6

The Concept of System and

the Paradigm of Complexity*
EDGAR MORIN

It is important to clarify the concept of system. Although the systems

theory has shown the generality of the concept of system, it has not gone
deep enough to discover the generative nature of systems.
As for the generality, all that was considered as matter in the 19th
century is now considered as system (an atom, a molecule, a star, etc.).
All that was considered as living substances have now become living
systems. However, what is social has always been considered as a
system. But this generality is not sufficient for recognizing the epistemo-
logical implications of the notion of system.
The systems theorists have thought that they have settled the prob-
lem. They believe that the new principle is holism, which seeks an
explanation at the level of the whole and opposes reductionism.
However, I would like to point out that their holism is based on the
same simplistic principle as the reductionism to which it opposes. As I
discusses elsewhere (Morin, 1977, p. 101), the systems theory has
neglected to scrutinize its own foundation and to clarify the concept of
system. Thus the paradigm of system remains underdeveloped,
atrophied, and stuck. The systems theory suffers from a basic weakness
and tends to fall repeatedly into reductionistic, simplistic, distorting,
and manipulating ruts, which it was supposed to have overcome and
from which it should have freed us.
On the contrary, the understanding of systems requires a new princi-
ple of knowledge different from holism. We must see systems not only
in a general term, but also in a generative way. We need a new
paradigm. "Paradigm" is defined here as a set of basic relations of
association and/or opposition among a limited number of key notions
that command or control all our thoughts, discourses, and theories.

* Translated from French by Magoroh Maruyama.

M. Maruyama (ed.), Context and Complexity 125

© Springer-Verlag New York, Inc. 1992
126 Edgar Morin

The notion of system has always been basic in designating the entire
set of elements constituting the whole. The notion becomes revolution-
ary only when it replaces the notion of thing or object, which was
conceptualized as having a form and a substance, decomposable into
primary elements, neatly isolatable in space, and subject to the external
laws of "nature." The new notion of system breaks away from the
classical ontology of object. We shall see that the object of the classical
science is a segment, an appearance, a simplified and one-dimensional
construct, mulilating and abstracting a complex reality that is rooted
simultaneously in physical organization and psychocultural organiza-
tion. We know the universality of the change that results from the
notion of system in contrast to the notion of object. But what we must
face is the radical ness of this change and the innovation which it might
bring about.

The Paradigm of System

The Macroconcept
The question of the system cannot be decided in the part-whole relation,
and the holistic paradigm forgets two very basic notions: interaction and
organization.
The whole-parts relations must be mediated by the notion of interac-
tions. It is incorrect to think that most systems consist of "parts" or
"elements," as this view is based on the notion of materials. On the
contrary, they consist of actions among complex units, which in turn
consist of interactions. An organism does not consist of celis, but of the
actions that take place among them. The set of these interactions
constitutes the organization of the system. "Organization" is the con-
cept that gives constructive coherence, rules, structures, etc. to the
interactions and regulates them. In fact, the concept of system is a
three-sided concept (Figure 1): system, which expresses the complex
unity and the phenomenal nature of the whole as well as the complex of
relations between the whole and the parts; interactions, which express
the set of relations, actions, and feedbacks that are interwoven in a

V
system; and organizaction, which expresses the formative character of

system organization

FIGURE I. interactions
 6. The Concept of System and the Paradigm of Complexity 127

