Manuel Vargas is the exponent of a new methodology for free will and moral responsibility called revisionism. Vargas contributed a chapter on revisionism to the book Four Views on Free Will (Blackwell, 2007) with John Martin Fischer presenting his semicompatibilism, Derk Pereboom, arguing for hard incompatibilism, and Robert Kane defending libertarianism. In the introduction to Four Views on Free Will, revisionism is described.
The core idea of revisionism is that the picture of free will and moral responsibility embedded in commonsense is in need of revision, but not abandonment. That is, the revisionist holds that the correct account of free will and moral responsibility will depart from commonsense. As is the case with libertarianism, hard incompatibilism, and compatibilism, this view can take a variety of more specific forms. (p.4)
Vargas describes his project as a critical look at the commonsense idea about "free will."
We tend to think of ourselves as having a powerful kind of agency, of the sort described by various libertarian accounts. That is, we see ourselves as having genuine, robust alternative possibilities available to us at various moments of decision. We may even see ourselves as agent-causes, a special kind of cause distinct from the non-agential parts of the causal order. Moreover, we tend to think of this picture of our own agency as underwriting many important aspects of human life, including moral responsibility. How we think about a range of social issues (crime and punishment, addiction, and even issues such as homelessness), and the social policies we construct around them, in part depend on the presumption of this picture of agency. The problem is that our self-conception is implausible and largely unnecessary. It requires a metaphysics of agency that we have no independent reason to believe in and it mistakenly holds that we cannot attain a range of important human and moral aspects of our life in its absence. What I will argue is that we can get by with a stripped-down conception of agency that avoids many of the problems that plague our libertarian self-conception. It does, however, require some revision in how we think about ourselves and how we understand the foundations of various moral practices. Importantly, revisionism does not, by itself, require that we jettison talk of moral responsibility, praise, and blame. Revisionism does not commit us to dismissing the pull of incompatibilist construals of our self-conception. Revisionism does not require that we deny that there is something right about compatibilist and libertarian claims that we are free and responsible. What revisionism does require is that we regard our intuitive, commonsense self-conception with a critical eye, giving up those parts that are least plausible or otherwise worth abandoning.(p.127)
As does John Martin Fischer, and following a trend started forty years ago by Peter Strawson, Vargas identifies the free will problem as primarily a question about moral responsibility.
It is not clear that there is any single thing that people have had in mind by the term "free will." Perhaps the dominant characterization in the history of philosophy is that it is something like the freedom condition on moral responsibility. Roughly, the idea is that to be morally responsible for something, You had to have some amount of freedom, at some suitable time prior to the action or outcome for which you are responsible. That sense of freedom – whatever it amounts to – is what we mean to get at by the phrase "free will." (p.128) I am somewhat more optimistic than Robert Kane is about the results of the arguments favoring incompatibilists... What makes these arguments powerful is not so much that they rule out the possibility of compatibilism but rather that they show how easily incompatibilism seems to capture ordinary ways of thinking about our own agency. (p.132) I believe that the traditional philosophical arguments generally favor an incompatibilist and alternative possibilities reading of our commonsense requirements for free will and moral responsibility. However, I think that there is a potentially more powerful way to show that commonsense thinking – what psychologists and others sometimes call "folk" thinking – is incompatibilist. We can examine experimental data. (For what it is worth, the label "folk" isn't supposed to be derogatory – it just refers to what we might think of as "ordinary folks.") I see no reason to think that philosophers are uniquely or even especially well equipped to determine the contents of commonsense beliefs. (p.136)
Vargas cites recent work in experimental philosophy by Nichols and Knobe confirming the commonsense view that we have free will.
A number of psychologists and empirically oriented philosophers have been doing experimental work relevant to these issues. One especially interesting set of results come from the work of Shaun Nichols and Joshua Knobe. In one experiment, they gave their subjects descriptions of two different universes, one in which everything is completely caused by whatever happened before it, and the other a universe in which almost everything is determined by whatever happened before except human decision making. Then, they asked their subjects to identify which universe is more like ours. Roughly 95 percent of respondents describe the second universe (the one in which human decision making was indeterministic) as the one most like ours. This result seems to strongly favor the view that our ordinary self-conception of human agency is incompatibilist (specifically, libertarian). It is difficult to imagine why we would suppose human decision making is exempt from determinism if it were not linked to our having free will. What the experimental data appear to show is that we really do imagine ourselves to be agents with genuine, metaphysically robust alternative possibilities, and we really do, at least in moments of cool, abstract consideration, tend to favor an alternative possibilities requirement on moral responsibility. So, the experimental data seem to be something of a victory for incompatibilist diagnoses of commonsense. (p.137)
Vargas then argues that the diagnosis (the description) of the commonsense view needs another look at what we would prescribe (in a normative sense) as a philosophical account.
