Larks comments on Efficient prestige hypothesis