Re: [Fis] Chemical information: a field of fuzzy contours ?
Dear Colleagues, taking the risk of repeating an idea that has been advanced here in FIS a number of times, I'd like to offer answers to Michhel's questions: Michel: " Now, I ask you the following: please can you provide an extremely simple example (the most simple you could imagine) of situation in which you can say: << in this situation, information is ... >>." Answer: Let me present a numerical table based on a+b=c, consisting of 136 additions (between 1+1=2 and 16+16=32) which is evaluated on 9 aspects of the additions (namely: a,b,a+b, 2b-a, b-a, 3b-2a, 2a-b, 17-(a+b), 3a-2b) and ordered on two of these aspects (therefore existing in 72 distinguishable collections of sequences of distances). This table gives rise to two concepts of Euclid spaces (consisting of 3 rectangular axes each) and two planes (with 2 rectangular axes each). The term "information" can then be used - as Michel asked for - in a deictic fashion by pointing to a collection of spatial points in both Euclid spaces and saying "<< in this situation, information is: >> whether we consider relevant the connection of these spatial point-collections with << this >> or << that >> collection of different spatial points which are connected to the presently pointed-at collection odf spatial points by either << this >> or << that >> re-orderings of the collection." while one points at two different ways of re-ordering the collection among the 72 ways of re-ordering the collection, and calling one of them << this >> and the other << that >>. It helps if you, dear Colleague, construct the above-mentioned table. Then it is irrefutalby clear that the meaning of the term "information" is indeed contained in the underlying rules that construct a+b=c as a logical procedure. The only innovation is that one does not ignore the differeneces between a1+b1=c vs. a2+b2=c (a1 # a2) as one was instructed at Elemenary School to do. Michel: "3. The comparison Pedro did with symmetry is of interest: can anyone define symmetry ?" Answer: Using the table constructed above, one may point to the two Euclid spaces and say: "The two spaces are << symmetrical >> and the term << symmetrical >> is defined by the following: a. the two spaces are interconnected by a point/plane which we call "symmetry centre/axe" and each collection of (a1,b1; a2,b2; a3,b3) which finds a spatial representation in both of the Euclid spaces is in a <> relation". Here, again, it helps if one constructs the above-mentioned table in visualising that a symmetry exists. I hope that these suggestions are both clear and understandable. It is, however, necessary to construct the table to be able to use the definitions. (Like one cannot explain the definition of sin(x) without having understood the construction of a trigonometric table). Karl 2011/9/23 Michel Petitjean > Dear FISers, > > Pedro raises several points. > Among them: > > 1. "Chemoinformatics" or "Cheminformatics" ? > Both terms are encountered. I would say that unless some authority > takes a decision, both terms will continue to be used. > > 2. Despite I gave an example of what could be cheminformation in a > concrete case, I did not tell what was exactly cheminformation in this > concrete case. I just asked the question of what it could be. > Now, I ask you the following: please can you provide an extremely > simple example (the most simple you could imagine) of situation in > which you can say: << in this situation, information is ... >>. > Chemical information is welcome, but an example from physics would be > great, too. However, please, no biology example, that will be dicussed > at the occasion of a future session. > These examples are expected to help us to define information in more > general situations. > > 3. The comparison Pedro did with symmetry is of interest: can anyone > define symmetry ? > During a long time, symmetry had in common with information that many > people attempted to define it in its own field, giving raise to many > particular definitions, but not to a common and widely accepted one. > Some years ago, although I needed to mention a definition of symmetry > in one of my papers, I was surprised that I could not find an unifying > one (symmetry is known since millenaries!!). Even in the book of Weyl > I did not find the expected one. > So, I decided to build my own one (Symmetry: Culture and Science, > 2007, 18[2-3], 99-119; free reprint at > http://petitjeanmichel.free.fr/itoweb.paper.SCS.2007). See also: > http://petitjeanmichel.free.fr/itoweb.petitjean.html > In fact, the group structure which is generally a priori imposed, is a > consequence of several properties that the definition should satisfy > to be in agreement with some obvious intuitive requirements (and so, > five different groups appear naturally, none of them being imposed a > priori). Of course, the proposed unifying definition applies to a > broad spectrum of situations, not only the geometric one: matrices, > functions, distributio
Re: [Fis] Chemical information: a field of fuzzy contours ?
