Convergent vs. Divergent Thinking
January 11th, 2009
I’ve written in the past about how I think that “data first” approaches are to be preferred over “structure first” approaches, but I could never come up with a sufficiently precise yet simple cognitive model of why.
Until very recently, when I came across tests that are meant to measure the ability for a person to ‘diverge’ from a given topic, other than ‘converge’ (or, in more familiar terms, ‘abstract’) from many to fewer.
Let me take a step back: standardized intelligence tests have been chronically focused on measuring the ability for an individual to perform convergence tasks such as abstraction, categorization, pattern matching, logic inference, series prediction, etc. All of them are based on the idea that intelligence is the ability to somehow process information and distill models about it. The bigger (and faster) the convergence (how many items you can fit in a category, how encompassing your model is), the better, the smarter.
This is obviously an outdated idea, but it’s still a surprisingly influential meme, and I could never find a better and more resolute answer to classification of intelligent thinking.. that was until I came across the idea of ‘divergent thinking‘.
While convergent thinking is about distilling few from many, divergent thinking is the opposite: start from few and come up with more.
Examples of divergent thinking are creativity, imagination, fantasy, flexibility, curiosity.
Let’s not make the mistake of thinking that rating creativity has anything to do with rating the ‘product’ of such creativity against others’, that’s a totally different thing: here, we’re merely trying to understand the ability, tendency or ease that a person has in accomplishing tasks that require divergent thinking.
Here one example of one such test: you’re told about an object, say ‘a screwdriver’, and it is asked of you to list all possible uses that you can think of of that object.
What was interesting to me when I came across that test was how poorly I performed at first.
I could not think about any use of a screwdriver other than ‘screwing/unscrewing a screw’ and, say, ‘open a can of paint’. It was only after reading other people’s responses such as “a pole for a little flag”, “a pin for a hinge”, “an ice pick”, “a weapon”, “a percussion instrument” that I realized that I had been automatically applying my convergent mind onto a divergent test. Once I knew that I had to use my imagination, that the uses only needed to be even just remotely plausible, not optimal, I started to perform a lot better.
It was this personal experience with such cognitive dissonance between convergent and divergent thinking that made me realize how different such modes of operation are and how they probably powered by very different areas of our brains.
It got me wondered again about the ‘data first vs. structure first’ dichotomy, which seems to emerge constantly around me in various aspects of my job, and made me realize that the ‘data first’ approaches put the emphasis on the divergent nature of the task, while ‘structure first’ approaches put it on the convergent one.
Considering my personal experience of automatically using a convergent mode to solve a divergent task and performing poorely (and with frustration), I wonder how many times tasks are designed by people using a convergent mode and given to people that naturally tend to use a divergent one. Could this be one of the reasons why most human-computer interfaces perform extremely poorly or create so much frustration to the average user? because their designers used one mode of thinking during the design and their users another to actually use the software?
I don’t have an answer for that, but thinking more about it can’t hurt.