Historian A. Roger Ekirch contends that in ancient times, people naturally had two sleeping periods at night- the first starts shortly after sundown and ends in an short break, where people were semi-conscious and would speak to each other about their dreams, or pray, or talk, or have sex. Wikipedia summary here. The second sleep would then end at dawn.
Here's a good NYT article by him.
The issue is, it seems like this is a pretty major claim- that the ancients had multiple sleep periods, yet there doesn't seem to be a lot of coverage of this theory, nor any academic critiques as far as I can find. Is this a legitimate theory, or does it have any possible problems?