 |
|
ISSN 1320-0682 | ||||
| Volume 4 | 1997 | ||||
Vladimir Dimitrov
School of Social Ecology
University of Western Sydney - Hawkesbury,
Richmond 2753, Australia
Fax: +61(45) 701531
Phone: +61(45) 701903
Email: V.Dimitrov@uws.edu.au
The application of fuzzy logic to social systems creates opportunities to examine:
Fuzzy logic is suited to studying such 'subtleties' in social systems because of its ability to:
Fuzzy logic provides an alternative way of understanding
uncertainty. From this new way of understanding can be derived innovative
approaches and strategies for working with the uncertainty that so often
characterises social systems.
Why is social complexity paradoxical? Because it is the source of many contradictory and opposing forces acting together and, at the same time, the product of the forces. Any attempt to unravel these forces create a circular process that can paralyse further individual or social actions.
The more the members of a group seek to pull the contradictions apart, to separate them so that they will not be experienced as contradictory, the more enmeshed they become in the self-referential binds of paradox' (Smith et al.).
To manage social complexity means to manage its inherent paradoxes and their effects avoiding the danger of double bind paralysis.
Why is social complexity chaotic? The dynamics of social complexity are chaotic in the sense that the aggregate fluctuations of any complex social process (or behaviour) represent an endogenous phenomenon that persists even in the absence of 'stochastic shocks'. The emergence of complex irregular behaviour depends on both the initial conditions under which the process dynamics evolve and the critical values of the parameters characterising this evolvement.
Every time we deal with mathematical representations of social reality as a whole, we deal with chaos. Closed social systems are dissipative; their dynamics are described by strange attractors (Goerner). Open social systems are symplectic (Dimitrov 1994); their dynamics are chaotic without the occurrence of strange attractors.
Emergent phenomena are typical of the dynamics of complexity. In social systems, emergent phenomena manifest themselves any time when collective behaviour transcends the behaviour of its 'components' behaviour; because it transcends individual behaviour, it cannot be 'in' it before the emergence takes place.
To manage social complexity means to manage its chaotic dynamics as well as its effects, thus avoiding the danger of destruction and collapse.
For example, let us consider the 'independency paradox': it is only when social dependencies are established that the interdependence emerges, and it is this collective interdependence (between people, and between people and their environment) that provides the notion of individual independence with meaning. This paradox becomes easily manageable any time we apply the following fuzzy rule:
IF there is interdependence between A and B, AND their relationship is one of a high enough degree of trust, mutual understanding and tolerance THEN both A and B are able to act quite independently.
Another example is the 'difference-similarities' paradox: while differences in interests and values are important for the survival of any human system, they have meaning only because of the similarities that also exist and provide a basis for any collective endeavour. This paradox can be handled if both similarities and differences are balanced by fuzzy logic in a way that excludes too much emphasis on them. When a suitable fuzzy framework is created, differences between people and the viewpoints they represent become complementary and necessary to one another, instead of being contradictory and opposite. In this way, the difference-similarity paradox simply is dissolved. One can find many examples of this transformation in the practice of negotiation.
A chaotic social reality paradoxically seems to unite people in a common desire to be different, unique and creative. Plural descriptions, open for change, based on people's personal experience and shared through dialogue, or derived by collaborative inquiry, are meaningful in such reality - they help to disclose the tensions in it, not to resolve them but to examine the contradictions and inconsistencies and, by the same token, to reveal the conditions which could trigger the emergence of some kind of order (transitory though it usually is) in social systems.
The principle of non-exclusion and non-isolation is fundamental for the use of fuzzy logic in any turbulent social (or socio-ecological) environment. Non-exclusion means that no options or alternative (however improbable it seems to be for inclusion in the future scenarios), should be excluded from consideration; it might turn out to be of crucial significance for the survival of society and its environment. Non-isolation means that chaotic behaviour does not privilege any economic optimality: to isolate only one option, alternative or strategy by describing it as the best, the optimal or the most efficient, from whatever point of view, is senseless; turbulent dynamics do not tolerate any pre-imposed isolation however 'optimal' it may appear to the decision-maker.
The paradoxical and chaotic nature of social reality causes a great deal of uncertainty and vagueness in human decision-making. Under conditions of uncertainty and vagueness, when no ultimate answers or best solutions exist, the search for understanding and consensus between people becomes crucial for the management of social complexity.
An irreducible indeterminacy constantly emerges when we explore more deeply both variety and uncertainty of group decision-making. Paradoxically, instead of consensus being the power house of common social action, it is 'dissensus' which operates in consensus seeking enterprise, permanently implanting chaotic vibrations in the process of communication.
However, the chaos does not cause the communication network to dissipate. Rather, it eventually gives birth to an emerging order in the form of a new type of dynamic consensus between stakeholders: consensus for seeking a consensus.
It does not matter that consensus in our society is 'condemned' to be momentary and transient - what can endure in time is human anticipation and aspiration, the impulse to act together, the natural desire to interact and communicate, to share with and care for others. In other words, not only a search for common actualisation of meaning but strong emotional factors (sharing and caring) catalyse the emergence of second-order consensus out of the chaos of dissent and disagreement, contradictions and conflict.
