Home > Philosophy > 1997 After Postmodernism Conference > Matsuno (specific 2)

Use of Natural Languages in Modeling Evolutionary Processes

Koichiro Matsuno
Department of BioEngineering
Nagaoka University of Technology
Nagaoka 940-21, Japan


Although formal language has been thought as a powerful tool for modeling any phenomenon in nature, it has a shortcoming of not being evolutionary. Any fundamental predicates appearing in arbitrary formal language are destined to be irreducible by themselves and not to be evolutionary. Such inadequateness of formal language is most keenly discernible when evolutionary processes are modeled.

In contrast, natural languages are competent enough to model evolutionary processes even though the former is no more than an outcome of the latter. Evolutionary competence of natural languages is found in the absence of irreducible fundamental predicates or in the approval of predications of infinite regression. Evolutionary dynamics modeled by a natural language materializes in reducing any dynamic predicate into others in a ceaseless manner. When they are taken to be representations in a natural language, the trio of natural selection, variations and adaptation are no exception in allowing their reducibility into others. The infinite reducibility of those dynamic predicates now yields an inevitable overlapping of their meanings among themselves since the whole vocabulary of any natural language remains finite. Separation between natural selection and variations, for instance, is not available when evolutionary processes are modeled by a natural language.

A perspective unique to modeling evolutionary processes by a natural language is to regard those predicates such as natural selection, variations and adaptation simply as a consequence of some action-oriented activities. The basic action-oriented activities we recognize in evolutionary processes are those that can generate the activities of similar nature in time domain indefinitely, as being in accordance with the occurrence of infinite regression in linguistic domain. A most pronounced action-oriented activity in evolutionary processes is the process for fulfilling the conditions of the conservation of energy, or heterotrophic activity in short.



Use of a formal language as a modeling tool seems to remain indisputable and invincible even for evolutionary processes. A greatest merit of formal languages is found within the symmetry among descriptive subjects involved. There is no privileged descriptive subject in the domain of formal languages thanks to the presence of firm and unshakable structures to rely upon. Any descriptive subject is supposed to describe an object unequivocally, and the author can remain anonymous. What concerns us here is whether the descriptive symmetry of the authorship latent in any formal languages could face up to describing and, if ever possible, modeling evolutionary processes at large.

As a matter of fact, the idea of symmetry is very old and still very powerful. Appreciation of the idea of symmetry is rooted deeply in our aspiration for identifying the invariant nature of the representation of an object of whatever kind, which may interest us either intellectually, artistically, or even aesthetically. Our affection toward the idea of symmetry has not, however, been free from its methodological limitation (Matsuno, 1985).

Observing a symmetry property on the part of the representation of any object requires complete separation between the subject observing the symmetry and the object itself. That is the Cartesian separation between subject and object. Unless an object is completely separated from the subject observing it, the invariant nature of the object would fail because of intrusion of the remaining agential capacity that has been unique exclusively to the subject. Agency as the capacity of distinguishing between before and after the events is generative and accordingly variable. Because of this, symmetry to be observed in the representation of an object in turn makes the object necessarily invariant and deprives it of any agential capacity (Matsuno, 1989). The Cartesian separation and the preservation of symmetry goes hand in hand.

A supreme example pointing the relationship between Cartesian separation and symmetry is found in physics. As assuming the presence of a subject, that is a physicist, who can describe the invariant object as it is, the Cartesian separation yields the representation of an object while assimilating the descriptive operation keeping the object invariant with a symmetry-preserving physical operation keeping certain physical entities invariant. Charges such as electric charges and the structural form of the law of dynamics are those entities that remain invariant through symmetric operations in a physical space (Wigner, 1964).

The Cartesian separation has been proved to be a superb descriptive vehicle for transforming the operation of symmetry-preservation in a physical space into the one in the linguistic domain. Identifiability of symmetric operations in physical space rests upon a descriptive object that can remain invariant during its description. In short, the likelihood of a descriptive object that can remain invariant during its description upholds a symmetric operation to be identified in the corresponding physical space.

So far, so good. Especially, as far as the ubiquity of the idea of symmetry is concerned, we cannot overemphasize the significance of the conceptual kinship between symmetric physical operations to be identified and the Cartesian separation as a methodology.

