Usless Science Forum
The Psyche => Science and Psychology => Topic started by: Matt Koeske on April 02, 2008, 12:33:48 PM
-
I just stumbled upon this while adding a Wikipedia source link to another post. Confabulation Theory (http://www.scholarpedia.org/article/Confabulation_theory) sounds like it could have some compatibility with Jungian thinking (or at least with my own twists on Jungian thinking).
The article linked above is dense, and my grasp of neuroscience is too limited to make a satisfactory evaluation of all of it. But I see (trumped up into a complicated scientific language) a lot of the same things I've grasped intuitively about cognition . . . especially while doing dream work.
For instance:
Formation of knowledge links involves a complex process of instantaneous, but temporary, knowledge link establishment at time of first co-occurrence; followed by repeated evaluations and strengthenings (if the evaluation is favorable) of each such individual provisional knowledge link over the following hundred hours or so. Sejnowski and Destexhe (2000) propose that cortical activity during sleep is suited for consolidating information in neural assemblies. Thus, confabulation theory proposes that knowledge evaluation and solidification processes that strengthen knowledge links are largely carried out during sleep and involve entorhinal cortex, hippocampus, and many other brain nuclei. Given that many tens of thousands of provisional knowledge links are often formed daily, it is no wonder that we must sleep a third of the time (see Hecht-Nielsen 2006a, 2006c for more information).
This is compatible with my current theory of dreaming (as I've elaborated elsewhere on the site). I will try to read more about this research (of the authors mentioned above) and hopefully benefit from more data.
Here's a passage about "Confabulation Theory in a nutshell":
Thinking is a phylogenetic outgrowth of movement. Multicellular animals began moving about 580 million years ago. The fitness benefits of purposeful movement rapidly drove the development of nervous systems. Soon, a new design possibility emerged: the elaborate neuronal machinery already developed for controlling movement could be applied to brain tissue itself. In particular, discrete brain structures, modules, emerged that could be controlled exactly like individual muscles. By manipulating these modules in properly coordinated 'movements' (thought processes), information processing (specifically, cognition vision, hearing, planning, reasoning, language, movement and thought process selection and inauguration, etc.) could be carried out – further amplifying animal competitive success and diversity. The purpose of each module is to describe one attribute that an object of the cognitive universe may possess. This description takes the form of activating one of a large number of symbols (each a small collection of specialized neurons) within the module. An individual axonal knowledge link (of which the average human adult possesses billions) unidirectionally connects a source symbol in one module with a second meaningfully co-occurring target symbol in a second module (a la Hebb). A module receiving knowledge links from symbols on other modules can be commanded to 'contract' (confabulate) yielding a single active conclusion symbol; namely, the symbol having the highest level of summed incoming knowledge link excitations. Mathematically, confabulation maximizes cogency (the conclusion selected is that which is most supportive of the truth value of the symbols supplying input excitation); which is the specific generalization of Aristotelian logic upon which all animal cognition is based. Multiconfabulation, in which multiple modules gradually confabulate, while mutually communicating via knowledge links, allows the parallel application of vast numbers of relevant knowledge links. Every time a module reaches a conclusion, action commands axonally connected from that conclusion (these learned connections are termed skill knowledge) are instantly launched. Action commands cause all non-reflexive and non-autonomic behavior (movement processes and thought processes). The extreme power of thought derives mostly from these last three factors: the massive parallel application of knowledge, the interoperability of knowledge links across attributes, and the context-sensitive triggering of new behaviors many times per second.
© 2006, Robert Hecht-Nielsen, All Rights Reserved.
See Also:
Robert Hecht-Nielsen's publications webpage (http://r.ucsd.edu/Publications.htm).
Mind fiction: Why your brain tells tall tales (http://www.newscientist.com/channel/being-human/mg19225720.100-mind-fiction-why-your-brain-tells-tall-tales.html)
-
A layperson's style newspaper article on the idea that sleep serves the purpose of memory consolidation and aids learning (citing the neuroscientists mentioned above). What the cognitive scientists call "memory consolidation" is the same thing as what I was elsewhere describing as self-organization of the brain-psyche's complex system with the goal of improving system efficiency. An interesting implication for depth psychologists who feel that dream phenomena is meaningful (which of course neuroscientists won't even address . . . yet!) is that not only the physiological process of dreaming is functional, but the stuff of dreams (in that this stuff represents the organization of memory or information with which we live and that we are) is also likely to have some kind of meaning. That is, if sleeping state memory consolidation aids learning, chances increase that consolidation or reorganization of memory as perceivable phenomena would bear the stamp of learning, as well.
That doesn't solve the problem of how to understand or interpret dream phenomena, but it puts a check in the column labeled "Dreams Are Meaningful". I've been wondering if the tendency of many (but by no means all) dreams to end with either a resolution to a problem or at least a leg up or better insight into how to face that problem might correlate with a process of more efficient organization or "memory consolidation". As I've written before, I think that effective dream work can help reinforce such consolidation and increased cognitive system efficiency . . . as psychotherapies (albeit imperfectly) have long been providing anecdotal evidence for. Anecdotal, but quite substantial. It it isn't of course, only depth therapists who have happened upon the notion that dreams can help us learn how to live more adaptively. Humans have harbored this as an intuitive folk belief probably as long as we've existed.
