Since the world is all that is the case, it is also a collective story. That does not meant hat just anything goes. You will still burn your hand if you touch the boiling kettle. — Banno

Yes, this is what I've been trying elucidate, but perhaps you're right and my use of 'fiction' here has only confused matters. There's perhaps a better way to explain.

The idea here (for me) is that despite having to posit hidden states as part of our informational meta-thoery (see my post to Srap above), these states can still be proper objects of reference. "the kettle" doesn't refer to my model of a kettle, it refers to (in the informational model) the hidden state itself. It's like us all speculating what's in the room next door. the subject of our speculations isn't our speculations, the subject of our speculations is what's actually in the room next door. As such, the best way I can find of 'translating' an active inference model to talk of "kettles" is to say that "kettle" refers (when it refers at all, that is - not all uses are referenential, of course) to the hidden state we're modelling, the contents of the room we're speculating about.

The difference is that we have data which more or less coheres with certain speculations (still on the 'room' analogy here) such that the speculation "nothing is in the room" would be difficult to cohere with a lot of noise and shouting coming from it - ie you still burn your hand when you touch the kettle.

So, perhaps more to @Luke than yourself (I may otherwise be preaching to the choir...)...

Kettles and water and boiling might be describable as collective fictions, but the part that needs emphasising (that I ought perhaps to have emphasised) is that they're purposeful fictions. They're not fantasies like the Lord of the Rings, where anything goes, they've a job to do - that of unifying, to an extent, our individual (neural) models so that we can co-operate, and not constantly surprise one another. This hooks them in to the actual external states we exist within in a way that actual fiction need not worry itself with.

(if that's cleared anything up @Luke then you can thank Banno)

If we have modeled imperfectly some detail of the hidden states, but we never encounter evidence that would encourage us to update our model, were we wrong? — Srap Tasmaner

There's a muddle of temporal terms here that I can't make sense of, You say that we "never encounter evidence...", but then ask "were we...". 'Were' from what temporal vantage point? Your first 'never' seems to disallow any point of reflection from which we can look back and ask the question.

can our model be properly said to supervene upon the hidden states? That is, can there be a change in our model without a change in the ("underlying") hidden states?

If the answer is "no," if our model is not so tightly coupled to the hidden states as that, what is the source of that relative freedom? And if our model is then, to some undetermined degree, independent of the hidden states, what entitles us to describe changes to our model as updates rather than just changes, which could, for all we know, be arbitrary, or, if not arbitrary, free? — Srap Tasmaner

Interesting question. Change and chaos. Or, less prosaically - (1) our models are based on priors and priors are necessarily based on historical inferences which will be constrained by previous states of the environment, not the current one, hence a source of de-coupling, our environment changes, our models run behind that change; and (2) there's a lot of noise in the system, neuron firing can be random, so a lot of what neural modelling does, one of the main reasons for backward acting suppression, is to make sure that noise is not mistaken for data, but this system isn't perfect, so sometimes it is. Obviously, there are then magnifying effects of both of these since we do not passively receive data from the external states, but rather we actively harvest data (and even manipulate those states) to match our models, so we're going to act in such ways as to re-affirm the model predictions insofar as that it possible, even if those model predictions have been affected by nothing but noise.

There is nothing, it seems, that we can point to as "evidence" that is outside the model, not even surprise; surprise is not a fact, but part of our model of ourselves. — Srap Tasmaner

Yes, that's right. We could model our 'surprise' as errors in our models, or as part of the narrative, as evil demons, as the tweaking of the matrix by our cruel AI overlords... The whole of active inference is a scientific model, an activity we undertake to help explain phenomena, just like philosophy. Some explanations seem better than others, but the reason why they do so can only ever itself be just another such explanation.

What becomes questionable is the claim that the "map" is not the territory but only a map, and the positing of a "genuine" territory out there, somewhere, that the "map" we wander around in is a copy of. That will surely strike most residents of the "map" as an article of faith. Anything can count as evidence for it, and nothing can count as evidence for it. — Srap Tasmaner

You're focussing on material construction, which is not that the model of active inference is about. The model is about data. It simply says that given a self-organising network of data nodes that is greater than a mere ring of nodes, there will inevitably be nodes which are inside a Markov blanket relative to the nodes which don't constitute the system. That has nothing to do with the material constituents of the nodes, only their informational relationship with one another. Thus, as a self-organising system, we must, by definition, have internal states, and boundary states (and there must exist external states). Without these three states we cannot say that there is a system at all, we cannot define it from 'not-system' without defining a boundary and (as far as data is concerned) that boundary must be Markov boundary if the internal network is any more complex than a single ring of nodes.

So the only way we could be informationally connected to 'the world' without a Markov boundary is if we say that we are the world, one integrated system. I don't think anyone is going there...

I don't think it will quite do to answer that "data underdetermines theory." What "data" there is, is not just theory-laden; it is crushed under the weight of the theory it's carrying on its back. It could, for all we know, be 100% theory. — Srap Tasmaner

That data underdetermines theory is implied by defining a system, as is raw data. As above, the mere definition of an informational system implies a boundary and external states. Complexity beyond a mere ring implies boundary nodes. Once you have those two elements, it is a necessary fact of the model that external states must be inferred by internal states from the states of the boundary nodes. If you introduce any variable whatsoever (active harvesting, ergodic feedback mechanisms, noise...) then it is a necessary part of the model that this inference will be prone to error - underdetermined by the external states.

So...

(1) nothing entitles us to make any claim that there is such a true state, or to make any claim about how close our conception is to it, — Srap Tasmaner

I don't think that's true. I've outline above how the very act of defining a system leads us to the conclusion that there are external, internal and boundary states and that external states can only be inferred by internal states from boundary states.

We have a model that is, for all we know, 100% mistaken, and at the same time, for all we know, all there is and no model at all. — Srap Tasmaner

Yep. I'm not sure what improvement in certainty you're looking for beyond that which we can rationally argue for.

Before you post "pragmatism" and count that as a job well done, plan on explaining exactly how pragmatism answers any of the questions I asked, or shows the questions to be ill-conceived. — Srap Tasmaner

Hopefully avoided pragmatism in my answer... but do check.

