Paradigm

Excellent technical video on the Amazon Kindle Fire split Browser/Cloud architecture; that makes the Amazon Web Services cloud computing strategy very clear, and brings up a new browser paradigm:

http://amazonsilk.wordpress.com/2011/09/28/introducing-amazon-silk/

Bertrand du Castel

Feelings and thoughts

I think that on average, women are more prone to show they feelings and conceal their thoughts, while men are more prone to conceal their feelings and expose their thoughts.

Bertrand du Castel

Human-Centered Oilfield Automation

My Human-Centered Oilfield Automation paper, originally published in the Schlumberger Journal of Modeling, Design, and Simulation, Vol 2, June, 2011, and now on the web puts, in a way, the two sides of my scientific life together.

On one side, the oilfield, its huge and intricate technology, global and remote geography, and fierce pursuit of business opportunities. On the other side, the mind with its most salient human revelations, from language to religion, sensation to sentiment, gesture to metaphor, and the stochastic consistency of its inner constitution.

Unifying both sides is the computer, which extends the mind and the knowledge, actions and operations, and constitutes the framework of our new social organizations. In Human-Centered Oilfield Automation, I show how quintessential neurocomputation, supported by stochastic grammars, allows bringing automation to the oilfield by providing extensions of the mind while supporting the preciousness of human innovation, understanding, and capability to act and react in a way that is in most cases well ahead of computer progress.

Bertrand du Castel

Zimmerman's theory as evolutionary psychology

As a reminder, Zimmerman's theory of the great leap forward some 50,000+ years ago is that a confluence of two cultural phenomena could be enough to explain it. In this case, I have ventured that it could be ecstasy and ritual, or, in other terms, religion. What was lacking in this overall explanation was a plausible brain evolution corresponding to that theory. Well, an article in Scientific American dated 20 July 2011 by Katherine Harmon, Fast Evolving Brains Helped Humans out of the Stone Age, sheds some light on the subject:

A traditional, more passive take on evolutionary psychology "fails to recognize that humans are changing their environment," and not at all randomly or haphazardly, Laland says. "We've built environments that are well suited to our biology, so we don't find ourselves massively maladapted for the contemporary world."

The view then, in my terms, is that the discovery of religion would itself lead to compatible changes in the molding of the environment. Henceforth, caves in the South of France some 30,000 years ago and cathedrals in the modern time.

Talking of Computer Theology, the picture in the article is that of a computer drawn over parietal art. Of course, parietal art is on the cover of Computer Theology. We missed the computer though ... reparation done.

Bertrand du Castel

Blending and the four-layer semantic theory of concepts of generics and metaphors

Interesting article by Karen Sullivan and Eve Sweetser: Is ‘Generic is Specific’ a Metaphor?  in Meaning, Form and Body, Fey Parrill, Vera Tobin, and Mark Turner (eds.), Stanford, CA: CSLI Publications: 309-327, 2009.

Their article doesn't provide a formal mechanism describing generics and metaphors in unison but our own article on the subject does by providing a Montague's type theory for a unification in Bertrand du Castel and Yi Mao: Generics and Metaphors Unified under a Four-Layer Semantic Theory of Concepts in the proceedings of The Third Conference on Experience and Truth, November 24, 2006, Taipei, Taiwan.

However, Sullivan and Sweetser provide a means to construct the formal model of our article, using  the blending theory developed by Gilles Fauconnier and Mark Turner: The Way We Think: Conceptual Blending and the Mind’s Hidden Complexities. New York: Basic Books, 2002.

So here is another unification. Gilles Fauconnier helped me back in the 70s by suggesting a facet of an article I wrote for Linguistic Inquiry: Bertrand du Castel, Form and Interpretation of Relative Clauses in English, Linguistic Inquiry Volume 9 Number 2 (Spring, 1978) 275-289, MIT Press. Here is note 1 in the article:

¹ I am greatly indebted to Gilles Fauconnier for having suggested, after reading an earlier version of this article, the ideas that are similar to those developed in Fauconnier (1976) on other grounds.

Well, life has interesting twists. We both worked for the same laboratory in France, and then we went our separate ways, both in the United States. And we've not seen each other since then except now we have again two complementary theories -- or so I think.

 Bertrand du Castel

About time

In Scientific American of 19 Jun 2011, Why Does Time Fly? by Martin P. Paulus:

Something moving toward you has more relevance than the same stimulus moving away from you: You may need to prepare somehow; time seems to move more slowly.

This would be consistent with a stochastic grammar model of the brain. A stochastic grammar weights events in as they occur to determine prior probabilities, which in turn predict next possible events (posterior probabilities). When new events fit the predictions of highest probability, little change occurs in the stochastic grammar processing. However, when the predictions of lowest probability occurs, such as looming danger, the stochastic grammar incurs marked modifications of its weights, that in turns reset new expectations.

This change of weight in the stochastic grammars corresponds in the brain to a change in the glial and neuronal activity, which involves chemical and electrical mechanisms that are not instantaneous. So, if the activity of the stochastic grammar would be the measure that the brain uses for evaluating time elapsed, that would create a measure of the effect at hand, and a means to actually experiment with the validity of the model.

Bertrand du Castel

Strategy and tactic

Some 20 years ago, I was at a social function with my wife, where we were seated at different tables. She reported to me that a discussion took on at her table on the subject of US Football and Everybody Else Football (which the US calls soccer). One of her fellow diners made the statement that soccer was uninteresting because it was all about tactics, not strategy, whereas US Football, obviously a superior product for that person, was about strategy as well as tactics. I have mumbled on that statement ever since.

The fundamentals finally downed upon me reading an Austin American Statesman article today about the 2011 Women's World Cup raise to team parity, which prompted in me the following thought:

If you don't understand the strategy, everything looks like tactics.

Actually, that maps very well with the model we developed with Yi Mao in Generics and Metaphors Unified under a Four-Layer Semantic Theory of Concepts. In technical terms, strategy is related to intensions (with an "s"), and tactic is related to extensions; variations in intensions modulate extensions by building on generics (the situation at hand) to expand metaphors (the situation in the making). In layman's terms, strategy is close to conception, tactic is close to action. I am ready to venture that this also maps the two paths of the brain to action. Tactic is the short loop that leads rapidly from perception in the (hypo)thalamus to action in the motor cortex ("jump when you see a snake"), and strategy is the long loop that adds another step, cognition, to that. In the long loop, perception goes to the (hypo)thalamus then to the pre-frontal cortex (cognition) and then only to the motor cortex ("calm down; the snake is not dangerous after all"). Naturally, both loops occur at the same time, with a delay for the long loop. Get out of a dangerous situation, but ponder whether it's the right move.

Well, football, whatever its incarnation, is the same thing. Create the situation (strategy) to profit from (tactic).

Bertrand du Castel

Mirror hurting

Extraordinary movie: Buck

The theme is that we hurt others because we hurt ourselves. It is a beautiful illustration of mirror neurons: internal hurting and external hurting are both using the same circuitry. If we can't tame it inside, we'll try outside. But it is, the same thing.

I have written the above very carefully, to give you opportunity to experiment now in the first person what I am talking about. Take the phrase above "we hurt ourselves." As you can perceive, it is ambiguous. Does that mean that we consider that we are somehow acting at self hurting ("WE hurt ourselves") or that something unspoken hurts us ("we hurt OURSELVES").

