What follows is a work in progress, and I would greatly appreciate any constructive feedback pertaining to my argumentation and the strength of my reply to the RPAP. As stated in the body of the post, additional information on the two primary articles cited can be found in the endnotes at the end of this post.
Frankfurtian-style counterfactual intervener scenarios of all different stripes hold a special place in discussions of free will and moral responsibility. In some situations, they are a necessary evil with which one must contend, and in others they are an insurmountable obstacle for some theories. Many journal articles and full-length books on these topics dedicate large sections of text to attempting to reconcile Frankfurtian-style counterfactual intervener scenarios (CIS) against Frankfurt’s modification of the Principle of Alternate Possibilities (PAP). I believe that, using Galen Strawson’s iteration of what he deems the “Basic Argument” for the impossibility of moral responsibility, I can at best obviate Frankfurt’s Revised Principle of Alternate Possibilities (RPAP) and at worst side-step the need for addressing the PAP/RPAP by way of positing a new principle based on Strawson’s Basic Argument, what I shall call the Principle of the Basic Argument (PBA).
For the purposes of this post, I should like to assume that most readers are at least passably familiar with the general trajectory of Frankfurt’s argument. If not, please see the endnotes for additional resources related to these two works. First I shall introduce Galen Strawson’s Basic Argument, and from there offer a (hopefully) pithy recap of Frankfurt’s final thought experiment and his re-envisioning of the Principle of Alternate Possibilities. With these two theories in hand, I shall be in a position to address how the very structure of the CIS ignores the inherent problem of distal versus proximal causes in considering moral responsibility and freedom of the will. I shall then demonstrate that any example utilized in the CIS model is ostensibly prone to dependence upon “the way a person is” and as such entails a lack of moral responsibility from the very beginning. Finally, I shall suggest a formulation of the PBA that will hopefully demonstrate the failings of the RPAP.
Galen Strawson’s Basic Argument
In writing “The Impossibility of Moral Responsibility” Galen Strawson aims to draw attention to well-established but often overlooked formulation contra the plausibility of moral responsibility. Calling it the Basic Argument, Strawson contends that it displays the impossibility of moral responsibility regardless of the truth of determinism. Further, he argues that it has been systematically underestimated and overlooked in classrooms across the world:
On the one hand, [the Basic Argument] convinces almost all of the students with whom I have discussed the topic of free will and moral responsibility. On the other hand it often tends to be dismissed, in contemporary discussions of free will and moral responsibility, as wrong, or irrelevant, or fatuous, or too rapid, or an expression of metaphysical megalomania.[1]
Throughout the article, Strawson employs multiple definitions of the Basic Argument (BA), ranging from one to nine premises, but all possessing the same conclusion. Because his article focuses on the need for stating the BA in a simple and straightforward manner, here is Strawson’s final iteration:
(1) You do what you do, in any situation in which you find yourself, because of the way you are.
So
(2) To be truly morally responsible for what you do you must be truly responsible for the way you are – at least in certain crucial mental respects.[2]
Strawson writes that (1) appears to be incontrovertibly true, and that only (2) holds any hope for reconciliation. Throughout his article, Strawson contends that true responsibility lies in the ability for an agent to be causa sui, and since none except possibly a deity can have such a quality, none but such a deity can be truly responsible. While he entertains a number of compatibilist and libertarian notions of moral responsibility, he dispatches them as not achieving sufficient responsibility to warrant being named a genuine alternative to the BA.
What of Harry Frankfurt’s famous counterfactual intervener scenarios, which allegedly overturn or modify the Principle of Alternate Possibilities (PAP) and demonstrate the possibility of moral responsibility despite an inability to have acted otherwise? There are at least two appropriate responses, I believe, to counterfactual intervener scenarios, one which demonstrates the failings of the Revised Principle of Alternate Possibilities (RPAP) to prove compatibility between an inability to have done otherwise and moral responsibility, and the other which precludes the need for directly discussing alternate possibilities in the first place. But first, a short recap of Frankfurt’s final thought experiment and a definition of the RPAP.
The Revised Principle of Alternate Possibilities
Frankfurt’s Revised PAP (RPAP) states that “a person is not morally responsible for what he has done if he did it only because he could not have done otherwise.”[3] This differs from the PAP in a critical way. The original PAP states that a person is not morally responsible for what he has done if he could not have done otherwise. Given the strength of counterfactual intervener scenarios (CIS), it seems a more fruitful endeavor to instead accept the revised PAP but question its applicability against Strawson’s Basic Argument rather than attempt to find a CIS which does not entail a rejection of the original PAP or a flaw in any and all CIS iterations that maintains the original PAP.
