摘要：This commentary argues firstly that we are so far from an account of what causes dream content that we should be cautious in linking that partial knowledge to a quest for function. Secondly, that it may be that all current data on the causes of dream content allow for the null hypothesis of dream function, that is, dreams do not have a function and dreaming has not been selected for during evolution. Thirdly, that the difficulty in investigating dream function is that the experimental designs currently used are of necessity correlational rather than a random assignment to dream content conditions. And, fourthly, that the continuity / discontinuity continuum needs to be augmented by an insight dimension.
