Dream experiment postmortem

A couple of years ago I attempted to duplicate the “experiment with time” described in J. W. Dunne’s book of that title. The experiment involves keeping detailed records of one’s dreams for a period of time and noting any resemblances between dream events and waking events which occurred shortly before or shortly after the dream in question. The higher the ratio P:F (the number of past resemblances to future resemblances), the stronger the evidence that apparently precognitive dreams are mere coincidences; the lower the ratio, the stronger the evidence that dream precognition is just as real as dream retrospection.

I posted the dream records I kept during the experiment at experimentsintime.wordpress.com. Relatively few of my dreams turned out to be discernibly connected to specific past or future events, and none of the resemblances I did notice were sufficiently strong to be truly compelling. Of the 23 dreams I recorded (or 23 nights’ worth of dreams, rather; I made no attempt to separate a given night’s dreaming into distinct dreams), 8 exhibited resemblances to specific past events (6 weak, 2 moderate), and 5 resembled specific future events (3 weak, 2 moderate). Thus my experiment would seem to be consistent with Dunne’s thesis that dreams are constructed from a roughly equal mixture of past and future components. (If all future resemblances were coincidental, we would expect past resemblances to outnumber them by at least an order of magnitude). However, the small number of resemblances noted, together with their overall weak quality, means that my experiment cannot be regarded as conclusive either way.


Apparently I am not a “good” dreamer in the sense that the experiment requires. That is, the raw materials from which my dreams are constructed tend to be integrated to a degree that it becomes difficult to identify the specific experiences on which the dreams are based. Dunne touches on this in his appendix to the third edition of An Experiment with Time.

It is a commonplace of psychology . . . that most structures of the imagination are ‘integration’ — blends of several images associated with several different waking impressions. And it is accepted generally that dream-images are mostly of the same character — certainly, it is very rarely that one comes upon one of these exhibiting an unmixed, photographic resemblance to any scene of waking life. Now, the possibility of discovering in one of these composite structures an element distinctive enough to be recognizable as pertaining to a chronologically definite incident of waking life depends, mainly, upon what may be described as the coarseness of the blending. The more intricate — the more fine-grained — is the integration, the more difficult becomes its analysis. And, with practice in constructing dream-images, just as with practice in waking imagery, the integrations become more fine-grained, more beautifully blended, and, so, less easy to associate with any chronologically distinctive waking incident, past or future (p. 137, Hampton Roads reprint).

I suppose I ought to be pleased that my dreams are so “fine-grained” and “beautifully blended” — that my dreaming self exhibits a high degree of creativity (in the Einsteinian sense of knowing how to hide your sources) — but mostly I am disappointed at my inability so far to evaluate Dunne’s thesis conclusively from my own experience.

Individuals apparently vary widely in the relative “coarseness” of their dreams. In an experiment involving six Oxford students, one of the subjects recorded 21 dreams, of which 18 resembled waking incidents, 6 of those resemblances being judged “good” or strong. Another recorded 16 dreams without finding a single resemblance of any value. It is my bad fortune to be closer to the latter end of the scale.

Another factor may be the regular — not to say boring — character of my day-to-day life, which would reduce the chance that anything in a dream would resemble a distinctive waking event belonging clearly to either the future or the past. Not wanting to burden a holiday with the troublesome task of keeping daily dream records, I ignored Dunne’s advice to carry out the experiment during a break from one’s usual routine — preferably during a vacation in an unfamiliar locale — and instead conducted it during a perfectly ordinary period of time. I shall probably be going to Australia next year for the first time, and may attempt a second experiment during that trip.


Some months after the end of my experiment, I did finally have a dream which bore a conclusively strong resemblance to a specific future event.

In my dream, I had the idea that I ought to write a book entitled Pineapples and Apple Pies. It would be an English textbook for use in Taiwan. The title was perfect because it illustrated the rules of compound formation in English: pineapple and apple pie were made up of the same two components (sic) but in a different order; also, pineapple was written as a single word with the stress on the first element, while apple pie was written as two with the stress on the second. I was delighted to have discovered two such common words which were thus related — much more natural than such strained pairs as songbird-birdsong and housecat-cathouse. As an added bonus, a pineapple was something typical of Taiwan, while an apple pie was a well-known symbol of America. (Upon waking, I naturally realized that the elements of the two compounds were not identical after all, that pine and pie were not the same.)

