Duolingo streaks, TikTok explainers, five-minute hacks: are these tiny lessons genuine learning or just feel-good trivia? Science has a surprising answer.
As a tech professional, I feel like I am always micro learning while looking up and figuring out how to do certain things on the fly. I appreciate this perspective and method especially as in my personal life I am also focusing on small wins each day and the bigger task in front of me can be more achievable that way. I am curious as I read this though, what your thoughts are on the flow state in terms of learning? I’ve noticed that while I sometimes set aside a small amount of time to watch a learning video at times I want to continue, like my mind actually wants to stay focused and immersed in one area for a longer period of time.
Thanks for your very thoughtful and well reflected comment. I absolutely see your point and this is what I partially tried to cover in the second point of the section “where it doesn’t work”. In any case, I think the key to understanding the micro vs macro debate is to understand the resources (both cognitive resources, as well as time and effort) that you as a learner have available and are willing to commit. As I understand that, studies show that microlearning works on a big scale because for most learners, it is true that reducing the cognitive load enables learning on the first place where it wasn’t possible before. But of course, if you have the chance to invest more resources and put in more commitment, your results will be better. If you’re interested on how exactly to go about that, you might find my article on memory encoding interesting.
Chunking and other strategies are cognitivist. Cognitivism is a model, but it doesn't predict learning. Eg. in language acquisition, cognitivism can explain this or that, but these strategies can't predict learning as much as socio-affective factors, aka motivation.
Hmmm... it's a bit unclear to me what exactly your point is.
In the article I link plenty of scientific research proving the efficacy of both chunking as a general technique and microlearning as a specific implementation. So, it does predict learning. The fact that these strategies "are cognitivist and cognitivism is a model" (whatever you take that to mean) doesn't invalidate that. Here's one of the articles I linked providing scientific evidence on chunking: https://pubmed.ncbi.nlm.nih.gov/29698045/
Also, I don't quite understand the difference you make between "it can explain this or that" and "it can't predict learning". What do you mean? It can explain why learning happens or doesn't, not just "this or that". It literally does explain precisely whether learning happens or not.
In the case of language, we're talking about a pretty complex, long-term process where many factors play a role. Language is, interestingly enough, at the same time a hard skill and a soft skill. For the hard skill side of things (grammar, vocabulary, syntax), microlearning does help. But language learning also includes soft skills (communication, expressing nuance, making presentations, reacting, negotiating, etc.). Chunking alone can't predict the success of the complex process of language learning, but it can predict the success of the achievement of partial learning outcomes it's meant to support.
The debate around language learning and language acquisition would be too long for a comment, but in a nutshell, acquisition isn't this "intuitive revelation"-kind of learning that many take it to be. It can also be influenced and boosted by learning design, and there chunking plays a big role as well as the factors you mention.
1. Language learning is 'emergent': it's non-linear and its the result of mass input.
2. What applies to langauge applies to everything else (unless shown otherwise), language is the paradigm case, cognitivism was initiated by Chomsky's Language Acquisition Device.
3. As a career English language teacher with a PhD in cognitive science, I don't buy any evidence 'proving the efficacy of learning'.
It's pretty much impossible to control, and performance on a test is meaningless (see Dreyfus below)
Most research studies can't be replicated, and cognitivist research has failed to translate into practical classroom techniques (chunking is as good as it gets and it's pretty disappointing, see below);
Schooling (instructed learning) is a total failure IMO, both in the classroom and in online apps.
Cognitivism is merely delaying the inevitable realisation that schooling needs to be dismantled because it's anti-human and anti-brain (it's basically brainwashing).
5. As Hubert Dreyfus argued: competence/expertise is not rule following but knowing how to respond to the situation, ie., leaving rules behind.
Holding on to rules/strategies is just postponing the maturation of competence.
See Dreyfus's five stages of skill acquisition.
