Are Logical Rules Universally Applicable?

[computerphys]

IMO, logic prevails universaly because it is better for survival, better for existence - it makes all work, it enables co-existence of countless of forms, from atoms to stars, from bacteria to humans, and beyond.

Yes, I agree with you. This is exactly the only pseudo-reason I've found to support a logical world against a non-logical one. But I think it is not a powerful reason, because it is very anthropocentric. Logic must exist for me to be able to understand a meaningful Universe.

If someone says to me that there exists a parallel universe where everything is chaos and no logical rules are fulfilled, I find no arguments against that idea. Nevertheless I feel a powerful rejection inside me to that point. I cannot use the anthropocentric argument against that point, because I don't live in that universe.

So logic could be better for existence, etc, but that doesn't imply it must be really universal.

If logic is not universal, that is, it coalesces along time or exists just in certain sub-universes, then, mathematics, physics and philosophy are also non-universal. Is that correct?

 

[apeiron]

Not making a choice is a way of saying you believe in fuzzy logic. Because the only way of accepting both logical systems in a compatible way is denying the contradiction principle, and that belongs to fuzzy logic.

Fuzzy logic is really just a subset of ordinary logic in my book. With vagueness, there is just pure indeterminancy rather than a statistically bounded uncertainty.

With fuzziness, you have a distribution of uncertainty about a set of microstates, but the macrostate - the global constraint - can be exactly specified. With vagueness, both micro and macro states would be undetermined.

We are not talking about a haziness or fuzziness but the most radical possible, maximal, kind of uncertainty or indeterminancy.

Well, at least I am. I agree that especially where people are talking about semantic vagueness (sorites paradox), they are thinking of fuzzy definitions with a gaussian distribution. I am talking here about a logic based on ontic vagueness where there is no boundary at all on uncertainty.

And indeed, the characteristic uncertainty that arises as the result of a development of a vagueness into some form of self-organising world is the pareto or levy distribution - we are talking powerlaws. This should be no surprise of course as self-organising development IS characterised by powerlaw statistics. And this is a logic that explains that dynamical world.

Not always. Descartes' cogito ergo sum is an ultimately knowable and absolute truth.

True. And therefore the reason why in the end I am agreeing that everything else becomes a matter of faith. The belief that we are modelling something "out there" for example. So I claim that models are justified by their utility (control over the world) and their confirmation via observation and measurement. But that in turn has to rely on my faith my impressions of a world are more than some private illusion.

Of course, in practice, I don't doubt the existence of reality, and so the need for good models, in the slightest. Cartesean doubt, while admitted, quickly becomes irrelevant.

Ok. "Shut up and calculate!‎" What stuff can be done using fuzzy logic that couldn't be done just using standard logic?

It is not fuzzy logic. It is a systems logic. And I use it for reasoning constantly. I analyse the world through its lens (much to the annoyance of almost everyone it seems ). So just read any post where I start talking about dichotomies that lead to hierarchies.

Vagueness is my story on initial conditions. Asymmetric dichotomies are my description of how initial conditions must develop to actually become something. Hierarchies are what then emerge as the persisting equilibrium outcome which results.

This is the same logical structure you will find in Anaximander's metaphysics. Aristotle's logic also covers it (though Aristotle was wrestling to separate the mechanical logic we now know from this more ancient view). Hegel had a version, though he ended with a spiral rather than a hierarchy. Peirce articulated almost the whole story but missed the vital scale aspect to it - the local~global asymmetry which is central to hierarchy theory.

Currently, I would say there are many people developing bits of the total package - feeling different parts of the same animal. So some work on vagueness, some on dichotomies, many on hierarchies. There has also been a big upsurge of interest in Peirce, which is where you will find the broadest understanding.

So this is work in progress right now.

 

[apeiron]

Also, if we say that in the very beginning of this Universe (around Plank's time) things weren't logical, nor illogical, then the fact that Universe today appears logical (via firm Universal natural laws), then this is easy to understand, that that which was illogical ceased to be and logic prevailed.

