Are we all natural learners?

How much knowledge would a child acquire if they never set foot in a classroom? And how much instruction from a teacher is too much? The answers to these questions underpin age-old debates around how teachers should teach, but John Morgan finds that the research paints a mixed picture on humans’ innate ability to learn
27th November 2020, 12:00am
Are We All Natural Learners?
John Morgan


Are we all natural learners?

In the Paleolithic era, children really understood the value of learning. Hunter-gatherer youths either learned the way back from the river to their hut or they ended up lost in the forest, meeting a sabre-toothed tiger for a one-sided lunch. Either they worked out how to locate a delicious woolly mammoth or they starved for want of woolly mammoth steaks as adults. Either they picked up enough of the local protolanguage to warn their parents that axe-wielding guys from the other tribe were approaching or the whole family felt the consequences of their failure to engage with self-directed learning.

As such, evolution, you'd think, would have made us into natural learners. But has it? Or has evolution actually only hardwired us to learn certain kinds of knowledge?

The distinction matters for education, and education in the Covid era in particular. The recent period of school closures during lockdown prompted many to comment that children seemed to have an innate capacity to learn on their own terms, away from the direction of teachers.

However, there are theorists who would say that was near impossible for certain areas of knowledge: that we have evolved an innate capacity for self-directed learning only when it comes to fundamental knowledge, like language; that with more recently developed forms of knowledge - like reading and much of what we learn in school - we have no evolved abilities to learn and we must be explicitly taught by direct instruction.

That's a controversial view among some teachers, but it is one that's increasingly gaining traction in schools and in policy. So is it backed by any science and, if it is, what does that mean for pedagogy?

The main proponent of a limit on natural learning is David Geary, a cognitive developmental and evolutionary psychologist who is a curators' professor at the University of Missouri. He defines his work, including his 2005 book The Origin of Mind: evolution of brain, cognition and general intelligence, as "evolutionary educational psychology".

Starting with an academic paper published in 1995, Geary began to form a theory for understanding how natural selection shapes children's ability to learn in school, which centred on a distinction between "biologically primary and biologically secondary" cognitive abilities (see box, below).

He argues that when it comes to "primary knowledge", which includes language, humans have evolved "built-in scaffolding and built-in biases that ensure children get the experiences they need to fully flesh out those abilities". His primary knowledge category also includes skills such as face processing and spatial navigation, and interest in the natural world.

Geary's category of "secondary knowledge" includes reading, writing and arithmetic, which he says are "evolutionarily novel".

"Reading, simple arithmetic - they emerged 5,000 to 10,000 years ago, depending on the culture," he says. He adds that even then, only a small proportion of populations were engaged in those activities. So there is "no strong evolved bias to be able to learn how to read or to be inherently motivated how to read," Geary argues.

Evolution of the species

Based on this work, it's clear that Geary has little time for the notion of children being natural learners in all areas.

"This kind of romantic, Rousseau-like [18th-century philosopher Jean-Jacques Rousseau] view - everybody is motivated to learn, they can learn anything, they are going to learn what they need to learn to do well - it's just nonsense," he says.

"It works fine if you're just living in the forest and you're hanging out with your tribe," Geary continues. But as humans have moved "to an evolutionarily novel context, that philosophy no longer works", he adds. "In my opinion, it has led to a fatal flaw in educational approaches - especially for kids who are struggling."

For years, Geary's work went under the radar in schools, having little direct impact on classrooms. But it was later used to underpin a theory that's now ever-present on the research lists of Ofsted documents, policy documents and teacher CPD: cognitive load theory (CLT). (For example, CLT received a reverential mention in the summary of the research underpinning Ofsted's inspection framework, and its fingerprints were firmly on the Department for Education Early Career Framework.)

CLT was developed by John Sweller, emeritus professor of educational psychology at UNSW Sydney. It holds that limits on our working memory mean that non-essential information must be minimised in teaching if children are to learn effectively, and that relevant information should be organised in the most efficient way possible.

Geary emphasises working memory as the key that gives us the ability to learn in "evolutionarily novel" fields - to pass the information learned from "effortful" problem solving into long-term memory. For Sweller, Geary provided the missing puzzle piece to complete CLT.

"For several decades, we consistently obtained results from randomised controlled trials indicating that explicit instruction in the form of worked examples (ie, teacher-led) was far superior to discovery learning (ie, child-led)," says Sweller.

But in answer to these findings, he and colleagues were "told to look at how people learn outside of school without explicit instruction"; for example, how "we learn to speak our native language without explicit instruction in how to organise our tongue, lips, breath and voice".

