Consciousness and the World

Paulh,

That must be a most interesting book. However, I wonder why the brain could not have or at least partly utilize a PDP architecture, and also operate as an edge-of-chaos dynamical system? Parallel distributed processing is a data processing architecture; chaos is a state-change effect frequently encountered in complex recursive systems whereby system behavior moves from steady-state to extremely complex patterns with increasingly random elements; and dynamics are the opposite of statics. From what I understand of data processing and massively parallel architectures (to which I am admittedly not an expert), there certainly is much reason to consider them highly dynamic; and given the many feedback loops and high complexity of a PDP-based human brain, there is every reason to suspect that the PDP brain would 'live on the edge of chaos', despite PDP's inherent mission as a pattern recognizer.

From what I've read, the brain probably isn't a unified PDP net; it certainly is compartmentalized into semi-specialized components that have higher-level communication circuits and participate in a variety of complex interaction mechanisms. Some of those components may be purely PDP based, others may be more Von Neumann - like symbol processors, still others may used mixed designs, e.g. symbol processors being emulated on a PDP base; and some may use a processing architecture not yet fully known, for all that I know (Clark was very careful in Microcognition to be open to this possibility).

Still, from what I've read thus far, it seems a fairly good bet that at least parts of the brain use a massively parallel architecture with pattern recognition and data abstraction capacities which are consistent with what is currently known from psychology regarding human sense-input / behavioral-output characteristics, and the reported mental states between these.

And there is no reason I could think of (again with apologies for not being an expert) that a PDP architecture, especially in the context of compartmentalization and higher level organization between compartments, would prevent the overall brain from being "always near chaos"; and sometimes even going over the edge, when input conditions go too far from what we were designed by natural selection to handle. E.g., when too much stress causes us to act erratically and unexpectedly, to have delusions, to have seizures, etc.

As to taking away the brain and the body and the world, causing the 'virtual machine' to stop; sorry, I must admit my ignorance here regarding what your point would be, beyond the fact that if the brain is disabled, or if the body is destroyed, or if the world disappears, the info processing activity that the environment / body / brain system supports would all be expected to come to a halt and be permanently obliterated, to the best of our experience and empirical knowledge regarding the physical world at this point in human history. Which, as science itself occasionally reminds us, is far from complete and ultimately reliable. Although I will certainly stipulate that a virtual computational machine such as the word processing application on your home computer will reliably come to a halt and have no further relevance if you destroy the motherboard and memory devices; ditto for the virtual cognition machines in human brains that allow thinking.

With thanks to all on this forum for the continuing banquet of food for thought, given that my own motherboard and memory devices and virtual machines still seem to be functioning, however intermittently,

Jim G.

... my own leanings at this point in time go slightly beyond that ... yes, connectionists have demonstrated remarkable achievements in pattern recognition - but these nets are engineered, and so even though they are examples of PDP they are heavily influenced by linear thinking ... rather than saying that the brain 'lives on the edge of chaos' I would hazard to say instead that the brain 'computes on the edge of chaos' - that human thought is more akin to a system actively jumping from chaotic attractor to chaotic attractor than it is a pattern recognizer ... from that perspective, human thought is qualitatively similar to the genetic network that computes a human body; human thought is qualitatively similar to the behavioral network of insects that computes the shortest path to a food source; human thought is qualitatively similar to the auto-catalytic chemical network that computes life ...



... I'm thinking more along the lines of the classic "brain in a vat" thought experiment ... if you took a human brain and put it in a vat and only provided it with the bare essentials of life support, would it be capable of consciousness? ... and if not, what sorts of inputs/outputs would be required to "scaffold" consciousness? a keyboard? a mouse? ... or a full-blown simulation of a human body embedded in the world?

eternalstudent

I am a fraid I have pulled a frightful error, in trying to respond to your post to that Byron quote, first my response came up under your name and then when I tried to remove it, I am fraid I deleted the entire post. I am new at this you might have gathered, at anyrate you have my apology, I fear that post is in cyberspace somewhere. If your interested, here is the intended response to your post.

Jim G,

The quote it is true has the individual in mind, when you speak of biological extensions/ scaffolding, forming culture you are speaking of the benifits to humanity in general, not really personal/ as of the individual. I would think however that an individual would have difficutly living life through an entirely secondary means--as often prisoners do. All in all I think Byron did pretty good for a brief oneliner. If you expect a oneliner to express an entire philosophy, you will often be disappointed. Again my apology!!!

"whereby system behavior moves from steady-state to extremely complex patterns with increasingly random elements"

What is a random element? I thought randomness was about perception and cognition abilities.

