A fascinating subject, James - why don't you send a group message, as it deserves as wide a distribution as possible!!! I wish I could go - my answer is, of course, that it REQUIRES metaphysics to avoid degenerating into naïve scientific realism... if it interests you personally, we could launch a thread here.

***

'Particular' sentence, eh? So, the proposition concerns a 'possible' sentence or a sentence 'type' - that would imply either that there are as many propositions as there are possible sentences or that there as as many propositions as there are sentence-types. I think there might be an answer lurking in a correct formulation of this intuition

On the conf - pity they haven't invited Craig Callender or Katherine Hawley. Dennett is a has-been, but - if I remember correctly - Friedman is pretty good. I see they've got Haris' friend, the good Doctor Ladyman... he's got some interesting ideas.

One might believe that there would be possible statements of a particular truth even in the absence of any actual language. Why would one believe that? I suppose faith in the power of reason might lead one to such a conclusion.

Personally I believe that although reason is powerful, it is not all-powerful... there may come a point at which words simply fail.

Before turning to James' remarks, a comment from our friend Tristan Vanrullen sporked from Facebook (come on, Tris, where are you when philosophy needs you?)

"a proposition is an expression having a "truth value".

"It's raining" is a proposition. But "give me the sugar please" is not.

About the truth value, it depends on the context expressed by the sentence, and also on underlying conceptual components, ontologies, subjective interpretations.

There are sentences looking like propositions, but told as 'assertions'. Assertions are propositions receiving a truth value affectation from the locutor. As so, they are told AS IF THEY WERE tautological propositions . "It is raining today" is naturally told to assert the fact that it is actually raining. A proposition allows people to discuss about it. "there is no limit to the universe", considered as a proposition, can be discussed and argumented. But told as an assertion, it would immediately need to first debate about the belief in such a truth value affectation."

So it might seem that saying "it is raining" is a proposition, but saying "it's raining... you &^%$ idiot!!" would be an assertion?

If we take the idea that the proposition expresses a real or possible state of affairs, then "it is raining" would correspond to a state of affairs - the assertion "it's raining... you &^%$ idiot!!" would rather seem to concern the evidence for the state of affairs being actual ;-)

Much like 'meanings' - or, for that matter, 'types'. I was re-reading Quine's notes on nominalism this morning (Oxford Studies in Metaphysics IV) and I was struck by the thought that we can't just *do away with* troublesome 'entities' like universals, sets, or propositions - the best we can hope is to reduce them to notions which are more susceptible to a nominalist reading (in the case of 'propositions', effecting a reduction to 'sentences', for example). I'll post a preliminary exercise later on today to show the kind of thing I mean.

As far as the "worrying characteristics" of prepositions are concerned:

I'll requote Zimmerman 2005: "at least in the case of a proposition that is not about a particular sentence or thought, it would have existed and been either true or false even in the absence of all sentences or thoughts"

As James remarked, the term 'particular' is important. A 'particular' sentence or thought in this case is surely one which has an instance – such "it's not raining". Some person utters or thinks the sentence "it's not raining" at some time, and there is some proposition which is somehow "about" the sentence which that person utters or thinks at that time. I have no idea what this "about" relation might be, unless it's somehow determined by the 'meaning' of the sentence.

Are propositions no more than "reified meanings"? I'm not at all sure that this is the case: propositions are held to be 'true' or 'false'; while 'meanings' might play a role in *determining* whether some proposition is true or false, when we talk about a 'true meaning' we're usually talking about the commonly-held sense of a term, not its truth conditions. Truth would seem to bear some relation to 'the world', even if that 'world' is *no more than* a structured collection of terms and beliefs about those terms. We can, for the sake of brevity, talk about "states of affairs" here, but with the provisos that David L evokes:

STATES OF AFFAIRS

Let us, for the sake of argument, accept the common-enough view that, if a proposition is true, then there is some 'state of affairs' to which that proposition corresponds. Now, given roughly, this asserts

i. (∀p) p is true → (∃f) f is a state of affairs

Now, (i) quantifies existentially over states of affairs: does this thereby imply that we're asserting the 'existence' of states of affairs? Logically, yes; but only within the 'theory' under examination. If the theory of propositions outlined by Zimmerman calls for 'states of affairs' or something similar, then states of affairs are posits of that theory. However, we might hold either that some other theory better explains whatever 'evidence' there is, or that 'states of affairs' can be reduced to or explained by some other, more fundamental, kind of entity: this is a matter of the indeterminacy of evidence to theory, and of our justifications for preferring one theory to another. (i) is a metaphysical axiom of a certain theory of propositions (there are others, but I'm not writing a thesis here);

