CLAUS EMMECHE a
aCenter for the Philosophy of Nature and Science Studies, Niels Bohr Institute, Copenhagen, Denmark
In this note some epistemological problems in general theories about living systems are considered; in particular, the question of hidden connections between different areas of experience, such as folk biology and scientific biology, and hidden connections between central concepts of theoretical biology, such as function, semiosis, closure and life.
The aim of this brief note is to consider partly hidden ideas about theoretical biology and its subject matter, living beings, organisms in their ecosystems -- which means beetles, cows, worms, bacteria cells, green algae, and dinosaurs, their history and interactions, their development and evolution, their structure and function, their origin, self-organization, the extinction of individuals as well as species, and the genesis of higher modes of life. In other words, an extremely multifaceted subject. First, however, recall an observation on the fate of general systems theory, which in the 1960s and 1970s had the ambitious goal of synthesizing the general fields of cybernetics, information theory, operation analysis, and specific fields, such as evolutionary theory and thermodynamics. That goal was not achieved and various reasons may be given for the failure, but an important factor might have been a too high level of theoretical generality in accounting for the highly different types of systems included in the ambitions of systems theory . With this in mind, we could ask for the possibility of facing a similar situation with respect to the current trends in systems thinking.
Experiential, experimental, and theoretical biology
We commence with the epistemology of evolutionary systems theory (complex
adaptive systems, developmental systems, self-organizing systems, etc.). This
idea may be called the hidden prototype fallacy. It focuses on the risk of
becoming seduced by our own theoretical creations, blinded by the life of
abstractions, and it asks if we all commit a fallacy of presupposing certain
characteristics of the class of systems under investigation (e.g., evolutionary
systems), even though these are not accounted for by the theoretical apparatus
of our theory. The argumentative rhetoric found in many discussions within the
field of self-organizing systems theory (and others) tends to hide basic
connections between folk biology, theoretical biology (e.g., evolutionary
systems theory), and experimental biology in one of its disciplinary
normal-science forms. We tend to neglect the deep role played by real
biology as a reservoir of experience and knowledge about living systems. I
claim that we should more explicitly cultivate such connections between
experiential, scientific and theoretical biology; but also that we should be
critical about the limits of such connections when they are hidden to explicit
discourse. To explain this idea, I have to give some definitions and
Folk biology: Even though one should be careful about scientism-related[*] connotations of this term (especially if one associates it with the term `folk psychology' within philosophy of mind, where it is occasionally supposed that in some distant future of neuroscience, folk psychological terms can be reductively eliminated when neoroscientific terms take their place), folk biology as used here is simply a name for the phenomenological fact that before anyone may learn scientific biology, we all experience the biology of real creatures, we see animals move by themselves, we see flowers grow and unfold, and we may even experience stories the adults tell about very small or invisible living beings like bacteria, amoebae and algae; creatures that some of us eventually come to see in school, or on television. The later fact should make it clear that the phenomenological experiential folk biology in a modern society is deeply influenced by the products of scientific instrumentation and science-derived notions and ideas about microbial life, cells, genes, molecules and so on. Even though the borders between folk biology and scientific biology are vague in this sense, we do indeed have both kinds of systems of experience, with the important historical asymmetry that the everyday life and language is primary (ontogenetically and historically) to scientific biology.