the interactions by composing, maintaining, protecting, regulating,

directing, and renewing them, thus serving as a backbone to the concept
of system.
These three notions are inseparable. The absence of anyone of them
seriously mutilates the concept of system. It is a macroconcept. In
contrast, we can now realize that the method of understanding by
simplification, with which we have been educated, has provided us with
atomic concepts instead of molecular concepts, static and isolated
concepts of chemistry instead of organismic concepts that generate one
another in the feedback relation of their interdependence.
The concept of organization originated in various sciences based on
structure. But structure is a static concept that falls back more on the
concept of order in the sense of invariable rules rather than on that of
organization. It tends to reduce the process of the system to the
structure that generates the process. It does not take into account the
mutual feedback between the whole and the emergences.
In most of the physical and chemical systems and in all biological
systems, the "organization" is active. It is "organizaction." It includes
supply, storage, distribution, and control of energy as well as consump-
tion and dissipation of energy in its work. The organizaction is a product
of entropy (degradation of the system and of itself) and a product of
negentropy (regeneration of the system and of itself). One must concep-
tualize the relationship between entropy and negentropy in a complex
way. These two terms are not dichotomously opposed, but they are tied
to each other (Morin, 1977, pp. 291-296).
We must conceptualize the organization (a) as a perpetual reorganiza-
tion of a system that tends to disorganize itself; and (b) as a perpetual
reorganization of itself, i.e., autoreorganization.
In living creatures, such an organization is doubly polarized: gener-
ativity (genetic organization containing the so-called genotype program-
ming), and phenomenality (the organization of activities and behaviors
of the phenotype). In other words, it is an organization that is auto-
(geno-pheno )-reorganizing. Moreover, such an organization involves
exchange with the environment. The environment is itself a macro-
organization in the form of ecosystem (a biocenosis inside a biotope),
and provides organization in the form of vegetables and animals as food,
and the potential for organization in the form of information. It is a
living organization that is at the same time the organization of closure
(maintenance of integrity and autonomy) and that of openness (ex-
change with the environment or ecosystem). In this sense it is an
auto-eco-organization. Thus from the simplest unicellular organisms to
human societies, all organizations are at least auto-(geno-pheno )-eco-
reorganization.
128 Edgar Morin

Therefore the problem of organization cannot be reduced to a small

number of structural rules. From the beginning, the concept of biologi-
cal and a fortiori social organization is a super-macroconcept, which is a
part of the macroconcept system-interactions-organization.
Organization is a concept of a superior paradigmatic character. The
paradigm of the classical science considered that explanation consists of
reduction to order (laws, invariances, averages, etc.). We are not
replacing order with organization, but are combining both into the
systemic organizational principle as a nonreducible principle of explana-
tion. The new principle includes disorder at the same time. The organ-
ization creates order (by creating its own systemic determinism) and
disorder. On the one hand the systemic determinism can be flexible and
allow for zones of uncertainty, play, and freedom. On the other hand
the organizing work produces disorder (increase of entropy). In the
organizations, the presence and continual production of disorder (de-
gradation, decay) cannot be separated from the act of organizing. At
this level the paradigm of organization involves also a new way of
thinking. Nevertheless, the new way of explanation should not exclude
disorder nor obscure organization, but most always consider the com-
plexity of the relation shown in Figure 2.

FIGURE 2. disorder - - - - order

The new paradigm includes uncertainties and antagonisms by combin-

ing terms that imply mutually. The new spirit of science, which began
with Bohr, consists in improving the explanation, not by eliminating
uncertainties and contradictions, but by recognizing them. It improves
knowledge by clarifying the zone of shadow that exists in all knowledge,
in other words by improving the ignorance, because acknowledged, duly
noted, and well-researched ignorance becomes qualitatively different
from the ignorance that does not know itself.
Finally, we must break away from the distorting notion that miscon-
strues the concept of system or organization by eliminating the notion of
being or existence. I have attempted to show that the concept of
self-organization generates being and existence (Morin, 1977, pp. 211-
215). This is of crucial importance and it sets two types of thought in
contrast: the one that functions only by obscuring the concrete beings
and existences, by condemning them to all types of manipulation; the
 6. The Concept of System and the Paradigm of Complexity 129

other that functions only by showing the reality of existential beings,

which is of crucial importance as far as living creatures and human
beings are concerned.
Thus, we see that a new knowledge of the organization creates a new
organization of knowledge. A new paradigm, consisting of the inter-
relations among the elements as shown in Figure 3, replaces the old
reductionistic and atomistic paradigm whose only principle of explana-
tion is order.

V
being < system

order
organization'

~
>
interaction

disorder
existence

*auto-(geno-pheno )-eco-re-organization
FIGURE 3.

Instead of the old solitary key word, we now have a macroconcept,

not only of the molecular character but also with circular relations
among the terms, in other words a macroconcept of a recursive
character.

