Autonomy as a property

that characterizes organisms

among other multicellular systems
La autonomía como
una propiedad que caracteriza a los organismos
frente a otros sistemas multicelulares
ARGYRIS ARNELLOSa,
KEPA RUIZ-MIRAZOa,b
ALVARO MORENOa
a IAS-Research Centre for Life, Mind and Society -

Department of Logic and Philosophy of Science.

b Biophysics Unit (CSIC-UPV/EHU)

University of the Basque Country (Spain)

Recibido: 4-2-2013 Aprobado definitivamente: 7-2-2013

ABSTRACT
Biology is full of examples of multicellular (MC) systems which may demonstrate some
organism-like properties but not all of them. Thus, it remains unclear if and when such sys-
tems should be considered as MC organisms, parts of organisms or groups of organisms. We
suggest the notion of autonomy as a possible candidate to ground conceptually MC organisms
and distinguish them from other forms of multicellularity. Considering unicellular systems as
autonomous organisms on the basis of the functional integration required for their metabolic
organization, we argue that MC systems should be also identified as autonomous, but on the
basis of exhibiting a special kind of functionally integrated and differentiated developmental
organization, which unfolds through a self-constructed set of mechanisms regulating the highly
plastic processes that bring about their own constitution as such MC entities.

© Contrastes. Revista Internacional de Filosofía: Suplemento 18 (2013), pp. 357-372. ISSN: 1136-9922
Departamento de Filosofía, Universidad de Málaga, Facultad de Filosofía y Letras
Campus de Teatinos, E-29071 Málaga (España)
358 argyris arnellos, kepa ruiz-mirazo, álvaro moreno

KEYWORDS
AUTONOMY, ORGANISM, MULTICELLULARITY, FUNCTIONAL INTEGRATION,
FUNCTIONAL DIFFERENTIATION, DEVELOPMENTAL CONSTRAINTS, EPIGENE-
TICS, SIGNALING NETWORKS

RESUMEN
La biología está llena de ejemplos de sistemas multicelulares que pueden mostrar algunas pro-
piedades de organismos, pero no todas. De este modo, no está muy claro si estos sistemas deben
ser considerados como organismos multicelulares y cuándo deben serlo. Aquí sugerimos que
la noción de autonomía es un candidato posible para fundamentar conceptualmente los orga-
nismos multicelulares y para distinguirlos de otras formas de multicelularidad. Considerando a
los sistemas unicelulares como organismos autónomos sobre la base de la integración funcional
requerida para su organización metabólica, argüimos que los sistemas multicelulares deben
ser también identificados como autónomos, pero sobre la base de que exhiben un tipo especial
de organización del desarrollo funcionalmente integrada y diferenciada, la cual se despliega a
través de un conjunto auto-construido de mecanismos que regulan los procesos enormemente
plásticos que dan lugar a su propia constitución como tales entidades multicelulares.
PALABRAS CLAVE
AUTONOMÍA, ORGANISMO, MULTICELULARIDAD, INTEGRACIÓN FUNCTIO-
NAL, DIFERENCIACIÓN FUNCTIONAL, CONSTRICCIONES DEL DESARROLLO,
EPIGENÉTICA, REDES DE SEÑALES

I. Introduction
Contemporary attempts to define biological individuality are based either
on physiological and genetic characteristics (see e.g. Santelices 1999), on the
cooperative/competitive behaviors of the entities involved (e.g. West and Kiers
2009; Strassmann and Queller 2010), on evolutionary concepts, like fitness and
adaptation (e.g. Queller and Strassmann 2009; Folse 3rd and Roughgarden 2010)
or on mixed approaches that favor specific aspects of functional integration,
such as germ-soma separation (e.g. Buss, 1987; Maynard Smith and Szathmáry
1995; Michod 2007), resulting in cohesive or policing mechanisms, so that
potential conflict among the constituting units is minimized (e.g. Frank 2003;
Godfrey-Smith 2009).
But each of these aspects does not apply in all cases un-problematically.
For instance, physiological unity is not present in every organism and genetic
relatedness is not strictly necessary for high cooperation, while genetic hetero-
geneity is not always a threat to multicellular integrity (Folse 3rd and Rough-
garden 2010). In many cases potential cooperation does not linearly transform

Contrastes. Revista Internacional de Filosofía. Suplemento 18 (2013)

Autonomy as a property that characterizes organisms...

