A functional account of degrees of minimal chemical life

Author(s): Mark A. Bedau

Source: Synthese , March 2012, Vol. 185, No. 1, PHILOSOPHICAL PROBLEMS ABOUT LIFE
(March 2012), pp. 73-88
Published by: Springer

Stable URL: https://www.jstor.org/stable/41411209

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms

Springer is collaborating with JSTOR to digitize, preserve and extend access to Synthese

This content downloaded from

159.65.238.93 on Fri, 23 Oct 2020 10:32:44 UTC
All use subject to https://about.jstor.org/terms
Synthese (2012) 185:73-88
DOI 10.1007/sl 1229-01 1-9876-x

A functional account of degrees of minimal chemical life

Mark A. Bedau

Received: 24 June 2010 / Accepted: 6 January 201 1 / Published online: 27 January 201 1
© Springer Science+Business Media B.V. 201 1

Abstract This paper describes and defends the view that minimal chemical li
essentially involves the chemical integration of three chemical functionalities: con
tainment, metabolism, and program (Rasmussen et al. in Protocells: bridging nonlivi
and living matter, 2009a). This view is illustrated and explained with the help of CM
and Rasmussen diagrams (Rasmussen et al. In: Rasmussen et al. (eds.) in Protocells
bridging nonliving and living matter, 71-100, 2009b), both of which represent the k
chemical functional dependencies among containment, metabolism, and program. T
CMP model of minimal chemical life gains some support from the broad view of l
as open-ended evolution, which I have defended elsewhere (Bedau in The philosophy
of artificial life, 1996; Bedau in Artificial Life, 4:125-140, 1998). Further support
comes from the natural way the CMP model resolves the puzzle about whether life
a matter of degree.

Keywords Life • Container • Metabolism • Program • Open-ended evolution •

Degrees of life

1 Introduction

Life seems to be one of the most fundamental categories in nature. A wild variety of
forms of life surround us, and we usually have no difficulty distinguishing the living
from the nonliving. A flower, a worm, and a bird are alive; a rock, a river, and a cloud

M. A. Bedau (13)
Reed College, Portland, OR, USA
e-mail: mab@reed.edu

M. A. Bedau
European School of Molecular Medicine, Milan, Italy

Springer

This content downloaded from

159.65.238.93 on Fri, 23 Oct 2020 10:32:44 UTC
All use subject to https://about.jstor.org/terms
 74 Synthese (2012) 185:73-88

are not. Any attempt to divide na

life and nonlife. Yet it is notorio
definition of life meets with ge
Luisi 1998; Bedau and Cleland 2
that are not themselves alive, a
living and nonliving matter diff
and nonliving things, or is this di
This paper is an attempt to addre
chemical twlife. My answer uses
life: CMP diagrams, and Rasmus
These questions must be faced
life in the laboratory (Szostak
et al. 2009a). The questions are d
oratory might be quite unfamili
(Rasmussen et al. 2003). How can
oratory meets the condition of b
definition of life? Yet that is exa

2 Minimal life as CMP chemic

One solution to this conundrum i

rials, and instead concentrate on
spatial isolation. Most chemical
ferent ways. So, a functional ac
chemical systems, and claim tha
share the right chemical functio
implementation vary from case
kinds of materials can achieve w
boundary conditions any materi
realization of minimal life.
The
functional approach2 that I
tocell
research community, whic
from nonliving materials. Acco

Astrobiologists who are seeking signso

face similar questions (National Researc
extraterrestrial life forms could have ori
uncritically assume that they will exhibit
or producing methane or formaldehyde
if liquid water is absolutely necessary for
to find alien forms of life elsewhere in th
most general, and the chemical details of
2 There is a vigorous literature about func
Tierra (Ray 1992). Langton (1989) famous
examples of life, and Sterelny disagrees
orthogonal to this earlier debate about A
diagrams are real, wet, chemical systems