Let us allow that (i) is an, if not the only, answer to the question of the relation between 'propositions' and 'the world'. Let us also allow that the 'meaning' of a sentence somehow describes a certain state of affairs, whether or not that state of affairs "actually obtains". If the state of affairs effectively obtains, then the sentence is 'true' – the 'meaning' picks out or indicates an actual state of affairs; if it does not, then the sentence is 'false'. A "state of affairs" would seem here to be a way the world COULD be, and as such is distinct from the way the world IS (which would be a proper part of "the way the world could be"). An actual state of affairs is a proper part of "the way the world is" – that is, it is compatible with and in some way implied by our overall beliefs about the world. We could say that an actual state of affairs is no more than the way we believe a certain portion of 'the world' to be, and as such, it seems rather inoffensive: we can, for example, accept that when Susan believes that it's raining she believes that a certain portion of the world is a certain way without accepting that there is some mind-independent 'world' that is actually and mind-independently 'a certain way' and that it is this 'mind-independent way the world actually is' that determines her belief. A commitment to "beliefs about things in the world" is not the same thing as a commitment to "things in the world" (this, I believe, is part of the sense of Quine's remarks on 'the myth of physical objects' in "Two Dogmas").

We can, therefore, accept that a certain theory of propositions requires "states of affairs" – but the ontology is relative to the theory. Our task is to suggest a rival theory, or rival understandings of the terms used in the theory, and to give sufficient justification for holding that our rival theory or understanding is preferable by applying Occam's razor.

***

I'l try and sketch out some ideas I've had for reducing the problem of propositions to more manageable proportions (nothing very elaborate, it's mere speculation without any background reading – but it might serve as an example)

David: agreed - what does it mean when we say that Galileo said that the Earth moves?

Is "that the Earth moves" a proposition? (not here it ain't - it's a dependent clause). OK, does "that the Earth moves" *express* a proposition? Can we say "Galileo made some utterance which has approximately the same meaning as the English utterance "the Earth moves"? Can we understand the 'that' as a kind of deictic that indicates the 'meaning' rather than the graphemic form? If we do, are we just shifting the problem from 'propositions' to 'meanings'? The questions are endless...

Yup, I'll buy that... all the same, the 'metaphysics' is interesting (more to come).

In an earlier message I roughed out a certain 'realist view' one could hold about propositions – that they are entities 'about' sentences; that they depend in some way on 'meanings' but differ from them in that they are susceptible to truth values; and that their truth values somehow depend on their being an 'actual state of affairs' to which the proposition 'corresponds'. I'm deliberately using the language of naïve philosophy here; we'll take a further look at the terms "proposition", "meaning" and so forth as we go along.

I gave a VERY crude representation of the 'true propositions correspond to states of affairs' view:

i. (∀p) p is true → (∃f) f is a state of affairs

Of course, this only says that for any true proposition there is a state of affairs (though it also implies that states of affairs would 'exist' even if there were no true propositions) – it gives no indication of the relation between the proposition and the state of affairs. This is a bit more complex – something like

i*. (∀p) p is true → ((∃!f) f is a state of affairs & uniquely describes )

The relation of 'x uniquely describes y' is, of course, very puzzling. Do we understand 'describes' as 'gives a full account of' or as 'picks out'? And what does either "giving a full account" or "picking out" actually *mean*? Here, we're right back with Frege and Russell (and perhaps Peirce).

As far as I know, only one kind of thing can "describe", "give an account of", and "pick out", and that is language. Unless we're going to imagine some kind of mystical relation (other than simple identity) between 'propositions' and 'states of affairs', language would seem to play a part somewhere. This is unsurprising: if we set aside what is given in the private contents of our states of awareness, the only way we can generate a publicly-accessible, unified, and coherent account of the world is through language (taking the larger sense of 'language' as a 'normative, codified, and structured symbolic system').

We can, I think, suggest a reduction of 'propositions' to 'meaningful sentences' without making too many assumptions about the "starry-eyed metaphysics of it all", but we might have to accept - temporarily and for pre-theoretical reasons - the unanalysed notion of 'state of affairs'. If the general lines of the reduction work, further clarification will depend very much on our metaphysical assumptions…

Is this 'state of affairs' the same as ***context?***

Dunno - what's a "context"?

