No, understanding mysteries of the mind requies finding fundamental principles of the mind and cognition, not neurons. The difference is like between elementary particle physics and thermodynamics. Einstein emphasized that thermodynamics has its own basic principles, which do not obviously follow from elementary particle interactions. For some (hypothetical and prooved) fundamental principles of the mind look in:

Perlovsky, L.I., Deming R.W., & Ilin, R. (2011). Emotional Cognitive Neural Algorithms with Engineering Applications. Dynamic Logic: from vague to crisp. Springer, Heidelberg, Germany.

I agree: we cannot "model our minds, by observing the neurons, brain, behaviors? ".

As already the fathers of the Scientific Revolution knew---the mind, unlike the brain, is not a *spatial* entity.

How we keep the experiences in memory, and why they go with a particular order I think is what matters.

well, partly agree, by doing research on how neurons work can also give us some insightful ideas of the supervenient level of mind, and for our understanding of cognition, a big part of our work depends on the research of neuroscience, right?

What i agree is, yes, we need to find the fundamental principles of our minds to reveal the mystery of minds.

However, if we say so, where do we start for our investigation of the fundamental principles? still need to gather the evidence from multiple areas.

I am trying to understand the analogy you are using, maybe for my understanding, you regard the domain of the mind as independent science, which has its own principles and fundamental laws? I do agree that there are some fundamental principles of the mind. but which direction should we look for? Not the neuroscience?

Thanks for the recommendation of your book, I get it from springerlink, which is really nice. Will read it soon.

There are some differences between our mind and our brain. Although uncertainty about future make me sure not about these, but I expect the future first to model neurons, second the brain and finally the mind. Because minding takes place in all of the which.

Up to now we can not even see exact responses of the brain areas concerning with what we see, think, smell or behave! So how could we speak about modeling for yet.

Bashir said: "I expect the future first to model neurons, second the brain and finally the mind. "

This is a widespread brute-force approach to the mind and will be even less successful than trying to use knifes to open the newly arrived cosmic ship from another galaxy, just because there are some protruding parts in it. ;--)

I'm reasonably sure that such approach has not and will not be successful in modeling the mind.

What has much more chances of success?

I believe one has to make a commitment to a particular informational process in Nature---my choice is induction---that is a reasonable candidate for the central informational process in Nature and proceed with its formal modeling.

However, probably the most important criterion for the proposed candidate model is that it should *immediately* tell us what has been missing from the big scientific picture. Why? This model *cannot fail to tell us something radically new in science*, because the integration of the mind into our scientific picture is not a matter of fact scientific business. In that sense, we do have a very strong selection criterion for various proposals, which is quite useful to have.

Dear Colleafgues,

it is a pleasure that my comments about fundamental laws of the mind received so many interesting replies. Few additional comments that might be helpful to some.

How can we decide what are the fundamental laws of the mind? If they could be deduced from newral architecture, then they would not be fundamental. Are there fundamental laws of the mind? The answer is not obvious. Some people believe that the world is understandable at every level. In other words there many levels with their own fundamental laws. It is likely that the mind cannot be understood from elementary particle physics. So, Newtonian laws of mechanics cannot be deduced from superstrings. At which level in the 'hierarchy' of the Universe there are fundamental laws? This is a matter of genuine scientific discovery. Myself and some other people think that there are fundamental laws of the mind. Why do I think that way? It is a combination of intuition and many years of experience and thinking. Finally, it has to be theoretical explanations of a vast extance of cognitive science from few basic-fundamental principles, explanations leading to experimentally verifiable predictions, and finally experiments confirming or disconfirming these predictions. Newton knew that his theory was correct because it explained a lot of things AND it satisfied his intuition. He even said: "I am not making hypotheses" - he believed in his intuition. Still he waited for years, untill new measurements of the Moon orbit confirmed his predictions. Same Einsteinian theory of General Relativity predicted so many things from a single fundamental law, and corresponded to his intuition, so that he could have been sure. Still he waited for measurements of light from stars deviation near sun to confirm his theory. Today, after many predictions have been experimentally confirmed I am more certain that fundamental laws of mind that corresponded to me intuition indeed are the fundamental laws.

Where these laws came from? Of course knowledge of neural architecture, psychology, behavior, are needed for understanding the laws of the mind. Still, decisions about what these laws are cannot be 'deduced' from other known things. When there is a theory based on few fundamental laws, explaining a vast amount of data in cognitive science, and in addition making new predictions, which are confirmed in experiments, than one can think that the intuition leading to these laws has been correct. That indeed there are fundamental laws of the mind. Among these laws there are mental representations as a foundation for perception and cognition, the knowledge instinct that drives the mind to match these representations to some parts of reality, aesthetic emotions corresponding to satisfaction or dissatisfaction of this instinct, and to the beautiful at the top of the mental hierarchy, the approximate hierarchy of mental representations, bottom-up and top-down signals at every level of the hierarchy, and 'vague-to-crisp' laws of interaction between these signals. Than, specific laws of interaction between cognition and language, which explain several mysteries about language (say why kids learn language at young age, but their cognition laggs behind...)

Best,

Is this of finterest?

I would appreciate comments.

Leonid

Dear Lev I think - as Dr. Leonid said - the general approach is similar to that of physics. Although we are not sure about Quantum Mechanics for now, but the measurements and calculations have satisfied our mind. I told in the last comment while we can not measure thinking, senses and minding through out imaging devices there is no hope to the procedure I said. But you are right this procedure is rare to be done but I think there is no probability equal to zero! But for sure, by today's knowledge, the answer is "no modeling can be done".

Dear Dr. Leonid

Do you think mind modeling takes the procedure such as quantum mechanics? I mean in quantum first is experiment then laws would be extracted from.

Dear Bashir,

There is an interesting discussion of this issue going on now in "Physics of Life Reviews" (an Elsevier journal).

Dr. Bernard Baars published a review 'against' Penrose and Hameroff view that quantum processes are essential for understanding consciousness. The main idea is that there is no experimental data suggesting a need for quantum processes in understanding consciousness.

In the February issue, Roger Penrose and Stuart Hameroff will publish a review rebuttling the Baars criticism.

I am looking forward to see this counterargument. This discussion between world experts on consciousness is fascinating.

Best

Leonid

Thank you Dr. Leonid. The answer was so interesting and I would like to keep on this issue.

Dr. Leonid, excuse me for asking too many questions, but your answers encourage me to ask again. Also I am a PhD student of Electrical Engineering and I like to define my proposal in common with Bioengineering and physics.

Is there any researches concern with implementation Artificial Neural Networks on devices? Such as CMOS, Optoelectronic devices etc.

I don't think quantum phenomena play a role in consciousness.

Consciousness must be understood from memory.

If conscious experiences would not be written down in some

kind of mental language, they would vanish immediately

and play no role. So memories of conscious events must

simply be memories, possibly with a label 'conscious'

attached. Nothing mysterious.

Regards,

Joachim

I don't think present physics can help us in this respect. Again, as I keep repeating, our science (physics) started on the basis of *spatial* motion of bodies. In that sense not much has changes with respect to its foundations.

With the mind, we are facing an informational phenomenon, not a spatial one. So we need fundamentally new formal language to address it.

Dear Bashir,

I am sure there should be h/w implementaion efforts.

Joachim,

I would add that it is necessary to consider interaction between conscious and unconscious processes.

Lev,

There is a lot of work done on consciousness, including predictions and verifications. I do not think a formal language could help. Mind is not a formal devic.

Bashir,

I am sure there is a lot of work in this area.

Joachim,

There is an interesting discussion of quantum processes importance for consciousness in "Physucs of Life Reviews" Elsevier journal

Lev,

I am not shure formal languages would help. The mind is not formal.

Thank you. It is a nice issue!!!!

