Of course, we think using language only....Without language, whatever comes in mind that is called as "Abstract thinking".....

Thinking is a symbolic process carried out in the NS of an organism.Any thing like ideas , associations, emotions....can function as symbols. In my view language is not necessary for thinking. But usage of language can add value to thinking as it can bring in additional universal cues related to one's memory and experiences .

There is something like 'comprehension before language'. Nevertheless,, when language appears it is an extremily important tool for comprehension and thought (and viceversa). See Vygotskij

What is first?

If our thinking and its necessity to talk about, or our voice and language whose make me think... (I think both) interact all time (it's impossible to not interact, but the necessity to talk about my thinking I think is the motive).

The complexity of our language is the reflect to our sophistication and capacity. (And language and though grow together, I think).

Thinking involves visualizing and/or remembering (sense signals from memory), and not only language. I think emotions produce patterns of thinking, and that language only expresses the visualizations or memories and help us communicate. However, within internal dialogue, using more accurate language can help improve our emotional state and thinking pattern.

We are mammalian, (the more sophisticated, but that is our base). We are not perfect, we are not unlimited, and we are in evolution. We have ears and voice, also eyes and skin. We grow with all our senses and they grow with our necessity. But language (not a perfect system, but one system) is an important input for our intelect now, for that response of though in a big percentage, and in the interact of thinking, relation, communication, etc. (We are our language, this evolves with us, is part important of us). The different languages input enrichment because more diversity about interprets of reality, for example, and this is good for evolution, for example. One man/woman cannot dominate all the human knowledge, all details of all brains, but can look the way to can understand in a simillar language, for all, the bases to can found or understand all others. (Metrical system, times, maths... etc.).

But we are in evolution. Language needs so many words to explain this, for example, and to expose feelings more quickly is possible the future. (Because we are in evolution, and at less all time we need to expose feelings and data that we need to transform to communicate to others).

Comprehension before language mentioned by Ilaria is a significant part of this mystery. We learn ideas from language because in language they are clearly formulated. Can we learn ideas from experience? Experience is continuous, there is no "ideas" in the world, we have to find them in our experience

Voice contains not only language ideas, but it also affects as emotionally. This point brought up by Ana Maria is interesting. Animal's voicing is emotional. Do emotions in human voice help or interfere with thinking?

Look a baby beginning to talk.

OUR ideas COME (be, born, etc.) in us because our experiences, because us and our "move"...

Without language, when you touch a velvet, you have many and nice "ideas"...

When you (YOU) smile (smell, sorry, if you smile a flower you have other concrete idea...) a flower (YOU need YOUR ideas about it).

A rose smells well although you remember a bad day with this rose. The ideas can be differents with differents persons because their experiences, but most smell well the rose.

But yes, we can learn and "make" our ideas because our experiences. (s-t limit). (moment+feeling)

Leonid, "Do emotions in human voice help or interfere with thinking?"

Of course.

Say the same word soft, or screaming.

Say a phrase looking for look up, or looking for outwit. (The "effect" often contrary...)

Say the phrase being a good and known person, or being a bad or unreliable person... (this is with more variables, the effect often contrary too).

"Thinking" is cognition outside the brain's default mode, and it involves problem solving. To the extent that some problem solving involves language, the answer is yes. To the extent that some problem solving involes images, the answer is no.

Ana Maria,

Very valid. I was also thinking about lower level emotionality of voice, which is always present unconsciously even without specific intent.

Ahed,

Do you think we would perceive all of these even if thee is no corresponding words in language?

When speaking about the language of the mind we should not cling too much on words. Words are only one manifestation of symbols. Symbols can also be images, sounds, smells, and many other. Up to 7+-2 symbols can be bundled in working memory to form a package that can later be stored as a whole in long term memory.

Symbols can also be generated internally. If you know Lisp: it's like the gensym function that creates a new symbol on demand. The expressive power of a symbol: a symbol can be associated with anything; this association may be used to define the semantics of the symbol.

For example: how many symbols are really necessary to express "Later I will go to the bakery and buy five buns"? It depends. Zero, if the action is implicitly clear; one if you have already a symbol associated with this sentence, or some more (e.g. for the number of buns). But I don't think anyone would have the complete number of 11 symbols in his head when thinking about it.

Astonishingly deaf-mute people, when inventing language, tend to use triples of words to form sentences, although 7+-2 would be possible. Maybe thinking works likewise.

Related articles:

http://www.newswise.com/articles/view/542102/

http://www.pnas.org/content/early/2008/06/30/0710060105

http://www.eurekalert.org/pub_releases/2009-02/uoc-dcu020309.php

http://rochester.edu/news/show.php?id=3610

http://www.pnas.org/content/102/7/2661.abstract

http://www.livescience.com/10628-brain-spots-handle-sign-language-speaking.html

http://phys.org/news121536271.html

Regards,

Joachim

Leonid, "lower level emotionality of voice... without specific intent".