Though naturally there is no pure randomness I can't see complexity adding to randomness in any way, just requiring a better understanding of the system to see that the randomness hasn't at all been influenced.

... yes ... deterministic chaos is, well, deterministic ... it only has the appearance of randomness because highly similar initial conditions typically lead to wildly different future states - so if you can't get an exact measurement of a chaotic system's initial conditions (which in theory you never can, let alone in practice), predictions of the system's future state diverge fairly rapidly from the actual system's future state (prediction error grows exponentially) ... you have to constantly recalibrate by taking new measurements and using those as new initial conditions for short-term future predictions ... it's like predicting the weather ... wait a minute - it is predicting the weather!!! :bigsmile: ...

I guess I have a problem with such a thing as deterministic chaos. I seem to have defined chaos as something absolute; a word for pure randomness, or at least very close to it.

Deterministic chaos implies there be knowledgeable chance of what the future holds in the system, which I can't see being the system causing consciousness; just as there can be no pure randomness.

Randomness implies infinite possibilities otherwise chance can be derived from it, right? Pure randomness is not just the inability to know the odds but also the inability to know the potentials so as to have the arrow in the first place (and we can say the arrow simply doesn't point to aything in particular, pretty obvious), but also for irrational chance, at least that would seem more appropriate anyways.

But since consciousness allows for perception of randomness and order and the cosmos are infinite, reality must rely on consciousness to exist and actuality as well, unless it is pure randomness, or pure order (being the same thing), thus one dimensional (which is the essence of actuality when you think about it).

Otherwise one of the premises are wrong. Either pure randomness exists, which blatantly we can say if potential exists pure randomness cannot; or the cosmos are not infinite.

But perhaps we can arrive at a mutual conclusion. Maybe there are divisions in the universe or cosmos that potential is completely untangled from one another.

I'd better stop, I'm becoming incoherent again. :depressed:

... no worries ... when you get to be my age and I'm off in a nursing home somewhere, I'm sure you and the rest of the philosophic community will be entertaining thoughts that I can't even dream of ... you lucky dogs!

Due to my lack of coherence I'm just a latent potential, at least you are a self styled super genius.

self-styled (slfstld) adj. As characterized by oneself, often without right or justification. (American Heritage Dictionary)

... better to have latent potential than none at all!

Yeah, I was being tricky with the word latent too.

latent - Definitions from Dictionary.com

For me to understand consciousness as such, it's a study of cause of creation, not effect. Walter Russell (philosophy.org) has written many books plus a home study cause which sits well with me. He states that consciousness is the unseen half of the cycle of creation, which is our inner being, where creation takes place.

Paulh,

Regarding connectionism and chaos: yes, connectionism as we know it is an artifact crafted with linear goals in mind, and is "fed" or taught along such lines. One can make a connectionist system do what is wanted because of its inherent flexibility, and thus reach the results that are desired -- akin to William James' concept of the psychologist's fallacy. The models our researchers study are still "toys" relative to what millions of years of natural selection has crafted within us, and we can't know for sure if what we are crafting is true to the path that mother nature took.

As to the brain computing on the edge of chaos -- ah, once again William James beat us there, with his "hair trigger" concept. In fact, here's a paper on Williams James, Chaos Theory, and Conscious Experience:

www.uoguelph.ca/~abailey/Resources/William%20James%20and%20Chaos%20The.pdf

My two cents is that this is a "both and" situation; parallelism and chaos coexist. Parallelism could well dominate in situations where recognition and fast response is needed; e.g., a rock is coming at fast speed towards your head. However, random walks around multiple state-space attractors could perhaps better describe our higher levels of mental activity, especially conflicted emotional-rational situations. "She loves me, she loves me not". The chaotic attractor effects could still be based around parallel architecture, albeit an architecture with much feedback and arranged according to multiple levels (ultimately forming the ad-hoc thalmo-cortical loopings that seem to drive executive decisions and motor actions, if I read Edelman right).

It might be interesting to think more about the concept of "self" in the context of a "strange attractor" within high-level cortex dynamics.

