Thursday 25 May 2023

Jason Stanley's context-independent cretinism.

Why did Jason Stanley turn into a virulently woke cretin?  Was it because of sinister forces operating on Campus? That may be part of the answer. However, bad logic- the 'intensional fallacy' in particular- had already vitiated the entire project of Analytical Philosophy. 

The fact is, when using a word whose meaning diverges- on the basis of our knowledge level or 'epistemic state', from all that it might be said to designate- claims re. sufficiency & necessity are vitiated for two different reasons. In the case of sufficiency, correlation or 'Granger causality' may provide an empirical or inductive justification for the claim. In the case of necessity, the thing is impossible for a sentential operator, unless there is a Truth predicate in the first order language. But, in that case, we have taken leave of  Tarski and are in an Intuitionist or Gentzen type realm. Nothing is gained over the formalism where Necessity is a predicate because the thing is otiose or an example of 'modal collapse'. One may as well be an Occassionalist and say, 'God is the only efficient cause. All that is, necessarily is, coz that's what God wants.' This is fine and dandy coz we know that nobody knows the mind of God. There is some spiritual humility to Occassionalism. But, if this essentially theological argument is used by atheists- you get a paranoid, totalitarian, ideology whose proponent gradually convinces herself that she and she alone has the  power of 'strong Kleene evaluation' or (δοκιμασία‎) dokimasiai, which determines eligibility for every class or category whatsoever.

If you believe this crazy shit, you will soon start thinking you can tell that all sorts of non-Fascists are actually Fascists and that they are probably setting up Gas Chambers all over the place even as we speak.

As a case in point, some 20 years ago, Jason Stanley wrote

Contextualism in epistemology is the doctrine that

the predicate 'knows' is context sensitive. The alternative is that 'knows' is either necessarily true or necessarily false regardless of context. But, for the reason given above, this can only be the case if there is 'modal collapse' in which case everything is necessarily necessarily necessarily true unless it is necessarily necessarily false. 

Alternatively, truth can be a first order predicate in a Gentzen type or other 'natural deduction system' with 'witnesses' or 'zero-knowledge' proofs (verifications). 

But any 'univalent foundations' or category theoretical approach will have this feature. The problem is that either 'knows' is 'primitive'- i.e. undefined rather than context sensitive- or else there is an infinite regress and so nobody 'knows' what contextualism really is. 

the proposition expressed by a knowledge attribution relative to a context is determined in part by the standards of justification salient in that context.

This is not contextualism. It is determinism of some sort. But 'standards of justification' only tell us what is justified not what is known. We know plenty of stuff which we wouldn't dream of going to the trouble and expense of justifying. Stanley is illicitly trying to replace 'contextualism'- which is fairly common sensical- with some absurd type of determinism whereby what can be justified magically determines what can be known. This means Jason can get to pretend that protocols of justification determine what is knowable. Thus if the law changes such that a particular type of evidence is no longer admissible, suddenly a rapist will no longer know he was raped by such and such Spice Girl despite having found a lot of evidence on Pornhub that something of that sort might well have happened.

The (non-skeptical) contextualist allows that in some context c, a speaker may truly attribute knowledge at a time of a proposition p to Hannah, despite her possession of only weak inductive evidence for the truth of that proposition.

No. A contextualist would not say 'truly'. They would qualify their assertion in some suitable manner- e.g. saying 'in a particular context or 'for a particular purpose', a speaker may consistently, or un-controversially, make such and such attribution.  

Relative to another context, someone may make the very same knowledge attribution to Hannah, yet be speaking falsely, because the epistemic standards in that context are higher.

Or lower or non-existent. In some context, every statement is true or false or meaningless or not a statement at all.  

Still, Jason is admitting that 'knowledge attribution' is 'intesional' and 'epistemic'- i.e. we don't know the 'extension' of the term because knowing is about discovering more about that extension. True, such knowledge may be arbitrarily 'buck stopped' by some particular system of protocols or else, be conjectured to have 'naturality' for some particular purpose.  But outside that specific context, 'knowledge attribution' does not obey Liebniz's law of identity. We can't perform any logical operation on it. To do so would be to commit the 'masked man' or 'intensional' fallacy. 

The reason this is possible, according to the contextualist, is that the two knowledge attributions express different propositions.

No. A contextualist is not committed to a belief that propositions are atomic or that they change when contexts change. Nor do they necessarily think justification matters at all save contextually and for some specific purpose.  It is a different matter that some person Jason considers a 'contextualist' believes, according to Jason, stupid shit. But this is an arbitrary assertion. 

