“In the beginning was the Word…”
John 1:1
New International Version
Series Editor
Jean-Charles Pomerol
First published 2019 in Great Britain and the United States by ISTE Ltd and John Wiley & Sons, Inc.
Apart from any fair dealing for the purposes of research or private study, or criticism or review, as permitted under the Copyright, Designs and Patents Act 1988, this publication may only be reproduced, stored or transmitted, in any form or by any means, with the prior permission in writing of the publishers, or in the case of reprographic reproduction in accordance with the terms and licenses issued by the CLA. Enquiries concerning reproduction outside these terms should be sent to the publishers at the undermentioned address:
ISTE Ltd
27-37 St George’s Road
London SW19 4EU
UK
www.iste.co.uk
John Wiley & Sons, Inc.
111 River Street
Hoboken, NJ 07030
USA
www.wiley.com
© ISTE Ltd 2019
The rights of Cédric Gaucherel, Pierre-Henri Gouyon and Jean-Louis Dessalles to be identified as the authors of this work have been asserted by them in accordance with the Copyright, Designs and Patents Act 1988.
Library of Congress Control Number: 2018960339
British Library Cataloguing-in-Publication Data
A CIP record for this book is available from the British Library
ISBN 978-1-78630-424-7
“The reader of this sentence will cease to
exist when he has finished reading.”
Metamagical Themas, Douglas Hofstadter
This book grew out of a meeting that brought together three scientists, who had different backgrounds, and who worked in different disciplines, but who all had this in common: their activities had to do with the living entity. Confronting each of them in their own way was the question: what is information?
Whether in ecosystems, human language, genetics or evolution of living organisms, one comes face to face with the idea that information forms the substance of the system(s) being studied, then reproduced, transmitted, coded and decoded, or manipulated in other ways. Such information may not be equally important to everyone, it may not be carried by the same channel, or perhaps even be of the same nature. Yet, one cannot afford to overlook the importance of the concept while continuing to speak of it without a proper understanding of what the much-used term means. The information that we exchange and talk about every day is coded in our language. Now, are other languages coded differently, say nature’s languages? Can we define them, analyze them and quantify them?
Quite possibly, part of what we are going to say here may appear obvious to some readers, but unacceptable to others. Considering the gap between these two reactions, however, it is reassuring that we come across both kinds of comment, perhaps in a sign that there is something non-consensual behind seemingly mundane ideas. Now, we are not quite done with this theme yet, and, in fact, this is going to open up further vistas of thinking. We hope that the publication of this essay will make possible an elaboration of the issue, and lead to fruitful discussions with those who find these questions interesting.
One difficulty that appears immediately pertinent has to do with the very nature of information. Is it material or is it not? The answer: while the channels used in transmitting it are material, information itself is not material.
Well, this is only the first of our difficulties. Our science does not readily accept non-material entities. Let us not forget that the absence of materiality was a major argument in the rejection of Newton’s theory: the force of gravity appeared quite esoteric, since the force is acting at a distance, without material support. In contrast, the solar turbulences mentioned by Descartes appeared to be much more real or tangible. Yet, while following the same Aristotelian line of thinking, up to the end of the 18th Century, it seemed reasonable to consider fire an element, to the point of giving it a name: phlogiston. A process received the status of matter through a distortion of reality. It was left to the times of Lavoisier to demonstrate the difference between a material entity and a chemical reaction, and thereby to bring to an end a misconception that had persisted for over 2,000 years.
Let us rid ourselves of this obsessive habit of seeing everything in terms of matter. Information, even though carried by material agents, is nevertheless different from them. Now, how can a non-material entity act on matter? While trying to resolve this apparent paradox, we need to remember that, while information is certainly non-material, it is active across channels of communication, themselves material.
Does not Léo Ferré, in his Il n’y a plus rien (“There’s Nothing Any More”), praise the non-material virtues of information?
“Peddle your ideas like drugs… You risk nothing at the borders. Nothing in your hands, nothing in your pockets, it’s all there in your head!
‘Anything to declare?’
‘Nothing!’
‘What is your name?’
‘Karl Marx.’
‘All right. Move on!’”
Indeed, Karl Marx probably produced nothing other than information. He bore no arms, raised no army. For all that, for this generation of ours, having lived in a world split into two blocs, through nuclear horror and the wars that raged from Vietnam to Nicaragua and Cuba and Afghanistan, and that reshaped (and continue to reshape) the geopolitical environment, how materially influential this information of Marx’s has been.
