Jason Stanley on 'intuitions' of 'truth-conditions'

Which Ancient Greek philosopher was from New York? Pyrrho? Agrippa? The truth is all Ancient Greek philosophers were from New York in one sense or another.

Competence in speaking and appreciating English of a certain kind could be measured by auditing a discussion on whether some or all philosophers come from New York.

Jason Stanley takes a different view-

Consider an utterance of the sentence "Some philosophers are from New York". If no philosopher in the world comes from New York, competent speakers of English know that it is false.

Surely this is nonsense. A Prof. of English is likely to be a competent speaker of English. How is he supposed to know whether there any philosophers- as opposed to tenured shitheads- in that City?

They also know that this utterance is true, if six philosophers in the world come from New York. In other words, competent English speakers have clear intuitions about the conditions under which what is said by an utterance of this sentence is true or false.

We have no such 'intuitions'. We genuinely can't tell if a guy is a shithead or a philosopher. Even if we were supplied with all sorts of tests which we could administer and came to the conclusion that this particular shithead passes a Turing test for 'being a philosopher', it would not be the case that we had the intuition 'he is a philosopher'. Rather, what we have is some laborious type of reasoning.

The apparent source of such intuitions is not difficult to locate. Competent English speakers know the meanings of the words in the sentence "Some philosophers are from New York."

The words are ambiguous. Philosopher may mean shithead or pederast or be a snide reference to the Philosopher King who leads the Free World.  The original New York is in Lincolnshire. There are 13 others around the world. We don't know whether 'is from' is to be taken literally or metaphorically. Even the word 'some' is ambiguous. 'Some man you are!' means 'you are no man at all.'

Competent speakers can speak competently. But this does not give them magical powers. Nor does it make the subject of a form of enchantment.

They also know how to combine the meanings of each of the words in this sentence to arrive at what is said by the utterance of the sentence, "Some philosophers are from New York." It is that linguistic competence that seems to be the source of their ability to report correctly about the truth of what is said by that sentence relative to different possible circumstances, e.g. the circumstance in which there are no philosophers from New York, or the circumstance in which six philosophers come from New York. So, the explanation for our ability to report about the truth and falsity of what is said by an utterance of "Some philosophers are from New York" in various possible situations is

that we feel it expedient to have this ability. If it turns out to be a nuisance we will start shrugging our shoulders and saying 'me no capisce'.

as follows. Competent English speakers know the meanings of the words used, and understand how they are combined. Their grasp of the truth-conditions of the utterance of that sentence

does not exist unless it is expedient and reinforced by some appropriate reward or punishment

is due to their ability to combine the meanings of the words, relative to the context of utterance. With this explanation in mind, consider an utterance of the sentence "Every philosopher is from New York", made at a small philosophy conference. It is natural to take this utterance to say something that is true if and only if every philosopher at the conference is from New York.

Nonsense! It may be merely a chauvinistic remark or else a statement of a particular type of non-Dualism. It is not natural at all to take it for a quite different statement, viz. 'I see that every philosopher here is from New York.'

If we cleave to the model of understanding just described, we will seek to explain our understanding of the truth-conditions of this utterance by appeal to a process of combining the elements of the sentence "Every philosopher is from New York", using our understanding of the words used in the sentence. But of course, there appears to be no expression in the sentence "Every philosopher is from New York" that corresponds to the understood constituent expressed by "at this conference".

Stanley thinks it would be natural for a shithead at a small conference of shitheads to point out that the affair is wholly parochial. If she feels obliged to advert to this humiliating circumstance, surely she will preface the remark with an expression of lamentation before proceeding to expatiate upon her own spiritual link to places of fairer renown.

Similarly, suppose, pointing at a 5 foot tall seven year old child, I utter the sentence "He is tall." I am most naturally understood as saying something that is true if and only if the child in question is tall for a seven year old child.

No. You are most naturally understood as saying you wish it to be thought true that you think the person you are pointing to is tall.

Preserving the model of understanding we began with, according to which our intuitions about the truth-conditions of an utterance are due to a process of combining meanings of the parts of the sentence uttered, would require us to find some constituent in the sentence that could be taken to supply the understood property of being a seven year old child. But again, it appears that the sentence "He is tall" contains no such constituent. So, we have a predicament.

