Learning as More Than Just Neurons Connecting
By: Curtis Kelly
Let’s look at memory and learning in terms of neuroscience. But to just say learning is neurons connecting via synapses, while true, is like saying the movie Barbie is pixels changing color. I would like to take you beyond simple memory formation into the complex and remarkable aspects of brain development. From infancy to adulthood, it happens in short spurts, in different areas, at different times.
The critical period for language acquisition
It has long been observed that children learn languages faster than adults. Children just seem to absorb new language naturally, while adults have to work at it. This led Eric Lenneberg to propose the Critical Period Hypothesis (source), sometimes also referred to as the Critical Age Hypothesis, a period starting at birth and ending around puberty. At about 12 years old, something odd happens to the brain that makes it less adept at just picking up new language, especially in terms of hearing foreign phonetic systems and pronunciation.
Lenneberg’s theory has been widely accepted in our field, inspiring numerous criticisms and refinements. When does the critical period really end, at puberty or in the late teens (source)? Is there no critical period for vocabulary (source)? Is it linked to lateralization of hemispheres (which we no longer believe, source). And are there different critical periods for different parts of acquisition, such as sound and grammar?
Singleton and others have also pointed out that adults, well past the critical age, are more adept at language learning if it is done through study, even if they are barred from mastering a foreign accent. They are also far less likely to suffer from the “use it or lose it” of children.
This theory is big. It is important. It means we can’t teach language to children and adults the same way. But one thing about it amuses me: how language teaching professionals often refer to the “critical period” in the singular, as if only one exists. Most are surprised, as I was, to learn from neuroscience that there are numerous critical periods for different functions, I suppose hundreds, that occur all throughout our lives, especially in the first twenty years. It is all about plasticity, one of the great discoveries of neuroscience in 1890, that really gained weight from 1960-1990[1].
Neuroplasticity
Neuroplasticity refers to the brain’s ability to reorganize and rewire its neural connections, and that is the basis of all learning. Not all areas of the brain have the same level of plasticity, nor is plasticity constant in the areas that have it. One part of the brain might be very plastic for one time period, and then harden in another (Buzsáki, 2020). Thus, we have time-limited critical periods and less rigid sensitive periods, when a part of the brain temporarily becomes super plastic. Greater plasticity, of course, means a greater tendency to change, which is why children with high plasticity in their language areas learn (and lose) language skills so quickly.
How, and whether, change occurs is highly dependent on environmental influences. The world shapes us in these critical periods, and if you think about it, that is why teachers are so important. We give learners the necessary experiences to feed the changes, the growth. Of course, parents, siblings, peers, play, and just rambling through the world do this as well.
Two examples of critical periods
Let’s look at a couple of examples. One of the earliest studies that shows how critical period work was conducted by Hubel and Weisel way back in the fifties. During their critical period for vision, they raised two groups of kittens, one group exposed to only horizontal lines and the other exposed to only vertical lines, even in the stripes in the caretakers’ clothing. The results of this experiment were staggering.
Each group of kittens became literally blind to lines running the other way. For example, cats from the horizontal lines group could see a chair seat, and jump up onto it for a nap, but kept banging into the chair legs (source). Not as staggering was that these two researchers got a Nobel Prize for this discovery. In short, Hubel and Weisel, exposed these kittens to a certain visual environment during a critical period of their ocular development. Visual input that fired neurons that processed lines going one way led to their being strengthened, while the unfired neurons, for lines going the other way, died off, never to return.
That helps explain how a critical period works. In a critical period, the brain reduces a brain “fertilizer,” brain-derived neurotrophic factor (BDNF), causing a neuronal famine. Much the same way the early Soviets did it with peasants, those neurons that work get fed, while the others starve. Critical periods are not permanent, and multiple periods exist for different kinds of development. Keep that in mind the next time you see:
- a child believing that if a line of coins is spread out, it has more coins in it than before (watch this)
- a child suddenly being able to read at age 5
- an adolescent brain developing super “social sensitivity”
- a person like Curtis Kelly failing in geometry in grade 9 and excelling in the same course a year later (story here)
- your husband finally starting to listen at age 40. (speculative, might be fiction)
(Okay, as Mirela Ramacciotti pointed out, the last two might not be related to critical periods, rather just normal development through connecting, but I could not resist.)