That ambiguity between we-as-agents and we-as-recipients in that one sentence is triggered by the mirror neurons mechanism. Mirror neurons trigger both action and reaction. In technical terms, the syntax represents the conduit (the mirror neurons) and the semantics represent the effect (action or reaction)?

This is, by the way, also a new and wonderful account of the relationship of syntax and semantics. I need to ponder some more on that.

Bertrand du Castel

Zimmerman's theory in Turkey

In National Geographic of July 2011, a great article on Göbekli Tepe, the excavation of a pre-agricultural set of temples from 11,600 years ago just next to where the oldest traces of agriculture have been independently found (the origin of wheat agriculture, south of Turkey also). So both religion and agriculture found in forms similar to today, and at the same place.

That's exactly what Zimmerman's theory of the great leap forward predicted, but interestingly enough, Zimmerman's theory was in answer to the puzzle of the human expansion some 70,000 years ago, not the recent post-glacial one. So what we'd find here is that Zimmerman's theory may apply to more than one expansion, an interesting proposition indeed.

Regarding the 70,000 years ago expansion, I reported in Zimmerman's theory and phonemes that new discoveries where emerging in that part of the prediction too, although in a more abstruse way yet.

Bertrand du Castel

Computer is a metaphor

Although in Computer Theology, we expanded at length on the role of metaphors in the parallel between the elaboration of human and computer networks, and although we also related the geometrical and emotional sensori-motor systems as the root of human metaphors, we missed mentioning by name the core geometrical sensori-motor system at the center of computer programming (we did consider the specific -motor system of a computer processing unit though).

Consider the following story:

Pick up a container;
Take the road at its beginning;
Pick up a flower you like;
At the first fork in the road, take the right if you have a flower, the left if you don't.
If you took the right:
Go to the end of the road;
Plant the flowers and stop.
If you took the left:
Keep going on the road;
You'll soon be back at the beginning of the road;
Start over.

Hopefully you'll be less picky about choosing a flower the second time. But you got the point. Here we have a physical story, which matches exactly a typical program, which I now write:

start();
x = new Container;
loop {
if (random())
x.push(new Flower);
stop();
}
}
x.pop()
stop();

I have shown with a very simple example that programming is based on the exact same geometrical concepts that underlie our behavior.

Examples multiply:

http://selficient.com/index.php5?query=computer+metaphor.

What's missing is the needed development of explicitly metaphorical programming languages. This may not be the next generation (I think the next generation will be a wave of multi-purpose parallel programming languages), but I would expect it to possibly be the next one.

Bertrand du Castel

Wake up

In 2007, I was talking with Mike Sheppard, leader of the US National Petroleum Council Carbon Capture and Sequestration subgroup regarding the conclusion of his report to the US government. I was very skeptical that humanity would be capable of the effort needed to avoid the catastrophic consequences of not acting in carbon sequestration. Here is his report, and now to follow is the just issued report on the ominous danger of inaction. The only thing I can do is just to give one more bit of advertisement to the voices who want to save us all.

Rogers, A.D. & Laffoley, D.d’A. 2011. International Earth system expert workshop on ocean stresses and impacts. Summary report. IPSO Oxford, 18 pp. Conclusion:

The participants concluded that not only are we already experiencing severe declines in many species to the point of commercial extinction in some cases, and an unparalleled rate of regional extinctions of habitat types (eg mangroves and seagrass meadows), but we now face losing marine species and entire marine ecosystems, such as coral reefs, within a single generation. Unless action is taken now, the consequences of our activities are at a high risk of causing, through the combined effects of climate change, overexploitation, pollution and habitat loss, the next globally significant extinction event in the ocean. It is notable that the occurrence of multiple high intensity stressors has been a prerequisite for all the five global extinction events of the past 600 million years (Barnosky et al., 2009)

As Mike just reminded me again, climate engineering (mitigating global warming by, for example, increasing the earth's reflection) can't solve the problem of ocean acidization. It's born of CO2, not of heat. That's the problem to address, and there is no time left.

Wake up.

Bertrand du Castel

Brain model and identity

I have finally figured out that we should take mirror neurons seriously. Here is a brain model, unconventional for sure, but aggressively considering the symmetry inherent to the brain's wiring.

Elementary brain modules (each line corresponds to one single module):

- See/Be seen
- Hear/Be heard
- Feel/Be felt
- Touch/Be touched
- Manipulate/Be manipulated
- Smell/Be smelled
- Taste/Be tasted
- And so on

Composite brain modules

- Relate brain modules, a few at a time
- Relate brain modules, all at once

Then all it takes to make an individual is identity, indistinguishable from consciousness.

Bertrand du Castel

Morphology as metaphor conveyance

One mystery of linguistics is the actual role of morphology (in layman terms, something akin to the formation of words).

It just dawned upon me that a likely role is conveyance of the original metaphor underlying the concept.

An obvious example is "understanding."

This fits very well with the observation that gestures accompanying speech also reflect original metaphors (say "I understand" and observe your own hand movements).

Bertrand du Castel

Cave of Forgotten Dreams

I saw Werner's Herzog beautiful Cave of Forgotten Dreams movie depicting the Cave of Chauvet.

Apart from Jean Clottes putting a touch of self-interpretation with his "fluidity" and "permeability," the movie is a wonderful matter-of-fact description, with no technical fault that I can see.

It makes me ask again the question: why are there only beautiful paintings of animals in pre-glacial rock art, not of plants, not of landscapes, not of people, with very very few exceptions such as the lower part of a woman in Chauvet? And in sculpture, why almost no men, only women?

I found two points in the movie that are new to me. One, the possible role of shadows of humans handling torches in the painting itself, two, the Australian rock artist saying that the spirit is holding his hand while retouching a painting.

Interpreting plants may have been not so important then (more on that later on in this discussion), interpreting landscapes neither, but interpreting animals certainly was paramount. Women were important then as they are now; if it was men who were making the paintings and sculptures, then women would be represented; and if it was women who were gathering, perhaps that would be a subject of lesser interest to men.

The hard question is why other men would not be represented in pictures: the only explanation would be that they were not nearly as predatory then (due to paucity) for these populations as animals. (Of course, men were represented in hand imprints, but that seems to be a different subject).

So, if all these hypotheses hold (but that's a lot of them), animals strategies would occupy much of the thinking of men, which would naturally lead to representation, for those having that capability; the rest, then, would be history ... in the proper sense of the term.

Bertrand du Castel

In Autism from the gut? I hypothesized that a source for autism should be looked at in the gut.

Jurgensen sent me a reference to Gut Bacteria Linked to Behavior: That Anxiety May Be in Your Gut, Not in Your Head (ScienceDaily, May 17, 2011) that in turns points out to The Intestinal Microbiota Determines Mouse Behavior and Brain BDNF Levels, by Emmanuel Denou, Wendy Jackson, Jun Lu, Patricia Blennerhassett, Kathy McCoy, Elena F. Verdu, Stephen M. Collins, and Premysl Bercik (Gastroenterology, Vol. 140, Issue 5, Supplement 1, Page S-57).

ScienceDaily mentions a possible relation to autism, but the article itself doesn't. What the article points to in terms of the influence of gut bacteria is:

Exploratory behaviour was reduced in GF NIH Swiss mice colonized with BALB/c microbiota, and increased in GF BALB/c mice colonized with NIH Swiss microbiota compared to mice colonized with their respective SPF microbiota

and

At 1 week post-transfer, mice colonized with BALB/c microbiota showed 44% reduction in hippocampal BDNF compared to mice with NIH Swiss microbiota, however at 3 weeks post-transfer, their hippocampal BDNF levels were similar

Regarding the first finding, my observation is that autistic subjects are both less exploratory (routines) and more (specific subjects). That would indeed fit .