6. Cognitive strategies research can explain, eg., why learners fail to perform on some measure.
Eg., I can explain pronunciation problems as being due to text-based learning which hides prosodic features of pronunciation, like intonation, linking, reducing, etc.
Teaching these features however does not predict that they will be acquired (though it might help).
On the other hand, learners can acquire these features without instruction.
Problem is, you can only teach a limited number of features per skill domain.
More instructed learning does not equal more learning beyond the very very basics, whereas cognitivism implies that you just need more 'strategies' and more instructed/intentional learning instead of 'just loads of input' which requires sheer motivation to interact.
7. Ergo, "it can predict the success of the achievement of partial learning outcomes it's meant to support" ... this is the problem. It's extremely partial, but it's all teachers and apps can do.
The problem of schooling (classroom or apps): If all you have is a hammer ...
8. Teaching or supported learning should aim only at this: comprehensibility, ie., disambiguation of input.
Cognitivism does not make this clear with it's concept of 'strategies'.
Unless the learner is highly motivated, the strategies do nothing.
If they promise that learning 'will' take place (in a deterministic way) that's false.
Moreover, strategy training often reduces motivation, eg., by insisting that this is the 'correct' way of doing something (eg., by insisting that there is a 'correct' accent) it's counterproductive. Motivation is all.
Your 'soft skills' imply correctness.
The whole 'skills' business is counterproductive in that way.
You don't get to decide what is a 'skill' (teacher/teaching centred perspective).
I mean, the schooling system insists on deciding, which is why it fails so badly, despite decades of cognitivist research.
8. Learning design: comprehensibility vs. chunking.
You have no way of deciding what is the 'correct' size of the chunk bc you have no way of knowing at what point in acquisition the learner is at a given point in time and it's a constantly moving target (monkey brain, learning is non-linear).
Why cognitivism is ultimately 'teaching centred' and not 'learning centred'.
I'm even skeptical about grading the input. Motivation is all.
See:
- John Schumann: acculturation hypothesis, neurobiology of learning, learning as foraging.
- Emegent grammar.
- Hubert Dreyfus: phenomenology of skill acquisition.
As a tech professional, I feel like I am always micro learning while looking up and figuring out how to do certain things on the fly. I appreciate this perspective and method especially as in my personal life I am also focusing on small wins each day and the bigger task in front of me can be more achievable that way. I am curious as I read this though, what your thoughts are on the flow state in terms of learning? I’ve noticed that while I sometimes set aside a small amount of time to watch a learning video at times I want to continue, like my mind actually wants to stay focused and immersed in one area for a longer period of time.
Thanks for your very thoughtful and well reflected comment. I absolutely see your point and this is what I partially tried to cover in the second point of the section “where it doesn’t work”. In any case, I think the key to understanding the micro vs macro debate is to understand the resources (both cognitive resources, as well as time and effort) that you as a learner have available and are willing to commit. As I understand that, studies show that microlearning works on a big scale because for most learners, it is true that reducing the cognitive load enables learning on the first place where it wasn’t possible before. But of course, if you have the chance to invest more resources and put in more commitment, your results will be better. If you’re interested on how exactly to go about that, you might find my article on memory encoding interesting.
Chunking and other strategies are cognitivist. Cognitivism is a model, but it doesn't predict learning. Eg. in language acquisition, cognitivism can explain this or that, but these strategies can't predict learning as much as socio-affective factors, aka motivation.
Hmmm... it's a bit unclear to me what exactly your point is.
In the article I link plenty of scientific research proving the efficacy of both chunking as a general technique and microlearning as a specific implementation. So, it does predict learning. The fact that these strategies "are cognitivist and cognitivism is a model" (whatever you take that to mean) doesn't invalidate that. Here's one of the articles I linked providing scientific evidence on chunking: https://pubmed.ncbi.nlm.nih.gov/29698045/
Also, I don't quite understand the difference you make between "it can explain this or that" and "it can't predict learning". What do you mean? It can explain why learning happens or doesn't, not just "this or that". It literally does explain precisely whether learning happens or not.
