Complexity and social systems David Manuel-Navarrete Approaches and Implications of using C

Approaches and Implications of using Complexity Theory for

dealing with Social Systems

David Manuel-Navarrete.

PhD candidate. Department of Geography. University of Waterloo. Email: david.manuel@campus.uab.es

Sociology can only describe society in society…It is a science of the social system and a social system of science. To make matters even more complex, as a science and, as a social system, sociology is also an internal observer of whatever system it participates in.

(Luhmann, 1994: 132-133)

ABSTRACT

There exist an increasing awareness within the scientific community for the need to deal with the complex dimension of social systems. This paper examines three approaches to incorporate complexity theory into the practice of social sciences. The first approach consists of supplementing the modernist program with chaos theory. The second one proposes a metaphoric application of complexity theory to describe social systems. The third approach is based on Post-Normal Science. Both the first and the second try to fit complexity theory into the paradigms used up to now in social sciences. Although this exercise can provide some interesting methods for understanding of social systems, it is argued that a fundamental change in the way Western society conceives science is necessary. Complexity analysis (or synthesis) should not only consist of adding more or different syntactical rules to the mathematical formal systems used to model the causal relations perceived in the outside world. Rather, it should imply a generalization of the scientific formalisms in order to include semantic relations, subjectivity, and context dependency. Even more, this generalization should go as far as to include into the umbrella of science a plurality of systems of knowledge in order to better understand the multidimensionality of social systems. The legitimation of a broader spectrum of formal systems of representation and communication of reality, will affect profoundly the collective and individual way in which Western societies perceive the world, and the very evolution of human beings as species.

Key-words: Scial systems, cmplexity, chas, pst-nrmal science, systems f knowledge.

Introduction: Why does complexity sound useful to social sciences?

The origin and development of Western Science has been dominated by modeling simple causal relations operating in physical systems (Back, 1997). As a result, social science, dealing with the complex two-way interactions of society, has until now been forced to use logical and mathematical instruments originally designed to deal with experimentation in hugely simpler systems (Eve et al., 1997).

1

Complexity and social systems David Manuel-Navarrete

Social scientist sought rescue in probabilistic models and developed statistics. Probability, originally describing the bases of reasonable judgment, became an aspect of the laws of nature. The normal distribution around a mean that occurred in so many conditions was seen as part of the living and social world. Moreover, experimental data included measurement and observation errors subject to the same probabilistic laws (Back, 1997). Other formalisms, like the comparative method, also emerged as a substitute for experimentation (when the number of cases is too small to permit statistical manipulation.

Although statistics allows the use of numbers to describe social systems, it gives a severely limited picture of social systems. Statistical inference is not about causality. It basically consisting of seeing which mathematical form fit available historical and longitudinal data sets. Although statistics was conceived as the study of randomness, even it must concede the existence of unexplained residual behavior. This is defined as noise, or variation in the dependent variable which is left over after independent variables have explained all they can explain. Variation in a given variable, then, can be divided into that which is explainable and which is not, and the later is not particularly interesting. In summary, statistics only serves to confirm or disconfirm commonsense hypotheses rather than to generate truly new ideas, perspectives, and knowledge (Marion, 1999).

Several myths have arisen with the intention of providing social sciences with the category of a formal system able to represent causal relations (Back, 1997). Field theory, sociometry, information theory, game theory, catastrophe theory, and fuzzy sets are some examples of efforts to model the behavior of social systems using the language of mathematics. These are representations, schemata, abstractions, simplifications, which does not conform completely to common experience.

Chaos and complexity theory is seen by some as the culmination of this progressive accumulation of new myths that create a new mathematics appropriate for dealing with social phenomena (Eve et al., 1997). There exist the hope that this new mathematical language will mirror the world as human beings experience it and even some of our reactions to it. This long awaited event becomes realizable through the simple recognition of complexity as a primary variable. Departing from this origin, the task of science can be seen as the description of complexity in systems and its changes, both empirically and mathematically. Going further on this direction, the analytical functions of traditional mathematics may be supplemented by fractal functions that represent the actual world with its irregularities. It looks like a step in the progression of social science, integrating the ideas of its predecessors in a new synthesis. However, complexity theory shows that it is not sufficient to have enough valid and reliable data, and correct methods of analysis for having a definite mathematical solution to a given problem. Put in another way, for many issues there is no sure way of finding a single and definite answer- no matter how well the researcher has proceeded (Eve et al., 1997).

2

Complexity and social systems David Manuel-Navarrete

In any case, the implications that will follow to the development of complexity theory for dealing with social systems are far from being a simple question. For example, Bausch (2001) categorizes the social aspect of over 30 major systemic theorist as falling into five broadly thematic areas: designing social systems, the structure of the social world, communication, cognition, and epistemology. This paper identifies three main approaches found in the recent literature on complexity theory applied to social systems: (1) the Supplementing the Modernist Program Approach, (2) the Metaphorical-Analytical Approach, and (3) Post-normal Science Approach.

