Defining Intelligence

There are many things I can do which impress people, some of these talents are abilities that people believe you need heightened levels of intelligence to be able to achieve.  That sounds arrogant to state, but the reason I have worded it this way is because I intend the statement to be contentious.  I take issue with the implication that knowledge and intelligence are equal.  I fervently oppose this sentiment because many of the things I can do that people find impressive or a sign of intelligence are not abilities that I would consider unique to myself, they're not even abilities I would consider a sign of intelligence. 

For example, I can solve a Rubik's cube which impresses some people.  The first thing I usually say to people is that they could learn how to do it too if they took the time and effort.  The reason I take issue with seeing this as a sign of intelligence is that the method of solving a Rubik's cube is essentially an algorithm - a series of instructions or steps that you follow in order to complete a task.  Anyone who takes the time to learn the steps to solving the cube will be able to do it.  You can even create a computer program that takes the state of a cube as an input, performs the algorithm on that input and produces a solved cube as an output.  Does that mean it is intelligent?  Well clearly the answer is no, it's just doing what you programmed it to do.  The same is true of any human solving the cube.  The true mark of intelligence is whether or not the person or the machine solving it figured out how to do it on their own or whether they were told how to do it.

That distinction is something that isn't easy to make when it comes to humans.  At least with machines you can examine the code of their program for signs that it was written by a human.  However with humans who can solve the cube, determining whether they figured it out themselves is a bit more complex.  You could pose questions about the underlying mathematics, but that assumes the person figured it out via a mathematical approach.  They may have discovered how to solve it without an understanding of the underlying mathematics.  You could ask questions about the mechanisms of the puzzle and its behaviour but again that assumes they studied these to determine how to solve it.  To draw a parallel this is like Maths class in school when you were asked to show your working out when solving equations - you might actually know the answer and how to do it, just have an inability to put that down on paper or explain it.  Therein lies the limitation of intelligence, that understanding of our own understanding often limits exterior examination.

You can use rote learning - the repetition of information until it is recalled effortlessly, but just because a pupil can recite their 12 times tables to heart doesn't mean they actually know how to multiply two numbers together.  The depth of their understanding is exposed when you ask them to do it with new inputs they have not seen before.  Asking a pupil to perform multiplication of 13 times tables for example will demonstrate whether they actually know how to multiply two numbers together.  In our Rubik's cube example this can be demonstrated with the Rubik's cube variants.  The traditional cube is 3x3x3 but there are other variants, 2x2x2, 4x4x4, 5x5x5 and above.  Give them a cube of a size they have not used before and if they truly understand the underlying mathematics or principles at work in the design of the puzzle then they will be able to solve it.

For the record I can solve the 2x2x2 and 3x3x3 variants, I can't reliably solve the 4x4x4 and above I can only get so far.  Therein lies the limit of my understanding of the puzzle.  The thing is, there are algorithms to solve the larger cubes, which I have read and used with success but I have never practised to the point where I can commit them to memory.  I choose not to for the simple reason that I don't want to learn how to do it in this way as there's no understanding of how and why what you are doing works.  I've mentioned abstraction before in previous posts on this blog, the idea that you don't need to know how something works you need only know how to use it, I don't like this approach when learning.  For the longest time in Maths class when learning about square roots I wanted to know how the square root was actually calculated, this wasn't covered until many years into my education, when I finally learned it's simply achieved through trial and improvement using various algorithms designed to speed up the process, I felt somewhat cheated.  If I had simply been told there was no other way than to use repetition to find it, I wouldn't have placed such value on knowing how to do it by hand.

Therein lies the crux of the distinction between intelligence and knowledge - not what you know but what you do not.  To put it more bluntly, intelligence is the approach to finding information you don't know and the extent of the structure of that approach.  Something that relies on pure random attempts - a random number generator - isn't intelligent, but something that takes a methodical approach to finding that number is more intelligent.  Those last two words are quite important to reinforce "more intelligent" i.e. intelligence is not absolute, it is not a state of true or false, but rather it is a scale.  Not surprisingly there exist many systems of quantifying where you are on that scale and attempts to measure it for example the Intelligence Quotient - IQ - method of grading intelligence.

Knowing a lot does not make you intelligent, how you use what you already know and how you try to find out that which you don't is what makes you intelligent.