Prototypes: Prototypes may not be the best notion to use in a 'scientific' context, because it describes a semantic feature of everyday language that, in some concept-fields, there is an instance (the prototype) that is more central in (seemingly) having most of the content-aspects of that notion than other instances. Thus, an apple is a more prototypical instance of the general everyday concept 'fruit' than, say, the nut in the fruiting body of the grass-like sedge plant -- even though botanically, both a equal instances, each with its own particular properties . It is a characteristic feature of most of the notions used within one of the frames of thought or paradigms of theoretical biology (such as evolutionary systems, developmental systems theory, biosemiotics) that they are rather abstract and vague. Abstract in the sense that they denote very general properties of life, and that they can be conceived of at some high or medium level of abstraction from concrete instances; and vague because it is seldom that the definitions of the notions in question entail clear-cut demarcation criteria to decide whether specific instances belong to the system type (or posit the property) in question. Here are some relevant examples of abstract and vague notions (some of which are closely related, almost by family-resemblance, and thus grouped together):
(a) self-organization (or emergence, autopoiesis, autocatalysis)
(b) evolution (or development)
(c) communication (or semiosis, information processing)
(d) living (or feeling, acting, learning)
To the extant that we, within a given paradigmatic frame, can use these terms
in a coherent way, we are also able to decide (within the limits of some
ambiguities) if a concrete specimen of life, or a physical concrete dynamic
system, instantiates one or more of these concepts. E.g., from my understanding
of the original theory of autopoiesis , I happen to be able to
decide that a single E. coli cell is an autopoietic system, whereas my
bicycle, a piece of cake, or my home city are not. I even know that a
multicellular organism may be more a problematic case, it may be a higher order
autopoietic system, but there is the possibility that "the observer is
mistaken" [Ref. 3, p. 108]. From reading selected thoughts about
self-organizing systems (from e.g., Simon, Prigogine, Jantsch, Salthe,
Kauffman, and Gell-Mann), I really know that a tornado, a bamboo plant, and a
city are examples of such systems whereas a watch, a dish, a rock, or a carbon
atom are not; and furthermore, that there seem to be borderline cases such as
the Earth, the solar system, a three-dimensional globulin macromolecule, and a
piece of crystal, all of which may be or may not be (considered as)
self-organizing, depending on the specific conditions. The reason for calling
attention to prototypes (which are normally considered as pertaining to pre- or
non-scientific contexts) is the suspicion that in some of the abstract
theorizing, the base exemplars are not described much better than by everyday
prototypes, and the pool of paradigmatic exemplars  indeed has a
Experiential biology: Now, what do I know more about the bamboo plant (as a concrete plant so to speak) when I can claim it to be a self-organizing system? Only in the technical sense of conceptual property-inheritance (known within programming and Artificial Intelligence) will this add to by knowledge of bamboo biology. However, one could also say that it is not really knowledge about the bamboo but about the abstract property of self-organizing systems, a property that all plants, including the bamboo, happen to share. The point is that in arguing about the theoretical details about evolutionary systems (living self-organizing evolving communicating systems) -- and about how to frame a general theory of such systems -- we take for granted the findings of experimental biology and a lot of (almost) tacit knowledge about what I would like to call experiential biology. This includes the domain mentioned above as folk biology (common, conventional, public, everyday notions of plants and animals) plus the subjective field of our own experiences of what it means to be a growing, feeding organism, a moving feeling animal, a sensitive human being. Thus in my use of the word, experiential biology includes folk biology (which again includes some, more or less non-darwinian, but very popular ideological notions of 'higher' and 'lower' forms of life, and so forth) but first and foremost, experiential biology includes a kind of subjective and qualitative knowledge of the feeling of life, of sentience, of the moods of passive laziness or active engagement, and so on. It is that part of a human umwelt that hardly can be realized by a robot . A passage that illustrates this qualitative aspect of knowledge of complex living systems very well -- especially because of its sympathetic sensuality and the concreteness in its description of self-organization -- is a beautiful quotation from Denis Diderot's Le Rêve de d'Alambert (1769), here quoted from Man's New Dialogue with Nature by Prigogine & Stengers [Ref. 6, p. 81], the passage where Diderot, in an imaginary conversation with d'Alambert, tries to demonstrate the failure of a mechanist explanation:
"What is in this egg? An insensitive mass before the germ is put into it ... How does this mass evolve into a new organization, into sensitivity, into life? Through heat. What will generate heat in it? Motion. What will the successive effects of motion be? Instead of answering me, sit down and let us follow out these effects with our eyes from one moment to the next. First there is a speck which moves about, a thread moving and taking colour, flesh being formed, a beak, wing-tips, eyes, feet coming into view, a yellowish substance unwinds and turns into intestines--and you have a living creature.... Now the wall is breached and the bird emerges, walks, flies, feels pain, runs away, comes back again, complains, suffers, loves, desires, enjoys, it experiences all your affections and does all the things that you do. And you will maintain, with Descartes, that it is an imitating machine pure an simple? Why, even little children will laugh at you, and philosophers will answer that if it is a machine you are one too!"Diderot's appeal to the experiential biology of a chicken in formation is of double interest here: It illustrates the point, that in the discursive context of very theoretical arguments about system types and how to explain them, we make use of more intuitive kinds of knowledge when we examine the merits or the failure of such theories (whether they are mechanist as in Diderot's case or not). The sentient living thing, like the chicken, or like you and me, becomes a prototype of a complex system that has not yet been explained by mechanistic principles. The fact, that even children can tell the difference between a watch or other artificial devices and living beings, is a fact of folk biology; and the capacity to make that distinction, to recognize a system as alive, is constitutive for the very concept of an organism. Furthermore, in other passages in Prigogine and Stengers' book, we can observe the tendency to blur that distinction, to hide the pheno-ontological difference between the prototypes of eddies, plants, and animals, by emphasizing that they all are 'dissipative structures'. This hints at a general point:
Is our concept of an organism a closed one?