However, once it is questioned how an invariant descriptive object could come into being, we shall be asked to face the subtle difference between simply accepting the object to be described on the one hand and getting into knowing it on the other. The Cartesian separation and the idea of symmetry for this matter, both dismiss the aspect of getting into knowing the object to be described. The Cartesian separation can remain immune to any likelihood for the Cartesian subject to be informed while describing the object.


Monologue versus Dialogue

Information as a capacity relating the difference between before and after the events is totally absent in the Cartesian separation between subject and object. In other words, the Cartesian description of an object is a sort of monologue in the sense that information about the object the descriptive subject assumes remains invariant during its description. Spatial horizon is infinite to a Cartesian subject involved in monologue. Monologue as a means of describing what remains invariant dispenses with the notion of information in the generative sense. The present close affinity between the Cartesian separation and monologue would however invoke a serious second thoughts once it is noticed that our linguistic exchanges are exclusively in the form of dialogue.

Dialogue is antithetical to the Cartesian separation between subject and object. One can observe that other parties involved in any dialogue can assume agential capacity making the distinction between before and after the events evident. Dialogue is information-generative in the sense that no party has any means for foretelling how other parties would respond. It is only after the events occurred when every party involved in a dialogue comes to know what other parties would say. Contrary to monologue, both spatial and temporal horizons are finite to participatory agents involved in dialogue (Matsuno, 1985, 1989; Rössler, 1987).

Dialogue is in fact ubiquitous in our material world. Of course, our colleagues are constantly involved in a sort of dialogue, for instance, by attending a meeting or a congress, but dialogue is not limited only to ourselves. Any biological organisms including even bacteria are constantly involved in dialogue or in mutual communication through feeding on or being fed on by others. Mutual communication between biological organisms through feeding on and being fed on is a form of dialogue because each party can detect what other parties have done only after the events. The same is true also to molecular communication or interaction proceeding inside a biological organism. Any molecule constituting an organism can detect what other molecules in the same organism are doing only afterward because the communication takes time no matter how small it may be.

Material communication or interaction of whatever kind is in a form of dialogue by letting the communication be locally sequential, instead of being globally simultaneous. Material communication is thus information-generative in the eyes of the participatory agents because of the dichotomy between before and after the events.

At this point, we come to see the contrast between the Cartesian subject to whom information remains invariant and the participatory agents to whom information is generative, while we face the same situation. Utility and versatility of the Cartesian separation remains unquestionable as a legitimate methodology, while the presence of participatory agents to whom information is generative is also indisputable. Symmetry preservation under invariant information is basic to the Cartesian subject, while information generation is intrinsic to the participatory agents. At issue is how to accommodate the conservation of information in the eyes of the Cartesian subject with the generation of information to the participatory agents of subjects.

In this regard, it would be helpful to consider the contrast between the conservation of energy in the eyes of the Cartesian subject, that is a detached and impartial physicist, and the energy gain and loss on the part of energy consumers as the participatory agents. The conservation of energy in the eyes of the Cartesian subject and the generative aspect of energy consumers as evidenced in evolutionary emergence of de novo biological species are completely compatible so long as we distinguish between the underlying descriptive stances. However, in order to maintain the compatibility, each stance is required not to overrun the limit of each competency.

Take, for instance, the principle of the conservation of energy. Although the principle can be derived as related to an entity that could remain invariant while being subject to a certain symmetric operation in a physical space, the choice of the space and the corresponding symmetric operations could remain arbitrary. Needless to say, the temporally translational symmetry of the law of dynamics gives rise to the conservation of energy insofar as charges and masses are imposed and guaranteed as invariant entities. However, it still remains to be seen whether charges and masses could completely be separated from the law of dynamics.

One indication suggesting the likelihood for the hidden coupling between the law of dynamics on the one hand and charges and masses on the other might be a splitting of charges and of masses due to certain symmetry-breakings applied to the original gauge symmetry. What is called to our attention is the relationship between charges and the law of dynamics under the burden of such charges. Physics has been very competent in identifying what all the charges would look like completely independently of the form of the law of dynamics that those charges are subject to. Nevertheless, the idea of charges imposed upon the law of nature remains as an artifact at best, because the symmetric operations maintaining the form of the law of dynamics invariant are kept apart from another symmetric operations that maintain the charges involved invariant. Electric charge cannot be merely a symbol e to be written down by a hand into the equation of motion.

This reflection, however, does not intend to undermine the well-established notion of electric charges or atomic masses in any sense. To the contrary, what is raised at this point is the likelihood that other entities may also become candidates of new charges.