Regarding the second finding, which has to do with learning and recall, I don't know. Cf. Hippocampal BDNF mediates the efficacy of exercise on synaptic plasticity and cognition, by Vaynman S, Ying Z, Gomez-Pinilla F. (Eur J Neurosci. 2004 Nov;20(10):2580-90).

Bertrand du Castel

Zimmerman`s theory and phonemes

Another one for Zimmerman's theory.

In Phonemic Diversity Supports a Serial Founder Effect Model of Language Expansion from Africa by Quentin D. Atkinson (Science 15 April 2011: Vol. 332 no. 6027 pp. 346-349), the author shows that language may have a common origin in the south of Africa.

Now, did that require a mutation, in which case Zimmerman's theory would be invalidated, or was that just a discovery, which would be compatible with Zimmerman's theory but would beg the question of whether that also means a single source for language-able humans (is there another kind?); otherwise it would be difficult to explain why it`s been invented only once.

Bertrand du Castel

Birds feeding chicks - again

I previously talked about birds feeding chicks in Programming a bird. One explanation of why birds have the same behavior when feeding themselves and feeding their chicks would be that they may consider chicks as parts of themselves. Then both activities would involve exactly the same neuronal circuits, with a minimalist change on where the mouth is.

When the chicks go away, they are not part of the parents anymore, and the feeding ceases.

Bertrand du Castel

Mirror neurons and reproduction

There is an exact parallel between the act of procreation and interaction of spermatozoids with the ovocyte. Which provides a natural evolutionary path, with a single neural mechanism at its root.

Bertrand du Castel

Where have you been?

Just a nifty application:


visited 58 states (25.7%)
Create your own visited map of The World

Bertrand du Castel

From Austin to Ushuaia

An article in the New York Time, Arrowheads Found in Texas Dial Back Arrival of Humans in America by John Noble Wilford today, finally provides a possible explanation to a puzzle. I was very surprised, going to Ushuaia, to see how close the indigenous culture (that died as recently as the 20th century) was to that of North America, although the distance is enormous. Coastal migrations discussed in the article to explain the presence of pre-Clovis artifacts in the hills of Austin (the migrations had to be coastal because of the ice barriers 15,000 years ago) can also explain, as the article mentions, the speedy population of Peru and Chile. While Argentina is on the other side, it's not that far away in the south and waterways are aplenty.

Bertrand du Castel

Zimmerman's theory revisited

It looks like Zimmerman's theory, that there was not need for a genetic event to explain the great leap forward of humanity in the past 70,000 years, but that just the convergence of two cultural events could do, just received some boost. In Classic Selective Sweeps Were Rare in Recent Human Evolution, by Ryan D. Hernandez, Joanna L. Kelley, Eyal Elyashiv, S. Cord Melton, Adam Auton, Gilean McVean, Guy Sella, and Molly Przeworski (Science 18 February 2011: Vol. 331 no. 6019 pp. 920-924), the authors present that no general genetic event may have occurred in humans in the past 250,000 years.

Bertrand du Castel

More discussion regarding Daryl J. Bem on precognition

Following are the emails I sent to Daryl J. Bem, and his response to the first of these, regarding his article Feeling the Future: Experimental Evidence for Anomalous Retroactive Influences on Cognition and Affect (Journal of Personality and Social Psychology, DOI: 10.1037/a0021524). Actually, I realized afterward that my explanation for his "precognition" experiment would also apply if it was a human presenting the screens instead of a computer. It seems likely that in the same way a computer primes its response to user input, as explained in my emails, a human would prime her/his response when presenting a question, in order to react more rapidly to the anticipated answer. If that's indeed the case, we should see other examples of "precognition" if the person asked is capable of perceiving the priming by the person asking the question.

--------------------------------------- emails
From: Bertrand du Castel
Sent: Monday, January 31, 2011 9:50 AM
To: Daryl J. Bem
Subject: RE: Your experiment vs. electronic forensics

I had read your article thoroughly and I have indeed appreciated the double random algorithm. This is because you've been so thorough that I undertook to write you (I am very familiar with randomization which is a key element of computer security).

I need to explain to you that contrarily to what you think, there is something telling happening inside the computer before the participant indicates her/his response. Let me explain to you why:

(1) When the user clicks on the answer, the computer goes into a series of instructions that then triggers the appearance of the screen of choice. It is perfectly true that this happens after the user has clicked, and therefore, this sequence of instructions has nothing to do with the appearance of precognition.

(2) However, what's important here is what happens before the user is presented with the choice. In order for the proper screen to be later shown by instructions (1) after the user clicked, the computer has to prime its program. The way it does it is by associating, before presenting the choice to the user, what is called in programming the "click event" to the proper screen, so that when the user clicks, the right set of instructions is triggered. The very fact that this priming occurs is source of a set pattern of radiations that could be detected by the user.

If you show me the program code, I'll be able to pinpoint you exactly to where in the code the priming is done before presenting the choice to the user. But you can do that by yourself: you just have to look in the program at what's called the "click event handler". What you'll see is that the event handler is initially set to be ready to go to the right screen according to the answer, and only after that is the choice presented to the user. The way it subsequently works is that when the user clicks on the answer, a "click event" occurs, and the preset "click event handler" triggers the right set of instructions. But again, for this to work, the "click event handler" had to be set properly before the choice is presented to the user.

Please let me assure you that I have no pre-conception here; I just want to inform you of a possible source of bias.

Thanks. Bertrand.

---

From: Daryl J. Bem
Sent: Monday, January 31, 2011 9:06 AM
To: Bertrand du Castel
Subject: Re: Your experiment vs. electronic forensics

Thank you for your informative comments. Your concerns would be particularly relevant if my experiments were testing real-time clairvoyance; they are less relevant because the design is set up to test precognition. Accordingly, nothing happens inside the computer until after the participant has already indicated his/her response. Moreover, as I discuss on pp. 11-15 of my article (attached) , I used both an algorithmic random number generator (RNG) and a true hardware-based RNG to rule out a number of possible alternative interpretations, such as real-time clairvoyance.

---

From: Bertrand du Castel

Sent: Monday, January 31, 2011 8:31 AM
To: Daryl J. Bem
Subject: Your experiment vs. electronic forensics

Executive summary: Computer programming such as that used for priming screen display creates radiations that may be detected by a subject.

Dear professor Bem,

In examining possible biases for your screen experiment, you may have forgotten one which is familiar to security programmers such as smart card operating system programmers. As a reminder, smart cards are security token subject to attacks because they contain money, or means to access power, for example, military or other places. One common attack thieves may use on smart cards is sensing the patterns of activation of the electronic circuits via equipment such as thermal sensors or electron beam probes. The most obvious example of defense actually used in smart cards, the first one taught to programmers, is that when writing in a program a "if" statement, the two sides of the statement should be of equal length, otherwise it is easy to detect which side of the statement has been taken by measuring timing of electronic activation. Equal length is obtained by padding with bogus statements the side which would otherwise be smaller.

