Will and the world: A study in metaphysics

CRITICAL STUDY WILL AND THE WORLD: A STUDY IN METAPHYSICS by N. M. L. Nathan. Oxford: Clarendon Press, 1992. 168 + x pages. In this ambitious and densely argued book, N. L. M. Nathan defends a general theory about t h e nature of metaphysics and metaphysical questions (in chapters 1, 2 and 10) and then applies the theory to two metaphysical questions about the will--its existence (in the form of the existence of "real volitions" (chapters 8 and 9)) and its freedom (chapters 3-7). Nathan argues that many metaphysical questions can be understood as conflicts between what persons want or desire and what they believe. Thus, they may want to possess a certain kind of freedom or free will, but believe they do no possess it because they believe in some form of determinism or fatalism, which seems to rule it out. Or they may want there to be objective moral values or certain foundations for knowledge or non-material selves (a real external world or a provident God), but believe, for one reason or another, that these things do not exist. Such want-belief conflicts, as Nathan calls them, may be first- or second-order. The conflict is first-order when something is desired (say, the existence of an objective good), but the agent believes it does not exist. The conflict is second-order when the agent wants to believe in something, but believes the evidence is insufficient for believing it. Wants and beliefs conflict in both cases, but in the first case the belief directly denies the existence of what is wanted, while in the second case the conflicting belief is that there is insufficient evidence for believing what one wants to believe. Resolution of the conflict in either case involves forsaking the want, or the belief which conflicts with it. If we look at metaphysical questions in this way, Nathan thinks we would be led to ask certain questions in the attempt to resolve 523 CR1TICALS'IUDY them. First, our attention would be drawn to the content of the desire or want that generates the conflict. Does that desire have a coherent content? Does the free will or objective goodness or disembodied ego we want to believe in make any sense? Second, we should ask why the disputed content is wanted. Do we want it for its own sake or for the sake of something else? If we want, say, free will for its own sake, then what kind of intrinsic value does it possess? If we want it for the sake of something else (say, moral responsibility), is it really a necessary condition for that other thing and does that other thing have intrinsic value? Third, we must ask whether the evidence for the existence of what is wanted really does tell against it or is insufficient to support it. Fourth, and finally, he says, "if it has somehow been decided that the conflict should not be resolved, then the victim's awareness that there are disadvantages in things being as he wants may at least diminish the force of his desire." (p. 20) In such a case, we look for what Nathan calls "palliatives"--something else the person may want to be the case that could be the case if the original disputed want there fulfilled. My first reaction to this interesting account of metaphysical questions was to think that, while it helps us to understand many traditional metaphysical conflicts, it does not apply to all of them. For example, metaphysical speculation can sometimes be provoked by conflict between different beliefs where the enquirer does not have a vested interest in the truth of either belief. This might be the case, for example, of Kant's first two antinomies, of Zeno's problems about the continuum, or various paradoxes, like the liar, the sorites, Prisoner's Dilemma or Newcomb's paradox. But Nathan accounts for this kind of objection in the final chapter where he sums up his view. Metaphysical speculation, he concedes, may be provoked by beliefbelief conflicts as well as by want-belief conflicts. But he argues that want-belief conflicts are the more "irksome." Or incompatible beliefs, we know that only one can be true. "But there is no such assurance in the case of wanting p to be true and belieing it false. Each attitude can be perfectly reasonable, and their joint possession a permanent affliction." (p. 161) Thus, want-belief conflicts turn out to be crucial for the construction of what Nathan calls "self-critical world views." A 524 CRITICALSTUDY world-view describes what somebody wants for its own sake and contains statements and resolve belie-belief and want-belief conflicts. He concludes that "every task traditionaly called philosophical finds its place somewhere in the construction or application of a selfcritical world-view." (p. 163) This broader view of the metaphysical enterprise--presented in the final chapter of the book--is more plausible than the original thesis that metaphysics concerns want-belief conflicts; and the broader view has much to recommend it. It shows why want-belief conflicts are crucial to metaphysics without being the whole of it. The view may still seem to leave no room for an utterly "dispassionate" search for truth. But I don't think this is the case; for in such a dispassionate search the desideratum of truth becomes the only thing wanted for its own sake in one's world view. Nathan's view also helps to explain the apparent unresolvability of many philosophical disputes. Conflicts over world views are not only about the facts but also about what is wanted (and what is worth wanting) for its own sake, something that the facts of the case may not settle. For all these