Wednesday, April 23, 2008

Creatively Speaking

Somewhere in my youth, I came into possession of a very simplified model of the human brain. I'm not sure from where, it was so long ago its roots were lost in time.

The idea is simple, the brain consist of three, and only three things. First is memory. We can memorize stuff. By that, I mean that we take some fact, associate it with a number of keys and then, on demand, we can return the fact as needed. It is committed to memory. It becomes a piece of knowledge, something we know. I'll quietly ignore long-term vs. short-term memory for this model.

The next thing is understanding. That is, we understand something. For this, I mean that we can take some set of facts that we learned, and at the appropriate time use them for some piece of higher reasoning. If you've just memorized a fact, that's nice, but until you understand it, you can't really take advantage of it. In some sense, understanding is really just how that set of facts is linked to other sets of facts that can in some way drive or change your behavior. You 'know' what an stove burner is (fact), and you 'understand' that once its turned on, it is hot and will burn you.

The very last thing is creativity. This, conceptually is a little more complicated. If a fact is a thing you remember, and understanding is how it is linked, creativity in some offbeat way is how you can create 'dynamic' linkages between non-related facts or understandings. You might have a pencil for instance, and you know how to use it to draw shapes and textures, but at the time you are working you're feeling blue, which is some intangible emotion. If you want to use the pencil and your ability to draw to express your emotion state by drawing a dark moody image of yourself, you are taking your pencil, and your self-view, both facts, applying your understanding of drawing and then 'combining' that with your current emotional state to 'create' something that on its own would not come into existence for anyone else. You are creatively expressing your emotions through a self-portrait with a moody overtone. The work is not reproducible.

Creativity then is this ability to combine two entirely unrelated things, in arbitrary, random, or mostly irrational ways. The base things are either facts or understandings.


A consequence of this simplified model is that the abilities of a human on one level are quite simple. We build up knowledge -- linking it to other bits -- in ways that we find useful. Indirectly, as if it were a computer program run amok, one can see creativity as some type of inherent 'flaw' in our brains. After all, connecting two things that "shouldn't" be connected is not rational or deterministic. It's like a computer programming randomly changing the data, it is outside of the norm. The brain shorts out, or something.

From a natural point of view, this model allows one to see human progress often as this long history of people gradually creating new ideas and memes by either observing them in nature or randomly combining together un-related bits. In a direct sense, all of our of knowledge beyond what is quantifiable, is accidental. We stumble into discovering it.

That's a little hard for many people to swallow, so it likely shows that this model is far to simplified to be absolutely correct, but still quite interesting. To get to the really high concepts, they have to be based on lower ones, that somewhere down the way have to be based on observable things, or creative leaps.

I like this model because it provides nice tidy explanations for many things. For instance, why somebody with 'photographic' memory does so well in their early university courses, but then has trouble later on; they never really understood the original stuff, just memorized it. The latter courses require understanding as well as memorization. You can't build on a foundation that you don't have.

Intelligence is not just our ability to remember facts, understanding and creativity play a huge role as well. So that means a brilliant person is one, then who has a lot of facts, and understands them well. While a genius is an intensely creative person, who many not actually sit on that big a reservoir of understandings or facts; but they can just combine what little they have into interesting new bits. There are a lot of brilliant people, while creativity, is a rarer thing.

This model harmonizes 'intelligence' and shows that one cannot easily compare their three values against other people's three values. With three dimensions, you cannot easily rank your intelligence. Often you can meet people who can quote tremendous amounts of knowledge, but seem to have trouble really putting the pieces together. Your typical absent minded professor may be incredible forgetful, but full of deep understandings. It is a great perspective on our abilities.


Now, all of that being interesting, what does it have to do with computer science? I read a excellent write up recently of someone talking about how they believe that math is actually creative. It is know as Lockhart's Lament and it is at:

This to me this is a hugely misunderstood characterization of what it means to be creative, and of what mathematics is, and why it is so special. In a way, relating it back towards creativity takes away from the true underlying awe and beauty of mathematics. It trivializes it.

In a very real sense, mathematics is an infinite set of axioms over an abstract space. It does not exist in our real world and because of its rigour, which could not exist either in this world, it is purely abstract. Math is sort of this pure crystallize abstract structure that we can perceive if we concentrate on it long enough. It exists, but without substance. A truly amazing thing.