This is a good point. We are rather stretching the word "logic" here - most would treat it as essentially a mathematical notion and this thread has really been about naturalising logic, seeing it in terms of actual physical processes. But I would call the early big bang vague, and the eventual heat death of the universe crisp. That is when everything was small (and so very hot) the state of anything was radically indeterminate (as QM models). Then when everything becomes as spread out and cold as possible, the state of the universe will be as definite everywhere as it can be (dark, featureless, sitting asymptotically close to absolute zero).

So at heat death, the universe will be at its "most logical" in the sense that it will have a single prevailing global state. The second law says entropy must be maximised. And by the heat death, that becomes a certain fact.

Right now of course, the universe is still patchy and has little bits of negentropic order all over the place - especially here on earth. Life and mind are "illogical" so far as the second law is concerned. Well, that is until you step back and see that negentropy exists because it accelerates entropification - so that makes life and mind again logical. Just so long as life and mind are a passing phase on the way to the proper destiny of a heat death.

 

[computerphys]

With fuzziness, you have a distribution of uncertainty about a set of microstates, but the macrostate - the global constraint - can be exactly specified. With vagueness, both micro and macro states would be undetermined.

I find this fuzziness and vagueness quite useful to describe mind states. I see that when you build an idea inside your mind, you first have vagueness, and later fuzziness, and been very lucky you may eventually reach a crispy state of knowledge about something.

What I cannot understand is how to apply this on a physical system ruled by laws of physics and more deeply by crispy logical rules.

So logic can be built inside our mind, from vagueness to crispness but not as a physical process based on rules. That final state, when our mind has concluded logical rules, I wonder about its universality. I mean, everybody under any circumstances would end up with the same logical system than Aristotle's. Did every human civilization share always the same logical system? If it is really universal, there should be a good reason for that. Probably it is the most basic possible set of self-compatible rules or something like that. In that case, mathematics would also be universal, isn't it?

 

[apeiron]

I find this fuzziness and vagueness quite useful to describe mind states.

So do I. Indeed, I was doing mind science and it was because a machine-style logic was clearly not up to the task of modelling the mind that I eventually came across this organic, developmental approach to logic.

Note the way Peirce reverses the usual approach to logic. He started with a description of how minds develop ideas (semiosis) and then generalised it to a description of how even universes develop (so instead of the universe developing laws, he talks about it developing habits - he uses psychological terms. Which is justified by the cogito - starting explanation from what we can be most sure about).

I mean, everybody under any circumstances would end up with the same logical system than Aristotle's. Did every human civilization share always the same logical system? If it is really universal, there should be a good reason for that. Probably it is the most basic possible set of self-compatible rules or something like that. In that case, mathematics would also be universal, isn't it?

If you check all major early philosophy, then you find that they have a similar organic approach (a vagueness that dichotomises and becomes hierarchically organised). So you see in the first greek metaphysics, and also Buddhist and Taoist philosophy.

For instance, Chris Lofting has looked at a dichotomising logic in the I Ching (which is not just some fortune telling nonsense) http://members.iimetro.com.au/~lofting/myweb/

But there are "two logics" even in the human mind.

The organic logic is how brains themselves are organised. The brain is a bunch of dichotomies leading to hierarchies. To explain the brain, we talk about attention vs habits, the ventral "what" or object recognition stream vs the dorsal "where" or spatial awareness stream, the left hemisphere focal style of processing vs the right hemisphere contextual style of processing, dopamine endogenous goal-directed attention vs norepinephrine exogenous vigilant attention...and so on and on.

But then humans also have speech. And this is based on a more machine like logic. Instead of holistic causality, there is a step by step, chain of cause and effect, style causality.

All languages are based on the same logical structure - subject-verb-object, a tale of who did what to whom. The serial nature of the speech act forces us to break the wholeness of the world up into an utterance which is based solely on effective cause (only one of Aristotle's four causes). If this, then that.

So you have an interesting situation. Our minds work organically and yet the habit of speech also means that we are used to analysing reality in mechanical terms.

The holistic view of the world is the truer - seeing the world as a whole, a web of self-organising interactions.

