Brain symbolizing circadian rhythms over a background split between dark starry night and bright cloudy day
Généré par l'IA / Generated using AI
π Neuroscience π Health and biotech
The secrets of sleep unveiled with science

What do we really know about our dreams?

Delphine Oudiette, Inserm Research Director and Co-Director of DreamTeam at Institut du Cerveau in Paris and Claudia Picard-Deland, Postdoctoral Researcher in the Department of Psychiatry and Addictology at Université de Montréal
On March 5th, 2025 |
5 min reading time
Delphine Oudiette
Delphine Oudiette
Inserm Research Director and Co-Director of DreamTeam at Institut du Cerveau in Paris
Claudia Picard-Deland
Claudia Picard-Deland
Postdoctoral Researcher in the Department of Psychiatry and Addictology at Université de Montréal
Key takeaways
  • Research on dreams shows that they begin as soon as we fall asleep, and it is possible that different regions of the brain do not fall asleep at the same rate.
  • Among the hypotheses on their function is the idea that dreams simulate threats and regulate emotions, allowing us to better manage situations or states experienced in dreams.
  • Studies on dreams are based, for example, on reports from dreamers or on laboratory experiments, but their level of evidence is generally low.
  • To better understand dreams, it is important for researchers to collaborate and share results on open databases.
  • Ultrasound imaging and recent discoveries about windows of connection with the outside world during sleep are avenues for better understanding.

A con­scious expe­ri­ence in the midst of sleep, dreams fas­ci­nate us as much as they elude us. Research is enabling us to bet­ter under­stand this rich cog­ni­tive phe­nom­e­non. Del­phine Oudi­ette, a researcher in the DreamTeam at the Paris Brain Insti­tute, and Clau­dia Picard-Deland, a post­doc­tor­al fel­low at the Cen­tre for Advanced Stud­ies in Sleep Med­i­cine at the Uni­ver­si­ty of Mon­tre­al, shed light on the subject.

Dream nights

A dream is defined as “any men­tal expe­ri­ence that occurs dur­ing sleep” explains Del­phine Oudi­ette. It can range from sim­ple thoughts to the most dream­like imag­in­ings, depend­ing on the stage and progress of the night. As Clau­dia Picard-Deland reminds us: “Dreams can be col­lect­ed from sleep­ers at any time dur­ing the sleep cycle.”

From the moment we fall asleep, we can observe abstract micro-dreams, with the per­cep­tion of sounds, hal­lu­ci­na­to­ry move­ments or images, vary­ing from one indi­vid­ual to anoth­er. This is one of Del­phine Oudiette’s research top­ics: “One hypoth­e­sis is that the dif­fer­ent areas of the brain do not all fall asleep at the same rate.” Depend­ing on the map­ping of this desyn­chro­ni­sa­tion, we encounter dif­fer­ent types of sub­jec­tive experiences.

As the night pro­gress­es, dreams become more com­plex. “They become more per­cep­tu­al, more vivid and involve more actions,” explains Clau­dia Picard-Deland. The sleep­er goes through sev­er­al cycles, each with the fol­low­ing phas­es: a tran­si­tion from wake­ful­ness to sleep; from light sleep to deep sleep to REM sleep. Dur­ing the light and deep sleep phas­es, dreams most often take the form of thoughts that stick to real­i­ty, but they can be enriched with images. REM sleep is the per­fect set­ting for mul­ti­sen­so­ry dreams, with char­ac­ters, actions and emo­tions asso­ci­at­ed with mem­o­ries. It can hap­pen that sleep­ers become aware that they are dream­ing with­out actu­al­ly wak­ing up: this is called a lucid dream. They are then able to observe their dream or even influ­ence it. This hap­pens more fre­quent­ly in REM sleep. Final­ly, in the tran­si­tion to wak­ing, we find per­cep­tions sim­i­lar to those of falling asleep.

Unless cer­tain regions of the brain are impaired, it is like­ly that we all dream, even when we have no mem­o­ry of them. This is sug­gest­ed by the obser­va­tion of patients with REM sleep behav­iour dis­or­der (RBD), who exter­nalise their dreams. “Some of them claim nev­er to have had any dreams. Yet we see them in action in their sleep,” says Del­phine Oudiette.

What’s the point?