Approaches to use complexity theory for the analysis/synthesis of social systems

It is the opinion of the author that the excessive reliability of social sciences on the paradigms and methods employed in physics have provoked a misrepresentation of what a social system is and what aspects or elements it includes. Before I describe the three approaches identified, I will comment on some aspects of social systems which have been poorly tackled in social sciences because of the “handicaps” imposed because of the kind of formal system applied:

1) In physics, different objects of the same type behave quite the same. The details of context or history in which the object finds itself does not usually determine its behavior. This is not an appropriate requirement for social sciences in which contextual details are a major factor. Complexity brings context to the fore. In human affairs, it is beginning to look as if history and tradition, even individual acts and decisions, are far more powerful determinants of how a society is organized than the economic and political “forces” that nineteenth-century social theory reduced to social laws.

2) In social systems cause and effect are not tied in such a way that individuals has no room for self-determination, or to affect the larger society. Power and knowledge are two aspects of the same process. Social systems carry information about themselves and their environments, and are able to act on such information (Marion, 1999).

3) It is argued that social systems are different to natural systems because in the former the observer forms part of the very system. It is arguable whether human beings are not in a similar way embedded and form part of ecosystems, but the relation observer-observed can be identified differently in both cases.

4) The question of agency has been broadly debated in sociology and political science. In human societies the origin of perturbations may be internal to society itself. In human systems, perturbations of far from equilibrium conditions can originate in the values and actions of humans themselves. This raises the distinction between teleology and causality. Teleological action requires a sort of control of success, which becomes the foundation of the interaction between subject and object. Goal-oriented or purposeful action is thus held together through the unity of consciousness. It effects a differentiation of the Ego from the object-world and constitutes a conscious interweaving of subjective energies with objective existence.

3

Complexity and social systems David Manuel-Navarrete

5) Consciousness is the next missing piece in the traditional social science puzzle. The requirement for unconsciousness refers to the difficulties posed for any experiments involving human beings by their human capacity of understand what is going on, attach meanings to it, and act according to their own purposes and meanings.

6) Social systems are products of not only physical aspects, and living dynamics, but also of symbolic communication. Consequently, the internal structure of a social system is typically far more ordered that is that of a purely physical system.

The use of complexity theory could help to incorporate some of these missing elements in the analysis/synthesis of social systems. The three approaches identified in this paper can be seen as ordered from less general/more specific to more general/less specific (see Rosen 1991). In the three cases, having complexity into account implies a generalization of the traditional scientific formalisms. Human beings respond to symbols (syntactic aspect of languages), such as words, ideas, concepts, opinions, emotions, projections, and beliefs. We join social movements because of abstract beliefs, certain words can galvanize us to group action, we join discussion groups to share ideas, marry because of love, create armies because of fear, and socialize because of friendship (Marion, 1999). We assign meaning to things that have no physical substance. These mental constructs interact with reality in complex ways and catalyze us to create complex webs of alliances that we call social structure. Is mathematics or complexity theory a suitable formalism for describing all these aspects of reality?

The study of social systems implies the use of systems of symbols suitable for describing or attaching meaning to the inner world processes and dynamics of individuals. Maybe the mathematical symbols are not the most appropriate to develop this task. Social sciences have imported lots of symbolisms used in science for describing the material world. This has favored the study of the material outer side of human beings, while societies have been depicted by the aggregated behavior of rational individuals. This material bias could change dramatically with the advancement of complexity.

First Approach: Supplementing the Modernist Program with chaos theory

According to this approach, applying complex theory consists basically of modifying the formal systems used to describe the outer world incorporating the mathematics of chaos. Thus, theories of chaos and complexity while recognizing some limitations of the modernist program provide the mean to ground social sciences within the “scientific” tradition. By correcting inadequacies in our scientific paradigm, we may appropriately and fruitfully continue to do “science” (Price, 1997). The basic assumption underling is to consider mathematics as isomorphic with the world as it is. Chaos and complexity theory provide new rules to enrich the old formalisms. We just need to introduce some “adjustments” for explaining the dynamics of certain phenomena. Chaos theory suggest that simple events generate behaviors so complex that one is tempted to call them

4

Complexity and social systems David Manuel-Navarrete

random, yet they are entirely deterministic and can be modeled with simple mathematical equations. In that way we have reached finally a real and legitimate “Social Science”.

