John Page

OK, yes, I think we agree there's a difference between control through intent rather than by mechanical law.

Agreed. Where I was going with this is that in humans (IMO) the "software + memory" portion are very intertwined. Data may be realized by a process for example. Example, two big data arrays x and y, too big to store in memory together. But y only differs from x by doing a 90 degree cockwise shift (same picture but on its side). Thus, y can be retrieved by transforming x and vice versa. I'm not saying this is a real model, just an example of how a "processesbase" can supplant data.

Miscommunication. Symbols are arbitrary in form but associated with the "real" data. Pointers tell you where the real data is.

Side point. I've always wondered how a Turing Machine is supposed tell symbols apart - that requires spatial recognition. Must have a very smart read head!

Difficult for me to argue since I don't know the answer either.

Agreed. Until we build a computer that can understand this, of course. Ditto agreed your comments that computers do not currently seem to understand concrete vs. abstract.

I guess we're back to layers. No big deal but I prefer to use the word perceive because each neuron in the chain acts on the data it receives, its not just a passive device.

Yes, consciousness is a conundrum for me. Lots to learn.

I'm sure some data is thrown away, otherwise even the capacity of the brain would become overloaded. I think we build up conceptual maps as we "grow up" so when we see a tree most of the actual data about that specific tree is discarded. All wse need to store is the space/time event and anything unique to that tree we notice. Of course, we could walk down the street and not notice the tree at all....

You're probably right. I was likely confusing myself also with trying to consider both the "operational mode" and the "self organizing" behavior that the brain likely undertakes. I must admit I'm rather taken with the concept of the "plasticity" of the brain where its supposed to "grow round the data".

The state of the dendrites belies the input data, axons output and the synaptic function triggers on certain conditions. So, the neuron is both data and process.

With up to 150K data inputs I think neurons detect conformance to complex structural templates. Perhaps you might consider them "presence detectors" or the ON unit of the Basic language.
This from The Scientist "Brain development is an awesome feat. Over a trillion neurons make connections with thousands of target cells to create coherent circuits for vision, language, movement, and memory. Most neurons migrate from their birthplace to their permanent home in the developing brain. Once they find a home, some cells send growth cones to different parts of the brain, creating networks with their axons."


I was just trying to clarify the point that just because something is abstract doesn't mean it doesn't exist in some form. I think you understand this, I wasn't trying to be obscure.

Right, this is where I think a conscious or rationalist mode of operation is more effective.

The consciousness can focus but does not seem to be able to probe down to low level cognition processes. Deception is not in what is received but in its interpretation, e.g. a stick appears to bend as it is inserted into water.

I don't think so. Give me any example and I'll try and parse it. I do try and use "external reality" as distinct from our minds internal representation of that reality. Is that internal representation real, sure (otherwise how could we know it), but is it accurate?

WRT an individual, I divide reality into quadrants. 1. External & Real, Known. 2. External & Real, Unknown. 3. Imaginary (or Abstract), Known. 4. Imaginary, Unknown (or Unimagined). My definition of reality includes all of this. Does this help?

I'm not saying a logical illusion isn't, at the end of the day, phenomenological.

Just like us! But good and bad are subjective moral judgements and I'm guessing as to whether dogs evaluate same.

It can be observed, but may not be.

Thanks for responding - I must get some sleeeeep now.

Cheers!

[ April 08, 2002: Message edited by: John Page ]</p>

owleye

John...

"Agreed. Where I was going with this is that in humans (IMO) the "software + memory" portion are very intertwined. Data may be realized by a process for example. Example, two big data arrays x and y, too big to store in memory together. But y only differs from x by doing a 90 degree cockwise shift (same picture but on its side). Thus, y can be retrieved by transforming x and vice versa. I'm not saying this is a real model, just an example of how a "processesbase" can supplant data."

The data/information processing model of human cognition is helpful, I think, because it supports a functional analysis of cognition. Certain processes transform data, others reduce it. Holding a pointer (symbol) to an object introduces certain efficiencies in that the mind does not need to bring to the surface all that is entailed by the concept being directed toward its object. As has been analyzed by Kant, for example, using his notions of extension and intension, every concept can be analyzed into its constituent parts (sub-species), which in turn can be further analyzed into its parts, without end (at least formally). Each sub-concept, however, has its own horizon of super-concepts that form a complex network that encloses it. The former is the intension of a concept, the latter is its extension (for Kant). (Much of this analysis was motivated by the advances in chemistry beyond the earth, air, fire and water model in the 18th century, though it applies to the organization of species and genera in living things as well.)

