3 ms·
I personally know people who are totally incapable of anything spatial but are very good with logic and maths. I score rather highly in standardised spatial tes
by fauigerzigerk 18y ago
I personally know people who are totally incapable of anything spatial but are very good with logic and maths. I score rather highly in standardised spatial tests myself, but I rarely think visually about algorithms or data structures. I found that doing so keeps me thinking inside the box, literally.
Many abstract problems are high dimensional and items are linked by functional relationships that are very difficult to express visually. I've seen so many people cripple their understanding of problems by forcing them into diagrams that are just plain wrong in some important respect.
I believe spatial thinking works very well for some things but completely breaks down for others. Just a very simple example. It's pretty easy to visualise a set of three numbers: {2, 7, 4}. Just put them into boxes. But now try to visualise the set of all primes or the set of all sets that do not contain themselves.
Even for sets that are finite at a particular point in time, like the set of adults, thinking of them as items in a box is very misleading. What would have to be visualised is really the membership criterion. But that set still contains items at a particular point in time, so how do you visualise the relationship between the membership criterion and those items?
I know what happens at that point. Vague visual analogies that keep readers of popular science magazines happy but have very little utility for actually doing research start to creep in.
But dont' get me wrong. I know very well that visualising things can help to understand many things very well. Just not all things, and the problem is that people don't know where to stop trying.
- whacked_new 18y agoI have some similar thoughts. For example, when people learn about 2-space and 3-space (or 2-d and 3-d arrays), the most common thing to do is to draw out the spatial analogue: a WxH table, a WxHxD cube. And so the examples always stop at 3. But in math you don't stop anywhere. I have wondered if people who have extreme finesse with math at arbitrary dimensions change strategies when dealing with higher dimensions, or if they never did this spatial conversion all along and deal with it completely symbolically. Then I tried to think of what certain things mean symbolically. In integration it's always the area. But in symbols it's just a cold, unfriendly product. I'm sure there are merits to doing it this way, but I have trouble seeing intuitive meaning in it. Perhaps the better mathematicians simply calculate and don't care about what sense it makes?
- fauigerzigerk 18y agoI'm not one of the better mathematicians so I cannot know, but I suspect you are right. And if you think of physicists, much of what they do is finding out what space is in the first place. So for them to rely on spatial metaphors or visualisations could be downright destructive to their thinking. And I believe the same holds for much of the logic we create in software.
- sfk 18y agoI don't remember the exact words, but in his video lectures about linear algebra Gilbert Strang says he has no real visual concept of higher dimensions.