In order to guarantee a full application of chaos theory, human systems must fulfill three conditions: (1) non-linearity, (2) iteration, and (3) sensitivity to initial conditions. The isomorphism is granted with the following arguments:

(1) Complexities of human behavior are sufficient for supplying any degree of non-linearity. There are linear predictable phenomena in social systems, but there are also many nonlinear processes.

(2) Social interactions are frequently iterative in nature. Values of the so-called explanatory social variables are influenced, at some point in time, by that we wish to explain.

(3) Prediction is abandoned, but we still can exert a certain degree of control through the description of possible trajectories and their associated strange attractors. This search of trajectories descriptions compels to strive for more precise measurements, and the collection of more accurate historical and longitudinal data.

The main shift consists of adding the notion of dynamic order (deterministic chaos) to the traditional descriptions based on premises of static order. It is argued that periodic and point attractor stability has blinded us to alternative perspectives of stability. Traditionally, a phenomena is either repetitive and stable, or random and without pattern. Non-linearity teaches us that randomness and unpredictability are givens, and that they can build order, albeit a rather complex one (Marion, 1999). As in statistics, this approach states that there is patterns to many seemingly unpredictable events. The problem is that the statistic analysis has been largely restricted to the study of static phenomena. There is much to be gained from studying the dynamics or trajectories of behavior.

Thus, deterministic chaos permits the continuation of the modernist program. Complexity is the modernist antidote against the post-modernist threat. The trick consists of finding order within chaos. In fact, chaos is identified as a crucial element in creating order. In the shift from linear reductionistic science to non-linear and emergent science. Some additional rules must be considered in the mathematical and conceptual models, but all the rest remains the same: (1) science still represents a privileged truth that is to be established through empirical investigation, (2) research is organized in tightly closed disciplinary areas, (3) science seeks to identify global properties (i.e. natural laws) (Price, 1997), and (4) the kind of formalisms applied force the use of numbers as the only way to describe what is real. The problems of reification and adequacy of measurements are considered solvable limitations. Generally speaking, what it is measurable is (quite freely) identified with what is real.

The central idea in the case of social science is to apply the quantitative methods or formalisms used in complexity/chaos in the physical sciences to the kind of quantitative descriptions of the social world that social scientists have available to them. Complexity-based research can be adapted to social scientific purposes, and the existing tools of social scientific research can be used as part of a complex program (Byrne, 1998). Social

5

Complexity and social systems David Manuel-Navarrete

systems are seen as consisting of a set of highly interdependent variables evolving over time. Rules can be specified to describe how the system changes from one state to another with respect to time. Now, the time axis is not Newtonian and continuous, but rather it represents transitions (Byrne, 1998). Time is irreducible, irreversible, and asymmetrical (Eve et al., 1997).

In addition to what it has been commented up to now, there are three basic conceptual contributions of applying a chaos/complexity analysis: (1) emergence, (2) phase transition, and (3) dissipation.

The incorporation of emergent processes in the analysis means to look for explanations of discontinuous and fundamental changes in the character of the social system as a whole. Once we express this process mathematically, emergence becomes an empirical reality. It is considered that social structures are actually composed of emergent properties. These are very simple rules for individual interactions which create chaotic formal systems (implying iterations of equations with feedbacks). Changes in those simple rules about how individuals interact with one another socially, politically, or economically might result in a completely different social structure after a few million cycles of interaction (Eve et al., 1997).

The second key-element of analysis is formalized under the concept of ‘strange attractors’. Mathematically, strange attractors “come to existence” as the result of allowing operating the feedback terms of the equation. They are the fractal form embedded in any nonlinear feedback process. Any of the millions of dots that compose a strange attractor represents a single solution to an iterative equation involving feedback (when chaos is present). In such a way, deterministic chaos/complexity theory constitutes the new vehicle for continuing the main endeavor of science, which is to find the equations describing the patterns of “reality”. In this exercise there is implied a certain notion of “pseudo-predictability”; we do not know what will happen, but we know there are only a set of alternatives (strange attractors) greater than one but less than too many (Byrne, 1998). An additional problem to prediction is imposed by the existence of bifurcation points, which determine trajectories leading to different strange attractors (Byrne, 1998). This situation of “soft predictability” forces the use of looser criteria when establishing modeling relations with natural or social systems. In that way, the formalization becomes more general (in the sense of Rosen) than the linear ones used within traditional science.

The third element, dissipative structures, is treated as something close to chaotic dynamics. Mathematically it means that parameters are determined not only by externalities, but also by other internal components inside the model (Lee, 1997). The example of the sand pile is used to illustrate the implications of considering social systems as self-organized structures: in spite of millions of chaotic events (the sand poured) that afflict individuals within society, the social structure itself (the sand pile) remains largely unchanged.

