Thursday, 17 May 2012

Outline of Chisholm’s “The Problem of Criterion”


Roderick Chisholm argues that we can successful arrive at a general criterion for distinguishing between cases in which we have knowledge and cases in which we do not if we begin with specific instances in which we clearly have knowledge, and then generalise from those cases to arrive at the principles of “good” belief. 


Step 1: Chisholm begins by adumbrating Montaigne’s formulation of the problem of criterion (590*).

Argument: We cannot know that our beliefs are true unless we have a criterion for distinguishing between true or false beliefs.  But we cannot know that a given criterion succeeds at distinguishing between true or false beliefs unless we can check to see if it yields the right results.  Moreover, we can check to see if a criterion yields the right results only if we already know which of our beliefs are true and which are false.

Upshot: When we attempt to find a criterion of knowledge, we find ourselves caught in a vicious circle.

Step 2: Chisholm describes the dogmatist and the sceptic as occupying opposite extremes on the continuum of how much we think we can know (590-591).

Argument: The dogmatist claims that we can know far more than we actually can.  The sceptic claims that we can know far less than we probably can.  Common sense suggests that the truth lies somewhere between these two extremes.

Upshot: We know less than the dogmatist claims, but more than the sceptic claims.

Step 3: Chisholm considers Cardinal Mercier’s tripartite criterion of truth (591-592).

Argument: A criterion of truth should be internal: the mind must be able to find sufficient reason within itself to hold that a proposition is true. A criterion of truth should be objective: it is not the subjective feeling of certainty, but the objective fact that causes the feeling that constitutes the reason for holding something true.  A criterion of truth should be immediate: If we are to find a criterion that avoids an infinite regress, then we must find grounds for assent that presuppose no other.

Upshot: A criterion of truth should be internal, objective and immediate.

Step 4: Chisholm considers Descartes formulation of the problem of criterion (592-593).

Argument: Beliefs are like a pile of good and bad apples in which the bad apples tend to infect the good.  We must therefore get rid of the bad apples; which is to say, the false beliefs. If we succeed, we would be left with a stock of true beliefs that we can rely on completely.  However, we can only tell which beliefs are the good ones if we have a method for doing so.  But given that there are good and bad methods, we must first find a method for deciding between good and bad methods. But in order to do that, we must first have a method for that method, and so on, ad infinitum. 

Upshot: The attempt to find a method to distinguish between good and bad beliefs generates an infinite regress.

Step 5: Chisholm reformulates the problem of criterion in terms of a pair of questions (593-594).

Argument: We may distinguish between the questions: (A) What do we know? What is the extent of our knowledge? and (B) How do we decide whether we know? What are the criteria of knowledge?  If we knew the answer to (A), we would be able to figure out the answer to (B).  Alternatively, if we somehow already knew the answer to (B), we would be able to figure out the answer to (A).

Upshot: There are two possible strategies for successfully answering (A) and (B).

Step 6: Chisholm distinguishes between methodist and particularist strategies for solving the problem of criterion (594-596).

Argument: Methodists believe that we already know the answer to (B), and that we can therefore use (B) to figure out the answer to (A).  Particularists  believe that we already know the answer to (A), and that we can use (A) to figure out the answer to (B). The sceptic claims that we cannot know the answer to (A) unless we first know the answer to (B), and that we cannot know the answer to (B) until we first know the answer to (A).   

Upshot: There are three possible approaches one can take to answering (A) and (B): the methodist, particularist, or sceptical.
Step 7: Chisholm sides with the particularist (596-597).

Argument: We do not need to apply a criterion to determine if we know things like this is a hand. There are certain clear cases of knowledge.  We are free to begin with such cases, and (based on these clear cases) opine what the criterion for knowledge should be.  For example, if we begin with clear cases of perceptual knowledge, we are led to embrace the criterion of accepting the testimony of the senses unless we have some reason to distrust them on a particular occasion.