But the mechanical approach, while more constrained (employing only efficient cause), has its own special power because it is combinatorial and open-ended. The subject-verb-object template of a sentence is a logical statement, and it can be used to state something false as well as true.

A holistic impression of the world is restricted to being true because it is self-constraining. It does not have the freedom to be false as a sum over a collection of interactions has only one outcome, one equilibrium balance. But a logical statement is an isolate act and so can make a leap, a postulate, hypothesis or abduction, in any direction. I can say the moon is made of green cheese - the seven words are an effortless assertion. Whereas it would be very hard for me to gain an impression the moon is made of green cheese unless it actually was.

So anyway, the human mind does in fact employ two logics - the holistic logic of a mind that perceives the whole of the world about it, and then the verbal logic which humans employ to reach beyond this immediate, constrained, state of experiencing to construct a realm of the postulated, the hypothetical.

One logic has become "logic" for most people. A logic based solely on efficient cause (if this, then that). A logic in which various formal constraints (the principle of the excluded third, the principle of contradiction) are taken as granted.

But people have forgotten about (or never actually understood) that there is also then the larger embedding logic of holism, the way the mind understands reality as a whole. And this should now itself be a focus for formalisation. We can construct a model of organic logic. Aristotle indeed did this - he was separating the two kinds of logic more clearly. Peirce took the project up again in modern times. In the 1960s and 1970s, science got quite close to doing so again with systems science, hierarchy theory, ecology.

The problem is that over the past 40 years, we have had the computer revolution, the information theoretic age. Machine logic has become so culturally dominant that organic approaches are these days treated as some kind of intellectual heresy. There is only now the machine approach - efficient cause. Holism is not even tolerated.

Again, these two logics are in fact both useful. Indeed, they are complementary, they are "part of each other".

But the proper relationship is {organic {mechanical}} and not mechanical => organic.

That is, the presumption of those who think that everything can be reduced to efficient cause, machine logic, is that anything organic, anything complex, can be constructed. So we have people trying to build artificial intelligence, artificial life, using mechanical principles. It is presumed that the organic can always be "built". And so the mechanical is what is foundational.

But I am arguing that the organic is the whole story, the mechanical is the local story. The organic is Aristotle's four causes, the mechanical is just his efficient cause. So if we are relating to two logics, one is the subset of the larger.

{final, formal and material cause {efficient cause}}

 

[Boy@n]

We can construct a model of organic logic.

So, could we put this model into a program/computer? Would this enable computers to become self-aware? If not, what is that which a computer lacks to be self-aware since it has almost perfect logic and memory? Also, how could we make a computer/robot to really feel (like humans and all living organisms do) and not just pretend to feel?

In the 1960s and 1970s, science got quite close to doing so again with systems science, hierarchy theory, ecology.

How far are we today in this field?

 

[apeiron]

So, could we put this model into a program/computer? Would this enable computers to become self-aware? If not, what is that which a computer lacks to be self-aware since it has almost perfect logic and memory? Also, how could we make a computer/robot to really feel (like humans and all living organisms do) and not just pretend to feel?

There are too many levels to this question to answer simply. Are we talking implementing in hardware or emulating in software, are we talking models or simulations, etc.

But if we are talking about mind science approaches that are based on the essential principles, then generative neural networks in the style of Stephen Grossberg are my idea of what has been closest.

However I wouldn't hold my breath for artificial intelligence. It is so easy to grow real intelligence, why would you even bother anyway? People are cheap to make.

How far are we today in this field?

I think it says something that all the people who seem to understand the best are in their 70s and emeritus now.

But there is some good new maths tools being developed such as scalefree networks and new approaches to entropy.

It is an interesting question. It could be said systems science really took off in the 1960s because of the idea of cybernetic feedback and its direct technical application in things like missile guidance systems. There is nothing like a good military application to make science immediately credible.

So if say neural networks were to suddenly take off (because Turing computation was running out of steam) then a more complex systems approach to logic might also take off. But regular computers are still on their Moore's curve of development. So there is nothing actually driving people to think about the world differently at the moment.