Don’t Think Too Hard.
I find it interesting that when we’re asleep—supposedly unconscious—we’re frequently fully conscious, mired in a nonsensical dreamworld of our own creation. There’s currently no universally accepted theory for the purpose of dreams—they range from cleaning up mental detritus to subconscious problem solving to cognitive accidents. On the other hand, we DO know plenty about what goes on in the brain during the dream state.
Studies show that in dreams, our thought processes are largely the same as they ones we use when we’re awake. The main difference seems to be that we don’t notice the insane world that we’re a part of. We reason perfectly normally based on our surroundings, we’re just incapable of reasoning about those surroundings—we lack metacognition when we’re dreaming. The culprit behind this is a brain area known as the dorsolateral prefrontal cortex (DLPFC). It’s responsible for, among other things, executive function (directing other brain functions), as well as working memory and motor planning. This combined with the fact that it’s the last brain area to develop (meaning it was the last brain area to evolve) suggests that it’s key in creating conscious, directed thought. And during sleep, it’s shut down, cutting off our ability to question the premises we’re given. So, barring entering a lucid dream state, we lack the mental hardware to recognize we’re in a hallucination when we dream—it seems perfectly normal.[1]
While we’re dreaming, a number of other neurological events are taking place. Long term memories are being accessed and replayed. Brain regions that are normally unconnected work in concert to unite disparate bits of information. Whatever the purpose of dreams is (if they indeed have a purpose at all), they appear to be a window to the strengthening of mental connections and creation of new ones that takes place while we’re asleep. And this behavior is necessary to maintain high-level mental functioning. People do better on tests after getting REM sleep, and people who are REM sleep deprived show extremely impaired memory and learning abilities.
Something similar occurs when our mind is wandering—unrelated brain areas are working together to develop new mental pathways (which is why talking about the importance of daydreaming is currently all the rage). The same thing happens when consuming alcohol—daydreaming and mental connection formation increases, and frontal cortex activity (and metacognition) decreases.
The implication here is that the creation of new cognitive pathways is something that takes place in the absence of conscious, directed thought. It passes the plausibility test—we have a limited amount of cognitive resources, so focusing our thoughts leaves fewer mental resources left over for other tasks. And the formation of new mental connections is extremely important—it’s essentially enlarging the search space our minds have access to when trying to solve a problem. Though it’s not under conscious control, it’s still a high-level function—young children and people with autism seem to have extremely muted dreams (if they dream at all), implying fewer mental connections are being formed. Our executive function is great at orchestrating different brain areas to find a solution to a problem, but it’s only able to look through the space of possible solutions that’s already been created.[2]
Conscious, directed thought—amazing as it is—is not the end-all, be-all of mental function. Humans are at the top of the intellectual food chain (not to mention the actual food chain), but a number of the things that make us ‘special’ are things we share with other animals. Plenty of them can pass the mirror test. Plenty have language, use tools and have complex social structures. Physiologically, what sets us apart is our processing power—the sheer volume of cognitive pathways we can create and sort through. Half of solving a problem is having a search space that contains the answer, and the human mind can create an ENORMOUS search space. But it does so without directed thought.
If a problem seems intractable, then, you may not be able to make headway by THINKING about it harder. That infamous burst of insight seldom seems to come while hunched over a desk or after that 10th straight hour in the lab—it comes “in a moment of distraction or else burst forth from the subconscious while we sleep” [3]. Though it’s obviously important to put the hours in to understand your subject (the brain can only work with what it’s given), a creative or insightful solution arises from those pseudo-random mental firings that are beyond our conscious control. The answer to a hard problem might be a mental path that your brain hasn’t formed yet, making trying to think your way through to it a fruitless endeavor. At a certain point, it’s important to step back, relax, and let your subconscious create more grist for the mill
[1] As an aside, the fact that the brain region responsible for working memory is shut down may be the reason why we generally don’t remember our dreams, and why writing them down and trying to remember them is an important step in learning to enter a lucid dream state (when the DLPFC is thought to be activated).
[2] Mind-wandering actually seems to be MOST effective when we’re at least partially aware we’re doing it—if you’re not paying any attention at all, something important could easily slip right past you. Something similar probably takes place during the lucid dream state, where we’re aware enough to direct the flow of the dream but not so aware that we wake ourselves from it (a frequent problem of beginning lucid dreamers).
[3] Though I suspect there is a selection bias at work here.
Doesn’t follow. First, that intense thinking primes the brain with the info it needs to correlate.
Second, I’ve often had insights on walks immediately following a working day.
Third, I’ve not uncommonly had insights during the actual working day as a result of hunching over and thinking really hard.
I wonder if your ability to consistently produce insight by thinking really hard is somewhat atypical. And I’m curious, how “verbal” is that process for you? The bulk of my surprise insights seem to come from face-to-face discussion forcing me to verbalize my thoughts on the spot.
Agree.