Leonid,

More accurately, by the "formal language" I mean fundamentally new formalism for structural object representation. We need to understand how the information about evolving objects is stored in Nature, i.e. how they are represented.

Lev,

in my model this is a process "from vague to crisp" and "from unconscious to conscious" (representations, plans, actions...). It has been experimentally demonstrated to be a good model to what actually happen in the mind-brain.

Brain (mind) use reality and fiction. And mixed. Sees reality, imagine and act. Have control of real situation (awake for example), and not. When begin to sleep you need not control. You must take into account this.

Some people think that reality exists in the world indepemdent of people's minds. Other people think that from psychological and cognitive points of view everything that exists in people's minds are parts of psychological reality, and "fiction" could be no less important for understanding mind than "reality." In this my short comment I addressed how the mind can create representations corresponding to events in the surrounding world.

Dear Lev

I agree that fundamentally new formalism for structural object representation is needed, maybe the generalization of real patterns of how we represent our inner experience and behaviors like what you mean, in my understanding.

Dear Joachim,

I agree that we definitely have our mental vocabulary. But I don’t think that the conscious experiences would vanish immediately if you don’t write down, some of them would store at somewhere, once they are triggered, you can recall them. And I think those are the experience that we can have access to.

dear Leonid,

if we can find out what the fundamental laws are for our minds, then how do you explain the mind and body problem. Do we still face the problem of overdetermination, the issue of mental causation? Or we can redefine the problem of mental causation?

and if a physical theory of the mind is possible, can we use that to address the causal dilemma we are confronting?

The physical instinct you proposed is very interesting.

Maybe a stupid question, can you explain more about the dynamic logic? how does the dynamic logic work? Like DL explains how classic logic emerges in the mind from illogical neural operations. I think it is a very interesting perspective to think about the minds.

also, when you mean illogical neural operations, does it imply that those operations are unpredictable ? If DL works well, how can we fill the gap between the illogical neural operations and the dynamic logic?

And do you think maybe there will be the ontological step for the emergence of novel logic system from the illogical neural operations?

I don't know too much of your work, but I want to understand more.Hope the questions are not stupid.

Variables are enormous. The strength of the experience at the moment of memory, the similar or different situation, and the combination with all brain all moment. It´s a big computer.

Dear Yanxia Feng,

Your questions are perfect, right about the most important issues.

The mind and body problem exists in classical logic (or simply logic). But logic is a very poor model of the mind. As Godel has shown, logic is self contradictory and "not logical." Related is the problem of mental causation. I published few papers on this topic - they are open, look for (perlovsky and free will; - e,g, in SCIRP, OJPP journal).

The physical theory of the mind does addresses the causal dilemma we are confronting, and some of these problems have been solved, other require more effort.

Dynamic logic is a process logic, a process from vague (mental representations, plans, memories...) to crisp ones. You can easily find several of my papers describing it. Much of these processes are accessible by consciousness. Only the final states of the dynamic logic processes (about 0.001% of the mind states), are approximately logical states and are accessible by consciousness. For this reason logic is a very poor model of the mind, it describes about 0.001% of the mind states, but we are conscious only about these states but consciousness convinces us that there is nothing else. These dynamic logic processes have been proven experimentally to be an adequate model of the actual mind processes.

By "illogical neural operations" I only mean that they are very different from simple "AI-type" logical descriptions. DL computations can be logically-computationally described, if you know your aim: DL. DL is computable in every short step, but one cannot compute a decision from neural firings - it is different but similar to "chaos" problem, small uncertainties accumulate. This is why the mind cannot be modelled from neurons. Like planet motions cannot be computed from superstrings dynamics. (For simple systems there are classical asymptotics in quatum mechanics, but for moderately complex systems it cannot be done).

Indeed DL "fills the gap" between the illogical neural operations and perception / cognition. It will take a lot of effort to simulate an ensemble of tens of thousands of neurons to get perception-cognition, it might be interesting, and by knowing the answer (DL) it might be possible over few levels in the mental hierarchy. Probably not from the very bottom, perceptual cortex is very complex, but say, from visual features (colors, lines, corners...) to objects, to scenes.

About "the emergence of novel logic system from the illogical neural operations" - this seems to be the dynamic logic. It describes a lot of things about mind, does not seem to predict wrong things, it is computable, it makes dozens of predictions, and some of them have already been confirmed experimentally.

I think we need a high-level systemic approach, i.e. a top-down approach, to be able to develop a holistic process model of the mind and understand high-level cognition. Such an approach may be complemented with a bottom-up approach in concretion of the model. E.g. modeling interaction (bottom-up) may give insights to the macro-level, furthermore the connection from the model of the mind to the model of interaction may give insights to social cognition ("society shapes self and self shapes social behavior" Stryker, 2006).

I think a neuron-based simulation (e.g. the human brain project) would not bring essential information how the system of or mind works. We need a model not a replica of our brain/mind.

I am also agree with Samer Schaat, a neuron-based simulation would not bring us light about how information is really processed as a whole. Neural circuitry are necessary for high-level cognitive processing, that is, those processes that reduce the large ammount of information of the environment but increases its relevance. The decission of crossing a street can be thinked of only one bit of information (cross/do not cross) but affected by a huge ammount of information (cars, movements, distances, our previous experiences....) and the importance of this decission is crucial for us !. But, how the consciousness is finally built? Maybe R. Penrose is correct and this last linking process (of a reduced ammount of high-level information) is not algorithmic...

Dear Samer and Joseph,

Einstein wrote that he valued thermodynamics because it was a science defined on an intermediate level, it was not based on physics of atoms and molecules, still it led to experimentally testable predictions. This is the definition of science.

The model of the mind has to use models at multiple layers. There can be no single-layer model from neurons to cognition. In this i agree with you. Still, it is possible to develop multi-layer models explaining much about mind.

The extension what missing from mathematical (even dynamic) logic are the clear conception and formulation of induction by generalization and analogy (see works of George Polya) as new reasoning methods (similarity based vague approximations), and the mathematical conceptualization of a multidimensional world. Multidimensional is not the same as multivariable as often stated in mathematics hence 'variables' are dependent in contingency, proximity, and time flow in the real world. That is why (example for analogy) automatic 2D to 3D transformation is a hard task but the brain; literature uses multivariable approximations for a multidimensional problem. Einstein's thermodynamic model is a fine example of generalization based induction which can not be inferred by well-known mathematical models from behavioral analysis of set of particles.

I presume based on our works on real time video based information capturing, if one can capture e.g. the notion of an apple whatever apple means and how may appear in a picture, in a video, or in a painting(!) then one have a chance to understand how information processing of the brain works.

About 'illogical neuron operations'. 'Illogical' things are probably still 'reasonable' ones only the 'reasons' which may contradict to logic are missing (see Asimov).

On the multi-layer models: a multi-layered model can be transformed into a single layered one in mathematical sense (by adding some 'if-then's, like an Universal Turing Machine) using the proof of equivalence between multi-world logic and first order logic. Am I wrong about that?

If start from the original question, I believe, that "by observing the neurons, brain, behaviors" we may model only "neurons, brain, behaviors" without any mind :-) I think that we may understand and model our mind only by means of two complementary tools: first - observing our mind (that possible only through introspection) and second - constructing theory of the observed phenomena on the basis of assumptions how these phenomena can occur - without reference to neurons or other biological structures (because they may only be a particular case of a much more general phenomenon). These particular cases may take us far from the essence of phenomena. Then we may discuss our models or theory with other humans for check if the theory is suit to their minds too, and we may experimentally check the theory by means of generating hypothesis (about behavior) from the theory and see if those hypothesis correspond to reality.

The opposite journey - from the superficial observation of specific system to general theory of essentials of general system seems to me less productive.