I have been arguing that there is good reason to think that an accurate diagnosis of commonsense will acknowledge the presence of incompatibilist elements in our thinking (minimally, metaphysically robust alternative possibilities). And, for some of the reasons I have presented, I doubt that we can make good on those elements. So, in broad terms, the revisionist proposal I am offering is a hybrid account: incompatibilism about the diagnosis and compatibilism about the prescription. Alternately, we might say the account is incompatibilist about the folk concept of free will and compatibilist about what philosophical account we ought to have of free will.
Vargas homes in on the reason that current event-causal libertarian accounts of free will are implausible. They are vague and confused about exactly how, where, and when indeterminism will occur in the brain to help with decision making.
Here's why I think libertarianism is comparatively implausible: libertarianism requires that indeterminism be present in our agency in a very particular way, at very particular times, in the process leading up to or in the decision about what to do. Just how the indeterminism operates varies by the particular theory, but all libertarian theories are committed to indeterminism showing up in the world at particular times and places. (I am largely ignoring the possibility of an uncaused event but the point applies to libertarian theories that appeal to uncaused events.) I am inclined to think that there are serious plausibility worries raised by any worked-out account of libertarianism, but for present purposes I am going to focus on [Robert] Kane's deservedly influential account of libertarianism to illustrate some of the specific worries about plausibility that can be raised against a libertarian account. On Kane's account, paradigmatic instances of free will, what he Calls SFAs, or "self-forming actions," are results of a particular kind of indeterministic brain process. The idea is that in moments of conflict or uncertainty, when there are multiple but mutually exclusive aims we would like to attain, this stirs up a chaotic system in the brain that becomes sensitive to lower-level indeterminacies in the brain. (As Kane himself notes, chaotic systems are usually understood to be deterministic.) These low-level indeterminacies (presumably at the quantum level) influence an agent's decision by affecting the sensitive chaotic system generated by the agent's desiring mutually exclusive aims. The result is a SFA, or an instance of free will. In connection with the prior criticisms about comparative empirical plausibility, it is worth briefly considering just how demanding the theory's commitments are: not only do agent mental processes have to turn out to be indeterministic, but they must also be indeterministic in a very particular way. If multiple mutually exclusive aims did not cause the brain to go into a chaotic state the theory would be disproved. If it tuned out that neurological systems weren't sensitive to quantum indeterminacies the theory would be disproved. If it turned out that neurological systems were sensitive to quantum indeterminacies, but not sufficiently sensitive to amplify quantum indeterminacies in a way that affects the outcome of choice, this too would disprove the theory. These are not marginal or insubstantial bets about what brain science will reveal to us. (p.141-143)
Vargas then asks the provocative question " What Does the Indeterminism Do, Anyway?"
Our question is this: what, precisely, is the work of the indeterminism? Is indeterminism required for control or is it required to elevate an agent that already has control into a free agent? Suppose that the work of the indeterminism is to bestow control. Consider, however, the nature of an agent's first moment of free will. Inevitably, that moment will not derive from prior free aspects of character, inclination, standing policies, and so on. It is, after all, the first free act. What then makes it count as free, as the kind of thing that could underwrite attributions of responsibility? Presumably, the causal forces that lead to that first willing will be constituted by a web of events, inclinations, character traits, decisions, and so on over which the agent had no control. Out of these things a first free act is generated.Vargas is correct that Kane's view of libertarianism adds some indeterminism to that first willing. And even Kane is worried that his view might be a hindrance to control.
So, if we do acknowledge that control can be attained out of elements that are not themselves controlled in their acquisition or content (as Kane acknowledges in chapter 1), it seems to me that we begin to sap some of the motivation for libertarianism as a prescriptive view. Here's why: it is hard to see what indeterminism adds to control, given that the options indeterministically available to the agent were all products of things beyond the control of the agent. In that first instance of free will, and in every instance that follows, what control the agent has is a function of what options the world bestowed on that agent (through experience, heredity, socialization, circumstantial luck, and so on). Any control the agent has must be built up out of those constraints. Given that even the indeterministic options are thus constrained, and the elements that gave rise to those options (experience, heredity, socialization, the circumstances one finds oneself in) were not in control of the agent, what does the indeterminism give the agent in the way of control? Why doesn't the indeterminism simply open up multiple paths to an agent, where the constitution and sources of those paths were not something over which the agent had control?This is actually very good. Indeterminism can help generate alternative possibilities for action, random combinations of past actions by the agent for example. Even small changes could lead to creative new thoughts. As long as the willing itself is adequately determined, and the randomly generated possibilities themselves are not the direct cause of action.
