Chapter 5: The Case of Forgetting
2. The Minimal Approval View and The Case of Forgetting
Consider the following case from George Sher’s book, Who Knew? Respon- sibility Without Awareness:
Alessandra, a soccer mom, has gone to pick up her children at their el- ementary school. As usual, Alessandra is accompanied by the family's border collie, Bathsheba, who rides in the back of the van. Although it is very hot, the pick-up has never taken long, so Alessandra leaves Sheba in the van while she goes to gather her children. This time, how- ever, Alessandra is greeted by a tangled tale of misbehavior, ill- considered punishment, and administrative bungling which requires several hours of indignant sorting out. During that time, Sheba lan- guishes, forgotten, in the locked car. When Alessandra and her chil- dren finally make it to the parking lot, they find Sheba unconscious from heat prostration.151
Assuming that Alessandra is a generally caring and thoughtful person who loves Sheba, it is not easy to get clear on our intuitive responses to such a case. On the one hand, we might feel bad for Alessandra for mak- ing such an upsetting mistake, one that does not on the face of it seem to reflect any sort of characteristically morally wrong personality traits. On the other hand, she seems to have acted negligently—we think she should still be held responsible in some sense for what happened. To bring out this intuition, imagine that Alessandra were to offer no apology to her children and family for what she had done.152 This, I think, would seem
151 Sher (2009): 24.
152 Although, perhaps apology is warranted even in cases that involve no responsibility. See, for example, Talbert (Forthcoming): 17, and Scanlon (2008): 150. Perhaps a similar in- tuition could be brought about, though, by considering the case in which Alessandra feels no remorse or special duty to comfort others affected as a result of her causal re- sponsibility.
highly inappropriate. While her family might rightly be sensitive to the fact that Alessandra herself might be suffering from the tragedy of the in- cident, we might not think them out of line to expect her to make amends of some sort. The verdict here is unclear, but there is at least a prima facie case for Alessandra’s being responsible in some way.
What does the Minimal Approval view say about Alessandra’s neglect of Sheba in this case as it pertains to her attributional-responsibility? It might be thought that the Minimal Approval view actually has resources that other views in the Deep Self family lack to show why Alessandra is, after all, attributionally-responsible for leaving Sheba in the car. In Chap- ter 3, I argued that agents could be attributionally-responsible for acting out of weakness of will. When Sam acts out of weakness of will and ends up going to a party rather than studying for her exam, she is attributional- ly-responsible for going to the party. It makes sense in these cases to also say that she is derivatively attributionally-responsible for not studying. If being attributionally-responsible for what you actually do can render you attributionally-responsible for what you fail to do as well, then we might think we should conclude from the fact that Alessandra is (let’s stipulate) attributionally-responsible for staying to talk to the school administrators, that she is also attributionally-responsible for leaving Sheba in the car.
In order to evaluate whether or not this is right, though, we’ll need to pay more attention to the principle that lets us move from Sam’s attribu- tional-responsibility for her action to her attributional-responsibility for her omission. One principle that might get us from Sam’s responsibility for going to the party to Sam’s responsibility for failing to study is the fol- lowing:
Modal Bridge Principle: If an agent, A, is attributionally-responsible for ϕ-ing-at-t, then A is also attributionally-responsible for not ψ-ing, where ψ-ing is anything else A could have done at t.
Adopting this principle would also give the result that Alessandra is re- sponsible for not tending to Sheba, since, arguably, she could have tended to Sheba rather than stayed to talk to the administrator. But notice that adopting this principle reopens the issue that Deep Self views were par- tially created to avoid. Given the fixity of the past and the truth of deter- minism, Incompatibilists will argue, there is nothing that Alessandra might have done at t other than what she actually did—stay and talk to the administrator. Given the hard work of maneuvering around these questions and arguing for their irrelevance in the case of attributionally- responsible action, it would be unfortunate if they were simply to reap- pear in the case of attributionally-responsible omission.
There is also a larger problem with this view. Consider the following case:
Oblivious Ollie: Ollie is walking to work and meets the conditions of Minimal Approval for walking to work. Unbeknownst to him, there is a small child drowning in the river behind him, but he never turns around and notices the child drowning. Is Ollie attributionally- responsible for failing to turn around and help the child? It would seem that he should not be, but given a plausible (compatibilist) con- strual of the Modal Bridge principle, he is attributionally-responsible for walking to work, and could have instead turned around and saved the drowning child.
So perhaps, instead, we ought to adopt a principle like the following: Evidence Bridge Principle: If an agent, A, is attributionally-responsible for ϕ-ing-at-t, then A is also attributionally-responsible for not ψ-ing- at-t where ψ-ing is any act that the agent has sufficient evidence is a potential course of conduct for her.