â Springer

This content downloaded from

159.65.238.93 on Fri, 23 Oct 2020 10:32:44 UTC
All use subject to https://about.jstor.org/terms
Synthese (2012) 185:73-88 75

system is a chemical system tha

such a way that they mutually s
chemical functionality is for the
ing all its components, concentr
operation from molecular parasi
to extract free energy from its
order to maintain and repair itsel
functionality is to put importan
information that is stored in the
during reproduction. These functi
in the sense that they are create
itself, rather than by some exter
for its own continual functioning
These three operational functio
systems: a metabolism that extr
ment, a program that controls c
binatorial information (the fun
the whole system together. In t
terms 'container', 'metabolism',
with minimal constraints on th
tainer will be an amphiphilic str
spatial localization might be ach
icle or micelle surface, or even u
(McCaskill 2009). Similarly, a me
(an alternative is energy from l
ATP and complex enzymes. Furt
control and inheritance without
(e.g., Sagre et al. 2000).
The functionally integrated CM
tionalities of container, metaboli
On top of that, they are chemical
continual operation of each com
with the two other component f
comes about because of chemica
ation is that containers can act a
conditions within a lipid aggreg
from those within bulk water, a
alytic. Another example is that
properties. One final example of
blocks of containers and program
of a metabolism. In each of the
system is supporting the operati

3 Gánti's chemoton (2003) is a CMP funct

support among C, M, and P are all stoich
self-assembly processes.

Springer

This content downloaded from

159.65.238.93 on Fri, 23 Oct 2020 10:32:44 UTC
All use subject to https://about.jstor.org/terms
 76 Synthese (2012) 185:73-88

Fig. 1 The elements of CMP

diagrams of minimal chemical (а) С
life. Chemical subsystems are
shown by capital letters, and
chemical support is shown by
arrows , with support of X M
from P
Y indicated by an arrow from Y
to X. a The three chemical
subsystems in minimal chemical
life: С Container, M
Metabolism, P Program, (Ь, О
b Circular arrows indicate that
each subsystem is chemically
self-supporting, с The CMP
diagram of the paradigm case of
minimal chemical life, in which M P
each of the three functional О о
subunits chemically supports
itself and the other two subunits

(С) О

/А

It is convenient to introduce CMP diagrams to distinguish different kinds of mini-

mal chemical systems. In these diagrams, the letters С, M, and P represent the chemical
functionalities of containment, metabolism, and program (Fig. la). CMP diagrams use
circular arrows to show that these three chemical subsystems are each self-sustaining
(Fig. 1 b). A fully integrated CMP system includes six straight arrows for binary mutual
chemical support among С, M, and P (Fig. lc). The CMP diagram in Fig. lc is the
paradigm case of minimal chemical life, according to the CMP model.
Part of the power of CMP diagrams is that they abstract away from a whole host of
complicated chemical details. The CMP model extends to cover any detailed chemical
story that fits the abstract CMP diagram. For some purposes, further chemical details
matter. Another, more detailed scheme for representing the forms of chemical func-
tional integration in minimal chemical living systems was developed by Rasmussen
et al. (2009b). Figure 2 compares CMP diagrams (left) and Rasmussen diagrams
(right). While CMP diagrams distinguish only three kinds of components (C, M, P),
Rasmussen diagrams subdivide chemical roles more finely: A, aggregate; I, informa-
tional structure; E, energy harvester; F, fuel; M, L, material precursors; pM, pE, pL,
precursors of M, E, and L; E*, energized form of E. In addition, while CMP diagrams
represent all kinds of functional support with just one arrow, Rasmussen diagrams dis-
tinguish a number of different kinds of functional dependencies: black arrow, chemical
reaction; blue circle dotted arrow, chemical catalysis; red double arrow, self-assembly
process; green diamond dotted arrow, energy transfer; yellow dotted enclosure, spatial
concentration. Some Rasmussen diagram arrows point to other arrows, if they catalyze

ô Springer

This content downloaded from

159.65.238.93 on Fri, 23 Oct 2020 10:32:44 UTC
All use subject to https://about.jstor.org/terms
 Synthese (2012) 185:73-88 77

(a> -1-pL

M hv^^hv*

p - frpM t>M P I

(b) fti ^ ~7'^

ö-^j ^ X- ^
Fig. 2 Rasmussen et al. (2009b) developed a more detailed representation of CMP systems, a The
Rasmussen diagram (right) of the three chemical subsystems in minimal chemical life, and in a CMP
diagram (left), b Minimal self-supporting functional triad of container, metabolism, and program, repre-
sented in a CMP diagram (left) and a Rasmussen diagram (right). The Rasmussen diagrams show a high-level
chemical representation of what underlies CMP diagrams