While we (the three scientists) were still getting to know one another, one of us expressed to Jean-Marc Levy-Leblond his desire to work on the concept of information. “Is that a concept?” he had asked. Indeed, information is not yet a concept. Could we build one out of it?
Cédric GAUCHEREL
Pierre-Henri GOUYON
Jean-Louis DESSALES
October 2018
Dear reader, do you know that an extraordinary happening is taking place at this very moment? Your eyes are taking in a linear message in a to-and-fro movement, line by line. The characters that make up the message are imprinting themselves on your retina, because of the sophisticated optical system that you possess. The retinal cells convert the stimuli into an impulse that then enters the brain. That organ then interprets the signal, and …, well, the information that the text had embodied has now entered your mind! Now, dear reader, would you please turn the page?
Thank you. Now, did you notice that something incredible just happened? A piece of information, non-material, was carried by a means, which itself was material (we will say more about this a little later), as described above, and produced a result which was physical. You moved your finger, and turned the page! The physical was put through a change by the action of a piece of non-physical information. That is the sort of thing we are used to. Indeed, our body is a kind of machine that converts information into action. Certain types of information, such as the recipes you find in cookery books, are clearly meant for use with material things.
This work will attempt to consider whatever could be treated as information from the point of view of science. Our times appear to be very much under the influence of that entity. One might ask if this information-age society of ours is seriously looking into what it is that it is talking so much about. The sciences speak about it. Physics and mathematics have developed a now-classical theory of information. As we will see shortly, these disciplines have only captured a fragment of the idea that we have developed in this presentation. Biology has been advancing increasingly with the study of genetics and epigenetics. Ecology takes account of information relating to different levels of organizations, individuals and communities of species. Information underpins new and emerging fields of study such as communication science and technology. Linguistics and the different branches of the humanities also speak of information, which lies at the heart of the network of relationships. Now, what is information about?
It seems advisable at this point to try to disambiguate the term “information”, this notional entity that reproduces itself, transforms itself, elaborates itself and expresses itself … There are to be found in history and the philosophy of science concepts that are both central and indefinable, such as time, which can be measured but not defined. On the other hand, we have an instinctive perception of what time is. Saint Augustine wrote around the year 400: “What is time? If I am not asked the time, I know it; but, if I wish to explain it, I don’t know it! And, yet – and I say this with all confidence – I know that if nothing happened, there would have been no past; if nothing is going to happen, there would be no future; and if nothing existed, there would be no present”. Right or wrong, we have the feeling that we know what time is about, and we are compelled to adapt our definition to particular needs.
It is much the same for the notion of information. Rather than look for a definition, should we not content ourselves with stating its characteristics, and search for a measure that will serve to identify it? The communications specialist studies the movement of information. The linguist studies how information is coded in languages, and how such codes evolve. The biologist investigates how genetic information interacts with the environment so as to ensure the production of an organism. The geneticist examines the transmission of genetic information, while the evolutionist devotes attention to learning how genetic information progressively led to the creation of widely differing organisms. What interests the ethologist is how information travels among a group of living beings. The ecologist quantifies the information contained in the diversity of a community of species, or within the spatial distribution of the components of an ecosystem. The science of nature constantly encounters this notion of information while dispensing with a definition of the term.
To return to our example – this book contains information. Strictly speaking, the information is not non-material in that a material entity is carrying it. However, in a certain sense, the material nature of that entity is unimportant. That material might be paper, or magnetic tape, a USB key, a CD, a hard disk, etc. If you were to read the text aloud to someone else, it would be the vibrations of air that would carry the information. If you were to learn the contents by heart, then your cerebral neurons would store it by a yet-unknown method. In none of these ways would the information be reducible to the material that carried it; the two entities are clearly different. Imagine that someone asks you for your impression of the novel you last read. You might answer somewhat like this: “It had a red cover, and had 250 pages of 100-gram grained vellum, with black-and-white printing”. This is what could happen if you confuse information and its material medium.
Information, which is carried by a material medium, and from which, however, it differs, possesses some remarkable characteristics. For one thing, it can be transferred from one sort of medium to another while remaining more or less unchanged, though not under all conditions. For another, it can be done into several copies or versions. Yet again, it can act on matter, as we have seen above. What a funny entity, what a funny concept!
Copy a text, photocopy it, scan it, photograph it, print it, read it, memorize it, etc., there are so many ways of transferring information from one medium to another. Some of these methods are rather good at making copies of data that will be close to the original, but to be able to do this you would need certain special kinds of medium. Other methods permit the data to be expressed, though necessitating a change in form.