Because we could never be surprised to find that Aragorn was not king of Gondor, or that "Aragorn was king of Gondor" is false. Surely we know our collective fiction (which is the model, which is the world) in exactly the same way, and with the same level of surety, that we know Aragorn was king of Gondor. So, whence surprise? — Luke

From hidden states.

Then the model is not equivalent to the world; there is a distinction between them. The world is not the model or a collective fiction, because the world can surprise us. — Luke

Who said anything about the world surprising us?

My point was that I'm not assuming, either. How am I begging the question — Luke

You said...

Since it is possible that our model could be false in at least some respects, and that we could be surprised, it follows that there is more to truth than a mere "collective fiction". — Luke

This is only true if the terms are interchangeable (that truth is about the model being surprising), otherwise your conclusion doesn't follow, hence you begged the question by assuming that relation in your argument for it.

I take the position of redundancy to be that there are no matters outside of language, and that the model is equivalent to the world, whether that is your personal view or not. — Luke

I really don't know where you're getting that idea from.

I don't read your posts — SophistiCat

Your reply is a performative contradiction.

don't care what you think, so don't jump up and down trying to catch my attention — SophistiCat

If tagging someone constitutes "jump[ing] up and down trying to catch [someone's] attention", then why are you tagging me when you "don't care" what I think? Seems a little undignified...

What I was referring as proof was against the argument from Tzeentch that:

Everybody and their dog knew it wasn't going to be a repeat of 2014, and that the Ukrainians would be prepared. — Tzeentch — ssu

So dodging the substantive question again, then?

Notwithstanding the fact that nothing in the article you cited comes close to refuting @Tzeentch's point. People assessing Russia's full invasion capacity to outmatch Ukraine is not even close to refuting the idea that everyone knew it wouldn't be a repeat of 2014.

There should be no surprises if the world is the model because you claim that the model is a collective fiction. — Luke

Why would that lead to a lack of surprise though? You're not joining the dots.

I already answered the question of why there should be no surprises using your analogy with "Aragorn was king of Gondor". — Luke

You didn't say why there should be no surprises using my analogy with "Aragorn was king of Gondor". You just declared that there should be none.

Why should there be any surprises if the world is the model and the model is a collective fiction? — Luke

Because the hidden states the world is a collective model of may be modelled imperfectly.

By that logic, you are also begging the question by assuming 'truth' does not refer to such hidden states. — Luke

I'm not assuming though. That conclusion doesn't itself form part of my argument for it.

You were one of those saying something "about there being 'nothing' outside of language". — Luke

Come on, at least the bare minimum of effort to fairly represent your interlocutors. It's literally written in the very quote you cited...

there's just 'the stuff that kettles are drawn from'. — Isaac

...so not 'nothing' then...

This is brilliant. So your 'proof' that Russia intended to take Ukraine is that some analysts thought that Russia could beat Ukraine in a full invasion. The same analysts you now want to claim were wrong? So the analysts who you want to claim got it wrong are the ones you want us to believe got it right?

And talking of...

inventing your own fabricated narrative that you then answer — ssu

...any intention of actually answering the very simple question I asked? Or just going to stick with fabricating your own narrative and answering that?

I made it clear to you more than once that you are not worth my time. — SophistiCat

No-one's seriously falling for that. Of course @boethius's comments are well worth your time, that's why you've just chosen to spend your time painting them as being beneath response. If they really were beneath response you'd just not respond. You respond because there's merit to them which you'd rather weren't seen as meritorious, that's the whole point of the 'your arguments aren't even worth responding to' tactic. It's verbatim from the playbook. If arguing against ideas you don't like isn't your scene then just don't, but let's not insult anyone's intelligence by pretending this is anything other than rhetoric.

If the world is the model, then there should be no surprises. — Luke

I still don't see how you're getting here. Why should there be no surprises if the world is the model?

it is possibile that our model could be false in at least some respects, and that we could be surprised, because you speak of the possibility of a better model. — Luke

Yes.

Since it is possible that our model could be false in at least some respects, and that we could be surprised, it follows that there is more to truth than a mere "collective fiction". — Luke

Only if you already beg the very question we're debating by assuming 'truth' refers to the hidden states that the model is of.

No, I'm saying that redundancy conflates the two. If "p is true" means no more than "p" and there is nothing "outside" language, then I don't see how it is possible for the fiction to fail in its task. — Luke

No one is saying anything about there being 'nothing' outside of language, I don't know where you're getting this from.

That proposition is part of the "collective fiction" model and it's not possible that it could be false. — Luke

I don't understand how it's not possible to be false. "Aragorn is the king of Mordor" is false.

Since it is possible that our model could be false in at least some respects, and that we could be surprised, it follows that there is more to truth than a mere "collective fiction". — Luke

Why? You're connecting 'truth' to surprise but that's the very connection in question - the degree to which the truth of "the kettle is boiling" is connected to the hidden states that might surprise me. I'm not denying that hidden states can cause surprise I'm denying the link (or the strength of it) between them and the semantic content of a speech act such as "the kettle is boiling".

I might have a model of my environment that I interact with and could be surprised by (if I get my predictions wrong, or fail to control it). Correspondence theory seems to want have it that our words somehow try to match that environment. I'm arguing that that's not what our words do. Truth is a property of statements, so the extent to which our words don't match an external world, is the extent to which the truth is unrelated to the external world.

None of which is related to the question of whether that external world can surprise us.

Wierd.

I asked "what is Ukraine's negotiating position?"

You answered...

The only way Russia is going to the negotiating table is... — ssu

And...

What Ukraine can do is — ssu

I didn't ask what would bring Russia to the negotiating table. I didn't ask what Ukraine could do in general. I asked what negotiating position you'd be prepared to advocate.

Did you have trouble understanding the question? Or do I take your inability to answer it as an indication that your position is exactly as incoherent as it seems.

It's a simple question. Ukraine comes to the negotiating table... what do they offer?

My guess is you would say that what makes one theory better than another is that it produces less surprises? — Luke

Basically, yes.

If the semantic content of expressions refers to a collective fiction, then how is surprise possible?

It is not possible that Aragorn was not king of Gondor, but it is possible that the kettle is not boiling.

That is, it is not possible that "Aragorn was not king of Gondor" is true, but it is possible that "the kettle is not boiling" is true.

