CONSTRUCTIVIST

FOUNDATIONS
Volume 1, Number 1
Date of publication: 01 November 2005

EDITORIAL
Alexander Riegler (Free University of Brussels)
Editorial.The Constructivist Challenge

OPINIONS
Radical constructivism:
Ernst von Glasersfeld (University of Massachussetts)
Thirty Years Constructivism
Second order cybernetics:
Diederik Aerts (Free University of Brussels)
Ceci n’est pas Heinz von Foerster

SURVEY
Enactive cognitive science:
Kevin McGee (Linköping University)
Enactive Cognitive Science.
Part 1: Background and Research Themes

CONCEPTUAL
Epistemic structuring:
Herbert F. J. Müller (McGill University)
People,Tools, and Agency: Who Is the Kybernetes?

EMPIRICAL
Radical constructivism:
Dewey Dykstra (Boise State University)
Against Realist Instruction.
Superficial Success Masking Catastrophic Failure and an Alternative

An interdisciplinary journal
http://www.univie.ac.at/constructivism/journal
 Advisory Board DESCRIPTION
William Clancey Constructivist Foundations (CF) is an independent academic peer-reviewed e-journal
NASA Ames Research Center, USA without commercial interests. Its aim is to promote scientific foundations and
Ranulph Glanville applications of constructivist sciences, to weed out pseudoscientific claims and to
CybernEthics Research, UK base constructivist sciences on sound scientific foundations, which do not equal the
Ernst von Glasersfeld scientific method with objectivist claims. The journal is concerned with the
University of Massachusetts, USA interdisciplinary study of all forms of constructivist sciences, especially radical
Vincent Kenny
constructivism, cybersemiotics, enactive cognitive science, epistemic structuring of
Inst. of Constructivist Psychology, Italy experience, second order cybernetics, the theory of autopoietic systems, etc.
Klaus Krippendorff
University of Pennsylvania, USA AIM AND SCOPE
Humberto Maturana The basic motivation behind the journal is to make peer-reviewed constructivist
Institute Matríztica, Chile papers available to the academic audience free of charge. The “constructive”
Josef Mitterer character of the journal refers to the fact that the journal publishes actual work in
University of Klagenfurt, Austria constructivist sciences rather than work that argues for the importance or need for
Karl Müller constructivism. The journal is open to (provocative) new ideas that fall within the
Wisdom, Austria scope of constructivist approaches and encourages critical academic submissions to
Bernhard Pörksen help sharpen the position of constructivist sciences.
University of Hamburg, Germany The common denominator of constructivist approaches can be summarized as
Gebhard Rusch follows.
University of Siegen, Germany • Constructivist approaches question the Cartesian separation between objective
Siegfried J. Schmidt world and subjective experience;
University of Münster, Germany • Consequently, they demand the inclusion of the observer in scientific
Bernard Scott explanations;
Cranfield University, UK • Representationalism is rejected; knowledge is a system-related cognitive process
Sverre Sjölander
rather than a mapping of an objective world onto subjective cognitive
Linköping University, Sweden structures;
• According to constructivist approaches, it is futile to claim that knowledge
Stuart Umpleby
George Washington University, USA approaches reality; reality is brought forth by the subject rather than passively
received;
Terry Winograd
Stanford University, USA
• Constructivist approaches entertain an agnostic relationship with reality, which
is considered beyond our cognitive horizon; any reference to it should be
Editor-In-Chief refrained from;
• Therefore, the focus of research moves from the world that consists of matter to
Alexander Riegler
Vrije Universiteit Brussel, Belgium the world that consists of what matters;
• Constructivist approaches focus on self-referential and organizationally closed
Editorial Board systems; such systems strive for control over their inputs rather than their
Pille Bunnell
outputs;
Royal Roads University, Canada • With regard to scientific explanations, constructivist approaches favor a process-
oriented approach rather than a substance-based perspective, e.g. living systems
Olaf Diettrich
Center Leo Apostel, Belgium are defined by processes whereby they constitute and maintain their own
organization;
Dewey Dykstra • Constructivist approaches emphasize the “individual as personal scientist”
Boise State University, USA
approach; sociality is defined as accommodating within the framework of social
Stefano Franchi interaction;
University of Auckland, New Zealand
• Finally, constructivist approaches ask for an open and less dogmatic approach to
Timo Honkela science in order to generate the flexibility that is needed to cope with today’s
Helsinki Univ. of Technology, Finland
scientific frontier.
Theo Hug
University of Innsbruck, Austria
Urban Kordes
SUBMISSIONS
Institut Jozef Stefan, Slovenia Language: Papers must be written in English. If English is a foreign language for you
Albert Müller please let the text be proofread by an English native speaker.
University of Vienna, Austria Copyright: With the exception of reprints of “classical” articles, all papers are
Herbert F. J. Müller “original work,” i.e., they must not have been published elsewhere before nor
McGill University, Montreal, Canada must they be the revised version (changes amount to less than 25% of the
Markus Peschl
original) of a published work. However, the copyright remains with the author
University of Vienna, Austria and is licensed under a Creative Commons License.
Author’s guidelines at http://www.univie.ac.at/constructivism/guideline.pdf
Bernd Porr
University of Glasgow, UK Send all material to Alexander Riegler ariegler@vub.ac.be.
John Stewart Important: Use “Constructivist Foundations” in the “Subject:” field.
Univ. de Technologie de Compiègne, France
Tom Ziemke For more information please visit the journal’s website at
University of Skövde, Sweden http://www.univie.ac.at/constructivism/journal

Cover page: Painting “Knidos” by Werner Horvath, New Austrian Constructivism.

With kind permission of the artist, http://members.telering.at/pat/werner.htm
 conceptual constructivism
EDITORIAL

Editorial. The Constructivist Challenge

Alexander Riegler A Vrije Universiteit Brussel, ariegler@vub.ac.be
capture their behavior in neat formalisms in
Purpose: This is an attempt to define constructivism in a pluralistic way. It categorizes order to make reliable predications is ren-
constructivist work within a three-dimensional space rather than along one dimension only. dered impossible. It is the responsibility of the
Practical implications: The interdisciplinary definition makes it possible to perceive the scientist to decide these problems: “Only
rather heterogenous constructivist community as a coherent and largely consistent scien- those questions which are in principle unde-
tific effort to provide answers to demanding complex problems. Furthermore it gives cidable we can decide” (Foerster 1991/2003).
authors of Constructivist Foundation the opportunity to locate their own position within The solution to such “big problems” in sci-
the community. Conclusions: I offer a catalogue of ten points that outline the constructivist ence simply cannot be delegated to nature as
program. Each of these aspects invites authors to extensively reflect on it and to approach the monolithic “objective arbiter.” Therefore
it from their disciplinary background to do work in any of the types of investigations the pluralistic perspectives are of utmost impor-
journal covers. Key words: constructivist approaches, interdisciplinarity, dogmatism. tance when scientifically approaching phe-
nomena of organized complexity.
The present first issue of “Constructivist
What is constructivism? Consider the
following example. Suppose that we
lized (cf. the “If it ain’t broken don’t fix it” syn-
drome, Riegler 1998, 2001b), caught in the Foundations” is the attempt to provide this
take a piece of chalk and write on a blackboard momentary situational context as deter- plurality. It creates an interdisciplinary forum
“A = A.” Now we point at it and ask, “What is mined by the way we have learned to deal with for authors and readers, philosophers and
this?” We may get one of the following things. But secondly, the analogy also warns engineers, academics and practitioners, to
answers. (a) White lines on a black back- us of authoritarian attempts to think there is approach the most challenging scientific
ground; (b) An arrangement of molecules of only true solution to a whatever we identify as problems from the constructivist perspective.
chalk; (c) Three signs; (d) The law of identity. a problem. Their appeal to “reality” as the This perspective, however, is not a monolithic
Regardless whether you are an art critic, a ultimate arbiter of (scientific) disputes gives building, nor is it the philosopher’s stone.
chemist, a philosopher, or a mathematician, it rise to the belief that there exists a mind-inde- Rather, it “is a way of thinking, not a collec-
is obvious that the answer will depend on pendent reality which defines what is true and tion of facts,” as Ernst von Glasersfeld (1985/
your educational background. At first sight what is not. However, as Mitterer (1992) 1992) once said about cybernetics which is
we may find this example amusing and harm- pointed out, isn’t claiming authority by refer- one of the major roots for several contempo-
less. We all know that personal preferences ring to an external truth the attempt to make rary constructivist schools. For many it
bias the way we perceive the world. But what one’s own point of view unassailable? The two sounds unthinkable to refrain from searching
about the following example, originally analogies above should make it clear that sci- for the correct answer, for the correct solution
attributed to Danish nobel prize winner Niels ence and philosophy gain from variety and to a given problem. Unfortunately, unambig-
Bohr and retold by Humberto Maturana? A the possibility of choosing from other uous solutions work for simple systems and
teacher “who asks a student to measure the options. simply problems only. We are used to the sit-
height of a tower with the use of an altimeter, Such variety and freedom of choice has uation in Newtonian physics that deals with a
may flunk the student if he uses the length of always been a major aspect a constructivist small number of entities subject to a few
the altimeter to triangulate the tower and philosophies and sciences. Heinz von Foer- forces only. And we are used to thinking that
obtains the height of the tower through ster’s (1973/2003, p. 227) “ethical impera- such situations can be extrapolated to mas-
geometry and not through physics. The tive”: “Act always so as to increase the number sively large inhomogeneous systems. How-
teacher may say that the student does not of choices” does not only anthropo-morphize ever, as it has turned out over the last decades
know physics” (Maturana 1978, p. 42). What W. Ross Ashby’s (1956) Law of Requisite Vari- these systems of organized complexity such as
this episode suggests is twofold. Firstly, by ety which states that the variety of actions human cognition, quantum physics, life,
focusing on one particular approach only we available to a control system must be at least economy, global weather, and many more,
will quickly get caught in ignorance and as large as the variety of actions in the system evade our attempts to generate simple and
denial of other approaches that might turn to be controlled (so by having more choice clear-cut answers. These systems call for
out much more fruitful. Of course, such is the you stay in control). Foerster’s imperative is interdisciplinary approaches, for open
human psyche: functionally fixed (Duncker also a reminder to the fact that most problems inquiries that enable investigators to escape
1935/1945). Once we have found a viable in science are undecidable in principle. These the confinements of a specific discipline and
solution (such as reading the display of an are problems of “organized complexity” to become aware of aspects that are necessary
altimeter) we tend to stubbornly apply the (Weaver 1948), characterized by a “sizeable to satisfyingly solve the problem.
pattern of our solution to all other problems number of factors which are interrelated into By now it should have become obvious
as well. In other words, our thinking is cana- an organic whole” (p. 539). Any attempt to that there is no simple answer to the initial

