Bernd-Olaf Küppers is a physicist, philosopher, and theoretical biologist, formerly at the Max Planck Institute for Biophysical Chemistry (Göttingen), then professor of natural philosophy at the University of Jena. His main research field is the origin and evolution of genetic information. He is author and editor of numerous books, including the monographs “Molecular Theory of Evolution” and “Information and the Origin of Life”. He has been recognized with, among other things, memberships of the German National Academy of Sciences and the Academia Europaea (London).

Küppers has published four books important for information philosophy, the 1983 Molecular Theory of Evolution: Outline of a Physico-Chemical Theory of the Origin of Life, the 1990 Information and the Origin of Life, the 2018 The Computability of the World: How Far Can Science Take Us? , and the 2022 The Language of Living Matter: How Molecules Acquire Meaning.

Küppers varied the famous quote by Theodore Dobzhansv "Nothing in Biology Makes Sense Except in the Light of Evolution" to become "Nothing in Biology Makes Sense Except in the Light of Information."

In the summary to his 1990 book Information and the Origin of Life, Küppers wrote...

The question of the connection between law and chance in theevolution of life is one of the central philosophical problems inbiology. In this book, fundamental aspects of this question have been treated from the viewpoint of modern evolutionary biology.

The investigation starts from the assumption that the object of its analysis counts as a well-founded part of natural science and does not need to be specially justified. It presupposes that evolution is an established fact, and that the synthetic evolution theory built upon the basis of Darwin's ideas provides an appropriate framework for the understanding and explanation of evolutionary phenomena (chapter 1).

However, the issue of the theoretical character of synthetic evolution theory is a problem in the philosophy of science that possesses a special significance with regard to the more general question of the connection between law and chance in evolution. Access to this problem is afforded in particular by the analysis of the so-called molecular theory of evolution, which links the principles of synthetic evolution theory with the fundamental laws of physics and chemistry. At the same time, many aspects of this connection throw new light upon the general problems of the formation of concepts and theories in the border region of physics, chemistry, and biology.

The molecular theory of evolution has been a result of rapid progress in biology and physics in the last two decades. It is based on the one hand upon the discovery in molecular biology that all basic phenomena of life such as metabolism and heredity can be traced back to regular interactions between biological macromolecules, and thereby to the laws of physics and chemistry, and on the other hand upon the discovery in physics of open systems that, far from equilibrium, can spontaneously assemble states of material order (so-called dissipative structures) that are also characteristic of living systems.

The molecular theory of evolution describes the origin and the early evolution of life as a process of material self-organization, in the process of which the two most important classes of biological macromolecule (nucleic acids and proteins) were able to organize themselves spontaneously into a system, governed by information, that possessed the three basic properties of living matter: metabolism, self-reproduction and mutability.

The modern theory of the origin of life is a physicochemical theory. As such, it attempts to base the historical process ofbiological self-organization upon its fundamental, unchanging principles and mechanisms. The theory does not claim to be able to reconstruct the process in its historical details.In the overall process of biological self-organization, three phases, each corresponding to the degree of optimization attained at thattime, can be distinguished ( chapter 2):

1. the phase of noninstructed prebiotic synthesis of biologicalmacromolecules,

2. the phase of self-organization of biological macromolecules to aself-instructed biosynthetic cycle, and

3. the phase of evolutionary optimization of the biosynthetic cycle.

The actual transition from inert to living matter took place during the second phase, and was essentially a transition from noninstructed to instructed synthesis of biological macromolecules. This phase of biological self-organization ended with the nucleation of a self-instructed biosynthetic cycle, which (after the formation of compartments) can be regarded as a primitive precursor of thel iving cell.

Every kind of instruction requires information. The information used by the self-instructed biosynthetic cycle is encoded in the detailed sequence of its nucleic acid components. These instruct the construction of a protein machinery that in turn catalyzes the reproduction of the entire cycle.

It has emerged that the information for the construction of a living organism is not simply encoded in the linear sequence of the monomers of its DNA molecules, but that this linear arrangement - as in human language -possesses a hierarchical superstructure. If we have on the whole restricted our consideration to the sequential aspect of information storage, this has been because the present investigation has been directed exclusively toward the initial phase of the origin of life. The hierarchical organization of gene structure, in contrast, is a comparatively late product of evolution. In particular, it encodes the information needed for the complex regulation and control phenomena of higher organisms.

The problem of the origin of life is clearly basically equivalent to the problem of the origin of biological information. In accordance with this, the idea of biological information emerges as the fundamental concept in the physicochemical theory of the origin of life.

