Content uploaded by Gerard Jagers op Akkerhuis
Author content
All content in this area was uploaded by Gerard Jagers op Akkerhuis
Content may be subject to copyright.
Found Sci (2010) 15:245–262
DOI 10.1007/s10699-010-9177-8
Towards a Hierarchical Definition of Life, the Organism,
and Death
Gerard A. J. M. Jagers op Akkerhuis
Published online: 9 June 2010
© Springer Science+Business Media B.V. 2010
Abstract Despite hundreds of definitions, no consensus exists on a definition of life or on
the closely related and problematic definitions of the organism and death. These problems
retard practical and theoretical development in, for example, exobiology, artificial life, biol-
ogy and evolution. This paper suggests improving this situation by basing definitions on a
theory of a generalized particle hierarchy. This theory uses the common denominator of the
“operator” for a unified ranking of both particles and organisms, from elementary particles
to animals with brains. Accordingly, this ranking is called “the operator hierarchy”. This
hierarchy allows life to be defined as: matter with the configuration of an operator, and that
possesses a complexity equal to, or even higher than the cellular operator. Living is then syn-
onymous with the dynamics of such operators and the word organism refers to a select group
of operators that fit the definition of life. The minimum condition defining an organism is its
existence as an operator, construction thus being more essential than metabolism, growth or
reproduction. In the operator hierarchy, every organism is associated with a specific closure,
for example, the nucleus in eukaryotes. This allows death to be defined as: the state in which
an organism has lost its closure following irreversible deterioration of its organization. The
generality of the operator hierarchy also offers a context to discuss “life as we do not know
it”. T he paper ends with testing the definition’s practical value with a range of examples.
Keywords Artificial life · Biology · Evolution · Exobiology · Natural sciences ·
Particle hierarchy · Philosophy · Big History · Astrobiology
G. A. J. M. Jagers op Akkerhuis (
B
)
Alterra, Centre for Ecosystem Studies, Wageningen, UR, P.O. Box 47,
6700 AA Wageningen, The Netherlands
e-mail: gerard.jagers@wur.nl
URL: www.hypercycle.nl
123
246 G. A. J. M. Jagers op Akkerhuis
1 Introduction
In a chronological overview of developments, Popa (2003) presents about 100 definitions of
life, meanwhile demonstrating that no consensus exists. Many classical definitions include
long lists of properties, such as program, improvisation, compartmentalization, energy, regen-
eration, adaptability and seclusion (Koshland 2002) or adaptation, homeostasis, organiza-
tion, growth, behavior and reproduction (Wikipedia: Life). Most properties in such lists
are facultative; it is still possible to consider an organism a form of life when it does
not grow, reproduce, show behavior, etc. The inclusion of facultative aspects is a source
of lasting difficulty in reaching consensus on a definition of life. Because of the seem-
ing hopelessness of the situation, certain scientists have adopted a pragmatic/pessimistic
viewpoint. Emmeche (1997) christened this viewpoint the “standard view on the defini-
tion of life”. He suggests that life cannot be defined, that its definition is not important
for biology, that only living processes may be defined and that life is so complex that it
cannot be reduced to physics. Others have warned that a comprehensive definition of life is
too general and of little scientific use (e.g. van der Steen 1997).
In their search for a definition, other scientists have focused on properties that are abso-
lutely necessary to consider an entity life. In this context Maturana and Varela (1980, p. 78)
have proposed the concept of autopoiesis (which means “self making”). They use the fol-
lowing definition: “An autopoietic machine is a machine organized (defined as a unity) as a
network of processes of production (transformation and destruction) of components which:
(1) through their interactions and transformations continuously regenerate and realize the
network of processes (relations) that produced them; and (2) constitute it (the machine) as
a concrete unity in space in which they (the components) exist by specifying the topological
domain of its realization as such a network.” Special about the autopoietic process is, that it
is “closed in the sense that it is entirely specified by itself (Varela 1979 p. 25)”.
The concept of autopoiesis has increasingly become a source of inspiration for discussions
in the artificial life community about how to define life (Bullock et al. 2008). Reducing the
number of obligatory traits defining life to just one, autopoiesis is a rather abstract concept.
People have sought, therefore, to describe some of the processes that underlie autopoiesis
more specifically. An example of such a description is a triad of properties defining cellular
life: container (cell membrane), metabolism (autocatalysis) and genetic program (e.g. Bedau
2007).
These descriptions, however, have not resulted in a consensus definition of life. This
has led Cleland and Chyba (2002, 2007) to suggest that a broader context, a “theory of
life”, is required. In line with a broader framework, life may be regarded as a special real-
ization of the evolution of material complexity. According to Munson and York (2003),
considering life in a general evolutionary context requires arranging “all of the phenom-
ena of nature in a more or less linear, continuous sequence of classes and then to describe
events occurring in the class of more complex phenomena in terms of events in the clas-
ses of less complex phenomena. “An important property of such a hierarchy would be that
“… an increase in complexity is coupled with the emergence of new characteristics …
suggesting that the hierarchical arrangement of nature and the sciences is correlated with
the temporal order of evolution”. Similar views for integrating material complexity and the
evolution of life can be found, for example, in the work of Teilhard de Chardin (1966, 1969),
von Bertalanffy (1968), Pagels (1985), Maynard Smith and Szathmáry (1995, 1999), and
Kurzweil (1999).
In contribution to these discussions, the present author has published an evolution
hierarchy for all “particles”. The latter hierarchy uses the generic word “operator” to address
123
Towards a Hierarchical Definition of Life, the Organism, and Death 247
Fig. 1 Using the operator
hierarchy to d efine life a nd
organisms. Arrows indicate how
closures create operators (more
information can be found in
Jagers op Akkerhuis (2008), and
the author’s website www.
hypercycle.nl)
MEMIC
ORGANISMS
CELLULAR
ORGANISMS
OPERATORS WITH A STATE OF
MATTER REPRESENTING LIFE
HADRONS
MOLECULES
ATOMS
CELL
EU-
KARYOTE
CELL
EU-
KARYOTE
MULTI-
CELLULAR
MULTI-
CELLULAR
(HARD-
WIRED)
MEMON
THE OPERATOR HIERARCHY
OPERATORS
REPRESENTING THE ‘DEAD’
STATE OF MATTER
cell-based
operators
atom-based
operators
hadrons
memon-based
operators
both physical (e.g. quark, atom, and molecule) and biological particles (e.g. prokaryote cell,
eukaryote cell, and multicellular). The word operator emphasizes the autonomous activity of
the entities involved, which “operate” in a given environment without losing their individ-
ual organization. The hierarchical ranking of all operators is called the “operator hierarchy”
(see Fig. 1; Jagers op Akkerhuis and van Straalen 1999; Jagers op Akkerhuis 2001, 2008
and the author’s website www.hypercycle.nl). Because the operator hierarchy is important
for the definition of life proposed below, the outlines of this theory are summarized in the
following lines.
