Starboy

Here is an interesting example of philosophical vagueness. The definition of philosopher list both realist and idealist as synonyms. Realist and idealist are listed as antonyms. Philosophy by definition is screwed up.

Starboy

Longbow

Well, I take it all back. It sounds like you are an epistemicist.

bd-from-kg

Longbow:

Yes, but this has nothing to do with “vagueness”, either of the philosophical or everyday sort. In your example, the meaning of “heap” clearly implies that some propositions of the form “X grains of wheat constitute a heap” are true and others are false (although there are also doubtful or “marginal” cases). Nothing like that is the case for Ockham’s Razor. The problem is not that, while there are some propositions that OR clearly entails are true and others that it clearly entails are false, there are “marginal” propositions whose truth depends on exactly how it’s interpreted. The problem is that there are no propositions whatever that it entails are true, nor any that it entails are false. In other words, it has no propositional content.

The reason for this is not hard to find. The meaning of non-tautological propositions lies in the fact that they assert something about the “nature of things”. Thus if P is a non-tautological proposition, it is true if the “nature of things” is this and false if it is that. Since language is fuzzy, for any statement of a proposition there are typically “states of affairs” such that it is unclear whether the proposition should be said to be true or false if that state of affairs obtains. But if one cannot give any description at all of a state of affairs in which the (alleged) proposition is true and another in which it is false, it doesn’t say anything at all about the “nature of things”, and therefore is not really a proposition at all.

But Ockham’s Razor is a completely general statement about how to decide whether any particular state of affairs obtains. Therefore it cannot entail any assertion to the effect that any particular state of affairs does or does not obtain. That would be like a rule of evidence (in positive law) entailing a statement to the effect that a particular defendant is innocent or guilty. Since the rules of evidence are part of a methodology for deciding just such questions, it would be completely inappropriate for them to entail answers to such questions. The same sort of thing holds for Ockham’s Razor. Since it is a (partial) statement of a methodology for deciding questions about what state of affairs obtains, it cannot entail answers to any questions about what state of affairs obtains. In other words, it is not a proposition, nor does it entail any propositions. It is properly regarded as a principle of action which any rational being follows. Properly stated, it doesn’t even look like a proposition; it says something like “Don’t multiply entities unnecessarily” or in a more modern formulation, “Choose the simplest hypothesis consistent with the known facts.” These are no more propositions than “Close the door” is a proposition.

I don’t understand why you are so resistant to this analysis. Once one accepts that it makes sense to speak of “valid principles of action” (or alternatively, “principles of rational action”) it opens the door to saying that moral principles are similarly not statements of propositions (and thus cannot be said to be true or false) but are valid principles of action. Thus the fact that they cannot be “proved” is natural and expected; after all, how would you go about “proving” Ockham’s Razor? And the fact that there is no such thing as a property of “goodness” or “ought-to-be-doneness” is no problem, since to say that an act is “right” is not to assert that it has some property. (Indeed, it is not to “assert” anything at all in the sense that a proposition does.) More generally, it allows one to say that there are objectively valid moral principles without being forced to claim that there is a mysterious “moral reality” about which they express true propositions.

[ October 24, 2002: Message edited by: bd-from-kg ]</p>

Longbow

Well, of course that's your contention. The question is why are you saying this? The statement of Ockham's Razor is not an imperative statement. It is not "Use the most parimonious theory." Or, if it is there is the statement "Only the most parsimonious theory is correct." It is this latter statement (or a declarative sentence like it) that is generally thought of as Ockham's Razor and such a statement seems to be true or false -- an assertion, not a recommendation.

It seems to me that such a statement as the latter, you have to be arguing, is not an assertion but more of the imperative formulation -- a command or recommendation. (Actually, a recommendation is really an assertion usually as well. But, it is clear that you do not think that the latter declarative statement has propositional content.) It is my contention that the real issue is philosophical vagueness, and we'll just have to see how this bears out.