If dreams vis­it each and every one of us, they must have some use. So, what ben­e­fits do we derive from this dream­like expe­ri­ence? “It’s very dif­fi­cult to dif­fer­en­ti­ate the effect of the dream itself from that of sleep in gen­er­al, so there are only the­o­ries,” says Del­phine Oudiette.

One of the best known is that dreams serve as a sim­u­la­tion of threats, allow­ing us to play out sce­nar­ios in com­plete safe­ty so that we are bet­ter equipped in the future. This hypoth­e­sis has been test­ed by exam­in­ing the dreams of first-year med­ical stu­dents who have passed a stress­ful com­pet­i­tive exam­i­na­tion. The peo­ple who had the most neg­a­tive dreams about the exam were the ones who did best1.

It is also pos­si­ble that we man­age to bet­ter reg­u­late our emo­tions2 thanks to these noc­tur­nal sce­nar­ios, in which we inte­grate the some­times neg­a­tive feel­ings of the day. We would thus relive them in a more neu­tral, even pos­i­tive con­text. The con­struc­tion of dreams is also a dri­ver of cre­ativ­i­ty. It com­bines ele­ments of our life in a strange or abstract way, trans­form­ing itself into a source of pro­lif­ic ideas and inspi­ra­tions3.

Clau­dia Picard-Deland empha­sis­es that “the dream is per­haps only a win­dow on cer­tain process­es that can take place dur­ing sleep,” such as mem­o­ry con­sol­i­da­tion. “It is pos­si­ble, how­ev­er, that dreams have a role to play,” she adds, “we still don’t have the means to prove it, but we can find cor­re­la­tions.” She mea­sured bet­ter per­for­mance in the com­ple­tion of a task learned the day before among par­tic­i­pants who had dreamed about this learn­ing4. She also observed an evo­lu­tion in the type of mem­o­ries that are inte­grat­ed into dreams dur­ing sleep. At the begin­ning of the night, these are recent ele­ments, from the day for exam­ple. The fur­ther we progress through the sleep cycles, the more dis­tant mem­o­ries, such as those from child­hood, are incor­po­rat­ed into the dreams5. “Dream­ing about the day’s learn­ing, cre­at­ing asso­ci­a­tions with oth­er mem­o­ries, would allow them to be inte­grat­ed,” she suggests.

In touch with dreams

What­ev­er the the­o­ries on the pos­si­ble func­tions of dream­ing, there is gen­er­al­ly lit­tle evi­dence to sup­port them. It should be said that it is dif­fi­cult to access this inti­mate expe­ri­ence at the moment it occurs. Some stud­ies are based on the a pos­te­ri­ori reports of dream­ers, which makes it pos­si­ble to tar­get a large num­ber of indi­vid­u­als. How­ev­er, tran­scrib­ing a mul­ti­sen­so­ry expe­ri­ence is not easy, espe­cial­ly as the brain likes to rein­tro­duce coher­ence into the nar­ra­tive, at the risk of cre­at­ing false mem­o­ries6.

“Anoth­er pro­to­col, the most com­mon in neu­ro­science, is that of ser­i­al awak­en­ings in the lab­o­ra­to­ry,” explains Del­phine Oudi­ette. The vol­un­teers are fit­ted with elec­trodes to record the stages of sleep (mea­sur­ing brain activ­i­ty, mus­cle tone, eye move­ments, etc.). Their sleep will be inter­rupt­ed sev­er­al times dur­ing the night by an inter­view. The clos­er it is to the occur­rence of the dream, the more like­ly the account will be accurate.

Some sleep dis­or­ders offer real-time win­dows of obser­va­tion into the dream. Som­nam­bu­lism, night ter­rors, and also RBD can break the mus­cle bar­ri­er that pre­vents us from mov­ing dur­ing sleep7. It is thus pos­si­ble to observe on video the actions of these patients phys­i­cal­ly liv­ing out their dreams8. How­ev­er, these noc­tur­nal behav­iours are rare, some­times dif­fi­cult to inter­pret, and it is less clear whether the results can be gen­er­alised to the gen­er­al population.

The fur­ther we progress through the sleep cycles, the more dis­tant mem­o­ries, such as those from child­hood, are incor­po­rat­ed into the dreams

Lucid dream­ers, who have the abil­i­ty to become aware that they are sleep­ing, are also involved in the study of dreams. They can com­mu­ni­cate from their sleep through pre-estab­lished signs, such as eye move­ments. They thus pro­vide infor­ma­tion in real time about their dream9. To max­imise the appear­ance of these win­dows of con­scious­ness, expert lucid dream­ers are called upon, as well as patients with nar­colep­sy, a sleep dis­or­der, one of the side effects of which is to pro­mote this phenomenon.