Typically, a non security-aware programmer doesn't know this, and would program your experiment without regard to patterns of electronic activation. The computer takes a different path when priming for one kind of screen or the other, because it associates the click instructions to one part of the computer circuitry or the other, which affects the internal pattern of the electronics in subtle ways, for example via caching algorithms of the operating system of the computer. It is entirely conceivable that a body primed for perception of erotic pictures via documented extra sensitization of the insula (cf. How do you feel — now? The anterior insula and human awareness (A. D. (Bud) Craig, Nature Reviews Neuroscience 10, 59-70, January 2009)) would be capable of detecting the differential radiations, and therefore make accurate predictions; since the signal is very weak, it is also very understandable that the deviation from the norm would be small, as it is in the experiment.

The way to avoid this bias is to have the computer itself connected remotely enough from the screen, while taking extra precautions in the insulation of the cable joining the computer and the screen. Your write-up doesn't say whether that was the case.

Regards. Bertrand du Castel.

Daryl J. Bem and computer security

In Feeling the Future: Experimental Evidence for Anomalous Retroactive Influences on Cognition and Affect (Journal of Personality and Social Psychology, DOI: 10.1037/a0021524), Daryl J. Bem presents his famous experiment:

This is an experiment that tests for ESP. It takes about 20 minutes and is run completely by computer. First you will answer a couple of brief questions. Then, on each trial of the experiment, pictures of two curtains will appear on the screen side by side. One of them has a picture behind it; the other has a blank wall behind it. Your task is to click on the curtain that you feel has the picture behind it. The curtain will then open, permitting you to see if you selected the correct curtain. There will be 36 trials in all. Several of the pictures contain explicit erotic images (e.g., couples engaged in nonviolent but explicit consensual sexual acts). If you object to seeing such images, you should not participate in this experiment.

In examining possible biases, he forgot one which is obvious to smart card operating system programmers. As a reminder, smart cards are security token subject to attacks because they contain money, or means to access power, for example, military or other places. One common attack thieves may use on smart cards is sensing the patterns of activation of the electronic circuits via equipment such as thermal sensors or electron beam probes. The most obvious example of defense, the first one taught to programmers, is that when writing in a program a "if" statement, the two sides of the statement should be of equal length, otherwise it is easy to know which side has been taken by measuring timing of electronic activation. Equal length is obtained by padding with bogus statements the side which would otherwise be smaller.

Typically, a non security-aware programmer doesn't know this, and would program Bem's experiment without regard to patterns of electronic activation. The computer would take a different path when priming for one kind of screen or the other, because it would associate the click instructions to one part of the computer circuitry or the other, which affects the internal pattern of the electronics in subtle ways, for example via caching algorithms of the operating system of the computer. It is entirely conceivable that a body primed for perception of erotic pictures via documented extra sensitization of the insula (How do you feel — now? The anterior insula and human awareness (A. D. (Bud) Craig, Nature Reviews Neuroscience 10, 59-70, January 2009)) would be capable of detecting the differential radiations, and therefore make accurate predictions; since the signal is very weak, it is also very understandable that the deviation from the norm would be small, as it is in the experiment.

Bertrand du Castel

Fantastic hypothesis

In my 40 years of studying grammar, from my PhD to my articles to my recent patent filings in the field, I have dreamed that somebody comes up with the hypothesis that explains it all.

Friedemann Pulvermüller and Luciano Fadiga have just done that in Active perception: sensorimotor circuits as a cortical basis for language (Nature Reviews Neuroscience 11, 351-360 (May 2010)).

Page 357:
The mechanisms underlying grammar and syntax may be domain general, applying to every kind of action. Indeed, similar to phrases in sentences, basic body acts are joined in action chains to form a meaningful goal-directed action sequence (drinking from a cup requires grasping, lifting, turning and so on). Importantly, even very complex types of syntactic structures have an equivalent in other action domains. The hierarchical structure of embedded or ‘nested’ sentences is paralleled, for example, in music and bodily interaction^111,126, as the following examples illustrate: a centre-embedded sentence (“The man {whom the dog chased} ran away”) has the same nested structure as a standard jazz piece (theme {solos} modified theme) and complex everyday action sequences (open door {switch on light} close door). In each case, a superordinate sequence surrounds a nested action or sequence (in the inner parentheses). Because language, music and body action have similar hierarchical syntactic structures, the principal underlying brain mechanisms might be the same⁵⁸. The domain-general role of Broca’s area, especially Brodmann area ⁴⁴, in the hierarchical structuring of actions (see the main text) could be derived from its evolutionarily earlier premotor functions in action control and action recognition. It will be a fruitful target of future research to clarify how syntactic processes and representations emerge from action–perception circuits and which properties of the human brain are important for building syntactic circuits¹²⁷.

In one sweep, the authors change the very nature of linguistic theory, which suddenly is thrown into considering metaphors not only in the build up of elements of language, but also in the build up of the structure of language itself. Grammaticality in this view reflects motor plausibility, and semantics is just a reflection of our everyday actions.

I think it will take some time for this to be both known and accepted (the earth goes around the sun, after all), but in my book, the authors got the prize.

In nominating for the price, I''ll add Robin Dunbar to Friedemann Pulvermüller and Luciano Fadiga; he anticipated this in his theory that language is an extension of grooming (Grooming, Gossip, and the Evolution of Language, 1996, Faber and Faber, London).

Bertrand du Castel

Neuroscience of language

Friedemann Pulvermüller and Luciano Fadiga have written Active perception: sensorimotor circuits as a cortical basis for language (Nature Reviews Neuroscience 11, 351-360 (May 2010)).

The title is clear. Digging further, I find another article by Friedemann Pulvermüller, Brain-Language Research: Where is the Progress? (BIOLINGUISTICS, Vol 4, No 2 (2010)), with the following quotes.

Page 259:
In one study, the semantic somatotopy could even be documented in abstract idiom processing (‘grasp the idea’, ‘kick the habit’; cf. Boulenger et al. 2009) consistent with an embodied, partly compositional view on abstract sentence meaning construction, to which lexical meaning contributes (Lakoff 1987, Barsalou 1999).

Page 268:

Gating experiments for example indicate that several competing hypotheses about possibly upcoming words are built, maintained and tested in parallel until one of them ‘wins’, a position immanent to models in the tradition of the cohort theory (Marslen-Wilson 1987, Gaskell & Marslen-Wilson 2002).

The first quote validates the physical grounding of metaphors, the second one sounds very much like the operation of a stochastic grammar.

Bertrand du Castel

Causality and Trust

Is there a difference between causality and trust?

If I don't trust that one thing can depend on another, there is no causality.

If one thing can't depend on another, there is no trust.

Bertrand du Castel

Stochastic grammar of stochastic grammar

Zellig Harris observed that the metalanguage is in the language; see for a recapitulation Harris' The Background of transformational and metalanguage analysis, in The legacy of Zellig Harris: language and information into the 21st century, Vol 1, 2002. (By the way, there is a very interesting article by Maurice Gross, Consequences of the metalanguage being included in the language, in the same book; the author finally debunks the use of VP in syntactic analysis, a long overdue pronouncement).

How would a stochastic grammar that contains itself look?

No, wait a minute, that's too easy. How would a stochastic grammar that can learn itself look?

That would resemble more to Harris' observation. I can tell a child, or a foreigner new with the language, "That's not the way this verb is used". The stochastic grammar should be able to say, "That's not the way a stochastic grammar is built". How do we say that in stochastic grammar land?

Bertrand du Castel

Darwin may have gotten it wrong after all

It sure took me some time to understand this one.