Upshot: We can solve the problem of criterion by beginning with particular cases of knowledge and generalising to formulate a criterion for true belief.

Step 8: Chisholm defines what he takes to be the key concepts in a theory of evidence (597-598).

Argument: A mental state, A, is intrinsically preferable to another, B, only if anyone who prefers B to A is mistaken in his preference.  A proposition, p, is beyond reasonable doubt only if believing p is epistemically preferable to withholding belief in P.  A proposition, p, is evident for S only if it is beyond reasonable doubt for S. A proposition, p, is acceptable if withholding belief in p is not preferable to believing p, and unacceptable if withholding p is preferable to believing p. A proposition, p, has some presumption in its favour only if believing p is preferable to believing not-p. A proposition, p, is certain if there is no other proposition that is epistemically preferable to believing p.

Upshot:  There are six key concepts that are relevant to the attempt find a satisfactory criterion of knowledge.

Step 10: Chisholm attempts to arrive at a criterion for certainty by considering the “first truths of fact” (598).

Argument: The “first truths of fact” refers to propositions about one’s mental states at a given time, such as one’s thinking certain thoughts, entertaining certain beliefs, or being in certain sensory or emotional states.  Such states are “self-presenting” in that it is necessarily true that if one is in that state at some time, t, then it is evident to one at that one is in that state at t.  Perceiving that p and remembering that p are not self-presenting states, but seeming to see that p and seeming to remember that p are. Moreover, seeming to see that p and seeming to remember that p is indirect evidence for p.

Upshot: If something is self-presenting, then it satisfies the criterion for certainty.

Step 11: Chisholm attempts to arrive at a criterion for certainty by considering the “first truths of reason” (599).

Argument: An axiom is a necessary proposition, such that one cannot understand it without thereby knowing that it is true.  A proposition is a priori for S at time t, if one of the following is true at t: (i) the proposition is an axiom for S at t, or (ii) it is evident to S that the proposition is entailed by a set of propositions that are axioms for S at t.  All a priori propositions are “first truths of reason”.

Upshot: If something is an axiom for one, then it satisfies the criterion for certainty.

Step 12: Chisholm weighs the merits of the claim that sources of indirect evidence should be treated as innocent (i.e., reliable) until there is positive grounds for thinking them guilty (599-600).

Argument: We can begin attempting to articulate principles for a theory of the indirectly evident by considering the following two claims: (M): If it is evident to S that she seems to remember that p, then it is beyond reasonable doubt for S that p, and (P): If it is evident to S that she seems to perceive that p, then it is evident to S that p.  However, (M) and (P) are too permissive, and a complete epistemology is needed in order to ad the appropriate qualifications.  Nevertheless, (M) and (p) are possible starting points.

Upshot:  We can, in principle, arrive at criterion if we paradigm cases of knowledge.

Step 13: Chisholm acknowledges that he has simply presupposed particularism when formulating his criterion of knowledge (600).

Argument: Given that our options are particularism, methodism, and scepticism, there is no non-question-begging way of choosing between the three.  However, this does not mean that the sceptic is right because if we assume particularism, we are able acknowledge an obvious fact that we are unable to acknowledge if we assume scepticism; namely, that we do know many things.  

Upshot:  We are no less justified in assuming particularism than any of the other two approaches to the problem of criterion.

Step 14: Chisholm maintains that his proposed criterion for knowledge satisfies Cardinal Mercier’s three requirements (600).

Argument: Since Chisholm’s criterion, which relies on self-presenting or axiomatic propositions as indirect evidence for claims about the world, does not appeal to any external authority, it satisfies Mercier’s requirement of being internal. Since Chisholm’s approach allows that a subject may be mistaken about whether one mental state is preferable to another, it follows that it is objective.  Since Chisholm’s criterion, if it is adequate, involves principles that are necessarily true, it satisfies Mercier’s requirement of being immediate.     

Upshot:  Chisholm’s criterion of knowledge satisfies Mercier’s requirements of being internal, objective and immediate.


