The world holds its breath.

An excuse for talking as much as I do about chess and baseball:

Chess, like language use most of the time, is turn-based. Baseball is kind of a hybrid -- there are things that happen simultaneously, but the structure of the game itself is so fundamentally turn-based that it has a turn-based feel to it, unlike, say, basketball or soccer.

A game of chess is a sequence of moves by both sides, and those moves together constitute a game. Even though there are skills you must have to play chess at all, and other skills you must have to play it well, those skills are not enough to have a game; you need an opponent, and the two of you together create a game. Baseball similarly has a myriad of skills that make up ability to play the game, but two teams must take the field and play together for their to be a game.

Both games have a natural back-and-forth, action-and-reaction rhythm. No action stands on its own, but is part of a phrase (as in music): he develops a knight and attacks the center, so I pin his knight with my bishop, so he kicks the bishop with his rook pawn, and then I retreat, having created a weakness I'll try to exploit later; the pitch is inside so the batter turns on it sending a high grounder toward third, which the third baseman has to take deep in the hole and, since he won't be able to make the throw to first in time, he pivots and throws to catch the runner heading for second, who slides wide to try to avoid the tag he knows is coming because the whole play's in front of him.

Both games show this move and countermove, move and response pattern, where each side's actions interlock with the other's side's previous actions, and together they form a meaningful whole. When you look at an action on its own, you can only imagine sequences in which it would make sense, but it doesn't on its own. The pitcher throws to first -- how? why? Was it a put-out? Or is he holding a runner? If holding a runner, should he? Is he screwing with the batter? Is he a nervous rookie? With chess it's even more obvious that a move on its own could mean almost anything, and what it means depends entirely on the course of the game it's part of. The "same move" (say, Re1) may be played multiple times in a game and have a completely different meaning each time.

What I find so uninspiring about Davidson is the choice of starting point: someone has blurted a sentence and I must interpret it. That's not how language is actually used at all. It's a back-and-forth cooperative behavior with participants contributing to a whole, much like chess or baseball. (Yes, chess and baseball are also cooperative, I hope in a perfectly obvious way, though also competitive, as conversation is sometimes too.)

All of this is rule-governed, but the rules don't tell you any of this. Nevertheless, the rules enable the complex behavior we get to enjoy. Of course, the rules of language, broadly construed, change more than almost anything else, including chess and baseball, but that doesn't mean that what rules there are don't function, perhaps temporarily, as the inner structure that supports the elaborate constructions of language use.

The bottom of page nine... — creativesoul

I've looked at the post a couple times, since you keep suggesting you provided all the answers there, and it's not doing much for me. On the one hand, sure it's reasonable to distinguish cases Davidson lumps together -- Lepore and Stone for instance argue that malapropisms, nonsense, and neologisms should all be treated quite differently. But just distinguishing cases he chooses not to is not enough; you also have to provide an analysis more compelling than his, and I don't see that on the bottom of page nine.

If what were this simple? — creativesoul

This:

The conventional rules of language use cannot take account of the success of unconventional use. — creativesoul

Stated as you have here, this sounds like a truism. But I see no reason to believe what you have written here, even discounting the meaninglessness of "take account of".

You're presupposing that the conventional discourse has 'it' right — creativesoul

That's not saying much, because the views that might plausibly be taken to have been attacked by Davidson include pretty much everybody's, and it's not like they all agree with each other.

If it were this simple, not only would Davidson's paper have only been one sentence, but it wouldn't have been needed in the first place because there'd be nobody holding the opposing view. It's not.

there is nothing in chess analogous to malapropistic expressions. — Janus

Actually there is! But these would be violations not of the pure syntax of chess, but either of its "school grammar", the received wisdom of how to play properly, or of its "real grammar", how to put moves together in a way that makes sense. For the former, you could look at ideas of "positional chess" as they've been understood through the game's history, the rise of the hypermodern school, that sort of thing, or even at a brilliancy that is "anti-positional" but works because of the specific position on the board.

