Abstract
This paper examines two arguments against the possibility of moral responsibility—the first directly from the work of Galen Strawson and the next inspired by Strawson’s argument. Both of these arguments are found wanting, and their shortcomings are used as a springboard to sketch a positive libertarian view of moral responsibility and defend that view against preliminary objections.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
‘Creatures like human beings’ here is meant to signify creatures that are psychologically like human beings in some important respects: the most notable are that they have only finite psychological histories; they have a variety of beliefs and desires; and they act for reasons.
Strawson has defended the argument I am concerned with in a number of different places, most notably in Strawson (1986, 1994, 2002). (It has also been discussed sympathetically by philosophers such as Saul Smilansky. See Smilansky (2000).) I will focus primarily on the version from Strawson (1986). Although it is the oldest formulation, it is many ways the most nuanced and detailed.
Strawson (2002), p. 451, emphasis in original. Clarke (2005) takes Strawson to task for supposing that an adequate theory of ultimate responsibility would have to justify potentially infinite rewards or punishments for human action. But we can see the point of Strawson’s remarks without having to follow him down the heaven and hell road, of course. As he queries earlier in the same article (p. 441), “are [humans] ever responsible for their actions in such a way that they are, without any sort of qualification, morally deserving of praise or blame or punishment or reward for them?” It is this emphasis on responsibility without qualification—not on infinite rewards or punishments—that is the real purpose of Strawson’s heaven and hell remark. (Strawson is thinking of heaven and hell as places where the only thing that plays a role in fixing reward and punishment is what one deserves as a result of what one is responsible for.)
This is a point made in a slightly different context by Mele (1995), pp. 223–224.
Strawson (1986), p. 30. Future references to this source will be parenthetical in the text.
If Strawson takes himself to be giving an analysis of a folk concept (as the previous passage seems to imply), the plausibility of that analysis is rendered suspect by recent findings in empirical philosophy, most notably those described in Nahmias et al. (2005). These surprising findings—although preliminary—do strongly suggest that folk conceptions of free will and responsibility are much more amenable to compatibilist analyses than has typically been supposed. For suggestive contrary findings, see Nichols and Knobe (2007).
I understand simple incompatibilism to be the view (roughly) that freedom and responsibility require simple indeterministic causation of actions by previous beliefs or desires—that is, causation that does not invoke any sort of agent. But deterministic causation by the same entities would prevent freedom and responsibility. A number of responses to Strawson have depended on a significant “watering down” of his notion of responsibility in either compatibilist or simple incompatibilist directions, including the ones in Mele (1995), Fischer (2006), and Clarke (2005), as well as—on certain interpretations—Clarke (1997). The same goes for newer work such as Vargas (2013) (see, e.g., p. 8) and Boxer (2013). This watering down may be appropriate—if, for instance, it turns out that Strawson has an overly ambitious understanding of what moral responsibility would require, as Boxer contends—but for my purposes I will work with Strawson’s own demanding characterization. Any weakening of the requirements would be good news for the defender of the possibility of responsibility, but since I believe no such weakening is needed, I will not help myself to it.
In this latter horn of the dilemma, A’s causing the formation of the intention is not ultimately to be understood in terms of some reason(s) causing A to form intention J. Rather, the agent causation involved is meant to be at least roughly similar to the views proposed in O’Connor (2000) or (2005).
Claiming that (3) is a metaphysically necessary may be controversial, in which case one could easily weaken the argument to accommodate the possibility that a backwardly immortal human being (or a human being capable of making infinitely many decisions in a finite interval of time) could complete the regress successfully. This is such a distant possibility (and moreover one that is of no theoretical interest to anyone presently engaged in the debate) that the damage to the heart of the argument would be negligible.
I also assume that the motive for an intention is not included in the intention. So, for instance, if I form the intention to eat because of my desire for food, the intention is an intention to eat simpliciter which is motivated by a desire for food. It is not an intention to eat-for-the-sake-of-satisfying-a-desire-for-food.
For the time being, I will remain agnostic as to the range of mental states that can play this role, though certain kinds of desires plainly can. I will also—again for the time being—leave the notion of a MOTIVE as a theoretical primitive. Relatedly, a potential motivating reason is a mental state that is capable of playing the role of a motivating reason, but may or may not actually do so.
