Nov 5: Epistemic justification
|
|
|
To do for next class - 1) Read chp. 4; 2) Second essay due. |
To do for next class - 1) Read chp. 5, pp. 114-125; 2) Try question 3 of the exercises. |
Introduction
Do you recall from last week's discussion the three different ways of analyzing a concept? (Definition, analysis of complex concepts, criteriological analysis). Do you also recall our attempt at performing a criteriological analysis like Chisholm's, of the concept of justification? What problems did we run into? (Could not find a definition that didn't either have nonevaluative terms in the antecedent, or we have difficulty defining prima facie in the consequent). Today will take a different tack and examine competing definitions for justification has a simple concept.
To begin, there is good reason to believe that epistemic justification is a normative concept. Why? (Because it is an "ought" claim is not an "is" claim, i.e. what we ought to do to avoid error). As a normative concept, it is not surprising that many definitions of it are deontological ones (i.e. theory or study of moral obligation - webster's). That is definitions which use terms like "duty", "obligation", "should", and "ought". In this case we're not discussing moral duties, etc. but rather epistemic ones. That is, duties which have as in and seeking truth and avoiding falsehood rather than doing right and avoiding wrong. In contrast, non-deontological definitions agree as to the normativity, but think that a believes being justified does not have anything to do with duty (like good cars, pigs, or birthday cakes). Thus justified are unjustified beliefs are simply good or bad with respect to the end of having true beliefs. There is no duty imposed on the believer to have true beliefs. Steup provides an example of this kind of theory by quoting Alston's definition.
Deontological justification
Both Descartes and Locke agree that as rational animals it is our duty to seek the truth. That is, it to believe only those things for which we have good evidence. If we believe things randomly we should be blamed for our irresponsible behavior. As Steup points out, Chisholm heartily agrees. What do you think? Is it our duty to seek the truth as it is our duty not to harm others?
What duty, precisely is it that this trio expects us to meet? Obviously it isn't that we should believe what is true and disbelieve what is false. There are far too many cases when we believe falsehoods which it only makes sense for us to believe the given the circumstances. However, this is the end we are striving for. Characterizing is end is not a simple as it seems. After nitpicking his way past one definition (that of Chisholm) Steup suggests the plausible "for every proposition it is my goal to believe it is an only if it is true". This sense reasonable, until we realize there are two assertions here. First, it is our goal to believe only what is sure but second it is our goal to believe everything that is true. However does it make sense to have as an end believing every truth? If it were, we would no doubt get lost in a slew of irrelevant truths. So rather, believing truly is to believe only propositions relevant to me if and only if they are true.
So what, then, is the duty? For Locke it is to "governed our assent as reason directs us". Steup also provides a second example, courtesy of Richard Feldman, he says one must "believe what is supported or justified by ones evidence and to avoid believing what is not supported by ones evidence". The question arises if this is simply a truism. That is, the do we mean by saying one ought to believe truly that one ought to believe in accordance with ones evidence? The answer is no. We could try to believe truly by reading tea leaves, or talking to our personal astrologer. There are, however to different ways of meeting this duty. Do you recall them? (The duty to believe p or the duty to refrain from believing q). So to say we are justified in believing something is a weak claim, it means only that it is not the case that I ought not to believe it. Which doesn't mean I ought to believe it.
This distinction lead Steup to identify two separate projects of the deontological characterization of justification. First is a weak one which we have seen, defining the meaning of the term. The second, stronger project, tries to determine when it is our duty to believe something. Is that difference a clear one? Consider the example of believing that there are four people in this room. According to the first definition I should not disbelieve this proposition. According to second it might be my duty to believe this proposition.
But it seems there might be a problem for both projects which arise from the involuntariness of belief formation. In particular, it does not make sense for a duty to apply to something which is involuntary. So, if our beliefs are involuntary it does not make sense to consider duties with respect to them. (Note that the word doxastic just means things which relate to belief). Can you think of some examples of involuntary believe? (Belief that the light is on, the blackboard is dark green, that you are thinking). Steup claims "these are beliefs that I could not manage to get rid of, no matter how hard I tried". Does this mean Descartes its method of doubt was impossible? Did he not really doubt the existence of the external world? Are these really beliefs that we can not possibly disbelieve? This is an important question because if we answer them positively the deontological project would be doomed. In particular, it would not be able to account for such beliefs at all. Beliefs which seem extremely common and which must be captured by any theory of justification. What do you think of the Elvis Presley counter example? What counts as voluntary and what does not seems the rather arbitrary, and up to Steup. Can you think an example which really seems involuntary? Note, that a lot of what matters here is what kinds of belief count a priori as justified. For Steup obviously introspective and perceptual beliefs do but this rests on his particular definition or theory of perception.
Quick Review: What two aspects are there to the deontological definition of the concept of epistemic justification? (One, we are committed to the end of believing truly. Two, this imposes on us that epistemic duty.) If something is a duty, what principle generally accompanies our acceptance of it? (The voluntariness principle). So, now that Steup has given us a counter example (an unjustified belief which meets the D1 critieria on p. 76) what are our options?
