9 Information and communication in living matter

Bernd-Olaf Küppers

Ever since the elucidation of the molecular basis of living systems, we have known that all elementary processes of life are governed by information. Thus, information turns out to be a key concept in understanding living matter (Küppers, 1990). More than that: the flow of information at all levels of the living system reveals the properties of communication. This means that the information stored in the genome of the organism is expressed in innumerable feedback loops – a process through which the genetic information is continually re-evaluated by permanent interactions with the physical environment to which it is exposed. In this way, the living organism is built up, step by step, into a hierarchically organized network of unmatched complexity.

The fact that all phenomena of life are based upon information and communication is indeed the principal characteristic of living matter. Without the perpetual exchange of information at all levels of organization, no functional order in the living organism could be sustained. The processes of life would implode into a jumble of chaos if they were not perpetually stabilized by information and communication. In this chapter, I should like to consider some of the consequences that follow from this for our philosophical understanding of reality.

9.1 About “information” and “communication”

In daily usage, the terms “information” and “communication” are not always clearly distinguished from each other. Yet, even the etymology of the two words indicates that the reference of the concepts cannot entirely overlap. The term “information” – following closely its Latin root informare – denotes primarily the formative, and thus instructive function of a message. By contrast, the word “communication” – derived from the Latin word communicare – denotes the process by which the sender and the receiver of information try to reach a common understanding. The subject of this understanding is a common evaluation of the information exchanged between the sender and the receiver. Alongside this, the bare instruction for which the word “information” stands seems like a command that results in the mechanical: that is, the unilateral transfer of the information from sender to receiver, without any aim of achieving a common or mutual understanding of the “meaning” of the information being expressed by its operative function.

Thus, if we wish to approach the concept of communication in living matter in its widest sense, we need to examine the relationship between information on the one hand and mutual or common understanding on the other. At the same time, we shall need to demonstrate that concepts such as “information” and “communication” can meaningfully be applied to natural processes. The latter task would seem to raise fewer difficulties for the concept of information than it does for that of communication.

Information, as suggested above, means primarily “instruction,” in the sense of a command or step in a computer program. This in turn has the function of imposing a selective condition on the possible processes that can go on in a system. In precisely this sense, living processes are “instructed” by the information that is contained in encoded form in genes. Expressed in the terms of physics: the genome represents a specific physical boundary condition, a constraint that restricts the set of physically possible processes to those that do actually take place within the organism and are directed towards preservation of the system (Küppers, 1992). Thus, the idea of “instruction by information” has a precise physical meaning, and in this context information can be indeed regarded as an objective property of living matter.

It is a harder task to demonstrate the universality of communication. One tends to assume intuitively that this concept is applicable only to the exchange of information between human beings. This assumption arises from the fact that the idea of a “common” understanding seems to make no sense outside the realm of human consciousness. However, this could be a false premise based upon a narrow use of the concept of understanding. Reaching a common understanding usually means reaching agreement. This in turn implies that one must understand one another in the sense that each party can comprehend what the other party intends to communicate. However, attaining a common understanding does not necessarily presuppose any reflections upon the nature or the subject of the communication process, nor does it imply the question of whether the contents of the communication are true or false. Rather, it requires only a mere exchange of information; that is, a number of messages to be sent in both directions – without, however, either party necessarily being aware of the meaning of what is being communicated.

There is thus a subtle difference in scope between “a reflected understanding” and “reaching a coordinated reaction.” If we are for a moment willing to put aside the highly sophisticated forms of human understanding, and to operate with a concept of understanding that encompasses only the objectives of achieving a coordinated reaction, then it becomes easy to see how this concept is applicable to all levels of living matter. We thus have to concede that molecules, cells, bacteria, plants, and animals have the ability to communicate. In this case, “communication” means neither more nor less than the reciprocal harmonization and coordination of processes by means of chemical, acoustic, and optical signals.

9.2 About “understanding”

The foregoing arguments have taken me along a path that some philosophers of science have branded “naïve” naturalism. Their criticism is directed especially at the idea that information can exist as a natural object, independently of human beings: that is to say, outside the myriad ways in which humans communicate. This charge of naturalism is heard from quite diverse philosophical camps. However, all such critics share the conviction that only human language can be a carrier of information, and that the use of linguistic categories in relation to natural phenomena is nothing more than a naturalistic fallacy. For representatives of this philosophical position, any talk of information and communication in the natural sciences – as practiced especially in modern biology – is no more than a metaphor that reveals, ultimately, a sadly uncritical usage of terms such as “language” and “understanding.”

