The disappearing agent objection to event-causal libertarianism
- First Online:
- Cite this article as:
- Pereboom, D. Philos Stud (2014) 169: 59. doi:10.1007/s11098-012-9899-2
- 373 Views
The question I raise is whether Mark Balaguer’s event-causal libertarianism can withstand the disappearing agent objection. The concern is that with the causal role of the events antecedent to a decision already given, nothing settles whether the decision occurs, and so the agent does not settle whether the decision occurs. Thus it would seem that in this view the agent will not have the control in making decisions required for moral responsibility. I examine whether Balaguer’s position has the resources to answer this objection.
KeywordsEvent-causal libertarianismNon-causal libertarianismDisappearing agent objectionIndeterminismMoral responsibility
Recent times have witnessed the explicit differentiation of three major versions of libertarianism, the event-causal, agent-causal, and non-causal types. In his Free Will as an Open Scientific Problem, Mark Balaguer endorses event-causal libertarianism. Critics of libertarianism have argued that if actions are undetermined, agents cannot be morally responsible for them. A classical presentation of this objection is found in Hume’s Treatise of Human Nature (1739/1978). In Hume’s version, the concern highlighted is that if an action is uncaused, it will not have sufficient connection with the agent for her to be morally responsible for it. I have argued that the best way to voice this type of concern is by the disappearing agent objection. I will contend that the DA objection indicates an advantage for agent-causal libertarianism over Balaguer’s event-causal libertarianism, and over a position—suggested by some of his remarks—that combines event-causal with non-causal components.
In event-causal libertarianism, actions, conceived as agent-involving events, that is, an agent making a decision or performing an action at a time, are caused solely by way of prior events, for example, an agent having a reason or a motivation at a time, and some type of indeterminacy in the production of actions by appropriate events is held to be a decisive requirement for moral responsibility (Kane 1996; Ekstrom 2000; Balaguer 2009, Franklin 2011). In alternative formulations, actions are indeterministically caused by states or property instances. This position has an ancestor in the Epicurean view according to which a free decision is an uncaused swerve in the otherwise downward path of an atom (Lucretius 50 BCE, De Rerum Natura).
According to agent-causal libertarianism, free will of the sort required for moral responsibility is accounted for by the existence of agents who possess a causal power to make choices without being determined to do so (Kant 1781/1787/1987; Reid 1788/1983 Taylor 1966, 1974; Chisholm Roderick 1964, Chisholm 1976; O’Connor 2000, 2008; Clarke 2003, 2008; Griffiths 2010). In this view, it is essential that the causation involved in an agent’s making a free choice is not reducible to causation among events involving the agent, but is rather irreducibly an instance of the agent-as-substance causing a choice not by way of events. The agent, fundamentally as a substance, has the causal power to cause choices without being causally determined to do so.
A third type of libertarianism is non-causal, and is advocated by Henri Bergson (1889), Carl Ginet (1990, 1996, 2007), Hugh McCann (1998), and Stewart Goetz (2007). Bergson argues that although action occurs in time, the time of conscious agency does not divide into the kinds of quantities or magnitudes that allow objects in space to be subject to causal natural laws. On Ginet’s conception, the key conditions for a basic action’s being free are that it is uncaused, that it has an agent as a subject, it has an actish phenomenological feel. In place of the actish phenomenological feel, McCann specifies that the action must be intentional, and intrinsically so. In the present philosophical climate, event-causal libertarianism is typically regarded as the most attractive of these views, all else being equal. So let us begin there, note its problems, and then see whether the non-causal and libertarian views are viable alternatives.
3 The disappearing agent objection
The objection is not that agents have no causal role in producing decisions, but that agents have insufficient causal role for the control that moral responsibility demands.
DA objection: Consider a decision made in a context in which moral reasons favor one action, prudential reasons favor a distinct and incompatible action, and the net strength of these sets of reasons are in close competition. On an event-causal libertarian picture, the relevant causal conditions antecedent the decision—agent-involving events—would leave it open whether the decision will occur, and the agent has no further causal role in determining whether it does. With the causal role of the antecedent events already given, whether the decision occurs is not settled by any causal factor involving the agent. In fact, given the causal role of all causally relevant antecedent events, nothing settles whether the decision occurs. Thus, plausibly, on the event-causal libertarian picture, agents lack the control required for moral responsibility.
This objection would appear to indicate an advantage for agent-causal libertarianism. What needs to be added to the event-causal libertarian account is involvement of the agent in the production of her decisions that would enhance her control so that she can settle which decision occurs, and thereby be the source of her decisions in a way that allows for moral responsibility. The agent-causal libertarian’s proposal is to reintroduce the agent as a cause, not merely as involved in events, but rather fundamentally as a substance. If the agent were reintroduced merely as involved in events, the DA argument could be reiterated with undiminished effect. What the agent-causal libertarian posits is an agent who possesses a causal power, fundamentally as a substance, to cause a decision without being causally determined to do so, and thereby to settle which of the antecedently possible decisions will actually occur.
