Two investigations into the Light Switch Phenomenon

Hearne, K. A light-switch phenomenon in lucid dreams. Journal of Mental Imagery, 5 (2), 97 - 100. 1981.


A previous extensive study of LDs revealed that some accounts of both NDs and LDs referred to an inability of the dreamer to switch on an electric-light in the dream scenery. This observation was pursued by giving the task to a number of LDers in order to ascertain the reliability of such reports.


Eight LDers (five females, three males) were requested to attempt to switch on an electric-light in a LD and report what happened. Subjects were naive as to the purpose of the experiment.


Five of them couldn’t turn the light on. The sixth couldn’t even find the switch. The seventh only got a sparking and flickering orange bulb. Only the eighth was able to switch the light on, but only after she covered her eyes first.


Subjects can routinely carry out pre-sleep instructions in LDs, yet the task of switching on a light presents a strangely consistent difficulty. Its widespread nature indicates a possible important limitation in the “dream engineering” process. The eighth case, who could switch on a light after covering her eyes and abolishing the imagery, is revealing when considered with the experience of another person who reported that a light may be switched off first, and then on, but not the other way round. This shows a possibility exists that there’s a ceiling limit on brightness in the dream imagery at any point (but this ceiling may vary over time), and any attempt to violate that level by manipulating the dream results in rationalized avoidance of the intended situation. If so, it would suggest that an autonomous dream-producing process operates, which has to maneuver the dream within such limitations of imagery. Other less plausible explanations are:

  • The phenomenon reflects the operation of a “sleep-maintaining process” which avoids dream scenes such as a light suddenly switching on, since that might normally waken a sleeping person irl (although a light can switch on from a reduced imagery as with the eighth case).
  • The phenomenon may symbolize the “powerlessness” or “lack of energy” of the sleep state (although its apparent specificity does not support that notion).

Moss K., Performing the light-switch task in lucid dreams: A case study. Journal of Mental Imagery 1989 Sum, Vol 13(2), pp. 135-137.


Hearne (1981) found that subjects were unable to turn on an electric light in a LD and concluded that this task was beyond the ability of the LDer (see above). In a follow-up study Hearne (1982) again reported this inability. He then hypothesized that there existed “a varying ceiling-level of imagery brightness”. Tholey (1983) also reported that subjects were unable to perform the light-switch task. However, he did find that some subjects were able to gradually increase illuminance by other methods.


The author himself was the subject of a three-month investigation during which he had 70 LDs. In a selected portion of them, he attempted the light-switch task or the control task of increasing luminosity by any method other than turning on an electric light. The subject reports a history of having 830 LDs since 1979. In some of them, he was able to gradually and suddenly increase luminosity by a variety of methods. However, prior to the current study, he never attempted to turn on an electric light in a LD.


The light-switch task was successfully performed in 11 out of 15 LDs. In five dreams, the resulting luminosity increase exceeded current and all previous luminosity levels for that dream. A total of 33 different lights were attempted. The light turned on during 20 of these attempts. On two occasions, the light turned on but was not sustained. The light had little effect on the environment four times. The other positive results were sudden, bright, sustained and effective. All five attempts to turn the light back off were successful.
In eight LDs the subject attempted to increase luminosity by a method other than switching on a light. In all eight LDs, luminosity was increased above the previous maximum level. In four of these dreams, luminosity was increased a second time. These results included both gradual and sudden increases. Examples of intentional luminosity increases included looking at a light source, opening a curtain and walking outside into the sunlight.


Several factors were conducive to successfully performing the task. In dreams that initially had a negative result, the subject was able to achieve a positive result if he persisted in attempting or changed to a different light switch. Another important factor was to find an angle of view that best facilitated the flow of action (switch to light to scene). Therefore, a view that included the switch and light in close proximity would be the most favorable to make the test work. In dreams where they were separate, this situation could be approximated if they were viewed alternatively. If only the light was viewed, it was helpful to have a good tactile sensation of the switch. Other views, such as looking only at the switch or at some part of the scenery, were especially vulnerable to failure. A good strategy was to attempt a luminosity increase by another method and then decrease it back to the baseline level prior to attempting a light-switch trial. Practice, motivation, frequency and the waking study and attention to the task may also be factors.
So despite encountering some problems, the subject was able to successfully complete the light-switch task. The assignment of finding and operating the switch and pairing it with the light chance can be difficult. However, other tasks can be performed in LDs with appropriately paired results. A voluntary sudden increase beyond an established luminosity level may be difficult at times but is not impossible, as indicated by the results of the control task. Hence the light-switch task involves a number of problems, none of which are impossible, but which together make it difficult.

Hmm… this is very interesting. but if there is a ceiling limit, then why was it that the subject could not turn the light on, but when he opened the curtains the sun didn’t turn off…? :eh:

Good question :smile:

The ceiling limit was proposed by Hearne (the first experiment), but it was kinda rejected by Moss (second experiment). IF Hearne is right, then the sun would only be turned off if the switch from dark to light is very quickly and very strong. However, I don’t know if Hearne has tried this. Moss however succeeded in opening the curtains and letting the sunlight in without any problem. Perhaps there IS an upper limit, but the transition during the opening of the curtain could be different in some way than switching on the light while in total darkness. If the sun was already visible through the closed curtain, this could also lower the chance that the ceiling limit would be reached upon opening the curtains.