"He had no answer to that perfectly reasonable critique," Sweller says.

What Geary's distinction between primary and secondary knowledge did was to provide "the critical jigsaw puzzle piece", explains Sweller. "We learn without explicit instruction outside school because we have evolved to acquire the relevant primary knowledge."

Like Geary, Sweller draws the implication that secondary knowledge can only be taught via explicit instruction: "It's foolish not to use the primary skill of communication to provide learners with the secondary skills for which schools were invented."

It's easy to see why CLT has become so important in the age-old debate between traditionalists and progressives - it appears to provide hard proof that the former is superior to the latter in terms of learning. Yet if CLT depends on Geary to truly answer its critics, how sure are we that Geary is right?

Peter Gray, research professor of psychology at Boston College and author of Free to learn: why unleashing the instinct to play will make our children happier, more self-reliant and better students for life, stresses his respect for Geary, but he sees Geary's theory as problematic.

If, like Geary, you study schoolchildren, "then, of course, direct instruction is going to work better than natural learning because natural learning can't happen in school" or, at least, the majority of schools as they are conceived today, says Gray.

However, the evidence from homeschooling, and from schools like Sudbury Valley in Massachusetts, which allows children to direct their own learning, is that when children "are in a literate, numerate society, they will learn to read and they will learn to use numbers to the degree that you need to use numbers for whatever they are doing in that society", he continues.

Abstract concept

Gray argues that Geary's division between primary and "evolutionarily novel" secondary knowledge just does not hold. Hunter-gatherers "were engaged in abstract thinking probably as much as any of us are", he says.

In a chapter published in Geary's book, Evolutionary perspectives on child development and education, Gray wrote: "Mathematics and science involve a variety of thinking that may not be fundamentally different from the thinking that hunter-gatherers engaged in regularly as they made and tested hypotheses about the movements of animals based on scant tracks, or the locations of tubers hidden underground during the dry season."

In that same chapter, Gray took issue with the idea that the brain is composed of "modules", which "came about to serve specific survival purposes", suggesting instead that our "innate learning mechanisms" may be "far more adaptable and flexible than" some evolutionary psychologists propose.

But another critique of Geary's ideas comes from George Ellis - emeritus distinguished professor of complex systems in the Department of Mathematics and Applied Mathematics at the University of Cape Town.

"There can be no innate modules of the kind he proposes for developmental, genetic and evolutionary reasons," says Ellis, co-author of Beyond Evolutionary Psychology: how and why neuropsychological modules arise. "We are natural learners who are hardwired to explore and seek for meaning. Both talking and reading are cultural inventions; it's just that one came later than the other. Talking and reading can be learned the same way by a process of successive approximation through guided trial and error in meaningful contexts."

Other academics are a little more accommodating of some of Geary's findings, though. Roberta Golinkoff, Unidel H Rodney Sharp chair in the School of Education at the University of Delaware and co-author of Becoming Brilliant: what science tells us about raising successful children, observes: "Everybody learns to walk, everybody learns to talk, everybody learns how to navigate in the world. But things like learning to play piano, learning algebra - all that stuff, we are not programmed for that. I would agree those are secondary."

However, Golinkoff - the founder and director of the Child's Play, Learning and Development Lab, whose research has looked at language development and early spatial knowledge - adds: "All children - whether they are raised in a tent, in a high-rise, in a hut, in a cave, I don't care - they are all programmed to learn. That is part of the adaptation that humans make to their environment."

Indeed, Fei Xu, professor of psychology at the University of California, Berkeley, whose lab focuses on knowledge acquisition, says: "It is clear that research from the last few decades on learning and cognitive development tells us that children are natural learners."

So, children can learn everything on their own? That's not quite what they are saying. Rather, Xu says that Geary's primary/secondary knowledge theory is a "useful distinction". Though children "can learn a great deal on their own", at the same time, "domains such as reading and writing appear to be quite difficult for many children and direct instructions may be necessary; similarly, symbolic mathematics is also hard for many children and good teaching appears to be necessary and important", she adds.

So, where does all of that leave teachers, then? The views of the academics broadly follow their views on the nature of evolutionary knowledge.

For Geary, the majority of children will "need much more structure, much more practice, much more teacher-directed things than most of these kids are getting" at present in schools. Learning secondary knowledge requires "focused effort, organised curricula and, in many cases, direct instruction, rather than a free-for-all, student-centred instruction," he argues.

