IFCoLog: THE INTERNATIONAL FEDERATION FOR COMPUTATIONAL LOGIC

 

A Working Document for Comment and Discussion

 

 

 

A MANIFESTO

 

Computational Logic has outgrown its humble beginnings and early expectations by far: with close to ten thousand people working in research and development of logic-related methods, with several dozen international conferences and workshops addressing the growing richness and diversity of the field, and with the foundational role and importance these methods now assume in mathematics, computer science, artificial intelligence, cognitive science, linguistics and many engineering fields -- where logic-related techniques are used inter alia to state and settle correctness issues -- the field has diversified in ways that the pure logicians working in the early decades of the last century could have hardly anticipated.

 

Dating back to its roots in Greek philosophy as presented in the works of Aristotle, logic has grown in richness and diversity over the centuries to finally reach the modern methodological approach expressed in the work of Frege. Logical calculi, which capture an important aspect of human thought, are now amenable to investigation with mathematical rigour; and the beginning of this century saw the influence of these developments in the foundations of mathematics, in the work of Hilbert, Russell and Whitehead, in the foundations of syntax and semantics of language, and in philosophical foundations expressed most vividly by the logicians in the Vienna Circle.

 

Picking up on these developments and on the early dreams of mechanised reasoning, the Dartmouth Conference in 1956 raised explicitly the hopes for the new possibilities that the advent of electronic computing machinery offered: logical statements could now be executed on a machine with all the far-reaching consequences that ultimately led to logic programming, deduction systems for mathematics and engineering, logical design and verification of computer software and hardware, deductive databases and software synthesis as well as logical techniques for analysis in the field of mechanical engineering. In this way the growing richness of foundational and purely logical investigations that had led to such developments as:

 

·  first order calculi

·  type theory and higher order logic

·  non-classical logics

·  semantics

·  constructivism, and others

 

was extended by new questions and problems in particular in computer science and artificial

intelligence, leading to:

 

·  denotational semantics for programming languages

·  nonmonotonic reasoning

·  logical foundations for computing machinery such as CSP, p-Calculus and others for program verification

·  logical foundations for cognitive robotics

·  syntax and semantics for natural language processing

·  logical foundations of databases

·  linear logics

·  probabilistic reasoning and uncertainty management

·  logical foundations and its relationship to the philosophy of mind,

 

and many others.

 

This growing diversity is reflected in the numerous conferences and workshops that address particular aspects of the fields mentioned.

 

For example, only twenty years ago, there was just one international conference on automated deduction (later to be called CADE). Today there is not only CADE but among others: LICS (Logic in Computer Science), RTA (Rewriting Techniques and Applications),LPAR (Logic Programming and Automated Reasoning), the TABLEAUX Conference, TPHOL (Theorem Proving in Higher Order Logic), UNIF (Unification Workshop), and FTP (First Order Theorem Proving), each of which is held regularly with its own set of proceedings and supported by a mature community. Frequently these conferences are backed up by dozens of national and international workshops, such as JELIA, CALCULEMUS, the Induction Workshops, PROOF.PRESENTATION, USER INTERFACES for ATP; the Nonmonotonic Reasoning Workshops, the Knowledge Representation Conferences, the Frame Problem meetings and many more.

 

A similar growth of meetings has been seen in the other areas mentioned before, for example logic programming with its main international conferences and workshops (more than two dozen regular meetings and events). Expansion and diversity can also be found in linguistics and natural language processing with their many conferences and workshops, as well as logic and the philosophy of science with its world conference: Congress of Logic, Methodology and the Philosophy of Science. Logical foundations of computer science and verification has seen major growth with its traditional conference LICS, nowadays represented also by CAV (Computer Aided Verification), FM-Europe (Formal Methods) and others, each of which is again accompanied by national and international  conferences, workshops and other events that reflect the growing industrial importance of these techniques.

