19-21 November 2005, ESF Exploratory Workshop

---

Call for Participation

ESF Exploratory Workshop on
UNDERSTANDING THE DYNAMICS OF KNOWLEDGE (UDK 2005)

---

17-19 November 2005, Certosa di Pontignano, Siena, Italy

Sponsored by:
- European Science Foundation
- University of Siena
- Institute for Cognitive Sciences and Technologies, CNR Roma

Workshop website: http://www.media.unisi.it/cirg/udk

On line registration: http://conference.unisi.it/gest-congressi/web/online_ing.asp?id_cong=658.
Please note that registration will be open until half october.

Overview

This ESF Exploratory Workshop aims to explore various co-related features of knowledge dynamics in human and artificial agents, including evolution of concepts, cognitive development and learning, and short-term dynamics such as belief change and information update. Both individual and social dynamics of knowledge will be explored, and their interplay analyzed. Formal and computational models will be compared with socio-cognitive theories of knowledge change, and with empirical findings in psychology, anthropology, and social sciences.

Scientific content and rationale

The dynamics of knowledge are currently the focus of a vast set of theories and models in a multi-disciplinary perspective: short-term dynamics of knowledge are studied in experimental and developmental psychology (e.g. false belief task, balance scale task, context effects, priming), but also within formal and computational frameworks in epistemology, logic, Artificial Intelligence and cognitive economics (e.g. belief revision, information update, argumentation, non-monotonic reasoning); long-terms dynamics of knowledge, like learning and development, are investigated by cognitive science and developmental studies (concerning both individuals and social groups, both humans and artificial agents), but also by comparative anthropology, cognitive ethology, social sciences and cultural studies; finally, evolutionary dynamics of knowledge (e.g. genetic heritage, socio-cultural transmission, phylogenesis of communication and languages) are studied by evolutionary psychology (e.g. cheaters detection and reasoning biases), cognitive ethology (e.g. evolution of adaptive traits in communication and cognition), cognitive economics (e.g. evolutionary game theory), social sciences (e.g. socio-biology), linguistic (e.g. gestures, gazes and facial expressions as precursors of language), but it is also a promising field of application and source of inspiration for Artificial Intelligence and computer science (e.g. genetic algorithms for neural networks, Artificial Life, co-evolving Multi Agent Systems).

In spite of the relevant results achieved in some specific areas (e.g. developmental approaches to mental disorders, relevance theory for utterances understanding, belief revision models, agent-based social simulation, connectionist architectures), all these models still suffer from a severe lack of integration: sometimes different approaches reflect different objectives of the investigation (e.g. developmental studies vs. evolutionary models of knowledge spreading in societies), sometimes they just express alternative viewpoints on the same phenomena (e.g. belief revision models vs. argumentation theories). In any case, the systematic interdisciplinary comparison between different approaches to knowledge dynamics has been so far too limited, both in extend (i.e. only few disciplinary perspectives were confronted on the same problem) and in depth (i.e. only surface similarities and contrasts were emphasized). On the contrary, the pervasive nature of the dynamics of knowledge compel to completely reverse this tendency: all relevant features and instances of knowledge change, development and evolution are closely tied together, therefore we should aim to provide an highly integrated framework for their explanation - a framework which is not expected to overrule more specific theories, but rather to enhance them through closer comparison with related approaches.

More specifically, this Workshop focuses on three major axes of integration, considered crucial for a better understanding of the dynamics of knowledge:

SHORT-TERM, LONG-TERM AND EVOLUTIONARY DYNAMICS OF KNOWLEDGE
Different dynamics of knowledge can be classified according to their time-scale: short-term dynamics (e.g. context effects, priming, belief revision strategies), long-term dynamics (e.g. learning and development), evolutionary dynamics (e.g. genetic heritage and cultural change). Although often fruitful in suggesting the most adequate approach to specific phenomena of knowledge change (like the use of epidemiological models to understand cultural transmission), this categorization cannot remain rigid and static. All these different layers of knowledge continuously interact with each other: short-term dynamics evolves dramatically during development (e.g. false belief attribution) and it is also affected by learning and cultural variation (e.g. default trust attribution to specific information channels is a direct effect of cultural values and educational practices); similarly, long-term dynamics like learning and development needs to be integrated into (and justified by) broader frameworks, respectively the socio-cultural context of individuals and the evolutionary history of species; finally, evolutionary explanations of knowledge dynamics must prove consistent with empirical evidences from the other levels - explaining why and how humans have evolved certain specific short-term and long-term dynamics rather than others, and what needs, pressures and preconditions have been answered by such dynamics. The first part of the Workshop (cf. Preliminary programme) will be especially devoted to explore this family of problems, fostering closer comparison and future integration between theories of knowledge dynamics at various time-scales.