If I take a fistful of sand in my hand, and let it go, it will fall on the ground with some grains on top, some grains on bottom, and a lot in the middle. Would Darwin have studied the evolution of sand, would he have suggested that grains on the top are survivors? Anthropomorphically so?

I dedicate this to Mike Sheppard, for telling me that I should read What Darwin Got Wrong by Jerry Fodor and Massimo Piattelli-Palmarini. I have not done it yet though, but the very title prompted in me the thought above.

Bertrand du Castel

Taxonomy of privacy

Derision is the mask of ignorance.

In his cryptogram this month, Bruce Schneier presents a revised taxonomy of social networking data, which I think is more generally the first attempt I know of in partitioning privacy from a data perspective. In short, his taxonomy distinguishes between (his terms):
1. Service data
2. Disclosed data
3. Entrusted data
4. Incidental data
5. Behavioral data
6. Derived data
That would apply to the brain as well as social networks.

Bertrand du Castel

Believing is already doubting

Believing is already doubting.

Ecstasy is the recognition of identity.

And the mammalian brain isn't. Read Our brains are more like birds' than we thought.

Bertrand du Castel

IBM, Watson, guts, and autism

In the New York Times' What is I.B.M.'s Watson, Thompson talks about the advantage of the best Jeopardy human contestants over the new IBM Jeopardy computer contestant:

So long as they have a good feeling in their gut, they’ll pounce on the buzzer, trusting that in those few extra seconds the answer will pop into their heads.

Here are the guts again, a center of feeling; as I said previously, that's a place to look in for autism.

Bertrand du Castel

Autism from the gut?

In Autism test could make the condition 'preventable', (U.K. Telegraph, 3 Jun 2010), Richard Alleyne reports:

Children with autism spectrum disorder (ASD) have a particular makeup of gut microbes and the researchers have found that this can be detected with a simple urine test.

I want to link this observation with the following, reported by Maura Lynch in Emotional Eating: The Brain-Stomach Connection (Elle, 18 May 2010):

Unglamorous as it may sound, the gut is a physical and emotional powerhouse: It’s estimated to contain more than 200 million neurons, more than the spinal cord has, and can do its work (i.e., digestion) independent of the brain.

Knowing that autism is particularized by an inability to read others' emotions, it makes me wonder whether this should be read as pointing to the gut for an origin of autism rather than to the brain.

Bertrand du Castel

Religion and science

It seems that the science of religion irks religion as much as the religion of science irks science.

Bertrand du Castel

Energy

In Stimulus Predictability Reduces Responses in Primary Visual Cortex (The Journal of Neuroscience, February 24, 2010, 30(8):2960-2966), Arjen Alink, Caspar M. Schwiedrzik, Axel Kohler, Wolf Singer and Lars Muckli say that:

We conclude from this finding that the human brain anticipates forthcoming sensory input that allows predictable visual stimuli to be processed with less neural activation at early stages of cortical processing.

That sounds very much like stochastic grammar processing, with the Bayesian priors with very low coefficients being unlikely to trigger much activity.

Bertrand du Castel

Don't know

Neuroscience Fourth Edition (Sinauer Associates, 2008) p. 203:

Despite the substantial advances in understanding the cellular and molecular bases of some forms of plasticity, how selective changes of synaptic strength encode memories or other complex behavioral modifications in the mammalian brain is simply not known.

Synaptic computation underlying probabilistic inference, by Alireza Soltani and Xia-Jing Wang (Nature Neuroscience, Vol. 13, N. 1, Jan 2010, p. 112):

Little is known about the neural computations underlying this cognitive ability of probabilistic reasoning.

The problem is clearly stated.

Bertrand du Castel

Artificial intelligence programming

A weird aspect of artificial intelligence programming is that the closer the program is to human behavior, the harder it is for the programmer to understand the program's behavior.

Bertrand du Castel

tibudopabikugolatudaropi

In Cracking the speech code: How infants learn language [Patricia K. Kuhl, Acoust. Sci. & Tech. 28, 2 (2007)] says:

Studies show that 8-month-old infants can learn word like units on the basis of transitional probabilities. Saffran, Aslin and Newport [26] played two-minute strings of computer synthesized speech (e.g., tibudopabikugolatudaropi) to infants that contained no breaks, pauses, stress differences, or intonation contours. The transitional probabilities were 1.0 among the syllables contained in four pseudo-words that made up the string, tibudo, pabiku, golatu, and daropi, and 0.30 between other adjacent syllables. After exposure, infants were tested for listening preference with two of the original words, and two part words formed by combining syllables that crossed word boundaries (for example, tudaro—the last syllable of golatu and the first two of daropi). The results show that infants learned the original pseudo-words.

I didn't understand what "0.3 between other adjacent syllables" meant, as somewhere probabilities must add to 1. So I went to the original article by J. Saffran, R. Aslin and E. Newport, Statistical learning by 8-month old infants [Science, 274, 1926–1928 (1996)]. Indeed, in that article, we find:

The only cues to word boundaries were the transitional probabilities between syllable pairs, which were higher within words (1.0 in all cases, for example, bida) than between words (0.33 in all cases, for example, kupa).

So that leads me to write the following stochastic grammar:

Word => 0.25 ti 1.0 bu 1.0 do
Word => 0.25 pa 1.0 bi 1.0 ku
Word => 0.25 go 1.0 la 1.0 tu
Word => 0.25 da 1.0 ro 1.0 pi
Start => 1.0 Sequence
Sequence => 0.99 Word 0.33 Sequence (see note)
Sequence => 0.1 End

Note: Actually, this stochastic grammar is not correct, because it doesn't reflect the fact that the experiment explicitly says that the same word cannot appear twice in sequence. That's what explains why the probability is 0.33 and not 0.25. However, to reflect that, a different stochastic grammar needs to be written. See at the end of this blog.

The next question is: how does the child derive that grammar from the observation of input; as the article says, a 2 minutes sequence is enough for the child to learn. How do they do it? While I know of universal grammar selection theory (e.g. Evolution of Universal Grammar, Martin A. Nowak, Natalia L. Komarova, and Partha Niyogi, Science 5 January 2001 291: 114-118), it seems that we are not at that level here, and that we should rather look at songbirds learning, e.g. Recursive syntactic pattern learning by songbirds, Timothy Q. Gentner, Kimberly M. Fenn, Daniel Margoliash & Howard C. Nusbaum [Nature 440, 1204-1207 (27 April 2006)], which would address another aspect of Patricia Kuhl's discussion in her article cited above:

In other species, such as songbirds, communicative learning is also enhanced by social contact. Young zebra finches need visual interaction with a tutor bird to learn song in the laboratory [32], and their innate preference for conspecific song can be overridden by a Bengalese finch foster father who feeds them, even when adult zebra finch males can be heard nearby [33].

Note that this article uses a regular grammar, not a stochastic grammar, and doesn't attempt to specify how the grammar is acquired. What the article does beautifully is demonstrating that an actual grammar is acquired by the birds, i.e, that concepts are built and sequentially and recursively organized.

Well, the problem I want to address is now clearly stated. If stochastic grammars are as general as I believe, I now need to think hard. There are two concepts, Word and Sequence, in the stochastic grammar I proposed above.

1. Are they the right concepts, in other words, is it the right stochastic grammar?

2. Whether they are the right concepts or not, how do such concepts emerge? Or don't they? Is answering question 2 also providing an answer to question 1?

This is somehow related to the binding problem, i.e. understanding how perceptual entities are actually linked to concepts in the brain.