We may successfully arrive at a criterion for knowledge by adopting a particularist approach which assumes that there are clear cases of knowledge, and generalises from such cases to arrive at broad principles.  The particularist approach Chisholm offers begins with propositions about which we are certain, and use such propositions as indirect evidence for propositions, which are thereby placed beyond reasonable doubt.


1. Do you agree with Cardinal Mercier's claim that a criterion of knowledge should be internal, objective, and immediate? Why or why not?

2. Do you think Chisholm's assumption that there are certain clear cases of knowledge is warranted? Why or why not?

3. What, according to Chisholm, is the difference between something being evident and something being beyond reasonable doubt? In your opinion, is being beyond reasonable doubt (as the expression is used by Chisholm) sufficient for a belief to be justified? Is it sufficient for a belief to constitute knowledge?

4. Do you agree with Chisholm's claim that we can be certain about "first truths of fact"? What about "first truths of reason"? Why or why not?

*Page numbers based on the, Epistemology: Contemporary Readings, anthology.


Russell Buehler said...

It seems I've found someone I like even less than Peirce. While the reading for tomorrow only makes it up to question 1 and without a great deal of argument given by Chisholm for his position, I have some initial reactions to the piece thus far. In regards to Mercier's claim that a criterion of truth should be internal, objective, and immediate, I must confess myself highly skeptical that such a criterion could exist, but accepting the supposition that it does, I'm in agreement only on the point of objectivity, and there not at all because of Mercier's short argument. Internal seems dubitable (unless Mercier uses internal in a sense which I'm not familiar with) based on the fact that many philosophers, as Chisholm himself points out, believe the criterion for truth lies in a correct relation with the external world (something external and thus an external criterion). Similarly, immediate seems to presuppose a great deal of theoretical framework, ruling out positions that seem at least prima facie feasible (coherentism).

On a more personal note, the consistent dismissal of skepticism on the part of some philosophers (e.g. "If you have a healthy common sense, you will feel that something is wrong with [skepticism]", pg. 591) has started to annoy me. Is the skeptical position really so abhorrent that it's acceptable to believe that anyone reasonable couldn't possibly accept it? The tacit assumption that there is something wrong with skepticism seems, at least to me, far more dogmatic than the acceptance of it.

Benjamin Altshuler said...

I appreciate Mercier's explicit portrayal of truth and I understand why these three designations are constructive. Objectivity seems logical for criteria, but I am not certain being internal and immediate are sufficient criteria for truth.

There are certain facts of the natural world to which I am not privy; my ignorance does not keep these truths from being so. Though the truth exists external to my experience, it can be conveyed to me through external explanation. No amount of introspection can provide the same explication. Therefore criteria for truth is not always internal.

The first time these facts of the natural world were presented to me, I did not grant that they were true right away. Rather, I weighed the consequences of their truth against my own skepticism and prior knowledge. Eventually, after sufficient testing, I recognized the truth of these propositions. The criteria for truth was not immediately obvious, but the result of reflection.

Certainly if routes to truth are internal and/or immediate, they are legitimate, but I contend other routes are possible. Stay tuned for an objection to objectivity, too.

Taylor ffitch said...

I'm also troubled by the internal aspect of the criterion for belief. It seems like Chisholm is setting "internal" in opposition to coming from an external authority like another person or book. It seems like he's trying to pick out the quality of being fully believed and reasoned by the knower, as opposed to being imposed by an outside force. However, requiring the criterion to be internal seems to me to be mutually exclusive with it being objective. I think his use of immediate is a little misleading, but I do agree with his argument that at least ideally the criterion for truth would not rest on a series of justifications, but be complete in itself.

Na'Quelle D. said...

I agree that a criterion of knowledge should be objective, but I'm hesitant to adopt the other two criteria. I don't fully understand what was meant by "immediate" and I cannot adopt a position on something I don't understand. I also have an issue with the internal portion of the criterion. Is he offering us a criterion of truth as possessed by the individual, or is he offering us a criterion of an external, objective truth that may or may not be possessed/believed by all? If he is looking at this from the individual aspect, then I can see how an element of the internal might come into play. But I still don't see how it would, necessarily, be one of the three tenets.