Frankfurt arrives at the above definition of the RPAP by way of the following thought experiment. As indicated above, this is Frankfurt’s final consideration of the thought experiment and so does not include the path by which Frankfurt arrives as this final example.
Black wants Jones4 to murder Sally, and will go to any length to ensure that this occurs. Because he does not want Jones4 to know his motives unless it is absolutely necessary, Black waits until Jones4 makes up his mind to take any action. In this way, no matter what Jones4 decides initially, he will end up murdering Sally. If Jones4 decides to murder Sally, Black will not act in any way. In this case, Jones4 will be just as morally responsible for what he does as if Black did not even want him to murder Sally. This is because Black played no role in leading Jones4 to act. But in this instance, Jones4 could not have done otherwise. If Jones4 decides otherwise, Black acts in such a way that Jones4 WILL murder Sally. Presumably this example is analogous with Jones2 and Jones4 is not morally responsible:
What action he performs is not up to him. Of course it is in a way up to him whether he acts on his own or as a result of Black’s intervention…but whether he finally acts on his own or as a result of Black’s intervention, he performs the same action. He has no alternative but to do what Black wants him to do. If he does it on his own, however, his moral responsibility for doing it is not affected by the fact that Black was lurking in the background with sinister intent, since this intent never comes into play.[4]
Having established Frankfurt’s final example, I am now in a position to explain its susceptibility to the Basic Argument.
The Basic Argument
The strength of the Basic Argument is, simply, that it subverts the RPAP by way of being able to ignore the intervener to a degree and instead rely upon a frequently overlooked aspect of Frankfurt’s RPAP, viz. the reliance upon personal preferences and predilections. Though the CIS is intended to draw a larger-scale corollary to deterministic systems wherein factors outside of the agent’s control result in the agent acting in a certain way, it also relies upon the inclinations of the agent, which lead us to a discussion of the Basic Argument.
Frankfurt’s CIS relies upon instilling in Jones “an irresistible inner compulsion to perform the act Black wants performed and to avoid others…manipulate the minute process of Jones’ brain and nervous system in some more direct way…in other words, let Black bring it about that those conditions [such that Jones cannot do otherwise] prevail.”[5] Let’s use three of the most disparate examples provided by Frankfurt in order to demonstrate that each has an effect upon Jones for the reason that they impact, or are based upon, elements of “how Jones is” and as such are outside of his control.
(1) “A terrible threat”[6] which guarantees Jones’ cooperation. Let us imagine Black threatens Jones’ only living family member with a long and painful death the likes of which has not been seen in this or any other possible world. For such a threat to have 100% efficacy, it must rely upon Jones’ pre-existing desires, i.e. his desires to preserve his genetic line, to spare his loved one physical or mental anguish and death, and so on. For, should we imagine that Jones is a sociopath or psychopath to the degree that he does not care for others in this way that moves him to act in accordance with Black’s wishes, then Black would utilize another threat, perhaps an equally painful death for Jones himself. Either way, for the threat to have causal efficacy it must be rooted in those factors of Jones’ existing mental, emotional, and/or psychological makeup.
(2) “a potion, or put him under hypnosis, and in some such way as these generate in Jones an irresistible inner compulsion to perform the act”[7] rely alternately upon Jones’ physiology and psychological willingness to be hypnotized. The potion example is easy enough, as Jones must be physiologically susceptible to the potion to a degree that its effects are guaranteed, for if Jones had a hardier constitution, a stronger or alternate potion must be used that is tailored to his biological makeup. For the hypnotism example, though I am certainly no expert, as I understand it no person can be hypnotized in any meaningful way if they are not comfortable with the process or open to such a system of suggestion in the first place. This would mean Jones would need to possess a desire or willingness to be hypnotized before Black began the process, otherwise it would not be effective.
(3) “Or let Black manipulate the minute processes of Jones’ brain and nervous system in some more direct way, so that causal forces running in and out of his synapses and along the poor man’s nerves determine that he chooses to act and that he does act in the one way and not in any other.”[8] This direct meddling with the very processes of Jones’ brain can be considered in two ways, as the thought experiment goes. First, Black can guarantee that by way of a counterfactual intervening scenario that the desired synapses fire according to Jones’ own “wishes” or innate inclinations, and if not, Black can alter the firing of the synapses such that the desired outcome instantiates.