Making identical copies seems simple enough. Just copy the stuff – as simple as that…or almost! Mind the almost. The fact is that we can show with the help of thermodynamics that it is impossible to make a perfect copy of a given message without a single error, within a reasonable amount of time and with limited resources. In other words, a copy is bound to differ from the original message, given the constraints mentioned. If I were only to change a single symbol in a novel, say by turning it into bold, the novel would remain the same, except when, through sheer happenstance, the changed symbol could have meant a total change in the narration. We can imagine a story like that, but it is not going to be easy. A copy of a copy of a copy of a copy … would you have a copy at some point that does not differ from the original? There is this game in which one individual tells a story to the next one, and the last version turns out to be wholly different from the first one.1 We shall see that evolution and the dynamics of evolution work this way in biology and, equally, in evolution. Now, when is one to say that the information is no longer the same? Is not that when it becomes possible to quantify the difference between the original and the copy?
We may wish to express a particular piece of information by moving it into another medium. For instance, when I read a printed text, the information passes from paper onto my brain. A major modification has taken place here. My brain has become the interpreter. That said, it must be added that I may change, to some degree, the information that was delivered by the text I read and learnt by heart so that I might remember it. Did we say to some degree? That implies a quantitative variation. Is there a value hidden behind this idea? If we were to ask a philosopher or a linguist if it is possible to measure the information content of a text, their first reaction might be to say that such a thing would not be possible. We shall return to this point in Chapter 4. Most would agree, in this context, that a given text suffers a loss in translation. Now, how on earth could one speak of loss if there were not a notion of quantity in the first place?
The problem arises especially from the fact that a transfer of information, such as that of text printed on paper being moved onto my brain, is more than just a transfer. An action such as this underpins the very existence of the particular information. If I were to open a book that contained signs I did not understand, I would not receive the information that the printed text carried. Printed text only carries information if I am able to understand that text, if I am able to make sense of that text, if that text is capable of action on a form of matter. What now?
When all is said and done, the information the novel contained, or the information the recipe book contained, does not exist only on the paper on which it was printed. It did not exist in my brain until I read the text. It was necessarily the result of an interaction between the two. It is an undeniable fact that I cannot define a piece of information independently of the system for which it is meant. Did I say meant? Is it always the case that every piece of information should be meant for a receiver? Certainly not!
When I see the sun rising, I know roughly what time it is and, again roughly, my direction. The information the sunrise conveyed was not meant for me. In fact, there would have been no information in that event if I had not interpreted the sunrise the way I did. The event would have taken place even if no one interpreted it, but it would not have contained the particular information. Of course, this would not have been the case with the printed text, which would not have come into being if no one were able to interpret it. Here information has been the result of the way in which I interpreted what I perceived in the environment. The environment contained information since there was in existence a reading system, mine. This is not something we can overlook, and we shall return to it. However, what interests us in the first instance is structured information, coded in a complex manner and meant to be interpreted.
We will distinguish three categories of elements that work in concert to make up information in the full sense of the concept. They are collectively present in all informational phenomena, and it is difficult to decide in which order they should be taken up. However, we will start in the order of their appearance in this introduction. We will call the ensemble MDE.
The illustration of the cookery book recipe appears particularly apt. We all know that the three categories of the elements are going to work. They will produce the recipe as their message. The recipe has a history behind it. The decoder is the person who will create the dish: the cook or the chef. This person himself or herself has a history of their own. Finally, the environment is made up of diverse elements, including the location where the action is taking place, the available ingredients, the time available, the psychological atmosphere and so on. The environment might have been influenced by the person working in the kitchen, and by the fact that he or she was going to use the recipe. Certainly, we might find among other elements things that arose out of recipes that were used earlier. We also recognize the possibility that a single recipe could yield dishes that differ among themselves depending on the decoder (the cook) and the environment (the style of cooking and the ingredients).
In biology, the coded message is what is called genetic information. The cookery book recipe was a metaphor for the concept. However, one might risk overlooking an underlying reality if one were to treat all comparisons as metaphors. A classic instance of this possibility would be the theory of natural selection that Charles Darwin proposed in 1859. Darwin was inspired by the notion of selection, which today is described as artificial, but found to be valid by cattle farmers and nurserymen, and it was seen that a similar process existed in nature. That was a metaphor for most philosophers, but it was an extension of a concept for most biologists. For the latter, we can give a formal definition of the concept of selection that should both satisfy the evolutionist and accord with nature. The same process is at work in both cases. The fact that we have used the word “process” earlier for a human activity does not implicitly confer primacy on that aspect of the concept.
