Yep, it's a simple point. So why all the fuss? — Banno

You seem to forget what you were originally trying to argue....

And that's the point I want to make; that when someone provides us with a definition we go through a process of verifying it; but what is it that we are verifying it against? We presume to be able to say if the definition is right or wrong; against what are be comparing it? Not against some other earlier definition, but against our common usage.

And if we already accept that this common usage is the test of our definition, why bother witht eh definition at all? — Banno

Clearly you now accept this was confused as we do seek definitions that introduce new measurables.

Our earlier "common usage" definitions come into question particularly when we come up against borderline classifications - like: "is a virus alive?". The vagueness or uncertainty we feel when answering is a sign that we now need to sharpen our definition by suggesting some new symmetry-breaking or dichotomous fork in the road by which we can measure what is what. An infected cell goes down this path to join the living, the virion fragment goes down the other path to join the class of the not alive, or whatever.

So we want to know as usual when facing indecision, what counts as the essential difference? What is the difference that makes a difference. And so, what were all the in fact inessential differences that might have been clouding our earlier "common usage" conception?

Excellent post!

So you say. Naive realist I'm fine with; but what is a transcendent solipsist? — Banno

You talk as if you can know the world without making measurements. One only has to look and one sees (ignoring the fact that seeing the world is the forming of a phenomenological state that is our interpreted sign of the noumenal - we can't in fact sneak a peek directly).

Tell me, Apo, how do you get on with Meta? I can't say I've paid much attention to discussions between you two. Are you in agreement as to the nature of essences?

Tell me, Apo, how do you get on with Meta? I can't say I've paid much attention to discussions between you two. Are you in agreement as to the nature of essences? — Banno

What does he think about essences? I can't say I've paid any attention to your discussions with him.

I've been discussing essences with him for years; it seems to me he has some sort of reified view of essences; although sometimes he talks about them as if they are no more than language conventions. He might join us here again.

For now I'm nonplussed that you think what I was saying in some way implies that a definition cnnot be improved. So I don't know what to make of the following:

Our earlier "common usage" definitions come into question particularly when we come up against borderline classifications - like: "is a virus alive?". The vagueness or uncertainty we feel when answering is a sign that we now need to sharpen our definition by suggesting some new symmetry-breaking or dichotomous fork in the road by which we can measure what is what. An infected cell goes down this path to join the living, the virion fragment goes down the other path to join the class of the not alive, or whatever. — apokrisis

This is not in contrast to what I have suggested; but I think it is in contrast to the notion of an essence, immutable and eternal, which seems to be what Meta has in mind.

Perhaps we have three views: Meta advocating essence as a real thing that we can set out in terms of the necessary and exclusive attributes; you, with some notion of an asymptotic essence that we can approach but never quite reach; and I, with the view that essences best ignored in favour of the examination of language use.

Just trying to make sense of your objection.

And yet you previously claimed: — Banno

As I explained earlier, there is no "correct" or "incorrect" to word usage unless we are talking about logic. And since "essence" is an aspect of correct usage, we only have essence in logic. So "essence" is a logical principle. When I explained that I was talking about epistemological principles here, you seemed to ignore that part of my post. I assumed you were to lazy to read my post, and I dropped out of the conversation.

So from where do you derive whatever you call the "essence"? — Banno

You derive the "essence" from studying the named object. Propositions of that essence are made, and when they are agreed upon we have acceptable premises for deductive logic. With deductive logic we can relate that object to other objects which have a defined essence. Without this defined essence, deductive logic is useless because there is ambiguity as to what the words refer to.

Tell me, Apo, how do you get on with Meta? I can't say I've paid much attention to discussions between you two. Are you in agreement as to the nature of essences? — Banno

As I told you earlier in this thread, I don't think of essence as an ontological entity, I think of it as an epistemological principle. I believe there is a large difference between the form which an object has, and the form which is understood by the human mind (and this understood form is what we call its "essence"). That is why human minds, and consequently "the essence" of objects is often mistaken. Without this separation there is no way to account for error. So every object has a form which is proper to it, and this is independent of the human mind, which itself assigns a form to the object as its "essence".

I've been discussing essences with him for years; it seems to me he has some sort of reified view of essences; although sometimes he talks about them as if they are no more than language conventions. He might join us here again. — Banno

That's about it, essences are language conventions. But, as I've been arguing in this thread, we can only assume the existence of such conventions where they actually exist. In common usage there are no such conventions. So we cannot refer to "common usage" as a premise for logical procedure. We need agreement on firm definitions (essences) to proceed with logic. Therefore we must get beyond common usage to produce some agreement before we have any sound premises for logical process.

This is not in contrast to what I have suggested; but I think it is in contrast to the notion of an essence, immutable and eternal, which seems to be what Meta has in mind. — Banno

This is clear evidence that you didn't read my post, and that's why I dropped out of the discussion back then, it appeared like you only read the first sentence, and responded to that. What would be the point to me wasting my time?