Constructivist Foundations 2005, vol. 1, no. 1 1

 conceptual
EDITORIAL
question “What is constructivism?” There are
many constructivist “schools” as constructiv-
ist concepts have been developed in various
scientific disciplines. In order to provide an
impression of how diverse constructivist
“schools” can be I shall sketch a few of the
constructivisms relevant for the journal.
The plurality of
constructivism
When some 30 years ago by Ernst von Glaser-
sfeld started publishing on a concept he called
“radical constructivism” (Glasersfeld 1995;
cf. also his recollection article in this edition)
he pioneered the philosophical-epistemologi-
cal approach. He maintains that knowledge is
not passively received but actively built up by
the cognizing subject (first principle of radi-
cal constructivism). Furthermore the func-
tion of cognition is adaptive; it serves the
organization of the experiential world, not
the discovery of ontological reality (second
principle). He calls his version “radical”
because he claims that constructivism has to
be applied to all levels of description. “Those
who ... do not explicitly give up the notion
that our conceptual constructions can or
should in some way represent an indepen-
dent, ‘objective’ reality, are still caught up in
the traditional theory of knowledge” (Glaser-
sfeld 1991).
Glasersfeld refers to the skeptic tradition
in philosophy, especially to Sextus Empiricus,
Berkeley (Esse est percipi, i.e., to be is to be per-
ceived), Vico (Verum ipsum factum, i.e., the
truth is the same as the made), and to Hans
Vaihinger’s (1911/1952) as-if philosophy. For
Glasersfeld, skepticism points the way to the
insight that whatever world view we construct
we do not have any means of validating it. He
also quotes Jean Piaget from whom he took
over the idea that the child constructs his or
her world by means of assimilation and
accommodation.
Another philosophically oriented per-
spective is Herbert Müller’s (2000; cf. also his
article in this edition) epistemic structuring (of
experience) approach. It assumes mental
structures to be tools for mastering unstruc-
tured experience. The principle of zero-deri-
vation claims that reality structures are not
derived from any given pre-structured enti-
ties inside or outside the subject thus obviat-
2 Constructivist Foundations
ing the need for belief in mind-independent
reality.
The cybernetic approach has a different his-
tory. Originally hired as editor of the proceed-
ings of the Macy-conferences on cybernetics
(cf. Pias 2003) the subject of which was “cir-
cular-causal and feedback mechanisms in
biological and social systems,” it soon struck
Heinz von Foerster that a “cybernetics of
observing systems” is far more interesting
than a “cybernetics of observed systems.” His
so defined second-ordered cybernetics became
the guiding paradigm of the Biological Com-
puting Lab (BCL) he was running for many
years.
Starting from the insight that nervous sig-
nals are merely electrochemical, Heinz von
Foerster formulated the Principle of Undiffer-
entiated Encoding: “The response of a nerve
cell does not encode the physical nature of the
agents that caused its response. Encoded is
only ‘how much’ at this point on my body, but
not ‘what’” (Foerster 1973/2003, p. 215). The
principle can be found in Maturana and
Varela’s assertion that the cognitive apparatus
is an “organizationally closed system” (see
biological approaches below).
It can be claimed that the psychological-
cognitive approach started with developmen-
tal psychologist Jean Piaget whose scientific
conviction can be summarized in his state-
ment “L’intelligence organise le monde en
s’organisant elle-même” (Piaget 1937/1954,
p. 311). In his theory of cognitive develop-
ment (e.g., he argued that in the beginning, a
newborn knows little about how to cope with
the perceptive impressions around her. Faces
might be funny or threatening colorful spots
and voices unknown sounds. In fact, she
doesn’t even know that these are colors and
sounds. Only by assimilation and accommo-
dation the child constructs a collection of
rules (schemata) during her ontogeny. Sche-
mata serve as a point of reference when it
comes to assimilating new experiences. If
impressions are too alien to be aligned to an
older, already assimilated experience, they are
either not perceived at all or give rise to the
accommodation of those existing schemata,
which are appropriately adjusted in order to
include the new “exotic” experience. With
each of these assimilating or accommodating
steps the child constructs another piece of
reality. Piaget’s theory has been interpreted in
a constructivist way especially by Glasersfeld.
constructivism
Psychologist Ulric Neisser (1975) devel-
oped a theory of schemata controlled informa-
tion pickup. A cognitive schema “accepts
information as it becomes available at sensory
surfaces and is changed by that information.
It directs movements and exploratory activi-
ties that make more information available, by
which it is further modified” (p. 55).
For Kevin O’Regan and Alva Noë (2001)
seeing is knowing sensorimotor dependen-
cies, and the brain is a device to extract alge-
braic structures between perception and
action. The authors refer to the work of
Donald MacKay (1969) on “sensorimotor
contingencies” and continued the work of
Paul Bach-y-Rita (1972) who pioneered with
work on sensory substitution. In particular,
Bach-y-Rita showed how a blind person could
gain some notion of sight by converting visual
camera images into tactile information, and
interpreted this as expression of brain plastic-
ity.
The theory of autopoietic systems formu-
lated by biologists Humberto R. Maturana
and Francisco J. Varela can be referred to as the
biological-neurobiological approach. Autopoi-
etic systems are a sub-class of self-organizing
systems which, if they exist in the physical
domain, are the class of living systems. For
them, the nervous system is a closed network
of interacting neurons where any change in
the state of relative activity of a collection of
neurons leads to a change in the state of rela-
tive activity of other or the same collection of
neurons. This is referred to as the “organiza-
tional closure” of the nervous system. It can
be argued that organizational closure repre-
sents the starting point for the formal inter-
pretation of radical constructivism (Riegler
2001a).
The work of neurophysiologist Rudolfo R.
Llinás (2001) provides empirical backing. He
too formulated a closed-system hypothesis:
“[The brain] is capable of doing what it does
without any sensory input whatsoever” (p.
94). According to his dreaming machine-
argument, we “are basically dreaming
machines that construct virtual models.”
Neurophysiologist Gerhard Roth (Haynes
et al. 1998) maintains that the limbic system,
the unconsciously working part of the brain
responsible for evaluations, is the ultimate
instance of volitional cognition. In their view,
consciousness is just a pseudo-ruling ego. It is
not the ego who constructs; it is constructed,
 conceptual
EDITORIAL
or as Wolfgang Prinz (1997, p. 155) put it, “We
do not do what we want, but we want what we
do.” In other words, this raises the question
who (or what) is responsible for the construc-
tions that form our cognition (Riegler 2003).
Later on, Varela together with Evan
Thompson and Elenor Rosch developed
another constructivist variant known as enac-
tivism or enactive cognitive science based on
key concepts such as autopoiesis, structural
determinism and structural coupling. In the
enactivist paradigm, experience is rooted
within the organizational autonomy of the
acting system and is considered fundamental
for social and cultural phenomena. As the
authors put it, it attempts to account for “how
action can be perceptually guided in a per-
ceiver-dependent world” (Varela, Thomspon
& Rosch 1991, p. 173; cf. also the McGee’s sur-
vey in this edition).
One could assume that the most “objec-
tive” of all disciplines, physics, does not con-
tribute to the constructivist spectrum. Inter-
estingly, however, arguing from the
background of physics, Olaf Diettrich (2001)
developed a constructivist evolutionary episte-
mology (or cognitive operator theory). He
claims perceived patterns and regularities are
just invariants of inborn cognitive (sensory)
operators. Therefore, laws of nature are
human-specific. A different set of cognitive
operators yields a different cognitive pheno-
type. Creatures equipped with such alterna-
tive phenotypes would be impossible to com-
municate with. Diettrich’s approach also
claims a homology between mechanisms gen-
erating mathematical terms and those gener-
ating observational ones, explaining thus why
mathematics is such an effective tool to
describe the world.
In his quantum-physical world view, Ger-
hard Grössing (2001) maintains that per-
ceived non-classical structure of space and
time in relativistic cases are human-specific
artifact based on neurophysiological pro-
cesses.
Paul Watzlawick’s well-known Palo-Alto
group (Watzlawick, Weakland, & Fisch 1974)
for family therapy uses constructivism to
make patients solve their interpersonal prob-
lems. Their approach can be called psychia-
trist-therapeutic. The basic therapeutic inter-
vention is to disrupt patterns of symptomatic
interaction by “reframing” a habitual situa-
tion, i.e., to “place conceptual and/or emo-
2005, vol. 1, no. 1
tional setting or viewpoint in another frame
which fits the ‘facts’ of the same concrete situ-
ation equally well or even better and thereby
changes its entire meaning” (p. 95). This
encourages the patients to find alternative
constructions of their worldview. In other
words, it helps to escape canalizations I
referred to in the beginning.
Psychologist George Kelly (1955) devel-
oped a challenging subjectivist theory, Per-
sonal Construct Psychology, that focuses on the
concept of anticipation. His “man as scientist”
metaphor expresses the idea that “a person’s
processes are psychologically channelized by
the way in which he anticipates events” (p. 46)
Human beings aim at a better control of their
world by predicting events and constructing
their reality. These constructions are con-
stantly subject to validation and subsequent
modification if necessary.
The list of constructivist approaches could
be even further extended. For example, there
is the literature-media science approach cham-
pioned by Siegfried J. Schmidt (1987), Geb-
hard Rusch (1987) et al. in the 1980s in Ger-
many. Part of the credit also goes to Wolfram
K. Köck who made excellent German transla-
tions of authors such as Maturana, Glasers-
feld, and Foerster, which triggered the great
impact of radical constructivism on the
humanities in German-speaking countries.
Further researchers in this area are Nancy
Spivey (1997) and Stefan Weber (2005) who
argues in favor of a non-dualistic media the-
ory as proposed by Mitterer (1992, 2001).
Building primarily on Maturana and Varela’s
autopoietic theory, Niklas Luhmann (1984/
1995) developed a system theoretical version,
which has found many followers especially in
Germany. Ernst von Glasersfeld and Leslie
Steffe (Steffe & Gale 1995) contributed a great
deal to implementing radical constructivism
in educational sciences. Former BCL member
Gordon Pask (1975) developed a constructiv-
ist theory of communication as applied to edu-
cation and extended by Bernard Scott (e.g.,
Scott 2001).
In reaction to mathematical Platonism,
mathematical constructivists such as L. E. J.
Brouwer, Arend Heyting (1975), and Jean
Paul Van Bendegem claim that mathematical
objects exist only if a method can construct
them. As a consequence they oppose, for
example, the notion of infinity, either by
denying the actual infinite or by denying both
constructivism
the actual and the potential infinite (Van
Bendegem 1999).
With Erlangen Constructivism Paul Loren-
zen and Wilhelm Kamlah (Kamlah & Loren-
zen 1967/1984, Lorenzen 1987) attempted a
circular-free foundation of sciences and sci-
entific languages. Its basis is twofold: a pre-
scientific vocabulary and standardized action
schemata to generate objects. Later, Erlangen
Constructivism was transformed into Meth-
odological Culturalism by Peter Janich (1996).
He claims a relativism on the fact that all jus-
tifications are based on pre-active and pre-
discursive consensuses, which are marked by
an already achieved cultural level. (“Alle
Begründungen und Rechtfertigungen finden
zulässige Anfänge in präaktiven und prädis-
kursiven Konsensen, die durch eine schon
erreichte Kulturhöhe ausgezeichnet sind.”)
As proponents of the computational
approach Steven Quartz (Quartz & Sejnowski
1997) and Gert Westermann (2000) could be
listed as well as Gary Drescher (1991) who
cast Piaget into algorithms.
Last but not least Constructionism (Harel
& Papert 1991) as an educational philosophy
should be mentioned. It emphasizes that in
order to learn about abstract concepts it is
necessary to create and experiment with arti-
facts. In this perspective, understanding and
experience are closely related in the sense that
learning is considered a process of active
knowledge construction rather than passive
knowledge absorption.
Does constructivism
matter?
Will constructivism change science? Carnap
discussed the effect of epistemology in his
well-known thought experiment of two geog-
raphers – a realist and an idealist – who travel
to Africa to investigate claims about an
unusual mountain. Carnap’s conclusion is
that the “two geographers will come to the
same result not only about the existence of the
mountain, but also about its other character-
istics, namely position, shape, height, etc. In
all empirical questions there is unanimity…
[The epistemological] divergence between
the two scientists does not occur in the empir-
ical domain, for there is complete unanimity
so far as the empirical facts are concerned.”
(Carnap 1928/1967, p. 334). Similarly, Hel-
3
 conceptual
EDITORIAL
mut Schwegler (2001) argues that all science
including physics is basically a language game
in the sense of Wittgenstein, i.e., scientists
communicate via language and work via these
communications. But in order to play this
language game correctly one doesn’t need to
adopt the constructivist world view. So, after
all, does a constructivist foundation matter?
As Glasersfeld said, knowledge is always the
result of a constructive activity rather than the
accumulation of propositional data (such as
position and heights of mountains). In other
words, constructivism shifts the focus of
attention from the propositional “knowing
that” to the pragmatic “knowing how.” In a
certain sense, scientists rather resemble shoe-
makers who have to work with their given
material. In the “realist mode” shoemakers
stick to the principles of shoemaking which
are believed to be true. Constructivist shoe-
makers, however, will more flexibly adopt
alternative approaches as for them the com-
mitment to a hypothetical truth is no longer
an essential criterion (Dewey Dykstra, per-
sonal communication). If this analogy is cor-
rect, then one of the advantages of a construc-
tivist-biased science certainly has more
potential to come up with new solutions.
The common
denominator
Let us pick up again the initial question, “What
is constructivism?” As argued above, giving a
one-dimensional answer does not only con-
tradict constructivist principles, it is above all
counterproductive for scientific and philo-
sophical endeavors. It would be difficult if not
impossible to lump together the many inde-
pendent disciplinary roots and proponents of
constructivism. However, it is possible and
desirable to distill their common denomina-
tor. From what has been said so far in this edi-
torial but without going into further details
(and thereby violating the idea of a denomina-
tor being wide enough to cover various para-
digms) I present the “constructivist program.”
It encompasses the following ten aspects.
1. Constructivist approaches question the
Cartesian separation between objective world
and subjective experience. As argued by Josef
Mitterer (2001), such dualistic approaches,
being the prevailing scientific orientation, are
based on the distinction between description
4 Constructivist Foundations
and object, and their argumentation is
directed towards the object of thought. His
thesis says: The dualistic method of searching
for truth is but an argumentative technique
that can turn any arbitrary opinion either true
or false. Therefore the goal of dualistic philos-
ophies, i.e., philosophies based on the sub-
ject–object dichotomy, is to convince a public
audience (readers, listeners, discussion part-
ners) of the truth. An example to surmount
the separation is the concept of “co-enaction”
(Varela, Thompson & Rosch 1991, p. 150)
according to which “...knower and known,
mind and world, stand in relation to each
other through mutual specification or depen-
dent coorigination.”
2. As a consequence of point 1, construc-
tivist approaches demand the inclusion of the
observer in scientific explanations. Foerster
(quoted from Glasersfeld 1995) summarizes
the crucial point in a single statement,
“Objectivity is the delusion that observations
could be made without an observer.” Mat-
urana (1978, p. 3) made it a dictum: “Every-
thing said is said by an observer to another
observer that could be him- or herself.”
3. Representationalism is rejected. Ques-
tioning Wittgenstein’s correspondence theory
of representation (“in order to tell whether a
picture is true or false we must compare it
with reality”) induced Glasersfeld to formu-
late the radical constructivist paradigm. In
the constructivist perspective knowledge is
the result of an active construction process
rather than of a more or less passive represen-
tational mapping from the environment of an
objective world onto subjective cognitive
structures. Therefore, knowledge is a system-
related cognitive process rather than a repre-
sentation (Peschl & Riegler 1999).
4. According to constructivist approaches,
it is futile to claim that knowledge approaches
reality. Instead, reality is brought forth by the
subject. As Glasersfeld (1991, p.16) put it,
“those who merely speak of the construction
of knowledge, but do not explicitly give up the
notion that our conceptual constructions can
or should in some way represent an indepen-
dent, ‘objective’ reality, are still caught up in
the traditional theory of knowledge.”
5. Constructivist approaches entertain an
agnostic relationship with reality, which is
considered beyond our cognitive horizon. Any
reference to it should be refrained from. This
position is not necessarily limited to skeptical
constructivism
philosophies. Positivist Rudolf Carnap
expressed the necessity of this aspect in his
1935 book saying that “we reject the thesis of
the Reality of the physical world; but we do not
reject it as false, but as having no sense, and its
Idealistic anti-thesis is subject to exactly the
same rejection. We neither assert nor deny
these theses, we reject the whole question.”
6. Therefore, the focus of research moves
from the world that consists of matter to the
world that consists of what matters. Since the
cognitive apparatus brings forth the world
out of experiences, our understanding of
what we are used to refer to as “reality” does
not root in the discovery of absolute mind-
independent structures but rather in the
operations by which we assemble our experi-
ential world” (Glasersfeld 1984). Or in the
words of Foerster, instead of being concerned
with “observed systems” the focus of atten-
tion shifts to “observing systems.”
7. Constructivist approaches focus on self-
referential and organizationally closed sys-
tems. Such systems strive for control over
their inputs rather than their outputs. Cogni-
tive system (mind) is operationally closed. It
interacts necessarily only with its own states
(Maturana & Varela 1979). The nervous sys-
tem is “a closed network of interacting neu-
rons such that any change in the state of rela-
tive activity of a collection of neurons leads to
a change in the state of relative activity of
other or the same collection of neurons”
(Winograd & Flores 1986, p. 42). This is a
consequence of the neurophysiological prin-
ciple of undifferentiated encoding: “The
response of a nerve cell does not encode the
physical nature of the agents that caused its
response.” (Foerster 1973/2003, p. 293).
Humberto Maturana (1978) suggests that we
can compare the situation of the mind with a
pilot using instruments to fly the plane. All he
does is “manipulate the instruments of the
plane according to a certain path of change in
their readings” (p. 42). In other words, the
pilot doesn’t even need to look “outside.” The
enactive cognitive science paradigm expresses
clearly: “...autonomous systems stand in
sharp contrast to systems whose coupling
with the environment is specified through
input/output relations. ...the meaning of this
or that interaction for a living system is not
prescribed from outside but is the result of the
organization and history of the system itself.”
(Varela, Thompson & Rosch 1991, p. 157)
 conceptual
EDITORIAL
8. With regard to scientific explanations,
constructivist approaches favor a process-
oriented approach rather than a substance-
based perspective. For example, following
Maturana living systems are defined by pro-
cesses whereby they constitute and maintain
their own organization. Their structure
refers to the “actual relations which hold
between the components which integrate a
concrete machine in a given space” (Mat-
urana & Varela 1979) while their organiza-
tion defines the “dynamics of interactions
and transformations” a system may undergo.
Material aspects are therefore secondary.
9. Constructivist approaches emphasize
the “individual as personal scientist”
approach as their starting point is the cogni-
tive capacity of the experiencing subject.
Sociality is defined as accommodating
within the framework of social interaction.
While social interaction is not considered a
new quality in contrast to interacting with
non-living entities, its complexity is
acknowledged. However, society is not a pri-
ori given, not the “social precedes the per-
sonal” (Gergen 1997). Rather, “society” must
be conceptually analyzed. Constructivism is
also rather pragmatic about “common
knowledge” such as texts. They “contain nei-
ther meaning nor knowledge – they are a
scaffolding on which readers can build their
interpretation” (Glasersfeld 1992, p. 175).
10. Finally, constructivism asks for an
open and more flexible approach to science
in order to generate the plasticity that is
needed to cope with the scientific frontier.
Also today’s knowledge-based society must
be assessed through its ability and willing-
ness to continuously revise knowledge.
Krohn (1997) refers to it as the society of self-
experimentation. Luhmann (1994) defines
knowledge as schemata that are regarded as
true but ready to be changed. Constructiv-
ism must be considered as a way to forgo the
dogmatism that prevents science from
becoming more fruitful and productive than
today.
This list is deliberately painted with a big
brush. Rather than limit future develop-
ments right from the onset, the list wants to
give the necessary latitude to future authors
in Constructivist Foundations to further
extend the constructivist program. This is
the constructivist challenge, and the journal
will be one of its main champions.
2005, vol. 1, no. 1
constructivism
NAVIGATING THE PLURALITY
Given the plurality of constructivist approaches it seems heretical to order them in one dimen-
sion only. Hence Constructivist Foundations will navigate the constructivist space in three
dimensions.
Dimension 1: Discipline
Along this dimension we find the following disciplines.
[ biological-physiological
[ cognitive-psychological
[ educational
[ engineering-computer scientific
[ historical
[ philosophical-epistemological
[ physical
Dimension 2: School
Since many scientists and philosophers have developed their respective version of constructivism
without necessarily paying much attention to historical or contemporary parallels a number of
labels for constructivist research have emerged.Therefore authors may align their submission to
Constructivist Foundation to any from the following (incomplete) list of “schools” (or paradigms).
[ constructivist evolutionary epistemology
[ cybersemiotics
[ enactive cognitive science (cf. McGee’s survey in this number)
[ epistemic structuring of experience (cf. Müller’s conceptual paper)
[ radical constructivism (cf. Glasersfeld’s recollection article)
[ second order cybernetics (cf. Aerts’s interview with Foerster)
[ theory of autopoietic systems
Dimension 3: Types of inquiry
As different disciplines prefer different types of inquiry, submissions to Constructivist Founda-
tions investigations too may focus on different ways of how to use their insights. Contributions
will be classified according the following dimension.
Opinions are written from the personal perspective of constructivist researchers and philoso-
phers (and are therefore subject to editorial editing only).
Surveys provide an extensive overview with the goal to bracket single insights and results to
provide a global picture.
Conceptual papers develop philosophical-argumentative support.
Empirical studies focus on psychological, biological, physical etc. evidence.
Synthetic studies try to turn conceptual or empirical insights of constructivist theories into
models, simulation, or hardware devices.
in a rather lucid and comprehensible way
The articles (being multilingual he has developed a pro-
in this edition found command of pragmatic language use)
and his personal account makes it easy to get
The present first edition provides a sample of a grip on his concepts even if one is meeting
the sort of articles that will be published in them for the first time.
upcoming editions of Constructivist Founda- The second opinion article is a voice from
tions. It starts with a recollection of Ernst von the past, an interview with Heinz von Foerster
Glasersfeld who summarizes the (personal) that he gave ten years ago at the large and
history of the radical constructivist paradigm. stimulating conference “Einstein meets Mag-
Newcomers to constructivism may find the ritte: An interdisciplinary reflection on sci-
text particularly appealing. Glasersfeld writes ence, human action and society,” which fea-
5
 conceptual
EDITORIAL
tured nineteen famous plenary speakers such
as Ilya Prigogine, Brian Arthur, Francisco
Varela, Chrisopher Langton, Julian Jaynes,
William Calvin, Bas van Fraassen, to name
but a few. The interview has never been pub-
lished before and is also available as audio file
for download from the journal’s web page.
Like Glasersfeld’s article it serves as a histori-
cal document for readers who want to get the
whole picture. Therefore, endnotes were
added that explain the relevance of people
mentioned by Foerster. He survived this
interview by seven years and died on 2 Octo-
ber 2002 in California (cf. memorial volume
in Riegler 2005).
The third contribution is the first part of
an extensive overview of the enactive cogni-
tive science (ECS) approach, mainly pio-
neered by Francisco Varela (1946–2001).
Kevin McGee is brilliant at pulling many
aspects together into a coherent survey of the
historical and conceptual background of ECS.
By outlining research themes he proves that
ECS is a fruitful research framework for the
future. The second part of McGee’s survey
will appear in the next edition of Constructiv-
ist Foundations.
Another type of paper published in the
journal are conceptual-philosophical articles
that provide the foundation for further theo-
retical reflections and practical empirical or
synthetic work. Herbert Müller’s conceptual
framework of “epistemic structuring of expe-
rience” is introduced and discussed in the
fourth paper. It opposes traditional meta-
physical ontology and focuses on the inver-
References
Ashby, W. R. (1956) Introduction to cyber-
netics. Methuen: New York.
Bach-y-Rita, P. (1972) Brain mechanisms in
sensory substitution. Academic Press:
New York.
Carnap, R. (1928) Der logische Aufbau der
Welt. Felix Meiner Verlag: Leipzig. English
translation: Carnap, R. (1967) The logical
structure of the world. Pseudoproblems in
philosophy. University of California: Ber-
keley.
Carnap, R. (1935) Philosophy and logical syn-
6 Constructivist Foundations
sion of thinking by thoroughly applying con-
structivism at all levels. As a result the author
claims that many “hard” problems in philos-
ophy such as the mind–body problem may
find easy solutions. The article not only intro-
duces a new constructivist variation, it has
also been shaped in a novel way. Originally
written as a target article for the world-wide-
web-based discussion forum “Karl Jaspers
Forum” at http://www.kjf.ca, it received such
a large number of comments (which in turn
spurred many responses by the author) that
the author wrote a revised version that
includes the criticism and support from the
comments. In other words, the paper has
undergone “public reviewing” which served
as a sufficient criterion for publication in
Constructivist Foundations. Also in future,
the editors of the journal intend to exploit this
mode as an alternative to the standard dou-
ble-blind peer reviewing used for other
papers in the journal.
The last paper in this edition is an empiri-
cal study of constructivist education that has
become a well-known education paradigm in
the US. Its author Dewey Dykstra dismisses a
number of allegations against constructivist
education and presents a new constructivist
alternative to the “elitist-realist paradigm.”
The selection of papers for this edition
reflects the flexibility of constructivist strate-
gies. It is evident that a broad variety of topics
and types of paper is difficult to find in most
other journals. Variety and diversity, however,
do not mean shallowness as the reviewed
papers show; they are distinct in their deep
tax. Kegan Paul: London.
Diettrich, O. (2001) A physical approach to
the construction of cognition and to cog-
nitive evolution. Foundation of Science 6:
273–341.
Drescher, G. L. (1991) Made-up minds: A
constructivist approach to artificial intelli-
gence. MIT Press: Cambridge, MA.
Duncker, K. (1945) On problem solving. Psy-
chological Monographs 58: 1–112. Ger-
man original published in 1935.
Foerster, H. von (1973/2003) On construct-
ing a reality. In: F. E. Preiser (ed.) Environ-
mental design research, Vol. 2. Dowden,
constructivism
discussion. It is the conviction of the editors
that carefully crafted conceptual, empirical
and synthetic articles as well as comprehen-
sive surveys yielding a global perspective and
personal opinions of senior scientists will
contribute to turning constructivist
approaches into a valuable ingredient of the
scientific endeavor as they provide new per-
spective, insights, and inspiration in areas
where conventional epistemologies have
proven increasingly insufficient research
strategies.
ABOUT THE AUTHOR
Alexander Riegler obtained a Ph.D. in artifi-
cial intelligence and cognitive science in
1995 from the Vienna University of Tech-
nology with a dissertation on constructivist
artificial life. His research interests include
cognitive science, philosophy of science,
and research in biological and cognitive
complexity. He worked at the department
of Theoretical Biology (University of
Vienna), and at the department of Com-
puter Science (University of Zurich). Since
1998 he has been a research fellow at the
Center Leo Apostel for Interdisciplinary
Research (Free University of Brussels). He
co-organized the interdisciplinary confer-
ences “New Trends in Cognitive Science”, in
1997 on knowledge representation and in
2001 on virtual reality.
Hutchinson & Ross: Stroudberg, pp. 35–
46. Reprinted in: Foerster, H. von (2003)
Understanding understanding. Springer-
Verlag: New York, pp. 211–228.
Foerster, H. von (1991/2003) Ethics and sec-
ond-order cybernetics. In: Inderstanding
understanding. French original published
in 1991 in: Ray, Y. & Prieur, B. (eds.)
Système, ethique. Perspectives en thérapie
familiale. ESF editeur: Paris, pp. 41–55.
Gergen, K. J. (1997) Social theory in context:
Relational humanism. In: Greenwood, J.
(ed.) The mark of the social. Rowman and
Littlefield: New York.
 conceptual
EDITORIAL
Glasersfeld, E. von (1984) An introduction to
radical constructivism. In: Watzlawick, P.
(ed.) The invented reality: How do we
know? W. W. Norton: New York, pp. 17–
40.
Glasersfeld, E. von (1985) Declaration of the
American Society for Cybernetics (Annual
Meeting of the ACS 1981). American Soci-
ety of Cybernetics Newsletter 24: 1–4.
Reprinted as: Glasersfeld, E. von (1992)
Cybernetics In: Negoita, C. V. (ed.) Cyber-
netics and applied systems. Marcel Decker:
New York, pp. 1–5.
Glasersfeld, E. von (1991) Knowing without
metaphysics: Aspects of the radical con-
structivist position. In: Steier, F. (ed.)
Research and reflexivity. Sage Publica-
tions: London, pp. 12–29.
Glasersfeld, E. von (1992) Questions and
answers about radical constructivism. In:
Pearsall, M. K. (ed.) Scope, sequence, and
coordination of secondary school science,
Vol. II: Relevant research. The National
Science Teachers Association: Washington
DC, pp. 169–182.
Glasersfeld, E. von (1995) Radical construc-
tivism. A way of knowing and learning.
The Falmer Press: London.
Grössing, G. (2001) Comparing the long-
term evolution of “cognitive invariances”
in physics with a dynamics in states of con-
sciousness. Foundation of Science 6: 255–
272.
Harel, I. & Papert, S. (eds.) (1991) Construc-
tionism. Norwood: Ablex.
Haynes, J.-D., Roth, G., Schwegler, H. & Sta-
dler, M. (1998) Die funktionale Rolle des
bewußt Erlebten [The functional role of
conscious experience]. Gestalt Theory 20:
186–213.
Heyting, A. (ed.) (1975) L. E. J. Brouwer, Col-
lected works 1. Philosophy and founda-
tions of mathematics. Elsevier:
Amsterdam.
Janich, P. (1996) Konstruktivismus und
Naturerkenntnis. Auf dem Weg zum Kul-
turalismus [Constructivism and the
knowledge of nature: On the path to cul-
turalism]. Suhrkamp: Frankfurt.
Kamlah, W. & Lorenzen, P. (1984) Logical
Propaedeutic: Pre-School of Reasonable
Discourse. University Press of America:
Lanham. German original “Logische
Propädeutik. Vorschule des vernünftigen
Redens” appeared in 1967.
2005, vol. 1, no. 1
Kelly, G. (1963) A theory of personality.
Norton: New York.
Krohn, W. (1997) Rekursive Lernprozesse:
Experimentelle Praktiken in der Gesell-
schaft. Das Beispiel der Abfallwirtschaft.
In: Werner Rammert, W. & Bechmann, G.
(eds.) Technik und Gesellschaft. Jahrbuch
9: Innovation – Prozesse, Produkte, Poli-
tik. Campus: Frankfurt, pp. S. 65–89.
Llinás, R. R. (2001) I of the vortex. MIT Press:
Cambridge.
Lorenzen, P. (1987) Constructive philosophy.
The University of Massachusetts Press:
Amherst.
Luhmann, N. (1994) Die Wissenschaft der
Gesellschaft (2nd edition). [The science of
society]. Suhrkamp: Frankfurt.
Luhmann, N. (1995) Social systems. Stanford
University Press: Palo Alto. German origi-
nal published in 1984 as: Soziale Systeme.
Grundrifl einer allgemeinen Theorie.
Suhrkamp: Frankfurt.
MacKay, D.M. (1969) Information, mecha-
nism and meaning. MIT Press: Cam-
bridge, MA.
Maturana, H. R. (1978) Biology of language:
The epistemology of reality. In: Miller, G.
A. & Lenneberg, E. (eds.) Psychology and
biology of language and thought: Essays in
honor of Eric Lenneberg. Academic Press:
New York, pp. 27–63.
Maturana, H. & Varela, F. (1979) Autopoiesis
and cognition. Reidel: Boston.
Mitterer, J. (1992) Das Jenseits der Philoso-
phie. Wider das dualistische Erkennt-
nisprinzip.. Passagen Verlag: Vienna.
Mitterer, J. (2001) Die Flucht aus der Beliebig-
keit. Fischer: Frankfurt.
Müller, H. F. J. (2000) Concept-dynamics and
the mind-brain question. Karl Jaspers
Forum TA32. http://www.kjf.ca/kjf/32-
TAMUL.htm
Neisser, U. (1975) Cognition and reality. Free-
man: San Francisco.
O’Regan, J. K. & Noë, A. (2001) What it is like
to see: A sensorimotor theory of percep-
tual experience. Synthese 129: 79–103.
Pask, G. (1975) Conversation, cognition, and
learning. Elsevier: New York.
Peschl, M. & Riegler, A. (1999) Does represen-
tation need reality? In: Riegler, A., Peschl,
M. & Stein, A. v. (eds.) (1999) Understand-
ing representation in the cognitive sci-
ences. Plenum Press: New York, pp. 9–17.
Piaget, J. (1937) La construction du réel chez
constructivism
l’enfant. Délachaux & Niestlé: Neuchâtel.
English translation: Piaget, J. (1954) The
construction of reality in the child. Ballan-
tine: New York.
Pias, C. (ed,) (2003) Cybernetics | Kybernetik.
The Macy-Conferences 1946–1953. Vol-
ume 1 Transactions/Protokolle. Diapha-
nes: Zürich, Berlin.
Prinz, W. (1997) Explaining voluntary action:
The role of mental content. In: Carrier, M.
& Machamer, P. (eds.) Mindscapes: Phi-
losophy, science, and the mind. Univer-
sitätsverlag: Konstanz, pp. 153–175.
Quartz, S. & Sejnowski, T. J. (1997) The neu-
ral basis of cognitive development: A con-
structivist manifesto. Behavioral and
Brain Sciences 20: 537–596.
Riegler, A. (1998) The end of science: Can we
overcome cognitive limitations? Evolution
& Cognition 4: 37–50.
Riegler, A. (2001a) Towards a radical con-
structivist understanding of science.
Foundations of Science, special issue on
“The impact of radical constructivism on
science” 6: 1–30.
Riegler, A. (2001b) The cognitive ratchet. the
ratchet effect as a fundamental principle in
evolution and cognition. Cybernetics and
Systems 32: 411–427.
Riegler, A. (2003) Whose anticipations? In:
Butz, M., Sigaud, O., and Gerard, P. (eds.)
Anticipatory behavior in adaptive learning
systems: foundations, theories, and sys-
tems. Lecture Notes in Artificial Intelli-
gence. Springer-Verlag: New York, pp. 11–
22.
Riegler, A. (ed.) (2005) Heinz von Foerster –
in memoriam. Kybernetes 34.
Rusch, G. (1987) Erkenntnis, Wissenschaft,
Geschichte. Suhrkamp: Frankfurt.
Schmidt, S.J. (ed.) (1987) Der Diskurs des
Radikalen Konstruktivismus. Suhrkamp:
Frankfurt.
Schwegler, H. (2001) Physics develops unaf-
fected by constructivism. Foundations of
Science, special issue on “The Impact of
Radical Constructivism on Science” 6:
241–253.
Scott, B. (2001) Gordon Pask’s conversation
theory: A domain independent construc-
tivist model of human knowing. Founda-
tions of Science 6: 343–360.
Spivey, N. N. (1997) The constructivist meta-
phor: Reading, writing, and the making of
meaning. Academic Press: San Diego.
7
 conceptual
EDITORIAL
Steffe, L. & Gale, J. (eds.) (1995) Constructiv-
ism in education. Lawrence Erlbaum:
Hillsdale, NJ.
Vaihinger, H. (1952) The philosophy of “as if ”
(Translated by C. K. Ogden). Routledge:
London. German original published in
1911 as “Philosophie des Als-Ob.”
Van Bendegem, J. P. (1999) Why the largest
number imaginable is still a finite number.
Logique et Analyse 42: 107–126.
8 Constructivist Foundations
Varela, F. J., Thompson, E. & Rosch, E. (1991)
The embodied mind: Cognitive science
and human experience. MIT Press: Cam-
bridge, MA.
Watzlawick, P., Weakland, J. & Fisch, R.
(1974) Change: Principles of problem for-
mation and problem resolution. W. W.
Norton: New York.
Weaver, W. (1948) Science and complexity.
American Scientist 36: 536–544.
constructivism
Weber, S. (2005) Non-dualistische Medien-
theorie. Eine philosophische Grundle-
gung. UVK: Konstanz.
Westermann, G. (2000) Constructivist neural
network models of cognitive develop-
ment. PhD thesis at the University of Edin-
burgh.
Winograd, T. & Flores, F. (1986) Understand-
ing computers and cognition: A new foun-
dation for design. Ablex: Norwood, NJ.
 historical
OPINION
Thirty Years Radical Constructivism
Ernst von Glasersfeld A University of Massachusetts, evonglas@localnet.com
L ooking back in order to unravel how you
came to where you now think you are, is a
step in the creation of history. It’s only a
beginning, because proper history requires
several people to compare recollections and
sort out what is compatible among them.
What I present here, is a personal account and
lays no claim whatever to objectivity. It is sim-
ply what I remember.
Although I didn’t know it at the time, I
became a constructivist towards the end of
the 1940s when I joined the interdisciplinary
ABOUT THE AUTHOR
Ernst von Glasersfeld was born in Munich,
1917, of Austrian parents, and grew up in
Northern Italy and Switzerland. Briefly
studied mathematics in Zürich and Vienna
and survived the 2nd World War as farmer
in Ireland. Returned to Italy in 1946, worked
as journalist, and collaborated until 1961 in
Ceccato’s Scuola Operativa Italiana (lan-
guage analysis and machine translation).
From 1962 director of US-sponsored
research project in computational linguis-
tics. From 1970, he taught cognitive psy-
chology at the University of Georgia, USA.
Professor Emeritus, 1987. At present
Research Associate at Scientific Reasoning
Research Institute, University of Massachu-
setts. Dr.phil.h.c., University of Klagenfurt,
1997. – Books (among others): Wissen,
Sprache und Wirklichkeit, Vieweg Verlag,
Wiesbaden, 1987. Linguaggio e comunicazi-
one nel costruttivismo radicale, CLUP, Mil-
ano, 1989. Radical Constructivism: A way of
knowing and learning, Falmer Press, Lon-
don, 1995 (German translation: Suhrkamp
Verlag, Frankfurt am Main, 1996; also Portu-
guese, Korean, Italian translations.) Grenzen
des Begreifens, Benteli Verlag, Bern, 1996.
Wege des Wissens, Carl Auer Verlag,
Heidelberg, 1997. Wie wir uns erfinden
(with H. von Foerster), Heidelberg: Carl
Auer, 1999. (Italian translation: Rome,
Odradek, 2001). More than 240 paper
publications since 1960.
Constructivist Foundations 2005, vol. 1, no. 1
group of friends who worked informally and
unpaid with Silvio Ceccato at the debunking
of traditional theories of language and
knowledge. Ceccato, who had studied music
at the Conservatory in Milan but, to please
his father, had also got a doctorate in law,
became curious about what it was that gen-
erated the appeal of musical compositions
that were deemed beautiful. He read all he
could dig up about esthetics but found no
satisfactory answer. He went on to read epis-
temology but found the accounts given of
knowledge and the functioning of language
equally disappointing. It was only when he
came across Bridgman’s operationalsm that
he saw a direction to develop his own think-
ing. To define the meaning of words by
means of other words, he realized, could
never show what meaning consists of. To
find out, you had to uncover how it was
made, you had to analyze the operations that
produced it. This was tantamount to analyz-
ing how knowledge could be constituted.
Ceccato spoke of operational awareness,
deliberately leaving implicit that he was
focusing on the construction of knowledge.
The word “knowledge”, he felt, had to be
avoided because it tends to lead to the notion
of representation in the sense of replica of a
pre-existing reality. Knowledge could mean
only one thing to him: the reconstruction of
results of prior operations. If the reconstruc-
tion matched them, it was true, if it didn’t it
was false.
In 1955 Ceccato went to a conference in
London and met Colin Cherry, the communi-
cation expert. Cherry suggested that the best
way for an unconventional theory of language
to make headway at that time was to apply it
to machine translation. Shortly after that,
Ceccato founded the Center for Cybernetics
at Milan University and asked me to help with
writing a proposal for research on machine
translation. The research was sponsored by an
office of the US Air Force, but lasted only for
three years. Some time after it ended I moved
to the United States with a small project of my
own, still based on Ceccato’s theory of lan-
guage (Ceccato 1969).
radical constructivism
One cornerstone of Ceccato’s theory was
that syntax could not be understood without
analysis of conceptual relations and that the
composition of sentences involved many
more such relations than were recognized in
traditional grammars. On the level of lan-
guage, most of these conceptual relations
were marked by prepositions, conjunctions,
and particles of that kind. In English, for
example, there are more than a hundred of
these “parts of speech” and most of them are
highly ambiguous with regard to the concep-
tual relations they indicate. In other lan-
guages the situation is similar, but the range of
ambiguity of prepositions is different in each
language. During the ten years that I worked
in computational linguistics, the problem of
translating prepositions kept turning up and
continually reinforced my notion that the
structure of the experiential reality we live in
depends in many ways on the language we
happen to grow up with. It was a simple con-
firmation of Whorf ’s contention that con-
cepts differ from one language to another
(Whorf 1956).
“You know, much of
what you suggest has
been said by Piaget”
Jean Piaget coined the expression “the
construction of reality” in the title of one of
his fundamental books on cognitive develop-
ment in children (Piaget 1937). It was ironical
that Ceccato, although vigorously opposed to
the Platonic notion of pre-existing eternal
ideas, never came to see the importance of
cognitive development for any theory of
knowledge. When I asked him about a Piaget
reference I had found, he told me not to
bother because Piaget was concerned with
children. Consequently I did not become
acquainted with the work of the inventor of
constructivism in the 20th century until my
research project in the United States was ter-
minate. Like many others it was the victim of
9
 historical
OPINION
an internal upheaval at the sponsoring Air
Force office. My team was adopted by the Uni-
versity of Georgia and I was assigned to the
department of psychology. Charles Smock,
one of the few members of that large depart-
ment who were interested in cognitive psy-
chology, very generously shared a couple of
courses with me to introduce me to teaching
(which I had never done except on ski slopes).
He had spent several years at Geneva and as
we began to talk about the problems of cogni-
tion, he said to me one evening: “You know,
much of what you suggest has been said by
Piaget.” So I began to read and to discover
Piaget’s “genetic epistemology”. Fortunately
Smock had nearly all the original French pub-
lications of the Geneva School and I could see
how pervasive the notion of individual con-
struction was in Piaget’s thinking. In the
English translations I later had to use in teach-
ing, this notion is much subdued if not elim-
inated, because most translators tended to
assimilate everything to their conventional,
realist views.
At about that time, Heinz von Foerster
sent me a batch of his publications. We had
met a couple of years earlier and found that
some of the problems we were interested in
were much the same. When I read his “An
epistemology for living things” (Foerster
1972), it struck me as extraordinary that there
were three people who had come to quite sim-
ilar conclusions by very different ways: Cec-
cato, who had come to see the activity of
knowing rather like a composer’s work of cre-
ating chords and then tunes by combining
notes in sequences that could be specified by
a relatively small number of elements; Heinz,
who came to constructivism by way of formal
logical and neurophysiological consider-
ations; and myself, who had been led to it pro-
saically by the differences among linguistic
realities. All three of us, however, had been
driven by unquenchable dissatisfaction with
the traditional accounts.
In 1974 Smock and I put together a report
with the title: “Epistemology and education:
The implications of radical constructivism
for knowledge acquisition”. I wrote a chapter
assembling some philosophical precedents
and presenting my interpretation of Piaget’s
theory. It was the first time the epithet “radi-
cal” was used. It was intended in the sense that
William James (1976) had used in his radical
empiricism, i.e., meaning “going to the roots”
10
or “uncompromising”. I chose it because at
the time many developmental psychologists
were mentioning Piaget’s constructivism but
without going into its epistemological impli-
cations. What they called construction
seemed to refer to the fact that children
acquire adult knowledge not all at once, but in
small pieces. I did not think that this was a
revelation and therefore called their approach
“trivial constructivism”. It was clearly no way
to gain the friendship of traditional psycholo-
gists but in the long run it did not do much
harm.
It was clearly no way to
gain the friendship of
traditional psychologists
but in the long run it did
not do much harm
Smock collaborated in research with Les
Steffe, who headed the Georgia section of the
Follow Through Program for children in the
first classes of elementary school. Steffe was a
Piagetian and his specialty was mathematics
education. During the first years of school,
“math ed” is of course all arithmetic. On the
basis of Piaget’s “clinical method”, which
included interviewing children about what
they were thinking, rather than merely
observing what they were doing, Steffe was
beginning to build hypothetical models of
how children go about solving simple prob-
lems of addition and subtraction. He called
the method “teaching experiments” and a
crucial part of it was the use of videotapes,
usually two or three minutes long. He, a grad-
uate student of his, the philosopher John
Richards, and myself would spend countless
hours viewing these tapes and trying to agree
on what we gathered from them. We had
heated arguments and for all of us it was a
powerful lesson, hammering in the funda-
mental fact that what one observer sees is not
what another may see and that a common
view can be achieved only by a strenuous
effort of mutual adaptation. In 1983 we
jointly published a book, which, by means of
many protocols, illustrated our view of how
children, through various stages of counting,
radical constructivism
develop a concept of number (Steffe et al.
1983).
The Follow Through report had practi-
cally no circulation, but the papers we had
published during the preparation of the book
and the book itself created a stir. In July 1987,
at the meeting of the International Group of
Psychologists in Mathematics Education
(PME), Mimi Sinclair (long-term Piaget col-
laborator) and I defended constructivism
against two prominent figures in the field.
Their attacks were so unscholarly and vicious
that they produced the opposite effect: they
helped to establish constructivism as a valid
basis for teaching mathematics.
As approach to a theory of knowing, con-
structivism became established at the Inter-
national Conference on the Construction of
Realities that Heinz von Foerster organized
with the help of Francisco Varela in San Fran-
cisco in 1978. There were very interesting pre-
sentations by Bateson, Goffman, Polanyi’s
daughter, and others, , and it’s a pity that
Varela, who collected all the texts, never got
round to publishing them. It was at this con-
ference that I met Paul Watzlawick. He had
developed his own constructivist ideas start-
ing from Schopenhauer and through his prac-
tice in family therapy. The problems he was
asked to solve, arose, he saw time and time
again, from the fact that the realities individ-
ual members of a family lived in were not
compatible with one another. Watzlawick
asked me to expand and translate the talk I
had given so that he could use it as the intro-
ductory piece to “Die erfundene Wirklich-
keit” (The invented reality), a book he was
about to publish in Germany. It came out in
1981 and has by now been reprinted 18 times.
It has done more than any other to spread the
notion of cognitive construction. An English
version of the book was published in the
United States by Norton in 1984, but as far as
I know the first edition was never sold out and
the book is no longer listed by the publisher.
The geographic difference in dissemination
has remained characteristic of constructivist
ideas in general. In the United States and
England, radical constructivism is mentioned
in the field of science and mathematics educa-
tion. In Germany, Austria, and Italy it has
become the subject of lively discussions
among philosophers, psychologists, educa-
tors, and therapists, and quite a number of
books have been published on the subject.
Constructivist Foundations
 historical
OPINION
The discrepancy of dissemination applies
also to the works of Humberto Maturana
(popular in Europe, but rarely if ever cited in
the US). He calls his subject “Biology of Cog-
nition” and defines it as “an explanatory prop-
osition that attempts to show how human
cognitive processes arise from the operation
of human beings as living systems. ” It falls
squarely among the efforts that try to explain
knowing as the result of self-organizing con-
struction, but he has never been comfortable
with the term “construction” and prefers to
speak of “bringing forth”.
We can only check the
coherence of our
constructs with other
experiences
Maturana was a student of Warren McCul-
loch and Jerry Lettvin and in collaboration
with them and Walter Pitts published the
paper “What the frog’s eye tells the frog’s
brain” (Lettvin et al. 1959). McCulloch called
it the first step into experimental epistemol-
ogy and it can be considered the first experi-
mental confirmation of radical constructiv-
ism. The frog builds up the image of a
desirable (i. e. digestible) insect from a pat-
tern of neural signals, regardless of what may
have triggered these signals.
Commentators, especially in Germany,
have somehow come to believe that Mat-
urana’s biological theory of cognition is the
source and empirical justification of radical
constructivism. This is a misconception.
When Protagoras wrote “Man is the measure
of all things” and added “being is the same as
to appear to someone” (Diels 1957) he sug-
gested that what we consider knowledge is of
our own making. It has to be, because we can-
not check what we experience with what lies
beyond the experiential interface; we can only
check the coherence of our constructs with
other experiences. Radical constructivism is
delighted to have biologists developing theo-
ries of autopoiesis and physicists developing
theories of quantum mechanics that are com-
patible with a constructivist theory of know-
ing, but the theory of knowing does not
depend on these more or less empirical find-
2005, vol. 1, no. 1
ings.1 It is based on the simple realization
that, as our thinking, our conceptualizing,
and our language are developed from and in
the domain of our experience, we have no way
of incorporating anything that lies beyond
this domain.
This last point is also what distinguishes
radical constructivism from the “methodical
philosophy” developed by Paul Lorenzen and
others at the University of Erlangen (Loren-
zen 1987). Much of what Lorenzen formu-
lated agrees with radical constructivism, e.g.,
“The theory of concepts I have sketched is not
a theory concerning existing things; it is not
an ontology. Concepts as described here
belong to our actions; they are interpreted
here functionally rather than ontologically”
(p. 12). But ontology still seems to enter into
his system when, in his essay on the “Logical
structure of language” (p. 105ff), he speaks of
objects being “part of nature” and of “per-
sonal observation” as though the observed
were externally given. That there are serious
disagreements between radical and “method-
ical” constructivism is indicated by the fact
that Peter Janich, the main present-day repre-
sentative of the Erlangen School, vigorously
attacked my way of thinking at the conference
on “Reality and Worldmaking” (Fischer &
Schmidt 1998, pp. 65ff).
…what we call
knowledge can be
“constructed” without
reference to anything
outside the experiential
confines
There have been many others whose
thinking contained constructivist elements –
Rorty, Feyerabend, Bruner, Dewey, Brouwer,
Fleck, Bogdanov, and probably some I have
never heard of; but, apart from the radical
constructivists I mentioned above, I know
none who tried to model the generation of
knowledge without reference to an ontic real-
ity. Yet, if the skeptics are right – and two
thousand five hundred years of Western phi-
losophy have not been able to prove them
wrong – neither the structure nor the texture
radical constructivism
of an ontic reality are accessible to human
knowing. It therefore seems legitimate to try
and conceive a model that may show how
what we call knowledge can be “constructed”
without reference to anything outside the
experiential confines.
The notion of model, however, inevitably
contains assumptions that lie outside the
domain the model may explain. In the case of
constructivism it is the assumption of a con-
sciousness that is able to remember, to reflect
upon experience, and to develop likes and dis-
likes. It is the least a model of cognition must
assume.
As for the future, I would suggest that
more work be done regarding the question
where the notions of society and of other con-
structing subjects come from. They are
needed to establish an intersubjective viabil-
ity of conceptions. I have sketched a way in
which a child could construct the notion of
another entity rather like itself (Glasersfeld
1979, 1989), and there are hints in Piaget’s
writings; but someone should systematically
ask three-year-olds “How is the cat different
from your teddy bear?” and six-year-olds
“How is your sister different from the cat?”
Clinical interviews would make the stories I
invented more plausible.
Note
1. Psychologists, have known since Johannes
Müller (1838) that the sensory signals in
the nervous system may contain quantita-
tive data about a stimulus but apparently
no qualitative “information”; it is a source
of puzzlement why so little effort has been
made to unravel how the qualitative as-
pects of sensation originate.
References
Ceccato, S. (ed.) (1969) Corso di linguistica
operativa. Longanesi: Milan.
Diels, H. (1957) Die Fragmente der Vorsokra-
tiker. Rowohlt: Hamburg.
Fischer, H. -R. & Schmidt, S. J. (eds.) (2000)
Wirklichkeit und Welterzeugung. Carl
Auer Systeme: Heidelberg.
Foerster, H. von (1972) An epistemology for
living things. In: Motin, E. & Piattelli-Pal-
marini, M. (eds.) L’unité de l’homme. Édi-
11
 historical
OPINION
tions du Seuil: Paris: (Reprinted as BCL
Report 9. 3).
Glasersfeld, E. von (1979) Cybernetics, expe-
rience, and the concept of self. In: Ozer, M.
N. (ed.) A cybernetic approach to the
assessment of children: Toward a more
humane use of human beings. Westview
Press: Boulder, CO, pp. 67–113. Also avail-
able at http://www.vonglasersfeld.com/
papers/056.pdf
Glasersfeld, E. von (1989) Facts and the self
from a constructivist point of view. Poetics
18: 435–448. Also available at http://
www.vonglasersfeld.com/papers/123.pdf
James, W. (1976) Essays in radical empiri-
cism. Harvard University Press: Cam-
bridge. Originally published in 1912.
12
Lettvin, J. , Maturana, H. R. , McCulloch, W.
S. & Pitts, W. (1959) What the frog's eye
tells the frog's brain. Proceedings of the
IRE 47: 1940–1959.
Lorenzen, P. (1987) Constructive philosophy.
MIT Press: Cambridge.
Müller, J. (1838) Prinzip der spezifischen
Sinnesenergie. In: Handbuch der Physiol-
ogie des Menschen (vol. 1). Hoelscher:
Coblenz.
Piaget, J. (1937) La construction du réel chez
l’enfant. Delachaux et Niestlé: Geneva.
Smock, C. D. & Glasersfeld, E. von (eds.)
(1974) Epistemology and education: The
implications of radical constructivism for
knowledge acquisition (Report #14). Fol-
low Through Publications: Athens GA.
radical constructivism
Steffe, L. P. , Glasersfeld, E. von, Richards, J. &
Cobb, P. (1983) Children's counting types:
Philosophy, theory, and application. Prae-
ger: New York.
Whorf, B. L. (1956) Language, thought, and
reality. MIT Press: Cambridge.
Links
Further material on Ernst von Glasersfeld can
be obtained from his new hormepage at
http://www.vonglasersfeld.com/
Received: 12 August 2005
Accepted: 30 August 2005
Constructivist Foundations
 historical
OPINION
Ceci n’est pas Heinz von Foerster
Diederik Aerts A Vrije Universiteit Brussel, diraerts@vub.ac.be
In 1995, the Leo Apostel Centre in Brus-
sels, Belgium, organised an interna-
tional conference called “Einstein meets Mag-
ritte”. Nobel prize winner Ilya Prigogine held
the opening lecture at the conference, and
Heinz von Foerster’s lecture was scheduled
last. He was the oldest among the invited
speakers, and Prigogine was the second-old-
est. Heinz von Foerster, having noticed this
arrangement, commented at the conference
dinner, “I am the oldest, and Prigogine is the
second-oldest. He opened the conference and
I will close it, that is really perfect”.
Heinz von Foerster was enchanted by the
conference theme “Einstein meets Magritte”
and – in the spirit of surrealist Belgian painter
René Magritte – had chosen an appropriate
title for his talk: “Ceci n’est pas Albert Ein-
ABOUT THE INTERVIEWER
Diederik Aerts is professor at the ‘Brussels
Free University (Vrije Universiteit Brussel -
VUB)’ and director of the ‘Leo Apostel Cen-
tre (CLEA)’, and interdisciplinary and inter-
university (VUB an the universities of Gent
and Leuven) research centre, where
researchers of different disciplines work on
interdisciplinary projects. He is also head of
the research group ‘Foundations of the
Exact Sciences (FUND)’ at the VUB. He is
secretary of the ‘International Quantum
Structures Association (IQSA)’ and editor
of the international journal ‘Foundations of
Science (FOS)’. He is a board member of
the ‘Worldviews group’, founded by the phi-
losopher Leo Apostel, which investigates
the possibility of constructing integrated
worldviews, taking into account the recent
scientific findings. He was the scientific and
artistic coordinator of the ‘Einstein meets
Magritte’ conference, where the world’s
leading scientists and artists gathered to
reflect about science, nature, human action
and society.The material of the conference
has been published in an eight-volume
series by by Kluwer Academic and VUB-
Press.
Constructivist Foundations 2005, vol. 1, no. 1
stein”. He asked to have a little picture of Ein-
stein printed next to the announcement of his
talk in the conference programme booklet, a
wish we were glad to grant. As he explains in
the interview below, over time he had become
convinced that semantics-related problems
are more important – even for the ontological
nature of things – than he estimated himself
in his earlier days. He saw a striking resem-
blance in the ways both Einstein and Magritte
revolutionized their respective fields of
inquiry and activity by making explicit the
influence of the semantic level on the syntac-
tic level. Like no one before, Magritte played
with the insight that ‘the model is not the
thing’, while Einstein shook the foundations
of all of physics with his theory of relativity.
Here, observation influences space as well as
time – two physical concepts that until then
were largely considered to be purely objective,
or even a priori.
The way Heinz von Foerster arrived at the
‘Einstein meets Magritte’ conference is itself a
surrealistic story of pure Magrittean quality.
It is worth telling it in some detail. On Friday
evening, the last evening of the conference,
there was to be a performance by the dance
and theatre group of Anne Teresa De Keers-
maker. To allow the troupe to prepare for their
performance, the large university hall, where
the Einstein meets Magritte conferences were
normally scheduled, therefore had to be made
available to them on Friday afternoon. That is
why the conference booklet announced that
the last three lectures of the invited speakers,
including von Foerster’s, would take place in
an old cinema building on the premises of a
military site close to the university campus.
The organising committee of the conference
had provided all invited speakers with vouch-
ers for free taxi rides in Brussels on the days of
the conference. The vouchers were also
intended to be used on arrival at the airport to
get to the location where the conference was
held. However, for some mysterious reason,
rather than taking him from the airport to the
university, the taxi driver drove Heinz von
Foerster to the university clinic, at the oppo-
site end of town. This was to be only the
second order cybernetics
beginning of von Foerster’s Magrittean odys-
sey. It took some time before von Foerster was
told that the university clinic was not the place
where he was supposed to be. He then took
another taxi, and showed the driver the con-
ference booklet in which his talk was
announced, thinking that this would get him
to the right place. He was right, for this time
the taxi did take him to the military site. But
this was not to be the end of his predicament.
The barracks site is not an area where civilians
can easily enter, and it was only after von Foe-
rster had marshalled his persuasive power,
wielding the conference booklet as conclusive
evidence, that the guards at the gate allowed
him to enter the military zone. He was
escorted to the cinema building, which had
conference announcements posted on its
walls. Indeed, some of the young conference
assistants, during earlier visits undertaken to
check that it could be used as an alternative to
the university hall that Friday afternoon, had
been eagerly putting up the posters at the cin-
ema entrance. But neither the military, nor
the guards at the gate, nor any of the cinema
staff watching a movie (“Dracula!”, von Foer-
ster would comment later in a mysterious
voice) now seemed to know anything about
the conference.
All in all, it took von Foerster six long
hours to get from the airport to the campus –
longer than it took him to cross the Atlantic
from the United States. I remember very viv-
idly the state in which he arrived just after his
unexpected adventure. He was all excited, a
bit angry, but fascinated at the same time. He
told about the amazing excitement he felt
when being walked several times past and
close by military horses standing in line.
“The horses snorted and stamped like hell
each time I walked by, they knew that this was
a person who did not belong there, but …
They definitely were the only ones to know
this for like three full hours… so you see
everything is relative, Einstein and Magritte
would agree, in certain circumstances a horse
can be more intelligent than a human being.
These horses would have let me escape imme-
diately.”
13
 historical
OPINION
Despite his six-hour odyssey and three-
hour captivity in Belgian barracks, he was
delighted to grant the organisers the following
interview, in which he tells us about an even
longer journey – that of his remarkable life and
scientific career. The transcription follows
Heinz von Foerster’s wording as closely as pos-
sible in order to capture his unique conversa-
tional charm. The endnotes were added later.
“Heinz, what you said
was very interesting
but how you said it
was abominable”
Heinz von Foerster: We are now supposed to
have an interview!?
Diederik Aerts: Yes.
HvF: Who is interviewing whom?
I am interviewing you.
HVF: You are interviewing me, okay. I am
ready for your interview.
Mr von Foerster, you are considered one of
the founders of the very important discipline
of “cybernetics”. It is true that this is also a
discipline now, but it’s kind of more interdis-
ciplinary than other disciplines, I would say.
Can you tell us a few words about when you
began with these things?
HVF: A long time ago.
And what inspired you?
HVF: Oh well, I was very much inspired by
other people. Now I tell you the story – it’s a
funny story. I left Vienna on an invitation of a
friend of my wife to come to New York. This
was in 1949. So I sailed with the Queen Mary
over the ocean, and unfortunately there were
tremendous storms and spring storms over
the Atlantic. Everybody was seasick. But I grew
up very close to one of the great amusement
parks in Vienna so I don’t get seasick at all. So
I was the only one, one of 2000 inhabitants of
Queen Mary, who was not seasick. I had about
20 waiters in the dining room waiting for me.
I arrived in New York healthy and well fed. By
a very peculiar accident, one very interesting
gentleman, an American neuropsychiatrist by
14
the name of Warren McCulloch1, one of the
leading scientists in neuropsychiatry, got hold
of a book – a small booklet I should say –
which I wrote in November of the previous
year on the quantum theory of memory. He
got hold of that thing. He read it and contacted
me in New York the third or second day I
arrived. “Mr Foerster this sounds very inter-
esting what you wrote. I’m the chairman of a
conference which will take place in New York
in about five, six days on circular causal and
feedback mechanism in biological and social
systems.” And I said “What?” I don’t under-
stand a word of it.” He said, “Okay; why don’t
you get yourself a book which has just
appeared in the bookstores. It’s called ‘cyber-
netics’ by Norbert Wiener2, a mathematician;
try to get that book and I’ll invite you to a con-
ference which takes place in 14 days from
today in New York, you have only to go”. I said:
“Wonderful, I shall do that”.
My vocabulary in English was then about
30 words – yeah, maybe 32; anyway I got that
cybernetics book by Norbert Wiener. I mean
when you read a scientific paper it’s always the
same; I mean there are the Latin words and
formulae you understand. I got the point of
what Norbert Wiener was talking about. And
14 days later I came to the conference which
was at a beautiful old fashioned hotel which
does not exist anymore, on 5th Avenue. I came
to the conference. To my surprise there were
only 20 people. It was not a circus for an audi-
ence. It was 20 people talking with each other.
And these 20 people were the crème de la crème
of American science. Warren McCulloch, one
of his guests was John von Neumann3, the
man who began the computer revolution,
Norbert Wiener himself was there, Gregory
Bateson4, the anthropologist, his wife at that
time Margaret Mead5, and I can rattle of
names and names. 20 absolutely outstanding
people. So I came in as a greenhorn, had to give
my story about …
How old were you then?
HVF: This is of course ‘49. I must have been 38.
So I give my story. And everybody was inter-
ested and I couldn’t of course tell my story
with 25 or 32 words in English.
And your story was about?
HVF: My story was about quantum theory of
memory. So fortunately there were 2 or 3
gentlemen who spoke fluently German and
second order cybernetics
English and whenever I was lost they took
over and translated what I said etc. etc. At the
end of the whole thing I had to leave the
room. I was called in after an hour and they
said “Heinz, what you said was very interest-
ing but how you said it was abominable. We
have thought a way how to teach you English
the fast way. We appoint you to become the
editor of our transaction.” – Can you imag-
ine? I mean, this only in America can happen.
I said, “For heaven’s sake, my English is not
good!” – “No, no, no that’s exactly why we
appoint you.” – I said, “But then you have to
change the title of that conference because I
can’t even pronounce the title of that confer-
ence!” And that is now an interesting point
and this is why I tell the story. I said, “Why
don’t you call the conference ‘cybernetics’?” I
just had read the Wiener book, I had the feel-
ing that what these people were doing was
talking cybernetics. Everybody thought that’s
a wonderful idea and applauded. And Norb-
ert Wiener was sitting next to me. They
applauded of course to Norbert. And he was
so moved had tears in his eyes that his col-
leagues accepted this funny word of “cyber-
netics.” He had to get up and walked out of
the room because he wanted to hide his emo-
tions.
So then they adopted the word “cybernet-
ics” for the conference, and with it my name
was immediately linked to the cybernetic
publication of the Josiah Macy foundation
meeting.
Because you did the editing?
HVF: I edited the first published conference
on cybernetics6.
And you learned a lot of English?
HVF: I learned English; I realised after about
two months I got 10 cm of manuscript which
was typed from a legal typist who was sitting
in the court – a stenographer – overtyping all
the conferences.
And you stayed in the States then?
HVF: Yeah, then I stayed with Warren McCul-
loch and met with people of the University of
Illinois. Since at that time they were looking
for a head of a particular section of research
and I happened to be just the man to do that,
I accepted that invitation and stayed with the
university of Illinois for the next 30 years or
something like that.
Constructivist Foundations
 historical
OPINION
That’s a beautiful story.
HVF: So I stayed there and my wife and chil-
dren came, after a couple of months to get
them a visa and permits etc. This was tremen-
dous bureaucracy, I had to climb over the
bureaucratic fences which took me about a
half a year more to get the permits and then
they could come.
And then you started to work?
HVF: Yeah, I was not permitted to do any-
thing during the time when I was only a visi-
tor in the United States. To wait until all these
things were cleared and then I took my job at
the University of Illinois as the director of the
electron tube research lab. And from there on
I continued my connection with the Macy
people and with the cybernetics people and
then after seven years … you know, six years
you have to be at a university then you can get
these wonderful sabbatical leaves. In this year
I thought I should learn more about biol-
ogy, neurology, neurophilosophy, and
neuroanatomy and went to McCulloch to
MIT, to the Massachusetts Institute of
Technology, for half a year. Then I was
working in Mexico
with a friend of
Norbert Wieners,
Arthur
Rosenblueth7, who
wrote with me
some early papers
about cybernetics
and I worked then
in Rosenblueth’s
laboratory in
Mexico City for
half a year. And
then I came back
and said now I
start a new lab and then I
started the Biological Computer Lab7 in 1958.
What were the problems that fascinated you
in this period?
HVF: At that time? Probably still the same
things. Today I feel more and more that
semantic problems are involved in our under-
standing of things. These are notions of
course which were already produced by Wit-
tgenstein and other philosophers who hap-
pened to be reasonably well known by me
because this is all coming out of the so called
Vienna circle. The language problem, the lin-
2005, vol. 1, no. 1
guistic problem, the understanding, the
hermeneutic problem – all these things are
relatively familiar to me. So I think there is a
point which I will also address perhaps
tomorrow on my paper because I will cele-
brate Magritte and Einstein as essentially
semanticists, not so much as physicist and
painter, but as people who were interested in
what is the meaning of a statement, what’s
going on when I make an explanation in
which we connect some points with other
points, what are the semantic links. So these I
think are the links between Magritte and Ein-
stein.
I didn’t think myself of this! You mean Ein-
stein when he was carefully reflecting on the
observers, how time and space got com-
pletely transformed by these indeed ‘seman-
tic’ reflections?
HVF: Exactly, precisely! I will bring about
tomorrow. I will not give it away at the
moment in our interview but I
found a wonderful perception of
Einstein in his early days when he
was wrestling with a notion which
is later found in the special theory of
relativity where
he came up
with a meta-
phorical state-
ment of his. If
you read that
today and you
think he had
told his notion
to a physician at
an insane asy-
lum he would
probably be put
inside that asy-
lum. This man is
crazy. He thinks
about things that don’t exist, he is talking
about them freely as if they were existing. I
thought this is a wonderful example of inge-
nuity.
You’ve made me curious. But I’ll have to wait
until tomorrow. Now, you are in a sense the
example of an interdisciplinary interested
person.
HVF: No disciplines, no disciplines, no disci-
plines … interdisciplinarity is already disci-
plinarity.
second order cybernetics
But I think your notion of transdiscipli-
narity is interesting, very interesting. Going
beyond the disciplinary segmentation and
fragmentation. Still interdisciplinarity says,
“I’m a chemist who is interested in physics” –
“Ah! That means you are a chemical physicist
or a physical chemist.”
Still disciplines. And it is indeed dangerous
that one should not make of the interdisci-
plinary a new discipline.
HVF: Precisely! Exactly! These are the seman-
tic traps. Language is taking over instead of
you taking over the language.
And the trap is very concrete the moment
you want to find funding.
HVF: Yeah! you must play the trap. Absolutely!
You had to put the trap to others who are set-
ting the trap. Otherwise they will be disap-
pointed like a guy who’s putting up a mouse-
trap and no mouse goes into it. The mouse has
to get into the trap to satisfy the trapper.
So you need to have a false mouse.
HVF: Yes, you have a very good point. You
need a false mouse. So they snap over the false
mouse and you can do your own thing.
But when we want funding we must present
it as a discipline of course, the discipline of
interdisciplinarity, but we would like to keep
it open and not fall into this trap.
HVF: Yeah! For many cases you could find a
good name which does not betray or belie the
point you are trying to make, but at the same
time would be understood by sponsors which
would be governmental agencies or founda-
tions – something like that. That would be
very important. A name for this global prob-
lem from a linguistic point of view.
Yes, it’s true. So okay, we’ve now arrived at
this period where you were in Illinois.
HVF: Yeah! As I said, I took a sabbatical and
started to make a foundation of the biological
computer lab. This was of course strongly
influenced from the perspective of this cyber-
netic group of McCulloch, Wiener, Von Neu-
mann. I had a very good rapport with all these
gentlemen. Von Neumann, for instance, who
was at Princeton at that time, invited me again
and again to come for several reasons. One is
of course that he went to school in Budapest,
which was part of the Austrian Kaiserliche und
15
 historical
OPINION
Königliche Habsburg monarchy. So he wanted
to know what is Vienna doing under Russian
occupation and was of course very worried
with members of his family in Budapest. So I
could tell him a little about what the Russians
were doing. They were occupying the region,
Budapest and Vienna. And Norbert Wiener’s
wife is from Silesia, from Schlesien. I was in
Silesia the last month of the existence of Ger-
many. So she wanted to know, “Heinz can you
tell be about this and that, how did the people
flee, how did they get out of the Russian occu-
pation?” etc. etc. So in both of these cases per-
sonal interests brought me closer and closer to
these fabulous people.
With Margaret Mead I had a very good
rapport. At once she was, I think, inventing
the idea of teaching Heinz English by making
him the editor of the transactions.
This was her idea?
HVF: Yeah, I think this was her idea, I have
this suspicion. It’s the woman who had this
crazy idea but the men went along, and I
admired her very much. I mean in this cyber-
netics meeting with all these brilliant men,
there is, of course, a tendency that they show
off their knowledge, show their peacock
feathers, and so on. But at the moment when
that happened Margaret put her fist on the
table, “We know that you can speak Greek!
Why don’t you talk about what we are talking
about!” It was fabulous how she was running
the show, the only woman in that group. And
with many others, Heinrich Klüver9, who
came also from Germany, I had a rapport
again and again. He was in Chicago. I was in
Illinois. We met each other again and again
and then I met some other friends of the early
Vienna period, Karl Menger10, a famous
mathematician, incredible mathematician,
the son of a very famous economist, Carl
Menger11, the papa with C, the son Karl is
with K, so you have a distinction between
papa and son. The son is dropping in the
alphabet from C to K, but anyway. Karl
Menger was one of my teachers at university –
a very young teacher. He was not very much
older than I was as a student. He must have
been 30 when I was 22 or something like that.
I was very fond of Menger’s notion of mathe-
matics. This was a confirmation of the con-
structivist notion of the 19th century.
The famous battle between two schools of
mathematics. One says mathematics is dis-
16
covered. The rules of mathematics, numbers
and everything is floating in the universe and
from time to time God permits me, takes off
the veil, and I see this incredible truth – this is
one school of mathematics. The other school
says it is our invention. We invented the num-
bers, we invented the calculus, it is our game
of doing things. Of course these two camps
cannot talk to each other. But I was immedi-
ately taken by the inventors and I followed the
inventor school of mathematics – the con-
structivist school of mathematics; in the 19th
century they were called the intuitionists. So
there were two big camps, the Kronecker12
camp and the Hilbert13 camp. Hilbert was the
formalist; he said, “I have nothing to do with
these guys and I do my formulae and I’m
working with them!” And Kronecker and his
people said, “You invent all that stuff, I mean,
it’s quite clear.”
In which camp was von Neumann?
HVF: Von Neumann was a constructivist,
yeah, yeah. So we had a magnificent time
with each other.
He had this machine
standing in his lab in
London and you were
allowed with a
sledgehammer to
destroy 500 tubes
At the computer lab, you were working on
many things – what would you yourself
point out as the most important things that
you were doing?
HVF: I will tell you the names and you will get
a little bit of the idea. One of the early partic-
ipants was Ross Ashby14, a British psychiatrist
who turned to cybernetics early in his life. He
realised that the standard notion of dysfunc-
tion–function of the brain is not the way to do
it. One must have a deeper knowledge about
the whole thing. So he became interested in
life systems which function even if you
destroy large sections of that system. He built
the first large machine out of, I think, 2000
vacuum tubes – at that time it was the only
second order cybernetics
way but today you would do it with transis-
tors. He had this machine standing in his lab
in London and you were allowed with a
sledgehammer to destroy 500 tubes. It was
doing some particular things – I don’t know
what but anyway – you went there with a
sledgehammer… for a moment the pointers
wiggled but then brrrr! moved into stability.
So you could see if systems are constructed so
that they can self repair – autodynamics etc
etc. – it is the key for the brain. He had a won-
derful example.
The example of a man who was working in
a rolling mill, a big rotary mill where, for
instance, steel rods are being built and what
you do is you come with a thick piece of metal,
which is red hot, go through the first wheels
where it’s squeezed to a smaller one and
becomes longer and is shot into the next mill,
shot and shot and shot, and becomes thinner
and thinner, longer and longer. At one instant
this rod – glowing – missed the next mill, shot
into the eye of a worker, penetrated the brain,
was sticking out at the other end. They
stopped the mill, carried the man, cut the
sticks, and brought him to Ross Ashby. The
man was talking! The man could talk! They
removed the rod because the man could die.
Of course the man had difficulties talking but
it was a functional creature. This is an exam-
ple of self repair, of switching certain func-
tions of the brain into the part which is still
intact, taking over what was lost, and the man
still functioned.
So Ashby became interested in these very
general systemic structures which maintain
themselves.
And then I had Humberto Maturana15 as
a guest for one or one and a half years who
brought in the autopoietic notion generated
in my lab. Things of that sort. And then at the
last contract with the institute of health, the
national institute of health, we worked on
haematology and population dynamics etc.
etc.
So we were playing in many different
domains but with the same, should I say
philosophical–epistemological attitude. It
produces a lot of different things and this I
think is what I enjoyed mostly.
Francisco Varela16 came in too?
HVF: Varela never was in, I only knew Fran-
cisco as a young man he was first staying in
Chile with…
Constructivist Foundations
 historical
OPINION
He had to leave today, I think, but he wanted
to meet you.
HVF: I met him this morning, we had break-
fast together. I was very glad to meet him. And
I invited him again to come see me in Califor-
nia, but he said he was so busy and he has dif-
ficulties to go there. Anyway Francisco is a
dear, dear old friend of mine, only that he was
never at the BCL. Maturana was at the BCL,
some other people from Chile. Ashby I
already mentioned. Then Gordon Pask17,
perhaps. He is well-known. Gordon was
spending good 2 years at BCL.
How did you manage to have this center?
You were the director and managing it?
HVF: I was managing the whole thing, not
only the intellectual and scientific part but
also the financial part. I did it all and I must
say I’m still very impressed how much money
I brought in.
There was constantly a group of 30 people.
May be 20 students, graduate students, some
of them making their Ph.D. thesis, the fasci-
nating thing was that they were paid for doing
their doctoral dissertation because they were
backed in the research projects.
I remember a young man, no, a professor of
the department of mathematics said, “Heinz
I know you have all these crazy people
Notes
1. Warren McCulloch (1899–1969) neuro-
physiologist and cybernetician working
on mathematically neural network model-
ing. Together with Walter Pitts he showed
in the 1943 paper “A logical calculus of the
ideas immanent in nervous activity” how
a neural network consisting of binary
threshold neurons can carry out logical in-
ferences at the level of first-order logic.
2. Norbert Wiener (1894–1964) became
eventually known as the founder of cyber-
netics based on this book (“Cybernetics or
Control and Communication in the Ani-
mal and the Machine”) which he pub-
lished in 1948.
3. John von Neumann (1903–1957) not only
made contributions to various mathemat-
ical fields but also had a major impact on
quantum physics, game theory, set theory,
computer science, and economics.
2005, vol. 1, no. 1
around you. I have a young man who twice a
week takes the train, goes to Chicago, sees his
psychiatrist, comes back, and is incapable of
doing anything. Would you like to see that
man?” – “Of course, send him over!” So he
sent him over here. Here came a very shy
man to me in my lab. I said, “You are a math-
ematician, what is your interest? Integrals?
Differential equations? Or what is it?” He
said this and that. “Wonderful! You are just
the right man!” I gave him the following
problem so I go to the board and say I have
no idea of how to go about that stuff. He said,
“Let me think about that!” So three days later
he came back. “I think I have a way of going
about that.” – “I thought you go to Chicago?”
– “No, no, I have no time I was working on
that problem.”
He was cured.
HVF: He had no time to see the doctor. My
friend the mathematician, “What did you do
to that man?” I said, “Nothing! I told him a
problem which I could not solve.” So this
immediately put him on the track. And I
really don’t know the mathematics; this guy
knows much more than I. But I know the
problem, I know who might be able to go
about.
4. Gregory Bateson (1904–1980) was a high-
ly interdisciplinarily working anthropolo-
gist, linguist, social scientist, and
psychiatrist.
5. Margaret Mead (1901–1978) was an an-
thropologist focusing on psychologically
oriented field work as well as building in-
terdisciplinary links between anthropolo-
gy and other fields.
6. “Cybernetics. Circular causal, and feed-
back mechanisms in biological and social
systems.” Transactions of the Sixth Con-
ference 1949, edited by Heinz von Foer-
ster. Republished together with the
following four transactions by Claus Pias
(2003).
7. Mexican physiologist Arturo S. Rosen-
blueth (1900–1970) researched the chem-
ical mechanism of nervous impulses
transmission at the Departments of Phys-
iology and Pharmacology at the Instituto
Nacional de Cardiología in Mexico City.
second order cybernetics
“What did you do to
that man?” I said,
“Nothing! I told him a
problem which I could
not solve”
That’s very nice, so you really had a very
good atmosphere there?
HVF: I had a wonderful atmosphere! The
people they stayed friends for the rest of their
lives, I’m sure.
But I know, because we have here also visit-
ing a composer from Illinois University and
he didn’t know you because he’s younger but
he says that your name is still well-known.
HVF: Yes, I wrote a book about music by com-
puters which was partly doing with people
from the music lab at the university.
We have to go to Pirsig18 now.
HVF: We have to go to Pirsig, let’s go! Okay,
let’s take some motorcycle and go to Pirsig!
Wonderful!
8. The BCL was founded at the University of
Illinois at Urbana-Champaign. See Müller
(2000) for a detailed article on the history
of the BCL.
9. Heinrich Klüver (1897–1979) made im-
portant contributions to visual psycholo-
gy, psychopharmacology, and animal
behavior research. Working as neurologist
at the University of Chicago, he intro-
duced Gestalt psychology to the US and
helped to formulate the discipline that is
today known as neuroscience.
10.Karl Menger (1902–1985), mathematician
and member of the Vienna Circle, known
for his work on various topics such as
curve and dimension theory, algebras,
probabilistic metric, and algebra of geom-
etries. In the 1930s he worked with intu-
itionist mathematician L.E.J. Brouwer. In
1936 he emigrated to the US to work from
1946 on at the Illinois Institute of Tech-
nology in Chicago.
17
 historical
OPINION
11.Economist Carl Menger von Wolfensgrün
(1840–1921) was the founder of the Aus-
trian school of economics.
12.Mathematician Leopold Kronecker
(1823–1891) is known for his statement
“God created the integers, all else is the
work of man” expressing his idea on a fi-
nite mathematics which should forgo
non-constructive existence proofs.
13.David Hilbert (1862–1943) favoured for-
malist mathematics according to which
mathematics can be defined using a finite
and consistent set of axioms. This was
proved impossible by Kurt Gödel’s In-
completeness Theorem in 1931.
14.W. Ross Ashby (1903–1972), was one of
the founders of cybernetics and general
systems theory, focusing on self-organis-
ing systems, information theory, and ma-
chine learning. From 1959 to 1970 he was
professor at the Biological Computer Lab-
oratory.
18
15.Humberto Maturana (1928–) is neurobi-
ologist at the University of Chile who de-
veloped the concept of autopoiesis.
16.Francisco Varela (1946–2002), a former
student of Maturana. Later he pursued his
own vision to develop a calculus of self-
reference.
17.Gordon Pask (1928–1996) developed a
Conversation Theory which he applied to
education. It grew out of his cybernetic
understanding of human-machine inter-
action as a form of conversation and dy-
namic process in which the participants
learn about each other.
18.Referring to Pirsig’s talk at the conferences
that started about when this interview end-
ed. Robert Pirsig (1928–) became famous
with his 1974 book called “Zen and the Art
of Motorcycle Maintenance: An Inquiry
into Value”, which introduced the “Meta-
physics of Quality” to account for the link
between quality, morality and reality.
second order cybernetics
References
Müller, A. (2000) Eine kurze Geschichte des
BCL. Österreichische Zeitschrift für
Geschichtswissenschaften 11: 9–30. Also:
http://www.univie.ac.at/constructivism/
papers/mueller/mueller00-bcl.html
Pias, C. (ed.) (2003) Cybernetics | Kybernetik.
The Macy-Conferences 1946–1953. Vol-
ume 1 Transactions/Protokolle. Dia-
phanes: Zürich, Berlin.
Further links
The entire interview can be downloaded as
audio file from http://www.univie.ac.at/con-
structivism/journal/1.1/interview.mp3
Paper received: 20 September 2005
Paper accepted: 20 September 2005
Constructivist Foundations
 engineering-computer scientific enactive cognitive science
SURVEY