But there's a fantastic illustration of the latter, a violation of chess's actual grammar. (Citation when I get home and can look it up.) This was a game between Kasparov and some candidate-level player. They were playing the first few moves noncommittally and indirectly, screwing around with move order before settling into a specific opening, the way grandmasters do, but at like move 3 or 4, the other guy made a transposition that actually hangs the exchange! Kasparov could have gotten a rook for a bishop and his opponent would have resigned at like move 5. Unheard of at this level! But Kasparov didn't even notice the mistake, not until after the game I think, because stuff like that just doesn't happen.

Sadly, google will not find it for me.

One last but of baseball lore, for those who like that sort of thing.

How to spot a rookie or an idealist: a fielder who, after making the last put-out of an inning, especially if it was a 1-2-3 inning, tosses the ball over to the mound or to a player of the opposing team coming onto the diamond to field.

No no no! That was your pitcher's ball. You have no idea what he did to it, especially if you had an up-and-down inning. No, that ball does not go to the opposing pitcher, but to the home plate umpire, who might look it over and keep it in play or start a fresh one.

I meant to mention baseball!

Baseball distinguishes between acceptable and unacceptable cheating.

A runner on second is expected to try to steal the catcher's signs and tip off the batter; but a team that puts someone with binoculars and a walkie-talkie in a box above center field is cheating.

Hiding the ball is part of the game, but there are some complications to that since sometimes the ball is live and sometimes it's dead. Deking -- I don't know how to spell that, it's short for "decoying" -- is when you pretend you're taking a throw to make a runner slide, and thus stop at whatever base you're covering when he could have continued running (if, say, the ball got past an outfielder on a bad hop); this is unacceptable because players risk injury by sliding. I've seen players come to blows over it.

And then oh my god there's pitchers and whatever they put on the ball...

There is a legitimate source of tension here.

Is playing chess just a matter of following the rules? In some sense, yes. Do the rules alone explain what happened in a game of chess? Obviously not.

Even in the improvised games of children, where the rules are modified while playing -- a better analogy for linguistic communication -- there are sometimes changes "proposed" that the other players will balk at. "You can't do that." "No, that's not fair." "If you can do that, then you just win." (I'm not guessing about this; I have six kids.) That sounds a lot like there are rules about what rules you can make. Of course, sometimes someone gets away with it.

What would Davidson say?

What we've got here is failure to communicate. — Cool Hand Luke

Surely there are cases where there is a failure to communicate that we'd be inclined to explain by either a speaker misusing a word, or by an interpreter misunderstanding it.

Davidson doesn't give any such examples, perhaps because he explains them as failure to come to agreement on a passing theory, rather than a lack of competence on the part of speaker or interpreter.

But it sure looks like that sometimes. Anyone got a nice example?

misattributing meaning to the words actually used — creativesoul

That's a funny thing. We can easily conceive of Mrs Malaprop having a conversation with her sister in which neither has any idea that they are using several words in ways the rest of the English-speaking community considers deviant. Their prior theories, we could hypothesize, would match, though it is a theory that differs systematically from the mainstream.

There's no point to this, of course, except humor, because the hypothesized matching of prior theories is just a fairy tale.

I'm going to have to work for a while, but quickly:

It seems you read Davidson as saying that conventions play no role in understanding what someone says. — Banno

I haven't figured out how to read him.

That's not how I read the article; it's rather that conventions are not the whole of what is involved. — Banno

Yes, you've said things like this before.

lexical meaning - convention - is insufficient; that language is more than following conventions.

Rule-based conventions are part of language, but not fundamental to it. — Banno

And I've asked before if this is saying linguistic conventions are necessary but not sufficient for communication. I don't understand what "not fundamental" means here.

And then shortly after you said:

language does not rely on rules. — Banno

And that sounds to me like saying linguistic conventions are not only insufficient for successful communication, but unnecessary.

What is Davidson's position?

I told my soccer story not because it's a good analogy for language use or communication -- it's not -- but because it's no longer clear to me what the engine of Davidson's argument is. Does it have anything at all to do with language? Or is the real argument to do with breaking rules and getting away with it, any kind of rules?

