Margaret BodenMargaret Boden's 2006 two-volume Mind As Machine is an extraordinary history of cognitive science, a field, she says, that
would be better defined as the study of "mind as machine".Boden documents the transition in psychology from the behaviorism of B. F. Skinner (in which the mind was treated as a "black box" or a "blank slate" on which environmental stimuli make their impressions - behaviors being the mechanical result of some responses to stimuli being reinforced) to the new field of cognitive science in the 1960's. Cognitive scientists explore the mind as containing parts that can be analyzed as logical systems much like the parts of digital computers. This is a very powerful metaphor, which Boden captures in her title, but man is not a machine in the Newtonian sense, as the reductionist behaviorists assumed, and the mind is not a computer, although like a computer, it is an information processing system which acquires, creates, stores, and manages the information needed to guide the actions of its body. Machines that compute and communicate provide fruitful analogies and metaphors that help us to understand the mind. In her 1990 book The Creative Mind, Boden explored the conflict between physical determinism and creativity or the freedom that creativity implicitly requires.
[Freedom] can be touched by science without being destroyed by it. A computational psychology can allow that much human action is self-generated and self-determined, involving deliberate choices grounded in personal loyalties and/or moral principles. Indeed, it shows us how free action is possible at all. To a significant extent, arguments defending freedom against science parallel arguments pitting creativity against a scientific psychology. Worries about predictability and determinism, for example, crop up constantly in such discussions. In either case, pure indeterminism gives mere chaos. To choose one's actions freely is not to hand responsibility over to the unpredictable...And in either case, predictability may be a virtue. One would not go for advice on a moral problem to someone whose judgments are not usually reliable, nor admire someone whose good deeds were always grounded in passing whims.On the other hand, Boden sees room for random events as bringing us thoughts, supporting the idea that our thoughts come to us, but our actions come from us.
Thoughts triggered by chance events, in the environment or within the brain, may be fruitfully integrated into the psychological structures of the mind concerned.Boden distinguishes "absolute randomness" (such as quantum physics provides) from mere "relative randomness" which may be predictable but is adequately random in some context (for example, pseudo-random numbers). She is not convinced of quantum mechanics and its absolute chance. But she is optimistic that her "A-randomness" could be harnessed by creative computers. [Indeed, the latest Intel computer processing chips provide a purely random source of noise.]
What is useful for creativity in minds and evolution is useful for creative computers too. A convincing computer model of creativity would need some capacity for making random associations and/or transformations. Its randomizing procedures might be A-random; for example, its instructions or associations might sometimes be chosen by reference to lists of random numbers. But they need not be: R-randomness would do. Indeed, some creative programs rely on random numbers at certain points, and genetic algorithms can produce order out of chaos. Moreover, some computer models spend their 'spare' time searching for analogies in a relatively unconstrained way. This computerized R-randomness could exist alongside more systematic (and somewhat more reliable) 'rules' for generating useful ideas.But Boden wonders whether absolute randomness exists and implies absolute unpredictability.
An event that is unpredictable in the strongest sense is absolutely unpredictable ('A-unpredictable'), unforeseeable in principle because it is subject to no laws or determining conditions whatever. In other words, A-unpredictability (like A-randomness) implies indeterminism of the most fundamental kind. Whether there are any genuinely A-unpredictable events is disputed. According to quantum physics, there are. Quantum physics claims that some physical events, such as an electron 'jumping' from one energy-level to another, are uncaused and therefore (at that level of description) unpredictable. The electron jumps this way rather than that for no reason at all. In the terminology introduced above, each individual electron-jump is A/E-random (it has no order, and no explanation).Boden took up the subject of free will in her very first publication (Mind, 1959, v. 68, pp.257-8) However, quantum physics also claims that large classes of supposedly A-random events are neither A-random nor E-random, and are in practice predictable. These large classes of sub-atomic events are not A-random, because they show order in the form of statistical regularities. They are not E-random either, because these regularities can be explained by the wave-equations of quantum physics. Moreover, these equations enable the physicist to make precise predictions about the long-term behaviour (the statistical distributions) of the relevant physical systems. Some people argue that quantum physics must be either incomplete or mistaken, because (as Albert Einstein put it) 'God does not play dice'. In reply, quantum physicists may admit that quantum theory could conceivably be mistaken, although they insist that only an irrational prejudice in favour of determinism would make anyone think so. However, they refuse to admit the possibility of its being incomplete in the way the determinist wants. They cite a mathematical proof that quantum theory cannot possibly be extended by adding hidden variables which obey non-statistical laws, because such extensions must alter the experimental predictions in specific ways. (Since this proof was offered, in the mid-1960s, almost all of the experimental tests have come out in favour of quantum theory.) In other words, if quantum theory is correct then the A-unpredictability of quantal events is indeed absolute. For our purposes, it does not matter whether quantum physics is correct or not. Granted, there may be quantum effects in the brain, triggering some of the ideas that enter the mind 'at random'. If so, then A-unpredictable individual quantum jumps might conceivably contribute to creativity (as other R-random events can do). But this does not put creativity 'outside science', any more than X-rays are. In short, quantum physics illustrates one of the ways in which unpredictability (even A-unpredictability) is not opposed to science.
Upholders of a free-will doctrine, rather than determinists, would, I think, want to claim a special sort of certainty in relation to decision-statements. They might not be prepared to say that we have free-will at all times, but would at least claim that there are many times when we can, by a free choice, direct our actions as we decide. Now if free-will is to be claimed at all, it is necessary that, at times when we exercise our free choice, we should do so consciously, and that we should have direct and incorrigible knowledge of what we are doing, in order that we may be held fully responsible for our voluntary acts. But in order to claim this incorrigible knowledge, and also to claim absolute certainty that our "act of will" will be followed by the appropriate voluntary act, a free-will theorist will have to distinguish between the voluntary instigation of a chain of events likely to lead to the desired result, X, and the later part of the chain, which is controlled causally. It is only over the first step of the chain that we can have control.In her massive Mind As Machine, Boden says that "cognitive science showed a way out of the impasse" between determinism and indeterminism, "by sidelining the determinism/
indeterminism issue and focusing instead on the huge cognitive and motivational differences between humans, dogs, and crickets."
John Eccles, for example, posited mysterious quantum phenomena acting on "critically poised neurones" in the brain. Why such phenomena are absent from the brains of dogs and crickets — or, for that matter, of human babies (who aren't yet free agents) — wasn't discussed. (Nor is this discussed when similar arguments are put forward today). Eccles's neurological hypothesis found few followers. But even if they found his "explanation" implausible, many people agreed with him that it's obvious that free choice is incompatible with determinism. The trouble was that the opposite seemed just as "obvious" to others. Compatibilists argued persuasively that indeterminism at the origin of action would destroy human freedom, because it would make it impossible to ascribe moral responsibility to people (e.g. Hobart 1934).But Boden goes on to approve Before the rise of cognitive science, this issue [freedom] was normally put in terms of determinism versus indeterminism. So the neuroscientist Daniel Dennett's two-stage model for free will. We have found that Dennett's decision model, and his six basic reasons for supporting it, are more cogent arguments for free will than any libertarian thinker has developed! Unfortunately, Dennett is a determinist and compatibilist who does not subscribe to his own arguments. Boden writes that
Dennett developed [his idea] in his essay 'On Giving Libertarians What They Say They Want' (1978b: 286-99), and especially in his book Elbow Room: The Varieties of Free Will Worth Wanting (1984a). This book was a philosophical analysis, deepened by a cognitive scientist's appreciation of computational architecture. (Many years later, Dennett would return to the topic — now, with a focus on the evolution of the relevant architecture: Dennett 2003a.) It provided a host of examples illustrating the psychological processes that underlie our freedom of action, and various (normal or pathological) restrictions on it. It made clear that only creatures with a certain type of cognitive–motivational complexity are capable of choosing freely. Or, rather, to choose freely just is to exercise those architectural features. In other words, determinism/indeterminism is largely a red herring. Although (Dennett argued) there must be some element of indeterminism, this can't occur at the point of decision — for that would give us a type of "free will" that's not one worth wanting. The indeterminism, then, affects the considerations that arise during deliberation. The person may or may not think of x, or be reminded (perhaps by some environmental contingency) of y — where x and y include both facts and values. The deliberation itself, by contrast, involves deterministic processes of selection, weighting, and choice. (Deterministic, but not universalist: moral principles, cultural conventions, and individual preferences are all involved.)