Enactive Cognitive Science.

Part 1: Background and Research Themes
Kevin McGee A Linköping University, kevmc@ida.liu.se

structivist thought, and as such, it is not

Purpose: This paper is a brief introduction to enactive cognitive science: a description of intended as a detailed introduction to (or
some of the main research concerns; some examples of how such concerns have been defense of) constructivism. (For an excep-
realized in actual research; some of its research methods and proposed explanatory mech- tionally lucid introduction to constructivism
anisms and models; some of the potential as both a theoretical and applied science; and and radical constructivism, see Glasersfeld
several of the major open research questions. Findings: Enactive cognitive science is an 1995.) A related point is that this paper is not
approach to the study of mind that seeks to explain how the structures and mechanisms an attempt to argue whether (or how) enac-
of autonomous cognitive systems can arise and participate in the generation and mainte- tive cognitive science differs from radical con-
nance of viable perceiver-dependent worlds – rather than more conventional cognitivist structivism. Rather it is a brief survey that
efforts, such as the attempt to explain cognition in terms of the “recovery” of (pre-given, attempts to answer the question, what does
timeless) features of The (objectively-existing and accessible) World.As such, enactive cog- cognitive science look like when conducted
nitive science is resonant with radical constructivism. Research implications:As with within a broadly radical constructivist orien-
other scientific efforts conducted within a constructivist orientation, enactive cognitive tation. For the purposes of this paper, the dif-
science is broadly “conventional” in its scientific methodology.That is, there is a strong ference between enactive cognitive science
emphasis on testable hypotheses, empirical observation, supportable mechanisms and and radical constructivism is that one can be
models, rigorous experimental methods, acceptable criteria of validation, and the like. a radical constructivist – even a philosopher
Nonetheless, this approach to cognitive science does also raise a number of specific ques- of mind – without being a cognitive scientist.
tions about the scope of amenable phenomena (e.g., meaning, consciousness, etc.) – and And, indeed, most of the literature on radical
it also raises questions of whether such a perspective requires an expansion of what is constructivism would not be considered a
typically considered within the purview of scientific method (e.g., the role of the observer/ direct contribution to cognitive science; this
scientist). Key words: cognition, radical constructivism. paper is intended to present a brief overview
of some work that is. This, in turn, raises
another issue: the emphasis here is on science,
Introduction of vocabulary (readers familiar with the stan- as commonly understood. Thus, although it is
dard constructivist caveats can safely skip this possible to have some distinctly relativistic,
The purpose of this paper is to introduce section). The paper proper begins in the next post-modern responses to constructivist
enactive cognitive science and to provide a section by presenting several examples of tenets – e.g., “science is a story that is one of
brief survey of some concerns and research results from cognitive research in perception; many possible stories, each with their own
work that could be said to fall within this these are intended to contextualize for the merits” – this paper does not enter into dis-
emerging discipline. reader why some researchers feel dissatisfied cussions or debates about such things as the
The paper is in two parts. The first part, with the dominant objectivist, cognitivist relative merits of scientific versus non-scien-
published here, provides some background paradigm, and why they are motivated to seek tific stories for explaining cognition.
and an overview of some major research or develop alternative paradigms such as For the purposes of this paper, a number
themes of enactive cognitive science. The sec- enactive cognitive science. This part then con- of terms are used in particular ways and an
ond part, to be published in a later issue, will cludes by highlighting several of the disci- attempt is made to situate various work
look more closely at some of enactive cogni- pline’s major research themes. within some meaningful categories. Although
tive science’s research methods and proposed In such a short paper, the topics and exam- such choices, and discussions about them, are
explanations, mechanisms, and models of ples are obviously meant to be illustrative an important part of the research process, it is
enactive cognitive science; more recent rather than fully detailed or comprehensive; not the purpose of this paper to enter into a
research; applications; and open questions the various references and bibliography debate about whether we should call a collec-
and areas of future research. should provide good indications of where to tive research effort enactive cognitive science or
The plan of this part is as follows. The look for more details. cognitive science in a radical constructivist vein
remainder of this section presents some brief Before starting, it is important to highlight or genetic epistemology – nor whether we
preliminary comments about constructiv- a few things that this paper is not. It assumes should classify a particular researcher as
ism, radical constructivism, and related issues a certain basic degree of familiarity with con- belonging to one approach or the other. To

