Externalism and Privileged Access

Externalism about mental content is the thesis that what one’s mental content (at least where atomic natural kind concepts are invoked) is individuated in part on the nature of that individual’s physical and social environment (what is out there and how words are used) – things external to the cognizer. In other words, we could fix everything internal to the cognizer, yet by changing things in her physical and social environment the content of at least some of her thoughts could be altered. Twin-Earth thought experiments are used to motivate externalism about mental content.

This line of thought seems to be problematic for the traditional doctrine of privileged self-knowledge. The idea here is that without the aid of empirical investigation, if one is thinking about p, then one can know that one is thinking about p. The claim is that I can know a priori that I am having a thought (when I am) and that my thought has the particular content that it does, for instance: that water is wet.

Here is an argument that externalism and privileged access are incompatible:

1. It is a priori knowable that that if one is thinking that water is wet, then one has the concept of water.
2. It is a priori knowable that the concept of water is an atomic natural kind concept.
3. It is a priori knowable that if a concept is an atomic natural kind concept then it is metaphysically impossible to possess it without having interacted with instances of that concept.
4. Therefore, it is a priori knowable that if one is thinking that water is wet, then one has causally interacted with instances of water.
5. It is not a priori knowable that one has causally interacted with instances of water.
6. Therefore, it is not a priori knowable by one that one is thinking that water is wet.

Premise (1) seems true since one must possess the concepts ingredient to the propositions that one thinks. I would fail to have the thought that water is wet if I failed to possess one or more of the concepts ingredient to the proposition water is wet.

Premise (3) follows from externalism about mental content. Which atomic natural kind concept one possesses is determined at least in part by external (physical and/or social) environmental factors.

Premise (4) is a sub-conclusion that follows from premises (1) – (3).

Premise (5) is clearly true. One cannot know a priori that water exists or that one has causally interacted with it. Empirical investigation is needed to know either of these claims.

The conclusion, line (6), is incompatible with the privileged access thesis. It states that what one is thinking is not a priori knowable to them in this case.

The compatibility of externalism and privileged access thus rests on premise (2). The claim is that it is a conceptual truth that water is an atomic natural kind concept. An atomic concept is one that lacks conceptual constituents. While it can often be known a priori that a concept is not atomic, it is by no means clear that one can know a priori when a concept is atomic. As such, this gives us some reason to doubt the second premise. In addition, it is much more difficult to see how one is supposed to know a priori that the concept water is a natural kind concept. While water is a natural kind, it does not follow that this fact is knowable a priori, and it seems quite likely that such proposition is not knowable a priori. After all, it is epistemically possible that water not be a natural kind, but rather not be a kind, like air or jade; or even fail to have an extension at all, like phlogiston. In other words, one must know that water is ‘Twin-Earth eligible’, and this is something that cannot be known a priori.

The incompatibilist can respond, as Boghossian does, by claiming that so long as a term aims to be denote a natural kind, something that is knowable a priori, then it must have a non-empty extension. However, it is not clear why externalists are committed to the claim that concepts that aim to denote a natural kind are only possessed by individuals who have causally interacted with the extension of that concept. In other words, why take the externalist’s thesis about natural kind concepts to also apply to concepts that aim to denote natural kinds? Since concepts that aim to name a natural kind can fail to name a natural kind, it is hard to see how this response helps the incompatibilist since it broadens the externalist’s commitments.

Since the externalist’s claims are particular to atomic natural kind concepts, then in order for the argument to go through from the fact that I have a thought involving a concept to the existence of an extension of that concept, one must know (a priori) that the relevant concept is of the relevant kind (an atomic natural kind concept). We have see reason to doubt that this can be known a priori, and if this cannot be known a priori, then the argument for the incompatibility of externalism and privileged access as typically construed fails.



Process reliabilism (hereafter reliabilism) is a theory of epistemic justification that claims that a belief’s justificatory status is determined by the reliability of the process that formed the belief. Here, a process is a functional operation that maps inputs to outputs (ie. input: experiential state; output: belief), and the processes we are concerned with are cognitive processes. A process is reliable if and only if it would yield a high proportion of truths over a wide variety of situations of the type we typically encounter (Alston). Beliefs that are the product of reliable belief-forming processes are justified, and justified to the extent that the process is reliable, while those beliefs that are not are unjustified.