into actual cooperation, and potential for low conflict does not mean actual
high integration or even actual low conflict (Strassmann and Queller, 2010).
Besides, there are cases of discordant selection, where selection operating at a
lower level may act against the selection at a higher level, even if the alignment
and export of the fitness interests of the component units of an organism could
also result in adaptation at the level of the (multicellular) organism (Gardner
and Grafen 2009; Folse 3rd and Roughgarden 2010). However, adaptation is
not an easy notion to explain, and is often taken to be demonstrated at a given
level of selection, thus resulting in different answers when it is the product of
group selection or of multiple levels of individual selection (Okasha, 2006). In
summary, there is a plurality of views and various difficulties around the pro-
blem of defining the biological individual and there are several criteria, whose
combination generates multiple different verdicts that do not necessarily overlap
and which fail to accommodate many examples we can think of in real biology
(Clarke 2011).
And yet, the clarification of the concept of an ‘individual organism’ is an
important philosophical and scientific problem, not only because the idea of
organism has played a key role in the history of biology (as a central part of
biological explanations ‒i.e., as the locus of mechanisms, of adaptations, of
selective-evolutionary dynamics); but also because, as it is argued in Ruiz-
Mirazo and Moreno (2011), without a strong idea of organism it would be very
difficult to provide a naturalized account of other fundamental concepts like
functionality, agency, autonomy, genetic information, etc.
Accordingly, we believe that a more comprehensive approach to biological
individuality is needed, and we suggest the notion of ‘autonomy’ as a possible
candidate for such an inclusive or comprehensive approach. Autonomy places
the individual organism at the centre of the stage and emphasizes that any
biological individual organism needs to realize the property of maintaining
itself as a metabolic system. Autonomous systems are not independent from
their surrounding, they critically rely on diverse features of the environment
(e.g.: general physico-chemical conditions for viability, energetic and material
availability, etc.) but they continuously generate and regenerate all the constra-
ints and mechanisms upon which the use and management of those resources
is based. Therefore, there is a continuous interplay between the organization of
processes constituting a relatively stable identity (or ‘self’), and the interactions
with the environment that this identity triggers and supports, which are crucial
for its maintenance. So autonomy must be conceived in terms of a particular
connection, or even collaboration with external systems. But, as in the case of
metabolism, it is impossible to talk of autonomy without the specification of
some form of self-constructed, individual identity. On the other hand, biology
shows many collective forms and groupings of entities, which may resemble

Contrastes. Revista Internacional de Filosofía. Suplemento 18 (2013)

360 argyris arnellos, kepa ruiz-mirazo, álvaro moreno

several composite forms of autonomy, but it is not clear whether those groupings
should be considered as full-fledged autonomous individuals or as just colonial
systems. Then, assuming that unicellular entities are autonomous organisms, the
question to address would be what sorts of MC systems, if any, meet equivalent
requirements and can therefore be regarded, themselves, as organisms –or, as
higher order autonomous organizations (Ruiz-Mirazo and Moreno 2011).
With this aim, taking into account the main results of current biological
investigations on the developmental processes of different types of MC systems,
we attempt to examine whether the concept of autonomy can account for other
functionally diverse but at the same time integrated forms of collective asso-
ciations of biological entities and processes. In other words, we shall suggest
when and how would it be possible to discern whether a group of cells is just
gathering together temporarily to improve their overall fitness, or irreversibly
becoming part of a higher-order autonomous entity. Certainly, the creation of
a higher order full-fledged autonomous entity requires some kind and degree
of functional integration, as an autonomous system’s creation is not possible
without a stronger subordination of the constitutive elements to the new functio-
nal requirements of the emerging global individual (Ruiz-Mirazo and Moreno
2011). Accordingly, we suggest that with respect to MC collections, autonomy
should be analyzed following at least a set of criteria related to the overall de-
velopmental coordination of the collection, the presence of differentiated cell
types and the regulation of their development, and their respective kind and
degree of functional integration necessary for the development and maintenance
of the organismal level (the level of the whole).
On those lines, in section 2 we discuss that, from an organizational point
of view, MC organisms should be identified by a set of criteria related to the
developmental mechanisms used by the collection of cells to coordinate their
differentiation pathways, in a way that the robust maintenance and potential
reproduction of the whole developing system is ensured through an emergent
functionally integrated organization. In section 3 we examine cases of advanced
bacterial multicellularity and of early eukaryotic multicellularity, where overall
developmental regulation is not that evident and a very limited degree of cellular
differentiation is at place, and we compare them with a case of early metazoans.
In section 4 we argue that although the former cases exhibit substantial coordi-
nation and cohesion (and functional integration surely at other organizational
levels), they do not qualify as autonomous entities, since full-fledged autonomy
at the MC level involves a special kind of regulatory control that provides the
potential for both functional diversification and functional integration required
for the proper and reliable unfolding of a process of development.

Contrastes. Revista Internacional de Filosofía. Suplemento 18 (2013)