I can't fully address your other post, but the passage you mentioned (pp. 263-264) comes after he matter-of-factly observes that prior theories aren't shared anyway (so we're casting about for something more general that might be). If prior theories aren't shared they never even met his criteria (1)-(3). He could have started there. Why is it only halfway through the paper that he mentions that the hypothesized candidate for explaining communication was always a non-starter?

aren't these aspects of the prior theory? — Banno

I think this is the form of the prior theory, and thus the form of the passing theory that actually does the work, according to Davidson.

I'm not sure how else to understand this:

Principles (1) and (2) survive when understood in rather unusual ways,

What's your take on the survival of (1) and (2)?

True story:

When I was in college, I drove over to Atlanta one day to go to the Ansley Mall Bookshop (RIP) and brought a stack of books by Wittgenstein up to the checkout counter. Guy there says, "Philosophy major, huh? I was a philosophy major. [Pointing] He was a philosophy major. She was a philosophy major too."

Guess what I do for a living now. Yes, that's right, I work at a bookstore.

Malapropisms introduce expressions not covered by prior learning, or familiar expressions which cannot be interpreted by any of the abilities so far discussed. Malapropisms fall into a different category, one that may include such things as our ability to perceive a well-formed sentence when the actual utterance was incomplete or grammatically garbled, our ability to interpret words we have never heard before, to correct slips of the tongue, or to cope with new idiolects. These phenomena threaten standard descriptions of linguistic competence (including descriptions for which I am responsible). — p. 255

The key term in this passage, the target of Davidson's argument, is linguistic competence.

If you have a glance at the cluster of related Wikipedia pages, you'll find that classical malapropisms are a type of speech error and linguists generally classify them as competence errors. Mrs. Malaprop is a sort of walking Dunning Kruger effect, who believes she knows more about some English words than she does. Her speech, on the usual view, is not riddled with simple performance errors such as slips of the tongue, but with perfectly deliberate utterances that betray a lack of understanding of what the words she's saying mean.

What Davidson notices is that she "gets away with it": her interpreters take her as saying what she thinks she's saying rather than as what she's actually saying. That this happens, is a fact. That it happens in real life, is a fact. So how are we to characterize these facts? Shall we say that the interpreters of Mrs. Malaprop, or of anyone who in real life misspeaks in any of a great variety of ways, have a competence as interpreters that can make up for the performance errors, of whatever origin, of speakers?

Let's suppose we do want to describe this as a linguistic competence. Davidson argues that such a competence would necessarily be ad hoc, a theory of meaning for a language with a vanishingly small field of application: what this speaker is saying on this occasion. It can be described formally using whatever sort of semantics you like, so that it looks like a semantics for a natural language, but the argument is supposed to have established that an interpreter will need a new semantics for each speech encounter, and that's not what anyone thinks of as the semantics, or theory of meaning, for a natural language.

Davidson's conclusion is that the idea of linguistic competence must be rejected:

The problem we have been grappling with depends on the assumption that communication by speech requires that speaker and interpreter have learned or somehow acquired a common method or theory of interpretation—as being able to operate on the basis of shared conventions, rules, or regularities. The problem arose when we realized that no method or theory fills this bill. The solution to the problem is clear. In linguistic communication nothing corresponds to a linguistic competence as often described: that is, as summarized by principles (1)–(3). — p. 265

That's the argument up to here, and it's clear enough how it works, and how we might accept or critique it. But the concluding paragraph continues:

The solution is to give up the principles. Principles (1) and (2) survive when understood in rather unusual ways, — ibid

(1) First meaning is systematic; (2) First meaning is shared. (1) and (2) can be taken as describing only passing theories -- this is the "unusual way" -- so is Davidson here endorsing passing theories as a genuine model of successful communication? This is what Lepore and Stone call "improvised meaning".

but principle (3) cannot stand, and it is unclear what can take its place. — ibid

The problem is that passing theories clearly cannot be learned in advance; you cannot be competent in the use of a passing theory before the theory even exists and it doesn't exist until the specific speech encounter in which you, as interpreter, produce it to cope with the specific utterance you are faced with. Thus, "conventions, rules, or regularities" are all out.

