Part IV: Freedomby David Foss May 6, 1994 Transuent Causes, 2nd Order Wills, and Doing OtherwiseWhat sort of cause am I anyway?Roberick Chisholm certainly seems to have something right when he asserts that the responsible agent must be some sort of cause of her actions. And there does appear to be some real problem in maintaining a picture of agency in which the [character] traits, desires, and motives of an agent, out of which spring her actions, find their origin in alien causes. For here, when an agent has not chosen to possess the desires, beliefs, or character which she does possess, and from which her actions do in fact arise, it would seem that she is no more responsible for the content of her action than had a foreign, hostile, and invisible force pressed her limbs into the service of some hideous deed against her will. The agent, for all her splendor, must choose to be who she is! In order that she be held accountable for the outcome of her actions, and to the extent that she may rightly be held accountable, she must be that spontaneous force which either directly affected the action in question, or brought about such tendencies in her own character that such action was unavoidable given her prior determinations of will. But this all seems a bit far fetched. Are we generally so restricted in our allocations of responsibility? I think not. There may be some such restraints, in terms of the degree to which the desire, belief, or character trait in question is in principle mailable (and perhaps, to what degree, or with what ease, it is so). But we generally do hold ourselves and others accountable for actions which clearly have been brought about by character traits (or beliefs, or desires) which we neither explicitly nor implicitly chose to have. The bigot might have been raised so as to never question or doubt his bigotry. But do we thereby excuse him of his bigotry? No. Indeed, by holding him responsible we do not so much seek to locate the cause of his bigotry in himself, as to affect in him the desire to remake himself: and indeed, to call such a desire the one he ought to have. There is, admittedly, more to the story. For now, however, these brief comments should be sufficient to suggest that Chisholm’s immanent causality may be asking us to look in the wrong place (if we are seeking the appropriate arena for attributions of moral responsibility). There is, nevertheless, a point Chisholm makes regarding the relative merits of agent versus event causality which deserves a brief comment. Agent causation, or “immanent causation,” is intuitively more immediate to our experience, and would seem therefore to be a (epistemologically) prior form of causation with respect to causation generally (including event, or transuent causation). Chisholm finds this epistemic priority (the fact that if there is any notion of causation we are entitled to believe, it either is itself, or depends upon, immanent causation) sufficient for a metaphysical priority (at least with respect to the certainty of the underlying phenomena. This is a notoriously weak argument, inferring ontological legitimacy from mere phenomenal acquaintance, and I will not get into the details here. But Chisholm does believe such a form of immanent causation entails, and must entail (in order for agents to be responsible), that the agent is (in some sense) an unmoved mover. His own attempt to rescue a notion of inclination from such a thesis is singularly unsatisfying. If the will may move without precedent or cause, even if such a power is only occasionally activated, then it would seem to entail that the will must be (in principle) undetermined entirely. The “compulsion” felt by a person’s incumbent desires and beliefs are no different than the compulsion felt by the environmental constraints (and causes) which may limit the free exercise of one’s will. In both cases, the will stands apart (or can stand apart) from the event-forces which determine the range of possible courses of action, and has the power to contravene upon them. In any case, the distinction between ‘resisting the temptation to bring about some state of affairs’ and ‘resisting the temptation to allow that state of affairs to happen’ (wherein Chisholm would like to call the failure of the former necessitating and the failure of the latter merely inclining the agent to will as she does) seems to collapse into a merely verbal distinction. The forces referred to by “temptation,” and the objects of such temptation, are not so different in kind as is needed to find an authentic freedom in the failure of either alone to determine the will. We may be merely inclined with respect to a particular desire or trait, without being thereby undetermined with respect to the sum total of our desires, beliefs, etc.. And if inclination is object-specific (i.e., the failure of compulsion at the level of a specific desire), a global necessity can not be ruled out. I may desire to take a bath, without it being the case that I do it. This hardly rescues my will from prior causes. Being “philosophical” in the