Autonomy as a property that characterizes organisms... 361

II. Organisms in an evolutionary and in an organizational context

Within the modern synthesis, an organism is considered as the central
point of reference for evolution by natural selection (NS). According to that
evolutionary theoretical framework, individual organisms could be all those
entities (genes, cells, groups, species) on which natural selection operates, in-
ducing the differential survival and reproduction of the respective population.
The Darwinian tradition proposes a natural explanation for the design of any
organism based on the mechanism of NS. From that conception, in which the
unit of evolution (or unit of selection) is what really matters, fitness and its maxi-
mization are normally taken as the fundamental criterion to define organisms
(Gardner 2009). Gardner argues that fitness-maximization is the key design
principle that explains how natural selection solves the problem of adaptation,
that is, the packaging of parts into units of common purpose (be them organisms
or watches) (Gardner and Grafen 2009). Accordingly, an organism is a whole
whose parts are all under selection to maximize its fitness.
However, there is a fundamental difference between living organisms and
artificial systems, regardless of their apparent operational or global similarities:
whereas the watch is formed by fixed components, fabricated beforehand and
later assembled, in the organism the parts are formed for and from the others,
some parts actually producing the others. The main issue in this difference is that
in living organisms, there is a constitutional relation among the parts to form an
organization. That is why, for the Darwinian tradition the identification of watch
and organism (even regarding only contrivance and relation among parts) is not
problematic, whereas from the organizational point of view a distinction should
be made. This distinction implies that there is a difference between explaining
‘function’ as deriving from the concept of organization, and as deriving from
adaptation. And that is also the main reason why Darwin’s theory did not solve
all the objections to mechanistic approaches, especially regarding the lack of a
complete, satisfactory account of the intrinsic functioning of organisms.
From an organizational perspective, an organism is functionally integrated
in the sense that it maintains a material dynamic organization, which plays a
fundamental causal role in the generation of the structural constraints that ac-
tually make it possible (Moreno et al. 2011). However, the notion of ‘functional
integration’ is difficult to be operationalized, as it implies the coordination
among complex dynamical parts, which is exhibited in different degrees and
types in each ensemble. One could even say that everything we find in nature
is functionally integrated, in some degree and in some respect (Wilson 2000;
Clarke, 2010). Slime molds are functionally integrated but not in the same degree
as an animal; the organs of a higher animal are causally and hence functiona-
lly integrated but so are the members of a good music band. Therefore, there

Contrastes. Revista Internacional de Filosofía. Suplemento 18 (2013)

362 argyris arnellos, kepa ruiz-mirazo, álvaro moreno

is a need for specification of the type (the strength, plasticity, complexity…)

of the interactions between parts in order for them to qualify as members of
a ‘single functionally-integrated’ whole ‒i.e. of a higher-order autonomous
organization.
Dealing with the problem of MC organisms, in this paper we focus on how
collections of cells are able to repeatedly develop into such highly organized
and integrated entities. Development is here considered as a process that makes
possible the generation of a MC system in which different types of cells perform
different tasks, and in which all of them come from a single, fertilized mother
cell. Now, it should be noted that, in general terms (keeping the organizational
perspective), a material structure is considered as functional if it is generated
within a system and it operates as part of a network of constraints that contribute
differently to the development and self-maintenance of that system at the level
of the whole (Mossio et al. 2009). Therefore, in a MC system, in particular, a
material structure would be considered as functional if it is created within the
system and if it operates so as to induce such cellular differentiations, the result
of which will contribute to the development and self-maintenance of the system
at the level of the whole. Accordingly, the kind of functional integration that
matters for our purposes should imply inductive and regulatory control on the
fate of the cells during the process of cellular differentiation: the MC system
drives the generation of its unicellular parts but, also and more importantly, re-
gulates its development in order to be able to re-construct the dynamic relations
between those parts, i.e. to re-construct its own MC organization. Thus, the
appropriate descriptive level to frame our problem is in the context of systems
that comprise a diverse network of cell-cell interaction mechanisms and display a
strong interweaving between their ontogenetic development and their subsequent
reproduction. And the idea is to do so from the point of view of autonomy: i.e.,
of the self-determination capacities of the MC system involved. Therefore, as we
shall elaborate below, we examine the developmental processes of MC systems
by focusing on the characteristic properties of intercellular constraints (‘signals’)
that are considered as functional because they are generated within the system,
and they are participating in a collective network of regulatory mechanisms
that are, in turn, capable of modulation of intracellular epigenetic mechanisms
affecting the fate of cellular differentiation during development, thus bringing
about a higher-order functional integration at the MC level.
So in the following section we will take a closer look at the process of
development of three MC systems: the cyanobacterium Nostoc.punctiforme
and the green algae Volvox.carteri, as a case of advanced bacterial and early
eukaryotic MC system, together with the echinoderm Strongylocentrotus.pur-
puratus (a sea urchin): a metazoan MC system. The MC systems chosen to be
examined here are all constituted by genetically homogeneous cells and they

Contrastes. Revista Internacional de Filosofía. Suplemento 18 (2013)

Autonomy as a property that characterizes organisms... 363

implement complex organizations with a relatively high degree of cohesion.