Not one of these three examples provides an adequate defense against the BA for the following reasons. (1) very clearly depends upon ‘the way Jones is’ prior to any meddling on the part of Black, and so Black’s meddling is really just finely tuned coercion. In the case of (1), Jones is behaving in accordance with his desire to avoid whatever malicious event Black has threatened, but more to the point, in accordance with his innate susceptibility to this form of coercion. Neither of these options fall within Jones’ control any more than resisting Black’s attempt to control him do. (2) similarly depends upon “the way Jones’ is” physiologically and psychologically to the extent that we can imagine a scenario in which case Jones is powerless to build an immunity to such a potion, or is born with a weakness to such drugs, etc. and this is just “the way Jones is” prior to Black’s meddling. For hypnotism, this relies also upon either Jones’ psychological willingness to submit to hypnotic suggestion by Black or weakness/susceptibility on the part of a more physicalist description of the reasons for Jones’ susceptibility to hypnotism. (3) quite literally alters the way Jones “is” in a way 100% outside of Jones’ control, thus forming perhaps the clearest example of Jones becoming or being a “certain way” and having no responsibility for this change.
Though Frankfurt includes the stipulation that his CIS thought experiment is flexible enough to include any sufficient form of coercion, persuasion, or manipulation outside of the examples included above, all of the examples he gives play , in some manner or another, upon “the way Jones is” or alter “the way Jones is” in order to force his behavior. Now that I have covered this issue, I shall re-examine the CIS model to demonstrate that, regardless of an intervener, Jones always acts in accordance with “the way he is” and in all hitherto conceivable cases he is not responsible in any meaningful sense for “the way he is.”
Using just one of the examples detailed above (although any example, even one not included above, would do equally well), let us consider the following iteration of the CIS:
(A) Black wishes for Jones to kill Sally. He threatens a gruesome and drawn out murder of Jones’ only daughter if Jones should fail to comply. Given Jones’ innate desire to protect his daughter from pain and death, Jones complies and murders Sally. Given the PAP, Jones is not morally responsible for murdering Sally.
(B) Black wishes for Jones to kill Sally. Jones has a deep hatred for Sally and stands to personally benefit from her demise. Black, fearing Jones may get cold feet at the last moment, positions himself to threaten a gruesome and drawn out murder of Jones’ only daughter if Jones should display second thoughts about murdering Sally. Unaware of Black’s presence, Jones is inclined overwhelmingly to murder Sally and does so. Contra the PAP, Jones is morally responsible for murdering Sally despite the inability to have done otherwise given the unseen but counterfactual and causally efficacious threat Black is prepared to make.
As with all of Frankfurt’s examples, we presume that the threat is causally efficacious in that it will never fail to coerce Jones if it is uttered.
Using the RPAP highlighted earlier, (B) is unproblematic from the perspective that Jones is not murdering Sally only because of the counterfactual threat lurking in the shadows – he is doing so in accordance with some set of desires, motives, inclinations, etc. As may be obvious to the reader at this point, given the nature of the Basic Argument, (B) is still problematic for moral responsibility because it is predicated on “the way Jones is” which is outside of Jones’ control. So, by way of threat or in accordance with his desires, Jones is never morally responsible for his behavior in any meaningful sense in all such examples. In this way, moral responsibility and determinism remain incompatible whether using the PAP or RPAP.
The Principle of the Basic Argument
If this is still not enough, we can offer a reframing of the PAP/RPAP in terms of the Basic Argument that obviates the need to discuss alternate possibilities directly, although I am sympathetic to the possible criticism that it implies or presumes the PAP (I would like to hear an argument for this, however.) For working purposes I shall refer to this as the Principle of the Basic Argument (PBA). Though capable of different iterations, the PBA is generally thus:
PBA: “A person is not morally responsible for her actions if those actions reduce to distal causes manifested in the way that person is, over and above subjective ownership of proximal causes.”
A few words of explanation on this iteration: not only does this formulation of the PBA include an element of Strawson’s Basic Argument, I have also elected to delineate distal and proximal causes. I do this because it highlights how a cause may stem from an agent and his desires while also indicated that such desires have causes which stretch back outside the control of the agent himself. So, although in (B) Jones’ desire to kill Sally is proximal and Jones may have chosen this desire (as much as is possible, which admittedly is not very much) the motivations for formulating such a desire ultimately track back along a path of causation to some distal cause(s) outside of the agent.