The symmetric operation that underlies the conservation of energy depends upon the nature of charges involved in the dynamics. If energy consumers whose representative cases are biological organisms happen to be emergent charges, the symmetric operations maintaining the law of dynamics in terms of those charges invariant may depend upon the nature of the energy consumers. Maintenance of the global conservation of energy depends upon how each energy consumer such as a biological organism feeds on and is fed on by others locally (Matsuno, 1995). The global symmetry latent in the conservation of energy turns out to be a consequence of the local act of each participatory agent for the sake of the global conservation of energy.

We thus come to recognize that the Cartesian subject whose existence could be invincible at least methodologically, legitimately rests itself upon the notion of invariance, instead of the notion of symmetry. The invariance of an object to the Cartesian subject could be identified as an entity remaining invariant through a symmetric operation only when the space in which the symmetric operation exercises itself and the charges that the operation acts upon are completely separated. Otherwise, the notion of symmetry to the Cartesian subject is merely derivative of the local act of participatory agents for the sake of the principle of invariance on a global scale.

The conservation of information to the Cartesian subject grounded merely upon the invariance of a descriptive object does necessarily render participatory agents to be information consumers involved in gain and loss of information locally for the sake of its global conservation. Energy consumers appeared in the form of biological organisms are in fact a particular instance of information consumers.


Information and Conservation

The relationship between symmetry and information is subtle. First of all, so long as the Cartesian separation between subject and object is employed, information about the object has to be conserved strictly on the methodological ground (Küppers, 1995). The conservation of information applies to a monologue of whatever sort even including the present writing by myself. Descriptive integrity of the intended discourse is just another name of the Cartesian subject and any author intending a monologue happens to be and has to be a Cartesian subject, even if he or she dislikes the very idea.

However, information to be conserved in the eyes of the Cartesian subject cannot perspicuously be represented as an invariant entity because the similar information to the participatory agents is viewed as being generative (Farre, 1994). This disparity between global conservation and local generativity exhibits a distinct contrast to the notion of symmetry, in the latter of which no participatory agents are allowed to intervene. This suggests to us an inverted view such that symmetry is an emergent property of information rather than the other way around (Gunji, 1995).

The conservation of information in the eyes of the Cartesian subject cannot be an entity to be described in a globally consistent manner, because otherwise its local generative capacity would be lost. Once the generative capacity of information on the part of participatory agents is focused, the distinction between information-in-itself and its representation has to be made. The representation of information is destined to be ill-fated because any representation assumes a perfect consistency with its local sub-representations, with the result that the local generative capacity latent in information would necessarily be lost. Nonetheless, if it is still tried to represent information, such attempt has to be local so as to keep the local generative capacity intact. In other words, any local representation of information is necessarily incomplete while admitting the generative capacity of revising its own representations indefinitely due to the involvement of participatory agents.

In contrast, symmetry-in-itself and its representation do not have such an incommensurability as the pair of information-in-itself and its representation have. The idea of symmetry does not rest upon participatory agents. Even if participatory agents are inevitable, the idea can arguably provide a legitimate excuse for how these participatory agents could be subsumed by a single Cartesian subject at least in its outlook. The difference between symmetry and information is on whether participatory agents are involved.

In view of the fact that the involvement of participatory agents is inevitable and indisputable, information is more faithful than symmetry to what is occurring in the actual material world though its complete representation always remains as an impossible dream because of its queer methodology. To the contrary, symmetry assumes its complete representation at the expense that the globally consistent representation is not generative anymore. Information is generative because it prohibits itself from being represented in a globally consistent manner. Information can be associated with symmetry only after the events have been fixed, because there is no room for participatory agents to survive in those events already finished. Exactly for this reason, symmetry is an emergent property of information. Generative aspect of information can be found within the process of symmetry-breaking, and the representation of symmetry-breaking could be possible only locally so as to keep its generative capacity intact.

Information is unique in urging us to talk about it only locally even if it remains invariant as a whole during the discourse. Although discourse sometimes could exercise such a brutal force as to ask for a global description, information has to be exceptional in restricting its description strictly to be local. A global representation of information is possible only through a form of symmetry as an emergent property of the former.

Natural language is for information in the making, while formal language is for symmetry in the product of finished events. An a posteriori formalization of the discourse in a natural language admits infinite spatial horizon to a descriptive subject only after the events, while the participatory agents being responsible for making frozen consistency are acting under finite horizons both in space and in time.