This means, of course, that it is completely wrong to think of the brain like we once did…as I once did…as being hard-wired. Although we are born with almost all our neurons formed, only 20% of the connections between them are in place. Then, in infancy we are forming over a million new neural connections every single second! (source) The brain is also pruning connections, especially after we hit a connections peak at two or so (source) where we lose about half of them. This is not a loss in intelligence; it is fine tuning for greater efficiency. Interestingly though, it is now theorized that schizophrenia might be caused by over-pruning, and autism by under-pruning–the latter explaining autistic sensitivity to noise and light (source).
We can see how pruning works in the next example of a critical period, my all-time favorite, illustrated by Patricial Kuhl’s research on how babies learn phonetic systems. Babies are born as she puts it, as “citizens of the world,” with the ability–meaning neural connections–to hear any phoneme in any language (Kuhl, 2004). However, in a critical period between the age of 8 and 12 months, the neural connections for sounds not reinforced by caregiver utterances are lost.
For example, Japanese generally can’t distinguish between L and R sounds. That is because, as this illustration shows, English speaking moms give lots of phonetic input for separate R and L sounds, while Japanese moms just give the Japanese R-L-D sound combo. As a result, brain scans show Japanese babies can hear R-L differences after birth, being “citizens of the world,” but they lose that ability between 8 and 12 months, a critical period for sound specialization, and instead become mother tongue specialists. Even the sound of their babbling changes (source).
In other words, our babies are Weisel and Hubel’s kittens. During that critical period, neurons for mother tongue sounds are activated and they survive and become even stronger. The rest disappear or are reassigned, and are very, very, very hard to get back.
And there is more. Like Japanese with R-L sounds, most Americans cannot hear Chinese shi-shu differences. But when Pat Kuhl put American babies in front of Chinese speakers, for a measly twelve hours of input, they kept that ability, as this chart shows. The American babies were just as good at hearing the difference as Taiwanese babies were. Keep in mind that babies of 8-10 months can’t understand any language yet, so they are just sampling sounds.
Even more fascinating is that American babies shown videos of those same Chinese speakers showed zero retention of the ability to hear that sound. Zero! The babies would crawl right up to the TV, showing interest, but that form of input had no effect. Take that Baby Einstein!
Interesting. Zero effect from video. That makes me wonder if that might be partially true for older learners as well. Is there something in our brains that makes us more responsive to true face-to-face encounters, or is that just true with babies? If there is, does that mean that online education might not be as effective, at least in some ways? I’d like to know.
But getting back to the topic, as we can see, there are many critical and sensitive periods when the brain becomes responsive to certain environmental inputs, whether sensory or motor. Here is a list of a few critical periods, all of which have lifelong consequences:
In infancy, 0 – 2.5 years: visual, hearing, language, touch processing, motor skills (movement and coordination), attachment to caregivers
In early years, 2.5 – 6 years: executive functions, stress response system (without early soothing, any stress can be disabling later), social skills acquisition, empathy, identifying threats and safety, conservation of mass
In childhood, 5 – 9 years: emotional regulation, improved stress response, symbolic proficiency (about the fourth grade), mathematical operations.
In adolescence, 10 – 19 years: frontal lobe development: social sensitivity, abstract reasoning
Connecting and pruning during critical periods is the main form of long-term brain development in the early years. That kind of change happens up through the 20s, but two other kinds of change start taking over in adolescents and adults, myelination and massive neural redeployment.
Myelination
This one is simple. Axons that fire a lot become wrapped in insulating myelin, the proverbial white matter, and that increases their transmission speed up to 100 times. That’s why intelligence peaks at middle adulthood, between 25 and 60. Certain processing areas change from dial-in modem speeds to broadband. It’s part of our beloved automaticity too.
Obviously, repeated practice, firing those neural combos again and again (but in spaced intervals) aids myelination.
Neural redeployment
This one is not quite as simple. To understand this kind of brain development, we need to understand two things the brain is truly good at: 1) focusing all its resources on solving a single problem, and 2) assembling ad hoc neural teams to do so. While focusing all its parts to work on a problem, the brain will call up neural groups serving other functions and “reuse” them to attack this new problem. For example, the bit of brain that keeps track of our fingers, since it is good at organizing, is also called upon for math tasks, like figuring out which number is larger (source).
Since the brain can identify which circuits are good at working on a problem, it puts them together into a temporary team. It might not end up being the same team for the same problem every time, but when a team works together often enough, the rule of “neurons that fire together, wire together” kicks in and their connections get stronger. In that way, very basic brain bits with simple functions wire together to form more complex processing areas.
