This is a discussion of creativity. Specifically, I want to look at whether or not creativity is something that we can teach students. Are we able to "create the conditions" for original thinking and if so, what are they? I don't work in research, I am not an artist or an academic, I'm just a teacher who is trying to get the best results for my students. The (yes, pretentious) title "Research and Practice", is just meant to suggest that in writing this I'm listening hard to arguments from smart people and applying their thinking to the school I work in and my own students. That's my job right? I think any academic institution should encourage the most rigorous and thoughtful questioning of its basic premises as frequently as possible. It is the sleep of reason, as Goya showed us, that produces monsters.
So what is creativity? Yup, that could be a whole blog in itself so I'm just going to define it in as uncontroversial and straightforward was as possible. Creativity is the expression of quality, original ideas. Creativity, therefore, has two components, the appearance of an original quality idea in consciousness and then the expression of that quality original idea in some form. If the quality original idea remains unexpressed it cannot be defined as creative. Creativity is about the selection of a quality original idea and its communication. The idea part is fundamental. There can be no expression of human creativity without the appearance of an idea in the consciousness of the person being creative. Given this, no theory of how to develop creativity in students can come without understanding under what circumstances quality and original ideas appear in consciousness.
In order to understand what is meant by an idea and to see how it relates to human awareness, let's look at an example of a highly creative artificial intelligence. The computer AlphaGo taught itself to play the ancient Chinese game of Go. Go is a much more complex game than chess. There are more possible combinations in the game than there are atoms in the known universe. It is both highly intuitive and highly creative. AlphaGo taught itself to play the game from scratch by playing itself over fourteen million times and storing these scenarios in neural networks. The world-class player that AlphaGo beat described some of the moves that AlphaGo played as being some of the most highly counter-intuitive and creative moves he had ever seen. Moves which should have put AlphaGo at a disadvantage allowed the AI to win and boundaries were broken in the game.
Can we consider this truly creative? We might argue that a computer cannot "have an idea", it is "merely" a stream of decisions that lead deterministically on from one another. It could not "choose" to throw the board down in a huff and stomp off, for example, unless programmed that way. The question for us is when or perhaps how does an idea "happen" on a computer. What is the computer version of an idea interpreted as being creative? Perhaps an idea is analogous with a decision. The computer doesn't have a frisson of excitement as it realizes it's about to do something pretty bloody brilliant. There's no obvious qualitative aspect. It would seem difficult to identify or even "see" exactly when the most creative decisions the computer took came into existence. Did they arrive all at once? Could you pinpoint the blip on the screen when the idea was taken (vs when it was "executed") Where are the "edges" of the ideas of the computer? Similar arguments can be used for human creative ideas. They arrive fully formed and yet are formless. Try "seeing" the beginning of an idea; it's impossible. Creative ideas that solve problems are representations that can only ever make sense from one perspective, our own. In some important sense then, computer creativity and human creativity aren't that different. The appearance of human ideas is a combination of the past and present resulting in an emergent property we call an idea and provoking a decision - some external manifestation - to be considered truly creative.
When people talk about an "inspired moment", or a poet senses a poem coming through the wind, or the Eureka effect, what they mean is that the source of their idea is unknown. Something clicked and the idea formed or began to form. Although we might explain how the idea came to be in our consciousness in a post hoc manner, we "rationalize it" the actual mechanism via which the idea appears to us is as hidden to our own awareness as the idea itself is from everyone else. What we have is awareness of the idea and then a decision on what to do with it. There is some peculiar combination of circumstances that has caused, perhaps after a long passage of thought, some original and creative idea to come about but it never clicks along in a nice logical fashion. That distinguishes creativity from the cold application of logic to give predictable answers. Creativity involves, in the human case, a leap of understanding that appears to shortcut our conscious processing power. I would guess that there is an evolutionary reason for the accompanying frisson of excitement and that's to let us know that ¨this idea was a good one¨. The limits of our reason are such that there is an emotional early-warning system that encourages us to proceed to the expression of this idea and not to discard it. This helps overcome the fact that because we don't really know where it came from we don’t know if it is a good idea.
