Computing The Creative Mind: How Margaret Boden Sails, then Scales, the Psyche
Computing The Creative Mind: How Margaret Boden Sails, then Scales, the Psyche
I originally chose to read Boden's book, The Creative Mind: Myths and Mechanisms, because I am drawn to the idea of creativity -- and because I am drawn to the elusiveness of its definition. What is creativity? Doesn't everyone define it differently, based on individual experience, shared cultural patterns, brain waves, and random phenomena? Why do we value creativity and is it biologically, genetically useful and productive? Is one called creative in comparison to others or based on his or her own merits and own judgment system? These questions are so potent in today's society, in every discipline and profession, that I want to know how they apply to science. In the context of this class, I'm excited about the possibility of tying creativity to biology, of placing creative behavior in a brain-based context or, conversely, showing how strange brain patterns, mistakes, or distinctions can account for outwardly creative behavior or even for the evaluation or appreciation of creativity. (Do different brains think different things are creative?)
Boden's book immediately situates itself in the realm of science, looking at creativity in a way that somehow seems to defy categorization. She does not talk only about MRIs or EKGs, nor does she only write about Byron or the sublime. Right from the start, she explains why "creativity" is a sticky term. Whether or not something is considered creative often depends on whether or not it is considered new, but Boden points out that newness is itself hard to define, since something can be new for the person conceiving of it without being new to the whole society (2). Boden further addresses the idea of newness by talking about types of creativity. Creativity is often seen as an unfamiliar combination of existing ideas. In order to appreciate that something is creative, she explains, someone has to be familiar with the ideas that are being put together and to recognize that the way they are being put together is new (3). Boden branches out from this conventional view of creativity by highlighting two other types: exploratory and transformative creativity, the former being a demonstration of new possibilities within an existing type or structure of thought, the latter being a type of idea that pushes the boundaries of a type of thought or creates an entirely new structure for thinking (4-5).
By clearly introducing several ideas for what makes something creative, Boden immediately catches my eye. She's right: we can think of creativity lots of ways, and these three are pretty useful. To lay them out right off the bat helps show that Boden is not interested in the simple idea that creativity exists -- she assumes we know that. Instead, she's going a step further, typing biology to behavior by asking how creativity relates not just to our brain but to the idea of a brain in general. The concept of artificial intelligence becomes, for Boden, not just a neat sci-fi exploration but instead a way of probing ideas about creativity and creative technology.
First, Boden explains that artificial intelligence provides some ways of showing how our brains might be structured -- linguistically, socially, grammatically. If exploring and transgressing existing structures of thought are two major ways humans can be creative, then we need to think about what those structures of thoughts are and, more important still, what creates and destroys them (can that creation be replicated?). If we can transform our own structures of thought, can the machines we make go a step further, changing themselves -- being creative even without our planning for that creativity? Boden introduces these questions (9), raising the idea of technological genealogy and the evolution of creativity.
Throughout the book, Boden introduces a variety of ideas about structured thought and how we have tried, over the years, to represent it. From semantics (107) to Euclydian geometry (120) and highly structured poetry (170), she demonstrates that humans have tried to map their own creativity throughout history -- that structuring structures of thought has been a major structure of thought! By placing ideas about structured thought in history, Boden shows how they have been useful previously as well as how they can be useful in new fields. She differentiates, on many an occasion, between what she calls H-creativity and P-creativity (43, 233, et al.) -- historical vs. psychological creativity, the former referring to ideas that are absolutely, historically new and the latter to ideas new (only?) to the person thinking of them. She points out that, in the grand scheme of things, historical creativity is much harder to predict, since it depends less on patterns among people and the way they think and depends more on patterns of large societies and on the evolution of all creative thought (233). How predictable is any unknown idea? she causes the reader to ask.
Boden looks at unpredictability and predictability in many contexts in order to show that just understanding our own human creativity is a large task. From associative memory and coincidence (233), to varying definitions of randomness (240) and what she calls the anti-creative idea of chaos (237), Boden shows that what we mean when we say "unusual" or "different" or "unpredictable" really varies depending not only on what we are referring to or who is talking but also on the type of language we are given and are using.
Having shown that humans are evidently quite imperfect when it comes to describing their own structures of thought and the ways those structures account for and are affected by creativity, Boden goes a step further and explains why humans' inadequacy at explaining their own creativity has led them to look at other means, other answers, other systems. With the "advance of scientific theory being matched by the retreat of anthropocentrism" (277), she says, humans have learned to test and re-test creativity, to find ways of trying to explain it -- and in many ways they have succeeded in showing that creativity exists. But, Boden explains, the idea of creativity is so tied up with the idea of wonder that many people fear that too much scientific treatment of it will end up categorizing or defining it too clearly, destroying their sense of wonder (278). As science has expanded, however, it has provided ways to increase, not destroy wonder; Boden points out that sometimes even a seemingly too-simple tool, like the compass, can provide wonder (280), and that wonder certainly does not have to be accompanied by confusion.
Beyond accepting the idea of trying to map creativity at all, Boden points out other obstacles to a strong bond between creativity and science. For one, she explains that the idea of uncertainty, unpredictability, and randomness is not unfounded or useless, even if it cannot be easily defined or pinpointed. Even if we can learn more about ourselves by trying to replicate our minds using computers, can we ever really expect a computer to replicate us appropriately? Can a computer ever really be creative, if it doesn't have the wide variety of input that humans do -- the everyday world, all the stimuli we receive, all our memories and thoughts and associations? Boden accepts this perceived limitation and simply explains that we have to be as creative as possible in creating creativity. The more concepts, conceptual spaces, and ways of maneuvering about (or changing, transforming, transversing) those spaces, the higher the potential for creativity a computer will have (294)!
But how -- she foresees critics asking -- can a computer achieve consciousness, even with lots of built-in ideas and idea-forming tools, if it is not really conscious the way a human is, if it doesn't really know that it is being creative? Boden again invokes the idea that "conscious" is itself a multivaried term (295), the definition of which cannot be easily decided upon and therefore cannot be relied upon as a variable in determining the potential for creativity. A creative system, Boden suggests, "must be able to ask, and answer, questions about its own ideas" (295). Computer programs, she says, are built on the idea of asking and verifying programs' ideas. As long as we program them well, then, and give them the capacity to test themselves and their environment, can't computers form a type of intelligence, a creative intelligence?
I'm very intrigued by Boden's book, which shows not only how and why computers might be able to be intelligent but also the huge implications of calling a computer intelligent or of defining creativity as something other than only human. After reading the book, I simply want to ask more questions. If we replicate creativity, wouldn't that mean owning it and destroying it? Would that destroy its wonder (its magic), or would it simply help destroy a silly, primitive, overly simplistic definition of wonder and help pave the way for new creations and new types of creativity? As I ask questions, I notice that I am thinking of ideas. I do not know if they are historically new, but I recognize their psychological uniqueness to me, and I acknowledge, thanks to Boden, that in asking questions about my own ideas I am functioning like a creative being. I then wonder whether all thought is creative -- or only new thoughts, only the thoughts that change us, only the thoughts that acknowledge their role as thoughts? Does creativity itself even merit such high regard, or is it just another structured system that we create in order to explain ourselves?
Boden, Margaret A. The Creative Mind: Myths and Mechanisms. 1990. 2nd edition. NY: Routledge, 2004.