A Very Short History of Dream Research

Sigmund Freud in 1906

While speculation about the nature of dreaming in the “western” tradition can be traced at least to Aristotle, serious study of the topic in the modern period clearly begins with Freud. Freud’s general psychology is today largely discredited, or at any rate a great deal less fashionable than it once was. The same can be said for his theory of dreaming, but nobody can deny that his initial efforts to untangle the mystery were monumental and enormously influential. Freud took dreams to be a “royal road” to understanding the unconscious mind, and developed a fairly complex theory for helping to interpret the “manifest” appearance of a dream into it’s secret “latent” content. In short, for Freud a cigar is rarely just a cigar. Dreams on this account are tricky, since they  simultaneously have to be the famous “guardians of sleep” (to keep us from waking up), yet also allow our minds to express unconscious wishes through “overdetermined” symbols. This is not the place to rehearse the edifice that is Freud’s theory – suffice it to say that his idea of dreams as masked communiques from the unconscious has had an inestimable impact on both psychology and ordinary life, and even if we reject the sometimes sex-obsessed interpretive lens of Freud himself this kernel of an idea remains extremely enticing. At the very least we must recognize (as Mathew Walker has recently pointed out) that Freud conclusively turned us away from the pre-modern idea that dreams were messages from the gods or prophecies of the future. Freud placed dreaming squarely where it belongs – in the mind/brain.

Nathaniel Kleitman, Aserinsky’s supervisor, in the 60’s. (Image from Smithsonian magazine, October 2003)

Of course, the biggest and most obvious difficulty with Freud’s approach is its total lack of scientific verifiability, or rather falsifyablity, which is the true test of a theory’s claim to “scientific” status. Even though cached in a scientific vocabulary, it seems impossible to imagine a counter-example to Freud’s theory that a clever Freudian can’t just explain away, which is why his work is so incredibly maddening to those in the “hard” sciences. For them, the real story of dreaming (and what is sometimes called the golden age of sleep research) doesn’t start until the mid 1950s when Eugene Aserinsky (one of my heroes, pictured at the top of the post in some VERY snappy research duds), Nathaniel Kleitmen and William Demmet began seriously studying the connection between “rapid eye movement” and dreaming. Subjects awoken during REM sleep reported having dream experiences seemingly all the time, while reports were negligible during non-REM (NREM) sleep. It seemed clear that dreams were somehow caused by (or otherwise connected with) REM sleep, and so the question became “what causes REM sleep?”.

Answering this question meant turning away from the psychological study of human subjects and their dreams and to the biology lab and a whole lot of unfortunate cats. By chopping away at cat brains, scientists (like Michel Jouvet) noted that you could actually remove a very great deal of the brain without impacting REM activity. Eventually they located a region in the lower, less evolved area of the brainstem (specifically the pons) that seemed to be responsible for REM sleep. That is to say, if the brain has damage to this region, REM activity ceases. While performing the same surgical tests on humans was obviously impossible, the results were confirmed by looking at subjects who naturally had damage to the same region.

The Pons! THIS is responsible for our dreams?

The results of such research were enormously destructive to Freudian theory. If dreams originate in the lower brainstem, it seems impossible to argue that any higher cognitive functioning of a repressed unconscious was involved – dreams, it seemed, were for the most part “mindless”. Theories like Allan Hobson’s went in an anti-Freudian, “epiphenomenal” direction, suggesting that dreams are merely an accidental physiological biproduct. On this account, the brain is “activated” by acetecholine triggers in the pons, and then attempts to “synthesize” the information it receives – hence the “activation synthesis” theory. Dreams were just brain’s way of coping with terribly bad signals arriving from primitive brainstem activities rather than regular sensory channels, and cigars became, once again, cigars. Since the publication of Hobson and McCarley’s paper outlining it in 1977, this theory was taken as gospel in non-psychoanalytic circles. In point of fact, Freud had already considered this view – describing (and rejecting) it as the “dreams are froth” theory.

However! The big “but” here is that the entire structure of this influential work rested on the correlation between REM and dreaming apparently established back in the ’50’s. Hobson and McCarley’s work, after all, was all about the causes of REM sleep, and it was just taken for granted that this was effectively the same thing as dreaming. This assumption turned out to be problematic.

While the early studies did suggest a correlation between REM and dreaming, it seems enthusiasm got the better of people and they ended up in a post hoc, ergo propter hoc fallacy. Much more recent work demonstates, rather interestingly, that even people who have sustained injuries to the REM-generating area of the pons still continue to have dreams! So even if this region IS responsible for REM activity (which seems pretty clear), it cannot be responsible for dreaming! And from the other direction, it turns out that normal (uninjured) people do indeed have dreams outside of REM sleep (even if NREM dreams are sometimes of a slightly different character). The early assumption a perfect correlation between the two needed re-thinking, and this error is an illuminating example of the kinds of mistake science can make when it follows only the more easily quantifyable, turning from human reports to cats under microscopes. Often enough the truths of science are just as much descriptions of the way scientists would like things to be as they are discoveries about the way things are, and I think dream science is a good example of the dangers of reductionist thinking.