"Miscommunication. Symbols are arbitrary in form but associated with the "real" data. Pointers tell you where the real data is."

Hmm. Perhaps I have been too quick here. I'd been thinking of signs, not symbols. A sign is a representation that refers to another representation, about which it is associated on a more permanent basis, through convention. A symbol may be a sign, but its use is a bit different. In the first place it is general, and we are not, by its use, specifically required to establish any concrete path to what it is a symbol of. In a mathematical expression the letter 'a' can be used in a formula, for example, to symbolize (or stand for) any constant real number, and 'x' to symbolize any variable that can range over real number values. As such, '+' and other operators, are better understood as signs. Of course, in pure (that is, formal) mathematics, all are symbols whose "meaning" is strictly derived from the logical rules of their use and construction.

"Side point. I've always wondered how a Turing Machine is supposed tell symbols apart - that requires spatial recognition. Must have a very smart read head!"

The general rule of recognition I find useful is that we don't know what something is (i.e., can't recognize it) unless we already "know" what it means to be it, and that this particular is capable of being picked out from this meaning. (One interlocutor of mine didn't care for this, and substituted "... unless we already know what it is to be it..." but I'm not yet convinced of the merits of this idea.) Note that given this, it implies that we don't really start the world with a blank slate, but have innate, though primitive, concepts (possibly even as significant as 'up vs. down') that allow us to make sense of the world.

Your citation:
This from The Scientist "Brain development is an awesome feat. Over a trillion neurons make connections with thousands of target cells to create coherent circuits for vision, language, movement, and memory. Most neurons migrate from their birthplace to their permanent home in the developing brain. Once they find a home, some cells send growth cones to different parts of the brain, creating networks with their axons."

The concept of a directed physical process is one that we must be careful not to get carried away with. (Recall my "as if" concession.) In any case, all development is based on current local conditions associated with each cell (considered as an atomic unit). It can only act on and be acted by its immediate environment. Obviously "migration" can be thought of as channeled and directed if we don't know the mechanisms that produce the channel and guide the development. In any case, I can accept this as a possible counter-example. However it remains within the domain of developmental biology, not cognitive science, I think. (It may, of course, help the computer scientist wishing to develop neural networks.) (And it may also help the neurologist.)

"Right, this is where I think a conscious or rationalist mode of operation is more effective."

Note that in humans, it is not merely a cognitive response that helps us. All things emotional, spiritual, intuitive, creative, and so forth, help in dealing with a world in which we are not quite in synch with.

"The consciousness can focus but does not seem to be able to probe down to low level cognition processes. Deception is not in what is received but in its interpretation, e.g. a stick appears to bend as it is inserted into water."

And this only shows that deception requires consciousness (including awareness and the creation of "appearances")

"I don't think so. Give me any example and I'll try and parse it. I do try and use "external reality" as distinct from our minds internal representation of that reality. Is that internal representation real, sure (otherwise how could we know it), but is it accurate?"

Parse the bent stick at the water line example.

"WRT an individual, I divide reality into quadrants. 1. External & Real, Known. 2. External & Real, Unknown. 3. Imaginary (or Abstract), Known. 4. Imaginary, Unknown (or Unimagined). My definition of reality includes all of this. Does this help?"

So far, no. But perhaps it will come out from your analysis of the example.

"I'm not saying a logical illusion isn't, at the end of the day, phenomenological."

Now that I think about it, I've been too hasty here as well. I should heed the advice of Kant and think that reason can be under an illusion, though it has different manifestations, depending on what faculty is being employed in making inadequate judgments. Judgements from pure concepts vs. Judgments of perception -- objective vs. subjective, or even aesthetic judgments. Illusions are a large topic in the Critique of Pure Reason. (Second part of the Analytic of Principles -- the Dialectic.)

"Just like us! But good and bad are subjective moral judgements and I'm guessing as to whether dogs evaluate same."

I would disagree here, since I believe we, unlike dogs, can act and judge on the basis of principles, and not just because we were conditioned to do so. A parent who is teaching a child that it is wrong to lie is teaching a principle. The question here is not how the child learns, which in part may be due to a conditioning process, but rather how the parent is able to consider that what she is teaching is a principle that the child must learn. The parent expects the child to learn that it is in principle wrong to lie and that punishment and rewards have no bearing on the principle itself. That is, one does not lie to avoid punishment nor avoid lying to receive some reward, but rather does so for its own sake. To achieve this level of maturity requires much more than what a dog is capable of. Indeed, not all children, nor all adults, achieve this. Moreover, it may come down to it mostly being an ideal, towards which we aspire, rather than something actually existing. But, we function not just in the real world, but also in the world as we would like it to be.

owleye

RandomGuy

This was posted a while ago. Forgive me for bringing it back to the top.