Using such an iteration of the PBA, we need not invoke alternate possibilities in any way, given that (A) and (B) above both rely upon “the way Jones is” and his inability to change this due to distal causal which are, by definition, outside of his control. This is because (A) relies upon external coercion which is efficacious only due to proximal causes, i.e. Jones’ inclinations and (B) relies upon proximal desires which instantiate due to “the way Jones is” by way of distal causes outside of his control. In this manner, Frankfurt’s RPAP fails to demonstrate the possibility of moral responsibility without alternate possibilities. Further, even if it could, we need not couch the discussion directly in terms of alternate possibilities to accurately explain the impossibility of moral responsibility, and can opt instead to use the PBA by way of Strawson’s work on the Basic Argument.
[1] Galen Strawson, “The Impossibility of Moral Responsibility,” in Free Will, ed. Gary Watson, 2nd ed. (New York: Oxford University Press, 2007), 214. Strawson also has a related post at The Stone that discusses an iteration of the Basic Argument and its relation to moral responsibility.
[2] Strawson, “The Impossibility of Moral Responsibility,” 219.
[3] Harry Frankfurt, “Alternate Possibilities and Moral Responsibility,” in Free Will, ed. Gary Watson, 2nd ed. (New York: Oxford University Press, 2007), 176. Can be found here for free but I obviously take no credit for the hosting or authorship of this article, and it is not the version I referenced for this post.
[4] Frankfurt, 174.
[5] Ibid. 173.
[6] Ibid.
[7] Ibid.
[8] Ibid.
It is important, I had been told by the academic staff, that a philosopher’s first presentation is to be made for the common man. All technical derivations should be introduced in the same way.
John,
I am unsure if your comment refers to Strawson’s Basic Argument or my post, but I shall assume the former. Please correct me if otherwise.
While Strawson’s argument is perhaps structured for the “common man” it would be incorrect to say it began as such. Rather, it would be more accurate to describe it as a simplified argument, derived from decades of trial and error in contemporary discussions of Free Will / Determinism, many of which are highly technical and certainly written for professional philosophers rather than, say, the readership of the NYT Best Seller list.
[...] have a new post up at the Florida Student Philosophy Blog, which can be found here. It concerns the Revised Principle of Alternate Possibilities (RPAP) and using Galen [...]
“Unaware of Black’s presence, Jones is inclined overwhelmingly to murder Sally and does so. Contra the PAP, Jones is morally responsible…”
I would think moral responsibility is meted out according to the information Jones had at the time of his decision to act. Say Jones was unaware of the fact that Sally had a fatal disease…
“capable of different iterations” -> “formulations”. Iteration implies to my mind the process of successive refinement, which hasn’t actually happened yet?
Not really relevant to your actual argument, but I get the impression that legal thinking about responsibility (noting Frankfurt’s “reasonable man” and whether coercion is truly irresistible) is that a reasonable person aware of a predisposition to murder in themselves should take steps to avoid exacerbating situations (eg recurrent drink-driving).
Hi David,
Thank you for the feedback on my use of “iteration” – you are correct, that what I present as the PBA is not a refinement but rather an initial formulation. I shall be more careful when using it in the future.
Moral responsibility in this case would seem to only passively rely upon the information at hand, given that in Frankfurt’s final example it is important that Jones be unaware of Black lurking in the shadows. As for Sally and her fatal disease, her death itself is not what demands a responsible party; rather, her murder does. It is the same as saying that, had Jones not murdered her, she would have accidentally stepped in front of a train and so Jones’ responsibility is mitigated somewhat.
Your point about legal responsibility is noted, but I do not see it playing a substantial role in this thought experiment for a few reasons. First, the Supreme Court has stated that, to use my terms, a robust sense of free will is a necessary underpinning of our criminal justice system. While instances of coercion and the like certainly play out in the courtroom to various ends, this is in part because the system itself presumes freedom of the will in a way that is consistent for a “could-have-done-otherwise” view of moral responsibility. Also, one could mount the defense that no reasonable person would seriously entertain the thought of murder as a legitimate option, much less carry it out. This would seem to be the case in a Kantian view, at least, since to do so treats others as a means to an end rather than ends in and of themselves.
Further, to run with your drunk driver example (and to carry on with our use of Frankfurt’s work) such a person could be classified as either a willing or unwilling addict. In either case, for reasons which should be obvious, she most likely does not have the ability to do otherwise, regardless of her second-order desires to avoid drinking and driving, or not to be that type of person.