Assuming one is fluent with the language/model, it is no surprise that Aragorn is king of Gondor, but it can be a surprise to find the kettle is not boiling.

If truth is no more than semantic content (i.e. if "p is true" is no more than "p"), then there should be no surprises. — Luke

I don't follow your argument here. I'm saying that the function of the collective fiction is to reduce surprise about each other's behaviour. Firstly one can still be surprised by that very behaviour if, for example, the fiction fails in its task. Second, one can still be surprised by one's environment. The actual response and the act of naming it are two different things. You seem to be conflating the two.

Taking the notion a step further, while it is your expectation, it is another’s anticipation. You expect me to understand; I anticipate I will. And vice versa. — Mww

I think so. I'd been using expectation as if it were synonymous with anticipation. I'm used to talking as if our brains are surprised by what our bodies do. Bayesian models and all that...

It's also true because an event occurred which was parsed as the kettle boiling. — fdrake

"Also"? It seems to be saying the same thing. After all if a different event had occurred, or no event at all, you wouldn't have used the words correctly...?

Effectively we're arguing about whether semantic content relates to appearance or phenomenon! — fdrake

I see it a third way (if that's allowed). Our phenomena are private, so we can't have a public language referring to them. But appearances (hidden states) are inaccessible except via our models, so we can't have a language that's in a one to one correspondence with them either. So to what does the semantic content of expressions refer? My answer is that they refer to a collective fiction. an agreed on, shared model. Just like the fact that we all 'know' Aragorn was king of Gondor. We can talk about Aragorn and his goings on and be right/wrong about them. Kettles are like that. A collective story about the causes of the sensations we all experience, kept consistent by repeated joint activity and repeated joint language use. Which leads directly to...

what do you see as causing phrases to have semantic content that we can collectively relate to and are approximately constant between people in many circumstances? — fdrake

I don't think we do. I think that the success of a expression is a post hoc story. I think we're very good a modelling other people's intentions based on their behaviour and the environment they're in. So when they say "Put the kettle on" we almost know already what it was they wanted done. That's why, if someone with some form of aphasia accidentally said "Put the cat on" in those same circumstances we'd pause only a breath before carrying out exactly the same instruction as if they'd said "put the kettle on". We already seemed to have a good prediction of what it was they meant by the expression before they even said it.

Would I be correct in stating that what can surprise this creature, as a hidden state, belongs to this normative functioning? Are hidden states thus bounded in this sense by the the aims of the organism in its niche? — Joshs

In a sense, yes, but I don't think 'bounded' is quite right, more fuzzy edged than that. A creature (bacterium in this case) has to cohere to survive, it has to resist entropy, forces which would cause it to disintegrate. In order to do that, it has to be able to make changes to its environment (and I'm including it's body here, anything outside of the system's Markov blanket). In order to do that it has to reduce surprise (surprise here is just inconsistency, randomness, entropy).

So yes, this activity will take place within it's niche and so in that sense you're right, but the main driver of this activity is the need to reduce entropy in order to remain a bounded organism (rather than just soup) and that is not bounded by it's particular aims, it's common to any self-organising system.

can we not consider language use as also normative practices of interaction with an environment that is itself ‘bounded’ by the purposes of the language user, even when they are surprised? — Joshs

Yes, I think we can here because language use is a social tool, it only works if other people in our community go along with it. It's a surprise minimisation tool, like any other, it's job is to reduce the surprise other people's behaviour might otherwise exhibit, but it works by us all agreeing, to an extent, on the functions of each expression, the means by which the surprise is reduced. In that sense, language is absolutely going to be bounded by the purposes of the users because we're only going to be able to share models we ourselves have some version of and we don't develop those models in isolation, we often 'pick them off the shelf' of models our society has available for us, most of which are stored and disseminated in the medium of language.

Do the words merely hook onto and describe an action, or are the words themselves actions , normatively guided forms of doing that aim to change an environment in anticipated ways that can be disappointed or invalidated as well as affirmed by the feedback from the environment they alter? — Joshs

Yes, I think it's both. Words (expressions) are definitely actions aimed at making an environment match more closely our expectation of it (the enaction side of active inference). But they only succeed in doing that (when they do succeed) because of the hook they have to other people's models, and this hook is only possible because we quite good at modelling (ie our models are quite accurate predictors of hidden states). If this latter weren't the case, then we'd find it very difficult to share terms, we'd have no common ground over which to share them (unless by complete coincidence!). Which, if I've understood you correctly, is almost exactly what you're saying with...

Agreement would be equally about material practices that are intrinsic to word use. Our words are not just accountable to the linguistic conventions of the group , but are directly accountable to the feedback from the modifications of material circumstances our words enact. — Joshs

...is that right?

I'm not sure what your point is. — ssu

I swore I wouldn't get into this nonsense again, but this one is so simply answered, and could well summarise half the thread in one sentence.

Here's a model negotiation...

"You do/stop X and we'll Y"

So with Ukraine to Russia it would be...

"You stop invading us and we'll Y".

What's the Y you'd be willing to advocate? Because apparently it's not ceding territory and it's not ceding any autonomy and you've just admitted that Ukraine are no threat to Russia. As @Benkei says, the sanctions are somewhat toothless. International condemnation seems irrelevant...

This is what @boethius and others mean by recognising that Ukraine cannot 'win'. They can't reach a point where they don't have to come up with a Y.

The whole argument we've been making here is to try and preempt what this Y is going to have to be as early as possible to avoid the destruction of war.

The counterargument has been that Ukraine ought resist all attempts to limit it's territory or autonomy at all costs (the 'existential fight' we keep hearing about).

But if it's not to cede territory, not to cede autonomy, nor can it threaten Russia itself...

...then what Y can it offer in negotiations?

Kettles don’t boil, even though that is the linguistic and therefore logical construct presented in the dialectic, which necessitates the unstated presupposition in order to validate the argument. — Mww

Ha! I hadn't even noticed. Which makes your comment all the more pertinent, I think. In my view, it goes back to what I was saying earlier about expectation and the use of language as a tool. I can communicate relatively well with someone who doesn't even share my language. I could say "the kettle is boiling", add a few gestures and, if I was in the right context, I could probably get the message across even if the other person had no idea what the words meant before our meeting. So what's happening here is not really to do with the semantic content of each word, or the order we put them in. It's to do with another person sharing my model, my expectations. Watching my behaviour, and using their own explanatory model of their own behaviour to predict what I'm thinking. The words are just me helping to facilitate that, but the process is happening anyway, facilitated or not.