Having considered a peculiar epistemological aspect of the abstract nature of
the new complex-systems theories and their relation to concrete instances of
known species of life, the focus can now be directed at a related though more
intrinsic-theoretical question about the attempts to characterise complex
Theoretical biology since Kant has invented several seemingly different but very general ideas about what constitute a living system, for example, metabolism, self-reproduction, evolution, hereditary information, code-duality, autopoietic semantic and autocatalytic closure, emergence, and functionality of parts in relation to a whole. I will now propose the apparently odd thought (1) that these notions refer to the very same feature (or property cluster), and furthermore, (2) that these general ideas are crucially dependent on the triumphant development of experimental molecular biology in the twentieth century.
As a caveat, it must be noted, that although a common idea or an everyday concept may easily be expressed in various natural languages such as Danish or English, it is not so evident that the same theoretical idea can be expressed in distinct theoretical settings. Several arguments exist against such a claim, for instance the notion of incommensurability between paradigms that allows only partial translation of concepts from one paradigm to a succeeding one, so the theoretical meaning of a term is not well preserved in the process of translation. Nevertheless, it is still possible to imagine, in a vague sense of sameness, that some fundamental characteristics of complex living systems, characterized by distinct conceptual tools, somehow refer to the same idea of what constitutes the most distinctive feature of a living system.
Another way to state this idea is to say that distinct theoretical perspectives abstract certain more or less crucial special features of 'real living systems', or (to state the same proposition in a less naive-realist mode) of some prototypical idea of living systems. We may posit a common deep structure of these prototypes (corresponding to a common unitary whole of characteristic processual properties of living systems) that we cannot see at the surface level of everyday concepts, but which nevertheless is constitutive of life as a generic phenomenon. Each specific theoretical paradigm then abstracts certain aspects out of this deep level cluster of characteristics, e.g., 'metabolism', 'the possession of a genetic code', 'autonomy', 'interaction with environment', 'interplay of self-organization and natural selection', or whatever. Furthermore, this abstraction process takes place in part by the hidden reference to certain system prototypes as explained above, e.g., the BZ-reaction (or Turing patterns) as a prototype for emergence or developmental order in embryogenesis, or the metabolism of a bacterial cell as a prototype for the autonomy (or autopoiesis) of life as such.
In this note I will only consider a sub-set of about four general notions related to the study of living systems:
1 Function as an explanatory tool of experimental biology and a philosophical argument for the autonomy of biology (or an embarrassment for some philosophers of science);
2 Emergence as an ontological notion covering all levels of organization from the physical to the mental, and considered to be an especially important aspect of understanding the origin of life because this process created special, new, unpredictable properties such as self-reproduction and evolution by natural selection;
3 Semantic closure and the concept of dual modes of complex systems in the theory of H. H. Pattee;
4 Biosemiosis as the defining feature of life, where biosemiosis means a sign interpretation process in living organisms .
