|
|
|
Social Software
The study of social interaction utilizing formal and/or systematic modelling techniques has in recent years seen a revival in the social sciences. This is due to a series of new developments and movements in decision theory, game theory and the methodology of model construction. These new developments offer promising novel solutions to, and illuminating perspectives on, the problems and paradoxes associated with the systematic study of social interaction. The new trends in
'social software' have significant implications for the understanding of the systematic approaches to social interaction and the understanding of actual social phenomena.
The conference brings together new foundational research as well as applications in social software from such diverse fields as philosophical logic, sociology, economics and computer science. The ambition is to assess the philosophical and sociological implications arising from the new movements, e.g.:
-
How do systematic techniques relate to real world phenomena, such as morality, social norms and language?
-
What are the consequences of the development of evolutionary game theory for game theoretical approaches to social interaction?
-
What do various results in philosophical logic and probability theory reveal about rationality, agent interaction and common knowledge?
Lectures on these and related questions will be given by some of the most prominent researchers in this new interdisciplinary field. All presentations will be of such a nature that they may be followed by advanced students and scholars in philosophy, economics, sociology and computer science provided with general knowledge of foundational issues.
Speakers
and Abstracts
Ken Binmore / The ESRC ELSE Centre at University College London (UK)
Simulation in
Game Theory: Uses and Abuses
This paper reviews the extent to which computer simulation has
proved useful in advancing ideas in game theory. The
conclusion is summarized by a remark made by the Victorian
astronomer, Arthur Eddington, "Never believe a new piece of
data until it as been confirmed by theory."
Wiebe van der Hoek / University of Liverpool (UK)
Social Laws
in Alternating Time
(based on work with Wojteck Jamroga, Mark Roberts and Michael
Wooldridge)
Researchers in computer science and artificial intelligence
have developed a range of logics for reasoning about the
properties of coalitions. These logics have been applied in
many different areas, and their computational and mathematical
(axiomatic) properties are increasingly well understood. In
the first half of the talk, we present some main developments
in this area. Starting from Coalition Logic, which provides
the basic logical expressions for reasoning about the powers
of teams of agents, we show how this logic can first be
enriched with temporal constructs, which make it possible to
reason about the power of teams over time, and then with
epistemic constructs, which make it possible to formalise and
study theinteraction between ability, cooperation, and
knowledge.
Then, we add deontic operators to the language, enabling one
to reason about what coalitions should or should not enforce.
We then make four key contributions to the theory and practice
of social laws in multiagent systems. First, we show that
Alternating-time Temporal Logic (ATL) provides an elegant and
powerful framework within which to express and understand
social laws for multiagent systems. Second, we show that the
effectiveness, feasibility, and synthesis problems for social
laws may naturally be framed as ATL-model checking problems,
and that as a consequence, existing ATL model checkers may be
applied to these problems. We also show that the complexity of
the feasibility problem in our framework is no more complex in
the general case than that of the corresponding problem in the
Shoham-Tennenholtz framework (it is NP-complete). Finally, we
show how our basic framework can easily be extended to permit
social laws in which constraints on the legality or otherwise
of some action may be explicitly required. We illustrate the
concepts and techniques developed by means of a running
example.
Magnus Jiborn
/ Lund University (Sweden)
Legitimacy
and Consent - the Idea of a Social Contract
Recent developments in evolutionary game theory has provided
new perspectives to the classical, hobbesian idea of a social
contract. Reinterpreting the idea of the social contract in
terms of a dynamic equilibrium selection problem, writers such
as Ken Binmore (1993, 1998) and Brian Skyrms (1996, 2004) have
greatly enhanced our understanding of how the social contract
of our society may have evolved and how it may be sustained.
Traditionally, however, social contract theory has been
thought to adress also issues of justification and legitimacy.
What implications does the modern game theoretical
interpretation of the social contract have for those normative
questions? This is the issue that I will discuss in my talk.
A standard objection against arguments that attempt to derive
legitimacy from a social contract is that the idea of society
being founded by contract is historically false, and that it
is hard to see how an agreement that we have not actually
made, but is merely hypothetical, could be normatively
binding. Such a contract, Ronald Dworkin says, is not simply a
pale form of an actual contract; it is no contract at all
(Dworkin 1989 p 18).
I will try to meet this objection, and argue that the modern
reinterpretation of the social contract might indeed be taken
as point of departure for a normative argument.
Rohit Parikh / City University of New York (USA)
States of
Knowledge and Social Software
In his Rational Ritual Michael Chwe argues that common
knowledge is needed for certain types of group actions like
royal processions, prisoner control and advertising. Of course
strict common knowledge is neither needed, nor, in general,
available. But approximations to it will often serve.
Is common
knowledge always what we WANT? The need to keep military
capabilities or social securty numbers confidential shows that
this is not always so. Lesser knowedge not only will serve but
is often desirable.
We state some
results about the connections between states of knowledge, the
flow of information, and group action.
Robert van Rooij /
ILLC - University of Amsterdam (The Netherlands)
Language as
Social Software: Language Use and Language Structure
Mainstream linguistics is still based on Chomsky’s influential
distinction between linguistic competence, which is regular
and orderly, and linguistic performance, such as actual
conversation, which he sees as ‘noisy’, full of errors.
Following this lead, syntax – which delineates the well-formed
expressions of a natural language – and semantics – which
specifies the (abstract) meanings of these expressions –, have
been studied independent of their actual use, because this
approach is regarded being more revealing. Of course, it has
often been recognized that actual use is much more important
for our understanding of language than this picture suggests.