INDIVIDUAL AND SOCIAL DYNAMICS OF KNOWLEDGE
Individual and social dynamics of knowledge cannot be understood separately from each other: the social creation, transmission and distribution of knowledge is an emergent phenomenon from the complex interaction of individual agents; at the same time, the cognitive processes of knowledge acquisition, interpretation, elaboration and generation are shaped and motivated by several social factors, e.g. cultural values, norms, reputation, communication protocols. How does social dynamics of knowledge emerge from individual processes of information update, belief revision, inter-personal argumentation? How these cognitive processes are in turn influenced and partially shaped by social pressures, with special reference to knowledge change, propagation and availability? Which paradigms and tools are most appropriate to simulate in artificial societies both the social emergence of large-scale knowledge dynamics, and their feedbacks on small groups and individual cognition? The second day of the Workshop will be mainly focused on such complex interactions between micro- and macro-level of knowledge dynamics (cf. Preliminary programme).

SOCIO-COGNITIVE THEORIES AND FORMAL AND COMPUTATIONAL MODELS OF KNOWLEDGE CHANGE, DEVELOPMENT AND EVOLUTION
Several aspects of knowledge change, like belief revision, information update and argumentation, have been extensively studied within strong formal and computational frameworks, both in logic, Artificial Intelligence and cognitive economics. While these approaches provide powerful tools of formalization and promising opportunities for practical applications (e.g. information retrieval, belief-based decision making, trust evaluation for security), they would in turn greatly profit from more in-depth cross-fertilization with other socio-cognitive disciplines, like cognitive psychology, social sciences, developmental studies, linguistic, evolutionary and comparative anthropology. This interchange would broaden the perspective of current formalisms, defining new relevant features of knowledge dynamics to be modeled and reproduced in artificial systems; on the other hand, socio-cognitive theories of knowledge change would equally benefit from updated logical and computational models, being able to test their predictions within better simulative frameworks (e.g. connectionist models for Artificial Life, multi-agent social simulations) and to study the interplay of more sophisticated and realistic artificial cognitive agents (e.g. fully autonomous, goal-oriented, belief-based agents). To this purpose, the third part of the Workshop will bring together world-leading experts in logic, cognitive economics, computer science and Artificial Intelligence, to face the challenge of outlining new formal and computational paradigms for human knowledge and its dynamics.