The Paradigm of Complexity

After the foregoing discussions, a fundamental term to clarify is com-
plexity. Most frequently what is considered complex is the complicated,
the entangled, or the fuzzy, which cannot be described because of the
astronomical number of measurements, operations, and computations
required to do so. But those who recognize the complexity of the
phenomena usually agree that it is possible to find a basic explanation,
by means of a small number of simple principles, that enables to make
almost infinitely varied combinations of some simple elements.
One might believe that an extremely complex discourse can be
explained beginning with the structural principles to combine phonemes
and words. One might believe to have found a key to biological
organization by showing the structure of the double articulation repre-
sented by combinations of four letters. Undoubtedly such explanations
can go quite far and enable us to understand both unity and diversity.
But they do not exhaust all of the problems of explanation. Structural
130 Edgar Morin

linguistics does not explain the meaning of the discourse. The genetic
algorithm explains neither the existence of the phenomena nor this
bundle of emerging qualities that we call "life." Molecular biology
explained the chemical machinery of life, but not life itself. It believed
that life is a mythological notion unworthy of science, and expelled life
from biology.
On the contrary, we must question the adequacy of all explanations
based on simplification of principles. Complexity is not a surface noise
of the real, but is the very principle of the real. The physical foundation
of what we call reality is not simple, but is complex. The atom is not
simple. The so-called elementary particle is not a simple primary unit. It
oscillates between the being and nonbeing, between wave and particle.
It may contain nonisolatable components (quarks). At the macro cosmic
level, the universe is no longer the ordered sphere that Laplace dreamed
of. It is at the same time dispersion and condensation, disintegration
and organization. Uncertainty, indeterminism, randomness, and contra-
dictions occur, not as residues to be eliminated by explanation, but as
non-eliminatable ingredients of our perception/cognition of the real. All
these ingredients, which ruined the principle of reductionistic explana-
tions, now nourish the complex explanations.
We must combine the notions that exclude one another in the
framework of the principle of simplification/reduction (Figure 4). The

~ ~
one multiple whole parts
~ ~
~
order/organization disorder
~
~
subject object
(observer) (observed system)
FIGURE 4. ~

establishment of mutual implications and therefore necessary conjunc-

tions among the notions that are disjoint in classical thinking is shown in
Figure 5.
It is complex because it introduces a complex causality, more speci-
fically an eco-auto-causality, where autocausality means recursive
causality in which the organizing process elaborates the products, ac-
 6. The Concept of System and the Paradigm of Complexity 131

system organization

\ ::interactiO';- I
FIGURE 5. .
eXIstence ~b'emg

tions, and effects necessary for its creation or regeneration, and where
autocausality needs causality from outside.

The Whole is Not a Collection

In opposition to the notion of general or specific theory of systems, I
propose a systemic paradigm that would have to be present in all
theories regardless of the field of their application.
Holism retrieves only a partial, unidimensional, and simplifying view
of the whole. It deforms the concept of totality to a concept, to which
other systemic concepts are reduced, whereas it ought to be a concept of
confluence. Holism arises from a paradigm of simplification, or a
reduction of the complexity to a key concept or a key category.
On the other hand, with regard to the new paradigm of system, Pascal
had already said: "I consider it impossible to know the parts without the
whole, not any more possible than knowing the whole without knowing
the specific parts." Out of this formulation by Pascal, we can make a
higher level understanding based on the constructive circularity of the
explanation of the whole by the parts and the parts by the whole, in
which the two explanations become complementary in the movement
associating them without losing their simultaneous and opposing charac-
teristics (Figure 6).

whole
~parts
FIGURE 6. ~
It is this active loop constitutes the description and the explanation.
At the same time, to keep a certain opposition and a certain play
between the two processes of explanation, which the simplistic logic
would regard as mutually excluding, is not vicious but productive.
Furthermore, the search for the explanation in the reverse direction
of one of the processes with regard to the other (parts ~ whole,
whole ~ parts) leads to the first introduction of the complexity at the
 132 Edgar Morin

paradigmatic level. This is because, as we shall see, we should not only

respect complexity at the level of phenomena but also at the level of the
principle, where it should be revealed in the first place.
At the same time, we must see the system not only as a global unity
(which is merely to replace the simple elementary unity of reductionism
with a macro-unity) but as a unitas multiplex, where antagonistic terms
are combined. The whole is a macro-unity, but the parts in it are neither
merged or confused. They have a double identity: an individual identity
that is not reducible to the whole, and a common identity of systemic
citizenship. Moreover, the atomic, biological, and social systems show
us that a system is not only a composition of unity from diversity, but
also that of diversity from unity. Examples are: Pauli's exclusion prin-
ciple, which creates electronic diversification around a nucleus; biologic-
al morphogenesis from an undifferentiated embryo to a complex adult
organism; societies that not only give a culture and a common identity to
diverse individuals, but also allows, by means of the culture, further
development of diversity. Here again, a loop is in operation between
two principles of explanation (Figure 7). The unifying thinking alone