So what is creativity? Yup, that could be a whole blog in itself so I'm just going to define it in as uncontroversial and straightforward was as possible. Creativity is the expression of quality, original ideas. Creativity, therefore, has two components, the appearance of an original quality idea in consciousness and then the expression of that quality original idea in some form. If the quality original idea remains unexpressed it cannot be defined as creative. Creativity is about the selection of a quality original idea and its communication. The idea part is fundamental. There can be no expression of human creativity without the appearance of an idea in the consciousness of the person being creative. Given this, no theory of how to develop creativity in students can come without understanding under what circumstances quality and original ideas appear in consciousness.
In order to understand what is meant by an idea and to see how it relates to human awareness, let's look at an example of a highly creative artificial intelligence. The computer AlphaGo taught itself to play the ancient Chinese game of Go. Go is a much more complex game than chess. There are more possible combinations in the game than there are atoms in the known universe. It is both highly intuitive and highly creative. AlphaGo taught itself to play the game from scratch by playing itself over fourteen million times and storing these scenarios in neural networks. The world-class player that AlphaGo beat described some of the moves that AlphaGo played as being some of the most highly counter-intuitive and creative moves he had ever seen. Moves which should have put AlphaGo at a disadvantage allowed the AI to win and boundaries were broken in the game.
Can we consider this truly creative? We might argue that a computer cannot "have an idea", it is "merely" a stream of decisions that lead deterministically on from one another. It could not "choose" to throw the board down in a huff and stomp off, for example, unless programmed that way. The question for us is when or perhaps how does an idea "happen" on a computer. What is the computer version of an idea interpreted as being creative? Perhaps an idea is analogous with a decision. The computer doesn't have a frisson of excitement as it realizes it's about to do something pretty bloody brilliant. There's no obvious qualitative aspect. It would seem difficult to identify or even "see" exactly when the most creative decisions the computer took came into existence. Did they arrive all at once? Could you pinpoint the blip on the screen when the idea was taken (vs when it was "executed") Where are the "edges" of the ideas of the computer? Similar arguments can be used for human creative ideas. They arrive fully formed and yet are formless. Try "seeing" the beginning of an idea; it's impossible. Creative ideas that solve problems are representations that can only ever make sense from one perspective, our own. In some important sense then, computer creativity and human creativity aren't that different. The appearance of human ideas is a combination of the past and present resulting in an emergent property we call an idea and provoking a decision - some external manifestation - to be considered truly creative.
When people talk about an "inspired moment", or a poet senses a poem coming through the wind, or the Eureka effect, what they mean is that the source of their idea is unknown. Something clicked and the idea formed or began to form. Although we might explain how the idea came to be in our consciousness in a post hoc manner, we "rationalize it" the actual mechanism via which the idea appears to us is as hidden to our own awareness as the idea itself is from everyone else. What we have is awareness of the idea and then a decision on what to do with it. There is some peculiar combination of circumstances that has caused, perhaps after a long passage of thought, some original and creative idea to come about but it never clicks along in a nice logical fashion. That distinguishes creativity from the cold application of logic to give predictable answers. Creativity involves, in the human case, a leap of understanding that appears to shortcut our conscious processing power. I would guess that there is an evolutionary reason for the accompanying frisson of excitement and that's to let us know that ¨this idea was a good one¨. The limits of our reason are such that there is an emotional early-warning system that encourages us to proceed to the expression of this idea and not to discard it. This helps overcome the fact that because we don't really know where it came from we don’t know if it is a good idea.
In an important sense then, the source of our ideas is as unknown to us as Alpha Go’s ideas or decisions are unknown to itself. The Ai "arrived to" the decision in a deterministic way similar to a human. The difference then, between AI creativity and human creativity, comes with awareness. We are aware of having an idea. Being aware of something isn't the same as knowing where or how it comes about. The actual idea is the result of an unconscious merging of a multiplicity of ideas.
Before getting further into the awareness of ideas, I believe there is an analogy between the source of our ideas and the behavior of sub-atomic particles. There is a thing called "spooky action at a distance" where particles at the most minuscule level seemingly interact with each other despite being separated in space and time. Nothingness is inferred to contain more energy than matter and "dark matter" refers to an as yet unperceived field required to explain the appearance of some freaky physical phenomenon. Inter Alia, how we could have obtained something out of nothing at the beginning of time and why particles behave differently when they are not being observed. Dark Matter is a field which is inferred to explain the appearance and behavior of certain physical phenomenon in the same way that the unconscious mind is inferred to explain the presence of mental phenomena and their concomitant behaviors. This for me is another argument for saying that what is sometimes calls divine inspiration is just deterministic unconscious processes mixing and blending concepts in ways we can't know about being brought out into our awareness by some external trigger.
