I don't think jumping to the Tao level is much of an answer — J

Yes, as per our PM conversation.

The Tao that can be named is not the eternal Tao... — T Clark

...says nothing. In explaining everything, the Tao explains nothing. There's still the work to do; we still carry water, gather wood. That's why this:

Everything in our minds is a blending of cognitive and non-cognitive states. — T Clark

contributes nothing.

...objective... — Red Sky

So hackneyed a term, given that no one seems to know what it means.

Are you looking for a mind-independent truth? But how could a judgement be mind-independent? Are you looking for something impartial or unbiased? But the whole point of attributing quality is to be partial and biased. So are you looking for values that exist despite opinion? What could that mean?

Are we left, then, with "intersubjective agreement", So that the bare fact that so many people eat McDonalds (or watch Star Wars) means it must be of a high quality?

Or is aesthetic judgement embedded in community and culture, tradition and workmanship, coherence and responsiveness; is it learned and communicable, an aspect of growth?

An activity rather than a thing.

A possible middle ground might be that there are no "entities" called reason and emotion, and that we can separate them only conceptually, not physically. If that's what sushi meant, I'd to hear more about the conceptual distinction. To what does it correspond? — J

The idea that we can seperate reason and emotion physically is surely a category error?

Hesperus and Phosphorus rigidly designate Venus. Two names for the same thing. Is the suggestion that reason and emotion are the same thing?

It would be interesting to see this filled out.

OK. But when I hear "There's a possible world in which P", I understand this to be equivalent to "It is possible that P". So far, I haven't identified any difference that matters in my world. Am I right? — Ludwig V

Yep. Not seeing the relevance.

Modal logic can be translated into FOL thus:
□A → ∀w′ (R(w, w′) → A(w′))
◇A → ∃w′ (R(w, w′) ∧ A(w′))
Where:
w and w′ are world variables,
R is the accessibility relation.

That is, necessarily A means that A is true in every accessible world, and possible A means that there is an accessible world in which A is true.

...the question where natural language sits in relation to the formal system is important. — Ludwig V

It's an important question. Note that there need be - indeed, there cannot be - a systematic relation between the two in the way there is between a system of syntax and a stipulated model for that syntax.

That's why the enterprise of the Tractatus couldn't work.

What we can do with formal logic is to show the coherence of some fragments of natural language.

Fair.

You may stipulate as you wish. It makes no nevermind.

My opine:

Seems to me that there is nothing that talk of qualia is about. In so far as talk of qualia is usable and useful, it is no different to talk of colours or tastes or what have you. In so far as something is added to the conversation by the addition of qualia, seems to me that Dennett is correct in showing that there is nothing here to see. — Banno

Deadnaming! :scream:

Yep.

I'm fairly well acquainted with some of the literature. My basic objection is that if they are private experiences then they are unavailable for discussion, and if they are available for discussion then they seem to be just what we ordinarily talk about using words like "red" and "loud".

So not of much use.

Given that I don't think the very notion of qualia can be made coherent, I oddly find myself agreeing with you, but for completely different reasons.

My laptop, even after all these years, still insists on quail over qualia.

That's the research?

There's a section on the method used to measure belief. If there is to be a critique of that idea, it ought start by explaining the process used.

Would that we had a neuroscientist on call.

a subject’s phenomenology can be mathematically formalised as a belief (i.e. a probability distribution) encoded by its internal states. The subject produces first person descriptions of phenomenology that can then be used to infer its lived experience... Bayesian mechanics affords a correspondence between internal dynamics and belief dynamics. This furnishes a generative passage if we assume that phenomenological content can be formalised as a belief — p.14

So there's the usual Bayesian analysis as a stand-in for belief. All sorts of things wrong with that, and foremost the presumed equivalence to which Way points.

But the notion of qualia being used - the word only appears in a footnote - remains obscure.

emotional thought — I like sushi

Not too sure what that is.

Why should we give the last word on this to neuroscience?

Lots in that.

So if something... some statement, be it form Descartes, Kant or Wittgenstein, is indubitable, will we count it as a belief? Seem to me we do. Should we? I'd have supposed that the statements of which we are certain form a subset of the statements which we believe. Am I mistaken?

You father seems an eminently sensible fellow. If being in an upper room causes anxiety, it would not be conducive to a good night's sleep. It would be irrational to do so.

