Are we intelligent or are we educable?
A brief Q&A on emerging theories as to what makes human thinking unique
Is the human capacity for being educable a cognitive gadget?
Wait, don’t click over TikTok! That might be the most wonky question I’ve ever typed, but it may also be important to understanding what makes human thinking unique and special, especially as compared to artificial intelligence. And I’m going to try to explain why via FAQ format, let’s see how this goes.
What’s a cognitive gadget?
Cognitive gadget is the term coined by Dr. Cecilia Heyes of Oxford to describe mechanisms of the human mind that arise primarily from cultural evolution, rather than through biological inheritance. Under this view, our cognitive architecture – the methods we use to think about the world – arise from specific cultural practices that evolve over time and are passed on from generation to generation.
Can you give me some examples?
Yes I can. One example is something cognitive scientists call “mindreading” or “theory of mind,” our ability to ascribe mental states to ourselves and others. This ability, Heyes argues, is taught to us by members of our social groups, and varies across cultures. Another example is learning by imitation. It was long thought that babies imitate behavior as a result of genetic programming, but there’s mounting evidence that it’s specific cultural experiences – such as playing with mirrors – that fosters this capability. And then there’s language itself, a cultural development that allows us to communicate abstract ideas and complex ideas to one another. (For more on cognitive gadgets, you can read my review of Heyes’s book here.)
Why should I care about cognitive gadgets?
The theory that we humans use cognitive gadgets to think is an attempt to explain what makes us unique as a species. And if the theory is correct – if it’s true that our minds are primarily shaped by our socio-cultural environment (rather than our genes) – it suggests our minds are quite pliable, and that we can improve our cognitive capabilities through specific practices, including education.
So what does “educable” mean?
Until about two weeks ago, I’d never heard this term myself. But it’s the word used by Dr. Leslie Valiant in his new book The Importance of Being Educable: A New Theory of Human Uniqueness. Valient, a computer scientist at Harvard (and Turing Award winner), argues that it is our capacity for being “educable” that makes humans special, and that we should replace the notion of human “intelligence” and use “educability” instead.
Right but what does educable mean specifically?
Valiant defines educability as “the combination of (a) learning from experience, (b) being teachable by instruction, and (c) combining and applying theories in both modes.” We explicitly pass along knowledge that has accumulated over many generations through education, and this knowledge gives us the unique ability to deal with novel, unforeseen situations -- in this sense, “education is different from training, which imparts the skill to perform a task that is foreseen at the time of training.”
That’s the quick summary anyway – at some point soon I plan to write a full book review, but if you want to go deeper now, this interview of Valiant is a good place to start.
Boy, educability sounds similar to “general intelligence” as defined by François Chollet that you wrote about recently, would you agree?
I sure would, and thanks for noticing the parallels. In particular, Chollet views our capacity to learn how to “extremely generalize” – that is, to handle “entirely new tasks that only share abstract commonalities with previously encountered situations, applicable to any task and domain within a wide scope” – as what makes humans unique. In contrast, artificial intelligence is mostly limited to task-specific skills – they are systems that have been trained rather than educated.
Or as Chollet recently tweeted, “General intelligence is *precisely* learning -- the ability to efficiently learn new things, beyond what your genes and past experiences prepared you for.” Sounds very educable. And cognitive gadgety.
Ok, so back the first question – is being educable a cognitive gadget? And does it matter?
Yes and yes, I think, though not everyone agrees on the first point. Valiant, for example, argues that educability is our “civilization enabler,” meaning, that our educable capacities have driven the development of our cultural institutions. When I interviewed him recently, I asked whether the causation might flow the other way – perhaps, per cognitive gadget theory, it’s our cultural practices that drive our educability? He seemed pretty skeptical, but I also was introducing him to the idea of cognitive gadgets for the first time, and in the context of a short phone call.
In any event, and however we might point the causal arrows, I think there is something harmonious happening across these insightful thinkers about how we think. Heyes, Valiant and Chollet are all pointing to our mental flexibility, our ability to learn and generalize, as what makes us unique and special. Educability is our superpower, and I’m prepared to argue it’s our most important cognitive gadget. Broadening understanding of this might help move us away from obsessing over “artificial intelligence” as something we train, and point us instead to developing “artificial educability,” as something that learns.
Easier said than done, of course.




Interesting. So, educable=able to learn new things or apply knowledge in new situations=general intelligence.
I don't much like the word "gadget" here - she also calls them "organs of thought," but I don't much like that either; I would just call them cognitive abilities that you can become skillful at. I would say cognitive architecture is rooted in genetics but developed by social interactions.
In your Education Next article where you review Heyes' book, you say that teaching programs are not interested in teaching the cognitive science of learning. That is so weird to me. To me, cognitive science is what pedagogy is all about. (But then, I must admit, most of my education profs were the worst teachers I ever had!?!) But when I was in college in the early 70's, Bruner was in and Skinner was out, and we learned things (mostly from the textbook!) like the cognitive principles you have on your Deans for Impact Science of Learning list, which were very helpful to me. The first 4 groups of principles on your list I would say are brain-based much more than culturally based. The ones in #5, on motivation, are more in the affective domain, which are still cognitive principles and very important for learning, but having more to do with the mind rather than just the brain, and more social-dependent.
The only thing on the list I don't agree with is "Students do not have different 'learning styles.'" I agree that it's not useful to give a survey and try to label the student and only teach them one way, but from 30 years of teaching high school and from observing my own kids, I certainly believe learning styles or modes exist, and they shouldn't be ignored.
My one daughter learned easiest from demonstrations. The other one learned easiest auditorally from lecture. They could learn other ways, but that was the easiest for them. Not that it isn't good to practice learning using the modes you aren't strongest in. I learn best by reading the theory behind something. My son is a tactile/kinesthetic learner. He's also dyslexic and on the autism spectrum, so he had a lot of trouble processing input, and needed time to figure out things in his own way. I had students who processed things better if they said them out loud in their own words, so I gave them opportunities to do that.
I think the best way to teach is to be sure to use the cognitive principles, like the ones on your list, that are known to work with most if not all students, but to use all the different modes possible for presenting and practicing concepts and skills. This catches what works best for each student, while giving them all extra contact and practice in a different way, which is better than repetition just one way. You also have to observe the students and see where they're having problems and try to figure out why and try different things. The other two things you have to keep in mind are (1) the teacher has to be sure they're not subconsciously assuming all students will learn best the way the particular teacher learns best and (2) not everything to be taught should be taught in the same way; sometimes the nature of the material or skill lends itself better to certain modalities.