Now consider the alternative, where indeterminism doesn't bestow control but rather adds freedom to an agent that already has control. (Recall that on the view Kane defends in chapter 1, indeterminism is actually a hindrance to control.) On such a picture, how could we make sense of control? Presumably, an agent could be said to have control by possessing some complex arrangement of agency, given a particular environment or range of environments. For example, control in an environment presumably relies on capacities to be sensitive or responsive to stimuli in the environment, the capacity to make decisions, the ability to reliably predict what effects one's actions will have on the environment and vice-versa. And, plausibly, none of these things requires indeterminism. (Indeed, this would seem plausible even if we weren't assuming that control does not require indeterminism). Indeterminism, then, is something superadded to control, something that transforms an already controlled agent into an agent with free will. This seems to preserve an important theoretical burden for indeterminism: it is the difference-maker between free and unfree action. However, consider an agent that had all of the requisite capacities for control but lacked the indeterminism. Call him Max. The libertarian would insist that Max would not satisfy the freedom condition on moral responsibility. But what exactly would the freedom given by indeterminism provide for Max? It would not provide an additional measure of control — this possibility was ruled out above. Indeed, we might imagine that Max has all the control that anyone can have. If so, it is exceedingly difficult to see what indeterminism adds to maximal control. We might even put things this way: Max has all the control required for moral responsibility. Like anyone reading this book, Max deliberates about what to do, decides some things are better and some worse, and decides to do some things rather than others. The only thing he is lacking is indeterminism. Were he to suddenly be bestowed with it (in whatever way the libertarian likes), this wouldn't change the way his deliberations appear to him. He would still be deciding between options. He would still (let us say) have just as much control as he had previously. And the mental elements out of which his control was constituted and out of which the indeterministic possibilities would be shaped would not suddenly become under his control if they were not already. So, whatever freedom it bestows on Max it is nothing that changes the way his deliberations will appear to him and it does nothing to change the control that he actually has. The work indeterminism does begins to seem ephemeral.Vargas' conclusion is pessimistic about adding any indeterminism to an agent with control. As with Fischer perhaps, Vargas does not give much importance to alternative possibilities that are indeterministic. They are not in the agent's control. But that is the idea of the "free" part of "free will." The alternative possibilities are "free." The "will" is adequately determined to exercise control over a decision consistent with character and value, and then has control to act on that decision with determination.
Free Will and Moral ResponsibilityManuel Vargas follows John Martin Fischer in connecting free will strongly to moral responsibility. Vargas says:
It is not clear that there is any single thing that people have had in mind by the term "free will." Perhaps the dominant characterization in the history of philosophy is that it is something like the freedom condition on moral responsibility. Roughly, the idea is that to be morally responsible for something, you had to have some amount of freedom, at some suitable time prior to the action or outcome for which you are responsible. That sense of freedom — whatever it amounts to — is what we mean to get at by the phrase "free will." However, there may be things for which free will might be important or other senses of free will that are independent of concerns about moral responsibility. For example, philosophers have worried whether free will is required for some human achievements to have a special worth or value, or for there to be values and valuing in any robust sense. Although I think much of what I will say can be applied to other aspects of thinking about it, I will primarily concerned with free will in its connection to moral responsibility, the sense in which people are appropriately praised or blamed. (Four Views on Free Will, p.128-9)Vargas then puzzles about children become free agents. At what age does the indeterminism appear that is needed for libertarian free will? [Note. This question goes back at least to Anthony Collins.]
Consider the question of how we go from being unfree agents to free agents. This is a puzzle faced by all accounts of responsibility, but there is something pressing about it in the case of libertarianism. As children we either had the indeterministic structures favored by your favorite version of libertarianism or we lacked them. If we lacked them as children, we might wonder how we came to get those structures. We might also wonder what the evidence is for thinking that we do develop said structures. Suppose the libertarian offers us an answer to these questions, and the other empirical challenges I raised in the prior section. We would still face another puzzle. What, exactly, does the indeterminism add? What follows in this section is not so much a metaphysical concern as it is a normative concern. It is a concern about what work the indeterminism does in libertarianism, apart from providing a way to preserve our default self-image as deliberators with genuine, metaphysically robust alternative possibilities. (p.148)Children have free will from birth. It is part of their biological makeup. It is the moral responsibility that they "come to get" at some age.