Imagine that Aristotle was in the hands the radio set Panasonic. He can watch his work, sounds, but does he even imagine any radio waves? Rather, he will build other hypotheses based on the form of details, wires between them and on the knowledge available to him - the earth, fire, air, etc. Radio waves have been understood by the other person - who watched not particular radio sets, but a simpler form of radio wave propagation and more general phenomena of nature.

I think that we may understand the mind, watching not for private and highly developed specific case of mind in humans, watching his neurons or the skull, but watching for the more general laws of nature, especially the laws of transformation of energy and information and the possibility of self-organizing systems based on these laws into something even remotely similar to mind.

So I agree with Josep L Rossello and Samer Schaat. We have to use "high-level systemic approach, i.e. a top-down approach, to be able to develop a holistic process model of the mind and understand high-level cognition".

Dear Zsoltd,

Dynamic logic models many mechanisms of the mind, and brain imaging experiments demonstrated that its predictions are correct, so tentatively, it is an adequate model.

There are many publications on dynamic logic and i would be glad to hear your opinion if you read them.

Dear Sergey,

I agree with your comments. I would add that in addition to introspection, psychology and cognitive science developed many methods for studying the mind.

Just think about unconscious mechanisms - they are not available to introspection, while they made up more than 99% of mental operations.

I would also add that both top-down and bottom-up mechanisms are important and well studied.

Leonid,

You mentioned a few philosophical principles which a science of the MInd like any other science has to follow. I agree with the ones you mentioned. When I started to conceive of a science of vision, I thought that a science of vision is possible, not only in the sense of a mere collection of experimental facts but in the sense of a well structured body of knowledge about images, visual , and neuronal phenomenology.

But there is one guiding philosophical principle that you did not mention for the establishment of any science in any phenomenal domains: STRUCTURAL HIERARCHY IS HISTORY. Any complex entity of the universe, from the global the universe itself as a whole, to the structure of the biological organisms or the structure of the classification of all present and ancient living organism HAS a hierarchical structure and this hiearchy correspond to the evolution of this phenomenal domain. Look at a cross section of a tree: the concentric structure you see there is the growth history of that tree. So if we go back to a model of the MInd, the very structure of this model, its hiearchical organization will have to mirror the evolution and ontogeny of the MIND. Any model of the functioning Mind is implicitly a model of the evolution of the Mind.

What do you think?

Louis,

I'm surprised you didn't mentioned our ETS formalism, in which your "guiding philosophical principle" is concretely realized. ;--)

Dear Louis,

This is a very interesting comment. On one hand i agree that mental representation(s) at the "top" of the mind unifying the entire mind experience have remnants of its origin in the minds of the most simple and ancient beings. On the other hand, at the very bottom of mental hierarchy, sensing abilities unify us even more directly with the most primitive beings. What would you say?

Lev, please remind what is ETS?

Lev,

I did not mentioned it because I did not want to divert attention to my own model. Yes I see similarities with your way of seeing and mine.

Leonid,

I did not look at your model in details. There is clarification of vocabulary that is necessary. You use the word "representation" and I am in the opinion that the evolution of visual perception lead to the evolution of a hieararchy of schemata image structure detection process which mirror the hiearchical structure of the image world. So all our sense-acting system are embodied platonic world of our relation with the world. I did not go in learning yet. So for me, the representation are implicit and not explicit. My geometrical framework is the scale-space and the scale-space-time framework. If you look at a human being body, the coarse-spatial scale feature of the human body are gradually detected in the coarse spatial scale image up to the more fine spatial scale image and the sequence of structure detection is enacted in the same sequence as the topological structure of this body surface, the same ontogenic sequence morphogenesis process. The structure of the visual system is a embodide platonic world. In this sense our body is a science of the world.

I answer yes to your question. I never pay attention to the neuronal phenomenology which is at the basis of your model. If my model has any validity and if your model has any validity, they should intersect somewhere.

Louis,

Thank you for this interesting explanation. Over years I moved from mathematical terms that you use to those that are used by cognitive scientists.

Your comment: " I did not go in learning yet" is fundamental. This is what set Aristotle apart from Plato. And this is why Artificial Intelligence was limited in engineering and could not explain mind.

My interest is in learning, so dynamics of representations is fundamental to my interests.

I am sure our models intersects somewhere.

Leonid,

I agree that learning is central to mammal brains, especially human brains. But the very structure of our body and of our brain in particular is our basic knowledge heritage from the evolution of life. I explain visual learning and mammal learning in general with a sense-acting learning taking place in dreams. And I explain the special human type of consciousness/thinking through a theory of imagination based on waked dreams: the splitting of the mammalian awareness into regular mammal awarness and self-enact (dream-like) awareness. My model is a neo-aristotlelian model. Like aristotle, the image-like forms are actualized. Our nervous system is an embodiement of the platonic world and self-awareness give us access to it. Socrates was saying that learning was a kind of rememberance of our past life. Our past life is the history of life on earth, it is our philogeny. Language evolution is a direct product of the splitting of awareness wich lead to theatrical communication and all the art and music particular: the connection between danse (body movement) and music . I am curious on your models.

I think that modeling the neuron configurations of the separate parts of our brain, and connecting them properly will require a new generation of computers, after all, our brains are not digital processors. But I suspect many problems will be explained upon the examination of such a system.

Dear Louis,

It is an interesting discussion.

My main interests are models leading to experimentally testable predictions. My main predictions are that

1) representations are vague

2) perceptions and cognitions (the Aristotelian actualization of Forms) are processes "from vague and unconscious to crisp and conscious"

3) representations of higher-levels in the mind hierarchy include vagueness about which lower-levels make them up (e.g. a situation representation is a collection of vague objects with vague certainty of belonging)

4) language-cognition interaction includes two parallel hierarchies of language and cognition; (4.1)language representations exist "ready-made" in surrounding language therefore they can be learned (become crisp) early in life (4.2)cognitive representations (especially more abstract than objects) do not exist “ready-made,” they have to be learned for long time (4.3)their learning is impossible from experience along, it involves experience guided by language.

5) emotionality of language prosody is fundamental to connecting sounds and meanings (5.1)it is different for different languages and depends on language inflections (5.2)languages with few inflections (like English) are low-emotional, good for science, but may lead the rest of culture to losing meaning (words are not emotional, and therefore meaningless, unless one heavily thinks) (5.3) languages with strong inflections are high-emotional, bad for science, but create strong meanings (not necessarily correct) in the rest of culture – we see it on TV everyday all around the globe (5.4)emotionality of languages determine evolution of cultures.

6) a fundamental cognitive function of music was a “mystery” (Aristotle, Kant, Darwin, Pinker); my model predicts that music helps holding contradictory cognitions (overcoming tendency to discard knowledge due to cognitive dissonance); this possibly explains a possibility of human evolution despite cognitive dissonance.

There are experimental confirmations for 1, 2, 3, 4, 6. More experiments are needed for 5, only tentative experimental evidence exist.

Do these make sense? Any overlap?

Leonid,

What do you mean by "representations are vague",

and

how can you test it, if we don't know yet what the 'actual' representations are?

Dear Duncan,

Following Einstein, I think that the world is understandable on multiple levels (= can be explained from few basic principles on each level separately).

You do not think that we need to explain the mind from atomic physics. Similarly, we do not have to explain high cognition from individual neurons and connections. It is important to know the main principles. But speed of computations today is already sufficient.

Lev,

Close eyes and imagine an object in front of you. Your imagination is vague, not as crisp in all details as you perception of the same object with opened eyes. We know that visual imagination of an object is a top-down projection on the visual cortex of the object representation.

A more detailed brain imaging experiment with identification of brain areas and the process timelines is reported in (M. Bar et al, 2006, PNAS)

Language is a sequencial way to express understand and emotions. When you understand quickly (or in a big area) you no need all the words, you need more the "image" than each word. (No the "name", but the meaning).