In a nutshell, any monologue within the scheme of formal language by a single author can remain objective thanks to the direct reference to the underlying invariant structure, while at the expense of evolutionary processes. In contrast, any dialogue within the framework of natural language participated by multiple authors is evolutionary because of the interplay among those participatory agents. It is this participatory capacity which makes natural languages be competent enough to cope with modeling evolutionary processes.


Evolutionary processes and information

Information is generative in the eyes of the participatory agents. That implies that information is causative in inducing generative activities in each of the participatory agents. Information to those internal agents is what is to be communicated among themselves, and what is communicated is a redistribution of actions (Leydesdorff, 1995). Moreover, each action expects its reaction as embodied in the law of action and reaction, that is to say, the third law of mechanics (Matsuno, 1989). The present relationship between action and reaction now invites at least two different viewpoints, depending upon whether information may be grasped in monologues of a formal language or dialogues of a natural one.

An external descriptive subject emphasizing a formal language maintains information to be conservative, and regards globally synchronous updating of all the pairs of action and reaction feasible because of the claimed infinite spatial horizon. Classical mechanics just happens to be this case. In contrast, the participatory agents conceived within the practice of a natural language take information generative, and see nothing other than locally asynchronous updating of action and reaction at work because of their finite horizons both in space and time. The present difference of perceptual horizons induces a qualitative difference in the nature of reaction (Matsuno and Salthe, 1995). Reaction produced under finite horizons both in space and time turns out to be action to others, because the reaction carries with itself a certain degree of contingency at any rate. Contingent reaction is nothing other than action.

Participatory agents conceived within a natural language thus constantly behave as active participants. When evolutionary processes are described as dialogues in a natural language, the source of activities can be sought in the participants, whether they are individual organisms or their constituent elements, among others. Needless to say, the trio of natural selection, variations and adaptation, for example, can be modeled in monologues in a formal language, with the consequence of establishing the global synchronization of the three. That is a neo-Darwinian framework guaranteeing synchronization between all the efficient or mechanistic causes on the one hand and their effects on the other. Nonetheless, global synchronization between causes and effects is simply not avbailable to dialogues in a natural language.

What makes evolutionary processes modeled as dialogues in a natural language unique is in the observation that evolutionary activities are sought on the level deeper than that on which natural selection, variations and adaptation could be conceived formally in a consistent manner. As a matter of fact, heterotrophic activity refers to an evolutionarily most pronounced process of updating action and reaction asynchronously. Evolutionary processes are significant in embodying those material processes of updating action and reaction asynchronously. Dialogues in a natural language are exclusively competent in describing and modeling asynchronous updating of those evolutionary participants.



Farre, G. L., 1994. Philosophical reflections on the question of emergence. In: G. E. Lasker & G. L.

Farre, (Eds.), Advances in Synergetics, Vol. I: Systems Research on Emergence, IIAS, Windsor Ontario, pp. 63-67.

Gunji, Y. -P., 1995. Global logic resulting from disequilibration process. BioSystems 35, 33-62.

Küppers, B.-O., 1995. The context-dependent biological information. In: L. Kruger & B.

Falkenburg, (Eds.), Physik, Philosophie und die Einheit der Wissenschaften. Grundlagender exakten Naturwissen schaften, Band 10. Spektrum Akademischer Verlag, Heidelber, in press.

Leydesdorff, L., 1995. Uncertainty and the communication of time. Syst. Res. 11, 31-51.

Matsuno, K., 1985. How can quantum mechanics of material evolution be possible?: symmetry and symmetry-breaking in protobiological evolution. BioSystems 17, 179-192.

Matsuno, K., 1989. Protobiology: Physical Basis of Biology. CRC press, Boca Raton Florida.

Matsuno, K., 1995. Consumer power as the major evolutionary force. J. Theor. Bio. 173, 137-145.

Matsuno, K. & Salthe, S. N., 1995. Global idealism/local materialism. Biol. Philos. 10, 309-337.

Rössler, O. E., 1987. Endophysics. In: J. L. Casti & A. Karlquist, (Eds.), Real Brains, Artificial Minds. North Holland, New York, pp. 25-46.

Wigner, E. P., 1964. Events, laws of nature, and invariance principles. Science 145, 995-999.

[After Post-Modernism Conference. Copyright 1997.]

All contents Copyright 2003 by The Focusing Institute
Email comments to webmaster