And here we are! It is precisely the vagueness that comes into play in what you just said. You just said that the fuzziness (combined with another assertion of yours essential about the nonexistence of synthetic a priori truths) is what implies that such an a priori subject that Ockhams Razor broaches cannot have propositional content. Are you not saying here that it is the fact that Ockham's Razor would have to be a priori, but since it is fuzzy, it is not tautological and so cannot have propositional content?

The key to this is how you are handling the fuzziness -- what you thik it means. That is, the key issue is how you deal with teh existence of philosophical vagueness. You seem to have no problem dealing with it when it comes to empircal matters. But, if it comes up a priori, then you are rejecting the possibility of propositional content.

Well that certainly is not true. I don't know how important this is going to be, but it is certainly possible to make a specific (and precise, for that matter) statement about one state of affairs that also implies a range of possibilities but otherwise leaves open and undetermined another state of affairs. You can do this empirically: "Open Box A. If box A says 'B' on the bottom, then open box B. Otherwise open box C." Here is a methodology. Implicit in this methodology is that you are not going to open box A last. It has already answered some questions about the particular state of affairs concerning what box you end up at. And for that matter, it has also constrained that outcome without determining it a priori.

What would be inappropriate is for the idea that we cannot end up at A to be the basis for figuring out a method to find which box to end up at. This would beg the question. Certainly, it can go the other way -- the methodlogy can entail facts about the end result of applying it. And it certainly can do so without being "transparent" -- that is, without it being possible to analytically determine the outcome in advance.

Nonetheless, I don't think this fact will slow you down much...

Well, that is the point isn't it? I am saying that it is a proposition and you are obviosly saying that it isn't. he problem for you is that Ockham's Razor isn't "Choose the simplest model." It is "The simplest model is the correct model." And if it isn't, then it should be. I want to establish here that it is not sufficient to say that this what you mean by Ockham's Razor or that it is a way of interpretting it. You are saying that interpretting the declarative sentence that it is formulated with as an assertion is not possible.

So, you have to show this. It sounds like your argument is that statements about philosophically vague ideas are not "analytic" and so must either be empirical or not have propositional content.

Well it should be clear then, that I am not doing it because I am attached to something like moral objectivity. The truth is, if you would like more insight into the intellectual development of your interlocutor, that I used to basically be a Logical Positivist. Back then, I didn't even know what the term meant. And I was more or less a moral subjectivist. So, I have never really had an agenda when it comes to something like establishing that moral philosophy is objective. In fact, if I could draw a conclusion like that, I could spend more time studying math and physics which is what I "really" like to do anyway. (Actually, I guess I like both.)

The reason I am so resistant to your approach is that I don't think it does make sense to talk about "rational principles of action". But, I do think that morality is just another a priori objective subject. So, the only reason a rational principle of action has any special significance is because it corresponds to some bonafide true proposition. Otherwise, it really is just a personal recommendation. I think you are trying to strike a balance of sorts. You are treating certain things kind of like personal recommendations, but insisting that they are stronger than just "You ought to try the lasagna."

In short, I really think that a sentence like "The simplest model is the correct model," is indeed just what it appears to be -- an assertion. And, by agreeing with such a statement, I am claiming something that has implications. If there were a fact implied by another model that has violated this principle but that was not implied by a more parsimonious model, then I would say that such a "fact" was actually a false proposition. In many regards such a distinction is purely academic -- what does it matter if I say that some such view is false while you say it is not "false" but perhaps something more like dismissable?

But, I do think that it is, in fact, more parsimonious epistemologically speaking (which doesn't matter since you don't think that epistemology has propositional content). And, I think that it does have a tendency to lead to potentially substantively different conclusions. Ultimately, the reason we seem to differ on this seems to be due to how we each react to the issue of philosophical vagueness. You take it to indicate the lack of propositional content in a priori subjects. I, on the other hand, reject the notion that precision is required for propositional content in a priori matters.

bd-from-kg

Longbow:

Because it’s true. Did you think I said it to win a bet?

Well, of course that's your contention. The question is why are you saying this?