Let us examine this controversy more closely and ask once more the question of what we actually understand by the word “understanding.” The tautological way in which I express this question indicates that one can easily get into a vicious circle when trying to approach the notion of understanding. This is because it seems generally to be the case that one can only understand something if one has understood some other things. This plausible statement is central to philosophical hermeneutics, the best-known and most influential doctrine of human understanding (Gadamer, 1965).

The hermeneutic thesis, according to which any understanding is bound to some other understanding, obviously refers to the total “network” of human understanding in which any kind of understanding is embedded. In other words: any form of communication presupposes some prior understanding, which provides the necessary basis for a meaningful exchange of information. In fact, there seems to be no information in an absolute sense – not even as a plain syntactic structure – as the mere identification of a sequence of signs as being “information” presupposes a foregoing knowledge of signs and sequences of signs. In short: information exists only in a relative sense – that is, in relation to some other information.

Thus, even if we adopt an information-theoretical point of view, there seems to be no obstacle to the hermeneutic circle, according to which a person can only understand something if he has already understood something else. Nevertheless, this perspective contradicts the intentions of philosophical hermeneutics, which puts a completely different construction upon the hermeneutic circle. Within the framework of this philosophy, the pre-understanding of any kind of human understanding is thought to be rooted in the totality of human existence. And this ontological interpretation is intended to lead not to a relative but to an absolute and true understanding of the world.

Moreover, because we use language to comprehend the world, the hermeneutic school regards language as the gate that opens for us the access to our existence. The philosopher Hans-Georg Gadamer (1965, p. 450) has expressed this in the often-quoted sentence: “Being that can be understood is language.” Even though some prominent philosophers of the hermeneutic school assign a special role to dialogue, their concept of understanding still fails to possess the objectiveness and relativity that characterize a critical comprehension of human understanding. On the contrary: a world view that rests its claims to validity and truth exclusively upon the rootedness of understanding in human existence has moved to the forefront and become the absolute norm for any understanding at all.

So, in contrast to the relativistic world picture offered to us by modern science, philosophical hermeneutics seeks to propagate a fundamentalism of understanding that is centered firmly on the philosophical tradition of absolute understanding. Moreover, if human language is considered to be a prerequisite for all understanding, human language becomes the ultimate reference in our relation to the world.

The thesis of this chapter, which intends to give language a naturalistic interpretation that allows us to speak of the “language” of genes, seems to be diametrically opposed to this position. According to the naturalistic interpretation, which is shared by other biologists, language is a natural principle for the organization of complex systems, which – in the words of Manfred Eigen (1979, p. 181) – “can be analyzed in an abstract sense, that is, without reference to human existence.” From the standpoint of philosophical hermeneutics, such use of the word “language” is completely unacceptable. From this perspective, biologists who think and speak in this way about the existence of a “molecular language” look like drivers thundering down the motorway in the wrong direction – ignoring all the signposts naturally provided by human language for comprehending the world.

9.3 The “language” of genes

Impressive evidence for the naturalistic view of language seems to be found in the language-like arrangement of genetic information. Thus, as is well known, the genetic alphabet is grouped in higher-order informational units, which in genetic handwriting take over the functions of words, sentences, and so forth. And, like human language, genetic information has a hierarchical structure, which is unfolded in a complex feedback mechanism – a process that shows all the properties of a communication process between the genome and its physical context.

Of course, the parallels break down if we try to use the full riches of human language as a measure of the “language-like” structure of the genome. But from an evolutionary standpoint, there are good grounds to assert that “language” is indeed a natural phenomenon, which originates in the molecular language of the genome and has found, in the course of evolution, its hitherto highest expression in human language (Küppers, 1995). For evolutionary biologists, there is no question as to whether languages below the level of human language exist; the issue is rather about identifying the general conditions under which linguistic structures originate and evolve.

The significance of the natural phenomenon “language” for the explanation of living matter was recognized and first expressed with admirable clarity at the end of the nineteenth century by Friedrich Miescher, the discoverer of nucleic acids. Asking how a substance such as a nucleic acid can generate the vast diversity of genetic structures, he drew an analogy to the principles of stereochemistry. In the same way – Miescher argued – that a narrow variety of small molecular units is able to build up large molecules of almost unlimited complexity that are chemically very similar, but which have very different structures in three dimensions, the nucleic acids are capable of instructing the vast diversity of genetic structures. This line of thinking led Miescher to the conclusion that the nucleic acids must be able to “express all the riches and all the diversity of inheritance, just as words and ideas in all languages can be expressed in the 24–30 letters of the alphabet” (Miescher, 1897, p. 116). Obviously Miescher’s view of living matter was that of a “linguistic movement” rather than that of a “clockwork machine.” However, the “linguistic movement” of living matter is not a dead system of rules, but a dynamic one.