The main advocates of contextualism

are not knowable till the end of time.  

have used the semantic doctrine of contextualism

which is that there is no sematic doctrine for contextualism. The thing may be pure pragmatics. 

to defend a certain response to the problem of skepticism. According to it, the force of the skeptical paradoxes is due to

our reluctance  to slap skeptics silly before telling them they imagined the whole thing and thus should not go crying to Mummy.  

the presence of unrecognized context-sensitivity in the language.

or our recognition that stupid people talk bollocks.  

When we are not discussing with skeptics, many of our ordinary knowledge attributions are true, because what is there at issue is a less demanding sense of knowledge.

No. What is at issue is utility. Pragmatics is about doing useful things.  

But the consequence of engagement with the skeptic is that the content of knowledge attributions shifts in a manner that is not recognized by the interlocutors.

Only if they are as stupid as shit. It is obvious that some sort of type theory applies. It just may not be worth our while to find out which one does.  

In such a “skeptical” context, knowledge attributions that may previously have expressed truths now express falsehoods.

Nope. They were propositions and propositions they remained. The predicate applied to the proposition may change from- 'true and verified as true', to 'true but not verified to be true', or 'probably true' etc, etc.  

The contextualist therefore seeks to explain the force of skeptical arguments by appeal to a feature of ordinary language.

Only if the contextualist is as stupid as shit or is being paid to talk bollocks. 

It is because knowledge-attributions are context-sensitive that we are fooled by skeptics into thinking that even in non-skeptical circumstances, many of our knowledge attributions are false.

No. There is no necessary connection between Knowledge and Justification or Proof or anything else. Why accept another's arbitrary stipulation when you can suggest that they only make it because they habitually devour dog shit?

Jason considers the following scenario created by Keith DeRose which does not deal with knowledge claims but rather with expectations.

Hannah and her husband are driving home on a Friday afternoon. They plan to stop at the bank on the way home to deposit their paychecks. But as they drive past the bank, they notice that the lines inside are very long, as they often are on Friday afternoons. Thinking that it isn’t very important that their paychecks are deposited right away, Hannah says “I know the bank will be open tomorrow, since I was there just two weeks ago on Saturday morning. So we can deposit them tomorrow morning.” But then Hannah’s husband reminds her that a very important bill comes due on Monday, and that they have to have enough money in our account to cover it. He says, “Banks do change their hours. Are you certain that’s not what is going to happen tomorrow?” Hannah concedes, uttering “I guess I don’t really know that the bank will be open tomorrow.”

Nobody does. The world may end tonight. Hannah's confidence in her expectations has decreased for a 'regret minimizing' reason. This is a story about expectations and risk aversion.  

This sort of example supports contextualism about knowledge ascriptions,

there are no 'knowledge ascriptions' here. It is obvious that we are merely speaking of expectations.  

because it suggests that the propositions expressed by one and the same knowledge-attribution may differ with respect to two different contexts of use, even though the two contexts are identical in all respects relevant for fixing the values of obvious indexicals. Here is the example in more detail. Consider a sentence like (1), as uttered by Hannah in the first situation: (1) I know that the bank is open tomorrow morning.

Nope. The sentence should be 'I know the bank was open last Saturday and all preceding Saturdays. I expect it will be open tomorrow'.  

Before she realizes the importance of having a bank account flush with resources by Monday, she utters (1). What she utters expresses a proposition that seems perfectly true.

No. We all know that the future hasn't yet happened and may not happen. We may expect something to happen but we don't know it will have happened.  

The proposition concerns a particular time, namely the next morning. She is then informed about the pressing need for a full bank account. She then utters: (2) I guess I don’t really know that the bank is open tomorrow morning. Again, it looks like Mary has expressed a proposition that seems perfectly true, one that concerns the same time as the proposition expressed by her previous utterance of (1). But (2) looks to be the denial of (1). If we take these intuitions at face-value, we obtain a contradiction.

No we don't. We know that the future hasn't happened yet. We don't know shit about it. But we do have expectations regarding it. For regret minimizing reasons we can alter our behavior even without any change in our 'Bayesian priors' simply because we are highly risk averse with respect to catastrophic outcomes. In this particular case, Hannah is also aware that the are all sorts of other reason why, even if the Bank is open tomorrow, she won't be able to deposit the check. Ultimately, if the couple drive back and wait in line at the Bank, it is because they want to avoid the discomfort of an anxious night.

Here are the two basic options one has to respond to this problem: (a) One can reject the semantic significance of one of the two intuitions. For example, one could deny semantic significance to the intuition that the proposition semantically expressed by Hannah’s utterance of (1) is true.