reactions. One important upshot is that many different Rasmussen diagrams collapse
to the same, single CMP diagram. The CMP model is a high-level abstraction.
Borderline cases are a fairly severe test of a theory. For theories of minimal chemi-
cal life, the most famous borderline case is the virus. Viruses typically are biologically
active only when they inhabit other cells as hosts; an available minimal chemical cell
host is shown in Fig. 3a. An isolated virus particle (Fig. 3b) is not considered alive by
most biologists today; note that the CMP diagram for the isolated virus (3b) is quite
different from the diagram of minimal chemical life (3a). This stark difference in the
virus diagram counts in favor of the CMP theory. But when a virus infects a cell and
inhabits it, and highjacks the cell's CMP machinery to copy the virus (Fig. 3c), then
the composite chemical system has a slightly more complex CMP diagram. Under
these conditions, the virus is part of a living system - another feature well represented
in CMP diagrams. So, the CMP model passes the virus test. The different states of a
virus (inert and infectious) are easily represented in CMP diagrams, and the diagram
of the infected host (3(c)) shows that the parasitic virus takes from C, M, and P but
gives them nothing back.
The virus example reveals something important about the CMP model. Viruses
are a classic borderline case, and borderline cases are sometimes due to insufficient
information. By contrast note that the CMP account of why viruses are borderline is
not epistemological but structural. The problem is not that we do not know enough

Springer

This content downloaded from

159.65.238.93 on Fri, 23 Oct 2020 10:32:44 UTC
All use subject to https://about.jstor.org/terms
 78 Synthese (201 2) 1 85:73-88

Fig.
pre-existing
3 A virus
form
depends
of (a)on
life, Oahost
its (b)
(a). A virus particle all by itself
(b) is not alive but inert. When
the virus infects the host, the
virus becomes part of a more
complex living system (c)
MÎ=^P
MÎ=^P 4% p*
и и

<c> Л

fá?
M > P
<J V

about viruses to tell whether they are alive. Rather, we know all t
viruses, and we can capture the relevant information in a CMP
isolated virus particle is an inert polymer (Fig. 3b), perhaps inside
this virus particle infects a host (represented as a CMP protocell,
system provides the chemical resources that the virus particle nee
although the virus fails to support the containment, metabolism,
(Fig. 3c). The whole chemical system diagrammed in Fig. 3c, in
its host protocells, is itself alive. Once we have sorted out the
virus infecting a host, is there any further question to resolve
viruses are alive? I cannot imagine what that question might be
is that minimal chemical life is a matter of degree because ea
located on a precisely structured landscape of systems that diff
their neighbors and that are more or less alive. I return to this to
At this stage one should ask whether our earlier puzzles about
guise. If there is no agreement about a definition of life, why s
giance to a model that says minimal cellular life is a chemically
triad of container, metabolism, and program? A host of furthe
Are all three functionalities really necessary? Are no further f
sary? One can imagine further candidate functionalities, such
ior, autonomous information processing, responsiveness to the
to promote self-interest, or behaving purposefully. Why are cont
and program by themselves sufficient? The rough consensus in
cell community around the functional triad view gives us reason
ously, but the agreement will not convince skeptics that the CM
especially since the "consensus" around the CMP model in the
community is not universal.
One way to reply to this worry would be to agree that other
autonomous information processing and sensitivity to the en
properties of living systems, but to hold that they can be explain

40 Springer

This content downloaded from

159.65.238.93 on Fri, 23 Oct 2020 10:32:44 UTC
All use subject to https://about.jstor.org/terms
 Synthese (201 2) 1 85:73-88 79

Perhaps the CMP model could exp

like autonomy, autopoiesis, and sel
of reply here, because that kind o
about the functional integration o
of the CMP capacities suffice to e
CMP capacities? A deeper explana
questions.