Constructivist Foundations 2005, vol. 1, no. 1 19

 engineering-computer scientific
SURVEY
the extent possible, an attempt has been made
to situate and describe researchers and their
work in ways that they themselves would rec-
ognize and accept. Of course, this is not
always possible; often it is the most interesting
and complex work that winds up in the
“boundaries” of our classifications – and we
wind up with such discussions as whether
Piaget is a realist or a radical constructivist,
what differences (if any) there are between
theoretical cognitive science and philosophy
of mind, and so on. The definitions and cate-
gories proposed here, then, are fluid and
pragmatic; the greatest hope for this paper is
that it helps facilitate substantive discussion,
cooperative research, and the sharing of
related research among those who are practic-
ing cognitive science with regard to the con-
cerns as they are broadly sketched here.
Having said that, there are some impor-
tant distinctions that will appear in this paper;
in particular, distinctions between the termi-
nology, the results, and the purpose of different
research that seems in one way or another
related to enactive cognitive science.
Right from the start, there is a source of
potential confusion related to the term enac-
tive as it is used within the constructivist tra-
dition. The term was initially popularized by
Jerome Bruner, a constructivist cognitive sci-
entist, who introduced to identify a particular
way to “translate experience into a model of
the world” (Bruner 1966, p. 10) – that is, there
are certain “mental models” (such as, know-
ing how to ride a bicycle) that seem to be the
result of action rather than, say, logical analy-
sis. The phrase enactive cognitive science, on
the other hand, was coined by Francisco
Varela, Evan Thompson, and Eleanor Rosch
(1991), who clearly wished to emphasize the
radical constructivist nature of actively
“bringing forth” and sustaining a viable world
that is inextricably related to the structure of
the knower. (We will have more to say below
about subsequent researchers who have con-
centrated on a single aspect of the enactive
agenda, with, for example, proposals for
active cognitive mechanisms that do not try to
address the issue of enacting a phenomenal
world.)
This example highlights the fact that the
central terms used in this paper are used in
many ways by many researchers, constructiv-
ist and otherwise. So, just as the use of the
term uncertainty principle does not guarantee
20
that the use is consistent with its typical usage
and meaning in quantum mechanics
(acknowledgments to Heisenberg notwith-
standing), even those who speak of “enactive
cognition” or “enactive mind” and explicitly
cite Varela et. al. as their predecessors, do not
always have the same radical orientation as
they do. Ami Klin and colleagues (Klin et al.
2003), for example, acknowledge them as the
source of inspiration for their own enactive
mind approach to the study of autism; how-
ever, their summary interpretation of Varela,
Thompson, and Rosch’s enactive model is
that “agents may vary in what they are seeking
in the environment, resulting in highly dis-
parate “mental representations” of the world
they are interacting with” (Klin et al. 2003,
p. 349). This, however, is something closer to
a subjectivist model of how individual inter-
pretations differ from the “intended message”
of the sender. This is not meant as a critique
of the work of Klin, merely to highlight that
such a model is both philosophically very dif-
ferent from more radical perspectives of such
things as learning (Glasersfeld 1995, Varela
1999a) – and scientifically a different model of
social cognition than, for example, the radical
one put forward in the theory of autopoiesis
(Maturana & Varela 1987).
Thus, as with all disciplines, there are
issues of terminology – where similar technical
terms or phrases are used in varying ways. In
this paper we will also encounter another kind
of distinction: scientific work that produces
descriptions, experimental results, or pro-
posed mechanisms that are useful for enactive
cognitive science – even when that work is not
explicitly conducted as part of an overall
attempt to do cognitive science within a radi-
cal constructivist framework. In other words,
there is a great deal of interesting research in
“conventional” cognitive science that can be
useful for explaining certain cognitive phe-
nomena in enactive terms – even though this
may not the intention of the researchers
themselves (and even in cases where they
explicitly reject the radical constructivist ori-
entation, e.g., Brooks 1991, Lakoff & Johnson
1999).
Throughout the paper, we will try to indi-
cate terms, such as embodiment and situated
action, that are used in different ways by dif-
ferent researchers in other research traditions.
Nonetheless, readers may find it most helpful
to keep in mind some of the basic tenets of
enactive cognitive science
radical constructivism while reading about
the research themes. Some brief background
about radical constructivism now follows.
From objectivism to
radical constructivism
In order to situate enactive cognitive science,
it will help to sketch some of the contributing
philosophical models: objectivism, subjectiv-
ism, realist constructivism, and radical con-
structivism. Briefly, objectivism takes as
largely unproblematic the belief that human
beings are objectively real cognitive agents
that exist as separate entities in an objective,
independently-existing real world with real
properties. Cognition, by this view, consists of
perceiving (or “recovering”) information
about that world, processing it, and acting
upon it. And cognition can be (more or less)
true or false to the extent that human percep-
tion and models, opinions, and beliefs corre-
spond to this world. In the transition from
philosophy to cognitive science, this model
has been largely couched in terms of percep-
tion and processing (“inner”) and features
(“outer”). Similarly, the issues of cognitive
change have been largely formulated in terms
of genetic and environmental determinism
(“nature versus nurture”) – and the degree to
which cognitive mechanisms are innately
given at birth, the result of maturation, trig-
gered by environmental phenomena, or oth-
erwise “emerge out of experience.”
One of the major alternatives to the objec-
tivist perspective is subjectivism, which leaves
the inside/outside dichotomy largely
untouched and simply reverses the emphasis
by arguing that “The World” is a “projection”
of mind. Objectivist concerns (and in certain
ways, subjectivist concerns as well) inform
much of conventional cognitive science – or
cognitivism – and cut across such distinctions
as preferences for employing “symbolic” or
“connectionist” computational models.
Another alternative to objectivism, con-
structivism, challenges a number of assump-
tions shared by both objectivism and subjec-
tivism – in particular, the notion that minds
and worlds are pre-given and that cognition
consists entirely of recovery (on the part of the
individual) or responding (to “transmissions”
from the environment). One of the key figures
in the constructivist tradition is Kant who
Constructivist Foundations
 engineering-computer scientific
SURVEY
argued that knowledge is not simply “given by
experience” (of an objective universe) nor the
result of abstract reasoning (about such a uni-
verse); rather, there are intermediate mental
structures, which he called schemas. One con-
sequence of Kant’s “Copernican Revolution”
in epistemology is that emphasis shifts from
debates about cognition as recovery of The
World to the issue of structures (schemas)
that inform the world as known and experi-
enced. For Kant, there is still some influence
of The World (so this is not simply solipsism),
but the world we live and know “bears the
mark of our structure.” Kant’s schemas were
simply theoretical structures; it was Jean
Piaget and others in the early twentieth cen-
tury who contributed to research on possible
mechanisms of active cognitive construction.
One way to describe Piaget’s work is that it
took Kant’s model seriously – and more
importantly, it tried to address the question of
how such an epistemological model can give
rise to our experience of a viable, reliable, and
solid world. Piaget’s proposal, in a nutshell, is
that “intelligence organizes the world by orga-
nizing itself ” (Piaget 1955, p. 311). Piaget
argued that although the mechanisms to cre-
ate individual schemas may be innate, most of
the actual schemas are not, that their con-
struction is not simply a matter of genetic
unfolding, and that the ones that are con-
structed are not directly determined by the
environment (and may not even correspond
directly to aspects of some independent envi-
ronment). (For a nice constructivist account
that sketches some connections from Kant to
related twentieth century proposals for cogni-
tive mechanisms, see Papert 1988.)
Within the epistemologically-oriented
constructivist tradition it is useful to distin-
guish between realist and radical constructiv-
ism.
The former is largely objectivist in its belief
in an external, objective, knowable world. By
this view, constructive mechanisms tend to be
“in the head” of the cognitive agent – and cog-
nitive construction is one way a cognitive
agent comes to have such things as interpreta-
tions, opinions, beliefs, and models of that
objectively existing world. By this view,
knowledge is still more or less a “mirror of
nature” (Rorty 1979), but some of that
knowledge is the result of active construction
by the cognitive agent. This realist orientation
tends to cut across such sub-tradition divi-
2005, vol. 1, no. 1
sions as “individual-based” constructivism
(often attributed to Piaget (Gruber &
Voneche 1977) or “social constructivism”
(attributed to Vygotsky 1978). It can even be
seen in the work of some less overtly objectiv-
ist constructivists, who say they “do not care”
about the ontological status of reality, but
rather about “reality as it is for the person.”
This realist version of constructivism has
informed work in a variety of domains, from
family therapy to pedagogy to the develop-
ment of pedagogical materials; it is also reso-
nant with a great deal of research in the scien-
tific study of mind.
“Radical constructivism, thus, is radical
because it breaks with convention and devel-
ops a theory of knowledge in which knowl-
edge does not reflect an “objective” ontologi-
cal reality, but exclusively an ordering and
organization of a world constituted by our
experience.” (Glasersfeld 1984, p. 24)
The latter, radical, version of constructiv-
ism holds that it is not just (some) “knowl-
edge of The World” that is constructed by the
cognizing agent, but rather, the agent’s knowl-
edge and its phenomenal world (i.e., the world
as it is for the cognitive agent) are both the
result of active construction. Ernst von Gla-
sersfeld coined the term radical constructivism
to distinguish it from more objectivist variet-
ies of constructivism. He has also nicely sum-
marized some of its major concerns (Glasers-
feld 1995, p. 51).
[ Knowledge is not passively received either
through the senses or by way of communi-
cation;
[ Knowledge is actively built up by the cog-
nizing subject.
[ The function of cognition is adaptive, in
the biological sense of the term, tending
towards fit or viability;
[ Cognition serves the subject’s organiza-
tion of the experiential world, not the dis-
covery of an objective ontological reality.
Radical constructivism as a philosophy of
mind in the western tradition can be traced
back at least 2500 years. As a particular orien-
tation, it has informed work in various
domains: pedagogy and therapy being among
the two most prominent. In recent years, it
has informed the scientific study of cognition
– as well as the development of materials
based on the insights of such study. It is pos-
sible, for example, to study how the world of
a cognizing agent relates to the agent’s struc-
enactive cognitive science
ture; similarly, it is possible to identify, pro-
pose, and model possible mechanisms of such
active construction, whether they are “in the
individual” or “in a system that includes (or
enacts) the individual.”
We now turn to look at some of the major
research themes of enactive cognitive science
that can be seen as broadly resonant with the
major concerns of radical constructivism.
Enactive cognitive
science
“…the enactive approach consists of two points:
(1) perception consists in perceptually guided
action and (2) cognitive structures emerge from
the recurrent sensorimotor patterns that enable
action to be perceptually guided. The overall con-
cern …is not to determine how some perceiver-
independent world is to be recovered; it is, rather,
to determine the common principles or lawful
linkages between sensory and motor systems
that explain how action can be perceptually
guided in a perceiver-dependent world.”
– Varela, Thompson & Rosch 1991, p. 173
During the 20th century, Jean Piaget, Lev
Vygotsky, Jakob von Uexküll and Maurice
Merleau-Ponty initiated pioneering research
to propose possible constructivist methods
for studying cognition – and to empirically
support possible constructivist models and
mechanisms for explaining cognition. Rele-
vant research also occurred in both the nat-
ural and artificial sciences: from biology,
neurophysiology, psychology, linguistics,
semiotics, sociology, and anthropology to
structuralism, cybernetics, systems theory,
artificial intelligence, robotics, and the more
recent “sciences of complexity” (e.g., artifi-
cial life). (For a popular account of this his-
tory that is sympathetic to the concerns of
enactive cognitive science, see Capra 1996;
for a treatment that concentrates on biology
and some implications for contemporary
work in robotics, see Sharkey & Ziemke
1998.) And, of course, during the same
period there have been more philosophi-
cally-oriented discussions of related topics,
e.g., James (1907), Dewey (1916), White-
head (1979), Heidegger (1962), Wittgenstein
(1963), Mead (1934), Gadamer (1976),
Goodman (1978), Rorty (1979), Glasersfeld
(1995).
21
 engineering-computer scientific
SURVEY
In order to contextualize some of the
research themes and mechanisms proposed
by enactive cognitive science, we will begin by
looking at a few examples of phenomena or
results that motivate this approach. Readers
familiar with the philosophical and epistemo-
logical history of radical constructivism may
feel that the issue is already well-motivated
from reason; nonetheless, in the turn to sci-
ence, it can also be helpful to find similar
motivations from cognitive research. Since
much of cognitivism is premised on certain
assumptions about interactions between
“inner mind” and “outer world,” we will look
specifically at some research in the area of per-
ception.
Note that it would have been possible to
describe motivating examples that highlight
problems with conventional objectivist
accounts of, say, social cognition. However, the
debate between proponents of “individual”
and “social” constructivism is among the
most controversial in the entire constructivist
tradition and there is a risk of devoting too
much attention to this one issue here. We will
return to the topic of social cognition in a sec-
tion below.
As we consider these and other examples,
it is important to keep in mind that even as we
discuss, say, biology, the results and subse-
quent hypotheses need not be taken as asser-
tions about some objective, material reality.
Rather, the relevance of such examples to
enactive cognitive science is that the experi-
mental results challenge certain notions
about the existence of an objective, timeless
reality; the proposed cognitive mechanisms
do not rely on a correspondence between an
agent’s cognitive model and an indepen-
dently-existing objective world; and some of
the proposed paradigms for interpretation
bypass certain traditional, objectivist
assumptions.
That said, we now turn to four brief exam-
ples of cognitive phenomena related to per-
ception: research on color-vision in pigeons,
on the workings of the olfactory system in
rabbits, on categories and concepts in
humans, and on the structure of the human
visual system.
Color vision
Humberto Maturana is a biologist by training
who participated in one of the most famous
pieces of biological research of the 20th cen-
22
tury, the study of frog vision (Lettvin et al.
1959). As a result of that work and his subse-
quent work on color-perception in pigeons
(Maturana, Uribe & Frenk 1968) he came to
adopt a radically different perspective on cog-
nition. His description of this transition in his
thinking is broadly similar to stories that
other researchers mentioned here have told
about their own work – their initial, conven-
tional convictions, their struggle to reconcile
research results with existing paradigms for
interpretation of those results, and their ulti-
mate conviction that an alternative paradigm
was required – so it is worth quoting him at
length.
“…[when we did our research on frog
vision] we did it with the implicit assumption
that we were handling a clearly defined cogni-
tive situation: there was an objective (abso-
lute) reality, external to the animal, and inde-
pendent of it (not determined by it), which it
could perceive (cognize), and the animal
could use the information obtained in its per-
ception to compute a behavior adequate to
the perceived situation.
“…[when we began to study color vision
in pigeons] it soon became apparent to us that
that approach leads to deep trouble. …There
are many visual configurations, with uniform
and variegated spectral compositions, in sim-
ple and complex geometrical forms, that give
rise to indistinguishable color experiences.
How should one, then, look for the invari-
ances in the activity of the nervous system, if
any, in relation to the perception of color?
“…What if, instead of attempting to corre-
late the activity in the retina with the physical
stimuli external to the organism, we did oth-
erwise, and tried to correlate the activity in
the retina with the color experience of the
subject? …the new approach required us to
treat seriously the activity of the nervous sys-
tem as determined by the nervous system
itself, and not by the external world …still
more fundamental was the discovery that one
had to close off the nervous system to account
for its operation, and that perception should
not be viewed as the grasping of an external
reality, but rather as the specification of
one…” (Maturana 1980b, pp. xiv–xv)
The main point here, and one we will see
repeated in many of the examples that follow,
is that even with rigorous, conventional
research methods and repeated experiment,
the researchers were unable to establish con-
enactive cognitive science
sistent correlations between percepts and sen-
sory-responses. They found that there could
be different responses to the same percepts –
and the same sensory responses to different
percepts. Being good scientists, they did
doubt their own results; they tried in various
ways to control their experiments and to
ensure there were not other complicating fac-
tors involved (something, again, typical for
many of the other researchers discussed here).
In the end, they felt compelled to entertain a
more radical reorientation.
As a result of this work, Maturana
(together with Francisco Varela and others)
went on to develop an autopoietic model of
biological functioning and cognition. By this
view, any physical structure with an autopoi-
etic – that is, self-producing and self-main-
taining – organization is both living and cog-
nitive. We will not defend this definition here
(although, for more detail, see Maturana &
Varela 1980, 1987, and for a critique, see
Boden 2000), but merely note that by this def-
inition, a biological cell is the smallest known
“unit” of life (and cognition). And, for the
remainder of this paper will use this defini-
tion as a way to loosely circumscribe the scope
of enactive cognitive science; that is, we will
consider cognitive phenomena from the cel-
lular to the social.
Olfactory system
The quotation from Maturana above does not
go into much detail about the particulars of
the difficulties of correlating sensory-infor-
mation with cognitive processing, so some
readers may not believe that there were
enough challenges to the objectivist model to
warrant a major shift in orientation. The fol-
lowing example serves two purposes. First, to
give a more textured description of the exper-
imental challenges for this type of research –
and some of the difficulties that arise in
assuming perception is some reasonably
straightforward form of veridical recovery
and processing of an objective and indepen-
dently-existing world. Second, to describe
some results of such research that suggest
actual problems for the objectivist assump-
tion, even after one has solved many of the
formidable research problems.
Walter Freeman is a neuro-biologist who
has been working for many years to study the
olfactory system of rabbits. In order to appre-
ciate his work, consider that conventional
Constructivist Foundations
 engineering-computer scientific
SURVEY
accounts, both popular and professional,
tend to present visual perception (and per-
ception in general) roughly as follows. Some
phenomenon in the world impinges (in the
form of a signal) on a sensory-surface con-
sisting of receptors. This signal is trans-
formed by the receptors into individual neu-
ral signals that are then transmitted (with
possible modulations) through a series of
relays to some physical brain module for fur-
ther processing. At which point, perhaps, the
result of processing is then relayed sequen-
tially to some motor-neurons. Although the
majority of work on perception is devoted to
vision, for our purposes there are some inter-
esting aspects about the olfactory system. In
particular, in mammals there is a single
major intermediary mechanism (the olfac-
tory bulb) between receptor-neurons for
olfaction and the cerebral cortex. One might
expect then, if there was good neurological
evidence that brain mirrors world, one would
find it here.
In many, many experiments over the years,
Freeman and his colleagues have attached
electronic sensors to the olfactory system of
rabbits and conducted extremely well-con-
trolled tests to measure patterns of neural
activity. These experiments have attempted to
determine such things as regularities in neu-
ral responses to different odorants, regulari-
ties in the recognition of odorants, and the
effects of learning to recognize new odorants
on the neural responses to known odorants.
Before we look at their results and conclu-
sions, assume that perception works roughly
as described above and consider some of the
conceptual and research challenges. The sen-
sory-surface consists of different types of
receptors (corresponding to different types
of molecules) which are randomly distrib-
uted on the nasal surface in non-uniform
concentrations and ratios (which differ from
individual to individual). A “single dose” of
the “same” odorant consists of many differ-
ent kinds of molecules that impinge upon the
sensory-surface in different concentrations,
at different time intervals, with different
velocities, upon different patterns of sensory-
neurons. These receptor-neurons then fire
(at different times, depending on when they
are triggered) in parallel to stimulate neurons
in the olfactory bulb. At the bulb (and
beyond) all the neurons are stimulated by
many other neurons, firing from many differ-
2005, vol. 1, no. 1
enactive cognitive science
ent “directions” – and they
each also stimulate many dif-
ferent neurons in many dif-
ferent “directions.”
As noted, the naive model
of perception tends to be one
of “one-way telegraph trans-
mission,” so it can be difficult
to visualize the complex and
active process described
above. However, having an
intuitive sense of this process
will help the reader under-
stand many other concepts
discussed in this paper, so we
will try to express all of this in
a vastly over-simplified sce-
nario in the style of Rube
Goldberg. Imagine that there
is a large, flat tent-canvas, sus-
pended parallel to the
ground; and, higher above the
canvas, there is a large, flat
Figure 1: Olfactory system (simplified from Freeman
mesh. We have, then, two lay-
2001, p. 87, with kind permission of the author)
ers of “sheets.” There are
“cups” of different shapes and
sizes attached upside-down to
the bottom of the canvas surface; the different the tubes extending up transmit little ping-
kinds of cups are randomly distributed on the pong balls each time a balloon lands in a cup.
surface, and even the ratios of different kinds The ping-pong balls fly up the tubes and hit
of cups are random. Each type of cup corre- the pouches. For every only one of the incom-
sponds to a particular type of balloon; if a bal- ing tubes coming from the canvas below, each
loon of the right size floats into a cup, the bal- pouch has on the order of ten other incoming
loon attaches briefly inside it. An odorant, in tubes attached to it; the other incoming tubes
this analogy, is a collection of balloons of dif- come from other pouches, some of them in
ferent sizes and densities. Two different doses the mesh (“parallel” to it) and some from fur-
of the “same” odorant can have very different ther up (“above”). Thus, each pouch gets hit
quantities of balloons – and different ratios of by “incoming” ping-pong balls from below,
particular kinds of balloons. Furthermore, from its neighbors in the mesh, and from a
two different noses can have different num- wide fan of pouches further up. Similarly,
bers of cups, with different proportions of dif- each pouch actually has several thousand out-
ferent kinds of cups in different positions going tubes; these fire ping-pong balls to
along the surface of the tent. The act of inhal- neighboring pouches in the mesh – and to a
ing, or bringing the balloons in contact with wide fan of different pouches further up. If
the surface cups, results in all these balloons enough balls hit a pouch at the same time,
bouncing into one another, hitting different then it lights up and sends a ping-pong ball of
cups at different times with different veloci- its own to every other pouch that is at the end
ties, and even causes the surface of the tent to of its outgoing tubes.
flap and move. The key thing to visualize is that the mesh-
On the top side of the canvas, there are pouches are active most of the time, whether
tubes, one extending up from the location of there are balls coming up from the tent sur-
each cup. Each of these tubes continues up to face or not. Furthermore, even when there are
make contact with a pouch, suspended in the balls coming up, most of the balls that hit a
mesh sheet. This mesh, extending widely over pouch at any given moment are from other
the tent canvas, has many such pouches. And pouches. And, finally, note that, unlike most
23
 engineering-computer scientific
SURVEY
artificial models of neural networks, the firing
of the ping-pong balls is not synchronized;
balls are being fired as a vast, irregular pande-
monium of activity.
This brief summary glosses over a great
number of additional complicating details
about the neurology – and it also does not
mention all of the many additional experi-
mental and conceptual difficulties; nonethe-
less, even without entertaining radical inter-
pretations of reality, it should still be clear
that the process of instantly “recognizing a
smell” is vastly more complex than any sim-
ple notion of recovering and accurately pro-
cessing an obvious and distinct signal.
In any case, to study how this system gives
rise to “recognizing an odor,” Freeman and
his colleagues carefully attached an array of
electronic sensors to the nasal bulb of living
rabbits to measure the neural responses of
neurons there. They wanted to get readings of
what happens to an “odorant signal” after the
neurons at the sensory-surface have been
activated. After the sensors were attached,
they conducted numerous careful experi-
ments in which they compared patterns of
activity in bulb-neurons by using well-con-
trolled samples of odorants. (Careful readers
will already notice that this description still
glosses over many issues about actual cogni-
tion, such as the fact that olfaction “in the
wild” does not have the benefit of odorants
distilled to some acceptable degree of purity
or consistency for purposes of comparison.)
In terms of the Rube Goldberg scenario, Free-
man and his colleagues were basically study-
ing the activity of the pouches in the mesh;
and the different overall patterns as different
pouches light up in different spatial and tem-
poral patterns. They were trying to deter-
mine whether and how the patterns correlate
(in some way) with (some reasonable
abstraction of) an odorant – and how/
whether any of this correlates with the rab-
bit’s experience of “recognizing that smell.” As
a result of this work, Freeman and his col-
leagues have managed to get reasonable read-
ings and develop a mathematical model. The
technical details are beyond the scope of this
paper, but they include dynamic topographi-
cal maps of chaotic (in the technical sense)
activity of olfactory neurons. This chaotic
activity is both necessary for healthy brain
functioning – and, more surprisingly, it is not
“noise that distorts the signal” of perception,
24
but rather participates crucially in the gener-
ation of perception. Freeman and Skarda
provide the best summary of their results and
their interpretation of them.
“For more than 10 years we tried to say that
each spatial pattern was like a snapshot, that
each burst served to represent the odorant
with which we correlated it, and that the pat-
tern was like a search image that served to
symbolize the presence or absence of the
odorant that the system was looking for. But
such interpretations were misleading. They
encouraged us to view neural activity as a
function of the features and causal impact of
stimuli on the organism and to look for a
reflection of the environment within by cor-
relating features of the stimuli with neural
activity. This was a mistake. After years of sift-
ing through our data, we identified the prob-
lem: it was the concept of representation.
“Our research has now revealed the flaws
in such interpretations of brain function.
Neural activity patterns in the olfactory bulb
cannot be equated with internal representa-
tions of particular odorants to the brain for
several reasons. First, simply presenting an
odorant to the system does not lead to any
odor-specific activity patterns being formed.
Only in motivated animals, that is, only when
the odorant is reinforced leading to a behav-
ioral change, do these stereotypical patterns
of neural activity take shape. Second, odor-
specific activity patterns are dependent on the
behavioral response; when we change the
reinforcement contingency of a [conditioned
response] we change the patterned activity.
Third, patterned neural activity is context
dependent: the introduction of a new rein-
forced odorant to the animal’s repertoire
leads to changes in the patterns associated
with all previously learned odorants. Taken
together these facts teach us that we who have
looked at activity patterns as internal repre-
sentations of events have misinterpreted the
data. Our findings indicate that patterned
neural activity correlates best with reliable
forms of interaction in a context that is behav-
iorally and environmentally co-defined…”
(Freeman & Skarda 1990, p. 376)
We have spent some time on the details of
this for a couple of reasons. First, it provides
more textured insight into the kinds of exper-
imental challenges Freeman and other
researchers have faced; the point is that they
did not shift orientation because it was “easy.”
enactive cognitive science
Similarly, the myriad variations of different
experimental parameters – and research
results that are difficult to reconcile with var-
ious objectivist interpretations – are fairly
typical for the kinds of research that have
motivated many enactive cognitive scientists
to consider radical alternatives to traditional
paradigms for interpreting their research.
This was the case for Maturana, and it was the
case, as we will now see, for Eleanor Rosch.
Categories, language, and concepts
The traditional theory of concepts and cate-
gories has manifested itself in a number of
ways in the history of western philosophy and
science. One manifestation is the assumption
that there are “natural kinds” – that is, that the
world (nature) is objectively categorized. The
related cognitive assumption is that concepts
mirror those natural categories. Freeman’s
work on the olfactory system is a particular
example of classification and categorization:
what, operationally, does it mean to recognize
(or classify) an odorant – or for one odorant
to be “the same” as another? Or, consider the
problem of categories from the various west-
ern philosophical perspectives sketched ear-
lier. Are colors categories “in the world”
(light, object-reflectance) or concepts “in the
head” (receptors, internal processing mecha-
nisms)? Are colors universal categories of the
world, universal (abstract) concepts of the
mind, or are they personally or culturally rel-
ative?
The cognitive equivalent of categories are
concepts, and as Eleanor Rosch so bluntly
puts it, “If ever there was a domain where
you’d think cognitivism could get it right, it
is concepts” (Rosch 1999, p. 61). Rosch is a
psychologist who, together with her col-
leagues, is widely credited with revolutioniz-
ing the study of concepts in cognitive science
through a pioneering, systematic, large-scale,
long-term, cross-cultural research effort
devoted to concepts, categorization, and lan-
guage. Through her research, she has
amassed enormous evidence to suggest that
cognitivism does not “get it right” when it
comes to concepts, and one of her interpreta-
tions of her results is that, given some of cog-
nitivism’s assumptions, it may not even be
able to get it right (for more detailed summa-
ries of this research and its implications, see,
e.g., Rosch 1999, Lakoff 1987, Rosch & Lloyd
1978).
Constructivist Foundations
 engineering-computer scientific
SURVEY
What are some of the cognitivist assump-
tions about concepts and categories – and
what does her research show?
Concepts as abstract universals . The classical,
abstract model of concepts and categories is
strongly influenced by formal logic. In
essence, it claims that categories have two
basic characteristics: categories have clear and
distinct boundaries – and “membership” in a
category is based on necessary and sufficient
“shared features.” In Rosch’s work, she has
shown significant exceptions to these require-
ments – and to virtually all of the major
entailments of such assumptions. Thus, she
has shown that the boundaries of categories
are not always well-defined, there are “degrees
of membership,” and that there are “best
examples” for categories (about which there is
wide consensus). To give a specific example,
consider the last point. In some of her work
she demonstrated that people have a notion of
a “best” example of a category, such as the
color red. But, as Rosch argued, if the classical
model of categories is correct, then any mem-
ber of a category (that is, any member that is
a member by virtue of having the same fea-
tures as another member) should be as good
a category-example as any other member of
the category.
A further entailment of classical theory is
that taxonomic systems of categories consist
of abstract, universal, hierarchical structures.
That is, there are universal concepts (such as
“object”), some are more general than others
(“object” is more general than “tree”), and
that these concepts existing in an unchanging,
hierarchical relationship to each other. Fur-
ther, these abstract categories are universal,
rather than specific to human experience.
Rosch showed that cognition relies crucially
on other taxonomic structures, that human
concepts are very dependent on human expe-
rience (a “table” is a concept that only makes
sense relative to the human activity of using
tables), and that although there is cross-cul-
tural agreement about what she, following
Berlin and Kay (1969, Brown 1958), calls
basic-level categories, these are closer to the
schemas of Kant than to pre-existing catego-
ries in the world or abstract universal con-
cepts in the mind.
Finally, one of the strong assumptions of
this abstract notion of concepts is that it
asserts a model of knowledge that is separate
2005, vol. 1, no. 1
from the particular “physical realization of
the knower” (e.g., embodiment). By this view,
human knowledge “inherits” the abstract,
rational, conceptual taxonomies – rather
than, say, those taxonomies being constructs
based on human experiences (such as having
bodies, taking physical actions in a world, and
the like). The abstract view even informs the
work of some constructivists, particularly
Piagetians who have studied the development
of various logical understandings in children.
Rosch’s research has demonstrated serious
flaws with each of these assumptions and con-
clusions. Indeed, her work strongly suggests
that concepts and categories must arise out of
particular experience (as Piagetians might
argue) – but that if children are asked to per-
form classification tests relative to basic-level
categories (rather than the more typical
abstract, formal categories of Piagetian exper-
iments), their performance and understand-
ing is much richer and more sophisticated
than indicated by many of the results on clas-
sification in the Piagetian literature.
One of the most important consequences
of this aspect of Rosch’s research is that it puts
into serious question any claim that humans
start with classical taxonomies upon which
are placed additional constraints. By this
view, infants would start with such abstract
taxonomies and additional mechanisms;
Rosch’s work makes a compelling case that
humans start with basic-level categorization,
and that the more abstract and general classi-
fication schemes come later, both temporally
and in terms of derivation.
Concepts as telative or unconstrained. If con-
cepts and categories are not abstract univer-
sals, could they then be “relative” (to personal
experience or cultural contexts)? There are,
for example, significant cultural differences in
the linguistic terms used for colors, the num-
ber of color-terms (some languages, for exam-
ple, designate all colors with just two terms),
and the like. Indeed, because of this, some
researchers have made various claims that
concepts must be relativistic. However,
Rosch’s work on both prototype effects and
basic-level effects raises serious questions
about all of these relativistic assumptions and
claims. For example, but she conducted
experiments in which she compared such
things as recall and identification of “best
examples” for different sets of invented color-
enactive cognitive science
terms. Her results showed broad similarity to
the recall and identifications made by people
in other cultures with other linguistic-terms
for colors. That is, each language had focal col-
ors that were broadly similar across cultures;
and members of different linguistic groups
identified the same focal colors, even when
they were using invented color-term vocabu-
laries.
Concepts as categories in the world . If color-
concepts are not abstract universals, nor rela-
tive to individual or cultural experience,
could they be derived from universal catego-
ries in the world? In fact, the vast cultural dif-
ference about color-terms and color-classifi-
cation models suggest that this hypothesis is
already problematic. To the extent that there
seems to be anything universal about color
concepts, it seems to be at the level of, say, cog-
nitive schema, rather than universal features
of an objective world.
Concepts in the mind . Finally, although con-
cepts may not be abstract universals of the
mind, couldn’t they be the result of individual
cognitive processes in some social, cultural, or
linguistic context? This is a variation on the
relativistic argument and one related cogni-
tive hypothesis is the Whorf-Sapir claim that
“language limits thought”; by this view, the
linguistic terms for colors would limit the for-
mation of color concepts. Rosch was able to
show, for example, that even in languages that
only have two words for different colors, the
native speakers were able to describe, indicate,
and remember a wider range of color concepts.
They were even able to learn and use new,
invented color-term vocabularies with more
terms in them. And, as noted above, the per-
formance of subjects working with invented
color-term vocabularies was broadly similar
to the performance of subjects from other
cultures using their native languages (or other
invented color-term vocabularies). In other
words, in certain essential ways, the limited
nature of certain native color-vocabularies
did did not prevent the identification of a
larger set of basic-color concepts, nor prevent
the learning of new color-vocabularies with
different structures, nor seem to correlate
with performance in remembering or identi-
fying basic-level terms in any of the vocabu-
laries. The main correlations in all cases
seemed to be at the basic-level.
25
 engineering-computer scientific
SURVEY
As a result of her research, she has come
more and more to the conclusion that catego-
ries are not pre-given in some objective world,
they are not pre-given universal cognitive
concepts, and they are not purely individual,
cultural, or relativistic. Rosch herself
expresses this quite eloquently,
“Because concepts and categorizations
play a vital role in bridging the mind-world
unit and in revealing situational contexts,
they may be able to provide a point of entry
for the study of situations. Instead of asking
how categories can be universal or how con-
cepts can represent an external world in an
internal mind, we could ask where categories
and category-systems come from in the first
place? … When mind and world are consid-
ered separate, causal or explanatory efficacy is
attributed either to the mind or the world.
Happenings in the world may be considered
stimuli to which the organism or person
responds, or the mind or person may be seen
as the source of desires, intentions, theories,
or actions on the things of the world. Polariza-
tion of these extremes leads to theories and
research programmes which go in and out of
fashion, but which do not progress. At best the
mind and world will be said to interact. The
new view requires rethinking how we want to
model causality and prediction altogether.
Mind and world occur together in a succes-
sion of situations which are somewhat lawful
and predictable. We want to be able to find
those laws and to find a level of description
which neither turns human actions into
something mechanical like engineering nor
something mental like fantasy.” (Rosch 1999,
pp. 74–75)
For Rosch then, as for Maturana and Free-
man, it is very difficult to account for the
research results in traditional objectivist, sub-
jectivist, or relativistic terms.
Visual perception
To conclude this brief survey of cognitive phe-
nomena that challenge cognitivist models of
cognition, we turn to research on the architec-
ture of the visual system.
We have already encountered problems
with the conventional model of perception-
as-recovery in the description of Freeman’s
work above. But there is a more startling chal-
lenge to the conventional transmission
model. Namely, that although most sensory-
neurons are not stimulated by other neurons
26
(that is, they respond only to surface stimula-
tion and relay electrical signals on to a differ-
ent “layer” of neurons) – the neurons at that
secondary layer receive something on the
order of 80% of their neural signals from other
neurons “within” the organism (Churchland &
Sejnowski 1992, Singer 1980). In other words,
neurons at this layer (and most “subsequent”
layers) do not act as “telegraph stations” in a
one-way relay from outside to inside, but
more like “polling stations” with contribu-
tions from all their neighbors, and the results
of such polls do not move in a single direction
along some path from outer to inner. This
does not even take into account that there are
many other kinds of neuronal interactions –
chemical releases, and the like – that may be
more cognitively relevant or significant. To
put it in the starkest possible terms, it seems
that most perception-based neural activity is
the result of processes that “spread outward”
rather than “proceed inward.”
Thus, even without entertaining radical
new paradigms of interpretation, conven-
tional neuroscience needs to develop some
plausible explanation of how this relates to
“perceiving things in the world.” It is impor-
tant to note that this observation is not con-
troversial; if anything, it is simply not
acknowledged in most conventional texts. To
the extent that it is acknowledged, the typical
response tends to consist of quickly mention-
ing it and then continuing with cognitivist
“proposals as usual.” The assumption seems
to be that someone else will solve this problem
in a way that is compatible with conventional
cognitivist models of perception (e.g., per-
haps by showing that different classes of neu-
rons have different significance in the trans-
mission and recovery of percepts).
As noted earlier, Maturana, Rosch, and
Freeman started with conventional assump-
tions, pursued their research along conven-
tional lines, and arrived at similar impasses
when they tried to reconcile their research
results with conventional, objectivist inter-
pretations. In each case, they concluded that
something was required other than tradi-
tional models of genetic/environmental
determinism and the long-held western alter-
natives of objectivism, subjectivism, relativ-
ism, and solipsism. In many cases, the partic-
ular research themes of enactive cognitive
science have arisen out of similar research
impasses.
enactive cognitive science
Research themes
We propose as a name enactive to emphasize
the growing conviction that cognition is not the
representation of a pregiven world by a pregiven
mind but is rather the enactment of a world and
a mind on the basis of a history of the variety of
actions that a being in the world performs
– Varela, Thompson & Rosch 1991, p. 9
The examples of research from the previous
section have a distinctly biological flavor, but,
as indicated earlier, the discipline is con-
cerned with a broad range of cognitive phe-
nomena. In their own work, Varela, Thomp-
son, and Rosch were particularly interested in
the development of methods for studying and
understanding first-person cognition (“con-
sciousness”), so that aspect has been further
developed in the years since then. Similarly,
there is now a more encompassing emphasis
on social aspects of cognition. Much of this
was implicit in their original proposal, but as
they themselves acknowledged, at the time
they were simply trying to sketch the broad
outlines of an emerging discipline. At this
point, we can say that a number of distinctive
research themes can be identified as central to
enactive cognitive science.
[ Umwelt, embodiment, and situated action
[ Autonomy, change, and creativity
[ Groundlessness, non-correspondence,
and viability
[ Emergence and self-organization
[ Consciousness and first-person cognition
[ Social cognition and intersubjectivity
[ Co-enaction
The remainder of this part of the paper –
borrowing freely from (Varela 1999b, Varela,
Thompson & Rosch 1991) – provides a more
detailed examination of these enactivist
research themes. In reading about the research
themes, it will also help to have some guiding
idea of co-enaction, a concept that is significant
in virtually all aspects of enactive research and
thought, but not always highlighted as such
(although, see Varela 1999b). This concept
emphasizes that the shift to an enactive per-
spective does not involve a simple change from
one absolute frame of reference to another
(e.g., internal to external, individual to social,
physical to linguistic, etc.). Rather, the shift is
to a perspective of co-specification and co-
determination. Susan Oyama (2000a, p. 180)
expresses this succinctly when she writes,
Constructivist Foundations
 engineering-computer scientific
SURVEY
“‘Natural’ biological persons are con-
structed, not only in the sense that they are
actively construed by themselves and others,
but also in the sense that they are, at every
moment, products of, and participants in,
ongoing developmental processes.”
This shift to co-enaction raises a number
of methodological issues, so after looking at
the other research themes, we will treat it last,
where it will serve also as a bridge to the sec-
ond part of the paper.
Umwelt, embodiment, and
situated action
This insistence on the codetermination or mutual
specification of organism and environment should
not be confused with the more commonplace
view that different perceiving organisms simply
have different perspectives on the world
– Varela, Thompson & Rosch 1991, p. 202
Objectivist models of cognition tend to start
from the assumption of an independent cog-
nizing agent recovering aspects of an inde-
pendently-existing world – and then using
some body-independent, universal rules to
reason about it. A central concept of enactive
cognitive science, on the other hand, is the
Umwelt, or the integration of agent and its
world that manifests itself in a lived-world.
This term was first used in this way by the
biologist Jakob von Uexküll, who was trying
to distinguish between the world (or niche) as
it is for the organism and some alternative,
objective world: “all that a subject perceives
becomes his perceptual world and all that he
does, his effector world. Perceptual and effec-
tor worlds together form a closed unit, the
Umwelt” (Uexküll 1957, p. 6). The main point
is that cognitive agents do not live in or expe-
rience some abstract definition of an objective
world, but a particular world. And that world
is variously related to the physical embodi-
ment (sensors and effectors), personal his-
tory, cultural contexts, and the like. Thus, for
example, we may see different species occupy-
ing the same part of a physical environment,
but from the perspective of their biology and
their behavior, the Umwelt of each species can
be almost entirely non-overlapping.
More recently, this notion is echoed in
descriptions of biological phenomena at
varying scales.
“…when scientists tell us that life adapts to
an essentially passive environment of chemis-
2005, vol. 1, no. 1
try, physics, and rocks, they perpetuate a
severely distorted view. Life actually makes
and forms and changes the environment to
which it adapts. Then that “environment”
feeds back on the life that is changing and act-
ing and growing in it. There are constant
cyclical interactions.” (Margulis 1989)
It also suggests one of the many reasons
that researchers in this tradition have turned
to philosophers such as Heidegger, with his
emphasis on being-in-the-world as a funda-
mental unit, rather than separate subjects and
objects (Heidegger 1962). In many ways, the
notion of an interpenetration of being and
world goes back at least to Berkeley and Kant.
One way it has manifested in cognitive science
is in Dewey’s (1896) proposal for a sensorim-
otor model of biological functioning which
was further elaborated as a foundational con-
cept for Piaget's theory. A similar concept
appears in proposals for cognitive models by
different researchers, such as Merleau-Ponty
(1963), Gregory Bateson (1979), Heinz von
Foerster (1984), and Kurt Lewin in his argu-
ments for understanding human cognition in
terms of a “life field,” e.g., “the person and his
environment have to be considered as one
constellation of interdependent factors”
(Lewin 1946, p. 793).
A key set of issues, then, involves the study
of how biology, personal history, and the
environment as it is for the organism co-spec-
ify each other; how they co-adapt and co-
evolve. Relevant work spans every scale of
biological phenomena, from the structural
coupling of biological cells (Maturana 1980a)
to cell-assemblies (Edelman 1992) to multi-
cellular organisms (Freeman 2001) to popu-
lations of organisms (Lewontin 2000); it also
raises serious questions about realist interpre-
tations of natural selection as “optimization”
(e.g., Varela, Thompson & Rosch 1991, chap.
9); and it challenges realist versions of co-
adaptation (e.g., Lumsden & Wilson 1981),
where attempts are made to resolve the debate
over genetic versus environmental determin-
ism by simply proposing that cognition is
some propitious “mix of the two” (for partic-
ularly lucid and rigorous critiques of these
attempts, and the entire “nature versus nur-
ture” impasse, see the work of Susan Oyama
2000a, 2000b).
Just as there are objectivist versions of con-
structivism, there is much work in cognitiv-
ism (especially robotics) that is arguing for
enactive cognitive science
the importance of embodiment (see Ander-
son 2003 for an overview). Much of this work
even references Rosch as a positive influence,
interpreting her results in ways that she her-
self does not. The perspective of many
researchers who now champion embodiment
tends to focus on what they see as the funda-
mental problem of abstract models in cogni-
tivism: the assumption of an independent
objective world leads to problems “ground-
ing” symbols in that world, e.g., (Harnad
1990). Their alternative is to develop embod-
ied models, where sensory inputs and motor
outputs play an important part in the cogni-
tive processing. Another alternative being
explored is to try and do away with internal
mediating mechanisms altogether – and
focus on models of “direct perception” or var-
ious behavior-based approaches in which the
environment functions as “its own represen-
tation.” (For an overview of much of the
recent work on robotics, see Pfeifer & Scheier
1999.) Notice that none of this requires any
radical reorientation away from objectivism,
and in many cases there is no such reorienta-
tion.
But the enactive emphasis on embodiment
is more than a physicalist assertion about per-
ception and motor activity. Rather it is about
how worlds are enacted, how those worlds are
for individuals, and how they relate to
embodiment. The key point for enactive cog-
nitive science is that the Umwelt is enacted
rather than given, independently existing, and
directly accessible.
As we will see below, this concept of the
Umwelt has a number of consequences, not
least for the notion of whether/how a cogni-
tive being “recovers” the world. To take one
example, the issue of cognitive activity can
look very different from an enactive perspec-
tive. The emphasis on cognitive activity has
been an essential focus for constructivist
models of cognition, e.g., Dewey (1896), Gru-
ber & Voneche (1977), Mead (1934), Vygotsky
(1978), Leont’ev (1978), Luria (1976). More
recent formulations tend to emphasize situ-
ated action which involves a shift of perspec-
tive from abstract, general (albeit, possibly
active) models of cognition to models that
involve moment-by-moment pragmatic
(James 1907, Dewey 1916) coping and adapt-
ing. As a result, researchers have started to
describe such situated (cognitive) activity in
such terms as readiness-to-hand (Heidegger
27
 engineering-computer scientific
SURVEY
1962), “tinkering” (or bricolage) (Papert
1980), and improvisation (Agre & Chapman
1990). However, as with many of the key
terms, it is used to variously emphasize every-
thing from non-planned action (Suchman
1987) to social situatedness (Clancey 1995)
(see Chaiklin & Lave 1993 for a discussion).
For the enactive tradition, situated action
is partly concerned with action rather than
“passive reception” or “imprinting” – and
partly concerned with situated coping rather
than with, say, general rules. But it is also con-
cerned with more than simply “actively cop-
ing” with an objectively-existing and dynamic
world; the bigger issue is how such situated
activity participates in the enactment of a life-
world.
“[‘Natural’ biological persons] are not
self-determining in any simple sense, but they
affect and “select” influences on themselves by
attending to and interpreting stimuli, by seek-
ing environments and companions, by being
differentially susceptible to various factors, by
evoking reactions from others.” (Oyama
2000a, pp. 180–181)
Lave and Wenger (1991), in their discus-
sion of situated action, propose a perspective
that is quite resonant with the concerns of
enactive cognitive science as described here;
we will return to this in the section on social
cognition below. Now, we turn to the issue of
deliberate autonomous action – and the
related issues of cognitive change and novelty.
Autonomy, change, and creativity
“…autonomous systems stand in sharp contrast
to systems whose coupling with the environment
is specified through input/output relations. …the
meaning of this or that interaction for a living sys-
tem is not prescribed from outside but is the result
of the organization and history of the system itself
– Varela, Thompson & Rosch 1991, p. 157
The concept of autonomy is largely absent
from the physical sciences – and to the extent
that cognitivism has followed this, it has
largely treated autonomy as an illusion or
epiphenomenon. The issue of autonomy has
entered certain kinds of work in cognitive sci-
ence, particularly work on autonomous
robots and agents inspired by the cybernetics
tradition. It is important, however, to clarify
what is actually at stake for the different kinds
of research. For much of the work in robotics,
the goal is simply to develop a device that can
28
work “unattended.” It is this sense that gave
impetus to the early technical developments
of cybernetics (Wiener 1948; Ashby 1956). A
separate, though related use, is where the goal
is develop a robot that can “arrive at its own
conclusions, decisions, and actions.” This
meaning is closer to our everyday use when
we say that we expect most children to grow
into adults who are autonomous. Although
this second meaning sketches a connection to
cognitive models, it is important to realize
that most work in robotics is not devoting any
major effort to model something approxi-
mating human autonomy (in fact, many in
the field would deny that autonomy, in the
everyday sense, is even meaningful within the
framework of conventional, deterministic
science). Typically, this second use of auton-
omy is to indicate that there are conditions for
which the robot cannot be pre-programmed,
so various techniques have to be implemented
that will allow it to adapt, succeed, and survive
“on its own.” Thus, even when work on robot-
ics includes a short “statement of belief ” that
robot implementations need to be “freed” of
the intentions of their designers (Pfeifer &
Scheier 1999), it is typically in this sense of
autonomy and freedom.
Within enactive cognitive science, how-
ever, a central tenet is that cognition involves
some substantive degree of autonomy; the
claim is that cognitive agents are not just buf-
feted by the twin forces of “nature versus nur-
ture” (nor do they simply “vanish” at the
intersection of such forces). Piaget, for exam-
ple, is perhaps most explicit in his attempt to
address it with his claim that development
involves a progressive “liberation” from sen-
sory-perception; in a larger sense, it seems to
inform his entire research, as when he repeat-
edly asserts that his model is neither environ-
mental nor genetic determinism. We will look
more closely at this topic in the second part of
this paper when we consider some of the
mechanisms proposed to explain and model
various aspects of enactive cognition.
Certain robotics research seems informed
by a perspective that is quite close to the enac-
tive notion of autonomy. In their discussion
of Umwelt, for example, Sharkey and Ziemke
clearly articulate a distinction about how cur-
rent robots may “be physically grounded but
…not rooted in their own world; they are
moving objects in our world” (Sharkey &
Ziemke 1998, p. 384). They are exemplary in
enactive cognitive science
highlighting these distinctions – and high-
lighting some of the potential risks of cogni-
tive science fooling itself with “wishful attri-
bution” if new terms such as embodiment are
used without actually changing any of its epis-
temological commitments. Nonetheless, as
we will see in part two of this paper, it is not
always clear how and to what extent such
enactive understanding translates into actual
research or implementation choices.
The issue of autonomy raises, by exten-
sion, the issues of cognitive change and cre-
ativity. Again, the standard view inherited
from physics is that both change and novelty
are illusions. Although these are early days
yet, the enactive claim is that genuine change
and creativity can be made reasonably rigor-
ous, rather than, say, being an alternative way
to talk about limited knowledge or predictabil-
ity (e.g., chaos theory). In the second part of
this paper we will look at several proposals for
possible formalisms, mechanisms, and mod-
els of innovation, e.g., dissipative structures
(Prigogine & Stengers 1984), symbiosis (Mar-
gulis & Fester 1991), autopoiesis (Maturana &
Varela 1980), a schema mechanism (Drescher
1991), dynamical systems models (Thelen &
Smith 1994), and self-organization (Kauff-
man 1993).
Groundlessness,
non-correspondence, and viability
“The worlds enacted by various histories of struc-
tural coupling are amenable to detailed scientific
investigation, yet have no fixed, permanent sub-
strate or foundation and so are ultimately ground-
less – Varela, Thompson & Rosch 1991, p. 217
Classical objectivist models of epistemology
tend to focus on aspects of knowledge related
to Truth; that is, whether some knowledge is
correct in its correspondence to what is True,
how one can verify such correspondence, and
the like. The influence of this tradition is felt
in many of the guiding assumptions of objec-
tivist science, with its attempt to identify the
ultimate elements of objective, material real-
ity – or to understand cognition by various
models of correspondence between mind and
world. Correspondence models of cognition
tend to focus on the level of perception or
concepts. At the level of perception, the prob-
lem is treated as: “how does sense-perception
retrieve raw sensory-data” (which leaves as a
problem how that sense-data is transformed
Constructivist Foundations
 engineering-computer scientific
SURVEY
into “what we experience”); the work of Mat-
urana and Freeman, sketched earlier, talks to
this. The second approach is to develop some
model for how concepts (in the head) corre-
spond to natural categories (in the world); the
work of Rosch raises serious questions about
that alternative.
In keeping with the radical constructivist
claim that human cognition has no access to
an objective reality, the models and explana-
tions of enactive cognitive science do not
assume ultimate foundations – and the mod-
eled systems do not rely on correspondence.
“It is clear that the notion of information pre-
existing in the world must be rejected” (Reeke
& Edelman 1988). By this view, cognitive
agents are not successful by virtue of their
access to an independent reality, nor do they
operate upon more or less accurate represen-
tations (or models) of such a reality, nor could
they even have such access. A major assump-
tion, then, is that although cognition is con-
strained and viable, it is largely provisional.
Concepts and categories and mind and world
change – sometimes in ways we barely notice,
and other times in ways that are radical reori-
entations. By this view, theories and models
are not provisional because “we have yet more
to learn about the full Truth,” but the pro-
cesses that give rise to the viable regularities of
our world also enact the challenges and dis-
crepancies we experience (in our theories and
our experiences). For many, this rejection of
absolute foundations tends to raise concerns
about relativism. But, as Mark Johnson notes,
“…this fear of relativism is predicated upon a
false assumption about the nature of objectiv-
ity – that either we have absolute foundations,
or there are not foundations of any sort what-
ever” (Johnson 1987, pp. 199–200).
Note that the turn away from ultimate
grounds and correspondence is not based
solely on a critique of objectivist attempts to
discover ultimate material foundations or ref-
erents. This point is not always so clear, since
some theorists critical of assumptions about
ultimate physical grounding have proposed
alternative “ultimate” grounds, such as soci-
ety, language, activity, interaction, conscious-
ness, or experience. Such a substitution, of
one ultimate ground for another, is not the
proposal made by enactive cognitive science.
Rather, the turn is to models and mechanisms
of viability. We will examine this topic more
closely in the part devoted to possible models
2005, vol. 1, no. 1
and mechanisms of cognition, but we will see
some of the relevant issues in the next section
on emergence and self-organization.
Emergence and self-organization
Cognition is enactively emergent co-determina-
tion of neural elements (local) and cognitive sub-
ject (global) – Varela 1999a, p. 81
Traditionally, science has tended towards
models of linear causality. With the rise of sta-
tistical mechanics and computational-based
approaches to science, a number of phenom-
ena gained prominence (e.g., probability,
concurrency, emergence, and the like).
Within objectivist approaches to cognition,
the study and modeling of emergent systems is
often concerned with how surprising and
complex results can emerge as a result of spec-
ifying and running simple rules – or, how cen-
tralized results can be produced by decentral-
ized activities. A related sense of emergence,
one more common in disciplines influenced
by systems theory, is that higher-level phe-
nomena can emerge (as “epiphenomena”)
out of the interactions between lower-level
entities (e.g., Holland 1998).
This second sense is close to the concerns
informing certain kinds of work in enactive
cognitive science. By this view, although
macro-phenomena emerge out of the micro-
scale, they are not entirely reducible to it; the
macro-level has certain characteristics and
regularities specific to it. (For brief history of
this version of the concept, see O’Connor &
Wong 2005). In some cases, this is expressed
in terms of the macro-level constraining the
micro-level (Pattee 1973, Freeman 2001), in
other cases, the claim is that the macro may
also have causal efficacy upon the micro (out
of which it emerges) (Martinerie et al. 1998;
van Quyen et al. 1997). In either case, the
claims such as, e.g., “wholes are not the simple
sum of their parts” (Lazlo 1972, p. 28), should
be resonant for phenomenologists who argue
that is not meaningful to discuss “properties”
(in the abstract) separate from the particular
whole that is of concern.
One of the key claims of structuralism,
Gestalt psychology (Köhler 1992), general
systems theory (Bertalanffy 1968, Lazlo
1972), and the more recent “sciences of com-
plexity” (Kauffman 1993), is that it is not
enough to consider constituent elements and
their causal influences, one needs to also
enactive cognitive science
include spatial and temporal relationships.
There is thus a shift to the analysis and
description of how systems are organized. For
physical or chemical systems, this may involve
various auto-catalytic (Prigogine & Stengers
1984, Eigen & Schuster 1979) or synergetic
effects (Haken 1987); for cognitive systems, it
can involve how such organizational relation-
ships are maintained (or change) over time
(Maturana & Varela 1980, Bateson 1979). One
consequence of this emphasis on emergent
phenomena is that, in addition to such forces
as genetic unfolding, environmental influ-
ences (“selection”), and the like, there have
been substantive proposals for various self-
organizing models and mechanisms (von Foe-
rster & Zopf 1962).
More recently, there is growing emphasis
on co-emergence. Prigogine, for example, in
discussing his work on dissipative systems,
writes “[there is not] any fundamental mode
of description; each level of description is
implied by another and implies the other. We
need a multiplicity of levels that are all con-
nected, none of which may have a claim to
preeminence” (Prigogine & Stengers 1984,
p. 300). Indeed, a wide variety of researchers
and theorists have argued for a new model of
reductionism (Fodor 1975; Churchland &
Sejnowski 1992; Thelen & Smith 1994), where
the notion of explaining a “level” of some
phenomenon does not proceed simply by ref-
erence to the next lower level. Thelen and
Smith, for example, express the co-emergent
perspective in their lucid phrase, “the power
of explanation is the dynamics of the pro-
cesses, in the view from below examined from
above. The explanatory power is the joint con-
sideration of the micro- and macrolevels”
(Thelen & Smith 1994, p. 39). In this spirit,
one recent, co-enactive proposal for a neuro-
logically-based hypotheses about perception
is Skarda (1999).
A related phenomenon is visible in the case
of multi-cellulars: replacement emergence.
That is, the cells that constitute bodies are
constantly being replaced. So, it is not just
remarkable that coherent and enduring bod-
ies emerge out of these dynamic parts – but
that the experience/perception of wholeness
and continuity exists at all. This is a variation
on the perception of continuity when watch-
ing a sequence of frames of a film; but in the
case of cognitive replacement, it is ourselves,
our bodies, our fleeting thoughts, impres-
29
 engineering-computer scientific
SURVEY
sions, and emotions that somehow enact and
maintain some sense of a unified and endur-
ing self. We now turn to examine this issue of
consciousness and self-awareness more
closely.
Consciousness and
first-person cognition
…if cognitive science is to include human experi-
ence, it must have some method for exploring and
knowing what human experience is
– Varela, Thompson & Rosch 1991, p. 23
Standard accounts in psychology texts tend to
assert that studies of consciousness (and the
use of introspective methods) were aban-
doned as unscientific at the end of the 19th
century – and that twentieth century psychol-
ogy had little to say about either conscious-
ness as some extra-physical mental phenom-
enon (“awareness”) or consciousness as the
experience of what it is to actually taste an
apple or see the color blue (“first-person
experience”). This version of history, how-
ever, glosses over a number of substantial
contributions to both theory and study of
consciousness during this same period (e.g.,
Vygotsky 1979). Phenomenology, for exam-
ple, was developed as a philosophy of lived
experience (Husserl 1962, Heidegger 1962)
and also formed the basis of cognitive
research, e.g., (Merleau-Ponty 1963). And
this notion of lived experience informs much
of the enactive concern with Umwelt – with
the phenomenal world.
As a number of commentators (e.g.,
Skarda 1999) have noted, the majority of
research in both neuroscience and psychology
still tends to avoid the question of how per-
cepts are transformed into experience. This
issue is independent of whether or not one is
a realist – indeed, it may be more challenging
for realist research that assumes, say, some
objective light-signals are transformed into
the colors of the phenomenal experience.
Indeed, attempts to explain color perception
and experience in terms of objective proper-
ties of light, materials, or biological mecha-
nisms are famously problematic. (For a thor-
ough analysis of all the different attempts to
resolve this in the area of color-perception, see
Thompson 1995).
In recent years, there has been a steady
resurgence of interest and study of the topic
(see, for example, the Journal of Consciousness
30
Studies; and for a collection of articles repre-
senting the diversity of approaches being
explored, along with peer-commentary and
critiques, see Varela & Shear 1999). Such work
attempts to address such issues as whether
consciousness can be studied scientifically,
how such research can proceed, what consti-
tute “results” of such research, the degree to
which such results can be shared, possible
applications of such research (e.g., therapy),
and the like.
Unfortunately, certain aspects of this issue
has informed much of the critical reaction to
enactive cognitive science. When Varela et. al.
proposed enactive cognitive science in their
book, The Embodied Mind, they did so by
describing one possible approach to the study
of consciousness and first-person experience,
namely methods and insights from Buddhist
meditation and epistemology – and also by
suggesting ways in which contemporary west-
ern cognitive science could inform Buddhist
epistemology. Much otherwise careful com-
mentary on the book seems to have not inter-
preted the portions devoted to Buddhism as
an extended example; instead, comments
have tended towards dismissing this aspect of
the book as championing some form of east-
ern mysticism.
The topic of consciousness not only raises
problems of method for its study, the role of
consciousness raises questions of method for
cognitive science. The role of the observer, for
example, has raised the familiar constructivist
critique, “objectivity is the delusion that
observations could be made without an
observer” (Foerster, quoted in Glasersfeld
1995). Taken seriously, the role of the
observer raises a number of challenging issues
in most of contemporary science. This is per-
haps most widely-known in the context of
quantum mechanics, but we will return to
this topic of the observer, and its possible
implications for method in cognitive science,
in the second part of this paper (but see, e.g.,
Maturana 1988).
Social cognition & intersubjectivity
…intelligence shifts from being the capacity to
solve a problem to the capacity to enter into a
shared world of significance
– Varela, Thompson & Rosch 1991, p. 207
As with each of the concerns described here,
there is cognitive research on “social cogni-
enactive cognitive science
tion” that does not seem to be constructivist –
realist or otherwise – in the epistemological
sense considered in this paper. Such work
tends to concentrate on such things as how
“culture, constituted of tools, artifacts, and
ways of thought, …carries the past history of
a society into the present, thereby both
enabling and constraining current thinking”
(Resnick 1991, p. 18).
When we turn to the constructivist tradi-
tion, we encounter what is perhaps its most
vexing and heated controversy: the debate
about “individual” versus “social” construc-
tion. The debate is complex and turns on a
number of different assumptions and varia-
tions on the general theme of whether society
derives from individuals – or whether indi-
viduals (and individual cognition) derive
from the social. One curious feature of this
otherwise constructivist debate is that each
side seems to feel they are providing a “correc-
tive” for the excessive subjectivist/objectivist
tendencies of the other. That is, much of social
constructivism’s critique of individual con-
structivism seems to turn on the claim that
individual-oriented constructivism is pro-
posing a mode of cognition happening “in the
head”; by this social constructive view, then,
individual constructivism is perceived as
something close to subjectivism. On the other
hand, proponents of individual constructiv-
ism often criticize the work of social construc-
tivism as being too realist in its model of
“transmission” between society and individ-
ual, as if the “society” now stands as a substi-
tute for The World and individuals somehow
have direct access to it.
Although a detailed discussion of this
debate is beyond the scope of this paper, we
will return to aspects of this topic in the sec-
ond part when we highlight some of the
research methods, interpretations of research
results, and proposals for cognitive mecha-
nisms and models. For now, this debate may
help to highlight the enactive perspective,
namely that the individual and social are co-
enacted. This perspective is present to varying
degrees in much of the founding thought on
social cognition (Dewey 1900, Mead 1934,
Vygotsky 1978, Heidegger 1962). However,
there are at least two aspects of social con-
structivism that are important to highlight in
this context: internalization and precedence.
As Lave and Wenger note (1991, pp. 47–
49), the issue of “internalization” is the source
Constructivist Foundations
 engineering-computer scientific
SURVEY
of a number of interpretations, ranging from
the overtly objectivist (“transmission mod-
els”) to more textured, relational models. It is
not always clear, for example, which version
of internalization Mead is proposing. In
much of his work, he seems clearly to be say-
ing that individuals become individuals by
absorbing the social – and for Mead, the
means for this is through social gestures (and
presumably, a model of perception that is
largely realist). In other cases, he seems to be
wrestling to express something close to the
more contemporary view from enactive cog-
nitive science, e.g., “No individual has a mind
which operates simply in itself, in isolation
from the social life-process in which it has
arisen or out of which it has emerged” (Mead
1934, p. 222).
The case of Vygotsky and the other Soviet
constructivists (e.g., Vygotsky 1978, Leont’ev
1978, Luria 1976) is even less clear – and this
is partly due to their subsequent commenta-
tors and interpreters. (Indeed, the issue of
“internalization” is equally unclear in the
Piagetian tradition – and for many of the
same reasons.) Certainly, some contempo-
rary social constructivists invoke the Soviet
constructivists in the same breath as Mead
and the pragmatists – and present them as
having largely the same epistemological ori-
entation. Others (e.g., Wertsch 1991) present
such work in a way that is closer to the “mutu-
ally constitutive” model articulated by Lave
and Wenger. Even contemporary social con-
structivists who have been vocal critics of
individual-oriented constructivism have
expressed views that can be interpreted as
broadly enactive, e.g., “a radical social con-
struction places [psychological functioning,
materialist metaphysics, and the existence of
macro-social institutions] presumptions in
brackets” (Gergen 1997).
One thing that seems more certain is that
for most social constructivists, the social is
primary and the individual is derived from it,
e.g., “consciousness is an emergent from such
[social] behavior; that so far from being a pre-
condition of the social act, the social act is the
precondition of it” (Mead 1934, p. 18) and
“the social precedes the personal” (Gergen
1997) (emphasis added).
As we should expect, the enactive perspec-
tive does not attempt to prioritize one over
the other. Lave and Wenger express this per-
spective nicely in their interpretation of situ-
2005, vol. 1, no. 1
ated action, which “dissolves dichotomies
between contemplation and involvement,
between abstraction and experience: persons,
actions, and the world are implicated in all
thought, speech, knowing, and learning”
(Lave & Wenger 1991, p. 52).
For enactive cognitive science, individuals
as individuals arise via the social – and vice
versa. To explain this, we can contrast non-
radical with radical versions of social con-
structivism. For non-radical social construc-
tivism, human society is the dominant envi-
ronment into which infants are born; this is
sometimes presented as a variation on the
work of Piaget, whose model is described (in
this context) as one in which children interact
with objects in the world, rather than within
social contexts. By this non-radical view of
social constructivism, then, the world of the
child is largely social, and it is within that
social world that active cognitive construc-
tion takes place. A more radical version of
social constructivism is one that emphasizes
the co-enaction of individual and society;
human consciousness and language as such
only arise out of this co-enaction at the social
scale.
Before ending this section, we will briefly
address the issue of intersubjectivity, or, how
we seem to know what thoughts or feelings
are in the minds of others. This issue presents
a theoretical challenge to much of cognitiv-
ism, which takes the perspective that minds
are “inside heads.” How do such models
account for knowing the contents of other
people’s minds?
Aside from the theoretical challenges,
there is a great deal of research documenting
phenomena that strongly suggest people do
know and act in ways that rely on their knowl-
edge of how others think, feel, and the like.
Research on intersubjectivity typically takes
place on the psychological or social level (e.g.,
Trevarthen 1993, Bruner 1986). There is,
however, also work to sketch explanations
from biology. Thus, although the fundamen-
tal unit of analysis in autopoietic theory is
clearly the individual (biological cell) and
clearly biological, by this theory, the group/
society is necessary for each type of individ-
ual, from single-celled organisms to inverte-
brates to humans (though, of course, the dif-
ferent structures participate in the co-
enaction of different forms of social/individ-
ual dynamics). Indeed, the model of con-
enactive cognitive science
sciousness and the particular form of “lan-
guaging” of humans proposed by Maturana
and Varela is one that is deeply intertwined
with the social. By this view, without such
social co-enaction, the human consciousness
of consciousness is not possible.
Intersubjectivity as formulated by enactive
theory is particularly complex and subtle, but
one of the main points is that “the minds of
others” are no more “separate and indepen-
dent” than is the world, social or otherwise,
that co-arises with the individual. In a sense,
for enactive cognitive science, intersubjectiv-
ity arises naturally out of co-enaction, where
“interactions between people are codepen-
dently defined, experienced, and acted out”
(Rosch 1999, p. 72). We now turn to this last
major research theme.
Co-enaction
…knower and known, mind and world, stand in
relation to each other through mutual specifica-
tion or dependent coorigination – Varela, Thomp-
son & Rosch 1991, p. 150
As suggested earlier, the history of objectivist
science is deeply informed by the idea of ulti-
mate foundations or causes. We thus see many
hypotheses and descriptions presented rela-
tive to some assumed absolute frame of refer-
ence. The move to enactive cognitive science
is stronger, and more radical, than some sim-
ple notion of relativism.
The idea of objective/subjective, physical/
social co-enaction is clearly articulated by
Heidegger (1962) – and Goodman (1978)
hints at something similar in his model of
worldmaking. However, it may be most
clearly illustrated by considering different
perspectives on evolutionary theory. In tra-
ditional accounts, there is adaptive evolution
of historical lineages because, in the succinct
formulation of the standard view, “the
organism proposes and the environment dis-
poses.” By this view, lineages progressively
adapt to a (more or less) unchanging, objec-
tive environment. Recent evolutionary the-
ory has modified this to a certain extent with
such proposals as the “Red Queen hypothe-
sis” (van Valen 1973). This perspective sug-
gests, by analogy to the eternal running of the
Red Queen in Lewis Carroll’s Through the
Looking Glass, that lineages are adapting to a
“moving target” – an environment that is,
itself changing.
31
 engineering-computer scientific
SURVEY
A more radical perspective is that lineages
and environments are changing because their
changes bring about changes in each other; in
other words, by this view, the evolving envi-
ronment of organisms is evolving partly
because of those organisms – because of bio-
logical transformations of materials and
energy, and because organisms actively mod-
ify their environments in different ways. As
Richard Lewontin (2000, p. 58) says, “the
constructionist view is that the world is
changing because the organisms are chang-
ing”. Similar examples of co-determination –
and co-regulation – are seen throughout the
life sciences, from the self-regulation of bod-
ies (homeostasis, Cannon 1932) to atmo-
spheric regulation at the global scale (Vernad-
sky 1998, Lovelock & Margulis 1974).
This radical perspective is resonant with
the more encompassing notion of co-enac-
tion, which may be one of the most distinctive
concerns of enactive cognitive science. The
idea is that different cognitive phenomena
across different scales are variously co-origi-
nating, co-generating, co-specified, co-deter-
mined, co-adaptive, co-constrained, co-
evolving, and co-emergent. Thus, society and
individual co-arise, mind and world co-arise,
micro and macro co-arise, and the like.
Throughout most of the history of con-
structivist cognitive science, attempts have
been made to indicate and formulate some-
thing similar, variously as upward and down-
ward causation, as circular causality, reactions,
or mechanisms, as reciprocal causality, as co-
arising or co-emergence, as mutual causality,
References
Agre, P. E. & Chapman, D. (1990) What are
plans for? Robotics and Autonomous Sys-
tems 6: 17–34.
Anderson, M. L. (2003) Embodied cognition:
a field guide. Artificial Intelligence 149:
91–130.
Ashby, R. (1956) Introduction to cybernetics.
John Wiley: New York.
Bateson, G. (1979) Mind and nature: A neces-
sary unity. Ballantine Books: New York.
Berlin, B. & Kay, P. (1969) Basic color terms:
Their universality and evolution. Univer-
sity of California Press: Berkeley.
Bertalanffy, L. von (1968) General systems
theory. Braziller: New York.
32
determination, selection, circularity, or genera-
tion, and the like. Cybernetics, for example,
has a long history of addressing questions of
circular-causality and even the role of the
observer in such systems (e.g., second-order
cybernetics, Foerster 1996).
One reason for choosing the term co-enac-
tion for this paper is that many of the other
terms tend to contribute to our difficulty in
thinking about and understanding simulta-
neous phenomena; we do not have very well
developed intuitions about them – and thus,
it is easy to assume that such proposals involve
such things as circular sequences of causality,
e.g., “first a has an effect on b, and then (as a
result), b has an effect on a.” This phenomenal
pattern is one possibility, but another involves
simultaneous acting-upon, constraining,
influencing, and, yes, even generating each
other.
More importantly, the term co-enaction
emphasizes that worlds-as-worlds (and indi-
viduals-as-individuals) are enacted – and that
ABOUT THE AUTHOR
Kevin McGee has a PhD in Media Arts and
Sciences from MIT. His main research inter-
est is the development of partner technol-
ogies based on enactive cognitive models.
He is currently affiliated with the Depart-
ment of Computer Science, Linkoping Uni-
versity and the Santa Anna Research
Institute, both in Linköping, Sweden.
Boden, M. A. (2000) Autopoiesis and life.
Cognitive Science Quarterly 1: 117–145.
Brooks, R. A. (1991) Intelligence without rep-
resentation. Artificial Intelligence Journal
47: 139–159.
Brown, R. (1958) How shall a thing be called?
Psychological Review 65: 14–21.
Bruner, J. S. (1966) Toward a theory of
instruction. Harvard University Press:
Cambridge, MA.
Bruner, J. S. (1986) Actual minds, possible
worlds. Harvard University Press: Cam-
bridge, MA.
Cannon, W. B. (1932) The wisdom of the
body. Norton: New York.
Capra, F. (1996) The web of life: A new scien-
tific understanding of living systems.
enactive cognitive science
such enaction is not localized to individuals,
social structures, language, or worlds. And, as
should be clear by this point, the enactive
emphasis differs radically from recent work in
cognitivism that proposes possible cognitive
mechanisms for coping with perceptual
changes that occur as a result of individuals
changing the state of an objective world (or
their perspective on it) by their actions.
Of course, one current difficulty with the
concept of co-enaction is that although it may
be possible to loosely describe certain phe-
nomena in these terms, and although these
descriptions may be be resonant with our
intuitions, it is, as yet, difficult to express this
in terms that are formally precise enough to
speak about hypothetical mechanisms – or to
implement as synthetic hypotheses, explana-
tions, or models. There are, nonetheless,
researchers attempting to articulate these
kinds of issues in the context of emergence
(and connectionist implementations) – or,
more recently, in terms of dynamical systems
models.
In part two of this paper, we will explore
this issue in more detail as well as examine
how the other research themes of enactive
cognitive science have informed the develop-
ment of relevant methods – and proposals for
different kinds of cognitive explanations,
mechanisms, and models.
Acknowledgments
For invaluable feedback and suggestions, I
would like to thank Alexander Riegler, Mario
Bourgoin, and the two reviewers of this paper.
Anchor Books: New York.
Chaiklin, S. & Lave, J. (eds.) (1993) Under-
standing practice: Perspectives on activity
and context. Cambridge University Press:
New York.
Churchland, P. S. & Sejnowski, T. J. (1992)
The computational brain. MIT Press:
Cambridge, MA.
Clancey, W. J. (1995) A tutorial on situated
learning. In: Self, J. (ed.) Proceedings of
the International Conference on Comput-
ers and Education. AACE: Charlottesville,
VA, pp. 49–70.
Dewey, J. (1896) The reflex arc concept in psy-
chology. Psychological Review 3: 357–70.
Dewey, J. (1900) Psychology and social prac-
tice. Psychological Review 7: 105–124.
Constructivist Foundations
 engineering-computer scientific
SURVEY
Dewey, J. (1916) What pragmatism means by
practical. In: Essays in experimental logic.
University of Chicago: Chicago, IL, pp.
303–329.
Drescher, G. L. (1991) Made-up minds: A
constructivist approach to artificial intelli-
gence. MIT Press: Cambridge, MA.
Edelman, G. M. (1992) Bright air, brilliant
fire: On the matter of the mind. Basic
Books: New York.
Eigen, M. & Schuster, P. (1979) The hypercy-
cle: A principle of natural self-organiza-
tion. Springer-Verlag: Berlin.
Fodor, J. (1975) The language of thought.
Harvard University Press: Cambridge,
MA.
Foerster, H. von (1984) On constructing a
reality. In: Watzlawick, P. (ed.) The
invented reality: How do we know what we
believe we know? Contributions to Con-
structivism. New York: W.W. Norton and
Company, pp. 41–61.
Foerster, H. von (ed.) (1996), Cybernetics of
cybernetics. Second edition. Future Sys-
tems: Minneapolis, MN.
Foerster, H. von & Zopf, G. W. (eds.) (1962),
Principles of self-organizing systems. New
York: Pergamon.
Freeman, W. J. (2001) How brains make up
their minds. Columbia University Press:
New York.
Freeman, W. J. & Skarda, C. A. (1990) Repre-
sentations: Who needs them? In
McGaugh, J. L., Weinberger, N. & Lynch,
G. (eds.) Brain organization and memory
cells, systems & circuits. Oxford University
Press: New York, pp. 375–380.
Gadamer, H.-G. (1976) Philosophical herme-
neutics. University of California Press:
Berkeley, CA.
Gergen, K. J. (1997) Social theory in context:
Relational humanism. In: Greenwood, J.
(ed.) The mark of the social. Rowman and
Littlefield: New York.
Glasersfeld, E. von (1984) An introduction to
radical constructivism. In: Watzlawick, P.
(ed.) The Invented Reality: How do we
know what we believe we know? Contribu-
tions to constructivism. New York: W.W.
Norton, pp. 17–40. Also available from
http://www.vonglasersfeld.com/
Glasersfeld, E. von (1995) Radical construc-
tivism: a way of knowing and learning.
Falmer Press: London.
Goodman, N. (1978) Ways of worldmaking.
2005, vol. 1, no. 1
Hackett Publishing: Indianapolis, IN.
Gruber, H. E. & Voneche, J. J. (eds.) (1977)
The essential Piaget. Basic Books: New
York.
Haken, H. (1987) Synergetics: An approach to
self-organization. In: Yates, F. E. (ed.) Self-
organizing systems. Plenum: New York.
Harnad, S. (1990) The symbol grounding
problem. Physica D 42: 335–346.
Heidegger, M. (1962) Being and time. Harper
and Row: New York.
Holland, J. H. (1998) Emergence: From chaos
to order. Helix Books: Reading, MA.
Husserl, E. (1962) Ideas: General introduc-
tion to pure phenomenology. Collier: New
York.
James, W. (1907) Pragmatism: A new name
for some old ways of thinking. Longman
Green and Co: New York.
Johnson, M. (1987) The body in the mind:
The bodily basis of meaning, imagination,
and reason. The University of Chicago
Press: Chicago and London.
Kauffman, S. A. (1993) Origins of order: Self-
organization and selection in evolution.
Oxford University Press: Oxford.
Klin, A., Jones, W., Schultz, R. T. & Volkmar,
F. (2003) The enactive mind, or from
actions to cognition: lessons from autism.
Philosophical Transactions of the Royal
Society: Biological Sciences 358: 345–60.
Köhler, W. (1992) Gestalt psychology: An
introduction to new concepts in modern
psychology. Re-issue edition. Liveright:
New York.
Lakoff, G. (1987) Women, fire, and dangerous
things: What categories reveal about the
mind. The University of Chicago Press:
Chicago and London.
Lakoff, G. & Johnson, M. (1999) Philosophy
in the flesh: The embodied mind and its
challenge to western thought. Basic Books:
New York.
Lave, J. & Wenger, E. (1991) Situated learning:
Legitimate peripheral participation. Cam-
bridge University Press: New York, NY.
Lazlo, E. (1972) The systems view of the
world: The natural philosophy of the new
developments in the sciences. Braziller:
New York.
Leont’ev, A. N. (1978) Activity, conscious-
ness, and personality. Prentice-Hall:
Englewood Cliffs, NJ.
Lettvin, J. Y., Maturana, H. R., McCulloch, W.
S. & Pitts, W. H. (1959) What the frog’s eye
enactive cognitive science
tells the frog’s brain. Procedings of the
I.R.E. 47: 1940–1959.
Lewin, K. (1946) Behavior and development
as a function of the total situation. In: Car-
michael, L. (ed.) Manual of Child Pychol-
ogy. Wiley: New York, pp. 791–844.
Lewontin, R. (2000) The triple helix: Gene,
organism, and environment. Harvard
University Press: Cambridge, MA.
Lovelock, J. & Margulis, L. (1974) Biological
modulation of the earth’s atmosphere.
Icarus, 21.
Lumsden, C. J. & Wilson, E. O. (1981) Genes,
mind, and Culture: The coevolutionary
process. Harvard University Press: Cam-
bridge, MA.
Luria, A. R. (1976) Cognitive development:
Its cultural and social foundations. Har-
vard University Press: Cambridge, MA.
Margulis, L. (1989) Gaia: the living earth. The
Elmwood Newsletter, 5(2).
Margulis, L. & Fester, R. (eds.) (1991), Symbi-
osis as a source of evolutionary innova-
tion: Speciation and morphogenesis. MIT
Press: Cambridge, MA.
Martinerie, J., Adam, C., van Quyen, M. L.,
Baulac, M., Renault, B. & Varela, F. J.
(1998) Epileptic crisis can be anticipated
by non-linear analysis. Nature Medicine 4:
1173–1176.
Maturana, H. R. (1980a) Biology of cogni-
tion. In: Autopoiesis and cognition: The
realization of the living. Reidel: Boston,
pp. 1–58.
Maturana, H. R. (1980b) Introduction. In:
Autopoiesis and Cognition: The Realiza-
tion of the Living. Reidel: Boston, pp. xi-
xxx.
Maturana, H. R. (1988) Reality: The search
for objectivity or the quest for a compel-
ling argument. The Irish Journal of Psy-
chology 9: 25–82.
Maturana, H. R., Uribe, G. & Frenk, S. G.
(1968) A biological theory of relativistic
colour coding in the primate retina. Arch.
Biologia y Med. Exp. 1: 1–30.
Maturana, H. R. & Varela, F. J. (1980) Auto-
poiesis and cognition: The realization of
the living. Reidel: Boston.
Maturana, H. R. & Varela, F. J. (1987) The tree
of knowledge: The biological roots of
human understanding. New Science
Library: Boston.
Mead, G. H. (1934) Mind, self, and society.
University of Chicago Press: Chicago, IL.
33
 engineering-computer scientific
SURVEY
Merleau-Ponty, M. (1963) The structure of
behavior. Beacon Press: Boston, MA.
O’Connor, T. & Wong, H. Y. (2005) Emergent
properties. In: Zalta, E. N. (ed.) The Stan-
ford encyclopedia of philosophy (Summer
2005 Edition). http://plato.stanford.edu/
archives/sum2005/entries/properties-
emergent/
Oyama, S. (2000a) Evolution’s eye: A system’s
view of the biology-culture divide. Duke
University Press: Durham, NC.
Oyama, S., (2000b) The ontogeny of informa-
tion: Developmental systems and evolu-
tion. Second revised and expanded
edition. Duke University Press: Durham,
NC.
Papert, S. (1980) Mindstorms: Children,
computers, and powerful ideas. Basic
Books: New York.
Papert, S. (1988) Introduction. In: Embodi-
ments of mind. Revised edition. MIT
Press: Cambridge, MA, pp. xxi-xxviii.
Pattee, H. H. (1973) The physical basis and
origin of hierarchical control. In: Pattee,
H. H. (ed.) Hierarchy theory: The chal-
lenge of complex systems. Braziller: New
York.
Pfeifer, R. & Scheier, C. (1999) Understanding
intelligence. MIT Press: Cambridge, MA.
Piaget, J. (1955) The child’s construction of
reality. Routledge: London.
Prigogine, I. & Stengers, I. (1984) Order out
of chaos: Man’s new dialogue with nature.
Bantam: New York.
van Quyen, M. L., Martinerie, J., Adam, C.,
Lachaux, J.-Ph., Baulac, M., Renault, B. &
Varela, F. J. (1997) Temporal patterns in
human epileptic activity are modulated by
perceptual discriminations. Neuroreport
8: 1703–1710.
Reeke, G. N. & Edelman, G. M. (1988) Real
brains and artificial intelligence. Daedalus
117: 143–174.
Resnick, L. B. (1991) Shared cognition:
Thinking as social practice. In: Resnick, L.
B., Levine, J. M. & Teasley, S. D. (eds.) Per-
34
spectives on socially shared cognition.
American Psychological Association:
Washington, DC, pp. 1–20.
Rorty, R. (1979) Philosophy and the mirror of
nature. Princeton University Press: Princ-
eton, NJ.
Rosch, E. (1999) Reclaiming concepts. In:
Núñez, R. & Freeman, W. J. (eds.) reclaim-
ing cognition: The primacy of action,
intention, and emotion. Imprint Aca-
demic: Thorverton, UK, pp. 61–77.
Rosch, E. & Lloyd, B. (eds.) (1978) Cognition
and categorization. Lawrence Erlbaum:
Hillsdale, NJ.
Sharkey, N. E. & Ziemke, T. (1998) Biological
and physiological foundations. Connec-
tion Science 10: 361–391.
Singer, W. (1980) Extraretinal influences in
the geniculate. Physiology Review 57: 386–
420.
Skarda, C. A. (1999) The perceptual form of
life. In: Núñez, R. & Freeman, W. J. (eds.)
Reclaiming cognition: The primacy of
action, intention, and emotion. Imprint
Academic: Thorverton, UK, pp. 79–93.
Suchman, L. A. (1987) Plans and situated
actions: The problem of human-machine
communication. Cambridge University
Press: Cambridge.
Thelen, E. & Smith, L. B. (1994) A dynamical
systems approach to the development of
cognition and action. MIT Press: Cam-
bridge, MA.
Thompson, E. (1995) Colour vision: A study
in cognitive science and the philosophy of
perception. Routledge: London.
Trevarthen, C. (1993) The self born in inter-
subjectivity: The psychology of an infant
communicating. In: Neisser, U. (ed.) The
perceived self: Ecological and interper-
sonal sources of self-knowledge. Cam-
bridge University Press: Cambridge.
Uexküll, J. von (1957) A stroll through the
worlds of animals and men. In: Schiller, C.
H. (ed.) Instinctive behavior: The devel-
opment of a modern concept. Interna-
enactive cognitive science
tional Universities Press: New York, pp. 5–
80.
van Valen, L. (1973) A new evolutionary law.
Evolutionary Theory 1: 1–30.
Varela, F. J. (1999a) Ethical Know-How:
action, wisdom, and cognition. Stanford
University Press: Stanford, CA.
Varela, F. J. (1999b) Steps to a science of inter-
being. In: Watson, G., Batchelor, S. & Clax-
ton, G. (eds.) The psychology of awaken-
ing: Buddhism, science and our day-to-
day lives. Rider: London.
Varela, F. J. & Shear, J. (eds.) (1999), The view
from within: First-person Approaches to
the Study of Consciousness. Imprint Aca-
demic: Thorverton, UK.
Varela, F. J., Thompson, E. & Rosch, E. (1991)
The embodied mind: Cognitive science
and human experience. MIT Press: Cam-
bridge, MA.
Vernadsky, V. (1998) The biosphere. Com-
plete annotated edition. Springer: New
York.
Vygotsky, L. S. (1978) Mind in society: The
development of higher mental processes.
Harvard University Press: Cambridge,
MA.
Vygotsky, L. S. (1979) Consciousness as a
problem in the psychology of behavior.
Soviet Psychology 17: 3–35.
Wertsch, J. V. (1991) Voices of the mind: A
sociocultural approach to mediated
action. Harvard University Press: Cam-
bridge, MA.
Whitehead, A. N. (1979) Process and Reality.
New York: Free Press, corrected edition.
Wiener, N. (1948) Cybernetics, or Control
and Communication in the Animal and
the Machine. John Wiley: New York.
Wittgenstein, L. (1963) Philosophical Investi-
gations. Second edition. Blackwell:
Oxford.
Received: 15 September 2005
Accepted: 15 October 2005
Constructivist Foundations
 philosophical-epistemological epistemic structuring
CONCEPTS