I conclude that there is no such thing as a language, not if a language is anything like what many philosophers and linguists have supposed. There is therefore no such thing to be learned, mastered, or born with. — ibid

And here it's clear that Davidson means to sweep up not just everyone who defends convention as the basis of linguistic competence, but all of generative linguistics. Davidson nowhere mentions Chomsky in the paper, but it's Chomsky's idea of "linguistic competence" (in distinction from "performance") that he is attacking: not just conventions have to go, but rules too and regularities, whether you learn them or are born with them. There is no such thing as linguistic competence of any kind under any description, although Davidson has a particularly dim view of convention, which I for one have been distracted by:

We must give up the idea of a clearly defined shared structure which language-users acquire and then apply to cases. And we should try again to say how convention in any important sense is involved in language; or, as I think, we should give up the attempt to illuminate how we communicate by appeal to conventions. — ibid

Why just "convention" here? What about the rules or regularities we might have been born with?

Another quick tour of Wikipedia makes it clear that Davidson is far from alone in critiquing Chomsky's distinction between competence and performance; but most of the critique from within linguistics has come broadly from pragmatics, from the anthropological approach, from people like Dell Hymes and his "communicative competence" or from functionalists, and those folks are going to tend to insist on the importance of culture and context, to drag in even more conventions to explain how linguistic communication works.

Davidson seems to want to cut off this approach at the knees too. His stand in for all of that, in this paper, is Paul Grice.

Davidson believes he has an argument that shows not just that linguistic competence, along Chomskyan lines, is too narrow a conception of competence, that, whether you're born with it or learn it, your specifically linguistic competence is never enough to explain how you communicate using language but relies on some further cultural competence, describable as mastery of another system of conventions, but that communication by means of language cannot be captured by any sort of convention or rule or regularity, because sometimes people break the rules, whatever rules, break them deliberately or inadvertently or through ignorance, and get away with it.

That is the argument of the paper as I understand it.

Is it a good argument?

When I was in high school I played soccer and the best player on our team, genuinely talented guy, also cheated now and then. I saw him do it. He could take a clear downfield from a defender by carefully, subtly, laying his hand right alongside his thigh to help cushion -- just for a second -- the fall of the ball and it would look, even to a ref standing right there, like he had gently taken the ball on the upper part of his thigh while running -- so of course his hand passes by his leg -- to bring it to the ground under his control. It was perhaps the most artful cheating I have ever seen. He got away with it. We would all laugh about it because Scott was both a fine player and an accomplished cheater.

Does it prove that soccer doesn't have rules?

I understand the theory. But we're talking about malapropisms and Davidson's claim that the intended meaning takes over from the literal meaning.

Is my T-sentence not what Davidson tells us would be part of the interpreter's passing theory? Did I misunderstand him?

"That's a nice soup latrine" is true IFF that soup looks like shit — Banno

I'm really not sure what you think you're demonstrating here.

Maybe it was a genuine malapropism; maybe it was, as I said above, a sort of pun, and some of the conventional meaning of "latrine" is brought along. There's not much to say for one interpretation or another without a little more context.

This is all beside the point though. If it's a malapropism, your T-sentence is

(T1) "That's a nice soup latrine" is true IFF that's a nice soup tureen.

And that's pretty clearly a problem. Davidson's whole point is that you could not possibly have learned such a rule in advance. I don't think any of us are contesting that -- of course you couldn't have. The question is whether you come up with this rule, a contextual definition of "latrine", on the fly and use it just the once, rather than dealing with the unexpected utterance of "latrine" in some other way.

The last couple of pages haven't added anything to the commentary. Indeed, they detract from it. — Banno

Are you not entertained?

There is an actual difference between interpretation and attribution. The latter gives rise to the former. — creativesoul

WeekWell that's a question, isn't it. What sort of thing is a Davidsonian method of interpretation?

Davidson is very clear that the literal meaning of each word is in play when an interpreter interprets a speaker; the interpreter is not just working out how the meaningful bits have been assembled and what the speaker means to say by having so assembled them.

Linguistics? — creativesoul

Yes, I just haven't bothered to look in a while, but I suspect what I described is the sort of thing linguists argue about -- how compositionality should be understood. But singesome of this, like there being stages, could be tested.