Hence, at least phenotypically speaking, they are potential candidates to be
characterized as MC organisms.
III. Developmental regulation in three multicellular systems

III.1. Developmental regulation in advanced bacterial and early eukaryotic

MC systems
Nostoc punctiforme is a filamentous photosynthetic and diazotrophic cya-
nobacterium with two cell types:1 vegetative cells, which are responsible for
photosynthesis, and heterocysts, which are responsible for nitrogen fixation.
The nitrogen products are passed along to the photosynthesizing cells and, in
turn, photosynthetic cells transfer to heterocysts several metabolites neces-
sary for their operation. Differentiation takes place at a frequency of about
8% of the total cells (Meeks et al. 2002), producing a semi-regular pattern of
morphologically and metabolically different cell types. Several models have
been proposed to explain the semi-regular pattern of heterocyst development
(see Kumar et al. 2010 for details). All cells in the filament detect the signal to
differentiate (nitrogen deprivation) but only some of them respond to it, leading
to a biased initiation process of differentiation. The earliest signal of nitrogen
limitation triggers the activation of the global nitrogen regulator, NtcA, in all
the cells that are at the appropriate cell stage. Activated NtcA then functions
as a transcriptional activator of other inducing or suppressing proteins of he-
terocyst differentiation, in a way that only a single heterocyst is present at a
given site in the filament. This can be visualized as the result of a competitive
interaction between an activator (HetR) and an inhibitor (PatS) of heterocyst
development. HetR is an intracellular activator of heterocyst development and
PatS is the diffusible inhibitor (see Kumar et al. 2010 for details). Cells in
the appropriate stage initiate development by synthesizing HetR. High HetR
activity induces cell differentiation, whereas high PatS decreases HetR and
leads to regression. HetR promotes its own synthesis (through autocatalysis)

1 In N. punctiforme the vegetative cells can also develop into akinetes, spore-like struc-
tures that are more resistant to cold and desiccation conditions, and into hormogonium filaments,
which lack heterocysts, have smaller size and a slow gliding motility, used for short-distance
dispersal. The development of akinetes does not seem to involve any intercellular signaling but
it is strictly triggered by environmental inputs. Hormogonium development will result in one
cell type in the filament and, again, this seems to be triggered by environmental signals followed
by multiple independent (intracellularly controlled) cell divisions for each one of the cells (see
Meeks et al. 2002 for details). It is for those reasons that we do not consider these modalities in
our description of N. punctiforme.

Contrastes. Revista Internacional de Filosofía. Suplemento 18 (2013)

364 argyris arnellos, kepa ruiz-mirazo, álvaro moreno

but also the production of PatS. On the contrary, PatS suppresses HetR. In
neighbouring cells, the entry of PatS will prevent the formation of HetR. In
more distant areas, diffusion of PatS may not be sufficient, so new centres of
activation can be formed. 2
A careful analysis of the details of heterocyst development (see e.g. Meeks
et al. 2002; Kumar et al. 2010; Maldener and Muro-Pator 2010) reveals that the
signaling network which guides the development of a differentiating cluster of
cells into one heterocyst at any developmental site in the filament operates un-
der a single constraint (PatS concentration) at the collective level. There seems
to be no generation of other compounds/structures (i.e., no synthesis of any
morphogen or some other kind of signal in the cells where HetR is suppressed
by PatS), which acts intercellularly on the phenotypic traits and organization
of the different cell that produced PatS ‒or on any other neighbouring cell. In
addition to that, heterocysts undergo terminal differentiation, as they lose the
ability to divide. In fact, whether a vegetative cell can be turned into a pro-
heterocyst is strictly connected to its life-cycle stage (Meeks et al. 2002). Now,
this implies a mechanism of differentiation that remains strongly coupled to
the core metabolic requirements of the process of growth and division of the
vegetative cells. In addition, heterocysts development is a terminal event be-
cause they ‘sacrifice’ their ability to reproduce in order to provide surrounding
vegetative cells with combined nitrogen. Moreover, there is a 24h delay in the
growth process of the filament after the development of functional heterocysts,
which is a result of the downregulation of genes related to vegetative growth and
division (Christman et al. 2011). Those two facts indicate that cell division and
cell differentiation cannot be modulated outside of the core metabolic context.
On the contrary, they imply a mechanism of the developmental regulation of
differentiation that remains strongly coupled to the metabolic requirements of
the vegetative cells.
Volvox carteri is a eukaryotic MC system constituted by unicellular algae
and has a developmental process that results in spheroid adults with two cell
types: large reproductive cells (gonidia) and small motile somatic cells. Somatic
cells do not divide but continue beating their flagella, thereby providing the
group with a capacity to swim. The rest of the cells in the group (the germ cells)
divide and produce progeny. Complete germ-soma separation derives from the
fact that, in Volvox carteri, germ cells directly become reproductive gonidia.

2 The developmental direction is also influenced by ancillary inducing and suppress-

ing signals. However, those signals are always acting intracellularly and downstream the main
activator and inhibitor, and independently within each cell. Therefore, they are irrelevant to our
analysis and will not be discussed here.