Dear FISers, I thank very much Robin, Xueshan, Stan, and Karl for their examples of information, that I summarized below: *** Robin: Of course, there is no "law" or formula that relates a bit of information to, say, quarks, spin, or whatever. These are different ways of looking at the same thing. Spin is a bit of information (I think it's just one bit, but I might be wrong, as I said, I'm no physicist.) Physical information is a re-conceptualisation of material form that allows it to be quantified. So, for example, physicists can (and do) say that information is generally conserved within black holes. (See the Black Hole Information Paradox, and the bet between physicists concerning it, http://www.theory.caltech.edu/~preskill/jp_24jul04.html) Now, there is obviously more to semantic information than material form, but it is my strongly-held belief that it should be possible to relate all other concepts of information back to physical information, and, in fact, I have proposed a way of doing that for semantic information, which I presented at the DTMD2011 workshop (I've also mentioned it in previous posts on this list), but I'll say no more about it here, because I think that's going too far off the current topic. Michel, maybe that was a bad example, misleading because of its binary nature. My understanding is that physical information is material form, re-conceptualised, and so the spin state, like every other physical attribute, not just the binary ones, IS information (non-semantic information), as and when it suits us to view it that way, i.e. to focus on form rather than substance. Historically, the concept of non-semantic information, or "pure pattern", arose in the context of information theory, but to focus on form is a basic human capacity, and given the concept of non-semantic information, however that arises, it is a small step to apply it to material form, which thus becomes pure pattern whose transformations are governed by the laws of physics. So material form is like data and the totality of physical laws is the program that operates upon it. The operations are, in principle and in general, reversible, and so physical information is conserved, like matter and energy. (I believe there is a strong consensus within physics that physical information is conserved in quantum mechanics.) In a certain sense the laws of physics "stand in" for substance, which is what constrains material form in our ordinary thinking. When we think in terms of pure patterns constrained by physics, every physical entity embodies its own description, and (which is to say almost the same thing) encodes the outcomes of all of its potential interactions. This is a very powerful way of thinking. Gavin: I agree with you that there is no such free-standing, "thing-in-itself" as information, but that doesn't invalidate the concept, far from it. Information is, in my view, basically form, and form doesn't exist without substance, but we work with form, ignoring substance, all the time, and achieve great things by so doing. *** Xueshan: 1. Chmoinformatics: A study about how to manage and compute chemical information, such as management of chemical abstracts, retrieval of chemical information through internet, molecules represented by graphs, data mining etc. there are many books like this in the bookstore. Of course, this may not be a subject that could arouse real interests among true information researchers, because there are thousands of applications of information technology in different areas, it is difficult for us to call all these applications of information technology as informatics or information science. 2. Chmoinformatics: A study about how chemical information function between two molecules or two supermolecules, according to the terms in biology and chemistry: between substrate and receptor, or in coordination chemistry: between donor and acceptor, or host and guest, we can only consider this thought as a conjecture which proposed by Jean-Marie Lehn of University Louis Pasteur——the noble prize winner of 1987. As a matter of fact, we all know that in the process of molecule reaction and recognition, an intelligent is in esse. This has been proved by Fischer’s lock-and-key model early in 1894. 3. Semiochemistry: A study about chemical information materials that mediate interactions between members of different species. This study consider pheromone, quinonyl compounds etc. as messengers. It is an interdiscipline of chemistry and biology. We especially want to know what advance about the second study about chemical information in chemists has made recent years. Because Lehn said in many places: “Supramolecular chemistry (chmoinformatics) has paved the way toward apprehending chemistry as an information science”. *** Stan: Would it not be the case that chemical information would relate to catalysts? That is, chemical scale configurations which have the property of forming enabling constraints for so
Re: [Fis] Chemical information: a field of fuzzy contours ?
