Ergebnis für URL: http://pespmc1.vub.ac.be/Papers/Workbook.html [Workbook1.gif]
Workbook of the
1st Principia Cybernetica Workshop
edited by Francis Heylighen
Free University of Brussels
PRINCIPIA CYBERNETICA
Brussels * New York
_________________________________________________________________________________
Published by Principia Cybernetica
Editorial Board:
Francis Heylighen^
PO-PESP, Free University of Brussels, Pleinlaan 2, B-1050 Brussels, Belgium.
Cliff Joslyn
Valentin Turchin
Copyright (c) 1991 by Principia Cybernetica, Brussels and New York
All rights reserved. No part of this book may be reproduced or utilized in any
form or by any means, electronic or mechanical, including photocopying or
recording, or by any information storage or retrieval, without permission from
the publisher.
_________________________________________________________________________________
Table of Contents
* [1]Preface
* [2]List of Contributors
[3]THE PRINCIPIA CYBERNETICA PROJECT
* Francis Heylighen: [4]An Evolutionary System Modelling Evolutionary Systems:
Introducing the Principia Cybernetica Project
* [5]The need for a Principia Cybernetica
* [6]Evolution and Constructivism
* [7]From philosophy to method
Cliff Joslyn: [8]General Notes about the Principia Cybernetica Project and
Related Initiatives
* [9]History of PCP
* [10]Relation to similar work
Valentin Turchin: [11]A Tentative Sketch of the Starting Nodes of PCP
Donald H. McNeil: [12]The Principia Project
[13]CYBERNETIC FOUNDATIONS
* Gordon Pask: [14]The Foundations of Conversation Theory, Interaction or
Actors Theory, all Cybernetic and Philosophically so (or 'Some Foundations of
Principia Cybernetica')
* Lars Löfgren: [15]Foundational Issues Addressed by Cybernetics
* Ranulph Glanville:[16] Excavation and Underpinning, Foundation and Building
* Gertrudis Van de Vijver: [17]Error: Epistemological Options in Cybernetics
[18]EVOLUTIONARY PHILOSOPHY
* Valentin Turchin: [19]Metasystem Transition as the Quantum of Evolution
* Cliff Joslyn:[20] Control Theory and Meta-System Theory
* [21]Powers' Control Theory
* [22]Definition of Control
* [23]Control as a Phenomenon
* [24]Powers' Control Systems
* [25]Constructivist Epistemology
* [26]Hierarchical Control
* [27]Learning and Organization
* [28]Memory and Imagination
[29]Challenges from Control Theory to Meta-System Theory and vice versa
Francis Heylighen: [30]Evolutionary Foundations for Metaphysics, Epistemology
and Ethics
* [31]A process metaphysics
* [32]A constructive epistemology
* [33]An evolutionary ethics
Marc E. Carvallo: [34]Self-organization, Evolution, and Religion: Some Notes on
Erich Jantsch's Theory of Religion
Alvaro Moreno, Arantza Etxeberria & Jon Umerez:[35] Biological Information: The
Causal Roots of Meaning
[36]KNOWLEDGE DEVELOPMENT
* Markus F. Peschl:[37] The Emergence of Symbols in Subsymbolic Neural
Representation Systems
* Charles Henry: [38]Non-Verbal Aspects of Language and Knowledge Structuring
* Elan Moritz: [39]Memetics: Introduction and Implication to the Evolution of
Knowledge
* J.L. Elohim : [40]Culture, Cybernetically Interpreted, is a Cybernetic
Reflection of Nature Altered by Culture
[41]COMPUTER-SUPPORT SYSTEMS
* Cliff Joslyn:[42] Software Support for PRINCIPIA CYBERNETICA Development
* [43]Desiderata
* [44]Structures
* [45]Methods and Technologies
Dirk Kenis: [46]MacPolicy: Delphi and Group Decision Support Ideas for Computer
Supported Cooperative Working
* [47]Group Decision Support Systems : State of the art
* [48]Computer Supported Cooperative Working projects at the V.U.B.
Francis Heylighen: [49]Structuring Knowledge in a Network of Concepts
* [50]Network representations of knowledge
* [51]Distinction and entailment types
* [52]Knowledge structuring
* [53]Node identification
* [54]Node integration
Robert Glück: [55]Metasystem Transition in the Machine and its Application to
Knowledge Systems
Peter Beyls: [56]Conceptual Exploration through Intimate Machine Interaction: a
statement *
Elan Moritz: [57]The Case for Imperfect Machines
_________________________________________________________________________________
Preface
This workbook contains abstracts and short papers selected for presentation at
the 1st Workshop of the Principia Cybernetica Project by the workshop scientific
committee. It is meant to give an overview of the work that will be discussed
during that workshop. As such it will allow the participants to prepare
themselves for discussion by examining the links, agreements, and differences,
between their ideas and those of the other participants. In a second stage it may
also function as a proceedings, providing a memory of what was presented there in
Brussels in July 1991.
The aim of the project, and the corresponding workshop can be summarized as
follows. Principia Cybernetica is an attempt by a group of researchers to
collaboratively build a system of cybernetic philosophy, moving towards a
transdisciplinary unification of the domain of Systems Theory and Cybernetics.
This philosophical system will be developed as a network, consisting of nodes or
concepts, linked by different types of semantic relations. The network will be
implemented in a computer-based environment involving hypermedia, electronic
mail, and electronic publishing. The project naturally splits into two issues:
1) development of the philosophy itself, which is systemic and evolutionary,
emphasizing the spontaneous emergence of higher levels of organization or control
through variation and natural selection. It includes: a) a metaphysics, based on
processes as ontological primitives, b) an epistemology, which understands
knowledge as constructed by the subject, but undergoing selection by the
environment; c) an ethics, with the continuance of the process of evolution as
supreme value.
2) development of computer-based tools and methods for collaborative theory
building (CSCW, groupware, SGML, knowledge acquisition...): many participants
with different backgrounds and working in different places exchange knowledge and
opinions about a common problem; their different contributions and reactions must
be integrated and structured, in order to form a coherent system of concepts and
values, transparently modelling the problem domain.
Both issues are united by their common framework based on cybernetical and
evolutionary principles: the computer-support system is intended to amplify the
spontaneous development of knowledge which forms the main theme of the
philosophy.
The contributions in this book have been classified in 5 sections. The first one
offers a general overview of the project, emphasizing its history, its main
philosophical positions, and its method. The second one addresses the issue of
foundations for cybernetics in general. The next section applies the concepts of
evolution and of the emergence of multiple levels to traditional philosophical
questions such as the origin of meaning and organization. The fourth section
emphasizes more in particular the development of knowledge and culture. The last
section studies different ways to use computers as tools to support the further
development of knowledge, in particular the knowledge system that will
incorporate the Principia Cybernetica.
Brussels, May 1991 F.H.
_________________________________________________________________________________
List of Contributors
Beyls, Peter; O. Van Dammestraat 73, 9030 Gent, Belgium.
Carvallo, Marc; Dept. of Philosophy of Religion, State University of Groningen,
Nieuwe Kijk in 't Jatstraat 104, 9712 SL Groningen, Nederland, E-MAIL:
marccarv@hgrrug5.bitnet.
Elohim, J.L.; Antonio Sola 45, Col. Condesa, C.P. 06140, Mexico D.F.; Fax:
+525-761-5023
Glanville, Ranulph; 52 Lawrence Road, Southsea Hants, PO5 1 NY, U.K., TEL. (+44)
(705) 737 779.
Glück, Robert; Technical University Vienna, Institut für Prakt. Informatik,
Argentinierstr. 8/180, A-1040 Vienna, Austria; TEL. (0222) 54 20 63, E-MAIL:
E1802DAA@AWIUNI11.Bitnet.
Henry, Charles; 217M Butler Library, Columbia University, New York, N.Y. 10027,
USA, TEL. 212-854-5477, E-MAIL:
henry@cunixf.cc.columbia.edu .
Heylighen, Francis; PO-PESP, Vrije Universiteit Brussel, Pleinlaan 2, B-1050
Brussels, Belgium; TEL. + 32-2-641 25 25, E-MAIL:
Z09302@BBRBFU01.BITNET.
Joslyn, Cliff; Systems Science, SUNY Binghamton, Box 1070, Binghamton NY 13901,
USA; TEL. +1 -607 729-5348, E-MAIL:
cjoslyn@bingvaxu.cc.binghamton.edu.
Kenis, Dirk; VTBP, Vrije Universiteit Brussel, Pleinlaan 2, B-1050 Brussels,
Belgium, TEL. + 32-2-641 2749.
Löfgren, Lars; Lunds Universitet, Systems Theory, Box 118, S-221 00 Lund, Sweden,
; TEL. +46-46 10 75 19 [office], + 46 - 46 12 88 27 [home], E-MAIL:
lofgren@dit.lth.se.
Moreno, Alvaro; Research group IAS (Information, Autonomy, Systems), Dept. of
Logic and Philosophy of Science, University of the Basque Country, Apartado 1249,
20080 Donostia-San Sebastian, Espana, E-MAIL: alvaro@fil.ehu.es,
Moritz, Elan; The Institute for Memetic Research, P.O. Box 16327, Panama City,
Florida 32406, USA, E-MAIL:
moritz@well.sf.ca.us.
Pask, Gordon; 48 North Street, Clapham Old Town, London SW4 0HD, UK; TEL. + 44 -
71 - 720 18 30 (home), +44 - 71 - 738 82 41 (home).
Peschl, Markus F.; Dept. for Philosophy of Science, University of Vienna,
Sensengasse 8/9, A-1090 Wien, Austria; TEL.: +43 222 42 76 01 / 41; E-MAIL:
a6111daa@vm.univie.ac.at
Turchin, Valentin; Computer Science Department, City College, CUNY, Convent
Avenue at 138th Street, New York NY 10031, USA, TEL. +1 - 201-337 1761 (Home),
+1-2126506178 [office], E-MAIL: TURCC@CUNYVM.BITNET.
Van de Vijver, Gertrudis; Seminarie voor Logica en Kennisleer, RUG, Lamoraal van
Egmontstraat 18, B-9000 Gent ; TEL. 091 - 64 39 61 [office].
_________________________________________________________________________________
THE PRINCIPIA CYBERNETICA PROJECT
____________________________________________________________________________
Francis Heylighen
PESP, Free University of Brussels
An Evolutionary System Modelling Evolutionary Systems: Introducing the Principia
Cybernetica Project
The need for a Principia Cybernetica
It is a common observation that our present culture lacks integration: there is
an enormous diversity of "systems of thought" (disciplines, theories, ideologies,
religions, ...), but they are mostly incoherent, if not inconsistent, and when
confronted with a situation where more than one system might apply, there is no
guidance for choosing the most adequate one. Philosophy can be defined as the
search for an integrating conceptual framework, that would tie together the
scattered fragments of knowledge. Since the 18th century, philosophy has
predominantly relied on science (rather than on religion) as the main source of
the knowledge that is to be unified.
After the failure of logical positivism and the mechanistic view of science, only
one approach has made a serious claim that it would be able to bring back
integration: the General Systems Theory (von Bertalanffy, 1968; Boulding, 1956).
Systems theorists have argued that however complex or diverse the world that we
experience, we will always find different types of organization in it, and such
organization can be described by principles which are independent from the
specific domain at which we are looking. Many of the concepts used by system
theorists came from the closely related approach of cybernetics: information,
control, feedback, communication... In fact cybernetics and systems theory study
essentially the same problem, that of organization, albeit with an emphasis on
either structures and models (systems), or on functions and communications
(cybernetics). In order to simplify expressions, we will from now on use the term
"cybernetics" to denote the global domain of "cybernetics and general systems
theory".
Though a lot of recently fashionable applications (e.g. artificial intelligence,
neural networks, cyberspace, man-machine interfaces, systems therapy ...) have
their roots in ideas that were proposed by cyberneticians, cybernetics itself
tends to stay at a distance from the mainstream scientific developments, and is
correspondingly not taken seriously by that mainstream. Moreover, though
cybernetics aims to unify science, it is in itself not unified. I wish to argue
that, instead of looking down on practical applications, cyberneticians should
try to understand how those applications can help them in their task of unifying
science, and, first of all, unifying cybernetics. It should look upon them as
tools, that can be used for tasks that may extend much further than the ones they
were originally designed for.
A similar situation arose around the end of the last century. Mathematics
proposed a great variety of very successful applications: geometry, calculus,
algebra, number theory, etc. Yet there was no overall theory of mathematics:
these different domains functioned mainly in parallel, each with its own axioms,
rules, notations, and concepts. Though most mathematicians would agree that these
subdisciplines had a "mathematical way of thinking" in common, one had to wait
for the classical work of Whitehead and Russell (1910-13), the Principia
Mathematica, before this unity could be clearly expressed. What was novel in this
work was that mathematical methods were applied to the foundations of mathematics
itself, formulating the laws of thought governing mathematical reasoning by means
of mathematical axioms, theorems and proofs. This proved highly successful, and
the Principia Mathematica stills forms the basis of the "modern" mathematics as
it is taught in schools and universities.
Our contention is that something similar should be done with cybernetics:
integrating and founding cybernetics with the help of cybernetical methods and
tools. Similar to the mathematical application domains (number theory, geometry,
etc.), the applications of cybernetics (neural networks, systems analysis,
operations research, ...) need a general framework to integrate them. Similar to
the integrating theories of mathematics at the end of the 19th century (Cantor's
set theory, formal logic, ...), the integrating theories of cybernetics at the
end of the 20th century (general systems theory, second-order cybernetics, ...)
are not integrated themselves.
Both mathematics and cybernetics are in the first place metadisciplines: they do
not describe concrete objects or specific parts of the world; they describe
abstract structures and processes that can be used to understand and model the
world. In other words they consist of models about how to build and use models:
metamodels (Van Gigh, 1986). Because of this mathematics and cybernetics can be
applied to themselves: a metamodel is still a model, and hence it can be modelled
by other metamodels, including itself (Heylighen, 1988).
In reference to Russell and Whitehead, the enterprise we propose is called the
"Principia Cybernetica Project" (Turchin, 1991; Heylighen, Joslyn and Turchin,
1991). The unified framework we wish to develop can be viewed as a philosophical
system: that is to say a global "world view" ("Weltanschauung"), which is clearly
thought out and well-formulated, avoiding needless ambiguity, inconsistency or
confusion. Starting from cybernetical concepts, it should try to integrate all
the different domains of human knowledge, experience, and action. It should
provide an answer to the basic questions: "Who am I? Where do I come from? Where
am I going to?" Like in traditional philosophy it should contain at least an
ontology or metaphysics (a theory of what exists in the world and where it comes
from), an epistemology (a theory of how we can know the world around us), and an
ethics or axiology (a system of goals and values that can guide us in our
actions).