Well, for example, for the statement "it is raining" its context would include when and where the statement is deemed to apply.

I do have my own ideas about what "states of affairs" might be, but I'd rather work from the assumptions underlying the view (as put foreward by such as C. I. Lewis or Chsholm) that "propositions correspond to states of affairs" without giving an a priori definition of "state of affairs"; an upfront definition would pre-colour my primary concern, which is to get a handle on propositions.

The "state of affairs" view is traceable to the first Wittgenstein ("The world is everything that is the case" etc.); I'd hold that it covers "fact"-based views as well. I'd like to return to other views (things like Fregean and Russellian propositions) later on...

From the work I've done so far, it COULD be the case that 'contexts' correspond to 'states of affairs', and particularly if a 'context' is in any way linked to a "time of speaking"... à suivre, as we say

We crossed messages there - yep, that sounds like a 'time of speaking enlarged to include the place of speaking.

I suppose context also includes the intent of the author. But, of course, that introduces a bit of fuzziness into the discussion. Clearly though, if the author intends to refer to the president of the Chamber of Commerce, it would be a mistake to apply what he said to the president of the Kiwanis Club simply because they are both presidents.

********

Bill--Just experimenting, my apologies if this is taboo, just trying to make a comment more directly on this comment of yours---

The normative definition of a sign logically must include the author. See my Fundamentals of Methodology II-Sign at ssrn.com/author=510356, pp4f, 17ff. I am impressed, btw, Bill. Few academic workers are commonsensical (sophisticated?) enough to appreciate that. (CSH)

For me, the tricky part of Zimmerman's comment is (abridging slightly) "propositions would have existed and been either true or false even in the absence of all sentences or thoughts". We can, I think, take it that an absence of "sentences and thoughts" implies an absence of language (if it doesn't, then Zimmerman has a lot of explaining to do

Setting aside for the instant whether they are 'true' or not, what could these "propositions that would have existed even in the absence of all sentences or thoughts" be? We can perhaps hazard an answer by drawing on the resources of our theories about *language* - and this in , despite of the old chestnut that "a proposition is not a linguistic entity": the only "examples" of propositions I've seen (other than the dummy ones of propositional logic) looked one hell of a lot like sentences; until we come up against a stumbling block (such as the task of determining replacement of that-clauses salva veritate in indirect quotation), I'd rather stick with the intuition that they're linguistic… let's see if I'm wrong.

The view under consideration holds that the number of propositions is greater than the number of "sentences that are uttered or thought". It doesn't, however, hold that the number of propositions is greater than the number of sentences that COULD be uttered or thought – indeed, on certain views, a large number of distinct sentences can express the 'same' proposition (and this question of when 'p' and 'q' are 'the same proposition' will return to haunt us later…), which would seem to entail that on THIS view there would be *fewer* propositions than "sentences which could be uttered or thought". Once again, we shall return to this – the matter hinges on the relation of 'sameness' (is this 'identity'?).

For my part, I was thinking in terms of the meaningful sentences that could possibly be generated in any natural language *given some kind of 'generative grammar'* (though, of course, we have to clarify what we mean by 'meaningful').

We can, I think, accept that in any natural language L, the resources of L's grammar allow us to generate an infinite number of sentences from a finite stock of lexical items; I think that – if 'meaningful sentence' has a sense – we can also accept that it can generate an infinite number of meaningful sentences. So, an infinite number of sentences *could* be generated, even if most of these sentences are never actually uttered or thought. While "possible meaningful sentences" are perhaps as worrying as propositions, at least we know what a sentence is and we have a reasonable understanding of the process by which they're generated. As for the appeal to 'meanings' - well, this is nearly as problematic as an appeal to propositions, but at least Quine gave us some indications of how to get a handle on the problem of how 'sentences relate to the world'. More of this later.

We can try reformulating the sentence "propositions would exist even if the corresponding sentences were never uttered or thought" by replacing the term "proposition" with "possible meaningful sentence" – this, of course, gives us :

ii. "possible meaningful sentences would exist even if the corresponding sentences were never uttered or thought"

Here, we can distinguish between two notions – either:

iia. "it is possible that there are an infinite number of meaningful sentences, even if an infinite number of sentences are never actually uttered or thought"

or

iib. "there are an infinite number of possible sentences, even if an infinite number of sentences are never actually uttered or thought"

The distinction is, of course, between de dicto and de re. (iib) makes an existential claim about 'possible sentences' – there *are* such things; while (iib) would make the 'existence' of the meaningful sentences subjunctive and dependent on the conditions of generation of such sentences. Thus, if the process *allows for* the generation of an 'infinite number of sentences' it is logically possible that an infinite number of sentences could be generated – and I think we could make a strong argument for this reducing to a statement of the basic assumption of generative grammar.