In which case, the speaker intended to say "latrine", and the conventional meaning of the term aligns perfectly with the speaker's intent. — creativesoul

Not sure it's that simple. When I was a boy scout, "latrine" meant a ditch or a hole in the ground; you won't see such a thing on anyone's dining room table. So it's still not just literal; an element of the meaning of "latrine" (container for excrement) is being borrowed on the pretense of the similar sound of the words "tureen" and "latrine". So, a sort of pun.

To continue speculating...

If you know that your dinner guest is looking at your dining table, and seems to be commenting on it, "latrine" is not one of the words you expect them to utter, because you know there is no latrine on the table; you do know there is a tureen on the table, so that word is on the list of words you expect to be uttered.

Whatever lexical lookup occurs, it seems not to range over all possible words, nor over all possible English words (respecting English morphology), nor even over all words we might think the speaker knows, but only over the words we think the speaker knows and expect under the circumstances. As @Isaac would say, we're predicting all the time.

Thus in normal conversation, it's possible to mis-speak and have the mistake corrected unnoticed.

It could be that malapropisms are noticed, when they are, because the substitution suggests an interpretation that could be relevant, as puns do. Thus "latrine" could be an innocent mistake, or a suggestion, conscious or not, that your soup is excremental.

they do have to knowingly misattribute meaning to "latrine" by virtue of drawing a correlation between it and the referent of "tureen". — creativesoul

If you were compiling a dictionary or a Tarski-style model of the English language, what would you do with the "latrine" sentence? If it's a one-off, the only attestation for "latrine" used as a synonym for "tureen", you could drop it from your sample.

But as Davidson would have it, the interpreter must assign a meaning to this sentence as it stands, and that means saying that in this sentence "latrine" is a synonym for "tureen".

Lepore and Stone say no: we'll take you not to have said "latrine" at all, but to have mispronounced "tureen". We could say, I think plausibly, that what makes the mispronunciation a malapropism is that the sound you make happens already to be a word. If you said "stureen" (perhaps because of the association between "soup" and "stirring") we're not really in dramatically different territory: it's still just an understandable mispronunciation.

But then must the listener treat "stureen" as a nonce synonym for "tureen"? I think Davidson says yes.

The real question, then, is what do we interpret? Do we assign meaning to the specific tokens you produce? Or is there a little preprocessing first, a little data-scrubbing?

This actually looks like a problem for Davidson, because we know, for a fact, that often mistakes go unnoticed, you read right over typos, and so on. That suggests the sentence interpretation stage comes after some initial classifying of the word tokens -- so by the time meaning gets assigned to the sentence, word-meaning has already, perhaps tentatively, been assigned to the word tokens, and sentence meaning is built out of those, not from the raw tokens.

(We know that the actual noises you make are classified as phonemes somewhat like this, and also somewhat early in the interpretation process.)

That word token processing stage seems to include some error correction, just as it does with phonemes. The dilemma of whether, in this sentence as it stands, we take "latrine" as a synonym for "tureen", may not arise.

But a whole lot of this post should be replaced by actual science...

Where, if at all, does explanation terminate?
— Janus

In action. — Banno

Out of context, that sounds weirdly fascist.

Eh. Probably just me.

repetition — Janus

Good point. Repeatability is crucial.

In general, though, I'd say the mountains of science about how children learn their native language should be the starting point, and all of this is armchair stuff.

So your imitation does not have to be correct in order to be understood? — Banno

I don't know what you were talking about, but I was talking about just plain imitating, like the babbling of pre-linguistic children.

Imitation is yet more rules. — Banno

Imitating is just imitating. If no one is judging whether you've correctly copied or how good the copy is, I'm not sure what you're worried about.

It's by imitation that conventions become established, not by people consciously seeing them as sets of rules to be followed — Janus

Imitation alone clearly couldn't establish language. That doesn't even make sense.

Big part of learning it, sure.

I'm not sure we ever squarely faced Davidson's central claim. Take Lepore and Stone's example:

That's a nice soup latrine.

I think everyone would agree

1. You said "latrine" when you meant to say "tureen".

Lepore and Stone describe the situation as

2. You mispronounced "tureen" as "latrine".

What's still not perfectly clear is whether

3. By "latrine" you meant "tureen"'.

That is, whether you were, consciously or not, assigning the meaning of the word "tureen" to the word "latrine".