Basic level of our voice I think said many about US. Not only level, ritm, pauses... but say too our way of looking, our positions. (What kind of person we are). Simillar voices can be in very different kind of persons (that we like or non). (A bad person that "I" known can make me reject a simillar voice in other). (Me). A voice acute, soft, is of a child or a woman, or a protectable person, normally, for example.

About "perceive more than words exist",

Yes, all time. Voice, words, language are only a color of all the colors of reallity, and not only we perceive more than words exist, we can transmit more than words exist, too. (With my bad English I make this all time...)

Joachim, only with three... (nice).

I need to summarize words because my bad English, and I habituate to this, and I found this a nice way to talk.

I have this custom before, because my way to write (beginning with poetry, many feelings in little words number). (People said to me "gran poder de síntesis").

It's a good way to accustom to reduce in little number of words "all" we can said, or we want to said. (I think).

Vigotski did a nice job relating language and thinking. His work is actualy more popular in west than in his country.

Dear Joachim,

What do you call symbols?

Sometimes "symbol" is used for street lights, or a mark on a paper. Sometimes "symbol" is used for an idea, which affected hundreds of millions of people over millennia (like crucified God). I think in these cases the same word is used for entirely different meanings. Do you agree? Please clarify what do you mean by symbol?

Ana Maria,

Of course you are right, emotions in a person's voice tell a lot about this person. This could be the main reason why people talk causally - not to find more about quantum physics, but to find out if they fill compatible with the other person.

Ahed,

it is a very interesting proverb. I do not know Arabic, but from what I know, it seems that Arabic is a highly emotional language. To put it another way, words in Arabic are strongly connected to thoughts (emotionally connected). Strength of this connection is different in different languages, it is weaker in English, a bit stronger in Spanish and German, even stronger in Russian, but I expect it to be much stronger in Arabic. Do you have your own feeling on this matter?

Darko,

I agree, Vygotsky was ahead of his time. It was a while since I read him. I do not remember if he discussed what is the difference between language and thinking, why children master language much earlier then they master thinking and how this is possible, and other topics we can discuss now.

Leonid,

I am glad that local spelling of his name was not a problem. Vygotsky was really popular mastermind but not of his time. It is unfortunate that he died so young. I was referring to his book Thought and Language. He proposes (observing children development) that thought is internalized speech. In young children you can observe that they need to verbalize what they are thinking and sequence of what they are doing. If I need to understand complex text I read it aloud. Bad readers move their lips while reading, indicating lack in internalization. The process is supposed to be as follows: meaning of the external symbols is provided to the child by their parents. Symbols are connected to external verbalisations. Firstly child is able to operate with external objects and verbalisations. Gradually words saturated with meaning become internalised symbols and child is able to operate with them in his mind, with less reliance on spoken words and external objects. It is fascinating when you recognise these processes in your own growing child.

Dear Leonid,

> What do you call symbols?

> Sometimes "symbol" is used for street lights, or a mark

> on a paper. Sometimes "symbol" is used for an idea,

> which affected hundreds of millions of people over millennia

> (like crucified God). I think in these cases the same word

> is used for entirely different meanings. Do you agree?

> Please clarify what do you mean by symbol?

my preference for a definition of symbols is:

A symbol is something atomic that can be used

to reference something else.

Like in a dictionary: the symbol is on

the left side, "something else" is something

that could be written on the right side,

whatever it might be.

Symbols have (by definition) no parts because otherwise

one would be able to take the parts and let them

reference further things. Chinese characters are

good examples of symbols. They aren't letters

nor syllables; they stand for things, ideas,

concepts, objects.

A symbol gets its meaning by assignment

and by agreement. Outside one's mind

symbols often have a meaning that was assigned

by tradition and convention. The meaning

of a symbol is defined by mapping. Since a

symbol might reference anything including

another symbol, there's no "social difference"

among symbols: symbols that reference divine

or philosophical concepts are of the same nature

as symbols that reference humble things.

Information theory and symbols: symbols must

be distinguishable. They are composed of bits.

How many? Enough. The amount of referencable

things grows exponentially with the number

of bits. About 20 or 30 bits for one symbol

should be enough to reference any memory

in the lifetime of a brain.

How the brain might spawn new symbols:

maybe by making a partial state an attractor,

or by using grandmother neurons.

Regards,

Joachim

Just to use analogy of any scientific field, if thinking can be regarded as theory then language is surely the mathematical repesentation of the theory. Maturity of any field of science is measured in terms of the sophistication of its mathematical repersentation

While thinking is by and large an idependent process, but there is another set of thinking required to fit that thinking into limited set of language. Many creative writers or speech givers almost always endup by surprising themselves by the quality of there speech or writeup. This process of creative expression is something enjoyed most by creative writers themselves and it is magic to most of them.