So, is this all just metaphoric speech? An outside observer, watching the disputants from a distance, might wonder what the controversy is all about, and might even suspect that it was a typical philosophers’ war over the meaning of words. Our observer would be certain to draw attention to the fact that we repeatedly take words out of their original context and transpose them into another, so that any discourse about the world of nature is bound to employ metaphors, at least to a certain extent.

Why, then, should we not simply regard terms such as “information,” “communication,” and “language” in biology as what they really are: namely, adequate and highly resilient media for the description of the phenomena of life? Do the recent spectacular successes at the interface between biotechnology and information technology not justify the use of these concepts in biology? The construction of bio-computers, the development of genetic algorithms, the simulation of cognitive processes in neural networks, the coupling of nerve cells to computer chips, the generation of genetic information in evolution machines – all these would scarcely be conceivable without the information-theoretical foundations of living matter provided by biology.

However, the foregoing questions cannot be disposed of with simple arguments. This is above all because “information,” “communication,” and “language” are charged with other notions such as “meaning,” “value,” “truth,” and the like. And this is where we run into the real nub of the discussion. Phenomena associated with meaning, as expressed in the semantic dimension of information, appear to evade completely all attempts to explain them on a naturalistic basis, and thus also to escape scientific description.

The right to interpret phenomena of meaning has traditionally been claimed by the humanities: especially by its hermeneutic disciplines. They have placed meaning, and thus also the understanding of meaning, at the center of their methodology; a clear demarcation against the natural sciences may indeed have been one of the motives for this. Whatever the reasons, the humanities have long gone their own way, have not considered it necessary to subject themselves to the scientific method of causal analysis – and have thus retained their independence for a considerable length of time.

The question of how broadly the concept of information may be applied is thus by no means a dispute about the content and the range of the applicability of a word. It would be truer to regard this question as the focal point at which philosophical controversies about the unity of knowledge converge – debates that have determined the relationship of the humanities and the natural sciences for more than a hundred years. The biological sciences, which stand at the junction between these two currents of thought, are always the first to get caught in the crossfire. This is because an information-theoretical account of living matter involving a law-like explanation necessarily introduces questions of meaning and, thus, the semantic aspect of information (Küppers, 1996). Furthermore, the introduction of the semantic aspect of information in turn leads to the most fascinating plan-like and purpose-like aspects of living matter, which have every appearance of overstretching the capacity of traditional scientific explanation. Are, then, physical explanations – and with them the entire reductionistic research program in biology – doomed to founder on the semantic aspect of information?

9.4 The semantic dimension of information

Our discussion up to now has suggested that semantic information is “valued” information. The value of information is, however, not an absolute quantity; rather, it can only be judged by a receiver. Thus, the semantics of information depend fundamentally upon the state of the receiver. This state is determined by their prior knowledge, prejudices, expectations, and so forth. In short: the receiver’s evaluation scale is the result of a particular, historically unique, pathway of experiences. Can – we may persist in asking – the particular and individual aspects of reality ever become the object of inquiry in a science based upon general laws and universal concepts? Even Aristotle addressed this important question. His answer was a clear “No.” For him – the logician – there were no general discoveries to be made about things that were necessarily of an individual nature, because the logic of these two attributes – general and particular – made them mutually exclusive. This view has persisted through to our age, and has left a deep mark upon our present-day understanding of what science is and does.

Under these circumstances, the achievement of the philosopher Ernst Cassirer appears all the more admirable. Opposing the Aristotelian tradition, Cassirer attempted to bridge the presumed gap between the general and the particular (Cassirer, 1910). Particular phenomena, he argued, do not become particular because they evade the general rules, but because they stand in a particular – that is, singular – relationship to them. Cassirer’s reflections may have been triggered by an aperçu of von Goethe (1981, p. 433): “The general and the particular coincide – the particular is the general as it appears under various conditions.”

According to Cassirer, it is the unique constellation of general aspects of a phenomenon that makes up its uniqueness. This is an interesting idea. It makes clear that an all-embracing theory of semantic information is impossible, whereas general aspects of semantics can very well be discerned. Taken for themselves, these aspects may never completely embrace the phenomenon in question. But through their unique interconnectedness, they allow the particular characteristics of the phenomenon to show through clearly. In other words: the unique properties of semantic information originate by superposition of its general disposition. The aspects that constitute semantic information in this sense include, among others, their novelty and their pragmatic relevance as well as their complexity (Küppers, 1996).