3 Life as open-ended evolution explains CMP functionality

I would explain the CMP model as a consequence of a more fundamental view of life,
a view that locates the essence of life as the exercise of the capacity to undergo autono-
mous proliferation and adaptation, ultimately involving to open-ended evolution. The
idea that life essentially involves the ability to evolve by natural selection has a long
pedigree (e.g., Maynard Smith 1975; Cairns-Smith 1985; Joyce 1994; Bedau 1996;
Ruiz-Mirazo et al. 2004). I want to defend a specific kind of view of life as primarily
identified with the holistic property of open-ended evolution (Bedau 1996, 1998).
One thing in favor of my view of life as open-ended evolution is that is explains
the central place of the CMP model. If life fundamentally involves the process of
open-ended evolution, then we should expect minimal cellular life forms to consist
of a functionally integrated triad of container, metabolism, and program. The process
of natural selection requires a varied population of reproducing entities with functional
properties under the control of heritable information. The heritable informational con-
trol of functional properties is what we have been referring to as the "program" in a
minimal cell. In order for programs to control cellular functionality, they need a source
of raw materials and energy from the environment, and that is provided by what we
have been calling a metabolism. Finally, what we have been calling a container is
needed in order to spatially concentrate and protect the molecular components of the
program and metabolism while the system assembles and organizes itself, grows and
finally reproduces.
Note also that, if there is a combinatorial^ large family of possible forms of infor-
mational control of metabolism and containment, and if information inheritance is
somewhat imperfect, then the process of evolution by natural selection would take
place. In time, natural selection might be able to improve those functionalities and
their integration. Figure 4 is a crude cartoon of how a population of reproducing CMP
systems could evolve by the process of natural selection. Truly open-ended evolution
might require an appropriate approximation of unlimited genetic variation, but I will
leave those details unsettled here, since they do not affect my overall argument.
Let me digress to prevent one misunderstanding of my view of life. I am not inter-
ested in the meaning of the word 'life' (or its translations in other languages). Anyone
who does not know the meaning of the word can look it up in a dictionary. I am asking
a different question, one about the true nature of the phenomena referred to by that
word. Furthermore, I am not interested in the analysis of our current conception of life.
Our concepts are full of preconceptions, and when viewed on a very long time scale,
they are continually in flux. There is every reason to think that the concepts of life

ô Springer

This content downloaded from

159.65.238.93 on Fri, 23 Oct 2020 10:32:44 UTC
All use subject to https://about.jstor.org/terms
80 Synthese (2012) 185:73-88

m==5P Д
m==5P
V AV

a
Л о

m==5P Д M2=^P Ä
m==5P M2=^P
uAw V

A' uAw V
MÏ Â Jp m! Ä Jp M" ДбЬ ¿P m?
MÏ

V ut) W

O
A I IV O Г) л л

A Ami А А А
м? - <>Р м- - IP yLLJr м 5 - Z? M? - ZP м 5 - <>Р
V W DD w fj
Fig. 4 Evolution of CMP struc
with arrows and stop signs. A l
by natural selection

actively used and distribu

This conceptual evolution
questionI am asking. I am
conceptualizations of it.
One way to construct a v
ply to collect examples of
But from my point of v
mistakenly ignore all tho
Iam trying to understand
encompass all possible fo
Such a theory of the nat

Springer

This content downloaded from

159.65.238.93 on Fri, 23 Oct 2020 10:32:44 UTC
All use subject to https://about.jstor.org/terms
 Synthese (2012) 185:73-88 81

But I believe that constructing

way to make progress toward be
among others).

4 A theory of life should explai

I believe that views about the na

ability to explain the fundamen
Any theory or model of minima
characteristic properties of life i
Which specific "fundamental ph
They certainly should include th
about life, including borderline
definition of life, there is broad
hallmarks. In addition, a numbe
resurface in discussions about wha
Various hallmarks of life have
e.g., Mayr 1997; Koshland 2002;
list from 1982 could be paraphra

1. All levels of living systems h

tion.
2. Living organisms are composed of a chemically unique set of macromolecules.
3. The important phenomena in living systems are predominantly qualitative, not
quantitative.
4. All levels of living systems consist of highly variable groups of unique individuals.
5. All organisms possess historically evolved genetic programs which enable them
to engage in teleonomic processes and activities.
6. Classes of living organisms are defined by historical connections of common
descent.
7. Organisms are the product of natural selection.
8. Living processes are especially unpredictable.