Conscious awareness of ideas in working memory cannot explain the source of our ideas, they must appear from "somewhere". The concept of the unconscious and memory is required to explain the fairly obvious fact that we remember things and have ideas. Here is where it gets interesting and where things are relevant for education and learning. If we cannot be aware of the source of our ideas, we can merely infer them, then we cannot explain the conditions under which those ideas come about. All we can do is control the inputs. What we might be able to do is explain how we use our awareness to slow down and modify ideas consciously. It's this ability to have some "control" over our ideas, to discard them, to act upon them that our freedom lies.
Nobody could identify with any certainty the "moment" when Alpha go took the decision to make the highly creative move. There are no "half-ideas". All of us have had the sensation of "knowing we know" something, but not being able to recall it. When the answer does come to us, it's often a little while afterwards. Creative and quality ideas don't follow a different pattern, they arrive fully formed, like sentences on the billboard of our awareness. Quality, creative ideas are complete thoughts that have quite literally been born-whole out of the emptiness of our unconscious minds. It seems to me therefore that there are two reasonably secure roots for encouraging more original and quality ideas which can be communicated effectively; training awareness and imagination and maximising conceptual inputs.
Yes, imagination is crucial to creativity, but it's limited. I can choose to imagine dragons, landscapes, smiling hordes of people. I cannot choose to imagine the solution to global warming, although I might be able to get a picture in my mind of a wind farm or some huge field of solar panels. This is because the source of my imaginings is limited to my experience. A three-year-old cannot imagine a fairy if she has never heard of a fairy. To link this to AlphaGo, AlphaGo is both highly intelligent and highly creative at Go however he cannot suddenly become a world-class painter or architect. We cannot draw into existence a creative idea we can merely control the inputs and then become aware of the appearance of the idea after an unconscious process. There is no backwards causality even if the beginning and the end of the highly creative idea appear simultaneously. AlphaGo had to play itself fourteen million times in order to remember enough different scenarios in order that the decision rules governing which moves to make were effective. What looks like intuition is millions of calculations saying if this then this happens and then, boom, the clock stops and the decision is taken. The moves of the other player stimulate the decision but the millions of board set-ups in memory are what allow neural networks to be constructed.
Experienced meditators can develop quite impressive control over their imaginative faculties. One Tibetan practice involves imagining a candle, holding that image in your mind and then freezing it, thawing it and manipulating it mentally. Doing this effectively takes years of deeply dedicated practice. Even the most experienced meditators, however, are limited to the imagining and manipulation of familiar objects by developing control over their awareness. Just the fact that something as simple as manipulating a visual object in working memory is so difficult points to the limitations of our ability to be intentional with awareness and imagination. Being aware of something normally gives highly constrained control over it. I might become aware of the presence of a spooky feeling in a room or someone's discomfort or indeed my own discomfort but I am limited in terms of what I can actually do with that awareness. Introspection and imagination are about developing some level of control over the products of our unconscious mind however they cannot in any sense overcome the limitations of what our consciousness is producing. Our awareness holds ideas, rejects some, draws from memory but it doesn't possess any magical elements which are not presented to it from the raw materials of our unconscious.
In short, the emergent representations of our minds and their underlying neurobiological processes cannot be created or destroyed but they can be changed or manipulated through our awareness and maximised by high-quality inputs. As a side note, on this point, Jung would disagree. Jung argued that humans DID contain a level of "intrapersonal" knowledge, his concept of archetypes was such that we all share a collective human bank of symbols that are stored deep within our psyches and impact on the nature of our conscious representations. Even if this was the case, it is perhaps even more of an argument for why we cannot consciously bring about original quality ideas, we can merely feed concepts into our unconscious. It is the case that highly creative people can be less knowledgeable in their field than their more expert colleagues however the argument here is that the reasons for this are unknown because we cannot control where ideas come from. We can merely control what we do with our awareness and slow down our imaginative capability to be more watchful for their appearance. Furthermore, it seems to be impossible to have any true creative breakthrough without actually having a deep, rich store of concepts which can be blended. Even if we are talking about cross-domain creativity the raw materials are so much more important than anything else if for no other reason than imagination and awareness are not, in themselves, able to impact causally on the past. Imagination and awareness work with representations yet do not cause them. Dominating in a domain, truly deeply understanding how it is connected together and engaging in some serious imaginative daydreaming about that topic strikes me as the best pathway towards enhanced creativity and not the notion of open-ended problems and group work. There is no reason to associate more conscious or deliberative thought with quality ideas or productions if there is not the rich conceptual understanding blending in the unconscious from which original and creative thoughts can appear. We know from the AlphaGo example that true creativity comes from mastery of a domain. The raw material of AI is a set of logical functions fuelled by learning and development of networks and yet the machine cannot program itself to do something different, it can merely intently master through the creation of knowledge. Creativity comes from the bottom up.