In part II we discussed the distinction between three dimensionsof information, termed syntactic, semantic, and pragmatic. The syntactic aspect of information comprises the relationship between the individual characters (chapter 3). The semantic aspect comprises the relationship between the individual characters and what they stand for (chapter 4). The pragmatic aspect comprises the relationship between the individual characters, what they stand for, and what action this implies for the sender and the recipient (chapter 5).

From this viewpoint, the three phases of biological self-organization (above) can be expressed as three phases of the origin of biological information:

1. the phase of the origin of syntactic information,

2. the phase of the origin of semantic information, and

3. the phase of evolutionary optimization of semantic information.

The synthetic aspect of the origin of biological information is unproblematic from a philosophical standpoint, since it is concerned solely with the formation of structures - for example, of macromolecules such as proteins or nucleic acids. In contrast, the semantic aspect raises many problems, since it refers to the content, that is, the sense and meaning of the information encoded in macromolecular structure. This raises the question of whether and, if so, to what extent "sense" and "meaning" can be objectified and studied scientifically.

As the example of human language shows, semantics cannot exist in an absolute sense, but only relative to a semantic frame of reference (chapter 4). Genetic information, too, possesses no absolute semantic value, but only a relative one, referred to the specific environmental conditions to which the organism in question has become adapted. The environment thus represents externally localized information, on the basis of which the semantics of genetic information are selectively evaluated.

According to Darwinian evolution theory, the purpose-directed, or information-controlled, element of biological structures possesses a certain function (and thus sense and meaning) with regard to the preservation of the dynamic order that is characteristic of living systems. The semantics of genetic information can thus be interpreted directly in terms of evolution: an organism is the better adapted to its environment (that is, the semantic information it contains has the greater value), the more accurately and rapidly it - for a given lifespan -reproduces its information content.

This principle provides a link with two philosophical theses of Carl Friedrich von Weizsacker, according to which the semantic aspect of information can only be objectified via the pragmatic aspect of information ("Information is only that which is understood"; "Information is only that which produces information").This thought runs through the entire discussion like a leitmotiv. It implies that even the "protosemantics" of biological information can be objectified by a dynamic criterion of the generation of information (see below).

The question of whether and, if so, to what extent the origin of biological information can be explained in terms of regularity within the framework of modern science is taken up in part III. Three suggested solutions are presented: (a) the chance hypothesis, in chapter 6, (b) the teleological approach, in chapter 7, and (c) the molecular-Darwinistic approach, in chapter 8.

The chance hypothesis interprets the origin of biological information as a singular event, whose description by the laws of physics and chemistry can only take place on the syntactic level, that is, on the level of chemical evolution. The semantic aspect of the origin of biological information, in contrast, is understood in the chance hypothesis as an epiphenomenon, which by its very nature is determined purely by chance and possesses none of the regularities of natural law. The chance hypothesis is justified by the assertion, at first seemingly contradictory, that within equilibrium statistics the probability of the chance synthesis of an information-carrying macromolecule is extremely low, because the information-carrying macromolecules represent only a minute fraction of their physically equivalent alternative sequences, and all these alternatives have under equilibrium conditions virtually identical prior probabilities. Consequently, in the context of the chance hypothesis, the origin of life is interpreted a posteriori as a singular random event, which because of its extremely low prior probability is unique and unrepeatable in the whole universe.

The teleological approach proceeds similarly from the explanatory deficit of equilibrium statistics, just described. In contrast to the chance hypothesis, which does not offer a causal explanation for the origin of biological information, the teleological approach postulates the existence of a goal-directed and at the same time irreducible, lawlike behavior, which narrows down the immense variety of possible macromolecular structures to those that are biologically relevant. In the teleological model of explanation, information-carrying biomolecules must be regarded as the material expression of a finalistic principle at work in Nature. The teleological approach and the chance hypothesis agree that the specific arrangement of the monomers of the DNA molecules in an organism, encoding regularity and purposiveness, are fundamentally inexplicable at the levels of physics and chemistry.

In the distribution of roles between law and chance in the origin of biological information, the molecular-Darwinistic approach adopts an intermediate position. In the molecular-Darwinistic model, the origin of biological information is derived from an interplay between undirected, random process (mutation) and regular behavior of matter (natural selection). The mathematical formulation of this approach is the molecular theory of evolution sketched out at the beginning. However, and here the molecular-Darwinistic approach differs fundamentally from its rivals discussed above, the interplay between mutation and selection is based in an essential way upon the principles of nonequilibrium thermodynamics. Further, the molecular-Darwinistic approach rests upon the working hypothesis that natural selection in the Darwinian sense appears already in the realm of inanimate matter.