Evolution and Constructivism
In addition to the traditional assumptions of systems theory, based on the
principle that organization is more basic than substance, we want to start from
the principle of evolution: systems are not given or fixed, they are the result
of a continuing process during which more and more complex forms of organization
emerge. This evolution does not have a final goal, it is directed only by the
trial and error process of natural selection. Different (re)combinations of
systems are formed by variation, but only those combinations are retained that
are stable, internally and with respect to the requirements of the environment.
The stability of the organization is what turns a mere assembly into a "system".
The variation process may be guided by knowledge acquired earlier, but in its
most basic form it is blind: it does not know where it is going, or which of the
variants it generates will be selected (Campbell, 1974). These principles are
sufficient as a basis for a complete metaphysics, epistemology and axiology, as
will be explained in a further contribution to this workshop.
Such an evolutionary philosophy is also constructive: it assumes that systems can
only be really understood by analysing the process through which they have been
assembled. The variation and selection mechanism continuously constructs new
systems from previous, usually simpler, systems. These building blocks themselves
have emerged from even simpler components, which are the result of combinations
of yet more primitives parts, ... The properties of the system cannot be reduced
to the properties of their components: they can only be understood as results of
the construction process itself, of the specific way in which the components have
been assembled.
In the limit, such a constructive view entails that one cannot be satisfied by a
philosophy which is based on "fundamental laws of nature", "first causes" or
"prime movers", that is to say on fixed foundations beyond further analysis.
Whatever principle or organization is at the base of a construction process, it
is itself merely the result of a previous construction and hence cannot in any
way be ultimate. The only "primitives" that can be accepted in a constructive
philosophy must be so simple as to be empty of organization. All others,
including the fundamental laws of physics, are to be viewed as the result of
evolution through variation and selection, and must be analysed as such. An
example of such an "empty" fundamental is the tautological principle of natural
selection: stable systems remain, unstable systems are eliminated (Heylighen,
1990).
Examples of foundational ontologies are proposed by Newtonian mechanics, which
sees hard, elementary particles moving in space according to deterministic "laws
of nature" as the essence of the world, and by the traditional monotheistic
religions, which see the world as created and governed by the God.
Such ontologies are not constructive: they explain the presence of properties
such as organization, stability, causality, or goal-directedness, by postulating
some unobservable fundamental causes (God, the laws of Nature) which by
definition already have the properties to be explained. In that way nothing is
really explained, the problem is merely pushed one level away, where it cannot be
further analysed. Indeed, in these ontologies it is impossible to ask where God
(or the Laws of Nature) came from, why He is permanent, why He is intelligent,
etc., because these facts are dogmatically or axiomatically established. In that
sense, such a position is not scientific, and we might even doubt to call it
philosophical, since philosophical thought by definition involves continuing to
ask questions.
In this sense, the constructive philosophy we propose is anti-foundational. Yet a
constructive philosophy can be considered foundational in the sense that it takes
the principle of constructive evolution itself as a foundation. This principle is
different from other foundations, however, because it is empty (anything can be
constructed, natural selection is a tautology), but also because it is situated
at a higher, "meta" level of description. Indeed, constructivism allows us to
interrelate and intertransform different foundational organizations or systems,
by showing how two different foundational schemes can be reconstructed from the
same, more primitive organization.
From philosophy to method
The Principia Cybernetica Project is distinguished not only by its philosophy
(the "content" of the project), but also by its method (the "form" of the
project). In accordance with the principle of the self-application of
cybernetics, both form and content will be constructive and evolutionary, based
on the development of higher levels of organization through the recombination of
simpler subsystems and the selection of those assemblies that are more stable.
The development of form and content, of method and theory, will hence occur in
parallel, with a continous feedback from the one to the other, so that each new
principle in the theory will be reflected in the method to further develop the
theory, whereas each improvement in the method will lead to the discovery of new
theoretical principles.
When constructing a cybernetic philosophy the fundamental building blocks we need
are ideas: concepts and systems of concepts. Ideas, similarly to genes, undergo a
variation-and-selection type of evolution, characterized by mutations and
recombinations of ideas, and by their spreading and selective reproduction or
retention (see the contribution of Moritz to this workshop). The basic
methodology for quickly developing a system as complex as a cybernetic philosophy
would consist in supporting, directing and amplifying this natural development
with the help of cybernetic technologies and methods.
It will require, first, a large variety of concepts or ideas, provided by a
variety of sources: different contributors to the project with different
scientific and cultural backgrounds. These contributions must be gathered and
stored in an easy and efficient way. Therefore we must use the most advanced
communication media, in particular electronic mail. The collected information can
then be kept in store on one or more central computers ("file servers") that can
be accessed from anywhere in the network of collaborators. In order to
efficiently find and use the information we need a system that allows the
representation of different types of combinations or associations of concepts.
This can be based on a hypermedia semantic network, with different types of
nodes, containing information in different formats (text, formulas, drawings,
...), connected by links. We further need selection criteria, for picking out new
combinations of concepts, that are partly internal to the system, partly defined
by the needs of the environment of people that are developing the system.
Finally, we need procedures for reformulating the system of concepts, building
further on the newly selected recombinations. Different ways to implement this
kind of interactive structuring and restructuring of concepts in a hypermedia
system will be discussed in the section on "computer support systems".
References
Boulding Ken (1956): "General Systems Theory - The Skeleton of Science", General
Systems Yearbook 1, p. 11-17.
Campbell D.T. (1974): "Evolutionary Epistemology", in: The Philosophy of Karl
Popper, Schilpp P.A. (ed.), (Open Court Publish., La Salle, Ill.), p. 413-463.
Heylighen F. (1988): "Formulating the Problem of Problem-Formulation", in:
Cybernetics and Systems '88, Trappl R. (ed.), (Kluwer Academic Publishers,
Dordrecht), p. 949-957.
Heylighen F. (1990): "Classical and Non-classical Representations in Physics I",
Cybernetics and Systems 21, p. 423-444.
Heylighen F., Joslyn C. & Turchin V. (1991): "A Short Introduction to the
Principia Cybernetica Project", Journal of Ideas 2:1, p. 26-29.
Whitehead A.N. & Russell B. (1910-1913): Principia Mathematica (vol. 1-3),
(Cambridge University Press, Cambridge).
Turchin V. (1991): "Cybernetics and Philosophy", in: Proc. 8th Int. Conf. of
Cybernetics and Systems, F. Geyer (ed.), (Intersystems, Salinas, CA).
Van Gigh J.P. (ed.) (1986): Decision-making about Decision-making: metamodels and
metasystems, (Abacus Press, Cambridge).
von Bertalanfy, Ludwig (1968): General Systems Theory, (Braziller, New York).
____________________________________________________________________________
Cliff Joslyn
Systems Science
State University of New York at Binghamton
General Notes about the Principia Cybernetica Project and Related Initiatives
History of PCP
The Principia Cybernetica project was conceived by Valentin Turchin, a physicist,
computer scientist, and cybernetician. He had developed a cybernetic philosophy
based on the concept of "metasystem transition", and wanted to further elaborate
it in the form of an integrated system with a hierarchical organization,
involving multiple authors.
In 1987, Turchin came into contact with Cliff Joslyn, a systems theorist and
software engineer. Joslyn suggested a semantic network structure using hypertext,
electronic mail, and electronic publishing technologies as strategy for the
implementation of Turchin's ideas for a collaboratively developed philosophical
system. Together they founded the Principia Cybernetica project and formed its
first Editorial Board. They wrote a first proposal, and a "Cybernetic Manifesto"
in which the fundamental philosophical positions were outlined. Joslyn began
publicizing Principia Cybernetica by posting these documents on the CYBSYS-L
electronic mailing list.
This generated a lot of response, including that of Francis Heylighen, a
physicist, cognitive scientist, and systems theorist. Heylighen had been
developing a very similar philosophy to Turchin's and had been thinking along the
same lines of creating a network of people who would communicate with the help of
various electronic media. He joined Turchin and Joslyn as the third member of the
editorial board in spring 1990.
Together they started to further develop their philosophical ideas, partly in the
form of "nodes" and publications, through elaborate electronic mail
conversations, complemented by personal meetings. They continued to attract other
people to the PCP idea through several activities: a sometimes heated public
debate on the CYBSYS-mailing list (winter 1990), a symposium in the context of
the Int. Congress of Systems and Cybernetics (New York, June, 1990), the
distribution of a leaflet, followed by the introductory issue of a newsletter, to
a mailing list containing journals, associations, electronic newsgroups and
individuals active in related domains, and the organization of a workshop in
Brussels. This led to the compilation of a continuously expanding mailing list of
people interested in collaborating in the PCP.
For the moment a specialized electronic mailing list, PRNCYB-L, is being set up
to facilitate the communication among that relatively large group of people.
Heylighen and Joslyn are also experimenting with the development of hypermedia
systems for supporting the development and organization of PCP concepts.
Relation to similar work
The PCP will be situated in the context of general intellectual history (Talmud,
Adler), and the history of systems science and cybernetics. In particular
different attempts to do similar work will be mentioned, such as Krippendorf's
Dictionary of Cybernetics, Singh's Systems and Control Encyclopedia, the work of
Troncale and Snow in the context of the International Society for Systems
Science, the Glossary on Cybernetics and Systems Theory developed for the
American Society for Cybernetics. A brief overview of links to current
development in computer systems (discussed in more depth in the section on
computer support systems) will be given.
____________________________________________________________________________
Valentin Turchin
Computer Science
City University of New York
A Tentative Sketch of the Starting Nodes of PCP
I first discuss the methodology of the construction of a system of nodes where
both the contents of nodes, and their relation and organization are tightly
interrelated. I propose to use the principle of stepwise formalization, on which
the whole edifice of science is built. In science we start with intuitive and
often imprecise concepts and on this basis create new models of the world which
are more formalized and more precise. Formalization may go in rounds, or levels,
becoming more intensive and extensive. Finally we reach a stage at which we
reinterpret those intuitive concepts that were taken for granted at the beginning
of the construction. Thus a clock, with its hands, becomes a structure of
elementary particles.
This, however, does not make unnecessary the usual notion of a clock, as well as
all other simple words we use in explaining physics. This is a hierarchy of
pictures of the world, where there are unbreakable ties between levels. Take
"simple" notions away, and the whole edifice will crumble. This method can be
referred to as the method of step-wise formalization.
I propose, therefore, that the nodes we are writing will be initially organized
according to the usual notion of their conceptual dependency understood
informally or semi-formally (the whole-part relation is also included, of course,
as a reason for siblings). As the collection of nodes grows, we give more time to
the work on formal semantics and the structuring of this accumulated material.
In this talk I present the result of my first attempt to sketch some basic
conceptual nodes of the Principia Cybernetica Project. Needless to say, it is
very imperfect, a very rough first outline of the top level of the system. But it
should allow us to start a discussion--and work. The present abstract includes a
list of nodes with their references up and down. For some of the nodes a brief
exposition of contents is provided.
node: PRCYB
Principia Cybernetica
references up: none
references down:
INTRO> Introduction
FORMAT> The format of the material
MAIN> The main node
REACT> Reactions, discussions, comments
NETWORK> Network of people
This is the head node of the system to which you address when you start examining
Principia Cybernetica. The head node includes an introduction INTRO, the FORMAT
describing the organization of the material, the main node MAIN which contains
subnodes which actually define our philosophical system, and the node REACT which
contains reactions to our work and discussions around it.
node: MAIN
The main node of Principia Cybernetica
references up:
PRCYB> Principia Cybernetica
references down:
KNOW> Knowledge
WILL> Will
FUTURE> Future
The contents of Principia Cybernetica follows the formula:
Our knowledge + Our will = Our future.
In our thought and language we distinguish two different classes of elements
about which we say that they exist: those expressing what we know, or think we
know, and those expressing what we are striving for and intend to do. We unite
the elements of the first class referred to as KNOWLEDGE, and the elements of the
second class as WILL. They are not isolated from each other. Our goals and even
our wishes depend on what we know about our environment. Yet they are not
determined by it in a unique way. We clearly distinguish between the range of
options we have and the actual act of choosing between them. As an American
philosopher noticed, no matter how carefully you examine the schedule of trains,
you will not find there an indication as to where you want to go.
We think about knowledge as a representation of the world in our mind.
Representation is the term used by Schopenhauer; the world for him is Will and
Representation.
Another way to describe the relation between knowledge and will is as a dichotomy
between not-I and I, or between object and subject. The border between them is
defined by the phrase "I can". Indeed, the content of my knowledge is independent
of my will in the sense that I cannot change it by simply changing my intentions
or preferences. On the contrary, I can change my intentions without any
externally observable actions. I call it my will. It is the essence of my 'I'.
The origins of this approach to all that exists are cybernetical. We try to
understand ourselves by building cybernetical creatures which model intelligent
behavior. The model of intellect such a creature has consists of two parts: a
device that collects, stores and processes information; and a decision
taker--another device that keeps certain goals and makes choices in order to
reach these goals, using the information from the first device. Thinking about
ourselves in those terms we speak about knowledge and will.
node: KNOW
Knowledge
references up:
MAIN>
references down:
EPISTEM> Epistemology: what is knowledge?
METAPHYS> Metaphysics: what is the nature of things?
CYBER> Cybernetics
MATH> Mathematics
NATURAL> Natural sciences
The first part of knowledge is, logically, the knowledge about knowledge itself:
what is knowledge? This part is known as epistemology (EPISTEM). Metaphysics,
informally, should answer to the question: what is the nature of things? Attempts
to understand this question in a more formal way and to give a satisfying answer
produced volumes of philosophy. We treat this problem from our cybernetical
positions--see METAPHYS. We divided the whole sum of exact sciences into
cybernetics (including the theory of evolution), mathematics and natural
sciences. It may come as a surprise that there is no place for humanities in this
node. They are found in the node Will. This does not mean that humanities do not
constitute knowledge--they certainly do. All the texts in Principia Cybernetica,
as any texts, represent knowledge; only actions, not texts, represent will. The
titles of our nodes must be understood as knowledge about human knowledge, and
knowledge about human will. Humanities, as one can see from the word itself, deal
with manifestation of human will.
node: EPISTEM
Epistemology: what is knowledge?
references up:
KNOW> Knowledge
references down:
MEANING> Meaning
TRUTH> Truth
THEORIES> Theories versus facts
SEMANT> Semantics
HEPIST> Historical review of epistemology
In cybernetics we say that a purposive cybernetic system S has some knowledge if
the system S has a model of some part of reality as it is perceived by the
system. But what is a model? The most immediate kind of a model is a device that
implements the concept known in mathematics as homomorphism. After some
generalization we come to the formula: a piece of knowledge is a hierarchical (or
recursive) generator of predictions.
node: METAPHYS
Metaphysics: what is the nature of things?
references up:
KNOW> Knowledge
references down:
KASCHO> Introduction. From Kant to Schopenhauer.