Balaguer defines a number of key notions in setting out his position. First,
the view does not involve any sort of irreducible agent causation, but it does hold that undetermined L-free decisions are (ordinarily) causally influenced by—indeed, probabilistically caused by—agent-involving events, most notably, events having to do with the agent having certain reasons and intentions. (2009, p. 67)
This sets up his characterization of libertarian freedom, or L-freedom, by way of a paradigm type of case:
A torn decision: a decision in which the agent (a) has reasons for two or more options and feels torn as to which set of reasons is stronger, that is, has no conscious belief as to which option is best, given her reasons; and that (b) decides without resolving the conflict—that is, the person has the experience of “just choosing.”
In turn, appropriate non-randomness combines authorship, and control; “in order for a decision to be L-free, it has to be authored and controlled by the agent in question: that is, it has to be her decision, and she has to control which option is chosen (2009, p. 83). Finally, Balaguer defines the notion of torn-decision indeterminism, or TDW-indeterminism:
A. If an ordinary human torn decision is wholly undetermined, then it is L-free—that is it is not just undetermined but also appropriately nonrandom, and the indeterminacy increases or procures the appropriate nonrandomness (2009, p. 78).
Thus in a torn TDW-indeterminist decision, there is no mismatch between the real underlying probabilities for the various options at the time of decision, and the probabilities based on the agent’s consciously available reasons.
TDW-indeterminism: Some of our torn decisions are wholly undetermined at the moment of choice, that is, the moment-of-choice probabilities of the various reasons-based tied-for-best options match the reasons-based probabilities, so that these moment-of-choice probabilities are all roughly even, given the complete state of the world and the laws of nature, and the choice occurs without any further input, that is, without anything else being significantly causally relevant to which option is chosen. (2009, p. 78)
Given that the account is event-causal libertarian, it’s crucial to see how this story is to be told in event-causal terms. Reference to agents being causally influenced by reasons, or agents causing decisions, is to be recast in terms of causation among events. Here are the relevant events:
Stated in ordinary language, Ralph is deciding whether to stay in Mayberry or move to New York. Favoring the move to New York are his desire to play for the Giants, and his desire to star on Broadway. Favoring staying in Mayberry are his desire to marry Robbi Anna, and his desire to manage the local Der Weinerschnitzel. Suppose Ralph makes the torn decision to move to New York—he just decides to move to New York.
E1: Ralph’s desiring at t1–tn to play for the Giants,
E2: Ralph’s desiring at t1–tn to star on Broadway,
E3: Ralph’s desiring at t1–tn to marry Robbi Anna,
E4: Ralph’s desiring at t1–tn to manage the local Der Wienerschnitzel,
E5: Ralph’s deciding at tn to move to New York,
E6: Ralph’s deciding at tn to stay in Mayberry.
E1 and E2 probabilistically cause E5.
The DA objection counts against the supposition that this account secures the control required for moral responsibility. Intuitively, this sort of control requires the agent to settle which of the options for decision actually occurs, and the event-causal libertarian view does not seem to allow for this in the case of torn decisions. For on this picture, given the causal influence of E1–E4, nothing settles whether it will be E5 or E6 that occurs. Thus the agent does not settle whether E5 or E6 occurs. The objection concludes that event-causal torn decisions cannot, in Balaguer’s terminology, be appropriately nonrandom, and the indeterminacy in question cannot increase or procure the appropriate nonrandomness, for the reason that appropriate control is missing: Ralph does not control which option is chosen (2009, p. 83). Moreover, authorship is missing if this sort of control is required for authorship.1
In his response to the rollback objection, Balaguer remarks:
Suppose that some agent S is torn between two options, A and B, and eventually chooses A in a torn-decision sort of way. And now suppose that God “rolls back” the universe and “replays” the decision… If the decision is undetermined in the manner of TDW-indeterminism, then if God “played” the decision 100 times, we should expect that S would choose A and B about 50 times each…. It seems to be a matter of chance or luck what she chose, and to the extent that this is right, it seems that S didn’t author or control the decision. At the very least, it seems that the element of chance or luck here diminishes S’s authorship and control.