Ockham’s Razor doesn’t say this, and if it did it would be transparently false. The most parsimonious theory is often not correct.

You can try all day to come up with a factual proposition entailed by OR, but you won’t succeed. In fact, I can make a stronger statement. For any factual proposition P, OR combined with P does not entail any factual proposition not entailed by P alone.

Not so. Ockham’s Razor is often expressed in the form of an imperative sentence. For example, we find:

Other statements of it are similar. For example:

This is essentially the same, grammatically speaking, as Mom’s “You are not to eat any of the pie until supper”. In other words, it’s an imperative expressed in a slightly different way. It obviously does not express a proposition.

As to whether Ockham’s Razor is best expressed in the form of a declarative statement, that’s a matter of taste. “My wife is the most wonderful woman in the world” is a declarative statement, but it is not ordinarily meant to express a factual claim. The real question is not what grammatical form OR should be couched in, but whether it expresses a proposition.

A factual assertion?! This one you’ll have to explain. Anyway, I have never said that OR is merely a recommendation. I once said that it’s more like a recommendation or injunction (two very different things) than a factual statement. But it isn’t quite like any of these things. Actually I don’t think principles of rational action are properly analyzed as special cases of something else. They can be compared to other kinds of statements to bring out how they are like and unlike them, but that’s all.

I really have no idea what you’re trying to say here. The statement “It’s dangerous to drive fast on an icy road” is fuzzy, but I would certainly not claim that it does not express a proposition, because it is easy to come up with clearly factual statements that capture at least part of its intended meaning. In other words, this statement can be reasonably interpreted as expressing factual claims, even though there are “marginal” cases of factual claims that it can reasonably be interpreted either as making or not making. That’s what is meant by a statement being “fuzzy”. On the other hand “The Absolute is green” has no factual content, because there is no factual claim that can reasonably be interpreted as being even part of its meaning. Thus “The Absolute is green” is not fuzzy, but meaningless. The same thing is true of statements such as “This world is but an illusion” or “Reality is one”. (Note: The fact that these statements are meaningless is entailed by the facts that (1) they are clearly intended to express propositions, and (2) they do not express propositions. I do not by any means take the position that all statements that do not express propositions are meaningless.)

Again, I would allow that a statement S has factual content if there is a factual proposition P which, combined with S, entails some factual proposition which is not entailed by P alone. The statements “The Absolute is green”, “This world is but an illusion”, and “Reality is one” fail to satisfy even this criterion. I don’t see how you can get more liberal than that without rendering the term “factual content” completely meaningless.

You have a weird notion of what “entailment” means. You might just as well say that “jump down, turn around, pick a bale of cotton” entails that a bale of cotton will be picked.

“Close the door” does not entail that the door will be closed. The most that can be said is that if the injunction “close the door” is obeyed, the door will be closed. Similarly, “Open box A first” does not entail that box A will be opened first. The correct statement is that if the methodology you describe is followed, box A will be opened first. But this is not a factual statement; it is an analytic statement – i.e., a tautology. It is not entailed by the methodology; it is simply true by virtue of the meanings of the terms involved.

In the case of Ockham’s Razor the corresponding claim would seem to be that, since it says to choose the simplest hypothesis consistent with the known facts, it entails that I will choose the simplest hypothesis consistent with the known facts. But this is nonsense; all that can be said is that if I choose the simplest hypothesis consistent with the known facts, I will have chosen the simplest hypothesis consistent with the known facts This is undoubtedly true, but it is not what Ockham’s Razor says; it is merely a trivial (and tautological) statement about the results of applying Ockham’s Razor.

Nope. Ockham’s Razor is not “The simplest model is the correct model”, it’s “Choose the simplest model.” If it were the former it would be trivially false.

Nope, it shouldn’t be. If it entailed any factual statement, there would be possible worlds in which it is false. If it were false in some possible worlds, we would have to know that this is not one of those worlds before we would be justified in applying it. But we cannot know that this is not one of those worlds before applying it, because Ockham’s Razor is an integral part of any rational methodology for determining the nature of this world.