ACTION> Action
FREEDOM> Freedom
GOD> God
SEMANT> Semantics
A metalanguage is still a language, and a metatheory a theory. Metamathematics is
a branch of mathematics. Is metaphysics a branch of physics? We argue that, in a
very important sense, it is.
node: MEANING
Meaning
references up:
EPISTEM> Epistemology
references down: none
Our definition of knowledge allows us to further define meaning and truth. When
we say or write something we, presumably, express our knowledge, even though it
may be hypothetical. Thus to be meaningful, a proposition must conform to the
same requirement as a piece of knowledge: we must know how to be able to produce
predictions from it, or produce tools which will produce predictions, or produce
tools to produce such tools, etc. If we can characterize the path from the
statement to predictions in exact terms, the meaning of the statement is exact.
If we visualize this path only vaguely, the meaning is vague. If we can see no
path from a statement to predictions, this statement is meaningless.
node: TRUTH
Truth
references up:
EPISTEM> Epistemology
references down: /* to cybernetic foundation of math. */
A piece of knowledge is true if the predictions made by the user of knowledge on
the basis of this knowledge come true.
node: THEORIES
Theories versus facts
references up:
EPISTEM> Epistemology
references down: none.
node: SEMANT
Semantics
references up:
EPISTEM> Epistemology
METAPHYS> Metaphysics
This node starts a hierarchy defining the meaning of the most general concepts
used in philosophy and science. People usually take them for granted. We want,
however, to define them as precisely as possible, and to derive their necessity
from the basic principles of epistemology and metaphysics. Success on this way
would confirm the validity of our epistemology and metaphysics. The main
principle we follow is this: our definition of meaning is tied to the concept of
modeling: the homomorphism picture. Therefore we should start any attempt to
formalize semantics with an analysis of various aspects of that picture and
various types of such pictures.
For the time being I do not break this node into subnodes. This is left for
future. At present, the following concepts have been (tentatively) analyzed and
defined:
1. State, physical and mental
2. Internal and External Knowledge.
3. Causality
4. Abstraction.
5. Prediction
6. Space and Time.
7. Observation
8. Object
9. Process
By the time of the conference I plan to write the following nodes:
* knowledge about knowledge (model of a model)
* real time versus model time
* historic record (the meaning of historical statements).
node: KASCHO
Introduction. From Kant to Schopenhauer
references up:
METAPHYS> Metaphysics
refrences down: none.
node: ACTION
Action, the ultimate reality
references up:
METAPHYS> Metaphysics
references down: none
Am Anfang war die Tat (Goethe.)
Will is manifested in action. If we are looking for the ultimate undoubted
reality of physics, we must turn to action, and not to the space-time picture of
the world. For a picture is only a picture, while action is an irrefutable
reality.
An action is a result of a free choice. The state of the world defines (or rather
is defined as) the set of feasible actions for each will. The act of will is to
choose one of these. We learn about action through our representations, i.e. our
knowledge about the external world.
____________________________________________________________________________
Donald H. McNeil
Developmental Systemologist
Philadelphia
The Principia Prospect
After more than fifty years the 'systems movement' still cannot agree upon its
subject matter or promulgate a coherent theory. As yet there is not even an
operational definition of 'system' acceptable to the majority of stakeholders.
The idea of a Principia project to remedy this disgraceful situation has merit.
Nonetheless, it matters how the project is put together, what it strives to do,
and why.
In the early days of 'systems thinking', people identified systems with forms and
formalisms, hence the emphasis on mathematics and hierarchies and structuralism
and--more recently--the 79 isomorphies of the ISSS. An alternative movement has
placed its emphasis on 'function', 'producer-product', operations research, and
living systems. At the same time, some independent thinkers have centered their
work on the fluzes of change and the flows of 'substance' as the essences of
systems. All the while, the notion that systems could best be understood through
cybernetics has attracted a well-organized following. The Principia Cybernetica
Project seems now to be allied with this latter view.
It is perhaps appropriate that distinct 'schools' of systems thinking have
divided themselves rather neatly into camps representing the four fundamental
aspects of any system: form, function, content, and control. These four, together
with the timing which inter-relates them, can be woven into the fabric of a
General Theory of Systems. A partiality to one aspect, e.g. control, may skew but
cannot completely subsume the other aspects. In a systemological schematic such
as that in the figure below, [picture deleted] the CONTROL aspect is represented
with emphasis but still in balance with the other three.
Even to make the simple drawing above presupposes a considerable amount of
generalized systemological conceptualization. The Principia Cybernetica Project
as currently described in its Newsletter #0 itself presupposes rather a lot about
cybernetics, systems, and philosophy. It is therefore of the utmost importance
here to think systemologically about the place, then meaning, and the method of
the Project. In these early stages, a Principia still has a chance to reflect
upon itself and to establish its mission accordingly. Let us proceed through the
highlights of the Project so as to clarify its positions and their relationship
to a systemological worldview.
Reference
Donald McNeil, "Systemology: The Fundamentals of a General Science of Systems",
1981. Masters Thesis, U. of Pa.
_________________________________________________________________________________
CYBERNETIC FOUNDATIONS
____________________________________________________________________________
Gordon Pask
CICT/OOC
University of Amsterdam
The Foundations of Conversation Theory, Interaction or Actors Theory, all Cybernetic and
Philosophically so
(or 'Some Foundations of Principia Cybernetica')
The intention of writing this paper is to show, albeit in a blinkered and limited
manner, that a philosophy of Cybernetics, encapsulated in the journal title,
"Principia Cybernetica" is not only justifiable, but necessary and in this day
and age, utterly essential. Have no qualms, it IS and that statement IS
significant. We need only to look out at this world, lying amongst many others of
similar and different kinds, to recognise this fact and, in doing so, to see that
factuality is fragile, parading, like a circus-procession, or a civic, mayoral
one with a lady drum-majorette in front, out of mind, into thought, from that
into utterance and inscription in words and text books.
One route of demonstration is by way of argument, to see that "proof", for
example, is a convention and not something sacrosanct. In order to grasp this
point, it is necessary to accept the existence of certain deeply embedded
disciplines, momentarily, at least. That these strange distinctions are fallible
becomes evident, yet they are hard to dispel, though they must be dispelled, with
the exception, perhaps, of the entrenched establishment of logicians,
mathematicians, psychologists and others. In such cases one is more likely to do
more harm than good, for their practitioners, logical text-book-writers,
mathematicians and others are folk who have invested much in terms of effort or
sheer labour, who wish to retire and, most surely, wish to do nothing which is at
all novel, nothing that might disbalance the boat of their nicely equilibrial
status quo.
This mode of argument can be given, for instance, but rather minimally so, by
comparing and contrasting "proof by reductio ad absurdum" (which calls paradox a
tautology, or a contradiction or else pushes it under the carpet, as disturbing
the status balance) with the proof form (I prefer demonstration), often known as
"productio ex absurdo", which uses the interesting fact of paradox as the
enticement to creative thought, new theorems, new ideas. These are minimal forms,
it would not be difficult to cite one hundred or one thousand or any, possibly
countable, number of more complicated, more illuminating, others.
Another method is, of course, by means of force majeur, of missiles and mortars,
taken as a duly academic metaphor. In order to pull out the plug in the bathtub
of science and philosophy, to empty the tub of bathwater without disposing of the
baby, also.
So, what do we do?? With good reason, after much contemplation, I submit that the
bathplug is called "time" and that the baby we retain is called "innovation",
creativity if you prefer that term. Of course, the operation is possible and
bound to succeed. It is, however, apparently destructive and I simply hate
destruction or demolition of any kind. As a result I have a preference for a
milder and more subtle approach, showing the multiplicity, the plurality of time
and the many facets of innovation in the slightly more restricted field, still
thoroughly Cybernetic and Philosophical, of Conversation Theory, Interaction of
Actors Theory, and the protologic or protolanguage which they share, Lp, by name.
In the sequel, this is the line pursued.
Notice, all the same, that the basic and partly enunciated theme may be expanded,
like a hydrogen balloon, into the entire extravaganza. Let us be clear about this
much, at least. We speak of theories, namely, C.T. (an abbreviation of
"Conversation Theory") of I.A.T. (an abbreviation of "Interaction of Actors
theory" and a deliberate word play or pun upon the much popularised I.A.
theory-or-not, but spawned from, rather than being the ancestor of, Cybernetics
itself) . There are more valid surrogates for Cybernetics under any label you
elect to take up as your own, freely chosen, particularity; there have been many,
like Bionics, Information Science, General and Special System Theory, heaven
knows what else; you choose whichever you like, it matters not a tittle or jot.
What I shall say, here summarise, remains invariant under any choice whatsoever.
Some 40 years past, in the context of the theatre, the laboratory and academia as
viewed by a research assistant, it became evident, to me at any rate, that the
search for a"scientific psychology "or a social science was but a fruitless
endeavour if we persisted in those still prevalent habits of apeing the
scientific by applying "scientific methods", like statistical techniques, to
entirely inappropriate data. In place of that, our group proposed and pioneered
several other frames of reference, anchored, for credibility, so far as possible
upon the existing paradigms adopted by science.
What does it mean to have a "Scientific Psychology", or to have a "Science of
Society"; that is, over and above the pseudo-sciences of inappropriate data,
smudged-splodged into a format which is superficially compatible with sciences
where the real data can, for example, be treated statistically as well specified
and independent event reckonings? Clearly, dependencies may exist. Clearly, also,
such more liberal dependencies can be accounted, if they ARE so simplistic, in a
manner not dissimilar to the bookkeeper's ledger, to be dealt with by accountants
and actuaries and the like. Lamentably, for some, or joyfully for most of us,
neither psychological nor social events; call them mental events, are NOT so
simple and cannot be recorded or manipulated in the ways suggested.
So what, being Cybernetists, are the scientific foundations of the mental events
with which we are so often, some of us most frequently, apt to deal? What
justifies the scientific flavour of the appellation "Principia", as in Newton's
or Russell's "Principia"? There are, of course, many possible replies to this
rhetorical question but, in this paper, I develop only one of them.
Let it be taken for granted, (failing which, you are welcome to a tedious but
more-or-less irrefutable demonstration of the fact), that Cybernetics is a
coherent and cohesive theoretical structure, this, in particular, being the case
for the so-called New-Cybernetics. Further, let it be taken for granted that
Cybernetics, a fortiori the New-Cybernetics and (not so much the System-Thinking
stuff) is sufficiently distinct to have an identity of its own, even though it
promotes interaction, itself, and may be regarded as positively engendering
interaction between superficially disparate fields.
So it appears as a coherent and cohesive system of analogy, of metaphor but
strict metaphor, designating analogies in which the similarities and the
differences are well specified. One asks, quite naturally, why this should be
deemed a"science" with pretensions to having firm "principles", rather, for
example, than an art or a philosophy or the logical aspect of a theology.
On this score, of being definitively "scientific", I am not so deeply convinced
as I am, dogmatically so, of the plain fact that Cybernetics most surely DOES
have PRINCIPLES which, for all their global breadth, maintain integrity. Perhaps
that is because I am not so convinced about any significant differences between,
say, science and art, believing that they must coexist together if either one or
the other is to make sense. However, it can be strongly argued that the
principles of Cybernetics resemble those of such disciplines as physics, biology,
cosmology, chemistry, molecular biology, microphysics, archeology, social
anthropology and geology. If the ossature of these disciplines is deemed to be
scientific, then, presumably, Cybernetics is scientific.
Upon these slightly tenuous grounds, let us survey some of the structural
similarities at hand. The list is by no means exhaustive at this moment, and it
is evolving.
(1) In real, rather than school-science, we seek appropriate hypotheses and data,
entities over which practitioners may agree or agree to disagree and know why
they do so. Admittedly, in school-science, many of us were TOLD that the testable
hypotheses emerged from great theories and that some even greater theory will be
revealed, but not until next year. Also, most of us were TOLD that scientific
data are repeatable, objective, causally mapped in progression as objects and,
later, unmentionable until next term, events.
Admittedly, if we were fools enough to accept these half true falsities as
anything other than the infrastructure of an elaborate, even if cost effective
examination process, then we may still entertain deeply ingrained concepts of an
unduly naive picture of science. But real and mature science is not, at all, like
that. Quite obviously, SOME, not ALL, data of reaction kinetics are inappropriate
to psycho-social-mental events. The search for "hard" scientific data branches in
different directions, appropriate to the field of enquiry, plural in any field of
enquiry. Thereby, hypotheses are posed, formulated, tested by appropriate data,
inductively verified or deductively falsified and theoretical structures erected.
Cybernetics admits, maybe preaches, all this. It also asserts that the kind and
the truth functional modality of the logics underpinning science are varied, like
the appropriateness of the domains which their logics generate. Fundamentally,
they are logics of many-sorted coherence, of many-sorted distinction, of self
reference and other reference, all of them are dynamic. The nowadays standard
Aristotelian view, is not denied. It is reified, locally, in those regions of a
manifold where there are valid metric space type representations. In this
respect, at least, the structure of Cybernetics resembles the structure of
science.
(2) It may be demonstrated, with passable elegance, that Cybernetics shares, with
science, certain skeletal principles. In many places, at least, these skeletal
attributes lie in one to one, isomorphic, correspondence. In other places, the
correspondence is, more likely, homomorphic and in others, it may only be
expressed by the category theoretic relations of functors between categories, or
their topological equivalents. However, so far as I know, there is no basic
dissonance. Amongst the principles involved are conservation, complementarity,
duality, exchange relations, parity, symmetry and symmetry breaking, uncertainty,
indeterminacy and the obvious mathematical or metamathematical properties of
distinction, of knottedness, of singularity in contrast to continuity, of various
types of demonstration, some being proof theoretic and others not so. To these it
is necessary to add a few others such as the void, the not void, the self and the
other. Science, itself, might benefit by their proper inclusion within its orbit.
Thus, in the classical sciences, we commonly revere the conservation of mass and
of energy, under the elegant equivalence of E = mc^2, c being the limiting
velocity of light, E = Energy and m = Mass, we have, in Cybernetics, several
conservations such as application (of procedures, complementary to products), of
procedures acting upon procedures (to produce and incidentally, reproduce them),
of meaningful information transfer and of distinction. For sure, they are not so
neatly related. But that is hardly surprising, once you keep in mind the scope of
Cybernetics which is so much more encompassing than that of classical science,
for instance, adumbrating scientists and the theories they develop.
The foregoings notions are intended to illustrate an evolutionary trend in
Cybernetics with which I am, personally, very familiar. The train of thought
could be extended further backwards and forwards (though "backwards" and
"forwards" are terms up for question in that self-same framework). However, using
these terms in the common language sense, (stripped, that is, of particular
formalities), I shall try to go by interpolation and by extrapolation in each
direction especially into what is, often brashly, called the future and in
serious Cybernetic Discussion, open to serious discussion.
____________________________________________________________________________
Lars Löfgren
Department of Information Theory
University of Lund, Sweden
Foundational Issues Addressed by Cybernetics
With its wholistic aims and its understandings of self-reference, cybernetics
addresses issues which have proved foundational not only for sciences and
information technologies but also for cybernetics itself.