The operator hierarchy ranks operators according to the occurrence of a circular pattern,
such as that which connects the beginning and end of a process or structure. Circular-
ity causes a closed organizational state, also referred to as “closure” (for discussions of
closure see, for example, Heylighen 1990; Chandler and Van de Vijver 2000). Because
closure causes a discrete “quantum” of organization (e.g. Turchin 1977, 1995; Heylighen
1991), the operator becomes an “individual entity”, a “whole” or a “particle”, while
still retaining its construction of smaller elements. Closure thus defines the operator’s
complexity level and sequential closures imply a higher complexity level. An operator’s
closure is the cause of its existence and typical for its complexity. This implies that com-
plexity is not measured in terms of the number of genes, functional traits or organs of
an organism, but in a very abstract way, in terms of the number of closures. Upon los-
ing its closure, the organization of the operator falls back to that of the preceding opera-
tor. The actual shape of a closure can differ. Biological examples of closure are the cell
membrane and the circle of catalytic reactions allowing the cell to maintain its chem-
ical machinery. It is essential for a strict ranking that a lower-level and a higher-level
operator always differ by exactly one closure level. The single closure (eukaryotic cell)
or a parallel pair of closures (autocatalysis plus membrane of the cell) that define the
next level are referred to as “first-next possible closure(s)”. A consequent use of first-
next possible closures allows physical and biological operators to be ranked according
123
248 G. A. J. M. Jagers op Akkerhuis
to the “operator hierarchy” (Fig. 1). The operator hierarchy includes quarks, hadrons,
atoms, molecules, prokaryotic cells, eukaryotic cells, multicellulars (e.g. plants, fungi) and
“animals”, the latter representing an example of the operators that possess a neural network
with interface and that are called “memons” in the operator hierarchy.
Due to its focus on closure, the operator hierarchy represents an idealization because it
excludes potential transition states in between two closures. For example, several hundreds
of metal atoms may be required before a functional Fermi sea transforms a collection of
single atoms into a metal grid. Also, the emergence of multicellularity (discussed in detail in
§3 below) may require a colonial, multicellular state in between the single cell and the mul-
ticellular operator. The above shows that transition states form natural intermediate phases
in the emergence of closures. The operator hierarchy does not include these transition states,
however, because its hierarchical ranking is exclusively based on entities that already show
first-next possible closure.
The main reason for writing this paper, and adding yet another definition of life to the
listings, is that the operator hierarchy offers several advantages in solving definition prob-
lems. First, the definitions of the operators are generally applicable because they focus on the
essences of organization. For example, demanding autocatalysis leaves open which specific
catalysts will perform the process. Second, the use of first-next possible closures ensures
a critical filtering of only obligatory properties from property lists. Finally, the use of the
operator hierarchy makes it easy to develop a hierarchy-based definition of life. In other
words, the operator hierarchy offers a novel path for structuring and simplifying discussions
about which entities are life.
The following paragraphs discuss different aspects of existing definitions of life and exam-
ine new ways to define the organism, living and death. At the end, a test of the practical value
of the present definitions for the solving of a range of classical problems, such as a virus, a
flame, a car, a mule and a mitochondrion, will be presented.
2 Defining Life and the Organism
Before discussing the use of the operator hierarchy for defining life, living and the organ-
ism, it is important to note that when talking about definitions, care should be taken that “a
definition is a series of superimposed language filters and only the definiendum (the term to
be defined) can penetrate it”(Oliver and Perry 2006). Problems may arise when the words
used for the definiendum and for the filter have a broad meaning or have different meanings
in different contexts. It is thus useful to elaborate on the current context for “life” before
continuing.
“Life” has different meanings in different contexts. For example, people refer to the period
between birth and death as their life (this is the best day of my life) even though lifetime would
be more correct. In addition, the experience of “being alive”, or “living”, also carries the label
of life (to have a good life). Other uses of life holistically refer to the importance of selective
interactions in ecosystems that over generations lead to better-adapted life forms (the evo-
lution of life). Ruiz-Mirazo et al. (2004) have proposed a definition of the latter type. They
statethatlifeis“a complex collective network made out of self-reproducing autonomous
agents whose basic organization is instructed by material records generated through the
evolutionary-historical process of that collective network”. In philosophy, life is sometimes
considered a graded concept for being because all what is, is alive in the measure wherein it
is (Jeuken 1975). Due to the contextual dependence of these and other interpretations, it is
123
Towards a Hierarchical Definition of Life, the Organism, and Death 249
improbable that a general definition of life can be constructed. van der Steen (1997) indicates
that even if such an overly general definition existed, it would probably be difficult to applie
it to specific situations.
To avoid problems with generality and multiple interpretations of concepts, the present
study adopts a limited viewpoint, presuming a one-to-one relationship between a definition
of life and a specific material complexity. In this context, life is an abstract group property
shared by certain configurations of matter.
The operator hierarchy offers a context for a general matter-based definition of life. Focus-
ing on all operators showing a complexity that exceeds a certain minimum level, the hierarchy
suggests a definition of life sensu lato as: matter with the configuration of an operator, and
that possesses a complexity equal to or even higher than the cellular operator. Only the pro-
karyote cell, the eukaryote cell, the prokaryote and eukaryote multicellular, the hardwired
memon and the potential higher-level operators fit this definition (Fig. 1). In addition to this
general definition, various specific definitions are possible by focusing on operators that
lay between a lower and an upper closure level. An example of a specific definition is one
describing cellular life (e.g. algae, plants and fungi) as: matter showing the configuration of
an operator, and that possesses a minimum complexity of the cellular operator and the maxi-
mum complexity of a multicellular operator. The latter includes only the cell, the eukaryotic
cell, the prokaryotic and the eukaryotic multicellular. It is possible to choose any of these
approaches for defining living as: the dynamics of an operator that satisfies the definition of
life.
The above approach results in a strictly individual based definition of life as a group
property of certain operators. This definition has the advantage, that it offers a solid basis
for defining the creation of offspring. Subsequently, the evolution of life can be dealt with
as an emergent process occurring in any system with interactions between individual living
entities that lead to differential survival of variable offspring, produced either without or with
recombination of parental information.
The organism is the key ontological unit of biology (Etxeberria 2004; Korzeniewski 2004)
and is also referred to as a “living individual”. Understanding the latter requires insight into
what is “living”, and what is an “individual”. By defining “living” as the dynamics of those
operators that satisfy the definition of life, the operator hierarchy uses operators instead of
individuals because operators define a being or an individual more strictly than the Latin
concept of individuum. The word individuum stands for an “indivisible physical unit repre-
senting a single entity”. This definition leaves a great deal of room for choice of the elements
that form the physical unit and for the rules that determine indivisibility. These indetermi-
nacies may be the reason for the discussion about whether certain life forms are organisms.
Townsend et al. (2008) use the phrase “unitary organism” to indicate the individual organ-
ism. However, certain jellyfish, for example, the Portuguese Man O” War (Physalia physalis),
look like individuals, but consist of differentiated individuals, each with its proper neural
network (e.g. Tinbergen 1946). In the operator hierarchy, the latter jellyfish are colonies, not
organisms, because each contributing individual has its proper neural network as its highest
emergent property, and the colony still lacks a recurrent interaction of the neural interfaces
of the individuals.
The operator hierarchy now suggests a way to create congruency between the definition
of life and the definition of the organism by accepting as organisms only entities that fit the
operator-based definition of life. For example, using the general definition of life, only the
cells, the eukaryotic cells, the prokaryotic and eukaryotic multicellulars and the memons are
organisms.