I don’t think it is even possible to formulate it as a declarative statement except by including an evaluative term like “should” or “best” or “prefer”. And I don’t say that it can’t be interpreted as an assertion; of course it can. I’m saying that it can’t be interpreted as a factual assertion – i.e., an assertion about the “nature of things”. “Ice cream is better than spinach” is not an assertion about ice cream and spinach; it is an assertion that I prefer ice cream to spinach. Similarly, “Knowledge is preferable to ignorance” is not an assertion about knowledge or ignorance. However, unlike the statement about ice cream, it is not merely an assertion about my preference for knowledge over ignorance; it’s really asserting that any rational being will prefer knowledge to ignorance. But this is not a factual statement. If we find that someone prefers ignorance to knowledge we will not regard this as falsifying the claim that knowledge is preferable, but as demonstrating that the person in question is not rational.

Well, so far as I’m concerned I haven’t made any claims about “philosophically vague” ideas, but it’s true that in general I think that philosophical ideas have no factual content. Nor should they. They are “about” providing a consistent conceptual framework for understanding and interpreting the world, not about the world itself.

Well, it doesn’t make sense to me to say that it doesn’t make sense to talk about “principles of rational action”. It seems self-evident to me that it there are principles, such as “choose the simplest model”, which rational beings will follow. How can this not make sense?

OK, here’s yet another principle of rational action: “Do not try to accomplish an objective by adopting a certain course of action if you know that it will not accomplish that objective”. What bona fide true proposition does this correspond to? Or how about this one: “Do not believe a factual proposition for which you have no evidence”. What bona fide true proposition does that correspond to?

Quite right. For example, I say that you should accept the conclusion of a valid argument whose premises you accept (or more precisely, you should be at least as confident that the conclusion is true as you are that all of the premises are true). The “should” here is much stronger than the “should” in “You should try the lasagna” and stronger even than the “should” in “You should not kill someone to get the money in his wallet”. In other words, it is not merely a recommendation, nor is it a moral “should”. It’s stronger than that. You might call it a “rational ‘should’ ”. I.e., it means “to do otherwise would be irrational.”

Sure. A false assertion.

True. But agreeing with it is completely irrational, since it is obviously false.

Then in a great many cases you’d be wrong.

You seem to be saying that the hypothesis that Ockham’s Razor expresses a proposition is the most parsimonious interpretation of it. But in the first place the principle of parsimony is out of place here. The best interpretation of a statement is the one that corresponds most closely to what the person who uttered it meant. It doesn’t matter whether this is simple or complex. In the second place, either Ockham’s Razor expresses a proposition or it doesn’t. It doesn’t seem to me that this is a matter of interpretation. But most importantly, the claim that OR expresses a proposition is absolutely fatal to it. If it expresses a proposition there are possible worlds in which it is false, and we have no way of knowing whether this is one of those possible worlds.

Longbow

It is not a methodological assumption. It is an epistemological contention. One that most philosophers agree with. If you are seriously trying to claim that philosophers do not contend this but rather just adopt it as a method, then you are just mistaken. You might believe that they are mastakenly asserting it, but it is clearly a contention not a command.

Right, I acknowledged that you said that there could be fuzzy empirical assertions. It is both the a prior nature of the statement and the fuzziness that seems to be causing you to reject the possibility of propositional content. In that case, the issue is still your reaction to the fuzziness not the a priori nature of it.

No, a statement being fuzzy just means that there is a Sorites Paradox in it (figuratively speaking). Your basic position seems to be that no a priori proposition can be this way. So, everything you are saying about that a "fuzzy" statement means that it has certain properties including that it states a factual claim (i.e. that it is empirical), is the point of contention. You are just restating your position. Basically, you are saying that these a priori statements cannot be propositions because they are fuzzy and fuzzy a priori propositions don't exist. I am saying that the fuzziness of a proposition doesn't require that proposition to be empirical, so I say these fuzzy statements are a priori propositions.

