Abstract
Implicit in any emergent phenomena is the presence of an observer to whom something unprepared or unexpected would emerge. Accordingly, the problem of emergent phenomena can be dealt with by explicating what sorts of observation would be available and intervene in the dynamics and in its representation. Once whatever dynamics is descriptively represented, the resulting description crystallized in an external record necessarily becomes global in the sense that there exists a point of view claiming a global perspective of the dynamics. Nonetheless, the descriptively global perspective does not imply that the dynamic process of identification proceeding on material grounds would also be global. If dynamic identification is global like in bird's-eye, the state dynamics based upon identifiability of the global state at every moment without utilizing any actual procedure of identification in the empirical sense would follow as in the case of classical and quantum mechanics. There would be no emergent phenomena within the scheme of state dynamics unless exerted upon externally. Identification of state and the resulting state dynamics dismiss the phenomena of being emergent altogether.
In contrast, once it is duly admitted that nothing propagates faster than light does, global identification of state will not become available any more. The dynamic process of identification turns out to be local, implying that there would be no center coordinating the development of the dynamics globally. The dynamics of local detection is thus of emergent character as contrasting the difference between what has been detected and will be detected. There is always something unexpected in the worm's-eye, while everything is seen as has been given in the bird's-eye. Emergent phenomena are possible only to those local observers of material origin that could claim the capacity of only local detection. Still, one requires a scheme of global description to see what is in the worm's-eye.
In order to transform the emergent character latent in the dynamics of local detection into the representation that is globally descriptive, it is required to identify a quantitative figure measuring the extent of the local detection that can remain fixed even if the global description is introduced. A candidate for that quantitative figure is the number of degrees of freedom. Emergent phenomena can be seen as an instance of changing the number of degrees of freedom through their association.
Thermodynamics as a many-to-one temporal mapping supplemented by random fluctuations provides a methodological framework that enables us to see what may look emergent locally in the worm's eye also as being emergent globally in the bird's eye. Emergent phenomena within the scheme of thermodynamics are seen as instances of net information generation accompanied by net energy dissipation.
Recent upheaval of the attention toward complexity and emergence urges us to take a new look at what the underlying dynamics is all about, because both complexity and emergence are the attributes of dynamic process par excellence. Biological evolution as a mode of dynamics seems full of complexity and emergent phenomena (Salthe, 1993). This quick perusal of dynamics leads us to ask ourselves what should be required of the basic framework of dynamics that can cope with complexity and emergence. At issue is the foundation of dynamics, no matter what it may refer to (Matsuno, 1989).
Underlying any description of dynamics is the activity of identifying the object to be described (Matsuno, 1989, 1993). However, descriptive identification unique to the subject who intends to describe whatever object raises its own problem. The activity of forming and identifying an object for description goes beyond what the description can give in the end (Pattee, 1993; Rosen, 1991). Object-oriented activities are more than what the resulting object tells about. Once the capacity of descriptive identification is taken to be an object to be described, an infinite regression would become inevitable such as asking the capacity of identifying the capacity of identifying the object.
In contrast, descriptive realization simply as a consequence of describing the object thus identified can stand on its own if there are no further object-oriented activities to intervene in the description. If the separation between descriptive identification and realization is completed with no remnant of object-oriented activities within the resulting description, one can see the description as a structured organization in which every participating component identifies itself and its place by establishing and referring to the firm and invariant relationship with all the others existing there (Farre, 1994). Each participant's identifying itself and others is by no means an actual identification in the empirical sense, but is already structured in the organization.
As a matter of fact, dynamic identification of material origin, of whatever sort it may be, is local and object-oriented, that is to say, detection of material origin is intrinsically local both in space and in time. Impossibility in identifying what will be detected before it has actually been detected or in structuralizing dynamic identification atemporally lets the participating dynamic detection be necessarily local both in space and in time. The resulting dynamics of local detection inevitably accommodates in itself object-oriented activities that cannot be structuralized. The problem of complexity and emergence within the scheme of the dynamics of local detection will have to be addressed accordingly.
Interestingly enough, however, the similar problem of complexity and emergence can also be dealt with by employing a form of structured dynamics supplemented by external disturbances or contingencies (Ehresmann and Vanbremeersch, 1987; Baas, 1994). The universal objective observer claiming an instantaneous bird's-eye view on the global scale, if available, can regard the consequence of the preceding structured dynamics perturbed by external disturbances as the emergence of a new structured dynamics. Crucial to the theoretical artifact of introducing the universal objective observer is that the capacity of identification on the part of the observer is neither structuralized in the dynamics nor taken to be object-oriented while exhibiting its observational activity. However, once the material basis of the capacity of observing is focused, the conception of the universal objective observer would turn out to be immaterial because of the absence of any material means claiming instantaneous detection or communication on the global scale. Examining material grounds of complexity and emergent phenomena is to place the dynamics of local detection at its inner-most core.