Feldman's solution
One obvious way of saving the deontological definition is to reject the voluntariness principle. On what basis could we reject this principle? If we can find an analogous situation in which obligations simply cannot be met but remain obligations, we can construct an argument from analogy. So, Steup presents the example of the student with a learning disability was unable to pass exams for a logic class. The question becomes, said the student get credit and for class without passing the example? Steup answers of course not. A second example is provided by a legal obligation. In essence, if someone enters into a contract and that they are unable to fulfill, they are nonetheless obligated to fulfill it (as in the case of a mortgage). So Feldman argues that the epistemic duty is similar. Whether or not we are capable of believing in accordance with evidence, it is still an obligation of ours. So this is one solution in to the problem with the deontological definition. What other problem does this also solve? (That of having positive epistemic duties)
Of course, there will be those who do not wish to reject the voluntariness principle. So, we could consider that principle to be a necessary condition on justification. In other words, someone is justified in believing something if and only if they voluntarily believe it and do not have an epistemic duty to refrain from believing it. This definition arises from our use of the term "justification" when discussing acts (though it sounds a little strange to say I am justified in walking from here to the bathroom). This definition then avoids the Elvis problem since the belief is not voluntary. What now of the involuntariness of perceptual introspective beliefs? Began, according to Steup's definition or theory of perception they will provide a large problem since they are justified and involuntary. But, according to this definition they are not justified. As Steup notes a deontologist might wish to distinguish between different kinds of involuntariness in order to save this definition. Do you recall the difference between soft and hard involuntariness? (Soft involuntariness is when we are still in able to change our beliefs using our intellectual faculties, that is given other evidence our beliefs might change; in other words the belief is under our intellectual authority). So, the deontologist has a difference solution with this distinction, that is softly involuntary beliefs can still have epistemic duties.
Truth conduciveness and probability
Steup presents an argument from Alston which intends to show that epistemic deontologist them is not truth conducive. What does truth conduciveness mean? Something is truth conducive if it makes it probable that propositions stemming from it are true. To give ourselves one more principle and, Steup introduces the "adequacy principle". It simply states that for a belief to be justified it must be based on an adequate ground. What is the ground? (Other beliefs or experiences). What makes a ground adequate? Truth conduciveness. As a criticism of the deontological argument, we can recast this as showing that deontological justification does not have to the truth conducive or make the beliefs probable.
What to kinds of probability that Steup discuss? (Epistemic and factual). What differences are there? To begin what are factual probabilities? These are probabilities whose value depends only on how things are in the world. That is, it depends only on the relative frequency of the given occurrence with respect to another occurrence (e.g. role of the dice, possible coin, deaths by banana, etc.). Epistemic probabilities, in contrast, our degrees of justification relative to a body of evidence and attach to propositions. Thus, the epistemic probability will depend on the evidence available to the agent. Note then, that Alston is using the term probable and factual sense. Thus, perceptual beliefs are typically probable and because there are laws of nature which determine that such beliefs and are likely to be probable.
With this kind of probability in mind and, however, Alston would not have agreement with the deontologists regarding his adequacy principle. The recall the example of the doctor in the remote countryside with the two kinds of medication? Essentially, based on her or evidence the best medication to provide is A but based on the factual status of the case the best medicine is clearly B. according to Alston she would not be justified in providing medication A. However this would put him in a position of asserting that no false beliefs could be justified. (Why would this be a bad position to be in? It makes requirements for justification to stringent) so the deontologist reasonable position of claiming that the doctor's belief was epistemically justified even no false and, a position not allowed by Alston's criteria.
What, then, if Alston try to use epistemic probability rather than factual probabilities? In that case, than the deontologist could reply that it is impossible for one to belief in in accordance with the evidence while not believing what is epistemically probable. To so in reply to Alston's original criticism that deontological justification is not truth conducive, the deontologist can reply that, in fact, it is epistemically truth conducive, simply not factually truth conducive which is an undesirable result anyway.
Internalism and externalism
The distinction just encountered between the deontologists and Alston lies in a yet deeper distinction between internalists and externalists. The internalists/externalist distinction cuts across epistemology, philosophy of mind, cognitive science and psychology. In all cases, internalists posit that only things inside the head in determining what is being discussed (e.g. justification, meaning, contents, etc.). Externalists posit the opposite. Notice that this characterization might be slightly different from that offered by Steup. He says a position is internalist if it allows any condition to be an internal one (this is a weaker definition). It is also not clear to a that coherentism and foundationalism are both internalist theories (will spend the next two classes discussing these theories). To make externalism seen little more reasonable than perhaps Steup allows it considered the following externalists explanation for meaning. (Give twater example, XYZ only formulaically different than H2O, Tyler Burge, Putnam, meaning depends on external characteristics).
Steup argues that internalism and makes more sense than externalism since justification is something which should only depend on information currently available to us. In other words, a duty can only be determined by information accessible at the time we must act. Thus our epistemic duty in can only rest on evidence currently available to us via reflection, according to the deontological theory. But, what if the meaning of all of these pieces of evidence depend on external factors?
To do for next class
1) Read chp. 5, pp. 114-125; 2) Try question 3 of the exercises.