Now, as I said earlier, the stochastic grammar I wrote above is not the right one anyway. It's time to stop this blog and go to a correct (if lengthier) explanation that I'll undertake in my Stochastic Consciousness knol. I'll see you there.

Bertrand du Castel

Computational Neuroscience

It sure took me a long time to discover that the field of Computation Neuroscience is what I have been after all along when looking at modeling brain behavior with computational methods. I now know where to focus. It came to me as a surprise that the founder of the field was a Frenchman, Louis Lapicque. Since everybody quotes his 1907 article Recherches quantitatives sur l'excitation électrique des nerfs traitée comme une polarisation (J. Physiol. Pathol. Gen., 9 :620–635, 1907), one would think that the article in on the web; I have not found it. I would love to read it so if you have it please send it to me at ducastel@slb.com. I can help by saying that J. Physiol. Pathol Gen. may well stand for Journal de physiologie et de pathologie générale. [I have sent a request to Nicolas Brunel, who has translated the article for the ACM, so I may get it this way].

Bertrand du Castel

Stochastic consciousness

In Scientific American's March 2010's issue (page 47), Marcus E. Raichle writes in The Brain's Dark Energy that

... the equivalent of 10 billion bits per second arrives on the retina ... six million bits per second can leave the retina ... 10,000 bits per second make it to the visual cortex ... the amount of information constituting that conscious perception is less than 100 bits per second ...

This is to be correlated with the following, in The role of the thalamus in the flow of information to the cortex, by S Murray Sherman and R W Guillery (Philos Trans R Soc Lond B Biol Sci. 2002 December 29; 357(1428): 1695–1708.doi: 10.1098/rstb.2002.1161.)

The lateral geniculate nucleus is the best understood thalamic relay and serves as a model for all thalamic relays. Only 5-10% of the input to geniculate relay cells derives from the retina, which is the driving input. The rest is modulatory and derives from local inhibitory inputs, descending inputs from layer 6 of the visual cortex, and ascending inputs from the brainstem. These modulatory inputs control many features of retinogeniculate transmission.

That latter citation has been used by Song-Chun Zhu and David Mumford in A Stochastic Grammar of Images (Foundation and Trends in Computer Graphics and Vision, vol 2, no 4, pp 2590362, 1006) [page 2] to motivate a stochastic grammar approach to human vision.

But stochastic grammars were first motivated by linguistic analysis, where they form a compelling improvement to the grammatical descriptions of Noam Chomsky [See for example my own (shameless publicity) Form and Interpretation of Relative Clauses in English, Bertrand du Castel, Linguistic Inquiry Volume 9 Number 2 (Spring, 1978) 275-289].

Stochastic grammars are not limited to language analysis nor vision, they extend also to the most intricate human activities, as shown by two World Intellectual Property Organization (WIPO) Patent Cooperation Treaty (PCT) patent applications: (WO/2010/010453) System and Method for Determining Drilling Activity, by Harry Barrow and Bertrand du Castel, and (WO/2010/010455) System and Method for Automating Exploration of Production of Subterranean Resources, by Bertrand du Castel and Harry Barrow.

This leads us to stochastic consciousness. While stochastic grammars in vision model our neuronal systems in space first (and then in time, although this is not modeled in the publication mentioned above), stochastic grammars in language and human activities exemplified by the patents above model our neuronal systems in time.

Let's take the time example, although the discussion below also applies in space. The essential of a stochastic grammar is a modular description of stochastic phenomena that extends both in width and depth in a recursive manner. That description amounts to a description of prior Bayesian probabilities that are then matched with posterior observations. Those observations allow updating the priors and then go to the next step. In layman's terms, we have a preconceived idea of what the world is like that we constantly update with what we observe, creating anew a preconceived idea of what would happen next.

What's very interesting is how it actually works. Most of the time, we would expect the world to behave our preconceived way. If the car is going on the highway, the next milliseconds it should be going as well. While the landscape is changing, it is also changing in preconceived ways, and we go our merry way as well.

But suddenly, a deer jumps on the highway. While this is part of our priors, i.e. our model of the world, the circuitry involved is in low state of expectation, because the probability of a deer jumping on the highway is low. But the posterior event, i.e. the jump of the deer, matches with the current priors and raises them considerably regarding a deer event. So the priors see a major shift in their probabilities with a new set of prior probabilities that will emphasize, for example, the likelihood that the car hits the deer. Such a major shift in circuitry involves, in the brain, an influx of excitatory neurotransmitter actions to the synaptic connections involved, while inhibitory neurotransmitter actions will lower down the previous priors. That looks very much like consciousness to me.

It is worth nothing that this takes very good care of the questions of top-down and bottom-up attention and their relationship to consciousness. A major adjustment of priors is equivalent to top-down attention, as it prepares the brain for a new set of events, and a major adjustment of posteriors is equivalent to bottom-up attention, as it triggers a major adjustment of priors. In both cases, neurotransmitters shift focus, and that's the feeling of consciousness. I need to look in the literature at whether somebody has correlated specific chemical releases to that feeling. Reading David J. Chalmers seminal The Puzzle of Conscious Experience, in Scientific American , December 1995 pp62-68 (see also the original version by Chalmers) where he posits the hard problem of consciousness, I can't help feeling that perhaps it's less a question of chemical releases than of their glial blood activation.

Let's take for example Chalmer's illustration of Mary, a neuroscientist who is the world's leading expert in color vision but who herself is color-blind (actually, Chalmer says that she has been living in a black and white world). He says that the easy problem of consciousness is the first part and the hard one the later. How to explain that she knows everything about color vision but color experience doesn't evoke a conscious feeling for her? Well, with a stochastic grammar, that's exactly what we would expect. In the former case, Mary has a set of priors that describe her preconceived expectations of a color theory, and that will be met with posteriors which are actual triggers of her theory (like, say, a wavelength measurement). In the latter, the priors are related to natural experience, that for her doesn't involve color. The posteriors don't evoke color either, so there is never a glial blood activation related to a change in the priors due to color. Conversely, for a person without color impairment, a particular set of colors may be different enough from prior expectations to reset priors in a significant way and therefore trigger a perceptible change in glial blood activation. At this point, I'd like to relate brain blood activation to a more immediately understandable phenomenon, that of, for example, blood activation such as blushing, or blood activation such as extra blood in the limbs. In those cases, blood is linked very directly with consciousness; it is natural to ask whether brain glial activation is similarly perceived. Well, I need another plunge into the literature to find out whether changes in the brain blood patterns have been measured while correlated with recognition of consciousness. What's for sure is that the reverse is true, as syncope is a loss of consciousness related to blood insufficiency.

Well, this blog is becoming too big. It's time I convert it into a knol.

Bertrand du Castel

Pliers, carrot, igloo

Interesting but actually misleading (more on that later) article explaining the fMRI studies show that the brain first categorizes nouns along 3 dimensions of food, shelter, and manipulation. When I read it, I immediately wondered if we could use that study to see metaphors plain in vivo.

So I tried to find the original work behind the article. Since the article references Carnegie Mellon's Center for Cognitive Brain Imaging, I went there. I didn't find the reference I was looking for but the work of the center is very interesting as they merge neuroscience with computer science. This is very much in the line of Computer Theology.

