The dreaming mind as a danger simulator
Antti Revonsuo and Threat Simulation Theory
Antti Revonsuo proposed that dreaming evolved as an offline world simulator in which dangerous events could be perceived and responses rehearsed without their real-world costs. Dreams contain threats, but does that establish an adaptive training function?
An evolutionary proposal
How to approach Revonsuo on dreams
Antti Revonsuo is a Finnish cognitive neuroscientist and philosopher known for research on consciousness and dreaming. In 2000, he set out the Threat Simulation Theory, often shortened to TST, as a testable evolutionary account of why dream experience has the form and content it does.
TST does not claim that every dream is a nightmare or that every threatening image predicts danger. Its stronger claim is functional: ancestral dreaming repeatedly simulated threats, rehearsing perception and avoidance in ways that improved survival and reproduction.
Finding danger in dreams supports a content claim. It does not by itself prove rehearsal, improved performance or evolutionary selection.
Six propositions behind the theory
1. A simulated world
Dreaming constructs an organised world with a body, places, objects, characters and unfolding events.
2. A threat bias
Threatening events and negative outcomes are proposed to be selectively overrepresented.
3. Waking activation
Real exposure to danger should increase the frequency and severity of dream threats.
4. Realistic engagement
The dream self should perceive danger and engage with it in a sufficiently realistic simulation.
5. Implicit rehearsal
Repeated simulation should strengthen threat perception and avoidance even without dream recall.
6. Evolutionary advantage
Better prepared ancestors should have gained a survival and reproductive advantage.
Consciousness inside a model
Dreaming as virtual reality
Revonsuo treats a dream as an immersive simulation rather than a collection of isolated pictures. The dream self usually has a body and occupies a three-dimensional world. Events seem to happen around and to that self, and the dream is commonly mistaken for reality while it lasts.
This world-simulation claim is broader than TST and can be accepted without agreeing that danger rehearsal is dreaming’s evolved function. A simulated world may enact relationships, memory, desire, routine activity, exploration, absurdity, and threat.
Dreaming can be a simulation without being a training exercise, just as a story can depict danger without having evolved to teach escape.
What counts as a threat?
Research using the Dream Threat Scale identifies an event as threatening when it could cause physical, psychological, social or material harm to the dream self or another character. Threats include attack, pursuit, accident, illness, loss, exclusion, failure and environmental danger.
Broad definitions capture much of ordinary dream adversity, but they also create a specificity problem. Missing an examination and escaping a predator are both negative simulations, yet they involve different systems, consequences and evolutionary arguments.
The evolutionary logic
Ancestral humans faced predation, violence, accidents, conflict, illness and social exclusion. Revonsuo reasoned that a mechanism repeatedly simulating dangerous situations could improve rapid detection and response without exposing the individual to actual harm.
For this to be an evolved adaptation rather than a useful side effect, threat simulation would need heritable variation and a reliable effect on survival or reproduction. Present dream reports can test predicted traces of the mechanism, but they cannot directly reconstruct ancestral fitness effects.
Current utility
A dream might incidentally prepare someone for a situation today.
Evolutionary function
Dreaming was selected because rehearsal increased ancestral fitness.
The first claim does not establish the second. Evolutionary explanations must show more than a plausible benefit.
A worked example
The narrowing path
I am walking along a cliff path when stones begin falling behind me. Ahead, a large dog blocks the path. I look for another route, climb onto a ledge and wait until the dog moves away.
This report contains environmental danger, an animal threat, rapid evaluation and an avoidance response. It fits the descriptive features of a threat simulation. TST would further propose that the simulation rehearses relevant perceptual and behavioural systems.
The report alone cannot show that rehearsal occurred or transferred to waking behaviour. A continuity account might instead relate the scene to current stress, while a metaphorical account might explore being trapped between two dangers. These readings can coexist at different explanatory levels.
Evidence from trauma-exposed children
Katja Valli, Revonsuo and colleagues compared dreams from severely traumatised and less traumatised Kurdish children with those of non-traumatised Finnish children. The severely exposed children reported more dreams, more threatening events and threats of greater severity.
The result supports TST’s activation prediction: exposure to real danger was associated with increased threat content. Yet the comparison involves populations differing in culture, living conditions, language and reporting context. It also does not demonstrate that the additional dreams improved coping or safety.
More threat after trauma is compatible with an activated rehearsal system, but also with continuity, intrusive memory and heightened emotional salience.
Are dreams more negative than waking life?
Valli and colleagues collected 419 dream reports and 490 daily event logs from 39 university students over two weeks, together with interviews about remembered past threats. Dream threats were more frequent and severe than current waking threats, and resembled past threats more closely than current ones.
This finding supports a selective negativity bias rather than simple copying of daily life. However, comparison units matter: a dream report and a waking event log are produced under different conditions, and memorable dream episodes may be more narratively concentrated than ordinary waking periods.
The response problem
A useful rehearsal should involve threat detection and a response that can be practised. Yet dreamers often freeze, fail, awaken, repeat an ineffective action or remain passive. Recurrent dreams and nightmares frequently end before successful avoidance occurs.
TST can reply that detecting and engaging with threat may itself exercise relevant systems, and practice need not end successfully on every trial. Even so, the transfer from dreamed response to improved waking performance remains the theory’s most important unverified link.
A direct contradictory test
Susan Malcolm-Smith, Mark Solms and colleagues compared participants living in a high-crime area of South Africa with participants in a lower-crime area of Wales. The South African group had greater recent exposure to life-threatening events but reported fewer threat dreams.
The groups did not differ in responses to dream threats. Fewer than 20 per cent of dreams contained realistic survival threats, and escape occurred in fewer than 2 per cent. The authors concluded that their findings contradicted key predictions of TST.