Any adequate theory of life should explain why such heterogeneous properties char-
acteristically coexist in nature, especially since each of the hallmarks can be possessed
by things that are not alive.
One important virtue of the view of life as open-ended evolution is its unified
explanation for Mayr's hallmarks of life; the view implies that we should expect those
heterogeneous-seeming properties to coexist in nature. If life consists primarily of
systems undergoing open-ended evolution, we should expect life to arise when natu-
ral selection produces complex adaptive organization in historically connected organ-
isms with evolved genetic programs. Furthermore, the random variation and historical
contingency in open-ended evolution explain why living phenomena are especially
unpredictable and involve unique and variable individuals. Finally, if open-ended evo-
lution is produced by a branching process involving birth, reproduction, and death of
individuals, then we can understand why it would give rise to a wealth of qualitative

â Springer

This content downloaded from

159.65.238.93 on Fri, 23 Oct 2020 10:32:44 UTC
All use subject to https://about.jstor.org/terms
82 Synthese (2012) 185:73-88

phenomena characterized by fro

The naturalness of these explanat
process of open-ended evolution.
of life, and the theory of life as
argument in favor of this view of
The phenomena of life also includ
ing these: Is the distinction betw
is life hierarchical, and how are
of life involve matter, form, bo
fundamentally related? There is
is life?" so controversial and diffi
These puzzles are controversial a
resolve the puzzles and also exp
elsewhere that viewing life as op
each of these puzzles (Bedau 199
theory of life should be evaluate
cases, and puzzles. To illustrate t
further look at one specific puzzle
or a scale. Seeing how the CMP
one final example of the way in

5 Minimal chemical life is a mat

Common sense leans towards the

A rabbit is alive and a rock is no
something being partly but not
stress by borderline cases like vi
by spores and frozen sperm whi
then "come back to life" when c
that the original life forms on Ea
If you focus on the chemical det
incremental chemical steps. It is
systems, then make a series of n
and end up with a chemically self
sorts of puzzles about whether or
signal strong support for a theory
According to the CMP function
chemical system containing thre
gram. In addition, each of those
as well as the operation of the t
there are nine distinct kinds of
container supporting the contain
supporting the metabolism, the co
the program, the container, or th
correspond to the nine kinds of

Springer

This content downloaded from

159.65.238.93 on Fri, 23 Oct 2020 10:32:44 UTC
All use subject to https://about.jstor.org/terms
 Synthese (201 2) 1 85:73-88 83

M s Á
- <>P м" - >ps M 5 м" À

M
/Д /Д' Â /Д'
e F M с P M tP M if
О V и V <J
Fig. 5 Diagrams of some of the ways in which CMP subsys
support each other. CMP diagrams can contain up to nine ar
metabolism, and program. In CMP diagrams none, one, tw
can have a circular arrow , independently of any other ar
are fully integrated functionally, so the six inner arrows f
arrows . For example, the arrows С -> M and С -> P regi
and program to function properly, because the container co
parasites and poisons

Now, consider simpler systems with a contain

in which some of those kinds of functional su
to count the number of kinds of such systems.
arrows is present or absent, so there are 29 = 5
As it happens, any single given arrow could be
ent kinds of chemical processes, so the figure o
number of different kinds of chemical systems
the nine-arrow CMP system. This correction only
argument.
The CMP diagrams that include all six inner a
arrows are shown in Fig. 5. One could argue t
is lost if any subsystem is missing a circular a
itself, mainly by repairing itself and by grow
arrow represents this self-reproduction. The p
(possibly with some error) to survive across m
represents this replication. The metabolism nee
have the materials and energy it needs to remain
that all of the circular arrows are needed to en
Are all of the straight arrows in the middle
CMP diagrams (omitting letters and circular ar
arrows are shown in Fig. 6, and those that om
Fig. 7.
The chemical systems shown in Fig. 6 are the cornerstone of an argument that life
comes in degrees. We would probably all (or almost all) agree that none of the CMP
systems in Fig. 6 are fully and simply alive. They are each missing something central

Springer

This content downloaded from

159.65.238.93 on Fri, 23 Oct 2020 10:32:44 UTC
All use subject to https://about.jstor.org/terms
 84 Synthese (201 2) 1 85:73-88

Д /Д /д' л /л /д
(а) (Ь) (с) (d) (e) (f)
Fig. 6 Six kinds of CMP diagrams with one
configurations of circular arrows, dependin
this figure implicitly assume that the cont
corner , and the program (P) is at the lowe
from container to metabolism; this means tha
the container. In diagram (c) the metabolism
help the proper functioning of the metaboli

Д Д Д Л
Д A A A

Л /д /А /д

Д Л 4L
Fi
ar

to
sy
al
li
ar
b
T
R
d
ar
re
If
d

â Springer

This content downloaded from

159.65:ffff:ffff:ffff:ffff on Thu, 01 Jan 1976 12:34:56 UTC
All use subject to https://about.jstor.org/terms
Synthese (2012) 185:73-88 85