When you said symbol I understand it and summarized like a chinese character "to look" this.

We can communicate without words but we need symbols too. More symbols we can mix, more rich our nuances in communication. I think this is important in our intelectual development. To "write" the symbols make us to "remember" easily while develop an idea. (We can look this in maths, with abstract and combinated ideas if we look too, not only think, we can understand better).

To write the symbols I think was very important in the intelectual development, (although we lost memory, because before we pin it all on thought).

Darko,

Thank you for reminding me about this Vygotsky thought.

Internalizing speech might be a significant part of what people are doing. Still, how is content of speech related to the world? How do new thoughts emerge?

Johann,

I agree we are smart and can understand a lot. But calling this "symbols" or "symbol manipulations" does not explain anything. May be you can explain what do you mean by this? That is exactly what I would like to understand, how our brain does it?

Joachim,

You suggest that a "symbol gets its meaning by assignment." This is exactly the process that I try to understand. One explanation you offer is a dictionary. But most kids relate words to objects without reading dictionaries. And in much of the world over thousands of years most of kids did not go to school. So, "assignment" does not sound as a convincing mechanism. It is a word that Locke used, it is a nice word used by many linguists. Still nobody can explain how it works.

Satinderjit,

You nicely expanded the question, but where is the answer? How people do it? Some are satisfied by saying: people have such an ability. I'd like to know a bit more about the mechanisms of this ability.

Ana María,

I just wanted to emphasize that the word "symbol" is used for entirely different things, for trivial objects and for psychological processes connecting the unconscious and consciousness. This creates a confusion, which proved to be very dangerous for science over the last 60 years.

Johann,

Our understanding is still too far from neurons. But it is sufficient to understand the difference between simple objects and complex psychological and cultural processes. Unfortunately even this basic difference is often unnoticed. Many people say "symbol" as if it explains something.

Joachim gave exact definition of his usage of "symbol." But then he discusses "symbolic ability", and this goes well beyond his definition of "symbol."

If of interest i wrote few papers on symbols:

Perlovsky, L.I. (2006). Toward Physics of the Mind: Concepts, Emotions, Consciousness, and Symbols. Phys. Life Rev. 3(1), pp.22-55.

Perlovsky, L.I. (2007). Symbols: Integrated Cognition and Language. Chapter in Semiotics and Intelligent Systems Development. Eds. R. Gudwin, J. Queiroz. Idea Group, Hershey, PA, pp.121-151.

Perlovsky L.I. (2010). Physics of The Mind: Concepts, Emotions, Language, Cognition, Consciousness, Beauty, Music, and Symbolic Culture. WebmedCentral PSYCHOLOGY 2010;1(12):WMC001374; http://arxiv.org/abs/1012.3803

Perlovsky, L.I. & Ilin R. (2010). Grounded Symbols in The Brain, Computational Foundations for Perceptual Symbol System. WebmedCentral PSYCHOLOGY 2010;1(12):WMC001357

Perlovsky, L.I. & Ilin, R. (2012). Mathematical Model of Grounded Symbols: Perceptual Symbol System. Journal of Behavioral and Brain Science, 2, 195-220; doi:10.4236/jbbs.2012.22024; http://www.scirp.org/journal/jbbs/

I know that "symbol" word has many meanings, and in my language a little more likely. But I know what Joachim was talking about with this "perfect" in my opinion word.

I tell you what I feel the "symbol" simillar to an image o picture, mentioned by Johann. But this can be more.

I feel a word, or a symbol (a word can say a thing, a feeling, a quatity, etc.etc.etc.) (A symbol can be a little phrase that represent a complete idea easy to transmit, for example). (A symbol can be a thing without words, or whose need many words to explain this only symbol, but can be a complete idea to use in the mind to create the way to explain our thought). (You can say "still not a symbol").

When I think to transmit I need to specify the chaos or mixture of my ideas, I can "look" like an image (an apple before to say apple), or feeling me in the garden if I talk about a rose, etc.

Because I draw and write normally when I feel a nice moment I look for the nice words or draw to put it, it's like a specialization, and I look for reduce the big moment in a little number of words.

Now I can think in words more than before, where I feel the same in a "mix of points and lines" (to say as it somehow).

(Hope that helps).

Some time ago I did observe my dog sleeping and found, many times, that he certainly was dreaming. I say that because he was like crying, or simetimes making sounds of happines, may be with soft barking. To me this means that at least dogs have there internal mental activity, and the extent of this activity as compared with humans nobody knows, I think.

