Your Mother and Modal Epistemology

Here is a problem with modal conditions on knowledge, as traditionally understood.  Some of the beliefs we form pertain to things that our own existence is ontologically dependent on.  Consider the following scenario:  

Two people, Timothy and Titus, look at a photograph of a woman and form the belief  She existed at some point.  Neither, let us suppose, know anything about the person in the photograph, however, as it happens, she is the mother of Timothy.

Were the belief false, Timothy would not exist.  As a result, in the closest world(s) in which the belief is false Timothy fails to form the belief, and there are no close worlds in which Timothy believes that proposition in which it is not true.  So Timothy’s belief is both sensitive and safe (according to traditional construals of sensitivity and safety), and necessarily so.  Yet, depending on contingent background facts about the photograph, there are situations in which Titus’s belief fails to be sensitive or safe.  So, according to traditional accounts of safety and sensitivity, the epistemic status of Timothy and Titus’s beliefs are different, but, according to common sense, this is not the case.

More on Sensitivity and Closure

Traditionally, sensitivity theorists deny closure.  In the last post I suggested that some anti-sceptical beliefs (e.g. I am not a BIV) which are often taken to be non-sensitive, are in fact sensitive, when the sensitivity condition is parsed so as to take into account belief-forming methods. This though doesn’t mean that there might not be some, more elaborately contrived, beliefs in which closure would fail, even given a version of sensitivity that takes into account belief-forming methods.

But closure failure could be avoided if we adopted a disjunctive account of knowledge, whereby knowledge is either sensitively formed true belief, or belief that is soundly inferred from a sensitively formed true belief.  Oftentimes disjunctive explanations (or disjunctive proofs) are seen as being less explanatory than non-disjunctive counterparts (as they are less unifying), but there is some virtue in this disjunctive account of knowledge.  It does justice to the holistic nature of our beliefs.  Beliefs, taken individually, may lack sensitivity or responsiveness to the world, but may constitute knowledge because of the way they are apperceptively integrated into a wider whole, of which some parts are appropriately responsive to the world.  It is well-known that coherentist constraints on knowledge, taken on their own, leave out the important thought that beliefs that constitute knowledge must in some sense be responsive to reality (“frictionless spinning in the void” and all that); but modal constraints on knowledge, taken on their own, may also leave out the important role that coherence-making relationships have with respect to knowledge.  It may be that both kinds of consideration must be built into an account of knowledge, but that neither can be understood in terms of the other.  In which case, a disjunctive account of knowledge would be in order.

Sensitivity and Closure

Kelly Becker, in his book Epistemology Modalized, gives a nice modal account of knowledge:

S knows that p iff:

1. p is true
2. S believes that p
3. S’s belief that p is formed by a belief-forming process or methodw that produces a high ratio of true beliefs in the actual world and throughout close possible worlds (reliability condition).
4. If p were false, S would not believe that p via the methodn S actually uses in forming the belief that p (sensitivity condition). (Epistemology Modalized, p.88)

Methodsw are individuated very narrowly, but not so narrowly as to include specific belief contents.  Specific belief contents are however included in methodsn.  Becker individuates a methodw as the narrowest specific-content-neutral method or process that is causally operative in belief formation.  An example of a methodw might be forming beliefs about which people are in the vicinity based on quick looks in at least dim lighting.  A methodn on the other hand might be something like If what I am looking at now has short legs and floppy ears (and such and so other features) then it’s a dachshund.

Becker also makes a serious and interesting case against closure under known entailment, and he takes it, as epistemologists generally do, that sensitivity is incompatible with closure:

The sensitivity component of our theory somehow predicts this result – I do not know not-[sceptical hypothesis] because, if it were false, I would believe it anyway. (Epistemology Modalized, p.120)

But in fact, it isn’t clear that his account of sensitivity is incompatible with closure.  Take a standard BIV case.  I believe I am not a BIV, yet 4 holds: if I was a BIV I would not believe that I was not a BIV via the methodn I actually use in forming the belief that I am not a BIV.  My method, after all, involves coming to know ordinary propositions about the world around me by interacting with it, and inferring from these ordinary propositions that I am not a BIV.  Since brains in vats cannot employ the same kinds of methods that embodied humans do, my belief that I am not a BIV is sensitive according to Becker’s analysis.

Safety, Sensitivity and the Value of Knowledge

Epistemologists who wish to provide a modal analysis of knowledge are divided over whether the safety or the sensitivity condition offers a correct necessary condition for knowledge.  The two are usually parsed in (something like) the following way:

Safety: S’s belief p is safe iff in nearly all (if not all) nearby possible worlds w in which S forms a belief p using the same belief-forming methods as the actual world, p is true.

Sensitivity:  S’s belief p is sensitive iff in the closest possible world in which S forms her belief using the same belief-forming methods as the actual world and in which p is false, S does not believe p.

Sosa suggested the following scenario, which has been influential in encouraging philosophers to favour safety over sensitivity:

On my way to the elevator I release a trash bag down the chute from my high rise condo.  Presumably I know my bag will soon be in the basement. But what if, having been released, it still (incredibly) were not to arrive there? That presumably would be because it had been snagged somehow in the chute on the way down (an incredibly rare occurrence), or some such happenstance. But none such could affect my predictive belief as I release it, so I would still predict that the bag would soon arrive in the basement. My belief seems not to be sensitive, therefore, but constitutes knowledge anyhow, and can correctly be said to do so. [‘How to Defeat Opposition to Moore’: 145-6]

Now, I’m not convinced that the condo owner really has knowledge in this case—as opposed to a true, justified belief, or knowledge of some closely related proposition, such as that it is highly probable that the bag is in the basement—and I also think that examples with the same structure as Sosa’s offer some evidence against the safety condition and in favour of sensitivity.

For instance, I once went to meet my supervisor on the fifth floor of our building, only to discover that his office was not there: he had moved to the sixth floor.  Two things strike me about this.  Firstly, it seems to me perfectly reasonable to hold that I didn’t have knowledge that he was on the fifth floor, even when this was true and I was justified (and safe) in believing it.  I was not properly connected to the facts: when the justified, true (and safe) belief became a justified, false belief, I was oblivious to the change.  Secondly, and relatedly, it is often thought that we value knowledge resides (at least in part) to its connection to felicitous action.  A mere true belief is, all things being equal, less valuable than knowledge because the vagaries of the world ensure that it has a tendency to turn into a false belief.  Known propositions are the sort that allow us to successfully navigate our environment, and false beliefs will (in ordinary circumstances) lead us to behave infelicitously (e.g. by going to the fifth rather than the sixth floor).  The thought here is that there are many scenarios akin to the one above, and, hence, sensitivity seems to be a necessary condition for the sort of belief that allows a person to successfully navigate her environment.