Within the hard sciences, like physics, foundational issues concern justification
problems. Which in classical physics are resolved by observation and
measurement-with a belief in a "detachable observer". In quantum mechanics, which
includes the measurement process, the justification problem becomes severe,
requiring a more abstract form of justification in terms of knowledge of
observability versus definability.
In cybernetic studies of knowledge of knowledge processes, the insight is gained
that such knowledge must be relativized to language, and that the "detachable
observer" be changed into a thesis of a "non-detachable language". This enforces
the autological predicament--to conceive of language in language--for which a
resolution in terms of a complementaristic conception of language has been
proposed. The complementarity may be conceived from various views. One is as a
tension between describability and interpretability within a language. Another,
in terms of degrees of partiality of self-reference within a language (where the
impossibility of a complete self-reference is synonymous with the
"non-detachability of language"). In cases where an object language has a
metalanguage, the complementarity of the object language is describable in the
metalanguage (but not in the object language). The complementarity is then said
to be transcendable, and the self-reference problem, that of describing a
language in the language itself, is "unfolded" (a characteristic cybernetic
justification of self-reference).
In particular, we discuss the Bohr-Pauli dialogue on a detachable observer, and
suggest complementaristic linguistic models for the self-referential measurement
problem in quantum mechanics.
We also suggest such linguistic models for the foundational problems of
probability theory, namely of how to conceive of models for probability--which,
as has been observed in particular for Kolmogorov's axiomatic approach, are not
describable within the theory. We attach to Josephson's view, concerning
strategies of science towards form versus meaning, that "the technique of
statistical averaging is especially irrelevant in the context of meaning, since
its influence in general is to transform the meaningful into the meaningless".
The problem of induction, foundational for most sciences, obtains a natural
explanation in the complementaristic conception of language. We suggest that
quests for inductive inferences of general laws from particular observations are,
and will forever be, in vain. Instead, an inductive inference is conceived as a
linguistic (mostly unconscious) process which utilizes not only particular
observations but also properties of the language which are beyond describability
(hidden) in the language itself. Thus, to be able to describe induction, as it
occurs in a language, we must have access to a metalanguage in which the object
language is describable. In reality, languages are themselves not produced from
descriptions, but are evolved.
The foundational problem of describing evolution, in biology as well as in
epistemology, is again conceived in terms of the complementaristic conception of
language--this time genetic language. In particular, we are able to give a
metamathematical argument for the higher force of an evolutionary process than
that of a planning process based on inductively generated descriptions in a
scientific language.
The impossible task of aiming at a complete description, in some language, of the
biological process of evolution is as we know replaced by aiming at less
ambitious goals. To a certain extent such goals can be analyzed in terms of goals
on higher levels. But the impossibility of reaching a complete description,
enforces a goal hierarchy with ultimate goals that are exempt from scientific
analysis--like ethical goals.
We analyze Moore's Principia Ethica and his concept of "naturalistic fallacy". In
particular we illuminate fallacies in trying to base ethics on evolution.
Reference
Löfgren, Lars (1991): "Complementarity in Language; toward a general
understanding," in Carvallo, M (ed.) Nature, Cognition and System II, Theory and
Decision Library, Dordrecht-London-Boston: Kluwer, 73-104.
____________________________________________________________________________
Ranulph Glanville
University of Amsterdam and Portsmouth Polytechnic
Excavation and Underpinning
Foundation and Building
In a number of earlier publications, I have examined both the nature of
fundamentals (in a belief system or thesis), and some of the fundamental concepts
of cybernetic systems (such as control, communication, variety, responsibility,
distinction, recursion and re-entry), especially in the light of, and as
generating, second order / the new / the cybernetics of Cybernetics.
In this paper I shall systematically consider the intension and extension of
other fundamental concepts from (especially Ashby's) early writings in
Cybernetics, both to consider of what they are made, and upon what they rest, and
to see how this casts them in a new light, particularly in view of the insights
we have gained in and through second order / the new / the cybernetics of
Cybernetics.
Thus, to use the architectural metaphor implicit in the title (the firmnesse of
Webb's original translation into English of Vitruvius's classic definition of
architecture--firmness, comodotie and delight), I examine the
Excavation and Underpinning
Foundation and Building
of cybernetics.
____________________________________________________________________________
Gertrudis Van de Vijver
Senior Research Assistant NFSR
University of Ghent
Error: Epistemological Options in Cybernetics
Error plays an important role in the ascription of teleological properties and
capabilities to systems. It is possible, on the basis of the meaning and the
place of error, to trace out the history of purposiveness. We aim at doing this
in going through the different cybernetic stages--the cybernetics of the first,
the second and the third order--and through the theories which were inspired by
cybernetics--connectionism and neo-connectionism--.
In first order cybernetics, and in most A.I. views, error is to be interpreted in
terms of the dysfunctioning of systems. Goal-directed behavior is always to be
interpreted on the basis of a 'goal-deficiency' model. Difficulties of the
missing goal-object, problems of circularity between the goal and the relevant
behavioral properties, arise in this context.
In an attempt to model certain properties of complex purposive systems, it became
clear that the possibility to behave in an erroneous way had to be build in. The
possibility of error is in this case linked with the possibility of building up a
representation in an inductive way. It is also brought in connection with a
relation of under-determination existing between a behavior (an idea, a theory)
and certain conditions preceding it.
How will the possibility to behave erroneously in this case be evaluated ? How
shall we make the possibility of going through a history, a history that is
characterized by an under-determination, into an integral part of an artificial
system ? What are the epistemological consequences of this ? We are confronted
here with specific epistemological difficulties which have to do with the
knowability of autonomous or self-organizing systems.
_________________________________________________________________________________
EVOLUTIONARY PHILOSOPHY
____________________________________________________________________________
Valentin Turchin
Computer Science,
City University of New York
Metasystem Transition as the Quantum of Evolution
Consider a system S of any kind. Suppose that there is a way to make some number
of copies from it, possibly with variations. Suppose that these systems are
united into a new system S' which has the systems of the S type as its
subsystems, and includes also an additional mechanism which controls the behavior
and production of the S-subsystems. Then we call S' a metasystem with respect to
S, and the creation of S' a metasystem transition. As a result of consecutive
metasystem transitions a multilevel structure of control arises, which allows
complicated forms of behavior.
In my book [1], I show that the major steps in evolution, both biological, and
cultural, are nothing else but metasystem transitions of a large scale. The
concept of metasystem transition allows us to introduce a kind of objective
quantitative measure of evolution and distinguish between evolution in the
positive direction, progress, and what we consider an evolution in the negative
direction, regress. In the present paper I outline the main ideas of this book,
and concentrate, in particular, on one of the aspects of biological evolution:
the appearance of human thinking.
When we speak of cybernetic systems, we can describe them either in terms of
their structures, or phenomenologically, in terms of their functioning, their
behavior. We cannot claim at the present time that we know the structure of the
human brain well enough to explain thinking as the functioning of that structure.
However we can observe evolutionizing systems and make conclusions about their
internal structure from a phenomenological description of how they function.
From the functional point of view the metasystem transition is the case where
some activity A, which is characteristic of the top control system of a system S,
becomes itself controlled as a metasystem transition from S to S' takes place.
Thus the functional aspect of metasystem transitions can be represented by
formulas of this kind:
control of A = A'
When a phenomenological description of activities of some systems fits this
formula we have all reasons to believe that this is a result of a metasystem
transition in the physical structure of the systems. Here is the sequence of
metasystem transitions which led, starting from the appearance of organs of
motion, to the appearance of human thought and human society:
control of position = movement
control of movement = irritability (simple reflex)
control of irritability = (complex) reflex
control of reflex = associating (conditional reflex)
control of associating = human thinking
control of human thinking = culture
In [1], I show how the most characteristic features of human thinking: creation
of tools, imagination, planning, overcoming the instincts, understanding of the
funny and the beautiful, creation of language, self-knowledge, can all be
understood as control of associating and its direct consequences. Then the
principle of metasystem transition is used for an analysis of cultural evolution,
and first of all, the development of science. We see that in the history of
science, as well as in the history of biological evolution, the major steps
forward are done through metasystem transitions. Looking even farther, we can try
to guess (and at the same time influence) the more remote stages of the evolution
of the mankind.
Reference
[1] Turchin V. (1977): "The Phenomenon of Science" (Columbia University Press,
New York)
____________________________________________________________________________
Cliff Joslyn
Systems Science
State University of New York at Binghamton
Control Theory and Meta-System Theory
This paper, as part of PRINCIPIA CYBERNETICA, is intended to be integrated into
the structure of that project. Therefore, we note potential links to the
following nodes:
Action, Behavior, Constraint, Constructivism, Control, Control System, Dreaming,
Dynamic Equilibrium, Emergence, Equilibrium, Evolution, Feedback, Freedom, Goal,
Hallucination, Hierarchy, Imagination, Intention, Knowledge, Life, Memory,
Purpose, Selection, Self-Organization, Stability, Thought, Variation, Will
The 1970's produced (at least) two great cybernetic meta-theorists: Valentin
Turchin and William Powers. In The Phenomenon of Science [TUV77] and Behavior:
The Control of Perception [POW73], respectively, they provide grand biological
and psychological theories resting on common principles: that evolved organisms
are hierarchically organized belief-desire control systems; that these cybernetic
systems are involved in cyclical modeling relations with their environments; that
blind variation and selective retention is a universal mechanism of both
biological and non-biological evolution; and that a consequence of these views is
that human freedom is necessary for social evolution.
While Turchin and Powers differ on the nature of control, and particularly the
origin of control systems, they share the great majority of a theoretical core.
Much of their theories are not unique in Cybernetics and Systems, or in general.
Indeed, the key aspects of their theories (e.g. the use of "hierarchy", "control"
and "purpose") are central to all of Cybernetics and Systems (e.g. [ASR52,
ASR56]). But in their work, these ideas have been developed in conjunction, and
have been successfully extended to produce elegant, consistent, general theories
of living systems in the context of Cybernetics and Systems theory.
In this paper we will examine Powers' Control Theory [POW73, POW89]. We will do
so from the perspectives of: Turchin's work, as expressed in the works of the
PRINCIPIA CYBERNETICA project to date--with which we assume the reader is
familiar [HEF90f, HEFJOC90, JOC88e, TUV77, TUV81, TUV87a, TUV90 ,TUV91a, TUV91b,
TUVJOC90]--and which we will call (for want of a better term) "meta-system
theory"; and the wider theories of evolving systems as developed by the
Cybernetics and Systems disciplines.
Powers' Control Theory
Powers' central thought is simple: all living systems are hierarchically
organized negative feedback control systems, where "feedback control system" is
essentially the same concept as that used in Control Engineering for the design
of regulatory mechanisms [MAO70, WIN48]. Thus, as in classical Cybernetics, the
simplest regulatory mechanisms, like thermostats, are prime examples. However,
Powers' intent is to claim that this theory of machines has universal
applicability to organisms. Thus control theory is an attempt to return
Cybernetics and biology to each other, as Cybernetics has lost biology for
engineering; and even the most sophisticated forms of theoretical biology
[EIMSCP79, NIGPRI89, VAFMAH74] have lost all concept of fundamental control
mechanisms as being the essence of life.
The following is an extremely terse outline of Control Theory.
Definition of Control
Control of an entity requires constraint, that is a selection or reduction in
variety of the possible states q of that entity. Typically, the exercise of
control will reduce the variation of possible states q to one, thus determining
the final state q* of the system.
But constraint is not sufficient for control. Constraint and determination
results from a variety of situations, many of which are not control, the primary
of which are stable equilibria. For example, supply-demand interaction in markets
stabilizes prices, but Adam Smith's "invisible hand" does not "control" the
market. Rather, control requires a constant and on-going interaction of the
controller with the controlled entity, such that continued constraint results in
sufficient stability around a state q* (or another kind of attractor) despite
perturbations and disturbances. Thus systems maintained at an unstable
equilibrium, such as an inverted pendulum or balanced broom, are exemplars of
control systems. Thus we arrive at the definition of control as offered by Rick
Marken:
A controlled event is a physical variable (or a function of several variables)
that remains stable in the face of factors that should produce variability.
[MAR88]
Control as a Phenomenon
Since any dynamical system which is being maintained at a state (or attractor)
which is out of equilibrium is under control, control theory legitimately
encompasses a great swath of current interesting work in Cybernetics and
Systems--in particular: most of the so-called "self-organizing systems" theories,
"far-from-equilibrium" physics [PRINIG72], synergetics [HAH78], and those
biological theories which focus on "metabolic" definitions of life [SC67].
Powers' Control Systems
The classical feedback control system is described by Powers as a
"stimulus-response", or S-R feedback controller. The topology of the system is a
throughput device with two inputs and one output, and an internal loop. A
feedback control system of Powers' design has the topology of a whole closed loop
with two inputs, the environmental disturbances and the reference level, and no
outputs.
Powers uses the following terminology:
Physical Quantity: That aspect of the environment whose variation is eliminated
in the face of disturbances, as in our definition.
Disturbance: Environmentally induced fluctuations of the physical quantity.
Output Function: The action, or behavior of the control system.
Error: Signal internal to the control system which directs behavior.
Comparator: Determines whether the perceived variable matches the reference
level, and generates an error signal if it does not.
Perceived Variable: The "appearance" of the physical quantity to the control
system. In neural organisms this is a "sensation" or "perception".
Reference Level: Similar to the role of the set point in an S-R controller. This
signal represents the controlled state of the perceived variable, or that state
of the perceived variable which produces no error.
Input Function: How the physical quantity is transduced in the control system
into the perceived variable.
Powers' argues that his view is superior to the S-R model in that the closed
environmental loop of his model is always implied in an S-R model. S-R models
purport to be control systems, but lack the controlled quantity, the entity whose
variation is eliminated despite environmental disturbance, and is in the
environment.
Constructivist Epistemology
Powers adopts a revolutionary view of control, in the context of constructive
epistemology, through two steps. First, we note that the controlled quantity is
in the environment, and assert as false the traditional control theoretic idea
that the output (behavior, action) of the system is the quantity under control.
The variation of action is rather large, on the same order as the variation of
the disturbance, and of opposite magnitude, in order to cancel out the effect of
the disturbance on the controlled quantity.
Second, we note the necessity that the input function mediates the appearance of
the environment to the comparator. We have to say that for the control system,
aspects of the environment only exist to the extent that corresponding input
functions exist, and we can effectively say that perceptions are the environment
for the control system. Assuming that the input functions are "good"--in the
sense of providing a relatively strong homomorphic mapping or model of the
environment, albeit of selected aspects--then when the variation of the physical
quantity is eliminated, then so is the variation of the perceived variable. Thus,
control of the physical quantity results in effective control of the perceived
variable.