Whether report­ed by nar­ra­tive or observed in real time, this infor­ma­tion about dreams can be cor­re­lat­ed with record­ings of brain activ­i­ty to study a dream sig­na­ture10. Arti­fi­cial intel­li­gence tools some­times help to decode this data11.

Towards a better understanding

“The sci­ence of dreams means many sleep­less nights for the researchers, some­times with few dreams col­lect­ed,” empha­sis­es Clau­dia Picard-Deland, “hence the impor­tance of col­lab­o­ra­tions and open data­bas­es for accu­mu­lat­ing record­ings.” Like DREAM, it is a col­lec­tion of dream reports and brain activ­i­ty record­ings col­lect­ed from around the world12. Lucid dreams are even rar­er, and try­ing to induce them is anoth­er way to increase the num­ber of sam­ples. This is anoth­er avenue she is exploring.

Gain­ing a deep­er under­stand­ing of dream­ing also involves improv­ing the qual­i­ty of the data col­lect­ed, thanks to more effi­cient mea­sur­ing instru­ments. Ultra­sound imag­ing is full of promise, accord­ing to Del­phine Oudi­ette: it would make it pos­si­ble to record brain activ­i­ty with a lev­el of pre­ci­sion and sig­nal qual­i­ty that has not been achieved until now. Cur­rent­ly only used on ani­mals and human babies13, it could soon be used on adults. “It’s quite excit­ing, this type of tech­no­log­i­cal advance would allow us to make a leap for­ward,” she enthuses.

“In dream neu­ro­science, there is every­thing to be done,” con­tin­ues Del­phine Oudi­ette. She has recent­ly observed, dur­ing cer­tain dreams, win­dows of con­nec­tion with the out­side world where sleep­ers are no longer cut off from the world. It is thus pos­si­ble to com­mu­ni­cate with them dur­ing their sleep. She is seek­ing to bet­ter under­stand how this phe­nom­e­non works. Does it pro­vide a clue as to how deep sleep is? Too many of these win­dows might explain the sub­jec­tive impres­sion of poor or lit­tle sleep in some patients suf­fer­ing from insom­nia. “These peo­ple may feel as if they have been awake when they have slept through the night,” con­firms Clau­dia Picard-Deland, who is study­ing the link between dream­ing and insom­nia. But her new project con­cerns a more social aspect: “Telling a dream is reveal­ing one­self to the oth­er, trust­ing them. Would there be any ben­e­fit to shar­ing it?” For the sake of sci­ence at least, telling our dreams will always be of interest.

