Overthinking, the curse of the rank beginner

09 Oct 2007 // science

Most Top 10 lists are gratuitous examples of loose and woolly thinking. Still, you sometimes come across a gem. Some of the most hard-boiled advice for creative work comes in the form of the Top 10 Things They Never Taught me in Design School courtesy of architect Michael McDonough. (I count science as a semi-creative endeavor.)

The point that struck me most was number 4: "Don't over-think a problem." McDonough's point was that all too often, we over-elaborate an idea into a needlessly baroque version of itself. "Every other critic I ever had always tried to complicate and prolong a problem when, in fact, it had already been solved. Designers are obsessive by nature. This was a revelation. Sometimes you just hit it. The thing is done. Move on."

But something about this piece of advice gnawed at me — there's more going on. In McDonough's complaint about overthinking, I think he's hinting at a much deeper phenomenon, namely that there are really two very different modes of thinking, and we often default to one mode when we should be operating in the other. For the purposes here, I'll call them overthinking, and bodythinking.

I remember when I was little, when I'd blurt out some kind of childish idiocy, my mother would, after a quick rap on the knuckles, remind me to think before I speak up. I remember thinking how that didn't make any sense: when we speak, we don't think about what we say before saying it. Thinking occurred in the very act of speaking. If I were to eliminate saying stupid things, it can only be out of habit, not through some kind of higher order cognitive process cutting off stupidity at the tongue.

For most of our education, we're encouraged to manipulate abstract rules, and sterile logical thinking. this is the essence of overthinking – the manipulation of totally abstract, and abstruse rules of logic. This is the domain where the IQ measure rules. IQ questions focus on spotting symbolic patterns, and making spatial relations.

IQ tests test abstract thinking, or overthinking. And certainly useful as an educational heuristic, it has very little place in the real world. Although I have friends who value their IQ quotient, I have never taken the test. Low or high, an IQ score tells us little about useful thinking. Real useful thinking is always contextual, bounded by a real world situation. Even in mathematics, arguably the most abstract of the academic disciplines, a high IQ does not necessarily qualify one to become a great mathematician. There is more to doing deep mathematics than being a member of MENSA.

In pretty much every introductory philosophy class, the case of the famous Wason four-card problem is always trotted out. This is a perfect example of overthinking. In the Wason problem, the students are first given a logical relation between numbers printed on one side of a card, and letters on the other side. Then, four cards are presented, such that only one card needs to be flipped over to check the consistency of the logical inference. The students are asked to reason about the logical inference and choose the card to flip. Of course, most students get it wrong.

But a later study found that the Wason problem misses how we really we think. The later study phrased the same logical relationship of the Wason problem, not in terms of an abstract relation between numbers and letters, but as an ordinary relationship between the age of a punter and underage drinking. Then, four punters are shown. Pretty much everyone works out which punter needs to be checked for illegal drinking. This shows that we can all reason logically, but we need a context to do it in. It is the abstraction – the overthinking – that trips us up.

Overthinking is the great divider between the pro and the neophyte, or in the case of science, the grad student and the postdoc. A student immersing herself into a field for the first time has no context – no knowledge of the principles, the jargon, the rules of the game. To come to any conclusion about a difficult scientific question, they run through the freshly crammed rules as if they were a set of abstract rules, just like an IQ test.

The expert works very differently. In his vastly entertaining study of snap judgements "Blink", Malcolm Gladwell shows that, for instance, art historians make complex decisions about the authenticity of a work in but a fraction of a second. They do not so much as think through a maze of inferences but feel the answer as a perception, even a pain in the gut.

In his memoir, G. H. Hardy likens the arduous process of solving a serious mathematical process to working on a loose tooth with the tongue. It is no accident that Hardy uses such a body metaphor to describe the process of thinking mathematically. Expert thinking is bodythinking. Bodythinking implies that the process of thinking is lived through, and felt, not reasoned through the conscious mind.

Perhaps a better example is the one that Hubert Dreyfus gives to explicate the Heidegerrian notion of "Being-In-The-World", which is a much more complete concept of bodythinking. Dreyfus too, tackles the difference between the beginner and the expert behavior, in two very different disciplines – chess and tennis.

In chess, Dreyfus summarizes the bountiful research on reaction times and playing behavior of chess players. The surprising thing is that during a game, both beginners and experts consider a similar number of moves. The expert does not grind through more calculations the beginner. Instead, the expert can see more in a board than a beginner. When shown a random snapshot of a game, the expert will typically be able to remember and recall it later with ease, like recognizing a face. Not so, for the beginner. The rules of chess have been embodied in the perception of the chess playing expert. However there are limits. When shown a randomly generated board of chess pieces, including positions that cannot possibly be generated in a real game, the expert will be unable to remember the position easily.

Dreyfus also studied the phenomenon of choking in tennis – how great tennis players sometimes break down inexplicably in a big game. Dreyfus argues that the transition from beginner to expert is to embody a complicated set of body movements – the tennis serve – from a set of disjointed analytical movements, into a unthinking action. The server embodies the motion. It is subconscious. Only if an action is embodied, can a athlete perform at high capacity. But in the heat of a big game, the athlete starts to worry about his skill, and in that moment, loses confidence in the embodied action. The athlete then starts thinking about the serve, and consequently tries to consciously adjust the action, thus bypassing the embodied action. He will then serve like a beginner again. Choke.

I find that now, after working over 10 years as a scientist, that reading a scientific paper is a much different activity than it was I when I started. Whereas before, I would struggle with opaque paragraphs, ill-defined terms, and incomprehensible manipulations, now, when I read a paper, I read by first putting my mind on automatic, then I mechanically read the prose of the paper, trying to keep as much of the text in my head at once, and wait for something interesting to jump out at me. I scan the lines in a text like a hunter trying to find the tail of the stag in a dense leafy forest. The difficulty is in the process of continually pushing through the text, not in identifying interesting, and important points.

The difference between expert and beginner, is embodiment. What, for the beginner, is a set of arbitrary but logical rules, becomes for the expert, a series of gut reaction. Working through a problem, is a slow body process, like digesting a full rich meal. Indigestion occurs when you overthink too much.