In the case of language, we're talking about a pretty complex, long-term process where many factors play a role. Language is, interestingly enough, at the same time a hard skill and a soft skill. For the hard skill side of things (grammar, vocabulary, syntax), microlearning does help. But language learning also includes soft skills (communication, expressing nuance, making presentations, reacting, negotiating, etc.). Chunking alone can't predict the success of the complex process of language learning, but it can predict the success of the achievement of partial learning outcomes it's meant to support.
The debate around language learning and language acquisition would be too long for a comment, but in a nutshell, acquisition isn't this "intuitive revelation"-kind of learning that many take it to be. It can also be influenced and boosted by learning design, and there chunking plays a big role as well as the factors you mention.
1. Language learning is 'emergent': it's non-linear and its the result of mass input.
2. What applies to langauge applies to everything else (unless shown otherwise), language is the paradigm case, cognitivism was initiated by Chomsky's Language Acquisition Device.
3. As a career English language teacher with a PhD in cognitive science, I don't buy any evidence 'proving the efficacy of learning'.
It's pretty much impossible to control, and performance on a test is meaningless (see Dreyfus below)
Most research studies can't be replicated, and cognitivist research has failed to translate into practical classroom techniques (chunking is as good as it gets and it's pretty disappointing, see below);
Schooling (instructed learning) is a total failure IMO, both in the classroom and in online apps.
Cognitivism is merely delaying the inevitable realisation that schooling needs to be dismantled because it's anti-human and anti-brain (it's basically brainwashing).
5. As Hubert Dreyfus argued: competence/expertise is not rule following but knowing how to respond to the situation, ie., leaving rules behind.
Holding on to rules/strategies is just postponing the maturation of competence.
See Dreyfus's five stages of skill acquisition.
6. Cognitive strategies research can explain, eg., why learners fail to perform on some measure.
Eg., I can explain pronunciation problems as being due to text-based learning which hides prosodic features of pronunciation, like intonation, linking, reducing, etc.
Teaching these features however does not predict that they will be acquired (though it might help).
On the other hand, learners can acquire these features without instruction.
Problem is, you can only teach a limited number of features per skill domain.
More instructed learning does not equal more learning beyond the very very basics, whereas cognitivism implies that you just need more 'strategies' and more instructed/intentional learning instead of 'just loads of input' which requires sheer motivation to interact.
7. Ergo, "it can predict the success of the achievement of partial learning outcomes it's meant to support" ... this is the problem. It's extremely partial, but it's all teachers and apps can do.
The problem of schooling (classroom or apps): If all you have is a hammer ...
8. Teaching or supported learning should aim only at this: comprehensibility, ie., disambiguation of input.
Cognitivism does not make this clear with it's concept of 'strategies'.
Unless the learner is highly motivated, the strategies do nothing.
If they promise that learning 'will' take place (in a deterministic way) that's false.
Moreover, strategy training often reduces motivation, eg., by insisting that this is the 'correct' way of doing something (eg., by insisting that there is a 'correct' accent) it's counterproductive. Motivation is all.
Your 'soft skills' imply correctness.
The whole 'skills' business is counterproductive in that way.
You don't get to decide what is a 'skill' (teacher/teaching centred perspective).
I mean, the schooling system insists on deciding, which is why it fails so badly, despite decades of cognitivist research.
8. Learning design: comprehensibility vs. chunking.
You have no way of deciding what is the 'correct' size of the chunk bc you have no way of knowing at what point in acquisition the learner is at a given point in time and it's a constantly moving target (monkey brain, learning is non-linear).
Why cognitivism is ultimately 'teaching centred' and not 'learning centred'.
I'm even skeptical about grading the input. Motivation is all.
See:
- John Schumann: acculturation hypothesis, neurobiology of learning, learning as foraging.
- Emegent grammar.
- Hubert Dreyfus: phenomenology of skill acquisition.