123
250 G. A. J. M. Jagers op Akkerhuis
3LevelsofLife
a. The cell. The most important properties of the cell are the autocatalytic set of enzymes
and the membrane. The autocatalytic set shows reproduction as a set. Every molecule
in the set catalyzes a reaction that produces some other molecule in the set until any
last reaction product closes the cycle. In different ways, reproduction as a set is part of
various theories about the origin of life (e.g. Rosen 1958, 1973, 1991; Eigen 1971; Gánti
1971; Eigen and Schuster 1979; Kauffman 1986, 1993; Bro 1997; Kunin 2000; Hazen
2001; Martin and Russel 2002; Hengeveld and Fedonkin 2007).
Autocatalysis demands that a cell can potentially autonomously sustain its catalytic
closure. Accordingly, if a cell allocates a part of its autocatalytic closure to another
cell, the cell is no longer an operator. An example of the latter is the mitochondrion. It
is generally accepted that mitochondria started the interaction with their host cells as
autonomous endosymbiontic α-proteobacteria. Over many generations, these bacteria
transferred more than 90% of their catalytic control to their host (Allen 1993; Berg and
Kurland 2000; Searcy 2003; Capps et al. 2003; Lane 2005). The loss of the potential of
autocatalysis implies that mitochondria have become a special kind of organelle.
In addition to autocatalysis, the operator hierarchy demands an interface because a set
of autocatalytic enzymes only gains the physical individuality that allows its maintenance
when it functions in a limited space, the limits being part of the system. The integration
of autocatalysis and the membrane is part of various important theories, for example, the
theories of autopoiesis (Varela 1979) and of interactors (Hull 1981).
b. The eukaryote cell. A single cell has two dimensions for creating a next closure. One is
to create cooperation between cells, which leads to multicellularity. The other is to create
an additional closure mediating the hypercyclic functioning of the cell in the form of
the nucleus. Interestingly, it is quite likely that the most important complexity boundary
in cell biology, that between prokaryotic and eukaryotic cells, thanks its existence to
the energy boost and genetic enrichment offered by endosymbionts. With respect to the
emergence of eukaryotic cells, theories roughly divide along two major lines depending
on whether the nucleus or the endosymbionts emerged first. In addition to other aspects,
support for the nucleus-first hypothesis comes from allegedly primitive eukaryotes that
show a nucleus without harboring endosymbionts. Genetic analyses (Rivera et al. 1998)
and observations of endosymbiont traces (Clark and Roger 1995), however, suggest
that the “primitive eukaryotes” are recent developments that lost their endosymbionts in
a process of evolutionary specialization. The endosymbiont hypothesis advocates that
a merger between a methanogenic bacterium that was member of the archaea and an
α-proteobacterial endosymbiont created the eukaryotic cell (Martin and Russel 2002).
Subsequent transmission of genes for membrane creation from the endosymbiont to the
host allowed it to produce membranes that formed the basis for the engulfment of the
nucleus. Whatever the actual path taken by evolution, the operator hierarchy focuses on
the occurrence of closure involving both structural and functional aspects of the host cell,
resulting in an internal interface for the autocatalytic set and the mediation of its function-
ing. Even though endosymbionts may become obligatorily integrated in the functioning
of their host cell by the transfer of part of their genetic regulation to the host cell, they do
not mediate the functioning of the autocatalytic set of the host nor form an interface for
its functioning. For this reason the operator hierarchy does not regard endosymbiosis,
but the nucleus as the relevant closure that defines the limit between prokaryotes and
eukaryotes.
123
Towards a Hierarchical Definition of Life, the Organism, and Death 251
c. The multicellular. When does a group of cells become a multicellular operator and,
according to the above definition, an organism? In the operator hierarchy, multicellularity
involves a structural and a functional component represented by structural attachment of
cells and an obligatory recurrent pattern of functional interactions between them. As such,
it is possible to define a multicellular operator (a multicellular organism sensu stricto)
as: a construction of mutually adhering cells showing obligatorily recurrent interactions
based on the same interaction type, that has the potential of maintaining its functioning
as a unit and that does not show memic structure.
Multicellularity has developed independently in many branches of the phylogenetic
tree (reviews by, for example, Bonner 1998; Kaiser 2001; Grosberg and Strathmann
2007) presumably because it is associated with a range of evolutionary advantages. Mul-
ticellularity increases mobility and access to resources, and reduces predation, and finally
yet importantly, the cells in genetically uniform multicellulars share the same genes and
do not have to compete with each other for reproduction. Willensdorfer (2008) indicates
that the alleviation of reproductive competition allows for a division of labor because
“cells can specialize on non-reproductive (somatic) tasks and peacefully die since their
genes are passed on by genetically identical reproductive cells which benefited from the
somatic functions”.
In some cases a multicellular organism results from the aggregation of individually
dwelling unicellulars (for example, true slime molds, Ciliates and Myxobacteria). More
generally, a multicellular organism develops when daughter cells cohere after cell divi-
sion. A simple, temporary form of multicellular life is present in slime molds. Here,
genetically-different, individually-dwelling cells aggregate and bind using membrane
proteins to form a colonial state in which the cells intercellularly communicate by dif-
fusion. At a certain moment, obligatory interactions between cells lead to the formation
of irreversible cell differentiation producing a reproductive structure. During this state,
the slime mold cells are temporarily a multicellular organism.
With the evolutionary development of plasma connections, advanced multicellular life
became possible. Plasma connections allow efficient and rapid intercellular communica-
tion, involving electrical signals, chemical signals and nutrient transport (Mackie et al.
1984; Peracchia and Benos 2000; Nicholson 2003; Panchin 2005). Plasma connections
have evolved in several lineages of multicellulars. Plasma connections between animal
cells depend on gap junctions, between plant cells on plasmodesmata, in blue-green algae
on microdesmata, and in certain fungi or in developing insect eggs on incomplete cell
walls. The evolution of gap junctions some 700 million years ago coincided with an
explosion of multicellular life forms.
Multicellular organisms may go through life stages that are not multicellular. For exam-
ple, sexual reproduction involves single-celled egg and semen. Furthermore, during the
two-, four- and early eight-cell stages most vertebrate embryos have loosely attached cells
without obligatory dependency. Accordingly, they represent a colony. When separated
from the colony, the cells show a normal development. Early separation of embryonic
cells is the reason why identical twins exist. Embryo cells in the early stages can even
mix with another embryo’s cells of the same age and develop into a normally function-
ing organism, called a chimera, in which some organs and tissues belong to a different
genotype than others. A definition of life should, therefore, respect that an organism’s
cells may differ in genotype. From the late eight-cell stage, the development of gap-junc-
tions marks the emergence of regulation as a unit, which makes the cellular colony a
multicellular.
123
252 G. A. J. M. Jagers op Akkerhuis
The realization of a multicellular’s potential for maintenance depends on prevailing
conditions. For example, a tree twig that is stuck in the ground may become a tree again
if the weather is not too warm, too cold, or too dry, etc. and if the twig has the genetic
potential for regeneration and is large enough, in good condition, etc.. Whether the twig
is an organism depends on its potential to show all dynamics required for being a multi-
cellular operator. This potential is in principle gene-based, but it depends on the condition
of the phenotype and the environment for its realization.