Words have been forced into Hannah's mouth. She is more likely to have said 'Hey, we can bank our checks tomorrow. I remember doing so on a Saturday morning.' She wouldn't have made a knowledge claim coz the rejoinder would have been 'who has seen tomorrow?' or words to that effect.  

 For the contextualist, then, knowledge relations come in higher or lower “strengths.”

No. Confidence in knowledge claims may have this property. Knowledge relations don't. 

Knowledge attributions are thus comparable to context-sensitive gradable adjectives, such as “tall” and “flat”.

But Epistemology is about knowledge attributions about knowledge attributions. The word fat isn't itself fat. We say 'x is fat' or 'x knows calculus' and that is fine if x is actually overweight and really knows calculus. There are grades of fatness and grades of knowledge of calculus. But there is no change in how fat relates to thin or 'knows' relates to 'ignorant of'. Thus knowledge relations aren't gradable. But some predicates have that feature. 

An attribution of tallness is sensitive to a contextually salient scale of height, as is an attribution to flatness. If what is at issue are basketball players, then that brings in one rather high standard for “tall”; if what is at issue are fifth-graders, then that brings in a considerably lower standard for “tall”. In this sense, one could speak of tallness relations coming in higher or lower “strengths” as well.

Nonsense! We don't compare fifth graders to basket ball players. A tall basket ball player stands in the same relation to a short baseball player as a tall fifth grader to a short fifth grader. 

IS “KNOW” GRADABLE? Contextualists typically tell us, when introducing the thesis, that it wouldn’t be at all surprising if predicates such as “knows that Bush is president” turned out to be context-sensitive in the ways they describe.

Sure. An old guy who thinks Bush Snr. is the current POTUS might be suffering from dementia.  

After all, we are told,

by homeless people- right?  

many natural language predicates are context-sensitive.

Coz that's how language works. 

As Stewart Cohen (1999, p. 60) writes: Many, if not most, predicates in natural language are such that the truth-value of sentences containing them depends on contextually determined standards, e.g. ‘flat’, ‘bald’, ‘rich’, ‘happy’, ‘sad’. ... These are all predicates that can be satisfied to varying degrees

not in natural language. There has to be some protocol bound discourse aiming at some specific objective for 'satisfiability' to arise. 

and that can also be satisfied simpliciter.

Nope.  Acceptation is not the same thing as satisfiability. 

So, e.g., we can talk about one surface being flatter than another and we can talk about a surface being flat simpliciter.

Nothing is perfectly flat. Einstein bent everything.  

For predicates of this kind, context will determine the degree to which the predicate must be satisfied in order for the predicate to apply simpliciter.

Nope. Context won't be enough. Knowing what coordination or discoordination game is going on is not enough to determine anything.  Knightian uncertainty is a feature of language. That's why semantics is shite. Still, the fact is we say 'I know Smith' if we've met Smith. We aren't saying we know much about him. Clearly that's 'gradeable'. But 'knowledge relations' aren't gradeable. Knowing Smith means saying Hi to Smith and Smith saying Hi to you. Knowing him well means something more- e.g. predicting what birthday present he'd like. 

So the context will determine how flat a surface must be in order to be flat.

No. We might ask 'how flat must a microchip be in order for it to simulate a particular type of anyon interaction?'. The context may be that of competitive enterprises in a high-tech branch of the engineering industry. But the answer might come from some arcane branch of String Theory.  On the other hand, it might turn out that the question is meaningless from the scientific point of view. Still the Marketing Department might latch on to it and run a campaign claiming that the superior flatness of the company's microchips means that it will very soon corner a lucrative market. This may boost the stock price. 

There is a great deal of evidence for Cohen’s claim that there is a kind of predicate, of which ‘flat’, ‘bald’, ‘rich’, “happy”, and “sad” may occur as a constituent, the semantics of which involve degrees or scales.

Only if there are specific purposes for which protocol bound metrics are established. But that is a utilitarian, not a semantic, consideration. Indeed, language may be wholly disintermediated from the validating process. A machine may reject microchips which aren't flat enough. Sad people may kill themselves without anybody ever describing them as sad. The fact may only become known at the inquest.  

The reason that contextualists appeal to such predicates is so that, given their frequency in the language, the claim that “know” is a predicate of this kind will be unsurprising.

These guys are academics. They set exams to discover how well their students 'know' the stupid shite they teach.  

How good is this prima facie case for contextualism about “know”?

It is as good as any other stupid shite these guys babble about.  