The discussion in part III makes it clear that a philosophical analysis of current hypotheses concerning the origin of biological information presupposes a formulation of the concept of law so general that even such diverse approaches as the chance hypothesis and the teleological approach can be related to one another, so to speak on a higher level of the formation of concepts and theories.

Now, a characteristic feature of a natural law is that it contains information about natural events in a compressed form. The logical next step is to look for a general formulation of the concept of law within the framework of information theory.

The information for the construction of a living organism is encoded in the detailed sequence of the monomers in its DNA molecules. This sequence can be found out, usually by experimental (chemical) sequence determination. Like any other empirical data in science, it can be represented by a finite row of ones and zeros (chapter 10, figure 19).

The representation of observational data as binary sequences leads directly to an information-theoretical interpretation of the concept of natural law. Clearly, a lawlike relationship is always contained in a series of observational data when there is a compact algorithm with the help of which the observed data can be represented in compressed form. The central feature of such an algorithm is the property that it can be represented (on the syntactic level) by using less information than the information needed for the representation of the data sequences that it generates. Conversely, a sequence of observational data must be regarded as random until an appropriate compact algorithm has been found. The term "random sequence" in this way acquires a precise algorithmic definition (chapter 9). However, the irreducibility, that is, the randomness, of a binary sequence is inherently incapable of proof, since the nonexistence of a compact algorithm capable of generating the sequence can never be proven (the randomness theorem).

From the randomness theorem, the roots of which go back to Gödel's incompleteness theorem, two fundamental limits of objective knowledge in science can be deduced (chapter 10). One such limit is due to the fundamental problems of biological discovery mentioned in part III: the chance hypothesis is fundamentally incapable of proof, while the teleological approach is fundamentally irrefutable. The other limit is related to the ability of natural law to describe reality as a whole: it can never be proved that a given algorithm is the smallest capable of describing a class of natural events in a lawlike way, since the smallest algorithm (in binary representation) is by definition a random sequence, whose irreducibility-because of the randomness theorem-cannot be proved. This is in particular true of the case when an algorithm is given for all classes of natural event. In other words, the completeness of a scientific theory, as would be expressed information-theoretically in the statement of a unique and irreducible algorithm, is fundamentally incapable of proof. It is only refutable by pragmatic means - by the statement of a new and more compact algorithm.

It is to be expected that the concept of natural law introduced on the basis of information theory will have implications for a number of other problems in the philosophy of science. The supposition has already been expressed in chapter 10 that there may be an intimate connection between the algorithmic definition of a natural law and the idea, developed by C.F. von Weizsäcker, of so-called proto-alternatives, as outlined in note 168.

In contradistinction to the chance hypothesis and to the teleological approach, the molecular-Darwinistic approach remains entirely within the confines of traditional scientific explanatory models. However, the molecular-Darwinistic approach is only valid if a Darwinian natural selection really does take place in the realm of inanimate matter, as a consequence of pure physicochemical regularity (see above).

The molecular-Darwinistic approach raises paradigmatically the question of the physical foundation of biology (the so-called problem of reduction). Consequently, the reduction issue is also the central point in the discussion of the possibilities and limitations of the molecular-Darwinistic explanatory model. It is considered in chapter 11, first as a philosophical problem. The selection principle is then shown to be physically justifiable (chapter 12). In chapter 13, the theoretical question of the explanatory structure of the molecular-Darwinistic approach is taken up.

The problem of reduction has long been the object of a vigorous debate between reductionistic and organismic biology. The reductionistic theory of living systems proceeds from the assumption that all life processes run in a strictly causal-genetic way and can be deduced on the sole basis of the material properties of their molecular carriers (the so-called genetic determinism). If limits are set to the physical description of living systems, then these are not fundamental in nature, but are due solely to the complexity of living systems. The organismic theory of living systems, on the other hand, considers the phenomena of life from a holistic point of view and postulates a connection between the level of integration of a living system and its resulting properties. This is above allt o be understood in the sense that an integrated genetic system at any level of organization can possess properties that cannot be explained by a physical and chemical analysis of its subsystems alone. In particular, organismic biology denies the possibility of a physical foundation of the principle of selection, as the naturals election is considered in organismic biology to be an irreducible property of systems that are already living.

Organismic biology is based upon two central theses: (1) the whole is more than the sum of its parts (this leads to the principle of emergence); (2) the whole determines the behavior of its parts (this leads to the principle of macrodetermination).