I had hoped to keep the god bothering at bay.

Everything that is consciousness is directedness. Ergo, there is always emotional content. What we feel is driven and what is driven is felt. — I like sushi

Does my air conditioner fit that, then? It intends to keep the room at 22ºc. Does it feel satisfied when it achieves it's goal, and frustrated by the frost?

A hackneyed argument to be sure, but it carries some import.

Perhaps we might avoid equating some brain state to "believing that..." at least until we have a clear way of setting out what a brain state is? No need to jump the gun.

Can any one provide a link to the research?

So what are the "qualia" that were measured?

That is too strong. — hypericin

Yeah, agreed.

So can we always seperate out the affective and cognitive aspects of a belief? Is there a method, rule or algorithm that does this for us? I'm thinking not.

Checking if I've understood... are you suggesting that belief is a propositional attitude, and that we also have affective attitudes, and that these are unrelated?

You and he are talking about completely different aspects of emotion and reason than I am. — T Clark

Yep. That'll be 'cause we're on the topic. And read more than just the abstract.

There's no conceptual work to do here?

But what of the issues raised in and ?

And in the cited articles? Or the SEP articles on belief and emotion?

If Bob has an irrational fear of snakes, does he believe snakes are dangerous? — frank

If his fear is irrational - he refuses to touch a Child's Python, perhaps - despite knowing that he will not be hurt - then isn't he is afraid, but does not believe the snake to be dangerous?

Thanks for this reply.

You ask if all beliefs contain doubt. The obvious counterexample is the hinge beliefs @Sam26 has urged on us, and with which I mostly agree. Do we say that, because these are undoubted, they are not genuine beliefs? Or do we separate the cognitive view that such hinges are indubitable from the connotative view that nevertheless, I might be wrong...

So if someone does not doubt that 2+2 is 4, do we discount this as a belief becasue it is indubitable?

How do we represent that Balthasar agrees the glass skywalk is safe but refuses to walk on it? If we say that since he holds that "the skywalk is safe", that he believes the skywalk is safe? Or do we say instead that because he will not act on what he holds to be true, that he doesn't really believe?

Charu's decision, let us supose, is against the odds - a bookmaker would say the lover will stray again. But Charu wants the bookmaker to be wrong, and so apparently acts irrational. Except that there is no possibility of the bookmaker being wrong if Charu does not trust her lover. Given her desire to stay with her lover, the decision to trust is rational.

David's belief is not to be subjected to doubt. What are we to say here - again, that it's not a proper belief becasue it is indubitable? Or is it, as is so often supposed, the very epitome of belief precisely becasue it is undoubted despite the evidence?

Plenty of material here, plenty to consider.

I recall Steven Pinker stating that we justify beliefs using reason, but we form them based on our affective relationships with the world. — Tom Storm

Yeah, pretty much. The belief is prior to the argument. But are we amenable to rational persuasion with regard to our beliefs? And to what extent? Should a mental state that is not amenable to persuasion based on evidence or justification properly be called a belief? That's the direction this discussion might go.

This is a fact rather than an idea. — I like sushi

Everything in our minds is a blending of cognitive and non-cognitive states. — T Clark

As says, how do we know? Let's aim not to make pronouncements but to map out the territory - what part of belief is cognitive, what is connotative, and how do they relate?

...we consider Reason (including logic, cognition and representation) as inseparable aspects of the same phenomenon, and not as separable states — Joshs

The obverse and reverse sides of a coin are inseparable, but that does not prevent us considering them separately as required. we might map how they relate and how they differ. We take the blanket statements and map out the where or how.

"Being referred to by a name is not part of the logical property structure -- it belongs to the semantic interpretation." — J

More accurately, being a name is not part of the syntax, but part of the semantics, of the model being used. The model may be formal, or it may be part of a natural language. Modelling is a part of logic.

I'll have to take it for granted that there are no other successful interpretations. — Ludwig V

A good point. If there were alternate explanation, what might they look like?

There's a generalisation of possible world semantics that turns out to be predicate calculus. That is, what can be expressed in modal logic can be expressed just in predicate calculus, after Montague and van Benthem. Strictly, modal logic is not an extension of predicate logic, but a fragment of it.

