Thomas Kuhn coined the term “paradigm shift” in his book, The Structure of Scientific Revolutions. (Here’s a synopsis.) If you want to understand how science really works, this is the first book you should read.
Think in terms of a hierarchy of ideas. At the bottom of the hierarchy is a hypothesis: a single proposition about reality. The hypothesis should be testable and subject to disconfirmation. The next step up is a theory: a larger, more abstract idea that generates a number of testable hypotheses. A paradigm is the next step up: it is the entire set of concepts and relationships that are assumed by a theory.
For programmers, there's a weak analogy between a paradigm and a graphics library. A graphics library contains a set of methods that create images: circles, lines, rectangles, and so forth. But the graphics library assumes a set of concepts: a coordinate space, a pen with its own attributes, shapes, buffered images, and so forth. The graphics library plus its associated concepts is rather like a paradigm.
A paradigm shift is an intellectual revolution. It’s not just a step forward — it’s also the forceful rejection of the current way of thinking. Suppose, for example, that you’ve been using a standard graphics library for years and you’ve gotten pretty good with it. You’ve mastered transform matrices and transfer modes and all the other fine points of the graphics library. You’re an expert.
But now somebody comes along and invents a completely new graphics library, called “Fourier Transform of Fractal Representation of 3D Objects”. Doesn’t that sound impressive? And suppose that this new graphics library can do fabulous things, allowing you to create images far beyond anything you’ve ever created before. Doesn’t that sound scrumptious?
But wait! You’ve spent years learning the old graphics library. You’re an expert with it. People turn to you when they have questions about it. This new graphics library puts you in the same position as all the other beginners. You’ll have to start from scratch and learn everything from the ground up. And this new paradigm — well, it’s just weird! It has all sorts of bizarre concepts that confuse you mightily. Studying it makes you feel just like you did years ago when you first started learning conventional graphics libraries — and that’s a very uncomfortable feeling.
Worse, adopting this new graphics library means throwing away all the expertise you’ve built up with the conventional system. That too is a most disconcerting proposition.
For all these reasons, experienced programmers would probably reject such a new graphics paradigm. Since they’re the recognized experts, they’ll steer people away from the new-fangled idea and recommend (in their professional judgement) that everybody stick with old idea.
Of course, the youngsters don’t have anything to lose by picking up the new graphics library; they’re just as ignorant of one as of the other. So they try out the new scheme and start doing impressive things with it.
This creates serious tension within the community. The old pros are telling everybody that the new scheme sucks, while the youngsters are praising it to the skies. How is this conflict resolved?
The slow solution is for the oldsters to die out and be replaced by the youngsters. This kind of thing happens all the time, in just about every professional field. It’s part of the evolutionary process. Sometimes, though, it can take several generations, because not all the youngster embrace the new scheme, and so the process gets spread out over many decades. Evolution can be a slow process.
But sometimes the transition is much sharper and more intellectually violent. This happens when there’s an obvious and serious flaw in old system. When people believe that the old system is broken, then they’re much more willing to move on to a new system.
The classic example of this is the problem of the speed of light and the birth of the theory of special relativity. In 1865 James Clerk Maxwell published his equations for the behavior of electricity and magnetism. Maxwell’s Equations did for electricity and magnetism what Newton’s Laws did for motion: they provided a complete solution to all such activity. In fact, Maxwell’s Equations were so powerful that they “predicted” light: they showed that electric and magnetic fields could combine in such a way as to produce a wave that was, well, exactly like light. Even more astoundingly, they showed that the speed of light could be calculated from the properties of electricity and magnetism — and the result was right on the nose!
There was an even more exciting possibility: Maxwell’s Equations predicted what THE speed of light was. Now, we know that the earth orbits the sun. That means that in January it’s moving in one direction and in July it’s moving in the opposite direction. That means that, if we can very precisely measure the speed of light on earth in January and July, and compare them, we should see the difference. Even better, we should see a difference between the average speed of light on earth and the “true” speed of light, then we’ll know how the earth is moving relative to the universe.
The idea here is there is some sort of absolute, motionless frame of reference for the whole universe, and the speed of light is measured against that frame of reference. So if we find the difference between our measured speed of light and the speed calculated from Maxwell’s Equations, then we’ll know how fast the earth is moving through the universal frame of reference. Perhaps we could even use this to identify the place that is NOT moving relative to the universal frame of reference: the center of the universe!
This was exciting stuff, and in 1876 two American physicists, Michelson and Morely, devised a brilliant experiment that carried out this measurement. Guess what? The speed of light was the same in January and July!
Well, this certainly threw things into a tizzie! How could the speed of light be the same when measured from differently moving objects? I won’t go into all the details of how physicists struggled with this problem, but they flopped hither and thither for nearly 30 years before Einstein came up with the solution. His answer was simple: the speed of light is the same in every frame of reference, and from that assumption, he worked out the consequences mathematically, which were that both space and time stretched or shrank depending on how you were moving relative to another person. In other words, if two fellows in rocket ships passing each other at high speeds compared their rulers and clocks just as they whooshed past each other, they’d each conclude that the other guy’s instruments were off.
Well, this was pretty weird stuff, but it was the inevitable consequence of the stipulation that the speed of light was constant in all frames of reference.
Here’s the kicker: physicists had been wringing their hands for thirty years over the problem of the speed of light. They knew that physics was broken: it just could NOT explain the Michelson-Morely results. And for that reason, they were willing to make the paradigm shift. Einstein’s results were resisted, of course, especially by the old fogeys, but all in all special relativity were accepted with amazing speed by the community of physicists — because they perceived that physics was broken.
In order to apply these ideas to interactive storytelling, we need to ask three questions:
1. What is the current paradigm?
2. Is it perceived to be broken?
3. What’s the new paradigm?