People, Tools, and Agency:

Who Is the Kybernetes?
Herbert F. J. Müller A McGill University, herbert.muller@mcgill.ca

5 Jaspers has emphasized that the mind

Purpose: This conceptual-epistemological paper deals with the old problem of inversion (consciousness, experience, awareness) is
of thinking, as typified by traditional metaphysics-ontology. It is proposed that a thorough encompassing. His observation can help to
constructivism – which views structures of mind, nature, and all, as not derived from (not identify the cause of the inversion, and also to
referring to) any pre-structured given mind-independent reality (zero-derivation, 0-D) – correct it. All mental entities, from smells and
can go beyond this conceptual impasse; it can also serve as a fall-back position for positive toothaches to physical objects, physiological
ontologies. Practical implications: The practical result of 0-D is that all structures of activity, theories of the universe, and divine
experience are understood as tools serving individual and collective subjects. revelation, are created and posited within
Conclusions: This conceptual correction results in a simplification for the understanding experience, which becomes structured by this
of some conceptual puzzles, such as the mind-brain relation, but also in a considerable activity only. Gestalt- and word-concepts are
increase of responsibility, because entities and agents formerly considered responsible, working-tools, and they do not “refer” to enti-
and outside the mind, are recognized to be extensions of the subjects. Key words: Inver- ties outside experience (zero-derivation, 0-
sion of experience, 0-D constructivism, problem of metaphysics-ontology, mind-brain D). They may be certified as real or true by
relation. trust that they are reliable. The encompassing
Note:This paper is a revised version of the Karl Jaspers Forum http://www.kjf.ca Target mind-and-nature-experience is the back-
Article 78. It is enlarged by the discussion of selected comments and responses in the forum. ground for phenomena (i.e., tool-structures)
and the only start-point for thinking. Work-
ing- or feedback-ontology can then replace
Mit allen Augen sieht die Kreatur 3 Rilke can perhaps be seen as a mystic, traditional inverted ontology, with the struc-
das Offene. Nur unsre Augen sind and van der Meijden (C1, C22) writes that his tures as tools. Since the 0-D view does not
wie umgekehrt und ganz um sie gestellt thinking cannot be used for “linear” (i.e., log- refer to ontological structures, it cannot be
als Fallen, rings um ihren freien Ausgang. ical) discussion. I think this is misleading. de-constructed in the ontological sense. The
(R. M. Rilke, Anfang der 8. Duineser Elegie) Mysticism (if that is the right word for Rilke’s MIR-view is then understood to be a modifi-
poetry) means a holistic view, and I want to cation of 0-D, a shortcut which bypasses its
show the relevance of Rilke’s holism for con- source. (More on the question of realism ver-
Overview ceptual thinking (and besides, van der sus constructivism can be found in Johnson
Meijden himself recommends “holograms”). 2004 and discussion; and in Müller R16 on a
1 Cybernetic concepts like feedback and sta- In his 8th Elegy, Rilke shows a central aspect of paper by Nola.)
bility of function apply to the subjective reality that is mostly overlooked by realistic 6 The subject or self, and his/her activity,
aspects of human experience as well as to metaphysicists. This can help to determine as well as the subject-object split and other
objective physical and biological systems. why traditional realistic thinking creates dichotomies, are also structured within expe-
Does this imply that we must understand our- problems; the main purpose is to find the rience as needed, alongside other tool-struc-
selves objectively, as machines in the physical connection between overall experience and tures including machines. The self may be
or biological sense? Does it mean that specific concepts. neglected or even denied, but it is in principle
machines will be in charge of us? And who sets 4 Inverted thinking sees reality as mind- always an aspect of experience and reality. – It
the goals? independently pre-structured (MIR; or also has become possible to understand much of
2 The relation between people and mind-exclusive reality, which would empha- the physiological activity that is needed for
machines resembles the one with earlier size the implied absence of the subject). Con- experience and mental structures to occur.
human tools like concepts, which are also cepts are said to refer to external entities, such But these explanations happen inside experi-
sometimes thought to have mind-indepen- as gods, beings-in-themselves, nature-in- ence; they are not identical with experience,
dent qualities. This idea expresses an inversion itself, etc. This inversion is particularly strong and they cannot replace it. For objectivists,
of experience, described by the poet Rilke as a in exclusively-objective views. It leads to con- this may be difficult to see: but because con-
problem for humans but not animals. He ceptual difficulties, most obviously in the sciousness is encompassing, no one can leave
spoke of traps that block our access to the open. mind-brain (or mind-body) problem, which its bubble, in order to look at it objectively,
It may be worth examining Rilke’s intuition. is intractable in any version of the MIR-view. from outside. Brain function is, like other