So if there are other ways of formalising modal logic, if they are as strong as possible world semantics, they will still be a subset of predicate calculus.

Roughly, any other alternative interpretation would be equivalent to possible world semantics.

...an apparent asymmetry between referring to something and being referred to by something. — Ludwig V

Yep. There's a difference between using a name to refer and inaugurating a name. Two different speech acts. Referring to something is a different sort of activity to making it that some word is used to refer.

The inauguration of a name involves a status function, one way or another. The word comes to "count as" the thing.

So now the question whether the concept of a rigid designator is part of the formal system and must be assessed in that context, or part of natural language and assessed in that context. — Ludwig V

My inclination is to think about the interpretation as something we do with the syntax, as an activity. If that's on track, then we use a name as a rigid designator. The name is a string of letters until used as a rigid designator. The string of letters is a part of the syntax of the system, the use as a rigid designator, to name something, is a part of the semantics of the system.

I somewhat regret suggesting a third level, since the gap between a formal modal and a natural language is no where near at the level of the gap between a syntax and a semantics.

We don't always find it easy to have a clear definition yes. — Jack2848

That's not the argument. The argument is that there is no "clear definition".

With the exception of merely stipulated technical terms, the way we use words precedes any definition. The word is used before it is defined. It follows that definitions are post hoc, with all the issues that involves.

Certainly this is true of "know", with it's etymology going back at least to the PIE root *gno- "to know."

Justified true belief was never going to be the whole of how we use "knowledge".

And even worse for knowledge as true information.

Excellent essay.

The usual suspects are here, bending over backwards to pretend that it doesn't apply to them. What a sad lot they are... the self-made man has a fool for his creator.

Well, you are not the first to present this sort of account, a scientists setting to rights the poor benighted philosophers.

It's easy to provide an answer when you haven't understood the issue.

Hang around for a bit, see if you notice anything odd or problematic in what you've decided. Then we might have an interesting chat.

Ok. Since you have it all worked out, I'll leave you to it.

, , , ...so one might find oneself at pains to demonstrate something that is not at issue.

The first argument:

I don't think dualism is an assumption. It's a description of a state of affairs. — karl stone

A description. So you are saying that it's an empirical observation? What is it we are observing here...

I understand what it means to observe the sky, or the horizon, or the sound of the sea. I'm not sure of what it could mean to observe internal and external worlds. I see the sky with my eyes, hear the sea with my ears - what sense do I use to observe my own mind? And who is it that is doing the observing of that mind, if not my mind...?

That all seems very odd. A long stretch.

The second argument:

Senses that are evolved to enable us to survive; and thus, demonstrably accurate to external reality. — karl stone

I'm not at all keen on Donald Hoffman, a chap with whom you have some points in common, but one point he makes is that there need be no relation between what the evolved mind presents to us and what is "out there". Indeed, his conclusion is quite the opposite. Evolution selects not for veridical perceptions, but for fitness-enhancing ones — and these two are not only distinct but often incompatible. We cannot assume that perceptual accuracy correlates with survival success.

When one decides on one's enemy - subjectivism, perhaps, whatever that is - one tends to see them everywhere. One might find oneself criticising an argument that hasn't been presented.

Isn't it rather that in order to make an observation at all, you become an observer seperate from what you are observing?

banno If I may chime in with a related question. I seem to be missing an understanding about what a property is. I can see that whatever name that object (i.e. the eiffel tower) has is distinct from any of the properties that it has qua physical object. But if "a" refers to a, then it seems to me to follow that a is referred to by "a". So if the former is a property of "a" then the latter must be a property of a? You and J seem to agree that "a" does have a property of referring to a, but that although a is referred to by "a", that does not constitute a property. I don't see why not. I do see that one could not use it to identity the reference of "a", because it would be circular, but that's a different matter, isn't it? — Ludwig V

So what is a property? Fair question. Logicians use "property" in a slightly different way to Aristotle.

What is "'a' refers to a" about? At a naive level we might suppose that since the sentence predicates "'a' refers to..." to a, then a is an individual with the property of being referred to as 'a'. It's not much of a property. Aristotle and others added various ad hoc ideas here in an attempt to make it work. Formal logic does something quite different.