Consensus-seeking differs from consensus-building. When seeking consensus, stakeholders do not necessarily look for a 'common ground' in the form of full agreement. On the contrary, they underline and study the differences between them, trying to understand social mechanisms which make stakeholders differ in their interests, values, goals, etc. No constraints on stakeholders' views and opinions, no changes of their values and beliefs are required as preliminary conditions for seeking a consensus.
The process is entirely open for emergence of new features and unpredictable situations - spontaneity is an important characteristic of this process. No preliminarily assigned goals exist - pre-imposed goals, constraints or requirements can narrow the scope of the stakeholders' search.
The search for consensus is motivated by the stakeholders' drive to be mutually complementary in their efforts to more fully understand the complexity of the issues and of their concern to find out how to act together in order to benefit from the differences in their knowledge. While conducting their inquiry, the stakeholders are aware of the irreducible fuzziness and uncertainty of this knowledge, yet they agree to explore it together and construct it anew. Thus, a second order, dynamic-type consensus emerges. This kind of consensus is not simply an overlap of stakeholders' interests, values, goals, positions, views.
Second order consensus means that there is a shared acknowledgement that there are diverse, changing and only partially shared views about complexity among the stakeholders, that it is full of zones of uncertainty ('value-dark zones') in which neither the causes nor the effects of what occurs is clear or even can be known. Also, there is an agreement to explore the complexity together in order to arrive at a better understanding of it by using not only your own but each other's experience, expertise and ideas, and, through this better understanding, arrived at an improved preparedness to act together; that is, to engage in joint, collaborative action to manage the complexity.
What occurs in the zones of uncertainty is at least in part influenced by the joint action of the participants (stakeholders), and so the complexity is being partly made by them. On the other hand, as the complexity evolves in time, it also exerts an influence on the stakeholders, and triggers them to reconstruct their views. What the new complex situation will be contains both intentional and unintentional elements as the participants in the joint action coordinate their activities and respond to each other's constructions of the reality of the situation and to each other's actions in it (Maturana, Turner).
What matters in the streamline of the dialogue is stakeholders' willingness to keep moving together - to explore options for consensus, to share knowledge and experience, to learn together how to create and implement group decisions when tolerating, appreciating and even 'celebrating' the differences in people's thoughts and actions.
In the free flow of unanchored, constantly changing and shifting individual values, beliefs and expectations, the higher the degree of willingness for dialogue, trustworthiness and creativity as expressed by a group of stakeholders, the higher their preparedness to act together.
Fuzzy Logic helps to transform the group profiles of the above consensus-seeking parameters into characteristics of stakeholders' preparedness to take actions together. Practically, this help is realised by means of properly chosen fuzzy logic rules.
IF W AND T AND C, THEN PAT
where W, T, C and PAT denote the following fuzzy classes:
W: Willingness to engage in dialogue
T: Trustworthiness
C: Creativity
PAT: Preparedness to Act Together.
Each fuzzy class is described using the following three lingustic variables:
L: low
M: moderate
H: high
It is assumed that in consensus-seeking practice, the values of the membership functions (see Appendix) to the above fuzzy classes, characterised by the linguistic variables 'low', 'moderate' and 'high', can be assigned by a facilitator who participates (observes, facilitates, helps) in negotiation between stakeholders.
The following fuzzy rules have been used in a software product called FLOCK (Fuzzy Logic-based Consensus Knowledge) (Dimitrov & Kopra), specially designed to help facilitators in practical realisation of the idea of second order consensus:
Example:
IF both W AND T = 'low' OR 'high', THEN PAT = 'low' OR 'high', respectively.
IF both W AND C = 'low' OR 'high', THEN PAT = 'low' OR 'high', respectively.
IF both T AND C = 'low' OR 'high', THEN PAT = 'low' OR 'high', respectively.
Example:
IF both W AND T = 'moderate' AND C = 'high' OR 'low' OR 'moderate',
THEN PAT = 'high' OR 'low' OR 'moderate', respectively.
IF both W AND C = 'moderate' AND T = 'low' OR 'high' OR 'moderate',
THEN PAT = 'low' OR 'high' OR 'moderate', respectively.
IF both T AND C = 'moderate' AND W = 'high' OR 'low' OR 'moderate',
THEN PAT = 'high' OR 'low' OR 'moderate', respectively.
Example:
IF W = 'low' AND T = 'moderate' AND C = 'high', THEN PAT = 'moderate'
IF W = 'high' AND T = 'low' AND C = 'moderate', THEN PAT = 'moderate'
IF W = 'low' AND T = 'high' AND C = 'moderate', THEN PAT = 'moderate'
IF W = 'moderate' AND T = 'low' AND C = 'high', THEN PAT = 'moderate'
IF W = 'high' AND T = 'moderate' AND C = 'low', THEN PAT = 'moderate'
IF W = 'moderate' AND T = 'high' AND C = 'low', THEN PAT = 'moderate'
In the context of fuzzy logic, such an actualisation of the meaning of 'being in community' makes sense and helps when dealing with the forces of social disintegration. Any act of acceptance of the other is preceded by some kind of fuzzification of any separating boundaries; this allows people to act together. Neither government nor business can function without accepting the rest of society as it is. Economic, social and political systems should be seen as evolving interrelated networks within society, and not as separated systems. National and global survival of humanity crucially depends on social integrity.