~
one diverse
FIGURE 7. ~
becomes increasingly homogenizing and loses diversity. The dif-
ferentiating thinking becomes catalogue-like and loses unity. Here
again, it is not a matter of "dosage" or "equilibrium" between these two
processes of explanation. One has to integrate both in an active loop in
which:
(the diversity organizes the unity which organizeS)

The central problem is not that of maintaining relations of the type

whole/parts or one/many, but one must also focus on the complex
character of these relations that can be summarized as follows (for
details, see Morin, 1977, pp. 105-128):
The whole is more than the sum of the parts. This principle is
intuitively recognized at all macroscopic levels, but also emergences
occur that are of new qualities and properties for the parts.
The whole is less than the sum of the parts because the latter lose some
of their qualities or properties under the constraining effects of the
organization of the whole.
 6. The Concept of System and the Paradigm of Complexity 133

The whole is more than the whole because the whole acts on the parts
that act on the whole. In other words, the whole is more than a global
reality. It is an organizational dynamism.
It is in this framework that we must understand the being, the
existence, and the life as emerging global qualities. These key notions
are not of the primary qualities, root, or essence, but of the realities of
emergence. The being and the existence are emergences of the entire
process that loops on itself (Morin, 1977, especially pp. 210-216). The
life is a bundle of emerging qualities resulting from the process of
interaction and organization between the parts and the whole. The
emerging bundle acts back on the parts and on the partial and global
interactions and processes that produced it. Hence the complex princi-
ple of explanation: One must not reduce the phenomenal to the genera-
tive, nor the superstructure to the infrastructure. But the explanation
must try to understand the process, in which the products or final effects
generate their own new beginning. This process is named here "recur-
sive" (Figure 8).

~
generatIve
phenomenal
~

~
infrastructure superstructure
FIGURE 8. ~

The parts are at the same time more than and less than the parts. The
most remarkable emergences in a very complex system such as a human
society occur not only at the level of the whole (the society), but also at
the level of the individuals. Awareness of the self occurs only in
individuals.
In this sense, the parts may become more than the whole. "The most
profitable control system of the parts should not exclude the bankruptcy
of the whole" (Beer, 1960). The "progress" is not necessarily the
formation of increasing totalities, but it may be in the freedom and
independence of small units. The richness of the universe is not in the
expanding totality, but in the reflective, deviating, and pheripheral
small units. This fact, noticed by Gunther (1962) and Brown (1962),
recaptures a statement by Pascal (1670): "When the universe would
134 Edgar Morin

crush the humankind, the human being would be nobler than the
universe, because he knows that he dies, and the universe knows
nothing of the advantage it has over the humankind."
The whole is less than the whole. In everything there are zones of
shadow, mutual ignorance, gaps and cracks between the repressed and
the expressed, the immersed and the emerging, the generative, and the
phenomenal. There are black holes in the core of all biological and
especially anthroposocial entities. It is not only the individual who is
ignorant and unaware of the social totality, but this social totality is also
ignorant and unaware of the dreams, aspirations, thoughts, love, and
hatred of the individuals, and billions of cells in each individual are also
unaware of them. If we include these concepts of black holes, zones of
shadows and gaps as well as the mutual ignorance in the systemic
paradigm itself, this latter leads by itself to modern theories of the
anthropological unconscious (Freud) and the sociological inconscious
(Marx).
The whole is insufficient, which follows from the above.
The whole is uncertain. We shall see later that there is no sure way to
isolate or close up a system among systems of systems of systems, to
which it is associated and in which it is inserted. It is also uncertain in the
sense of poly totalities in the living universe, in which every polytotality
can be thought of at the same time as a whole and as a part. As far as
Homo sapiens is concerned, what is the system, the society, the species,
the individual?
The whole is conflictual. I have attempted to show (Morin, 1977, pp.
118-122, 217-224) that in all systems there are forces antagonistic to
their perpetuation. These forces are activated, cancelled, constantly
controlled, suppressed (by negative feedback), or used constructively.
For example, in astronomy, implosions and explosions constitute spon-
taneous regulating of the organizing type. The living organizing forma-
tion is intelligible only in terms of permanent disorganization, which
destroys molecules and cells that in turn are continuously replaced with
new ones. At the level of human societies, we must understand systemi-
cally the concept of Montesquieu, according to which social conflicts
have been sources of not only the fall but also the greatness of the
Roman empire, as well as Marx's theory that ties the concept of society
organized in classes to the concept of antagonism between the classes.
Thus we must base the concept of system on a non totalitarian and
nonhierarchical concept of the whole, on a complex concept of unitas
multiplex, open to poly totalities. This preliminary paradigmatic view is
of great practical and political importance. The paradigm of holistic
simplification leads to a neototalitarian functionalism and adequately
merges with all modern forms of totalitarianism. In any case it leads to
manipulation of units in the name of the whole. In contrast, the logic of
 6. The Concept of System and the Paradigm of Complexity 135