And then, finally, there's the question of whether the interpreter must say

4. In this sentence the word "latrine" means "tureen".

It’s just that two clouds being clouds requires the existence of couldhood — Tristan L

Here's some stuff about clouds:

In meteorology, a cloud is an aerosol consisting of a visible mass of minute liquid droplets, frozen crystals, or other particles suspended in the atmosphere of a planetary body or similar space.[1] Water or various other chemicals may compose the droplets and crystals. On Earth, clouds are formed as a result of saturation of the air when it is cooled to its dew point, or when it gains sufficient moisture (usually in the form of water vapor) from an adjacent source to raise the dew point to the ambient temperature. — Wikipedia

There's a definition grounded in science, so it's a little more precise than the everyday understanding of "cloud", and some explanation -- tells you what clouds are actually made of, for instance, which is not perfectly obvious from their appearance, and why they form.

What would the addition of the term "cloudhood" add to this discussion? What would it enable you to know or to say that meteorologists and ordinary folk cannot?

Interesting response from Ernest Lepore and Matthew Stone:

So, what, then, is our take-home lesson? It is that, as semanticists, we should reject Davidson’s explanations in “Nice Derangement.” There is no reason to believe in Davidson’s passing theories, i.e., in improvised meanings! — Convention before Communication

I am aware.

But to me it seems obvious that getting around the regress problem involved with rules cannot simply involve the application of more rules. — Janus

No, of course not.

This getting a little far afield, nevertheless...

You can also describe the regress as needing first to understand the language in which the rules are expressed, and needing before that..., etc.

For all that it's still clear to me that rules play an enormous role in language use, so we need a way around the regress.

The beginning of that is to note that the 1 or 2 year old learning 'apple' (or being taught it) is doing something noticeably different from the 13 or 14 year old learning 'momentum'.

(Recommended: "Some Reflections on Language Games" by Wilfrid Sellars.)

At any rate, Davidson does not make a regress argument against conventions or rules, so perhaps he too feels the regress can be gotten around.

There's this idea of a language war between descriptivists and prescriptivists, but that would give you two insane positions: either "there are no mistakes," or "usage doesn't matter". Basically, you just navigate a linguistic landscape, accept some rules (and maybe internalise them, or maybe just pay lipservice), and discard others (maybe as a deliberate choice, or because your word habits are too strong and you just forget). Rules can be internalised from (3) into (2), and (1) can make hypotheses about when that happens (but I'm not sure how good (1) is at that currently). — Dawnstorm

Yeah that's a really curious point: debates about theory have a natural analog in our linguistic behavior, in part because using language seems always to lead to some personal theorizing about it, but also in the practice itself.

(I found myself going the other way too: Davidson starts with examples where people piss on convention from a great height, and then in his theory ...)

What's the relationship between "first/literal meaning" and "lexical meaning"? — Dawnstorm

I've been pretty loose about this too.

I expect that rigid structure to be some sort of ideal type of a structuring principle; something people use to both create utterances and compare other people's utterances to, and something that will on occasion fail: people make mistakes, people don't find the words to express what they want to say and approximate with the best words they can find (and on failure to communicate try alternate ways of expressing themselves)... and so on. — Dawnstorm

I think this is absolutely right, and it's a curious thing. The abstract model has three functions:

1. It is a scientific model a linguist might use to predict behavior.
2. It is physically instantiated within an agent and actually produces the behavior we observe.
3. It is an ideal that language users are aware of and use to guide their (perhaps few) conscious choices about linguistic behavior.

(Logic, decision theory, and game theory all have a similarly peculiar status.)

That the model is an ideal that a language user refers to, sometimes, in choosing what linguistic act to undertake, I find particularly interesting: to choose your words carefully is sometimes to imagine a sort of ideal speaker of your language, who would choose the correct words, and to choose the words they would choose. ("Imbue.")

It seems pretty clear though that (3) there, the ideal we're aware of, is not the same thing as (2), and where (1) fits in is unclear. The structure of (2) we are largely unaware of, so it's more likely that (3) is something else that (2) generates alongside linguistic behavior. And (3) can readily grow from simple correctness to the art of rhetoric.