Since output is not controlled, and even the physical quantity is not controlled,
since the physical quantity only exists for the organism in virtue of mediation
through perception, we arrive at the revolutionary idea that it is the perception
that is in fact controlled; that, for the organism, it is the input which is in
fact the controlled quantity. Thus, the novel title of Powers' first book
[POW73]: it is not, as the received behaviorist tradition would have us believe,
that perceptual stimuli allow an organism to correctly control its behavior, but
rather the organism's behavior which allows it to correctly control its
perceptual stimuli.
Hierarchical Control
Control systems are hierarchically nested when the output function of a "higher"
control system does not affect the physical quantity, but rather serves to set
the reference level of a "lower" one. The higher level system has the lower level
as its environment, and, speaking loosely, controls the lower level system. The
multi-level control system has the same topology as the single level: a closed
loop with two inputs. The lower level system must necessarily act at a faster
temporal scale than the higher level.
Powers identifies nine levels of hierarchy in human control systems, which can be
outlined according to the following schema:
"Systems concepts" require control of principles;
which require control of programs;
which require control of relationships;
which require control of sequences;
which require control of transitions;
which require control of configurations;
which require control of sensations;
which require control of intensities.
A "systems concept" is a unifying conceptual and ideological system of thought,
such as a religion, or the "scientific method". The hierarchy extends downward
towards more specific perceptual categories, since in control theory it is
perceptions that are controlled, not actions.
Learning and Organization
Learning and change is provided by introducing a meta-control system which stands
aside the entire control hierarchy. While the perceptual control hierarchy acts
in real time, and is the result of learning, this second, "organizational" layer
acts on the perceptual layer over a longer time than the behavior, and affects
changes on the perceptual control system--in short, learning. This
"organizational system" is genetically innate, and unchanging itself.
The entire perceptual control system is stimulated by and acts on the
environment, but the environment also makes physiological affects on the
"intrinsic", or physiological state of the organism. Genetically determined
structures produce intrinsic perceptual signals, such as hunger, thirst, lust,
and pain; but also emotions such as satiation, satisfaction, joy, anxiety, etc.
Either positive or negative signals could require learning, to increase or
decrease the intrinsic perception respectively. This is mediated through an
intrinsic error signal. Output of the organizational system is directed at the
perceptual control system, and produces change in it. This change can be either
random, blind variation, or somewhat directed (meta-learning). In either case, a
null intrinsic error level results in selection and retention of the new
configuration.
Memory and Imagination
Powers asserts that memory is uniformly distributed not only among all levels of
the control hierarchy, but also in each control system at each level. Thus, a
sensational-level control system might "memorize" a color; while a program-level
control system might "memorize" a Bach etude. The simpler model of the control
system is now modified so that memory is addressed from the output of an
upper-level control system, and provides the immediate reference signal to the
control system. Perceptions in turn are stored in that memory.
The final additions are two switches on both the reference signal and the
perceived variable. Each switch can be either off or on. When the perception
switch is on, perception proceeds normally; when it is off, it is the memory
signal which is transferred to higher levels. When the memory switch is on,
action proceeds normally; when it is off, action is disabled, but a signal of the
memory is transferred to the perceptual signal, if it is prepared to receive it.
There are four cases:
Input On Off
On Control Automatic Actions
Output
Off Observation Imagination
At various levels, "imagination" can be sleep, dreams, hallucinations, or
thought.
Challenges from Control Theory to Meta-System Theory and vice versa
In considering control theory from the perspective of meta-system theory and
PRINCIPIA CYBERNETICA, and vice versa, we are first very pleased with the
opportunity to examine a full-fledged cybernetic and evolutionary theory which is
very similar in spirit, but not in detail, to meta-system theory. The interaction
of these two schools of thought, and their independence from each other, must
continue, to the advantage of both.
For example, consider the great similarity, yet also the great conceptual
differences, between Powers' perceptual control hierarchy and Turchin's
evolutionary control hierarchy:
Culture is control of thought;
which is control of associating;
which is control of complex reflex;
which is control of simple reflex;
which is control of movement;
which is control of position.
Aside from the fact that Turchin's hierarchy is in terms of actions, while
Powers' is in terms of perceptions, there is clearly a similar intent behind each
one: to provide a consistent and elegant cybernetic treatment of organisms from
the perspective of control hierarchies. Undoubtedly both require significant
revision, but the overall program remains clear.
More specifically, we will consider some points of comparison:
Control Specifics: A great advantage that control theory holds for meta-system
theory is that it greatly specifies and clarifies what is meant by the definition
of the meta-system transition: that "the top control system of a system becomes
itself controlled" [TUV91a]. Powers gives this an operational definition, and
allows meta-system theorists the opportunity to match their more abstract,
philosophical theory more closely to the phenomena as revealed by our specialist
colleagues.
Evolution: Both meta-system theory and control theory adhere to Campbell's
[CAD74] view of "blind variation and selective retention" as a universal
mechanism for all kinds of evolution, including genetic, learning, and social
development. But an advantage of meta-system theory is that it is primarily
interested in these evolutionary steps, and intends to explain the evolution of
all emergent levels. Thus, it asks the questions: what are the "essences" of
physical phenomena, life, genetics, sex, multi-cellular organisms, social
organization, and intelligence? Although in his later works [POW89], Powers is
expanding to consider social organization and the origins of life as control
phenomena, this is being done in a somewhat piecemeal manner. Although ultimately
the conceptual unification of control theory should succeed, in terms very
similar to meta-system theory, meta-system theory pursues these subjects at its
most basic task.
The Uniquely Human: What explains humans as unique animal forms? Hunger and lust
are output functions of the organizational system, and provide inputs directly to
the relationships, or perhaps programs level. Which level is unique to humans?
How can its origin be explained in evolutionary terms? How can linguistic
ability, humor, and aesthetics be explained in control theory terms? These are
all addressed directly by meta-system theory, but are underdeveloped in control
theory.
Imagination: Meta-system theory would agree with control theory that imagination
is central to the selection of goal states, and describes these as acts of Will.
But meta-system theory asserts that imagination is unique to humans, and this
ability to control associations of mental representations is the essence of
intelligence. But in Powers' model both imagination and memory are inherent at
all levels of the perceptual control hierarchy. Perhaps there is empirical
evidence to support one view over the other, or a conceptual unification of the
two.
Meta-System Transitions and Ultra-Meta-System Transitions: Clearly, from the
perspective of meta-system theory, each level of the control hierarchy indicates
a meta-system transition. But meta-system theory also involves ultra-meta-system
transitions, which are incorporations of entire meta-system hierarchies in
another at a qualitatively higher dimension, allowing unlimited replication of
the now lower level meta-systems. It seems that Powers' organizational system is
an ultra-meta-system transition, yet applied to only a single meta-system
hierarchy. Further, especially the transition to thought and rationality (the
program and principle levels?) should allow these kinds of ultra-meta-system
transitions, which are evidenced in human linguistic systems and their unlimited
ability to generalize [TUV77].
References
[ASR52] Ashby, Ross: (1952) Design for a Brain, Wiley, New York
[ASR56] Ashby, Ross: (1956) An Introduction to Cybernetics, Methuen, London
[CAD74] Campbell, Donald T.: (1974) "'Downward Causation' in
Hierarchically-Organized Biological Systems", in Studies in the Philosophy of
Biology, eds. F.J. Ayala and T. Dobzhansky, U. California Press, Berkeley
[EIMSCP79] Eigen, M, and Schuster, P: (1979) The Hypercycle, Springer-Verlag,
Heidelberg
[HAH78] Haken, Herman: (1978) Synergetics, Springer-Verlag, Heidelberg
[HEF90f] Heylighen, Francis: (1991) "Cognitive Levels of Evolution", in: Proc.
1990 Int. Congress of Systems and Cybernetics, ed. F. Geyer, Intersytems,
Salinas, CA
[HEFJOC90] Heylighen, Francis; Joslyn, Cliff; and Turchin, Valentin: (1991) "A
Short Introduction to the PRINCIPIA CYBERNETICA Project", Journal of Ideas, v.
2:1, pp. 26-29
[JOC88e] Joslyn, Cliff: (1988) "Review of the Works of Valentin Turchin", Systems
Research, v. 4:4, pp. 298-300,
[MAR88] Marken, Richard S: (1988) "The Nature of Behavior: Control as Fact and
Theory", Behavioral Science, v. 33, pp. 196-206
[MAO70] Mayr, O: (1970) Origins of Feedback Control, MIT Press, Cambridge MA
[NIGPRI89] Nicolis, G, and Prigogine, Ilya: (1989) Exploring Complexity, WH
Freeman, New York
[POW73] Powers, WT: (1973) Behavior, the Control of Perception, Aldine, Chicago
[POW89] Powers, WT ed.: (1989) Living Control Systems, CSG Press, Gravel Switch,
Kentucky
[PRINIG72] Prigogine, Ilya, and Nicolis, Gregoire: (1972) "Thermodynamics of
Evolution", Physics Today, v. 25, pp. 23-28
[SC67] Schrodinger E.: (1967) What is Life?, Cambridge U., Cambridge
[TUV77] Turchin, Valentin: (1977) The Phenomenon of Science, Columbia U., New
York
[TUV81] Turchin, Valentin: (1981) The Inertia of Fear and the Scientific
Worldview, Columbia U. Press, New York
[TUV87a] Turchin, Valentin: (1987) "A Constructive Interpretation of the Full Set
Theory", J. of Symbolic Logic, v. 52:1
[TUV90] Turchin, Valentin: (1990) "Cybernetics and Philosophy", in: Proc. 8th
Int. Congress of Systems and Cybernetics, ed. F. Geyer, Intersytems, Salinas, CA,
NOTE: Draft
[TUV91a] Turchin, Valentin: (1991) "Metasystem Transition as the Quantum of
Evolution", in: Workbook of the 1st PRINCIPIA CYBERNETICA Workshop, Heylighen F.
(ed.), Principia Cybernetica, Brussels-New York
[TUV91b] Turchin, Valentin: (1991) "A Tentative First Sketch of the Starting
Nodes of PCP", in: Workbook of the 1st PRINCIPIA CYBERNETICA Workshop, Heylighen
F. (ed.), Principia Cybernetica, Brussels-New York
[TUVJOC90] Turchin, Valentin, and Joslyn, Cliff: (1990) "The Cybernetic
Manifesto", Kybernetes, v. 19:2-3
[VAFMAH74] Varela, FG, and Maturana, HR et. al.: (1974) "Autopoiesis: The Origin
of Living Systems, its Characterization, and a Model", Biosystems, v. 5, pp.
187-196
[WIN48] Wiener, Norbert: (1948) Cybernetics, MIT Press, Cambridge
____________________________________________________________________________
Francis Heylighen
Free University of Brussels
Evolutionary Foundations for
Metaphysics, Epistemology and Ethics
A process metaphysics
Philosophies traditionally start with an ontology or metaphysics: a theory of
being in itself, of the essence of things, of the fundamental principles. In a
traditional systemic philosophy "organization" might be seen as the fundamental
principle of being, rather than God, matter, or the laws of nature. However it
still begs the question where this organization comes from. In a constructive
systemic philosophy, on the other hand, the essence is the process through which
this organization is created.
There have been several attempts at building a process metaphysics, by
philosophers such as Whitehead (1926) and Teilhard de Chardin (1959). However,
these early process philosophies are characterized by vagueness and mysticism,
and they tend to see evolution as goal-directed, guided by some supraphysical
force, rather than as the blind variation and selection process that we
postulate. They are thus not constructivist in the radical sense as defined in my
first paper in this book.
The ontology we propose would start from elementary actions or processes, rather
than from static objects or particles. These processes are the primitive
elements, the building blocks of our vision of the universe, and therefore remain
undefined. In fact they can be modelled in such a way that they are in themselves
completely empty (Heylighen, 1990). Relatively stable "systems" are automatically
constructed by such processes through the mechanism of blind recombination and
selective retention of stable combinations (Heylighen, 1991b).
This leads to a self-organizing evolution of the universe as a whole. It is
characterized by the spontaneous emergence of more complex organizations (cf.
Simon, 1962) during evolution: from space-time and elementary particles, to
atoms, molecules, crystals, dissipative structures, cells, plants, animals,
humans, society, culture... In this hierarchy of system types (Boulding, 1954),
cybernetic models typically start from about the level of thermostats or
dissipative structures. Yet a constructive systemic approach can also be used at
a much lower level, for example to reconstruct the elementary structures of
space-time (Heylighen, 1990), or the fundamentals of set theory (Turchin, 1987).
A reconstruction of the most important stages in this global evolution should
allow us to answer the questions: "Where do I come from? Who am I?"
Processes of emergence are the "quanta" of evolution: discontinous transitions
which do not change just the state of a system but its organization itself. They
lead to the creation of a new system with a new identity, obeying different laws
and possessing different properties (Heylighen, 1991a). In such a system, the
behaviour of the whole is constrained by the parts (a "reductionistic" view), but
the behaviour of the parts is at the same time constrained by the whole (a
"holistic" view) (Campbell, 1974a).
Perhaps the most important type of emergence is the "meta-system transition"
(Turchin, 1977). Examples of metasystem transitions are the emergence of life, of
multicellular organisms, of the capacity of organisms to learn, of human
intelligence... A metasystem transition is characterized by an increase of the
variety of possible actions (freedom) at the object level (usually through the
assembly of a multiplicity of object systems), together with the emergence of a
situation-dependent control at the metalevel, which coordinates, and chooses
from, the variety of actions available at the level below (Heylighen, 1991a).
A constructive epistemology
Evolution can be likened to a problem-solving process searching through trial and
error for an answer to the question: how to build a system that will survive in a
maximum variety of situations? Knowledge is one of the results of that search: a
mechanism that makes systems more efficient in surviving different circumstances,
by short-cutting the purely blind variation and selection they have to do
(Campbell, 1974b). The appearance of knowledge in the hierarchy of metasystems
corresponds roughly with the emergence of life. Knowledge functions as a
vicarious selector (Campbell, 1974b) which selects possible actions of the system
in function of the system's goal (ultimately survival) and the situation of the
environment. By eliminating dangerous or inadequate actions before they are
executed the vicarious selector foregoes the selection by the environment, and
thus increase the chances for survival of the system. Vicarious selectors are
organized in a hierarchy of control levels (Campbell, 1974b), in accordance with
our metaphysics based on metasystem transitions.
A vicarious selector can be seen as the most basic form of a model: an abstract
system representing processes in the environment. A model is necessarily simpler
than the environment it represents, and this enables it to run faster than, i.e.
anticipate, the processes in the environment (Heylighen, 1990). It is this
anticipation of interactions between the system and its environment, with their
possibly negative effects, that allows the system to compensate perturbations
before they have had the opportunity to damage the system.