Sometimes two multicellular organisms show symbiosis, such as plants living in close
association with mycorrhiza fungi in their roots. As the fungus and the plant already are
multicellular on forehand, a plant with mycorrhiza represents an interaction between two
multicellular o rganisms.
d. The memon. Attempts to define life frequently focus on the typical properties of the
first cell. The underlying assumption may be that all organisms consist of cells and that,
for this reason, the definition of the living properties of cells will automatically cover
other, more complex organizations. According to the operator hierarchy, this reasoning
is incomplete because, with respect to artificial intelligence, it unsatisfactorily excludes
technical life apriori. The reason is that the fundamental construction of the brain is not
principally different when built from cellular n eurons, technical neurons (small hardware
acting as a neuron) or programmed neurons (virtual devices modeled to act as neurons).
Even though all organisms on earth currently consist of cells or show neural networks
that consist of cells, the fact that technical memons may, one day, have a brain structure
similar to cellular memons implies that a general definition of life must consider the
possibility of technical memons.
Memons show a neuron network and a sensory interface. The basic neuron-units have
been named categorizing and learning modules or CALMs and allow for a recurrent
network of CALMs (Murre et al. 1992; Happel 1997). The interface includes sensors
that allow the memon to perceive its body and environment, and effectors that allow it to
move the cellular vehicle it resides in. The interface and vehicle co-evolved during the
evolution of neural networks. In principle, it is possible to construct a functional memon
from any kind of technical hardware that provides the required neural architecture. This
is the reason that the study of neural networks in biology shows a fundamental overlap
with research on technical artificial intelligence. The recognition that memons show a
recurrent network of CALMs surrounded by an interface allows Siamese twins with sep-
arate brains to be classified as two memons sharing the same vehicle and showing in this
vehicle a partial overlap of their interfaces.
4 No Life, no Reproduction
According to some authors (e.g. the von Neumann and Burks 1966) reproduction is a pre-req-
uisite for life. Like the chicken and the egg problem, it can also be said that life is a pre-requisite
for reproduction. Clearly, any decision on this matter critically depends on the context that
is used to define life. If the operator hierarchy is used, the least complex life form is the
prokaryotic cellular operator. Two arguments currently suggest that life is a pre-requisite for
reproduction. The first states that even though all other organisms originate from the first cell
by reproduction, the first cell itself had an inorganic origin. The emergence of the fi rst cell thus
shows that life does not obligatorily result from reproduction. The second argument posits that
organisms do not need to show reproduction, i.e., producing offspring, to comply with the
operator-based definition of life; The operator-based definition demands that organisms show
123
Towards a Hierarchical Definition of Life, the Organism, and Death 253
two closures: autocatalysis and a membrane. Autocatalysis can be regarded as reproduction
without creating offspring. As Jagers op Akkerhuis (2001) pointed out, a utocatalysis implies
that a cell autonomously creates a structural copy of its information, a process that is called
“structural (auto-) copying of information”. Before answering the question of whether the
structural (auto-) copying of the cell’s information means that it must reproduce, it is impor-
tant to detail the concept of information. For the latter, I suggest applying Checkland and
Scholes (1990) definition of information to the autocatalytic set. These authors have defined
information as data with a meaning in a context. In line with this reasoning, Kauffman (1993)
proposed that, by selecting the autocatalytic process as the context, every catalytic molecule
becomes a data-unit with a catalytic meaning (the “purpose” mentioned by Kauffman 1993,
p. 388) and represents a part of the information of the autocatalytic process. Following one
round of autocatalysis, or more rounds to account for the loss of enzymes over time, the cell
contains copies of all of its information. At that moment, it has autonomously performed
structural copying of information and fulfills all the requirements of the operator hierarchy,
even when it does not produce an offspring. Based on this reasoning, the capacity of autocata-
lytic maintenance is an obligatory requirement for cellular life and reproduction is a possible
consequence.
The above implies that it is not relevant for a general definition of life to distinguish
between life forms with or without replication, as Ruiz-Mirazo et al. (2004) has suggested.
The latter authors distinguish “proto-life stages” that do not show a phenotype-genotype
decoupling (soma with genes) from “real life” with genes. In line with the operator hierarchy
based definitions, Morales (1998) warns that “if reproduction is required: This is a trou-
bling development, because it means that we could not tell whether something is alive unless
we also know that it is the product of Darwinian evolution.” The operator-based definition
considers life as a prerequisite for reproduction instead of reproduction as a prerequisite
for life. Consequently, worker bees, mules, infertile individuals and other non-reproducing
organisms and/or phenotypes are life. This point of view also solves problems that may arise
when demanding that memons be able to reproduce as a prerequisite for recognizing them
as life forms. In fact, none of the cellular memons living today shows reproduction, at least
not reproduction of their neural network structure determining their closure. The things they
pass on during reproduction are the genes of their cells, allowing the development of a mul-
ticellular organism with a neural network, capable of learning but devoid of inherited neural
information other than reflexes.
5 Life Holding its Breath
The above chapter shows that reproduction is not a prerequisite of life but a possible
consequence of it. Going one step further, it can also be concluded that metabolism is not
a prerequisite for life. Many taxa such as bacteria, protozoa, plants, invertebrates and ver-
tebrates have developmental stages showing natural inactivity (seeds, spores) or reversible
inactivation when submitted to desiccation, frost, oxygen depletion, etc. The inactive state
carries the name of anabiosis, after the process of coming to life again (for a review of “via-
ble lifelessness” concepts, see Keilin 1959). Another type of reversible inactivity showing
marked similarity with anabiosis is the state of neural inactivity in memons following anes-
thesia. An anesthetic that blocks the transmission of signals between neurons while leaving
the remaining metabolic activity of the neurons intact causes a reversible absence of neural
activity that corresponds to an anabiotic state of the memon.
123
254 G. A. J. M. Jagers op Akkerhuis
Even in the early days of the biological sciences, scholars discussed whether dried or
frozen anabiotic stages are alive at a very slow pace, or whether they are truly static states
of matter. In 1860, the famous Société de Biologie in Paris wrote a lengthy report on this
subject (Broca 1860–1861). Quite importantly, this report concluded that the potential to
revive an anabiotic stage is an inherent aspect of the organization of the material of which
the object consists and that it is equally persistent as the molecular state of the matter form-
ing the system. In short, the Société de Biologie found that “la vie, c”est l”organisation
en action”. Additional support for this conclusion came from Becquerel (1950, 1951)who
subjected anabiotic stages to a temperature 0.01 degree above absolute zero, a temperature at
which no chemical processes can occur, even not very slowly. Becquerel demonstrated that
structure alone is enough to allow revival at normal temperatures. Anabiosis from absolute
zero or complete desiccation has led to the conclusion that “The concept of life as applied
to an organism in the state of anabiosis (cryptobiosis) becomes synonymous with that of
the structure, which supports all the components of its catalytic systems”(Keilin 1959), or
that “life is a property of matter in a certain structure”(Jeuken 1975). With respect to the
question of: what certain structure?, the operator hierarchy suggests that all operators with a
complexity similar to or higher than the cell answer this question.
6LifeasWedoNotKnowIt
Considerations about “life as we do not know it” depend on assumptions. As a context for
such assumptions, the operator hierarchy offers two advantages. First, the operator hierarchy
has its basis in the general principle of first-next possible closure. Second, the rigid internal
structure of the operator hierarchy offers a unique guide for assumptions about life that we
do not yet know.
