Learning by doing - some thoughts

Read an article on "How we teach undergraduate computer science is wrong". Article was forwarded to me by the ever-enthu teacher - Ashish. The crux of the article is this - "learning by doing & discovering" doesn't always work. The ideas appealed to me because it was demolishing a dogma without creating a new one. Often "dogmas" are fairly useless things. However, I dont mind so long as they are used as a checklist and perhaps even as some sort of a guide.


I have nothing against "Learning by doing & discovering". I only think it is not like a cure-all. It will have its place and its limits. I learnt cooking by doing. My mother stood next to me and gave me instruction while I faithfully put in all the ingredients to make a dish that was as good as my mothers. That was learning by doing. I couldn't have learnt to cook half as easily had I only read about all the recipes and never physically cut the vegetables and done all the other stuff. But to be a good cook it is important to know the concepts -  the effect of ingredients on the taste of a dish ; the effect of a cocktail of ingredients on the taste, etc. Concepts are important. In cooking these concepts are gained from one's own experience and also that of others (all passionate cooks share notes about their dishes!). Learning by doing is a great way to learn cooking .


How about learning physics the same way. Lets teach a person how to measure & calculate speed. A child understands the "idea of speed" just like he understands "good taste". But he must now measure & calculate speed.


Measuring is simple. Hand over a ruler and a clock and the child can measure the distance and the time taken to move that distance. Simple. Is it? The child has yet to calculate "speed". That is simple too. Just divide distance by time and you get the speed. It looks simple but it really isn't. If you ask the child to divide distance by time she can carry out the division flawlessly. The child can get the answer but how does she know that the answer gives the speed?  To my mind the child will not learn the meaning of "distance by time" during this "doing" exercise. This meaning has to be taught earlier (in a "non-doing" setting). It has to be taught through the "idea" of division. 


Division is a practical exercise. Divide a piece of cake into 4 parts each person gets 1/4th of the cake. Divide distance of 100 m into 10 parts, each part is 10 m long. Divide 60 minutes into 10 parts you get 6 parts. Thus to teach division is it  good idea to bring a cake or a clock to the class and show how it works? I think that would be a disaster. A clock is not just an object that tells us the time. It is an object with a design, and so many parts - dull or attractive, etc. It has too many elements that will distract from the concept of "division using a clock". It just makes more sense to draw a clock on the board and go ahead from there. There is another stumbling block to trying to get too close to reality. It takes one away from the "concept". A "concept" pervades examples and is not localized within that example alone. So, why make a song and dance about "learning by doing". The important thing is the "thinking". You cant learn the concept without thinking. Certainly, examples must be given and concepts will get planted in the brain only after studying variety of examples. That is just the traditional way. But the concept will take much longer to get implanted by insisting on "Learning by doing" (as i said there will be too many distracting elements competing for attention and the concept of division is likely to find it hard to survive this competition).


This is all for today. I think I have more things to say on the matter but that's for later. But, i dont want to be misunderstood so let me add here. I am not a critic of "learning by doing". However, I wont hold it as a dogma. It has to be used and has to be used at the right opportunity. That may be a subject matter of one of the later posts.