To jump to the conclusion of this comparative analysis: It is conjectured that
biosemiosis presupposes functionality, that functionality is only possible
under a closure of operations, and that this closure is an emergent phenomenon
of a semiotic character. (If this is so, a synthesis is needed, and also an
Thus, the null hypothesis to test, so to speak, alternative to the conclusion just mentioned, is that these four notions of life are either irreducibly incompatible or express four independent characteristics of life. We go step by step:
(i) Biosemiosis presupposes functionality. The first step is to assume that we have defined life in a semiotic way in some sense [7, 14].
To explain how, let me combine Peirce's notion of a sign with Bateson's notion of information [cf., Ref. 15 and 25] to emphasize the relational character of a biosemiotic process: Life entails semiosis as the action of signs, where a sign is a first, i.e., a representamen that stands (by a code or a habit) in such a relation to a second, its object, so as to determine a third, its interpretant, to take the same relation to that object (that the representamen takes) and thereby effecting that interpretant so that this effect is significant (potentially or actual) to that interpretant's interpreter organism, in the sense that it is a difference that makes a difference to the interpreter. The interpreter must be an organism, a part of an organism, or an organism-like entity, and the effects on that organism's parts, to be significant (i.e., `to make a difference'), cannot be merely physical, because by definition, the difference, if any, they make, is of potential or actual purport or relevance to the organism in question, which means that they concern the organism's chances of finding food or other sources of energy, or that they ultimately concern its chances of surviving and reproducing.
Admittedly, this is a more restricted sense of sign action that connects the semiotic interpretant more closely to a material interpreter-organism (than Peirce's sense of sign action; or than that found in Ref. 16; cf. Ref. 17). Nevertheless, it shows the relation between, the parts of the semiotic process, and the functionality of the organism as a whole in relation to that organism's biophysical parts: Were there no organisms -- or no functional parts of the organism that contributes to maintain the organism as a whole (where each part, as it were [and as it evolutionary is], exists for the purpose of the whole, as Kant pointed out) -- there would be no action of signs. In this sense, sign function and biofunctionality of organisms are intrinsically related. (Of course, on higher levels of organization, anthroposemiotic processes may display a semiotic functionality that only indirectly presupposes biofunctionality; for example, the Internet as a growing semiotic web functions to connect many computers in a huge virtual library, that may be described linguistically or sociologically, and only ontologically presupposes the biofunctionality of the brains of their human designers and users). The hidden prototype of a basic biosemiotic system is the simple prokaryotic cell.
(ii) Functionality is only possible under a closure of operations. In this note I cannot comment upon the body of philosophical research done on the notion of function in biology (but see Ref. 18), and I shall simply assert that basically, the notion of function in biology is the teleological (and mereological) notion of "a part existing for the good of the whole", or "having the purpose of" doing something in relation to the whole. This is the case disregarding whether the whole is intentional or not, and disregarding whether or not the notions of purpose, and `the good of' (and the related complex of function-ideas) in principle, theoretically, may be reduced to mechanistic causal explanations in a historical setting, e.g., in a neo-Darwinian setting. There are serious reasons for assuming such a reduction as impossible, even in principle, but I do not go into that here.
Instead, I will claim that the reason why functionality is only possible under a closure of operations is astonishing simple (if I have fooled myself here, I hope to learn how). Only when the causal chain from one part to the next closes or feeds back in a closed loop -- at once a feed-back on the level of parts and an emergent function defined (as mentioned) as a part-whole relation -- can we talk about a genuine function. In other words: It is because function is the function of a part that works effectively to produce (part-part efficient causation) influences on other parts within the same whole (the same form, the organism's) -- where each part is constrained by the same whole (formal causation) -- the total of parts interacting under these constraints in a coherent emergent pattern is the whole organism, whose maintenance (final causation) as form is the goal of each part. Here, final causation -- i.e., the dual process of downward constraints (formal cause) on the behaviour of the parts and the emergent pattern of the parts forming a functioning organism (final causation), which is made of parts (material causation) -- is the causation of a physical part within a biological whole being committed to a specific role in the internal organization of that whole, thus the internal ascription (de re) of a role to the part is the emergence of that part's function. Consider a newly transcribed polypeptide chain in a cell before it folds (by self-assembly or with the help of protein chaperones) into an enzyme: This physical thing has not yet an actual function, only a potential one as a finished part, that is, an enzyme. As a physical thing, the enzyme is just a complex molecule; as a biochemical thing, the enzyme is a functional part of the cell's metabolism which, in addition to the membrane and the DNA code, makes up the very cell! Function can only be cell-organized, so to speak, not simply 'physically self-organized'. (Again we can observe: The hidden prototype is [in the twentieth century] the cell, or [in the twentieth and nineteenth century] the organism).