Okay -- I'll agree with that...so?

You have no reason to say this. "The Absolute is green," is meaningless because the concepts of absolute and green aren't comparable. But, the other statements, although you might think it turns out that they lack meaning, they certainly are intelligible in a way that "The absolute is green" is not.

So you are going to have to explain this one a lot better than this. Saying that they are meaningless because they aren't empirical would be begging the question.

That is what it means when you say that a method entails something. I don't think that it is hard to glean that from the discussion.

Nope. Epistemology is a subject in which people make contentions not issue orders. And, all of such thngs as Ockham's razor are not suggestions, commands, recommendations or anything like that. The philosophers that come up with them are arguing in defense of a position. It is absurd to think otherwise.

The best I can make of your defense would be the outrageous assertion that philosophers make statements about epistemology (statements like Ockham's Razor) as intentionally not assertions but rather as something else similar to recommendations.

By "factual", you mean empirical. I am saying that it is an a priori assertion. So, I do not interpret it as being empirical. On the other hand, when you say that a factual assertion is just a statement about the "nature of things", then any assertion is a factual assertion -- even a priori ones. Or, if you want to interpret "things" to mean physical objects, then we are back to your saying it isn't empirical which I agree that it isn't.

Well, this is pretty nearly a tautology. Rational beings almost by defintion persue knowledge. But that is beside the point, what is most interesting about such a statement is that it seems to fit into your "objectively rational principle" notion, but I think it is subjective. In other words, I don't think that "Knowledge is preferable to ignorance" has any more propsoitional content than that it states your preferences and that beyond that there is no other objective content.

Probably, the simplest interpretation of it would be that a belief in an empirical proposition without evidence is not justified. This is an epistemological assertion concerning whether or not someone has one of the pieces of knowledge, namely justification.

Okay -- a false assertion, then. It is nonetheless clearly a priori, and you have just agreed to it being an assertion.

This is what is so ridiculous about your interpretation of epistemology as not being a bunch of propositions that philosophers argue over but rather something else. The philosophers all think they are propositions, so you aren't adhering at all to how these statements are intended. Do you really think that the philosophers that state things like Ockham's Razor do not intend them as assertions?

Seriously?

In any case, the issue of parsimony is less irrelevant than you think. As a model of the phenomenon of philosophical inquiry, my views are more parsimonious than yours. I have this one epistemological object -- the proposition. You have two distinct objects handled in two very different ways. One is the proposition and the other is the rationally compelling principle. This is another reason you have the burden of proving that we need this distinction to properly account for everything. I think it can all be done with propositions. Furthermore, if you reject epistemology lock, stock and barrel, then that is one thing. You can just say that it is all "in the way of a recommendation". As soon as you admit that any of it isn't, then it is irrational (whether that is a proposition or "in the way of a recommendation") to have it divided needlessly into two parts -- that which is propositional and that which is "in the way of a recommendation".

What? If it expresses a proposition, then such a proposition would be a priori under my contention. You should know by now that I do not agree with you that a proposition must express something about the empirical world or be vacuously implied by the definitions of the terms that it is composed of to be a proposition. I can't quite recall off hand, was it your contention that all a priori propositions were contained in math or formal logic. If that is your position, then you know that I don't agree with you. I am claiming that philosophy -- things like Ockham's Razor, morality, etc -- is all a priori and all propositional.

bd-from-kg

Longbow:

At this point we seem to be talking past one another to such an extent that continuing to reply directly to your comments would be a complete waste of time. You completely misinterpret almost everything I say to such a degree that trying to clear up your misconceptions would take up far more time than I have available. So let me just try to get to the heart of the matter.

First off, regardless of your obsessive belief to the contrary, our disagreement has nothing to do with “vagueness” or “fuzziness”. The problem is not that statements like “This world is but an illusion” or “Reality is one” are “fuzzy” or “vague”, but that they are completely meaningless. You seem to think that I say this because neither you nor anyone else has been able to explain exactly what they mean, but in fact I say it because no one has ever been able to give the slightest indication of even some part of what they mean. I have no idea how to give them even a partially intelligible interpretation – at least, if I’m restricted to interpreting them as expressing propositions.