Goldman gives the following reliabilist account of justification:

1. If S’s belief that p at t results (immediately) from a belief-independent process that is (unconditionally) reliable, then S’s belief that p at t is justified.

2. If S’s belief that p at t results (immediately) from a belief-dependent process that I (at least) conditionally reliable, and if the beliefs (if any) on which this process operates in producing S’s belief that p at t are themselves justified, then S’s belief in p at t is justified.

A belief-independent process is a process that does not have a belief state as an input. A belief-dependent process has at least one belief state as an input. A process is conditionally reliable when a sufficient percentage of output beliefs are true given that the input beliefs are true.

Goldman complicates this account somewhat in trying to account for defeaters. He claims that the justificatory status of a belief is also a function of the processes that could or should have been employed. Thus, for a belief to be justified it must also be that there was no reliable or conditionally reliable process available to the agent which had it been used in addition to the process actually used would have had the result that S did not believe that p at t.

At first glance, reliabilism is promising. For many justified beliefs one is able to identify a reliable process type that brought it about, and for most unjustified beliefs one is able to identify an unreliable process type that brought it about. Although this account of epistemic justification has some initial intuitive appeal, it faces substantial criticisms. Perhaps the biggest problem facing reliabilism is the Generality Problem. Any particular belief-forming process is a token of an indefinite number of types, and these types often differ radically in their reliability. The problem is that in order to evaluate the claim of reliabilism we need an account of which process type(s) is the relevant type – the type which determines the justificatory status of the belief.

For example, my belief that there is a computer in front of me is the result of a process of all of the following types: visual processes, done on Wednesday processes, processes that lead to true beliefs, processes done at 4:30pm, processes done by a Canadian. Understood as the result of some of these processes, my belief is unjustified; yet understood with other process types as being relevant, my belief is justified. The challenge for the reliabilist is to give a principled account that tells us what the relevant process types are. This task is difficult since if the relevant process types are construed too narrowly, there will only be one token of the relevant type. As such, all true beliefs would be justified and all false beliefs would be unjustified (since reliability would be determined on the basis of only one case). But this is clearly not the case. There are justified false beliefs and unjustified true beliefs. On the other hand, if the relevant process type is construed too broadly, then too many beliefs will inherit the same justificatory status, though the resulting beliefs seem to be of differing statuses. For instance, if perception were a relevant process type, then all beliefs that resulted from perception would be equally justified, but clearly this is not the case. Perception leads to better justified beliefs when the cognizer is in good lighting and looking at medium sized objects (in contrast to tiny objects in the dark).

Alston responds to this problem by defending the claim that for each belief forming process there is only one process type that is ‘psychologically real’, and it is this process type that is relevant for assessing justification. The process that is psychologically real is the one which corresponds to the actual inputs and outputs of the belief forming process. If we are concerned with my belief that there is a computer in front of me, this is the actual output, and the input was aspects of my particular perceptual experience. Alston’s claim is that there is only one function that maps this input to this output – that function is the relevant belief forming process type.

Although this response limits the possibilities for which process type is relevant, Alston has failed in reducing the list to one candidate. There are multiple operative psychological types since the operational function can broadened by broadening the input and outputs (without changing what the inputs and outputs are). The output that I believe there is a computer in front of me is also the output of a function that takes the same input and whose output is either this belief or my belief that there is a typewriter in front of me. Both the input and the output of a function can be broadened via disjunction (among other ways) and this allows for more and more functions (process types) to be active in a case of belief formation. We still need a way of determining which of these ‘active’ functions or processes is the relevant one – the one by which we assess the resultant beliefs justificatory status. Until we have such an account, reliabilism cannot get off the ground because we cannot see what it says about various cases.


Strong Foundationalism

Foundationalism is a theory of epistemic knowledge and justification that provides an answer to the infinite regress problem of justification. According to foundationalism, all beliefs that are epistemically justified are either basic, or trace back in a suitable way to a basic belief. Basic beliefs are beliefs that are epistemically justified independent of any justificatory relations to other beliefs. These basic beliefs serve as the stopping point on the chain of epistemic support so as to avoid an infinite regress regarding justification.