This study must be all bullshit, unless there is something special about me…

They may have in the past heard about the light switch thingy (it must have existed before the first experiment, or else the experiment wouldn’t have taken place). Which makes the experiment kindda redundant…

Don’t forget these experiments took place in the 1980s. Perhaps it seems bullshit at first, but even now from time to time there’s always someone on this forum questioning why the light-switch task doesn’t work in his/her dreams. Nevertheless, simply because YOU never have problems with the light-switch task, it doesn’t have to mean it was completely ridiculous to investigate this phenomenon…

Why is it you are able to read something, (if only once, but still capable) but clocks are filled with mumbo jumbo? It would seem that either text would be the same way, or the time would just change when your looked back at it. Has anyone read a theory or study on this phenomenon?

I know letters and numbers are registered differently in our brains, but this is still quite perplexing.

That is another myth… Clocks also usually works just fine.

I think it is all about expectations… Most people who claim light switches and such did not work, had already read that they are supposed to not work.

Although I usually use an analog clock when reality checking, it often displays a time, but then changes, instead of looking absurdly wrong (which it does sometimes). But again, text is often illegible as well.

I think just about everything in dreams is expectation. If you are scared in a nightmare–the lights might not work. But if you have become lucid, it may work. A lot of people have trouble with different things.

Like Tomas states, it IS all expectations. When I first LDed…I couldn’t get some things to work, like light switches. But…when you start making your own beliefs, they show up in your dreams. So if you go to a light switch and truely believe that it will turn on, it will! Same goes for everything, such as flying. Some people can’t fly, because they simply go hey, this is impossible. But once those people adapt to lucid dreaming, and obtain their own new fresh beliefs, there are no more limits.

This is why the new people are having so much problems on this forum. They keep reading things like LDing is hard, or even the small things like light switches dont work, or its hard to fly. So then they start saying to themselves that it really is hard, or things dont work. Which prevents them from doing what they really used to desire to do in a LD.

I don’t think everything is based on expectations. A few nights ago I had a non-lucid dream in which I had got a job offer in the mail to teach a ballet class for children, (in an airport no less.) and I was looking over the mail to find out what room number it was. The first time I checked it, the room number was a string of numbers. I scanned the doors, and each one had long string of numbers outside the doorway, but none of them matched the ones I had just read. After going past doors, I rechecked my paper. It was filled with a different string of numbers. I was frustrated, and instead of realizing this was a dream sign, I asked one of the security guards. Nonvital information to my point, but I was on the wrong floor. Don’t you hate when that happens?

I was expecting the numbers to stay the same because I didn’t realize I was dreaming, and they still changed. Explain this?

I think it is a bit different with long strings of txt/numbers and such, cause it will be alot of information for your brain to remember. Dreams are indeed usually quite a bit more unstable than waking state in some aspects, no matter what you expect.

What i react to is that there is so many articles that claim that is simply impossible to use a lightswitch, read and such in dreams, even though alot of dreamers here is a living proof that this is not true at all.


That’s something your dream person in a ND was expecting…you didn’t truely believe in what would happen conciously [so it would seem].

At the time of the experiments, researchers already knew dreams are highly influenced by expectations, associations and thoughts in general. So yes I think they agreed expectations play a big role in the light-switch task. However, the question was “could there be an upper-limit to this expectation”? If you were able to perform the task with success, then it’s easy to dismiss these experiments. But to me this question is very well worth to investigate. If everything you tried in a dream worked fine based upon expectations, even then, how can you really be sure there isn’t such thing as a ceiling limit to certain expectations, which would only become visible under certain special circumstances (just like Moss also acknowledged)? And that’s exactly what they were trying to investigate using the light-switch task.

Clocks have all the right figures for me. The only difference they have from real clocks is that their time changes when I look away and back again. However, I’m expecting them to change (it is, a RC) which is why they do. If I wanted to, I could make it remain the same quite easily.

And about that ND you posted up. Sounds to me like your sub-c was giving you a DS :wink:.

I think what you are saying is quite valid, however, if the task were around expectations wouldn’t they use a blind sample and compare it against another sample that was given the expectation? If they were trying to reach an understanding of ceiling limits for luminosity, surely there are better methods.

Hm I agree there are better ways to investigate solely the impact of expectations within dreams :smile:

Nevertheless, this investigation raises an important issue about possible ceiling limits in dreams. In one of the papers it was mentioned more experiments were under way, but there was no reference… I’ll look for it though :smile:

Great… I will look forward to reading those (and any others!) :grin:

Keep up the good work… I for one really appreciate these postings on experiments!!


If only I could find the papers describing these other experiments…

In any case, somewhere in the next few days you can expect a paper about experiments done to determine whether or not DCs have consciousness :smile:

I had an interesting dream the other night that must have been sparked by this post… and yet I didn’t think to go lucid!

So I would like to share this dream here!

To my recall, I have never dreamt about turning on lights etc, but I can now say I have. I was in my apartment and I went to turn on all 4 light switches for the lights (in RL there are only 3) One light went on (the furthest away from me), the next light went on, but flickered like the light bulb was about to die… this was the second closest to me, and the closest didn’t go on at all!!! The light that didn’t work… I repeatedly flicked the switch several times until the lights worked (cause in RL those friggin lights do that :grrr: ) The light that was flickering, I got that to work by grabbing a chair, standing on it, and then tapping the ceiling around the light… it blustered a little more light, flickered… died… and then came to full life. The dream ended with me content that I got all the lights to work.