My point was along the lines of “Jones is unaware of Black”: his degree of responsibility is a function of his knowledge of the situation, specifically what alternative actions he actively chooses amongst, not what alternatives are actually open to him. This alIows deceit and manipulation to affect the outcome in the usual common-sense way. In 4B, it seems to me that Black is completely superfluous to the question of whether Jones can be morally responsible, since he does not impinge on Jones’ decision making processes.
I chose a fatal illness because this removes agency ie Jones can get his desired outcome without anyone being responsible.
On rereading, I don’t think you spend enough time on the proximal and distal causation material and your PBA. Even libertarians would agree that your desires have antecedent causes (viz experimental evidence for unconscious perception and motivation), and that they lead to (are necessary but not sufficient causes of) actions. The question is how action in the face of conflicting desires and beliefs arises.
The drink driving and crime passionel examples are interesting to me because of the diminished acceptance of diminished responsibility, due to social change. The point of view in the former is that it is quite possible to be a non-driving alcoholic, whether your addiction is willed or unwilled. Still, I think most courts would hold Jones to have diminished responsibility in 1-3,4A. Their arguments would involve whether Jones ever had an opportunity to subvert Black’s plans (which by design of the thought experiment is none), and whether an “ordinary reasonable person” (average loving parent etc) would have found the coercion (or other intervention) irresistible.
Hi David,
Ah, I see. Please allow me to clarify Frankfurt’s purpose in structuring the thought experiments as he did, although obviously a thicker description would come from his article itself (see the footnotes for an online version):
Prima facie it is a fairly unproblematic statement that moral responsibility (MR) requires the ability to have done otherwise. I see this lurking implicitly in your statements that we should evaluate Jones’ MR based on those options he perceived open to him, given that the coercion and other acts by Black are purposefully irresistible. What Frankfurt was replying to in his original work was the notion that there cannot exist MR in situations where a person could not have done other than s/he did. To that end, Frankfurt structures a set of thought experiments that culminate in a final iteration (proper use this time!) that demonstrates how we can still find Jones morally responsible despite there being only one outcome that can obtain.
So, while you are correct that it is relevant in a legal sense whether the coercion or what have you was legitimately irresistible by a reasonable person, this is not the purpose of the thought experiment. Rather, as outlined above and in the post, the purpose is to give at least one instance where a person could not have done otherwise but is nevertheless responsible for that behavior. Frankfurt, and others, believe that counterfactual intervener scenarios (CIS) provide such examples.
With respect to 4b, Black’s presence is quite important for the structure of the thought experiment itself. Again, as outlined above, this is a CIS. For, if Black was not waiting in the wings to tinker with Jones’ brain, we would merely have a situation in which Jones commits murder and seemingly had the ability to otherwise. This example is not very interesting, because it does not pull any philosophical weight with respect to the Principle of Alternate Possibilities – for, unless we further modify it, what is to stop us from saying that Jones was responsible because there were no other factors determining his behavior or the outcome of his actions?
Being a bit dull, I don’t see much difference between 4B and my example where S dies from a disease if J does not choose to kill her. The novelty is simply the speed at which the alternative action is forced to occur by Mr B, since coercion/manipulation etc are accepted by most as reducing J’s moral responsibility (providing “he could not be expected to act otherwise under the circumstances”). That is, changing the wording to “a person is morally responsible for what he has done if he could have done otherwise at the time he formed the intention to act” cleaves closer to what I would see as the usual interpretation. This also is more useful when thinking about MR for interrupted acts, such as a foiled attempt at murder. These are actually very similar conceptually, except Mr B always interferes to stop the murder iff J decides to go through with it.
Hi David,
One of the many differences between 4b and your illness example is that the purpose of the CIS is that we are not focusing on Sally’s death but rather Jones’ murder of Sally. In 4, Jones murders Sally no matter what occurs. In your example, Jones can either kill her himself or wait for her to succumb to her illness. It should suffice to say that this is a very different situation than that which is presented in the CIS.
Jared, this thread is a year old, but there is a new argument that claims to overturn Strawson’s basic argument made by a neuroscientist called Tse in a MIT press book called ‘the neural basis of free willL criterial causation.’ The basic idea appears to be that neuronal events can recode each other without making each other fire. This recoding changes the info that will drive a neuron in the future. This gets around the basic argument because the neural code is not organized so that mental events change their own physical neural basis, but instead they set parameters on what mental events can possibly be realized by future neural events. It is a pretty interesting set of arguments, and as much philosophy as it is neuroscience. –Shinya