An account of basic beliefs must state which beliefs count as basic, and in virtue of what are these beliefs epistemically justified. Among foundationalists there are disagreements regarding both of these issues (as well as the issue of which relations are suitable for conferring justification to non-basic beliefs). Timothy McGrew defends a strong foundationalist account of basic beliefs. McGrew identifies three classes of basic beliefs: (i) second-order beliefs about beliefs (the belief: ‘this is a belief of mine’), first-order beliefs about immediate sensations (the belief: ‘I am experiencing this’), and beliefs regarding the content of one’s memory (the belief: ‘this is how it seems to me (that it was)’).

The claim is that these beliefs are justified independent of any justificatory relations to other beliefs since they are incorrigible or infallible or certain – one cannot be mistaken about their truth (or, necessarily, if believed, then they are true). The truth of these propositions is guaranteed in believing them through the way that their content is accessed by the believer – by being referentially formed (this prevents all beliefs in necessary truths from being basic). If there were nothing for ‘this’ to refer to in the belief ‘I am experiencing this’, such a belief could not be formed (a proposition would not be there); and when ‘this’ does refer, the belief must be true in virtue of reference (what one is experiencing will be picked out by ‘this’, and one is experiencing what one is experiencing).

Several problems emerge for this strong foundationalist account of basic beliefs. First, it seems that beliefs about one’s inner states (current experiences, beliefs, memories) are rare. We typically form beliefs like ‘there is a computer in front of me’ and beliefs of the form ‘I am experiencing this’, yet these latter beliefs are to be the foundation of our justified empirical beliefs on the strong foundationalist account. If these beliefs are rare, then there is not much of a foundation to ground the justification of non-basic beliefs. As such, the strong foundationalist account of basic beliefs would have the result that not many of our beliefs about the external world are justified.

The strong foundationalist can respond by claiming that such basic beliefs are typically present, but that they simply are not often consciously entertained. As such, these beliefs would be tacit or dispositional or simply subconscious. This response brings in complicating details regarding the nature of belief, but an argument is still needed regarding why one should think that there are such tacit or dispositional or subconscious beliefs. McGrew’s claim is that having an experience is equivalent to or entails having a tacit referential belief about that experience.

This claim is hard to accept. Having an experience or a belief is a distinct mental state from having a belief about that experience or a belief about that belief. As such, it seems to be at least logically possible for the two distinct states to come apart: that one could have the experience without the belief, or that one could have the belief without the experience. Either possibility seems to lead to trouble for the strong foundationalist.

If one can have the belief about the experience without the experience itself (or alternatively, the belief about the belief without the belief itself), then these beliefs are not incorrigible. These beliefs about experiences can go wrong since they can be had without the relevant experiences. However, although one may be able to believe the same proposition as the referentially formed belief without having the relevant experience, one is not able to have that belief in the appropriate indexical fashion. As mentioned above, is there is no experience for the ‘this’ to pick out, there is not a proposition there to be believed. As such, the defender of strong foundationalism can still claim that the reflectively formed beliefs are incorrigible despite the fact that the same proposition, if believed under a different guise would not be incorrigible (propositional incorrigibility is to be distinguished from doxastic incorrigibility).

A deeper problem lies in the possibility that one may have the relevant experience without forming the referential belief. Since the belief and the experience seem to be distinct states, this too seems to be a logical possibility. It could be that two individuals have the same experience of a book and each form the belief that there is a book in the room, yet only one of them forms the relevant tacit and reflectively formed belief about that experience. In such a situation, strong foundationalism has it that only the individual that forms such a belief can be justified in the belief that there is a book in the room. This seems to make an individual’s epistemic justification an accident of an individual’s psychology. In this case, each individual seems to have as good of evidence for the belief that there is a book in room as the other, and it is hard to see how a tacit belief could make such a justificatory difference. Given such a possibility, the strong foundationalist account of basic beliefs seems to get it wrong.


Want to Feel Challenged?

Here are two challenging words I've come across recently:

1. From John Piper: Ever wonder why 1 Peter 3:15 (Always be prepared to give an answer to everyone who asks you to give the reason for the hope that you have) has lost its applicability? When is the last time someone asked you the reason for your hope within? The reason no one is asking is that it looks like we are hoping in the same things that they are hoping in (jobs, savings accounts, family, friends, etc.). We need to live lives and utilize our money/time/family/etc. in such a way as it makes people ask because it is clear that our hope is not in what their hope is in.