the paradigm of complexity leads not only to a "truer" knowledge, but

also to the search for a practice and a policy, both of which are complex.
I shall elaborate this point later.

The Psychophysical Character of the Systemic Paradigm

The old paradigm of simplification made it necessary for us to choose
between two orders of the systemic reality: (a) a system is a measurable
physical category that presents itself to the perception of the observer
who must endeavor to make an accurate description of it; or (b) a
system is a mental category, or mentally constructed heuristic and/or
pragmatic model used for controlling, understanding, or simulating the
phenomena.
The new complex conceptualization of the system cannot be
accommodated in either of the alternatives. The system is a concept
with double entry:
Physis ;;:::: psyche

The concept has physical feet and a psychic head. It is:

physical psychic
by formation and by existence by distinction and isolation, by the
(interaction, combination of choice of successively nesting concepts
ecological circumstances, energetic (subsystem, system, suprasytem,
and thermodynamic conditions and ecosystem) .
operations). Even a system of thinking
has a physical component
(biochemicophysical processes related
to cerebral activities).

from which we obtain:

a principle of the art (of diagnostics);
a principle of critical reflection
(on the relativity of the notions of the
boundaries of systems);
a principle of uncertainty.

Hence the indivisibility of the psychic/physical character of the sys-

tem, and the indivisibility of the relation subject/object or observer/
observed, and consequently the necessity to include the observer in the
observation instead of excluding him from it.
It is necessary to elaborate a metasystem of comprehension in which
the system of observation/perception/concept-formation has to be
observed, perceived, and conceptualized in the observed system which
is observing itself. Therefore we have a chain of consequences that leads
us to complexify our mode of perceiving and conceptualizing the world
136 Edgar Morin

of phenomena. We must proceed to a paradigmatic and epistemological

reform that is more important than what we saw until now, because the
articulation between the knowledge of organization and the organiz-
ation of knowledge calls for a reorganization of knowledge by intro-
ducing a second-degree reflectivity, i.e., a knowledge of the knowledge.
At the same time, the dichotomy between physical sciences and
mental sciences, between the sciences of nature and the science of the
culture, between the biophysical sciences and the anthroposocial sci-
ences seems to us to be a premature mutilation and an obstacle to all
serious knowledge. If the desire to articulate these disjoint sciences
seems grotesque, then the acceptance of this disjunction becomes even
more grotesque.
If we are still incapable of making the new articulation needed, we
must at least put the following side-by-side:
the observer the observed system
the subject the object
the culture (which produces a physical the physis (which produces biological
science) organization that produces
anthroposocial organization, hence
culture)
The operation of distinguishing, which is basic to all cognitive acts,
becomes complex. It appears to be a result of a transaction between the
observer and the observed world, during which one of the partners can
very well fool the other. The way of making cognitive distinction is
embedded in a culture which provides paradigms permitting and requir-
ing the distinction. It has therefore an ideological characteristic among
other characteristics. Even though one should not reduce science to
ideology (in other words to see science merely as an ideological product
of the society), one must he aware that an ideological component enters
any scientific knowledge. We cannot afford to skip examining the
ideological aspect of the scientific knowledge including one's own know-
ledge. This applies also to thoese who believe that they have the true
science and denounce the ideology of others.