Jason fell into a pit he himself dug. The fact is natural sentences are not protocol bound, juristic, or alethic in nature. If a child is tall, there is little point saying he is tall save for some other purpose- for example to show admiration or, it may be, betray a mere absence of mind such as is befitting a King or a chatterbox.

If we look at certain sentences, there seems to be a clear and elegant explanation of why we have the intuitions we do about the truth conditions of utterances of those sentences.

Fuck would this shithead know about 'clear and elegant explanations'? The whole point about intuitions is that they need no such things. On the other hand counter-intuitive mathematical proofs or scientific equations may have elegant and illuminating explanations.

But if we consider utterances of other sentences, the explanation appears to break down. The first response to this predicament is to attempt to preserve the clear and elegant explanation in the face of the apparently recalcitrant data. The second is to abandon the clear and elegant explanation of the source of our truth-conditional intuitions in favor of a different one.

Why not simply avoid the predicament altogether? It is easily done. Don't say 'intution makes us think people truly said something they didn't actually say at all'. A guy who says 'all philosophers are from New York' isn't saying all the philosophers here are from New York. Nor is a guy who says 'that child is tall' actually saying 'that child is tall for his age'.

My concern with the second response to the predicament is that the suggestions I am aware of for dealing with the additional complexity essentially end up abandoning the project of giving a systematic explanation of the source of our intuitions. They invariably involve appeal to unconstrained and non-explanatory notions or processes (cf. Stanley (2002a)). I have therefore been inclined to pursue the first of these options (cf. Stanley (2000)). My purpose in this paper is to continue the project of defending the clear and elegant explanation of the source of our intuitions about the truth-conditions of utterances. I will do so by considering some replies to previous arguments in favor of it. I will argue that proponents of abandoning the clear and elegant explanation have not yet made their case.

 So this is an argument between fools as to who is capable of the greater folly.

Section I. The Challenge from Context-Sensitivity On the simple explanation of the source of our intuitions about the truth-conditions of utterances of sentences we understand, it is due primarily to a compositional process of interpretation.

The principle of compositionality fails for idiomatic natural language. For a protocol bound system of interpretation- e.g that of juristic hermeneutics- there will always be 'hard cases' and every attempt at 'harmonious construction' is going to seem forced or bizarre to some concerned party.

It may be that, there are ex post 'univalent foundations' which everybody would consider true and just, as Light dies and the owl of Minerva takes flight. But then, it may equally be that there is one proposition which logically entails every other true proposition. But, what would that have to do with us?

Our knowledge of meaning, together with our knowledge of relevant contextual facts, allows us to assign meanings to the parts of a sentence, and the intuitive truth-conditions of an utterance of that sentence are what results from combining these values. Somewhat tendentiously, I will call proponents of the simple explanation, semanticists. Innumerable researchers from pragmatics have challenged the semanticist's model. Here is the form of the standard challenge. First, a linguistic construction C is produced that appears intuitively to have a certain reading R. Secondly, the researcher claims that the readings cannot be due to the semantics of that construction. That is, the claim is that R cannot be due to the compositional semantic interpretation of C, relative to the envisaged context of use. The conclusion the researcher draws is that the assumption that the intuitive truth-conditions of a sentence relative to a context are due to semantics is incorrect. A large number of researchers opposed to the semanticist employ arguments of this sort

Why don't the semanticists make their theory bullet proof by appealing to a non deterministic decision process? The world has moved on since the time of Frege.  There's probably some co-evolved process which some 15 year old is simulating with an app for your phone which will look pretty much like an i-language type program.

So the question posed by such examples is how to establish that the intuitive readings of the problematic sentences in (1)-(10) are due to linguistically determined content. Section II. Responding to the Challenge I adopt the conception of semantics at work in Stanley (2000) and spelled out in detail in King and Stanley (2004). The semantic content of a sentence relative to a context is derived by taking the semantic contents of the parts of that sentence, relative to that context, and composing them in accord with the composition rules governing the syntactic structure of that sentence.

Why restrict semantic contents in this manner? Why not let it be whatever it 'ought' to be- using Nmatrices or whatever- to make the theory come right?