 

This diversity is not necessarily disadvantageous, as every community that has evolved addresses its own important set of problems and issues, and it is clear that one group cannot address them all. However, fragmentation can carry a heavy price intellectually -- as well as politically -- in the wider arena of scientific activity where, unfortunately, logical investigations are often still perceived as limited in scope and value.

 

For these and other reasons we hereby propose to establish an international federation -- IFCoLog -- to be registered as a legal entity and possibly accepted as a member society in the International Council of Science (ICSU). The members of the International Federation for Computational Logic (IFCoLog) will be the communities associated with the major conferences and logic societies, and they in turn will encompass the ten thousand or more individual members working on logic-related topics.

 

 

THE FEDERATION

 

How can these dozens of societies, sociologically evolved communities and conference affiliates be re-united without losing their historical identities? One possible solution is inspired by the manner in which the European AI societies are organised into ECCAI (European Coordinating Committee for Artificial Intelligence): there is one registered society, namely ECCAI, whose members are the European national AI-Societies. With the growing unification of Europe there are currently about 25 members who represent all European AI researchers and whose representatives meet every two years at the time of ECAI, the European Conference of Artificial Intelligence.

 

So the idea for IFCoLog is as follows: An International Federation for Computational Logic (IFCoLog) will be created and legally registered, whose members are the current (and future) communities related to computational logic. Currently, this would include the groups and their respective representatives who are listed below for the Board of IFCoLog. Some of these are actually organised into legal societies, others are simply associated with a conference, but nevertheless form a scientific community of considerable size and importance. To make this workable, it will be required to form an organisational structure that does not infringe on the interests of the individual communities but nevertheless ensures maximum cohesion. The following organisation is therefore proposed:

 

THE GENERAL ASSEMBLY

 

The General Assembly will consist of the representatives of the individual member societies, i.e. each society that has been accepted as a member by the board of IFCoLog will elect a representative for the General Assembly for IFCoLog. We expect the General Assembly to consist of several hundred members eventually. The General Assembly will discuss most issues related to the federation electronically and possibly hold meetings in conjunction with one of the main conferences, such as FLoC, or later the joint  conference, which most members are likely to attend any way.

 

The General Assembly will elect the Board, whose task is similar to a government of a state, whereas the General Assembly would roughly correspond to the elected parliament.

 

THE BOARD

 

The Board will be composed of elected representatives from the member societies and communities. In order to maintain a functional size it is currently restricted to twenty members, which may grow over the years to at most twenty five members. Thus the (dozens of) member societies, represented in the General Assembly will be partitioned into areas, currently represented by the following groupings:

 

·  Automated Deduction (CADE, AAR and JAR, TABLEAUX, FTP, LPAR etc.)

·  Formal Methods (FM Europe, etc.)

·  Fuzzy Reasoning

·  HOL (Conferences and Society)

·  Knowledge Representation (KR, FAPR, JELIA, SIGFAI, etc.)

·  Logic in Computer Science (LICS, EACSL, etc.)

·  Logic and Databases (LID etc)

·  Logical Foundations and Methodology (DLMPS)

·  Logic and Language (CoLi, LCI, ESSLLI, etc.)

·  Logic and Machine Learning (ILP)

·  Logic Programming (ALP, CL 2000, Logic in DB, LP workshops: LOPSTR, etc.)

·  Nonmonotonic Reasoning (NMR, RMS, etc.)

·  Symbolic Computation (JSC)

·  Symbolic Logic (ASL, national logic societies)

·  Term Rewriting (RTA, UNIF, etc.)

·  Verification (CAV, FM-Europe, etc.)

 

Others will be added as needed.

 

We expect this partitioning into areas to be handled without too strict a borderline and for current purposes the General Assembly as a whole will elect the board members under the given understanding that each area has a fair chance to be represented. If necessary areas may later be established within the General Assembly on a more formal basis.

 

The idea is that this should be an open process with as little factionalism as possible: the main motivation is not to alienate, but to unite.