If anything, introducing a chaos theory into the naturalistic paradigm only means that our views are now held only held to on because of what nature has directly caused us to believe, and what we are chanced to believe in. This includes thoughts that have no meaning what-so-ever (ie, thoughts determined by chaos). So I think our thoughts in a naturalistic paradigm are still determined.

In short, whether determined by nature, or determined by chance (ie, chaos)-- our thoughts are determined still.

[ April 10, 2002: Message edited by: RandomGuy ]</p>

Wyrdsmyth

The title of this thread brings up the most interesting question...

How can we test the hypothesis that we have free will? Or, that do not have free will?

There's just absolutely no way to do it. At least I can't think of a way. We can't rewind the universe, and play it out again to see if we'd keep doing the same things "over and over." And even if we could -- what would that mean, really? If you choose chocolate ice cream over any other flavor, and we keep rewinding that event and you keep choosing chocolate ice cream no matter how many times we play it out, does that mean at time t you couldn't do otherwise than choose chocolate ice cream? Does that make you a chocolate ice cream-eating robot?

I don't even know what such an experiment would prove... to your subjective point of view, where you only "experience" that particular event once, you certainly feel as if you are free to choose chocolate ice cream, or not. But to us observers, on this metaphysical ledge, with time-rewind remotes in our hands, it looks amazingly apparent that you can't do otherwise at this particular moment.

Is this what we mean by free will? The power to do X or ~X at a given time, t? The power to be causally indeterminate? Random, somehow, at the physical level of existence -- spontaneous?

Or, do we mean the subjective sensation of "feeling free," feeling unconconstrained, not coerced, no pressured -- a subjective experience in which we don't detect any outside influences? But what are "outside influences"? How do we separate "you" from all of the factors "outside of you" that have influenced who you are, and are still influencing you right now? Isn't that a somewhat arbitrary barrier?

John Page

My delay is due to my interest in this topic, not otherwise! Here goes:

An issue here is one of hierarchy, the concept leading through to theories of type etc. I believe the only hierarchy at play is one of levels of abstraction.

As you observe, "typing" reality itself results in a complex network of relationships that seemingly defy Third Normal Form. Persisting in this approach results in anomalous results including Russell's Antinomy, for which my theory of abstraction provides a formal explanation.

I do not see how a formal system that involves infinite regress (in this case of super-concepts and sub-concepts) can be proven. Simply, one cannot visit all the instances of concepts and prove them. I believe the mind is finite, limited by its supporting physical mechanism, thus all the concepts contained therein might be mapped and explained.

The only formal system that I know of that can be reliably treated with extensionality is mathematics, a science in which each unit is of equal value by definition. However, numbers themselves are abstract entities so our exploration of their properties and behavior is limited only by our imagination. As before, our freedom is limited by the constraints on our mental faculties.

I think we're getting closer in our terminology. The symbol + is used to denote a function or operation. This seems to me different than a pointer which is just meta-data and not a process.

Obviously the brain contains some hard wiring but we do have an ability to learn. It seems to me that all our concepts come from experience - if they were purely hard-wired we would have very limited ability to adapt to different environments. My theoretical model of the mind contains template concepts (axioms for matching identities). These concepts are interrelated so they form a working hypothesis about the outside world. In turn we may then act with seeming foreknowledge, enabling our ability to plan creating the "illusion" of free will.

Agreed, I was supposing that the "migration" of nerve cell connections might be a part of the mapping of correlations found in reality. I forget exact sources, but isn't there evidence that the number of new connections in the brain decreases rapidly during adolesence, prompting the thought that we've "built" our map of things?

Just a note that I regard emotional activity as a kind of thought.
I think its the other way round, consciousness requires a degree of deception!
OK, this was in relation to illusion. The first time we encounter the eye's image I assert that we will process it literally. Through experience, e.g. repeatedly taking the stick in and out of water, we learn to compensate for the literal image. Through this compensation we "know" that there is something fishy going on, that there is something illusory because bending a stick takes force and unbending a stick is often impossible. So, cognitive dissonance belies the illusion. Further investigation reveals the cause of the illusion. (I'm not sure I did a great job on this).