Based on the general principle of first-next possible closure, the operator hierarchy shows a
strict sequential ranking of the operators. Assuming that closures act as an absolute constraint
on all operator construction, the operator hierarchy then has universal validity. Support for
the latter assumption comes from the observation that, as far as we know, all operators with
a complexity that is equal to or lower than the molecules seem to have a universal existence.
If this universality extends to the biotic operators, the material organization of higher-level
operators, such as cells and memons, may then possibly be found in the entire universe. Such
universality would significantly assist in the search for exobiotic life forms because alien life
may show similar organization to the life we do know, at least with respect to the first-next
possible closures involved. The demand of closure still leaves a good deal of freedom for
the physical realization of operators. On other planets, different molecular processes may
form the basis of the autocatalysis and interface of the first cells. Similarly, the operator
hierarchy poses no limits to the actual shape, color, weight, etc. of exobiotic multicellular
organisms. Furthermore, even though the presence of neural networks may be required for
memic organization throughout the universe, the operator hierarchy does not restrict the kind
of elements producing these networks, or the details of the neural network structure other
than demanding hypercyclicity and interface.
The rigid internal structure of the operator hierarchy allows predictions about the con-
struction of life forms that have not yet evolved on Earth. Of course, any discussion of this
subject involves speculation, but the operator hierarchy may well offer a unique starting
point for such a discussion. In an earlier publication (Jagers op Akkerhuis 2001), I have
indicated various future operator types with a higher complexity than the cellular hardwired
memon. To minimize the aspect of speculation, I would like to discuss here only the memon
123
Towards a Hierarchical Definition of Life, the Organism, and Death 255
immediately above the cellular hardwired memon (see Fig. 1), the so-called “softwired me-
mon”. According to the operator hierarchy, this type of memon should be able to copy
information structurally. This means that the organism should be able to copy all of its infor-
mation by copying the structure of its neural network. At a lower level in the hierarchy, cells
do this by copying their genetic molecules. Softwired memons can also do this. They are
based on a virtual neural network that resides in computer memory arrays. During their oper-
ation softwired memons continuously track all their neurons, neural connections, connection
strengths and interactions with the interface. It is therefore only a small step for softwired
memons to read and reproduce all the knowledge in their neural network by copying these
arrays. On these grounds, it may be deduced that softwired memons (or still higher complex-
ity memons) form the easiest way to satisfy the demands of the operator hierarchy for the
autonomous, structural copying of information. The operator hierarchy suggests therefore
that life as we do not know it will take the shape of technical memons.
The above reasoning shows that the operator hierarchy offers clear criteria with respect
to different forms of “artificial life”. The acceptance of an artificial entity as life is only pos-
sible when it shows all of the required properties of an operator. Referring to the difference
between strong artificial life and weak artificial life, which do and do not consider a-life
entities as g enuine life, respectively, it would be fully in line with the present reasoning to
consider as genuine life all a-life entities that fulfill the requirements for being an operator.
7 On Life and Death
Given the present focus on states of matter, it is quite simple to define dead matter as: all
operators that do not fit the general definition of life. It is more difficult, however, to define
death.
Given the current point of view, death represents a state in which an organism has lost
its closure. The use of closure in this definition helps prevent that “…. the properties of an
organism as a whole [would be confused] with the properties of the parts that constitute it”
(Morales 1998). However, organisms also loose their closure during transitions that are part
of life cycles and that are not associated with the organism’s death. For example, the closure
of the organism is lost and a new closure gained when the zygote exchanges its unicellular
organization for the multicellular state of the embryo and when the multicellular embryo
develops to a memic state. Is it possible to specify the loss of closure during death in a way
that excludes closure losses during life cycles?
With respect to the above question of how to exclude the loss of closure during transitions
in life cycles when defining death, the general process of deterioration offers a solution.
During their lives, organisms deteriorate because of injury and ageing. The loss of closure
marking death is always associated with the organism’s irreversible deterioration. Demanding
irreversible deterioration, therefore, helps to prevent that one would be tempted to consider,
for example, a caterpillar as having died, when its tissues are reorganized during the transition
via the pupae to a butterfly. Accordingly, it is possible to describe death as: the state in which
an organism has lost its closure following irreversible deterioration of its organization.
Using the above definition, death may occur in either an early or late phase of the deterio-
ration process, and following the death of multicellulars, a short or long period may pass until
the organism’s body parts become dead matter. The latter has its cause in the hierarchical
construction of multicellular organisms. Accordingly, the loss of the highest closure implies
a classification of the remaining body as an operator showing the first-next lower closure.
123
256 G. A. J. M. Jagers op Akkerhuis
Death depends on the loss of closure. To illustrate the contribution of this statement to
the analysis of death, the death of a memon can be used. Due to the memon’s strongly inte-
grated organization, death may occur at various levels that affect other levels. For example,
the multicellular regulation may be the first to collapse due to the loss of liver functions.
After a certain period, this will cause failure of neural functioning, the latter marking the
memon’s death In another situation, the neural functions may be lost first, and the memon
is the first to die, dragging its body with it in its fall. However, sometimes enough neural
activity may remain for a vegetative functioning of the memon’s body as a multicellular
unit. The vegetative state cannot maintain itself autonomously (in principle, a requirement
for a multicellular organism) but it may continue given the right medical care. If this care is
withdrawn, the multicellular body will start deteriorating after which the cells in the organs
and tissues will start dying at different rates. At a certain point, the multicellular closure
is lost, and separately surviving cells have become the next level operators to die. Physio-
logical differences between cells now determine the period during which they can survive
in the increasingly hostile habitat of the dead memon, which is cooling below the normal
operating temperature of cells and which shows many adverse chemical changes such as the
lowering of oxygen levels, the release of decay products of dead cells, etc. Shortly after the
memon’s death, it is possible to take intact body-parts, organs and cells from its body and
sustain their functioning following transplantation to a favorable environment. For example,
the offspring of cells from the cervix of Henrietta Lane are still cultured as He La cells in
many laboratories.
8 The Inutility of Property Lists
The above arguments and examples have explored the possibilities of using the operator
hierarchy for creating coherent definitions of life, the organism, living and death. However,
how should the outcome be evaluated? Have the present attempts led to definitions that could
be generally accepted in the field? A way of evaluating this that has become rather popular
is to check the results against lists of preset criteria. Those who want to evaluate the present
approach in this way may want to examine the following lists of criteria.
Morales (1998) has published a list of properties for a definition of life that includes the
following criteria: 1. Sufficiency (Does the definition separate living entities from non-living
ones?), 2. Common usage (simple classification of easy examples), 3. Extensibility (Does the
definition deal with difficult cases, such as viruses, mules, fire, Gaia, extraterrestrial life and
robots?), 4. Simplicity (few ifs, buts, ands, etc.) and 5. Objectivity (Criteria are so simple that
everyone applies them with the same result). Emmeche (1997) offers another criteria list for
a definition of life that includes the following: 1. The definition should be general enough to
encompass all possible life forms. (The definition should not only focus on life as we know
it.), 2. It should be coherent with measured facts about life, (It should not oppose obvious
facts.), 3. It should have a conceptual organizing elegance. (It can organize a large part of the
field of knowledge within biology and crystallize our experience with living systems into a
clear structure, a kind of schematic representation that summarizes and gives further structure
to the field.), 4. The definition should be specific enough to distinguish life from obviously
non-living systems. Emmeche (1997) furthermore states that a definition “should cover the
fundamental, general properties of life in the scientific sense”. Korzeniewski (2005)has
also proposed a list of criteria for a cybernetic definition of life, and Poundstone (1984)has
extracted further criteria for life from the work of von Neumann and Burks (1966). Oliver and
123
Towards a Hierarchical Definition of Life, the Organism, and Death 257
Perry (2006) have suggested a list more or less similar to that of Emmeche (1997) focusing
specifically on properties of a good definition.