 

 

 

 

THE EXECUTIVE COUNCIL

 

This is the actual executive that would run the everyday business of the Federation, and it should be comprised of:

 

·  The President

·  Five Vice-Presidents

·  An Executive Officer

     ·  The Founding President
     ·  A Treasurer     
     · The PR Officer (for journals, WWW, press releases, etc), and

·  A full-time co-ordinator, a halftime secretary and further technical staff as needed to carry out the everyday business (such as website, email listings etc.).

 

1. The President. This should be an outstanding scientist with appropriate presidential personality who can unite and bring together the many factions. He or she is not necessarily active in the day-to-day running of the Federation. The President will be elected by the Board and the Executive Council and should serve for six years.

 

2. The Vice-Presidents. There should be five vice-presidents elected by the Board and the Executive Council for a limited period of time (say six years). Ideally the five VPs should represent the major scientific subareas, such as symbolic logic, logic programming, automated deduction, logic in computer science and artificial intelligence, formal methods and verification, logic and language as well as logic and philosophy. The vice presidents should also come from the major geographical regions of the world i.e. at least: North America, Europe, and the Pacific Rim should each be represented. This reflects the current shape of the global village, which is likely to remain dominant for the first part of the next century at least. One Vice-President should come from the Network of Excellence in Computational Logic (CoLogNET) as long as it provides initial support and resources.

 

MAIN PROMOTIONAL ACTIVITIES OF THE FEDERATION

 

Inasmuch as the Federation aims to counterbalance the growing division in the field and to represent it once again in its entirety, it is deemed to work on the following four major goals:

• Information
• Represenation
• Promotion
• Cooperation

More specifically it will be activ in order to :

• influence funding policy
• increase international visibility
• set up concrete educational curricula
• set up special chairs in computational logic
• encourage high-quality teaching materials (books, videos, etc)
• maintain an active information policy
• create an infrastructure for web sites and links
• maintain a register of individual and corporate e-mail addresses
• establish an informal journal, (such as AI Magazine, "CoLogNET Newsletter" or others)
• found a formal scientific journal. The ACM Transactions in Computational Logic (ToCL) has been proposed and essentially accepted as the official journal, but this still needs formal approvement of the EC and the Board
• foster information sharing and exchange with other providers such as DBLP, CoLogNet, etc.
• establish its own set of scientific awards
• provide an insurance policy for the individual affiliated conferences in case of loss
• hold and support one major federated conference every three years (FLOC)

THE JOINT CONFERENCE

 

Every four or five years, the member communities agree to hold one major conference -- IJCoLog: The International Joint Conference of the Federation for Computational Logic -- that consists of the back-to-back conferences of the individual members, similar to FloCS which is held every two or three years, while individual conferences like CADE, LICS, CL 2000 etc. will be held yearly or biannually. This joint conference, with probably more than a thousand expected participants, will be a major show of strength, unification and cross-fertilisation, and will ensure the overall visibility of the Federation.

 

THE EXECUTIVE COUNCIL

 

For the initial period, the proposed officers are:

• President: Moshe Vardi
• Vice-President: Dov Gabbay
• Vice-President: Alan Bundy
• Vice-President: N.N.
  
• Vice-President: Johan van Benthem
• Vice-President: Jörg Siekmann
• Founding President: Dana Scott
  
• Executive Officer: Werner Stephan
• Coordinator CoLogNET: Heike Scheuerpflug
• Treasurer: N.N.
• Secretary (halftime): Helga Kochems

 

 

THE BOARD

• Samson Abramsky
• Krzysztof Apt
• Marc Bezem
• Bruno Buchberger
• Edmund Clarke
• Tony Cohn
• Jens Erik Fenstad
• Ulrich Furbach
• Koichi Furukawa
• Georg Gottlob
• Deepak Kapur
• Claude Kirchner
• Donald Martin
• John Mitchell
• Tobias Nipkow
• Erik Sandewall
• Jeannette Wing

 

 

Special requests and queries should be sent to the Executive Officer, Werner Stephan: 

or to the Co-ordinator, Heike Scheuerpflug : .