Friday, 27 April 2012

Outline of Feldman's and Conee's "Internalism Defended"


Richard Feldman and Earl Conee defend the claim that S’s beliefs are justified only by things that are internal to S’s mental life (henceforth, J-internalism) against two types of objections: (1) those that claim that there are some justified beliefs for which there are no internal justifying states, and (2) those that claim that there are insurmountable difficulties in specifying the nature of the connection between beliefs and their corresponding justifying internal states.  Feldman and Conee argue that neither class of objections pose an unanswerable challenge to J-internalism.


Step 1: Feldman and Conee begin by distinguishing between two types of J-internalism: accessibilism and mentalism (pp. 1-2*).

Argument: Accessibilism holds that the epistemic justification of S’s beliefs depend solely on factors to which S has reflective access.  Mentalism holds that the epistemic justification of S’s beliefs depend solely on factors that are internal to S’s mental life. Accessibilism and mentalism are coextensive only if we assume that S has reflective access to every aspect of her mental life relevant to justification.

Upshot: There are at least two broad kinds of J-internalism: accessibilism and mentalism.

Step 2: Feldman and Conee characterise mentalism in terms of a supervenience and a mental duplicate claim (p. 2).

Argument: The supervenience thesis (S) holds that the justification of S’s beliefs supervene on S’s occurrent and dispositional mental states, events and conditions.  The mental duplicates thesis (M) holds that if two individuals, S1 and S2, are exactly alike mentally, then they are exactly alike justificationally.  Since the claim that justification supervenes on mental states entails that two individuals with the same mental states have the same justification, (S) entails (M).

Upshot: Mentalism may be characterised in terms of a supervenience thesis and a mental duplicates thesis, where the former entails the latter.

Step 3: Feldman and Conee argue that mentalism is the preferable characterisation of J-internalism (pp. 2-3).

Argument: In the philosophy of mind, internalism is the thesis that the content of S’s attitudes depend solely on factors internal to the cognitive apparatus of the agent.  This is meant to exclude social and environmental factors as determiners of the content of a subject’s mental states.  In short, mental content supervenes on a subject’s inner states.  Likewise, mentalism is meant to exclude plainly external factors as determiners of justification.  In short, justification supervenes on a subject’s mental states.

Upshot: Mentalism preserves the parallel between J-internalism and internalism in the philosophy of mind.

Step 4: Feldman and Conee consider five examples in which a change or difference in the mental life of subjects result in a change or difference in justification (pp. 4-5).

Argument:  In the first four examples the justificatory standing of an agent changes depending on what items of information are within the subject’s ken. In the fifth example, a purely internal difference decisively determine the subject’s justificatory standing.

Upshot: While the five examples considered do not establish that J-internalism is true, they suggest that J-internalism is consistent with our pre-theoretical intuitions.

Step 5: Feldman and Conee distinguish between two broad classes of objections to J-internalism (p. 5).

Argument: The first class of objections (i.e., those put forward by Alvin Plantinga and Alvin Goldman) claim that there are some justified beliefs for which there are no internal justifying states. The second class of objections (i.e., those put forward by William Alston and Ernest Sosa) claim that there are insurmountable difficulties in specifying the nature of the connection between beliefs and their internal justifying states.

Upshot: The defender of J-internalism must respond to two broad classes of objections if the plausibility of her belief is to be maintained.

Step 6: Feldman and Conee consider Alvin Plantinga’s objections to evidentialist versions of J-internalism; the thesis that an agent is justified by her evidence (pp. 5-6).

Argument: Plantinga maintains that evidence must either be propositional, sensory, or impulsional. Since any type of evidentialism that held that evidence constituted by beliefs (i.e., propositional) or experiences (i.e., sensory) could not account for mathematical knowledge, J-internalism must also include evidence that is constituted by a sense of conviction (i.e., impulsional).  However, since a sense of conviction accompanies all beliefs, allowing for impulsional evidence would result in all beliefs being justified.

Upshot: There cannot be a plausible account of evidentialism.

Step 7: Feldman and Conee respond to Plantinga’s objection to evidentialism (pp. 6-7).

Argument: Feldman and Conee point out that: (1) We often believe things we find counterintuitive or implausible and disbelieve things we find intuitive and plausible. So it is not true that all beliefs are accompanied by a sense of conviction (or at impulsional).  (2) Even if all beliefs were impulsional, there may still be defeaters that outweigh the impulsional evidence of a given belief, rendering the belief justificationally impotent. (3) According to some plausible views, we have an aprori insight that allows us to grasp simple mathematical propositions.  This allows us to say that mathematical beliefs are justified without implying that all beliefs are justified.

Upshot: Plantinga’s objection to evidentialism fails.
Step 8: Feldman and Conee consider Alvin Goldman’s objection to J-internalism (p. 7.

