Wednesday, August 25, 2010

A Propositional Attitude

Propositional attitudes are states of mind like beliefs, hopes, and expectations that have "propositional content". That is, they are attitudes directed at something that can be true or false (propositions are things that can be true or false). I believe that the sun will come up tomorrow; I hope that it won't rain; and I expect that Susan will bring the potato salad. Each of these mental states have a proposition as its object, and that proposition may turn out to be false. But my attitude is directed at their truth. For example, I might be worried that it will rain; my worry is directed at the proposition that it will rain just as my hope was directed at the proposition that it wouldn't. Either way, I have an attitude about the truth of a proposition.

In this post I want to appropriate this term of art from analytical philosophy and redeploy it in the context of writing instruction. I want to talk about what it means to have a "propositional attitude" in the sense in which we might talk about having a positive or negative attitude, i.e., as a disposition. I want to talk about it as a sort of intellectual posture. Academic writing, I want to say, revolves around propositions, even if it does not always propose them, and sometimes altogether eschews proposing them. An academic piece of writing is, to a great extent, an arrangement of propositions.

How do you cultivate a suitably propositional attitude in your writing? Well, the first thing you can do is identify the propositions that are relevant in your research. For each writing project, try to write 20 or 30 sharp, declarative sentences whose truth-value matters in the context of that project. These sentences may express commonly held views that you want to call into question. Or they may more straightforwardly express your views, i.e., things you believe. They will normally include statements of theory, statements of method, analytic conclusions, and even common knowledge. The essential thing is that whether or not they are true is important to you, as the writer, and therefore, implicitly, to your readers. You expect your readers to care about the truth of these sentences.

You don't always expect them to agree with you. In fact, you are often well aware of specific readers who will disagree with you because you know they think differently about the propositions in question. What I'm calling a propositional attitude is simply a good-natured willingness to discuss the truth of these things or a steadfast refusal to do so. The first set of propositions define the subject matter of your field, the second defines your foundations. Some truths are not going to be called into question in the paper you are working on and your attitude is that anyone who wants to do so should go elsewhere.

Other truths are simply not going to be involved at all. This group of propositions (true sentences that are not relevant to your topic) is of course infinitely large. It can help, however, to identify a few of these just to mark out territories that you will not enter.

A final word to deconstructionists: even if your paper is engaged primarily in "questioning", which is a perfectly respectable activity, then it is nonetheless "about" the propositions you are questioning. Moreover, the tradition of deconstruction is itself built around a set of propositions about the nature of language and writing, and these propositions, however ironic you may be about them, are also capable of being written down. You do well to articulate them at least for yourself at the outset of your writing project. Irony, too, is a propositional attitude.


Presskorn said...

I like the analogy between propositional attitudes and propositions that sum up a research project. Perhaps the analogy is already strained, but I think it could strained further :-)...

It is somewhat incomplete to say that a propositional attitude (in the analytic sense) is an ”attitude about the truth a proposition”. Rather it is having an attitude about the truth of a proposition UNDER A CERTAIN DESCRIPTION. This is what accounts for the referential opacity of sentences with propositional attitudes. If it was just an attitude about the TRUTH of a proposition, the principle of substitution would apply the phrasing of the proposition – which is not the case.

But this fits the analogy nicely: When writing down the 20-30 sentences that sum up a research project, one should pay attention to the phrasing of the sentences. Ideally, the 20-30 sentences are not just suppose carry the tagline “I believe in something roughly like this.”, but rather “I believe in something exactly like this.”

On a slightly different score, I could also phrase a question to you in terms of propositional attitudes: In speech act theory, we say that propositional attitudes have “directions of fit”. Either they have a ‘word-to-world’-fit, i.e. the proposition fits a fact, or it has a ‘world-to-word’-fit, i.e. the proposition wishes or command a fact to obtain.
So, should the propositions that you advice us to draw up for our research project have the status of ‘word-to-world’, i.e. “These proposition accurately sum up my present research project” OR rather the status of ‘world-to-word’, i.e. “I command myself to do a research project that can be described by these propositions.”?

Both, I guess. But thank you for giving me an analogy to toy around with.

Thomas said...

I've always taken a proposition to be more abstract than that. In the terms you suggest, a proposition is an ideal entity that corresponds exactly with a real (i.e., far from ideal) sentence in a natural language under a description. But you're right that it was imprecise of me to talk about an attitude "about the truth" of a proposition. I meant an attitude about the content of the proposition, i.e., about the fact that is the case if the proposition is true. Also, I've always thought that the sentence (in a natural language) and the fact (as, say, given in experience) have the proposition (in the mind of the knower) in common. I'm not sure if it's a good reading of the Tractatus, but I usually credit Wittgenstein with making me see it.