face of adversity.Is the free determination of the will merely a matter of a well-ordering of an agent’s desires, wants, and beliefs? There is a school of thought, dominated by Harry Frankfurt and friends, and nicely refined by Michael Slote (even if he does ultimately find the general approach inadequate), which clearly locates freedom in the context of a broad “rationality constraint.” The will is free (or more precisely: an agent is free) if and only if there is a substantive correspondence between higher order desires and values, and first order affective desires (or values). Duress and compulsion are occasions where either one’s second order desires are directly contradictory (eg., wanting to desire X, and wanting to not desire X), or where one’s first order desire contradicts her second order desires et al (eg., wanting to not desire X, but desiring X anyway). This seems to capture much of what we have in mind by compulsion, and cases in which free will is clearly not present. Indeed, as Slote suggests, it can even explain the phenomenal dissonance a bank clerk may feel in having to give the bank robber his cash: there is a substantial conflict in terms of the rational coherence of the clerk’s second order (affective) desires. [and so on...] May 10, 1994 Fourth Assignment: Who’s fault is it anyway?
Let us not mince words. If determinism, or something like it, were true, what would be different about our world? Not very much. We would still go about using moral language, ascribing praise for laudable acts, and blame for horrid ones. Indeed, if determinism were true, and if we still go about using such language, it would seem pure nonsense to claim that the truth of determinism and ascriptions of moral worth are incompatible. If determinism is true, it has always been true, and always will be true. Similarly, we have used moral language for a very long time. Indeed, moral language is arguably the most primitive (if not also the oldest) sort of language in humans. If we have been determined to live out our lives as we have, then we have equally been determined to use our language in such a way. Moral language would be just one more bit of the deterministic world. I use the word “ought” because I must. I blame, and praise, according to the weight of history, which presses each endorsement or rebuke from my lips. What a strange world it might be! But how different would it really be from our own? There is, of course, a problem with such a response to the “threat” of determinism. For those of us who worry about the degree to which our concepts about the world coherently reflect the substance of the world, it is not at all obvious that the notion of “freedom,” a central player in our current moral discourse, can remain tenable under the “realization” of determinism’s truth. If I do not stand, in some sense, as an autonomous moral player capable of directing my activities by the intervention of reasons, and not merely by the necessities of material motions, how could I coherently hold myself accountable? How can our moral language contingently locate in me a causal role I could not but have? And worse, if determinism is true, the very notion of being guided by reasons becomes, at best, a sham. If I am necessarily driven by particular, discernible, and specifiable material causes, there is no guarantee such causes will even resemble reasons, except insofar as I merely call my causes (or the causes which communicate motion through my body and mind) “reasons” retroactively. And finally, if determinism is true, and I could not but be how and who I am (i.e., I could do or be nothing other than I do or am), and I fully recognize this fact, how could I consistently distinguish between occasions when I am praiseworthy (or blameworthy), and occasions when circumstances “excuse” my actions, whatever content they may have. In the following comments, I will attempt to reconcile these difficulties with the obvious fact that, if determinism were true, then the widespread utilization of moral language is (or has been) a necessary component of human life. This should be sufficient to show the futility in any effort to claim that a modern recognition of the truth of [psychological] determinism necessarily entails a rejection of the moral concepts of freedom, responsibility, and so on. My view, therefore, is something of a compatibilist view with respect to the modern “issue” of “freewill and determinism” (although I wish to remain generally agnostic on the matter of whether determinism is indeed true). Before I proceed any further, however, it is essential that a few preliminary definitions be made perfectly clear. A great deal of confusion has resulted from an ill-formed appreciation of the precise content of the thesis of determinism, in all of its relevant manifestations. And likewise, on the matter of our moral notions of freedom, responsibility, and such, a great deal of unnecessary hysteria has formed as the result of a general failure to appreciate the categorical distinction between reasons and causes, which divides (or in a