Leonid,

Yes you want mechanism of ability to think and using language. I bet if anybody can have clear answer of that. is there only one answer? Does all minds works in same manners? Even if somebody comes out with some answer I doubt if there exsists a mechanism to validate it. The plain truth is that the working of human brain was more or less unknown. I have an even more fundamental question - Is it even possible for brain to understand itself fully?-which I have raised in this forum.

The question , however has engaged the mind of various scholars throughout the generations. Here is one non-mathematical take on it by French writer Porust. He of course limited himself to the process of writing and how the magic works...

"The beautiful things we shall write if we have talent," Proust says, "are inside us, indistinct, like the memory of a melody which delights us though we are unable to recapture its outline. Those who are obsessed by this blurred memory of truths they have never known are the men who are gifted ... Talent is like a sort of memory which will enable them finally to bring this indistinct music closer to them, to hear it clearly, to note it down ..."

But to get the exact answer in mathematical notation or any form......, I think is tall order. Are you sure mind can work itself out?? I mean in mathematics or othervise you cannot use the function or concept to define itself.

Leonid,

> But most kids relate words to objects without

> reading dictionaries.

kids create dictionaries. We are able to store and recall

patterns. This is done with autoassociative memory.

Autoassocative memory means that the key is the part of

the whole pattern. To recall a pattern, a fraction of

it must be known and fed into the memory device. If enough

is known (say 1/7 of the area), the rest of the pattern

is retrieved. BTW, this is, IMHO, the reason for our

working memory capacity. A symbol must be enough of a

fraction of a pattern to conjure up the rest of

a symbol tuple.

So "dictionary entries" are implicitly made by storing

patterns of symbols that occur in everyday life, but

might also spring from thinking.

A study that goes into that direction:

http://www.upenn.edu/pennnews/news/penn-research-overturns-theory-how-children-learn-their-first-words

Another approach to assign meaning to symbols is with

functions. If s is a symbol, there could be one single

function f that calculates the meaning of a symbol f(s).

Not very different from the dictionary method.

f(ELEPHANT) would determine the meaning of the symbol ELEPHANT.

Or there could be as many functions as symbols. For every

symbol there could be a function that determines the meaning

of the symbol: ELEPHANT(), CAT(), DOG().

The meaning of most symbols is context dependent. So we

could introduce a parameter for the context c:

ELEPHANT(c), CAT(c), DOG(c), but we might as well omit

the context parameter, since it's implicit and always needed.

A third method of assigning semantics to symbols is via

networks. One of my research interests. Suppose we encounter

a new symbol s. If s has a meaning similar to t,u,v,w then

the meaning of could be defined by connection s with these:

s-t, s-u, s-v, s-w.

It's actually a recursive definition: the semantics of a

new symbol is defined by other symbols whose semantics

is already settled.

Like here:

http://robots.net/article/3500.html

http://gallantlab.org/semanticmovies/

Other literature:

http://www.sciencedaily.com/releases/2012/11/121108140847.htm

Regards,

Joachim

Johann,

Crucified God is a complex idea because it has been at the center of our civilization for two thousand years. This idea is central to thousands of books, and to many topics of our culture. Our entire culture evolved the way it is today much under the influence of this idea. It is just one example of using a word "symbol" for complex socio-cultural processes. For example, "free will" is another complex socio-cultural idea, a symbol. "Human rights" is another example. In all these cases a word "symbol" is applicable, yet it is entirely different from a mark on a paper, which also can be called "symbol"

My point is that the word "symbol" is used with significantly different meanings.

It is a great simplification to say "symbol" and to hope to explain much, or to spell crucifiction in an attempt to understand 2000 years old cultural history of the western world

Ana maria,

You nicely amplified one psychological and motivated aspect of the word "symbol", but you left out another one: "symbol" as an arbitrary unmotivated pointer to something else.

Sachchidanand,

You gave a nice list of different things. I would like to emphasize that they are all different and require separate understanding.

Satinderjit,

It is an excellent point, understanding the mind and language is a difficult problem. This is why we all discuss it for so long. Proust is a great writer and a deep thinker. But here we do not go to the full depths of analyzing the mind the way he did. We try to understand the mind scientifically. How far we can get and where is the limit - nobody knows. We certainly can understand scientifically a lot of things about the mind.

Joachim,

"kids create dictionaries... with autoassociative memory." This might sound more scientific, but in fact, exisiting mathematical algorithms or neural networks called autoassociative memory do not accomplish what you assume they do.

A simple explanation why they cannot do it is as follows: there are thousands of words and thousands of objects. The number of possible associations between words and objects is too large. It is much larger than the lifetime (measured in seconds or in milliseconds). Therefore there is never enough experience to learn associations by using any auto-association memory.

Nevertheless we do learn to speak and to understand what we say. My interest is how this happens. And what exactly do we learn?