Constructivist Foundations 2005, vol. 1, no. 1 35

 philosophical-epistemological
CONCEPTS
structures, formed (and to be understood)
inside consciousness, not vice versa. This sug-
gests a reformulation of the mind-brain ques-
tion: how can physiological knowledge help
us to deal better with ourselves?
7 Adams claims (C15, §11) that I “sud-
denly introduce the subject, without defini-
tion”; this is not correct, since I introduce it
from the beginning, as one of the structures
created in experience. He also writes that
“there is no need for a subjective aspect”
(Adams C15, §22), and that the physical body
is “inexplicable” (Adams C15, §15) because it
is one of the structures of the world. These
statements reflect an MIR-view on the part of
this commentator. They must be rejected, as
both subject and world (or object, including
the body) are parts of experience. The sub-
ject-construct forms pragmatically like every-
thing else, such as the subject-object split (cf.
§6), and is an aspect of all experience.
8 Awareness of the lack of outside sources
furthermore implies that individual and col-
lective selves are in principle in charge, even
though they may not be equipped to deal with
some situations, or incompletely formed, or
weak. In that case the leadership (goal-setting
and feedback-correction) falls back, by
default or by design, onto other agents:
humans, organizations like governments, or
sometimes machines. But although such
abdications of agency are common, there is
no reason to think that non-human agents
can or should take over, either to dominate us
or to relieve us from responsibility. Kant’s cat-
egorical imperative can provide a guideline
for ethical goal setting.
9 Some of these points will be compared
with the traditional naturalist (realist, meta-
physical-ontological) view of John Searle, as
he describes it in his recent book “Mind”
(2004).
First part:
People and tools
Cyborgs and cybernetics
10 The philosopher Ian Hacking, formerly of
Toronto and now in Paris at the Collège de
France, has recently given a presentation at
the Université du Québec in Montreal on the
role of the cyborg in everyday life. He pointed
out that every human being with an artificial
36
electronic implant or other aid (a cardiac
pacemaker, or a hearing aid, for instance) is a
cyborg. I am fairly deaf, and function better
with a hearing aid – but I don’t like the idea
that I depend on it, and others may have sim-
ilar feelings, but such gadgets are part of life.
And besides, what about airplanes and their
pilots, cars, bicycles, eye glasses, forks, knives,
and toothbrushes (with or without electron-
ics)?
11 The word cyborg was coined by Man-
fred Clynes in 1960, from “cybernetic” and
“organism,” in collaboration with Nathan
Kline, at the Rockland State Hospital in New
York State, USA, in work to devise ways for
humans to function in space exploration;
they presented this at a NASA conference.
There is even a cyborg manifesto on the inter-
net (http://cyborgmanifesto.org/ – with the
same general message as that of Hacking),
which one can support or oppose.
12 The idea of cyborgs has at present a
somewhat magical aura, it seems; a bit like
Harry Potter, or the notion that green men
from outer space will take over the earth. Peo-
ple like such phantasies because they take
them out of the confines of “official” reality,
truth and thinking, and this without respon-
sibility. But it may be of interest to look at this
in a less melodramatic context: in a wider
sense, some aspects of this question are quite
old.
13 The term cybernetics in turn was pro-
posed by Norbert Wiener in 1948 for “the
study of information and control in humans
and machines” (Wiener 1954), utilizing the
Greek word kybernetes, for helmsman. He
appears to have developed this concept dur-
ing World War II, while studying anti-aircraft
fire control. It is generally accepted by now
that cybernetic principles (such as feedback
and stability) also apply to biological systems.
Wiener’s work was also of influence on Heinz
von Foerster, one of the originators of con-
structivism, where it extends to psychology as
well.
Objective and subjective systems
14 At this point cybernetic views become rel-
evant for what happens in subjective experi-
ence, and this results in some conceptual puz-
zles. What is a “system”? The usual meaning is
objective, that is, a functional unit of technical
type which people have made and use (for
instance a missile guidance system, or a ther-
epistemic structuring
mostat), or a natural (mostly biological) unit
which one observes. An objective question is
for instance, how a visual unit emerges from
the different excitatory states of retinal recep-
tors (Foerster & Glasersfeld 1999, pp. 111–
117). But how can one understand oneself as
a functional system with cybernetic proper-
ties? Am I a system? (The problem is that one
is “one” to start with; see §56 ff below.) To ask
that is a bit confusing (how can one become
an object?) and therefore it is often skipped,
perhaps implying that there is no such ques-
tion (as in “don’t look now, maybe it will go
away”). But there is.
15 In some areas one can see mechanisms
of this type both objectively, from the outside,
and subjectively, from the inside. The herd
instinct serves animals to decide on behavior
choices (there may be a lead animal that for
instance decides whether to fight or flee).
Zebras attacked by lions tend to flee, in which
case one or several of them may get killed but
the rest of the herd is saved; this can be under-
stood as an inherited behavior pattern with a
built-in target of survival for the group. But
herd reactions can be manipulated for
instance by hunters who drive a herd of buf-
faloes over a cliff, and most of them perish.
16 At times one is personally in situations
of this type (intellectual, religious, or political
fashions, garment fashions, investment fash-
ions), and the result may be beneficial, neu-
tral, or detrimental. The aim here is mostly to
avoid mistakes of random behavior, due to a
need for company, or to be “with it,” on the
assumption that the group cannot be entirely
wrong. But in either case one should not
abandon personal responsibility to judge and
decide, by relying on input from what others
do, or from what some leader says one should
do.
17 One re-considers in taking a distance
from what goes on and judging it objectively,
that is as-if it had a built-in structure and
meaning, not dependent of any one individ-
ual subject or group. This is often a helpful
procedure, because one can then re-assess the
situation, taking a fresh look, and perhaps re-
structure it, and also avoid being overtaken by
what happens. Re-setting effects can also be
obtained by other methods, like meditation,
and on a more directly physiological basis, by
yawning or sleep. Such procedures are similar
to re-starting a computer which has devel-
oped program entanglement.
Constructivist Foundations
 philosophical-epistemological
CONCEPTS
Second part:
Tool problems
Subjects, objects, Descartes, leaps,
and Achilles
18 But the distancing (objectivation) proce-
dure is commonly over-interpreted beyond
the distancing, in the sense that the structure
(thing, process, or situation) in-itself has a
mind-independent or even absolute (ontolog-
ical = metaphysical) structure, existence, and
significance. This notion tends to promote a
belief in a primary or ontological split between
subject and object, as it has for instance been
formulated by Descartes. His procedure
(Augustinus’ dubito ergo sum → cogito ergo
sum) has had a peculiar effect: his starting
point was subjectivity (“res cogitans,” perhaps
implying that Descartes saw himself as a
thinking object, a sort of internalized outside
agency). He posited this subjectivity as a
doubt-free basis of certainty (or stability in the
cybernetic sense) for all knowledge, but the res
cogitans remained so fuzzy that for instance
Hume and Kant questioned its validity.
19 Perhaps as a response to this problem of
fuzziness, there are debates like whether the
self is “diachronic” (in terms of a self-pro-
duced “narrative,” that one posits as one’s
essence) or “episodic” in time (involving
memory to a lesser extent) (Strawson 2004).
Others, like Crick (1994), went further, deny-
ing the self altogether; his former collaborator
Koch (2004) has recently mitigated this opin-
ion by saying (in agreement with Searle 2004,
cf. §22 ff) that though “consciousness is part
of the ordinary physical world,” yet “it is not
ontologically reducible to … brain processes”
and that, to resolve this paradox, “an ultimate
theory of consciousness” is needed. The fuzz-
iness endures, evidently.
20 In contrast, Descartes’ second notion,
of objective outside reality (“res extensa”),
remained mostly unchallenged, although it is
no less nebulous. It is commonly assumed to
be the basis for a postulated mind-indepen-
dently pre-structured reality (MIR, tradi-
tional metaphysics-ontology) – and also often
seen as basis for all knowledge – to which not
only philosophers (since at least Parmenides)
but also many scientists still subscribe, often
only implicitly. Consider a simple example:
you look at a roof and say it is red, although it
appears black to everyone (i.e., at night). You
2005, vol. 1, no. 1
mean that it will look red in the daytime, or
with artificial light, which in an ontological
leap of faith you posit to be its real MIR-state.
21 The use of ontology (the leap to MIR-
belief) continues to be the conceptual Achilles
heel of science and philosophy. Realists, and
“hard” scientists in particular, often deny that
they use metaphysics, and may be offended
when it is pointed out that they do (some even
become unfriendly). In recent years though
they commonly use the word “ontology” as if
that were the most self-evident term to
employ, perhaps unaware that it is a branch of
metaphysics. Conventional metaphysics thus
has a comeback in unexpected places. The
problem here is that an ontologist is someone
who is able to find a non-existent black cat (or
even two of them) in a room without light (cf.
§69).
Ontology according to Searle
22 The recent book called “Mind” by Searle
(2004) is a clearly written exposure of a tradi-
tional ontological view of the philosophy of
mind. This text can help for a comparison and
contrast of realism with the 0-D view, which
is important to do.
23 Searle offers “the solution to the mind-
brain problem” (pp. 111 f.), saying he tries to
“just state the facts,” namely that subjective
feelings (though they possess “subjective or
first-person ontology”) are “part of the real
world,” by which he means “facts” (which are
“third-person ontology”). He calls this his
“basic ontology” (p. 133). The connection
between his first- and third-person ontologies
did not become clear to me from his descrip-
tion; the proposed “solution” appears to be
yet another re-wording of the mind-brain
problem in the non-functional MIR-view; all
his argumentation in this book is MIR-belief-
based.
24 He says (pp. 13, 131 f.) that Descartes’
dualism is wrong, but his main objection is
apparently not the dualism per se, but that, as
a consequence of his dualism, Descartes was
skeptical about the possibility of knowing
reality. I would think that doubt (skepticism)
is an essential part of thinking and knowing;
certainty invites problems.
25 And also, Searle’s use of traditional
ontology is not possible without dualism
because it refers to MIR, which pre-supposes
S–O dualism. He uses the terms “facts” or “the
world” in the sense of traditional realism,
epistemic structuring
implying the ontological S–O split. He uses
the term “ontology” – without questioning it,
as an aspect of “reality” (see §20 and §69).
That means that he believes in MIR (in fact,
he calls it “observer-independent phenom-
ena” (p.7), a somewhat puzzling term, since in
the philosophical meaning, phenomena are
descriptions of the “observer’s” experience,
and thus cannot be “observer-independent”).
Thus he practices dualism despite his state-
ment that it is wrong.
26 He writes (p. 2) that, for talking about
mind and consciousness, he has eliminated a
great number of “isms,” including material-
ism (pp. 84–106), as possibilities for a con-
ceptual framework. (But he does not mention
constructivism; considering it might have
changed his conclusions). Instead he defends
naïve realism (pp. 274–277), not by proving
it, but by pointing to the “unintelligibility of
its denial.” He calls his own realistic position
“biological naturalism” (pp. 113–115). Let
me add that realism cannot be proven because
the knowledge of metaphysical entities which
realism implies is impossible. The problem
here is that metaphysical realism is not intel-
ligible (despite the fact that the MIR-shortcut
cf. (§55) is common practice in daily life and
in science). Traditional MIR-belief results
from misinterpretation of the reliability of
many word-concepts as referring to extra-
mental entities. But there is no need to deny
realism so much as to change its understand-
ing: to transform MIR to working-MIR (or
realism to as-if realism); and this is quite intel-
ligible.
27 Searle makes the important point
(p.202) that science is defined as a method or
procedure rather than by ontology, with
which I agree entirely. Now if he could elimi-
nate MIR-ontology (traditional metaphysics)
altogether – it is superfluous and causes prob-
lems. Thinking involves procedures, use of
mental tools, including temporary pragmatic
stabilizers as needed. That is to say, ontology-
metaphysics should be seen as working- (or
as-if-) ontology-metaphysics. He has also
shown (as others have) that machines do not
(have to) understand symbols they deal with
(his Chinese room argument, pp. 89 ff), and
again I quite agree. We, not machines, are the
ones that understand symbols, including the
metaphysical ones. Jaspers called them
“ciphers”; we create and use them as tools:
guideposts and stabilizers.
37
 philosophical-epistemological
CONCEPTS
28 He finishes the book by saying (p. 304)
that “we do not live in … two worlds, a mental
world and a physical world … Rather, there is
just one world; it is the world we all live in, and
we need to account for how we exist as a part
of it.” I agree, with one qualification: the
world we live in is not discovered by us in a
pre-fabricated state, but we structure it (cf.
§50). Our possibilities of structuring are lim-
ited by the viability of the structures as deter-
mined when trying them out, that is, by feed-
back-in-experience, and in some areas the
choices are very limited (like for persons, or
gravity, or counting). Structuring mostly
does not imply “make up” or “build physi-
cally,” but it does always mean that all mind-
and-world structures are the result of individ-
ual and collective structuring and trust,
within (“given”) unstructured and undivided
experience (cf. also Wood C12, §35).
MIR, existence, falsification,
and de-construction
29 A start-point for discussion of the MIR-
tool can be Kant’s observation that phenom-
ena – rather than MIR itself (noumena or
onta) – are all one can ever know. This prob-
lem was also the start-point for the phenom-
enological-existential philosophers since
Hegel. But for stability reasons, and chiefly
because they did not renounce the ontological
subject-object split explicitly, they had a dis-
concerting tendency to fall back onto MIR-
ontology, which they paradoxically implied
could be done on a phenomenological basis
(they missed the point that phenomena are
constructed). Heidegger even wanted to write
a “fundamental ontology.” Logical and other
positivists, despite their anti-metaphysical
start, had an even quicker relapse into MIR-
metaphysics (if they had ever left it). Realists
simply disregard Kant’s finding (and the les-
son from Plato’s cave parable) until now and
stick with MIR-belief; this can lead to notions
like Minkowski’s 4-dimensional MIR-“block-
universe” and pre-destination, a particularly
striking side-effect of thought-inversion. This
paradox vanishes with change from MIR to
working-MIR (see §62 ff).
30 Constructivism is an attempt to
improve on phenomenology in overcoming
this problem, but the possibility of MIR-
relapse remains here as well.
31 All ontological truths are subject to fal-
sification or de-construction qua onta, by
38
showing that they are posited, i.e., they are not
truths in the ontological (MIR-)meaning
(Popper’s falsification and impossibility of
ontological verification immediately follow
from this). They can, however, be trans-
formed into working-structures; then they
have a pragmatic and limited viability. Some
of the more recent changes in world-view are
related to giving up unwarranted assump-
tions. For instance the beliefs in MIR-absolute
space and time, which in relativity theory
were, because they had no operational mean-
ing, replaced by a working assumption of con-
stant speed of light independent of the speed
of the emitting body but not of that of the
observer (though speed of light is often inter-
preted as MIReal as well).
32 I agree with Glasersfeld (C8, §2) that
“time” too is a created structure (see also §75).
Wood (C12, §9, §12) says that “time is more
than a mental form” because it is used by oth-
ers; non sequitur: time is a collectively used
construct.
33 On the other hand, Kant’s point (cf.
§29) is compatible with the constructivist
assertion that we build structures (for self
and world and all) inside our non-structured
encompassing experience: the created struc-
tures (and nothing further) are the phenom-
ena we can know, and use as working tools.
34 Some structures, which are found (or
wished) to be particularly reliable, are
invested with trust and may be called real or
true, and then they may in turn be used as
sources of strength. For instance the moon is
real because one expects it to be visible the
next time one looks, and when it is not, one
has some explanations: clouds obstructing
the view, the moon descending below the
horizon because it circles the earth, etc. Other
times had other explanations, for instance in
old Egypt an ad-hoc goddess (Nut) swal-
lowed the sun in the evening and gave birth
to it in the morning. And, due to their reli-
ability and importance, sun and moon have
become gods, or close to that, in some cul-
tures.
35 But that some-one or some-thing
“exists” means only that he, she, it “stands
out.” It does not say why it stands out, nor
whether or not it is mind-independent, or
guaranteed in some way (cf. §42). For
instance, existence can simply be posited
pragmatically, as working-existence (as 0-D
proposes, cf. §62 ff).
epistemic structuring
The problems of qualia and ineffables
36 Nor does this imply a different status, for
instance, for qualia, as compared to other real-
ities, such as – preferably physical – objects
(which are in MIR-belief thought to be the real
reality), shapes or numbers. Qualia present
conceptual problems in an MIR-framework,
because they cannot easily be MIRs. Realists
(naturalists) discuss for instance whether qua-
lia like color, smell, or pain “exist”; but this
question has meaning only in a traditional
MIR-view; in 0-D, qualia-creation requires
the subjects’ activity, as do all other mind-and-
nature-and-all structures.
37 Wood (C12, §13) writes that pain dis-
appears with anaesthesia but the fear of its
return does not; I agree, these are different
questions.
38 For structuring and stability thinking
needs overall encompassing structures. How-
ever, these are not provided by ordinary expe-
rience, including perception, but rather by
extrapolation from scientific theories, by
intuition, or by religious revelations.
39 We are inside the encompassing (expe-
rience or structures), and thus these struc-
tures cannot be defined inside experience;
they are what is often called “ineffable.” If one
tries to comprehend such (non-)entities (cf.
§56) as-if they were circumscribed MIR-enti-
ties, one faces the conceptual paradoxes of
theism and of various other types of MIR-
concepts-of-everything. Though they cannot
become circumscribed entities, they can have
names (cf. §63), which makes this more puz-
zling (but “nothing” also has a name).
40 Most of us are born into one or another
tradition of the many practiced religious or
semi-religious creeds; thus for most people
the question of choosing between them does
not come up. Still, because they are arbitrary
in the sense that there are so many competing
offers for help with the need for structure,
they need and receive social enforcement for
persistence and effectiveness. In theistic reli-
gion the emphasis is on establishing the belief
with help of an ontological leap to MIReality,
and thereafter its reinforcement by indoctri-
nation, preaching, rituals, and punishment
for disagreement. In any of these views, strong
beliefs tend to cause conflict with experiences
that cannot be understood within them.
41 In science and philosophy, discussions
can in contrast result either in strengthening
or in weakening of the discussed structures.
Constructivist Foundations
 philosophical-epistemological
CONCEPTS
The source of the MIR-problem
42 The MIR-problem of existence stems from
(a) the explicit or implied non-functional
belief in an ontological or primary subject-
object split, which is a pre-supposition for
MIR-belief (since there can be no mind-inde-
pendence without it). But the S–O split itself is
often not seen as a problem and therefore not
discussed; but precisely this causes a problem,
because the ontological leap (cf. §20) then
goes unnoticed and is naïvely taken for
granted. And (b) the belief in the ontological
S–O split implies the notion, also often not
even mentioned, that word-concepts, or even
gestalt-formations alone, either have or “refer
to” a mind-independent already-structured
existence of some type (cf. Glasersfeld 2004).
43 Concerning this point, Glasersfeld (C8 ,
§1) comments that “ … relinquishing the two
notions [subject and object] does not get rid of
a S–O split. In my view, words do have mean-
ing, but what they denote are not objects of
reality but pieces of the language user’s experi-
ence… I would say one can confront only
items that one considers separate from one-
self.” – I agree; one cannot confront experience
as a whole because one is inside it. “Self ” is a
(pragmatic) structure inside experience like
all others, and so is the S–O split; experience is
encompassing but “the self ” as a created struc-
ture is not (cf. §56). The self can confront
objects as “other,” but cannot confront the
encompassing experience (see §39, §63, §65,
§87), inside which the self is created. This
point is tricky, because one can “confront” the
created word-concepts, such as “encompass-
ing,” almost like objects, even when one can-
not confront the experience itself. That can
mislead into the belief that the experience
itself as well can become a circumscribed
object (similar to “nothingness,” or “infinite,”
or “God,” about which much has been said and
written, although the contents of these con-
cepts are “ineffable”). Neglecting the distinc-
tion between self and encompassing can also
lead to notions of omnipotence. Aside from
this, language cannot be the “house of being”
(Heidegger) or, as Glasersfeld has also empha-
sized, the “cause of consciousness” (Mat-
urana), despite its central role as a human tool.
The problem of stability
versus existence
44 Since one can be certain only to a limited
degree about the reliability of structures in
2005, vol. 1, no. 1
experience, word-concepts (or their postu-
lated fictitious referents) may be wished to be
more reliable (real or true). One may want
mental structures to be guaranteed by a
mind-external agency like God or Nature,
although both of these are actually unifying
extrapolations from structures created in
experience. You feel more comfortable deal-
ing with a pre-fabricated outside world than
with only the word-concepts you trust – but
the outside world is itself a word-concept,
made “real” via trust. This step to MIR-belief
is often but not always supported by feedback
experience (including feedback from group
thinking and behavior) while using it –
though usefulness and reliability do not mean
MIR-existence.
45 Because the S–O split is secondary (or
pragmatic), there can be neither MIR-ontol-
ogy nor subjective or solipsistic ontology (cf.
§23). This may sound simple but is difficult to
maintain in practice; there is a strong ten-
dency to fall back into MIR-thinking, because
it feels safer, and is easier to do.
46 Sufficient stability is needed for suc-
cessful function, but it does not require MIR-
ontology. The ontological leap (cf. §20) pro-
vides stability (particularly when the ontology
is viewed as absolutely valid), but also causes
problems, such as rigidity or fanaticism. One
can instead use working- (or as-if-) ontology,
where ontological (= metaphysical) entities
are not assumed to be absolutes, but posited
as working-tools. Such as-if-metaphysical
tools work. Although there is no such thing as
a banana-in-itself, one can use that concept
for practical purposes, as-if it existed. For
instance in advertising, as in “The Delicious
Popocatepetl Banana.” (I mentioned to the
chief of a marketing firm that – as every reader
who has followed me up to here will surely
agree – she uses metaphysical methods, but I
received the answer that I must be on drugs. I
swear I was not – any more than I should be.
This left me with the impression that I had not
achieved an instant market breakthrough; but
mind you, I did not get a chance to explain
working-metaphysics to her.)
47 Dykstra (C32) emphasises, as a guide-
line to education, the ability of all people to
create structures. The question of the actual
effects of various education-methods may
become central for future planning. 0-D ver-
sus MIR-belief is a question of whether reality
is to be found, as MIR-belief assumes, or con-
epistemic structuring
structed, as in 0-D (C37; the practical ques-
tion is further discussed in R9 and R12 to
Dykstra).
48 In McCarthy’s opinion (C2) equating
consciousness with the encompassing makes
knowledge merely subjective (also in C17,
§12). This statement is misleading; encom-
passing experience, and its knowledge func-
tion, are subject-inclusive, but not solipsistic.
49 McCarthy (C5, §3) also expresses his
uneasiness with the concept of “trust”: “I am
not happy with the word trust … I would seek
a greater certitude than this implies.” I think
to be absolutely certain is not a recommend-
able wish. It is dangerous, and it can be
desired only if you postulate MIR; to be
clearer for this particular aspect, one could
perhaps call it mind-exclusive reality. McCar-
thy suggests that some phenomena “are spe-
cial” (non-mind-independent reality, nMIR),
as suggested in Goethe’s Anschauende Urteils-
kraft (or I would add, Husserl’s Wesens-
Schau) – but this does not change the con-
struction aspect, and Husserl fell back into
MIR. I don’t believe that nMIR can help here.
McCarthy (C5, §3) says that thinking is not
subjective nor objective but “hyperjective” –
this term may, or may not, be compatible with
the notion of the pragmatic (rather than
ontological) S–O split.
50 The answer to the question “How do we
move from 0-D to … mind-inclusive experi-
ence?” (McCarthy C17, §4) must be: by struc-
turing self-world-and-all within the other-
wise unstructured. But McCarthy’s further
questions like “How does knowledge arise?
What is the criterion of epistemic closure?
What sort of feedback do we need? … the con-
tent confronting it?” (C17, §5) suggest MIR-
belief on his part. “We have to see the neces-
sary postulate at the theoretical level by a pure
act from the depths of our own selves (“This”
would make closure) and find its equivalent
in our actual experience.” (C17, §6) – In my
opinion things are less complicated: we create
structures, and accept them if they fit well,
some of them with a compelling feature of
closure, but this closure does not imply an
equivalent everywhere. – That my definition
of consciousness “locks knowledge into sub-
jectivity” (C2, §1) etc., corresponds to a fre-
quent objection to constructivism; it neglects
the encompassing quality of consciousness,
within which both subject and object are
structured, with no ontological S–O split.
39
 philosophical-epistemological
CONCEPTS
51 Another expression of the same type of
misunderstanding is Wood’s opinion (C12,
§30) that “… experience [is] … not only
encompassing but also encompassed by real
possibilities”: but reality is encompassed by
experience, and not vice versa as implied in
Wood’s MIR-view.
52 The constructed as-if-MIR can be
understood in analogy to some mathematical
concepts where this feature is more plainly
seen: the infinite, or imaginary numbers,
which work although they are obviously
man-made (“fictitious”) and are not usually
said to exist mind-independently. To claim
that they are would mean trying to turn math-
ematics into traditional metaphysics; the
Pythagoreans had already tried this, without
success. The created mathematical tool-enti-
ties are not “wrong” or “impossible” (includ-
ing for instance imaginary numbers) if one
knows their functions (cf. also Lakoff &
Núñez 2000). Likewise, the working-ontol-
ogy tools are working-structures, they are nei-
ther wrong nor impossible; it is their tradi-
tional MIR-equivalents that are wrong and
impossible.
Strategy-feedback for problems
53 In working-metaphysics, ontological enti-
ties become working-structures that are
applied to unstructured experience, some-
what like (a grid of) markers to uncharted ter-
ritory. They can be known only as working-
propositions (mental tools). Therefore the
earlier naïve-ontological view of the subject-
object split is transformed into a pragmatic
one, in which working-structures are cor-
rected by feedback during use (one might call
it feedback-ontology). Ongoing experience
(the mind, consciousness) is the only avail-
able start-point for any conceptual structure
(theory) that would deal with the mind-brain
relation.
54 The term “feedback” is used here in a
more general sense than usual. It goes beyond
quantitative influences on functions (change
of magnitude or direction), and on to the
question of adequacy of the utilized tools
(functional entities) themselves, or even to
changes in overall strategy. This situation is a
bit like having a flat tire: you notice (“feed-
back”) that the car malfunctions, and to deal
with that you need to inflate or change a tire,
or perhaps a tow-truck is needed, or some
other tool or strategy. This requires the avail-
40
ability of an arsenal of tools and strategies, or
creation of new ones.
Problems with traditional realism as
shortcut for 0-D
55 The traditional realism (MIR-metaphys-
ics-ontology) can be understood as a special
instance of working-ontology, obtained from
it by using a shortcut: by shunting its origin in
structuring activity, or more commonly by
neglecting or even denying it. But although
this works for many purposes, such neglect
does not make it right, since it mistakes tem-
porary fixations, within experience, for per-
manent, or even absolute, outside realities.
This may result in difficulties when trying to
deal with problems.
Third part:
Remedy. 0-D tool
structuring within
encompassing
experience
The encompassing mind and
subject-inclusive structuring
56 Consciousness (awareness, experience,
mind) is encompassing (or perhaps more to
the point: we are in the encompassing, and we
ourselves are the encompassing; (as Jaspers
1947, p. 39, put it, “das Umgreifende, in dem
wir sind, und das wir selber sind”). One is
usually focused on some entities of mind-
and-world and not on the encompassing, but
nothing can occur outside it – notions that
there are things outside experience are extrap-
olations from experience. I suggest that, to
clarify the question of the subject, it is helpful
to distinguish between the two aspects men-
tioned by Jaspers: (a) consciousness as “the
encompassing” and (b) self (“I” or “we”),
which is the encompassed subject-structure
inside this, our own, encompassing mind.
The relation between the two aspects can
present conceptual difficulties but is funda-
mental; in mysticism and similar states they
can become the same (cf. §43 above).
57 Adams (C15, §24) writes that “0-D
apparently stands for “no explanation.” … I
am disappointed to find nothing beyond the
endorsement of social construction.” Conse-
epistemic structuring
quently, he wants to know what 0-D contrib-
utes beyond that. In my opinion, social con-
struction is only one aspect of epistemic
construction; all mental tools are con-
structed, including those which are private to
one person, though many structures have in
addition a social input. My main aim is an
access to the mind-brain question, and con-
structivism is the only one I know of so far;
since this implies changes in conceptual view,
it has general ramifications. The “no explana-
tion” statement is correct, it expresses an
expectation based on MIR-belief. Explana-
tions (C15, §5) can only be structures inside
experience; attempts to explain experience
itself imply reduction of experience to some-
thing inside experience, which is self-contra-
dictory. As Adams writes (C15, §24), 0-D is
phenomenal, descriptive.
58 I don’t say that “structured mind and
world exist,” as Adams claims (C15, §3) that I
do; rather the structures are created, and if
they are invested with trust they may or may
not be said to exist (C15, §25); Adams’ state-
ment as quoted would, in contrast, imply
MIR-belief.
59 Although we can (actually have to)
consider and handle ourselves as finite subjec-
tive entities, our mind also encompasses all of
self-and-world. In other words, the subject or
self (which we posit as our personal identity),
is structured inside experience (which one
confronts as given in unstructured condition)
as are also objects, the subject-object and sim-
ilar dichotomies, qualia, numbers, universal
and theistic word-concepts. This makes expe-
rience subject-inclusive (but not subjective-
only, i.e., not solipsistic – constructivism is
commonly mis-interpreted as solipsistic).
Structuring the self tolerates a greater degree
of arbitrariness than structuring in many
other fields of knowledge does. This has
prompted the statement that “we invent our-
selves” (Foerster & Glasersfeld 1999).
60 The encompassing aspect of conscious-
ness is an important consideration, although
Jaspers did not quite manage to leave tradi-
tional ontology behind. Since he was aware of
the impossibility as well as of the necessity of
metaphysics (which he called “ciphers” (Jas-
pers 1947, p. 1030)), he finally decided on
what he called “peri-echontology” (p.158),
which is still a kind of MIR. It seems to me
instead more helpful to deal with this paradox
in terms of “working-metaphysics” (as-if-
Constructivist Foundations
 philosophical-epistemological
CONCEPTS
MIR) as a human instrument (see Müller
2003) rather than as “being” or “peri-being.”
Working-metaphysics deals with both these
aspects: impossibility of metaphysics as
affirming MIR-existence, as well as need for a
tool of this kind (as with mathematics, cf.
§52).
61 Nixon (C33) objects that an as-if-abso-
lute is a contradiction in terms. This is true,
and actually points to the main difficulty with
using working-ontology; it frequently causes
MIR-relapse. It stems from the conflict
between wanting certainty and the insight
that none is available: that doubt is present,
and actually desirable, at all times. We may
have come to a point where we need to accept
this conflict as a fundamental condition of
thinking (cf. §46 ff), rather than comforting
ourselves with absolute beliefs of one or
another kind. A sort of consolation may be
that this kind of problem is easier to compre-
hend than the notion of absolute but inacces-
sible truth and reality in traditional meta-
physics. As I mentioned in R13, the relation
between traditional and 0-D epistemologies
is asymmetrical, since 0-D supports MIR-
belief as a makeshift procedure, but MIR does
not support 0-D.
0-D (working ontology) mid-course
correction in tool-use
62 The mind (or experience or awareness or
consciousness) is the encompassing matrix or
background or envelope, unstructured and
undivided except for our automatic and/or
deliberate structuring, for all structures (qua-
lia, self, S–O split, others, brain, world, uni-
verse, gods, etc.). They are structured inside it
as needed (working structures), not from any
somehow given ready-made structures (zero-
derivation, 0-D). Awareness of this relation
could be a start-point for an “ultimate theory
of consciousness” as requested by Koch (2004;
and §19 above).
63 Although the encompassing matrix can
be named (as a not-structured item, within
which structures are formed), it cannot itself
be encompassed, that is, it cannot become an
entity within itself (see §43). But further-
more, since the 0-D view is a negative asser-
tion, denying pre-structured entities, it can-
not be ontologically de-constructed. (It could
only be falsified by proof of MIR (= ontology-
metaphysics), which has been tried unsuc-
cessfully for a long time). However, it can serve
2005, vol. 1, no. 1
as fall-back position for other (i.e., positive)
epistemologies, and help avoid MIR-relapses;
this is a positive assertion, open to falsifica-
tion by counter-examples that cannot fall
back on 0-D (see also §22–§28 above, on
Searle’s book). The change from MIR-ontol-
ogy to working-ontology might be seen as a
sort of mid-course correction for conceptual
thinking (see also my R14, §13 ff; and discus-
sion of Pivnicki’s
ˇ comment C24 in §111 ff
below).
64 The correction implies acknowledg-
ment that the matrix (consciousness) cannot
be a part of the ordinary physical world which
is a multitude of encompassed entities (as
Searle 2004, p.17, also points out). Words like
“mind” or “consciousness” refer to the
encompassing (cf. §56 above), and therefore
to something “ineffable.” Mind (conscious-
ness) cannot be reduced to MIR-entities,
including objective processes like evolution.
That is a likely conceptual reason for Jaspers’
insistence that the essence of man cannot be
found in evolution, and also for the curious
denial of biological evolution by some reli-
gious groups (cf. Wood, C40). This was, how-
ever, not Jaspers’ opinion (see Müller R8): “Es
ist gewiss, dass der Mensch im Ganzen mit
biologischen Mitteln nicht erfassbar ist, – dass
er aber bis in alle seine Realitäten hinein
zugleich eine biologische Realität ist und biol-
ogisch, d.h., mit den Kategorien fassbar ist, in
denen alles Leben der Tiere und Pflanzen
erforscht wird.” (My translation: “It is certain
that man as a whole cannot be grasped with
biological methods, – but that he is into all his
realities at the same time a biological reality,
and can be grasped biologically, i.e., with the
categories in which all life of the life of animals
and plants is studied.”) (Jaspers 1949, p. 47).
65 The many attempts to structure the
whole of consciousness, e.g., “God,” or “All,”
or sometimes “Nature” or “Universe” (i.e., in
case they include the subject) are necessarily
limited to mystical or paradoxical beliefs,
because the encompassing cannot be encom-
passed by concepts. In contrast, exclusively-
objective theories of everything cannot work
because they exclude the mind (subject). That
does not imply that mental functions do not
depend on brain function, but rather shows
the irreversible relationship of consciousness
to the structures (e.g., brain or brain-func-
tion) that can “exist” only within it, not inde-
pendently of it.
epistemic structuring
0-D can help with some puzzles
66 A person’s experience is from the begin-
ning shaped by, among other things, input
from other people (to varying degree – autis-
tic persons use less of that, due to impaired
brain function). Thus we know others’ minds
from the outset, by empathy and with the help
of verbal communication. The so-called
“problem of other minds” is an example of an
artefact caused by the belief in pre-structured
MIR, for which only bodies or observable
behaviors exist (are real); other minds do not
qualify for MIReality.
67 There are further puzzles (as listed by
Searle 2004, pp. 9–32) that have similar char-
acteristics, and I want to suggest that a good
part of them is also gratuitous, an artefact of
the MIR-view. To the extent that they are, they
disappear in the 0-D view, which also accom-
modates much of what Searle calls “features of
consciousness” (pp. 134–145). I will briefly
mention a few of these in the following.
68 The mind-body problem. If only MIR is
real, the mind cannot be real, because the
mind cannot be mind-independent; in 0-D,
objective concepts like brain are acknowl-
edged to be tools of and within the mind. In
my opinion, this provides an intelligible and
contradiction-free connection between mind
and physical reality. See §80 ff, and for further
details, Müller (1997, 2000, 2001).
69 The problems of doubt or skepticism
about knowledge of the external world, and the
analysis of perception (a “disaster in the his-
tory of philosophy,” according to Searle 2004,
p. 23): in my opinion skepticism (or doubt) is
an essential feature of thinking and knowl-
edge; naïve as well as explicit MIR-beliefs are
dead ends, certainty is foolish. Or, in case you
like the term “disaster in the history of philos-
ophy”: it is rather that for over 2500 years phi-
losophy has not really ceased to chase and find
the non-existent ontological black cat (MIR)
(cf. §21), despite a few false starts of efforts to
do without it. In fact, recent writings are often
less restrained and more naïve about that kind
of endeavour than the critical philosophy of
the 18th and 19th centuries.
70 One might add here the Question of
perception (Searle 2004, pp. 259 ff). Do you
know what you see, or do you see what you
know, both, or neither? For instance in TV car
commercials the wheels often turn backwards
while the car moves forward – which way does
naïve realism (p. 274) turn? In 0-D one would
41
 philosophical-epistemological
CONCEPTS
say that the immediate perception (turning
backward) is a non-functioning structure
which is quickly invalidated via feedback
from memory of other experience, to the
effect that wheels turn forward when cars
move forward, and that the immediate naïve
perception is the outcome of a technical arte-
fact. Objects and functions are entities struc-
tured within the encompassing mind-and-
nature experience, they are not MIR.
71 The qualia problem (cf. §36). In case
only supposedly pre-structured MIR-objects
are believed to be real, qualia may not be con-
sidered real. In 0-D there is no such question;
real are structures that one invests with trust
while using them.
72 The problem of free will. This too is a
problem only in the inverted reasoning of
MIR- and mechanical-cause-primacy. In 0-D
we are free to structure our experience and to
act within the trusted reality, though within
limits that can differ greatly between various
fields of experience, as determined by feed-
back experience when using the structures. If
one reasons within exclusive objectivity, the
subject is omitted; but objectivity itself is a
specialized view within subject-inclusive
experience, which acts as-if there were no
subject. (The term “as-if ” is here used in a
wider sense than by Vaihinger.) Causes and
logic are mind-and-world sequences.
73 The problem of self and personal identity
(see the mind-body problem §68 and §80 ff).
74 Do animals have minds? This often
involves theological arguments about the soul.
In 0-D, the main difference between higher
animals and humans is the availability and use
of human language, which greatly extends the
possibilities of gestalt-function and action.
75 The problems of sleep, unconsciousness,
the unconscious. MIR-belief would require
persistence of consciousness (similar to per-
sistence of the moon) if it were to be MIR-real.
For 0-D in contrast this is not a relevant con-
ceptual question, because the self-agency
structure does not per se imply an assurance
or requirement of uninterrupted presence.
(“Time” of course is a construct like other
structures; it structures the flow aspect of
experience, cf. §31.)
76 The problem of intentionality. This is
only a problem if MIR is assumed; it disap-
pears if reality is structured in experience,
since we mostly know our structures and do
not intend something outside experience.
42
77 Mental causation and epiphenomenal-
ism. Another typical MIR-belief problem; if,
as in 0-D, the subject is a part of all reality
there is in principle no such difficulty. MIR-
causality is a pragmatic assumption, since it is
helpful to regard physical events like that; it is
economical to omit the subject in practice
(but not in principle).
78 Psychological and social causation.
Idem; this is a question of individual and
social experience and initiative. (Most of
these problems are traps in Rilke’s meaning,
in the sense of the intial quotation from his 8th
elegy.)
In summary
79 Although conscious experience can only
take place on the basis of brain function (e.g.,
Crick’s (1994) “neural correlates of con-
sciousness”), brain function is not MIR. The
only access to (= knowledge of) brain func-
tion is through the encompassing conscious
experience, inside which structures are cre-
ated, from no given referents (0-D). The expe-
rience-matrix has to be the start point for
“theories of consciousness” like for every-
thing else. That would imply circular reason-
ing only if consciousness were seen as an MIR-
object (as in “levels of consciousness” for
instance), but not consciousness as encom-
passing, experience-as-a-whole, inside of
which objective events take place.
Fourth part:
The kybernetes
The real self as helmsman
80 But now really: is the mind real or not?
Briefly, the mind (experience, consciousness)
is the only available access to reality, which is
constituted within it, and this includes the
structured self. The self is an item (agency)
structured inside experience which itself is
ineffable (it cannot be defined, cf. §39),
although it can be named.
81 But in case reality is defined as mind-
independent, the mind cannot be real,
because evidently the mind cannot be mind-
independent. This may sound like a joke, but
that question has caused much conceptual
confusion, and will continue to do so until
reality is understood in a more appropriate
manner (§44 ff).
epistemic structuring
82 Damasio (1994) thinks Descartes erred
in proposing an ontological separation of
mind and body; he proposed that the mind
must be seen as related to the whole organism.
That mind and brain should be studied
jointly is indeed an important point. But the
concept of “embodied mind,” which Damasio
proposed to counteract the Cartesian S–O
split, is not helpful; it is a pseudo-solution, an
attempt to provide a verbal answer to a con-
ceptual problem that is largely caused by
words to start with.
83 The notion of embodied mind has in
recent years been used by a variety of authors,
but it is a conceptual non-starter, because it
implies that for instance the mind = the brain,
and vice versa. This is not possible, because
one can have a mind without knowing any-
thing about the brain, a dead brain has no
mind, etc. And this self-contradictory idea is
also expressed by other opaque mantra-
like3formulae like “the mind–brain.” The
cause and/or result of using such terms are
attempts to objectify the subject; in other
words, instead of overcoming the S–O split,
one tries to shift the subject from Descartes’
subject pole to the object pole. (That is, to
convert the subject into an object, partly in
efforts to make the study of consciousness
“scientific” – this term can also become a
mantra, especially if it is not defined).
84 Efforts of this type still result in a great
deal of literature; they also imply in effect that
there is no subject. But instead one can see
both S and O as pragmatic and heuristic
structures, created simultaneously, within an
originally undivided (and also generally
unstructured) ongoing experience. The brain
as well as the self are structures within the
matrix of experience (the mind, awareness,
consciousness), which is originally neither
structured nor even sub-divided, and thus the
brain is in the mind, not vice versa. The S–O
split as well is in that case a pragmatic (sec-
ondary, and not ontological or primary)
structure.
85 Mantras or incantations are verbal
methods used for the purpose of reaching
sources of power. They have for instance reli-
gious uses like in attaining a personal medita-
tion state (as in “om mani padme hum”), or to
invoke the help of, or to unite with, a posited
outside divine source. For such aims, they can
be helpful. But they ought to be distinguished
from specific conceptual tools, used for more
Constructivist Foundations
 philosophical-epistemological
CONCEPTS
circumscribed tasks. For instance, “science”
refers to methods of study of specific ques-
tions that use reliable tools (including con-
cepts), that produce replicable results in the
statistical sense (see Müller 2001, §8c); the
term should not be used as a mantra, for
instance to invoke the authority of science, as
sometimes happens (see also §92 and §105
ff). Terms like “embodied mind” and “mind-
brain” have no procedural meaning, save as
incantations, conjuring a solution to the
mind-brain relation problem.
86 If one wants to use a conceptual frame-
work (theory) to address the mind-brain rela-
tionship, the S–O question (see §18 to §42)
must be addressed. Specifically, one has to
change from a non-functional belief in a pri-
mary ontological S–O split – whether naïve or
explicit – to a secondary pragmatic one
(inside undivided and unstructured experi-
ence). The repudiation of the ontological S–O
split, and of traditional ontology in general, in
favor of a change to working-ontology and a
pragmatic S–O split, is a prerequisite for a
“theory of the mind.” And further, since the
mind-brain relationship is a rather funda-
mental conceptual problem, it can be consid-
ered a test question for the overall usefulness
of epistemologies.
87 Glasersfeld (C8, §3) writes: “What
would theories of consciousness be if con-
sciousness is not able to reflect back upon
itself as though it were an object?” I suggest
that the “self,” a structure within conscious-
ness, can to some extent be constructed, but
that consciousness, although it can be named,
is not an object or entity (it is “ineffable”),
because of its encompassing quality. The
question is similar to: what are the infinite, the
all, God? The encompassing quality is funda-
mental, but in object-oriented studies it is
inevitably neglected. Theories try to provide
structures for enquiry, but they too have to be
based in unstructured experience, and must
take account of it, if they want to avoid blind
alleys (as they are encountered in objective
mind-exclusive theories of everything). This
is a likely reason for the invention of the term
“transcendence”: these meanings transcend
all definable (circumscribed) items (cf. §43).
88 McCarthy (C17, §10) suggests that “…
with an S–O split, consciousness begins …
therefore consciousness cannot be identical
to experience”: that is a question of definition,
I would think. To postulate a discontinuity at
2005, vol. 1, no. 1
this point implies a primary (MIR-ontologi-
cal) S–O split. S–O structuring happens
within experience, and in addition some peo-
ple claim they can revert to an empty state of
consciousness (nirvana), that is to before the
S–O split.
89 Reality-fixation by trust applies not
only to objects but also to the subject, and as
well to religious and similar structures. There
is in principle no reason why the subject (the
self-agency tool) should be less real than other
structures (such as object-tools), although
the self can be weak for people with poor self-
esteem, or those indoctrinated into submis-
sion. In other words, although consciousness
cannot encompass itself, it can encompass a
self-agent (or subject) that becomes struc-
tured inside consciousness. (Whether it sees
itself as more diachronic or as more episodic
in Strawson’s (2004, §19 above) meaning is
secondary to the agency question per se.)
90 The helmsman is the self, once it (she,
he) has become sufficiently structured and
vested with trust. It is not functional if it is
weak or not structured at all, and then the
steering goes back to other agencies by
default: to other selves (like parents, partners,
community agents, religious or political lead-
ers, or also to manipulators of various kinds)
and other posited authorities, for instance
religious or political entities and texts. And
always, objective biological pre-self “systems”
are at work, either simultaneously with the
self-agent and other-person-agents, or some-
times more autonomously, as in some new-
born animals (newly hatched fish are on their
own from the start). There is a need for coor-
dination between the various agencies, and in
the mature person this is largely a task of the
pre-frontal cortex together with the left hemi-
sphere language functions. Mental as well as
physical tools are extensions of the self.
On the use of tools with ascribed
autonomy and authority
91 Over the millennia, mental tools were
developed, such as perceptual gestalt forma-
tions (as animals already have), and in addi-
tion specifically human tools like words, to
elaborate and fortify these formations, in
word-concepts, ideas, and numbers. The lat-
ter start out as words for counting activity and
its results; machines like computers have been
added. Some of the tools have turned out to
be useful, both for individuals doing things,
epistemic structuring
and for communication, with the result that
much of experience has a communal aspect.
In addition, there were words for events that
were difficult to deal with, they were called
gods, or forces of nature. Early on, people saw
themselves as being directly influenced by
such forces, for instance in the Iliad. The sense
of self-agency and communal action devel-
oped gradually, and they still develop individ-
ually and world-wide; many relapses occur
into earlier ways of thinking, with abandon-
ment of responsibility to fictional outside
authorities.
92 People are usually aware that their
powers to influence events are limited, but
may try to join forces with more powerful
MIR-entities. But they have themselves struc-
tured these entities: They believe in human
authorities (leaders), supernatural human-
like entities (gods), or in word-concepts
(ideas). The Pythagoreans did this, they wor-
shipped numbers, or also Marx, who used the
science mantra3 for reasons of authority: he
claimed that his Hegelian ideas must not be
doubted because they were no longer philo-
sophical speculation but science, and what
could be better than a scientifically proven
political dogma and system? All you need is an
interpretation of a text. In these instances,
people put their own creations in charge of
them. The same happens when “nature” is
ascribed the status of mind-independent real-
ity (nature tells scientists what is real, they
may insist). People using word-concepts are
the precursors of the cyborgs and of users of
complex machinery, but there is no need to
think that tools of either type inevitably
become our masters (§95 to §104 below).
Complexity and your certainty:
stability versus absence of doubt
93 There is a constant alternation between
having control and relinquishing it to other
agents. This varies a great deal, not only
between persons, but also for one person
between situations. If you take an airplane, you
transfer control to a large system of people and
machines of which you suppose that they func-
tion adequately, while you relax and have a
drink (thinking about the many uncertainties
might turn you into a non-flyer, even in the
absence of terrorists). Or in adhering to a reli-
gious, political, or scientific opinion-set people
may trust that they base themselves on a
doubt-free given outside MIReality (§24, §69).
43
 philosophical-epistemological
CONCEPTS
Fanaticism is often used to compensate (and
over-compensate) for confused thinking and
other personal weakness. And some – your fel-
low humans, including your employees, like
politicians, kings, and clerics – may abandon
thinking to belief in authority, proclaiming
that they possess the truth, and that doubt can
be dispensed with – you may suffer the conse-
quences. (The story of the man-made decline
of Easter Island by deforestation, related to
excessive ancestor worship of the inhabitants,
is relevant here; cf. Wright (2004), who also
points out that the several thousand years of
civilization amount to a tiny proportion of the
millions of years of use of physical tools.)
Your levels of biological regulation
94 One should remember that something
similar happens within oneself. There are
many self-regulating systems in the body:
sub-cortical steering of heart function, blood
pressure, metabolic and hormonal require-
ments, and also at the cellular level, etc., all
with set targets, and feedback and stability
mechanisms. They may be more, or less, ade-
quate, and they have to be coordinated with
each other too. The overall steering agency
shifts to some extent during life, from mainly
sub-cortical control early on, to an additional
super-ordinated pre-frontal control later.
And in humans the latter is closely related to
the left hemisphere language function, which
among other things helps with social interac-
tion and coordination. This is clearly a very
complex task, more so for humans than for
animals, due to the greatly increased social
and intellectual possibilities. Much of psychi-
atric disability turns out to involve impair-
ment of pre-frontal function. Being in charge
requires using the available functions effec-
tively.
Resuming steering (with our tools)
Und wir: Zuschauer, immer, überall,
dem allen zugewandt und nie hinaus!
Uns überfüllts. Wir ordnens. Es zerfällt.
Wir ordnens wieder und zerfallen selbst.
Wer hat uns also umgedreht, daß wir,
was wir auch tun, in jener Haltung sind
von einem welcher fortgeht? …
(Vom Ende der 8. Elegie Rilkes)
95 Who has turned us around like this? The
word-conceptual tools (including religious
and scientific ones) may impose their own
44
steering functions, if we let them, and more
so if they are believed to have powers that
transcend individual and collective human
experience and control. The main difficulty is
the ascription of final authority to (postu-
lated outside) MIR-agents, rather than the
alternation per se between the various cyber-
netic systems. This authority ascription away
from the subject results in an inversion of
thinking, and can make a return to one’s own
agency and free interplay difficult.
96 Through human history, this situation
has shifted. One may suppose that when
word-concepts became tools for communi-
cation and steering, it turned out that they
and their meanings could be easily changed
and manipulated, resulting in arbitrariness
(sometimes mis-named “relativity”) of
thought and action, and thus in uncertainty
about one’s own behavior and that of others.
More certainty was needed, and belief sys-
tems for thought-and-behavior were devel-
oped, adopted and enforced. These systems
helped to improve individual and collective
stability by limiting the range of thinking and
actions that were to be expected, and to help
with this the rules were furthermore often
ascribed to postulated outside authorities
that must not be doubted.
97 The reference values are provided by a
person (shaman, guru, etc.) and/or by narra-
tives or texts that are declared to be binding;
thinking and behavior was guided by them.
Individual and social stability can to some
extent be obtained in that way, but subse-
quently the restrictions for thinking may
prove to be a handicap as well, for various
endeavors, such as science, and sometimes
for socially important questions which need
to be dealt with (such as presently genetic
research, etc). Historically, the idea of Nature
as MIR developed from that of God as MIR.
98 Chumakin (C3, C13, C21) largely
agrees with my paper, though coming from a
different point of view, and though he
appears to consider “mind” to be pre-struc-
tured (C13, §2). He suggests that “Science
will do science and constructivism will do
humanities – we do not need the 0-D
approach, as we will not have to relate mate-
rialism (realism) and idealism” (C13, §11).
Here I disagree, since this implies a split of
experience and activity into two areas that
have nothing to do with each other. 0-D, as I
propose it, is an overall view for all experi-
epistemic structuring
ence; and besides, the term idealism for 0-D
is erroneous: ideas are tool-structures, not
absolutes (cf. §101 and R2). One can have an
idea of a chair when not dealing with a chair,
but the viability of an idea is tested by trying
it out. For instance one assumes the chair to
be real when, in conformity with the idea of
a chair, it also meets criteria like solidity and
perseverance. But hat point may be close to
incomprehensible if one bases oneself on the
notion of a primary mind-independently
pre-structured (metaphysical) reality, in
which case what Chumakin calls the “blind
giant of science” (C3, §7) takes over.
99 Are God or Nature mind-independent
or not? The most common opinion in both
instances is dualistic belief in MIR (tradi-
tional ontology = metaphysics, and this
despite the denial by many who use it). The
traditional belief in mind-independent
nature has a long history too, and in Des-
cartes’ error it was only formalized; it devel-
oped from MIR-God; indeed Descartes
retained God as a guarantor for his proposal.
At times God is presented as identifying him-
self by saying “I am who I am,” which leaves
the question floating. Mystics, on the other
hand, tend to feel that something is missing,
and aim for union of self with God or the uni-
verse. And in atheistic religions the question
is also mostly seen in a unitary way.
100 Wood (C4, §0.1 ff) says that my view
is immanent rather than transcendent –
assuming that he refers to God, I agree. I
think for the reasons mentioned in §99, that
in case for religious purposes you use a belief
in God, it is more helpful not to see Him as
MIR. But because they want certainty from
outside themselves, many theists might insist
on that (this is the Vatican’s opinion, for
instance).
101 For scientific work it is a good idea to
remain aware that what we call MIR results
from investment of our own structures with
our own trust, i.e., reality-belief (as-if- or
working-MIR). This awareness can restore
the agency to the scientists. That does not
mean that they will then become idealists, or,
beyond that, even solipsists. It does mean
that objectivity is a tool rather than referring
to a mind-independent source of truth, and
that the subject is a part of all experience.
This means that exclusive (MIR-) objectivity
is an error, not that all experience is only sub-
jective (cf. §11 to §14, §56, §107).
Constructivist Foundations
 philosophical-epistemological
CONCEPTS
102 Absolute beliefs can strengthen ethi-
cal conviction, since they eliminate doubt
(and facilitate fanaticism as well). But the rec-
ognition that there are no ethics absolutes
does not imply moral arbitrariness (or “rela-
tivism”), which would mean a lack of obliga-
tory guidelines. The reference values are col-
lectively posited and made obligatory, and
although social construction is only one of the
forces at work in structuring, these days this
becomes a world-wide collective task. Tradi-
tional sacred texts are among the sources for
such guidelines.
103 As it turns out, however, humans re-
discover their leadership – actually that there
is no choice now but to reclaim it, because
absolutes are blind alleys. In a constructivist
framework, Kant’s categorical imperative (or
the Golden Rule) can provide the method for
ethical goal setting. Communal and global
agreements on such maxims become impor-
tant and difficult tasks.
104 McCarthy (C5, §17) says that in con-
structivism duty is imposed from outside – I
am not sure why he thinks so; invoking
instructions from outside contradicts the
basic tenets of constructivism.
Physiology, scientific studies of
consciousness, artificial intelligence,
and subjective experience
105 So what should we make of the numerous
attempts to reduce subjective experience to
physiological events in objective systems
(such as the activity of neuronal networks,
synapses, dendrites, intracellular tubules, 30–
40Hz electrical activity of the cortex, quan-
tum events in the neurons, etc., etc.)? (See
Hameroff (2003) for a list of candidates for
Crick’s “neural correlates of consciousness”).
106 All self-declared “scientific studies of
consciousness” that I am aware of, including
the detailed conceptual studies of Metzinger
(2003), imply the naïve ontological S–O split,
and are therefore, in my opinion, not able to
deal with the mind-brain question. Chalmers
(1995) had labeled it “the hard problem of
consciousness” – “how physical processes in
the brain give rise to conscious experience.”
But it is not just a hard problem; it is the
wrong question and cannot be solved in prin-
ciple (cf. also Müller (2001), and §85 above).
The aim of such attempts is always a so-called
“scientific” replacement of subjective experi-
ence by objective events, on the mistaken
2005, vol. 1, no. 1
assumptions (a) that mind and brain are pri-
marily (ontologically) separate, à la Des-
cartes, and (b) that science is defined by MIR-
objectivity, that is, by traditional metaphys-
ics. The subject then disappears necessarily.
107 If in contrast we see the separation as
pragmatic (secondary), as a working-propo-
sition, the subject automatically is and
remains a part of all experience. And studies
of brain activity occur in the mind, not the
mind in brain activity, whether or not it is a
topic of study. For objectivists, that may be an
unfamiliar aspect: but because one cannot
leave the bubble of encompassing experience
(consciousness), it is impossible to look at
the mind from outside. The realist’s inability
to understand this is a striking consequence
of thought-inversion.
108 Suppose that in some time from now
it will be possible to provide a complete
simultaneous demonstration of the objective
events going on during subjective experience
(such as electrical and chemical brain activity
both overall and at the cellular and molecular
level). Will these events be identical with the
subjective experience? To ask this question is
to deny it (that is, in my opinion, but as men-
tioned in §83 some insist that they are iden-
tical). The only possible relationship between
the two is that awareness of brain activity
(including one’s own) occurs inside experi-
ence, which is the only available starting
point. In 0-D, physiological knowledge can
help to understand and deal with ourselves
better: a formulation of this or similar type
could replace the intractable mind–brain
problem (cf. §68, §86, §106), an artefact of
the MIR-view.
109 Artificial intelligence research created
much discussion in the 1950s and 1960s, and
many enthusiasts are still now at work in this
field. Recently the evaluations are more
restrained, confined to specific tasks, and
some philosophers have even declared the AI
proposition to be incoherent, impossible, or
also immoral (cf. McCarthy 2004). AI efforts
are undertaken as extensions of human intel-
ligence, the aims and means of such machines
and programs are determined by people who
design and finance them.
Conclusion: Reality is labor-intensive
110 Human goal-setting, decision-making,
and responsibility are not going to be taken
over by machines or other tools any time
epistemic structuring
soon, and this is what we should concentrate
on (or else we may fall prey to electrical
power failures, to someone’s programming
errors, or to various absolute beliefs). It
implies taking on some responsibilities that
have traditionally been relegated to posited
outside authorities. This is a matter of world-
wide negotiation rather than primarily of
interpretation of sources of assumed cer-
tainty, such as religious or political texts, or
scientific knowledge, which may contradict
each other on some central questions. Reality
is a continuous task for structuring, not
something one can find ready-made.
Epilogue: Footnotes to Plato?
111 Pivnicki
ˇ (C24, §1) asks whether TA78 is
a presentation of contemporary philosophy.
No, this would require several books to do,
and could obviously be much better achieved
by a philosopher than by me. My presenta-
tion is an attempt to deal with the mind-
brain relation and related conceptual diffi-
culties, and refers to a few authors only. Can
we improve on Plato? (C24, §3) (Cicero’s
question: num eloquentia Platonem superare
possumus ?) – I am not sure that this should
be the principal goal, but Plato, like other
philosophers, has left us with a difficult prob-
lem: what is metaphysics, why is it there, and
what should we do with it?
112 One can examine the metaphysics-
question with respect to the first philosophy
(that is, metaphysics) of Plato’s student Aris-
totle; this leads to a re-wording of some of the
themes of the present paper. For Aristotle,
physics (ta physika) are by nature (te physei)
later, but for us (hemin) earlier. Therefore,
because of the way we think, Aristotle pre-
sents the origin (tas archas), namely the “first
philosophy” or metaphysics, after his writ-
ings on physics. According to the studies of
Reiner (1954; see Müller R6), the order of his
books was not accidental, as had traditionally
been thought, but followed the sequence of
thinking as Aristotle understood it. He saw
the causally first thing, our understanding, as
being accessible to us only through (and
therefore after) the causally later thing, that
is, natural entities.
113 This was, it would appear, a key for-
mulation of the inversion of thinking (in the
sense of Rilke), which has had an important
effect on the later philosophy of the Arabs,
scholastic thinking (God as first MI-cause,
45
 philosophical-epistemological
CONCEPTS
Aristotle’s unmoved mover), and modern
occidental philosophy (nature as MIR-
source). The aim was always to find a com-
plete and certain outside system to rely on
and to be governed by. Even nowadays, theo-
ries of everything (in physics and cosmol-
ogy) are often viewed as having, or represent-
ing, MIReality. They were, and still are,
ascribed this role despite their heterogeneous
and often mutually contradictory contents.
114 Aristotle was, like Plato, a dualist,
who believed in nature-in-itself; both fol-
lowed here one of the threads of the revela-
tion to Parmenides by a goddess (“it is and
cannot not be”), while neglecting another
important one (“knowing and being are the
same”), which could have allowed consider-
ation of the structuring aspect. This omission
resulted in a long-term derailment of con-
ceptual understanding (no wonder White-
head observed that all occidental philosophy
amounted to footnotes to Plato). Consider-
ing the events of the 20th century in science
and elsewhere, we cannot afford that; and at
present this causes so many problems as to
make a correction (from metaphysics to
working-metaphysics, from MIR to as-if-
MIR) mandatory.
115 In the corrected (0-D) view, one can
start with the same sequence of thinking, as
seen by Aristotle, from physics to general
understanding. Here the meaning would
simply be that we structure specifics first,
many of them with a compelling feature of
perceptual gestalt-closure; only later can we
deal with general principles.
116 The difference between 0-D and the
Aristotelian MIR-view is firstly that no longer
do we try to find pre-existing entities outside:
we create, within experience, structures that
are initially determined by more automatic
completion and stabilisation (such as percep-
tual gestalt-closure). Only then, by means of
extrapolation from gestalt-thinking, and
with the help of language (word-concepts),
can we transcend ordinary experience and
create more general structures like God and
Nature, which are also needed but require
more deliberate stabilisation (by doctrines
and theories) and often enforcement by
46
indoctrination. – And secondly, no longer do
we claim that the later general structures are
the “origin” or “cause” or “mover” of the ear-
lier specific ones, but we use them as addi-
tional tools: for the overall integration of
thinking (working-metaphysics, working-
religion). – Although arbitrariness of struc-
turing is more evident in the case of the gen-
eral structures, all mental (mind-and-
nature-and-all) structures, including the
most automatic ones, are secondary to expe-
rience, and ad-hoc in essence. The – always
implied – ontological leap (to MIR-belief) is
reversible by conceptual analysis (fall-back
onto as-if-MIR), or also by techniques like
meditation; these methods complement each
other.
117 One needs to acknowledge that meta-
physics-ontology is present in all thinking,
and that it must be addressed in some way to
get out of the conceptual impasse, which is
perhaps most directly evident in the mind-
brain question.
118 The preceding presentation is in part
a mish-mash of phenomenology, psycholog-
ical explanation, and a discussion of objective
functions like electronics or physiology. I
want to apologize for this only mildly: it can
be handled if one keeps in mind that all con-
cepts are human tools, phenomenological
concepts just as objective ones, and also phys-
ical tools with and without built-in self-guid-
ance. In case this origin is kept in mind, there
is no principal reason why, say, phenomeno-
logical and objective concepts should not be
used together, although one has to pay atten-
tion to the differences, since objectivity
implies an assumption of (as-if-)MIR, while
phenomenology does not.
119 But I do want to apologize for repeti-
tiveness in my presentation (it is done in an
effort to minimize misunderstandings); and
for any possible errors in interpretations, on
my part, of the work of the authors I have
mentioned.
120 One might compare the theme of the
8th Duinese Elegy with the one of an earlier
poem by Rilke (at age 24), from his “Stunden-
buch” (1899/1936), which shows no concern
with inversion:
epistemic structuring
Ich kreise um Gott, um den uralten Turm,
und ich kreise jahrtausendelang;
und ich weiß noch nicht: bin ich ein Falke, ein Sturm,
oder ein großer Gesang.
121 In my translation:
I circle about God, about the age-old tower,
and I circle for thousands of years,
and I know not yet: am I a falcon, a storm,
or a great song.
ABOUT THE AUTHOR
Herbert FJ Müller, born 1924 in Köln
(Cologne). After school and (1942–1945)
military service, studied medicine at Univer-
sität Köln (1945–1951, Dr.med.); further
study of psychology and philosophy (Köln
and Bonn). From 1951, medical internship
in New Jersey, postgraduate training in psy-
chiatry, neurology, and electro-encephalog-
raphy at New York University, Universidade
do Brasil (Rio de Janeiro), Düsseldorf, and
(since 1956) McGill University in Montreal
(now Associate Prof. of Psychiatry). Clinical
work at Douglas Hospital, Montreal, since
1959, with emphasis on EEG and gero-psy-
chiatry; presently still part-time. Published
over 70 papers in these fields. – Studying the
mind-brain relation, it became clear to me in
1994 that for an access to this question, the
notion of pre-structured mind-independent
reality must be abandoned. As this requires a
more general review of concept use, and of
other fields of knowledge, I started editing the
Karl Jaspers Forum http://www.kjf.ca in 1997;
my aim was to examine this question and its
implications (Seven Target Articles in KJF, many
discussions). In 1999, I became aware of rad-
ical constructivism (chiefly the work of Ernst
von Glasersfeld), which has many features in
common with my present view. A Symposium
on the mind-brain relation, in which Glasersfeld
participated, took place at the Douglas Hos-
pital, Montreal, in September 2001 (the
papers and discussions are in KJF). My present
work concerns the conceptual basis of this
point of view – which I label structuring with
zero-derivation or zero-reference – and its
relation to other views.
Constructivist Foundations
 philosophical-epistemological
CONCEPTS
Public reviewing
Note:This is the list of comments and selected
responses of the author. that were used to aug-
ment the line of argumentation of the paper.
Since not all comments were used and the
numbering should stay in sync with the original
number in the Karl Jaspers Forum, the number-
ing here is non-consecutive.
C1 van der Meijden, A. (2005) “The in-verse.”
http://www.kjf.ca/kjf/78-C1MEI.htm
C2 McCarthy, M. (2005) On the kybernetes.
http://www.kjf.ca/kjf/78-C2MCC.htm
C3 Chumakin, M. (2005) How ideal phe-
nomena are real. http://www.kjf.ca/kjf/78-
C3CHU.htm
C4 Wood, G. C. (2005) Transcendence, etc.
http://www.kjf.ca/kjf/78-C4WOO.htm
C5 McCarthy, M. (2005) Steering through the
ranges of knowledge. http://www.kjf.ca/
kjf/78-C5MCC.htm
C8 Glasersfeld, E. von (2005) Subject/object
split. http://www.kjf.ca/kjf/78-
C8GLA.htm
C12 Wood, G. C. (2005) Reality as I see it.
Comment to R3. http://www.kjf.ca/kjf/78-
C12WO.htm
C13 Chumakin, M. (2005) A realist approach
to mind/brain problem. Comment to R2.
http://www.kjf.ca/kjf/78-C13CH.htm
C15 Adams, W. A. (2005) Four premises in
search of an explanation. Comment on R3
to C8 by Glasersfeld. http://www.kjf.ca/
kjf/78-C15AD.htm
C17 McCarthy, M. (2005) Why consciousness
should not be equated with experience.
Comment to R1. http://www.kjf.ca/kjf/78-
C17MC.htm
C21 Chumakin, M., Models. Comment to
R2. http://www.kjf.ca/kjf/78-C21CH.htm
C22 van der Meijden, A. (2005) Rilke, mys-
tics, and anaesthesia. http://www.kjf.ca/
kjf/78-C22ME.htm
C24 Pivnicki,
ˇ D. (2005) Response, but no
question. http://www.kjf.ca/kjf/78-
C24PI.htm
C32 Dykstra, D. I. Jr. (2005) The problem of
incommensurability in debates on educa-
tion (and other issues). Comment to R9.
http://www.kjf.ca/kjf/78-C32DY.htm
C33 Nixon, G. (2005) 0-D and absolutes in
education. http://www.kjf.ca/kjf/78-
C33NI.htm
C37 Dykstra, D. I. Jr. (2005) Challenge of
2005, vol. 1, no. 1
incommensurability in educational set-
tings. http://www.kjf.ca/kjf/78-
C37DY.htm
C40 Wood, G. C. (2005) Biological diverticu-
lum. Comment to R15. http://www.kjf.ca/
kjf/78-C40WO.htm
R2 Müller, H. F. J. (2005) Reality and ghosts.
http://www.kjf.ca/kjf/78-r2chu.htm
R6 Müller, H. F. J. (2005) Reiner on Aristotle’s
metaphysics. http://www.kjf.ca/kjf/78-
R6PIV.htm
R8 Müller, H. F. J. (2005) 0-D for evolution
and religion (The Vatican and Jaspers).
http://www.kjf.ca/kjf/78-R8WOO.htm
R9 Müller, H. F. J. (2005) The 0-d umbrella in
education (Absolutes if necessary but not
necessarily absolutes). http://www.kjf.ca/
kjf/78-R9POR.htm
R12 Müller, H. F. J. (2005) Education: Con-
struction by individuals versus structure-
teaching. http://www.kjf.ca/kjf/78-
R12DY.htm
R13 Müller, H. F. J. (2005) Epistemological
asymmetry. http://www.kjf.ca/kjf/78-
R13NI.htm
R14 Müller, H. F. J. (2005) Mid-course cor-
rection (on Diettrich’s epistemology).
http://www.kjf.ca/kjf/78-R14DI.htm
R16 Müller, H. F. J. (2005) On the anti-con-
structivist view of Robert Nola. http://
www.kjf.ca/kjf/78-R16NO.htm
References
Chalmers, D. J. (1995) The puzzle of con-
scious experience. Scientific American
273: 80–86.
Crick, F. (1994) The astonishing hypothesis.
Scribner’s: New York.
Damasio, A. R. (1994) Descartes’ error. Emo-
tion, reason, and the human brain. Quill,
Harper Collins: New York, pp. 245–252.
Foerster, H. von & Glasersfeld, E. von (1999)
Wie wir uns erfinden. Carl-Auer-Systeme
Verlag: Heidelberg.
Glasersfeld, E. von (2004) The constructivist
approach: Toward a theory of representa-
tion. Karl Jaspers Forum TA73. http://
www.kjf.ca/kjf/73-TAGLA.htm. This is my
translation, corrected by the author, of a
lecture given by von Glasersfeld at the Uni-
versité du Québec à Montréal on 21
November 1985, and published as : Glaser-
epistemic structuring
sfeld, E. von, (2004) L’approche construc-
tiviste: vers une théorie des
représentations. In: Jonnaert, P. & Mas-
ciotra, D. (eds.) Constructivisme, choix
contemporains. Hommage à Ernst von
Glasersfeld. Presses de l’Université du
Québec: Sainte Foy, pp. 213–224.
Hameroff, S. (2003) Heading in the wrong
direction. Review of Cristof Koch’s “Quest
for consciousness.” Science & Conscious-
ness Review 2. http:// www.sci-con.org/
articles/20041102.html
Jaspers, K. (1947) Von der Wahrheit. Piper:
München, Zürich.
Jaspers, K. (1949) Vom Ursprung und Ziel der
Geschichte. Fischer Bücherei: Frankfurt &
Hamburg. Reprinted in 1955 by R. Piper &
Co: München. English translation by M.
Bullock: (1953) The origin and goal of
History. Yale University Press: New Haven.
Johnson, D. K. (2004) The view from some-
where: A philosophical critique of radical
constructivism. KJF Forum TA 75. http://
www.kjf.ca/kjf/75-TAJOH.htm
Lakoff, G. & Núñez, R. E. (2000) Where math-
ematics comes from. How the embodied
mind brings mathematics into being.
Basic Books: New York.
Koch, C. (2004) Thinking about the con-
scious mind. A review of Mind –An intro-
duction by John Searle. Science 306: 979–
980
McCarthy, J. (2004) What is artificial intelli-
gence? http://www-formal.stanford.edu/
jmc/whatisai/whatisai.html
Metzinger, T. (2003) Phenomenal transpar-
ency and cognitive self-reference. Phe-
nomenology and the Cognitive Sciences 2:
353–393. Reprinted in: (2004) Karl Jaspers
Forum TA67, http://www.kjf.ca/kjf/67-
TAMET.htm
Müller, H. F. J. (1997) Is the mind real? Karl
Jaspers Forum TA1. http://www.kjf.ca/kjf/
1-TA12.htm
Müller, H. F. J. (2000) Concept-dynamics and
the mind-brain question. Karl Jaspers
Forum TA32. http://www.kjf.ca/kjf/32-
TAMUL.htm
Müller, H. F. J. (2001) Brain in mind. Karl Jas-
pers Forum TA45. http://www.kjf.ca/kjf/
45-TAMUL.htm
Müller, H. F. J. (2003) Effect of working ontol-
ogy on some conceptual Puzzles. Karl Jas-
pers Forum TA57. http://www.kjf.ca/kjf/
57-TAMUL.htm
47
 philosophical-epistemological
CONCEPTS
Reiner, H. (1954) Die Entstehung und
ursprüngliche Bedeutung des Namens
Metaphysik. Zeitschrift für philoso-
phische Forschung 8: 210–237. English
translation: (1990) The emergence and
original meaning of the name “metaphys-
ics.” Graduate Faculty Philosophy Journal
13: 23–53.
Rilke, R. M. (1936) Das Stundenbuch. Insel-
Verlag: Leipzig. [These poems were writ-
ten in 3 volumes, from 1899 to 1903. The
first volume, from which the quoted poem
is taken, has the title “Vom moenchischen
Leben”. The poems were dedicated to Lou
Salomé.]
48
Rilke, R. M. (1949) Die Achte Duineser Elegie
(7–8 February 1922). Insel-Verlag: Reut-
lingen. German: http://www.phil-fak.uni-
duesseldorf.de/germ/germ4/gedichte/
riduin08.htm – English translation by Ali-
son Croggon: http://www.thedrunken-
boat.com/rilke.html – English translation
by Robert Hunter: http://www.dead.net/
RobertHunterArchive/files/Poetry/Ele-
gies/elegy8.html
Rioux-Soucy, L.-M. (2004) Mon voisin est un
cyborg. Le Devoir (Montréal), 5 Nov 2004.
Searle, J.R. (2004) Mind: A brief introduction.
Oxford University Press: New York.
epistemic structuring
Strawson, G. (2004) A fallacy of our age. Not
every life is a narrative. Times Literary
Supplement, 15 October 2004, p. 13.
Wiener, N. (1954) The human use of human
beings. Second Edition Revised. Double-
day Anchor Books: Garden City.
Wright, R. (2004) Fool’s paradise. Easter
Island’s unlearned lesson. Times Literary
Supplement, 19 November 2004, p. 16.
Received: 18 September 2005
Accepted: 19 October 2005
Constructivist Foundations
 education radical constructivism
EMPIRICAL