Well, looking further, here is the press release the article is referring to. And here is the original publication (PLoS ONE 5(1): e8622. doi:10.1371/journal.pone.0008622 13 January 2010) by Marcel Adam Just, Vladimir L. Chrkassky, Sandesh Aryal, and Tom M. Mitchell, entitled A Neurosemantic Theory of Concrete Noun Representation Based on the Underlying Brain Codes.

It turns out that the authors studied 60 words such as pliers, carrot, igloo, and found out that the brain uses manipulation, eating, and shelter-entry for classification. So what it's all about is actually confirming that the brain can manage its natural classification: pliers are tools, carrot is food, and igloo is shelter. It's amazing how distorted things become when we move away from the original publication. To make sure, here it is again.

Bertrand du Castel

Uneasy religion

Interesting chart of the religiousness of the United States (measured by Gallup), showing that religiousness correlates with social uneasiness (measured by percentage of population impoverished, murder rate, occurrences of theft, divorce rate, and health and contentment ranking).

Interestingly, that may contradict other Gallup data showing a relationship between religiousness and suicide rates of countries. The more religion, the less suicides; that would seem to also be a measure of social uneasiness. What we need here to be sure is the suicide rates of the United States to make an apple to apple comparison.

- But if the data of the chart hold scrutiny, would the explanation be that social uneasiness increases the need for trust systems, and henceforth religion, as per Computer Theology (p. 2)?

Would then that be an explanation for the difference found more generally between countries, as illustrated by Gallup?

Bertrand du Castel

Future forward, or backward?

Tim pointed my attention to an article of the New York Times referencing the article Moving Through Time by Lynden K. Miles, Louise K. Nind and C. Neil Macrae in the January 10 issue of Psychological Science.

The authors say:

Our findings demonstrate that mental time travel has an observable behavioral correlate—the direction of people’s movements through space (i.e., retrospective thought = backward movement, prospective thought = forward movement).

What would be very interesting is to know whether the Aymara people of Bolivia and Peru, whose language considers that the future is behind a person and the past in front, would recline towards the back when thinking about the future and incline towards the front when thinking of the past.

What would make that particularly interesting is that, if positive, it would go towards the Sapir-Whorf Hypothesis of language embodiment.

Bertrand du Castel

Shed the mask

In Auditory Masking with Complex Stimuli (pp. 343-352 of The Cognitive Neurosciences, Fourth Edition, MIT Press, 2009), Virginia M. Richards and Gerald Kidd, Jr. study the effect of perturbations on auditory signal perception. They call those perturbations masking, and they make the distinction between physical and informational masking.

An example of physical masking is noise from the street that makes the speech hardly audible. An example of informational masking is interfering conversations that make it hard to make sense out of the speech. The authors study those distinctions for a better understanding of human auditory processing.

The question I'm wondering about is how prevalent is masking, both physical and informational, in our perceptory system and its accompanying processing capabilities. The shadow of night is physical masking, and the combination of odors is informational masking. Venetians masks are both physical and informational maskings?

A person with a mask has an identity, just as the same person without a mask. Those identities are at the same time same and different. In terms of Computer Theology (p. 304), the explanation is that the two different identities are those of the person with the mask and of the person without the mask. We call those identities differential. But there is another identity, which we call experiential. In this case, there is one experiential identity for the person in front of you, whether carrying or not carrying a mask.

Back to the auditory system. Is physical masking experiential and informational masking differential? Let me know what you think.

Bertrand du Castel

Olfaction or consciousness?

In Olfaction: From Percept to Molecule (pp. 321-342 of The Cognitive Neurosciences, Fourth Edition, MIT Press, 2009), Yaara Yeshurun, Hadas Lapid, Rafi Haddad, Shani Gelstien, Anat Arzi, Lee Sela, Aharon Weisbrod, Rehan Khan, and Noam Sobel present a correlation of two first principal components. One, a first principal component of olfactory perception, which they name pleasantness, and two, a first principal component of physicochemistry, which they name molecular weight.

Their interest is in identifying vectors of research similar to those of other sensations, like pitch and frequency for auditory signals, in order to support a parallel investigation for olfactory signals, even though those latter signals are processed by the brain differently from other signals; olfactory signals go directly to the cortex before going to the thalamus, whereas other signals go to the thalamus first.

My interest here is different, and I definititely plead weirdly ignorant here. There is something in my own feeling of consciousness that is reminiscent of my feeling of odor. This is just a hunch. When I say "I feel good this morning", it that pleasantness related to consciousness; and if it is, is that pleasantness scale similar to that of odor in terms of its relationship to a physical vector which would be that of consciousness?

Bertrand du Castel

Rituals and attention

In Selective Attention Through Selective Neuronal Synchronization, Thilo Womelsdorf and Pascal Fries say (p. 298 of The Cognitive Neurosciences, Fourth Edition, MIT Press, 2009):

Therefore these findings suggest that inter-areal communication during attentional top-down control is conveyed particularly through rhythmic synchronization in a high beta band, either in addition to or separate from the frequency of rhythmic interactions underlying bottom-up feedforward signaling.

Top-down attention is directed attention, as in "look at that cat," while bottom-up attention is triggered attention, as in "something is amiss in this picture." Attention and rhythm; that forms a nice underpinning for social behaviors.

Bertrand du Castel

Zimbabwe Inanke

In the Wall Street Journal today Jan 23, 2010, a nice description of Zimbabwe's Inanke cave. The landscape, animals, shamanistic references are so similar in design and significance to the Pecos river cave paintings that the continuity between the two civilizations has to be recognized; but from South Africa to the south of the United States, is that longest of trek testimony to a single religious culture?

I note again that Zimbabwe's paintings are said to be post-glacial (5,000 to 10,000 years ago) by the article. Quite different from the pre-glacial (up to 35,000 years ago) cave paintings of the south of France and north of Spain, which I still don't believe are comparable in terms of religious expression. Where are the pre-glacial African cave paintings? Are there any?

Bertrand du Castel

Mirror neurons and consciousness

Interesting talk just posted on Ted of VS Ramachandran's The Neurons that shaped civilization.

What he considers is mirror neurons in my prefrontal cortex allowing me to sense when somebody touches something. They mirror the neurons that are fired when I (not somebody else) touch something.

Ramachandran says that when I actually touch something, my hand (say I am touching with the hand) sends signals of touching to my prefrontal cortex and that's how I differentiate direct neurons from mirror neurons.

Now Ramachandran says that if my arm is anesthesied, my direct neurons don't get the usual signals but are still looking for them so they get them from me seeing the other hand. Therefore my brain concludes that I am touching with the hand of somebody else. Ramachandran says that I have just assumed somebody else's consciousness.

But Ramachandran doesn't say: what happens with the mirror neurons then?

I think that the answer is that there are not both direct and mirror neurons, but rather one set of neurons that reacts to touching. However, that set can be linked to a representation of my hand (that's for direct touching) or/and to the representation of somebody else's hand (that's for mirror touching). If I see both my hand and that of somebody else each touching something, my neurons are activated, with a link to my hand (strong because reinforced by my hand signals) and another link to the other hand (weak). If my arm is anesthesied and the strong signal disappears, the only signal around is that from the other hand, even if weak, and it takes the place that would normally be taken by my own hand's signal.

That assumes that seeing my own hand, anesthesied or not, primes my neurons for searching signals from it. That sounds right, but now somebody needs to tell me if my explanation fits the neurological evidence. From what I have read of the original litterature on mirror neurons (see the reference in Computer Theology, p. 420: Mirrors in the Mind, by Giacomo Rizzolati, Leonardo Fogassi and Vittorio Gallese, Scientific American, November 2006), it is. In their original discovery, the authors had a monkey's neuron monitored by sound when the monkey touched a banana. One day the operator, instead of the monkey, touched a banana and the monkey's neuron resonated. I was actually the same neuron, which is what I say in my explanation.

