relationship which includes sons in our sense,’ and next aver that ‘sons in our sense, are regarded as real sons of the group, not of the individual--because of a past stage of promiscuity making paternity indiscoverable.’