I agree. However, we could draw inferences about the nature of reality by examining the past, and apply that analysis (that model of reality) to making predictions. This is, of course, the nature of physics. — Relativist

Yes, and as I said earlier, such examining is all we have to go on, and so it is rational to base our inferences on that observation and its understanding and apply that analysis to making predictions. This is the nature of, not only physics, but science generally. I think this puts to rest the problem of induction.

Note that your examples concern our beliefs. There's a difference between the past constraining the future, and the past constraining our beliefs about the future. Bayesian calculus only allows the latter.

The other is Gillian Russell's recent work on logic, just mentioned. That is about the world rather than about our beliefs. — Banno

the idea that the past constrains the future relies on the idea that the '"laws of nature" may evolve over long time periods, but will not suddenly alter. — Janus

It occurred to me when I wrote the above that I am addressing only our ideas (beliefs). I could have written 'invariances' instead of "laws of nature". Do you think it is reasonable to say that if the past constrains the future it follows that nature's invariances do not suddenly or randomly alter?

I'm not familiar with Gillian Russell's work...will check it out.

I'm not familiar with Gillian Russell's work...will check it out. — Janus

Came across this...

This book’s proof of the Strong General Barrier Theorem is a landmark achievement in twenty-first century philosophy. Not since Ludwig Wittgenstein’s Tractatus (1921) has such an important contribution been made to philosophical logic. — Barriers to Entailment by Gillian Russell

It occurred to me when I wrote the above that I am addressing only our ideas (beliefs). — Janus

Yep. So Bayesian Calculus is about belief, but Russell's work is to do with models, and so truth. Looks pretty cool. It is a formalisation of the problem, and the "barrier to entailment" mentioned in the OP.

But I need to get into the detail.

I could reframe the question—if in logic something's being true entails other things being true (at least sometimes) can the same be said of reality? That is if something obtains is it ever necessary that other things will also obtain?

Sure. Our words are about the world. The true, ones, at least...

But I wouldn't put that in terms of necessity. Too loaded.

Bayesian calculus deals with our beliefs, such that given some group of beliefs we can calculate their consistency, and put bets on which ones look good. But it doesn't guarantee truth. So what it provides is rational confidence, not metaphysical certainty. It's in line with Hume's scepticism.

I suspect we are emphasising different aspects of the same issues, and that we do not have an actual disagreement. What do you think?

I could have written 'invariances' instead of "laws of nature". Do you think it is reasonable to say that if the past constrains the future it follows that nature's invariances do not suddenly or randomly alter? — Janus

If you are asking me, then invariance is the globally constraining symmetry of the Cosmos. And so what we mean by talking about Nature’s laws.

And there is evolution of these laws as constraints produce freedoms. And those freedoms can reorganise the general state of those constraints, so producing a new state of cosmic order.

Furthermore, the change is in fact often abrupt. As in the phase transitions of the kind when the temperature or pressure drops, causing steam to turn to water and then ice. A change in topological state from gas to liquid to solod.

The Big Bang was a story of at least five or six such major reorganisations in just its first billionth of a second. Events like inflation, a reheating dump, the Higgs crack, the CP violation phase. All added constraints to the previous physics to lead us towards the world as we now know it.

It took three minutes to get to a state of a hot atomic plasma of electromagnetic radiation. Then 380,000 years for the next major phase change that was the cooling of that plasma - that radiation soup - to the point that it broke into a cosmic microwave background filled with the condensed crud of a gravitating dust of atomic matter. Electromagnetism neutralised and so gravity able to start sweeping the atomic dust into stars and galaxies.

So the laws of nature did evolve through many stages. There was the one prevailing evolutionary logic. The Big Bang was a fireball cooling itself by expanding. But that same cooling-expanding exposed new ways that a suitably cooled and spaced out state of material being could find more complex ways to become thermally organised, As a dissipative structure, it could reorganise itself into richer forms with their ever more detailed or localised laws.

Luckily for the existence of us. As otherwise we wouldn’t be here.

And do things persistently in that rich state. At least until the current high water mark of cosmic complexity starts to eat even itself and erode back to the generalised nothingness of a Heat Death void. The ultimate inversion of the Hot Big Bang where it all started,

So the story of the laws of nature is that they started ultimately simple, became interestingly complex, and then eventually are going to degenerate back to the ultimately simple. An ultimate simplicity that is kind of the same, just the dichotomous or inverted form of that initial symmetry. The anthesis to the thesis.

In the end, nothing will be left. But it will also be so eternally everywhere. And it’s rules will be as simple as possible as by then, possibilty will likewise have become as simple as it can get.

I suspect we are emphasising different aspects of the same issues, and that we do not have an actual disagreement. What do you think? — Banno

I think that is probably right. I've been watching a lecture by Russell on YouTube—finding it interesting, but there's a lot to wrap my head around.

That's an interesting account that certainly seems to make sense. If we are in the evolutionary middle, so to speak, does it seem plausible to think we in a stable era where the invariances are not likely to suddenly radically change?

It's not easy stuff, but to my eye it's by far the most interesting thing going on in philosophy at present.

It certainly seems to me something interesting to explore, but as you no doubt know, I am not well-schooled in formal logic.