Let me throw in that these complex processing areas later link up with others giving us even higher-level skills, and so on. That is why children need to be able to add before they can multiply.
When you look at an everyday object, say a fork, the basic neurons for processing horizontal lines, shiny surfaces, curves, smoothness, etc. are activated. That particular combination is what makes up our neural representation of a fork, which stimulates more complex circuits for texture, metallic taste, fork affordances, foods to stab, plate sides, or maybe that time your sister jabbed you with her fork. And that’s how we identify (and define) forks: as simple neural circuits, mainly sensory at the bottom, connected as a fork representation. Part of that representation includes the word “fork.” Voila! We do not have internal dictionaries to define words, just particular sensory and motor routines tied together because they have previously fired together so many times.
Lakoff (2014) shows how this kind of neural connecting explains how metaphors work–another cool trick of the brain. For example, why do we see any increase of just about anything associated with an upwards direction, or decrease downwards: “crime rates are rising,” “TESOL conference attendance is a dive,” “insurance claims have been bouncing up and down all year?” This happens because when we were babies, we saw someone pouring juice in a glass and the basic brain bits for processing amount (increase in this case) and direction (going up) fired together. Then decrease and downwards fired together when the juice got drunk. These areas, so often firing simultaneously, wired together. And so, even in adulthood–and this is universal in all cultures–we associate a change of amount with a direction. Has your understanding of this metaphor gone up?
This connecting of simpler functions to make more and more complex ones is also what lets us understand abstract ideas. It seems that all abstractions are ultimately based on simple, old physical experiences. This is also the process that allows children to do certain mathematical operations at later ages, like multiplication (ages 6-10), algebra (ages 11-13), and geometry and calculus (ages 14-18). The brain must develop certain structures–teams of simpler functions–before that can happen. The same is true for reading, as this illustration shows. All the indicated brain areas must be developed and connected before a child can read, almost always at age 5. Neural reuse gives us unbelievable abilities like calculating math problems and reading.
I suspect offering learners desirable difficulties aids neural team building and solidification (human as well). A desirable difficulty is a learning task that requires considerable effort, leading to deeper learning, but is desirable at the same time. The task must be a challenge that learners wish to achieve and it must be achievable, such as solving a mystery. And of course, most computer games utilize this concept.
What does this mean for teaching?
Once you have an understanding of how the brain is jumping from one development to another all the way up to middle adulthood, you can finally free yourself from the one painful fallacy that has dominated education, and really, all society: that youth who perform well are smart and those who don’t are stupid. More than raw intelligence…if there really is such a thing…the ability to do something at a higher level is usually, especially with children, just an indicator of whether or not the underlying brain structures have yet developed or not.
In fact, the smart vs dumb performance notion might just be the opposite. Kids who “don’t get it” are probably developing more slowly, which is actually a sign of a higher intelligence down the road. According to Giedd, those slower kids often end up with highest IQs at age 19 (related article). That is not always the case, genetics and brain structure plays a role as well, but it bothers me that we automatically treat such slow developers as inferior, something I like to call “better brain punishment.” Sufferers of it include Newton, Darwin, Edison, Einstein, Churchill, Rowling, Spielberg, Jobs, many of you, and, well, I could go on for pages.
Instead of slower, different development occurs as well, as with autistics and ADHDers, for whom our normal way of teaching does not fit.
Let’s keep in mind that children who struggle with math, language, or whatever might just be encountering those educational topics too early, while those who flourish in them might merely be in the middle of a developmental stage that needs to be fed by that kind of input. It is timing that counts, not intelligence. Keep an eye out for signs a child is experiencing a period of brain change and fuel that change.
For older learners, let us take advantage of myelination by reinforcing learning targets through spaced repetition: homework, review, quizzes, and tests. That is why mastery-oriented educational systems, like those in Finland, are so much more successful than mere delivery-oriented systems, where the deluge of new information is usually counterproductive. Think of the brain as a muscle, where repeated practice, especially retrieval practice (such as self-quizzing), and the deep processing of desirable difficulties build those muscles up.
I hope this article helps you do a workout on your mental muscles as well, the ones you use to implement brain-compatible learning in your classroom.
[1] Come to think of it, that’s when I gained weight too.
Curtis Kelly (EdD) is an author, speaker, and founder of the BRAIN SIG and this magazine. He is fascinated with how the brain changes throughout our lives or even in a single day, such as the way memories reorganize to become more useful.