6

Complexity and social systems David Manuel-Navarrete

Finally, it is important to point out how this approach seeks legitimacy in the use of computational strategies that were not available a few years ago: “Now we have not only the technology- the computer, capable of endless, fast, accurate iterations of operations involving mutually dependent elements- but also the theoretical machinery of fractal mathematics and chaos science. We can now create facsimiles of reality by successive tweaking of the variables and the connections among them. This reverses the top-down theory to phenomena approach. Not just describing the universe, but creating possible universes and then selecting the one closest to the actuality” (Eve et al., 1997: XXV).

Second approach: Metaphoric-analytical application of complexity for describing social systems

As in the case of the first approach, the Metaphoric-analytic approach proposes a “tool-kit” to solve some drawbacks, and allow the continuation, of the modernist program. This modification could be described as a reconstruction of the modernist program (in contrast to the deconstructive approach of postmodernism). A new philosophical standpoint, Realism, emerges from the tension between modernism and post-modernism. Realism is seen as a philosophical ontology which can be complemented with complexity as a scientific ontology (Byrne, 1998). This indicates clearly that science is considered as the truly way of representing reality and not just a language or system of knowledge among others.

In contrast with the first approach, now complexity and self-organization are considered as more important theoretical concepts than chaos itself. We can build models of social systems as self-organizing systems in which the mathematics of chaos are useful to describe certain social dynamics (i.e. to describe some patterns of behavior). Complexity is seen as a middle way between order and chaos, and it is argued that social systems dynamics are usually complex rather than chaotic or stable. In social systems we usually deal with very different sorts of data (many units, few observations versus many observations of single units over a long time period), and therefore we should focus more in complex transformation of state (evolutionary development) rather than on chaos. Change in self-organized systems is adaptive; that is, relationships between components at one level change in response to changes taking place in other levels. Adaptive systems are both changing and stable.

The “modification” of the linear formal premise for modeling reality is more radical in this approach than it was in the previous one. Here, it is promoted the use of non-mathematical categories for representing breakdowns of linearity, even though the production rules will be still essentially of a mathematical character. Complexity helps to legitimate qualitative approaches. Nicolis remarks on the impossibility of a full quantitative understanding of complex phenomena (Nicolis, 1995). Under this point of view, the disagreements between positivists and adherents to the strong interpretative program of qualitative social science are irrelevant and pointless (Byrne, 1998). Mathematics, which depends on underlying continuity is not easily isomorphic with qualitative distinctions, but it can still be used as analogy for the world as it is (Byrne,

7

Complexity and social systems David Manuel-Navarrete

1998) (i.e.: strange attractors is not only a matter of a formula with solutions, but the concept is used for analogies and for understanding phenomena from a different perspective or view).

One of the most obvious mathematical metaphors for social phenomena according to this approach is the use of strange attractors for representing different state of the system (Eve et al., 1997). Here, the idea of attractors is based on resonance or sync-ness: Individual particles (people) interact with one another, and their behavior correlate as a result. The system crosses over an invisible boundary and the landscape of attractors alters dramatically. As in the first approach, such changes are called bifurcations. These “metaphorical attractors” can be detected through techniques such as cluster analysis of data sets (Byrne, 1998). The justification for the analogy of strange attractors with social system is summarized in the following points:

• Attractors are stable, but (as social systems) their trajectory never repeats itself. They have the capacity to change and adapt.

• Both can grow or shrink to encompass a broader or a narrower range of behaviors, can alter their appearance, and convert to a completely different state or even fad away.

• They can learn and carry information about their past, anticipate the future, and reproduce.

Such metaphorical comparisons do not prove that social behavior are strange attractors, but if social systems looks like attractors and behave like attractors, why do not see if attractors will tell us something about social systems.

Implicitly, there the idea of discovering the “thermodynamic state variables” of social systems, and then try to find the equations, through cluster analysis, which relate these variables in general, or for specific situations: “What matters is not the individual trajectory of social atoms, but rather the changing characteristics of the complex social order within which those trajectories occur. We have to understand how the micro is aggregated into something beyond the sum of its parts, to understand society as constituted by sets of attractors within the range of possible condition spaces, and to understand how changes in controlling variables for the whole system can come to reconstitute the form of that attractor set ” (Byrne, 1998: 71).

The definition of time is similar to the one given in the previous approach, but now the evolutionary dimension (accounting for developmental and phase-shift processes, and fundamentally historic changes) receives more attention than the continuous (although non-linear) conception of time (Byrne, 1998).

It is important to mention that the idea of promoting the application of chaos and complexity concepts metaphorically does not exclude the development of more analytical applications. In fact, the role of numerical data is still crucial in any representation of the real world, and, furthermore, it is argued that the very data can generate a reflexive process in which: “the theory serves as a basis for the organization of the model but the data itself is also used to generate ideas in an exploratory way“ (Byrne, 1998: 67).