The day after the dream, I was teaching E., a child whom I tutor privately. At each of our sessions, he has to sign a record sheet, and, with a young child’s normal love of nonsense, he often writes a random word or two (often ghost or pig) on the paper after signing his name. This time, the random “word” he wrote was — applepine. When I said, “Applepine? What’s that?” he said, “You know, like pineapple but the other way!” Of course I had never in my life encountered the “word” applepine before, and the odds of running into it by chance the day after my dream are effectively zero. In fact, E.’s choice of that word is so bizarre and unaccountable that I’m almost more inclined to consider it evidence of mind-reading on his part than of precognition on mine.

Unfortunately, even this does not really count as conclusive evidence, since it occurred outside the formal experiment. The key thing is the ratio of clearly precognitive dreams to clearly retrospective ones, and I wasn’t keeping any records of the latter at the time. However, it did serve to keep me interested in the question, and I shall probably conduct further experiments in the future.


Dunne also describes similar experiments in waking precognition, something I may try in the future, since my dreams seem to be of such poor quality, and since I have experienced some success with waking precognition in the past, before I ever encountered Dunne’s ideas. The following is taken from an email I wrote to a family member.

Here’s an experiment you can try. Go somewhere dark, roll your eyes up and to the side like people do when they’re trying to remember something, and “try to get something.” Just try to pick something up, like tuning a radio — easier done than explained. When I tried this, the first thing I “got” was the nonsense words “wudder-wudder-wudderfly” followed by a rapid succession of mental images: a green parrot viewed in profile; a short, wide, yellow tin; and something else which I no longer remember. Hours later, at work, one of my coworkers who had just come back from abroad (the Philippines, if memory serves) gave me some snack food from that country — in a package decorated with a green parrot in profile. Another colleague, a Japanese teacher, was doing some sort of cooking activity with her students and has brought some kind of Japanese bean paste or something in a yellow tin like the one I had seen. I can’t remember now what the third image was — this was years ago — but it also “came true.” (Nothing came of “wudder-wudder-wudderfly,” though.) I haven’t tried that particular exercise again — partly because I didn’t know what I was “tuning in” to and it seemed a little creepy, and partly because I didn’t want to ruin the magic by trying it again and not getting anything.

Given that, against the background of Dunne’s theories, precognition no longer seems “creepy” to me, I may try an organized experiment in waking precognition in the near future. Of course it will be impossible to calculate a ratio of precognitive to retrospective images, but if the results are sufficiently striking they may be conclusive anyway.


Filed under Anecdotes, Dreams, Precognition / Prophecy

11 responses to “Dream experiment postmortem

  1. Thanks for this report and summary.

    I tend to assume that there is ‘something in it’ about this field – on the basis of past societies taking dreams very seriously. And my assumed mechanisms are a mixture of telepathic and divine. For instance, your pineapple/ applepine instance might be explained by some attuning of yours and the student’s mind.

    What I do find strange/ impossible is a purely secular materialist explanation of genuine pre-cognition; because I would suppose that humans either could or could not do this, and that the kind of enigmatic, mixed and garbled precognition which Dunne describes would not be a plausible thing.

    If, on the other hand, we acknowledge human free will and autonomous agency, then it seems to me plausible that divine communications would, in most people, end up being garbled somewhat (except when the recipient was an ‘infallible’ prophet – which is perhaps impossible) – because, I mean, that the mixture of wholly accurate divine communication and the errors entailed by human reception/ understanding/ expression would produce the garbling.

    This, then, leads onto that old question of divine pre-cognition – and whether it is essentially predictive or out-of-Time. Since I (currently) tend to believe in a more Mormon idea of God (in Time, not out-of-Time) then I suppose that even a divine communication can never say that a precise thing will happen for sure – at least not when human free will stands between the prediction and the occurence, since the outcome of human free will cannot (intrinsically) be predicted for sure – not even by God.

  2. While it is obviously a possibility that at least some precognitions are divine communications, I have a hard time interpreting as revelations those which Dunne relates. Aside from being “garbled,” they are just all so pointless. Virtually all of the precognitive dreams described in Experiment deal with events of no importance (such as seeing a woman walking down the street holding an unusually-designed umbrella upside down), or at least of no relevance to the recipient (e.g. a volcanic eruption on the other side of the world). It is hard to imagine what possible purpose the Deity could be furthering by these scattershot revelations of seemingly meaningless facts. Subtle is the Lord, but surely there are limits!

    As I’m sure you will recall, Dunne himself proposes a theory of Time which purports to explain precognition (including its garbled nature and its tendency to occur more often in dreams than in waking life) in naturalistic terms. Having just now finished my first reading of Experiment, I have not yet had time adequately to process and evaluate Dunne’s theory, but it appears promising. What is your own opinion of it?

    • Well, Dunne would not be a candidate for divine communication – I don’t think he was any kind of theist.