Mikaël Mayorgas
1Arnulf, I., Grosliere, L., Le Corvec, T., Gol­mard, J.-L., Las­cols, O., & Duguet, A. (2014). Will stu­dents pass a com­pet­i­tive exam that they failed in their dreams? Con­scious­ness and Cog­ni­tion, 29, 36‑47. https://​doi​.org/​1​0​.​1​0​1​6​/​j​.​c​o​n​c​o​g​.​2​0​1​4​.​0​6.010
2Maran­ci, J.-B., Nigam, M., Mas­set, L., Msi­ka, E.-F., Vion­net, M. C., Chaumereil, C., Vidail­het, M., Leu-Seme­nes­cu, S., & Arnulf, I. (2022). Eye move­ment pat­terns cor­re­late with overt emo­tion­al behav­iours in rapid eye move­ment sleep. Sci­en­tif­ic Reports, 12(1), 1770. https://doi.org/10.1038/s41598-022–05905‑5
3Horowitz, A. H., Esfa­hany, K., Gálvez, T. V., Maes, P., & Stick­gold, R. (2023). Tar­get­ed dream incu­ba­tion at sleep onset increas­es post-sleep cre­ative per­for­mance. Sci­en­tif­ic Reports, 13(1), 7319. https://doi.org/10.1038/s41598-023–31361‑w
4Picard-Deland, C., Aumont, T., Sam­son-Rich­er, A., Paque­tte, T., & Nielsen, T. (2021). Whole-body pro­ce­dur­al learn­ing ben­e­fits from tar­get­ed mem­o­ry reac­ti­va­tion in REM sleep and task-relat­ed dream­ing. Neu­ro­bi­ol­o­gy of Learn­ing and Mem­o­ry, 183, 107460. https://​doi​.org/​1​0​.​1​0​1​6​/​j​.​n​l​m​.​2​0​2​1​.​1​07460
5Picard-Deland, C., Konkoly, K., Raider, R., Paller, K. A., Nielsen, T., Pigeon, W. R., & Carr, M. (2023). The mem­o­ry sources of dreams : Ser­i­al awak­en­ings across sleep stages and time of night. Sleep, 46(4), zsac292. https://​doi​.org/​1​0​.​1​0​9​3​/​s​l​e​e​p​/​z​s​ac292
6Con­way, M. A., & Love­day, C. (2015). Remem­ber­ing, imag­in­ing, false mem­o­ries & per­son­al mean­ings. Con­scious­ness and Cog­ni­tion, 33, 574‑581. https://​doi​.org/​1​0​.​1​0​1​6​/​j​.​c​o​n​c​o​g​.​2​0​1​4​.​1​2.002
7Broughton, R. J. (2022). The Para­som­nias and Sleep Relat­ed Move­ment Disorders—A Look Back at Six Decades of Sci­en­tif­ic Stud­ies. Clin­i­cal and Trans­la­tion­al Neu­ro­science, 6(1), 3. https://​doi​.org/​1​0​.​3​3​9​0​/​c​t​n​6​0​10003
8Siclari, F., Val­li, K., & Arnulf, I. (2020). Dreams and night­mares in healthy adults and in patients with sleep and neu­ro­log­i­cal dis­or­ders. The Lancet Neu­rol­o­gy, 19(10), 849‑859. https://doi.org/10.1016/S1474-4422(20)30275–1
9Konkoly, K. R., Appel, K., Cha­bani, E., Man­gia­ru­ga, A., Gott, J., Mal­lett, R., Caugh­ran, B., Witkows­ki, S., Whit­more, N. W., Mazurek, C. Y., Berent, J. B., Weber, F. D., Türk­er, B., Leu-Seme­nes­cu, S., Maran­ci, J.-B., Pipa, G., Arnulf, I., Oudi­ette, D., Dresler, M., & Paller, K. A. (2021). Real-time dia­logue between exper­i­menters and dream­ers dur­ing REM sleep. Cur­rent Biol­o­gy, 31(7), 1417–1427.e6. https://​doi​.org/​1​0​.​1​0​1​6​/​j​.​c​u​b​.​2​0​2​1​.​0​1.026
10Siclari, F., Baird, B., Per­ogamvros, L., Bernar­di, G., LaRocque, J. J., Ried­ner, B., Boly, M., Pos­tle, B. R., & Tononi, G. (2017). The neur­al cor­re­lates of dream­ing. Nature Neu­ro­science, 20(6), 872‑878. https://​doi​.org/​1​0​.​1​0​3​8​/​n​n​.4545
11Horikawa, T., Tama­ki, M., Miyawa­ki, Y., & Kami­tani, Y. (2013). Neur­al Decod­ing of Visu­al Imagery Dur­ing Sleep. Sci­ence, 340(6132), 639‑642. https://​doi​.org/​1​0​.​1​1​2​6​/​s​c​i​e​n​c​e​.​1​2​34330
12Wong, W., Andrade, K. C., Andrillon, T., De Arau­jo, D. B., Arnulf, I., Avvenu­ti, G., Baird, B., Belle­si, M., Berg­amo, D., Bernar­di, G., Bla­grove, M., Collins, M. B., De Gen­naro, L., Decat, N., Demirel, Ç., Dresler, M., Eichen­laub, J.-B., Elce, V., Gott, J. A., … Tsuchiya, N. (2023). DREAM : A Dream EEG and Men­ta­tion data­base. https://​doi​.org/​1​0​.​3​1​2​3​4​/​o​s​f​.​i​o​/​69e43
13Deffieux, T., Demené, C., & Tan­ter, M. (2021). Func­tion­al Ultra­sound Imag­ing : A New Imag­ing Modal­i­ty for Neu­ro­science. Neu­ro­science, 474, 110‑121. https://​doi​.org/​1​0​.​1​0​1​6​/​j​.​n​e​u​r​o​s​c​i​e​n​c​e​.​2​0​2​1​.​0​3.005

Our world explained with science. Every week, in your inbox.

Get the newsletter