(iii) Closure is an emergent phenomenon. This is in part implied in the previous development: Only by analysing a system in terms of minimally a higher and a lower level (a whole with some parts) can one identify functionality as based on the causal closure of operations; this functional causality being emergent (as defined in Ref. 19) upon the local part-to-part interactions of the individual parts (efficient causality) within that whole. A further comment: Closure has been used as a predicate in quite different circumstances (see other contributions to this volume), and denote quite different concepts. Here it is used and defined in the biological realm, and is not merely informational, or organizational, but also material and energetic, and thus biologic closure is never perfect (compare Ref. 20).
(iv) Closure in biosystems is of a semiotic character. As an imperfect emergent phenomenon, the closure of operations between parts in organisms is producer of differences that makes differences to the parts as well as to the whole organism. Thus, endosemiotic sign links can be analysed as causal links between functioning parts that regulate the entire organic machinery of the body. But, one could object, by what necessity should an organic complex device with the property of emergent functionality be of a semiotic character? This is the deepest and most difficult question and I shall only sketch two possible answers, one of metaphysical necessity, the other of a fact-like or law-like necessity: (1) the Peircean (metaphysical) answer: Wherever Nature takes on so complex habits that allow for the existence of living feeling, the intensity of mind has grown high enough to generate the action of thirdness characteristic of genuine triadic sign action [16, 21]. (As few scientists are familiar with Peirce's philosophy, few will accept this answer I'm afraid). (2) the von Neumann-Schrödinger answer (termed so to emphasize that it is a reconstruction of their original arguments combined with contemporary molecular biology): The minimal complexity of a system which does not spontaneously degenerate but complexifies further by open-ended evolution [Ref. 22, p. 80] is exactly the first self-organizing system (`feeding on negative entropy') with an internal semiotic structure, an internal 'code-script' . The endosemiotic nature of the code is the fact that it (biochemically) embodies an internally defined mapping from a nucleotide sequence space to a protein sequence space within the system. Were there no code (and memory) of this kind, there would be not enough biochemical specificity ('information') in the closed structure of reactions, and the system would fall below the threshold and degenerate. This may be related to unknown 'laws of complexity'. Due to lack of present knowledge about primitive kinds of metabolisms (covering the continuum "no life--primordial life--life"), this answer may be read as a not yet proven hypothesis open to scientific investigation.
To sum up: biosemiosis presupposes functionality, and functionality is only possible under a closure of operations. The prototype organism at this level of research is a single cell, its parts forms the endosemiotic network, and the membrane parts make that organism receptive also to changes in the ambient reservoir of significant influences. This closure is an emergent phenomenon of a semiotic character, and as a closure, it is only partial, imperfect, relatively open. Therefore we can conclude (a) Synthesis is needed. (b) Further epistemological clarification of these concepts is needed too. (c) A null hypothesis -- that the four notions of life, they are, biosemiosis, functionality, emergence, and closure, express four independent characteristics of life -- has been refuted.
I'll express my thanks for inspiration to contributors of the evolutionary systems volume  as well as the participants in the closure workshop in Ghent in May 1999. I claim no special novelty of the ideas expressed in this note and take the sole responsibility for its opaqueness and incompleteness.
