Prerational Intelligence: Adaptive Behavior and Intelligent Systems Without Symbols and Logic, Volume 1, Volume 2 Prerational Intelligence: Interdisciplinary Perspectives on the Behavior of Natural and Artificial Systems, Volume 3

1585 Pages·2000·38.37 MB·English

by  Holk Cruse

PRERATIONAL INTELLIGENCE:
ADAPTIVE BEHAVIOR AND INTELLIGENT SYSTEMS
WITHOUT SYMBOLSAND LOGIC,
VOLUME 1
STUDIES IN COGNITIVE SYSTEMS
VOLUME 26/1
EDITOR
James H. Fetzer, University ofMinnesota, Duluth
ADVISORY EDITORIAL BOARD
Fred Dretske, Stanford University
Charlcs E. M. Dunlop, University ofMichigan, Flint
Ellery Eells, Univeristy ofWisconsin, Madison
Alick Elithom, Royal Free Hospital, London
Jerry Fodor, Rutgers University
Alvin Goldman, University ofArizona
Jaakko Hintikka, Boston University
Frank Keil, Cornell University
William Rapaport, State University of New York at Buffalo
Barry Richards, Imperial College, London
Stephen Stich, Rutgers University
Lucia Vaina, Boston University
Terry Winograd, Stanford University
PRERATIONAL INTELLIGENCE:
ADAPTIVE BEHAVIOR AND
INTELLIGENT SYSTEMS
WITHOUT SYMBOLS AND LOGIC,
VOLUME 1
Edited by
Holk Cruse
University of Bielefeld, Bielefeld, Germany
J effrey Dean
Cleveland State University, Cleveland, Ohio, U.S.A.
and
Helge Ritter
University of Bielefeld, Bielefeld, Germany
Springer-Science+Business Media, B.V.
A C.I.P. Catalogue record for this book is available from the Library of Congress.
ISBN 978-94-010-3792-1 ISBN 978-94-010-0870-9 (eBook)
DOI 10.1007/978-94-010-0870-9
Printed with the support of thc University of Bielefeld. Germany
Gedruckt mit Unterstützung der Universität Bielefeld
Printedon acid~fi·ee paper
All Rights Reserved
© 2000 Springer Science+Business Media Dordrecht
Originally published by Kluwer Academic Publishers in 2000.
Softcover reprint of the hardcover 1st edition 2000
No part of the material protected by this copyright notice may be reproduced or
utilized in any form or by any means, electronic or mechanical,
including photocopying, recording or by any information storage and
retrieval system, without written permission from the copyright owner.
TABLE OF CONTENTS
Preface IX
PART 1: INTRODUCTION TO VOLUME 1
HOLK CRUSE, HELGE RITTER, and JEFFREY DEAN I
Introduction to Volume 1 3
PART II: INTELLIGENCE AND PRERATIONAL INTELLIGENC E
PETER LANZ I Introduction to Part II 7
PETER LANZ I The Concept oflntelligence in Psychology and
Philosophy 19
DAVID MCFARLAND I Rational Behaviour of Animals and Machines 31
IPKE WACHSMUTH I The Concept of Intelligence in AI 43
PART III: A CASE STUDY
WALTER HEILIGENBERG I The Jamming Avoidance Response (JAR)
of the Electric Fish, Eigenmannia: The Processing of Sensory
Information and Motor Contra! 59
PART IV: INTELLIGENT SENSING
HANS-ULRICH BAUER, GEORG HARTMANN, and KLAUS OBERMAYER I
Introduction to Part IV 87
FRAN(:OIS CLARAC, DANIEL CATTAERT, and ALAIN MARCHAND I
Peripheral Sensory Modules Controlling Motor Behavior 91
JOACHIM ERB ER and BABETTE PRIBBENOW I Antennal Movements in
the Honeybee: How Complex Tasks are Solved by a Simple
Neuronal System I 09
V
vi TABLE OF CONTENTS
GEORG HARTMANN I lnvariances in Visual Systems ofRobots and
Animals: Functional and Architectural Consequences 123
CHRISTIAN WEHRHAHN I Retinal Coding for Vernier Acuity
and Motion 151
HANS-ULRICH BAUERand JOCHEN BRAUN I ls There Parallel
Binding ofDistributed Objects? 163
REINHARD ECKHORN and MICHAEL STÖCKER I Synthesizing Camplex
Perceptions: I. Concepts ofVisual Feature Associations
Basedon Neural Synchronization and Related Experimental
Results in the Visual Cortex 175
MICHAEL STÖCKERand REINHARD ECKHORN I Synthesizing Camplex
Perceptions: II. Models ofVisual Feature Associations
Basedon Neural Synchronization 191
BARTLETT W. MEL I A Neurally-lnspired Approach to 3-D
Visual Object Recognition 205
KLAUS OBERMAYER I Modelling the Formation of Sensory
Representations in the Brain 215
PART V: INTELLIGENT MOTOR OUTPUT I
SENSORIMOTOR COUPLING
Low Level:
HOLK CRUSE, JEFFREY DEAN, and WILLIAM B. KRISTAN I lntro-
duction to Part V 235
WILLIAM B. KRISTAN, Jr. I Distributed Processing vs. Dedicated
Neurons in the Production of Simple Behavioral Acts 243
ANSGAR BÜSCHGES, ARNE E. SAUER, and ULRICH BÄSSLER I
Flexibility of a Proprioceptive Feedback System Results From Its
"Parliamentary" (Distributed) Organization 267
JACQUES DUYSENS I Load Detectors and Their Reflexes During
Gait in Cat and Man 287
STAN GIELEN I Coordination of Redundant Manipulators: Reduction
ofDegrees ofFreedom 305
TABLE OF CONTENTS VII
SIMON GISZTER and WILLIAM KARGO I Movement Organization
in the Frag Spinal Cord: Prerational Intelligence? 323
High-Level:
JACQUES PAILLARD I The Neurobiological Roots of Rational
Thinking 343
JOSEPH AYERS I A Behaviour-Based Controller Architecture for
Reactive Underwater Robots 357
AVIS H. COHEN I Control ofMovement is Distributed and
Heterarchical 3 71
JEFFREY DEAN I Locomotion in Insects: Patterns Generated by
Interacting Oscillators 3 91
HOLK CRUSE, CHRISTIAN BARTLING, GENNADY CYMBALYUK,
JEFFREY DEAN, MARTIN DREIFERT, THOMAS KINDERMANN,
JOSEF SCHMITZ, and MICHAEL SCHUMM I A Modular Artificial
Neural Net for Controlling a Six-Legged Walking System 407
JOSEF STEUER, HANS-JÜRGEN WEIDEMANN, and FRIEDRICH PFEIFPER I
The Control System ofthe Six-Legged TUM Walking Robot 431
BAREARA WEBB I An Arbitrary Architecture for an Artificial