With respect to the use of criteria lists, I agree with other authors (Maturana and Varela
1980; van der Steen 1997) that it is not necessarily an advantage if a theory performs well
or a disadvantage if a theory performs poorly according to a list of criteria; an approach’s
value does not necessarily correspond to its performance in these types of checklists. The
match depends on the similarity in major goals and paradigms and the creator’s influence on
the selection and definition of criteria in a given list. In addition, the selection of “favo rable”
lists can lead to false positives.
For the above reasons, I am convinced that it is only possible to evaluate the currently
proposed definitions “the hard way”, i.e., by critically examining the internal consistency
and transparency of their logic. In this respect, the present approach has the advantage of a
fundamental bottom-up construction. It starts with defining elementary building blocks, the
operators, and their hierarchical ranking in the operator hierarchy. To recognize and rank the
operators, the operator hierarchy uses first-next possible closures. In the resulting hierarchy,
the definition of every higher-level operator depends, in an iterative way, on a lower-level
“ancestor” until a lowest-level ancestral system is reached, which is presumably the group
of elementary particles that according to the superstring theory may have a common basic
structure. The result is a strict, coherent and general framework that is open to falsification:
the operator hierarchy. Subsequently, the operator hierarchy offers a fundament to define a
range of secondary phenomena, such as life, the organism, living and death. Because of the
reference to the operator hierarchy, the present definitions are short, logical statements that
show a high specificity with respect to whether a certain entity satisfies the definition (list of
examples in the following section).
9 Testing the Definition of Life
When using the operator hierarchy as a context for a definition, it is easy to conclude that
viruses, prions, memes or replicating computer programs are not forms of live. Both a virus
with a surrounding mantle and a viral strand of DNA or RNA are not operators, thus not life.
Prions are molecules, thus not life. Memes, such as texts and melodies, are pieces of coding
that memons can decode and replicate (Dawkins 1976). Accordingly, memes are not opera-
tors, thus not life. Ray (1991) has created computer programs that can replicate themselves
onto free computer space, show mutation, and modify and compete for the available space in a
virtual world called Tierra. Since its start, this virtual “ecosystem” has seen the evolution of a
range of different computer programs. In the same way as molecular viruses depend on cells,
the programs in Tierra depend on a computer to copy and track their structure. Accordingly,
they are not operators, thus not life. Sims (1994) has used genetic algorithms for evolving
virtual computer creatures with body parts and a neural network with interface. The simu-
lation of these animal-models allows virtual movement such as finding and grasping virtual
food items. Sims’s programmed creatures may possess hypercyclic neural networks and on
these grounds show similarity to softwired memons. According to the operator hierarchy, a
softwired memon should autonomously be able to copy its information structurally. Although
I am not an expert in this field, it seems to me that Sims’s organisms do not themselves keep
track of their arrays with information about their interface and neurons, neural connections,
and connection strengths, and that they do not autonomously organize their maintenance.
Assuming that the latter interpretations are correct, Sims’s computer animals are not yet life.
123
258 G. A. J. M. Jagers op Akkerhuis
The use of the present definition also allows the effortless rejection of other systems that
are not operators and sometimes receive the predicate of “borderline situations”, such as
flames, whirlwinds, crystals, cars, etc. Technical, computer based memons, however, such as
robots, can be operators when they show the required structure.
To summarize the practical applicability of the present definition of life, I include a list
of the examples that were discussed in the text and supplement them with some additional
cases. The examples in this list form three groups depending on whether the entities involved
are operators or not, and whether they show a complexity that equals or exceeds that of the
cellular operator. In the text below I use the concept of “interaction system” (e.g. Jagers op
Akkerhuis 2008) for all systems that are not operators because the interactions of their parts
do not create a first-next possible, new, closure type.
Group A. Systems that are not life because they are not an operator
1. An entire virus particle with external envelope (represents a simple interaction system)
2. A computer virus based on strings of computer code
3. A flame
4. A tornado
5. A crystal
6. A car
7. A bee colony (The colony is an interaction system, and the bees are organisms.)
8. A cellular colony not showing the requirements of multicellularity (The individual cells
are organisms and thus represent life.)
9. A colony of physically connected cellular memons (As long as the individuals lack the
required memic closure.)
10. A robot (As long as it is a non-memic technical machine.)
11. Computer simulations of organism (including memons) that depend on external “orches-
tration”
12. A cutting/slip of a plant that cannot potentially show autonomous maintenance given
the right conditions (It lacks the closure required for multicellularity.)
13. A separate organ, such as a liver or leg (Not potentially capable of autonomous main-
tenance.)
14. Endobiontic bacteria having lost genes that are obligatory for autonomous mainte-
nance. The transfer to the genome of the host of DNA coding for enzymes required
in autonomous maintenance implies a partitioning of the aucatalytic closure between
the endobiont and its host,. Because of this, the endobiont is no longer an autonomous
organism but has become a special kind of organelle.
Group B. Systems that are operators but that are not life because their complexity is lower
than that of the cellular operator
1. A prion
2. Self-replicating DNA/RNA particles (Catalyze their own copying in a solution containing
the right building materials.)
3. A DNA or RNA string of a virus that is copied in a cell
Group C. Operators representing life
1. A cutting/slip or other plant part that can potentially maintain itself given favorable
environmental conditions
123
Towards a Hierarchical Definition of Life, the Organism, and Death 259
2. Anabiotic organisms (The fact that they are dried, frozen, etc. does not take their required
closure away.)
3. Fully anaesthetized animal supported in its functioning by a mechanical heart-lung sup-
port and showing no neural activity (This can be regarded as a form of memic anabiosis
with the potency become active again.)
4. A computer memon or other technical memon (A memic robot.)
5. An artificial cellular operator constructed by humans
6. A exobiotic cellular operator with another chemistry than that found on earth
7. Sterile or otherwise non-reproducing organism (e.g. a mule, worker bee, sterile individ-
uals)
8. Endoparasites or endosymbiontic unicellular organisms living in cells and still possessing
the full potential of autocatalysis
10 In Conclusion
1. Overviews of the definitions of life from the last 150 years show that no consensus def-
inition on life exists. In the light of the continuous failure to reach consensus on this
subject, certain scientists have adopted a practical viewpoint, accepting, for example,
the use of property checklists for identifying living systems. Others have advocated that
the need for a generally accepted definition remains acute. Amongst the proposals for
solving the problem is the suggestion to construct a broader context, a “ theory of life”
before continuing with attempts to define of life.
2. Inspired by the latter suggestion, the present paper invokes a classification of the gener-
alized particle concept, called the operator hierarchy”. This hierarchy has several advan-
tages for defining life: first, it offers a general context for including and differentiating
between life and non-life, and second, it offers the unique possibility to extrapolate
existing trends in the evolution of material complexity and to use these as a guide for
discussions about “life as we do not know it”.