Perhaps we have three views: Meta advocating essence as a real thing that we can set out in terms of the necessary and exclusive attributes; you, with some notion of an asymptotic essence that we can approach but never quite reach; and I, with the view that essences best ignored in favour of the examination of language use. — Banno

I take a broadly Peircean or systems view of essence. So it is real enough as the formal and final cause of being - the constraints or habits that shape material being. What's the great difficulty exactly?

Perhaps your misunderstanding is the reductionist one of thinking the essence of things is some mysterious substantial property hidden within - like a spirit stuff. Have you studied metaphysics much?

...I, with the view that essences best ignored in favour of the examination of language use. — Banno

The problem being that you refuse to examine how language is used in deductive logic. If you would get over this refusal, you would see that examining language use, and ignoring essences, presents a contradictory proposition.

"But I am talking about life and mind as a semiotic process where the hardware isn't deterministic. In fact, it mustn't be deterministic if that determinism is what the information processing side of the equation is hoping to supply." — Apokrisis

I simply cannot get away from the idea that the material instability you describe (providing a mechanism for information to express through) is actually deterministic causation expressing itself in a complex way which only gives the appearance of indeterminacy.

The "wobbly switch" is an interesting concept, but I view them foremost as "sensitive switches". There is some unreliability in these switches (and in channels of data transmission/expression), but at the very least they have consistent rates of failure. Although the pseudo-randomness of these unreliable switches can be incorporated into the functions of the data directly, (innovating new data through trial and error for instance (a happy failure of a set of switches)) at some level these switches must have some degree of reliability, else their suitability as a causal mechanism would be nonexistent.

Computers already do account for some degree of unreliability or wobbliness in their switches. They mainly use redundancy in data as a means to check and reconstruct bits that get corrupted. In machine learning individual "simulated neuronal units" may exhibit apparent wobbliness owing to the complexity of it's interconnected triggers or owing to a psudeo-random property of the neuron itself which can be used to produce variation.

Biological life is mechanism all the way down (in scale) is something which I fundamentally agree with, but the further down we go, the more different the mechanisms become (from the overall organism) until we reach points of abstraction which then test the limits of observation.

At the top of the hierarchy of networks of interactions would be an organism like a human. As we zoom into the body, multi-cellular structure gives way to single cells and inter-cellular mechsnisms, which then gives way to intracellular mechanisms, then to the mechanisms of DNA and RNA, and then to the molecular and atomic world.

The physical laws of atomic and molecular systems are what governs the behavior large complex molecules such as DNA which forms the base unit of data in the semiotic exchange that you're identifying. We could seek to understand individual genetic molecules as complex and emergent behavior resulting from physical laws of it's sub-mechanisms, but that doesn't add anything to the semiotic description of life in the same way that individual physical switches in computer memory do not add to a similar description of a hypothetical learning (and sentient) machine.

Consider the hierarchy of mechanisms found in biological life: DNA is it's base unit and all it's other structures and processes are built upon it using DNA as a primary dynamic element (above it in scale). A human mind with it's interconnected neurons uses the neurons themselves as a base unit upon which conscious thoughts are produced (through layers of complex connections and hierarchical networks no doubt). In a hypothetical truly sentient and learning computer, the acting base-unit which gives rise to semiotic exchanges isn't the binary bit, it's specific clusters of connected bits which produce the dynamic "simulated neuron" where all relevant data to the functions of the AI are actually stored. Like the human mind and also DNA, sentient machines have a "base unit of data" (apply charity here) which although itself is composed of smaller bits and smaller mechanisms, does not carry any semiotic meaning from the interactions of elements beneath it's own scale. The complex hierarchy of biological life that is built upon DNA is by far more stunning and complex than human consciousness or a true AI (body and all), but minds composed of hierarchical networks and layers of interconnected data are directly analogous to the physical hierarchy of mechanisms found in biological life which stun us so profoundly, wobbliness included.

I'm not sure where this puts our disagreement (if we indeed have one). I suppose my main difficulty is assenting to indeterminism as a property of living systems for semantic/etymological/dogmatic reasons, but I also cannot escape the conclusion that a powerful enough AI built from code (code analogous to DNA, and to the structure of connections in the human brain) would be capable of doing everything that "life" can do, including growing, reproducing, and evolving. Specifically the self-organizing property of data is what most interests me. Natural selection from chaos is the easy answer, the hard answer has to do with the complex shape and nature of connections, relationships, and interactions between data expressing mechanisms which give rise to anticipatory systems of hierarchical networks.

I suppose my main difficulty is assenting to indeterminism — VagabondSpectre

Indeteminism is a property of matter, as Heisenberg showed when he discovered the uncertainty principle. Nothing anywhere is fully determined or deterministic, I think the only motivation for believing that it is, is to escape the anguish of uncertainty.