Dear Leonid,

Dynamic logic (as all kinds of mathematical logic) are correct and valid for reasoning. We have used it for software design verification. Logics are efficient tools, no doubt they are designed that way. My argument is all about we are missing key factors in formal models; analogy and generality based induction. Induction exists in all sorts of logic without the use of generality. We use analogies for reasoning, that is why reasonable acts are not necessarily logical ones. That is why we can choose general good (e.g. long term freedom) instead of immediate selfishness (e.g. wealth). How that can be described and be interpreted by currently known formal models? If we do not know how they may formally work, then how supposed to find the mechanism in our brain? Is it possible to learn how an alien space ship work without discovering one?

That is, we won't be able to understand or analyze the brain as a whole without a fundamentally new mathematical background unless we accidentally discover something landsliding. We will always be able to describe phenomena, to analyze certain experiments with illogical or unusual behavior, we will have some theories and models which compete to each other in some way and in different scenarios. Just like in psychology. Nevertheless, growing number of new models and theories are not about understanding but good guessing. It is like using civil codes in science and arguing in front of a jury who is right which btw is not about truth but the proper application of known rules.

Leonid:

"Close eyes and imagine an object in front of you. Your imagination is vague, not as crisp in all details as you perception of the same object with opened eyes."

My "imagination is vague" because---in the language of the ETS formalism---typically, I do not store all the events (or features) necessary for the fuller reconstruction of the object. That doesn't mean that my representation is "vague". For example, right now my ability to differentiate Chinese faces is poor, but after spending several years in China this ability will improve dramatically. Why? Because I would engage more and more events/features in my memory for the classification (and hence representation) of faces.

So our representations are not "vague" at all but *economical*: Why from the very beginning bring in unnecessarily large arsenal of events/features, instead of bringing them in gradually, as the corresponding classification need arises.

Dear Leonid,

Dear Leonid,

Here are a few comments expressing the resonance of my views with your views.

Hopefully we will move from a vague to a more crisp conception of this intersection.

1) representations are vague

When we imagine a human, think about a human , the enacted awareness is more vague than the visual awareness. Most of the details of perceptions are useless for thinking. If our thinking/imagination would be as clear as our perception we would be in the situation of the schizophrenic: confusing what is imagined with what is there. Self-enacted perception should not be confused with perception. Smell and taste cannot be imagined nor dreamed. In my model, what we can imagine is what was originally dreamed by all mammals for learning from their experience. It is mostly visual, and motor learning that can be learned by mammals.

2) perceptions and cognitions (the Aristotelian actualization of Forms) are processes "from vague and unconscious to crisp and conscious"

Cognition is not separated from Perception. For a phenomenal illustration of this refer to the excellent books of Rudolf Arnheim: visual thinking, the art of visual perception. The myth of the big divided between perception and cognition was created by Parmenides and consolidated by Plato/Cognition/idea/language/unchanging oppose to perception/visual art/change. Aristotle restablish the continuity. But Galileo/Descartes/Newton restablished both atomist and Platonist by the geometrisation of the world within the Cartesian space/time theatre. My biggest discovery has been to find the mathematical meaning of what the Aristotelean actualization of a form means and how it is related to the image world as a embodied platonic world. How the similarity of form correspond to the similarities in the biological evolution of the forms: their common descend. All image structure are hieararchy, taxonomy, with the line, circle, triangle, rectangle at the most abtract level: the perfect forms of Plato and euclidiean geometry. Actualization of a form is like in Aristotle’s model: the ontogeny of the form and like all ontogeny it proceed from the most generic to the more detail aspects. What is conscious is this surface in 3D. But all the similarities we vaguely perceived and used in our ability to make metaphor are still vaguely conscious and are in fact based on common descent in ontogeny.

4) language-cognition interaction includes two parallel hierarchies of language and cognition;

From the evolution of imagination :the spitting of the awareness between regular external awareness and self-enacted attention based awareness create a double reality: the actual and the fictuous reality of the waked dream. The fictuous self-enacted reality can be physically enacted by body movement because most of it is actually made of self-enactment of body movement. So early humans learned to split their perception of other human movement between the regular vs self-enact movements. Symbolism/Theatriical communication was created. The next few steps will proceed by standardization of mimicry into mimic words and the incorporation of voice signal for urgent alarms and for doing it in parallel of doing their regular activities. This will acceleration the genetic transformation of the larync through under the Baldwin effect which is a culture based evolutionary crane.

(4.1)language representations exist "ready-made" in surrounding language therefore they can be learned (become crisp) early in life

Language standardization has created specific languages and this has created the fragmentation of humanity into cultural groups able to communicate that became nations much before civilization. What an irony that in order to improve communication we have created boundaries where no communication occurs!. Most words correspond to pointers in our internal visual percept hiearchies. Analogy are possible because of the common descent of concepts. Words are externation of mind. Words do not express things explicitly but indirectly by allowing us to build them from our internal. Language expression by other correspond to our internal visual self-speech and correspond to our explicit memory which correspond to our ablily to create story, fiction, myths. Living in a language community means living in a culture which is like living into a common fiction where we play a definite role in the cosmos. A culture is a form of external memory and learning of such community is cumulative and depend on the efficiency to store knowledge in myths. Later the invention of writing will create an incredible increase in the efficiency of the external cultural memory and learning and permit science and philosophy which require library and large scale cultural communication.

Every big gift comes with its curse. The problem with learning language is that it transfer knowledge that we did not personnally discovered into our nervous systems. This knowledge get stored in the same way as the one we have learned or inherited genetically. It is imcompleted, sometime false, policatically motivated, cosmologically reinforcing, fragmented and render us disconned with nature. In our adult stage of development, we reminber with nostalgia time of our youth, a time of oneness where the external culture did not fragmented us. All our life we strive to unify ourself with nature and be one again and the only way to achieve this is to create an unified cosmology.

(4.3)their learning is impossible from experience along, it involves experience guided by language.

Completely agree. Imported culture is also a poison and only a great dose of creativity can transmut it. Imported culture help us think (tools of the mind) but think in the way allowed. It is a kind of prison of the mind. It almost think for us. The pressure to think along the standardization of language is very high. A culture is a cosmology.

5) emotionality of language prosody is fundamental to connecting sounds and meanings (5.1)it is different for different languages and depends on language inflections

Once language moves from gestual to mostly oral and that the standardization into more and more abstract words. The limited memorization, and the necessity/imperative to optimize the storing of oral cultural knowledge into stories has made it necessary to use speech similarities to stabilize memory. There is an intimate connection between control of the body and music and it enter into speech. I do not get your concept of prosody but it resonate strongly inside me, it sounds true.

(5.4)emotionality of languages determine evolution of cultures.

Evolution of culture is the evolution of the cultural technologies : storage (writing) and broadcasting technologies. We do not think as individual because our thinking tools are cultural and thus shared by a community. The evolution of humanity is the gradual evolution of the thinking together, collective intelligence and of the technologies which make collective intelligence more efficient. I am presently developing ideas on how to improve storage and search of knowledge on the web: Platonia intitiative.

6) a fundamental cognitive function of music was a “mystery” (Aristotle, Kant, Darwin, Pinker); my model predicts that music helps holding contradictory cognitions (overcoming tendency to discard knowledge due to cognitive dissonance); this possibly explains a possibility of human evolution despite cognitive dissonance.

I have been developing ideas on music for a long time going in the same direction but with an emphasis on the mathematical structure of the motor control of the body. It is the connection between danse and music that have suggested this line of thought. I am not yet able to flesh out these ideas.

Regards,

- Louis

Ana Maria,

On Language. It might be great if we could move our thoughts and emotions from one brain to another. But we are limited to language (+ art). May be this limitation is important for some reasons, which have not yet been well explored.Note a two-step process in used in genetic evolution: genotype and phenotype.

