Why the simple REM-emotion, SWS-facts split fails
The idea that REM sleep handles emotional memories while slow-wave sleep (SWS) handles neutral facts is intuitive but not well supported by direct experiments. In a daytime-nap study, researchers used targeted memory reactivation—a technique that quietly replays sounds associated with learned material during sleep—to cue emotional memories in either REM or SWS [1]. Contrary to the researchers' hypothesis, reactivating emotional memories during REM led to worse recall, not better. In the same study, reactivation during SWS enhanced memory, and the benefit was strongest in people who had spent more time in both SWS and REM [1].
A later study by the same group reached a similar conclusion: SWS and REM appear to have complementary roles in emotional memory, with REM potentially involved in forgetting emotional material rather than strengthening it [2]. Neither study tested neutral or factual memories, so they cannot confirm what SWS does for neutral content.
REM sleep may prune emotional memories rather than preserve them
The finding that REM reactivation impaired emotional memory is consistent with the idea that REM sleep actively weakens or eliminates emotional memories—unless they were recently reactivated during SWS [1]. In the nap study, the product of time spent in SWS and REM predicted how much reactivation helped: people with more of both stages showed better memory for cued emotional items and worse memory for non-cued ones [1]. That pattern suggests sleep may both preserve and discard emotional memories depending on whether they are reactivated.
An earlier study using selective REM-sleep deprivation also found that REM sleep contributes to emotional memory consolidation, but its abstract reports that emotional memory bias was greater when both SWS and REM were present compared to SWS deprivation alone [3]. That is not a clean endorsement of REM as the emotional-memory stage; it points to an interaction between the two stages.
The evidence for SWS supporting neutral facts is indirect
The claim that SWS specifically supports neutral factual memory comes mainly from studies that were not designed to test that question directly. One selective deprivation study reports that SWS supports neutral memory consolidation [3], but its abstract is fragmentary and does not describe the neutral-memory result in detail. A meta-analysis of emotional recognition memory found evidence for a selective enhancement of emotional over neutral memory after REM sleep, but it does not establish that SWS is the neutral-memory stage [4].
A neurophysiological study found that REM sleep preferentially facilitates emotional memory consolidation, but it did not examine SWS's role in neutral memory [5]. So the neutral-facts side of the split remains plausible but poorly tested by the available evidence.
What this means for the original question
The evidence supports a revised picture: SWS is not just for neutral facts—it strengthens emotional memories, especially when they are reactivated [1][2]. REM sleep, rather than simply enhancing emotional memory, may help weaken or forget emotional material that is not reactivated [1][2]. The idea that SWS handles neutral factual memory is not disproven, but the studies that speak to it are indirect and do not provide a clear, consistent answer [3][4].
Sources used in this answer
Emotional memories are enhanced when reactivated in slow-wave sleep but impaired in REM
Reactivating emotional memories during REM sleep impaired recall, while reactivation during slow-wave sleep enhanced memory, and the benefit depended on time spent in both stages.
Both slow wave and rapid eye movement sleep contribute to emotional memory consolidation
Slow-wave sleep and REM sleep appear to have complementary roles in emotional memory consolidation, with REM potentially involved in forgetting emotional memories.
The effect of selective REM-sleep deprivation on the consolidation and affective evaluation of emotional memories
Selective REM-sleep deprivation and slow-wave-sleep deprivation affected emotional and neutral memory differently, with emotional memory bias greater when both stages were present.
Sleep's impact on emotional recognition memory: A meta-analysis of whole-night, nap, and REM sleep effects
A meta-analysis found evidence for a selective enhancement of emotional over neutral memory after REM sleep.
REM sleep, prefrontal theta, and the consolidation of human emotional memory
REM sleep preferentially facilitated emotional memory consolidation, but the study did not examine slow-wave sleep's role in neutral memory.