Preliminary schedule

Thursday 17 November 2005
8.00	Workshop opening
8.30-13.30	Short-term dynamics of knowledge: Cognitive and computational models of belief change	Chair: Elizabeth Robinson (University of Warwick)
	Cristiano Castelfranchi (ISTC-CNR, Roma)	From knowledge to action: Reason-based belief dynamics and belief-based goal dynamics
	Robert French (University of Liège)	TBA
	Natalie Gold (Duke University)	Belief dynamics, framing effects and decision-making
	Boicho Kokinov (New Bulgarian University)	TBA
	Fabio Paglieri (University of Siena)	Data, beliefs, and acceptances: Ontology and dynamics of doxastic states
14.00	Lunch
15.00-20.00	Dynamics of knowledge in childhood: Cognitive and developmental	perspectives Chair: Boicho Kokinov (New Bulgarian University)
	Kristien Dieussaert, Deborah Vansteenwegen, An Van Assche (University of Leuven)	On the stability of instruction and experience based beliefs
	Vittorio Girotto (IUAV, University of Venezia), Michel Gonzalez (CNRS / University of Provence)	Young children's intuitions about posterior probability
	Han van der Maas (University of Amsterdam)	Cognitive development: The balance scale task
	Elizabeth Robinson (University of Warwick)	Belief formation and change in human development
	Stella Vosniadou (University of Athens)	Knowledge acquisition and conceptual change in childhood
20.30	Dinner
Friday 18 November 2005
8.30-13.30	Evolutionary and socio-cognitive dynamics of knowledge	Chair: Cristiano Castelfranchi (ISTC-CNR, Roma)
	Peter Gärdenfors (Lund University)	The evolution of a theory of mind, common knowledge and cooperation
	Maurice Grinberg (New Bulgarian University)	Evolution of conceptual maps as a result of learning
	Brian Loasby (University of Stirling)	Knowledge in economics
	Antoine Billot (PSE-ENPC, CORE, Paris), Jean-Christophe Vergnaud (CNRS-Eureqa, Paris), Bernard Walliser (PSE-ENPC, EHESS, Paris)	Multi-player belief revision and information value in games
	Paolo Turrini, Mario Paolucci, Rosaria Conte (ISTC-CNR, Roma)	Word of mouth: The added value of beliefs' dynamics
14.00	Lunch
15.00-20.00	Formal models of knowledge dynamics: Comparison with social and cognitive theories	Chair: Peter Gärdenfors (Lund University)
	Johan van Benthem (University of Amsterdam / Stanford University)	Cognition as interaction
	Andreas Herzig, Benoit Gaudou, Dominique Longin (IRIT-CNRS, Toulouse) Analyzing communication in a logic of grounding, belief and intention
	Hans Rott (University of Regenburg)	Shifting priorities: Simple representations for twenty-four iterated theory change operators
	Krister Segerberg (Uppsala University), Hannes Leitgeb (University of Salzburg)	TBA
	Frank Veltman (University of Amsterdam)	Mixed moods and unmixable modalities. Modelling beliefs and intentions
21.00	Social dinner in Siena (only for invited speakers)
Saturday 19 November 2005
8.00-14.00	Computation and applications: Knowledge dynamics in logic and Artificial Intelligence	Chair: Robert French (University of Liège)
	Leila Amgoud (IRIT-CNRS, Toulouse)	A unified setting for inference and decision: An argumentation-based approach
	Aldo Franco Dragoni (Technical University of Marche)	Distributed vs. centralized belief revision
	Wiebe van der Hoek (University of Liverpool)	A survey of dynamic epistemic logic
	Henri Prade (IRIT-CNRS, Toulouse)	Possibilistic logic and knowledge dynamics
	Rineke Verbrugge (University of Groningen)	Agents changing their minds about the others' minds: Belief revision in multi-agent systems
14.30	Lunch
15.30-19.30	Changes in view: future developments and priorities in the study of knowledge dynamics Chairs: J. van Benthem, C. Castelfranchi, P. Gärdenfors, R. French, B. Kokinov, L. Robinson Open debate among workshop participants, oriented towards the definition of future scenarios and innovative approaches to knowledge dynamics, in order to envision the research agenda in this field for the next decade and to foster further cooperation among the participants (joint projects, exchange programs, excellence networks, etc.)
19.30		Closing remarks
20.30		Farewell dinner

Organising Committee

Co-Chairs:

* Cristiano Castelfranchi
ISTC-CNR
Institute for Cognitive Sciences and Technologies
National Research Council
Via S. Martino della Battaglia 44
00185 - Roma, Italy
cristiano.castelfranchi at istc.cnr.it

* Boicho Kokinov
Central and East European Center for Cognitive Science
Department of Cognitive Science and Psychology
New Bulgarian University
21 Montevideo Str.
Sofia 1618 Bulgaria
bkokinov at nbu.bg

Local Organizer:

* Fabio Paglieri
(Corresponding Organiser)
Department of Communication Sciences
University of Siena
Piazza San Francesco 8
53100 Siena
paglieri at media.unisi.it

UDK Invited Speakers

Leila Amgoud (IRIT-CNRS, Toulouse FR)
Johan van Benthem (Universiteit van Amsterdam NL)
Cristiano Castelfranchi (ISTC-CNR, Roma IT)
Rosaria Conte (ISTC-CNR, Roma IT)
Kristien Dieussaert (Katholieke Universiteit Leuven BE)
Aldo Franco Dragoni (Politecnico delle Marche IT)
Robert French (Universite de Liege BE)
Peter Gärdenfors (Lund Universitet SE)
Vittorio Girotto (Università IUAV, Venezia IT)
Natalie Gold (Oxford University UK)
Maurice Grinberg (New Bulgarian University, Sofia BU)
Andreas Herzig (IRIT-CNRS, Toulouse FR)
Wiebe van der Hoek (University of Liverpool UK)
Boicho Kokinov (New Bulgarian University, Sofia BU)
Hannes Leitgeb (Universitat Salzburg AU)
Brian Loasby (University of Stirling UK)
Han van der Maas (Universiteit van Amsterdam NL)
Fabio Paglieri (University of Siena IT)
Henri Prade (IRIT-CNRS, Toulouse FR)
Liz Robinson (University of Warwick UK)
Hans Rott (Universitat Regensburg DE)
Krister Segerberg (Uppsala Universitet SE)
Frank Veltman (Universiteit van Amsterdam NL)
Rineke Verbrugge (Rijksuniversiteit Groningen NL)
Stella Vosniadou (National and Kapodistrian University of Athens GR)
Bernard Walliser (Ecole Nationale des Ponts et Chaussées, Paris FR)