2. This, from Anthony Bradley: He wonders how there can be 115,000 orphans in the US while there are 224 million Christians.


Yeah, Sarcasm is Really Terrible

I notice a lot of people saying that they do not want their kids to be sarcastic, and they give advice as to what to do to prevent your kids from becoming sarcastic. What's the big deal about being sarcastic? And no, that was not sarcastic.

Sarcasm is a linguistic device and it is hard to see what separates it from others that accounts for why it is to be avoided. Let's compare as we dare sarcasm with onomatopoeia. The latter is the formation of a word via the imitation of a sound ('boom', 'plop', 'zing', etc.). I take it that there are not people out there trying to get their kids to avoid onomatopoeia.

There are contexts in which sarcasm is inappropriate, yet the same is true of onomatopoeia. In fact, there is a great overlap as to which contexts these are. So, this cannot be the reason that sarcasm is to be avoided.

Sarcasm, you might think, has an annoying sound to it when said with that sarcastic tone. First, it is not clear that sarcasm requires such a tone. Second, many instances of onomatopoeia are instances of annoying sounds. After all, who likes to hear 'plop' or 'squish'? So, this cannot be the reason that sarcasm is to be avoided.

It may be true that sarcasm is often employed in a negative way -- used to disrespect someone. It can't be the mere fact that sarcasm can be used in such a way that it is to be avoided since many linguistic devices that are presumably not to be avoided *can* be used in this way. Onomatopoeia can be used in this way. Further, sarcasm needn't be negative in this way. I could utter to a fellow employee, "yeah, your not going to get the promotion" sarcastically, yet what I say is far from negative, in fact it is a complement to my associate. So, sarcasm needn't be negative in any relevant way. We are left with the claim that sarcasm is often used in a negative way, but why should this affect whether or not it should be used at all? If most people used pencils to poke themselves in the eye, it would be a bad thing, but nothing would follow regarding my using a pencil.

As a result, I can see no reason why sarcasm ought to be avoided. It should be responsibly used, but this is a different issue -- one that takes a lot more thought.


Is Knowledge Closed Under Known Logical Implication?

Closure is the epistemic principle that for any propositions p and q, if someone knows that p, and that p entails q, then that individual is in an epistemic position to know q – that individual has sufficient evidence to know q. In other words, were that individual to believe q on the basis of her knowledge that p and that p entails q, she would thereby know q. The basic idea is that we can add to what we know by deducing things from what we already know. Such a principle seems obviously true and denying it appears to be an extravagant cost to any epistemological theory. Nonetheless, closure has been under attack in recent literature in epistemology. Closure has principally been denied as a way of avoiding skepticism. The skeptic cannot get to the conclusion that I do not know that I have hands, even if I do not know that I am not radically deceived by an evil demon, if closure is false (if it is false that we have knowledge of known entailments of what we know).

Fred Dretske gives a case like the following as a purported counterexample to closure. Joe goes to a cage marked ‘zebras’ at the zoo. Joe looks at the striped animals in the cage and believes they are zebras. In fact, it seems that Joe knows that they are zebras. For an animal to be a zebra, it cannot be a cleverly disguised mule. In knowing what a zebra is, Joe knows this fact as well. Does Joe know that the animals before him are not cleverly disguised mules? Dretske says no, and he claims that this is the intuitive answer. In supporting this claim, Dretske states that all of Joe’s evidence in favor of the animals being zebras has been neutralized and does not count in favor of the animals not being cleverly disguised mules, since Joe would have the same perceptual evidence in that case as well. In this case, (a) Joe knows there are zebras before him, and (b) Joe does not know that these are not cleverly disguised mules, even though (c) Joe knows that if an animal is a zebra then it is not a cleverly disguised mule. If (a) through (c) hold, then this is a counterexample to the closure principle. [for the contemporary relevance of this example see this].