At the beginning of the 1950s, the philosophers and logicians Yehoshua Bar-Hillel and Rudolf Carnap (1953) tried to quantify the meaning of a linguistic expression in terms of its novelty value. This idea was a direct continuation of the concept developed within the framework of Shannon’s information theory, where the information content of a message is coupled to its expectation value: the lower the expectation value of a message, the higher its novelty and thus its information content. This approach takes care of the fact that an important task of information is to eliminate or counteract uncertainty. However, the examples adduced by Bar-Hillel and Carnap are restricted to an artificial language.

A more powerful approach to measuring the semantics of information is that based upon its pragmatic relevance. This approach has been described in a paradigmatic way by Donald MacKay (1969) in his book Information, Mechanism and Meaning. The pragmatic aspect of information refers to the action(s) of the receiver to which the information leads, or in which it results.

For some time now, my own efforts have been focused on a new approach, intended to investigate the complexity of semantic information (Küppers, 1996). Unlike the approaches described above, this one does not seek to make the meaning of information directly amenable to measurement. Rather, it aims to demarcate the most general conditions that make up the essence of semantic information. Investigations of this kind are important because they afford a more general insight into the question of the origin of information, and therefore have consequences for major fundamental problems of biology such as the origin and evolution of life (Küppers, 2000a).

9.5 How does information originate?

Let us consider the relationship between semantic information and complexity in more detail. Information, as we have said, is always related to an entity that receives and evaluates the information. This in turn means that evaluation presupposes some other information that underlies the process of registration and processing of the incoming information. But how much information is needed in order to understand, in the foregoing sense, an item of incoming information? This question expresses the quantitative version of the hermeneutic thesis, according to which a person can only understand some piece of information when it has already understood some other information.

At first sight, it would seem impossible to provide any kind of answer to this question since it involves the concept of understanding, which, as we have seen, is already difficult to understand by itself, let alone to quantify. Surprisingly, however, an answer can be given, at least if we restrict ourselves to the minimal conditions for understanding. To this belongs first of all the sheer registration by the receiver of the information to be understood. If the information concerned conveys meaning – that is, information of maximum complexity – then the receiver must obviously record its entire symbol sequence before the process of understanding can begin. Thus, even the act of recording involves information of the same degree of (algorithmic) complexity as that of the symbol sequence that is to be understood.

This surprising result is related to the fact that information conveying meaning cannot be compressed without change in, or even loss of, its meaning. It is true that the contents of a message can be shortened into a telegram style or a tabloid headline; however, this always entails some loss of information. This is the case for any meaningful information: be it a great epic poem or simply the day’s weather report. Viewed technically, this means that no algorithms – that is, computer programs – exist that can extrapolate arbitrarily chosen parts of the message and thus generate the rest of the message. But if there are no meaning-generating algorithms, then no information can arise de novo. Therefore, to understand a piece of information of a certain complexity, one always requires background information that is at least of the same complexity. This is the sought-after answer to the question of how much information is needed to understand some other information. Ultimately, it implies that there are no “informational perpetual-motion machines” that can generate meaningful information out of nothing (Küppers, 1996).

This result is the consequence of a rigorous relativization of the concept of information. It is a continuation of the development that characterized the progress of physics in the last century: the path from the absolute to the relative. This began with the abandoning of basic concepts that had been understood in an absolute sense – ideas such as “space,” “time,” and “object” – and has since led to well-known and far-reaching consequences for the foundations of physics. Whether the thorough-going relativization of the concept of information will one day lead to a comparable revolution in biological thinking cannot at present be said. This is largely due to the fact that the results up to now have been derived with respect to the semantic dimension of human language, and it is not yet clear to what extent they are applicable to the “language of genes.” For this reason, questions such as whether evolution is a sort of perpetual-motion machine must for the present remain open.

At least it is certain that we must take leave of the idea of being able, one day, to construct intelligent machines that spontaneously generate meaningful information de novo and continually raise its complexity. If information always refers to other information, can then information in a genuine sense ever be generated? Or are the processes by which it arises in nature or in society nothing more than processes of transformation: that is, translation and re-evaluation of information, admittedly in an information space of gigantic dimensions, so that the result always seems to be new and unique? Questions such as these take us to the frontline of fundamental research, where question after question arises, and where we have a wealth of opportunities for speculation but no real answers.