Let us think, your explanation "with autoassociative memory" does it explain why kids learn language early in life, whereas similar mastery of understanding of the world comes much later?

What exactly comes later? We know that kids can talk like adults without 'understanding like adults.' What exactly this mean? In which way adult understanding is different from understanding of the child? These are simple questions, but they defy simple answers.

Joachim,

Let us see if your network approach explains thinking. "If s has a meaning similar to t,u,v,w then the meaning of s could be defined by connection of s with these:

s-t, s-u, s-v, s-w. "

The very first question is how to define and measure a similarity of a new idea to old ones? This is a big problem, the entire thinking process might be needed. But let us leave this difficulty for now.

It seems that every idea is a combination of old ideas. Creation of 'really new' idea is not possible. Of course not everybody has to be creative. Even explaining 'simple' thinking with old ideas is a difficult problem. And then where the original 'old' set of ideas came from?

I appreciate these are complex questions. But there should be some approaches to solving them.

Leonid,

Ofcourse, we want to study the mechanism of human mind in scientific manner. But we should take input in this regard from work done in some other fields ; be they may not be following scientific notation, terminology and discipline. I think theology is full of work done to understand,tame, control and manage mind. The sufi tradition in larger Islamic world is great example of body of work dealing with minds. The literature is full of such issues dealing with ambiguties and uncertinty of mind.

We in the scientific world are so absorbed in our own little world that we tend to ignore the large body of work done by others , on the same isues that we are struggling to find foot on. Part of it is arrogance of scientific world but larger part of it due to the fact that in moderen world we have done well in studying human exsistence by dividing it into some fields or subjects. But we have forgotten the art of combing this division......

Leonid, with language you need a context because if not you need infinites words, and this not help to comprehension.

I speak Spanish, all this words are new for me, and I only can say that when you said "symbol" I understand exactly what you are talking about, better than image, picture, feeling, word, phrase. (The same that when I read "information" the first time about a fine data of Universe).

About child in the beginning of use the language, when you begin to use (or to make with your use) a thing or a method, you need to repeat first to "look" what is this method.

Simillar than when you want to draw horses, you must copy first many horses to can make they without copy later.

Creation always has a base.

Leonid,

> Let us see if your network approach explains thinking.

> "If s has a meaning similar to t,u,v,w then the meaning

> of s could be defined by connection of s with these:

> s-t, s-u, s-v, s-w. "

> The very first question is how to define and measure a

> similarity of a new idea to old ones? This is a

> big problem, the entire thinking process might

> be needed. But let us leave this difficulty for now.

I'd prefer speaking of symbols. Ideas are complex

structures possibly comprising a bunch of memories.

The relatedness of symbols could arise from being

activated at the same time. For single neurons the

Hebbian rule states that connections are strengthened

if the neurons fire at the same time.

For the brain it's an easy task to translate

between a distributed representation and a

single neuron representation (grandmother neuron).

http://www.eurekalert.org/pub_releases/2009-07/uol-ols072009.php

http://www.newswise.com/articles/view/512626/

http://www.jneurosci.org/content/26/40/10232.full

On the other hand there's the working memory.

It probably does not work with single-neuron symbols.

Rather, it seems to use a distributed code.

The advantage of the working memory is that it is

not necessarily subject to external stimuli: it's

in our hand to fill it with a symbols of our choice.

That means we can associate symbols at will,

and so define similarity.

http://www.mpg.de/5005704/brain_oscillations_memory

http://www.eurekalert.org/pub_releases/2012-11/niom-ibw110212.php

http://medicalxpress.com/news/2012-11-brain-code-visual-memory-deciphered.html

> It seems that every idea is a combination of old ideas.

> Creation of 'really new' idea is not possible.

> Of course not everybody has to be creative. Even explaining

> 'simple' thinking with old ideas is a difficult problem.

> And then where the original 'old' set of ideas came from?

The linguists restrict themselves to a relatively

small set of words that can be heard spoken in

the outside world. Say about some thousands of words.

Researchers that investigate the language of the

mind should reckon with millions of symbols since

the brain is able to generate a new symbol every

couple of seconds. A new symbol is not an idea, it

is at first devoid of any meaning, but the meaning

can later be assigned.

Regards,

Joachim

Johann,

I am not playing any gym games. I am trying to understand how the brain-mind works.

I do not think it is useful to use words ambiguously on purpose. Please explain why do you think this is a good idea for any purpose?

To some extent I agree with you, in this discussion the word symbol has not been fully defined - to the extent that we do not have everybody's agreement.

Nevertheless, we should not ignore the fact that Joachim gave his definition quite exactly. And I gave mine may be less precisely, because in my mind it is more complex, may be too complex for a short discussion, therefore I gave references for few of my papers on this topic.