Chaos theory applied to social systems builds a general picture of society as being constituted by interactive 'fuzzy dynamics' giving rise to a range of emerging types of social behaviour (Goerner).
We observe that a multitude of transient equilibria emerge in social dynamics, that there is a constant evocation of increasingly complex forms of order in evolving systems in the life of the society, and that spontaneity is inherent in the temporality of this life. All of these effects are a manifestation of what we term 'the divergence syndrome' (Dimitrov et al.1996) where dramatic social changes can emerge as a result of seemingly insignificant socio-political actions. The divergence syndrome describes the 'self-feeding' acceleration of energy flow that takes place in social systems.
The world remember that 'small' change in the ruling politburo of the former Soviet Union in the mid-eighties, when Gorbachev was appointed as a Secretary-General of the Soviet communist party; the effects were shocking.
Another socio-political illustration of the divergence syndrome: slight changes in the ideological platform of a political party or a politician, expressed by means of seemingly insignificant fuzzy hedges (that is, words used in statements or narratives to intensify or dilute the fuzzy set's membership functions or to change the degree of fuzziness in a fuzzy set, such as: 'more or less', 'very', 'quite', 'somewhat', 'slightly', 'extremely', 'positively', 'generally', 'around', 'about', 'near', etc.) can bring forth enormous changes in their interpretation as a basis for action. For example, during an election campaign, politicians can promise 'to keep the defence potential of the state at a more or less stable level', but once in power, they can use this statement as justification of a large program for testing new nuclear weapons with enormous negative consequences for people and environment.
Fuzzy logic helps in describing, analysing, understanding and eventually dealing with the paradoxical and chaotic dynamics of social systems.
Fuzzy logic also offers a useful framework for understanding the oscillations between disaggregation and communion in society: what matters in a 'healthy community' is the mutual acceptance of people such as they are. This is the starting point for any consensus seeking process.
A fuzzy set (class) A is characterised by the membership function m (A), which takes values in the interval [0,1], that is, m(A): U-> [0,1], where U is a universe of discourse in which A is defined (Zadeh, 1965).
For example, if the universe of discourse U includes all proposals (ideas, statements) generated by a group of stakeholders participating in a consensus-seeking enterprise in response to the question: 'How to move out of the present state of standstill in negotiation?', then the fuzzy class of stakeholders' creativity can be described using a membership function equal to 0 for stakeholder(s) unable to produce any single proposal, and then increasing to 1 towards stakeholder(s) with maximum number of proposals (say 10).
The fuzzy class can be characterised by a linguistic variable corresponding to a set of values of the membership function. In the above example, if the ideas generated by each stakeholder varies from 0 to 10, then the 'creativity' of stakeholder A with only one proposed idea can be characterised by the linguistic variable 'low', while the creativity of stakeholder B with 8 generated ideas can be characterised by the linguistic variable 'high'.
The linguistic variables have fuzzy boundaries; that is, two 'neighbour' variables always have a 'non-zero overlap'. In the above example, if stakehoders C and D offer 4 proposals each, their creativity can be considered both as 'medium' (or 'moderate') with a degree of membership, say 0.8 and 'low' with degree of membership, say 0.3. This occurs because the neighbour linguistic variables 'low' and 'medium' has a non-zero overlap. In the software facilitation tool FLOCK mentioned in the paper, the facilitator assigns the values of the linguistic variables at each stage of the negotiation process.
Fuzzy logic helps us to quantify also the degree of truth of a fuzzy statement (a 'fuzzy statement' is a statement which contains linguistic variables). The standard definitions in Fuzzy Logic are:
where x and y represent some fuzzy statements.
For example, if the degree of truth of the statement x = 'Creativity of stakeholder A is low' is 0.9, then the degree of truth of the statement NOT x = 'Creativity of stakeholder A is NOT low' will be 0.1; that is, 1.0 - 0.9 = 0.1. If the degree of truth of the statement y = 'Creativity of stakeholder B is high' is 0.9, and the degree of truth of the statement z = 'Stakeholder B's willingness to participate in dialogue is high' is 0.6, then the degree of truth of the statement (y AND z) = 'Stakeholder B is with high creativity AND his(her) willingness to participate in dialogue is also high' = min(0.9, 0.6) = 0.6.
Fuzzy Rules are of the form IF... THEN..., where both IF and THEN terms are natural language expressions of some fuzzy classes or their combinations. Fuzzy Logic provides powerful computational techniques for manipulations with these classes aimed at specific problem-solving. More about Fuzzy Rules, their application and frequently asked questions about Fuzzy Logic can be found at http://www.cac .usl.edu/~manaris/ai-education-repository/fuzzy-resources.html.