For one thing, the expectations, goals and preferences of the
interlocutors determine the actual interpretation just as much
as syntactic and semantic rules do. For another, if syntax and
semantics come first, and use comes later, an explanation of
how linguistic rules have arisen and how they change is hard
to get.
Only recently, the way language use influences linguistic
interpretation rules has been studied systematically. In
historical linguistics and psycholinguistics it is argued that
conversational implicatures play a crucial role in the
emergence and acquisition of grammar. A rigorous formal
account, however, is still lacking. Artificial Intelligence
has given rise to an experimental approach towards the
evolution of meaningful communication. It studies how
linguistic rules (a shared lexicon, and some rudimentary
combinatorial principles) can spontaneously emerge from actual
communication. The results of these experiments are very
appealing, but also in need of theoretical explication.
In this talk I will argue that game theory provides us a
precise and testable framework in which both language use and
language evolution can be studied successfully. Moreover, I
will argue that the study of language use is important for the
study of language structure.
Wlodek Rabinowicz / Lund University (Sweden)
Voting
Procedures for Complex Collective Decisions - An Epistemic
Perspective
This presentation, which is based on a joint work with Luc
Bovens, addresses a problem for theories of epistemic
democracy. In addressing a complex issue which can be
decomposed into several sub-questions, a collective can use
different voting procedures: Either it can let its members
vote on each sub-question and then use the voting outcomes as
premises for its conclusion on the main issue (premise
based-procedure, pbp), or it can let the members directly
vote on the conclusion (conclusion-based procedure,
cbp). The procedures can lead to different results, but which
of them is a better truth-tracker? On the basis of Condorcet's
jury theorem, we show that the pbp is clearly superior if the
objective is reach truth for the right (= correct) reasons.
However, if the goal instead is to reach truth for whatever
reasons, right or wrong, there will be cases in which using
the cbp turns out to be more reliable, even though, for the
most part, the pbp will retain its superiority. In that
connection, we also consider the truth-tracking potential of a
'sophisticated' variant of the pbp, which is sensitive to the
size of the majorities supporting each of the premises.
Hans Rott / University of Regensburg (Germany)
Preferences
and Programs for Rational Agents
Social choice theory holds that what should be done in a
community depends on the preferences of the community's
members. These preferences may in addition be thought of as
governing the individual actions taken by the members of the
community. Rational choice theory can be interpreted as
claiming that agents behave as if they were programmed to
maximize the satisfaction of their preferences (much in the
same way as chess computers do). In this picture, the
existence of personal preferences forming the basis for
rational programs of individual and social action is taken for
granted. In my talk, I first review some well-known empirical
work of Debreu, Tversky, and others that appears to call in
question the existence of suitable preferences. I ask whether
the phenomena identified by these researchers transfer to the
realm of `cognitive decisions', and to what extent the
normative enterprise of philosophy is affected by observations
of such a kind.
Brian Skyrms / University of California, Irvine (USA)
Learning to
Network: Some Dynamic Models of Social Network Formation
I discuss some models in which formation of social networks is
driven by learning dynamics. Interactions are modeled as
games, and payoffs feed into the learning model. Issues such
as clique formation and equilibrium selection are investigated
for some simple games.
Program Committee
Pelle
Guldborg Hansen / Roskilde University /
pgh@ruc.dk
Vincent F.
Hendricks / Roskilde University /
vincent@ruc.dk
Stig Andur
Pedersen / Roskilde University /
sap@ruc.dk
Organization
The conference is organized by
and partially
sponsored by the Danish Research Council for the Humanities
Registration
Please write the
Pelle Guldborg Hansen to register:
Department of Philosophy and Science Studies
Roskilde University, P6
P.O. Box 260
DK4000 Roskilde, Denmark
Phone: (+45) 4674 2540 Cell: (+45) 2334 2175 Fax: (+45) 4674 3012
Email:
pgh@ruc.dk
A small conference fee is to be paid upon final registration
(Thursday, May 27, 2004). The conference fee is 75,00
Danish kroner a day, thus participation for the entire
duration of the conference (Thursday, May 27 - Saturday 29,
2004) is 225,00 Danish kroner. The conference fee
covers the conference booklet, tea and coffee during the
breaks.
If email is used include 'SOCIAL SOFTWARE - DK' in the subject
entry. All questions pertaining to registration and
accommodations should be directed to Pelle Guldborg Hansen.
Lunch and Dinner
Arrangements
On Thursday,
Friday and Saturday participants may choose to order lunch
through the conference organization; payments for these
arrangements are due during final conference registration on
Thursday, May 27. Each lunch costs 70,00 Dkr and includes a
buffet and free beverages.
A conference dinner is scheduled for Friday, May 28 at 19:00,
Restaurant A Porta in the center of Copenhagen. The dinner
costs 400,00 Dkr and includes a three course meal and wine.
Only a limited number of seats is available.
Participants interested in lunch orders and/or conference
dinner participation should notify Pelle Guldborg Hansen no
later than Friday, May 21, 2004. Please let it be known
whether you are vegetarian. Only cash payments are accepted
and no later than upon final conference registration during
Thursday, May 27. Undue cancellation of dinner and/or lunch
orders is subject to full charge.
A sight-seeing tour of Copenhagen is planned for Thursday, May
27. The tour will cost 50,00 Dkr which includes
transportation, guides and passes. Signing up for the tour is,
as for the lunch and dinner arrangements, no later than
Friday, May 21, 2004, and payment is due no later than
upon final registration.
|
|
|