8

Complexity and social systems David Manuel-Navarrete

One of the main endeavors of applying this approach to social systems is the identification and description of interactions: “In social science we still lack insight into the nature of social interactions, what the variables are, and what the functional relationships between them may be. To show that chaos theory can explain events in a social science setting, we must begin to understand the nature of those variables, interactions, and relationships”(Marion, 1999). Special attention is devoted to the micro (individuals)- macro (society as a whole) interactions: “Conceptualizing the relationship between the conscious agency of individual and/or collective social actors and the social conceived of in terms of social structure” (Byrne, 1998). Society overall behavior can only be understood as an emergent consequence of the holistic sum of all the myriad behaviors embedded within. Because of the dynamic interactions among individuals, a system emerges. It is not deliberately created, it may, and very often does, just happen. The system exerts a commanding influence on the behavior of the individuals, thus further assuring its vitality and survival: Social systems as self-organized emergent entities (Marion, 1999).

Departing from the consideration of individuals as the building blocks of social life, the task becomes either modeling the connections between levels or figuring out general bridging laws able to connect statements about regularities observed at different levels. Both activities require taking measures at different hierarchical levels to account for micro-macro relations. As suggested by Byrne (1998), the data collected can be organized into contingency tables (n dimensional condition space within which cases (individuals) are found in certain sub-domains and not in others). This accounts for the social system in terms of elements within the system, but we also may have measures of the state of the whole social system in terms of key descriptive attributes. Sometimes the system measures are an aggregate value of individual measures, but its systemic implication is different from that of any of the individual components which makes it up. It is important to note that we may consider the condition state of a social system as a whole in terms of a set of indices. Here we can use time series data to describe whole societies over periods of change and see if the changes are non-linear (Byrne, 1998).

Finally, this approach establishes a very clear dichotomy between holism (looking at the whole society) and reductionism (looking to aspects of society or of individuals). In lots of its aspects the approach is clearly declared as anti-reductionist. However, there is not any awareness about (1) the reductionistic risks of holism (i.e. reduction to the whole), and (2) the possible inconveniences of still being applied by sociologists, within social sciences, and considering exclusively traditional social variables.

Third approach: Post-Normal Science

The previous approaches described chaos and complexity theory as a formal system that, with the supplement of reductionism (or vice versa), would provide a way of describing the whole reality as an ultimate truth. The Post-Normal Science Approach questions the

9

Complexity and social systems David Manuel-Navarrete

ability of any specific formalism (or system of knowledge) to account for the whole reality. This serious doubt brings the suggestion of transforming our conception of what science is, including under the label of “Scientific Knowledge” a multiplicity of Systems of Knowledge (SK). A SK can be defined as a fundamental code of a culture which construct the episteme, and that determine the empirical orders and social practices of a particular group in a particular historical era. An episteme is an organization of relations that allows a discourse to make sense.

An important implication of this broad generalization of the scientific method is to face the possibility that mathematics is not sufficient to encompass the human and social world (Back, 1997). We may then turn to other aspects of imagination and creativity (i.e. art and literature). Complexity theory may help in this direction, but we must consider the possibility that some aspects, which are relevant for understanding social systems, cannot be encompassed by this formalism, even though it is more general than the used traditionally by reductionistic research. For instance, we could question whether Western thinkers have blinded themselves to important understanding in their zeal to remove spiritually from the gaps in our understanding. Nature is not a predestined automaton; rather it is capable of free will, of creativity, of teleology. God cannot, after all, be banished from the gaps, for the gaps are axiomatic in nature (Marion, 1999). The question is whether complexity theory is an appropriate way for incorporating the spiritual aspect relevant for the description of social systems. Generally speaking, if an aspect is considered important for the description of social systems, and this is a decision rather than a technical issue, then we have to look for an appropriate way of representing it. In this context, ‘appropriate’ refers mainly to the ability of communicating the information linked to the aspect to be considered.

To adopt a pluralism of SK involves a lesser emphasis in measurements. Numbers are not the only way of representing reality and are not more objective than other kinds of encoding. Scientists are not someone collecting numbers, but someone interacting with the other individuals, in the society, as an individual knower.

The crucial shift comes from the acceptance of truth as just a convention that allows us to communicate with the rest of individuals and within different parts of our inner world. We cannot know analytically whether there is a reality out there. This is a question of both faith and practicality, which opens the discussion about what languages are more suitable in any given situation for any given purpose. As a consequence, the starting point of any scientific endeavor consists of identifying the personal or collective purpose underlying any specific research exercise.