The predicates mentioned by Cohen – the “kind” of which he speaks in his second to last sentence, are not a disjunctive sort.

They can be disjunctive. In India we speak of 'tall leaders' who, generally speaking, are rather short and squat.  

They are gradable adjectives.

Anything can be gradable for some particular purpose. The Nicaraguan horcrux of my neighbor's cat is more clammy than its Guatemalan horcrux- which is actually quite surprising if you think about it.  

Most of the gradable adjectives listed by Cohen are (I believe) context-sensitive, such as “flat”, “tall”, and “rich”.

They may be interpreted as such. But they could also be seen to be a subpoenas issued by the DOJ to the Nicaraguan horcrux of my neighbor's cat.  

In talking about buildings, “is tall” may express a property that it doesn’t express when talking about people.

Or the reverse may be the case. 

Furthermore, this sub-class of gradable adjectives are context-sensitive in just the way that Cohen and DeRose claim that knowledge-ascriptions are.

An adjective is just a word. It isn't sensitive at all. It is a different matter that the word can be used with a degree of precision for some useful purpose. But, speaking generally, a loose, discretionary, 'economia' is better than rigid 'akreibia'.

According to Cohen and DeRose, knowledge-ascriptions come in varying degrees of strength.

No. A knowledge ascription is merely a bunch of words that is interpreted as such. There are people who say 'you know deep down that you want to suck me off. Why not just do so already? Must we play mind games?'  We don't consider them to be making 'knowledge-ascriptions'. They merely want a b.j. 

In other words, knowledge-ascriptions are intuitively gradable.

No. They are merely a bunch of words. Our intuition only comes into play where some purpose of our own is served.  

Contextualists speak, as their theory suggests, of higher and lower standards for knowledge.

But those standards are arbitrary assertions. The epistemology attached to a discipline is not itself part of that discipline.  It is stupid, arbitrary, shite.

Gradable adjectives are the model for gradable expressions.

Everything is gradable for some specific purpose. So what?  

It is therefore no surprise that epistemologists since Unger (1975, Chapter 2) and Lewis (1986) have been exploiting the analogy between “know” and context-sensitive gradable adjectives such as “flat” and “tall”.

Adjectives solve coordination games. They give information about Schelling focality. But hedging on discoordination games may also occur.  

But, as I will argue in this section, the attempt to treat “know” as a gradable expression fails.

Not if it is useful. Pragmatics matter. Semantics is stupid shite.  

This casts suspicion upon the contextualist semantics for knowledge ascriptions. First, it shows that one cannot appeal to the context-sensitivity of words like “tall”, “flat”, and “rich” to justify the context-dependence of knowledge-ascriptions.

But Jason just mentioned a couple of guys who do exactly that! I suppose he means 'one should not appeal to xyz'. But 'should not' does not mean 'can not'.  

Secondly, it casts doubt upon the claim that knowledge comes in varying degrees of strength, a core claim of contextualism.

Claims about knowledge aren't knowledge. They are stupid shit. The context is the utter cretinism that prevails in Jason's own branch of Psilosophy.  

There are two linguistic tests for gradability.

But grading can be done by machines! Jason & Co may have to hand out grades to cretins so as to get paid. Do they use 'linguistic tests'? Perhaps. But the test may involve asking for a b.j.  

First, if an expression is gradable, it should allow for modifiers.

Anything at all 'allows for modifiers'.  

For example, predicative uses of comparative adjectives allow for modification, as in: (5) a. That is very flat. b. That is really flat. c. John is very tall. d. John is really tall.

But the predicative use of meaningless terms also allows for modification- e.g. 'Uqwkqw is very tall' or 'tall is very uqwkqw'. 

Secondly, if an expression is gradable,

e.g  Uqwkqw which is less meaningful than Urquharty

it should be conceptually related to a natural comparative construction.

there are no 'natural comparative constructions'. Category theoretical 'naturality' is probably a chimera- we can't be sure. 

What obtain are focal solutions to coordination games. But it is sensible, or regret minimizing,  to hedge on a discoordination game.  

So, for “flat”, “tall”, and “small” we have “flatter than”, “taller than”, and “smaller than”.

just as we have more Urquharty or less Uqwkqwian. 

Both of these features are to be expected, if underlying the use of the relevant expression is a semantics involving degrees or intervals on a scale.

Semantics doesn't underlie shit. Utility does. For some specific purpose, the pragmatics of an expression- even one whose meaning is unknown or unassigned- may be protocol bound in a mertrizable manner. Thus an AI may look at the use of the word and come to formulate an 'intension' for it. It often happens that a term of disapprobation- e.g. 'Tory' which originally an Irish Catholic Rebel-  comes to have a positive meaning for those originally slandered by its use. 