To give a short definition, the word symbol is used in two ways, one I prefer to call "sign", it is an arbitrary, non-motivated pointing to something else (Joachim uses this, and it is in line with Peirce), and another one that I prefer is that symbol is a motivated psychological process connecting unconsciousness and consciousness (this is in line with Saussure and Jung).

I do not think that crucifiction is "an abstract depiction of the religious part of 2000 years western history". Hundreds of thousands or may be millions of people gave their lives for better understanding for the meaning of this symbol. And if today some people may think that these previous generations have been less sophisticated than we are today - rightly or wrongly, agree or disagree - it is not an abstract depiction.

Maybe today some people fully and consciously understand the meaning of the crucifiction and this is why they are not as much affected by it and not unconsciously driven by it. Or the opposite is also quite possible - some people have lost understanding of this symbol, and do not understand what it meant for hundreds of millions of people over two millennia.

Joachim,

I agree you touched on some of the most interesting topics. But they could be more complex than it seems. I do not think it is easy for the brain to establish connections between two ideas (representations or whichever). Because the very question is where these ideas or representations came from?

One can say "grandmother neuron", but what is the mechanism that creates it? So far there is no robot with anything like human "grandmother neuron".

Similar is "Hebbian rule" - it is easy to connect two random unrelated objects, which happen to be next to each other, say a rock on the road wayside and a piece of garbage next to it. This association would not create a meaningful 'situation' out of these two random pieces.

" it's in our hand to fill it with a symbols of our choice... and so define similarity." - I agree Joachim, it is an excellent example why meaningful similarity measure is not a trivial mechanism.

If of interest, I have published few papers on why this problem have not been solved. The difficulty is related to a most fundamental mathematical difficulty identified by Godel. Overcoming this difficulty requires a correspondingly fundamental mathematical idea.

Leonid,

> One can say "grandmother neuron", but what is the mechanism

> that creates it? So far there is no robot with anything like

> human "grandmother neuron".

the advantage of grandmother neurons in live brains is that

semantic similarity can be expressed as geometric vicinity.

This makes spreading activation energy efficient. I mentioned

grandmother neurons only as an example for how the brain

could represent symbols. In robotics, semantic networks

don't play a role yet. Symbols can be expressed in the

C language as strings, enums, integer constants. No neurons

needed.

> Similar is "Hebbian rule" - it is easy to connect two random

> unrelated objects, which happen to be next to each other,

> say a rock on the road wayside and a piece of garbage next

> to it. This association would not create a meaningful

> 'situation' out of these two random pieces.

A related question is whether the combination of the rock

and the garbage deserves a symbol, say S20130408. Suppose the

observer encounters this combination more often. Then it might

be worth to have a recognizer for it. Such a recognizer would do

nothing else than to scan the current content of working memory,

and upon discovering the combination of rock and garbage

shout "S20130408".

This is the mechanism of chunking. It permits compression of

a seen scenery by replacing a couple of symbols by a single

one. Couldn't be compression of scenes the driving force

behind situation acquisition?

> If of interest, I have published few papers on why this

> problem have not been solved. The difficulty is related

> to a most fundamental mathematical difficulty identified

> by Godel. Overcoming this difficulty requires a

> correspondingly fundamental mathematical idea.

It sounds interesting. My opinion on the Godel Incompleteness

Theorem is that it doesn't set humans and machine minds apart,

for various reasons. E.g. since humans don't apply much

logic in everyday life. May I ask for your opinion?

Regards,

Joachim

Johann,

I am glad we agree on a need for unambiguous definitions.

I am not sure symbols are vector spaces (if this is what you meant by directed symbols. Emotions possibly could be modeled this way, but symbol-processes could be more complex than vector spaces.

Joachim,

I agree with you, the Godel Incompleteness Theorem doesn't set humans and machine minds apart. Just the opposite, the mind is not "logical" and logic is not "logical."

But it is essential to appreciate the logic "does not work." This have not been quite appreciated. For example, most learning algorithms use logic in the learning process, e.g. "this is a chair" is a logical statement, and algorithms using similar learning processes will not work. It is not easy to notice presence of logic. Neural networks have been designed to overcome logic limitations, fuzzy logic had similar aim. Yet, logic made its way into algorithms using neural networks and fuzzy logic.

Johann,

First of all of course finalize the paper and publish it. This is important. Next, you could look at my papers on symbols and on dynamic logic. From cognitive perspective they model mathematically advanced psychological theories. From engineering perspective they solve problems that have not been solved for decades (with limited effort).

Cognitive:

Perlovsky, L.I. (2007). Symbols: Integrated Cognition and Language. Chapter in Semiotics and Intelligent Systems Development. Eds. R. Gudwin, J. Queiroz. Idea Group, Hershey, PA, pp.121-151.