Dear Michel, It seems to me that Shannon's formulas are mathematical and yet content-free. By the specification of a system of reference they can be provided with dimensionality and then also meaning. For example, in the case of the momenta and positions of particles H is multiplied with k(B) [S = k(B) *H] and thermodynamic entropy [Watt/Kelvin] can thus be defined. Momenta and energy are in this case exchanged upon collisions. S measures the dissipation in the non-ideal case. This is a specific (physical) theory of communication. If molecules are exchanged life can be generated (Maturana); if atoms are exchanged, chemical evolution can be expected. It seems to me that the general scheme is the specification of (i) WHAT is being exchanged -- this specifies the domain theory -- and (ii) HOW one expects to be exchanged (e.g., dissipative or conservative, recursive or incursive, etc.), and (iii) WHY. The why question bring us to evolution theory; for example, the selection environments for the variation (uncertainty) can then be specified as hypotheses. Best wishes, Loet -Original Message- From: fis-boun...@listas.unizar.es [mailto:fis-boun...@listas.unizar.es] On Behalf Of Michel Petitjean Sent: Monday, September 26, 2011 1:39 PM To: fis@listas.unizar.es Subject: Re: [Fis] Chemical information: a field of fuzzy contours ? Dear FISers, I thank very much Robin, Xueshan, Stan, and Karl for their examples of information, that I summarized below: *** Robin: Of course, there is no "law" or formula that relates a bit of information to, say, quarks, spin, or whatever. These are different ways of looking at the same thing. Spin is a bit of information (I think it's just one bit, but I might be wrong, as I said, I'm no physicist.) Physical information is a re-conceptualisation of material form that allows it to be quantified. So, for example, physicists can (and do) say that information is generally conserved within black holes. (See the Black Hole Information Paradox, and the bet between physicists concerning it, http://www.theory.caltech.edu/~preskill/jp_24jul04.html) Now, there is obviously more to semantic information than material form, but it is my strongly-held belief that it should be possible to relate all other concepts of information back to physical information, and, in fact, I have proposed a way of doing that for semantic information, which I presented at the DTMD2011 workshop (I've also mentioned it in previous posts on this list), but I'll say no more about it here, because I think that's going too far off the current topic. Michel, maybe that was a bad example, misleading because of its binary nature. My understanding is that physical information is material form, re-conceptualised, and so the spin state, like every other physical attribute, not just the binary ones, IS information (non-semantic information), as and when it suits us to view it that way, i.e. to focus on form rather than substance. Historically, the concept of non-semantic information, or "pure pattern", arose in the context of information theory, but to focus on form is a basic human capacity, and given the concept of non-semantic information, however that arises, it is a small step to apply it to material form, which thus becomes pure pattern whose transformations are governed by the laws of physics. So material form is like data and the totality of physical laws is the program that operates upon it. The operations are, in principle and in general, reversible, and so physical information is conserved, like matter and energy. (I believe there is a strong consensus within physics that physical information is conserved in quantum mechanics.) In a certain sense the laws of physics "stand in" for substance, which is what constrains material form in our ordinary thinking. When we think in terms of pure patterns constrained by physics, every physical entity embodies its own description, and (which is to say almost the same thing) encodes the outcomes of all of its potential interactions. This is a very powerful way of thinking. Gavin: I agree with you that there is no such free-standing, "thing-in-itself" as information, but that doesn't invalidate the concept, far from it. Information is, in my view, basically form, and form doesn't exist without substance, but we work with form, ignoring substance, all the time, and achieve great things by so doing. *** Xueshan: 1. Chmoinformatics: A study about how to manage and compute chemical information, such as management of chemical abstracts, retrieval of chemical information through internet, molecules represented by graphs, data mining etc. there are many books like this in the bookstore. Of course, this may not be a subject that could arouse real interests among true information researchers, because there are thousands of applications of information technology in different areas, it is difficult for us to call all these applications of information technology as in
Re: [Fis] Chemical information: a field of fuzzy contours ?