reasons I think Nathan has given us an interesting and provocative view of the metaphysical enterprise that rings true in many ways and deserves to be widely discussed by philosophers. Less successful, it seems to me, is his treatment of the free w i l l issue. To be sure, Nathan's chief interest is not in resolving this perennial issue, but in showing how it illustrates his general view of metaphysics. And, in fact, I think that traditional debates about free will do support his view of metaphysics in many respects, though I also think that his account does not do justice to some features of these debates. Nathan's discussion of free will is dense, and while it contains many insightful arguments, the plethora of distinctions sometimes obscures central issues, let me illustrate with several examples. Of the many kinds of freedom Nathan distinguishes, two are most important for contemporary debates about free will. The first of this is what he calls conditional--or C-freedon--the kind of freedom favored by many compatibilists--according to which the power to choose otherwise is spelled out in terms of hypothetical propositions of the form "if the agent had wanted (or desired, or believed...) 525 CRITICALSTUDY otherwise, he would have chosen otherwise." Such a freedom, Nathan acknowledges, would be compatible with determinism. The second important freedom is what he calls "*-freedom": " N ' s decision at t to A was *-free if and only if (i) it was either C-free (or free in another sense distinguished by Nathan) and (ii) had N decided not to A, the state of the world just before that decision would have been exactly as it actually was." (p. 31) By virtue of clause (ii) *-freedom is incompatible with determinism, according to Nathan, and would be kind of freedom favored by incompatibilists. In accordance with his general approach to metaphysical questions, Nathan asks of each of these freedoms why in fact we want, or should want, a freedom of this kind. The question is indeed central, but in the case of both freedoms, it seems to me, his answer leaves something to be desired. Regarding (compatibilist) C-freedom, he says that people may want it because knowledge of conditionals of the form "If N had wanted... X, N would have decided to do Y" helps us to predict human behavior. "Since we are liable to want to predict human decisions, we are liable to want those conditionals to be true, knowledge of which helps us to make predictions." (p. 38) But, while we may in fact want such predictive powers, they are scarcely the only, or even the most important, reasons why people have wanted Cfreedoms. Conditional freedoms are favored by compatibilists because such freedoms represent the absence of various impediments to free activity such as physical constraint, coercion, psychological compulsion and political oppression. The attraction of freedom from "constraints" of all these kinds has been the chief motivator of conditional accounts of freedom for modem compatibilists since Hobbes and Hume. Prediction has played a role, but only a subordinate one. Of more importance is Nathan's account of why we want (incompatibilist) *-freedom. For it is *-freedom, he argues, that creates the basic want-belief conflicts concerning free will. Cfreedoms are wanted, but it is also reasonable to believe that we sometimes have them. By contrast, many have doubted the very possiblity of coherence of *-freedoms. But the more immediate question, from Nathan's point of view, is why we want *-freedom in the first place. To explain this, he focuses on what P. F. Strawson has 526 CRITICALSTUDY called the "reactive attitudes" (gratitude, resentment, admiration, and the like). Using gratitude as an example, Nathan asks us to imagine that we are grateful for another person's decision, but later find out that the person could not have done otherwise. Our attitude would change, he thinks; we would be less grateful. But what does "could have done otherwise" mean in this context? He thinks it must have the meaning of *-freedom. For "suppose that, full of gratitude for N's decision to A, you then come to believe that although he could in some sense have decided not to A, he could not, as things were just before his decision to A, have decided not to A. Then your attitude alters. You are less glad that the decision has been taken." (p. 46) As Nathan sees it, the italicized clause expresses what is required by *freedom and explains both why *-freedom is wanted and why it is incompatible with determinism. But his intuitions on these points are by no means universal and require further support. First, many philosophers would insist that gratitude and other reactive attitudes do require that an agent could have done otherwise, it is far from clear that "could have done otherwise" must be interpreted in a manner incompatible with determinism, such as *-freedom. What is so important, many would ask, about the requirement that, bad the agent done otherwise, "the state of the world just before.., would have been exactly as it actually was"--such that our gratitude (or resentment) should hang in the balance? This requirement of *-freedom is abstract and by itself provides few clues as to why anyone should want a freedom satisfying it. To unearth the proper clues, I think one has to dig deeper into the motivational roots of free will beliefs than Nathan does in his discussion of *-freedom. Earlier in his discussion, he passes over several ideas that seem to me essential for understanding what lies behind the intuition about gratitude. One of