The semantic value of a basic constituent of a sentence is what is determined by speaker intentions together with features of the context, in accord with the standing meaning of that lexical item. Given this conception of semantics, the position of the semanticist is then that the source of our intuitions about the truth and falsity of utterances relative to various possible circumstances is due to semantics. When faced with the claim that a certain construction C has a reading R that prima facie does not seem traceable to the semantics, the semanticist has three options. The first option is to establish that the alleged reading is not part of the intuitive truth-conditions of an utterance of that sentence, but is instead due to the pragmatics (as in the discussion of (7), above).

If we evolved, then intuitions evolved. But the fitness landscapes dictates hypersensitivity to certain types of 'truth-conditions'- like a lion's roar- and not giving a shit about idle chatter. But this would scarcely extend to the propositions of a deterministic semantics. The regret minimizing course would be to leave oneself wriggle room rather than gain a Momus window into one's own soul.

The second option is to argue that the claim that reading R is not due to the semantics is due to an overly simplistic conception of the semantic content of some elements of C. When the correct semantics for the relevant expression is given, the reading does emerge from the semantics (cf. King and Stanley (2004, Section 5), on "implicature intrusion").

Why stop there? Why not say the correct semantics for the relevant expression is a 'slingshot' . So tightly is the web of predication drawn, to know one thing truly, as the Jains say, would be to know all things.

The third option is to argue that the claim that reading R is not to the semantics is due to an overly simplistic conception of the syntactic structure of C. In fact, C contains covert structure, and once this is recognized, reading R does emerge from the semantics (cf. Stanley and Szabo (2000) on domain restriction, and Stanley (2000) for discussion of other constructions). So, when faced with a list such as that given in the previous section, the semanticist has, in each case, three alternatives. The first is to reject the semantic significance of the data, the second is to give an alternative semantic assignment to some overt element, and the third is to argue for covert syntactic structure.

Alternatively, these guys could do something useful like make Alexa better using Nmatrices and neural networks and the sort of thing VeeCees used to throw shedloads of money at.
What is silly is to think one has to stick with a deterministic Research Program of an obsolete type.

As I have indicated, it is a construction specific matter which of these options is preferable. The difficulty facing the semanticist's opponent is that she must establish, for each case, that none of the three very different alternatives is available as an account of the data.

If you are an opponent of 'Semanticists', your best course is to say they are useless. Their research doesn't 'pay its way'. If it actually does pay its way, then you look a fool for opposing them.
Suppose some 'Semanticists' are doing something useful which 'pays for itself', then whatever novel methods they are using fall outside the scope of the 'three different alternatives' Jason believes he has discovered. His work is useless. It does not 'pay its way'.

 Conclusion. My purpose in this paper has been to defend the claim that the intuitive truth-conditions of an utterance are due to semantic interpretation.

This is a merely semantic claim. It has no 'intuitive 'truth-condition'. It can only arise in pragmatics as part of a useless methodenstreit. It does not 'pay its way'. These guys may as well spend their time saying 'coo!' and 'boo!'

Many of those who have objected to it have done so by saddling the position with absurd theoretical commitments, such as the position that all information conveyed in any discourse is due to the semantics.

This too is merely a matter of semantics. One could always say 'Semantics' means much more than you think it means- or indeed, what can be thought, it means.'

Part of my goal in this paper has been to explain what costs the semanticist's position incurs, and what costs it doesn't, by elucidating the target concept of semantics.

Costs should be computed, but only if Gains are more substantial. What matters is the net profit on the activity and whether this is more than any other activity the agents could be engaged in.

As I have argued, in certain cases (e.g. that of deferred reference), the semanticist must make decisions about the defiendum that are informed by theoretical considerations. But this is the ordinary practice even in the human sciences. 

But 'human sciences' either pay for themselves or are declared a waste of money and type of fraud. That is why the State funds a lot of research on Economics and Psychology but none on Astrology or Alchemy.

It is true that Philosophy still gets a little money. Jason thinks he repays this subvention by foaming at the mouth about Fascism. Perhaps he is right. But a cheaper alternative would be to publish the equally paranoid ravings of genuine psychotics who happily do the job for free.