This appears to be the case most clearly in Bernstein (2005). Bernstein implies that the challenge posed by Strawson’s argument is to make sense of true responsibility without appealing to occult entities like Cartesian egos, noumenal selves, or agent causation. But really, the challenge is meant to be broader; the challenge is meant to be making sense of true responsibility even helping oneself to these kinds of entities. As Strawson repeatedly emphasizes (not only in Strawson (1986)), his argument is meant to be purely a priori, and arguments based on empirical unrespectability are assuredly not purely a priori. Even analyses as crisp as Mele (1995) or Clarke (2005) seem to ignore this other fork. In all fairness to these critics, though, in some moods Strawson is much less interested in this other horn of the dilemma, in Strawson (2002) even offering rough sketches of arguments against the coherence of agent causation. And in other moods he makes it sound as though, at bottom, the other fork is going to be dealt with in pretty much the same way as the more well-trodden reason-cause fork. (This occurs even in Strawson (1986)—see p. 40 for instance).
Self-determination is, by definition, what is required in action for true responsibility (26).
Motivating reasons for an intention here are presumably closed under the motivation relation. So, e.g., if R is a motivating reason for J, and Q is a motivating reason for the intention to adopt R as a potential motivating reason, then Q is a motivating reason for J. This is how there could be further motivations for J beyond R. (It is not wise to suppose that the selection of R in forming intention J requires an intention to select R, since this would presumably require the selection of further motivating principles. This could generate an infinite regress and challenge the coherence of the agent causation picture sketched here, irrespective of any concerns about its relationship to responsibility. But I see no obvious conceptual difficulty in supposing that agents can select motivating reasons directly—which in turn directly generate intentions—without having to form further intentions to select those motivating reasons.)
We know that such an agent cannot self-determine with respect to any further motivating reasons because of regress worries, any more than it can self-determine with respect to the initial reasons. Consequently, it is useless to posit their existence if we are trying to understand how an agent could be truly responsible.
I have done my best here, but as I mentioned it is not entirely clear which of the concepts that figure in this argument are stipulated and which are not. (The only completely obvious case is SELF-DETERMINATION, which is stipulated. Based on the textual evidence discussed previously, it is pretty certain that TRUE RESPONSIBILITY is not stipulated, or at least not intended to be. Moreover, if TRUE RESPONSIBILITY were stipulated, the argument would lose most of its interest.) If I am wrong in my interpretation of Strawson and (3′) or (4′) contain substantive claims, my criticisms can easily be reformulated to address whichever one of these premises is then relevant.
For the original discussion of this sort of worry, see Carroll (1895).
There is also a burden of proof issue. It is often supposed that individuals trying to establish the falsity of an existential modal claim have the burden of proof over those trying to establish its truth, because of the vastness of modal space. In other words, philosophers of this mindset think that it seems like a more sensible methodology in typical circumstances to require someone arguing that something is impossible to establish that in none of the vast array of possible worlds do we find the entity in question, rather than require his opponent to establish that somewhere in the vast array we do find such a thing. (On this point in a somewhat different context, see Lycan (2003).) If this an appropriate way to look at the situation, it provides a further problem for Strawson, since he is the one aiming to establish the universally quantified negation (i.e., that necessarily for all x such that x is an agent cause who selects motivating reasons, it is not the case that x is truly responsible). Strawson must then provide a more detailed argument.
Smilansky (2000), p. 66.
Strawson (2002), p. 457.
See also his example from the footnote on p. 40.
On this and related points, see Schapiro (2009). Schapiro thinks of the relevant sort of desire (what she calls “inclination”) as having an animal pull on us, as calling us the way an imperative calls us rather than a normative claim. This premise can also be defended on views of desire more extremely anti-rationalist than Schapiro’s, though there are serious theoretical problems with those views. For an alternative view of desire (incompatible with both the soundness of this argument and the truth of the particular theory I sketch in response to it), see Scanlon (1998).
O’Connor (2000), p. 67.