I simply cannot get away from the idea that the material instability you describe (providing a mechanism for information to express through) is actually deterministic causation expressing itself in a complex way which only gives the appearance of indeterminacy. — VagabondSpectre

Well there are two levels to the issue here.

What I was highlighting was the surprising logic (for those used to expecting a biological requirement for hardware stability) that says in fact life requires its hardware to be fundamentally bistable - poised at the critical edge of coming together and falling apart. That way, semiotics - some message - can become the determining difference. Information can become the cause of thermal chaos becoming more efficiently organised as an organised dissipative flow.

So regardless of whether existence itself is "deterministic", biology may thrive because it can find physics poised in a state of radical instability, giving its stored information a way to be the actual determining cause for some process with an organised structure and persistent identity.

Then there is the question of whether existence itself is deterministic - or instead, perhaps, also a version of the same story. Maybe existence is pan-semiotic - dependent on the information that can organise its material criticality so that we have the Universe as a dissipative structure that flows down an entropic gradient with a persistent identity, running down the hill from a Big Bang to a Heat Death.

I realise that metaphysical determinism is an article of faith for many. It is part of the package of "good ideas" that underpins a modern reductionist view of physical existence. Determinism goes with locality, atomism, monadism, mechanicalism, logicism, the principle of sufficient reason. Every effect must have its cause - its efficient and material cause. So spontaneity, randomness, creativity, accident, chaos, organicism, etc, are all going to turn out to be disguised cases of micro-determinism. We are simply describing a history of events that is too complicated to follow in its detail using some macro-statistical level method of description.

So we all know the ontic story. At the micro-scale, everything is a succession of deterministic causes. The desired closure for causality is achieved simply by the efficient and material sources of change - the bottom-up forces of atomistic construction.

Now this is a great way of modelling the world - especially if you mostly want to use your knowledge of physics to build machines. But even physics shows how it runs into difficulties at the actual micro-scale - down there among the quantum nitty-gritty. What causes the atom to decay? Is it really some determining nudge or do we believe the strongly supported maths that says the "collapse of the wavefunction" is actually spontaneous or probabilistic?

So it is simply an empirical finding - that makes sense once you think about it - that life depends on the ability of information to colonise locations of material instability. Dissipative structure can be harnessed by encoded purpose, giving us the more complex phenomenon we call life (and mind).

And then determinism as an ontic-level supposition is also pretty challenged by the facts of quantum theory. That doesn't stop folk trying to shore up a belief in micro-determinism despite the patent interpretive problems. But there are better ontologies on the table - like Peircean pragmatism.

In brief, you can get a pretty deterministic looking world by understanding material being to be the hylomorphic conjunction of global (informational) constraints and local (material) freedoms.

So when some event looks mechanically determined, it could actually be just so highly constrained that its degrees of freedom or uncertainty are almost eliminated.

Think of a combustion engine. We confine a gas vapour explosion within some system of cylinders, valves, pistons, cranks, etc. Or a clock where a wound coiled spring is regulated by the tick-tock of a swivelling escapement. A machine can always just spontaneously go wrong. The clock could fall of the wall and smash. Your laptop might get some critical transistor fried by a cosmic ray. But if we are any good as designers - the people supplying the formal and final causes here - we can engineer the situation so that almost all sources of uncertainty are constrained to the point of practical elimination. A world that is 99% constrained, or whatever the margin required, is as good enough as ontically determined.

So that would be the argument for life. Molecular chemistry and thermodynamics doesn't have to be actually deterministic. It just has to be sufficiently constrained. The two things would still look much the same.

But there is an advantage in a constraints-based view of ontology - it still leaves room for actual spontaneity or accident or creative indeterminism. You don't have to pretend the world is so buttoned-down that the unexpected is impossible. You can have atoms that quantumly decay "for no particular reason" other than that this is a constant and unsuppressed possibility. You can have an ontology that better matches the world as we actually observe it - and makes better logical sense once you think about it.

Although the pseudo-randomness of these unreliable switches can be incorporated into the functions of the data directly, (innovating new data through trial and error for instance (a happy failure of a set of switches)) at some level these switches must have some degree of reliability, else their suitability as a causal mechanism would be nonexistent. — VagabondSpectre

See how hard you have to strain? Any randomness at the ground level has to be "psuedo". And then even that psuedo instability must be ironed out by levels and levels of determining mechanism.

But then why would life gravitate towards material instability or sources of flux? It fails logic to say life is there to supply the stabilising information if the instability is merely a bug and not the feature. If hardware stability is so important, life would have quickly evolved to colonise that instead.

My ontology is much simpler. Life's trick is that it can construct the informational constraints to exploit actual material instability. There is a reason why life happens. It can semiotically add mechanical constraints to organise entropic flows. It can regulate because there is a fundamental chaos or indeterminism in want of regulation.

