Verified by 
There are three schools of thought or conjectures regarding the role, if any, that dreams might play in memory consolidation: 1) they play no role; 2) they reflect more fundamental memory consolidation processes and 3) dreams are required for memory consolidation.
In favor of conjecture 1 is the fact that people who do not dream or at least who recall few dreams seem not to be particularly impaired in terms of memory consolidation—though it has to be admitted this issue has not been studied extensively. Animals probably do not experience the same kinds of narrativized phenomenal experiences we experience during REM sleep though they very likely do experience dream images. They seem to function very well without narrativized dreams.
When we look directly at dream content in humans, episodic elements that make up episodic memories are largely absent in dreams (Fosse et al., 2003). But we should not expect to find fully formed memory episodes in dreams if dreams are causally required for encoding of informational elements into long term memory stores. Instead we should find exactly what we do find in dreams: fragmented images, thoughts, units or features of multi-modal episodes such as characters, background settings, colors, and especially emotions. Emotions are crucial for consolidation of memories as they tell the memory system what informational elements should or should not be integrated into memory.
Most dream scientists appear to favor conjecture 2—the idea that dreams merely reflect more fundamental memory processing. In favor of conjecture 2 there is the dream lag effect (Nielsen et al., 2004) wherein images about experiences that took place around a week (5–7 days) prior to the dream re-appear in the dream. Presumably it is these images from experiences in the past week that are selected by the system to be stored in long-term memory. (It would be interesting to study whether the images that re-appear in dreams are accessible to recall some months later.)
I myself believe that dreams are causally implicated in memory consolidation—at least for emotionally significant experiences. In short, that conjecture 3 is correct. The success of image rehearsal therapy for frequent nightmare sufferers suggests that alteration of dream imagery and dream scripts is necessary if emotionally charged images are to be integrated successfully into long-term memory. The dream/nightmare images themselves matter. So while conjectures 1 (dreams play no role in memory) and 2 (dreams merely reflect memory processes) cannot be ruled out, a restrictive version of number 3 (dreams are required) is becoming more plausible.
Several authors (Llewellyn 2013; Levin and Nielsen, 2007; Horton and Malinowski; 2015) have recently presented models of how dream images could facilitate memory consolidation processes. I have discussed the Llewellyn and the Levin and Nielsen models previously on this blog so wish to mention here briefly the Horton and Malinowski model, as they attempt to take into account the fact that dreams are very obviously concerned with autobiographical experiences and memories. If we want to discover if and how dreams participate in the memory consolidation process the most obvious place to look is at autobiographical memory.
The basic mechanism that Horton and Malinowski propose (as far as I can see) is that complex multi-modal elements (complete autobiographical experiences) are decomposed during the dream state into decontextualized elements/images that are then recombined in novel ways to support consolidation and creativity. They write that “The rebinding of memory sources occurs when the brain enters a hyperassociative, reflexive state during both REM and non-REM sleep.” The hyperassociative state is thought to be due to the relaxation of frontal executive control processes during sleep (especially REM). Horton and Malinowski point out that decontextualization process leads to the ability of all these fragmented / disconnected images/elements can be more easily consolidated into long term memory because they can now relate or semantically connect up with any number of other elements (hyperassociativity). All this makes sense to me and is consistent with Llewellyn's associative model and Nielsen and Levin’s model of the integration of emotional and traumatic memories into long term memory. Repetitive nightmares can be conceived as a breakdown in the decontextualization process. The multimodal original autobiographical experience is not effectively decomposed so it sits in working memory in the dream state continually attempting integration into long term memory but failing. But it loses none of its emotional or traumatic force so the individual suffers its experience over and over again.
These recent models of the role of dreams in consolidation of emotional and autobiographical memories need empirical testing. if we find that dreams are in fact causally necessary for memory consolidation then we will have at least one empirically confirmed function for dreaming: it facilitates memory consolidation. The decontextualization process in particular needs intensive investigation as it is central to the memory consolidation process and may be crucial for selection of what does and does not get into long term memory.
Horton C. L. and Malinowski, J.; (2015) Autobiographical memory and hyperassociativity in the dreaming brain: implications for memory consolidation in sleep. Front Psychol. 2015; 6: 874. doi: 10.3389/fpsyg.2015.00874 PMCID: PMC4488598
Llewellyn S. (2013). Such stuff as dreams are made on? Elaborative encoding, the ancient art of memory, and the hippocampus. Behav. Brain Sci. 36, 589–659. 10.1017/S0140525X12003135 [PubMed] [Cross Ref]
Levin R., Nielsen T. (2007). Disturbed dreaming, posttraumatic stress disorder, and affect distress: a review and neurocognitive model. Psychol. Bull. 133, 482–528. 10.1037/0033-2909.133.3.482 [PubMed]
Nielsen T. A., Kuiken D., Alain G., Stenstrom P., Powell R. A. (2004). Immediate and delayed incorporations of events into dreams: further replication and implications for dream function. J. Sleep Res. 13, 327–336. 10.1111/j.1365-2869.2004.00421.x [PubMed] [Cross Ref]
Fosse M. J., Fosse R., Hobson J. A., Stickgold R. J. (2003). Dreaming and episodic memory: a functional dissociation? J. Cogn. Neurosci. 15, 1–9. [PubMed]