Whether normative beliefs could provide their own kind of non-animal causal “oomph” is a difficult question. If they could, then the agent causation theorist could take completely at face value the idea that agents always function as partially but not wholly moved movers; she could then get away with avoiding the claim that agents can function as unmoved movers, where “unmoved mover” has the sense of something that generates all the causal oomph without any contribution from its motivating reasons. (I do caution against believing in “unmoved movers” in another sense, where being an unmoved mover implies that an agent acts unmotivated—i.e., without motivating reasons. That is not a consequence of the picture I have sketched, though. There is a difference between an agent providing all the causal oomph and an agent acting without a motivating reason.)
Other beliefs can, of course, instrumentally extend these normative beliefs in the same way that beliefs can instrumentally extend desires. So, for instance, I could believe that causing another person extreme pain is wrong except in the most unusual circumstances, and this fundamental motivating reason could be extended by a belief that if I hit John over the head with a lead pipe, I would cause him pain, and that these would not be the sorts of unusual circumstances referenced in my normative belief.
The indeterministic causation in question here is notoriously difficult to precisely contrast with inclination (as it figures in agent causation), but there is nonetheless a very clear intuitive difference. (A first pass would be to say that x indeterministically causes y in the relevant sense only if x is the sole causal contributor to the bringing about of y, but this of course ignores background factors that could be described as causal contributors. What we need is an adequate way to distinguish background factors from causal ones in order to make a suggestion like this work.)
At least in his 1986 formulation of the argument, which has been my central focus here. At other points, he has been more dismissive of agent causation, sometimes even offering brief arguments against its coherence (e.g., in Strawson (2002)), though at no point suggesting that his overall case against responsibility in any way hinges on the claim that agent causation is incoherent. In any event, these arguments are very cursory, and (coming from a philosopher as skilled as Strawson) could not seriously have pretenses to engage with the arguments from the literature for the view’s coherence.
O’Connor (2000) does sympathetically discuss the prospects for developing an emergentist view that fits a bit better with naturalistic worldviews than (e.g.) a Cartesian dualist picture does, but he is still quite far from achieving what would be considered ontological respectability in naturalistic circles.
INCLINATION here is a fundamental notion, and should not be confused with what people mean when they use ‘inclination’ as a synonym for a specific kind of desire. What I mean by ‘inclination’ in this context is a fundamental ability that a potential motivating reason has to make itself seem attractive to the agent. (This ability is at least partly independent of the content of the desire or belief, if beliefs have the power to incline.) When an agent is inclined in this sense toward a potential motivating reason, this results in the selection of that motivating reason being made more probable than it otherwise would be.
It is important to be careful here, though. The normative belief cannot merely instrumentally extend the desire—this would make only the desire a motivating reason in my sense. (Only the fundamental motivators are motivating reasons; instrumental extenders do not count.)
Though any justifiers that are empirical are probably susceptible to similar skeptical worries.
There is a similar sort of worry that is sometimes raised about instrumental reasoning. (See, for instance, Dreier (1997) and Railton (1997).) The same basic kind of response applies there—what is before the agent (in typical cases) is the selection of a motivating reason from among a set of potential motivating reasons. Often, no potential motivating reason that is available would generate the intention to check on the justification of one’s practical reasoning, nor require such a thing as part of respecting one’s normative representation(s).
It is perhaps a matter of some controversy among agent causation theorists whether selection happens at the level of motivating reason or directly at the level of intention. For a picture suggestive of the intention alternative, see O’Connor (2005). O’Connor seems to prefer the direct intention picture because he is worried about agents being able to stipulate, for all intents and purposes, that their real motivating reasons were the most noble ones available. There may be other ways to handle this potential problem, though.
It should be clear already, but it is worth emphasizing that what I have been calling “motivating reasons” are motivating reasons at a fundamental level. So, e.g., if the overall structure of my motivation is a desire for X and a belief that Y is the most effective way to achieve X, only the desire functions as a motivating reason in my sense.
Interestingly, though, it may not be essential that these desires suggest a substantively different action (where actions are individuated here at the level of intention), just that the desire present a feasible alternative motivation for the formation of the same intention.