***

If grammar allows illimitable recombination, it is at least *logically* possible to generate an infinite number of sentences. It's metaphysically possible, too: one could imagine an entity that was capable of uttering an infinite number of sentences or – if one prefers a slightly less extreme form of mysticism – we could for example understand a 'possible' sentence as a sentence that is uttered in some possible world. I'd like to look at this second point in my following message

If the intent is private, can it be included in the context?

DH: "I'd like to look at this second point in my following message"

DH: "If the intent is private, can it be included in the context?"

You seem to have failed to carry through with your plan...

DH: "If the intent is private, can it be included in the context?"

If the intent is completely private, no. But the intent may frequently be determined from the context. Now, I am aware that one can end up going in circles with the context dependent on the intent and the intent being determined by the context. That's life, is it not --- a bit fuzzy right down to the core?

I chose "following" rather than "next" to allow at least for the possibility of answering some smart alec who might point out that my 'next' message was not ACTUALLY about the matte in hand - "following" here implied a sequence based on the arguments under development rather than on the fundamental temporal and graphemic sequence of "any message I might post". I should, of course, have made this explicit.

Dang!

Clearly there are infinitely many possible English sentences...

... The number Pi is approximately 3.

... The number Pi is approximately 3.1.

... The number Pi is approximately 3.14.

... The number Pi is approximately 3.141.

... The number Pi is approximately 3.1415.

... The number Pi is approximately 3.14159.

... etc.

DH: "I chose "following" rather than "next" to allow at least for the possibility of answering some smart alec who might point out that my 'next' message was not ACTUALLY about the matte in hand - "following" here implied a sequence based on the arguments under development rather than on the fundamental temporal and graphemic sequence of "any message I might post". I should, of course, have made this explicit."

Looks like you got my number...

Dang!!

What I like about you, Bill, is that you oblige me to be VERY careful with my formulation - even to the level of replacing naïve set-theoretical notions with explicit mereological notions.

BTW, Bill: thanks for an excellent example of an effectively infinite sequence of "possible sentences"....

David,

My pleasure!! I have given the matter a lot of thought. It all goes back to Alexander Pruss and his 'possible worlds' theory.

One thing to note is that the sentences are not simple variations on a theme. One might say, for example...

... A is A.

... A is A and A is A.

... A is A and A is A and A is A.

...

Although the sentences are all different, they would all have the same meaning. Not so for the previous example.

As we shall see... and I shall return to "meanings".

To return to the problem in hand – the uncomfortable implications of Zimmerman's remark "at least in the case of a proposition that is not about a particular sentence or thought, it would have existed and been either true or false even in the absence of all sentences or thoughts"

I remarked that, if grammar allows illimitable recombination, it's metaphysically possible to generate an infinite number of sentences: one could, for example, understand a 'possible' sentence as "a sentence that is uttered in some possible world".

Now, I think the ontological price we pay by introducing 'possible worlds' at this point is offset by the benefit we get from reducing 'propositions' to 'sentences'. The problem is not really whether using possible worlds adds to our ontology or not – it's more a question of whether it EXPLAINS anything about propositions, and here we have to bear in mind that not only are they the kind of things that correspond to the "that" clause of indirect quotation, but they're also the kinds of thing that we can have ***attitudes*** about, and any workable reductionist theory would have to show how THIS kind of thing can be explained in terms of 'sentences' (and it's the "attitude" angle that is, I think, important in the context of Zimmerman's comments, which roughly concern "irreducible temporal attitudes").

Anyway, let's look at how the "possible world" approach might work, and then see if we can actually use it:

If we take 's' to be any member of the 'infinite number of sentences that could be generated in a given language', we could give something like:

iii. (∀s) (∃w) [(∃x) x is in w & utters]

Now, in itself (iii) is unobjectionable to a nominalist understanding of propositions and to a de dicto understanding of PWA as the modal (or – more correctly – "alethic") context doesn't fall within the scope of an alethically-unmodified existential quantifier: informally, we're not saying that ***there is*** some sentence such that that sentence is uttered in some possible world, but just that sentences in general are the kind of thing that ***could*** be uttered.