I am also interested in semiotics, but having a few other non-philosophical interests and commitments which are important to me, what I really need is more time if I want to gain more than a superficial understanding of these things.

Bayesian calculus deals with our beliefs, such that given some group of beliefs we can calculate their consistency, and put bets on which ones look good. But it doesn't guarantee truth. So what it provides is rational confidence, not metaphysical certainty. It's in line with Hume's scepticism. — Banno

I think you neatly demonstrate the pitfalls of relying on words when doing serious metaphysics - as in here, trying to say something useful about epistemic method without merely restating the bleeding obvious.

You’ve just twisted a lot of words to meet your sociological ends. And I’m sure you feel that is a watertight verbal construction. There is some chain of entailment that was constructed to close your linguistic sketch.

But oh what a leaky boat. It never left the safe harbour of self-centred idealism. Which would be the only reason it felt like it floats.

Cheers. Semiotics seems to me to miss the point by treating all causal talk, all meaning, as merely codes or signs floating in abstraction. It's what we do!

You don't actually say anything here about why I'm wrong. That's why I tend not to reply to your posts.

You can't be claiming that Bayesian calculus is not about belief. So, what?

You don't actually say anything here about why I'm wrong. — Banno

I already pointed out the issue of priors. Others have noted your failure to cash out semantics in stabilising ontic commitments.

You can't be claiming that Bayesian calculus is not about belief. So, what? — Banno

It is about the openness of beliefs closed under ontic commitment. Best inference constrained by the reality to be encountered at its end.

Your arguments are so sloppy. You point to a SEP page and say “see!”. You mumble to your class about maybe having to start a new thread on that and then womble off to lunch. Apparently forget that immediately and wander back in chewing on a sandwich.

But anyways. Bayesian reasoning + dissipative structure theory = a biosemiotic model of life and mind. Friston’s Bayesian mechanics.

That is what Bayesianism closed under thermodynamics looks like. A world that sets the weights on a mind’s priors in useful fashion.

You should really try to catch up with where epistemology is at.

Cheers, Apo. More about me. Thanks!

It is about the openness of beliefs closed under ontic commitment. — apokrisis

:wink: The grand edifice is tinsel.

Friston’s Bayesian mechanics, like any Bayesian scheme, formalises rather than solves Hume's answer to his scepticism.

Friston describes how an adaptive system maintains itself by predicting and minimising surprise, treating say the nervous system as a hierarchical Bayesian network that continually updates its internal model of the world to reduce the gap between expected and actual sensory input. Neurons encode probabilistic beliefs; learning occurs through adjusting these beliefs (priors) to improve future predictions. This formalises pattern recognition as an inferential process: perception and action both serve to confirm or refine the brain’s generative model.

Relating this to Hume’s scepticism, Friston doesn’t refute it so much as operationalise it. Hume doubted that we can justify inferring the future from the past; Friston shows how organisms predict the future by continuously revising expectations in light of prediction errors. The model gives a pragmatic, mechanistic account of such learning, not a logical justification for it.

Friston’s Bayesian mechanics is widely influential but still speculative. It’s accepted in the sense that its core idea, the brain as a prediction machine that minimises error, has strong support across neuroscience, psychology, and AI. But the claim that all cognition, life or the universe as a whole can be explained as “free energy minimisation” is speculative, overly abstract or perhaps even unfalsifiable.

So his ideas are accepted as a powerful framework for modelling cognition and perception but speculative as a general theory of life, the universe and everything.

Adding Pierce and such looks good, but lacks substance.

Aha! AI now writing your posts. That will solve one of your problems. :lol:

Now you just need to learn to write honest prompts.

Anyways, as AI replies….

Friston's Bayesian mechanics learns from experience by using prediction errors to update its internal models of the world, a core component of the Bayesian Brain Theory and Active Inference. Incoming sensory data is compared to the system's predictions, and any discrepancies drive changes to the probabilistic beliefs and generative models, allowing the system to adapt to its environment and improve its simulations of reality.

These prediction errors are crucial because they are used to update the internal models and beliefs. This process of updating probabilistic models based on new sensory evidence is the core of Bayesian inference. Through this ongoing process of prediction, comparison, and updating, the organism constructs and refines its "reality model," which enables adaptive behavior in a complex environment.

So you forgot the power of recursion.

And also the power of attention. The ability to shift from responding habitually for as long as the future is resembling the past, to attention the moment it no longer does.

Friston's Bayesian brain model explains attentional processes by proposing that attention acts to estimate and manage uncertainty during hierarchical inference, thereby controlling the flow of sensory information. By dynamically adjusting the "precision" of prediction errors (how much weight is given to sensory input vs. prior beliefs), the brain can focus processing resources on the most informative parts of the sensory scene, a process which naturally accounts for phenomena like salience and selection

So priors can be suppressed and new ones rapidly prototyped by dynamically fixating on some narrowed part of the information space. Letting that spotlit part of the world now be what constrains the system’s abductive reasoning.

Brains are very clever in their design. Might be worth focusing on how they actually function for a change,

AI now writing your posts. — apokrisis

Something you'd never do...

I'll leave you to it, insults and all.

...the openness of beliefs closed under ontic commitment. — apokrisis

:nerd:

Have fun.

Caught red-handed!

See the paper Hume’s Law and other Barriers to Entailment

Page seven is pretty clear.