[*] [note added in proof]: I mean scientism in the usual sense of "these doctrines, outlooks and associated forms of social practice extending to science an authority beyond its legitimate scope" (from the entry on scientism in: W.F. Bynum, E.J. Browne & Roy Porter, eds. (1981): Dictionary of the History of Science. Macmillan Press, London.)
1. LILIENFELD, R. 1978. The Rise of Systems Theory. Wiley. New York.
2. RORSCH, E. B. AND C. B. MERVIS. 1975. Family resemblances: studies in the internal structure of categories. Cognitive Psychology 7: 573-605.
3. MATURANA, H. R. & F. J. VARELA. 1980. Autopoiesis and Cognition. D. Reidel Publishing Company. Dordrecht, Holland.
4. KUHN, T. S. 1962. The Structure of Scientific Revolutions [2nd. ed. 1970]. The University of Chicago Press. Chicago.
5. EMMECHE, C. (submittet). Does a robot have an Umwelt? Reflections on the qualitative biosemiotics of Jakob von Uexküll. Semiotica .
6. PRIGOGINE, I. AND I. STENGERS. 1984. Order Out of Chaos: Man's New Dialogue with Nature. Bantam Books.Toronto.
7. HOFFMEYER, J. 1996. Signs of Meaning in the Universe. Indiana University Press.Bloomington.
8. HOFFMEYER, J. & C. EMMECHE. 1991. Code-Duality and the Semiotics of Nature. In On Semiotic Modeling. M. ANDERSON & F. MERRELL, Eds.: 117-166. Mouton de Gruyter. Berlin.
9. PATTEE, H.H. 1977. Dynamic and Linguistic Modes of Complex Systems. Int. J. General Systems 3: 259-266.
10. EMMECHE, C. (1999). The Sarkar challenge to biosemiotics: Is there any information in a cell? Semiotica 127 (1/4): 273-294.
11. GELL-MANN, M. 1994. The Quark and the Jaguar. W.H. Freeman. New York.
12. McSHEA, D. W. 1991. Complexity and evolution: What everybody knows. Biology and Philosophy 6 (3): 303-324.
13. EMMECHE, C. 1994. The Garden in the Machine. The emerging science of artificial life. Princeton University Press. Princeton.
14. EMMECHE, C. 1998. Defining life as a semiotic phenomenon. Cybernetics & Human Knowing 5(1): 3-17.
15. EMMECHE, C. 1990. Det biologiske Informationsbegreb. Kimære. Århus [Ph.D. dissertation from 1989, in Danish].
16. MERRELL, F. 1996: Signs Grow: Semiosis and Life Processes. University of Toronto Press. Toronto.
17. EMMECHE, C. (in press). Transdisciplinarity, theory-zapping, and the growth of knowledge. Semiotica. (review of Ref. 16).
18. ALLEN, C., BEKOFF, M. AND LAUDER, G., eds. 1998. Nature's Purposes. Analyses of function and design in biology. The MIT Press. Cambridge, Mass.
19. BAAS, N.A. AND EMMECHE, C. 1997. On emergence and explanation. Intellectica 25 (2): 67-83.
20. LEMKE, J. 2000. Opening up closure: semiotics across scales. Ann. N.Y. Acad. Sci. 901: [this volume].
21. PEIRCE, C. S. 1931-1958. Collected Papers of Charles Sanders Peirce. Vol.1-8. C. Hartshorne, P. Weiss & A. Burks, Eds. Harvard University Press. Cambridge, Mass.
22. VON NEUMANN, J. 1966. Theory of Self-Reproducing Automata. Edited and completed by A. W. Burks. University of Illinois Press. Urbana.
23. SCHRÖDINGER, E. 1944. What is Life? The Physical Aspect of the Living Cell. (Reprint 1962). Cambridge University Press. Cambridge.
24. VAN DE VIJVER, G., S. N. SALTHE & M. DELPOS, Eds. 1998. Evolutionary Systems. Biological and epistemological perspectives on selection and self-organization. Kluwer: Dordrecht.
25. BATESON, G. 1972. Steps to an Ecology of Mind. Ballantine Books. New York.