But maybe it goes in the other direction! Maybe (2) is the habituation of practices first achieved through the conscious building up of (3) -- the usual System 2 to System 1 pipeline. It's just that there's probably, if the sorts of things Chomsky and Pinker say are right, some structure in (2) waiting to find out how it should arrange itself. That still leaves a lot of options for (3): vestiges of the early training, rationalized into a system; a system rationalized out of the behavior we currently engage in we know not why; a rationalized and simplified systematization of some of what (2) is up to passed along to consciousness because it seems to want it.

that there are no propositions is a proposition — Tristan L

Question begging.

Propertyhood is a property. — Tristan L

Question begging.

with what right can we anaphorically say that Bob is thinking the same thing as Alice? Only by accepting that both are thinking about one and the same abstract entity, right? — Tristan L

We can say that Bob is thinking the same thing as Alice if Alice is thinking it's going to rain and Bob is thinking it's going to rain. I don't need any more justification than that. If to say there is an abstract entity, what they are thinking, is only to say that we can consider what they are thinking as if it were an independent object, though it isn't, then I have no beef with abstract entities. (I'm happy to leave the argument between nominalism and Aristotelianism for another day.)

But I left off the first part of your point:

if there is no causal connection between Alice’s thinking and Bob’s thinking — Tristan L

You claim that there is a causal connection between them, and that this is because they are both causally connected to something, a proposition, that is "not-spatial, not-tidesome (not-temporal), not-physical, not-mindly, and onefold (simple)"?

If that's the case, I don't know what you mean by "causal".

From the pen of Joss Whedon:

Why should a passing theory be called a theory at all? For the sort of theory we have in mind is, in its formal structure, suited to be the theory for an entire language, even though its expected field of application is vanishingly small. The answer is that when a word or phrase temporarily or locally takes over the role of some other word or phrase (as treated in a prior theory, perhaps), the entire burden of that role, with all its implications for logical relations to other words, phrases, and sentences, must be carried along by the passing theory. Someone who grasps the fact that Mrs Malaprop means ‘epithet’ when she says ‘epitaph’ must give ‘epithet’ all the powers ‘epitaph’ has for many other people. Only a full recursive theory can do justice to these powers. These remarks do not depend on supposing Mrs Malaprop will always make this ‘mistake’; once is enough to summon up a passing theory assigning a new role to ‘epitaph’. — pp. 261 - 262

The role that "epithet" plays in the language: that's its lexical meaning. (See PI §43.) Not only is Davidson not rejecting lexical meaning, his whole argument depends upon it.

Is his argument a reductio of the idea that there are such roles? I think, rather, he claims that we cannot, with certainty, know in advance what word may, even temporarily, be called upon to play a given role, because someone might utter not the standard word but a different one. That doesn't change the fact that we will describe the role taken on by the uttered word anaphorically: 'epitaph' here means what 'epithet' means. Not only do I see no alternative to that being offered, I don't know what such an alternative could possibly be.

This passage is actually the fulcrum of the entire argument. When Davidson said

The widespread existence of malapropisms and their kin threatens the distinction, since here the intended meaning seems to take over from the standard meaning. — p. 252

this is where he was headed: the intended meaning (what 'epithet' means) takes over from the standard meaning (what 'epitaph' means). This is the point at which he substantiates "seems to" as "does", by his lights.

I'm not addressing here whether this is a reasonable account of malapropisms and their kin; my point is only that Davidson is nowhere giving up the idea of lexical meaning, only prying it apart from "what is conventional or established" (252).

But does he? How do we know what 'epithet' means and what 'epitaph' means?

*

A lot of the milk into town and it was pretty crestfallen when you want or go on the road. — Srap Tasmaner

According to you, this means nothing, and

reinforces the point that language does not rely on rules. — Banno

You want to explain that?

Outstanding. Thanks for calling this to my attention.

I thought I knew what was going on and it's so much worse.

Lovely. In what sense "convenient fiction" is a distortion, is exactly what I wanted to know. Thanks so much for the Aristotle lesson.