The Systemized Theories

The system is a generic concept better than a general concept. It is
generic in a new mode of thinking useful in a general way. On the other
hand a general theory of systems is not needed if one looks for general
applicability. The systemic organizational dimension ought to be pre-
 6. The Concept of System and the Paradigm of Complexity 137

sent in all theories on physical, biological, anthroposociological, and

conceptual universe. These theories, if they were branches of a general
theory of systems, would reduce the diverse phenomena captured by
human senses to one of the systemic dimensions without making con-
nections among many dimensions. But each of the phenomena has its
physics, chemistry, thermodynamics, nature, organization, being, and
existence. We must therefore differentiate between theories relating to
types of phenomena.
We must add that the general systems theory as applied to biological
and social systems, but based only on the notion of open system, is
completely insufficient. Therefore it seems necessary to return to the
physical, biological, and anthroposociological theories, to go deeper in
their systemic-organizational dimension, and to articulate them (a) in
the key concepts of organization, and (b) in a way of thinking capable of
operating the loops among terms that are at the same time com-
plementary, concurrent, and antagonistic.
Otherwise, one falls back into the vices of reduction, homogeniza-
tion, and abstraction, which the theory of systems claims to remedy.

Conclusions
1. The system is not the masterword for the totality. It is a root word
for complexity.
2. One must elevate the concept of system from the theoretical level to
the paradigmatic level. The same can be said of the concept of
cybernetics of machines.
3. It is not a matter of making a general theory including the atom, the
molecule, the star, the cell, the organization, the artifact, and the
society. The problem is to reconsider them all in a richer way in the
light of the systemic/organizational complexity, including our own
reality.
4. Under the domination of the paradigm of simplification/disjunction,
the being, the existence, and the life dissolve into systemic abstrac-
tion which perpetuates all the abstractions obscuring the richness of
the real and resulting in unrestrained manipulation. On the con-
trary, under the effect of the development of the complex concept
of system/organization, the being, the existence, and the life sud-
denlyemerge.
5. As long as the systemic idea remains theoretical, it does not change
the paradigm of disjunction/simplification which it believes to have
overcome by overcoming the reductionistic atomization. However,
its "holism" becomes reductionistic by reducing everything to the
138 Edgar Morin

whole. It is only at the paradigmatic level where its complexiety

truly fiorishes that the systemic can open itself to a new complex
organization of thought and action.
6. A new rationality shows a glimpse of itself. The old rationality
sought only to harvest order in nature. But it fished fishbones, not
fish. The new rationality enables one to perceive fish and ocean by
allowing to conceptualize organization and existence; in other
words, it enables one to perceive that which cannot be fished.
7. The old theory organized by orders from top down. But one must
begin with organization, in other words with the play of interactions
between the involved parts and the whole. In this sense, "to
organize" must replace "to order." The more complex the organ-
ization, the more it contains disorder called "freedom."
8. Organization is not institution. It is a regenerative and generative
activity always present at all levels, based on computation, elabora-
tion of strategies, communication, and dialogue.
9. The systemic paradigm requires us to master, not the nature, but
our own art of mastering. This opens to us forms of action contain-
ing the awareness and control of the self.
10. Such a principle opens a way to a responsible, liberal, radically
different, and community-oriented practice. Each of these terms is
transformed by its interaction with the others. This practice leads to
a rediscovery of the problem of wisdom and the necessity to build
our wisdom. The search for this wisdom is in this sense a search for
a way to overcome the fissure that developed in the Western world
between the universe of meditation and that of social practice.

References
Beer, Stafford. (1960). Below the twilight arch. In General Systems Yearbook.
Brown, Spencer. (1972). The Laws of Form. New York: Bantam Books.
Gunther, Gottard. (1962). Cybernetical ontology and transformational opera-
tion. In Yovits, Marshall C. and Cameron, Scott (Eds), Self-organizing
systems, New York: Pergamon.
Morin, Edgar. (1977). La Methode. Paris: Seuil.
Pascal, Blaise. (1670). Pensees. (New edition 1976, Paris: Mercure de France).