Exactly what I would expect. I like to mix and match. I benefit from concentrated work and then I have my biggest breakthroughs in the shower or while running.
That you have to be in the shower or something to have an idea is probably indication that in the lab you are doing it wrong. Maybe there is a good manual on how to think productively? It’d be a right topic for a post.
I think this might be about right. I rarely have interesting insights while doing other things, but there was this one time where I was trying to fall asleep and suddenly figured out what the bug was in the code I had been working on a couple hours before. I had it down to the exact line that was wrong. But that’s very much the exception for me. I’d be interested in such a post.
I’ve spent some time studying research on how to think productively and I can say with some confidence that I am in fact doing it all wrong. Of course, I also have enough knowledge to be a master seducer and know all the techniques of a martial artist far beyond my meagre belt. Alas...
Some attempts at writing a good manual describe ways to promote the kind of relaxed dreamlike thinking described here and how to exploit it to best effect.
I really need to fight that instinct to make the grand sweeping conclusion—it’s always getting me into trouble.
Agree, that’s what I meant by ‘putting in the hours to understand the subject’
My best guess is that not all insight requires new mental connections—focused, analytical thinking may be capable of raising the availability of mental pathways that are already formed. This is dangerously close to equally explaining all outcomes, but I’m not proposing this as a fully general theory of problem-solving.
What evidence do we have for it? There seems to be gradual deterioration of language learning facilities, and at very old age mental facilities are generally diminished, but other than that, is it true than children learn significantly faster than non-senile adults?
You know, I’m not even sure about that. It takes at least 6-8 years for most children to gain something approaching adult fluency in language. For most adults putting continuous substantial effort into learning a second language (for instance in an immersive environment), it only takes around 2 years to gain similar fluency. It may be more difficult for many adults to learn to hear and pronounce foreign phonemes thoroughly, but many adults don’t have problems with that, and almost all can improve greatly with a bit of speech therapy.
I’ve certainly seen significant evidence that adults with comparable motivation to learning a language actually learn quicker than a child does. While this motivation is rather rare I’ve been convinced that the critical period idea is a myth.
I think there’s high variation in adults’ ability, low in children’s ability.
A million or more Jews came to Israel from the USSR in 1990-2 (my family among them). Virtually none of them knew any Hebrew (studying or teaching it had been illegal in the USSR), or any other language spoken in Israel (English, Arabic). The immigrants had to learn Hebrew to be able to get jobs, attend school/university/retraining, and interact with local authorities. Local Russian-speaking communities did not yet exist on the requisite scale to take in more than a few newcomers. There was enormous pressure, and all the motivation you could want.
My family was among these immigrants (I was 6 years old at the time), and I’ve met hundreds or thousands of others. There is a very strong negative correlation of age and fluency in Hebrew, even today, after 18 years of immersion. People who immigrated at age 10-15 or younger are fluent speakers, have little to no accent, and it is very remarkable to meet an exception. Among people who immigrated as adults, most are fluent but make grammatical, pronunciation or spelling mistakes, or have vocabulary problems. Among people who immigrated at age ~~ 50-60 or older, many (I estimate at least 20-30%) never progressed beyond rudimentary skills necessary for e.g. shopping, cannot read or write fluently.
There is of course strong correlation between environment post immigration (school, work, non-Russian-speaking workplaces) and ability, but I’ve met enough such people to know that the causation goes both ways.
I suggest that causation operates via the very real change impact on motivation and the willingness to change one’s identity. 18 years of dedicated practice is enough time for nearly anyone to become an expert in nearly anything to a far greater degree than mere competence.
Studies find that adults actually learn languages faster than children do and older children faster than younger. Here is a more subtle investigation of how the motivation of adults interacts with circumstance and language in natural circumstances.
Possible confounding factors:
old people have higher discount rates (‘if I’m going to die next year, why bother?’)
less time to benefit from learning period (years, as opposed to many decades like a kid)
old people have less gain (both from #1 & #2 and because they have support structures which speak their native tongue—their relatives and friends)
old people are just stubborn and dislike the new country and its language (did the old people want to leave? Unlikely).
There’s motivation and then there’s motivation. Children are too young to respond to the requirement incorrectly enough to completely fail at it. If your System-2 must learn a new language, but you can still System-1 speak your native tongue within family or with friends, it is a different situation from where a child has no way of communicating at all, and communication is not just an instrumental requirement, it is a powerful instinct. Of course, some older adults will have damaged learning ability and won’t be able to learn a new language, but it’s an entirely different issue.
You’re saying all adults could learn a new language and become fluent if the motivation was on the level of “not being able to communicate with anyone at all”. Is there hard evidence of this? In particular, once such people achieve basic communication, do they reliably move on to fluency?
But have you seen evidence that all such adults succeed in learning the language, just as all children successfully learn their first one?
They don’t. The social and personal influences to encourage such efforts are not all that common. Most adults don’t master most areas of expertise.