Before getting further into the awareness of ideas, I believe there is an analogy between the source of our ideas and the behavior of sub-atomic particles. There is a thing called "spooky action at a distance" where particles at the most minuscule level seemingly interact with each other despite being separated in space and time. Nothingness is inferred to contain more energy than matter and "dark matter" refers to an as yet unperceived field required to explain the appearance of some freaky physical phenomenon. Inter Alia, how we could have obtained something out of nothing at the beginning of time and why particles behave differently when they are not being observed. Dark Matter is a field which is inferred to explain the appearance and behavior of certain physical phenomenon in the same way that the unconscious mind is inferred to explain the presence of mental phenomena and their concomitant behaviors. This for me is another argument for saying that what is sometimes calls divine inspiration is just deterministic unconscious processes mixing and blending concepts in ways we can't know about being brought out into our awareness by some external trigger.
Conscious awareness of ideas in working memory cannot explain the source of our ideas, they must appear from "somewhere". The concept of the unconscious and memory is required to explain the fairly obvious fact that we remember things and have ideas. Here is where it gets interesting and where things are relevant for education and learning. If we cannot be aware of the source of our ideas, we can merely infer them, then we cannot explain the conditions under which those ideas come about. All we can do is control the inputs. What we might be able to do is explain how we use our awareness to slow down and modify ideas consciously. It's this ability to have some "control" over our ideas, to discard them, to act upon them that our freedom lies.
Nobody could identify with any certainty the "moment" when Alpha go took the decision to make the highly creative move. There are no "half-ideas". All of us have had the sensation of "knowing we know" something, but not being able to recall it. When the answer does come to us, it's often a little while afterwards. Creative and quality ideas don't follow a different pattern, they arrive fully formed, like sentences on the billboard of our awareness. Quality, creative ideas are complete thoughts that have quite literally been born-whole out of the emptiness of our unconscious minds. It seems to me therefore that there are two reasonably secure roots for encouraging more original and quality ideas which can be communicated effectively; training awareness and imagination and maximising conceptual inputs.
Yes, imagination is crucial to creativity, but it's limited. I can choose to imagine dragons, landscapes, smiling hordes of people. I cannot choose to imagine the solution to global warming, although I might be able to get a picture in my mind of a wind farm or some huge field of solar panels. This is because the source of my imaginings is limited to my experience. A three-year-old cannot imagine a fairy if she has never heard of a fairy. To link this to AlphaGo, AlphaGo is both highly intelligent and highly creative at Go however he cannot suddenly become a world-class painter or architect. We cannot draw into existence a creative idea we can merely control the inputs and then become aware of the appearance of the idea after an unconscious process. There is no backwards causality even if the beginning and the end of the highly creative idea appear simultaneously. AlphaGo had to play itself fourteen million times in order to remember enough different scenarios in order that the decision rules governing which moves to make were effective. What looks like intuition is millions of calculations saying if this then this happens and then, boom, the clock stops and the decision is taken. The moves of the other player stimulate the decision but the millions of board set-ups in memory are what allow neural networks to be constructed.
Experienced meditators can develop quite impressive control over their imaginative faculties. One Tibetan practice involves imagining a candle, holding that image in your mind and then freezing it, thawing it and manipulating it mentally. Doing this effectively takes years of deeply dedicated practice. Even the most experienced meditators, however, are limited to the imagining and manipulation of familiar objects by developing control over their awareness. Just the fact that something as simple as manipulating a visual object in working memory is so difficult points to the limitations of our ability to be intentional with awareness and imagination. Being aware of something normally gives highly constrained control over it. I might become aware of the presence of a spooky feeling in a room or someone's discomfort or indeed my own discomfort but I am limited in terms of what I can actually do with that awareness. Introspection and imagination are about developing some level of control over the products of our unconscious mind however they cannot in any sense overcome the limitations of what our consciousness is producing. Our awareness holds ideas, rejects some, draws from memory but it doesn't possess any magical elements which are not presented to it from the raw materials of our unconscious.