3. In close association with the reviewed literature, the use of the operator hierarchy allowed
the following definitions to be suggested:
A. From the viewpoint of the evolution of material complexity, life is: matter with the
configuration of an operator, and that possesses a complexity equal to or even higher
than the cellular operator.
B. Living describes the dynamics of an operator that satisfies the definition of life.
C. The definition of unitary organisms can take the form o f: the operators that fit the
definition of life.
D. A multicellular organism (the cellular operator showing the multi-state) is: a con-
struction of mutually adhering cells showing obligatorily recurrent interactions
based on the same interaction type, that has the potential of maintaining its func-
tioning as a unit and that does not show memic structure
E. Dead matter applies to all operators that do not fit the definition of life.
F. Death is: the state in which an organism has lost its closure following irreversible
deterioration of its organization.
4. From the discussion of examples in the literature, it was concluded that the present set
of definitions easily distinguishes life and non-life regardless of whether this is tested
using the “obvious examples”, the “borderline cases” or “life as we do not know it”. This
suggests that the present approach may well offer a practical step forward on the path
towards a consensus definition for the states of matter representing “life”.
123
260 G. A. J. M. Jagers op Akkerhuis
Acknowledgments The author would like to thank Herbert Prins, Henk Siepel, Hans-Peter Koelewijn, Rob
Hengeveld, Dick Melman, Leen Moraal, Ton Stumpel and Albert Ballast for constructive discussion and/or
comments on the manuscript and Peter Griffith for English correction.
References
Allen, J. F. (1993). Control of gene expression by redox potential and the requirement for chloroplast and
mitochondria genomes. Journal of Theoretical Biology, 165, 609–631.
Becquerel, P. (1950). La suspension de la vie des spores des b actéries et des moisissures desséchées dans
le vide, vers le zéro absolu. Ses conséquences pour la dissémination et la conservation de la vie
dans l”univers. Les Comptes rendus de l”Académie des sciences, Paris 231, 1274; 1392–1394.
Becquerel, P. (1951). La suspension de la vie des algues, lichens, mousses, aux zéro absolu et role de la
synérèse réversible pour l”existence de la flore polaire et des hautes altitudes. Les Comptes Rendus
de l”Académie Des Sciences, Paris, 232, 22–25.
Bedau, M. A. (2007). What is life. In S. Sarkar & A. Plutynski (Eds.), A companion to the philosophy
of biology (pp. 455–603).
Berg, O. G., & Kurland, C. G. (2000). Why mitochondrial genes are most often found in nuclei. Molecular
Biology and Evolution, 17, 951–961.
Bonner, J. T. (1998). The Origins of Multicellularity. J.T. Bonner. Integrative Biology, 1, 27–36.
Bro, P. (1997). Chemical reaction automata. Precursors of artificial organisms. Complexity, 2, 38–44.
Broca (1860–1861). Rapport sur la question soumise à la Société de Biologie au sujet de la reviviscence
des animaux desséchés. Mém. Soc. Biol., 3me Série, II, 1860, 1–139.
Bullock, S., Noble, J., Watson, R., & Bedau, M. A. (Eds.) (2008). Artificial Life XI: Proceedings of the
Eleventh International Conference on the Simulation and Synthesis of Living Systems. Cambridge,
MA: MIT Press.
Capps, G. J., Samuels, D. C., & Chinnery, P. F. (2003). A model of the nuclear control of mitochondrial
DNA replication. Journal of Theoretical Biology, 221, 565–583.
Chandler, J. L. R., & Van de Vijver, G. (Eds.) (2000). Closure: Emergent organizations and their dynamics.
Annals of the New York academy of Sciences 901.
Checkland, P., & Scholes, J. (1990). Soft systems methodology in action (p. 329). Chichester: Wiley.
Clark, C. G., & Roger, A. J. (1995). Direct evidence for secondary loss of mitochondria in Entamoeba
hitolytica. Proceedings of the National Academy of Sciences of the USA, 92, 6518–6521.
Cleland, C. E., & Chyba, C. F. (2002). Defining “life”. Origins of Life and Evolution of the Bio-
sphere, 32, 387–393.
Cleland, C. E., & Chyba, C. F. (2007). Does “life” have a definition?. In T. Woodruff, I. I. I. Sullivan, &
J. A. Baross, Planets and Life: The Emerging Science of Astrobiology. : Cambridge University Press.
Dawkins, R. (1976). The selfish gene. Oxford: Oxford University Press.
Eigen, M. (1971). Molekulare Selbstorganisation und Evolution (Self organization of matter and the
evolution of biological macro molecules). Naturwissenschaften, 58, 465–523.
Eigen, M., & Schuster, P. (1979). The hypercycle: A principle of self-organization. New York: Springer.
Emmeche, C. (1997). Autopoietic systems, replicators, and the search for a meaningful biological definition
of life. Ultimate Reality and Meaning, 20, 244–264.
Etxeberria, A. (2004). Autopoiesis and natural drift: Genetic information, reproduction, and evolution
revisited. Artificial Life, 10, 347–360.
Gánti, T. (1971). The principle of life (in Hungarian). Budapest: Gondolat.
Grosberg, R. K., & Strathmann, R. R. (2007). The evolution of multicellularity: A minor major transi-
tion?. Annual Review of Ecology, Evolution, and Systematics, 38, 621–654.
Happel, B. L. M. (1997). Principles of neural organization: Modular neuro-dynamics. PhD thesis, 125 pp.
Hazen, R. M. (2001). Selective adsorption of L-andD-amino acids on calcite: Implications for biochemical
homochirality. Proceedings of the National Academy of Sciences, 98, 5487–5490.
Hengeveld, R., & Fedonkin, M. A. (2007). Bootstapping the energy flow in the beginning of life. Acta
Biotheoretica, 55, 181–226.
Heylighen, F. (1990). Relational Closure: A mathematical concept for distinction-making and complexity
analysis. In R. Trappl (Ed.), Cybernetics and Systems “90 (pp. 335–342). Singapore: World Science.
Heylighen, F. (1991). Modeling Emergence. World Futures: The Journal of General Evolution, 31
, 89–104.
Hull, D. L. (1981). Units of evolution: A metaphysical essay. In U. J. Jensen & R. Harré (Eds.), The
philosophy of evolution (pp. 23–44). New York: St. Martins Press.
123
Towards a Hierarchical Definition of Life, the Organism, and Death 261
Jagersop Akkerhuis, G. A. J. M. (2001). Extrapolating a hierarchy of building block systems towards
future neural network organisms. Acta Biotheoretica, 49, 171–189.
Jagersop Akkerhuis, G. A. J. M. (2008). Analyzing hierarchy in the organization of biological and physical
systems. Biological Reviews, 83, 1–12.
Jagersop Akkerhuis, G. A. J. M., & van Straalen, N. M. (1999). Operators, the Lego–bricks of nature:
Evolutionary transitions from fermions to neural networks. World Futures, the Journal of General
Evolution, 53, 329–345.
Jeuken, M. (1975). The biological and philosophical definitions of life. Acta Biotheoretica, 24, 14–21.
Kaiser, D. (2001). Building a multicellular organism. Annual Review Genetics, 35, 103–123.
Kauffman, S. A. (1986). Autocatalytic sets of proteins. Journal of Theoretical Biology, 119, 1–24.
Kauffman, S. A. (1993). The origins of order. Oxford: Self-organization and selection in evolution. Oxford
University Press.