This does not appear to be an isolated instance of ambiguity. A natural process often cannot be fully understood if human industry has invented something to mimic it. It is said that William Harvey, circa 1628, had visualized the natural pacemaker for the heart in an age when pumps were being widely used. At the same time, a metaphor need not conceal an extension of a concept. In the following sections, a return to the cookery book recipe, and the other metaphors, need not hide the fact that the idea in the message applies equally well to the genome, and there will always be a decoder (human or not) and an environment to receive the message.
Now it is not a question of giving a complete definition of information as it is of producing a plain outline of its characteristics that we consider of interest to us. Every sign can represent an element of information and generate an action. From a falling body that conveyed information as to the direction of its weight, to another whose state indicates temperature and to signals that can set off more or less automatic or reflex reactions, all will be considered to represent information.
What is relevant to life sciences is information coded in a message made up of a series of signs none of which means anything by itself, but an almost unlimited combination of which can design a multitude of things. Such information, as we have seen, is distributed between the message itself, the decoder and the environment. This information can replicate itself fairly faithfully and induce an action through the decoder. One such action, predictably and significantly, is the replication of the message itself. The languages out of those that have been found and studied are the human languages and those of the nucleic acids (DNA and RNA). We will see when we return to the concept of language that there are several other languages in nature. In particular, the functioning of ecosystems might conceal languages through which ecosystemic information might be constructed.
We have seen that a message is carried by a material entity. In the case of a message in a human language, the medium involved could be one of many types: the brain, a surface (stone, paper and so forth), a magnetic medium or data processing. The medium is much more homogeneous in the case of the genetic language. As far as we know, it is a nucleic acid: DNA or RNA. Technical progress in biology and data processing has in recent years enabled transfer of genetic information to a computer. Sequences of the DNA of bacteria have been read and stored in a computer’s memory. This is what happened when they were modified and bundled into a new genome by engineers in Craig Venter’s group. A genome was thus “written” into the computer, so that it synthesized a carrier chromosome of this genome. The chromosome was then introduced into a bacterium, and decoded in the usual manner and used. While a bacterium has not yet been synthesized, a genome has been. The point to note is that genetic information has been able to move through a computer, then to return in the DNA, and then serve as it would normally do.
At the base of the human language the elements are sounds, phonemes and signs written on media. Elementary sounds vary in number and content depending on the language. Linguists study all these aspects of language. As for the language of genetics, the four nucleotides of DNA and the four of RNA are still (to our knowledge) the same for all forms of terrestrial life (humans, in their usual humility, call them “universals”). These four signs, which collectively form the message of a nucleic acid, are the four nucleotides currently designated by the letters A, T, C and G for the DNA, and by A, U, C and G for the RNA. Molecular biology describes the functions of these elements.
The decoders of the human language are, of course, the human beings themselves. The collection of mechanisms with which the human being decodes a message and is able to transform it into action flows from a huge ensemble of disciplines: physiology, biochemistry, neurobiology, linguistics, psychology, biomechanics and so on. The mechanism by which the human being acquires the ability to achieve a particular task involves biological and social components, and, specifically, training in the language in which the message is coded. Another type of reading system developed with robotics. Robots are devices that make it possible to carry out tasks following human commands, which are in fact information programs. For the language of genetics, the reader is the living entity, and allows its genome expression. Elucidation of part of this decoder was a major scientific adventure of the 20th Century. The discovery of the genetic code (in the former sense, to which we shall return later) permits one to read the information carried by the DNA through a transcription in the RNA, and then a translation in protein. The discovery also allowed us to see how hereditary information was formed, transmitted and expressed. Let us, however, bear in mind that the information is collectively held in the ensemble MDE, which is made up of the message (the genome or the document), the reading system, which can decode the message (the living entity or its components) and the environment. Richard Charles “Dick” Lewontin called this the triple helix.
In this book, we start by examining information in the context that is the most natural for us: the human language. It would be interesting to begin by examining the prism of human and social relations, and then go on to the less familiar variations in animal communication. It seems reasonable that genetic information should open the discourse with different examples. We will try to stay clear of whatever new things biological languages might have brought to the concept of information. This position will encourage a fresh perspective on studies of ecological information. Ecosystems do carry information, but can we identify hidden ecological languages? Could languages of a new kind help us understand and manage our ecosystems better? We will end with a brief overview of information and the languages that code it. In this context, natural languages, refining and transmitting information, have sustained the various advances that have marked the different epochs of our planet, from its ecosystems right down to its denizens, and in particular, human beings and their cultures.
We would like to warmly thank Alan Bailur for his careful English translation and editing of a preliminary version of this text.