Which begs the question of the basis on which dogs can and do act and judge, and how this is different from the use of "principles".

Thanks, Cheers!

John Page

I think the difficult one to measure is when someone thinks they have freedom to choose. We need to know more about how the mind works. I'm hoping we discover an axiom of freedom that enables us to act "altruistically" by logically cancelling out deterministic effects. Whether this is rational or not I don't know....

Cheers!

owleye

John....

"An issue here is one of hierarchy, the concept leading through to theories of type etc. I believe the only hierarchy at play is one of levels of abstraction."

Does this mean, for example, that the concept of cat has a lower level of abstraction than the concept of mammal?

For another example, would the above mean that the concept of law has a lower level of abstraction than the concept of rule, assuming laws are rules of a certain kind. Is it possible to situate the level of abstraction of the concept of justice within that of the concept of law, or vice-versa,?

"As you observe, "typing" reality itself results in a complex network of relationships that seemingly defy Third Normal Form. Persisting in this approach results in anomalous results including Russell's Antinomy, for which my theory of abstraction provides a formal explanation."

I'm not familiar with Russell's Antimony. I confess I must have missed your formal theory of abstraction, probably because I wasn't paying enough attention. There are more than the two kinds of organizations (hierarchical and network), of course, but assuming only two, how is your theory of abstraction both hierarchical, as you indicated previously, and network-based, as you indicated here?

"I do not see how a formal system that involves infinite regress (in this case of super-concepts and sub-concepts) can be proven."

Today, I suppose what would be needed is a kind of mathematical induction. Kant would have made use of a kind of the recursive quality to species and sub-species to demonstrate it. Note that 'infinite regress' does not mean that the regress is a completed infinity. This, for Kant, is merely an idea of reason. It only means that any finite number to the chain of sub-species cannot represent the entire chain. Alternatively, for every species, there exists, in priniciple, sub-species. Alternatively, there can be no infima sub-species. This is because it is in the nature of a species to have variation, and whenever there is variation it is possible to make another division, which in turn can make another division.

I take Kant's understanding as interesting but only so as a starting point. During his "transition" (post-critical) period, collected together into what is known as his Opus Postumus, in which he came to understand the advances made in chemistry, he realized there was a "gap" that needed to be filled. He had worked out a top-down approach, as well as a bottom up approach, but realized that unless he could understand how these two approaches could be unified (i.e., meet in some common principle), his earlier work would not be completed. Unfortunately he died before he could complete it. It is just as well, though, since the advances in chemistry (principally by Lavoisier) would be inadequate by today's standards (e.g., Kant was trying to find a place for ether, during this period).

"Simply, one cannot visit all the instances of concepts and prove them. I believe the mind is finite, limited by its supporting physical mechanism, thus all the concepts contained therein might be mapped and explained."

We disagree here. The mind, I believe, is infinite, principally because language is infinite. There is no limit to the number of different sentences that can be constructed. Those philosophers who have sought to physicalize the mind have come under severe criticism on this basis.

"The only formal system that I know of that can be reliably treated with extensionality is mathematics, a science in which each unit is of equal value by definition."

As you may know, Hilbert thought he could formalize arithmetic, only to be shot down by Godel. In any case, you need to define your use of 'extension' before you go further. I took pains to indicate that the extension I was using in the above was Kant's. (Note that 'extension' also has the meaning associated with 'spatial extension' or even duration. However, neither of these quantitative senses are intended here.)

Extension in its modern use can mean the set of individuals that Kant would say (and we might agree) that is subsumed by (or falls under) a given concept. Another use is that aspect of meaning which refers to the individual itself. The extension of 'cat' to "this cat," for example. It is to be distinguished from the intensional meaning through what is meant by a 'cat' -- i.e., it is directed to the concept of a 'cat'. Frege differentiates these by the terms 'sense' and 'reference'.

"However, numbers themselves are abstract entities so our exploration of their properties and behavior is limited only by our imagination."

Interestingly, Kant has a view of number so that it represents a magnitude in general. He doesn't have any understanding of set theory or type theory in our modern sense. Indeed, he has only subject-predicate logic to work with. You have a different understanding because you are a product of a modern theory of relations. However, the philosophy of mathematics is a rich and interesting area, and by no means settled. If you regard mathematics in strictly formal terms, the idea of a 'unit' may give you a problem. To even posit a unit may put you already into an "anti-formalist" position.

"I think we're getting closer in our terminology. The symbol + is used to denote a function or operation. This seems to me different than a pointer which is just meta-data and not a process."