Chaotic attractor effects admittedly are prevalent within control devices in complex on-going natural systems, e.g. insect networks and auto-catalytic chemical networks in the body. However, these control devices and their exploitation of attractors (especially the more complex attractor types) are generally quite slow. I am now reading "Kinds of Minds" by Daniel Dennett, who points out that the brain evolved to speed up things and increase precision, and thus increase survival strength. One great advantage of parallelism is speed, precision, flexibility and robustness. Chaotic attractors still have to deal with . . . well, with chaos! (i.e., randomness). The brain might well exploit the simpler forms of attractors, and the more complex and strange ones probably occur within it due to the brain-body's complexity and highly recursive operating nature. I'm just saying (again, as a layman) that chaotic attractors are not the entire show under the cranium, they probably share the stage with parallelism, and also with forms of linear data processing too. They certainly do allow a much more sophisticated understanding of the brain's operation, and the nature of resultant mental states. I'm glad that you shared this idea, it has turned on some light bulbs amidst my own haphazard "mental scaffolds", which I so often trip on!

As to the brain in the vat -- certainly a full blown simulation, including the ability to precisely mimic the brain's chemical reactions and correlations to the various states that it reports to the brain, would be needed. Per Dennett, the brain is inherently designed around existing control systems embedded in the body. These control systems are older, evolution-wise, and are slower and less precise control systems (depending on signaling by hormones and other molecular devices); but the brain heavily depends upon their operation. They and the resultant blood chemistry responses would require an extremely complex mimicking system in order not to throw the brain out of its design parameters and cause it to malfunction. It's possible as a thought experiment, but I doubt if it will be realized any time soon.

Very interesting, though.

Jim G.

So the brain probably relies on intermolecular forces in order to function at it's full potential, and since we can agree strongly that consciousness has a physical side, consciousness requires the molecular makeup of which we have, perhaps?

I mean, its possible that we could artificially mimic these forces with other elements like more of the metals, considering that quantum mechanics provides more freedom, so cybernetics is possible, and to a degree fully conscious machines not relying on our form.

H20310401,

With regard to randomness, I was not thinking of the maximum metaphysical concept of complete lack of order. In that case, a continuing phenomenon would be completely unbounded; its upcoming values and states could then change to any state or number allowed in our universe (or NOT allowed, if COMPLETELY COMPLETELY unbound -- a universe that allows all and indeed requires infinities). I am using the more limited concept of randomness that is used in statistics, i.e. randomness that is bounded by higher-order regularities, e.g. probability distributions. This kind of randomness says that there is no way to know just what the next value or state is going to be; but there are ways to determine the probability of any possible future state. In a number chain, this means that the only way to tell someone else how that chain of numbers has turned out is to sent all of the numbers produced. There is no way to shorten it, no way to convey it with fewer digits of information. By contrast, take the case of the number sequence 1, 1, 2, 3, 5, 8, 13, . . . . Say that you've let it go out to 1000 numbers; do you need to send enough digits to convey 1000 individual numbers? No, you just say "starts with 1, then 1, then every following number equals the addition of the past two, continue 998 times." Much more economical - and not random.

However, even in the case where there is no way of saying what the next number will be based upon previous number events (i.e., randomness as I am using the term), there may still be probabilities and aggregate relationships. E.g., if the series is a random coin toss where 0 is heads and 1 is tails, and the coin is weighted such that heads comes up 60% of the time and tails occurs 40% of the time, you can't say what the next number will be based on the chain of past numbers, but you can say that the chances of a 0 are 60%, and the aggregate count of 0's relative to all numbers is expected to be close to 60% (closer and closer as the total of all numbers increases).

As to complexity and this kind of limited randomness: my rough intuition is that a more complex mechanism or process gives increased opportunities for a steady-state system to enter a more highly random state; again, this randomness being the mild version, and not the "whole world has gone crazy" variety. A rough comparison is a fast flowing, unobstructed river. It's pretty easy to predict the progress of any particular water molecule when the river has no obstacles. The river is a fairly simple system at that point. Put in one or two obstructions, e.g. a bridge pier, and the water flows can still be easily understood. But make the river go over a precipice with rocks and boulders scattered on it, and water starts splashing in all sorts of directions. Imagine a string of water molecules following each other down the river, nice and orderly; but they go in different directions once they reach the "complex" part, i.e. the boulder-strewn waterfalls. One goes one way, the one right behind goes another, no one can say why. In the aggregate, on average, you can say that they keep going forward, they eventually continue downstream from the waterfalls; but each molecule's individual path through the rocks and falls cannot be predicted based on where the ones ahead of it went. The rocks and precipice has put the water / river system into temporary chaos. That, roughly speaking, is an example of complexity causing chaos and increased randomness.

(Please excuse the 101 level of these explanations; you've heard them before, I'm repeating them mostly for my own sake).

The brain might be compared to that river, a river of information signals (mediated by the firing of ion potentials through neurons and across synapses). In some instances, the flow is slow and relatively unimpeded, and the randomness is low. But increase the flow and direct it through complex junctions and processes, and a point of chaos transition can occur. E.g., put a person under a lot of pressure and confusing circumstances, and their behavior can become unpredictable.

