Editable Google Slide (CC-BY-4.0)
One of the greatest pleasures in life is finding things out. Especially finding how things work.
We believe there are two general strategies / methods for active learning that we (humans) use. Note the word ‘active’. Passive learning is different and in our experience dramatically less effective. Now back to the 2 methods of active learning.
1. Taking things apart
This is the obvious one: if you want to know how something works – take it apart and see what it’s made of.
Taking things apart has allowed us as a species to dive deep into what matter is made of. We’ve been creatively breaking things to figure out what they are made of and how they work for quite a while. Small children start doing that way before they learn how to speak.
Yet there are a few shortcomings to this method. Let us highlight 2 that we find most interesting:
- First, sometimes as soon as you take a thing apart – you break it completely and all the traces of behavior you were interested in – are gone.
- Second, sometimes you simply can’t figure out how a thing works by taking it apart. Because the exhibited behavior of the thing and of its parts (in a rich and complex environment) seems to be too complex to give you hints of how the whole thing actually does what it does. Human brain is a good example – even with the recent advances of mapping activities to individual neurons fundamental questions about how we think are still up there, largely not answered. This is where Computational / AI methods of research come into play. Methods that instead of taking the real brain apart, try to put an artificial one together and through that learn how thinking in general works.
2. Putting things together
Putting things together to learn how something works – this may sound counterintuitive and yet is extremely powerful. Consider a toddler, trying to learn how language works. Taking language apart and identifying underlying grammatical structure is beyond her capability of concentration and logical thinking. Yet she is fully capable of (gradually) grasping the language by putting what she already has (sounds of her baby-speak) together in a variety of ways and applying it to where adults usually apply language. Over time and after much trial and error – she gets it. Just like toddlers, scientists now have to use this method more and more whenever they are faced with behaviors that are too complex to analyze by taking them apart. Human brain? A lot of our current understanding of how it works comes not from just taking it apart, but from putting crude models together and seeing where they fit and where they break.
‘Putting things together’ method is, in our view, undervalued. Given Braitenberg’s law of ‘uphill analysis and downhill synthesis’, it can often get us to better results more efficiently.
The major problem with this method though is that one can always put up a ‘qualia’ argument. How do we know that your model that seems to exhibit the same behavior as our object of study, is internally similar as well? There is no answer. People around you may well be zombies pretending that they think, but ‘actually’ not doing it at all.
Personally we have a pragmatic viewpoint on this. If it walks like a duck and talks like a duck, then most likely it’s a duck after all.
Now the most important thing.
We believe active learning at its best is always a combination of 2 approaches described above. The process can start anywhere, but has to go through the cycles of ‘putting things together’, ‘taking them apart’ and in always looking carefully at how things brake (because those edge cases are pointing us to the essential components of complex behavior that we are studying).
As we learn how to think and investigate the world better (and teach our kids to do it as well), let us not forget about the power of putting things together as a way of learning how things work. Let’s build models (if only mental ones) & watch them brake. Making something may very well be one of the most powerful ways of learning how it works.
P.S.
As you may have noticed, this approach is applicable not only to finding how existing things work, but also to designing new things. Because essentially design process is an exploration of how something (that doesn’t exist yet) should work. And if viewed from this angle, design process is simply learning how the thing you are trying to design works. You can take a vague idea (or a possibility) of a new thing and try to take it apart in your mind’s eye. Then you can switch over to putting (already available) things together in a way that would crudely represent the thing you are trying to design. Then you see where your model fits and pay particular attention to where it breaks. This gives you more insights about how the ‘real’ thing should work. Repeat this process over and over and you may end up with a good new thing.
But that’s a different topic altogether.
Read more:
Lev Vygotsky, Jean Piaget, Valentino Braitenberg, Seymour Papert