Nevertheless, after all is said, it seems hard to believe that ceremonial capture does not sometimes have a more real significance. Often it may symbolize the coyness or mark the subjection of woman. More frequently it may stand as a mere general reminiscence of the good old times of force and lawlessness. Still it would be strange, indeed, if it did not also appear as a direct survival of actual wife-capture.[551] Granting this, however, the significance of capture _de facto_ remains the same. We perceive more clearly that it has very widely prevailed; yet it must still be regarded as a mere incident of war and pillage. It has nothing whatever to do with the institution of marriage. It could never on any wide scale have been the normal manner of procuring wives. To assume that wife-stealing has been a universal phase in the evolution of marriage is not one whit more reasonable than to hold that robbery has been a normal stage in the evolution of property.[552] In spite of Hobbes or McLennan, it remains to be proved that a state of chronic hostility was ever a general phase in the history of mankind. Such a state is inconsistent with the prevalence of the blood-feud.[553] Even the rule of exogamy among primitive peoples does not harmonize with general wife capture. For the coexistence of clan-exogamy and tribal endogamy means, under normal conditions, a tendency toward peace within the tribe.[554] There is strong reason to believe that in every period of social development consent and contract, in some form, have been the cardinal elements of marriage. Captured or stolen women have usually become slaves or concubines; and, except in rare instances, the relatively small number of them made wives must always have been insignificant as compared with the number of wives obtained in other ways. Thus the solution of the problem of so-called marriage by capture appears to be similar to that of polygyny. The practice of taking several wives is exceedingly common; but on close examination we discover that polygyny is relatively unimportant, and that it has never been able to displace monogamy as the normal type. So it is with the practice of capturing women for wives. However prevalent the custom, it does not seem ever to have greatly influenced the natural laws or modified the fundamental motives upon which marriage and the family rest. But the value of the evidence upon which this conclusion is based can be thoroughly appreciated only after we have traced the origin of contract in marriage. Let us begin with wife-purchase, especially in its relation to the custom of capturing women. [551] This is in effect conceded by Spencer. While rightly rejecting the theory of systematic foreign wife-capture, as a general phase in the development of marriage, he holds that the symbol of rape may sometimes result from struggles for women within the tribe, or from the resistance of the father and male relatives of the bride.