Presskorn said...

Somewhat irrelevant note:

I employ the exactly the same use (i.e. meaning :-) ) of the word ”proposition” as you do: it is ideal entity corresponding to a fact, which can be realized/denoted in several different sentences in natural language(s).

However, the main reason that propositional attitudes have bothered logicians so much is that propositional attitudes cannot take such ideal entities as their object, since what comes after the that-clause following a propositional attitude cannot be rephrased by a different sentence in natural language. (What follows the that-clause is, as logicians say following Quine, ‘referentially opaque’.)

Take the example of two extensional sentences containing no propositonal attitudes:

(1) Oedipus killed Laios
(2) Oedipus killed his father.

These two sentences are equivalent, they denote the same proposition, the same ideal entity corresponding to a fact… no problem.

But this, of course, is not the case for sentences containing propositional attitudes:

(1) Oedipus believed he married Iokaste
(2) Oedipus believed he married his mother.

Here we get a change in truth-value from applying the principle of substitution, even though ‘he married Iokaste’ and ‘he married his mother’ are extensionally equivalent and presumably denotes the same ideal proposition…. Result: logical headache and a problem for the whole theory of ideal propositions.

In any case, what follows the that-clause in a sentence containing a propositional attitude is highly sensitive to the particular description employed. And that is what I meant by saying that a propositional attitude is an attitude about the truth of a proposition under a certain description.

BTW, Wittgenstein deals with this problem in Tractatus §5.541ff, but I never quite understood how what he says is suppose to solve anything…

Presskorn said...

I should have written "Oedipus believed THAT..." etc. in order to make the talk of that-clauses sensible...

Thomas said...

This will probably get us out of my depth philosophically (and WAY off the topic of the post), but here goes...

Though I understand the argument you are making, I have never believed that (1) and (2) state the same proposition. To believe you have married Sally and to believe you have married your mother are two entirely different things. Even if, I should emphasize, you know that Sally is your mother. This idea of "extensional equivalence" is complete hokum.

"Venus" is simply not extensionally equivalent to "the Evening Star". The first refers to a planet, the second to a pin-prick of light in the sky at evening.

Intension covaries with extension and vice versa. There is no such thing as equivalence of reference with inequivalence of sense. Frege was wrong.

I do realize that I'm just making assertions there, and I'll think a bit more about why I think they are true. It's been awhile since I've had to make this argument.

Presskorn said...

In my opinion, the problem, in general, is more with the notion of a ’proposition’ (let’s purge these damn ideal things from our ontology!) than with the distinction between extension and intension....

One problem with giving up the notion of 'extensional equivalence' is that it becomes hard to explain what a 'fact' is - if it's not the the sort of thing that different sentences can share an extensional reference to...

But I’d like to hear your argument, when you’ve brushed off the dust from it...

Thomas said...

Oh, no! I think we really need those propositions (though we can argue about whether they belong properly in our ontology; in a sense, I believe they belong there too.)

On the point at issue: We will agree that "Oedipus married Iokaste" is as true as "Oedipus married his mother" because Iokaste is Oedipus's mother. Nonetheless, we will also agree that the two sentences differ in sense. They don't mean (in one sense) the same thing.

Now, you want to say they refer to the same fact. There is a good reason for thinking so. After all, we are also going to agree that facts are "truth makers"; propositions are true or false in so far as they refer to facts that do or do not obtain. And (here's the would-be clincher) the "Oedipus married Iokaste" is (as we said above) as true as "Oedipus married his mother".

I want say that the the "as true as" needs to be qualified also with the following warrant (in Toulmin's sense): "Iokaste is Oedipus's mother." This is a statement of fact (not a logical truth).

That means there is a third fact, namely, the fact about Oedipus's birth to Iokaste. I want to say there is a comlex relationship between these three facts and therefore no referential equivalence between the terms "Oedipus's mother" and "Iokaste".

That's what I've got for now.

Presskorn said...

Well, that is indeed a bit of a clincher. Off the top my head, talk of "as true as" is a grammatical mistake. I.e.

"True" has no defined use (has no sense) as a comparative adjective in the language game of speaking of the truth of propositions.

Which however is not to say that I can't see the drift of what you're saying. More later perhaps.

((I also wonder if you wouldn't generally discourage students or researchers from using "truer" or "as true as", if they came to see you seeking your help with their prose? And your reason would probably be that is the comparative use is vague.))