I think it may be "eternally" true to say "I boiled the kettle on the 3rd of March 2022", since I did indeed boil the kettle on that day. — fdrake

Did 'boiling' involve getting all the water to 100C, a rolling boil, the first bubbles, too hot to touch ("that's boiling!")... I don't see how we can establish the truth of such a vague and contextualised notion as 'boiling', even if we pin the event right down to the millisecond.

We could say that it's true that you did something which matches the description. But that just gets us back to where I started (or was it another thread?), where the truth of "I boiled the kettle" amounts to little more than whether you've used the words correctly in your language. "I boiled the kettle" is true because the thing you did is one of the things the expression could rightly be used to describe.

it's really weird to make statements like "The kettle boiled and I poured cold water from it immediately"/ "The kettle boiled and the water stayed still". You convey an expectation of behaviour and then negate it; the meaning of the sentence part "the kettle boiled" is in conflict with the second bit precisely because the first bit doesn't necessarily fix all the events around it, or which can be embedded in relevant descriptions of causal chains involving it, but nevertheless constrains expectations of other sentence parts. — fdrake

Yes, I think this is right, it's (for me) an example of the way that hidden states constrain our models of them. We can have a range if modelled expectations for the entailments of 'boiling a kettle', but none of them can have cold water come out. None of them can result in ice. The hidden states we're trying to reduce surprise in are real and so have constraints. What I'm arguing here (though mostly paraphrasing Ramsey) is that because hidden states are not themselves models, nor bounded in any way, no 'natural kinds', there's no right model. There's only wrong ones. Truth (as correspondence) seems to need a right model.

As a rough summary by example, ""the kettle is boiling" is true" makes sense at its level of descriptive granularity because: ( 1 ) the definite article "the" picks out a specific kettle in the environment ( 2 ) that kettle is individuated from its environment by parsing it into salient objects with distinct patterns of behaviour ( 3 ) "is boiling" states a type of environmental pattern the kettle partakes in ( 4 ) under our level of demarcation, it's "only" the kettle that could boil, not the electrical currents or the plug socket despite both partaking in the boiling process ( 5 ) the kettle exhibits the parsed expectations which constitute (somewhat fuzzily!) boiling ( 6 ) that makes "the kettle is boiling" true. — fdrake

This is a good framework from which to progress, it gives us something to work with. The quibbles...
In (1) the definite article acts as an agreement that we will treat a part of the environment as a kettle it doesn't need us to treat exactly the same part of the environment that way, only similar enough that I'm not going to surprise you and vice versa, which constrains the choices to one which is going to respond roughly the same way. So (2) I have little trouble with except to add that we enact those objects, we can create as well as curate, but that still probably gets us to the same place. (3)and (4) I have no issue with. At (5) I think we miss a step. So at (1) we agree to treat a part of the environment as a kettle, at (3) we do the same for 'boiling', but the theory that the kettle at (1) is exhibiting the pattern at (3) is still, like any theory, subject to underdetermination. Something as simple as 'the kettle is boiling' admits of very little wiggle room for such, but still an important point with regards to 'truth' because it means that even the process-derived truth at (6) remains somewhat agreed on. We don't escape the need for us to socially agree in order for something the have a truth value by this means, it's just that we're constrained in what we could ever possibly socially agree to and still function.

I thought we were discussing what "is true" does, not what "X" (or "the kettle" or "the kettle is boiling") does. — Luke

The argument is that the predicate '...is true' cannot be analytical of X by correspondence if X has no fixed extension. Hence the discussion about X's extension.

Here's a link to a post of mine about this. If you clink on that link, it takes you right to what I said. In this context, we could say it refers to what I said. Following that link is how you get the job done of finding out what I said. — Srap Tasmaner

Very nice. But you didn't construct the link in order to get me to find out what you said. You don't care if I know what you said. You care what I do next. The expression was designed to convince me of an argument. I'm not going to go into the psychological theories about why we do such things, suffice to say the end goal isn't just that I know what you said. So the link may refer, but the words don't, they effect. It's the effect you're interested in, the effect which is the reason you choose them.

If words worked like your computer links, then I think you could claims that they have the effect they do because they refer. But that's the very argument Ramsey is making. That words (or propositions, rather) don't refer like computer links. there's no UUID, they're never specific enough to meet Russell's criteria for meaning 'there is such a thing X and it has property Y'.

I think that's part of what makes semantic content work through; it holds environmental objects equivalent through how we access them, how they function, and how we expect them to respond to manipulation. — fdrake

Roughly. By which I mean that the importance of recognising the functional role of this semantics is that it only need identify a similar enough environmental object to get the job done. we don't need to know if we're including the errant screw to get the expression "put the kettle on!" to work. It'll do it's job even if I'm not sure if 'the kettle' even exists. Even if I've never seen 'the kettle', but merely assume there is one in the kitchen. In this latter case, by 'the kettle' I simply mean 'whatever it is in your kitchen you use to boil water'. I'm not (yet) seeing how such vague and ephemeral environmental objects can be amenable to analysis of their properties to make "the kettle is black" something which can be eternally, objectively 'true', outside of the language game in which it was used. Something like a pragmatic view of truth is the closest here, I think, I can see a possible route to such an approach.

For semantic content be informative about an object's state, there must be an association between that object's state and the language about the object. — fdrake

As with "put the kettle on" above, semantic content doesn't seem to always need to be informative about an object's state. I might not even know of the existence of an object in that expression, so I can't see how my use of the term 'the kettle' could carry information about it's state?

Understanding those functional roles is accurate when it mirrors the developmental trajectories of the kettle, so when someone says "the kettle is boiling", it's true to say that when the kettle is boiling... Because the switch was flipped and the water was boiling etc... Something happened to make it true in context, and the pattern of language tracks the properties through shared expectations of environmental development. The association becomes a causal history of interacting environmental events and language. — fdrake

This makes a lot of sense. It does point, though, to something more of a pragmatic view of truth, rather than a correspondence view. True here being contextual, being about a model of events which works as an explanation. I have some sympathy for that position, but my quibble is that we simply don't always use the word that way.

