Main authors discussed: Keith DeRose, John Hawthorne, Jason Stanley, and Timothy Williamson

Hawthorne and Stanley (2008) defend the Knowledge-Action Principle:

(KA) It is proper to treat p as a reason for action (for some p-dependent choice) iff you know p.

Mikkel Gerken defends the Warrant-Action Principle:

(WA) In the deliberative context, DC, S meets the epistemic conditions on rational use of (her belief that) p as a premise in practical reasoning or of (her belief that) p as a reason for acting (if and) only if S is warranted in believing that p to a degree that is adequate relative to DC (2).

Gerken takes WA to be an alternative to KA because he assumes (reasonably) that warrant does not suffice for knowledge.  He is using ‘warrant’ as a term that could cover ‘justification’ and ‘entitlement’ and believes that there can be warranted false beliefs and Gettiered beliefs.  He brackets the question as to whether anything known to be true has DC-adequate warrant, but notes that some examples of Jessica Brown’s cause trouble for (KA) on this front.  His discussion and this discussion will be focused the claim that something less than knowledge is needed for S to be in a position to treat p as a reason for action in such a way that this is epistemically proper.

Gerken argues that Gettier cases and cases of warranted but false belief cause trouble for KA but not for WA.  He also thinks the motivation for KA is problematic.  Hawthorne and Stanley offer some examples in support of (KA) such as this one:

Suppose, for example, that Hannah and Sarah are trying to find a restaurant, at which they have time limited reservations. Instead of asking someone for directions, Hannah goes on her hunch that the restaurant is down a street on the left. After walking around for some amount of time, it becomes quite clear that they went down the wrong street. A natural way for Sarah to point out that Hannah made the wrong decision is to say, “you shouldn’t have gone down this street, since you didn’t know that the restaurant was there (Hawthorne and Stanley 2008: 571).

Gerken thinks that our use of ‘knows’ in such contexts is often imprecise, so these examples offer little support to (KA).  He also notes that if we modify the details of the case, it seems we can use this sort of case against KA.  Rather than have Hannah act on a hunch, suppose she has exceptionally good evidence but mistakenly believes that the restaurant is where it’s been for years.  It seems reasonable for Hannah to go down the street although she does not know that the restaurant is there. This version of the restaurant-case appears to be a counter-example to The Knowledge-Action Principle. Hannah reasonably treats p as reason for acting although she does not know that p. Had Sarah complained, “you shouldn’t have gone down this street, since you didn’t know that the restaurant was there,” she would be unreasonable. Note that this verdict is based on a relevantly similar judgment about a case that is relevantly similar to the original version of the case (Gerken forthcoming: 5).

My initial reaction is similar to Gerken’s.  It would be inappropriate for Sarah to complain, upbraid, blame, etc… if she knows that Hannah’s belief was supported by the evidence.  However, that might be because Hannah should be excused for what she did and not because she rightly or permissibly reasoned from a false premise with regrettable consequences.  It isn’t clear that KA delivers us the wrong verdict.

(Just a quick sidenote.  Gerken thinks that Gettier cases cause trouble for KA, too.  Boiled down to its essentials, the idea is this.  In (some) Gettier cases, it seems proper for the subject to treat p as a reason for action when p is justifiably believed on the basis of good reasons but only accidentally connected to the truth.  KA delivers the verdict that it is impermissible to treat p as a reason for action in such cases, but that seems wrong.  Here, I think Gerken is on good ground.  Those who defend KA can say that you are merely excused for treating p as a reason for action when your belief that p is Gettiered, but I don’t feel the pull of the intuition that there’s something to excuse that is wrongful.  I’ll focus on the case of warranted, false belief from here on out.)

Gerken argues that it’s a mistake to classify cases of reasoning from false beliefs as merely excused. (One of the reasons this paper comes highly recommended.)    Suppose someone defend KA by saying that the subject should be excused if she has every reason to believe she knows p.  Gerken thinks that this is problematic:

Not all agents are capable of such higher-order thoughts … The more principled problem is that many subjects who may act or deliberate rationally on the basis of a warranted false belief are incapable of thinking the second-order thought. For example, a 30-month-old toddler may see a wax apple on the table and believe that there is an apple on the table. He may then use that belief as a premise in basic practical deliberation. Likewise, a chimp that sees a wax banana may falsely believe that there is a banana and act on the belief. But there are empirical reasons to suppose that 30-month-old humans and chimps cannot competently employ higher-order thoughts, such as I know that there is an apple on the table. If some subjects that may be rational in acting on a warranted false belief are incapable of forming the higher-order thoughts required for secondary propriety, the appeal to secondary propriety/excuse is less appealing (Gerken forthcoming: 11).

Those who defend KA might say two things here.  First, that there are two ways to show that blame isn’t fitting.  There is a distinction between excuses and denials of responsibility (or exemptions).  You offer excuses for wrongs committed by those that can be held responsible.  You offer denials of responsibility for individuals that lack the capacities that allow them to assume responsibility for their deeds.  It doesn’t follow from the fact that someone doesn’t meet the requirements that would allow them to assume responsibility for their deeds that their actions cannot be excused.  It doesn’t follow from the fact that they do not meet the conditions necessary for excuse that they must be held responsible for any breach of any norm.

Second, they can say that what matters is not whether the subject can form beliefs about the reasons and entertain propositions about what they can know but whether the reasons are there for them to make a reasonable judgment about what should be done, may be done, etc…  Somewhere between chimphood and voting age there may well be stages where children can be held accountable for their deeds where they may well not entertain propositions about what they know but may know various things, know various things about right and wrong, and if they have the sorts of reasons that put them in a position to know that they would act rightly, why not excuse them for their wrongful deeds?  Gerken anticipates this second response and raises the worry that it would force us to say that there can be warrants we have for beliefs in propositions we cannot entertain, grasp, or believe.

Perhaps the most promising thing to say in defense of KA is that the subject should be excused for treating p as a reason for action provided that the subject satisfies the requirements imposed by WA.  Gerken’s response:

[T]his approach is little but a collapse of the knowledge account into a warrant account, such as WA. It is compatible with a warrant approach to hold that the degree of warrant normally required for knowledge is what is normally required for action/practical reasoning. But a central reason for distinguishing between “warrant normally required for knowledge” and “knowledge itself” is this: In certain epistemically abnormal cases, the degree of warrant that would have met the warrant condition on knowledge in normal circumstances is insufficient for knowledge. In some such cases, the warrant may nevertheless be sufficient for rational action/practical reasoning. As mentioned, cases of warranted false belief and Gettier-style cases are paradigms of such epistemically abnormal cases. So, to say that knowledge is normally required for action/practical reasoning but that warranted belief will do in abnormal cases amounts to accepting a version of a warrant account compatible with WA (Gerken forthcoming: 14).

Although I think that there are ways to meet Gerken’s challenge, he has shown that it will be difficult to bring (KA) in line with intuitions about Gettier cases and cases of warranted, false belief.  The excuse maneuver that seems so tempting will be difficult to develop in a satisfactory manner.

There is much in this paper that I couldn’t cover because of space constraints, but I recommend it to those interested in debates about the epistemic norms governing action, assertion, and belief.

Reviewed by Clayton Littlejohn
University of Texas, San Antonio

__

Hawthorne, J. and J. Stanley.  2008.  Knowledge and Action.  The Journal of Philosophy, CV, 10: 571-
90.