As to your musings about the ULTIMATE nature of randomness and the ULTIMATE nature of the universe and the ULTIMATE nature of consciousness - well, I'm not sure that I completely understand, but I am interested. My rough "folk" notion is that consciousness requires a mix of randomness and determinism - i.e., "deterministic chaos". Or, coming at it the other way, chaos creating order through emergence. This is something of a complementarity of contradictions, a yin-yang process of state transitions.

The purest of randomness would arguably require some forms of infinity, which doesn't seem to exist in our universe, either on the big side (the relativism limit to light speed) or the small (quantum limits on smallness). The purest of determinisms, the Laplacian block universe, would arguably not require consciousness, given the general ("folk") association of consciousness and free will on an essence/essential level. We perceive a world with conscious awareness but no infinities, with randomness and yet with finite predictabilities. Bounded potential and unconstrained possibility ARE tangled, and are only unfolded in the realm (i.e., abstract "state-space" that is non-4D or trans-4D) of consciousness. And in that process, in that tension between the random and the determined, somehow consciousness gives the physical entity that hosts it information about "the value of being" (with apologies for lack of a better, less New-Agey phrase), information that arguably has survival value for the being and its species. IMHO.

Jim G.

H20310401,

Interestingly enough, Daniel Dennett, who I cite regarding the brain's strong interdependence with body control processes and with the environment that supports human life, is also a bit of a functionalist, toying with the notion that consciousness can be created in silico, in chemical / physical circumstances much different than Earth-life based carbon chemistry. IIRC, John Searle has flip-flopped on this point. Douglas Hofstadter, who has written some books and articles with Dennett, strongly supports functionalism, but also says that whatever consciousness might develop in a silicon-based machine (or other data processing structure) might not be anything like ours. We might not even be able to recognize it per Hofstadter (again, IIRC).

Personally, I think that the "essence" of consciousness involves something beyond chemical and physical distinctions, such that a consciousness based on other chemistries and environments (e.g., life from other solar systems) will ultimately share with us something about "being itself". I'm getting close to Hegel and Heidegger territory here, yikes!

Jim G.

Perhaps ion potentials using sodium and potassium ions are cause for gradients in triggering memory, lol. (I know this is stupid nonsense here).

And then if we added some other elements that seemed apt to be linear like boron and aluminum ions, lol, to the action potential.

And maybe we rely on radioactive elements such as barium and strontium for randomness, such on the quantum scale that such gradients can produce mutative and creative results to the mind....lol.

But yes, chemical and physical aren't the only distinctions for the essence of consciousness.

I always wondered, if the universe was created by light, a completely acausal, atemporal, nonlocal concept, that consciousness is created by such a phenomenon involving light in our heads. Maybe there is this little area in our heads where some light is flowing back and forth in a rhythmic flow that sequentiates for our consciousness and reality,:drinking:.

But point being that randomness is the same as absolute order. everything is the same as nothing. 0 (in the sense of not being anything) is the same infinite.

... would you settle for electro-chemical impulses flowing back and forth in a rhythmic flow? ... some recent research into brain behavior on scales larger than just individual neurons or specialized neural nets point to an emergent synchrony that flows back and forth in the brain during mental activity ... one hypothesis is that this is the telltale sign of deterministic chaos at work, the implication being that human thought is less like connectionist models and more like nonlinear dynamics ...

Wow, thats fine, but I feel it necessary to have chaos by something as orderly and linear as light, being that it is much the same thing.

I would have thought that the creation of cause would have been of linearity rather than of nonlinear dynamics, oh well.

I mean perhaps I have this reversed. I think that an opposite should hold true to reality as the other opposite. We view the universe in such symmetry. And so, nonlinearity would be appropriate for causing such a linear abstract.

... the creation of engineering is linearity - without linearity (where the whole is literally the sum of its parts) we'd likely still be in the iron age ... but linearity is a double-edged sword - while it brings engineering into the realm of the possible for us humans, at the same time it renders our engineered creations lifeless ... nonlinearity is a fundamental creative genius of nature - if it weren't for nonlinearity, we wouldn't be here ...

My understanding of creation is, inner thinking sets energy in motion, that motion will always be in the form of cycles, each cycle being a complete life time within itself, eg as I breathe in I compress, this is the generative half the cycle,(living half) as I breathe out this is the radiating half (death half) all creation uses this principal from the microscopic to the universes. the whole of creation relies on the Rhythmic Balanced Interchange between opposites.