Dear Zsolt,

Dynamic logic is an experimentally proved model of actual processes (in some cases). But the name "Dynamic logic" is not copyrighted. people use the same name for a lot of things. From your description I see that you did not read my papers on dynamic logic. many of them are freely available here on RG

Lev,

Dynamic logic was inspired by studying the mind processes, and it has been demonstrated experimentally that its predictions about workings of the mind are pretty good, even when entirely unexpected.

May be ETS is good for certain problems, and from what you described it seems to be reasonable. Does it make predictions for unexpected unknown properties of brain-mind? I would add that dynamic logic was applied to many engineering problems and it works better than other best algorithms in these areas by 100 times or better. May be ETS is even better?

Louis Brassard

Very interesting. Let me comment along your text.

1) representations are vague

"f our thinking/imagination would be as clear as our perception we would be ... confusing what is imagined with what is there."

Very interesting comment. Steve Grossberg published something on the mechanisms involved in separating imagination with reality, but I do not remember if the issue was completely resolved.

"Smell and taste cannot be imagined nor dreamed."

This is another extremely interesting point. It might be that smell and taste "representations" are inborn, not learned, so imaginations and dreams are not needed. I wounder if professional gourmets "tasters" and "noses" might have dreams and imaginations about tastes and smells.

"In my model, what we can imagine is what was originally dreamed by all mammals for learning from their experience. It is mostly visual, and motor learning that can be learned by mammals."

My model is similar in this regard, may be a difference is in that representations can be learned from "scratch". An image of a piano is unlikely to exist in a newborn mind.

2) perceptions and cognitions (the Aristotelian actualization of Forms) are processes "from vague and unconscious to crisp and conscious"

"Cognition is not separated from Perception."

Completely agree. Your discussion beginning from Parmenides is very interesting. Somehow my mind remembers passages that inspire me and ignores those that do not.

"My biggest discovery has been to find the mathematical meaning of what the Aristotelean actualization of a form means and how it is related to the image world as a embodied platonic world. "

Very interesting, let's compare our models. Please email me your paper to

leonid at seas dot harvard dot edu.

"How the similarity of form correspond to the similarities in the biological evolution of the forms: their common descend."

Here might be a difference, in my models mental processes are described by dynamic logic, which is different from standard genetic evolution algorithms. On the other hand, I do not believe that the current standard understanding of genetic evolution is correct. For several years I am looking for a geneticist that will find genetic mechanisms "implementing" dynamic logic. Years ago it was a tabu topic. Nowdays geneticists can consider this idea of "gradient ascent" type genetic evolution.

"All image structure are hieararchy"

Here we agree.

"..., taxonomy, with the line, circle, triangle, rectangle at the most abstract level: the perfect forms of Plato and euclidiean geometry."

Here we likely differ. I think that the Platonic perfect forms are results "from vague to crisp" dynamic logic processes.

"Actualization of a form is like in Aristotle’s model: the ontogeny of the form and like all ontogeny it proceed from the most generic to the more detail aspects. What is conscious is this surface in 3D. But all the similarities we vaguely perceived and used in our ability to make metaphor are still vaguely conscious and are in fact based on common descent in ontogeny."

Very interesting.

4) language-cognition interaction includes two parallel hierarchies of language and cognition;

"From the evolution of imagination :the spitting of the awareness between regular external awareness and self-enacted attention based awareness create a double reality: the actual and the fictuous reality of the waked dream. The fictuous self-enacted reality can be physically enacted by body movement because most of it is actually made of self-enactment of body movement. So early humans learned to split their perception of other human movement between the regular vs self-enact movements. Symbolism/Theatriical communication was created. The next few steps will proceed by standardization of mimicry into mimic words and the incorporation of voice signal for urgent alarms and for doing it in parallel of doing their regular activities. This will acceleration the genetic transformation of the larync through under the Baldwin effect which is a culture based evolutionary crane."

Very interesting.

I think there is a big evolutionary gap between human language and primate vocalizations. Primate larynx is controlled from the limbic system (by involuntary emotions). In humans there are emotional centers in cortex that are partially voluntary. This neural rewiring was necessary before (in parallel?) with language emergence. Recently I came to a conclusion that there was another BIG evolutionary change necessary for language - it is emergence of musical ability (see my papers on music origin and cognitive function). These two big evolutionary changes might be the reason why it took 25 million years from a mirror neuron system in monkeys to language (if this MN hypothesis is correct). But may be you correctly described the prior step of language evolution. Certainly the Baldwin effect could make complex evolution fast.

(4.1)language representations exist "ready-made" in surrounding language therefore they can be learned (become crisp) early in life

"Language standardization has created specific languages and this has created the fragmentation of humanity into cultural groups able to communicate that became nations much before civilization. What an irony that in order to improve communication we have created boundaries where no communication occurs!. Most words correspond to pointers in our internal visual percept hiearchies. "

I would agree about words for objects

"Analogy are possible because of the common descent of concepts."

I could see common decent of languages but I do not think, let me repeat, that an image of a piano exists in a newborn mind.

"Words are externalization of mind."

If I correctly understand, i would disagree" mental cognitive concepts are learned as "internalization" of words (and language).

" Words do not express things explicitly but indirectly by allowing us to build them from our internal. Language expression by other correspond to our internal visual self-speech and correspond to our explicit memory which correspond to our ablily to create story, fiction, myths. Living in a language community means living in a culture which is like living into a common fiction where we play a definite role in the cosmos. A culture is a form of external memory and learning of such community is cumulative and depend on the efficiency to store knowledge in myths. Later the invention of writing will create an incredible increase in the efficiency of the external cultural memory and learning and permit science and philosophy which require library and large scale cultural communication. "

I think the process in the individual mind is in the opposite direction, from language to cognition. It explains a lot of things, which are difficult to understand otherwise.

". In our adult stage of development, we remember with nostalgia time of our youth, a time of oneness where the external culture did not fragmented us."

Indeed language fragments the unity of psyche. Both: diversification and unity are essential for survival. I think music helps unify differentiated psyche.

" All our life we strive to unify ourself with nature and be one again and the only way to achieve this is to create an unified cosmology."

I would think that unified cosmological myths are results of ancient undifferentiated consciousness. This corresponds to history.

(4.3)their learning is impossible from experience along, it involves experience guided by language.

"Imported culture is also a poison and only a great dose of creativity can transmut it. Imported culture help us think (tools of the mind) but think in the way allowed. It is a kind of prison of the mind. It almost think for us. The pressure to think along the standardization of language is very high. A culture is a cosmology."

Agree. Abstract concepts are learned from language.

5) emotionality of language prosody is fundamental to connecting sounds and meanings (5.1)it is different for different languages and depends on language inflections

" I do not get your concept of prosody "

Prosody contains emotionality, that is motivation, a motivation to connect sounds with meanings.

(5.4)emotionality of languages determine evolution of cultures.

"Evolution of culture is the evolution of the cultural technologies ".

Agree and the knowledge is the most important technology.

6) a fundamental cognitive function of music was a “mystery” (Aristotle, Kant, Darwin, Pinker); my model predicts that music helps holding contradictory cognitions (overcoming tendency to discard knowledge due to cognitive dissonance); this possibly explains a possibility of human evolution despite cognitive dissonance.

I have been developing ideas on music for a long time going in the same direction but with an emphasis on the mathematical structure of the motor control of the body. It is the connection between danse and music that have suggested this line of thought. I am not yet able to flesh out these ideas.

Wish you success in this.

Best

Leonid

Dear Leonid,

Your ideas on similarity driven mathematical models are great, and it reflects more cleared thoughts than we have about what we (also) are trying to achieve. Thanks for the advice to read them and more thanks to you to write them. I am not absolutely sure that I perfectly captured your ideas, or even understood the whole point. I am sure that I need more time to process them and to do some field tests to verify my preliminary thoughts.