Argument: Goldman argues that internal states cannot account for the justification of stored beliefs.  At any given time, there are many things we know that we are not considering. Since we know them, we also believe them, and the belief in question is justified. But given that the belief in question is nonocurrent, it could not be justified by a conscious mental state.

Upshot: J-internalist are stuck with the unacceptable result that stored beliefs are not justified, unless some internal justifier can be found.

Step 9: Feldman and Conee respond to Goldman’s objection to J-internalism (pp. 7-8)

Argument: Feldman and Conee argue that there are occurrent and dispositional senses of justified, just as there are occurrent and dispositional senses of belief.  Nonoccurrent beliefs may enjoy dispositional justification.  Moreover, there is no reason why the numerous ordinary justified beliefs a subject is not consciously considering could not justify her stored beliefs.

Upshot: Goldman’s objection to J-internalism fails.

Step 10: Feldman and Conee respond to Goldman’s objection to J-internalism based on the possibility of forgotten evidence (pp. 8-10).

Argument: Goldman points out that an agent’s belief may remain justified long after she has forgotten the original source of justification.  However, Feldman and Conee point out that: (1) the vivacity (or sense of certainty) that accompanies such a memory belief may provide justification for that belief, and (2) the agent’s justified beliefs about the reliability of her memory, may also provide her with justification for a memory belief even after she has forgotten the original justificatory source of the memory belief.

Upshot: The forgotten evidence objection to J-internalism fails.

Step 11: Feldman and Conee consider William Alston’s objection to J-internalism (p. 11).

Argument: Alston argues that if an agent believes some proposition, P, based on some evidence, E, then E can provide justification for P only if the agent has the higher-order belief that E provides justification for P.  However, typically, when we believe some proposition, P, based on some evidence, E, we never actually stop to consider if E justifiers P.  Consequently, if we buy into this view, then many of the beliefs we ordinarily take to be justified turn out not to be. 

Upshot: Given that J-internalsim requires higher-order beliefs about the justificatory efficacy of our evidence, many beliefs we ordinarily take to be justified turn out not to be.

Step 12: Feldman and Conee respond to Alston’s objection to J-internalism (pp. 11-13)

Argument: According to the evidentialist account favoured by Feldman and Conee, possessing the right evidence by itself secures the justification for the corresponding beliefs, independent of whether or not the agent considers the justificatory efficacy of said evidence.

Upshot: Alston’s objection to J-internalism fails.

Step 13: Feldman and Conee consider Ernest Sosa’s objection to J-internalism (p. 13).

Argument: Although some experiences directly fit the introspective beliefs that describe them (such as my experience of a white triangle against a black background), other experiences do not (such as my experience of a white 25-sided figure against a black background).  This poses the question: why does the experience of a triangle justify the introspective belief while the experience of a 23-sided figure does not.

Upshot: If the J-internalist is unable to explain this difference, then it remains unclear how, by their lights, any experience may be said to justify a belief.

Step 14: Feldman and Conee respond to Ernest Sosa’s objection to J-internalism (pp. 13-14).

Argument: Feldman and Conee point out that while we have the ability to perceptually recognise a triangle, we lack the ability to perceptual recognise a 23-sided figure.  However, if we imagine an agent who had the ability to perceptually recognise the latter, then that agent’s experience would be able to justify the corresponding introspective belief.  However, since such an agent would not be internally identical to us, this does not pose a problem for the claim that agents that are alike internally are alike in terms of justification.

Upshot: Sosa’s objection to J-internalism fails.


Given that J-internalism, broadly construed, suffers from no fatal defects, philosophers may move beyond the debate regarding whether or not J-internalism is feasible to flushing out a specific, detailed account of internalist justification.

*Page numbers based on the American Philosophical Quarterly article (Volume 38, Number 1, January 2001)


Russell Buehler said...

I don't have anything particularly insightful to say at this point; in general, j-internalism seems plausible enough, so I think Feldman and Conee succeed in at least doing that. In general, most of the objections to j-internalism didn't seem particularly good with Feldman and Conee's responses being...well...expected. In general, though, I tend to find j-internalism theoretically pleasing, so my thoughts aren't particularly surprising.

Benjamin Altshuler said...

The authors show that mentalism relies on the properties of supervenience and the causally entailed mental duplicate thesis. This means that as justification of a subject's beliefs change, so too does their mental properties. Can two individuals with the same justification differ in their mental states? I believe they can and further, I do not support the metal duplicates thesis.

F & C claim (rightfully) that their theory relies on the mental duplicates concept, though they do not successfully show that the thesis holds. My doubts stem from 1) the purely theoretical nature of such an argument and 2) the intuitive disjunction of beliefs and motivation.