For instance, the semantic effect of a modifier such as “very” on a word like “tall” is to increase the contextually salient degree on the scale of height that an object must exceed in order to satisfy the predicate.

Nonsense! We Indians may have said 'Kharge is a tall leader'. Now, since Kharge engineered a stunning victory in Karnataka, we say 'Kharge is a very tall leader. Sadly, Congress has a tall poppy syndrome. Who wants to bet the fellow won't be abandoned and ignored this time next year?' 

The claim that knowledge ascriptions are gradable fits elegantly into the contextualist explanation for, say, DeRose’s bank case discussed in the previous section.

What mattered in that case was ignorance ascriptions. Nobody knows what tomorrow will bring. Don't procrastinate.  

For in Hannah’s final utterance, she claims that she does not really know that the bank is

will be, not is 

open. It is natural to read “really” here as a degree modifier, as in the examples in (5).

Nobody really knows what will happen tomorrow. We have expectations, not knowledge.  

That is, it is natural to read this discourse as providing evidence for the gradeability of “know”. Over the course of the discourse, Hannah asserts that she knows that the bank is open,

will be open. Hannah is talking about the next day 

but also asserts that she doesn’t really know that the bank is open.

will be open. 

That is like someone asserting that Bill is tall, but conceding that Bill is not really tall.

No. It is like someone asserting that Bill is tall today. Tomorrow, he will be short because the person making the assertion is a maniac and plans to cut off Bill's legs tonight. 

But the explanation of the bank case that appeals to the gradability of knowledge is not correct.

Because we don't know the future. We have expectations about it.  

As the above example suggested,  negations of degree-modifier uses of “really” can be conjoined with assertions of the unmodified forms without inconsistency: (6) a. John is tall, but not really tall.

He is tall in one sense but not in another sense which is more 'real'.  

b. Michigan is flat, but not really flat.

because of the curvature of the earth. Also it does have some hills.  

In contrast, the same facts do not hold of the use of “really” when appended to a knowledge-ascription (“#” expresses oddity): (7) # If the bank is open, then John knows that the bank is open, but doesn’t really know that the bank is open. The sentences in (6) are perfectly natural. In contrast, (7) is very odd.

It may be or it may not be. It may be part of John's duties to verify that the Bank is open before sending his assistant with the day's cash receipts to the branch in question. Here 'really knows' means 'verified according to set protocols'. As a matter of fact, the Bank branch was open and so no great harm was done. Still, John may receive a rebuke from his Supervisor who warns him to be more scrupulous in future.  

This suggests that the “really” that occurs in the above description of the bank case is not a degree modifier.

Yet, there is an easily imaginable scenario where it was in fact a modifier.  

Indeed, prima facie, propositional knowledge ascriptions are not gradable.

The reverse is the case because of verification protocols. As a private individual, John is welcome to treat 'expectations' as 'knowledge'. But, his terms of employment, may require him to do verification in his official capacity.  

First, knowledge ascriptions do not seem to allow for modification: (8) a. *John very knows that penguins waddle.

This may not be grammatical but we understand that there has been an elision of 'well' after 'very'.  

b. *John knows very much that penguins waddle.

the elision is 'about penguins and thus knows'  

Second, there is no natural comparative conceptually related to “know”.

Yes there is. Knowledge and verification go together.  

The following locutions are deeply strained: (9) a. ??John knows that Bush is president more than Sally knows it.

But this is perfectly informative. John is passionate about American politics. Sally is currently off her head on ayahuasca and has been living in a commune in Brazil for the last twenty years.  

b. ??Hannah knows that Bush is president more than she knows that Clinton was president.

Coz Hannah is five years old. 

If the semantics of “know” did involve scales, it would be mystery why there wouldn’t be a comparative form of “know” available to exploit the scale.

but there was for the older, more inflected, Indo-European term from which it evolved. Some older folk still use terms like cannier or canniest. 

It has been noted before that ascriptions of propositional knowledge are not gradable (cf. Dretske, 1981, Chapter 5).

Yet Jason & Co spend a lot of time giving out grades on the basis of which student has more or less of the stuff.  

However, the data surrounding knowledge ascriptions is more complex than these prima facie considerations suggest.

No. Data is simpler than the inchoate notions with which we approach a subject. 