Although the view sketched here is not widely held or discussed in contemporary philosophy, it is by no means original in its broad structure. Campbell (1967) sketches a roughly similar view (though without really engaging the impossibilist arguments), and numerous historical figures have done the same. (While there may be superficial similarities to Kant’s views—at least in certain of Kant’s moods—I do think this sort of view has marked differences from the ones Kant defends. These differences are not merely differences of detail.)
There are also important questions about the contents of motivating reasons and their relationship to the contents of the beliefs that instrumentally extend them, as well as to the contents of the intentions they lead to.
References
Bernstein, M. (2005). Can we ever be really, truly, ultimately, free? Midwest Studies In Philosophy, 29, 1–12.
Bonjour, L. (1985). The structure of empirical knowledge. Cambridge: Harvard University Press.
Boxer, K. E. (2013). Rethinking responsibility. Oxford: Oxford University Press.
Campbell, C. A. (1967). In Defence of free will with other philosophical essays. London: George Allen & Unwin.
Carroll, L. (1895). What the Tortoise said to Achilles. Mind, 4, 278–280.
Clarke, R. (1997). On the possibility of rational free action. Philosophical Studies, 88, 37–57.
Clarke, R. (2005). On an argument for the impossibility of moral responsibility. Midwest Studies in Philosophy, 29, 13–24.
Devries, R., & Triplett, T. (2000). Knowledge, mind, and the given. Indianapolis: Hackett.
Dreier, J. (1997). Humean doubts about categorical imperatives. In G. Cullity & B. Gaut (Eds.), Ethics and practical reasoning (pp. 81–100). Oxford: Clarendon Press.
Fischer, J. (2006). The cards that are dealt you. The Journal of Ethics, 10, 107–129.
Lycan, W. (2003). Free will and the burden of proof. In A. O’Hear (Ed.), Minds and persons (pp. 107–122). Cambridge: Cambridge University Press.
Mele, A. (1995). Autonomous agents: From self-control to autonomy. New York: Oxford University Press.
Nahmias, E., Morris, S., Nadelhoffer, T., & Turner, J. (2005). Surveying freedom: Folk intuitions about free will and moral responsibility. Philosophical Psychology, 18, 561–584.
Nichols, S., & Knobe, J. (2007). Moral responsibility and determinism: The cognitive science of folk intuitions. Nous, 41, 663–685.
O’Connor, T. (2000). Persons and causes: The metaphysics of free will. New York: Oxford University Press.
O’Connor, T. (2005). Freedom with a human face. Midwest Studies in Philosophy, 29, 207–227.
Railton, P. (1997). On the hypothetical and the non-hypothetical in reasoning about belief and action. In G. Cullity & B. Gaut (Eds.), Ethics and practical reasoning (pp. 53–79). Oxford: Clarendon Press.
Scanlon, T. (1998). What we owe to each other. Cambridge: Harvard University Press.
Schapiro, T. (2009). The nature of inclination. Ethics, 119, 229–256.
Sellars, W., 1956. Empiricism and the philosophy of mind. In Herbert Feigl and Michael Scriven (Eds.), Minnesota Studies in the philosophy of science, Volume I (pp. 253–329). Minneapolis: University of Minnesota Press.
Smilansky, S. (2000). Free will and illusion. Oxford: Oxford University Press.
Strawson, G. (1986). Freedom and Belief. Oxford: Clarendon Press.
Strawson, G. (1994). The impossibility of moral responsibility. Philosophical Studies, 75, 5–24.
Strawson, G. (2002). The bounds of freedom. In R. Kane (Ed.), The Oxford handbook on free will (pp. 441–460). New York: Oxford University Press.
Timpe, K. (2007). Free will: Sourcehood and its alternatives. London: Continuum.
Vargas, M. (2013). Building better beings: A theory of moral responsibility. Oxford: Oxford University Press.
Acknowledgements
Many thanks to Audre Brokes, Ginger Hoffman, and Todd Moody, all of whom read earlier drafts of this paper and offered helpful suggestions.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Corabi, J. Two Arguments for Impossiblism and Why It isn’t Impossible to Refute them. Philosophia 45, 569–584 (2017). https://doi.org/10.1007/s11406-017-9847-7
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11406-017-9847-7