A property in extensional formal logic is just a set of things. So the property of being red is just the grouping of individuals that are red, the property of being in Australia is just the things that are in Australia, and that's all. The property of being red = the set of all red things. This captures only which objects fall under a predicate, not why or how.

The predicate “is referred to by ‘a’” defines a singleton set: {a}, so in this stripped-down view, it could be called a property — because a is in the set. However, singleton sets carry no information or weight. It's trivial, it's not discoverable by empirical means, it's neither intrinsic or necessary that "a" refer to a.

Do we count it as a property, then?

The point made earlier, in amongst a bunch of other stuff, is that in predicate calculus, properties are thought of as represented by the letters f,g,h... and so on. So f(a) says that a has the property f. Such systems are given an interpretation by assigning individuals to the individual variables.

We have the list of predicate variables, "f","g","h"; and the list of individual variables, "a","b","c"...

We assign a property, say f, to some individual, say a, by writing f(a).

We assign an interpretation to this syntax by assigning an individual to each of the individual variables, a to "a", b to "b", and so on.

So, assigning a property to an individual happens in a different part of the logic to assigning a name to an individual.

For these two reasons, having a name is not usually considered as having the property of having that name. Being referred to by a name is not part of the logical property structure — it belongs to the semantic interpretation.

So I still want a way to characterize the difference between saying "The Eiffel Tower is tall" and "That object [pointing] is called 'the Eiffel Tower'". Yes, the first is a property and the second is not, but where do these statements fall on the syntactic/semantic spectrum? — J

Seems fair to say that in natural languages the distinction between semantic and syntactic is fluid, far more so than in a formal language.

As I understand it, Kaplan's approach is to explain the character and the content of an indexical separately. The character is a function that in a given context yields the object being pointed to. So the character of "I" yields me; the character of "that" yields what is being pointed to, and so on for different indexicals. The content is the individual involved.

So “that object [pointing]” has an indexical with character λc. demonstratum(c), and content (in c) = Eiffel Tower. There's plenty more formalism that can be dropped in here, but the idea is basically that each different indexical has a character that returns an individual as the content. "I" returns the speaker, "that" returns the thing being pointed to, "you" returns the person being addressed, and so on,

The content of "that" is a rigid designator, and so in predicate and modal logic can be an individual variable.

This is all semantics. It's about the things, not about the strings.

So "That is tall", indicating the Eiffel tower, and "The Eiffel tower is tall", are about the same thing, referring to it using a rigid designator.

But "That object [pointing] is called 'the Eiffel Tower'" has an indexical that returns a rigid designation that is then equated to a name -- hence 'the Eiffel Tower' is in quotes. Being in quotes indicates that it is part of the metalanguage, that it's about the interpretation of the language and not a sentence int he language.

Is there something here that this misses?

Oh, boy. there's a lot in that post.

Relevant logic uses a variation of the accessibility relation from possible world semantics to model relevance relations between worlds. Together with Kripke's strict conditional this is used to group conditionals together in a way that seeks to overcome the so-called paradoxes of implication.

I'd have to do a bit more work before claiming to see how it works in the examples you give. My thought at the time I made the suggestion was that perhaps the relevance relation might be used to clarify the difference between, say, "heat is molecular motion" and "temperature is molecular motion"; that temperature might be relevant to molecular motion in a way that heat is not. it'd be something like that from a world in which we have temperature, we might only be able to access worlds in which we have molecular motion, but from a world in which we have heat, it might be that we can access a world without molecular motion. It might follow that temperature, and not heat, relates necessarily to molecular motion. I do not have sufficient grasp of the machinery of relevance logic to follow through on this.

What this shows is that we've been ignoring a bit of the theory of possible worlds that is becoming increasingly relevant here. We've been saying that ☐A is understood as "A is true in every possible world". This needs some qualification. These statements are indexed to possible worlds and to the accessibility relation. So a more accurate account is that ☐A in world one is understood as "A is true in every world that is accessible from world one".

We get away with the cut down version by assuming that we are working with the syntax of S5, in which every world is accessible from every other world.

And there's another qualification that is needed, as to the difference change in the use of "contingent" in Kripke's semantics. Previously, "contingent" had a sense of "dependent", so something was contingent if it's being true was dependent on some other fact. That is not so much the case in possible world semantics. A statement will be necessarily true in w₁, as explained, if it is true in every world accessible form w₁. A statement will be possible in w₁ if there is at least one world, accessible from w₁, in which the statement is true. It follows that if a statement is necessary then it is possible. A statement will be impossible in w₁ if there is no world accessible from w₁ in which it is true.