It seems (to me, and sorry if I miss something) dynamic logic (DL) in your terms is a very very efficient multivariable clustering tool. Why do I say that? The background is all about likelihood to known, fully described patterns without the ability of forming new concepts. For calculating the likelihood DL uses expectation model and a fuzzy class membership on *all* available data. It definitely could (and it does) identify seperated schemes/patterns without context but I have some doubts about how it fits for complex structures like striped T-shirts, Picasso's apple, anaphoric references, genitive phrases, interpreting associations like in "Washington sent an official protest to the North Korean government". I guess, and it is just a guess, it won't work. Moreover, it uses variable based fuzzy likelihood which does not allow, or it at least may not reflect the real total or partial ordering between objects, and even limits the possibility to categorize them.

What is missing? Capturing objects and processing language assumes at least a *purpose*, a *function* (aka role), a *context* in which a *subset* of multisensory information can be measured, interpreted and compared to known ones. Some information become omitted depending on the purpose (or the function in a purpose), some are guessed or replaced (e.g. facial observations, object rotations, Picasso's apple etc.), yet all are perceived.

* Sometimes the difference between known and observed phenomena (e.g. when we are getting aware moods even in written languages or through chatting),

* sometimes the context (association of meanings in different places hold information) - the model itself in DL,

* sometimes the purpose (e.g. body language affects us more than verbal ones because our main purpose to capture intentions/threats/etc, and that is why we know almost nothing about the surrounding environment when watching TV) - it could be the selection of attributes in DL??,

* sometimes the object itself is the information.

DL tries to capture the latter one "only". In other words, it seems to me that DL ideally produce answers for 'What?' questions.

In image processing background elimination is the information supression tool, or the differential imaging captures the information of movements by eliminating irrelevant parts. Context (called selection of region of interest) is a hard task in images (or videos) since there is no brain to interpret what are seen or how image was captured. Purpose is predefined, usually we tries to identify/measure/classify/etc. the object through visible signals, approximations, and a possible set of geometric transformations. Living objects may have more than just geometric transformations like changes in body, having defects, being half-eaten, or through aging. They would require to handle analogies and generality based inductions. I don't see at this point how DL could overcome these problems.

Capturing and formulating analogy and generality based induction seem missing from DL. In fact, I maybe overlooked it but I did not find any "inference rules". How one acquires new knowledge? For example, in order to invent the clockwork one had to discover the wind/watermill. Would DL find an analogy between the two or could DL formulate it without limiting the number to those specific attributes? Could it capture the (functional) similarity? I think not. Or at least not without firmly focusing on the purpose. Similarly, all sharp objects are similar (or some identical) to the other if one tries to hurt someone; the function in a purpose is the same. We must face here a generalization (hurting people can be done by sharp objects), an observation (thin objects like sharp ones if they are fast) and an induction depending on the context (to scratch my hand I need that paper). In addition, there must be an ordering on functions, in this example to measure how badly we want to hurt someone.

Dear Zolts,

Thank you for the good words. I certainly appreciate your interest. You are right, dynamic logic can be applied to clustering. It has also been applied to a dozen of classical engineering areas (detection, classification, fusion, tracking, spectrum estimation, processing of signals from various sensors,...) with results exceeding all previous algorithms by 100 times and more. It also has been used for starting new engineering areas, such as web search engines with language understanding, and others.

Another view on Dynamic logic is that it is a model of the mind processes. It has been used to model processes of perception, cognition, emotions, and the entire hierarchy of learning concepts from objects, to situations, to abstract concepts, up to the very top of the mental hierarchy (the meaning and purpose of life, emotions of the beautiful and sublime, ...); language, interaction of cognition and language (the dual hierarchy of both, effects of this interaction in several areas of psychology, including Kahneman and Tversky theory, Barsalou PSS theory. It led to theory of aesthetics solving many problems that have not been understood (e.g. what is beautiful?). Emotionality of languages and their influence on language evolution, cultural evolution, evolutionary models. Models of the cognitive function and origins of music, parallel evolution of consciousness, music, and cultures, interaction between music and cognitive dissonances... In all of these areas these dynamic logic models made many predictions. Some of these have been experimentally tested and validated. Others are being tested by several experimental cognitive scientist groups around the world.

I appreciate it might sound a bit unbelievable, but when model predictions are experimentally tested, these are the ultimate scientific "proofs". You could read papers in the areas of interest to you and judge for yourself. A recent review of some of these is available here on RG:

Perlovsky, L.I. (2012). Brain: conscious and unconscious mechanisms of cognition, emotions, and language. Brain Sciences, Special Issue "The Brain Knows More than It Admits", 2(4):790-834. http://www.mdpi.com/2076-3425/2/4/790

A lot more is available in 4 books, the latest one:

Perlovsky, L.I., Deming R.W., & Ilin, R. (2011). Emotional Cognitive Neural Algorithms with Engineering Applications. Dynamic Logic: from vague to crisp. Springer, Heidelberg, Germany.

Again,

I appreciate your interest and am looking forward to future discussions.

It could turn out that what you suggest in the second question is just for erroneously occupying oneself. For the first question, I would say that it seems you need to use your mind to model the mind but that is like using your car to model a car.

Colin,

please repeat what you call the 1st question and the 2nd question.

About modeling the mind you can look in

Perlovsky, L.I., Deming R.W., & Ilin, R. (2011). Emotional Cognitive Neural Algorithms with Engineering Applications. Dynamic Logic: from vague to crisp. Springer, Heidelberg, Germany.

Thanks, Johann. It is better to ask about the origin of the mind, instead of my original question. I think, maybe for now we can just focus on the different pieces related to the brain or mind. Then if the time is right, then we can put all the things together with an appropriate approach. Hopefully we can see that coming.

Situation awareness, that is interesting. Do you work on the mental models of metaphors too?

I am not familiar with your work yet. Will read the papers you mentioned above.

Talk more soon.

Best regards!

Yanxia

Dear Yanxia,

Biology is only replicating the physics of the phenomena, it is not the physics itself.

There is enough pictures of neurons out there, the physics of what they accomplish is what has to be understood.

More over, mapping/modeling a particular person's brain, would only lead to the model of that particular person's brain.... meaning: you can not use that model to speak of other brains. Of course, there are regions that follow special tasks, but those regions are numerically immense (millions of neurons) and would depend on each individual life experience. Then, if we add plasticity, now you have no particular order or fractal-like pattern to follow.

The latest intend by IARPA to do this modeling approach is IARPA-BAA-12-05.

The study is full of inconsistencies. Don't get me wrong, I wish them the best, but the reality is: you do not understand the logic behind a computer by taking pictures of the hardware.

Hi Frank, I didn't understand then, why cognitive scientists are doing all kinds of experiments to get to know better about our brain, or some particular brain. I understand the part of "mapping/modeling a particular person's brain, would only lead to the model of that particular person's brain". However, by observing the brain, we can get to know some common features of every brain has, then maybe some certain recurring patterns can be generalized?

what do you think? It is kinda of strong to state that claim, I think.

Dear Colleagues,

The best way to understand workings of the mind is (1) to model these mechanisms, and (2) to test predictions of these models experimentally. A lot of scientists are doing this, and this way we gain understanding.

Dear Yanxia,

When I read again the claim I wrote, I have to agree with you, it sounds very strong, even disrespectful in some hidden degree. But please understand, it was not my intention to insult anyone or render their efforts pointless.

Science is cruel, it doesn't serve those who love it; it only serves itself.

This is a cataclysmic outcome that all of us who work in research (the new or failing sciences) must painfully understand. There are stories in mathematics where brilliant men failed instantly in the most brutal of ways trying the serve their very love who had no fear of betraying them.

I think the honor in science is not in achieving, but in trying.

During my theoretical research, I heard sooo many different ideas about the relationship between mind and matter, matter and mind. Most of those ideas were arguments and claims in the absence of a model, let alone a theory.