these ideas is the notion of "sole authorship," which Nathan attributes to me, (p. 36) Sole authorship implies, as he says, the absence of any control over the agent's choices by other agencies, whether the control is overt or covert, constraining or non-constraining. But, on my view of it, sole authorship also means, in a broader sense, that the ultimate or final explanation for an agent's decision is to be found in the agent's will 527 CRITICALSTUDY alone, and in no one and nothing else. If the explanation for the choice could be traced to something else beyond the agent's will, then the responsibility would devolve to that other thing and the agent would not be ultimately responsible for the choice. (That is why all sorts of control by other agencies, overt or covert, constraining or nonconstraining, are excluded by sole authorship.) Nathan suggests that whatever virtues may lie in such a notion of sole authorship would be captured by his account of "could have done otherwise" in terms of *-freedom; and so he quickly passes over sole authorship to get to a discussion of *-freedom. My own view, to the contrary, is that the idea of sole authorship is more fundamental for understanding incompatibilists intuitions than *-freedom or even "could have done otherwise." If we w a n t freedom in an incompatibilist sense, I think it is because we want to be ultimately responsible in the sense required by sole authorship for at least some of the choices and actions through which we shape our own characters and motives (and this same requirement lies, I think, behind incompatibilists intuitions about the reactive attitudes, like gratitude). Such authorship confers on us the dignity of being to some degree original creators of our own destinies rather than subjects determined by forces or influences beyond our control. In other words, it is sole authorship which explains the importance of *freedom (and its requirement of indeterminism), not the other way around. To be sure, sole authorship and ultimate responsibility are elusive notions, which conjure up images of "prime movers unmoved," "noumenal selves," "transempirical egos," and other obscure forms of agency or causation that have been part of traditional accounts of indeterminist free will. There is no guarantee that sole authorship and ultimate responsibility are attainable or even intelligible, however desirable they may appear to the untutored imagination. In Nathan's terms, if we want such things, our wants may lack a "coherent content." Nathan argues that his *-freedom is a coherent content and might exist for all we know, since the deterministic theses that oppose it have not been shown to be true. (Chapters 4 and 5) He also argues that undetermined choices need not be random, irrational or motiveless, as is often asserted. I think these claims are right; yet 528 CRITICALS12JDY there are deeper problems about the coherence or intelligibility of undetermined free choices which he does not discuss, and which would have come to the surface if more attention had been given to sole authorship and ultimate responsibility. I also think these deeper problems can be solved and have spent the better part of the past twenty years trying to make sense of notions like sole authorship and ultimate responsibility without appealing to obscure or mysterious agencies or causation (noumenal selves, etc.) that have been the stock in trade of traditional libertarian views of freedom. The point I am making here is simply that Nathan's passing so quickly over the notion of sole authorship in order to get to *-freedom renders his account of free will importantly incomplete. This failing, however, does not invalidate the claim that most concerns him: the free will issue does exemplify to a significant degree Nathan's account of metaphysical enquiry in terms of wantbelief conflicts and self critical world views. Free will, in the sense of sole authorship and ultimate responsibility, is something many persons want, but there are also many reasons to believe that they do not possess such a power, or cannot possess it, or even that it is not intelligible. Such reasons have led many philosophers today to give up on free will in the strong incompatibilist sense and settle for less problematic compatibilist accounts of freedom and responsibility. In short, the relevant want has been extinguished. To his credit, Nathan (though he does not take a firm stand one way or the other) does not think the game is over yet regarding this particular want-belief conflict; and I agree. There is still some juice left in the venerable issue of free will, and still the responsibility that the beliefs against it, and not the want for it, will be extinguished. In addition to the discussions of metaphysical questions and free will, Nathan includes, in chapters 8 and 9, challenging defense of volitions (against criticisms or Ryle and others). He argues for the existence or real volitions, thereby completing the discussion of metaphysical questions concerning the existence of the will and its freedom, and shows how debates about the reality of volitions also fit the pattern of want-belief conflict. In summary, this is a book willworth reading, especially for its account of metaphysical enquiry in terms of want-belief conflicts arts self-critical world views, which is, 529 CRITICALSTUDY I think, an original and significant contribution to current debates about metaphysics, and about the the nature of philosophy generally. ROBERT KANE THE UNIVERSITY OF TEXAS AT AUSTIN AUSTIN, TEXAS 78712 U.S.A. 530