Contrastes. Revista Internacional de Filosofía. Suplemento 18 (2013)

Autonomy as a property that characterizes organisms... 365

A healthy culture of V. carteri will cleave 11 times and will result in di-
vision of labor among (approximately) 2000 somatic cells and 16 germ cells.
During embryogenesis some of the cleavage divisions are asymmetric, produ-
cing large/small sister-cell pairs (Kirk 2005). At the end of embryogenesis the
volume of the gonidial initials is about 30-fold larger than that of the somatic
initials. However, at that stage, cells are only different in size. Then, by a still
unknown mechanism, the size of each sister cell leads to the activation of ei-
ther a somatic or germline program (Kirk 2005). Thus, small cells develop as
biflagellate somatic cells for motility, biosynthesis of the extracellular matrix,
and phototaxis, and large cells develop as non-motile, germ cells specialized
for growth and reproduction.
In the developmental process of V. carteri (see Kirk 1998; 2005 and Hall-
man 2011 for details) the initiation of somatic or gonidial developmental process
is explicitly dependent on the size of each cell and cellular differentiation is
achieved by intracellular cell fate specification. Specifically, V. carteri cells that
are below the threshold diameter of 8 μm at the end of cleavage will activate the
somatic-cell program of differentiation, while cells above that threshold activate
the gonidial program. And this happens even if all of the cytoplasm that they
contain is cytoplasm that would normally have been found in somatic cells. In
addition, the way V. carteri achieves its complete germ-soma separation prohibits
any further flexibility and variability in the genotype-phenotype mapping, as
discussed thoroughly by Nedelcu and Michod (2004). Thus, in V. carteri –even
more explicitly than in N. punctiforme, given the complete germ-soma separation
of the former– cell division remains either totally decoupled (in somatic cells) or
strongly coupled (in germ cells) to cell growth and global system reproduction.
Therefore, we can conclude that the development of cellular differentiation in V.
carteri takes place independently of any constraining signaling structures pro-
duced from other cells in this MC system. Furthermore, as in the N. punctiforme
case, development in V. carteri is also strongly coupled to the core metabolic
requirements of the processes of growth and division, preventing potential me-
chanisms for meta-cellular modulation of the developmental process and the
adaptive evolution of further cell differentiation patterns.
III.2. Developmental regulation in metazoans: the case of a sea urchin
The sea urchin (SU) Strongylocentrotus purpuratus is a small invertebra-
te3 that belongs to the echinoderm phylum. Although it is a relatively simple

3 The sea urchin is an excellent system to study fundamental mechanisms underlying

the development of metazoan embryos, since it is relative simple, and its genome has been se-
quenced. The construction of its endomesoderm gene regulatory network (GRN) pertaining to
both intra- and inter-cellular signals up to 30 hours post-fertilization is currently the largest and

Contrastes. Revista Internacional de Filosofía. Suplemento 18 (2013)

366 argyris arnellos, kepa ruiz-mirazo, álvaro moreno

metazoan, it shows a very interesting developmental process leading to diffe-

rentiated tissues and organs. Sea urchin embryos develop into free-swimming
pluteus larvae consisting of only five major tissues in about 3 days. The period
of cleavage lasts almost 9 hours post fertilization (hpf) and the embryo hatches
from its fertilization membrane at about 18 hpf. Mesenchyme blastula begins
at 24hpf. By the time gastrulation begins (around 27-30 hpf) skeletogenic,
mesoderm, anterior and posterior endoderm and ectoderm cell lineages have
been specified.
One of the most interesting features of the SU development (see Peter and
Davidson 2010 for details) is the intercellular interactions that dynamically
modulate key aspects of the process. The development of SU is characterized
by signals (maternally provided or/and created within the system) that constrain
the organization of other cells, so that their developmental fate is appropriately
specified and ensured (initialization of the development of the endomesoderm;
timely separation between mesoderm and endoderm specification and between
anterior and posterior endoderm specification, initiation of gastrulation, etc). And
such a kind of intercellular signaling continues throughout the developmental
process, at least up to the pluteus larva stage. As a matter of fact, with the con-
tinuous increase in the number and links among these intercellular signals, the
S. purpuratus embryo develops an amazing set of features and physiological
and anatomical properties in just 3 days. There is a highly elaborated set of
intercellular and intracellular signals that endows the embryo with an amazing
capacity for regulating its own development. Indeed, cells in the SU embryo are
continuously signaling to each other to ensure that tissues critical to survival of
the embryo are present (Angerer and Angerer 2012).
There are three characteristic properties of these signals: a) they establish
intercellular mechanisms that regulate the developmental process by trigge-
ring, activating and suppressing intracellular processes responsible for the
specification of the developmental fates of the respective cell lineages and for
the modulation of other intercellular mechanisms that will further regulate
the developmental process; b) there are different combinations between inter-
cellular constraining signals, which result in different types of intercellular
mechanisms and consequently, in qualitatively different ways of developmen-
tal regulation; and c) the intercellular operation of these signals is uncoupled
from the reproductive and self-maintaining intracellular processes of the cells
whose development they regulate. More specifically, through the intercellular
mechanisms that coordinate and regulate the embryo’s developmental pro-
cesses, firstly, the specification state of each cell lineage is spatiotemporally

most detailed network described in any embryo, which is constantly updated (for more details,
check the website: http://sugp.caltech.edu/endomes/#UpTo30NetworkDiagram).