Caleb R said...

I also looked at Mercier's conditions for a criterion of truth. I agree with his formation of the first two conditions, internal and objective, but found myself not accepting of the third; immediacy. I understand in proposing this third condition that Mercier is trying to avoid the problems of infinite regression, i.e. the circular arguments that Chrisholm puts forth earlier, however I am hesitant to accept that this means we should have an immediate criterion of certitude. On a basic level, if I want to be sure that something is in fact the truth, do I want to base that truth on something immediately decided upon? or would I rather base it on discussion and debate that has some sort of range of regress instead?

Russell Buehler said...

My initial response to the rest of Chisholm's article is to question why he constantly refers to notable philosophers; at points he uses it to move the topic forward, but it's never particularly required, and--since I'm not feeling very charitable--I'm beginning to wonder if the intent--conscious or otherwise--is to invoke a subtle argument from authority or argument from numbers fallacy in favor of his position. He certainly concedes that there seems to be something deeply troubled about his position (I hold that any defense of a position that begs the question--as Chisholm admits he does--is more than a little troubled); moreover, his last few sentences seem to, again, play into the conception that skepticism can't be taken seriously, that it simply isn't a viable option (I certainly don't see how it begs the question in the same manner that his two options do).

In general, I must admit that I'm more than a little disappointed with the proposed analysis; so far as I can see, Chisholm rarely offers any real support for his conclusions beyond 'look this philosopher agrees' and while he makes substantive claims at points, he spends little time filling them out or considering possible objections; instead choosing to simply move on to topics that--in all truthfulness--make no sense until you have a convincing account of the prior principles (which he does not). So, on that note, I'll just point out that Chisholm's assumption that there are certain clear cases of knowledge is warranted only in so far as it refers to unobjectionable cases (e.g. 'I am currently in pain') and that anything beyond these cases is simply assuming what he intends to prove (thus, the question begging). Chisholm's article, then, only seems convincing if you already agree with all of his conclusions.

Benjamin Altshuler said...

There is perhaps a problem with the apple metaphor. Let us adapt: Like people and beliefs, there are a few types of apples: Good apples and bad apples have already been established. However, rotten apples are sometimes indistinguishable from good because rot begins at the core. So let's also consider apples that appear good, but are rotten from the inside. We cannot cut open all apples to check, just as we cannot dissect beliefs for truth.

We must establish a regimen for recognizing and sorting apples before they are shipped to stores in order to minimize the number of bad apples that reach shelves. (This sounds like chicken sexing) Just as you cannot know which apples are good or bad, you cannot know which methods are successful.

Although it might be possible to become adept at this and other unprovable skills, it is never possible for the task to become self evident or for a practitioner to make a designation with confidence beyond reasonable doubt. If such certainty were possible, it would qualify as justification, in my opinion. In approaching the asymptote of certainty (yet never reach it), I feel a methodist technique will produce surer results. This is because individual apples, beliefs, and chicks differ, but knowledge for the process can discriminate on the basis of good/bad rather than arbitrary individual differences.

Taylor ffitch said...

I'm personally stuck on first truths of fact. The explanation of Moore's logic is somewhat helpful, but I'm still not convinced that we can take statements like "this is a hand" as true. Chisholm makes a decent argument for it being more likely than not that certain things are true, but to me that's not enough to base criteria for truth on. We would need something that is absolutely, unquestionably true in order to do this. As much as it pains me to agree with Russ's skepticism, I think we remain stuck on "the wheel" and Chisholm's excuse that there's no way to make an argument without begging the question isn't a solution. It seems like an admission that in this case the skeptics might just be right.

Andrew Cely said...