9.6 The world of abstract structures

Finally, I should like to return briefly to the question with which we began: Are the ideas of “information,” “communication,” and “language” applicable to the world of material structures? We saw how difficult it is to decide this on a philosophical basis. But it may also be the case that the question is wrongly put. There does indeed seem a surprising solution on the way: one prompted by current scientific developments. In the last few decades, at the border between the natural sciences and the humanities, a new scientific domain is emerging that has been termed “structural sciences” (Küppers, 2000b). Alongside information theory, it encompasses important disciplines such as cybernetics, game theory, system theory, complexity theory, network theory, synergetics, and semiotics, to mention but a few. The object of structural sciences is the way in which the reality is structured – expressed, investigated, and described in an abstract form. This is done irrespectively of whether these structures occur in a natural or an artificial, a living or a non-living, system. Among these, “information,” “communication,” and “language” can be treated within structural sciences as abstract structures, without the question of their actual nature being raised. By considering reality only in terms of its abstract structures, without making any distinction between objects of “nature” and “culture,” the structural sciences build a bridge between the natural sciences and the humanities and thus have major significance for the unity of science (Küppers, 2000b).

In philosophy, the structural view of the world is not new. Within the frame of French structuralism, Gilles Deleuze took the linguistic metaphor to its limit when he said that “There are no structures that are not linguistic ones … and objects themselves only have structure in that they conduct a silent discourse, which is the language of signs” (Deleuze, 2002, p. 239). Seen from this perspective, Gadamer’s dictum “Being that can be understood is language” (Gadamer, 1965, p. 450) takes on a radically new meaning: “Being” can only be understood when it already has a linguistic structure. Pursuing this corollary, the philosopher Hans Blumenberg (2000), in a broad review of modern cultural history, has shown that – and how – the linguistic metaphor has made possible the “readability” (that is, the understanding) of the world. However, the relativity of all understanding has of necessity meant that the material “read” was reinterpreted over and over again, and that the course of time has led to an ever more accurate appreciation of which “readings” are wrong. In this way, we have approached, step by step, an increasingly discriminating understanding of the reality surrounding us.

References

Bar-Hillel, Y., and Carnap, R. (1953). Semantic information. British Journal for the Philosophy of Science, 4: 147.

Blumenberg, H. (2000). Die Lesbarkeit der Welt. Frankfurt/Main: Suhrkamp.

Cassirer, E. (1910). Substanzbegriff und Funktionsbegriff. Berlin: Bruno Cassirer.

Deleuze, G. (2002). À quoi reconnaît-on le structuralisme? In L’Äle Déserte et Autres Textes, ed. D. Lapoujade. Paris: Minuit, 238–269.

Eigen, M. (1979). Sprache und Lernen auf molekularer Ebene. In Der Mensch und seine Sprache, eds A. Peisl and A. Mohler. Frankfurt/Main: Propyläen Verlag, 181–218.

Gadamer, H.-G. (1965). Wahrheit und Methode, 2nd ed. Tübingen: J. B. C. Mohr.

Küppers, B.-O. (1990). Information and the Origin of Life. Cambridge, MA: The MIT Press.

Küppers, B.-O. (1992). Understanding complexity. In Emergence or Reduction? Eds A. Beckermann, H. Flohr and J. Kim. Berlin: de Gruyter, 241–256 [reprinted in Chaos and Complexity, eds R. J. Russell, N. Murphy and A. R. Peacocke (1995). Vatican City State: Vatican Observatory Publications, 93–105].

Küppers, B.-O. (1995). The context-dependence of biological information. In Information. New Questions to a Multidisciplinary Concept, eds K. Kornwachs and K. Jacoby. Berlin: Akademie Verlag, 135–145.

Küppers, B.-O. (1996). Der semantische Aspekt von Information und seine evolutionsbiologische Bedeutung. Nova Acta Leopoldina, 294: 195–219.

Küppers, B.-O. (2000a). The world of biological complexity: Origin and evolution of life. In Many Worlds, ed. S. J. Dick. Pennsylvania: Templeton Foundation Press, 31–43.

Küppers, B.-O. (2000b). Die Strukturwissenschaften als Bindeglied zwischen Natur- und Geisteswissenschaften. In Die Einheit der Wirklichkeit, ed. B.-O. Küppers. Munich: Fink Verlag, 89–105.

MacKay, D. M. (1969). Information, Mechanism and Meaning. Cambridge, MA: The MIT Press.

Miescher, F. (1897). Die histochemischen und physiologischen Arbeiten. Bd. I. Leipzig: F. C. W. Vogel.

von Goethe, J. W. (1981). Werke, Hamburger edition vol. 12. München: C. H. Beck.

      Information and the Nature of Reality: From Physics to Metaphysics, eds. Paul Davies and Niels Henrik Gregersen. Published by Cambridge University Press © P. Davies and N. Gregersen 2010.