London Ed sends his thoughts on language and reality. My comments are in blue.
Still mulling over the relation between language and reality. Train of thought below. I tried to convert it to an aporetic polyad, but failed. The tension is between the idea that propositions are (1) mind-dependent and (2) have parts and so (3) have parts that are mind-dependent. Yet (if direct reference is true) some of the parts (namely the parts corresponding to genuinely singular terms) cannot be mind-dependent.
How about this aporetic hexad:
1. Propositions are mind-dependent entities.
2. Atomic (molecular) propositions are composed of sub-propositional (propositional) parts.
3. If propositions are mind-dependent, then so are its parts.
4. In the case of genuine singular terms (paradigm examples of which are pure indexicals), reference is direct and not mediated by sense.
5. If reference is direct, then the meaning of the singular referring term is exhausted by the term's denotatum so that a proposition expressed by the tokening of a sentence containing the singular referring term (e.g, the sentence 'I am hungry') has the denotatum itself as a constituent.
6. In typical cases, the denotatum is a mind-independent item.
Note that (3) is not an instance of the Fallacy of Division since (3) is not a telescoped argument but merely a conditional statement. London Ed, however, may have succumbed to the fallacy above. Or maybe not.
Our aporetic hexad is a nice little puzzle since each limb is plausible even apart from the arguments that can be given for each of them.
And yet the limbs of this hexad cannot all be true. Consider the proposition BV expresses when he utters, thoughtfully and sincerely, a token of 'I am hungry' or 'Ich bin hungrig.' By (4) in conjunction with (5), BV himself, all 190 lbs of him, is a proper part of the proposition. By (6), BV is mind-independent. But by (1) & (2) & (3), BV is not mind-independent. Contradiction.
Which limb should we reject? We could reject (1). One way would be by maintaining that propositions are abstract (non-spatiotemporal) mind-independent objects (the Frege line). A second way is by maintaining that propositions are concrete (non-abstract) mind-independent objects (the Russell line). Both of these solutions are deeply problematic, however.
Or we could reject (3) and hold that propositions are mental constructions out of mind-independent elements. Not promising!
Or we could reject (4) and hold that reference is always sense-mediated. Not promising either. What on earth or in heaven is the sense that BV expresses when BV utters 'I'? BV has no idea. He may have an haecceity but he cannot grasp it! So what good is it for purposes of reference? BV does not pick himself out via a sense that his uses of 'I' have, that his uses alone have, and that no other uses could have. His haecceity, if he has one, is ineffable.
So pick your poison.
By the way, I have just illustrated the utility of the aporetic style. Whereas what Ed says above is somewhat mushy, what I have said is razor-sharp. All of the cards are on the table and you can see what they are. We seem to agree that there is a genuine problem here.
- There is spoken and written language, and language has composition with varying degrees of granularity. Written language has books, chapters, paragraphs, sentences and words. The sentence is an important unit, which is used to express true and false statements. [The declarative sentence, leastways.]
- Spoken and written language has meaning. Meaning is also compositional, and mirrors the composition of the language at least at the level of the sentence and above. There is no complete agreement about compositionality below the level of the sentence. E.g. Aristotelian logic analyses 'every man is mortal' differently from modern predicate logic. [Well, there is agreement that there is compositionality of meaning; but not what the parsing ought to be.]
- The meaning of a sentence is sometimes called a 'proposition' or a 'statement'. [Yes, except that 'statement' picks out either a speech act or the product of a speech act, not the meaning (Fregean Sinn) of a sentence. Frege thought, bizarrely, that sentences have referents in addition to sense, and that these referents are the truth-values.]
- There are also thoughts. It is generally agreed that the structure of the thought mirrors the structure of the proposition. The difference is that the thought is a mental item, and private, whereas the proposition is publicly accessible, and so can be used for communication. [It is true that acts of thinking are private: you have yours and I have mine. But it doesn't follow that the thought is private. We can think the same thought, e.g., that Sharia is incompatible with the values of the English. You are blurring or eliding the distinction between act and accusative.]
- There is also reality. When a sentence expresses a true proposition, we say it corresponds to reality. Otherwise it corresponds to nothing. So there are three things: language, propositions, reality. The problem is to explain the relation between them. [This is basically right. But you shouldnt say that a sentence expresses a proposition; you should say that a person, using a declarative sentence, in a definite context, expresses a proposition. For example, the perfectly grammatical English sentence 'I am here now' expresses no proposition until (i) the contextual features have been fixed, which (ii) is accomplished by some person's producing in speech or writing or whatever a token of the sentence.]