As much as I want to like Chisholm's argument, I really can't seem to embrace it much as others that we've come across. Going off of Russ's comments, he really does seem to be dismissing scepticism as a totally useless method for finding knowledge, since, as he states, it would continue to keep asking for legitimacy for all methods and the methods of methods and so on for an infinite amount of time. This seems to be clever at first, but ultimately it seems to simply shrug off sceptic arguments with a slightly unconvincing reason so that he doesn't have to deal with objections to his findings. As much as I dislike scepticism at times for similar reasons, I respect its ability to find flaws in other people's arguments, and you cannot claim to disarm an entire set of questions aimed at your paper simply because you do not wish to deal with them.

I'm also in agreement with most people's analysis of the rest of his paper. I too find that objectivity is an important part of what determines knowledge, since different circumstances should not make a difference to the individual pieces of knowledge we each can possess. That being said, the other two aspects, that knowledge is internal and that it is also immediate seem to be somewhat laughable suggestions. In one sense, he is trying to tell us that knowledge is supposed to be objective, invariable, but in the same breath also acceptable to the individual mind. Some people may not find the same things to be believable for cultural reasons, where they have been raised to believe a certain thing even though all evidence points to the opposite. It appears ridiculous to bring the ideas of internalism and objectivity together in the way that Chisholm does. Add to this the fact that it must be immediately believable, where we must be able to grasp it with a specific method quick enough without getting caught in the infinite regression of skepticism. So basically, we're supposed to impulsive and shallow enough in our decisions that we are able to think of these in a way our mind's can grasp, but at the same time we must also be correct on an objective scale. The combination of those seems to bring inherent disaster.

I feel like my agreeability with Chisholm really takes a turn for the worse since he believes in the particularist approach, which I believe to be fundamentally flawed in and of itself. The idea that we can generalize the foundation of knowledge from specific examples and situations that match in terms of solvability seems to me to be a very inefficient and almost ignorant method of attacking the issue of finding a definition for knowledge. Whereas methodists and sceptics focus on all aspects of where knowledge can be defined, particularism just seems to pigeonhole itself based off of the subjective nature of its evidence. Personally, I disagree with this aspect of finding criterion for knowledge, and as such, the rest of his findings seem to hold no weight with me. I suppose after writing all of this, I must incline to retract my first comment - it seemed I had no intention, or ability for that matter, to agree with Chisholm's propositions.

Na'Quelle D. said...

I can agree with Chisholm that there are clear cases of knowledge that either can't be questioned or that would be completely pointless to question, pragmatically. For example, if someone says they are in pain, on what grounds can you question that given that there is no definitive external means to determine whether a person is in pain?

And from what I get from Chisholm's argument, there is no difference between something being evident and something being beyond a reasonable doubt. In fact, he uses one term to define the other. Given his definition, I think that a belief being "beyond a reasonable doubt" is sufficient for justification even though I would never say that it is sufficient for a belief to constitute knowledge. In fact, based on Chisholm's definition, a belief being "beyond a reasonable doubt" doesn't even necessarily mean that the belief is true, just that it is "epistemically preferable".

I have nothing to say about his first truths of fact and first truths of reason argument. I had completely checked out by then.

Caleb R said...

In class I was a little thrown off by the Modus ponens and Modus tollens examples when we were arguing against the stance of the skeptic when we were unsure what to believe. I think I understand the basis for the assumption that If SK, then not -k(p) is more right when looking at the second propositions; SK or k(p), because I know that I have a body, but I do not know that I am a brain in a vat.

This being said, I also brought up in class my "skepticism" about Chrisholm's and Mercier's conditions for knowledge. I just find it hard to find conditions or rules to agree on since in class we seem to always find a counter example that discredits them. This goes well with the "wheel" that is created when looking at the "methodists" and "particularists." I just don't see how you can arrive at one without the other in terms of a basis for that person's beliefs.

Anonymous said...

the cornerstone of chisholm arguement is in step 7... "There are certain clear cases of knowledge." my pet turtle's poo could explode this pathetic maxim.