The causal history; making event-patterns of language co-occur with event-patterns of environments; means both reciprocally inform, and in many use cases reciprocally co-determine - we make our environments navigable and manipulable. — fdrake

Again, I think this approach is attractive in that it gives some explanation of why an expression might work, why that particular collection of words might get a job done that some other collection would not have. Because it's tracked the shared pattern of events. I can communicate remarkably effectively with someone whom I've never met an share no common language with. The can 'see' what I'm trying to get them to do, and vice versa. I could get a villager in Morocco to put the kettle on despite having no shared language and never having met. I can do this because that villager can make good predictions about what my behaviour indicates simply by virtue of their brain having made similar predictions about their own body (but this is all an aside). The point being that our shared history of interacting with the environment creates similar models of it which we can then use to infer intent in others. Language merely expediting that process.

Semantic content is then a historically informed behavioural expectation of the environment, whose developmental trajectories are demarcated through current and prior expectations of development. The causes in the present in both language and the world resemble the causes in the past - the former is a criterion of iterability (like the private language argument against privation), the latter is a criterion of publicisability (like the private language argument against the beetle's wiggling being determinative of sense). — fdrake

Yes, I really like this. Linguistic acts used to denote, and keep consistent, shared expectations, and do so by repeated successful use. Definitely moving toward pragmatism though, if we're wanting this to end up with a definition of 'truth'. We spoke before, I think, about the idea of my wanting my model of what a kettle is to be similar to yours in order to reduce surprise when interacting with you. The way I can use things like ostension and language. I think what you're saying here ties into that nicely. The barrier still in place against correspondence, though, is the lack of specificity. I only need 'the kettle' to be sufficiently similar in our shared expectations about it to maximally reduce surprise. Too specific a object won't do that, it actually needs to be vague to have a chance of my having unsurprising expectations of it.

The causal patterns of language use grow to resemble the environment modulo perceptual individuation. Feedback lets the former and the latter have reciprocal impact; so much so that in many circumstances we can append "I think" to to a phrase, like "I think the kettle is boiling" and convey the same behavioural expectation of the environment but indexed to an agent. But the distinction between the two is precisely useful because the environment's patterns are shared. The causal histories differ, so the behavioural expectations in the environment and in language differ, so the meanings differ. One is true when you think it, one is true when the kettle boils.

The mirroring+coupling of causal patterns of language use and environmental comportment/expectations of development is what sets up the remarkable agreement obtained on whether the kettle is boiling... When it is in fact boiling or not. The causal history of language and environment over time, through constant work and tailoring, becomes discriminative on both environment and language. — fdrake

If I've understood you correctly here, this is similar to what I was saying earlier about the environment constraining what can be said. It sets limits on what will work because regardless of out models of it, it is set out in such and such a way and it's not a homogenous soup which we can make of what we will. There may be a wide range of values which will make "the kettle is boiling" true (in that sense), but they will not be infinite. "the kettle is boiling" won't work given certain environmental constraints. Again, pointing to a rough correspondence, but one insufficiently specific to be amenable to the sorts of truth analysis direct correspondence would seem to need.

I think where we differ is that I interpret the pragmatic context as part of the function, and the function itself isn't situated within a body, it's situated between bodies, in the environment, and within bodies - like with Srap Tasmaner 's comment about externalism vs internalism of semantic content. I don't think "the science" sides with either side on that, at least not yet, so it remains a site of substantive philosophical disagreement. — fdrake

As I said above to Srap, The idea that I can change the external world by some vocalisation (same for doing so by some gesture) to others does indeed rely on the notion that those other sufficiently share my models, and that co-operation is sufficiently part of their policy, for those gestures to work. But I don't agree that it requires a set of shared 'meanings' which are then reified to some objective status with sufficient specificity to be amenable to truth analysis. We can invent gestures on the hoof and still be understood. If there's a language barrier, certain words are quickly learned (and what is learned, is what the word does). Communicating with someone who doesn't share my language is less efficient, but still very possible and we can carry out many basic co-operative tasks, we don't seem to need an already prepared external system of word and reference.

So I think, yes, this is all about our shared would, but I don't think the co-operation this is all here to allow requires an actual set of word>reference facts that are external to our intentions. It simply requires that we're similar enough in intentions and co-operative enough in policy that we can see evidence, in another's behaviour, of what we need to do to bring about the state of the world which includes helping the other.

I'm not clear where this is going. I don't think there's anything about saying that "the kettle" is defined functionally which renders it victim to Oliver's dad's joke position.

If I say "Pass me the kettle", I'm just using an expression which I've learned is a tool to get something done. In my terminology, I have an expectation that the world be such that I can fill something with water and I'm interacting with the world via language to make it match my expectation. It's a prediction of what action will make it that way. As Wittgenstein has it at the beginning of the PI, I could have just said "kettle!", or simply pointed to it and clicked my fingers.

Nothing here defines what "the Kettle" means in any specific way. It's a tool I reach for as part of a strategy to get some change in the world enacted, and it's non-specific. So long as it gets the job done. It's sufficient, it seems to keep these expressions vague, relying on compound ones to be more specific "that kettle over there, the red on, not the black one..."

The point of all this was to say that there's no eternal, external, definition of what constitutes "the kettle" that anyone could use to determine the truth (by correspondence) of "'the kettle is black' is true". There's no corresponding external world object to "the kettle"

"The kettle" is a linguistic act. Saying that something corresponded to it would be like saying that something corresponds to my extended finger when I'm pointing at the kettle. If I point and say "pass me the kettle", nothing different has happened than if I point and click my fingers, motion with my had that you're to pass me the object I'm pointing to. But we don't say that my action with my hands 'corresponds' to the object in question, so why should my actions with my voice box do so?

With the pointing and gesturing, I rely on the fact that you share sufficient aspects of my expectations, and that my goals are sufficiently part of yours, that you'll see my actions as evidence for your policies. The same seems the case with language acts.