Computers already do account for some degree of unreliability or wobbliness in their switches. They mainly use redundancy in data as a means to check and reconstruct bits that get corrupted. In machine learning individual "simulated neuronal units" may exhibit apparent wobbliness owing to the complexity of it's interconnected triggers or owing to a psudeo-random property of the neuron itself which can be used to produce variation. — VagabondSpectre

Yep. Computers are machines. We have to engineer them to remove all natural sources of instability. We don't want our laptop circuitry getting playful on us as it would quickly corrupt our data, crash our programs.

But biology is different. It lives in the real world and rides its fluxes. It takes the random and channels it for its own reasons.

You then get the irony of neural network architectures where you have fake instability being mastered by the constraint of repeatedly applied learning algorithms. The human designer seeds the network nodes with "random weights" and trains the system on some chosen data set. So yes, that is artificial life or artificial mind - based on pretend hardware indeterminism and so different in an ontologically fundamental way from a biology that lives by regulating real material/entropic indeterminism.

...which then gives way to intracellular mechanisms, then to the mechanisms of DNA and RNA, and then to the molecular and atomic world. — VagabondSpectre

But you went sideways to talk about DNA - the information - and skipped over the actual machinery of cells. And as I say, this is the big recent revolution - realising the metabolic half of the cellular equation is not some kind of chemical stewpot but instead a highly structured arrangement of machinery. And this machinery rides the nanoscale quasi-classical limit. It sits exactly at the scale that it can dip its toe in and out of quantum scale indeterminacy.

This is why I suggest Hoffman's Life's Ratchet as a read. It gives a graphic understanding of how the quasi-classical nanoscale is a zone of critical instability. You get something emergently new at this scale which is "wobbling" between the purely quantum and the purely classical.

So again, getting back to our standard ontological prejudices, we think that there are just two choices - either reality is classical (in the good old familiar deterministic Newtonian sense) or it is weirdly quantum (and who now knows how the fuck to interpret that?). But there is this third intermediate realm - now understood through thermodynamics and condensed matter modelling - that is the quasi-classical realm of being. And it has the precise quality of bistability - the switching between determinism and indeterminism, order and chaos - that life (and mind) only has to be able to harness and amplify.

It is a Goldilocks story. Between too stable and too unstable there is a physical zone where you can wobble back and forth in a way that you - as information, as an organism - can fruitfully choose.

So metaphysics has a third option now - which was sort of pointed to by chaos maths and condensed matter physics, but which is all too recent a scientific discovery to have reached the popular imagination as yet. (Well tipping points and fat-tails have in social science, but not what this means for biology or neuroscience.)

Consider the hierarchy of mechanisms found in biological life: DNA is it's base unit and all it's other structures and processes are built upon it using DNA as a primary dynamic element (above it in scale). — VagabondSpectre

This just sounds terribly antiquated. Read some current abiogenetic theorising and the focus has gone back to membrane structures organising entropic gradients as the basis of life. It is a molecular machinery first approach now. Although DNA or some other coding mechanism is pretty quickly needed to stabilise the existence of these precarious entropy-transacting membrane structures.

I suppose my main difficulty is assenting to indeterminism as a property of living systems for semantic/etymological/dogmatic reasons, but I also cannot escape the conclusion that a powerful enough AI built from code (code analogous to DNA, and to the structure of connections in the human brain) would be capable of doing everything that "life" can do, including growing, reproducing, and evolving. — VagabondSpectre

I do accept that we could construct an artificial world of some kind based on a foundation of mechanically-imposed determinism.

But my point is that this is not the same as being a semiotic organism riding the entropic gradients of the world to its own advantage.

So what you are imagining is a lifeform that exists inside the informational realm itself, not a lifeform that bridges a division where it is both the information that regulates, and the persistent entropic flux that materially eventuates.

My semiotic argument is life = information plus flux. And so life can't be just information isolated from flux (as is the case with a computer that doesn't have to worry about its power supply because its humans take care of sorting out that).

Now you can still construct this kind of life in an artificial, purely informational, world. But it fails in what does seem a critical part of the proper biological definition. There is some kind of analogy going on, but also a critical gap in terms of ontology. Which is why all the artificial-life/artificial-mind sci-fi hype sounds so over-blown. It is unconvincing when AI folk can't themselves spot the gaping chasm between the circuitry they hope non-entropically to scale up and the need in fact to entropically scale down to literally harness the nanoscale organicism of the world.

Computers don't need more parts to become more like natural organisms. They need to be able to tap the quasi-classical realm of being which is completely infected by the kind of radical instability they normally do their very best to avoid.

But why would we bother just re-building life? Technology is useful because it is technology - something different at a new semiotic level we can use as humans for our own purposes. So smart systems may be just smart machines ontically speaking, not alive or conscious, but that is not a reason in itself not to construct these machines that might exploit some biological analogies in a useful way, like DeepMind would claim to do.