significant manner distances) the moral from the empirical. I take [physical] determinism to be the claim that given the state of the universe S1 at time t1, and assuming the laws of nature L, we can calculate the state of the universe S2 at any other time t2, by the appropriate application of those laws upon the ordered pair <S1, |t1 - t2|>. Obviously, there are some fairly serious constraints upon this thesis which make it difficult to demonstrate (and possibly even unlikely to be true — leaving aside matters of morality). In particular, the laws of the universe must be functional. That is, given any state of the universe, application of the laws of nature will yield one state of the universe for each distinct time differential. In addition, it does not matter whether the calculations are humanly possible, but only that a true measure of any given state of the universe, coupled with the actual laws of nature, will yield absolute values for the state of the universe across any given time differential. It is important to recognize the independence between the truth of determinism and potential epistemic limitations in order that we quickly dismiss two confused (albeit unfortunately common) objections.[1] First, the fact that we may not ever be able to predict the future is no argument against the truth of the broader thesis of determinism. Second, our inability to confirm with certainty that our understanding of the laws of nature is correct, even if such an inability were absolute (i.e., if the measurements needed to confirm outright a proposed formulation of the laws were in principle beyond our capacities of discernment), it would not threaten the potential truth of the over-all point that some laws (the correct ones) do determine the dynamic “life” of the universe (given some “initial” state). Of course, this last point does raise the spectre that (as may be suggested by a glance into “chaos” theory, which is anything but non-deterministic) no amount of evidence could ever demonstrate the validity of the deterministic thesis outright. But this point is not really important here. We only need suppose that determinism is a likely picture of the way the universe works. And more particularly, for our purposes, we need only suppose that it holds true for us. That is, [psychological] determinism need only be taken to suggest that, given the state of our person M1 at any time t1, and assuming the relevant psychological or physiological laws, the state of our person M2 at any other time t2 is absolutely determinate (by the appropriate application of those laws, and so forth). And here, we need only notice that we have not got much of a reason to reject outright the possibility that such laws do exist, or that, for any given moment, a complete description could be given of our “state.” What can we say about freedom? If [psychological] determinism is true (and I have suggested that it is not obviously false), then it would seem that “freedom,” in at least one sense, cannot be true of us. If it is true that there is only one [naturally] possible course of action, only one possible way in which things will actually turn out, then it cannot be the case that anything (including, most importantly, our actions and choices) might have been any different than the way they are (or have been, or will be). It is critical to notice that this is a very narrow sense of “possible.” This is not a strictly “logical” necessity, in the sense of ‘following absolutely from the laws of nature in-themselves.’ Logical form only necessitates the absence of outright contradictions (i.e., the presence of consistency). Neither is this a strictly “moral” necessity, in the sense of ‘following absolutely from some given set of “natural” imperatives’ (perhaps derived from the nature of rationality or of personhood). Indeed, it is the rather complex phenomenon of “natural necessity,” in the same sense in which we say, “This sugar, being water soluble, and now placed into that cup (which is filled with water), shall necessarily dissolve in that water.” And this is a form of necessity which does not obviously entail any other form of necessity (although it is entailed logically from the laws of nature coupled with occurrent contingencies), without which it is no more a threat to moral judgment than to logical analysis. The fact that we often refer to occurrent conditions, when making any judgment of natural necessity, as contingent, may appear to beg the question with respect to the appropriateness of calling those aspects of the situation “possible” (in any sort of strong sense). But the traditional distinction between “necessities” and “contingencies” here is merely a reappearance of a distinction between “universals” and “particulars,” where the peculiarities of any given situation, which make it that situation, are called contingent. That is, they are contingent upon prior “contingencies” being present, and not merely determined by the structural constraints imposed by the natural laws. (Still, it might be asked, what other sort of “possibility” or “contingency” could there be?) I take moral necessity and possibility to be another matter altogether, reading each of these more closely as “obligatory” and “permissible.” But this is not being entirely fair to the potential hazard posed by natural, or psychological, determinism. In order that an act be morally necessary (obligatory), it must be materially possible (naturally contingent). At least this is the way we hope it will be. The sort of freedom that it seems we need is the capacity to act, or not to act, as we ought (in accord with moral necessity). To have this capacity, we must have the power to do, or refrain from doing, what reason commands. I suspect that it is this last point which led many thinkers, from Abelard to Chisholm, to suppose that the will (the power of doing, or refraining from doing) must intercede after the deliberative process has delivered a verdict as to the proper course of action in a given situation. In other words, the activity of the [free] will must govern, not only the enactment of one’s rational decisions, but also the ‘acceptance’ of the decisions themselves. By locating “freedom” after the process of deliberation has settled upon one recommendation, however, a serious mistake is being made as to the content of this general observation. The power to do, or not to do, as reason proposes (or as moral necessity would command), is not a power with respect to the determination of reason (or the command of moral necessity). Rather, it is a power with respect to natural or physical potentialities. It is an explicit recognition of the intersection of moral necessity, and natural or physical possibility. It is an expression of the hope that ought really does imply can. “But,” an incompatibilist will surely object, “what good does it do us to be ‘able’ to refrain from lying, where there is no contradiction with any law of nature (given our general physique, social constitution, etc.), if the material conditions of the world are such that we never find ourselves telling the truth.” Indeed, this is a mightily pitiful sense of “can,” if it may turn out that the laws of the universe (or physiology) coupled with occurrent conditions make it, absolutely, false that I ever will. If we read “false” as “[naturally] impossible” here, the absurdity of asserting any enduring or relevant sense of “can” seems clear. Daniel Dennet, in I Could Not Have Done Otherwise: So What?, proposes that the sort of “can” being thrown about here is quite irrelevant to our moral judgments, about others and about ourselves. Unfortunately, he is not so successful as he appears to think in refining the sense of possibility needed for moral coherence. Possibility, for Dennet, is not eliminated from our conceptual baggage, as a hard determinist might propose, but is restricted to matters of local indeterminacy (what he calls “pockets of local fatalism”[2]). Such a sense of possibility is needed, he admits, in order that our moral language still make sense. Unfortunately, the boundaries of locality are notoriously difficult to fix, and the criteria under which various contexts count as similar to the one at hand are exceedingly difficult to discern. Clearly, if we pay too much attention to the details of any situation, any knowledge gained by such an investigation will be utterly useless: the details will likely never recur exactly as they have before. If the purpose of an investigation into the circumstances of a particular decision or event is to generalize the phenomenon, such that we might thereafter accommodate ourselves to the real likelihood of recurrence under quite different “global” conditions, then it would seem that too much detail in our account of the phenomenon will only hopelessly restrict the generalization. So we ought to keep our eyes open to the general pattern of behavior being exhibited at any given moment, of any given agent (or event). Possibilities, which subject an agent to moral valuation, are therefore those generalizable elements of a particular occurrence or action, such that given approximately similar (or not so similar) world-conditions, the phenomenon or act will recur. But, does this adequately account for the sense in which local (not merely spatiotemporally) contingencies can rob us of our moral liberty? Even Thomas Reid recognized that we often say of a good person, or that the good person often says of herself, that she could not have done otherwise (with respect to some laudable deed), without therewith tempering our praise of her person. For Reid, this could only be correctly said of the [truly] good person figuratively. Dennet takes the stronger position that this indeed must be true, that the agent really could not have done otherwise, in many cases of actual praise. That is, in order that many cases of praise be appropriate, we would like to think that the agent (whether someone else or ourselves) could not have acted against the command of reason. We say they did it for the right reasons. The added claim that ‘they could not but do it given those reasons’ only contributes to our high regard for them. Unfortunately, Dennet is confusing moral necessity with natural necessity. This is not the same “could have done otherwise” that so tickles the fancy of “those who hold the CDO [Could have Done Otherwise] principle.” Rather, it is a recognition of what I have been calling moral necessity, or the morally obligatory. He states, seemingly quite correctly, “the general capacity to respond flexibly in such cases [where the moral imperatives strike us as absolutely clear] does not at all require that one could have done otherwise in the particular case, or in any particular case, but only that under some variations in the circumstances — the variations that matter — one would do otherwise.”