As such, all math exists and has always existed. It always will, with or without us.

The branches, axioms, primitives etc. of the various different types of math are our expressions of the that abstract foundation. It is out there, right now. All of our existing math -- a drop in the overall bucket -- and all of the stuff we still do not know. Mathematicians set out on a quest to 'find' new mathematics. They find it, prove it is rigorous, and then explain it to the rest of us.

In that sense, mathematics as we know it, is a huge set of facts and a significant amount of understanding about those facts. In its abstractness, it is consistent, and contained. It is not contaminated by the real world, where entropy and chaos dominate. It exists in a pure state, perfect, rigorous and eternal.

Knowledge and understanding then easily define our existing view of mathematics. They exist, so there is little that is inherently creative in their existence. Learning mathematics is not a creative exercise. You study the existing works, memorize the facts and work on problems until the understanding of the facts kicks in. If you are being led down the garden path, then as exciting as it is, learning about some new set of axioms and how to use them really isn't your creative spark, is it?

If learning mathematics is not particularly creative, then the big question must be one about 'finding' the ideas initially; whether or not that is a really 'creative' endeavour?


To digress just a little bit. Software development is in its very own way a form of expressing a set of instructions to a computer that in its abstractness and rigour are similar to the act of creating a mathematical proof. Expression, be it about what you had for lunch, the weather or some specific algorithm for searching is a form of communication. The underlying topic may change, but choosing English, French, C or Java for your expression is still a way of communicating something. For English, it is non-rigorous and it goes to another human being. For C, it is very rigorous and it goes to a computer, but if both are about algorithms, then they are just different ways of expressing the same underlying thing. We only choose to express vastly different things in the different mediums.

There is, of course, a question about the expression of intangible quantities. It is one of those that Richard Gabriel in his book "Patterns of Software" says architect Christopher Alexander called: "the quality without a name". We have to categorize the world into those things that are determinable, tangible in some way, and those that are not. Tangible usually means, some form of physical existence. For the intangible, mostly it comes to us, personally in some what that is spiritual to some degree. By that I do not evoke any higher meaning, other than to describe 'spiritual' as a mood that is specified by humans. It is one of our many emotions. In Gabriel's book he talks about the feelings one gets when they go into these huge architectural wonders, often churches, but not always. It moves people in some intangible way. That unnameable quality then to my mind is just another intangible human emotion. Perhaps awe, or something similar.

That, fortunately makes for a nice and tidy taxonomy, since I've often described fine art -- for example -- as having an embedded emotional context, which is what separates it from merely being graphic design. If it stirs you in some way, it is art. If it is just pretty, then it is only graphic design. The same is true for music, it moves you or it is 'pop' music.

What we can perceive with our intelligence is tangible, and what we can experience with our emotions is intangible.

And so, in my earlier example with the drawing, the self portrait combines two very distinct things, one quite tangible, the representation of a human -- with the other, which is intangible -- the emotion of feeling somewhat blue. The creative 'short' or spark that drives that is not natural to most people, so although afterwards we (some of us) can see what the artist was getting at, and are moved by it. Although left in the same circumstances, we -- not being that specific artist -- would not have been creative in that same way.


Coming back to mathematics: we know that the math is out there, and that it is abstract, yet not intangible. It is a thing, and if you search hard enough, eventually you will come upon it. I am not saying that in any way it is easy, or that it would take humanity anything less than thousands of years of culture before they might stumble on a complexity concept like negative numbers, which is exactly what happened. It also took a long time for us to figure out zero, complex numbers, calculus, etc. Some humans clearly can get there faster, but the road that needs to be traveled needs to be traveled by all, collectively.

We could not have complex branches of mathematics for instance without having an understanding of negative numbers. The intuitive leap that one and only one person can make, can only be so far. We are limited by our life spans. The imaginary of standing on shoulders of giants is exactly correct if you realize that the giants themselves are only just other people standing on other people's shoulders. And on it goes.