Arthropod 44 3
List of Authors 461
Subject Index 465
Author Index 475
PREFACE
The present book is the product of conferences held in Bielefeld at the
Center for interdisciplinary Sturlies (ZiF) in connection with a year-long ZiF
Research Group with the theme "Prerational intelligence". The premise ex
plored by the research group is that traditional notions of intelligent behav
ior, which form the basis for much work in artificial intelligence and cog
nitive science, presuppose many basic capabilities which are not trivial, as
more recent work in robotics and neuroscience has shown, and that these
capabilities may be best understood as ernerging from interaction and coop
eration in systems of simple agents, elements that accept inputs from and act
upon their surroundings.
The main focus is on the way animals and artificial systems process in
formation about their surroundings in order to move and act adaptively. The
analysis of the collective properties of systems of interacting agents, how
ever, is a problern that occurs repeatedly in many disciplines. Therefore,
contributions from a wide variety of areas have been included in order to
obtain a broad overview of phenomena that demoostrate complexity arising
from simple interactions or can be described as adaptive behavior arising
from the collective action of groups of agents. To this end we have invited
contributions on topics ranging from the development of complex structures
and functions in systems ranging from cellular automata, genetic codes, and
neural connectivity to social behavior and evolution. Additional contribu
tions discuss traditional concepts of intelligence and adaptive behavior.
1. INTELLIGENCE, INTELLIGENT BEHAVIOR AND RELATED
STRUCTURALFEATURES
Here is not the place to attempt to define intelligence, but rather we will
briefly consider some of the features we associate with the concept of in
telligence or intelligent behavior. What do we mean by intelligence? What
actions by an animal or a machine would we consider to show true intelli
gence? If, as we will argue, intelligence can be seen as a continuum, what
features are useful for ordering degrees or kinds of intelligence?
If we can agree on a working definition of intelligence, then it is a simple
matter to define prerational intelligence as intelligence that does not rely on
ix
H. Cruse et al. (e ds.), Prerationallntelligence: Adaptive Behavior and Intelligent Systems Without Symbols
and Logic, Volume 1, ix-xix.
© 2000 Kluwer Academic Publishers.
X PREFACE
rational thought, symbolic representations and logical reasoning. The ulti
mate goal is to consider whether there are common organizational principles
in systems showing such behavior.
Standard dictionary definitions of intelligence high light the ability to deal
with new or trying situations but, at the same time, they link intelligence
strongly to rational thought. The viewpoint adopted here is that this linkage
represents an anthropocentric bias emphasizing highly advanced human ca
pabilities. This anthropocentric bias influenced much early thinking about
artificial intelligence, leading to a focus on knowledge and cognition, on
symbolic representation and on Jogical manipulation ofthese symbols in for
mally described, closed systems. Similarly, control of movement relied on
exact analytical solutions providing reference signals for servocontrollers,
while perceptual processing attempted to interpret sensor data by applying
stored knowledge.
However, a truly complete explanation and simulation of these rational
abilities must extend to the axioms, such as the origin of the symbols and
the logical rules. Consideration of the foundations for these higher func
tions from evolutionary and neurophysiological perspectives requires a dif
ferent approach, one suggesting that many behaviors that humans naturally
describe as a logical process in fact are realized in biological systems using
quite different mechanisms.
Thus, the Research Group places less emphasis on the use of reason and
symbols, and more on the function. As a starting point, it characterizes
intelligence as "the ability to deal with new situations". This is the nominal
use of intelligence; it clearly is an attribute of a behaving agent and most
clearly relates to action selection. In addition, there are the adjectival and
adverbial uses of intelligent: usually with an implicit value judgement, about
the characteristics and mode of operation of a particular behavior, action,
design or process. The Research Group feels that the latter use can profitably
be extended to modules or subprocesses below the Ievel of an autonomaus
agent, such as perceptual processes that identity features in the environment
or motor processes organizing movements for example.
To begin, we will briefly consider what features might characterize in
telligence. First of all, it is important to emphasize that intelligence is an
attribute of an agent acting in a context. Our judgements of intelligence are
affected by aspects of both the agent and the context. Intelligence resides
in the relation between the set of behaviors or outputs that an agent could
produce in a given context and the set it actually produces.
As an attempt to provide axes for the space of intelligence, we suggest the
following six attributes characterizing the relationship of agent and context.