At this point, it is easy to get stuck into the debate around the existence of an objective world, but the point is that even if such a thing exists, there would not be a unique way of grasping the whole reality. Conversely, if an objective reality does not exist, then we also need to construct different formal systems and conventions for allowing communication among humans. A strategic decision is to see reality as an elusive commodity. The best we can do is closer and closer approximations. We may never arrive at absolute and total reality. SK serve to manipulate parts of reality, understand what it is made up of, and

10

Complexity and social systems David Manuel-Navarrete

communicate with others our own personal perceptions and experiences, but none formalism should not be confounded with ‘the reality’. Knowledge is perspectival, requiring multiple viewpoints to interpret a heterogeneous world. Every individual perceives the world singularly. This provides any individual with an intrinsic value. But in order to communicate, we need to agree on a common representation of our different perceptions.

Two important assumptions of this approach are that: (1) the observer cannot be detached from the observed, and (2) situations are unpredictable in themselves, not only by virtue of the limits of their observers. This unpredictable character of reality allow us to devote our efforts to other ways of understanding, where surprise can be seen as something positive and creative. Freedom recovers its meaning as a word usable by science and philosophy (Eve et al., 1997).

These assumptions imply a new science, new scientific institutions (if any) moving from a scientific public administration to a publicly administered science. The new Science might deal with diverse kinds of systems and situations (not only the simplest physical ones) through the use of different languages and formalisms. The price is to abandon the pretension of objectivity. This brings to the fore the need for a new conception of quality control of the scientific discourse. It becomes essential to understand the formal mechanisms operating in any SK, so that we can check the coherence of contextual applications with regard to the internal formal rules and symbols (semantic checking). More importantly, a crucial aspect of research becomes the understanding of the deep implications of choosing among different SK on (1) our way of perceiving the world, (2) our social structures and organizations, (3) the power relations in our society, and (4) our structures of values, among other aspects. Part of this understanding comes through the study of the history and philosophy of the scientific pursuit. It is not only important to know how to use tools of analysis, but also what does to use these tools mean, where they come from, which significance have in our lives and in our way of representing reality. It is questionable whether current research institutions are prepared for facing this challenge.

According to this approach, complexity should not be based on an unfolding of the modern paradox into dichotomies of local versus global, individual versus social, order versus disorder, or human versus non-human (Lee, 1997). The idea of complementarities could be a good candidate for substituting the breakdown of linear dichotomies. Thus, nearness and distance, subjectivity and objectivity, individual and societal, and so on are not seen as mutually exclusive opposites but rather as a relation of mutual improvement (reciprocal causation); that is, the development of one side of a dichotomy is a precondition of the development of the other side (i.e. exchange is not the addition of two processes of giving and receiving but rather a new, third one which emerges while each of the two processes is, absolutely simultaneously, the cause and effect of the other (Eve et al., 1997)). The world becomes gray rather than black and white. Whether it will be considered as black or white will depend on the SK used to describe it.

11

Complexity and social systems David Manuel-Navarrete

Complementarity of dichotomies does not prevent SK of having associated values (i.e. influencing the decision on which aspects of reality focus our inquiry) that should be described and taken into account. Apart of these subjective values, there is another kind of values intrinsic to the analysis of social systems, which emerge as a result of the balance among power relations, determining the decision of using a specific SK over another. This latter kind of values is very close to ideologies (a form of conceptual simplification, prescribing certain notions and proscribing others). The difference is that while SK has the main goal of explaining one part of reality (discerning truth), ideologies seek for changing one part of the reality in a certain direction. It could happen that certain values intrinsic to a certain SK fit better into the scheme of a certain ideology. This would promote the use of that specific SK in order to articulate the discourse within that specific ideology. However, this is different than saying that all SK are necessarily ideological and political. On the other hand, SK influences our perception of the world, our sense of reality, and this has an effect in the way we organize society (i.e. power structures). In any case, we are speaking about very complex relationships.

Post-Normal Science shifts radically the main focus of the first and second approaches. The point is not to discern whether complexity theory represents a complement of the scientific program or a transforming paradigm within this program that will allow a better representation of truth. The relevant issue now is what are the implications of applying a specific SK (i.e. complexity theory) to describe a specific and context-dependent situation. The main endeavor of science becomes making available to individuals the plethora of SK that can be relevant in a specific situation. Therefore, the key question to address here is; how does this approach evaluate the implications of applying complexity theory to the specific context of ‘social systems’?

From the stand point of the Post-Normal Science approach, two main decisions must be taken in order to answer the former question: (1) what aspects of social systems should be represented using complexity theory, and (2) which is the purpose of the description? (i.e. if the purpose is to get a paper published, then you are going to use the language which has a highest degree of legitimacy in the journal you are writing to).