When I did find a model or theory to try to make sense of the mind—note, I wrote "the" mind—I found that the modeler or theoretician left themselves out of the model or theory.

That's a no-no. If one wants to understand "the" mind, that means that individual is not longer thinking of how to model "my" mind. That is, the correct model or theory of mind will require and involve a first-person perspective. Any other model of mind is suspect from the get-go.

As part of my theoretical research, I read Searle, Eccles, Edelman, Crick, Delbrück, Chalmers, Flanagan, Penrose et alia, but none of those folks had an adequate, complete model of my mind. I would have to say that all of my reading about neurons and the brain indicated that those things were distal to my mind. Behavior is a physical manifestation of my mind.

In this regard, my mind was one of the last things I positioned in theory. That theory proved—much to my surprise and intrigue—was that what I call "my mind" and what you call "my mind" is the same mind.

The theory proves that my mind is what astrophysicists call dark matter—exotic particles of matter found within the photon and also what we call the unlit areas of "space"—is undetectable by experimentation, cannot be isolated, and thus has been difficult to model. The closest anyone has come to detecting dark matter is the quark, which, like I mentioned above, can't be shown to exist in isolation.

Here is the model for my mind and proof for the unity of my mind:

https://www.researchgate.net/publication/233865314_Complete_and_Consistent_Theory_of_the_Universe?ev=prf_pub

Conference Paper Complete and Consistent Theory of the Universe

Hi

Many questions are asked about the mind and consciousness, "what is consciousness", "can we make a model of the mind", "Consciousness and unconsciousness", etc ...

I followed lots of discussions on these subjects, and read several theories (Sigmund Freud, Carl Jung, Alan M. Turing, Patricia Smith Churchland, Daniel Dennett, Bryan Kolb, Crick, ...), there is always a something that escapes us, something we can not touch, something we can neither quantify nor qualify lacking.

I think this is due to our current knowledge(today state of science), which is not as advanced as we think to enable us to establish a complete and flawless theory that can be verified.

Dear Erik,

If my Mind or any Mind could be completely modeled it would mean that it is an illusion, that it is not a center of decision. A model by definition is functional and so is a non decisional mechanism. So the alternative view is that there is complete model of any natural entities in the Universe. We know already that we cannot model completely an electron or any elementary particles. These are irreducible realities. What do we think that an amoebia or a human being could be completely models? Where this strange thought come from? Maybe it comes from the success of science at modeling very limited aspects of complex systems. Modeling aspects is not modeling completely the system. The map and the territory. Even a solar system with three bodies cannot be modeled in such a way that we can predict with an illimited precision the future position of the bodies.

What does it mean to understand mind?

The best scientific understanding possible, the foundation of the scientific method is a model making experimentally verifiable predictions, and then experimental verification of these predictions.

The model could be classical or quantum, exact or approximate - these are secondary to the very question of understanding. It is good if the model can specify accuracy of its predictions. But even qualitative predictions is a strong indication of understanding. When there is no predictions at all, this indicates that there is no understanding.

Does it make sense?

Dear Vilson,

Thanks for the comments. However, please allow me to ask some questions, I don't quite understand what you mean by "In fact, our bodies are configured by our choices, through self-PNL or inductively for media/culture actions."... can you explain more or is there any related concepts or paper that i can read more?

I want to understand what you are saying.

Thanks!

yanxia

If we can understand the heart (and helps) we can intent to understand the mind... All time we intent to understan "us" in our actual partiality. The mind is part of our "understand" but heart too. All depends of levels.

I think the brain has a high plasticity, about its functionality, and how keep the ideas and use this in the differents moments is what we need to know.

I think the next pass is to change words for meanings in our way to experiment the knowledge.

hello, how to change words for meanings, such as? Vilson, your idea is kinda of different, what does it mean though? :D

We have to study the biological aspects of thinking and see how the brain solve different problems.

We also need to take into account another dimension of the question which is that of modelling.

Besides the biases induced by the environment which modify our behaviour which in turn modifies chemical reactions, etc., we also need to look at the bias induced by our modelling.

Modelling is just a simplification of the real phenomena. It all depends on how simple or complex is the model. If we complicate the model it lacks predictive value, but if we simplify it too much we may loose important details about the underlying phenomena. The selection of model also depends on the biases induced by the underlying beliefs held by the scientist. This is especially true for example in AI where we try to model human behaviour, underlying brain function, etc. In this regard I always like to look at the book " Artificial Minds" by Stan Franklin which tackles a fundamental problem in this area which is the Mind Body problem. In this type of question we have three major types of views which are the:

1) Mentalist

2) Dualist

3) Materialists

Depending on the view the researcher has, you will see them try to prove or disprove a model and its underlying theory.

Hi Yanxia

My approach to the mind has been to start with observations that are shareable. Then name, order and inter-relate those observations. The result is a taxonomy with an unexpected structure. But it is homologous to the Periodic Table of chemical elements. The Taxonomy includes itself and all activities in relation to it. So it naturally includes scientific inquiry and gives insights into its relations with other aspects of social life: aspects science often tries to screen out. This notion of mind assumes committed-action is the basic feature of human functioning.

If you are interested, visit: http://www.thee-online.com

Note that the taxonomy is itself atheoretical and includes all fundamental (irremoveable) theoretical and philosophical positions. So it is not really a model. Like the Periodic Table, it invites theoretical explanation (i.e. scientific understanding), and those hypotheses (or models of what underpins the structure), possibly neuroscientific in nature, would need to be supported by experimentally verifiable predictions, and testing. This would be a more profound science than currently exists, because it would be investigating an identifiable and articulated field.

Have a look into my robot brain. I model the minimitating groups of neurons. As a result there is a fuctioning brain that learns simple taskson its own.

http://www.intelligent-systems.com.ar/intsyst/robotBrain.htm

We have to model the biophysical structure and dynamics of credible cognitive brain mechanisms and systems and then test relevant implications of these theoretical mechanisms. For examples of this approach, see here:

http://people.umass.edu/trehub/thecognitivebrain/chapter16.pdf

http://people.umass.edu/trehub/YCCOG828%20copy.pdf

http://theassc.org/documents/where_am_i_redux

http://evans-experientialism.freewebspace.com/trehub01.htm

Arnold,

Excellent. I completely agree. By modeling we can find out about the mind same way as we have found out about the law of gravitation from models of planet motion. We compare model predictions to observations.

The only problem is in developing good models resulting in verifiable predictions.

We already have some models and predictions confirmed in experiments.

And this is a normal scientific research.

Best,

Leonid

Arnold,

Leonid,

that's totaly my oppinion. I agree. A network of researchers with similar research should co-operate.

The tricky part is agreeing on what is the appropriate level of modeling to capture the desired phenomenon.

In my taxonomy, the criterion is whether the model is relevant to achieving something i.e. of practical use. This then makes it more likely to be a product of biological evolution. It also gives meaning to the experiment. A lot of neurobiology experiments have no meaning—except within the confines of the discipline i.e. they make a grant/tenure more likely. We need to do experiments that relate to how people actually create and organize their lives, especially social life.

What this means is that we need two levels of modelling, each with their own assumptions. The neuronal-biological-physical level and the psychological-social level. The latter model can be conceived as an illusion -- it probably is. But that does not matter: all of science is an illusion. It is based on beliefs that are simply accepted. The illusion that the taxonomy models is the one that people use on themselves and each other. It matters greatly.

Physical-causal models must then be tested (experimented with) using the model of human illusions (consciousness). That will create the needed integration.