Perlovsky, L.I. (2006). Toward Physics of the Mind: Concepts, Emotions, Consciousness, and Symbols. Phys. Life Rev. 3(1), pp.22-55.

Perlovsky, L.I. & Ilin, R. (2012). Mathematical Model of Grounded Symbols: Perceptual Symbol System. Journal of Behavioral and Brain Science, 2, 195-220; doi:10.4236/jbbs.2012.22024; http://www.scirp.org/journal/jbbs/

Engineering:

Deming, R.W. and Perlovsky, L.I. (2007). Concurrent multi-target localization, data association,

and navigation for a swarm of flying sensors, Information Fusion, 8, pp.316-330.

Perlovsky, L.I. and Deming, R.W. (2007). Neural Networks for Improved Tracking, IEEE Trans. Neural Networks, 18(6), pp. 1854-1857.

Perlovsky, L.I. & Ilin, R. (2010). Neurally and Mathematically Motivated Architecture for Language and Thought. Special Issue "Brain and Language Architectures: Where We are Now?" The Open Neuroimaging Journal, 4, 70-80. http://www.bentham.org/open/tonij/openaccess2.htm

http://www.benthamscience.com/open/tonij/articles/V004/SI0026TONIJ/70TONIJ.pdf

Johann,

the problem with writing something on three-valued logic is

that there is a bunch of already existing papers. It's a work

of minutes to create a truth value table (like the attached),

so it's difficult to find something really new.

It's an empirical fact that we can store patterns of symbols

as a tuple. Here we are back to language: a tuple of symbols

corresponds to a sentence in the language of the mind.

There's a limit of about 7+-2 symbols for a tuple. Some say

it's rather about 4 or so. We can "tag" a tuple by adding

more symbols if there's a place left. Such an additional

symbol could be a truth value (true, fals, unknown), but

also a probability information. Or additional context info.

Regards,

Joachim

Indeed I used ternary values. Maybe in a different context than

you intend. In order to choose a goal, the Go program has several

"experts". Each expert can do a comparison of goals with a function

that returns +1, 0, or -1 to indicate whether a goal is better, equal,

or worse than another. The experts all belong to a hierarchy level.

Two goals are compared by calling this function of each of

these experts. If all experts vote for +1 or 0, the first goal is chosen.

If all experts all experts vote for -1 or 0, the second goal is considered

better. If some experts vote for +1 and some for -1, then additional

experts of a lower hierarchy level are consulted. At the lowest

hierarchy level is an expert that decides randomly.

Regards,

Joachim

Both are very much related. Without a Language in mind how will think? Symbolically?? and again the vice versa.... You can't use a Language with nothing in your mind.

Thinking consists of state transitions of the brain.

But what's the most appropriate level for description?

Atoms? Neurons? IMHO the content of the working

memory best describes the current state of the mind.

One could imagine a square with words in each

corner or a triangle or a tetraeder or other structures.

But the least common denominator is a tuple of symbols.

About 7+2 symbols, or just only 4.

It is not required that words in out minds are

well-formed with respect to a grammar.

That's not a condition for thinking. Simply fill

symbols with meaning and juggle them around.

Regards,

Joachim

See the Sapir-Whorf Hypothesis, which postulates that we tend to think in the languages we master.

http://en.wikipedia.org/wiki/Linguistic_relativity

Johann,

Thank you for your comment.

I am skeptical about multivalued logics because they are not different in principle from classical logic. But written papers should be published.

In my papers I answer the question to some extent, and this answer has several unexpected aspects, so it is worth continued discussions - as witnessed by our discussions here. It also opens the whole field for experimental tests.

Joachim gives another good reason. Just a review of what have been done would take months.

When you have time, look at my dynamic logic. It is a relatively new technique, using it we have solved problems that remained unsolved for decades, and it would take a month or two to write a paper.

Saumendu,

It is a great comment. This is a reason for this discussion.

Joachim describes the essence of many approaches.

The difficulty that I see is in how many ways 7 can be selected from 1,000 words, it is 7^1000, many many more than all interactions in the Universe. There is no way to learn which combinations are useful and which are not.

Bachan,

This is a good illustration of where the difficulty starts. Kids learn language by 5 or 7. This is possible because language exists "ready-made" in surrounding culture.

But thoughts do not exist "ready-made" in surrounding world.

Barry,

You are right. SWH is a separate very interesting topic. For some in-explainable reason it is a tabu for many linguists. (By the way, if of interest, I suggested that emotionality of languages affects cultures possibly even stronger than semantic contents of words).

But SWH is a consequence of language directing thinking, it is not an explanation. So the question remains, how exactly language directs thinking? A naive view is that language 'points' to concepts. But this idea is not sustainable: concepts do not exist in the world "ready-made", still this idea dominates today.

What you think?