And a statement will be conditional in w₁ if there is at least one accessible world in which it is true, and at least one accessible world were it is false.

The dependence on "dependence" drops out, along with a whole lot of metaphysical baggage and a few proofs that god exists. That's part of the reason there is some resistance against this logic from those of a naive theological bent.

So to the metre rule, and area that is fraught with misunderstanding. Consider this, from your Malcolm quote: "Certainly there is no requirement to hold that if the sentence is not a contingent statement then it must be a necessary statement."

This might be so if we think of contingency as dependence. The length S is dependent on the length of the rod, in the base example.

But in a possible world semantics, if a statement is true, and not contingent, it follows immediately that it is necessarily true. Malcolm, from Kripke's perspective, is mistaken: there is a requirement to hold that if the sentence is not a contingent statement then it must be a necessary statement.


And to god. So for Malcolm, the concept of God, like seeing a material thing, is not inherently self-contradictory. While specific reasoning may be invalid, a general demonstration of non-contradiction is not possible. Both concepts are integral to human thought and life.

See my comments in for a bit more on Anselm. While we might not agree that the idea of a something a greater than which cannot be conceived is self-contradictory, it's not clear that it can be made coherent, either. There is the problem of how to deal with a necessary being without the consequence of modal collapse.

But that's enough for now. Thanks again for your posts.

He's trying to lay out the requirements for a consistent picture, not choosing among pictures — J

That's the idea.

It's not that we can't, or even shouldn't, choose amongst the metaphysical theories. It's that if some metaphysics is not in harmony with the best thinking on modal logic or logic more generally, we ought treat it with scepticism; it has issues to be addressed.

He did at one stage describe himself as an anti-realist, but that seemed to be a result of his toying with truth, a seperate issue to possible worlds.

The women in this party are so assertive now that we may need some special rules for men to get them pre-selected. — Alan Stockdale, to NSW Liberal’s Women’s Council

:roll:

,

Kripke: Identity and Necessity

Kripke asks if this lectern could have been made of ice. His answer is that it is entirely possible that the lectern before us is made of ice, but that if this were so it would be a different lectern. — Banno

I don't think I've bought into Cartesian assumptions — karl stone

I'll try to explain the assumption again. It's ubiquitous, and so can be difficult to see.

Descartes supposes that we have on the one hand, mind, and on the other, the stuff of the world; roughly the presumption is that of dualism, of a divide between what is subjective and what is objective. Immediately on making this supposition he, and we, are faced with the problem of how the stuff of the mind interacts with the stuff of the world. Descartes solution was god, your solution is observation.

But what if that supposition, that schism between thought and thing, were a mistake?

I'll leave that hanging. Thoughts?

This makes the issue much more precise, thanks. — J

Cheers.

The difference between syntax and semantics is very clear in formal logic. Less so in natural languages. The following is probably familiar.

Let's look briefly at propositional logic. It includes just letters, p,q, and so on, as many as you want, and a couple of symbols, usually v and ~. To this we add formation rules that tell us what we are allowed to write. First, we can write any letter by itself. So we can write "p", or we can write "q". Then, we are allowed to put a "~" in front of anything else we are allowed to write; so we can write "~p" and "~q" and so on. Then, for two things we can write, we can join them with a "v". So we can now write "~pvq". From this, we can set out a system that shows how some strings of letters are well-formed - they follow the formation rules - while others are not... if we follow those rules we can never write down "pvvq~", for example. (I've left out brackets just to keep things simple. Also, ^ and ⊃ can both be defined in terms of v and ~, so they are not needed here)

All we have here is a system of syntax. It is purely a set of rules for stringing letters together in a specific way. In particular, it tells us nothing of what "p" and "q" stand for, and so nothing of which of our strings of letters might be true or false.

We add an interpretation to this syntactic system by ascribing "T" or "F" to each of the letters, together with a rule for the truth functionality of "v" and "~". A string beginning with "~" will be T if and only if the stuff after the "~" is F, and a string joined by a "v" will be T if and only if the stuff on either side of the "v" is also T.