In each of the different plays of the decision, it is Ralph who does the choosing (2009, p.93)
However, the “it is Ralph who decides” answer faces a dilemma. In the event causal picture, there are two ways to construe the claim that Ralph decides to go to New York:
…the most we could hope for, vis-à-vis authorship and control is that it be Ralph who does the just-picking… Ralph chooses—consciously, intentionally, and purposefully—without being causally influenced by anything external to his conscious reasons and thought. Thus it seems that in this case, we do get the result that it is Ralph who does the just-choosing. And so it also seems that in this scenario, we procure as much authorship and control for Ralph as we can, given that he is making a torn decision. (2009, p. 97)
3.1 First Horn
But this analysis makes the position vulnerable to the DA objection, and arguably can’t secure appropriate non-randomness. On this option, the objection that Ralph has insufficient control over which decision is made retains all of its original force. Balaguer claims: “if the just-choosing were done by anything other than the agent, the she would lose authorship and control.” (2007, p. 97). But the concern is that if the just-choosing is what gets Ralph the control, and, in the spirit of event-causal libertarianism, control is a causal matter, then it seems that what is being specified is that a causal relation obtains between Ralph himself and the decision. However, the event-causal libertarian allows only causal relations among events, and not a fundamental causal relation between agent and event.
The pure event-causal libertarian possibility is that
Ralph decides to move to New York
is to be analysed as:
Ralph-involving events E1-E2 probabilistically cause Ralph-involving event E5.
3.2 Second Horn
In accord with this possibility, Balaguer might explicitly invoke a theory with an event-causal component and a non-causal component, one that affirms (i) a probabilistic, indeterministic causal relation between E1–E2 and E5, together with (ii) the agent’s being the subject of her action, and the action’s having an actish phenomenological feel (Ginet 1990, 1996) or being intrinsically intentional (McCann 1998), and that it is these components that result in the agent’s having the requisite level of control. This makes the crucial control relation non-causal, contrary at least to the spirit of event-causal libertarianism. On this option, control and authorship in a decision are ultimately a matter of a relation between an agent and a decision-event, where that relation is not causal, but rather being-the-subject-of, as in Ginet’s view (1990, 1996, 2007).
The other salient possibility is that
Ralph decides to move to New York
specifies a non-causal relation between Ralph and a decision, perhaps the relation of being-the-subject-of.
In Kane’s account, the phenomenology of decision-making process is decisive or at least counts heavily for ascertaining whether an agent is morally responsible for an undetermined action.
Now indeterminism may in some instances undermine choice… We imagined that Jane had reached a point in her deliberation at which she favored vacationing in Hawaii when, owing to a quantum jump in her brain, she found herself intending to vacation in Colorado. The case was odd because she did not have the sense of voluntarily doing anything… she would be reluctant—and we would be reluctant—to say she chose anything in such a case… So indeterminism can sometimes undermine choice. But there is no legitimate reason to generalize from cases like Jane’s and say it must always do so. Consider the businesswoman by contrast. Her experience, unlike Jane’s, is of consciously and voluntarily choosing to follow her moral conscience and to return to help the victim, thereby resolving a preceding uncertainty in her mind. Also, in the businesswoman’s case, unlike Jane’s, the indeterminate process discovered by the neuroscientists immediately preceding the choice was experienced by her as her own effort of will, not merely as a random occurrence in her brain that happened to influence the outcome. Given these circumstances, it would be hasty, to say the least, to lump the two cases together and draw conclusions about the businesswoman’s case from Jane’s… Why would the businesswoman conclude that she did not really choose in such circumstances (rather than that her choice was undetermined) just because, under very different circumstances, Jane did not really choose? (Kane 1996, pp. 182–183)
But any incompatibilist has reason to be wary of this response, since a compatibilist could as easily appeal to this same sort of phenomenological consideration in response to the objection that agents cannot be responsible for causally determined actions (Pereboom 2001, 2007). A compatibilist might argue that if an agent experienced her causally determined decision as resulting from an outside determining force, she would have good reason to believe that she was not making choice for which she was morally responsible. If, by contrast, a causally determined decision were experienced as voluntary and resulting from the agent’s effort of will, she would have a strong reason to believe she was morally responsible for it. However, incompatibilists generally would reject this compatibilist defense for the reason that a metaphysical fact about the causal history of the action—that the decision is causally determined—rules out the agent’s moral responsibility, regardless of its phenomenology. It would be implausible to claim that the phenomenology can carry more weight when the threat to moral responsibility is indeterministic instead.
But there is an element of Balaguer’s view that would render the phenomenology mistaken, and this is the additional supposition that Ralph’s decision has a purely event-causal libertarian etiology, by contrast with an agent-causal or non-causal account. And under this supposition the disappearing agent argument (DA) challenges the claim that the phenomenology of control is veridical.
…if my decision was undetermined in this way, then it was not causally influenced by anything external to my conscious reasons and thought. And this seems to eliminate the only worry we might have about the accuracy of the phenomenology of my decision. If we know for certain that once I moved into a torn state and was going to make a torn decision, nothing external to my conscious reasons and thought had any causal influence over how I chose… then what else could lead us to say that the phenomenology was mistaken? The answer, it seems to me, is nothing.