There are several constructions that suggest that knowledge ascriptions are, despite initial appearances, gradable. In the remainder of this section, I provide a complete case that knowledge ascriptions are not gradable,

but that case must be 'non-informative' or wholly imperative and arbitrary, otherwise it could itself be graded.  

and draw some morals for contextualism about “know”. One might think that knowledge-ascriptions are gradable on the basis of the obvious felicity of the following sort of construction: (10) a. John knows Bill better than Mary does. b. Hannah knows logic better than John does. But in the sentences in (10), “know” does not express a relation between a person and a proposition. These sentences are not propositional knowledge ascriptions; rather the occurrences of “know” in them express the acquaintance relation, what would be expressed in German by “kennen” rather than “wissen”.

This became 'ken' and 'wot' in older forms of English with the former relating to perception and the latter referring to understanding. But, in this case, neither is relevant.

There are a number of facts about Bill and there are a number of facts about Logic. John knows more facts about Bill while Hannah knows more facts about Logic than John. This is because John and Bill are sharing a prison cell. Hannah is a Professor of Logic and Epistemology. It may be that Hannah says 'I think Logic is completely empty. It has zero informativity. There is nothing in it to understand' whereas John thinks that logic can prove he was innocent of the murder for which he has been sent to jail. After all, Bill can testify that John was on a spaceship being anally probed by Aliens when the murder took place.  

It is only the gradeability of propositional knowledge ascriptions that is at issue in contextualism in epistemology.

If knowledge was 'extensional' and intrinsic to the subject of knowledge, we wouldn't need contextualism. It is right and proper to kill everybody from a particular minority because each and every one of them, regardless of any evidence to the contrary, is obliged by their detestable religion to kill and eat Christian children once every week. You may say, 'in the context of winning the war, why devote scarce resources to gas chambers?' This proves you are yourself descended from those evil cannibals.  

However, “know” can marginally occur with “very much” or less marginally with “very well”, as in: (11) a. ?I very much know that Bush is president. b. I know very well that Bush is president. But it is doubtful that these occurrences of “very much” and “very well” are genuine semantic modifiers of the knowing relation, rather than pragmatic indicators.

Pragmatic indicators can get baked into the semantics of the 'ideal speaker'.  

In this sense, these constructions are similar to: (12) 2 is very much an even number.

No. When it comes to such numbers, there is Kripkean 'buck stopping' and a well defined extension.  No number is more even than any other. Intensional objects have an epistemic component and can be graded on the basis of verification. 

Decisive evidence for this comes from several sources. First, note the unacceptability of negating the constructions in (13): (13) a. *I don’t know very much that Bush is president.

Again this is a case of elision. The person may be understood as saying 'I don't know very much about American Politics, Indeed, I'm not even sure Bush is the President. '  

b. *I don’t know very well that Bush is president. The unacceptability of the sentences in (13) contrasts with the naturalness of negating the verb phrase in a case in which “very much” is clearly modifying the verb: (14) I don’t like Bill very much.

Here 'very much' tones down the informational content. It makes you appear polite and civilized. It isn't that you dislike Bill. It's just that you don't lurve him as a Christian should.  

Secondly, “know” is only with great awkwardness combined with “very well” in non-assertoric speech acts. Contrast the sentences in (15) with (16): (15) a. ??Do you know very well that Bush is president?

I suppose there are some people in America who think Trump is the actual President. Biden is illegally squatting in the White House. People of this sort may know all sorts of things 'very well'. The mentally ill often do.  

b. *Do you know very much that Bush is president? (16) Do you like Bush very much? So, the sentences in (11) are clearly not cases where the degree of knowing is operated on by “very much” or “very well”.

There are people who know very much about how Biden stole the election from Trump who is the genuine POTUS.  

Defenders of contextualism might hold that “better than” rather than “more” is the comparative relevant to “know”, as in (17) Hannah knows better than anyone that she is poor. But here again, the construction means that Hannah is familiar with the fact more than anyone else – e.g. she lives with the consequences.

Not necessarily. Hannah has a PhD in Poverty Studies. She is currently working long hours for less than minimum wage in the hope of getting tenure. She actually knows better than anyone else that by every metric in current use, she is classed as 'poor as shit'.  

4 More importantly, “better than anyone” is idiomatic. For example, consider the oddity of: (18) a. ??Hannah knows better than three people that she is poor.

Those are the three people blocking the pay rise Hannah is demanding for people like herself. The irony is that they are all tenured Professors of Woke, Virtue Signaling, shite.  

b. *Hannah doesn’t know better than anyone that she is poor. So, “better than anyone” is an idiomatic construction, one from which we can infer little about the semantics of “know”.