С I ^ .О | Г ~ Nfrv
-■-»PM --»pM - -op M

-•~*pEH>E ' E* - -t>pE-ÌEv E* -;-OpE->E4 E*

hv* hv* hvT hv* ¡г/ Miv*
ч

.С ^ ^ ^
" -*>PM - "»pM - -op M

~ -opE -^E E* - -opE-£E E* - -opE4>E E*

ч -opE

Yef hv* h уГ Ahv* ¡л? ЛЬу*

Л | ^ | , ( |
- -QpM - ppM & - QpM

'ÇPE h/h/
EXE Ahv*
Ahv* ^ Vehv^
hv^ EWE hv* ^hv^
hv* Ve hv^Xhv*
eWE Xhv* ^
Fig. 8 A set of Rasmussen diagrams which differ
for minimal chemical life only by omitting one catal
canonical Rasmussen diagram is alive, then each of

Figure 9 shows the space of all possible

cases of systems that are considered to b
cases of systems that are not alive (far le
of functional support are not alive (Rasm
nine kinds of functional support (the ni
keep a system from being alive. Betwee
living is a grey zone of CMP systems tha
A qualification is in order. Perhaps not
by the CMP diagrams in Fig. 9 are chem
impossible to create in the wet lab. Furth
are independent; the existence of some a
of some other arrows. There is a huge ne
program systems that have integrated c
one of these chemical systems can be lo
systems. Systems are located on that map
this network of possible CMP integrated
mally, in only one arrow. Any two syste
apart, measured by the minimal number
system into the other (their Hamming d

Springer

This content downloaded from

159.65.238.93 on Fri, 23 Oct 2020 10:32:44 UTC
All use subject to https://about.jstor.org/terms
86 Synthese (201 2) 1 85:73-88

space of minimal CMP di

Fig. 9 The space of all possible 29 = 512 mi

in this space differ in only one arrow. At
things as inert vesicles or plasmids; they ar
CMP diagrams represent the simplest proto
today. In between these two poles is a grad
spectrum is given a precise and complete c

We have identified no sharp line

randomly picks a CMP diagram fro
fully alive nor fully nonliving; it
any number of precise distinctions t
integrated systems, and one could
fully alive," but I recommend agai
drawn, and I see no principled re
line. CMP diagram space consists
integrated CMP systems; and the s
of kinds of chemical systems that
Some are clearly not alive, and oth
in between. That is all there is to
shades of grey in the space of 5 12
life/nonlife distinction beyond thi
distinction. Identifying a precise d
the living and the nonliving is to i
in nature.

4 Of course, smoothly changing a paramete

one classic example is the emergence of a
evidence that the life/nonlife distinction inv

Springer

This content downloaded from

159.65.238.93 on Fri, 23 Oct 2020 10:32:44 UTC
All use subject to https://about.jstor.org/terms
 Synthese (2012) 185:73-88 87

6 Conclusion

This paper connects two very different perspectives on life: (1) the CMP model
of minimal chemical life as involving the chemical integration of containment,
metabolism, and program (Rasmussen et al. 2009a), and (2) the broad view of life
as the process of open-ended evolution, which I have defended previously (Bedau
1996, 1998). Constructing minimal chemical forms of life in the laboratory requires
one to target some particular kind of chemical systems, and the protocell research
community is largely aiming to make systems that chemically integrate containment,
metabolism, and program, and thereby are able to grow, reproduce, and in the long
run evolve. In this way, CMP systems are the minimal chemical requirements for a
population of CMP systems that reproduce and evolve by natural selection. For this
reason, the CMP model is a natural ally of the view of life as open-ended evolution.
Cmp and Rasmussen diagrams crisply identify the kinds of functional support
involved in these minimal chemical forms of life. Furthermore, the space of CMP
systems provides a way to categorize and grade minimal chemical systems on the
degree to which they are alive. This resolves one of the long-standing puzzles about
whether or not the life/nonlife distinction is a dichotomy. And all of this illustrates the
view that explaining life hallmarks and puzzles is centrally involved in defending any
theory of life (Bedau 1998, 2007).