Bring this back to 'Truth', the notion that "X is true" can be checked by examining the properties of X relies on 'X' referring to some fixed set of properties. But 'X' doesn't refer to a fixed set of properties. 'X' doesn't refer at all, it's a type of action that gets a job done, it doesn't refer any more than lifting my arm does.

all that's wrong: knowledge doesn't have parts, not truth, not belief, despite entailing both truth and belief; and the explanation of action solely in terms of narrow conditions, as the internalist would have it, is weaker than the explanation of action in terms of wide conditions, as the externalist would. — Srap Tasmaner

Go on...

The upshot here is that you successfully refer to the kettle in the kitchen despite possibly holding a false belief about it, perhaps many (what brand is it? when did you get it? didn't you have to replace it and this is the new one, or was that a different kettle?) and your intention should be taken, in proper externalist fashion, to be toward the actual object, not toward your possibly mistaken idea of the object. — Srap Tasmaner

I'm with you so far, but all this seems to make our vague picture sufficient to get the job done. I'm not seeing the link to it being sufficient for the analysis of truth.

In order to get you to make tea, my picture of the kettle can be vague (not even deciding if it includes the screw or not), it can be mistaken (I could think al along that it's a mine when it belongs to you)...it doesn't matter one jot to get the job done since you can infer my intent sufficiently.

But in order to check the truth of "the kettle is boiling", it's insufficient. Unless all you want to be true is "the kettle {the picture I have of it at the time I'm speaking this sentence} is boiling {the idea of 'boiling'' that I have at the time of speaking this sentence}". That, I suppose, could be true by correspondence, but only the speaker would know and only at the time of speaking, so I can't see that being the story of 'truth' the correspondence theorists are looking for here.

The point I was making was that if we cannot collectively and permanently agree on what a kettle is, the we cannot asses the truth value of any statement about it by correspondence. Correspondence to what?

You can successfully refer to George Washington even if everything you think you know about him is false. — Srap Tasmaner

I don't even agree here, but ready to be schooled. In what way 'successful'? If I think GW is cow an what her to be brought to me for milking and say "fetch me George Washington". When I'm brought a US president, I'm certainly not going to think my reference was successful.

The kettle is not just any vessel for boiling water, but the one in the kitchen, the one you mean, the one you have an intention toward. — Srap Tasmaner

Yes. I used 'the one which boils the water' as shorthand. we could add 'the one which boils the water, the one which I get to take home without conflict, the one which I can reach when I sat on the rocking chair...' We can still make a functional account. It just takes imagination, because writing it all down longhand is tiresomely time-consuming.

You don't have intentions toward any such idea -- that's the lesson above -- but toward what you have ideas about. — Srap Tasmaner

I agree. I'm not sure if you're think I don't, but, for clarity, I do.

I don't know the kettle has been fixed, though I have a true belief that it has been fixed. That's epistemic luck. I handle the kettle as if it's been fixed and have no trouble; I might even attribute my successful endeavors with the kettle to my having fixed it, even though our assumption here is that it would have made no difference if the kettle had still been unfixed. — Srap Tasmaner

I think this just shows the problems with attributing the idea of 'knowledge' to truth. Fraught with such problems. Knowledge as justified true belief doesn't really make any sense. But that's definitely another thread, one that I think already exists even...

Admiration for the screw example. It makes it so clear that what counts as a part of the kettle is up to us. — Banno

Thanks.

the truth of a sentence often depends on more than just its meaning; it often depends on a material object, or on a mental phenomenon, etc. — Michael

I'll reanimate a previous example. Those of us with an upstairs will perhaps have a landing light which has a switch upstairs and one downstairs. The landing light can be switched on or off by either switch.

We can't say, though, that whether it's on or off depends on either the upstairs or the downstairs switch. The downstairs switch could be either in the up position, or the down position and the landing light still be on. Same for the upstairs switch.

My suspicion is that there is a gap brought about by there being a difference in kind between neural networks and truth statements. — Banno

Yep, mine too. Basically, the gap is 'black-boxed' out. I have one model where hidden states are inferred by neural networks and then acted upon (to reduce surprise), then another where in real life we name those hidden states by their collective function (by how we together act on them). So 'kettle' refers to the hidden state we treat as a kettle. Whether our treatment will be successful is between us and our models. But where 'truth' might fit is opaque. Obviously it can't reference the hidden states, that would be futile, nor the model (we don't even have access to those ourselves, only their output in terms of action). so I can't see anything in the model of how we interact with the world that 'truth' could possibly refer to... hence my preference for redundancy.

In more Wittgensteinian terms, there is an active intent that makes the kettle a kettle. The kettle exists as a result of our treating it as such; which is not to deny that our intent is constrained, Isaac's hidden states. But it is constrained by the kettle; that seems to be what we have decided to call some of the hidden states.

So, Isaac, perhaps those states are hidden from our neural nets, but not from us — Banno

I should have read on, I could have just agreed rather than write it all out longhand...

I think 'hidden states' is a confusing term. I would prefer it weren't the one used, but it's become a technical term now, so we're stuck with it, but too many think it means hidden as if the states were just behind that rock, or round the corner. all it means is that they are in connection with nodes at a Markov boundary of a network and thereby, in some sense, 'hidden' from the nodes within that network (ones that are obviously only connected to the boundary nodes). so yes. There's no reason to think they're hidden from us in the common sense. We name them, and we make tea with them.

I take it that Isaac has a strong position that they're linguistic all the way down and what they count as publicly is what they are, and the reference mechanism actually references an entity conjured up by collective agreement rather than some concrete fact. The referent of "the kettle" is a collectively enacted categorisation of the environment, rather than some environmental object. — fdrake

That's pretty close, but I've maybe clarified a bit in my reply to Srap above. That categorisation is about function, not spatio-temporal locations. We're not collectively declaring that that collection of matter-soup there is a 'kettle', so much as declaring that whatever collection of matter-soup is boiling the water is a 'kettle' (plus a boatload of other functional requirements adding specificity - so 'my kettle' is 'whatever aspect of the environment boils water and I can determine where it goes without any counterclaim... 'the black kettle' is 'whatever aspect of the environment boils water and which would be difficult to see against my stove in the dark'... and so on)

Let's say that the kettle is boiling is true, what would the proximate cause of that expression's truth be? My intuition for that is that the kettle really did boil. I think Banno, @Srap Tasmaner and @Michael would agree (though possibly for different reasons), though I suspect @Isaac would have a strong quibble. — fdrake

You'd be right. I couldn't see it as a cause, so much as a repetition of the sentence you're analysing. If the expression in question is "the kettle is boiling", then looking at what 'causes' it to be true seems a little disconnected. I could understand the question of what cases us to say it's true, but not of some state of the world causing an expression to have some property...

...unless maybe utility. Which (much to many people's distaste, no doubt) is the other route I'm tempted by in discussions about truth. I think the hidden states of the world constrain what we can collectively enact, which is where I diverge from the more radical idealist interpretations of model-dependence. So we might say that "the kettle is boiling" would be a useless expression unless the world were in some state.

I love the fact that I always get a free anecdote along with the philosophy in your posts. I immediately tried the trick on my son, who was visiting. We were most amused.

But to the meat of it...

If, for instance, that screw holds one end of the handle in place, you know whether and how the handle can be used. It will be important for me to have that knowledge too in order to put the kettle on. (I have a dozen or so possible scenarios in my head now, but I assume you don't need any of those spelled out.) — Srap Tasmaner

Yes, that makes sense. In my model, I would see that as an expectation I have, a belief about what using the word "kettle" will do in the context. A belief about your beliefs, if you like. Given that the target of all this is 'truth', though, and 'truth' being traditionally a component of knowledge. I might say, for clarity, that neither I not you need to 'know' any of this. It's sufficient that we believe it.

the fact that there are multiple options doesn't mean you didn't have something specific in mind — Srap Tasmaner

Indeed. But never specific enough, is Ramsey's point, to make propositions about which can then be objects of the sort of analyticity that questions of Truth put them under. We might, this way, end up with a kind of private correspondence theory of truth "the kettle is black" is true for my kettle (the one I had in mind). I suspect most purveyors of truth-theories, would be dissatisfied with that.

If every object we were concerned with carried a UUID, and we could keep track of those, we could use those to end up in the same place. — Srap Tasmaner

I think that's right, but not in a concrete world-outside-language sense. The object that I'm referring to when I say "put the kettle on" may or may not have the errant screw. I may not care. my picture of it may simply not be in sufficient detail to even decide if it has the screw or not. And I think this is because the "kettle" bit of the sentence doesn't refer to an object by material composition, it refers to an object by function. What I'm referring to with "kettle" there is 'whatever it is that boils the water', not 'that collection of fundamental particles there'.

...but then, that referent is awfully hard to use as an object of correspondence, since lots of potential states answer to it.

The truth value changed because I painted the kettle red. — Michael

Is that true?

I can't see how it's coherent to escape the way meaning determines truth by claiming some intervening statement is true anyways.

If you're going to do that, you might as well say "the kettle is black" is true because you painted it black.

"I painted it red" isn't an un-interpreted, raw fact of this world-outside-language. It's just another already interpreted fact of the world, just like the kettle's blackness. In order to agree that you did, indeed, paint the kettle red, we need to agree what the kettle is and what 'red' is (and what 'painting' is, but we can leave that for now).

You're introducing it to the argument as if it were a purely material fact, but it's a fact of exactly the same kind as the one we're troubling over.

2.

Your argument seems the equivalent of...

If the {on/off state of a light} can change without {switch for that light} changing then the {on/off state of a light} depends on more than just the {switch for that light}.

Yet in the above scenario I'm not prevented in any way from still using the switch to entirely determine whether the light is on or off, right? It's not like it requires {switch} plus {other factor} to be on. The switch still turns the light on or off. It's just that something else does too.

We could end up in a kind of war of attrition between me constantly flicking the switch to turn the light on, and this other factor constantly turning it back off again. As I said to @Srap Tasmaner above...

If we want an ephemeral, relativist 'truth', then sure we could compare the 'kettle' of any given conversation to the 'black' in that same conversation. — Isaac

If we have to agree to the meaning of "the kettle is black ", then we have to also agree that your actions at T1.5 constitute "painting the kettle red". Thus it turns out this {other factor} is linguistically determined too.

Did you paint the kettle red? Only if we agree about 'kettles' and 'painting' and 'red'.

you're rejecting my claim that the sentence "the kettle is black" means the same thing at T1 and at T2 — Michael

No, I agreed to that. You stipulated that it does for the sake of your thought experiment.

If it's not me painting the kettle red that changes the meaning then what does change the meaning? — Michael

Nothing, we chose not to, according to the rules you stipulated. You stipulating such rules is what determined the truth value at T2

...

T1 - we decide the meaning of "the kettle is black" is such that it is true.

T1.5 - You paint a kettle.

T2 - we decide we decide the meaning of "the kettle is black" is such that it is false (for the newly painted kettle).

Our action at T2 is completely unconstrained by your action at T1.5.

Our action at T2 is, of course, constrained by you constraining it for the sake of argument. That's just tautologous and doesn't tell us anything about what is necessarily the case.

Are you saying that me painting the kettle red changes the meaning of the sentence "the kettle is black"? — Michael

No. I don't see how you're getting that.

The meaning of the sentence "the kettle is black" didn't change at T2 — Michael

...because we stipulated it wouldn't.

So...

its truth value depends on — Michael

...us stipulating the meanings of the expressions under consideration.

The expression at T2 could be either true or false depending entirely on our stipulation. So you can't conclude that it (as a fact about what is the case) depends on you painting the kettle red. It's immaterial whether you paint the kettle red.

After T1, there are two possible scenarios.

Scenario 1 - you paint the kettle red (in this world-outside-language) and we stipulate the meaning of "the kettle is black" such that it's false.

Scenario 2 - you paint the kettle red (in this world-outside-language) and we stipulate the meaning of "the kettle is black" such that it's true.

Either scenario is possible. So painting the kettle red (in this world-outside-language) has no determining relevance.

If you artificially constrain the situation to only allow scenario 1, then all you're doing is determining the truth value of the statement at T2 by eliminating the other option. You've made it true because truth is binomial and you've excluded one option.