I assert that most people can master almost anything if they spend 4 hours every day in dedicated practice for 10 years (ref). I do doubt that most people will perform such a remarkable feat (papers in the same reference traced much of the heritability of expert talent to inherited motivational factors.)
Still in the end most children will learn their first language faster than most adults will learn a second.
There’s fairly decent evidence for it.
Did you refer to a child learning their first language? That’s an unfair comparison. With no language to think in, a child is much less intelligent and capable than any adult. It’s not surprising that adults can use verbal reasoning/thinking skills, and the similarities with the languages they already know, to learn a new language much faster than a child would.
The comparison should be made between adults and children fluent in one language (so at least 6-8 years old) learning a second one. In this situation I expect high variation among adults, some being much better than the typical child, and some much worse.
Well, it might be unfair in some contexts, but the point was that since children don’t learn their first language faster than adults learn their second, we can’t really draw any conclusions from language learning about the overall speed of learning of children.
If we knew whether older children learn their second language faster than adults, then we might draw conclusions.
We know that. (They don’t. But earlier exposure to a second language does predict that they will continue to higher levels of proficiency.)
That’s very enlightening, thanks. Then the case of Russian immigrants in Israel may indeed be due to insufficient pressure to learn and other factors.
Hrmmm, there doesn’t seem to be any (though google turns up suprisingly little in either direction). So much for ‘common knowledge’.
I’ve removed the reference, thanks for the catch.
Even for language learning, it doesn’t seem that children learn faster. They just have nothing else to do, while adults typically only work on language some portion of the day/week. A full-time effort to learn a language will allow an adult to learn a new language to fluency in a few months, that is much faster than children do. Another matter is anchoring created by previous languages, that results in accent and other linguistic quirks.
I have no doubt this is true for especially talented language learners. I am not one of those people, and most people aren’t either. Language learning ability has a very high variance. 1-2 years seems like a reasonable median.
Have you actually tried this?
If I did, it wouldn’t be a particularly useful piece of info (anecdote). I didn’t, and I read about such programs.
I was conversationally fluent in Spanish after traveling in Spanish-speaking countries for six months, despite studying the grammar for only a week and spending most of my time speaking English. I can only imagine how fluent I’d be if I had actually devoted myself to learning instead of, well, doing what I like to call “stupid things in dangerous places”. (In all fairness, Spanish is pretty easy to learn from an English base, especially if you’ve studied Latin. I imagine Chinese or Swahili would be a lot harder.)
It would be nice to know, eirenicon, whether you had any competency in any other languages besides English before you learn Spanish.
Ah, of course. No, English was my only language at the time. I studied French in grade school but have no more than a few words of it left—that said, the underlying grammar, which is similar to Spanish, probably didn’t just disappear. I also took a couple Latin courses in high school, but never became very proficient and again, only retained a few words and a rough understanding of structure. When I began learning Spanish it was all very new and quite difficult at first. I do think my strategy was a good one, though. The week I spent taking private lessons was devoted to grammar and grammar alone, on my insistence, and it paid off. En mis viajes fue muy útil.
Difference?
I’ve thought on this a bit. I’m not sure directed thought on a hard problem is wasted, even if it’s clear that an intuitive insight is required to solve it. It might be that long, directed thought on a problem primes the rest of the mind to gear its resources to exploring the problem space as well.
When I spend all day working on something intensely, it’s not uncommon for me to have likewise intense dreams about it. Interestingly, the illogical mishmash that forms the dream is more jarring and disorienting than normal, perhaps because my brain is primed to be looking for incorrect ideas, since that’s what I was probably doing all day.
I’ve got mixed opinions on this. On the one hand, I agree that thinking “harder” doesn’t do anything directly useful, even if it maybe primes later useful thoughts. I even agree that creating anything genuinely new is heavily dependent on coincidence, in the sense that accidents can shorten a search in ways that intentional searching cannot.
However, on the other hand, it’s also clear that failure to solve problems or get good ideas is very often a failure to consciously ask the right questions of your subconscious. A lot of times, the thing stopping people is that they simply haven’t bothered to ask their brains for a solution in the first place… or that, having asked, they don’t keep System 2 quiet enough to prevent verbal overshadowing of the search results coming back from System 1. (i.e., the real reason why brainstorming isn’t supposed to include criticism.)
It’s not my experience. I usually find solutions in intuitively comprehended (visual) imagery that is purposefully developed (learned) to model the problem under study. The easiest example is geometry problems, other kinds of problems need translation. Of course, the solution itself isn’t arrived at according to an understood algorithm, but your description makes the process less controllable than it seems to be. I believe the imagery is just another paintbrush handle, just as words are, but drawing with it is more of a craft than mysterious inspiration.
Once in a while I seem to become aware that I’m dreaming. This is usually followed by “deciding” to solve whatever situation I’m dreaming about by waking up.