The universal validity of these two theses, and in particular their relevance to the phenomena of life, are not questioned here. What we do criticize is the claim that reductionistic biology is fundamentally incapable of explaining holistic phenomena, such as the emergence of macrodetermination, within the framework of its analytical method. However, the analysis of concrete examples from physics and chemistry shows that in the branches of science to which the reductionist program attempts to reduce the phenomena of life, the concept of system has all along been a genuine component of explanations. Therefore, the concept of system cannot be used as a criterion for distinguishing between between animate and inanimate matter. On the contrary, the special position allotted to the "system" within organismic biology is beginning to allow a methodological mysticism to creep back into biology, at a time when this had only just begun to be eliminated by the development of molecular biology. It can be shown that a number of barriers to research in biology have been erected merely by the unreflecting adoption of holistic thinking. A principal victim of this is organismic biology itself, which is now confronted with several self-sustained contradictions (chapter 12).

After showing in chapter 11 that the methodological position of organismic biology is already anchored in the reductionistic program, so that it does not present a methodology that goes beyond that of reductionism, it was demonstrated in chapter 12 that one of he central assertions of organismic biology is incorrect: natural selection is not an irreducible property of living systems, but is demonstrably a physically justifiable extremum principle that already appears in inanimate material systems as long as certain material and physical prerequisites are met.

The physical justification of the selection principle provides profound insight into the mechanism of evolutionary processes. It is seen that the Darwinian concept of adaption is related primarily to the functional aspects of the adaption and less to the structural in agreement with the thesis with which we started, according to which the semantic aspect of biological information is objectifiable only via a dynamic criterion of generation of information.

When related to the phenomena of life, the concept of constraints has an expanded meaning - in comparison with its traditional application in physics. In this case it embraces both biological constraints, as laid down in the primary structures of biological macromolecules, and also the physical environmental conditions under which the biological macromolecules operate as carriers of information or of function, respectively. It is primarily the biological constraints that are incorporated into the explanandum in evolutionary explanatory models, while the physical environment is given as an antecedent condition.

The molecular-Darwinistic model possesses one important feature in which it differs from traditional models of physical explanation:while in traditional explanatory models the constraints present contingent quantities, which are imposed upon the explanans as irreducible antecedence conditions, the molecular-Darwinistic model considers the constraints to be a part of the explanandum (chapter 13).

The issue becomes more complicated in consequence of the fact that on the evolutionary level the biological constraints and the system dynamics on which they operate are interdependent. The feedback between the constraints of a system and the system dynamics induced by the constraints is so characteristic for all self-organization processes that this phenomenon can be used to define the term "self-organization" (chapter 13).

In the historical origin of life, the development of self-organizing systems began from unspecific constraints, as they must have been at the end of the phase of chemical evolution, when macromolecules did not yet embody any genetic information in their primary structure. During the period of self-organization, the unspecific constraints were then transformed successively into biological ones, that is, ones encoding genetic information. The biological constraints are thus noncontingent quantities insofar as they represent a limited and, from an evolutionary standpoint, regular selection from an immense number of physically equivalent alternatives.

The characteristic features of the evolutionary optimization process can be illustrated with the help of a simple model (chapter8, figure 16). If all combinatorially possible sequences of a biological macromolecule are used as coordinates in a "sequence space," then the process of the origin of biological information resembles a walk through a multi-dimensional landscape, the profile of whichi s determined by the selection values associated with a value coordinate (value profile). The route taken is subject only to the rule that it always leads from a low peak to a higher one, that is, from a low (local) maximum to a higher (local) maximum.

Thus only the sign of the gradient of the optimization route is determined by natural law, not its detailed path. There are basically two reasons for the indeterminacy of the detailed path. One is that the "direction" taken by the optimization process depends on genetic variation, and this in turn is the result of fundamentally indeterminate genetic mutations. The other is that the structure of the value profile depends upon the individuals taking part in the evolutionary process. Each differential shift in concentration distribution deforms the value profile itself, so that it comes to influence the gradient and the "direction" of the optimization route. Goal and goal-directedness in a process of biological self-organization are indissolubly interconnected and interdependent. On the basis of natural law, therefore, it is possible to predict that biological structures exist, but not what biological structures exist.The structures that are found reflect the historical uniqueness of living systems, and the details of their origin are in principle inaccessible for description in terms of natural law. This means: the origin of biological information can indeed be explained as a general phenomenon, but the concrete content of biological information can not be deduced from the laws of physics and chemistry.

Information and the Origin of Life, pp.168-177