Acknowledgments The methods for diagramming CMP systems, and the emphasis on the CMP model of
life, are deeply indebted to the representations of protocell chemical organization in Rasmussen diagrams,
which were developed in collaboration with Rasmussen et al. (2009b). My colleagues are not responsible
for how I have applied our earlier work. Thanks also to the audience at Artificial Life XI, August 2009,
when an earlier version of this work was presented. For helpful discussion, thanks to Kate Elgin.

References

Bedau, M. A. (19%). The nature of life. In M. Boden (Ed.), The philosophy of artificial life (pp. 332-
357). New York: Oxford University Press.
Bedau, M. A. (1998). Four puzzles about life. Artificial Life, 4 , 125-140.
Bedau, M. A. (2007). What is life? In S. Sarkar & A. Plutynski (Eds.), A companion to the philosophy
of biology (pp. 455-471). New York: Blackwell.
Bedau, M. A., & Cleland, C. E. (Eds.). (201Ü). The nature oj lije: Classical ana contemporary
perspectives from philosophy and science. Cambridge: Cambridge University Press.
benner, s. A., Ricardo, A., & uarngan, M. A. is mere a common cnemicai moaei юг me m
the universe? Current Opinion in Chemical Biology ; 5, 672-689.
Cairns-Smith, A. G. (1985). Seven clues to the origin of life. Cambridge: Cambridge University
Deamer, D. (2005). A giant step towards artificial life? Trends in Biotechnology, 23 , 336-338.
Farmer, D., & Belin, A. (1992). Artificial life: The coming evolution. In C. Langton, C. Ta
J. D. Farmer, & S. Rasmussen (Eds.), Artificial life lì (pp. 815-840). Redwood City, CA: Ad
Wesley.
Gánti, T. (2003). The principles of life, with commentary by James Griesemer and Eörs Szathmáry. Oxford:
Oxford University Press.
Joyce, G. F. (1994). Forward. In D. W. Deamer & G. R. Fleischaker (Eds.), Origins of life: The central
concepts (pp. xi-xii). Boston: Jones and Bartlett.
Koshland, D. E., Jr. (2002). The seven pillars of life. Science, 295 , 2215-2216.
Langton, С. u. (1У8У). Artificial Ute. in U. u. Langton (eas.j, Artijiciai nje [¿ama re msiuuie siuaies m
the sciences of complexity, proceedings vol. IV) (pp. 1-47). Redwood City, CA: Addison- Wesley.

Springer

This content downloaded from

159.65.238.93 on Fri, 23 Oct 2020 10:32:44 UTC
All use subject to https://about.jstor.org/terms
 88 Synthese (2012) 185:73-88

Luisi, P. L. (1998). About various defin

28 , 613-622.
Luisi, P. L. (2006). The emergence of l
Cambridge University Press.
Maynard Smith, J. (1975). The theory
Mayr, E. (1982). The growth of biologi
Mayr, E. (1997). This is biology : The sci
McCaskill, J. S. (2009). Evolutionary
S. Rasmussen, M. A. Bedau, L. Chen,
Bridging nonliving and living matte
National Research Council of the Nat
life in planetary systems. The Nationa
php?record_id=11919. Accessed Sept
Rasmussen, S., Bedau, M. A., Chen, L.
(2009a). Protocells: Bridging nonlivi
Rasmussen, S., Bedau, M. A., McCask
S. Rasmussen, M. A. Bedau, L. Chen,
Bridging nonliving and living matte
Rasmussen, S., Chen, L., Nilsson, M
ter. Artificial Life, 9, 269-316.
Ray, T. S. (1992). An approach to the
& S. Rasmussen (Eds.), Artificial life
proceedings vol. X) (pp. 371-408). Re
Ruiz-Mirazo, K., Pereto, J., & More
open-ended evolution. Origins of Lif
Sagre, D., Ben-Eli, D., & Lancet, D. (2
in mutually catalytic noncovalent as
USA, 97 , 4112-4117.
Sterelny, K. (1995). Understanding life
Philosophy of Science, 46, 1 15-183.
Szostak, J. W., Bartel, D. P., & Luisi,
Taylor, C. (1992). "Fleshing out" artifici
(Eds.), Artificial life II (pp. 25-38). R
Wimsatt, W. C. (1987). False models
(Eds.), Neutral modes in biology (pp

Ô Springer

This content downloaded from

159.65.238.93 on Fri, 23 Oct 2020 10:32:44 UTC
All use subject to https://about.jstor.org/terms