Specifically the self-organizing property of data is what most interests me. Natural selection from chaos is the easy answer, the hard answer has to do with the complex shape and nature of connections, relationships, and interactions between data expressing mechanisms which give rise to anticipatory systems of hierarchical networks. — VagabondSpectre

As I say, biological design can serve as an engineering inspiration for better computer architectures. But that does not mean technology is moving towards biological life. And if that was not certain before, it is now that we understand the basis of life in terms of biophysics and dissipative structure theory.

you refuse to examine how language is used in deductive logic. — Metaphysician Undercover

Hu?

Hu? — Banno

English seems to have been now completely deducted from the statement as it first appeared. Curious. Perhaps English wasn't the language of logic after all?

But now we have to figure out what "hu" means in some private language. Guesses anyone? Could it be...

Hu (ḥw), in ancient Egypt, was the deification of the first word, the word of creation, that Atum was said to have exclaimed upon ejaculating or, alternatively, his self-castration, in his masturbatory act of creating the Ennead.

A masturbatory exclaimation? Well, quite possibly. So not hu? but hu! ;)

Where do you look, in order to determine that metabolism and replication are necessary and sufficient for life?

Presumably, at things that are alive.

It follows that you already know which things are alive before you set out this posited essence. — Banno

This works fine when we have essentially one instance of life and everything that evolved from it. But as Bitter Crank pointed out, we cannot wield our common usage intuition when we go to Mars and decide if something is life. A formal set of guidelines would really help, but it also must be flexible. Such guidelines are probably not forthcoming until we have several examples to compare (as opposed to the one we know now) and we have a reasonable data set from which common traits might begin to stand out. Who knows, it might turn out that we don't qualify ourselves.

Much of the critique here rolls around on my use of common. I meant it as shared, not mundane.

So there is nothing here to stop out common use of "life" being extended - indeed, I have several times explicitly said that definitions can be extended.

Not worthy of you, Apo.

Can you explain what Meta meant? Is he just claiming that deductive logic relies on explicit definitions? What is it that I am missing?

O:)

This is exactly why we need to find the essence of a living being! With this precedent, what is to stop anyone else from granting human rights to any other objects? Note, I am all for protecting nature, just not to the level of granting human rights.

So there is nothing here to stop out common use of "life" being extended - indeed, I have several times explicitly said that definitions can be extended. — Banno

Yeah. Except rather than extended, they need to be differentiated. And so they can no longer be shared - being a new choice.

This goes back to what seems your fundamental misunderstanding about language use. A word does not have a definitional essence in as ostensive sense. It instead functions as an apophatic constraint on uncertainty.

So a word like "life" or "cat" is already extended in that it covers anything even vaguely living or catlike. The word, as a sign, points not at some definite collection of particular instances and nothing beyond that. It instead constrains our understanding in some generalised way that could be cashed out in any number of restricted sense. Many of which will be differences that don't make a difference.

So if I am talking about some cat, it could be large all small, black or tabby, male or female, etc, unless I feel the need to specify otherwise, adding more words and thus more constraints on your state of interpretation.

Thus there is some essence in play - the purpose that is my communicative content (as much as that is ever completely clear and not vague to oneself, even in some propositional statement). But the word can't carry some exact cargo of meaning from me to you. All we share is some history of learning to have our uncertain interpretations constrained to be near enough similar while still remaining creatively open-ended.

The advantage of my semiotic or constraints approach is that it accounts for how meaning can be formed and conveyed without something specific, particular or actual existing by way of a referent. I might actually have in mind a black, male moggy. You might have in mind a tabby female. But it doesn't make a difference until it makes a difference.

And that view has important consequences for truth theories, among other things. It also should explain why definitions matter as the way to bring out putative differences. We can't actually be agreeing in some positive fashion - as opposed to some accidental and undisclosed fashion - unless we have discovered and articulated a possible point of disagreement.

This cat we are talking about - what's its colour, age and gender? Let's see if we still have the same referent in mind. And if it doesn't matter, then it is not essential. The essence remains at the greater level of generality which is simply what we mean by "cat".

So the existence of essence is demonstrated by applicability of generality. Reference can be open-ended or "already extended" because - dichotomously - it is also anchored by an apophatic generality. We know that cats aren't dogs or fungi or rocket-launchers as those other general alternatives are ruled out by some abstracted cat-essence.

And while common usage does seem to get catness by perceptual abstraction (some acceptable combination of traits), science can pin that down with greater ontological rigour. It can say that evolution actually does create genetic lineages - actual constraints encoded in genetic information. So we can start to measure cat-ness in a way that can be quantified as some distance separating cats and pumas, and then more generally, leopards and panthers (although confusingly - hu! - leopards are phylogenetically panthera rather than leopardis).

It is actually very important that - from Aristotle on - we seek to name the forms of the natural world in this essentialist fashion. The subsumptive hierarchy always seemed completely logical. And so it was discovered to be. Evolution reads like a forking tree of differences that made a difference because something must divide one species from another at the level of actual historical information. We don't just socially construct the meanings of words. We can hope to asymptotically approach the world's essential divisions by seeking out the constraints that got refined by differences that made a difference.