Complexity theory can contribute in the representation of social systems as nested hierarchies composed by: structures of communication, systems of meaning, or discourse, ideologies, roles of individuals (i.e. “priests” in charge of preserving the current structure), power relations, values, individual perceptions, technology, knowledge, configurations of energy, matter, money, information fluxes, human time allocation structures, rituals, and others. It is arguable that complex theory is useful to represent all of these aspects. Nevertheless, this is not a decision to be taken in the frame of this paper. The rules governing this decision are neither universal nor grounded in structures of the mind. Rather they are historically contingent and situationally specific. This recognizes the connections between SK (discursive phenomena) and non-discursive phenomena (institutions, political events, economic practices and processes). Objects of knowledge (i.e. social systems) may have both discursive and non-discursive conditions of existence. The decision will depend on the person(s) who are deciding. With this respect, science must be seen as a part of art. There are some techniques that are useful, but the artist has

12

Complexity and social systems David Manuel-Navarrete

to look inside and express or project aspects of the self. Then the techniques (Systems of Knowledge) are just means for developing the artistic activity and for communicating the results. This artistic activity must be seen as a self-referential exercise carried out by individuals (i.e. through a social process of individualization) and societies (i.e. through institutional changes). We build up our social system by shaping our relationships with the environment. A crucial component of every social system is the internal structure of the self.

In summary, this approach proposes to change the rules of quality governing science. The shift is from the current strict formal rules derived from mathematical formalisms, to pluralistic sets of rules including at the same time very general and very specific formalisms. To do so, it is argued that what is important is the individual process of decision or the social process of negotiation that must lead to the use of a specific set of SK in any specific situation. Within the academic world there is a strong resistance to recognize that any objective system of quality has only legitimacy inside its formal system of knowledge. On the other hand the alternative of leaving the evaluation of quality exclusively to a process of negotiation has the strong inconvenience of depending on the existing power structures and beliefs (i.e. the already existing legitimacy among the currently used systems of knowledge) within the process of negotiation takes place.

Some implications of the ideas presented

The alternative use of the different approaches explained above have profound implications in the way we think about the world and the way we use mathematics as a description of it.

The resulting picture coming from the application of a reductionistic paradigm shows a world of regular changes, functioning like a mechanical machine, and giving intellectual support to the rise of technology and industry. Order is traditionally seen as the fruit of work: “our culture is characterized by an almost primal or subsconscient acceptance of the work ethic, the conviction that good results from effort” (Marion, 1999: XII). This misunderstanding about the emergence of order has been a powerful metaphor for promoting interventionism oriented to stabilize and make systems more predictable. The scientific endeavor has been shaped according to this objective, namely to describe the stable patterns operating in the physical, biological, and social systems.

Complexity theory brings into fore an important concept: self-organization. According to it, order emerges naturally because of unpredictable non-linear interactions. Self-organization implies a balance between order and chaos. This way of describing reality could be more helpful for explaining adaptation, deliberative behavior, reproduction, and evolution. Social systems must be able to map their past (memory) and consolidate gains (order), but at the same time stimulate novelty to adapt to changes in their environment.

According to an approach based on supplementing the Modernist Program, the new nonlinear science would suggest that the answer, if there is one, might lie in a totally

13

Complexity and social systems David Manuel-Navarrete

unexpected direction (i.e. emergence of a new dimension of interdependencies). Application of complexity theory to the social sciences may be difficult because of our frequent inability to impose rigorous controlled conditions for the conduction of many studies (Eve et al., 1997), but chaos theory gives hope for attaching quantities to variables meaningfully. In this way, we are be able to finally forecast processes at the social level through chaos theory. Chaos can lead to order and indeed to a form of stability that gives us an improved ability to forecast and even control the future. Chaos is first and foremost deterministic, and it is this determinism that is illuminating, not indeterminacy. This view could lead, for instance, to a representation of society where people organize themselves into an economy through a myriad of unconscious individual acts of buying and selling; it happens without anyone being in charge or consciously planning. In social systems, the mechanism for selection of one social alternative over another would be “communicative success”.

In the viewpoint of the Metaphorical-analytical Approach, any system that obeys rules- even if the behavior is chaotic- can be controlled once the rules are known (Byrne, 1998). Change in self-organized systems is adaptive; that is, relationships between components at one level change in response to changes taking place in other levels. On the other hand, the system under study has chaotic dynamics, but human actions and policies are still seen as ordered rational actions using the information made available by science (being science the application of mathematical formalism, although now we speak about non-linear equations). The lessons of nonlinear dynamics indicate that through the manipulation of relevant politic, economic and social inputs, desirable social outcomes may not be realistically attainable (Byrne, 1998).