(iii) simply allows that we could generate an infinite number of sentences and that it would be possible for each of those sentences to be uttered; but it doesn't allow that those sentences be 'meaningful'. Now, rather than develop a complex theory of meanings at this point, I'd like to turn to another unanalysed relation: that of 'truth' (Zimmerman characterises propositions - whether they are uttered or nor - not only as 'existing' but also as being susceptible to fixed truth values). For the moment, I'm going to write as if the notion of 'truth' had some sense (we can return to this later), and take it for the sake of argument that a 'meaningful' sentence is one that is susceptible of being either true or false (and – for the sake of argument – tertium non datur).

Nonetheless, if we introduce 'truth' into the model given in (iii), we're already faced with a theory-dependent choice: should we prefer

iv. (∀s) (∃w) [(∃x) x is in w & (utters & (s is true ∨ s is false)]

or

v. (∀s) (s is true ∨ s is false) & (∃w) [(∃x) x is in w & utters]

(iv) places the truth of 's' within the alethic context; if we follow our reasoning, this should imply that (s is true-in-w ∨ s is false-in-w). Now, this would imply that truth is relative to worlds and it's not at all certain that we could export the notion to alethically unmodified quantification. It follows that if a sentence's being true or false somehow contributes to its meaning, this would imply that 's' has ***meaning-in-w***; and I'm not sure that this is a strong enough notion to allows us to found the idea that, potentially, there are an unlimited number of meaningful sentences.

(v) is, I would say, closer to Zimmerman's conception of propositions: they are entities which are true or false regardless of whether they are uttered or not. However, (v) can be seen as allowing that an unlimited number of ***meaningful*** sentences could be uttered and that there 'meaning' would hold of the actual world, and not just of some possible world..

This distinction is, I think, very important. Do we consider that we could produce an unlimited number of sentences which would ***in themselves*** be true or false of the actual world, or should we hold that those sentences would only be susceptible of truth values if they were uttered (or thought)?

The metaphysical assumptions underpinning this require that we return to the notion of 'facts' or 'states of affairs'; this will be the subject of my next posting IN THIS SEQUENCE.

(I posted that because I just wanted to sketch out the general model for your criticism before turning to "interpretations")

"The world is everything that is the case" (Tractatus Logico-Philosophicus, I)

'WHAT IS THE CASE', 'FACTS', 'STATES OF AFFAIRS'… POSSIBLE WORLDS?

Evidently "everything that is the case" is not "one thing" (this would resemble the slingshot - http://en.wikipedia.org/wiki/Slingshot_argument), so we should understand this as "all those things which are the case". Wittgenstein's initial statement would therefore suggest that the world is the "sum of all those things that are the case". For the first Wittgenstein, "a thing that is the case" is a 'fact', and a 'fact' is an "existing state of affairs".

'States of affairs' are interesting here, as they would seem to constitute a superclass of which 'facts' are a subclass. Of course, 'existing' could be given a restrictive sense, but in that case 'state of affairs' would simply reduce to 'fact'; and this leads to two very interesting questions:

1. If 'facts' are a subclass of 'states of affairs', what is the status of those states of affairs which are NOT facts?

2. What is it that makes a given 'state of affairs' into a 'fact'?

That which is the case is certainly actual, not merely possible.

Thank you, Bill - the answer in a nutshell.

#H1-215. David Hirst ,Nov 11,2009 10;02 I liked this. - Visual observations do also communicate, don.t they. They may not be able to build reasoning in the mind of the observer, but they do are capable of giving rise to emotions that can even lead to subsequent-thoughts and/or actions, mostly latter.

It is possible/true only "linguistic thoughts" can generate theory, but "sensory perceptions" do also seem capable of instinctual response and feelings. Why we would not treat the theory and feeling, both as equally important? Is therefore possible to find some structure that can represent T/F and performatory/nonperformatory answers; so that, this can include propositions as well as other sentences. Sentences are enough?

#H1-218. Ref David Hirst, Nov 12, 2009 9:28 pm. "generating" versus "uttering" sentences. Do we consider some "unauthored" sentences too? "generated" looks somewhat like mathematicalised idea of sentence, "uttered" looks like more physical and real. A computer or pick-of-lots generated sentence, if it makes sense, we have no way of differentiating it from uttered sentences. And an uttered sentence if a mere jumble words, then that too wouldn't make sense, and that wouldn't be a sentence. From what we were deeming as "sentence" in common parlance, a "meaningless sentence" is an oxymoron.

#After (2 pages, 38 posts), the discussion where underground to?

I have only added a student's write-down-to-understand.