Now a final point. My touching neurons get input from both the image of my hand and that of somebody else, but how do they themselves communicate where the input was from? The only explanation I see to that is that the recipients for the strong signal out (which means that my touching neurons think, rightly or not, that I am actually touching something) and those for the weak signal out (which means that I am watching somebody else touching) are, in the former case, far from my touching neurons, and in the later, close to them. Attenuation would make the difference. I have no idea whether that's a realistic hypothesis. If not, what else?

Bertrand du Castel

Saliency and attention

In The Quest for Consciousness (Roberts & Company Publishers, 2004, p. 161), Christoph Koch gives the following title to a section of the text:

Salient Objects Attract Attention

Maybe he should have said instead:

Attention Makes Objects Salient

That would be consistent with the computer theory of exceptions. Exceptions are raised when the brain cannot otherwise handle the situation; a situation that can't be handled by the regular process is salient.

Or an otherwise unremarkable object, i.e., an object fitting the current process, may be subject to attention. It then becomes salient, and therefore triggers an exception.

As with computers, the exception is an opportunity for the exception-handling procedures to examine the process that raised the exception. A process that can so be examined is called reflexive in computer science. So the brain uses the exception-handling procedure to investigate itself. That's what consciousness is made of.

You may ask at this point: how does a computer bring an object to attention? Technically, it sets for that object an explicit Throw procedure. That procedure goes out of the current process to find the next Catch procedure, that manages the exception thrown. In the absence of attention, the Throw procedure is implicit and simply triggered when the on-going process finds it can't handle the case at hand (it has found a salient object).

How does the brain know where to set up an explicit Throw procedure. Well, we've now moved the discussion of consciousness to that of attention. The way a computer does it is while it is in exception mode, i.e. when it is capable to look at a reflexive process, it can intervene on that process as dictated by the exception procedure.

Add to this recursivity and voila. We've built a conscious brain.

Bertrand du Castel

Clerics and war

Reflecting on the past decade, it dawned upon me that clerics, from shamans to priests to imams, have historically be fond of blessing acts of war (and sometimes war itself) and celebrating their consequences, while they don't appear to participate in the same way in acts of police.

In the computer theology (Computer Theology, Midori Press, 2008, p. 373) recursive hierarchy of trust and policy infrastructures, that would mean that wars involve trust infrastructures, while police involves policy infrastructures.

So war would be about trust, or the lack thereof. One analysis would be that as long as two parties' trust infrastructures are compatible, everything goes fine, but when the context (or the trust infrastructures themselves) changes, incompatibilities demand resolution.

In Computer Theology (p. 141), we emphasize that, under breach of the trust that consolidates it, the social ecosystem is quick to revert to the physical ecosystem. That would reinforce that analysis. War would be an act of trust, albeit externally oriented, police an act of policy, internally oriented. Trust being a religious matter, clerics come into play on the war theater.

That begs the question of today's context or trust infrastructure changes. Might be one of those context changes the radical shift in information sharing brought in the past two decades by the Internet, that shaped a new perception of economic realities?

Bertrand du Castel

Symbolism vs. number of neurons in the brain

Christoph Koch writes in Biophysics of Computation (Oxford University Press, 1999, p. 87):

Given an approximate density of 100,000 cells by mm³ in the primate, a synaptic density of 6 x 10⁸ per mm³, a total surface area of about 100,000 mm² for one hemisphere, and an average thickness of about 2 mm, the average human cortex contains on the order of 20 billion neurons and 240 trillion synapses (2.4 x 10¹⁴) [...]

This reasoning is not yet complete as while the number of synapses falls from the argument, the number of neurons doesn't. However, in The Quest for Consciousness (Roberts & Company Publishers, 2004, p. 71 note 5), the same Christoph Koch says:

Given a packing density of 50,000 cells per mm³, a total surface area of 2 x 100,000 mm², and a thickness of about 2 mm, the average human cortex contains on the order of 20 billion neurons and 200 trillion synapses (2 x 10¹⁴).

Also in The Quest for Consciousness (Roberts & Company Publishers, 2004, p. 26), Koch says:

All the visual information that the brain can access is implicitly encoded by the membrane potentials of the more than 200 million photoreceptors in the two eyes.

Since photoreceptors are neurons, those facts combined provide a formal proof (we insist on the terms "formal proof" here) of the elaboration of symbols by the brain. The proof is that if each image is using 200 million neurons, only 100 images (100 x 200 million = 20 billion) can be stored by the entire brain, obviously insufficient. Information needs to be reduced (which we actually know it is, but that's besides the argument here).

This sets the scene for a formal study of symbol construction, starting with a very small number of photoreceptors and going up from there.

Bertrand du Castel

Cloud Computing Definition

Computer Theology's definition: Cloud Computing is the global optimization of computer resources.

Bertrand du Castel

Christof Koch and Roger Penrose

In Computer Theology (2008, Midori Press) we wrote (page 40):

There is indeed a tendency to observe a reducing perspective on computers, and particularly on computer systems. For example, Roger Penrose seems to suggest as much when he bases his analysis on an equation of computing with Turing machines in The Emperor’s New Mind, even considering the extreme example of a Turing machine with infinite memory, infinite state variables and unlimited computing time; there is more to computing than theoretical equivalence.

In The Quest for Consciousness (2004, Roberts & Company), Christof Kock writes (page 8, note 12):

Penrose's books (Penrose, 1989, 1994) are among the most lucid and best-written accounts of Turing Machines, Gödel theorems, computing, and modern physics I have read. However, given that both monographs nominally deal with the human mind and brain, they are equally remarkable for the almost complete absence of any serious discussion of psychology and neuroscience.

This is most remarkable in that, as we observed in Computer Theology, we should extend Koch's mention of a lack of serious discussion ... to computer science. Turing machines, as important as they are, are a minuscule part of any cursus in computer science, or, for that matter, of theoretical computer science. But that's all that Roger Penrose and Christof Koch speak about when they think of computers. In the case of Roger Penrose, it's just sad. In the case of Christof Kock, it's also surprising, as his most wonderful book of neuroscience is entitled Biophysics of Computation (1999, Oxford University Press) and is so exquisite in its description of neural computation. But again, all he seems to know about computer is Turing machines (he also make some references to computer hardware, but they are even more of a case of ignoring mainstream computer science). See for example page 469:

The brain has frequently been compared to a universal Turing machine (for a very lucid account of this, see Hofstadter, 1979). A Turing machine [...]

and then Koch explains differences between Turing machines and the brain's operations. What's remarkable is that it's obvious to a computer scientist like me that Turing machines are the wrong metaphor. I'd rather think of the theory of exceptions when considering consciousness, or of object-oriented concepts when examining miror neurons, or more generally, in mapping the brain, of ontologies, of Bayesian networks, of stochastic grammars, and many other computer science constructs. That neither Roger Penrose nor Kristof Koch, great scientists in their fields and purporting to address computer questions, would stop at computer science 101 reflects the dire state of interdisciplinary studies. If this note encourages them to dig into a science that would indeed benefit so much from these most important scientists' insights if they were more informed, I'll be happy.

Bertrand du Castel