Keilin, D. (1959). The Leeuwenhoek lecture: The problem of anabiosis or latent life: History and current
concept. Proceedings of the Royal Society of London, Series B, Biological Sciences, 150, 149–191.
Korzeniewski, B. (2005). Confrontation of the cybernetic definition of a living individual with the real
world. Acta Biotheoretica, 53, 1–28.
Koshland, D. E., Jr. (2002). The seven pillars of life. Science, 295, 2215–2216.
Kunin, V. (2000). A system of two polymerases. A model for the origin of life. Origins of Life and
Evolution of the Biosphere, 30, 459–466.
Kurzweil, R. (1999). The age of spiritual machines. When computers exceed human intelligence.New
York: Viking.
Lane, N. (2005). Power, sex, suicide (p. 354). Oxford: Mitochondria and the meaning of life. Oxford
University Press.
Mackie, G. O., Anderson, P. A. V., & Singla, C. L. (1984). Apparent absence of gap junctions in two
classes of Cnidaria. The Biological Bulletin, 167, 120–123.
Martin, W., & Russel, M. J. (2002). On the origins of cells: A hypothesis for the evolutionary transitions
from abiotic chemistry to chemoautotrophic prokaryotes, and from prokaryotes to nucleate cells. Philo-
sophical Transactions of the Royal Society of London, Series B-Biological Sciences, 358, 59–83.
Maturana, H. R., & Varela, F. J. (1980). Autopoiesis and Cognition. The Realization of the Living.
Dordrecht: D. Reidel (also in Boston Studies in the Philosophy of Science, 42).
Maynard Smith, J., & Szathmáry, E. (1995). The Major Transitions in Evolution. Oxford: Freeman & Co
(now Oxford University Press).
Maynard Smith, J., & Szathmáry, E. (1999). The origins of life. Oxford: Oxford University Press.
Morales, J. (1998). The definition of life. Psychozoan, 1(a strictly electronic journal), 1–39.
Munson, J. R., & York, R. C. (2003). Philosophy of biology. Encyclopedia Britannica. Bri-
tannica 2003 ultimate reference suite. (http://www.compilerpress.atfreeweb.com/Anno%20Munson
%20EB%20Philosophy%20of%20biology.htm).
Murre, J. M. J., Phaf, R. H., & Wolters, G. (1992). CALM: Categorizing and learning module. Neural
Networks, 5, 55–82.
Nicholson, B. J. (2003). Gap junctions—from cell to molecule. Journal of Cell Science, 116, 4479–4481.
Oliver, J. D., & Perry, R. S. (2006). Definitely life but not definitely. Origins of Life and Evolution of
the Biosphere, 36, 515–521.
Pagels, H. R. (1985). Perfect symmetry: The search for the beginning of time. New York: Simon and
Schuster.
Panchin, Y. V. (2005). Evolution of gap junction proteins—the pannexin alternative. Journal of Experimental
Biology, 208, 1415–1419.
Peracchia, C., & Benos D. J. (Eds.) (2000). Gap Junctions: Molecular Basis of Cell Communication in
Health and Disease (Current Topics in Membranes, Volume 49). Academic press, New York.
Poundstone, W. (1984). The Recursive Universe: Cosmic Complexity and the Limits of Scientific Knowl-
edge. New York: William Morrow.
Popa, R. (2003). Between necessity and probability. Searching for the definition and the origin of life.
Advances in Astrobiology and Biogeophysics
. Berlin: Springer.
Ray, T. S. (1991). Evolution and optimization of digital organisms. In K. R. Billingsley, E. Deroh-
anes, & H. Brown, III (Eds.), Scientific excellence in supercomputing: The IBM 1990 contest Prize
Papers (pp. 489–531). Athens, GA: The Baldwin Press, The University of Georgia.
Rivera, M., Jain, R., Moore, J. E., & Lake, J. A. (1998). Genomic evidence for two functionally distinct
gene classes. Proceedings of the National Academy of Sciences of the USA, 95, 6239–6244.
Rosen, R. (1958). A relational theory of biological systems. Bulletin of Mathematical Biophysics, 20,
245–260.
123
262 G. A. J. M. Jagers op Akkerhuis
Rosen, R. (1973). On the dynamical realization of (M,R)-systems. Bulletin of Mathematical Biophys-
ics, 35, 1–9.
Rosen, R. (1991). Life itself. A comprehensive inquiry into the nature, origin and fabrication of life.New
York: Columbia University Press.
Ruiz-Mirazo, K., Pereto, J., & Moreno, A. (2004). A universal definition of life: Autonomy and open-ended
evolution. Origins of Life and Evolution of the Biosphere, 34, 323–345.
Searcy, D. G. (2003). Metabolic integration during the evolutionary origin of mitochondria. Cell
Research, 13, 229–238.
Sims, K. (1994). Evolving 3D Morphology and Behavior by Competition. In Brooks, M. (Ed.), Artificial
Life IV Proceedings, (pp. 28–39). Cambridge, MA: MIT Press.
Teilhard de Chardin, P. (1966). Man’s place in nature. The human zoology group. Editions du Seuil VIII,
Paris, 1949.
Teilhard de Chardin, P. (1969). The future of man. Editions du Seuil V, Paris, 1946.
Tinbergen, N. (1946). Inleiding tot de diersociologie (Introduction to animal sociology). J. Noorduijn en
Zoon, Gorinchem.
Townsend, C. R., Begon, M., & Harper, J. L. (2008). Essentials of ecology (3rd ed.). Oxford: Blackwell
publishing.
Turchin, V. (1995). A dialogue on metasystem transitions. In F. Heylighen, C. Joslyn, & V. Turchin (Eds.),
1995, The Quantum of Evolution. Toward a theory of metasystem transitions. (Gordon and Breach
Science Publishers, New York) (special issue of “World Futures: The journal of general evolution,
Vol. 45).
Turchin, V. (1977). The phenomenon of science. New York: Columbia University Press.
van der Steen, W. J. (1997). Limitations of general concepts: A comment on Emmeche’s definition of
“life”. Ultimate Reality and Meaning, 20, 317–320.
Varela, F. J. (1979). Principles of biological autonomy. New York: North Holland.
von Bertalanffy, L. (1968). General systems theory, foundations, development, applications. Middlesex,
England: Penguin Books Ltd. Harmondsworth.
von Neumann J., & Burks, A. W. (1966). Theory of self-reproducing automata.
Willensdorfer, M. (2008). On the evolution of differentiated multicellularity. arXiv 0801.2610v1
Author Biography
Gerard A. J. M. Jagers op Akkerhuis graduated (Cum Laude) in entomology, nematology and animal
physiology at the Wageningen Agricultural University, The Netherlands, in 1986. He received a PhD in eco-
toxicology at Wageningen Agricultural University in 1993. Between 1992 and 1993 he worked as a post doc
in ecotoxicology at the Vrije Universiteit Amsterdam. From 1996 to 1999 he worked as a guest scientist/mod-
eler at the National Environmental Research Institute (NERI) in Silkeborg (Denmark). Since 1999 he works
at Alterra, Wageningen UR. In 2009 he held a temporal position at the Radboud University Nijmegen. His
research activities are focused on ecology, ecotoxicology, system science and evolution. He developed the
‘Operator Hierarchy’ as a comprehensive meta-evolutionary approach for the evolution of particle types, from
elementary particles to animals.
123