Against Realist Instruction:

Superficial Success Masking
Catastrophic Failure and an Alternative
Dewey I. Dykstra, Jr. A Boise State University, ddykstra@boisestate.edu

and because the development of understand-

Purpose: Often radical constructivists are confronted with arguments why radical con- ing is not central to education as we know it,
structivism is wrong.The present work presents a radical constructivist alternative to such his mentors, though sincere, were unable to
arguments: a comparison of the results of two instructional practices, the standard, realist- assist him in some way that would settle his
based instruction and a radical constructivist-based instruction, both in physics courses. dissatisfaction.
Design: Evidence from many studies of student conceptions in standard instruction (Duit
2004) is taken into account. In addition, diagnostic data, pre and post instruction, were
collected from over 1,000 students in multiple institutions across the U. S. over a period Understanding
of about 15 years via an established diagnostic of conceptual understanding of motion and
force. Findings: Evidence from many studies of student conceptions in standard instruction When students can repeat something verbatim,
(Duit, 2004) is that little or no change in student conceptions happens in standard instruc- it is obvious that they have learned it. Whether
tion. About half the students in the particular study reported, all science and engineering they have understood it, is a question these tests
majors, experienced standard, realist-based instruction and show an average effect size of avoid. (Ernst von Glasersfeld 2001)
0.6 standard deviations and an average normalized gain of 15%.The other half of the stu-
dents, none of whom were science and engineering majors, experienced radical construc- What might be meant by understanding? Von
tivist-based instruction and show an average effect size over 2.5 standard deviations and an Glasersfeld suggests that understanding is
average normalized gain over 60%. Diagnostic pre scores were nearly the same for both avoided in typical test results. Gardner makes
groups. Practical implications:The outcome, that students, neither science nor engineer- a kind of operational definition.
ing majors, made changes in understanding foundational topics in physics far greater than “…students who receive honor grades in
science and engineering students, poses (1) an ethical challenge to the continued adherence college-level physics courses are frequently
to standard, realist-based instructional practices and (2) an intellectual challenge to the unable to solve basic problems and questions
usefulness and appropriateness of the elitist-realist paradigm on which such standard encountered in a form slightly different from
instruction is based. Conclusions:This radical constructivist argument uses the effect of that on which they have been formally
paradigms to judge their pragmatic value, not their truth-value. Based on pragmatic value, instructed and tested.
radical constructivism results in superior outcomes when applied to physics instruction. “If, when the circumstances of testing are
The approach to instruction can be applied generally in education. slightly altered, the sought-after competence
Key words: elitism, physics, paradigm, realism. can no longer be documented, then under-
standing—in any reasonable sense of the
term—has simply not been achieved.”
Prologue Introduction (Howard Gardner 1991), pp. 3 and 6)
In Radical Constructivism (RC) (Glasersfeld In the Fall of 1969 a young man started teach- The orientation to the meaning of under-
1991) we judge our understanding by what ing high school physics. He believed that the standing in the present work is focused on the
can be accomplished using that understand- students should leave an instructional experi- nature of a person’s understanding, not on the
ing. Realists do not understand the notion ence understanding the phenomena studied nature of what might be claimed to be inde-
that RC is not about “truth,” but about fit to differently than they began the instructional pendent of that person. Hence, if one observes
experience. In this article, we look at evidence experience. As it turns out, this was naïve, but another to act in a certain way in some context,
in educational settings. While the evidence for him it was the point of teaching and edu- one can formulate an explanation, a con-
will be mostly from physics education, what is cation. He quickly realized it was not happen- structed understanding, under which the
being described is not unique to the results of ing in his own classroom and, as we shall see, other person seems to be operating by a pro-
physics instruction. Education as we know it later he and others found it does not happen cess known as abduction. (Peirce 1955) If, later
today is based in realism. How do the results in most classrooms. As a new teacher, he had in another context, one’s explanation of the
of this education as we know it compare with mentors who tried to help. Because at the time other fits the other’s understanding, then it is
results of education based in RC? he could not articulate what the problem was reasonable to be able to predict the behavior of

Constructivist Foundations 2005, vol. 1, no. 1 49

 education
EMPIRICAL
the other. If the other person does indeed
behave in the fashion predicted, then one can
make the claim that it is as if the constructed
understanding is present in the other. If the
observed behavior differs from the prediction,
then one can make the claim that the con-
structed understanding does not appear to be
present in the other. These constructed mental
models of the understanding of others are the
closest we can come to knowing the under-
standing of others. Descriptions of such
understandings that can be seen to be explana-
tions of the behaviors of others in the case of
force and motion are given later in this article.
On the prevalence of
change in understanding
in physics instruction
Early work
By 1980, this same young man had taught
high school for 4 years, completed graduate
work in Physics and taken a position in a uni-
versity Physics Department. At about the
same time he earned his doctorate, articles
were beginning to appear in journals describ-
ing students’ understanding of topics in phys-
ics. In some of these articles the following
observations were expressed:
Kinematics-velocity. “Our research also has
provided evidence that for some students cer-
tain preconceptions may be remarkably per-
sistent. As mentioned above, even on post-
course interviews, when difficulties occurred
they could be traced to the same confusion
between speed and position that had been
demonstrated during pre-course interviews.
The belief that a position criterion may be
used to compare relative velocities seemed to
remain intact in some students even after sev-
eral weeks of instruction.” (Trowbridge &
McDermott 1980)
Kinematics-acceleration. The conceptual dif-
ficulties with acceleration that were encoun-
tered by the students in our study appeared to
be very persistent. Often, as illustrated by the
pairs of interview excerpts on Acceleration
Comparison Task 1, the procedures used by a
particular student were the same before and
after instruction. … A significant number of
50
students from a wide variety of courses con-
fused the concepts of velocity and accelera-
tion. … At the completion of instruction,
fewer than half of the students demonstrated
sufficient qualitative understanding of accel-
eration as a ratio to be able to apply this con-
cept in a real situation. Even with assistance in
making the necessary observations, these stu-
dents were unable to combine this informa-
tion in a manner that permitted successful
comparison of two accelerations.” (Trow-
bridge & McDermott 1981)
Electric circuits. “We have examined students’
explanations of an extremely simple electric
circuit, one that involved only three major
components. We found that many students
were unable to interpret the circuit correctly.
… One suspects, therefore, that a significant
proportion of students in physics courses will
have this type of difficulty. Even more dis-
turbing is the fact that the misconception per-
sisted in some students who had been
through a calculus-based course in electricity
which included five experiments on electric
circuits.” (Fredette & Clement 1981)
Real image formation. “It was clear from the
interviews with the post-students that it is
probably not uncommon to emerge from an
introductory physics course without under-
standing the essential role of a converging lens
or a concave mirror in the formation of a real
image… There is often a tacit assumption that
students who have performed satisfactorily in
the geometrical optics portion of an intro-
ductory physics course can respond correctly
to the basic questions presented at the begin-
ning of this paper. The discussion above dem-
onstrates that, although they might have been
able to give correct verbal responses to these
questions, the students who participated in
our study were frequently unable to relate
their knowledge to simple, but real, optical
systems.” (Goldberg & McDermott 1987)
Scope of findings
By 1990 many such articles had been pub-
lished in many journals and books were being
written on the topic of students’ conceptions
in science. Several groups had been maintain-
ing bibliographies of these works in the mid-
dle 1980’s including our young man, now
older. These efforts were combined and can
be found in a regularly updated bibliography
radical constructivism
now including more than 6,400 entries (Duit
2004). All of the entries that document
change in students’ conceptions reveal that
little or no change happens when students
experience even the best of standard science
instruction, not just physics. The items in the
bibliography come from a variety of coun-
tries, in both hemispheres.
Entries in this bibliography now extend
back to 1904. What can be called person-on-
the-street (pots) conceptions of natural phe-
nomena have been documented in student
behavior and interviews over a full century.
Instruction has changed little since well
before that time—it still follows the standard
inform, verify, practice model. It is difficult not
to conclude that…
…in all science instruction for more than a
century, the result has been little or no change
in student understanding of the phenomena
studied.
An insidious change in understanding
– the affective side
While standard physics teaching seems to be
leaving students’ conceptions of the physical
world unchanged, it is not leaving students
unchanged in other important respects. Only a
tiny percentage leaves such instruction with
positive beliefs about either themselves or the
field of physics.
“On est frappés par la récurrence des mots
qui désignent l’expérience des mathéma-
tiques et ses souvenirs: dictature, répulsion,
terrorisme, couperet, cauchemar, mathopho-
bie; et en même temps: inintérêt, application
mécanique de regles, ennui profond. Il en va
largement de même pour les sciences, en par-
ticulier pour la physique, que touts les
enquêtes désignent comme la discipline ayant
laissé les plus mauvais souvenirs et provo-
quant apres coup le plus de réactions hostiles,
voire agressives.”
“One is struck by the prevalence of partic-
ular words which describe the experience of
mathematics and memories of it: dictatorial,
repulsion, terror, nightmare, math-phobia
and at the same time: disinterest, mechanical
application of rules, profound boredom. It is
largely the same in the case of the sciences, in
particular with physics, which all the inter-
viewees describe as the discipline that gives
the worst memories and provokes the most
hostile, even aggressive, reactions.” (Astolfi
1997, translation by the author)
Constructivist Foundations
 education
EMPIRICAL
Very successful students, as judged by their
high school physics teachers, speaking near
the end of their high school physics course:
“I used to love math and science. ... Now I
just want to get through. I am always being
told what to do, what to think. There’s no out-
let. I am supposed to absorb someone else’s
information and then I realized it’s not for
me.
“I listen all week, then when we do the lab,
there are really no surprises. ... It took me a
real long time to get into physics. It almost
seems that in physics you can figure out the
lab without actually doing it, which isn’t very
motivating. It just seems like, maybe it’s the
way it’s set up, but I pay attention all week and
I have a general idea of what’s going on. The
lab is on a Thursday, toward the end of the
week, so...we build up to the lab. ... We, my
group, we use what we learned in our notes,
the equations and stuff, to fix up our lab
results. Most of the time we read the lab back-
wards. I don’t know if that’s cheating but he
[the teacher] sets himself up that way.”
(McDonnell 2005, p. 584).
College students responding about their
experience in introductory physics and chem-
istry courses at the university level:
“I think the students around me are having
the same sort of thought-provoking questions
about the material that I put into my journal,
but under time pressure they don’t pursue
them, [and] eventually they learn to disregard
“extraneous” thoughts and to stick only to the
details of what they’ll need to know for the
exam. Since the only feedback we get is on the
homework assignments, the students cannot
help but conclude that their ability to solve
problems is the only important goal of this
class.
“[Another criticizes] …a course design
that assumes that everyone in the class has
already decided to be a physicist and wants to
be trained, not educated, in the subject…”
(Tobias 1990, pp. 37 and 41)
The last four of these comments were col-
lected in studies involving students with cre-
dentials typical of students who would do well
in the science and engineering. Clearly they
left their experience with a less than positive
attitude about physics as a field of study.
Sadly, the vast majority of those who expe-
rience instruction in science leave the experi-
ence believing they are not good at science,
physics in particular. In fact our system is so
2005, vol. 1, no. 1
radical constructivism
effective at convincing people early of this physical world. They are expected to carry out
characterization that few ever experience laboratory activities in which what they have
instruction on topics in physics by someone been informed is supposed to be verified.
who specializes in teaching such topics. Just They are expected to solve homework prob-
on the order of 25% of high school graduates lems or exercises in which they are to practice
take physics in U.S. high schools. As evi- what they have learned.
denced in the above comments, even taking
physics from such a specialist may only make 1. The diagnostic. The diagnostic, the Force
the result worse. What is of fundamental and Motion Conceptual Evaluation (FMCE)
importance here is not the flow of people into (Thornton & Sokolof 1998), is a set of multi-
the profession of physics, but the negative, ple-choice questions in which the questions
elitist lesson nearly all of the students con- and the sets of choices have been crafted to
clude about themselves—a lesson as we shall reveal students’ conceptions about force and
see is questionable at best. its relationship to motion and their concep-
tions about motion. The diagnostic has the
A closer look at the cognitive aspect purpose of discerning the nature of the stu-
During the 1990’s several diagnostics of stu- dent’s conceptions of force and motion, not
dent conceptions concerning various topics whether a student knows the “right” answers
in physics were developed. One of these was according to a physicist. The process of devel-
used before and after science and engineering opment, involving several thousand students,
students studied motion and force in intro- included collecting free-form responses to
ductory physics courses from institutions questions. Individual interviews were con-
across the U. S. over a period of a dozen years. ducted with students concerning their under-
Most of the institutions from
which the data was received are
large state supported universi-
ties of the sort producing the
bulk of the engineering and sci-
ence graduates in the U. S.
Pre and post data were pro-
vided from both of two different
levels of introductory physics at
the university level. One level of
course is one involving only
algebra and trigonometry and is
typically taken by majors in biol-
ogy, geology, kinesiology, con-
struction management, and pre-
health professions, such as pre-
medical. The other level of intro-
ductory physics course involves
the calculus and is taken by
majors in physics, chemistry,
geophysics, and engineering.
The same topics are treated and
similar laboratory exercises and
homework problems are carried
out. The significant difference
between these two courses is the Figure 1: This is a poster made by students at the very
level of mathematics. The teach- beginning of their study of the nature of force. Most
ing practices in the courses are students regardless of the introductory course, whether
essentially the same. Students they are in secondary school or college make very
are expected to attend lecture similar posters in that they express the person-on-the-
and read a textbook by which street (pots) view of the nature of force.
they are informed about the
51
 education
EMPIRICAL
Figure 2:This poster was made by non-
science/non-engineering students in a
conceptual physics course at the end of their
study of the nature of force in an alternative
approach to physics teaching. It reveals that
these students appear to understand the
Newtonian-like (New) view of the nature of
force and are aware of the distinction between
it and the pots view. Typical science or
engineering students at the end of standard
physics instruction on the nature of force are not
likely to depict a New view, nor be aware of the
distinctions between the views demonstrated
by the students who made this poster.
standing of the questions and reasons for the
choices they selected. The FMCE has ques-
tions concerning velocity, acceleration, and
what physicists would refer to as Newton’s
first and second laws of motion, Newton’s
third law of motion, and mechanical energy.
On each topic there are at least 5 questions
and several have in excess of 10 questions.
There is a mix of questions involving graphs
of either force or motion and questions that
do not involve graphs.
Of the 21 questions on Newton’s first and
second laws, 17 were used to formulate two
15-point scales. One of these corresponds to
the choices made by the student consistent
with the typical person-on-the-street (pots)
view of force and motion. The other 15-point
scale corresponding to the choices made by
52
the student consistent with a Newto-
nian-like (New) view of force and
motion.
2. Two views of force and motion. These
two conceptions of force and its rela-
tionship to motion can be briefly
described in the following ways. In the
pots view, force is the explanation of
motion or velocity. In this view there is
always a force in the direction of motion
and the magnitude of the velocity varies
as the magnitude of this force. Figure 1 is
from a poster made by a group of four
students at the beginning of their study
of the nature of force. In the New view,
net force is the explanation of accelera-
tion. The term, net force, refers to the
aggregate effect of all the forces that hap-
pen to be acting on an object at any point
in time. In this view the acceleration is
always in the direction of the net force
and, as the magnitude of the net force
varies, so does the magnitude of the
acceleration. Figure 2 is from a poster
made by non-science/non-engineering
students at the end of their study of the
nature of force. It suggests that an
understanding of these two views of
thinking about force and how these two
views contrast is present in the group of
four students who made the poster.
view do not generally make much
conceptual distinction between motion
and changing motion; that is, between
velocity and acceleration. Acceleration
for them is a kind of special case of
motion: velocity in which the magnitude of
the velocity is increasing. In this view deceler-
ation is a special case of velocity in which the
magnitude of the velocity is decreasing. When
the velocity is zero or constant, there can be
neither acceleration nor deceleration, hence
both have magnitudes of zero.
Persons who appear to use the New view
parse motion differently. For them all changes
in motion, that is, changes in velocity, are in
some sense equivalent and distinct from the
motion or velocity itself. This is much like the
distinction between a function and its deriva-
tive in the calculus, but familiarity with the
calculus is unnecessary to form this distinc-
tion about motion.
The two views are conceptually funda-
mentally different from each other in that the
radical constructivism
New view rests on this distinction between
velocity and this particular notion of acceler-
ation as any change in velocity. In the pots
view such a distinction does not exist. The
nature of force in the two views is very differ-
ent because what is being explained by the two
views of force, the velocity in one and the
acceleration in the other, is profoundly differ-
ent. In the actions of a person holding the pots
view of force, the notion of acceleration used
in the New view of force does not appear to
exist. Trowbridge and McDermott (1981)
refer to this in the quotation from their paper
about student conceptions of acceleration
given earlier in this article.
3. Evidence of change in science and engineer-
ing majors in standard physics instruction.
Table 1 gives some results from this study.
Only data from students for whom there was
both pre and post data was included. The
table indicates the type of physics course, the
general location of the institution, the year in
which the data was collected and the number
of students in each course.
It is clear that the initial (pre) person-on-
the-street (pots) view average score is rela-
tively high (about 10 out of 15) and the initial
(pre) Newtonian-like (New) view average
score is very low (0.6–2.6 out of 15). Scores in
these ranges might reasonably be expected in
the pre diagnostic, if there had been no previ-
ous instruction. But, one should keep in mind
that these students experienced the typical
curriculum in the U. S. They experienced one
or more instructional sequences on forces:
first, in elementary school (grades 1–6, ages
about 6–12), another in 8th or 9th grade (ages
about 14–16) and at least 25% of them
received instruction on these topics in high
school physics (normally 12th grade, age
about 18). Because the students in this part of
the study are all science or engineering
majors, it is probable that more than 25% of
them took a physics course in high school.
The changes from pre to post scores are
not particularly large. The pots view scores
drop from 9 or 10 down to 7 or 8 out of 15.
The New view scores rise from 1 or 2 up to 3
or 5 out of 15. This final outcome does not
convince one that any significant number of
the students in these courses leave with an
understanding of the New view concepts
explicitly taught by Ph. D. physicists and their
graduate students. These students still appar-
Constructivist Foundations
 education radical constructivism
EMPIRICAL