While it impractical to consider a case of two mentally identical individuals, it provides an interesting case study. The truth is that it is impossible to have two separate individuals under identical mental states and conditions. Without actual subjects, conclusions remain hypothetical. The nearest approximation is to look at one person under nearly identical mental and physical conditions. Although distanced in time, When Johnny gets up for school on proximate days, he is in the same location, frame of mind, level of awareness, and presumably, he has the same mental states. All this and yet his justifications differ, as evidenced in that Johnny does not hold the same things that he knows to be true from one day to the next. Johnny's confidence and motivation also differ daily. Although Johnny remains the same in mental respects from one day to the next, these qualities do not appear to supervene on his justification.

Taylor ffitch said...

I'd like to reply to Ben. Your point is interesting but I'm not convinced that identical mental states don't produce identical beliefs. Your example of the same person waking up on different days is actually very far from being an example of nearly identical mental states. Think of all the things that happen during the day to change your mental state before you wake up the next morning. A huge amount of new information has been processed between waking up on day 1 and on day 2, even if environmental factors are constant. This is why a person has different beliefs and justifications on different days. It would be impossible to have two people in the same precise mental state, you're right, but this doesn't mean that a theoretical understanding of what that would mean isn't valid.

As for internalism in general, I'm pretty okay with it. The one thing would say is that since(if I understand correctly) externalism can include internalism within it, I'm not sure I see the point in narrowing things down and making life more difficult by excluding an entire realm of justification.

Benjamin Altshuler said...

Ah, thanks for the response. I am still not convinced that it is practical to base a theory in part on a completely theoretical hypothesis, but I see that it is productive to conduct the thought experiment. If I may amend, then, and ask what about an individual's decisions on a single day. In any given moment, Johnny (or anyone) makes decisions, feels emotions, and acts. If he was to relive that moment, I wonder if he would necessarily act exactly the same.

Such a question, like mental duplicate thesis, is purely theoretical, but certainly provides fodder for thought. Certain media examples explore this phenomenon called a time loop in a more casual way with "choose your own adventure" books and parallel universe reality plots. Characters have a certain quantity of knowledge, and likely the same basis for justification, yet small differences in action alter the course of the story arc.

Andrew Cely said...

This conversation got pretty lively all of a sudden, so I'll leave my two cents here.

I like Ben's point about waking up on separate days and occasions, as well as the whole reliving argument that shows how even in nearly perfect similar situations, a person would not have the same mental states if lived twice. However, I would ask this - does the person who is involved in this "Groundhog's Day" experience remember that they have already relived this time frame? If so, then how can you assume that things won't turn out in exactly the same way under exactly the same circumstances. Technically, you start out with the same mentality waking up, so your thought process, dreams, and beliefs are all technically the same in every reliving. In this sense I look at Taylor's approach and understand her convictions that such an exact precise mental state is not only unattainable, but also not totally invalid.

I appreciate internalism for what its worth, but overall, I find that I have more sound evidence for knowledge when viewing things from an externalist standpoint.

Russell Buehler said...

First, I would like to plead sleep deprivation on my first post. Second, I disagree with everyone...okay, I mostly agree with Taylor, but only mostly. Now that that's out of the way, I'll take each post in turn. In response to Benjamin's first post, I must disagree with the contention that the purely theoretical nature of an argument should be a reason to doubt it; although my intuitions may be skewed, (I believe that) purely theoretical arguments often have a force that empirical ones do not. Consider, for example, mathematical proofs or the argument that 'all bachelors are unmarried men' is true in virtue of its meaning. Both seem excellent candidates for purely theoretical arguments, but I struggle to believe that there is something questionable about them merely in virtue of this. Of course, you may very well consider this a purely theoretical argument...

In response to Benjamin's consideration of two mentally identical individuals, I must submit that it is not actually impossible--at least not broadly logically impossible (or, to my mind, even nomologically impossible). There is no contradiction implied in such a situation. It may be impractical, but impossibility seems far too strong. In terms of the remainder of Benjamin's first post, I find that I agree whole-heartedly with much of Taylor's response; that is, of course, until the impossibility of two people having precisely the same mental states is broached again. As far as reasons for preferring internalism even though it is a subset of what's available to externalism, I believe there are a number of intuitive reasons for preferring the least powerful solution, the smallest number of justifiers we may admit to achieve a correct analysis. This is, for example, the same reason why it's more satisfying to prove geometric facts without invoking unnecessary and powerful set theoretic axioms.

Finally, for Benjamin's second post and Andrew's post, I must again plead a difference of opinion; in particular, it seems quite intuitive to me that our repeating individual will be justified in believing exactly the same things if their mental status has not changed. Note, however, that this doesn't mean they will necessarily act in the same way; after all, if I take two mentally identical people, and then start chasing one with a baseball bat, I wouldn't expect the other to also start running around. Different external stimuli would very well be expected to produce different responses. In Benjamin's thought experiment, then, it's necessary to both stipulate and be aware that, from the moment our repeater starts, there can be no difference in external stimuli. Keeping that in mind, I struggle to see where any difference between the first and second runs could arise from. Last, but not least I would push that the suggestion in the paper wasn't for knowledge internalism, but rather for justification internalism.