Given a set of available degrees of freedom in motion, for instance, the underlying dynamics comes to face the problem of how each degree of freedom determines its value in time. Noting that dynamic identification of material origin is necessarily local, one is led to observe that the identification remains indefinite about what will be identified, whereas definite about what has already been realized. The temporal asymmetry between the possible yet to come and the actual already established now lets the law of motion be of a one-to-many temporal mapping as contrasting the indefiniteness in the future to the definiteness in the past (Matsuno, 1989).
We have noted that the dynamics of local detection holds the law of motion of a one-to-many tempopal mapping with regard to its constituent degrees of freedom. In contrast, thermodynamics is unique in maintaining the law of motion of a many-to-one mapping type guaranteeing approach toward a thermal equilibrium at the asymptotic time limit. The many-to-one temporal mapping inherent in thermodynamics is, however, of imposed character due to the external observer who limits the available dynamic variables only to the thermodynamic ones. Reduction of an enormous number of degrees of freedom with the original dynamics of local detection to a set of extremely limited ones of the thermodynamic variables necessarily yields a many-to-one temporal mapping with an irreversible characteristic letting each different input transform into a unique common output with the elapse of time. The factors other than those taken up as thermodynamic variables are regarded simply as random forces acting upon the variables and causing fluctuations in the latter. The thermodynamic framework of a many-to-one temporal mapping supplemented by random fluctuations as a means of representation serves simply as a theoretical artifact, while the dynamics of local detection to be represented is of material origin.
Thermodynamics as a many-to-one mapping supplemented by random fluctuations is at best methodological in taking constituent individual degrees of freedom as given. Methodologically, there is allowed no net information generation within the framework of thermodynamics at thermal equilibrium. It is the dynamics of local detection that is responsible for constructing and altering each individual degree of freedom. A possible enhancement of specificity, that is to say, information generation that is of course of emergent character, may be sought within constructing and rearranging individual degrees of freedom to be actualized in the dynamics of local detection.
Thermodynamics at thermal equilibrium would neither enhance its own specificity by itself nor increase the complexity accompanying emergent phenomena insofar as its constituent degrees of freedom remain fixed. The invariant nature of the constituent degrees of freedom provides the thermodynamics with a methodological reference against which whether or not there would be any information generation or enhancement of specificity can be identified. Crucial to the presence of such a methodological reference for identifying whether there could be any emergent phenomena is the nature of each constituent degree of freedom for the thermodynamics especially with regard to whether or not it could remain immutable.
If immutable, the resulting thermodynamics could hold no capacity for generating emergent phenomena. This is because whatever spontaneous emergences of local character, even if excitable as fluctuations, would come to be mutually cancelled with each other in the long run due to the presumed persistence of a thermal equilibrium. Accordingly, emergent properties are to reside within the mutability of each degree of freedom that constitutes a thermodynamics as a consequence. It is the dynamics of local detection, instead of thermodynamics itself, which specifies how individual degrees of freedom could be mutable.
In particular, neither thermodynamics of a many-to-one temporal mapping supplemented by random fluctuations nor mechanics of a one-to-one mapping can address itself to the problem of how emergent phenomena could be possible physically or ontologically. In order to face this problem squarely, one requires a firm ontological ground other than a one-to-one temporal mapping on a microscopic scale or a many-to-one mapping on a thermodynamic scale. As a matter of fact, the one-to-many mapping underlying the dynamics of local detection that we have observed can cope with emergent phenomena. Its methodological competence is rooted in the ontological basis admitting that nothing propagates fater than the speed of light.
If global identification were the case like in the bird's eye, the state on the global scale could be claimed without taking the actual procedure of observation in the empirical sense. On the other hand, if local detection is taken seriously as it should be, the source of being emergent can be sought within the contrast between what has been detected and will be detected. Furthermore, the problem of how to represent the emergent phenomena will need to be focused in its turn, since the representation completed again points to the global situation. The representation is global in the record. What can be in the worm's eye has to be translated in the manner that the bird's eye can grasp.
The problem of emergent phenomena is intrinsically convoluted. If the global representation of a dynamics is taken before it is examined whether the actual dynamic process of identification could be accomplished globally in an instantaneous manner, there would be no emergent phenomena unless exerted upon externally. If the local process of identification is taken before the global representation of the dynamics is tried, on the other hand, it would be required for the representation to coordinate various local perspectives in the manner such that emergent phenomena may be brought about on the global scale as the effect. The number of degrees of freedom and its change through their association or dissociation just fulfill the condition for connecting emergent phenomena in the making in the worm's eye to the similar phenomena in the product in the bird's eye.
References