ently think about the world in terms of the

Whole class scores Average Scores Effect Size Normalized
pots view.
Effect size refers to the size of the difference Pre (0–15) Post (0–15) (st dev) Loss Gain
in the class average diagnostic scores, post
Year Term N pots New pots New pots New <L> <g>
minus pre, in units of the standard deviation
of the scores. An effect size of 0.6, the average Algebra–Trig Level Intro Physics
here, is often considered in the moderate West Coast Public Univ. A
range for educational research. Given the
actual final performance of the students, such 1990 99 10.1 1.5 8.5 3.3 –0.47 0.59 –0.16 0.13
an effect size can hardly be called laudable, “Prairie State” Public Univ.
especially in the New view score.
Normalized gain and loss are measures of 2002 SP 112 10.3 0.9 9 2.7 –0.40 0.66 –0.13 0.13
the fraction of the possible gain or loss that Calculus Level Intro Physics
could occur in the scores. In Table 1 the nor-
malized loss is calculated on the pots view North East State Public Univ.
scores. Typically for a whole class average in 1998 72 9.6 1.7 8.5 3.5 –0.30 0.47 –0.11 0.14
these examples of standard physics instruc-
West Coast Public Univ. B
tion, the pots view score drops. The normal-
ized loss is calculated to give a negative result 1999 Wint. 87 9.3 2.6 6.5 5.4 –0.62 0.60 –0.30 0.26
when the pots view score drops. The normal-
1999 SP 73 9.1 2.3 7.6 4.0 –0.36 0.38 –0.17 0.13
ized gain is calculated on the New view scores.
A typical normalized gain of 0.15, the average 2000 SP 115 9.2 2.4 7.2 4.8 –0.50 0.59 –0.21 0.19
seen here, or about 15% might be acceptable, West Coast Private Univ.
if the pre New view score were high. Since this
is not the case, a normalized gain in New view 2000 SP 38 9.8 0.6 9.6 1.7 –0.08 0.54 –0.03 0.09
score of 15% is wholly unacceptable. We need
Table 1: Pre–Post Data, Measures of change in normal instruction, science and
to be seeing effect sizes and normalized gains
engineering majors. pots: person-on-the-street view of force and motion; New: Newtonian-
that are many times larger, if the standard
like view of force and motion; N refers to the number of matched pairs of data from each class.
deviations remain similar.
Effect size is the post score minus the pre score, the quantity divided by the pooled standard
These results are consistent with reports
deviation. Normalized loss is the post pots score minus the pre pots score, the quantity divided
in the bibliography. They appear to be repro-
by the maximum the pots score could go down (to zero). Normalized gain is post New score
duced routinely every semester in most loca-
minus the pre New score, the quantity divided by the maximum the New score could go up
tions, in the U. S. and in many other locations
(15 minus the pre New score).
around the world. What is taught in standard
physics instruction is not understood by an
overwhelming majority of the students. It is intent, maybe we should look at what is hap-
important to remember the changes seen Realism in instruction pening:
here are the third or fourth attempt to teach [ Very little change in understanding of
these ideas to students considered to be, as Evidence calling for explanation physical phenomena occurs as a result of
science and engineering majors, among The tacit assumption and sometimes physics teaching.
those capable of learning this material. explicit characterization of physics teaching [ Most people we subject to this instruction
Apparently most students find the experience seems to be that we present content so that leave with an unrealistic view of the enter-
of physics instruction distasteful and dis- students can receive it with the idea that they prise of physics, that it is all mathematics
couraging. All of the students learn from hold on to it. The drive is to present the con- and completely determined by measure-
these experiences that there are a very select tent to the students so that they can have it. ments.
few who can make sense of physics, but the The implications being that (1) they must be [ Most leave the instruction believing they
vast majority cannot. presented the content in order to have it and are not capable of understanding physical
The outcome of standard instruction in (2) that the content can be presented in such phenomena. They must rely on those
physics is a spectacular failure and has an a way that it is possible for the students to who are capable of such understanding
appalling effect on society in general. We fail receive it. for knowledge of the truth about the phe-
to teach what we intend. Instead, we manage One could characterize physics teaching in nomena.
to teach most people they are on the lower this view as content-driven. But, if this really is [ Typical classroom activities and exams in
rung of a caste system in which they are the case, why is it that so many students have physics do not reveal the presence or
dependent on a higher caste for declarations failed to get it for so long with nothing being absence of changed understanding of the
of the truth. done about it? Instead of asking what is the phenomena.

2005, vol. 1, no. 1 53

 education
EMPIRICAL
[ Since around 1980, what is now referred to
as the physics education research (PER)
community has been bringing to our
attention the finding that most students
leave physics instruction not understand-
ing what has been taught. In addition to
the sources already cited, one can find
papers and sessions presented at meetings
of the American Association of Physics
Teachers, the American Educational
Research Association and the National
Association for Research in Science Teach-
ing since the late 1970’s on this issue.
[ As a consequence of physics teaching hav-
ing been the same, inform, verify, practice
method for numerous generations,
change in understanding in physics
courses has been lacking possibly for cen-
turies.
[ In that time many sincere, diligent, very
intelligent people have taught physics, yet
standard classroom activity and exams in
physics have been crafted which do not
reveal whether or not change in under-
standing has occurred. None of these peo-
ple seem to have noticed the general lack of
change in understanding of the phenom-
ena about which they teach.
[ Since the late 1970’s a number of very vocal
members of the physics teaching commu-
nity have openly dismissed the research
results and alternative teaching practices
showing vastly improved learning results,
for example Geilker (1997), Erlich (2002),
Cromer (1997) and Aldridge (1995).
[ Alternative approaches to teaching physics
demonstrated to result in significant
change in understanding are ignored and
resisted. Physics is still mostly taught as it
was for centuries before 1980.
An explanation
It seems in spite of the stated intent, what is
actually happening is better described in
another way. A better description would be to
describe the teaching of physics as:
Physics teaching is the presentation of the
established canon by approved methods
for the benefit of the deserving.
Embedded in this program is a realist
notion of the nature of the knowledge that
constitutes the canon.
“…we postulate the objective existence of
physical reality that can be known to our
54
minds…with an ever growing precision by
the subtle play of theory and experiment.”
(Torre & Zamorano 2001, p. 103)
That this knowledge can be transmitted is
clear in that it is to be presented. Apparently
approved methods have passed the criterion of
being effective transmissions of knowledge. In
this program such knowledge apparently can
exist in the symbols (words, sounds, gestures)
used in the presentation. It also apparently
exists in nature independent of the student
since what is presented is to be verified in lab-
oratory experiments or exercises.
It is acknowledged that not all can receive
the transmitted knowledge effectively. To
account for this the construct, deserving, is
applied. If one is deserving, then one can effec-
tively receive the transmitted knowledge. To
be deserving one must first have the mental
capacity and then one must work diligently
enough to be successful at “getting” what has
been transmitted or can be seen in nature.
In this program the teacher’s responsibility
is to present the established knowledge by
approved methods. This is frequently put as to
expose the students to the knowledge. At this
point the teacher’s job is essentially com-
pleted. Whether or not a student “gets” the
knowledge is out of the teacher’s hands. The
student is either deserving or not. Maybe the
teacher can influence students to be diligent or
work hard, but the mental capacity part was
set before the teacher comes in contact with
the student.
It is important to notice that this program
also implies a concept of the nature of people.
A few people are deserving, but most are not.
This is an elitist notion of people. Some people
can “get” it but most cannot – but that’s okay,
we can’t all be physicists (sic).
This program is an expression of teaching
within a realist, elitist paradigm. Being a par-
adigm in the Kuhnian sense, it explains all rel-
evant observations. The paradigm defines
what observations have sufficient status to be
addressed in the paradigm and what observa-
tions do not. It is a complete system within
itself. There is no need to ask why so few “get”
the transmitted knowledge. Von Glasersfeld’s
question: “…but do they really understand?”
is irrelevant and non sequitur. To question the
approved methods or not be driven to present
the canon is heresy within the paradigm. This
paradigm construct explains the observations
listed above.
radical constructivism
Situations describable as realist paradigms
are very possibly unique. The underlying
beliefs and characterizations of the world in
such paradigms are considered statements of
“objective truth.” Hence, the whole system of
such paradigms is not considered a construct
by the true believers. Instead, it is the truth.
Such paradigms are not ideologies according
the their practitioners, because the elements
that constitute the system are statements of
truth. As truth, once established, it is not to be
questioned.
We see then that the meaning of content-
driven as applied to this description of physics
teaching is the drive to present the content.
One must cover the subject. It is not about stu-
dents “getting” the content. Some more consci-
entious of the practitioners of the paradigm
may tweak the methods and take very small
liberties with what portions of the canon are
presented to see if a few more of the deserving
can be uncovered. This experimentation is
limited. One who goes too far runs the risk of
being accused of heresy. Such pressure is
always carried out in the name of objectivity,
since the ideology of the paradigm is that there
is no ideology to the paradigm.
To prepare a physics teacher in this para-
digm, we must first make sure that person is in
possession of the canon. Without this, what
would be presented is false, corrupt or incom-
plete. In the U. S. we expect the potential
teacher to take as many as possible of the phys-
ics courses a “real” physics major takes. Then,
we spend a semester teaching this person
approved methods of presentation. This is
called a “methods” course. We give them a lit-
tle practice and a chance to show they can exe-
cute the methods that have been taught. This
is called “student teaching” in the U.S. When
teacher candidates can repeat back the canon
including the proscribed skills and execute the
methods of presentation, then we certify them
to be teachers of physics. We have an approved
practitioner of the paradigm.
From within the paradigm just described,
using T. S. Kuhn’s terms, the normal science is
that this is the way things are. We cannot, but
continue to refine our present understanding
as we approach ever more closely the truth. We
are closer now than we were a decade past.
Things are just this way.
We can judge this paradigm by its effect on
society. Its system fails students. Students leave
instruction with the same understanding of
Constructivist Foundations
 education radical constructivism
EMPIRICAL

the phenomena as they began the instruction.

Most of the students learn that they are not Whole class scores Scatter Plot Averages Effect Size Normalized
among the deserving. It fails society in that it Pre (0–15) Post (0–15) (st dev) Loss Gain
promotes elitism, creating an artificial caste
Year Term N pots New pots New pots New <L> <g>
system, and renders most members of society
intellectually stunted or handicapped. Conceptual Physics, College Level
Intermountain State University
An alternative paradigm 2000 FL 90 9.3 0.8 2.5 9.2 –2.20 2.50 –0.66 0.59
2001 SP 87 9.8 0.8 2.2 9.6 –2.40 2.40 –0.74 0.62
As it turns out, the young physics teacher, now
older, started out, without realizing it, on the 2002 FL 66 9.4 0.8 2.2 8.8 –2.19 2.26 –0.72 0.57
outside of the established paradigm. He mis- 2004 SP 69 7.9 2.2 1.8 10.8 –1.85 2.31 –0.78 0.67
takenly thought the point of teaching was that
students develop new understanding as a 2005 SP 53 9.9 0.8 1.5 11.1 –3.08 3.38 –0.85 0.73
result of their experience in the classroom. High School Level
For him the typical outcomes of conventional
teaching were disturbing. Without yet being North Central State High School
able to articulate the nature of this anxiety, he 2001 FLa 23 11.3 0.6 0.6 13.3 –5.4 6.3 –0.95 0.89
searched for an answer. He found it when the
2001 FLb 24 10.6 0.9 0.8 13.1 –3.7 6.1 –0.93 0.86
work of the Genetic Epistemologist, Jean
Piaget, was described so that he could see it as
a theory base from which to operate in the Table 2: Pre–Post Data, Measures of change in alternative instruction, non–
classroom. (Fuller, Karplus & Lawson 1977) science, non–engineering majors.
In a sense this theory base would enable him
to do science as he taught with the goal of A radical constructivist paradigm that give such declarative statements meaning
empowering his students to develop new This alternative teaching practice is embed- to the maker of such statements. This explan-
understanding. ded in a radical constructivist (RC) paradigm. atory knowledge exists only in the mind of
(Glasersfeld 1991) In this paradigm the each individual as a constructed mental
A different teaching practice nature of knowledge is incommensurate with entity. As such, this knowledge cannot be
Fortunately in addition to being able to study that of the realist-elitist paradigm described transmitted. It is a consequence of the condi-
Piaget’s ideas, the young man benefited from above. In this RC paradigm, knowledge can tion that meaning exists nowhere but in the
close mentoring contact with a number of be divided into two types. One is experience, mind of the individual, that for the meaning
colleagues. As of this point in time one result experiential knowledge, and the other is to arise in the individual, the individual must
is an alternative practice of physics teaching explanation, explanatory knowledge. (Jam- construct it. Hence, the label, constructivism,
that can be called student understanding- mer 1999) This explanatory knowledge can- describes the consequence of the fundamen-
driven. This teaching practice can be not be judged any other way than for fit to tal nature of knowledge employed in radical
described as: experience. The degree of fit does not convey constructivism.
Physics teaching is the process of in any way the status of true description of an For a realist, everything breaks down at
engaging students in developing independent reality or of being closer to such this point, if it has not already. We are all iso-
new understanding of physical phenomena truth. Such truth for explanatory knowledge lated and incapable of communicating with
has neither existence nor status in this RC par- each other, if meaning cannot be transmitted.
The focus here, the central object of adigm. This is one of the fundamental points For the radical constructivist, nothing could
manipulation, is neither the canon, nor the of incommensurability between radical con- be further from the case, but in RC commu-
phenomena, nor the apparatus. It is the stu- structivist and realist paradigms. nication has an entirely different explanation.
dents’ explanatory schemes, their concep- Experiential knowledge is the experience For the realist, the transmission of realist-type
tions concerning the phenomena, their itself, hence experiential knowledge cannot knowledge cannot be dissociated from com-
understanding of the phenomena. The expe- be transmitted via language. Students must munication. For the radical constructivist,
riences, for which these explanatory schemes have their own experiences. Without these communication is the individual construc-
are developed, play the important role of experiences there is nothing to explain, no tion of meanings to be associated with sym-
checks on the explanations for fit, but it is the need for explanatory knowledge. bols and combinations of symbols from
explanatory conceptions that are the main The explanatory knowledge is not declar- someone else. At an early age, one constructs
focus of attention. These conceptions consti- ative statements but the meaning of such the notion of “other” based on patterns of reg-
tute a student’s understanding, hence the statements, the understanding from which ularity of experience. Later, one modifies the
practice is student understanding-driven. the statements are generated, the conceptions construct “other” to endow it with cognizing

2005, vol. 1, no. 1 55

 education
EMPIRICAL
capacities one is aware of in one’s own con-
sciousness. By experience, trial and error, and
reasoning, one is continually building and
modifying a kind of look-up table that con-
nects the symbols of language with meanings
that appear to fit experience (Glasersfeld in
press). Through copious interaction with
“others,” we develop look-up tables that work
sufficiently well that we can take our own
look-up table as shared with “others.” This
taken-as-shared communication process
enables us to interact enough to decide cer-
tain experiences can be taken-as-shared and
that explanations can be taken-as-shared.
The ability to make these mental construc-
tions is considered a capacity of all human
beings. Elitism plays no role in this paradigm,
either in the teaching practice or in explaining
the outcomes of the teaching.
Piaget describes a mechanism that drives
this process of meaning construction. What
drives meaning construction is the need or
desire for equilibration between one’s explan-
atory mental constructs and one’s experi-
ences. (Piaget 1985) One moves to modify or
construct new explanation when one per-
ceives one’s existing mental constructs do not
fit experience, i.e., when one disequilibrates.
Because the resolution of the disequilibration
is new mental constructs that do fit experi-
ence, the resulting accommodation is always
one that fits a greater range of experience,
hence, a kind of pragmatic progress. In a nut-
shell, for change in understanding to occur,
the teacher first needs to engage the students’
attentions in comparing their existing con-
ceptions with some behavior of the phenom-
enon that likely does not fit those concep-
tions.
Some results of this alternative
teaching practice
Using the same diagnostic, the FMCE,
described previously in this work, conceptual
change on force and motion was studied in a
course for non-science/non-engineering
majors. The teaching practice used is the one
described above from within a radical con-
structivist paradigm. Fewer of these students
are likely to have had physics in high school.
Most college science faculty imagine these
students to be in the category: less deserving.
As such the learning results would be
expected to be inferior to that of the science
and engineering majors.
56
A high school teacher trying out the alter-
native teaching practice for the first time also
used the FMCE diagnostic. The only modes of
communication between this teacher and the
author were electronic mail and telephone.
The teacher was conducting a project as part
of the requirements for a master’s degree in
science education. Table 2 displays data col-
lected in these two different classroom set-
tings.
A comparison between Table 1 and Table 2
shows a marked difference in conceptual
change on the diagnostic scores. The initial
average scores for each view in Table 2 are not
particularly different than those in Table 1.
This is because these students experienced
similar standard instruction in elementary
school and in the 8th or 9th grades to that
experienced by the science and engineering
majors in Table 1. The magnitudes of the
changes in table 2 are much larger. The effect
sizes are larger and the normalized changes
are larger. The effect sizes easily meet Bloom’s
challenge of a change of two-sigma over the
results of normal instruction. (Bloom 1984)
Clearly, non-science, non-engineering
majors in this alternative teaching practice,
guided by radical constructivism, consis-
tently and routinely change their understand-
ing of these phenomena by an amount several
times larger than science and engineering
majors in standard physics instruction guided
by an elitist realism. This is far beyond statis-
tical significance. Typical statistical signifi-
cance in educational research is claimed for
the kind of changes (0.5 standard deviations)
seen in Table 1. Yet, we see that the level of
understanding actually accomplished in
Table 1 post-scores is so small as to be imprac-
tical as a justification for the instruction.
Two reasons can be imagined to explain
that standard instruction is so entrenched
and so widespread, but with such little actu-
ally learned. The first comes out of the
broader explanatory scheme used here. The
standard elitist-realist paradigm, as is a char-
acteristic of paradigms, has developed an
explanation for everything it deems relevant.
Such outcomes as seen in Table 1 are just the
way things are. Very few really are “deserving,”
so we spread a wide net to catch the few
“good” people. Add to this an assessment
scheme which consists of (1) checking to see
if the catechism can be recited (assessment of
the students) and (2) looking to see if every-
radical constructivism
one is complacent and satisfied (assessment of
the teachers) and one hardly needs any more
explanation. The second, not completely
independent of the first, is that the system we
have works out really well for those consid-
ered deserving. To perpetuate this caste sys-
tem, a system is needed to convince people,
even the undeserving, that this is just the way
things are. In that way, the deserving do not
have to defend their special status, all of soci-
ety will do this for them. Hence, the preserva-
tion of inflated egos of the “deserving” can be
seen to be a factor is preserving this status quo
in physics instruction.
Teaching within a radical
constructivist paradigm
In this alternative program of physics teach-
ing, the teacher plays a fundamentally differ-
ent role.
“…a physics major has to be trained to use
today’s physics whereas a physics teacher has
to be trained to see a development of physical
theories in ... students’ minds.” (Niedderer
1992), p. 151)
Having the students read a standard text or
the teacher present the canon, not only is a
waste of time; it stifles the process of develop-
ing new understanding. In standard instruc-
tion there is a text to be read and relied upon
and most class time is taken up by instructor
lectures, yet we see no useful change in under-
standing in Table 1. Instead, most of the class
time needs to be occupied with students
explaining to each other their conceptions,
discussing how well the various conceptions
fit the experiences with the phenomena, plan-
ning with each other what adjustments might
be called for when the fit to experience is
found lacking, and discussing the results of
tests of these accommodations against further
experience with the phenomena.
In order to see the development of physical
theories in students’ minds, the teacher must
have access to copious amounts of student
explanations and predictions concerning the
phenomena being studied. The teacher needs
to be familiar with ways of thinking about the
phenomena the students are likely to have. A
teacher candidate can begin developing this
familiarity by studying the efforts of others
who have examined students’ conceptions.
The bibliography (Duit 2004) is a major
source in such study. Ultimately, it is neces-
sary to listen to and watch many students as
Constructivist Foundations
 education
EMPIRICAL
they demonstrate their understanding of the
phenomena and as they evolve their under-
standing. This has to happen in the class-
room.
In a RC paradigm, teaching cannot be
about the teacher confronting the misconcep-
tions of students and correcting them. This is
the typical, very logical response of those in
the elitist-realist paradigm who deign to look
at the student conceptions research in the bib-
liography. In RC a student’s conception is not
a misconception. It fits the student’s experi-
ence sufficiently that the student perceives
equilibrium between the conception and
experience. It is the student’s perception of
equilibrium or disequilibrium that plays the
central role. The teacher cannot give the stu-
dents new conceptions because the teacher
cannot transmit meaning. Only the student
can change his or her own conception. This
only happens when the student perceives
some disequilibration, lack of fit, between
personal conceptions and personal experi-
ence. All a teacher can do is to set up condi-
tions in which students are more likely to
make changes.
In order to influence whether or not the
students make any changes to their concep-
tions, the teacher needs to engage the students
in a series of processes:
1. Elicitation: First, the students need to be
engaged in examining their own beliefs about
the phenomenon at hand. Each student needs
to make these explicit to her or himself by
writing and then talking about them. This
process is often called the elicitation of initial
conceptions. Normally it is not necessary in
everyday life to make such things explicit to
oneself, nor is it called for in normal school-
ing; hence it is not a practice most are com-
fortable with or skilled at. In fact, in typical
schooling students learn at a very early age
that it is not wise to express one’s own ideas,
but to focus on guessing what the teacher
wants someone to say.
To accomplish elicitation in the face of
these challenges, students can be engaged in
making a prediction. They are greeted with an
actual example of the phenomenon and asked
what they think would happen if a certain
change were made. In addition to the predic-
tion, an explanation that makes sense to them
is asked for. Students are asked, first, to write
this down without discussion. Then, they are
asked to share their ideas with a small group
2005, vol. 1, no. 1
of other students. In this sharing discussion,
they are asked to interact and try to under-
stand any new ideas or new nuances of ideas
they encounter and make notes about these.
The point here is not whether the prediction
is accurate, but that the students make explicit
to themselves and each other the nature of
their conceptions.
2. Comparison: Until this point they are
generally restrained from actually trying to
see what will happen. The central object of
manipulation here is neither the apparatus
nor the phenomenon itself. Instead, it is the
students’ understanding, their explanatory
conceptions, of the phenomenon. To try
things first generally drives these conceptions
deeper making them more difficult to elicit
and explicitly examine. This latter is the func-
tion and purpose of putting the elicitation
phase first. Once the elicitation is completed
and all understand the explanations deemed
reasonable, it is time to check to see if experi-
ence fits any of these explanations. The stu-
dents are asked to carefully observe and faith-
fully record what is observed with respect to
the particular prediction at hand. They need
to make note of what fits the predictions and
what does not fit the predictions. In the case
of the latter they are asked to make specific
notes about the nature of mismatch between
the experience and the predictions.
It should be noted that since the teacher is
trying to establish conditions in which con-
ceptual change would occur, the teacher
should select specific examples in which what
the students will predict does not match the
experience they will have. This requires the
teacher to have constructed a sufficiently reli-
able mental model of the students’ mental
models in order to make such selections. It
also requires that the teacher have a broad
knowledge of the details of experiences possi-
ble with the phenomena to be studied. Note
that the canon of physics has not been men-
tioned here. It is very difficult for teachers to
have these skills unless teachers have explicitly
participated in the same sorts of processes to
accomplish change in understanding them-
selves.
3. Resolution: When the anticipated dise-
quilibrated state has been achieved by the stu-
dents, given an intellectually safe environ-
ment, many begin to critically analyze their
initial explanatory knowledge and the nature
of the misfit between it and their new experi-
radical constructivism
ence. The students are encouraged to con-
struct possible modifications to those initial
conceptions or whole alternatives. To achieve
an accommodation, it is necessary to test
these modifications or alternatives. Such tests
are carried out by first working out predicted
outcomes based on the proposed changes and
then checking to see what happens. Iterations
are continued until most students report sat-
isfactory equilibration.
4. Application: The testing of possible
accommodations constitutes a nesting of
additional phases (1) & (2) repeated within
the third phase. Alternatively it can be seen as
a kind of 4th phase, one of application in which
not only the testing of potential modifications
to explanation is conducted, but the phenom-
enon is further explored, using apparently
successful explanatory schemes. In effect then
the phenomenon is seen through the new per-
spective made possible by the new explana-
tion. In the process how well and broadly this
scheme applies to the phenomenon is deter-
mined. Often, new aspects of the phenome-
non are discovered and deeper understanding
of the explanatory scheme is realized.
Superficial success:
Training and indoctrination of scientists in
the elitist-realist paradigm is accomplished
It takes 3–5 or more cycles of standard
instruction for understanding to be
accomplished
Spectacular failure:
For nearly all of society there is
no change in understanding concerning the
phenomena studied
In any science
In any country
For the last century or more
Students, including those who become
scientists, learn
Only a few special people can “understand”
science.
We must rely on “scientists” for “scientific
truth.”
Table 3: Outcomes of elitist–realist
science instruction. Not unique to science
instruction
57
 education
EMPIRICAL
Student understanding-driven, not
canon or content-driven
The established canon does not drive the
ordering or development of the predictions
that are used. It could be allowed to do so, but
the learning results suffer when this is done.
Before the semesters of the college level con-
ceptual physics course shown in Table 2. The
series of predictions the students were asked
to engage in was still partially canon driven.
Equal time in the study of motion, kinemat-
ics, was given to position, to velocity and to
acceleration. The standard paradigm holds
that one cannot really know velocity until one
really knows position and so forth for acceler-
ation. It was noticed from the diagnostic data
that whether or not a student demonstrated
understanding of velocity was not a predictor
of change in understanding with respect to
force.
Under this canon-dictated, equal-time-
for-all-three-topics design, the typical effect
sizes were slightly under 2 standard devia-
tions and the normalized gain was slightly less
than 0.5. When all but two of the ten activities
on position and velocity were dropped and
the time gained was used to examine acceler-
ation more deeply, results changed. The result
of abandoning the canon and allowing one’s
understanding of the students’ understand-
ings drive the process can be seen in Table 2.
There was an additional 0.5 standard devia-
tion effect size and about 0.15 normalized
gain. This departure from the canon results in
an additional change essentially equal to that
of the standard instruction in total.
Conclusions
Standard physics instruction is effectively
described as the presentation of the established
canon by approved methods for the benefit of
the deserving. It runs on an inform, verify,
practice cycle. Teaching practices based in this
description result in almost no practical
change in understanding of the phenomena
studied on the part of the students. On the
other hand, there is change in understanding
as a result of this instruction. Unfortunately,
the change is that students learn a caste system
based on who can “understand” the phenom-
ena in standard instruction and who cannot.
Since most of them leave the instruction not
understanding, most decide they are not in
58
radical constructivism
the caste of those who can understand. It is ing practice runs on an elicit, compare, resolve,
those in the caste who can “get” the canon apply cycle. Students considered less deserv-
who are considered the deserving. Further- ing, less capable in the other paradigm, but
more, all learn that this is just the way things taught in this alternative practice, are shown
are—the ideology-less ideology of objectivity repeatedly to be capable of making far greater
in realism. This program of teaching and the change in understanding than science and
elitist-realist paradigm on which it is based engineering majors taught in the elitist-realist
can be seen to explain the spectacular, wide- paradigm of standard instruction. This result
spread, and long-term failure of standard demolishes the “objectivity” of the deserving/
physics instruction and its destructive influ- less deserving explanation for the fact that so
ence in society. few students “get it” in standard science
If there were no examples of effective alter- instruction. As a result the “objectivity” of the
natives, this state of affairs could be argued as whole elitist-realist paradigm fails.
truth,. An alternative approach to teaching In RC one cannot claim radical construc-
based on a radical constructivist paradigm is tivism to be “The True paradigm.” One can
shown to be one such alternative. This teach- only show that RC is the basis for a paradigm
FUTURE WORK
Regardless whether one agrees or disagrees with the above line of reasoning, one still faces a
moral, ethical dilemma in terms of social justice in education.The outcomes of standard instruc-
tion are unsatisfactory and destructive. At least one alternative with good outcomes has been
demonstrated.
In the name of social justice in education, are we not obligated to respond in a careful, but serious,
reasoned way, that rises above sectarian bickering?
Are we not obligated to end, with all possible haste, the negative outcomes of standard
instruction? Are we not obligated to end, with all possible haste, the training of teachers to inflict
such intellectual and social damage by thoroughly revising their training to equip teacher can-
didates not to inflict such damage? Unless we accomplish paradigm change from the elitist-
realist paradigm, that paradigm will remain hegemonic and the destruction will continue. Got
change for a paradigm?
Paradigm change occurs when people become dissatisfied with things as they are. One can
reasonably argue that our young man started outside the prevailing paradigm.What drove him
to develop an alternative practice of teaching and to consciously define for himself his paradigm
was his disequilibration over the discrepancy between his expectations about the outcomes of
teaching and his observations of the outcomes when he began teaching. Can teacher candi-
dates be engaged in the same discrepancy? This may be what Niedderer (1992) was referring
to when he wrote: “…a physics teacher has to be trained to see a development of physical
theories in ... students’ minds.”
It is possible to imagine a RC-based course of study for teacher candidates in any subject.
This course of study would have as its central focus; the evidence of a person’s understanding
in whatever subject is to be taught. Surrounding this central focus should be the examination
of how, why and under what circumstances understanding appears to change and methods of
facilitating this change process in the students.The presentation of the established canon of the
subject and approved methods would cease to determine anything in the course of study for
teacher candidates, though the established canon might remain a presence at the periphery in
the course of study for teachers.
Students who learn in this radical constructivist paradigm not only develop significant,
deeper understanding of the phenomena studied, but also develop a different self-image. This
self-image is positive and empowering. Students come to see the value in understanding the
point of view of others and develop skills for working with others to create better common
understanding of issues they face. Every student in every classroom can do this. Wouldn’t the
world they create be far better than the one we have now? When do we start changing how
we teach? If not us, who? If not now, when?
Constructivist Foundations
 education
EMPIRICAL
that yields far more favorable results in
instruction. On the other hand, the failures of
instruction in the elitist-realist paradigm and
its failure to fit experience, in terms of who
can and who cannot “get” instruction in phys-
ics, do enable us to draw the conclusion that
the elitist-realist paradigm fails on the
grounds of outcomes and logical integrity. It
should be either abandoned or substantially
modified, if such is possible. Until and unless
a satisfactory modification is demonstrated, it
should not be allowed to drive what it calls
education. What it calls “education” is not
education. At best, it is training and indoctri-
nation in a rationally and ethically unsup-
portable paradigm. At worst, it is ideological
indoctrination and is a destructive institution
in our society. It has no place in the education
of our society.
The line of reasoning presented is about
the relative usefulness of realism vs. radical
constructivism. Radical constructivism leads
to outcomes much more desirable than elitist
realism in the context examined. While one
References
Aldridge, W. G. (1995) Invited Paper. The Sci-
ence Teacher. October: 8.
Astolfi, J.-P. (1997) Preface. In: Trabal, P. (ed.)
La violence de l’enseignement des mathé-
matiques et des sciences: Une autre
approche de la sociologie des sciences.
L’Harmattan: Montreal.
Bloom, B. S. (1984) The 2 sigma problem: The
search for methods of group instruction as
effective as one-to-one tutoring,” Educa-
tional Researcher 13: 4–16.
Cromer, A. (1997) Connected knowledge:
Science, philosophy and education.
Oxford University Press: New York.
Duit, R. (2004) Students’ and teachers’ con-
ceptions in science: A bibliography. Avail-
able at http://www.ipn.uni-kiel.de/
aktuell/stcse/stcse.html
de la Torre, A. C. & Zamorano, R. (2001)
Answer to question #31. Does any piece of
mathematics exist for which there is no
application whatsoever in physics? Ameri-
can Journal of Physics 69: 103.
Ehrlich, R. (2002) How do we know if we are
doing a good job in physics teaching?
American Journal of Physics 70: 24–28
2005, vol. 1, no. 1
can see direct application to physics teaching,
the conclusion and applications apply much
more generally. This is justification for trying
to understand radical constructivism instead
of trying to prove it wrong.
Epilogue
Presently our young physics teacher, now
much older, is accomplishing his goal of engag-
ing students in developing new understanding
of physical phenomena. His still developing
understanding of the work of Piaget and of
Radical Constructivism play significant roles
in the on-going development of this successful
teaching practice. The results for college stu-
dents in Table 2 reveal that his students make
changes in their understanding of the phe-
nomena in quantity far superior to that of stu-
dents experiencing standard instruction. He
continually works at engaging the unengaged
in his classrooms in order that even more
achieve the results of the rest of the class.
Fuller, R. G., Karplus, R., Lawson, A. E. (1977)
Can physics develop reasoning? Physics
Today 30: 23–29.
Fredette, N. H. & Clement, J. J. (1981) Student
misconceptions of an electric circuit:
What do they mean? Journal of College
Science Teaching 10: 280–285.
Geilker, C. D. (1997) Guest comment: In
defense of the lecture-demonstration
method of teaching physics. American
Journal of Physics 65: 107.
Glasersfeld, E. von (1991) Knowing without
metaphysics: Aspects of the radical con-
structivist position. In: Steier, F. (1991)
Research and reflexivity (Inquiries into
social construction edited by Frederick
Steier). Sage Publications: London. Avail-
able at http://www.vonglasersfeld.com
Glasersfeld, E. von (2001) Radical construc-
tivism and teaching. French original pub-
lished in: Perspectives 31: 191–204.
Available at http://www.vonglasers-
feld.com
Glasersfeld, E. von (in press) The constructiv-
ist view of communication. Presented at
the 2003 Memorial Meeting for Heinz von
Foerster, Vienna. Available at http://
www.vonglasersfeld.com
radical constructivism
ABOUT THE AUTHOR
Dewey Dykstra grew up in Maryland, USA.
In junior high and high school he entered
and won prizes at Science Fairs. He gradu-
ated with a B. S. in Physics from Case Insti-
tute of Technology in 1969 and began
teaching high school in the inner city of
Cleveland, OH that same year. After three
years he moved back to Maryland and
taught 9th grade physical science and 12th
grade physics for a year. He entered gradu-
ate school at The University of Texas at Aus-
tin and earned a Ph. D. in condensed matter
Physics in 1978. He has been on the Physics
faculty at Oklahoma State University and
Boise State University. He is privileged to
have had as mentors: Robert Fuller, James
Minstrell and Ernst von Glasersfeld, among
others. – “In addition to teaching and
research in physics, I have conducted
projects investigating the nature of learning
in physics problem solving and developing
physics instructional materials with guid-
ance from evidence collected concerning
students' conceptions of physical phenom-
ena. Over the years I have held several posi-
tions in the American Association of Physics
Teachers, both elected and appointed and
is a regular contributor to meetings of this
organization. I enjoy travel and have been
pleased to share my work or taught physics
in Canada, Germany, India, Italy, Mexico,
Scotland and Trinidad. In India I had the priv-
ilege of engaging Tibetan Buddhist scholar
monks in inquiry concerning physical phe-
nomena. When I get a chance, I enjoy visit-
ing archaeological sites, playing bagpipes,
and cooking.”
Goldberg, F. M. & McDermott, L. C. (1987)
An investigation of student understanding
of the real image formed by a convex lens
or concave mirror. American Journal of
Physics 55: 108–119.
Jammer, M. (1999) The concept of force.
Dover: New York. Originally published in
1957 by Harvard University Press: Boston.
McDonnell, F. (2005) Why so few choose
physics: An alternative explanation for the
leaky pipeline. American Journal of Phys-
ics 73: 283–286.
Niedderer, H. (1992) What research can con-
tribute to the improvement of classroom
59
 education
EMPIRICAL
teaching. In: Nachtigall, D. K., Bartsch, H.
& Scholz, C. (eds.) Proceedings of the
international conference on physics teach-
ers’ education. University of Dortmund:
Dortmund, pp. 120–155
Peirce, C. S. (1955) Abduction and induction.
In: Buchler, J. (ed.) philosophical writings.
Dover Publications: New York, pp. 150–
156.
Piaget, J. (1985) The equilibration of cogni-
tive structures: The central problem of
intellectual development. (Translated by
T. Brown and K. J. Thampy). The Univer-
60
sity of Chicago Press: Chicago. Originally
published in French 1975 by Presses Uni-
versitaire de France.
Thornton, R. K. & Sokoloff, D. R. (1998)
Assessing student learning of Newton’s
laws: The force and motion conceptual
evaluation and the evaluation of active
learning laboratory and lecture curricula.
American Journal of Physics 66: 338–352
Tobias, S. (1990) They’re not dumb, they’re
different: Stalking the second tier.
Research Corporation: Tucson AZ.
Trowbridge, D. E. & McDermott, L. C. (1980)
radical constructivism
Investigation of student understanding of
the concept of velocity in one dimension.
American Journal of Physics 48: 1020–
1028.
Trowbridge, D. E. & McDermott, L. C. (1981)
Investigation of student understanding of
the concept of acceleration in one dimen-
sion. American Journal of Physics 49: 242–
253.
Received: 17 September 2005
Accepted:13 October 2005
Constructivist Foundations