Under the Post-Normal Science Approach, freedom implies discoverable meaning in an act (distinguishing an act from an event). A free act is unpredictable but retrodictable after it has occurred in that it makes sense. Choosing is the kernel of scientific activity (rather than discovering). To explain an event in terms of free choice is to turn one’s research efforts to the history of semantics of the complex personal and social feedbacks that underlie it (Eve et al., 1997). The decision of broaden the concept of science has far-reaching implications as, for instance, to question the primacy of Western Systems of Knowledge, or the suitability of using Systems of Knowledge related with the outer material world as a basis for building the knowledge in other fields.

Conclusions

Science based its authority on experiment and observation, but it needed something to counteract extreme raw empiricism. This non-empirical role has been played by mathematics as the ultimate source of authority. It was claimed that mathematics is not only a convenient way of representation, but it may represent an ultimate reality. Mathematics conform a myth for perfection (analytic-synthetic model). The procedure of science was based on the assumption that we can assign numbers to any state of affairs, that ever-increasing accuracy in measurement is possible and will lead to better theories, and that we can handle the requisite number of variables. Any state of affairs that cannot

14

Complexity and social systems David Manuel-Navarrete

be handled in this way must either be transformed into a manageable form or they must lie outside the purview of science.

Chaos theory is attractive to some social scientist partly because it offers a hope of bringing the attributes of a mature science to an area that until now has seemed intractably complicate except in descriptive terms. The main goal is to find regularities through approaching the precise functional relationships between variables, which is the goal of a mature “mathematized” science. Chaos theory is embraced by social scientist owing to the hope of mathematical validity or perhaps, better, mathematical validation (Eve et al., 1997). But chaos and complexity theories are just two symbolic representations of the world among others.

In an unchanging universe, systems with negative feedbacks are more stable. But in a universe like ours, a positive feedback system, with unpredictable new properties being produced all the time, may well have a better chance of running just to stay in place. Traditional science has developed lots of negative feedback mechanisms, but the primacy of this system of knowledge is not sustainable any longer. In order to allow human knowledge to run in harmony with the environment, we need to preserve our own cultural reserves of unpredictable positive feedback so as to be able to respond appropriately to those that surround us (Eve et al., 1997). The big shift consists of building a pluralistic (and by extension less corporative) science. This is already happening and the best proof is the lack of trust in those institutions built on negative feedback processes.

It is important to conciliate contextual research and general formalisms. There is a broad range of systems of knowledge from more general to more specific, each one being suitable to account for a specific aspect of reality. None of these formal systems should be applied universally. Here it is important do not confound global with universal, and universal with general. Even the most general formalism should be applied contextually. There is not only a question of spatial scale.

Institutions have to change in parallel with the development of a new science. Part of this change consists of giving a greater emphasis by government on the cultivation of virtue in the population, and a somewhat lesser priority on the immediate fixing of problems.

15

Complexity and social systems David Manuel-Navarrete

References

Back, K.W. (1997) “Chaos and complexity. Necessary myths”. In Eve, R.A., S. Horsfall, and M.E. Lee (Eds.) (1997). Chaos, Complexity and Sociology. Myths models and theories. (39-51). Sage Publications. USA.

Bausch, K.C. (2001) The Emerging Consensus in Social Systems Theory. Kluwer Academic Publishers. Dordrecht.

Benathy, B. (1996) Designing Social Systems in a Changing World. Kluwer Academic Publishers. Dordrecht.

Byrne, D. (1998). Complexity Theory and the Social Sciences. An introduction. Routledge. London / New York.

Eve, R.A., S. Horsfall, and M.E. Lee (Eds.) (1997). Chaos, Complexity and Sociology. Myths models and theories. Sage Publications. USA.

Funtowicz, S., and J. Ravetz (1995) “Science for the post normal age”, in L. Westra and . Lemons (eds.), Perspectives on Ecological Integrity, 34-48. Kluwer Academic Publishers, The Netherlands.

Lee, M.E. (1997) “From enlightenment to chaos. Toward non-modern social theory”. In Eve, R.A., S. Horsfall, and M.E. Lee (Eds.) (1997). Chaos, Complexity and Sociology. Myths models and theories. (15-30). Sage Publications. USA.

Luhmann, N. (1994) “’What is the case?’ and ‘What lies behind it?’: The two sociologies and the theory of society. Sociological Theory, 12 (2): 126-139.

Marion, R. (1999). The Edge of Organization. Chaos, complexity theories and formal social systems. Sage Publications. USA

Nicolis, G. (1995). Introduction to Nonlinear Science, Cambridge University Press. Cambridge.

Pierce, B. (1997) “The myth of postmodern science”. In Eve, R.A., S. Horsfall, and M.E. Lee (Eds.). Chaos, Complexity and Sociology. Myths models and theories. (3-14). Sage Publications. USA.

Rosen, R. (1991) Life Itself: A comprehensive inquiry into the nature, origin, and fabrication of life. Columbia University Press. New York.

16