We get left then with the understanding that finding new forms of mathematics is a search, and it is one that comes from a limited distance from our last existing searches. With each new generation we can cover new ground, but only just barely. In that, although for one human, they learn the existing knowledge and then the existing understandings, they can only start searching from that point on. To them it is an open un-explored field with hidden surprises, to humanity however, it is just part of an ever going permutation through all of the localized space relative to the last set of solved problems. So, in the end, the best that a mathematician can do -- in that sense -- is to find or get to their discovery before the hoard of people following them do. Speed is important, but so is being in the right place at the right time, as the field is constantly filling up with others overturning the same rocks and prodding the same bushes too. If you don't discover it, they eventually will.

All of this really comes back to whether or not the idea of someone prodding around in a big field looking for interesting bits of mathematics is in fact being creative or not. And, to some very large degree, if mathematicians are creative in what they find, then by virtual of their similarity, computer scientists -- at least for the very first time they are writing something really unknown -- are equally as creative.

As if you needed more twists, we still need to try and relate this situation back to something more physical.

The simplest mapping would be to look at a graphic artist. They are by virtual of their training able to produce things that are visually appealing. Yet, as we are awash in pretty graphic design, you know that their work is not stirring in any fashion.

It looks nice, but it doesn't not give rise to emotions. That's true because while they have taken their facts, and taken their understanding -- in this case abilities and skills at things like drawing or matching pleasing colors -- they have not crossed the gulf and gone on to combine these with some intangible emotional context or any other thing that is out of the ordinary. Their skills may be beyond what some in the 17th century would have been able to do -- appearing back then to be 'art' -- but here and now, in this time they are applying the things they learned in school to a specific end. Understanding, yes, creativity, probably not.


So going back to our mathematicians, as they ponder and search throughout the field, are they really being creative, or are they not just applying the skills thy were taught towards the goal of identifying previously unknown bits of mathematics? There may, I admit be some truly creative spark there that helps particularly gifted mathematicians to get to a specific point in the field faster than others and allows them to identify their thing rapidly. How else, can you explain Einstein? Although his leaps where related to a specific science, not general abstract mathematics, the distance that he appears to have travelled vs. his contemporaries at the time seems to be enormous. But again, what he found was there plainly for all to see, if they wanted to, or were willing to look. Relativity always was, and always will be. You just have to want to see it. But still, the leap was big and it took twenty years for his fellows to catch up with him and get to that particular place.

And so, for me, I'd have to say that most of the leaps that are going in mathematics are not leaps, but a natural progression of steps. And that these steps are an expected consequence of the knowledge and understanding that come from studying the field and really understanding it. That sometimes, but very rarely, a few make creative'ish giant leaps, but in general people are just working towards the goals of their field. If any one famous mathematician had died early, it may have taken us longer to get to their understanding, but we would still get there. The results might be slightly different, possibly composed of a different set of primitives, but it will still be functionally equivalent. As mathematicians create proofs, so too do computer programmers create code. It is exceptionally less complex, and far more mundane, but still shares the same type of relationship.

The code exists, we just have to find that sequence of instructions that expresses the solution to the problem in the most elegant fashion. So, we have to ask again: is programming really creative? The answer would have to be rarely. Only at that moment where we are combining two unrelated things in a truly unique way are we actually allowing our brains to take creative leaps. Otherwise we are just applying our understanding of how to code.

If your just doing another day's worth of problem solving, then those are exactly the skills programmer's should have and were taught. Software design has creative moments, and often the design is left unspecified so that some of the effort is required by the programmers, but mostly the business domain, technical domain and best practices remove a significant amount of the degrees of freedom. The interfaces share the same creative tones as graphic design. Most of it is really the application of an understanding.

Is software development art? Never. There is virtually no place in software for the expression of intangible emotions, I can barely imagine what that might look like. Tools are just tools. But, to slightly confuse the issue, the content of the programming could be art. That is, you could create a masterpiece of art using code as the means of expression. Games designers and many of the early pioneers of tools like flash succeeded or came close.

But really, a tool that moves you to tears is somewhat counter-productive, yes? It would be difficult to be both stirring and functional. Although, I do have to admit that some big company software products have caused me to cry, but really, that is only due to dealing with their crap, and my inability to be able to find a better replacement. The code did not stir the emotions, the angst of dealing with it did.