      I thought Dunne’s theory was not true – And I don’t see how you could get a theory of physics from the dream data he had. It is a kind of Platonic Theory, but like all such it leaves unexplained the ‘interface’ between serial and non-serial time – which has the effect of kicking the can down the road from one un-understandable thing to another.

      I think Dunne’s ideas were popular because they opened a door to taking dream content seriously in an obejctive sense (whereas Freud had made dreams of purely subjective importance). But Jung is the most influential ‘objective’ dream theorist of the twentieth century – and there is some truth (I believe) in Jung’s general idea of a kind of human group mind. However, this does not provide the ‘meaning’ that Jung seemed to believe it did – because it yields a purely instrumental/ therapeutic concept of dreaming.

  3. Samson J.

    Interesting that you write about this just now. Without divulging specifics, my wife and I have recently had several bits of very important news, and it so happens that each of us had had a dream about one of the news items. Possibly coincidence, but do you ever get that feeling, that intuition that you just *know* something is more than coincidence?

    I’m keen to try your “try to get something”-in-the-dark experiment.

  4. Bruce, if you agree with me that Dunne’s precognitive dreams were not of divine origin, then they must have been caused by something else. You say you assume “a mixture of telepathic and divine” mechanisms, but several of Dunne’s results are clearly not telepathic in the sense of having been communicated to him by another person. You will recall the volcano dream, in which the number he dreamed of was the number of casualties as he misread it in a newspaper article; and the wasp dream, in which the precognition hinged on a mental association which he and only he made.

    Samson, if you do decide to try any experiments in precognition, I should be most interested to know the results. E-mail me (wiltyc at gmail) if you get anything.

  5. Samson J.

    Well, if you’re seriously interested, I think I should probably put a bit of care into making the experiment a little more scientifically “rigorous”. I’ll think about how to do this, and feel free to offer me any suggestions!

  6. Pingback: Dunne’s experiments in waking precognition | Bugs to fearen babes withall

  7. @WmJas – I have pondered your point about these particular detailed precognitive dreams, that are also seemingly meaningless. I conclude that insofar as they cannot be explained by telepathic means or coincidences, then I just don’t believe them. In effect, because I can’t make any sense of of them, I don’t believe they happened! I am not making this as an argument, but a statement of my own response.

  8. How very Humean of you, Bruce!

    I think people very often reject information for similar reasons (see my post on talking statues), though they’re rarely so honest about what they’re doing.

    Nor is it necessarily invalid to do so. A “miracle” (defined here as an alleged occurrence which it is impossible to make sense of within the context of your current understanding of the world; so Jesus’s walking on water would not be a “miracle” from the point of view of a Christian) should be accepted only if you judge it more likely that your entire worldview is wrong than that the miracle story is bogus.

    Since I am already quite sure that my current understanding of time is completely inadequate, I am relatively open to “miracles” which would call that understanding into question. (When it comes to talking-statue yarns, on the other hand, I am shockingly dismissive!) Your situation vis-a-vis such “miracles” would naturally be different, and would depend on what exactly your worldview is and how confident you are that it is correct.

  9. @WmJas – This is the way that scientists work – or at least theoretical scientists. (I found this encapsulated in Francis Crick’s autobiography What Mad Pursuit, which I read in c 1988 during the years when I was winding-up to shift from lab science to theoretical)

    They (we, I) *never* believe *all* the data – but work on the assumption that some of the data are wrong. The job is to pick out the correct data and incorporate only that into the theory – then use the theory to guide further observation.

    A good theory will, eventually, explain why the wrong data were wrong (of course this assumes that the data are honestly derived as well as competently generated – but of course that would be a stupid assumption in research nowadays). But initially, and perhaps for some considerable while, the assumed-wrong data will simply have to be ignored – even though it cannot conclusively be proven wrong.

    I have brought this ingrained habit/ trained skill into theology – so that I always expect any source or set of data to contain errors, to have contradictions. This means that any theory which accounts for *all* the data *must* be wrong (because some of the data will certainly be wrong).

    I know that some/ most theologians regard their data (especially scripture) as 100 percent accurate – IF properly understood – but this merely has the effects of making all disputes into wrangles over how things are to be understood, and generates a vast – almost unconstrained – diversity of scriptural and other evidential ‘understandings’.

    This is another of the things I find rings true about the frank Mormon admission of current fallibility in the scripture, prophets, understandings etc – and the aimed-at process as one of increasing but never complete understanding/ knowledge – because this is much more the way that things worked in the (now gone) era of progress in real science.

  10. Pingback: Waking precognition: a trial run | Bugs to fearen babes withall

Leave a Reply to Samson J. Cancel reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s