At any rate, starting from this new perspective, scientists like Mark Solms began using the same strategy that Jouvet had used on his cats (though obviously without the invasive surgery). He looked at people with various brain lesions to see what regions were necessary for dreaming to take place, and this time the results were quite different. It turns out that the lowly brainstem is not particularly relevant here – only when certain areas of the forebrain are damaged does dreaming stop (specifically the “occipito-temporo-parietal junction” at the back of the forebrain, or the “limbic frontal what matter of the ventromesial quadrant of the frontal lobes”). Furthermore, the relatively recent technology of PET scanning (and, later, MRI scanning) helped scientists (like Alan Braun) get a better picture of what areas of the brain are active during sleep and dreaming.

The first “dream realization”? Early PET scan of the brain during REM – from Alan Braun’s (et. al.) 1997 paper.

Contra the activation-synthesis theory, the entire forebrain is not “activated” during REM – only very specific regions are. In particular, we see increased activity in the visuospatial regions at the back of the brain, the autobiographical memory center in the hippocampus, the motor cortex, and perhaps most importantly, the entire limbic system (the name for our entire emotional regulation structure), particularly the amygdala and the cingulate cortex. The emotional limbic system is, in fact, up to 30% more active during REM than when awake! In addition, scientists noted the distinctive lack of activity on both sides of the prefrontal cortex, the area Walker likes to describe as the “CEO” region of the brain which is responsible for logical decisions and ordered thought. This all strongly suggests that dreams are not created by merely random, frothy activity in the brainstem, but by crucial higher-order systems that most importantly include autobiographical memory and emotion, all while the logical control mechanisms are switched off.

Illustration from Solm’s book (listed below). The top illustration is the “dreaming brain”, the bottom the “emotional brain”.

Does this mean that Freudian ideas can slip back in, so to speak? Only sort of. I for one can understand those from the hard-nosed Hobsonian camp, because like them I have never been able to stomach the idea of smug Freudian know-it-alls passing down their wisdom from on high, or of giving up my own capacities for interpretation to any kind of guru who all-too-often abuses the power we grant them. I think much of Hobson’s frustration came less from the absurdities of Freudian theory than from the sense that too many psychoanalysts seemed to function as charlatan-guru-priests whose authority was largely undeserved. At the same time, I find some of Hobson’s writing bordering on vitriol, and it makes one wonder just why it all upsets him so much. Lots of people drive me nuts, including most bewitching, know-it-all gurus, but I don’t feel they merit the launching a rhetorical crusade against them. 

At any rate, today it seems that any hard-core epiphenomenalist approach has been almost as discredited as the Freudian model it gleefully displaced, and with both theories off the table a much more balanced and nuanced picture can begin to emerge. This, so far as I can tell, is the happy position of contemporary dream scholarship. Inquiring into dreaming requires, perhaps more so than any other science, a “dual aspect” approach, whereby we stop pretending either that the mind is “really” just the brain, or that the brain is irrelevant to theorizing the mind. From this sensible position, we have been able to decipher several central consequences of dreaming (and here I mean dreaming, not just REM sleep). For starters, and perhaps unsurprisingly, dreaming provides enormous assistance to our emotional lives. Dreams present us the opportunity to process, work through, and re-experience emotion, particularly anxiety and stress, from the comparatively “safe” place of sleep. We’re “safe” both physically (we’re cozy in our beds, not actually naked in front of a large audience), and neurochemically, because of the lack of norepinephrine (brain adrenaline) in our systems. So in our dreams we can “safely” explore emotional situations that have caused us anxiety, rehearse them, put them in their proper contexts, and ideally get over them. (In this regard, for example, Rosalind Cartwright demonstrated that people who had gone through emotionally traumatic divorces recovered much better if, and only if, they had dreams about the situation…).  In a related vein, dreaming has also been determined to help us with empathy, particularly in recognizing the emotions of others and interpreting facial expression. Without dreaming, our social

Without dreaming, you wouldn’t be able to tell this woman is unhappy with her date. Also, the man is a cactus.

skills would be pretty lousy, and we’d all be pretty terrible on a date. Finally, in accord with the sage old advice that recommends “sleeping on it,” dreaming can powerfully help with creative problem solving. Because the brain’s logical “CEO” is temporarily out to lunch, new, unexpected connections between seemingly unrelated ideas can be formed and radical solutions present themselves. In the process of comparing recent experience with past memory, we discover past events that provide relevant ideas  our conscious selves have only difficult access to.

Much more information on dream science is available in very accessible, non-specialist formats. The above text is largely a summary of Chapter 6 of Mark Solm’s and Oliver Turnbull’s excellent “The Brain and the Inner World” (2002, Other Press), which itself summarizes the story told in more detail by Andrea Rock’s “The Mind at Night” (2004, Basic Books). I also borrow heavily from Chapter 10 of Mathew Walker’s “Why we Sleep” (2017, Penguin), and here and there from everything else I’ve read and learned. Some other reading material suggestions can be found here.