There's very much a law that tells us how much energy it takes to transmit one bit of information. It has been used in radio astronomy for years. Regards Gavin Dear Michel, It seems to me that Shannon's formulas are mathematical and yet content-free. By the specification of a system of reference they can be provided with dimensionality and then also meaning. For example, in the case of the momenta and positions of particles H is multiplied with k(B) [S = k(B) *H] and thermodynamic entropy [Watt/Kelvin] can thus be defined. Momenta and energy are in this case exchanged upon collisions. S measures the dissipation in the non-ideal case. This is a specific (physical) theory of communication. If molecules are exchanged life can be generated (Maturana); if atoms are exchanged, chemical evolution can be expected. It seems to me that the general scheme is the specification of (i) WHAT is being exchanged -- this specifies the domain theory -- and (ii) HOW one expects to be exchanged (e.g., dissipative or conservative, recursive or incursive, etc.), and (iii) WHY. The why question bring us to evolution theory; for example, the selection environments for the variation (uncertainty) can then be specified as hypotheses. Best wishes, Loet -Original Message- From: fis-boun...@listas.unizar.es [mailto:fis-boun...@listas.unizar.es] On Behalf Of Michel Petitjean Sent: Monday, September 26, 2011 1:39 PM To: fis@listas.unizar.es Subject: Re: [Fis] Chemical information: a field of fuzzy contours ? Dear FISers, I thank very much Robin, Xueshan, Stan, and Karl for their examples of information, that I summarized below: *** Robin: Of course, there is no "law" or formula that relates a bit of information to, say, quarks, spin, or whatever. These are different ways of looking at the same thing. Spin is a bit of information (I think it's just one bit, but I might be wrong, as I said, I'm no physicist.) Physical information is a re-conceptualisation of material form that allows it to be quantified. So, for example, physicists can (and do) say that information is generally conserved within black holes. (See the Black Hole Information Paradox, and the bet between physicists concerning it, http://www.theory.caltech.edu/~preskill/jp_24jul04.html) Now, there is obviously more to semantic information than material form, but it is my strongly-held belief that it should be possible to relate all other concepts of information back to physical information, and, in fact, I have proposed a way of doing that for semantic information, which I presented at the DTMD2011 workshop (I've also mentioned it in previous posts on this list), but I'll say no more about it here, because I think that's going too far off the current topic. Michel, maybe that was a bad example, misleading because of its binary nature. My understanding is that physical information is material form, re-conceptualised, and so the spin state, like every other physical attribute, not just the binary ones, IS information (non-semantic information), as and when it suits us to view it that way, i.e. to focus on form rather than substance. Historically, the concept of non-semantic information, or "pure pattern", arose in the context of information theory, but to focus on form is a basic human capacity, and given the concept of non-semantic information, however that arises, it is a small step to apply it to material form, which thus becomes pure pattern whose transformations are governed by the laws of physics. So material form is like data and the totality of physical laws is the program that operates upon it. The operations are, in principle and in general, reversible, and so physical information is conserved, like matter and energy. (I believe there is a strong consensus within physics that physical information is conserved in quantum mechanics.) In a certain sense the laws of physics "stand in" for substance, which is what constrains material form in our ordinary thinking. When we think in terms of pure patterns constrained by physics, every physical entity embodies its own description, and (which is to say almost the same thing) encodes the outcomes of all of its potential interactions. This is a very powerful way of thinking. Gavin: I agree with you that there is no such free-standing, "thing-in-itself" as information, but that doesn't invalidate the concept, far from it. Information is, in my view, basically form, and form doesn't exist without substance, but we work with form, ignoring substance, all the time, and achieve great things by so doing. *** Xueshan: 1. Chmoinformatics: A study about how to manage and compute chemical information, such as management of chemical abstracts, retrieval of chemical information through internet, molecules represented by graphs, data mining etc. there are many books like this in the