However, (a) through (c) do not hold in the above example, this is not a counterexample to closure. Several objections can be brought to Dretske’s account. First, Joe’s evidence that there is a zebra before him is also evidence that there is not a cleverly disguised mule in front of him. Dretske seems to assume that Joe’s evidence for the former claim is solely his perceptual experience, but this is a mistake. Joe will not be justified in believing that there is a zebra before him without further background evidence regarding the nature of zoos, zookeepers, and the difficulty in so disguising a mule (that zoos are not in the buisness of trickery, that zoos often have actual zebras, etc.). Without such background evidence that is assumed in the case, it is false that Joe is justified in believing that there is a zebra in front of him. This can be seen by imaging a case where Joe has the same perceptual experience as of a zebra, but is in a context completely unfamiliar to him. In such a case it is far from clear that Joe knows that there is a zebra before him.

However, when coupled with the relevant background evidence, Joe’s perceptual evidence is also evidence that there is not a cleverly painted mule before him. Considerations concerning the nature of zoos, zookeepers, and the difficulty in disguising a mule, make the belief that there is a zebra before him the belief that Joe epistemically ought to have. What more is needed for knowledge? Joe’s belief is justified, true, and not subject to any Gettier considerations. Therefore, if this evidence is good enough for Joe to know that there is a mule before him, it is also good enough to know that there is not a cleverly disguised mule before him. Thus, closure is preserved.

Dretske might reply that a child lacks this background evidence, yet still knows that the animal before her is a zebra. If one lacked this background evidence, then one would not be justified in believing (and thus would not know) that the animals were not cleverly disguised mules, but in such a case the individual would also not know that there were zebras before her. It is by no means clear that a child in such an epistemic situation actually knows that there is a zebra before her. Even if the child can distinguish between ‘is a zebra’ and ‘merely looks like a zebra’ without the relevant background evidence, she has insufficient reason to affirm that the animal is a zebra. In addition, even if it is granted that the child knows that the animal is a zebra, it is doubtful that she can draw the inference regarding cleverly disguised mules, and if she cannot do so, then the case does not bear on the closure principle.

Second, the intuitions that Dretske appeals to do not support a denial of closure. Though people typically do claim to have seen zebras in the zoo, and though when confronted with the possibility that what they saw was actually cleverly disguised mules these people may deny that they know that what they saw was not cleverly disguised mules; when this denial is made, the former self-attribution of knowledge is often withdrawn, at least temporarily. This response is the response dictated by the closure principle. It seems that at no one time do individuals claim knowledge of a proposition and simultaneously deny knowledge of one of that propositions known entailments. As such, intuitions are on the side of closure, not against it. We have not been presented a counterexample to the closure principle.


Epistemic Contextualism

The skeptic claims that we know much less than we think that we do. Take the following argument for skepticism:

(1) I do not know that I am radically deceived by an evil demon.
(2) If I do not know that I am radically deceived by an evil demon, then I do not know that I have hands.
(3) Therefore, I do not know that I have hands.

This type of argument can be run for nearly every contingent proposition, thus threatening our belief that we know a great deal of propositions. The premises seem true, yet the conclusion seems false – the puzzle of skepticism is the result.

Epistemic contextualism is an attempt to offer a satisfying response to this puzzle in having the result that most of our everyday knowledge ascriptions are correct and by giving an explanation as to why the puzzle arises.

This solution claims that the word ‘know’ is context sensitive – that ‘know’ means different things in different contexts or that it has a different semantic value in different contexts (its contribution to the proposition expressed varies according to context). In this way ‘know’ is like the comparative adjectives ‘tall’, ‘rich’, and ‘flat’. The standard that these terms invoke is dependent upon the context of the attribution – it is the ascriber’s standards that matter. In an everyday context when someone utters the sentence ‘a 6’ person is tall’ he expresses a truth, whereas when someone in a basketball context utters that same sentence he expresses a falsehood. The standard picked out by ‘tall’ differs in these contexts.

The idea is that in cases where S has a true belief that p, the truth-value of the sentence ‘S knows that p’ can vary according to the epistemic standard invoked by the context of attribution. Contexts will differ as to the relation one must be in to a proposition (how justified one must be or how reliable one must be, etc.) in order to count as knowing a proposition. In everyday contexts, the standard for knowledge is relaxed so that most of our knowledge attributions are true. The sentences, “S knows he has hands”, “S knows he ate cereal for breakfast”, etc. all come out true. In epistemological contexts where skeptical worries are present, the standards for knowledge are much more stringent so that those same knowledge attributions are false, those same sentences now express falsehoods.