My short analysis would suggest that '+' is not normally considered to be a symbol. Rather it is a sign. Denotation itself a feature of signs. That is, given a denotational meaning of a sign, the sign signifies what the sign denotes, in this case, it denotes a mathematical operation. I don't think it makes sense to say the '+' symbolizes a specific mathematical operation. If we regard '+' as a symbol, it would tend to represent a certain class of (additive) operations generally, which, of course, is perfectly acceptable, mathematically. Of course, as a symbol, it could also take on elaborate ritual meanings associated with it being a 'mark' -- the mark of the cross, for example. The meaning in this case, then, is the set of all practices and institutions associated with a group of one sort or another. In its ordinary context, however, I would think '+' refers to a specific mathematical operation, known as 'addition' and for this reason, I think of it as a sign, and is subject to being tokenized.

"Obviously the brain contains some hard wiring

I don't know this to be the case. How is it obvious? Indeed, how is it true? To be innate doesn't mean "hard-wired." Well, maybe I'm being misled by our different understandings of being hard-wired. Hard-wired implies (at least to me) being unalterable (unrewirable).

"but we do have an ability to learn. It seems to me that all our concepts come from experience - if they were purely hard-wired we would have very limited ability to adapt to different environments."

I gather you are using 'hard-wire' in a flexible way to permit gradations of hardness.

My theory is that we begin with a certain neuronal configuration (following a period in infancy in which we adapt to our immediate environment but subsequently forget what we've adapted to and have to learn everything from "scratch" so-to-speak). This initial state, I would claim, consists of a set of initial categories that organize the initial set of experiences. If we didn't have this initial capability, we would have no way to organize anything and it would be a total chaos. This initial state, though quite general, is not immutable. Indeed, different cultures categorize things quite differently. It is said, for example, the the east (Chinese in particular) develop process ontologies, while the west has developed substance ontologies.

"My theoretical model of the mind contains template concepts (axioms for matching identities). These concepts are interrelated so they form a working hypothesis about the outside world. In turn we may then act with seeming foreknowledge, enabling our ability to plan creating the "illusion" of free will."

Much more is needed, I think, to move from "concepts are interrelated" to "working hypothesis" to being "about the outside world", to "acting with foreknowledge" to "planning", and to "free will". Each of these steps is in itself a huge leap (or requires a lot of work to fill in before the linkage is made).

"Agreed, I was supposing that the "migration" of nerve cell connections might be a part of the mapping of correlations found in reality."

There certainly are neural connections between the brain and the feet, to give but one example, such that when our feet get hot, our brain informs us of this by generating the proper pain response which is somehow directed to our feet. This "somehow" is undoubtedly complex. Indeed, it is not uncommon to have a "pinched nerve" in our lower lumbar region of the spine cause us to feel heat in one or another of our feet. How would you model this?

"I forget exact sources, but isn't there evidence that the number of new connections in the brain decreases rapidly during adolesence, prompting the thought that we've "built" our map of things?"

Our entire body is mapped in the brain. We lose a limb but still think it's still there. Actually we can extend beyond this as well. Merleau-Ponty has written much on this. Our mind extends to include, for example, the feather on our hat so that we have to duck to avoid bending it when we pass under a low beam. But maybe this isn't what you mean. Do you mean our visual or auditory field?

"Just a note that I regard emotional activity as a kind of thought."

What kind of thought is it?

"I think its the other way round, consciousness requires a degree of deception!"

I personally don't think this makes sense, but since it does to you, what do you mean by deception? What is the deception?

"OK, this was in relation to illusion. The first time we encounter the eye's image I assert that we will process it literally. Through experience, e.g. repeatedly taking the stick in and out of water, we learn to compensate for the literal image. Through this compensation we "know" that there is something fishy going on, that there is something illusory because bending a stick takes force and unbending a stick is often impossible."

What makes it possible for us to "know" that there is something fishy going on? Let me consider that this isn't even necessary to your point. Why can't this alleged compensation take place without our being aware of it? My claim would be that in order to be aware of it, to notice something "fishy", we need consciousness. It is only through a reflection of our own perceptions that allows this thing to be an illusion. That is, since in order to make use of information derived from photons of light, it has to be transformed into visual data about objects in the world. What makes this compensation that you call it, any different than any other type of transformation?

"So, cognitive dissonance belies the illusion. Further investigation reveals the cause of the illusion. (I'm not sure I did a great job on this)."