What about the predictability of the moon's orbit around the earth?

At the Newtonian scale, how can indeterminacy be observed?

Remember LaPlace's Daemon?

We may regard the present state of the universe as the effect of its past and the cause of its future. An intellect which at a certain moment would know all forces that set nature in motion, and all positions of all items of which nature is composed, if this intellect were also vast enough to submit these data to analysis, it would embrace in a single formula the movements of the greatest bodies of the universe and those of the tiniest atom; for such an intellect nothing would be uncertain and the future just like the past would be present before its eyes.

I think this kind of thinking has had a disproportiantely large influence on post-Enlightenment thought. A lot of people still think like that - but Heisenberg (et al) showed that at the most fundamental level, it simply isn't so. Uncertainty and the probablistic nature of physics really does torpedo that. (That was subject of a lot of philosophical debate in the early 20th century by the likes of Arthur Eddington, Enst Cassirer, not to forget Heisenberg himself, and also Niels Bohr (who incorporated the ying-yang symbol into the family Coat of Arms.))

That is what allows for the element of creativity, of serendipity, of things that just happen for no apparent reason. Whereas Western culture seems to retain a belief that at bottom, what is real are 'bodies in motion' that are determined by physical forces. But physics itself has shown that really, in the memorable phrase by James Jeans, 'the universe is more like a great mind than a great machine'.

This goes back to what seems your fundamental misunderstanding about language use. A word does not have a definitional essence in as ostensive sense. It instead functions as an apophatic constraint on uncertainty. — apokrisis

Stop there. I don't disagree with that. Of course words do not have an "essence in an ostensive sense". While I don't much like your use of "apophatic constraint', I don't see that what you are suggesting differs in any substantive way from what I have proposed.

What you call a constraint on a definition I would describe as an additional term, changing the application.

So quolls are referred to as tiger cats. They are marsupials. We had one a year ago that would come once a month and have takeaway chicken, curtesy of my coop. When the Girl said things like "That cat took another chook last night", the meaning was clear.

But one might add to the definition of cat "...and is not a marsupial", thus ruling out the use of "cat" to refer to quolls.

Sure that "apophatic constraint" works for certain purposes, but it rules out a useful way to use the word "cat"; it would be improper to say that one use was "the correct use of cat".

There is no essence of cat here; only different uses.

Thus there is some essence in play - the purpose that is my communicative content (as much as that is ever completely clear and not vague to oneself, even in some propositional statement). — apokrisis

I gather from the parenthetic comment that you are yourself not too happy with this terminology.

So let me prod that a bit: What is the "communicative content", apart from the utility of the conversation?

What you call a constraint on a definition I would describe as an additional term, changing the application. — Banno

I prefer my precise terminology. It makes it clear that adding constraints is the subtraction of possibilities. We are talking about the intersection of sets, not the union of sets - if one must resort to set theoretic talk.

Your way of putting things is ambiguous. The change could be either logical-or or logical-and.

So quolls are referred to as tiger cats. They are marsupials. We had one a year ago that would come once a month and have takeaway chicken, curtesy of my coop. When the Girl said things like "That cat took another chook last night", the meaning was clear.

But one might add to the definition of cat "...and is not a marsupial", thus ruling out the use of "cat" to refer to quolls.

Sure that "apophatic constraint" works for certain purposes, but it rules out a useful way to use the word "cat"; it would be improper to say that one use was "the correct use of cat".

There is no essence of cat here; only different uses. — Banno

Cute story but full of holes. Just look how fast you slid from "tiger cat" - a common colonial term - to "cat". So quoll might equal tiger cat as a valid translation between mispronounced aborigine and settler coinage. Both would point at the same animal. But to call a quoll or tiger cat a cat is another whole can of worms.

The quoll is "sort of like a cat, but not really". We would have to be appealing to some more general notion of the essence of catness to create a union of two sets of observations. So rather than getting more precise - adding constraints to produce an intersection - we would be relaxing constraints to produce a union at a higher level of generality. It is the more abstracted essence of catness that we must have in mind to justify this turn of speech.

So sure, the correct use of "cat" is flexible. We can step back to higher generality in a way that allows union operations - hey, quolls are rather cat-like in look and habit (or more like cats than rabbits, goats, chickens, and other animals we know from our homeland). Or we can add constraints to do the opposite. We can talk about all the cats that are also marsupials - and find the intersection is in fact the null set.

Language is great because it doesn't get too caught up in levels of generality and particularity. Although it does of course employ pronouns and qualifiers (like -like and -ness and -icity) to add this logical distinction as necessary.

But still, the Peircean approach does see the metaphysical essence of things as speaking to their formal and final cause. What unifies particulars is their purpose and rational organisation. So quolls would be like cats because the same body form is good for the same purpose, the same ecological niche. There actually is something in common that we might want to capture as a general X-ness. The needs of a small nocturnal carnivore is a constraint that acts on the genetics of both.