Models are not static reflections or homomorphic images of the environment, but
dynamic constructions achieved through trial-and-error by the individual, the
species or the society. This construction of models is similar to the continuous
construction of systems by variation and selection that takes places everywhere
in the universe. What models represent is not the structure of the environment
but its action, insofar as it has an influence on the system. They are both
subjective, in the sense of being constructed by the subject for its own
purposes, and objective, in the sense of being naturally selected by the
environment: models which do not recursively generate adequate predictions are
likely to be later eliminated. There is no "absolutely true" model of reality:
there are many different models which each may be adequate in solving particular
problems, but no model is capable to solve all problems.
The most efficient way to choose or to construct a model which is adequate for
the given problem is by reasoning on a metacognitive level, where a class of
possible models can be analysed and compared. This requires a metasystem
transition with respect to the variety of individual models.
An evolutionary ethics
The evolutionary philosophy can also be used for developing an ethics or system
of values. The basic purpose here would be the continuation of the process of
evolution, avoiding evolutionary "dead ends". Natural selection entails survival
and development (growth, reproduction, adaptation...) as the essential value.
However, the idea of an evolutionary ethics has not been very popular until now,
and we will therefore go into a little more detail about this aspect of our
philosophical system. Evolutionary ethics got a bad reputation because its
association with the "naturalistic fallacy": the mistaken belief that human goals
and values are determined by, or can be deduced from, natural evolution
(Campbell, 1978). Values cannot be derived from facts about nature: ultimately we
are free in choosing our own goals (Turchin, 1991).
However, we must take into account the principle of natural selection, which
implies that if our goals are incompatible with the conditions necessary for
survival, then we will be eliminated from the natural scene. Of course, there is
no natural law or absolute moral principle which forbids you to commit suicide,
but you must be aware that this means that the world will continue without you,
and that it will quickly forget that you ever have been there. If we wish to
evade this alternative, this means that we will have to do everything for
maximising survival.
A second fallacy to avoid is the naive extrapolation of past evolution into the
present or future. The mechanisms of survival and adaptation that were developed
during evolution contain a lot of wisdom--about past situations (Campbell, 1978).
They are not necessarily adequate for present circumstances. This must be
emphasized especially in view of the creativity of evolution: the emergence of
new levels of complexity, governed by novel laws.
For example, biological evolution, based on the survival of the genes, has
favoured egoism: maximizing one's own profit, with a disregard for others (unless
those others carry one's own genes: close family). In a human society, on the
other hand, we need moral principles that promote cooperation, curbing too strong
selfishness. Once the social interactions have sufficiently developed the
appearance of such moral principles (e.g. "thou shalt not steal") becomes
advantageous, and hence will be reinforced by natural selection, even though it
runs counter to previous "egoistic" selection mechanisms (Campbell, 1978). The
development of human society is an example of a metasystem transition, which
creates a new system evolving through a mechanism which is no longer genetical
but cultural (Turchin, 1977).
One of the implications of that transition concerns the interpretation of
survival. Although the death of individual organisms may be useful for the
renewal of the gene pool, making it easier for the genes to adapt to changing
circumstances, it is no longer necessary for cultural evolution. In biological
evolution survival means essentially survival of the genes, not so much survival
of the individuals (Dawkins, 1976). With the exception of species extinction, we
may say that genes are effectively immortal: it does not matter that an
individual dies, as long as his genes persist in his off-spring. In
socio-cultural evolution, the role of genes is played by cognitive systems
("memes", Dawkins, 1976), embodied in individual brains or social organizations,
or stored in books, computers and other knowledge media. However, most of the
knowledge acquired by an individual still disappears at biological death. Only a
tiny part of that knowledge is stored outside the brain or transmitted to other
individuals. Further evolution would be much more efficient if all knowledge
acquired through experience could be maintained, in order to make place only for
more adequate knowledge.
This requires an effective immortality of the cognitive systems defining
individual and collective minds: what would survive is not the material substrate
(body or brain), but its cybernetic organization. This may be called "cybernetic
immortality" (Turchin, 1991). We could conceive its realization by means of very
advanced man-machine systems, where the border between the organic (brain) and
the artificially organic or electronic media (computer) becomes irrelevant. The
death of a biological component of the system would no longer imply the death of
the whole system.
Cybernetic immortality can be conceived as an ultimate goal or value, capable to
motivate long-term human action. It is in this respect similar to metaphysical
immortality (Turchin, 1991): the survival of the "soul" in heaven promised by the
traditional religions in order to motivate individuals to obey their ethical
teachings (Campbell, 1979), and to creative immortality (Turchin, 1991): the
driving force behind artists, authors or scientists, who hope to survive in the
works they leave to posterity.
Another basic value that can be derived from the concept of survival is
"self-actualization" (Maslow, 1970): the desire to actualize the human potential,
that is to say to maximally develop the knowledge, intelligence and wisdom which
may help us to secure survival for all future contingencies (Heylighen, 1990).
Self-actualization may be defined as an optimal, conscious use of the variety of
actions we are capable to execute.
However, if that variety becomes too great, as seems to be the case in our
present, extremely complex society, a new control level is needed (Heylighen,
1991b). This may be realized by a new metasystem transition, similar to the one
mentioned in the section on epistemology, leading to a yet higher level of
evolution. A more detailed understanding of this next transition may help us to
answer the question "Where are we going to?".
The main remaining problem of an evolutionary ethics is how to reconcile the
goals of survival on the different levels: the level of the individual (personal
freedom), the society (integration of individuals), and the planet (survival of
the world ecology as a whole). It is an open question whether the "cybernetically
immortal" cognitive system that would emerge after the next metasystem transition
would be embodied most effectively in an individual being ("metabeing",
Heylighen, 1991b), or in a society of individuals ("superbeing", Turchin, 1991).
It is clear that the different levels have very complicated interactions in their
effect on selection (Campbell, 1979), and hence we need a careful cybernetic
analysis of their mutual relations.
References
Boulding Ken (1956): "General Systems Theory - The Skeleton of Science", General
Systems Yearbook 1, p. 11-17.
Campbell D.T. (1974a): "'Downward causation' in Hierarchically Organized
Biological Systems", in: Studies in the Philosophy of Biology, F.J. Ayala & T.
Dobzhansky (ed.), (Macmillan Press), p. 179-186
Campbell D.T. (1974b): "Evolutionary Epistemology", in: The Philosophy of Karl
Popper, Schilpp P.A. (ed.), (Open Court Publish., La Salle, Ill.), p. 413-463.
Campbell D.T. (1979): "Comments on the sociobiology of ethics and moralizing",
Behavioral Science 24, p. 37-45.
Dawkins R. (1976): The Selfish Gene, (Oxford University Press, New York).
Heylighen F. (1990): "A Structural Language for the Foundations of Physics",
International Journal of General Systems 18, p. 93-112.
Heylighen F. (1991a): "Modelling Emergence", World Futures: the Journal of
General Evolution, (Special Issue on "Creative Evolution", G. Kampis, ed.) (in
press)
Heylighen F. (1991b): "Cognitive Levels of Evolution: from pre-rational to
meta-rational", in: Proceedings 8th Int. Conf. on Cybernetics and Systems (Vol.
II), F. Geyer (ed.), (Intersystems, Salinas, California).
(in press)
Maslow A. (1970): Motivation and Personality (2nd ed.), (Harper & Row, New York).
Simon H.A. (1962): "The Architecture of Complexity", Proceedings of the American
Philosophical Society 106, p. 467-482.
Teilhard De Chardin (1959): The Phenomenon of Man, (Harper & Row, New York).
Turchin V. (1987): "Constructive Interpretation of Full Set Theory", J. of
Symbolic Logic 52:1 , p. 172-201.
Turchin V. (1991): "Cybernetics and Philosophy", in: Proc. 8th Int. Conf. of
Cybernetics and Systems, F. Geyer (ed.), (Intersystems, Salinas, CA).
Turchin, V. (1977): The Phenomenon of Science, (Columbia University Press, New
York ).
Whitehead A.N. (1929): Process and Reality: an essay in cosmology, (Cambridge
University Press, Cambridge).
____________________________________________________________________________
Marc E. Carvallo
State University of Groningen,
The Netherlands
Self-organization, Evolution, and Religion:
Some Notes on Erich Jantsch's Theory of Religion
In his theory, Erich Jantsch uses both terms 'religion' and 'religio'. The former
is usually valued pejoratively as ideological, institutionalized or at least as
belonging to the structural-functional order that not only basicly but also
'surplus-ly represses life, dissipation and creativity of thee fluctuational
order. This type of religion is characterized as the established traditional,
western, monotheistic, and dualistically engined (comp. Jantsch 1980: 73, 177,
181, 241, 249, 257, 264). Only very seldom religion is valued positively, viz.
that of cultures which are predicated upon paradigms essentially different from
that of the above established one, as one might come across the buddhism, the
mysticism, etc. (comp. Jantsch 1950: 303). This type of religion is an expression
of what he calls 'religio' which generally means 'linking backward to the
origin', 'restoring the broken symmetry' etc. (comp. Jantsch 1950: 216 ff., 264,
300-311). 'Religio' defined as such is the very evolution in Jantsch's vision.
Consequently, the second type of religion is one of the vortices, splashes or
ripples of the stream which is called 'religio' or evolution.
Evolution in Jantsch's vision is principally non-darwinistic and is characterized
by :
1. non-dualism, coherence and self-consistency;
2. indeterminism and openness;
3. dissipative self-organization.
In this paper we will try to critically assess these principles and look for
their possible theoretical and practical viability. Which condition should be
fulfilled by the principle of self-consistency in order to also be 'ante-hoc' or
'a-priori' valid? What is the exact nature of the relationship between the future
and the past according to Jantsch? Is it symmetric as has been propounded by e.g.
Spinoza, Hegel or Kierkegaard or is it asymetric as was asserted by some modern
panentheists e.g. Hartshorne (1973)? And from our present position of being the
world of symmetry-breaks: to what or whom are we exactly or ultimately linking
backward? Or does the endless 'religio' constitute the very ultimacy and infinity
of ours? These are some profound questions surreptitiously hidden between the
lines of Jantsch's evolutionary vision.
References
Hartshorne, C. (1973), The Logic of Perfection, La Salle, Illinois: Open Court
Publ. Co.
Jantsch, E. (1980), The Self-Organizing Universe, Oxford/New York: Pergamon.
____________________________________________________________________________
Alvaro Moreno, Arantza Etxeberria & Jon Umerez
Dpt. of Logic and Philosophy of Science
University of the Basque Country, Spain
Biological Information: The Causal Roots of Meaning
From an epistemological point of view, Life involves two kinds of processes that
are, until now, irreducible one to another: the process of materially causing an
effect and the one of representing or controlling another process. Semiotics
draws a borderline between "semiotic" and "pre-semiotic" phenomena distinguishing
between a "natural" meaning (that a sign must possess in respect to its referent
by reason of a causal relationship between them) and a "non-natural" one
(established by mediation of an interpreter and being the binary sign/referent
relationship arbitrary without it). We would like to argue that even if this
dichotomy is widespread in science nowadays, biological systems present phenomena
of a mixed nature, where some relationships among components, even if being
intrinsically causal, will not be established without the concurrence of a third
instance that regulates them and could, therefore, be consider an interpreter of
them.
Our hypothesis will be the following: Natural meaning in Biological Systems has
to do with cause/effect relations at a certain level of organization that revert
in "emergent" configurations at a higher level that, on the one hand, fulfill
some functional action and, on the other, are unpredictable from the lower level.
In this way, a relationship that is diadic (cause/effect) becomes triadic if we
take into account the functional interpretation of it that occurs in the higher
level; inversely also accomplishes a regulating action (boundary condition or
constraint) over the processes taking place in the lower level.
We intend to discuss the following points to argue in favor of this hypothesis:
1) Causality and Determinism; even if often treated as synonyms there are
important epistemological differences between them and most of the biological
phenomena are causal not being deterministic.
2) Causality between different levels of biological organization will be
characterized as forms of emergence. We will take into account three forms of
observation of emergent phenomena:
a) Epistemological; in the sense of a deviation of the behavior of a system in
respect to a model of it. Consequences are novelty production and
unpredictability.
b) Ontological; from a bottom-up perspective there appears a great simplicity in
the upper level, in contrast to the variety of the lower one, which has a
regulating or controlling effect on it. Selection of equally viable alternatives
can also be taken as a case of ontological emergence.
c) Methodological; phenomena studied in a) and b) revert in problems for
classical tools of system description. The main difficulties are the modelling of
the variability of relevant components in biological processes and the necessity
of an-always-changing dynamics that stems from it.
3) From these points we can describe two types of information in biological
systems (information1 and information2).
a) Information1 is characterized by self-referentiality.
More specifically, it is a form of organization characterized by the construction
(starting from the lower dynamical level) of a network formed by components
constituted in sequences of metastable structures which produce and inverse
transformation discrete to continous by the effect of the dynamical components of
the network. The result of this network is an overdetermination on the dynamical
organization of lower level. Some degrees of freedom dicrease at this level and
the action of the upper metastable level creates new functional components if
necessary for the maintenance and reproduction of the network as a whole. The
upper metastable structures (discrete) can be characterized as a
"self-descriptive" information within the system (for example, genetic
information).
b) Information2 grasps the notion of "knowledge" and its referent is external to
the system. The way of providing information2 of a qualitative and semantic
content is not to make it a constituent of models or descriptions independently
constructed in priviledged spaces (minds, brains, computers or libraries).
Instead it is important to collect its semantic content from the active/causal
role it plays within the system itself, in an intrasystemic way. Information2
requires a more complex type of network than information1: it involves a
transformation of external physical patterns into sequences of metastable units.
The action of these latter is functionally evaluable by a loop that ensures the
reproductive identity of the system (the network described for information1) and
its causal action consists in the establishment of a functional correlation
towards the environment through some specific control action on the network.
References
Cariani, P. (1990) "Adaptivity and Emergence in Organisms and Devices" . To be
published in World Futures: the Journal of General Evolution, (Special Issue on
"Creative Evolution" G. Kampis, Ed.) .
Csanyi V. (1989) Evolution in Biological and Social Systems. A General Theory.
Duke University Press.
Eco, U. (1976): A Theory of Semiotics; Bloomington, Indiana University Press.
Fernandez, J; Moreno, A. & Etxeberria, A. (1990): "Life as Emergence: the Roots
of a New Paradigm in Theoretical Biology". To be Published in World Futures: the
Journal of General Evolution, (Special Issue on "Creative Evolution", G. Kampis,
Ed.).
Heylighen, F. (1989) "Causality as Distinction Conservation: a Theory of
Predictability, Reversibility and Time Order". Cybernetics and Systems, 20, pp
361-384.
Kampis, G. (1990). Self-Modifying Systems in Biology and Cognitive Science.
Pergamon Press.
Klee, R.L. (1984): "Micro-Determinism and Concepts of Emergence". Philosophy of
Science, 51, pp. 44-63.