This principle both correctly exempts Ollie from attributional-respons- ibility, and, at least on the face of it, might even avoid the incompatibilist worry. But consider the explanation it gives as to why Alessandra is re- sponsible for leaving Sheba in the car. She is responsible for talking to the school administrators when it is the case that she also had evidence (to which she failed to attend) that she could go tend to Sheba to avoid catas- trophe at that moment instead. But what role does the existence of the ev- idence to which Alessandra failed to attend play in explaining why she is responsible? Alessandra may fail to attend to the evidence due to a ran- dom misfiring in her brain, and so the fact that this evidence existed doesn’t tell us anything about what Alessandra is like agentially. It may be true that it tells us that a surface-level normative fact is true about Ales- sandra: there was evidence that an alternative course existed, evidence to which she ought to have been responsive. But grounding the explanation of an ascription of attributional-responsibility in a normative fact like this would require accepting a sort of deep foundational asymmetry between attributability for acts and omissions. Since attributability for actions on the Minimal Approval view is determined by pure metaphysical condi- tions of agency, it would be odd if first-order normative theorizing needed to take place in order to determine the conditions for attributability for an omission,.
Instead, I think the Minimal Approval view should be coupled with a principle that ties attributional-responsibility for omissions more closely to the fact that the mechanism of the agent’s action relates to the fact that the agent would approve to some degree of what she does do instead of the omitted action. This will explain why Sam is attributionally- responsible for failing to study, but will have the result that Alessandra is not attributionally-responsible for failing to tend to Sheba. This is the principle I think the Minimal Approval theorist ought to adopt:
ly-responsible for ϕ-ing-at-t, then A is also attributionally-responsible for not ψ-ing-at-t, when ψ-ing is an alternative considered in the worlds in which A meets the conditions of minimal approval.153
Sam is attributionally-responsible for not studying, not because she could or should have studied, but because of a fact about her agency. Sam min- imally approves of going to the party even though she knows that she could be studying, and this is related to what leads her to go to the party. Notice that, given this bridge principle, Sam’s responsibility for her omis- sion actually has something to do with her approval. She’s responsible for going to the party because she minimally approves of it, and she’s respon- sible for not studying because she minimally approved of doing some- thing else instead. Her responsibility for her omissions comes from the fact that we know something about what she is like when she is faced with the
153 What grounds these bridge principles? One possibility is that they are derived from principles that would explain why agents are responsible not just for their actions, but al- so for the consequences of their actions. On this strategy, the fact that an agent does not ψ instead when she ϕs is just a consequence of ϕ-ing like any other. Different principles of this form correspond with the modal, normative, and contrastive approval bridge princi- ples. For example:
Modal: If an agent, A, is attributionally-responsible for ϕ-ing-at-t, then A is also attributionally-
responsible for all of the consequences of ϕ-ing that she could have foreseen, (possibly including
the fact that A will not ψ instead).
Evidence: If an agent, A, is attributionally-responsible for ϕ-ing-at-t, then A is also attributional-
ly-responsible for all the consequences of ϕ-ing that she had enough evidence to foresee, (possible
including the fact that A will not ψ instead)
Normative: If an agent, A, is attributionally-responsible for ϕ-ing-at-t, then A is also attribu-
tionally-responsible for all of the consequences of ϕ-ing that she should have foreseen (possibly in-
cluding the fact that A will not ψ instead).
Contrastive Approval: If an agent, A, is attributionally-responsible for ϕ-ing-at-t, then A is also
attributionally-responsible for all of the consequences of ϕ-ing she foresees in the worlds in which
she reflects on which of her motivations to act on at t and approves of ϕ-ing, (possibly including
choice situation in which the option of studying features—there’s a part of her that wants to go to the party anyway.
The same can’t be said of Alessandra. Her minimal approval of talking with the administrator doesn’t tell us anything about what she cares about with respect to Sheba, since considerations about Sheba aren’t among those that would factor into her agential psychology whatsoever when she so approves.
Note that Alessandra’s failure to think about Sheba’s being in the car is plausibly one of the things we ought to hold fixed when assessing wheth- er or not she has the disposition required for being attributionally- responsible for talking to the administrator according to the Minimal Ap- proval view. To see why this should be so, notice that people are plausibly attributionally-responsible for ϕ-ing even if there are other options that they do not remember that would silence their desires to ϕ were they to come up with those options. If I had remembered that I had spaghetti in the pantry I would never have given any weight to a desire to go to the grocery store, but this doesn’t mean that I’m not attributionally- responsible for going to the grocery store.
Alessandra never considers Sheba, and so there is no bridge that shows that just because she is attributionally-responsibility for talking to the administrator, she should be attributionally-responsible for forgetting Sheba. This leaves us needing to look elsewhere to explain the intuition that Alessandra is in some sense responsible for failing to tend to Sheba.