[3] It should be noticed that this is a transformation of the sense of “could” very much like that proposed by George E. Moore (cf. Free Will), where it reduces to some expanded notion of “would, under relevantly similar conditions, follow the [differs] choice of the agent.” This is a reading of “could” (like Moore’s) which directly concerns, not the determinations of the will directly, but the exercise of that will upon the world. But the “could” so beloved by the proponents of CDO concerns, not only the dependence of a particular action upon given reasons and world-conditions, but the absolute (or occasional) openness of the world with respect to the determinations of reason. The world (and our physiology) must be open to the content of moral necessity in order that such necessity be found compelling, and in order that we be held accountable to it. Dennet has not dismissed the relevance of such openness. He has only renamed it ‘local indeterminacy’, or ‘dependence upon unrepeatable aspects of circumstance’. So, even in Dennet, there is a surviving sense in which it really must be true that, in order that I be “responsible” for a given act, I could have done otherwise. And this could is not an openness with respect to the content of moral necessity. It is not an openness with respect to the determinations of “reason” (whatever that faculty may be), such that I must have the capacity to “choose” other than I do, which is necessary for moral judgment. Rather, it is an openness with respect to the world, in light of reasons, which is required. It must be the case that my reasons make a difference in how I act, and what I do, in order that I may be held accountable according to reason. And this openness is clearly represented by Moore’s reading of “could.” It must be the case the had I decided or chosen otherwise, I would have done otherwise.[4] Dennet helps us realize that it is generally pointless (from a moral point of view) to ask whether we could have chosen otherwise. The match we are really concerned with is that between the determinations of reason, and the exercise of those determinations. Of course, the reasoning process is sometimes flawed, and I may not choose according to the right reasons. But all that is needed here, in order that I be blameworthy for such rational failures, is that had I reasoned correctly (and thereby would have chosen correctly), I would have acted correctly. Surely, it will be objected, that had I not the power to reason other than I do, then it would be absurd to blame me for the [rational] decisions I do make. But this is not quite right. We may temper the way in which we rebuke moral transgressions in light of the judgment that an agent could not reason correctly, as is reflected in the distinction between manslaughter and murder. But we rarely excuse an agent outright for a failure to deliberate appropriately. The sort of openness required, then, is material or natural contingency, in light of determining reasons. We need a notion of natural possibility which might allow us to coherently demand of ourselves and others, that the right reasons lead us to act. The intelligibility of such a demand, and the causal or empirical effectiveness of such “rational” decision making systems, as distinct criteria on the coherence of our moral language, each rest upon a rather incontrovertible fact: it makes sense to praise and blame. I am not uttering complete nonsense (or no more than when I utter any intelligible sentence) when I tell the child, “you ought not lie.” And it is clearly not futile to demand that I and others act from good reasons: when good reasons lead one to act, one tends to do better in the world. If determinism is true, if we have behaved the way we do by the sheer weight of history, then the notion of rationality must comply with it. It would indeed be a strange world if the way things could really be were irrational. But even if this were true, and that natural necessity did not respect the rule of reason, would we have any cause to suppose that we had a reason to abandon our moral concepts? To what would we abandon them? In order that natural necessity pose an actual threat to moral necessity (or to the rational use of moral concepts), it must be rationally ordered, in which case it would pose no threat at all, being of an lesser rational constraint than reasons themselves. |