- In particular, what is it that language signifies or means? Is it the proposition? Or the reality? If the latter, we have the problem of explaining propositions that are false. Nothing in reality corresponds to 'the moon is made of green cheese'. So if the meaning of that sentence, i.e. the proposition it expresses, exists at all, then it cannot exist in mind-independent reality. [This is a non sequitur. It can exist in mind-independent reality if it is a Fregean proposition! But you are right that if I say that the Moon is made of green cheese I am talking about the natural satellite of Earth and not about some abstract object.]
- But if a false proposition suddenly becomes true, e.g. "Al is thin" after Al goes on a diet, and if when false it did not correspond to anything in external reality, how can it become identical with the reality? And we say that such a proposition was false, but is now true, i.e. the same thing that was false, is true. But if the reality is identical with the proposition that is now true, and if the same proposition was once false, it follows that the proposition, whether true or false, is not identical with anything in external reality. [One issue here is whether a proposition can change its truth-value. Suppose we say that a sentence like 'Al is fat' is elliptical for 'Al is fat on Jan 1, 2015.' The latter sentence expresses a Fregean proposition whose TV does not change. Fregean propositions are context-free: free of indexical elements including tenses of verbs. And who ever said that correspondence is identity?]
- It follows that the relation between language and reality is indirect, i.e. always mediated by a proposition. A sentence, to be meaningful at all, signifies or expresses a proposition, and a relation between the proposition and reality exists if the proposition is true, but not when the proposition is false. [I'll buy that.]
- But what sort of thing is a proposition? It is a publicly available object, i.e. available to the common mind, not a single mind only, but not part of external mind-independent reality either. [You are asking a key question: What is a proposition? It is a bitch for sure. But look: both Fregean and Russellian propositions are parts of external mind-independent reality. Do you think those gentlemen were completely out to lunch? Can you refute them? Will you maintain that propositions are intentional objects?]
- We also have the problem of singular propositions, i.e. propositions expressed by sentences with an unquantified subject, e.g. a proper name. It is generally agreed that the composition of singular sentences mirrors the structure of the corresponding proposition. In particular the singular subject in language has a corresponding item in the proposition. Thus the proposition expressed by 'Socrates is bald' contains an item exactly corresponding to the word 'Socrates'.
- But if propositions are always separate from external reality, i.e. if the propositional item corresponding to 'Socrates' is not identical with Socrates himself, what is it? [You could say that it is a Fregean sense. But this is problematic indeed for reasons I already alluded to anent haecceity.]
- Russell's answer was that singular sentences, where the subject is apparently unquantified, really express quantified propositions. If so, this easily explains how the proposition contains no components identical with some component of reality. [Right.]
- But it is now generally agreed that Russell was wrong about proper name sentences. Proper names are not descriptions in disguise, and so proper name propositions are not quantified. So there is some propositional item corresponding to the linguistic item 'Socrates'. [And that item is Socrates himself! And that is very hard to swallow.]
- But if the proper name is not descriptive, it seems to follow that the singular proposition cannot correspond to anything mental, either to a single mind or the group mind. Therefore it must be something non-mental, perhaps Socrates himself. [Or rather, as some maintain, the ordered pair consisting of Socrates and the property of being bald. You see the problem but you are not formulating it precisely enough. When I think the thought: Socrates is bald, I cannot possibly have S. himself before my mind. My mind is finite whereas he is infintely propertied.]
- This means that sentences containing empty names cannot be meaningful, i.e. cannot express propositions capable of truth or falsity. [I think so.]
- This is counter-intuitive. It is intuitively true that the sentence "Frodo is a hobbit" expresses or means something, and that the meaning is composed of parts corresponding to 'Frodo' and 'is a hobbit'. But the part corresponding to 'Frodo' cannot correspond to or signify anything in external reality, i.e. mind-independent reality. [Yes]
- So what does 'Frodo' mean? [You could try an 'asymmetrical' theory: in the case of true singular sentences, the proposition expressed is Russellian, while in the case of false singular sentences the proposition expressed is Fregean. Of course that is hopeless.]