Contextualists differ as to what features of contexts are responsible for the disparate standards which they invoke. Some appeal to which alternatives are relevant, others to whether the belief in question tracks the truth, but these details need not bother us here.

Applied to our argument above, the contextualist can affirm (1) – (3) in some contexts (high-standards contexts) while denying the conclusion, (3), in ordinary contexts. (1) is true in high-standards contexts, since in such contexts the very possibility of error is enough to destroy knowledge and it is possible that we are radically deceived by an evil demon, but it is false in everyday contexts. Since the mere contemplation or assertion of (1) is often seen as sufficient to raise the standards of the context by contextualists, this explains why (1) seems true to most people – where (1) is asserted the context shifts so as to make it express a truth. (2) is true in all contexts, according to the contextualist, and this also accords with common intuitions. There is a slight nuance, however, since (2) is true relative to a context, since in any context where one does not know the antecedent, one also does not know the consequent. (3) is true in high-standards contexts since the standards of knowledge are high and there is a possibility that I do not have hands. However, in ordinary, low-standards contexts, (3) will be false since one is sufficiently justified in such contexts to know that one has hands. This accords with our intuitions that we typically do know a lot of things. Whereas other responses to skepticism are forced to deny at least one of the propositions (1) – (3), contextualists can affirm (1) – (3), with some context sensitivity additions, while at the same time denying the skeptical conclusion (3), as it concerns our everyday knowledge ascriptions.

Thus, according to contextualism, the skeptic’s denial that we know various things is actually compatible with our ordinary claims to know those very things. The claim is that an appreciation of the context shifts makes the very puzzle disappear – everybody wins. In addition, contextualism credits most of our knowledge attributions with truth. In ordinary contexts, assertions that we know a great variety of things are true, and in skeptical contexts the assertion that we do not know these very things is also true. As such, this answer takes seriously the use in speech and thought of the word ‘know’, and claims its correctness in part on this basis.

While contextualism does have its appeal, it is not a satisfactory response to the skeptical puzzle. First, there are worries concerning the plausibility of this response to our puzzle. Contextualism has the result that many English speakers are not competent in their understanding and use of the word ‘know’. This can be seen in their reactions to sentences they once affirmed yet now deny. If I affirm the sentence, “a 6’ person is tall” in an ordinary context and then deny that sentence in a basketball context, there is no fear that I have contradicted myself and it is clear that I meant different things when I uttered this sentence in different contexts (similarly for other indexicals). Further, it is quite easy to clarify what I meant in uttering the sentence in each context (He is tall for a human, he is not tall for a basketball player). However, if I affirm the sentence, “I know I have hands” in a ordinary context and then deny that sentence in a high-standards context, I do feel as though I have contradicted myself. Further, there is no obvious way of clarifying my earlier claim so as to demonstrate that it was correct and distinguishing it from my later claim. The modifiers, “by low-standards”, “pretty-much knows”, “knows for certain”, all seem inadequate to the task. As such, contextualism seems forced to posit semantic blindness regarding the working of ‘know’ in the English language. This is a significant cost.

Aside from worries about the plausibility of contextualism, there are worries concerning its treatment of skepticism. In the first place, contextualism concedes to the skeptic that in some contexts we don’t know very much at all. This is a significant concession, and it is far from clear that it is one that should be made. Further, contextualism fails to address the heart of skepticism. The interesting skeptical challenge is not that we do not know very much in extremely high-standards contexts, but that we do not know very much according to ordinary standards for knowledge. Here, the contextualist merely asserts that we do typically meet such standards and that our everyday knowledge ascriptions are correct. However, no account is given as to why it is that we meet these standards. A successful response to skepticism would show how it is that we typically meet the ordinary standards for knowledge, yet such an account is not given by contextualists. As such, contextualism fails to address skepticism in an important way.

What is Epistemology?

Epistemology is the sub-field of philosophy which is my primary area of concentration. Epistemology is the study of knowledge and justified belief. It looks into the fundamental questions regarding these issues. In what follows I will be posting weekly summaries on some contemporary issues in epistemology in part to prepare for my comprehensive exam.


I am now on Facebook, and Rate-My-Professor.