From a linguistic perspective, the SWH is seen as problematic because it postulates a strong version of the idea that language influences cognition. To draw on just one example, Whorf noted that Hopi lacks tense, which led him to speculate that the Hopi lack a sense of time. (It seems that he was unaware that Chinese also lacks tense.) He failed to notice that if the Hopi truly lacked a sense of time they would not have been able to plant crops or understand that people grow older. As Pinker (2005) pointed out, there is no culture anywhere that does not have a sense of time.

As for the idea that "thoughts" do or do not exist in the child's world, surely some do. Concepts of right and wrong are cultural, and although some studies have suggested that children are born with a moral compass, that compass must be fine-tuned to the moral and ethical standards of their society. In addition, as Searle (1983) explained quite elegantly, some thoughts/concepts are social constructs. As I recall, money was one of his primary examples.

Leonid,

> The difficulty that I see is in how many ways 7

> can be selected from 1,000 words, it is 7^1000,

> many many more than all interactions in the

> Universe. There is no way to learn which combinations

> are useful and which are not.

in some situations we know that a combination

of symbols is important. We can then actively

improve memorization by repetition. If a phone

number or an inscription is important for us,

we can keep it in mind by repeating it periodically.

There is also a passive learning mechanism.

It's a statistical argument. A combination of

symbols is useful if the probability is high that the

same combination will occur many times again.

Last Wednesday a study was published (Anna

Shapiro et al.) where subconscious learning

of symbol combinations was demonstrated.

http://www.princeton.edu/main/news/archive/S36/56/09E19/

Regards,

Joachim

To speak properly, we have to monitor our thoughts. However, individuals sometimes say things without thinking. In this case, our brains dominate our words and we can get in trouble and hurt others. However, the monitor process in our brain allows us to say that thinking does play a role in language.

Let's look at parrots. Do they really think when they are talking? We can watch the following video to illustrate my thoughts:

1. http://www.youtube.com/watch?v=hp0gkjJLlVk

2. (Intelligent animals) http://www.youtube.com/watch?v=145ow8YCP7g

Albert Einstein Reads ‘The Common Language of Science’ (1941)

http://www.openculture.com/2013/03/listen_as_albert_einstein_reads_the_common_language_of_science_1941.html

"My original decision to devote myself to science was a direct result of the discovery which has never ceased to fill me with enthusiasm since my early youth - the comprehension of the far from obvious fact that the laws of human reasoning coincide with the laws governing the sequences of the impressions we receive from the world about us; that, therefore, pure reasoning can enable man to gain an insight into the mechanism of the later. In this connection, it is of paramount importance that the outside world is something independent from man, something absolute, and the quest for the laws which apply to this absolute appeared to me as the most sublime scientific pursuit in life". Max Planck in Scientific Autobiography.

"The words or the language, as they are written or spoken, do not seem to play any role in my mechanism of thought. The psychical entities which seem to serve as elements in thought are certain signs and more or less clear images which can be 'voluntarily' reproduced and combined. .... This combinatory play seems to be the essential feature in productive thought before there is any connection with logical construction in words or other kinds of signs which can be communicated to others". Albert Einstein in a letter to Jacques Hadamard.

Feynman continues: "What I am really trying to do is bring birth to clarity, which is really a half-assedly thought-out-pictorial semi-vision thing. I would see the jiggle-jiggle-jiggle or the wiggle of the path. Even now when I talk about the influence functional, I see the coupling and I take this turn - like as if there was a big bag of stuff - and try to collect it in away and to push it. It's all visual. It's hard to explain."

Schweber: "In some ways you see the answer - ?"

Feynman: "The character of the answer, absolutely. An inspired method of picturing, I guess. Ordinarily I try to get the pictures clearer, but in the end the mathematics can take over and be more efficient in communicating the idea of the picture. In certain particular problems, that I have done, it was necessary to continue the development of the picture as the method before the mathematics could be really done."

James ,

You gave an excellent example of one person suggesting a wrong hypothesis 70 years ago. This hypothesis (that Hopi understand time differently from Englishman) can be easily tested scientifically. But SWH is not about one example, it is a fundamental scientific idea about interaction between language and cognition. It is at least 700 years old. Scientific evidence is overwhelming. Linguists (that you mention) could study evidence, like all scientists do (and like many linguists do). I think people in the world can think differently than American professors do, it is quite a racist view to assume that everybody have to think like Americans. There is a lot of interesting research on how language influence thinking, I would like to see more.

Joachim,

very interesting.

Ahed,

It could be an interesting line of research.

Johann,

Very interesting, indeed this idea could make dynamic logic intuitively clear.

Johann,

My approach is to publish ideas, if some colleagues will take it without reference, so be it.

Jose,

Parrots are an interesting topic. There is insufficient scientific research on their brain mechanisms.