The "in-built reward systems" that apply when developing language and social skills are just not present when we "learn quadratic equations and algebra", when "you have to engage these effortful, attention-demanding, often boring systems to develop these competencies," he adds.

However, Ellis says that Geary "does not appreciate the active exploratory role of children in learning" and "does not take into account the emotional systems that underlie our enthusiasm to learn, or not".

Golinkoff is, again, a little more accommodating of Geary but still has disagreements: she says that guidance is needed, but that it does not have to be direct instruction. "That's where I disagree with my friend Dave…I am a champion of playful learning," she says.

Golinkoff advocates "guided play" (or "engaged learning" for older children), where "the child ends up asking the questions and almost requesting the information you are trying to get across".

She was co-author on a study that sought to evaluate the results of guided play, free play and "didactic instruction" in teaching four- and five-year-olds. Each of the groups was introduced to geometric principles of shape via one of the three pedagogical approaches before all the children were asked to look at, for example, typical, atypical and "non-valid" triangle shapes, then asked to sort them into "real" and "fake" triangles.

The free-play group ended up comfortably behind. But the guided-play group - where the teacher asked for the children's help in discovering "the secret of the shapes" and they all put on "make-believe detective hats" - was found to outperform the direct instruction group. The direct instruction group "appeared to learn that counting [sides] was important, but not why it mattered for determining the shapes' properties", the research paper said.

"You need guidance to learn these secondary things," says Golinkoff. "But the big difference between the guided play and the direct instruction group was the guided-play group transferred their knowledge to shapes they hadn't seen at a rate 30 per cent greater than those who had been told."

Xu also takes a midway view: "Since children are both natural learners and active learners," perhaps "a mixture of direct instruction and free choice and exploration would work best", she suggests.

The lockdown effect

Of course, the earlier lockdown may give us a real-world insight into these arguments: at least at the start of the lockdown, direct instruction was minimal and self-directed learning more common. What might that period show?

"We'll find out where that's going to put us when the assessments start back up," says Geary. "If advocates of child-centred learning are correct, and kids are motivated to learn, have their own agency, can pick it up as long as they are getting the stuff - in this case online - then there should be no drop-off. We should be A-OK: no increase in variance [in achievement between different social groups], everything is good, and I'm wrong."

But if "secondary skills really do require what I say they require, then we should see drops in competencies, or not much development," he adds. So, "if what I said has any truth to it, then we have a problem".

But Gray sees Geary's argument on lockdown as an extension of the "myth of the summer slide" - the idea that pupils lose academic ground over summer holidays - highlighting studies that actually show "that when children are out of school doing real things in the real world, if they are in a literate and numerate environment, they are becoming better at reading and becoming better at real use of numbers".

When it comes to pupils forgetting things that they previously learned over the holidays or during lockdown, Gray takes that as evidence of the failures of direct instruction - that we forget things drilled into us by rote without engaging our understanding. "This is why almost everybody forgets most of the math they learned [at school] a year or two after they graduate," he adds.

But for Golinkoff, the focus on lockdown is probably not going to tell us much. She cautions against getting too hung up on what is innate and what is not innate when it comes to learning. "There is never, never a time when nature operates independently of nurture," she says.

Golinkoff's view is that it's "natural for children to learn anything that is presented in a way that they can assimilate it". And that goes for learning in general - lockdown or not, she says.

"There is absolutely no reason why school has to be a slog," she adds. "I love Dave Geary - but there's no reason why teachers have to talk at kids and lecture the kids."

So, is the learning of a child who's studying the definition of a subordinate clause online during lockdown connected to, or totally different from, the learning of the Paleolithic child who avoided walking into the sabre-toothed tiger's claws?

I am afraid there is no definitive answer that we can come to: when it comes to how evolution has shaped children's capacities for learning today, there are different tribes of education researcher, with conflicting theories and interpretations. The best that can be said is that teachers should view both this research and that connected to it - like CLT - within this context.

And be happy that the battle between these tribes is with words, not axes.

John Morgan is a freelance journalist

This article originally appeared in the 27 November 2020 issue under the headline "Are we natural learners?"

You need a Tes subscription to read this article

Subscribe now to read this article and get other subscriber-only content

  • Unlimited access to all Tes magazine content
  • Exclusive articles and email newletters

Already registered? Log in

You need a subscription to read this article

Subscribe now to read this article and get other subscriber-only content, including:

  • Unlimited access to all Tes magazine content
  • Exclusive articles and email newsletters