So "apophatic constraint" doesn't in fact rule out the creative use of language. Instead it underpins it. And this is how I know you don't actually get it. It is only this kind language use that remains open-ended even when constraints are combined. Constraints merely limit proper interpretation.

If we are talking about black cats, we might still be speaking of Miles Davis. "Black" and "cat" can have a whole host of associated meanings according to the communicative context. This essential open-endedness of a sign is not a problem unless you are wedded to a clunky set theoretic view of meaning where words must refer to some definite collection of things. Constraints can only reduce uncertainty, they don't ever eliminate it. That is why Peircean logic employs vagueness as a modality. It explains the inherent flexibility with which even the strictest syntax determines meaning. Semantics is irreducibly open-ended - yet also perfectly ameniable to being apophatically bounded.

I gather from the parenthetic comment that you are yourself not too happy with this terminology. — Banno

There was a spelling mistake there. I meant communicative intent and not communicative content.

So again this relates to the Peircean view that essence is final cause or the purpose that shapes things. And the parenthetical point was the positive assertion that even speakers may be vaguer than their rather definite sounding speech acts imply.

Speech is a creative act and syntax imposes apophatic constraint. We simply have to eliminate a lot of possible qualifications and hesitations we might have in mind to actually say something out loud in a communally acceptable fashion. And in contrary fashion, stating something aloud gives a proposition a crispness that may suddenly make us feel we are thinking with wonderful clarity now. We got our meaning exactly right. We were vague, but now we are not. Our intent is clear to us too because of the way grammar eliminates imprecision ... apparently.

The change could be either logical-or or logical-and. — apokrisis

Indeed; that is quite on purpose, so that it corresponds to a conjunctive or disjunctive normal form - as any definition must. But of course we can specify the OR- or the AND-, so need be no ambiguity: AND NOT marsupial.

An interesting post.

First off, as a curtesy, let's avoid phrases like "...you don't actually get it"; To my eye, you don't actually get it, either. That's rather the point of having a conversation.

Secondly, amI right to detect a normative approach to language here? For you, is there a correct way to use words? Is seeking more and more constraint a good thing?

But we are making progress. So essence is final cause or the purpose that shapes things. It should be fun to fill that in a bit. Purpose for whom? The speaker, or their communicative intent? The community? Some transcendent mind?

English seems to have been now completely deducted from the statement as it first appeared. Curious. Perhaps English wasn't the language of logic after all? — apokrisis

Yeah, I got that in my notification. The fact is, that Banno's claimed "examination of language use" is very selective. There is a refusal to examine how language is used in deductive logic, where unambiguous definitions (essences) must be adhered to, or else the logic is rendered useless. Banno just insists that we should refer to "common" use.

It is possible that we might examine numerous instances of "common" use, in the method of Platonic dialectics, to determine if they all have something in common. From this we could synthesize a definition (essence).

Can you explain what Meta meant? Is he just claiming that deductive logic relies on explicit definitions? — Banno

Correct, and doesn't a definition give one the essence of the thing? You may have an aversion to the word "essence", but your "common use", which is full of ambiguities and equivocations is quite useless for any deductive logic.

But there is an advantage in a constraints-based view of ontology - it still leaves room for actual spontaneity or accident or creative indeterminism. You don't have to pretend the world is so buttoned-down that the unexpected is impossible. You can have atoms that quantumly decay "for no particular reason" other than that this is a constant and unsuppressed possibility. You can have an ontology that better matches the world as we actually observe it - and makes better logical sense once you think about it...

...See how hard you have to strain? Any randomness at the ground level has to be "psuedo". And then even that psuedo instability must be ironed out by levels and levels of determining mechanism...

...My ontology is much simpler. Life's trick is that it can construct the informational constraints to exploit actual material instability. There is a reason why life happens. It can semiotically add mechanical constraints to organise entropic flows. It can regulate because there is a fundamental chaos or indeterminism in want of regulation. — apokrisis

Where information begins to regulate chaos is the semiotic ground level (how that base dataset first emerges aside), but there's no difference in practice between "pseudo-randomness" and "actual randomness". Indeterminism as a property of a system is indistinguishable from looking at an open system without understanding the nature of the energy which flows into or out of it. Given that life is an open system, and that the dissipative structures to which you allude depend on an influx of energy (in order to resist the second law of thermodynamics), where does hard indeterminism actually benefit the model?

Dissipative structures and engines are great for modeling the energetic batteries of life which keep the complex structures in shape and interacting, but the self-organizing property of the data itself is what most fundamentally interests and astounds me. Dissipative engines and structures which hold back life from reaching states of thermodynamic equilibrium are required for it to perpetuate, but they aren't the fundamental processes I'm interested in. The electron transport train is what keeps life warm so to speak, but the self-organizing property of life's data goes beyond that to provide innovative direction well beyond mere random variance. True or pseudo randomness may be involved at some base levels, but as you say, tiers of informational networks constrain such uncertainty (or incorporate it at a base level) and refine it into something not random, but anticipatory.