In short, the emergent representations of our minds and their underlying neurobiological processes cannot be created or destroyed but they can be changed or manipulated through our awareness and maximised by high-quality inputs. As a side note, on this point, Jung would disagree. Jung argued that humans DID contain a level of "intrapersonal" knowledge, his concept of archetypes was such that we all share a collective human bank of symbols that are stored deep within our psyches and impact on the nature of our conscious representations. Even if this was the case, it is perhaps even more of an argument for why we cannot consciously bring about original quality ideas, we can merely feed concepts into our unconscious. It is the case that highly creative people can be less knowledgeable in their field than their more expert colleagues however the argument here is that the reasons for this are unknown because we cannot control where ideas come from. We can merely control what we do with our awareness and slow down our imaginative capability to be more watchful for their appearance. Furthermore, it seems to be impossible to have any true creative breakthrough without actually having a deep, rich store of concepts which can be blended. Even if we are talking about cross-domain creativity the raw materials are so much more important than anything else if for no other reason than imagination and awareness are not, in themselves, able to impact causally on the past. Imagination and awareness work with representations yet do not cause them. Dominating in a domain, truly deeply understanding how it is connected together and engaging in some serious imaginative daydreaming about that topic strikes me as the best pathway towards enhanced creativity and not the notion of open-ended problems and group work. There is no reason to associate more conscious or deliberative thought with quality ideas or productions if there is not the rich conceptual understanding blending in the unconscious from which original and creative thoughts can appear. We know from the AlphaGo example that true creativity comes from mastery of a domain. The raw material of AI is a set of logical functions fuelled by learning and development of networks and yet the machine cannot program itself to do something different, it can merely intently master through the creation of knowledge. Creativity comes from the bottom up.
We hear a lot about how we need to give open-ended problems and get kids to put their brains to use in working out the very best solutions in order to generate a kind of creative mindset. Here's the thing, doing this fulfils none of the criteria for developing creativity spoken about above. Domain-specific creativity such as the Go example requires a level of mastery of the domain however even creativity which crosses domains would seemingly require a high level of technical knowledge in at least one domain to bring value to interdisciplinary collaboration. One cannot get better at being interdisciplinary, one can get better at a domain, or several domains, and then work on connections. Without the domain, there can be no connections. The raw materials of AI are a set of logical functions, essentially yes or no answers, a binary system. Very simple rules. From these decisions and decision rules that are modified due to feedback, there is emergent intelligence. There is the ability to have "insight" into something. Nothing mystical, it's a cold hard calculation of all the options based on having run them through yourself. AlphaGo played itself 14 million times. That would take a lifetime for a human. The point is, the goal is very specific - learn how to play go - and the inputs are pretty well controlled and yet great creativity is possible. Open-ended problems don't cause improved thinking skills, expertise and memory encourage creative problem-solving. Millions of scenarios were run over in infinite speed and the best was chosen based on running through what would happen different scenarios given the best choices of opponent. The more we specialise the greater the chances of breaking new ground are and most importantly the more we will be able to see possibilities in the other domains that are important to us.
I think this is "scientifically" accurate: "Yes, imagination is crucial to creativity, but it's limited."
ReplyDeleteConsider this "thought experiment": everything subtracted from everything will amount to nothing. Now count how many "complete annihilation attempts" have been successful throughout the observed universe. In other words, the limit of our "exclusive predictions" is "always" inferior to that of our "additional observations". So, identifying the first law of thermodynamics (as Schrodinger and Einstein do in their equations): the subject of our predictions (fiction) cannot be the object of our observations (as conserved).
The "moral" of our stories, then, remains: "working together disproves individual, cultural and economic solipsisms" In other words: general contrasting predictions (identifying the first law of thermodynamics) originate by predicting the future in general (as multiplies itself in contrast to being created at some point in the past and destroyed at some point in the future). physicswithoutborders.org
In "more" rather than "never" practical terms (explaining Cantor's diagonal argument): exclusive predictions direct additional observations (proof):
--> one exclusive point will avoid another exclusive point, excluding zero points <--