Moreno, A.; Fernandez, J. & Etxeberria, A . (1990): "Biological Computation and
the Emergence of Cognition"; Presented in the "Symbols and Dynamics Workshop" in
Storrs (Connecticut). Submitted to Systémique.
Polanyi, M. (1968). "Life's Irreducible Structure". Science.160, pp.1308-1312.
Pattee, H.H. (1982) "Cell Psychology. An Evolutionary Approach to the
Symbol-Matter Problem". Cognition and Brain Theory.5(4), pp.325-341.
Rosen, R. (1985)".Organisms as Causal Systems Which Are Not Mechanisms. An Essay
into the Nature of Complexity". in R. Rosen. Ed. Theoretical Biology and
Complexity. Academic Press.165-203.
Sercarz E.E.; Celada, F.; Mitchinson, A.A. & Tada, T. (1988): The Semiotics of
Cellular Communication in the Immune System. Springer.
_________________________________________________________________________________
KNOWLEDGE DEVELOPMENT
____________________________________________________________________________
Markus F. Peschl
Dept. for Philosophy of Science,
University of Vienna
The Emergence of Symbols in Subsymbolic Neural Representation Systems
Knowledge representation is one of the central problems in the investigation of
cognitive processes, cognitive science, AI and cognitive modeling. In the
traditional approach of orthodox (i.e. symbol manipulating) AI symbols are
assumed to be the ultimate or atomic representation structures. As will be
discussed in this paper, it turns out that, if we are assuming a more
epistemological perspective, the assumptions being made in this traditional
approach are not adequate for achieving a deeper understanding of cognitive
processes. Orthodox AI and cognitive science are mainly interested in technical
and computer science issues; the naive understanding of (natural) language and
its generality as a representation system is not reflected--this will be done,
however, in this paper in order to show the basic problems of this approach.
In traditional Artificial Intelligence and cognitive science the central problem
of knowledge representation is very much reduced to technical issues and symbol
manipulation. This paper discusses some problems arising, if neither
epistemological nor neuroscientific issues are considered in the field of
cognitive science and of investigating knowledge representing and knowledge
processing systems. An alternative approach is presented: computational
neuroepistemology; it tries to consequently and interdisciplinarily integrate
epistemological, neuroscientific, second order cybernetics as well as computer
science (Parallel Distributed Processing) issues. Some methodological issues of
this approach will be presented: it is based on the assumption that (scientific
as well as common sense) knowledge develops in a cybernetical feedback process of
speculation, construction, empirical investigation and verification; computer
science plays the important role of integrating these two poles by applying its
simulation techniques (i.e. neural computing).
Both natural language and formal symbols are assumed to be one of the most
important representation structures in natural as well as in artificial cognitive
systems--I am trying to differentiate between various levels of representation in
an epistemological investigation considering both traditional (i.e. symbol
manipulation) and neurally inspired simulation methods of AI and cognitive
science. The pros and cons are discussed; it turns out that a symbolic
representation system can be integrated and embedded in the more general neural
representation system if we are considering constructivist and second order
cybernetics concepts (of language, knowledge, etc.; Maturana, von Glasersfeld,
von Foerster, Varela,...).
The neural representation system as well as language are understood as a system
of references; generally spoken, a certain pattern refers to another pattern or
state by an artificially generated and constructed relation. It turns out that
both have a constructivist character which means that knowledge and language are
the result of a process of construction both being realized in neural processes.
Language is understood as one special and very complex form of behavior which is
generated, as all other behavior, by the nervous system. What we are calling
symbols (in our language, in computers, etc.) are emergent properties of the more
general neural representation and reference system.
Symbols and language have to be understood as a highly specialized system of
references following rules which we describe as the grammar of a language--it is
important to see, however, that the grammar of a language is only one possible
way of describing laguage on a very superficial level--computational
neuroepistemology suggests a bottom-up approach being determined by the neural
dynamics rather than by artificial "systems of explanation". The implications of
such a view on the development of a model of cognition will be discussed in
detail. A model of cognition being based on these assumptions is presented.
____________________________________________________________________________
Charles Henry
Butler Library
Columbia University, New York
Non-Verbal Aspects of Language
and Knowledge Structuring
Any cybernetic system that relies on the semantic relationship of words as part
of its structuring needs to confront an inherently paradoxical aspect of
language. This paradox, which concerns the relationship between verbal and
non-verbal components of language, involves the structuring of knowledge as well,
and represents a kind of 'covert' act of intellection that has recently become
the focus of cognitive studies. The premise underlying this paradox can be
summarized as:
1) language engenders images in the mind, whether the language is written or
spoken
2) words or phrases that are unrelated etymologically and have no syntactic,
phonetic, or semantic correlation can nonetheles produce identical images
3) the images thus produced as mental representations can in fact contradict or
oppose the apparent (linguistic) meaning of the text
4) in some cases the meaning of a word or phrase can only be understood by the
recognition and analysis of these images
5) the end result of this process can be the acquisition of new knowledge
6) elucidation of the non-verbal aspects of language in comparison to the
linguistic models of language sheds new light on the mind/brain question
The purpose of this paper is twofold: to elucidate the non-verbal aspect of
language from concrete examples separated by milennia in order to underscore that
this aspect of language is universal, not limited to a particular language,
historical period, linguistic structure, or theme. Secondly, to show the
implication this phenomenon holds for cybernetic design, touching upon
contemporary research in cognitive science, artificial intelligence, and the
physiological properties of the brain. Linguistic examples are drawn from the
ancient Egyptian Book of the Dead, the Hebrew Genesis, and a contemporary poem by
P. Neruda to underscore the universality of the premise.
____________________________________________________________________________
Elan Moritz
The Institute for Memetic Research
Florida
Memetics: Introduction and Implication to the Evolution of Knowledge
Memes are information clusters whose patterns and meanings provide selective
advantage for their replication and spread. In the context of human society,
memes can be regarded as units of cultural transfer. Examples of simple memes are
hair and clothing fashions, slogans, certain religious beliefs, popular music
tunes, certain graphic designs [e.g., the "peace" symbol, "male" and "female"
symbols, the multiple orbits symbol for "atomic" related equipment or hazard].
The attributes that characterize memes are their preferential copying [with a
high degree of fidelity to the original version], by many individuals, as
compared to other informational entities. More complicated meme-like constructs
are also possible. These may be collections of simple memes. Examples of
'meta-memes' are scientific theories, religions, movies, musical symphonies, etc.
In the first part of this paper we present the basis for and recent progress of a
quantitative science of memes [memetics] that combines a decriptive calculus for
memes, principles of population dynamics, information theoretic measures with
physics based least action principles. In the second part of the paper we discuss
the implications of memetics for the evolution of knowledge.
With respect to the objectives of the Principia Cybernetica Project [PCP], and
the interest of developing computer based linking of knowledge, several mappings
of mental memes, or ideas, to physical representations, are discussed. In
complete analogy to the biological gene - genetic engineering metaphor, it is
possible to utilize the PCP framework to construct new knowledge using meme
mutation, combination, and spread. If we categorize PCP participants as humans
[H] and machines [M; e.g. computers, books, videotapes, or any non-biological
information capture/manipulation devices], then new knowledge can emerge by one
or more of the following interactions: H-H, H-M-H, H-M, M-M. Estimates of
quantity of new knowledge [though not necessarily correct knowledge] generation
and spread can be obtained given empiricaly available memetic relationships.
Results of simulations using a Zipf Law [inverse frequency] meme spread
activation are presented. Suggested resource [e.g., time, energy, memory, space]
cost metrics for PCP interactions are described. Results of meme spread and new
meme generation simulations are interpreted in terms of the suggested resource
cost metrics.
References
Moritz E. (1990): "Memetic Science: I - General Introduction", Journal of Ideas
1, p. 1.
Moritz E. (1990): "Replicator Based Knowledge Representation and Spread
Dynamics", in: Proc. IEEE International Conference on Systems, Man & Cybernetics
[Nov 4-7, 1990 Los Angeles, California], p. 256-259
____________________________________________________________________________
J.L. Elohim
Seccion de Graduados e Investigacion
ESIME, Instituto Politécnico Nacional
Mexico
Culture, Cybernetically Interpreted, is a Cybernetic Reflection of Nature Altered by
Culture
Apparently, hominids first and human beings later, since those very early times,
after their emergence as "quite cleverer animals", while they were engaged in
"searching out" how to survive, were increasingly obliged to reflect into their
minds diverse aspects of that "reality" where they were located, most often
without realizing that they were also constituent parts of such reality.
Gradually these beings had to learn how to get better reflections of everything
in their surroundings.
Nowadays, many of us are certain that decision-making--aiming to organize
consciously what we assume our respective performances should be in the future,
at least for surviving (physically, emotionally or intellectually)--must be
supported indispensably by our thoughts constituted by relatively suitable
images, i.e. reflections of particular aspects of the reality; aspects that we
are capable to perceive and whose images in our minds are judged useful for such
purpose.
When analyzing the kind of thoughts expressed during mankind's history, it seems
proper to assert that a lot of them constitute the main source of information
that has made possible to build a certain artificial world, not another. This
"world" inserted into the natural one makes up the "human" civilization.
But this artificial world has never been fully conceived in advance; it has never
been designed as a whole nor the whole set of available reflections of the
reality has been implemented at the same time. On the contrary, civilization has
always been a set of facts and events that have come out of chaotic combinations
of quite dissimilar processes:
* progression, retrogression and deterioration
* evolution and involution
* innovation and tradition
* peace and war
* cultural changes and social bonds
Yet it cannot be denied that civilization is the outcome of an increasingly
improved comprehension of the evolution of real phenomena which comprises both
the natural ones and the others invented by men, which are in fact the essence of
the artificial world already mentioned.
Recently, during the last two decades, a generalized cybernetics has emerged and
developed as an alternative guide that, no doubt, has greatly improved such
comprehension, and that has gradually allowed us to consciously conduct the
dynamics of phenomena belonging to diverse realities: inanimated, living or
artificial.
In accordance with this cybernetics, I would claim that every object of the
natural world is in fact a subject which can be seen as a relatively
well-structured system that "exists" by itself, and has a place in space, while
its "performance" is a function of a certain degree of autonomy in an environment
which is under the influence of many other subjects. This is a relative autonomy
that arises as the outcome of diverse cybernetic relations among the subject's
elements. These relations "help" the subject to organize its "performance" by
itself and create suitable conditions allowing the subject to learn how to take
into account effects of the expected performance on many other subjects, while
finding out how to perform "freely".
Stones, amoebas, plants, animals, ... which emerge as effects of particular
involutionary and evolutionary natural processes, are clear evidences of the
infinite number of possibilities that arise from the manifestation of these
cybernetic relations. The natural emergence of adaptive, prospective, intuitive,
... processes, awareness, consciousness etc., are also evidences of such kind of
possibilities.
Human beings, which emerge as well from natural processes, have apparently
reached the highest level of autonomy by means of their thinking, which offers
them the possibility of getting a proper cybernetic understanding of everything
that moves in time and in space. Such understanding is knowledge that quite
circumstantially becomes the source of cultural actions. These actions become as
well sources of specific technological, economic, political, educational, ...
actions which in accordance to the way they have been developed can be considered
as systems that manifest themselves first as intellectual possibilities and
become later societal phenomena.
I would claim through this paper that any kind of intellectual systems, and human
culture in general, are necessarily particular reflections, relatively faithful
(though sometimes distorted on purpose), of cybernetic possibilities intrinsic to
the dynamics of Nature, which is increasingly altered by an artificial world that
so far is built rather unconsciously by men.
_________________________________________________________________________________
COMPUTER-SUPPORT SYSTEMS
____________________________________________________________________________
Cliff Joslyn
Systems Science
State University of New York at Binghamton
Software Support for PRINCIPIA CYBERNETICA Development
PRINCIPIA CYBERNETICA [1] is a project to develop a collaborative, consensually
based, constructive, philosophical system. Essential to such a project are
computerized tools to aid in system construction. Such tools and technologies as
hypertext, hypermedia, electronic mail, and textual markup, would allow the
construction and publication of structured, non-linear, multi-dimensional
semantic systems and documents by a collaborative group of spatially separated
contributors in a hybrid natural and formal language environment.
We will consider the purposes (ends) and the architecture of a possible computer
system (means) through which these goals could be approached.
Desiderata
* The constructive development of a system of philosophy.
* The collaborative development of a system of philosophy.
* To support the development of consensually-held views among a number of
researchers.
* The dynamic development of a system of philosophy.
* To support the process of semantic analysis.
* To fully represent the semantic relations among the components of the system.
* Easy movement between natural language, formal language, and mathematical
notation.
* To publish the whole or portions of the system through traditional means.
Structures
* Strict and loose hierarchical structures.
* Semantic categories for content development (e.g. part-whole relations).
* Syntactic categories for formal development (e.g. bibliographical references,
time-flow of argumentation).
* Various publishing models (e.g. dictionaries, encyclopedias, reference books,
compilations, arguments, textbooks).
Methods and Technologies
* Hypertext and Hypermedia systems.
* The necessity of ASCII source files and hypertext markup languages.
* The role of the SGML standard for general textual markup and electronic
publishing.
* The role of the HyTime standard for hypertext markup.
* Commercial SGML and HyTime products to support PRINCIPIA CYBERNETICA.
* Electronic mail, electronic journals, and the PRNCYB-L mailing list.
References
[1] Heylighen, Francis; Joslyn, Cliff; and Turchin, Valentin: (1991) "A Short
Introduction to the PRINCIPIA CYBERNETICA Project", Journal of Ideas, v.2:1, pp.
26-29
____________________________________________________________________________
Dirk Kenis
Policy Consulting Services
Free University of Brussels (VUB)
MacPolicy: Delphi and Group Decision Support Ideas for Computer Supported Cooperative
Working
The first part of this paper will provide a state of the art in Group Decision
Support Systems (GDSS) research and related fields. The second part will give an
overview of the results of preliminary research. The third part will describe the
aims of the ongoing research at the V.U.B. on Computer Supported Cooperative
Working.
Group Decision Support Systems : State of the art
With the boom of networking and the necessity for sophisticated well-structured
communication and discussion software, offering more than just passive data
transfer like in traditional mailing systems, we expect the GDSS ideas to become
implemented more and more in enhanced network communication software, that
enables collaboration and interactive simulation.
An important shift in recent GDSS theory is making it clear that the traditional
narrow approach of developing network applications for GDSS-rooms to ameliorate
decision-making sessions directed by 'animators' is replaced by a more general
interest in what is described best as Computer Supported Cooperative Working.
Hence, we witness also the recent emergence of GDSS related fields:
* Computer Conferencing (especially Computer scientists);
* Shared Information Technology (especially Management Information scientists);
* Computer-aided (or -facilitated) Group Decision Making ;
* Computer Supported Cooperative Working (especially Knowledge Scientists,
Cybernetics);
* Computer Based Group Communications and Computer Mediated Communication
(especially Communication scientists).
Computer Supported Cooperative Working projects at the V.U.B.
