Are there good learners and bad learners?

30 October 2019

Interview with 

David Mets, University of California San Francisco

Finch.png

Birds learn best when presented with a song tuned to their abilities

Share

Some people are better learners than others. Or so we thought. Actually, it turns out that there might just be a disparity between the way they learn best and the way the information is being presented. Speaking with Chris Smith, UCSF’s David Mets explains what he's been finding with his songbirds…

David - Some people learn better, and others learn worse. And there's some evidence that some of that is genetically-driven. But we know that some of that can be environmentally driven. So one question is, can we provide a better environment for some genetic makeups to increase learning outcomes.

Chris - In other words, if I'm a genetic poor learner, can you nonetheless compensate by changing the environment to one to which I am better adapted to learn in?

David - That's right and, additionally, you might not be genetically a poor learner you just prefer to learn with a specific type of presentation, and so you might seem poor because the stimulus isn't presented in a way that is tuned to your preferences. But, if you tune it more accurately, you're actually quite a good learner.

Chris - And how have you been looking into this?

David - The system where we study this is in songbirds, and they learn their vocalisations in a process very similar to how humans learn speech. They listen to their dad early in life and then they go on to practice, practice, practice and, ultimately, produce a very complex vocalisation. And this is a powerful system, because we can vary the environmental parameters and we also can vary the genetics, so we can ask very specific questions about Is this the right environment for this individual based on their genetic makeup, or is it the wrong environment.

Chris - How did you actually do that then? How did you vary the genetics and vary the environmental parameters in order to test that and tease apart the two?

David - So in this particular population of finches where we work we know that some individuals are biased genetically to sing faster and some are biased to sing slower. We can go in and take individuals we know to be genetically biased to sing faster and we can take some individuals and present them with a tutor song that is similarly fast, or a tutor song that is at sort of an average rate, or at a very slow rate. And then we can see how well they learn in those three different environments. And we can do that for fast learners, medium learners, and also slow learners.

Chris - So this is a bit like I go to a lecture at medical school and some lecturers teach with overheads, some teach with PowerPoint, and some are in the dark ages with talk and chalk, and it suits some of the class better than others. And so you're varying the different teaching styles. But how about the genetics, because you mentioned you can also mix that up?

David - Yeah. So in this case we basically took a population that is a genetically heterogeneous population, so there's lots of different individuals with a different genetic makeup, and we can measure how that genetic bias works. So, basically, we take individuals from all different genetic backgrounds and we provide them with one tutor song and we see what they end up being biased to sing. Basically we have a fixed environment and we see if they behave as their father behaved, and so we can then estimate whether that bird is biased to sing fast or slow based on their genetics.

Chris - And then the obvious question you're going to be asking is "right, okay, a bird that doesn't learn well in one setting, if we vary the conditions, it might be a slow learner in one setting but then we switch modality and it learns much better?"

David - Yes that's right. One possibility is that you see better and worse learning and the best stimulus will be the average song. So we provide everybody with an average song and some individuals appear to be poor learners, some individuals appear to be good learners. The birds who in that context appear to be poor learners we take those guys and we say oh what's their genetic bias, or these guys were all genetically biased to sing slowly, and so we'll provide them with a slow stimulus and we'll see what happens. And for all of the birds, for slow birds members and fast birds, providing a stimulus matched to their genetics increased their learning and, in many cases, it increased it to the point of being essentially equivalent across the different genetic backgrounds.

Chris - So it really is horses for courses when it comes to learning isn't it! Why do you think this relationship exists?

David - I think that this relationship exists mostly because there is genetic diversity. Individuals are different and their brains are built in different ways and they're more adept at thinking about things in this dimension or that dimension. To some degree, that's determined genetically, but the mind has such capacity for adaptation and plasticity and taking in information that, really, these genetic influences are tamped by that ability. But, tuning it a little bit to be slightly more appropriate results in better learning, and so I think it really is just a product of having - you know - organisms that are to some degree influenced by their genetic structure and then also an incredibly adaptive neural system that that helps them learn.

Chris - The parallels with humans - now you talk about it, and as I come to think about it - are really potentially very striking aren't they? When you think "how does advertising work? What's the best way to teach children in a classroom? Are influencers on YouTube influential because of whom they're talking to and the way they're doing it?"

David - Yes, it's an excellent question. We focus in the paper on the case of trying to adapt learning stimuli to individuals to increase outcomes, and we have been actually quite slow at applying that kind of idea in in educational settings, but Google and YouTube and Facebook and so on are extremely good at figuring out our individual proclivities, and providing you with an ad that matches you perfectly; matches your biases and shifts you even further in a specific direction...

Comments

Add a comment