More generally, we can all agree that when we make decisions we often have the phenomenology of control in acting. But in the free will debate, serious questions are raised about whether a theory of agency that is causally deterministic or one that endorses pure event-causal indeterminism can accommodate this phenomenology of control in acting—or perhaps what is really the phenomenology of agency together with certain beliefs about the control we exercise in acting. In absence of an answer to the objection, citing the phenomenology in support of control should not be effective.
3.2.1 A mixed theory
Can this proposal secure the control whose absence the DA objection highlights, or is agent-causation perhaps required?
(NC1) Ralph is the subject of a conscious, actish-feeling, purposeful, intentional decision to move to New York, where Ralph’s decision’s being conscious, actish-feeling, purposeful, and intentional does not involve Ralph’s causing the decision to occur.
Hugh McCann (1998, p. 180) holds that an agent’s exercise of active control has two aspects. Any basic action is:
On Ginet’s view, each basic action is characterized by an actish phenomenal quality, its seeming to the agent as if she is directly making happen the event that is her basic action. Ginet stresses the “as if” nature of this appearance; it does not, he says, literally represent to the agent that she is causing the event in question… Might the actish feel of some occurrence itself constitute that event’s activeness, or the agent’s exercise of active control? The answer seems plainly to be negative. Indeed, as Ginet (1990, p. 9) appears to allow, an event with the indicated intrinsic quality might be brought about by direct stimulation of someone’s brain, in the absence of any relevant desire or intention on the part of that person. An occurrence produced in this way and in these circumstances would hardly seem to be an exercise of the subject’s agency. (2008, cf. 2003, pp. 19–21)
a spontaneous, creative undertaking on the part of the agent, and
intrinsically intentional. The intentionality of a basic action is a matter of its being intrinsically an occurrence that is meant, by the individual undergoing it, to be her doing.
Clarke’s objections seem strong to me, and I think they show that supplementing event-causal libertarianism with a non-causal component is not a promising way for Balaguer to address the DA concern.
Where intentionality is divorced from an appropriate causal production, it does not seem that it can, by itself, even partly constitute the exercise of active control. If intrinsic intentionality is at all possible, then it would seem possible for an event with this feature to be brought about in the manner and in roughly the circumstances (in the absence of any relevant desire or prior intention) just considered in discussing Ginet’s view [i.e., by direct stimulation of someone’s brain]. Again, such an occurrence, even if intrinsically something the individual undergoing which means to be an exercise of agency, hardly seems to be an exercise of agency at all. (2008, cf. 2003, pp. 20–21)
Against the non-causalist one might object that the making relation is just the causal relation. After all, isn’t causation fundamentally just making something happen or producing something? Clarke says, for example:
[Making] It was up to me at time T whether that event would occur only if I made it the case that it occurred and it was open to me at T to keep it from occurring (2007, p. 245)
And Ginet himself remarks:
An event that nondeterministically causes another brings about, produces, or makes happen that other event, though it is consistent with the laws of nature that the former have occurred and not have caused the latter (2003, p. 33)
Thus when Ginet states “I made it the case that the event occurred,” this would be equivalent to saying “I caused the event to occur.” Alternatively, one might propose with David Lewis:
To suppose it is possible for there to be indeterministic causation is to suppose that causation does not reduce, Humean fashion, to universal regularity but is rather a brute relation among particular events, a relation of production, a relation that may be impossible to specify in non-synonymous terms. (2007, p. 244)
But Ginet’s [Making] would seem to be at least roughly equivalent to: it was up to me at time T whether that event would occur only if I made it the case that it occurred, and I made the difference as to whether it occurred. On a ‘difference making’ account of causation, its being up to me whether that event occurred would then involve my causing it.
We think of a cause as something that makes a difference, and the difference it makes must be a difference from what would have happened without it. (Lewis 1986)
my sense is that what is being claimed might well be incoherent. Since (NC1) and (NC2) are in trouble, it may be that the libertarian needs to appeal to agent-causation to answer the disappearing agent objection.
(NC2) Ralph is the subject of a conscious, actish-feeling, purposeful, intentional decision to move to New York, where Ralph’s decision’s being conscious, actish-feeling, purposeful and intentional does not involve Ralph’s making the decision happen,
Randolph Clarke (2003) points out that on event-causal libertarianism, in addition to agent’s involvement in the antecedent events or states, there is a further respect in which the agent might be thought to contribute to a decision, and that is in the causing of the decision by the antecedent agent-involving states and events (2003, p. 74). The DA objection also challenges the supposition that this agent-involving causing of a decision provides for the agent’s moral responsibility. On the event-causal libertarian picture, the causal conditions antecedent to the causing of the decision will leave it open whether the causing of Ralph’s deciding to move to New York, or else the causing of Ralph’s deciding to stay in Mayberry, will occur. With the role of the antecedent conditions already given, nothing settles which causing will occur, and thus Ralph as agent does not settle which causing will occur.