The Dallas Morning News
August 28, 2001, Tuesday
Maybe brain needs some Zs to help it learn its ABCs
BYLINE: By Tom Siegfried
Birds do it. Bees do it. Humans can't seem to get enough of it.
It must be good for something. But nobody really understands for sure why sleep is so important.
Sleep isn't simple. From the outside, the body seems to relegate consciousness to an offstage role while cycling through several stages of slumber. But inside the brain, nerve cells continue to fire electrical signals, usually slowly, with occasional brief interruptions for rapid-fire messaging.
Sometimes the rapid signaling takes over. The eyeballs dart around, and a perceptual mishmash of apparent sensations tells the brain a story or scares it half to death. "During sleep," write neuroscientists Terrence Sejnowski and Alain Destexhe, "the internal activity of the brain has a richness that defies explanation."
Still, sleep researchers keep trying to explain it. Why should animals as ancient as insects, as modern as humans, and all evolutionary intermediates spend so much time doing something so dangerous? Sleep must have some purpose that serves the goal of survival. Otherwise, the awake would have inherited the Earth by now.
Experts often argue about why sleep is necessary. Recent studies have begun to build a case, though, that sleeping plays an important part in learning and memory.
Experiments in rats, mice and people all suggest that efficient learning of new skills requires sleep. Nerve-cell activity during rapid-eye movement, or REM, sleep mimics what those nerves do while practicing a task, brain scans reveal. For humans learning visual patterns, non-REM sleep early in the night and REM later both seem needed to remember the patterns. Computer simulations show how cycles of electrical activity during sleep might reflect processes that store memories. And analyzing brain chemistry during sleep and during learning shows several suspicious coincidences.
"There is growing evidence that memory consolidation occurs during sleep," Sejnowski, of the Salk Institute in California, and Destexhe, of the French research agency CNRS, wrote last year in the journal Brain Research.
Training rats and mice to learn a new task, for example, increases the amount of REM sleep they engage in at specific times after the training. Rats swimming in a tank of murky water can learn to remember where a submerged platform is in relation to visible signposts. But the rats can't find the platform very quickly if you deprive them of REM sleep for four hours after the training session. (Other rats, deprived of just as much sleep but at a later time, find the platform faster than the rats that are REM-deprived.) Apparently REM sleep at just the right time helps rats store a memory of the platform's location.
Making such memories permanent, other studies show, requires chemical signals that "turn on" certain genes. The active genes produce proteins that reshape synapses, the junctions where nerve cells communicate.
When scientists block the chemical reactions that turn on the genes, or stop protein synthesis directly, the rats fail to remember what they're supposed to _ but only if the blocking is done during the time when REM sleep would have been consolidating memory.
Further chemical evidence for the sleep-learning link comes from studies of a brain region known as the hippocampus, which plays an essential role in storing long-term memories. During REM, amounts of the messenger chemical acetylcholine go up in the hippocampus. Levels of another chemical, serotonin, go down. Mice genetically engineered to respond less to serotonin engage in more REM sleep, one study showed; another study found that such mice had better memory for new tasks. The chemistry of sleep and the chemistry of learning seem to be tightly intertwined.
But why should sleep be important for memory? Sejnowski and Destexhe's computer simulations suggest that early stages of sleep prepare nerve-cell circuits for learning. Then key parts of the brain juggle their jobs to switch from the bustle of daily life to the maintenance work of nighttime.
During the day, the thalamus _ sort of the brain's central switchboard _ relays input from the senses to the brain's wrinkled outer layer, or cortex. The cortex, where thinking goes on, tries to make sense of the messages and decides what to do about them. But at night, the cortex loses interest in the outside world, sending signals to the thalamus, which feeds the cortex's own messages back to it.
"In a sense, during sleep, the cortex no longer listens to the outside world, but rather to itself," the scientists write.
Meanwhile, the hippocampus gets into the act. In REM sleep, the cortex sends messages to the hippocampus, similar to the signals it sent there during the day. In slow-wave sleep, the hippocampus fires signals back to the cortex _ perhaps a distillation of the day's sensory impressions.
This back-and-forth signaling with the hippocampus at night can reshape synapses in the cortex, solidifying memories in the nerve circuits that had been active during daytime training.
While you're awake, too many nerve cells are activated at the same time to make precise long-term memories. You wouldn't want to redo the architecture of the brain while it's busy any more than you'd reconfigure your hard drive while answering e-mail.
"Implementing ... network reorganization must necessarily take time and be performed during a state in which normal processing _ such as sensory processing _ should not occur," write Sejnowski and Destexhe. "This may ultimately be the primary reason why we need to sleep."
Of course, this picture is drawn from a mix of physiology, computer simulation and speculation. Even if learning is an important reason for sleeping, it may turn out not to be the only one. When it comes to sleep, science still has a lot of learning to do.
___
(Tom Siegfried is science editor for The Dallas Morning News. Write to him at: The Dallas Morning News, Communications Center, Dallas, TX, 75265.)
___
(c) 2001, The Dallas Morning News.
-
See Also: Why Do We Sleep? (http://www.nbb.cornell.edu/neurobio/linster/BioNB420/pdfs/sejnowski_destexhe_2000.pdf) [.pdf]