Frames, Multi-Agents and Good Behaviours in Planning Rationales

Spatial planning still lacks of robust scientific attention to knowledge and knowledge-in-action coordination in multiagent environments (Faludi, 1973, 1987; Friedmann, 1987). Intuitively, this limitation is particularly invalidating, as the current generation of spatial plans aims at democratizing its traditional expert and top down approach and enhancing its knowledge contents and multilogic potentials (Forester 1989, 1999; Friedmann, 1987; Healey, 1997; Sandercock, 1998). At the forefront of knowledge engineering, distributed and multiagent intelligence, unfortunately, when paying attention to coordination of multiagent microtasks in task accomplishment is still short in the elaboration of the integrated social thoughts that are prerequisites of the new generation of knowledge-based interactive spatial plans (Ferber, 1997). However, in knowledge-based spatial planning engineering there is increasing awareness of the typical rational and computational complexity of multiple source knowledge integration (MSKI): problems like contradictions in beliefs and intentions, semantic redundancies and uncertainties, and other theoretical and practical inconsistencies definitely hamper the spreading new age democratic planning arenas, making their assumptions and tools largely ineffective (Borri, 2001). The new strategic, interactive, and strongly future-oriented and visionary socioenvironmental planning, in which through cognitive sessions and forums a multiplicity of agents (stakeholders) interact to set and solve complex problems is an interesting challenge to multiagent coordination in knowledge engineering (Avlijas, Borri, Monno, 2005). Mainstream strategic planning has a typical iterative organisation, which consists of a number of stages, all of which are assisted by a knowledge engineer (as individual or group figure) performing roles of facilitator, mediator, or reflexive agent: (i) preliminary organisation of the cognitive interaction (problem definition, selection of stakeholders to be involved), (ii) implementation of CI (facilitation-mediation-negotiation assistance by the knowledge engineer, going from the traditional Delphi sessions to the recently adjusted ones characterized by recognizable agents who think and act in destructured and often conflictual ways, so being in tune with real processes of building of social knowledge), and (iii) finalized elaboration of the experimental outcomes in terms of problem setting and/or problem solving. An evident current methodological trend is increasing multiagent protagonism and responsibility in all phases of social elaboration of MSKI – according to the new social knowledge style which pervades spatial planning – and enhancing endogeneous knowledge potentials. Cognitive interventions by knowledge planners along these stages of multiagent interaction can assume active modes, insofar as planners give definite structure to (i), make strong efforts to focus knowledge in (ii), and make interpretive more than descriptive use of the experimental outcomes in (iii), or passive modes, insofar as they facilitate problem emersion from stakeholders-agents who freely enter the arena in (i), do not focus knowledge in stage (ii), and make descriptive – more than interpretive – use of the experimental outcomes in (iii). Another important notation about multi-agent cognitive experiments deals with their two alternative motivations: per se problem complexity requires support by and coordination of a multiplicity of cognitive agents; problem complexity is sufficiently small to allow that a cognitive agent can set-solve alone the problem, but cognitive democracy – more than one agent involved – is required because of the existence of a moral aspiration to it. So, being sufficiently aware of MSKI complexity, we’ll deal only with a limited number of aspects of strategic interactive planning, assumed as particularly interesting for their potential developments in the near future: emergence of new knowledge from the experiments and practical relevance of the cognitive experiments in problem setting and/or solving.