There is a lot of assumptions that are being made here (good ones, but which need to be fleshed out). The first that comes to mind is that there are causal relationships that are assumed by your analysis. If you are a Humean, you would not, through reason alone, be able to determine that because something is, something must follow. There is no reason, for example, why the stick should not bend and straighten in accordance with your taking it in and out of the water, unless you make assumptions about the stick itself. The "famous" computer scientist Judea Pearl of UCLA, has made great strides in the area of causal reasoning. This may allow us to identify illusions of reasoning, though I feel confident that without consciousness, the bent stick in the water line would not be considered an illusion. It is a fact of consciousness, not of reasoning.

"Which begs the question of the basis on which dogs can and do act and judge, and how this is different from the use of "principles".

Have you not been paying attention? Dogs and cats would act on the basis of conditioning. Humans can act on principles. They can lift the concepts used to discriminate from that context to a more general context such that humans can know they have made a mistake. You are judging that a dog or a cat knows that they have made a mistake based on the behavior of that dog. But I contend that this is projection of our own sense of making a mistake. The dog reacts the way it does on the basis that it has been conditioned to do so by a system of rewards and punishments. For example, if it behaves in a way that makes it appear that it is cowering for example, there is the possibility that it learned to do this because it might mean to us that we need not further punish them for what they did (which to the dog means that there is less likelihood of further punishment). I think children and to a lesser degree adults are "trained" in this way. But because we learn to develop principles of action, or reasons for acting, we alone (or almost alone) are capable of acting on principles.

owleye

Wyrdsmyth

I don't see why acting altruistically is incompatible with determinism. But then again, I don't see why all forms of free will have to be incompatible with determinism, either. Like I said, they're pretty nebulous concepts. It's not always clear to me what people mean by "free will" or "determinism." The terms sometimes get tossed around like it's obvious what they mean, but I don't think it's always obvious, or that people all mean the same thing by them. Is free will the ability to do X or not-X at time t, to be indeterminate at the quantum level? Is 'indeterminacy' important or even relevant to free will? Or, is 'unpredictability' the important component? What if we are determined in our actions (physically, causally determined), but fundamentally unpredictable -- even in principle... is that good enough for us to say we're free-willed? Or, is free will just a subjective sensation, a feeling of not being coerced or pressured in any way, a kind of liberated 'I can do anything I want' feeling? You could have that feeling, and yet still be determined at the physical, quantum level of matter and energy...

But let's say certain interpretations of Quantum Mechanics are true, and reality is fundamentally indeterministic? Does that even help the case of free will at all? Or hurt it? Does that just make us QM pachinko machines? Quantum dice rolling in our heads, at the sub-atomic level, which manifest to us as decisions, choices, and intentions...

[ April 11, 2002: Message edited by: Wyrdsmyth ]</p>

John Page

Hi Owleye, I'll respond to the abstraction questions first.

</strong>
The abstract entity such as "the concept of cat" does not remain constant. Why? Because our concept of cat and its "place in the world" becomes more sophisticated over time as we learn.

This makes it impossible for me to give you a reliable answer to your specific question, I do not know how the contents of your mind are organized.

In some cases, I think we will find that a "higher level" concept cannot be constructed without without first building a "lower level" concept. For example, I cannot imagine how one could conceive of a square root without concepts of quantity (number) and division.

Back to the cat. If I truly believe that all cats are mammals then the concept of mammals would be at a higher level of abstraction. However, its amusing to think that in a different part of our mind we might retain the remnants of a structure from the time we knew cats but the notion of mammals had not been introduced. Accordingly our minds might, under certain circumstances, present us the scenario where mammals are types of cat. This could explain explain how confusion arises in our minds.

Indeed, what I am proposing is that because the mind does not know beforehand how to structure its concepts of the environment it places concepts at the layer of abstraction that is most coherent. Formalism exposes this process to scrutiny.

As a footnote, since the relationships between concepts are more likely to be complex "network like" and subject to change I am not proposing that there are strict abstraction levels 1, 2, 3. I can't draw a picture here but with the mental image of a sophisticated back prop network you could end up with relative levels of 1, 1.2 etc.

If this is unclear it because I'm hypothesizing. Maybe I just need to let my brain restructure a little more....

I would apply the same kind of remarks as above except I'm not comfortable with concepts being contained within other concepts. I think an assumed 1:n relationship can give this appearance of "containment".

Cheers!

DRFseven

And where does the imagination come from? Experience, which is "replayed" with other associated replays. And all experience depends upon prior experience. People just don't seem to get that we can't form thoughts out of nothing.