Instead of wasting your time trying to “explain” why I say this or challenging me to “prove” that these statements are meaningless, it would be more productive if you were to explain what propositions they express. I’m very flexible here. Rather than trying to explain their meaning in full, you can just specify some propositions that they entail, or even more liberally, some propositions that are entailed by them together with other propositions which are not entailed by the other propositions alone. Or, if you say that they don’t express propositions in any sense, then what is their function and purpose; in what context would the statement “reality is one”, for example, be relevant or appropriate, and what would be the function or purpose of making it? You say that these statements are “intelligible”, but I don’t find them so. Enlighten me.

Second, I find your concept of “synthetic a priori” propositions incomprehensible. As I pointed out to you before, Kant’s use of this term is understandable, although I don’t really agree with his calling the things in question “propositions”. But I understand why he does. The important thing is that he understood his “synthetic a priori propositions” as statements about limitations on human understanding, or about how (given the way our minds work) we must necessarily perceive things. This is quite a different thing from claiming that there are “a priori” truths about the world which are true independently of who is perceiving it or the perceiver’s built-in conceptual framework.

The basic problem is this. In order for a statement to be said to express a proposition, it must be meaningful to say that it is true or false. Now I understand what it means to say that an analytic statement is true: it means that it follows directly from the meaning of the terms involved. And I understand what it means to say that an empirical statement is true: it means that it divides the class of logically possible worlds into two subclasses, WT and WF, and asserts that this world is in WT. But you say that there is a third category of statements that express propositions: ones that are not true (or false) by virtue of the meaning of the terms involved, but are nevertheless true in all logically possible worlds. Thus a statement in this category does not divide the class of logically possible worlds into two subclasses, and saying that it is true cannot be interpreted as meaning that this world is in one of these subclasses rather than the other. But in that case I’m at a loss as to what it could possibly mean to say that such a statement is true. And until you can clarify what it means to say that such statements are true, I must continue to regard them as not expressing propositions – i.e., as statements that it is not meaningful to call “true” or “false”, although they might still be meaningful in some other way.

Note that this objection to your notion of “synthetic a priori propositions” does not rest on logical positivism. I place no conditions on what sort of criteria may be used to define WT and WF, or on what sort of things might properly be said to “exist”. I simply insist that in order to be said to express a proposition, a statement must divide the class of logically possible worlds into two classes, WT and WF, such that it is true iff this world is in WT. But if you can give some other plausible account of what it means to say that a statement is true or false, I’d be happy to consider it.

Again, the problem is not that I find such statements vague, but that I have no idea how they mean. That is, in what sense can they mean anything at all? Or equivalently, what does it mean to say that they are true or false?

Third, you seem to think that my notion that Ockham’s Razor does not express a proposition is “ridiculous” – that this is a complete misreading of how philosophers in general understand it. I disagree. I think that this is pretty much how most philosophers understand it. They may not agree with my interpretation, but they do not regard it as expressing a proposition. At any rate, this interpretation is certainly not ridiculous, nor is it completely out of the philosophical mainstream. I already cited two reputable sources that define Ockham’s Razor in terms of imperative statements. The only other way that it is commonly phrased is that one “should” choose the simplest hypothesis. But saying that one “should” choose the simplest hypothesis is advocating or recommending or prescribing or enjoining doing so. (Note that the sense of “should” here is clearly not moral.) In other words, Ockham’s Razor is commonly understood as prescriptive.

In fact, although Ockham’s Razor is a particularly clear example, I think that the great majority, if not all, metaphysical claims are properly understood as prescriptive. Here is an excerpt from the Encyclopedia Britannica article on metaphysics that explains this idea much better than I could. I quote it here both because it expresses my own view almost exactly and because it shows that such a view is (to say the least) philosophically respectable.