Life as a dissipative system is a useful description that helps us understand why we don't breakdown in entropic heat-death, but it doesn't do a whole lot to help us understand the complex organizational structures (the data) which yes exploit disspitative engines, but then use the energy they get from them to expand and increase their own sophistication and anticipatory power.

...But my point is that this is not the same as being a semiotic organism riding the entropic gradients of the world to its own advantage....

...My semiotic argument is life = information plus flux. And so life can't be just information isolated from flux (as is the case with a computer that doesn't have to worry about its power supply because its humans take care of sorting out that)....

...Now you can still construct this kind of life in an artificial, purely informational, world. But it fails in what does seem a critical part of the proper biological definition. There is some kind of analogy going on, but also a critical gap in terms of ontology. Which is why all the artificial-life/artificial-mind sci-fi hype sounds so over-blown. It is unconvincing when AI folk can't themselves spot the gaping chasm between the circuitry they hope non-entropically to scale up and the need in fact to entropically scale down to literally harness the nanoscale organicism of the world....

...As I say, biological design can serve as an engineering inspiration for better computer architectures. But that does not mean technology is moving towards biological life. And if that was not certain before, it is now that we understand the basis of life in terms of biophysics and dissipative structure theory.... — apokrisis

If we boil this down, life is self-organizing information (and consumes energy to do it, and so requires abundance of fuel). If we forget about external power input as a part of AI as an organism, or we include power generation (and self-maintenance) as a function of a sufficiently advanced computer-AI organism, then it would also seem to fully qualify as a dissipative system. It's electron transport chains are much more uniform (where ours incorporate peculiarly designed (naturally evolved) bucket lines of diverse carriers), but the feat of informational self-organization which these electrons fuel is the same kind of complexity which I thrust as most central to the behavior of life.

Most of the complexity in such a computer AI organism is contained in the structure of it's data networks (and it's inputs/outputs), where the complexity of biological life is spread out across a diverse ladder of interacting parts (which can also be considered a part of/expression of data contained in DNA). Learning digital information networks are also physical structures which give rise to physical complexity that can rival the complexity found in nano-scale biological machinery. Even though it all exists materially as stored charges (what we abstract as bits), the connections and relationships between these parts can grow in complexity by more efficiently utilizing and ordering it's bits rather than by acquiring more of them (although more bits doesn't hurt).

We don't have an AI yet capable of taking control over it's own existence (in the way that biological life does as a means of perpetuation), but I think that chasm is shrinking faster than most people realize. Simulations aren't real in the sense that their products are abstract representations of real things, but what happens when the simulation looks back at you, starts asking questions of it's own, and starts thinking? It's not a simulated thought, it's that self same property of information/informational structure somehow giving rise to spontaneous anticipatory sophistication we typify as intelligence. It cannot be explained solely as a dissipative system or as the direct mechanisms through which information expresses (wobbly switches), it is rather a feature of the structure and content of the data itself, like a pattern which builds upon itself to achieve greater depths of complexity (and function).

What I'm interested in specifically is how the structure of information guides it's own development toward more sophisticated and anticipatory formats. Evolution through successive trial and error is a major help toward understanding where the DNA and other biological mechanisms of biological life get their anticipatory power, and it can perhaps be useful for understanding how a mind emerges from neurons (both biological and digital minds), but it doesn't quite shed any detail on what these regimes and structures actually look like.

I think this kind of thinking has had a disproportiantely large influence on post-Enlightenment thought. A lot of people still think like that - but Heisenberg (et al) showed that at the most fundamental level, it simply isn't so. Uncertainty and the probablistic nature of physics really does torpedo that. (That was subject of a lot of philosophical debate in the early 20th century by the likes of Arthur Eddington, Enst Cassirer, not to forget Heisenberg himself, and also Niels Bohr (who incorporated the ying-yang symbol into the family Coat of Arms.))

That is what allows for the element of creativity, of serendipity, of things that just happen for no apparent reason. Whereas Western culture seems to retain a belief that at bottom, what is real are 'bodies in motion' that are determined by physical forces. But physics itself has shown that really, in the memorable phrase by James Jeans, 'the universe is more like a great mind than a great machine'. — Wayfarer

I think that to us humans there will always be de facto indeterminacy primarily due to limits of measurability and the physical horizons of knowledge they imply. Heisenberg didn't prove indeterminacy, he proved uncertainty inherent in certain kinds of measurements.

If the true quantum realm cannot be physically accessed, then any of it's internal goings on might as well be their own separate universe, and all we see are the peculiar phenomenon that those goings on manifest as at larger scales and groupings of matter (I.E, particle-wave behavior and the collapse of it's wave-like properties).

Indeterminism is an epistemic reality of human knowledge, but not necessarily a property of the physical world.