You agreed above that we can decide what words mean. So, for the sake of this example, we decide that the screw in the draw is not part of the kettle, and to use a spectrophotometer to measure the kettle's colour, agreeing which range of results indicates the kettle being black and not-black. — Michael

Yep. Look at the bolded bits. The activities described are those of a living language.

you posit a stable neural network, or some such thing, as a stable "thing", which would support repetition of the same, or similar mental activity, constituting the thing which others might call a "belief". — Metaphysician Undercover

Not at all. Just like a 'race' is any kind of activity which has a start, a finish, and some competitive element, a 'belief that the pub is at the end of the road' is any mental arrangement which results in a tendency to go to the end of the road when wanting to get to the pub.

For the sake of argument we have fixed the referent of the phrase "the kettle" (and "black", and "red") such that the truth value of "the kettle is black" is unambiguously true at T1 — Michael

Doesn't that just beg the question a little? Ramsey's concern about propositions is exactly that we just can't do that.

The meaning of the sentence didn't change at T2 but its truth value did. Therefore, the truth value of the sentence depends on more than just its meaning. — Michael

But the meaning only didn't change because you said it didn't. Again, this misses the main objections (Ramsey's propositions and Wittgenstein's private rules). If you declare that the meaning of an expression is just exactly what you say it is, then I think you might possibly be able to make the move you want to make. But then you'd have a private rule concerning the meaning.

The sentence "the kettle is black" is true at T1. I paint the kettle red at T2. The sentence "the kettle is black" is false at T1. — Michael

Is it? Does "the kettle" include the screw in the drawer or not? Does it include it at T1, but not at T2, or vice versa perhaps? Did you paint the screw. Was it a really dark red that I'd call black?

if it is so identified, identified by the use of language, and by our forms of life more broadly, as the man said, then it is the thing in that sense identified by our use of the expression "the kettle". — Srap Tasmaner

Absolutely. You've hit the nail on the head.

It cannot be be the thing in that sense identified by our use of the expression "the kettle" because no single, agreed on thing (matter, particles, hidden states) fits that bill.

What the expression "the kettle" does, changes from use to use.

If it's not, then there has been no collective identifying of something by use of the expression "the kettle". — Srap Tasmaner

Sufficient to get a job done though. If I say "put the kettle on" I don't need you to know if that includes the screw in the drawer. I assume you gather my intent. I could probably have just said "tea time!"

If we want an ephemeral, relativist 'truth', then sure we could compare the 'kettle' of any given conversation to the 'black' in that same conversation.

But if we want a 'truth' that gets outside of these conversations... Which use are we going to pick?

But it's still the case that whichever matter we decide 'counts' as being the kettle must exist for the sentence "the kettle exists" to be true. — Michael

Yes. But since it could be literally any matter at all, to claim that the truth of any sentence involving kettles depends on this fact would render all statements about kettles always true, since there's always some matter.

That's for us to decide. — Michael

Yep. Using language.

The truth of "the kettle is black" cannot be determined by hidden states because nothing in those hidden states determines that they should be a kettle, nor exhibit the property 'black'. We determine that by language use.

Did I? Where? — Michael

I asked...

Any nonlinguistic feature? — Isaac

...and you answered...

Sure — Michael

But if it...

was a mistake. — Michael

...then my original question stands unanswered. Does this particular matter the truth about the colour of the kettle depends on, include the screw in the drawer or not?

A sentence like "the kettle is black" isn't made true by another sentence, but by the existence of a particular material object — Michael

You just agreed the contrary. You said "yes" when I asked if the material particular matter was any particular matter. So it isn't made true by the existence of a particular material object, since any material particular matter will do, it's always true.

It's 4 I'd quibble over. I'm not sure in what sense we can say the kettle is a concrete object if we can't agree on what that concrete object constitutes (and yet still unproblematically use the expression 'the kettle'). If we can use "the kettle" without issue, and yet can't even say whether the thing includes the screw in the drawer or not, it is hard to see how the kettle could be a concrete object.

But it depends how you're using 'concrete' here. I think the world consists of those objects we, collectively, identify with our forms of life (our language, for modern humans). So the kettle is definitely an object in the world, in that sense. But that's not this world-outside-language that @Luke and @Michael seem to be reaching for.

That world seems closer to what I would call 'hidden states'. But hidden states are a hypothetical notion in a scientific model. There is (according to the model) a relationship between hidden states and our shared objects, but it's a constraining one, not a determining one.

still requires that there is something in addition to the sentences "it is raining" and "the kettle is black" for these sentences to be true. Truth depends on more than just language. — Michael

Absolutely. And we're agreed there. But if what it relies on can't be specified (does it include the screw or not?), then it can't act as truth-maker. Worse, if what it relies on merely need be something, but not any specific thing, then it drops out of conversation. Which is all redundancy is saying.

Therefore, there are no boiling kettles outside of language, either? — Luke

No. Language is what delineates 'kettle' as an object. Without it, there's just 'the stuff that kettles are drawn from'.

The kettle itself — Luke

So, outside of our talk, is the screw in the drawer part of the kettle or not?

I don't believe there's much controversy about what a kettle is. — Luke

So is the screw in the drawer part of the kettle or not?

Boiling point — Luke

... is the scientific definition. It take a colloquial definition (one where I need to see a good volume of bubbles before I'll say the kettle is boiling). what fact of the world-outside-of-language, tells you I'm wrong?

Redundancy without realism leads to relativism and a disconnection of language from the facts of the world. If you accept realism, then you also accept some form of facts, correspondence and truthmaking. — Luke

Not at all. I laid this out (you don't seem to be actually reading the things I'm writing - if I'm not being clear, perhaps you might say so). What is real might well constrain our language. That is is not specific enough to act as truth-maker, doesn't mean it doesn't exist.

what is the point of testing a theory in science? — Luke

To get a better theory?

kettle being black depends on the existence of particular particles at particular locations in space. This has nothing to do with language (even if language is required to talk about it). — Michael

What particular particles? Do they include the screw in the drawer or not?

the kettle being black depends on the occurrence of a particular sensory experience — Michael

What sensory experience? The one I say is that of a kettle, or the one you say is that of a kettle?

get wet when I stand out in the rain — Michael

Do you? Or do you get damp when you stand out in drizzle? If you're wearing a coat are you still getting wet? Does the sentence "I didn't really get wet, just a bit damp" make no sense to you?