Cargo Interactive

This is one of the tenets of behavoural psychology. If you ignore all elements of cognition, and only look to the relationships between stimulus and response, you can avoid trying to explain all the complex stuff that goes on in your head. And for a long time this was extremely successful. We were able to gleam a whole field of information about how animals might learn to respond to certain situations. But when you start to get into murky concepts like categorization, or identifying attributes, this get’s really hard.

Take, for example, salty and sweet. Two tastes you would assume are quite easy to discriminate. It’s actually quite hard, and I struggled for a long time with one boy trying to get him to tell the difference. You can’t say things like, “You have sweet things for desert,” because sometimes you might have potato chips after dinner. Is this really desert? Or would you consider it an after dinner snack? Well, even that discrimination could be a nightmare to teach to someone who can’t easily tell the difference to begin with.

So I stopped, and I decided to start with two given axioms. Crackers are salty. Chocolate is sweet. I told him, “Don’t worry about why this is true, just assume it and memorize it.” Then I asked him, is chocolate cake like chocolate, or like crackers? It’s like chocolate, and chocolate is sweet, so chocolate cake is sweet. How about potato chips, are they like chocolate or crackers? Well, potato chips are like crackers, so potato chips are salty. Now he had some conjectures to work with.  French fries are like potato chips, it’s salty. Ice cream goes with cake, it’s sweet.

Suddenly I took the guess work out of it. The boy didn’t have to memorize large sets of information, and he went from getting 50% of the answers right to about 90%, in a matter of minutes. It’s not perfect, you can always get some strange associations. Sometimes, pretzels are covered in chocolate. But for those special cases, if that’s all you have to worry about, you can memorize it brute force. If you have to learn a lot of discrete information at once, you’re going to start mixing up what goes where. Anyone who’s tried to cram for a test has encountered this. But if you can use an analogy or a metaphor which accounts for the majority of cases, you can take out the guess work. Of course, the hardest part is finding that analogy to begin with.

The solution to insight puzzles often relies on restructuring. This usually means that the problem-solver has made a mistake somewhere in their interpretation of the problem, and before they can make a solution assumptions need to be changed. For example, take this insight puzzle:

A man has married over forty people in his life. He has never been divorced and he is not a polygamist. How can this be?

Like most people, I would get tripped-up by the word ‘married.’ Marriage the most important concept in the puzzle, and it helps define all the constraints: getting married more than once implies divorce or polygamy. Of course, a simple solution is to say, ‘He murdered all his wives.’ Technically, this solves the problem, but nobody would admit that the solution took any insight. The real solution is that the man is a priest, and by ‘married over forty people’ we mean he oversaw the ceremonies.

What characterizes this difference between the cop-out answer (murder) and the valid answer? Is it the change in meaning for the word ‘married’ which causes that ‘a-ha’ feeling?