Contrastes. Revista Internacional de Filosofía. Suplemento 18 (2013)

Autonomy as a property that characterizes organisms... 367

stabilized; and, secondly, the continuation of the specification process is un-

coupled from the respective intercellular mechanisms ‒among other things
because their characteristic/operational time scales are different. The previous
three properties make the developmental process of S.purpuratus qualititively
different, because there is an effective regulatory control system that operates
intercellularly, at the level of the MC organization, and which autonomously
coordinates its development. This regulatory control system consists of interce-
llular mechanisms sufficiently independent from the intracellular processes of
the assembling units, and which can be varied without disrupting those more
basic intracellular processes, but which are still able to be linked to parts of the
cellular epigenetic mechanisms, modulating their operations. In other words,
there is a dynamic decoupling between the regulatory control mechanism (or
set of controlling subsystems) that act(s) intercellularly in the developmental
process and the controlled processes of the intracellular epigenetic machinery,
responsible for the implementation of the actual state of each cell lineage. This
‘regulatory control system’ constitutes an endogenously created set of specific
functional signals that belong to the (newly developed) MC organizational level
and it is through their constraining action that such a complex developmental
process is effectively driven and stabilized. In addition, the set of developmental
signals gets ‘recursively’ generated along the developmental process itself: that
is, part of these intercellular signals constraining intracellular processes, which
initiate and eventually determine the formation of the diverse cell lineages, also
affect (by inducing or suppressing) the production of other intercellular signals
which, in turn, will constrain the intracellular processes of other cells in the
embryo, and so on and so forth.
In all, S. purpuratus shows a new developmental logic that coordinates
cellular differentiation and integration processes, based on a regulatory control
system which operates in a distinct hierarchical level. This hierarchical level is
decoupled from the underlying metabolic, intracellular needs and is comprised
by several intercellular mechanisms that determine the fate of different cell
lineages, while allowing new possibilities of cell differentiation, and brings
forth in this way a new domain of possible forms of MC organization. In the
next section we briefly argue why this type of MC system with a higher-order
complex regulatory control system, specifically applied to the control and re-
gulation of its developmental activity, is a single functionally integrated and
differentiated developmental unit, and should be conceived as a full-fledged
autonomous organization.

Contrastes. Revista Internacional de Filosofía. Suplemento 18 (2013)

368 argyris arnellos, kepa ruiz-mirazo, álvaro moreno

IV. Developmentally regulated and functionally integrated MC organi-

zations as autonomous systems

As we explained in the previous section, not all developmental MC sys-

tems constituted by genetically homogeneous cells exhibit the same kind and
degree of functional diversity, plasticity and integration in their organization.
MC systems like N. punctiforme and V. carteri, for instance, do not demonstra-
te a MC organization with the capacity to impose regulatory control over the
epigenetic dynamics taking place in the cells. Of course, in both MC systems
(Nostoc and Volvox) there are far-from-trivial cell-to-cell interaction processes
necessary for their operational coordination (division of labor) and collective
pattern generation or global behavior. As an example, in N. punctiforme there
is a rich exchange of metabolites between vegetative cells and heterocysts,
which is necessary to meet the needs of the two cell types. So, they do have
intercellular mechanisms with a constraining effect on intracellular dynamics;
however, these are not sufficiently diverse and recursive to open a new functional
and hierarchical domain that could lead to highly integrated and differentiated
developmental organizations. In other words, although there is cooperation
among cellular entities, there is no an overall control system to coordinate and
regulate development, ensuring its maintenance and reproduction and, thus,
integrating these cells together as a robust entity and a unit of selection in its
own right. Thus, from this theoretical standpoint, N. punctiforme and V. carteri
should not be regarded as autonomous systems.
On the contrary, in S. purpuratus, there is an operational combination of
different types of regulatory mechanisms acting intercellularly, and controlling
the epigenetic intracellular processes, in a way that cellular differentiation is
enhanced and immediately channeled into what could be considered a full-
fledged functionally integrated organization. This form of organization pro-
vides the capacity for robust self-construction of the collective system, which,
through an internally regulated developmental process, is able to reproduce
itself, as well. More precisely, a set of intercellular material structures shaping
developmental processes are endogenously created by the system (in a similar
way as enzymes are endogenously produced by each cellular metabolism),
so that they regulate, modify and control both self-production processes and
processes of exchange with the environment. In other words, in surprisingly
close analogy with a unicellular autonomous organism, this type of MC system
becomes a functionally integrated entity: i.e., its dynamic organization plays
a causal role in the generation of the material structures that actually make it
possible. It is the production of those regulatory control mechanisms that will
trigger off and regulate the development of the functional relations among its
autonomous unicellular parts what turns the whole system into a self-maintaining

Contrastes. Revista Internacional de Filosofía. Suplemento 18 (2013)