Interpretive balance: differences in collection methods, culture, exposure types, and threat definitions may help explain the conflict between studies. They do not justify ignoring a well-designed negative result.
What does the evidence support?
| Claim | Current appraisal | Reason |
|---|---|---|
| Dreams create organised simulated worlds | Well supported descriptively | Dreams commonly contain an embodied self, settings, characters and unfolding events experienced as real. |
| Dream content has a negativity or threat bias | Substantial support | Several content studies find aggression, misfortune, and negative emotions to be more frequent than their positive counterparts. |
| Waking trauma can increase dream threats | Supported but inconsistent | Traumatised child samples support the prediction, while the South Africa and Wales comparison did not. |
| Dreamers efficiently avoid simulated threats | Weak or mixed support | Failure, passivity and awakening are common, and successful escape can be rare. |
| Dream rehearsal improves waking threat skills | Not directly demonstrated | There is no decisive evidence linking an individual’s threat dreams to later performance gains. |
| Threat rehearsal is dreaming’s evolved function | Plausible but unproven | Present content cannot establish ancestral heritability, selection and fitness effects. |
Three claims that must remain separate
Content
Dreams frequently represent threats and negative events.
Process
Representing threat exercises perception and response systems.
Function
That exercise improved ancestral survival and caused the system to be selected.
Evidence becomes progressively thinner from left to right. A sound assessment should not use strong evidence for threat content as if it equally supported adaptive evolutionary function.
The ancestral and modern mismatch
Modern threat dreams often concern examinations, lost phones, missed transport, workplace failure or public embarrassment. TST can interpret these as a flexible ancient system responding to current threats, especially threats to status and belonging.
The flexibility also makes the theory harder to falsify. If ancestral dangers, contemporary worries, realistic threats and symbolic failures all count, almost any negative dream may be classified as simulation. A useful theory must specify what evidence would count against it.
From threat to social simulation
Revonsuo and colleagues later developed Social Simulation Theory. Dreams are highly social, and many simulate attachment, cooperation, conflict, mind-reading and relationships rather than physical danger. The later theory broadens the evolutionary programme beyond threat.
Research supports a strong sociality bias in dream reports, but adaptive rehearsal again remains disputed. Social content may reflect the importance of people in memory and waking concerns without having a specialised training function.
Using the theory responsibly
- Describe the threat precisely before interpreting its purpose or symbolism.
- Notice how the dream self detects, appraises and responds to danger.
- Distinguish successful action from freezing, failure, awakening or unresolved repetition.
- Explore relevant waking threats without assuming that the dream is useful rehearsal.
- Consider continuity, trauma, memory and social simulation as alternative or complementary accounts.
- Do not treat recurring threat dreams as proof that the person unconsciously wants or needs further exposure.
Nightmares and clinical limits
TST is a theory of dream function, not a treatment for nightmares. Recurrent nightmares can impair sleep, increase anticipatory anxiety and maintain daytime distress. Their possible evolutionary history does not make current suffering adaptive or beneficial.
Evidence-based approaches such as imagery rehearsal therapy intentionally alter a nightmare script and rehearse a different version while awake. Its clinical evidence does not confirm TST, because deliberate waking rescripting differs from spontaneous threat dreaming.
Threat as metaphor
The landscape of possible harm
Threat dreams turn abstract vulnerability into spatial relations: danger closes in, support disappears, ground gives way, an exit narrows, or an unseen force pursues. The imagery can model how danger is organised even when it does not rehearse an effective response.
A metaphorical exploration asks what kind of danger is pictured and how the dream self is positioned. TST asks why a biological system would repeatedly create such situations. These are related but distinct questions.
Questions for examining a threat dream
- What exactly is threatened: body, attachment, status, belonging, property or future possibility?
- How quickly is the threat detected, and which cues reveal it?
- Does the dream self act, freeze, hide, negotiate, seek help or awaken?
- Is the response effective within the dream?
- What current or past waking threats resemble the dream relation?
- What would count as evidence that this dream rehearsed a skill rather than represented a concern?
- What meaning does the dreamer find, and which imposed explanation do they reject?
Research and further reading
- Revonsuo, 2000: The Reinterpretation of Dreams.
- Revonsuo, 2000: response to critiques of Threat Simulation Theory.
- Valli and colleagues, 2005: dreams of traumatised children.
- Valli and colleagues, 2008: dreams are more negative than real life.
- Malcolm-Smith and colleagues, 2008: Threat in Dreams, An Adaptation?.
- PubMed record for the South Africa and Wales comparison.
- Zadra and Donderi: threat perception and avoidance in recurrent dreams.
- Germain and colleagues: typical themes and the specificity problem.
- Franklin and Zyphur, 2005: broader evolutionary rehearsal hypothesis.
- Helminen and Punamäki, 2008: dream imagery in children exposed to military trauma.
- Domhoff and Schneider, 2018: comparison of social simulation and neurocognitive theories.
- Tuominen and colleagues, 2022: social seclusion, dream content and Social Simulation Theory.
- Dresler, 2015: dreaming as multifunctional simulation.
A final perspective
Threat Simulation Theory made dream function experimentally discussable by deriving predictions about content, trauma and response. Evidence supports organised dream worlds and a substantial threat bias, but findings about trauma exposure conflict and the crucial transfer from dream rehearsal to waking advantage remains undemonstrated. The theory is strongest as a research programme, not as a settled explanation of every threatening dream.
This page is for education and critical discussion. Dream interpretation is exploratory rather than diagnostic. Disturbing, recurrent or trauma-related dreams may warrant support from an appropriately qualified health professional.