The opposite is the case. We can infer that the semantics of know implies that it is gradable  

Furthermore, none of the non-philosopher informants I asked found the following acceptable, though they disagreed amongst themselves which was worst: (19) a. ??John knows that Bush is president better than Mary does. b. ??John knows that Bush is president better than Bill knows that Clinton is a Democrat. Furthermore, all of my informants reported a strong difference in acceptability between these sentences, on the one hand, and the perfectly acceptable: (20) a. John likes Bill more than Mary does. b. John likes Bill more than Mary likes John.

Physicists often ask non-physicist informants whether they think an anyon based String Theory would be better than one which features Cheese straws.  

So “better than” is not a natural way to express comparisons between levels of epistemic position with “know”.

It is perfectly natural to say 'I know x better than you'. Incidentally, philosophers are supposed to know philosophy better than non-philosophers. Then the discipline turned to shit. It became adversely selective.  

If the semantics of “know” did involve scales of epistemic strength, then there should be uncontroversially non-idiomatic comparisons and modifications.

Schools and Universities grade students on the basis of 'epistemic strength'. The thing isn't 'idiomatic'. It is nomothetic and protocol bound- though absurd in the case of worthless disciplines.  

6 One might think that these facts about ‘know’ have syntactic rather than semantic explanations. Perhaps sentences like (8) and (9) and those in (13) are deviant because verbs that take sentential complements grammatically do not allow for comparisons or intensifiers.

Syntax is about protocol observance. But protocols may be non-linguistic. For a specific purpose, everything may be gradable.  

But consider “regret”, a factive verb in the same syntactic category as “know”: (21) a. Hannah very much regrets that she is unemployed.

There is a large technical literature on 'regret minimization'. Hannah, sacked from the University for being a TERF- she used a Feminist pronoun to refer to herself- shows more regret than any of her colleagues who were also sacked for the same offense. How do we know? The answer is that Hannah alone is undergoing gender reassignment surgery. 

b. Hannah doesn’t regret very much that she is unemployed.

Because she is part of a class-action suit which will soon see her rolling in moolah. Also not having to teach cretins is a big relief.  

c. Hannah regrets very much that she is unemployed. d. Hannah regrets that she is unemployed very much. Here, the degree of regret clearly seems to be modified by “very much”. Furthermore, “regret” easily allows for comparisons:

Coz that's how words work. 

(22) Hannah regrets that she is unemployed more than she regrets that she is unpopular.

Hannah knows she is unemployed more than she knows she is unpopular because she smells like shit. Could somebody please have a quiet word with her?  

This shows that the lack of straightforward comparatives or degree modifiers has nothing to do with the syntax, or even the factivity, of “know”.

The reverse is the case. I know more about this than you. I regret this more than you do. This hurts me more than it hurts you. 

There are syntactically similar expressions whose link to degrees and scales is far more plausible.

There is always some scenario where it is plausible that to say 'John is more urquharty than Bill' is meaningful and will get you a reputation for sagacious and incisive social observation. 

It is also worth mentioning that other expressions upon which one might be tempted to base the context-sensitivity of “know” are, unlike “know”, also gradable. So consider epistemic modals.
Epistemic modality is the kind of necessity and possibility that is determined by epistemic constraints. Sadly, no such constraints are necessary or possible. Epistemology isn't knowledge. It isn't informative. It is a waste of time. 
A contextualist might understandably wish to appeal to the apparent context-sensitivity of epistemic modals to justify the apparent context-sensitivity of knowledge-ascriptions. But one problem with this strategy is that epistemic modals, unlike knowledge-ascriptions, are intuitively gradable:

No. They are useless and stupid. There are no synthetic a priori judgments. Phenomenology is tomfoolery.  

(23) a. It is very likely that I will publish more papers on this topic.

This is not an 'epistemic modal'. It is a statement about a future action which is not constrained by any epistemic considerations. Most philosophy papers are shit. Whether a particular shithead publishes more worthless shite has to do whether a particular Academic Ponzi scheme can keep drawing in suckers. 

b. It is more possible that Hannah will become a philosopher than it is that she will become a mathematician.

Because philosophy is adversely selective.  Also the subject lends itself to virulent wokeness. 

So the analogy between epistemic modals and “know” is almost as strained as the analogy between context-sensitive gradable adjectives and “know”.

We know epistemic modals are shit. We know that adjectives are 'context sensitive'. That's what makes them useful. 

The evidence concerning gradeability is more complicated when one considers the deverbal adjective “known”.