What could be of more practical use to us than to have an internal representation of the world we live in that is presented from our own egocentric perspective -- i.e., subjectivity, our phenomenal world, consciousness -- and to have the cognitive-motor mechanisms that enable us to understand and change this world to our advantage? Evolution has given us the brain mechanisms and systems that enable us to experience our current world this way and to forge new worlds. The theoretical goal is to detail the structure and dynamics of these causal brain mechanisms. Our theoretical models should be described in sufficient detail so that we can perform empirical tests of their implications to confirm their validity. The retinoid model of consciousness is one such model. To my knowledge, the retinoid model is the only theoretical model so far that has been presented in sufficient detail to explain and successfully predict many different kinds of previously inexplicable conscious experiences (e.g., the moon illusion, seeing-more-than-is-there, Julesz random dot stereograms, the Pulfrich illusion, etc). Does anyone here know of a different theoretical model that can explain these phenomena without invoking a homunculus?

The mind and thought work by analogy and contrast; we cannot suppose to model the mind because a true model would not be visible. I think an important question is whether modeling attempts approach as a limit the real thing? I do not think so...

Arnold,

I know about Steven Lehar's vision model. Your retinoid model seems at first glance to have some similarities with Lehar's vision model.

Stevan Lehar claims that the world is represented in the head. On that we agree. But Lehar's vision model does not detail the biological mechanisms that can do the required subjective-representational job. Since the Lehar model is under-specified, it cannot explain or predict such phenomena as the moon illusion or size constancy without putting the explanatory burden on a resident homunculus in the brain.

This is a great question. It is probably THE great question for researchers in cognitive science and artificial intelligence. There are many ways to try to answer it, including those already contributed by others here. My attempt is as a computer scientist and a psychologist and reflects some of the 'top-down' arguments, but with the caveat 'down to where'?:

If the models are to be based on brain physiology, as your 'neurons and brains' question implies, two assumptions are at work. The first is that higher-level models of cognitive functioning (such as the experience of surprise) can be generated bottom-up from models of lower-level physiology (such as the sensation of touch), as long as enough is known about brain function in between. The 'brain function in between' will probably includes the functional role played by brain structures as well as transformational effects of distributed neural processing.

A second assumption seems to be that higher-level cognitive functions cannot be realised in other information systems (whether biological, electronic or who knows what) without lower-level functioning somehow being replicated or emulated on those dissimilar architectures. I do not find such an idea to be persuasive and, as a result, feel that modelling efforts are better placed closer to functions that the ordinary observer* would recognise as belonging to the human mind. In other words, modelling what minds *do* rather than what happens in brains. This is closer to the 'behaviour' part of your question.

From a computer science perspective, the hardware may constrain the performance of a logical system but it does not typically delimit what it is possible to do, given infinite processing time and memory. In other words, the logic and data representation can be realised on any hardware, either directly or via some machine-specific translations. In more familiar terms, you can run several web browsers (Firefox, Opera) on Apple, Dell or Oracle hardware, regardless of using Mac, Windows or Linux. So if 'mind' were like a program, it could be made to run on any computer. Thus it is hard to see the value of modelling lower-level neurophysiological processes in order to get to the interesting stuff: as you say, how our 'mind works'. However, It may in the end be true that the human mind is what it is by virtue of its instantiation in the brain as a biological mechanism, and that it simply could not 'run' in any other mechanism. This is an open question. Minds have a special property known as intensionality**; it is conceivable that minds have this property just by virtue of their biological implementation. Were that true, it is unlikely that any efforts to model and simulate the mind in other media, including box-and-line diagrams in scientific papers, would fail.

I encourage you to consider what 'mind' means to you, in order to decide what it is you wish to model and hence how likely it is that a particular approach will deliver meaningful results.

* The 'ordinary observer' is simply a reference to the Turing Test of intelligence.

** John R. Searle 1983 'An Essay in the Philosophy of Mind' Cambridge University Press http://ebooks.cambridge.org/ebook.jsf?bid=CBO9781139173452

Thanks for the great comments, Leno Watts. I think I prefer the first assumption. That is what I am doing right now. The second assumption sounds like, it is anti-reductionism. Would rather see what happens in brain. As the definition of mind is blurry. What is a mind?

I think you are right, what "mind" means to you is more important.

However, it is not surprised that you refer to the Turing test as a computer scientist.

Thank you!

Leon,

You wrote: "Minds have a special property known as intensionality**; it is conceivable that minds have this property just by virtue of their biological implementation. Were that true, it is unlikely that any efforts to model and simulate the mind in other media ..."

The hallmark of "intensionality"/"aboutness" is *subjectivity*, the spatiotemporal experience of objects and events "out there" from a fixed perspectival origin. If so, it seems to me that any kind of hardware instantiation of a mind must have a mechanism that can represent the volumetric space in which it exists, including its own locus of fixed perspectival origin in its surrounding space. I have asked many knowledgeable investigators if they know whether such an artifact has been built and none have been able to point to one.

The retinoid model of consciousness is a theoretical model of a system of neuronal brain mechanisms that can represent the world space within which the brain exists, and do so with a fixed neuronal locus of perspectival origin; i.e, a neuronal embodiment of subjectivity. In terms of its operational logic, it necessarily operates with both propositional and analogical neuronal mechanisms. So when you ask "Down to where?", my answer is down to the level of neuronal brain mechanisms. To get a better idea of the retinoid model of consciousness, you might take a look at these publications:

http://people.umass.edu/trehub/YCCOG828%20copy.pdf

http://theassc.org/documents/where_am_i_redux

http://evans-experientialism.freewebspace.com/trehub01.htm

Leon,

All levels are complimentory to each other. Experimental psychology of vision build phenomenal vision models. In the case of vision, it is important to model the image stimulus as well. Modeling the used of visual information in the overall activity of the organism is also important. Modeling how all these models could have evolved biologically is important at a functional and physiological levels. All these different modeling approach are necessary and should constraint and help each other.

Until proven otherwise Mind can only exist in biological organisms. And so far, although we can model many aspects of the relations arising between measurable aspects of our conscious experience, we have no clue why this information has to be consciously experienced in the first place. We have do not have a scientific answer for the fundamental role of consciousness. If I ask the man on the street, the answer is obvious: Experience or consciousness is necessary for us to take decision otherwise we would be zombies. But the act of taking a decision is very difficult to model scientifically. The very idea of a model of decision making denied any need of consciousness because anything that can be modeled just need the model and is simply a zomby. What the woman on the street easilty conceive from her experience and her notion of free will would demand a scientific notion of downward causation which is currently no available.

I think we do have a scientific answer for the fundamental role of consciousness. Without consciousness we could not adapt to the world around us, we would only be able to adapt reflexively to sensory stimulation of our body. Consciousness is our personal phenomenal representation of the world we live in.

Our phenomenal world is such an omnipresent and intimate presence that we fail to see it as the fundamental referent of our concept of consciousness. From the subjective first-person perspective (from within the brain), it is simply my being here in this world with all of its present and possible contents. From the objective third-person perspective (from outside of my brain), it is a transparent brain representation of the world from a privileged egocentric perspective. The scientific problem is to show how the brain is able to create this singular kind of representation. -- our phenomenal world.

Arnold,

''Consciousness is our personal phenomenal representation of the world we live in.''

The world we live in reveal itself in our perceptual body experience but not only that, our body reveals its needs (nutritional, confort, etc) and our aspirations of life also reveal themself through the multiple emotions we experience.

The third person perspective, the perspective that we communicate to each other using language and which can only exist to each of us through its translation into our first subjective perspective. Only the aspects of a first perspective that can be transduced into a language form can become a third person perspective. The language of mathematic has been invented as the ultimate language for a 3th person perspective and it is why it is the ultimate scientific language.

Louis and Lev,

I don't think that mathematics is the ultimate scientific language. Mathematics is a formal language. All formal languages are meaningless without a social language. All social languages are meaningless without the metalanguage of the semantic networks in the cognitive brain. So the semantic mechanisms of the cognitive brain provide the ultimate language, scientific or otherwise. For more about this see "Building a Semantic Network", here:

http://people.umass.edu/trehub/thecognitivebrain/chapter6.pdf