A useful way to understand this is that p,q, and so on denote either T or F. We've moved from syntax to semantics.

We can expand our syntactic system by allowing ourselves to write not just p's and q's, but also "f(a)" and 'g(b)" and so on, in the place of those p's and q's. We can add rules for using ∃(x), but still at the syntactic level - just setting out what is well-formed and what isn't. This gives us a bigger system.

And to that we add more interpretation, were the letters "a","b" and so on stand for a and b, respectively, and "f" stands for some group of such letters, perhaps "f" stands for {a,c,e} while "g" stands for {b,c} or whatever. We then get that f(a) is true - "a" is in the set {a,c,e}, while f(b) is false - "b" is not in the set {a,c,e}. Similar rules apply for interpreting the quantifiers. And this gives us predicate calculus.

We can then expand the system once more, adding the operator "☐" outside of all of the stuff in the syntax for predicate logic, together with a few rules for how we can write these. This gives us the systems S1 through S5. These are just ways of writing down strings of letters, with ever more complicated permutations.

In order to give a coherent interpretation to these systems, Kripke taught us to use possible world semantics. In a way all this amounts to is a process to group the predicates used previously. So we said earlier that "f" stands for {a,c,e}, and to this we now add that in different worlds, f can stand for different sets of individuals. So in w₀ "f" stands for {a,c,e}, while in w₁ f stands for {a,b}, and so on in whatever way we stipulate - w₀ being world zero, w₁ being world one, and so on. Now we have added a semantics to the syntax of S4 and S5.

So reference-fixing is giving an interpretation, yes? — J

Exactly.

We have two levels, if you like, for each system. At one level we just set out how the letters can be written out, what sequence is acceptable. That's the syntax. At the next level, we add an interpretation, what the letters stand for. That's the semantics. So for propositional logic, the letters stand for T or F, and for predicate logic, we add individuals a,b,c... and for modal logic we add worlds, w₁, w₂ and so on, in order to get out interpretation, our semantics.

And to this we might add a third level, where we seek to understand what we are doing in a natural language by applying these formal systems. So for propositional logic, we understand the p's and q's as standing for the sentences of our natural language, and T and F as True and False. For predicate logic, we understand a,b,c as standing for Fred Bloggs, the Eiffel tower and consumerism, or whatever. And in modal logic, we get Naming and Necessity, where we try to understand our talk of modal contexts in natural languages in terms of the formal system we have developed.


I left out brackets, truth tables, domains, and accessibility, amongst other things, and only scratched the surface of extensionality. But I hope I've made clear how clean the distinction is between syntax and semantics in formal systems.

If reality's nature is not such that eyes can give us valid information about it, then I would expect reality to have evolved some other system to do so. — Patterner

There's something deeply problematic about using evolution to explain away ontological problems. Evolution assumes a degree of realism in assuming that there is a deep past in which there are things that could evolve, so of course it is consistent with realism. But it would be a mistake to think that evolution demonstrates realism.

None of which is to say that evolution didn't occur. It's just not methodologically a good way to try to dispose of solipsism or idealism.

Now it seems to me that despite his protestations against Cartesianism, @karl stone is buying in to many of the assumptions that Descartes made. He wants to find firm foundations and build a system from those foundations, a very Cartesian method. Sure, instead of the cogito he wants to use perception as that foundation, but it isn't going all that well.

Of course I agree with @Sam26 that a response is found in a treatment of what it is to doubt, along the lines of Wittgenstein's discussion of hinge propositions, but unlike Sam I reject idealism, along with certain sorts of realism, as a false juxtaposition.

But the devil is in the detail, and the way forward is to keep struggling with the analysis.

I wonder how we make sense of such claims as "if I were you then ...." — Michael

@J gave a pretty clear account of this, don't you think? Together with @frank's account of how we identify an individual with their origin, which is the approach Kripke is arguing for in Naming and Necesity.

Seems to me you are correct that @karl stone hasn't succeeded in casting aside the sceptic.

I'm pleased to see so much analytic work going on. Working through the issues is the only way to work out how to fit all these pieces together - if that is possible.

This is core analytic philosophy - looking closely at how the terms involved are being used, comparing them with formal systems we know are consistent, seeing what works and what does not. Bread and butter stuff. It's hard conceptual work.