In fact, I would go further than saying merely that most philosophers do not interpret OR as expressing a proposition; I would say that it cannot be reasonably interpreted as doing so. In support of this claim I note that you (like everyone else who has tried) have failed to give a remotely reasonable account of what sort of proposition it expresses. It certainly does not mean that the simplest hypothesis consistent with the facts is always the correct one, or that the simplest hypothesis will in fact be selected or preferred. So even if it were true that most philosophers interpret Ockham’s Razor as expressing a proposition rather than being prescriptive, it wouldn’t follow that we must take them at their word and interpret it accordingly. I agree that this would be what we ought to do if it could be done in a reasonable way. But unfortunately interpreting OR as a factual assertion is logically incoherent given its purpose and function. So we have no alternative but to try to understand what the philosophers (if any) who understand OR as expressing a proposition are really getting at, and find a meaningful interpretation of it that captures their intent as well as possible.

Now for comments on a couple of your replies.

That’s really an interesting answer, because it’s not far from the way I would analyze this statement. It does not express a proposition, yet it expresses much more than a personal preference. It expresses a preference that any rational being will necessarily have. Or to put it another way, it expresses a principle of rational action. That is, any rational being will regard the choice that he would make if he had enough knowledge and understanding as more “valid” than one he would make with less knowledge and understanding. But it is not a tautology that the former choice is more “valid” than the latter. And while it may be a tautology that any rational person will prefer the former choice because this is part of what it means to be “rational”, the definition of “rational” is not arbitrary. Being “rational” actually means something, although it is impossible really to define what it means. A definition that includes things like preferring choices based on knowledge and understanding to those based on relative ignorance is trying to capture some of what “rational” means; it is not merely arbitrarily lumping together a bunch of disparate qualities and assigning them the label “rational”. In other words, a definition of “rational” that includes things like this is descriptive rather than stipulative.

Finally:

But this “bifurcation” is completely natural, and we are all perfectly familiar with it. A rational agent has to decide what to do; “principles of rational action” are guidelines for such decisions. One of the most important (though secondary) kinds of “what to do” questions are “what to believe” questions; principles like Ockham’s Razor apply to this kind of decision. And of course the subjects of “what to believe” questions are propositions. Ockham’s Razor is not the subject of a “what to believe” question, but is rather a guideline for answering “what to believe” questions.

In any case, as I commented earlier, applying the “principle of parsimony” is out of place in this kind of context. One doesn’t decide whether there are “principles of action” on the basis of the principle of parsimony any more than one decides whether there are two different kinds of proposition (analytic and synthetic) on the basis of this principle. This is what empiricists do, but it is a mistake. There is obviously a fundamental difference between analytic and synthetic propositions, and it is absurd to say that a theory that denies this distinction is superior on the grounds that it is more parsimonious. Similarly, if it can be seen on reflection that there are such things as “principles of action” which cannot be reduced to propositions, a theory that denies that they exist is not superior because it is more “parsimonious”.

And in fact principles of action cannot be reduced to propositions. Propositions simply do not go to “what to do” questions. Thus the propositions “X will make you happy” and “Y will make you miserable” have nothing to say about whether to do X or Y. For that you need a principle of action, such as “Other things being equal, do what makes you happy rather than what makes you miserable”. This particular principle can perhaps not be called a “principle of rational action”, because it is perhaps conceivable that some rational being might prefer being miserable to being happy, but it is certainly a principle of action that is universally valid for humans. Again, suppose that B is a factual proposition for which you have no evidence. You now are faced with the decision whether to believe B. Now you can say “believing in something without evidence is not justified”, but this merely leads to the conclusion that believing B is not justified; it still doesn’t answer the question of whether to believe it. For that we need a principle of action such as “Do not believe any factual proposition without evidence”. This tells us what to do. Thus principles of action have a function that simply cannot be fulfilled by propositions.

Thus there clearly are such things as principles of action, and they cannot be reduced to propositions. So the “principle of parsimony” does not apply. A theory that achieves simplicity by ignoring important logical distinctions is not preferable on that account.