Long term (2-3 years) interdisciplinary research programmes, of which one in
collaboration with a software house, with in total 3 full-time social scientists
and 4 full-time computer scientists will start mid '91 at the Free University of
Brussels, in order to find out ways to enhance group-work through structured
(network) communication.
In a first phase we will further develop and experiment with a GDSS based on the
principles of a--in human sciences--succesful research method, the (Policy)
Delphi method. Our GDSS, with HyperCard as interface on top of a powerful
database, will be enhanced with several operational research techniques (Multiple
Criteria Decision Aids) and, after substantial testing in real world settings,
will be rewritten through an object-oriented approach as an independent software
application. Through Technology Assessment and Analogy Methods applied to similar
communication technological innovations, combined with group-dynamic experiences
in our experimental GDSS-setting, we will try to find out the conditions to
improve 'Computer Supported Cooperative Work'.
On the basis of these findings we want to develop a flexible meta-tool, a
hypermedia environment where 'groupwork' applications can be easily created or
adjusted. This 'flexible meta-tool' will enable us to develop
network-applications in the field of interactive simulation (i.e. preparing
answers for opposing questions on an important meeting through interactive
gaming), or with expert systems enriched instruments for group planning (i.e.
aids to construct scenarios in a network session), etc.
____________________________________________________________________________
Francis Heylighen
PESP, Free University of Brussels
Structuring Knowledge in a Network of Concepts
The basic evolutionary-systemic and constructive principles that have been
discussed in my two previous contributions to this volume can be directly applied
to the design of a computer support system that would help Principia Cybernetica
collaborators to develop a coherent system of philosophical thought. In fact the
same type of support system might be applied to any complex problem domains where
on the basis of a lot of ill-structured, ambiguous and sometimes inconsistent
data a more or less simple and reliable model is to be built. The problem we are
speaking about is one of applied epistemology. A good epistemology, offering a
concrete and general theory of how knowledge develops during individual or
cultural evolution, should also be useful as a guide when a new model is
practically to be developed.
Network representations of knowledge
I start from the assumption that a lot of knowledge is already available, in
literature and in the heads of different (potential) contributors to the project,
but that that knowledge must be integrated into a coherent and transparent model.
The knowledge will be assumed to be written down in the form of "chunks",
containing text, formulas, drawings, sound, ..., whatever media are most
appropriate to express the underlying ideas. I further suppose these chunks to be
split up into distinct "ideas" or "concepts", such that one chunk should define
not more than one concept.
Of course, these different concepts will be related and one chunk will in general
contain references to several other chunks. For example, the chunk denoting the
concept "dog" might contain the following sentence: a dog is a carnivorous
mammal, with a protruding snout. This means that the concept dog has associations
with a least the concepts mammal, carnivorous and snout. If these concepts are
also available as chunks, then we might create a link from the dog chunk to the
mammal chunk and so on. Computer applications that allow such an easy
representation and manipulation of chunks connected by links are called
hypermedia systems. The chunk with its text and graphics can be shown in a window
on the screen, and it suffices to click on one of the links to show the next
chunk to which the link is pointing (Heylighen, 1991).
Hypermedia system are useful for storing a large amount of complex, interrelated
information (e.g. an encyclopedia) in a easy to handle way. However, there is an
inherent ambiguity involved, since it is not a priori clear what a link is
supposed to mean: any kind of association, as well causal, as logical, as
intuitive as spatial, ..., might be represented by a link. Therefore we need a
better structured system if we want our networks of concepts to support us more
efficiently. By introducing different types of chunks (nodes) and links we may
turn our hypermedia system into a semantic network: the different types of links
will determine (part of) the meaning of the concept to which they are attached.
The problem with semantic networks for knowledge representation is still that of
ambiguity: there is an unlimited number of link and node types that may seem
appropriate, and their interrelationships will in general be very unclear. In
order to limit the set of types, we need an unambiguous, fundamental
interpretation of what concepts and links in our network really stand for. I will
now propose such an interpretation with the corresponding types, and show how it
can be applied to the structuring of knowledge.
Distinction and entailment types
A concept (node) is supposed to represent a distinction: a way to separate
phenomena denoted by the concept (belonging to its class or extension), from
phenomena that do not belong to its extension. Defining a concept means proposing
a procedure for explicitly carrying out that distinction. Definition will be
assumed to be a bootstrapping operation: a concept is always defined in terms of
other concepts, that are themselves defined in terms of other concepts, and so
on. In general there is no primitive level of meaningful concepts in terms of
which all other concepts can be defined. This is in accordance with my
constructive philosophy, stating that any foundations of a conceptual system must
be empty of meaning in order to be acceptable as basis for a complete
philosophical explanation (Heylighen, 1990b).
One way to define a concept is by listing the set of concepts that it entails
together with the set of concepts entailed by it. By entailment I mean an
"if...then" relation, which is more general than the logical (material)
implication. For example, if a phenomenon is a dog, then it is also a mammal: dog
-> mammal. It means that a phenomenon denoted by the first concept cannot be
present or actual, without a phenomenon denoted by the second one being
(simultaneously) or becoming (afterwards) actual.
In order to derive fundamental types of distinctions (concepts, nodes) and links
(entailments), we will posit two basic dimensions of distinction: stability (or
time) and generality, with the corresponding values of instantaneous - temporary
- stable, and of specific - general. The combination of these 3 x 2 values leads
to 6 types of distinction (see table).
time\generality general specific
__________________________________________________________________________________________
stable class object
temporary property situation
instantaneous change event
For example, an object is a distinction that is stable (it is not supposed to
appear or disappear while we are considering it), and specific (it is concrete,
there is only of it). A property is a distinction that is general (several
phenomena may be denoted by it, it represents a common feature), and temporary
(it may appear or disappear, but normally it remains present during a finite time
interval). An event is instantaneous (it appears and disappears within one
moment), and specific (it does not denote a class of similar phenomena, but a
particular instance).
With these node types we can now derive the corresponding link types by
considering all possible combinations of two node types. There is one constraint,
however: we assume that a more invariant (stable or general) distinction can
never entail a less invariant one. Otherwise, the second would be present each
type the first one is, contradicting the hypothesis that it is less invariant
than the first one. For example, a class cannot entail an object, a situation
cannot entail an event. Yet it is possible that concepts with the same type of
invariance, (e.g. two objects) might be connected by an entailment relation. All
remaining possible combinations can now be summarized by the following scheme
(the straight arrows represent entailment from one type to another (more
invariant) one, the circular arrows entailment from a concept of a type to a
concept of the same type):
[Workbook2.gif]
For example when an object A entails a class B, A -> B, then A is an Instance_of
B. When an object A always entails the presence of another object B, then B must
belong to or be a part of A. When a change A entails another change B, then A and
B "covary" and hence A can be interpreted as the cause of B. When an event A
entails a situation B, then A must be simultaneous with or preceding B in time.
The advantage of this scheme is that most of the intuitive and often used
semantic categories (objects, classes, causality, whole-part relations, temporal
precedence, etc.) can be directly constructed from it, in a simple and uniform
format. Complementarily, given some of those everyday categories, we can use the
scheme to reduce them to simple entailment links between nodes of specific types.
In fact the types themselves can be represented as nodes, and each node of a
particular type will have an entailment link to that 'type'-node. This allows us
to reduce a complicated set of semantic categories to an extremely simple formal
strcuture.
Knowledge structuring
Given that structure, consisting of a list of nodes and entailment links between
them, we can now start to formally analyse the network. Define the input and
output sets of a node:
Input: I(x) = { y | y -> x} = "extension" of concept x
Output : O(x) = { y | x -> y } = "intension" of concept x
The meaning (definition, distinction) of x can be interpreted as determined by
the disjunction of its input elements, and the conjunction of its output
elements. Our previous remark about definitions can now be reformulated as the
following bootstrapping axiom (Heylighen, 1990ab):
two nodes are distinct if and only if their input and output sets are distinct:
x != y I(x) != I (y), O(x) != O(y)
However, such a complete definition assumes that all concepts allowing to
distinguish between x and y are present in the network. In practice, the network
of concepts we are building by writing down our knowledge in the form of
connected chunks, will be incomplete in some respects, redundant in other
respects. Instead of using the axiom as a static description of how a complete
network should be structured, we can use it as a procedure to find ways to make
the network more adequate, by adding missing concepts, or by deleting redundant
ones. We can distinguish the following two main techniques (cf. Heylighen, 1991;
Bakker, 1987; Stokman & de Vries, 1988):
Node identification
When input and output sets of two nodes x and y are identic or similar, the
computer support system may propose the user to either identify (merge) the two
nodes, and replace them by one single node, or to add new nodes or links that
would more clearly differentiate between x and y. An algorithm may test the
identity or inclusion of the input and output sets, and according to the results,
propose the following possibilities to the user:
1) I (x) = I (y):
a) O (x) = O (y) => Identify (or distinguish) x and y
b) O (x)[subset of]Ì O (y) => Identify x and y, or distinguish I (x) from I (y)
2) I(x) )[subset of] I(y):
a) O(x) = O(y) => Identify x and y, or distinguish O(x) from O(y)
b) O(x))[subset of]Ì O(y) => Identify x and y
c) O(y) )[subset of] O(x) => Connect x to y, x -> y
Node integration
When a cluster of nodes have a common set of "external" input or output nodes
(that is to say nodes that do not belong to the cluster), then from the point of
view of those external nodes, the nodes inside the cluster are indistinguishable.
Hence the nodes, though not strictly indistinguishable according to the
bootstrapping axiom, behave indistinguishably from a certain viewpoint.
From that point of view, the cluster may be called closed (Heylighen, 1990a) and
it might therefore be replaced by a single "integrated" node. The integrated node
"summarizes" the cluster nodes on a more abstract level, and may hence simplify
the conceptual model. Similar to the case of node identification, the external
indistinguishability of clustered nodes may be spurious, and this should prompt
the user to add additional distinguishing links and nodes.
There are different types of closure, with different meanings and formal
properties, depending upon which sets of external input or output nodes are
common among the cluster, for example: transitive closure, equivalence, cyclical
closure, ... If the closure is only approximative (the cluster nodes have several
external neighbours in common, but these do not form a complete set of any
specific type), then this method is similar to the one called "conceptual
clustering" in machine learning, where the boundaries between clustered and
non-clustered nodes become fuzzy, and depend on the treshold chosen for the
number of common neighbours.
In conclusion, the present set of concepts and techniques, when implemented on a
computer through a suitable intuitive interface, should enable an individual or
group of users to elicit and structure their knowledge about a domain under the
form of a network of concepts connected by entailment links, and support them to
minimize the redundancy, complexity and incompleteness of their model.
The introduction of new nodes and links by the user corresponds to a form of
variation by recombination of concepts. The recognition of a closed cluster of
nodes by the system corresponds to the selection of a distinction that is more
stable or invariant than the distinctions between the internal concepts of the
cluster (Heylighen, 1990a), with closure as fundamental selection criterion. The
elicitation and structuring of concepts in this manner hence follows the general
evolutionary mechanism that was postulated in my previous papers about
evolutionary philosophy.
References
Bakker R.R. (1987): Knowledge Graphs: representation and structuring of
scientific knowledge, (Ph.D. Thesis, Dep. of Applied Mathematics, University of
Twente, Netherlands).
Heylighen F. (1990): "A Structural Language for the Foundations of Physics",
International Journal of General Systems 18, p. 93-112 .
Heylighen F. (1990): "Relational Closure: a mathematical concept for
distinction-making and complexity analysis", in: Cybernetics and Systems '90, R.
Trappl (ed.), (World Science, Singapore), p. 335-342.
Heylighen F. (1991): "Design of a Hypermedia Interface Translating between
Associative and Formal Representations", International Journal of Man-Machine
Studies. (in press)
Stokman F.N. & de Vries P.H. (1988): "Structuring Knowledge in a Graph", in:
Human-Computer Interaction, Psychonomic Aspects, G.C. van der Veer & G.J. Mulder
(eds.), (Springer, Berlin).
____________________________________________________________________________
Robert Glück
Institut für Prakt. Informatik
University of Technology Vienna
Metasystem Transition in the Machine
and its Application to Knowledge Systems
We propose to incorporate the notion of metasystem transition (MST) in knowledge
systems and to provide tools capable of performing MST with regard to a knowledge
system, an idea first stressed in [5]. In particular, we propose to investigate
MST for systems supporting the Principia Cybernetica Project, a project dealing
with cybernetic philosophy in which the very concept of MST plays a fundamental
role. This is supported by very promising applications of the concept of MST, in
the form of the Futamura Projections (FMP), to compiler-construction and
-generation, central fields of computer science [1]. Why not apply and take
advantage of the benefits of the MST for knowledge systems implemented on a
machine?
What follows is a review of the principle of MST and the formulation of two
potential application to knowledge systems.
* MST is a transition from one system S to a metasystem S*. It expands the
hierarchy of systems by adding a new level of control. In principle, adding a
new level of control can be done repeatedly. In case the same system S is
added as metasystem S*, it is called self-application of S.
* A knowledge system (in general) comprises two parts: a knowledge base which
is structured according to some knowledge representation scheme (e.g.
semantic network) and a mechanism maintaining and organizing the knowledge
base. More importantly this mechanism (inference engine) is capable of
answering questions by examining (reasoning about) the knowledge base. Such
engines may e.g. incorporate fuzzy or non-monotonic logic (of course not
restricted to those).
The application of MST to knowledge systems can be described formally: Let infer
be an inference engine, k the knowledge base and q a question. This will be
formalized as follows (the notation is the same as in [2]).
"Run the inference engine infer to answer the question q by
examining the knowledge base k".
Definition: A program alpha is a program specializer (e.g. partial evaluator,
supercompiler) iff for all programs p and arbitrary values x, y and the
metavariable Y the following characteristic equations holds:
(1) = =
Formula (1) represents the first MST: the knowledge base and the inference engine
become objects under the control of alpha. Note that p-x is a program that is
fixed to the value x. In addition note that the expression is metacoded
([[arrowdown]] = arrow down) [4]. The operation inverse to [[arrowdown]] will be
denoted by [[arrowup]] (arrow up). It is obvious that p can be substituted by
infer in formula (1).
(2) = =
(3) infer-k ::=
The program infer-k represents a program which is capable of answering questions
about k without the need for examining and interpreting the knowledge base. All
actions needed for interpreting the knowledge base have been removed so that q
can be answered more efficiently.
The knowledge base may be large and change more frequently than the inference
engine. In this case it may take some time for alpha to analyze infer and k in
(3). Doing one more MST we get the following formula by applying alpha to the
right side in (3), where gen is a program that is constructed by the second MST
according to the semantics implemented by the inference engine (4)
Usage: http://www.kk-software.de/kklynxview/get/URL
e.g. http://www.kk-software.de/kklynxview/get/http://www.kk-software.de
Errormessages are in German, sorry ;-)