No. What was known is like what is known and what will be known. New data can throw anything into doubt.  

But even here, there does not appear to be a good case for a semantics involving a scale of epistemic strength.

Yet, we say things like 'it was well known in ancient Greece that a Tyrant might manage things better than a Democracy would have done'. New data may prove us wrong. A trove of Celtic runes is discovered in Galicia. It turned out that Celtic spies had been closely monitoring public opinion in various Greek polities. It turns out that the texts on whose authority we relied were known to have been composed for money by desperate savants.  

This adjective, unlike its verbal relative, does give rise to comparisons and modifications. But they are not of the relevant sort. So, for example, consider: (24) a. That broccoli is low-fat is better known than that broccoli prevents cancer. b. That broccoli is low-fat is well known.  (24a) does not mean that there is more evidence that broccoli is low-fat than that broccoli prevents cancer; rather, it means that the fact that broccoli is low-fat is more widely known than the fact that broccoli prevents cancer.

It may mean precisely that. What matters is the context. If a professional scientist is speaking, we assume that hard and fast scientific evidence is being discussed. If the remark is made by some vegan Trustafarian at a dinner party we assume that what is being remarked on is the beliefs of that section of the population which gives credence to the thaumaturgic powers of Crystals. 

Similarly, (24b) means, not that there is a lot of evidence that broccoli is low-fat, but that it is widely known that broccoli is low- fat. Evidence for this hypothesis comes from the fact that while (25a) is perfectly acceptable, (25b) sounds quite odd: (25) a. That broccoli prevents the flu is well-known, but illunderstood. b. ?That broccoli prevents the flu is well-known, though few people know it.

In context, we can repair the above sentence to read 'well established scientifically though few are aware of the result' possibly because of the mischievous activities of Big Meat, not to mention the scoundrels who sell Cough Syrup.  

Furthermore, as Tamar Gendler has pointed out to me, instances of (26) are quite odd: (26) It is well known that p, and less well-known that q, but more people know that q than know that p.

In context, we can repair the sentence to read 'p is well established' but this truth has been kept from most people  

This data is explicable on the assumption that the only available reading for “well-known” is widely known.

We are welcome to supply any reading we like. Epistemic modals don't actually exist.  

So, while the data is more complex here, the adjectival relative of ‘know’, on the rare use of it where it expresses propositional knowledge, does not

a word is missing here. Is it 'appear'? 

to be an obvious candidate for analysis via degrees on a scale of epistemic strength.

But if the thing can be usefully done for some specific purpose, it will be done. Suppose some justiciable matter is involved. A Court may 'read in' just such a scale so as to ensure justice is done.  

Another potential source of evidence for the gradability of “know” comes from its use in certain kinds of embedded questions. Consider, for example: (27) a. John knows how to swim well. b. John knows how to ride a bicycle better than Mary does.9 c. Hannah knows where Texas is better than John does. It is quite plausible that these are attributions of propositional knowledge.

 If propositional knowledge is involved in swimming (though fish do it) or bicycling (though BoJo does it) then- sure. Hannah is a homing pigeon from Dallas. Release her anywhere in the continental United States and she takes the most direct route to Texas. John, who belongs to the Bush dynasty, is the junior senator of the Lone Star State. He thinks Texas is just north of the Yale Campus. 

10 If so, one might think that this suggests that knowledge is gradable after all.

For any specific purpose, it- like everything else- is gradable. Jason's stupid shite is stupider than Tim Willamson's stupid shite.  

However, in these cases, what is being compared are answers to questions.

But everything can be an answer to a question- at least that is the case with my farts.  

In each case, one person is said to have a better answer to a certain question than another; the answer Hannah has to the question “Where is Texas?” is better, or more complete, than the answer John has.

Is it more useful? That is what matters. We release Hannah the pigeon and very soon we see her flying off in a South Westerly direction. Texas is that-a-way. John Bush, meanwhile, is headed towards the Canadian border.  

So, embedded questions do not provide evidence for the gradeability of knowledge claims.

If the question is asked for some useful purpose, then such evidence arises on the basis of verification. Hannah the pigeon has got to Texas while John Bush has been arrested by the Mounties for being coked to the gills.  

I don't know if Jason even pretends to being doing analtickle philosophy any more. Still, we can see that his turn towards 'Virulent Wokeness' was based on a belief that knowledge is context independent, 'natural' or 'non arbitrary', and lodged 'in rebus' in things themselves. This meant that Jason could identify as Fascist all sorts of people who were no such thing purely on the basis of his being 'woke' or 'antifa'. 

No comments: