For a topic once immortalized in Milton's oft-quoted words (Paradise Lost, Book 2):

Others apart sat on a hill retired,
In thoughts more elevate, and reasoned high
Of Providence, Foreknowledge, Will, and Fate–
Fixed fate, free will, foreknowledge absolute,
And found no end, in wandering mazes lost.

philosophical discussions of freedom and determinism are these days among the more boring and off-putting. I am afraid that this collection does not do much to make them more approachable.  Its contents seem mostly to be revised versions of papers given at the Inland Northwest Philosophy Conference in 2001, though only one contribution has taken the trouble to recognize or respond to points in the other discussions.  The editors offer a brief introduction that may do something to explain the present state of affairs.  They first set out the standard positions given in introductory texts: incompatibilism, the view that determinism conflicts with the possibility of our being free in whatever way is necessary for us to be responsible for our actions or decisions; compatibilism, the view that determinism does not so conflict.  On the incompatibilist side of the fence there are the libertarians who go on to affirm that we are (at least sometimes) contra-causally free and the 'hard' determinists who go on to affirm the general truth of determinism and thus the non-existence of any freedom or responsibility.  On the compatibilist side there are 'soft' determinists who try to give an account of what human freedom amounts to that is consistent with the truth of universal determinism. 

None of these options has won anything like universal acceptance and there does not seem much new to say about them.  As the editors note, several assumptions generally accepted by all players have now been shown to be dubious, if not simply false, so a lot of the discussion has shifted to these issues.  One is the status of determinism itself.  Notoriously quantum mechanics is often taken to assert a fundamental kind of indeterminacy.  In the present volume, John Earman, our man in CERN or rather wherever theoretical physics is practiced, offers a survey of the relations between different physical theories and determinism.  Most readers will, I suspect, have to take most of what he says on faith.  One curious claim is that Newtonian theory is less hospitable to determinism than, say, special relativity.  I am inclined to say, tell that to Laplace, but I am also prepared to believe him.  What is perhaps a pity is that his strictly theoretical discussion does not address the question (raised in Honderich's contribution) of the relationship between the commitments of fundamental physical theory and those of our more homely psychological, sociological, etc. thought (at that level, no one would be happy to say, oh well, Bush's election, or his decision to invade Iraq, just happened without any causal conditions).  Nor does he address the pragmatic issue raised by Roberto Torretti (The Philosophy of Physics, Cambridge University Press, 1999, p. 143):

The existence of deterministic developments with free beginnings is taken for granted in the daily practice of laboratory physics.  The experimenter sets up and sets going of her own free will an almost closed physical system whose evolution is governed to a good approximation by this or that system of differential equations, and lets it run until she intervenes, again freely, to practice some measurement or to wilfully alter its course.

Another traditional assumption that has come under sustained pressure is the principle of alternative possibilities: the idea that if I am to be morally responsible for doing X it must be possible for me to do something other than X.  In 1969 Harry Frankfurt offered thought experiments in which this principle seems falsified and they and it have been the focus of much discussion ever since.  In this volume several contributors accept them as refuting the thesis, while Peter van Inwagen, in a curiously Caesarean style, suggests that it may be meaningless.  He does, however, offer other principles that are in the same ball-park, though he hopes invulnerable to the sort of counterexample Frankfurt offered.  (If you are annoyed that I have not indicated what sort of case Frankfurt used, you will share my annoyance at van Inwagen's own way of proceeding in discussing MacKay and Johnson's counterexamples to one of his own proposals [pp. 223-4].)  So let me sketch Frankfurt's idea:  A is contemplating whether to do X or Y.  B wants A to do X and is in a position to ensure that, however A's deliberation proceeds, A will decide to do X and do it (the mechanism here is of course science fiction).  A decides to do X and does it.  A is, we think (don't we?), morally responsible for doing X, but as described, A could not have not done X; it was not possible for A to do Y. 

Personally I am not sure what to think about Frankfurt cases.  Perhaps they show the principle is false; perhaps they reveal our unpreparedness to consider nonexistent procedures of decision-implanting.  If you like playing with science fiction cases, Ishtiyaque Haji has a paper here exploring several variations on a case of someone who sets up a bomb that will go off at some later time.  The overall aim of the paper is to replace our usual assumption that we can only blame people for what they are doing or have done with prospective blame or responsibility.  Haji notes that this idea requires a different conception of responsibility: a self-disclosure view.  I intend to water the garden when I get home and thereby reveal what I think of concerns to preserve a scarce resource, so Haji can blame me now, before I have spilled a drop of water on my garden. 

Todd Long has a paper that might address my concern about nonexistent mechanisms.  He notes that the usual Frankfurt story uses two possible mechanisms: A's usual deliberative processes and B's science-fiction way of making A decide to do X.  Long's idea is that he can tell Frankfurt-type stories in which every option involves only A's usual deliberative processes.  In effect he raises the question of what conditions we want to impose upon the inputs to deliberation in order that it or its outcome in action be morally responsible.  (Or simply free — I do not understand the hang-up about moral responsibility.  Torretti's experimenter is acting freely, whether or not any moral issues arise in her experimenting.)

There is another venerable tradition which avoids all these worries by saying simply that acting freely is a matter of doing what you want to do, or some more complicated story about wants, desires or preferences.  This type of view is represented here by Keith Lehrer's paper.  Unfortunately, while the initial idea is attractive, the convolutions that are now apparently necessary (Lehrer speaks of a preference structure you prefer to have and that you have because you prefer to have it, and would have had a different one if you had preferred to have that [cf. p. 54]) may make one wonder whether mere mortals ever act freely.

As the editors note, recent discussion has been much influenced by van Inwagen's arguments for incompatibilism: the idea that I have a choice, or that I can do X and I can do Y (where Y is incompatible with X), seems to require a kind of openness in the future that determinism rules out; if it is true that I can do Y which I am not doing then I can somehow change the past or the laws of nature, which of course I can't, so I can't do Y (this is not van Inwagen, just my gloss).  John Perry has a clear discussion of some of the issues here.  He faults van Inwagen for in effect equating 'can do X' with 'does X' — which is certainly not how the informal logic of 'can' behaves.  Perry argues that what he calls a weak account of 'can' eludes van Inwagen's argument, and that a weak account of laws of nature (roughly the traditional understanding of Hume's view that laws simply express constant conjunctions) would also do so.  Perry, like many others, wants his laws to be a bit more meaty than a weak theory would permit (but his own talk of structural principles doesn't seem to require any more meat — why should I think an account of structure has more umph than a contingent universal truth?) so he opts for the weak reading of 'can'.  His weak reading amounts to saying that I can do X if I have the ability to do X, and that in turn amounts to my having in my repertoire of basic actions some movement such that executing it will bring about X; I can of course have such a stock of abilities without exercising them (cf. p. 245).  By contrast, Perry's strong reading of 'can' requires that if I can do X at a particular time then it wasn't settled or fixed at any earlier time whether I would do X (p. 241), which demands the falsity of determinism to be applicable.

Another clear and focused paper is Richard Feldman's comparison of 'contextualist' accounts of knowledge and freedom.  Roughly, the contextualist about knowledge says that in some contexts we operate with one set of standards for what to count as knowledge while in others, and in particular in a philosophy seminar, we operate with other more stringent standards.  So ordinarily I know that I am typing this review on a computer in my office, but when an epistemology student comes in and we start reflecting on skeptical possibilities I no longer know it, since I can't rebut the bizarre and usually unconsidered possibility that I am really a brain in a vat, or whatever.  Applied to freedom the corresponding thesis would be that ordinarily we only take certain causal factors into consideration when wondering whether someone is acting freely.  So once we have checked that no one is pointing a gun at me, or has given me a post-hypnotic suggestion to do X, and have ruled out a few other possible factors, we grant that I am free in doing X.  But of course we haven't looked at every causal factor that relates to my doing X, in particular internal neurological features. But if we were to do that, we would conclude that I am not free.  While the idea of there being various fields with respect to which we judge of the existence or otherwise of obstacles to our doing things seems to me defensible, Feldman wants to reject contextualism in both forms: he offers alternative explanations of the kind of linguistic data that have been cited in support of it, and notes that it concedes too much to skepticism/incompatibilism.  His major objection, however, is that it does not address what he thinks the main issues: do we meet the usual standards for knowledge? What conditions are sufficient for free action?

An assumption underlying most discussions is explicitly denied in Honderich's somewhat disjointed paper: that we have a unitary conception of freedom.  Rather, Honderich claims, we have both a compatibilist conception (exemplified in our faltering attempts to make punishment rehabilitate criminals) and an incompatibilist conception (agents as first originators of their actions, involved in our retributive thinking, for instance).  So any attempt to say that our conception of freedom is so-and-so is mistaken since we have two distinct conceptions.  One problem then is which one fits the facts?  Honderich is clear that it is the compatibilist picture, so our next problem is what to do about the incompatibilist conception and the attitudes it imbues.  Can we somehow excise them from our lives and our normal reactions to people?  Here he offers two paths.  The modest one is to be optimistic and hope that we can; the ambitious one is to suggest that radically new conceptions of consciousness, causation, and explanation will shed light where now there is only obscurity.

There are several other papers here that I have not mentioned, including a curious exploration of Buddhist views (by Nicholas Gier and Paul Kjelberg) that made me aware that there are claims in that tradition that bear on our topic but that I am still unable to articulate.  Robert Kane has a paper on libertarian theories of free will; Carl Ginet one on trying to act; Dana Nelkin on the sense of freedom; Nomy Arpaly on varieties of autonomy; and John Fischer on the transfer of nonresponsibility.  They are all competent and pertinent for those wandering in these mazes; but the common reader had better pass this volume by.

© 2007 Ed Brandon

Ed Brandon is, by training, a philosopher, and now is working in a policy position in the University of the West Indies at its Cave Hill Campus in Barbados.

Categories: Philosophical