Autonomy as a property that characterizes organisms... 369

and self-reproducing integrated organization, becoming itself autonomous, but

at a different hierarchical level. This is essentially why S. purpuratus can be
considered, from this perspective, a true MC autonomous system.
V. Conclusion
There might be an intuitive grasp on the properties that our common sense
takes as important about organisms, but this fails to accommodate many exam-
ples observed in real biology. Then, a more careful and elaborate theoretical
approach is needed to assess thoroughly the concept of organism: an approach
in which the ‘organizational-systemic’ perspective complements the historic
and collective/collaborative dimension. In this perspective, building upon the
autopoietic conception (Maturana and Varela 1973), it has been suggested that
unicellular systems with the capacity for self-construction, which are able to
produce those constraints that ensure the maintenance (and eventually, the re-
production) of their organization, are already autonomous in a basic or minimal
sense (Ruiz-Mirazo and Moreno 2004). Thus, at the unicellular level, a prokar-
yotic or eukaryotic cell, as an autonomous organism, implies a type of material
organization in which its structural components (as ‘parts’) and the cell (as the
‘whole’) are interdependent, maintaining an intricate dynamic relationship.
In other words, a cell is a functionally integrated entity because its dynamic
organization plays a causal role in the generation of the functional structures
that actually make it possible, that actually regulate the (molecular) develop-
ment of its organization. According to a parallel line of argument, candidate
MC systems, in order to become autonomous, must produce (as a result of the
interactions among all the parts involved) a network of functional constraints
that regulate the developmental processes of each of the differentiated parts in
a highly reliable and reproducible way. And, indeed, biology shows that there
are complex MC systems that achieve a full-fledged functionally integrated
and differentiated organization precisely through the regulatory control on its
developmental dynamics. These autonomous MC systems fully deserve to be
regarded as organisms.

Acknowledgements
This work has been supported by grants from the Ministerio de Ciencia e
Innovación FFU2009-12895-CO2-02, Ministerio de Economía y Competitivi-
dad FFI2011-25665 and Gobierno Vasco IT 505-10. Argyris Arnellos holds a
Marie Curie Research Fellowship (IEF-273635). Authors would like to thank
also the organizers of the first conference of the AIFIBI (Valencia, Spain, Nov.

Contrastes. Revista Internacional de Filosofía. Suplemento 18 (2013)

370 argyris arnellos, kepa ruiz-mirazo, álvaro moreno

2012), where this work was presented, and Antonio Diéguez in particular, for
the opportunity to share these ideas with the Ibero-American community of
philosophers of biology and get insightful feedback from them.

References

ANGERER, R.C., ANGERER, L.M., (2012) «Sea Urchin Embryo: Specification of

Cell Fates». eLS John Wiley & Sons Ltd, Chichester. doi: 10.1002/9780470015902.
a0001513.pub3
BUSS, L. W. (1987). The evolution of individuality. Princeton University Press.
CLARKE, E. (2011). «The problem of biological individuality». Biological Theory
5:4, pp312-325
CHRISTMAN, H.D., CAMPBELL, E.L., MEEKS, J.C. (2011) «Global transcription
profiles of the nitrogen stress response resulting in heterocyst or hormogonium
development in Nostoc punctiforme». J. Bacteriology 193(24): 6874-6886.
FOLSE 3rd, H. J., and J. ROUGHGARDEN. (2010). «What is an individual organism?
A multilevel selection perspective». The Quarterly review of biology 85, no. 4:
447-472.
FRANK, S. A. (2003) «Repression of competition and the evolution of cooperation».
Evolution 57, no. 4: 693–705.
GARDNER, A. (2009) «Adaptation as organism design». Biology Letters 5(6): 861-
864
GARDNER, A., and A. GRAFEN (2009) «Capturing the superorganism: a formal
theory of group adaptation». Journal of Evolutionary Biology 22, no. 4: 659-
671.
GODFREY-SMITH, P. (2009) Darwinian Populations and Natural Selection. OUP
Oxford, March 26.
HALLMANN, A. (2011) «Evolution of reproductive development in the volvocine
algae». Sex Plant Reprod 24: 97-112. doi: 10.1007/s00497-010-0158-4
KIRK, D.L. (1998) «Volvox: molecular-genetic origins of multicellularity and cellular
differentiation». Developmental and cell biology series. Cambridge University
Press, Cambridge.
_____, (2005) «A twelve-step program for evolving multicellularity and a division of
labor». BioEssays 27: 299-310
KUMAR. K., MELLA HERRERA, A.R., GOLDEN, W.J. (2010) «Cyanobacterial
Heterocysts». Cold Spring Harb Perspect Biol 2010. doi: 10.1101/cshperspect.
a000315
MATURANA, H., and VARELA, F. J. (1973) De Máquinas y Seres Vivos -- Una
teoría sobre la organización biológica. Editorial Universitaria S.A., Santiago
de Chile.

Contrastes. Revista Internacional de Filosofía. Suplemento 18 (2013)

Autonomy as a property that characterizes organisms... 371

MAYNARD SMITH, J., and E. SZATHMÁRY (1995) The major transitions in

evolution. New York: Freeman.
MEEKS, J.C., CAMPBELL, E.L., SUMMERS, M.L., WONG, F.C. (2002) «Cellular
differentiation in the cyanobacterium Nostoc punctiforme». Arch Microbiol 178:
395-403
MICHOD R. E. (2007) «Evolution of individuality during the transition from unicellular
to multicellular life». Proceedings of the National Academy of Sciences USA 104
(supplement 1): 8613-8618.
MORENO A, RUIZ-MIRAZO K, BARANDIARAN, XE (2011) «The impact of the
paradigm of complexity on the foundational frameworks of biology and cognitive
science». In Hooker CA, Gabbay DV, Thagard P, Woods J (eds) Handbook of
the Philosophy of Science (Vols. 1-10, Vol. Philosophy of Complex Systems, pp.
311–333).
MOSSIO, M., SABORIDO, C. and MORENO, A. (2009) «An Organizational Account
for Biological Functions», British Journal for the Philosophy of Science 60 (4):
813-841.
NEDELCU, A.M., MICHOD, R.E. (2004) «Evolvability, modularity, and individuality
during the transition to multicellularity in volvocalean green algae». In: Schlosser,
G., Wagner, G., (eds) Modularity in development and evolution. University of
Chicago Press, pp 468-489.
OKASHA, S. (2006) Evolution and the levels of selection. Oxford University Press.
PETER, I., DAVIDSON, E.H., (2010) «The endoderm gene regulatory network in sea
urchin embryos up to mid-blastula stage». Dev Biol 340(2): 188-199
QUELLER, D. C, and STRASSMANN, J. E (2009) «Beyond society: the evolution of
organismality». Philosophical Transactions of the Royal Society B: Biological
Sciences 364, no. 1533: 3143.
ROSSLENBROICH, B. (2008) «The theory of increasing autonomy in evolution - a
new proposal for understanding macroevolutionary innovations». Biology &
Philosophy 24: 623-644.
RUIZ-MIRAZO K, MORENO A, (2004) «Basic autonomy as a fundamental step in
the synthesis of life». Artificial Life 10(3):235–259
RUIZ-MIRAZO, K. & MORENO, A. (2011) «Autonomy in evolution: from minimal
to complex life». Synthese (2). doi:10.1007/s11229-011-9874-z.
SANTELICES, B. (1999) «How many kinds of individual are there?». Trends in Ecology
& Evolution 14, no. 4: 152–155.
STRASSMANN, J. E, and QUELLER, D. C (2010) «The Social Organism: Congresses,
Parties and Committees». Evolution 64, no. 3: 605–616.
WEST, S. A, and KIERS, E. T (2009) «Evolution: What Is an Organism?». Current
Biology 19, no. 23: R1080–R1082.
WILSON, J. (2000) «Ontological butchery: organism concepts and biological
generalizations». Philosophy of Science 67: 301-311

Contrastes. Revista Internacional de Filosofía. Suplemento 18 (2013)

372 argyris arnellos, kepa ruiz-mirazo, álvaro moreno

Argyris Arnellos es Marie Curie Research Fellow en el Departamento de Lógica y Filosofía

de la Ciencia UPV/EHU, IAS Research Center for Life Mind and Society

Líneas de investigación:
Filosofía de la biología, sistemas autónomos y complejos, cognición y emoción.

Publicaciones recientes:
ARNELLOS, A. & MORENO, A. (2012) «How functional differentiation originated in
prebiotic evolution». Ludus Vitalis XX(37): 1-23.
ARNELLOS, A., BRUNI E., L., EL-HANI C. & COLLIER, J. (2012). «Anticipatory
Functions, Digital-Analog forms and Biosemiotics: Integrating the tools to model Information
and Normativity in Autonomous Biological Agents», Biosemiotics, 5(3): 331-367.

Dirección electrónica: [email protected]

Kepa Ruiz-Mirazo es Investigador Permanente del Departamento de Lógica y Filosofía de

la Ciencia (UPV/EHU) y de la Unidad de Biofísica (CSIC-UPV/EHU).

Lineas de Investigación:
Origen de vida, filosofia de la biologia, sistemas complejos, modelos protocelulares.

Publicaciones recientes:
RUIZ-MIRAZO, K. & MORENO, A. (2013) «Synthetic Biology: challenging life in order
to grasp, use or extend it». Fothcoming (in Biological Theory).
MAVELLI, F. & RUIZ-MIRAZO, K. (2012): «Theoretical conditions for the stationary
reproduction of model protocells». Integrative Biology (ya accesible online -- DOI: 10.1039/
c2ib20222k)-

Dirección electrónica: [email protected]

Alvaro Moreno es Cetedrático de Lógica y Filosofia de la Ciencia en el Departamento

de Lógica y Filosofía de la Ciencia UPV/EHU y en el IAS Research Center for Life Mind and
Society

Lineas de Investigación:
Filosofia de la Biologia, Sistemas Complejos, Origen de la Cognición

Publicaciones recientes:
RUIZ-MIRAZO, K., & MORENO, A. (2012) «Autonomy in evolution: from minimal to
complex life». Synthese, 185(1): 21–52.
ARNELLOS, A. & MORENO, A. (2012) «How functional differentiation originated in
prebiotic evolution». Ludus Vitalis XX(37), 1-23.

Dirección electrónica: [email protected]

Contrastes. Revista Internacional de Filosofía. Suplemento 18 (2013)