CAUSATION AND THE ORIGIN OF LIFE. METABOLISM OR
Department of Chemistry, Ben-Gurion University of the Negev, Beer Sheva, 84105, Israel
(Received 7 January 2003; accepted in revised form 27 January 2003)
Abstract. The conceptual gulf that separates the ‘metabolism first’ and ‘replication first’ mechan-
isms for the emergence of life continues to cloud the origin of life debate. In the present paper we
analyze this aspect of the origin of life problem and offer arguments in favor of the ‘replication first’
school. Utilizing Wicken’s two-tier approach to causation we argue that a causal connection between
replicationand metabolismcan onlybedemonstrated ifreplicationwould havepreceded metabolism.
In conjunction with existing empirical evidence and theoretical reasoning, our analysis concludes
that there is no substantive evidence for a ‘metabolism first’ mechanism for life’s emergence, while a
coherent case can be made for the ‘replication first’ group of mechanisms. The analysis reaffirms our
conviction that life is an extreme expression of kinetic control, and that the emergence of metabolic
pathways can be understood by considering life as a manifestation of ‘replicative chemistry’.
Keywords: causation, chemical evolution, metabolism first, molecular replication, origin of life,
replication first, teleology, teleonomy
The resurgence of interest in the origin of life on earth that began early in the 20th
century has been largely driven by the conviction that all living beings evolved
from inanimate matter, and that life is no more than a complex set of physico-
chemical processes (for recent reviews see: Fry, 2000; Lahav, 1999; Orgel, 1998,
1992; Lifson, 1997; Wächtershäuser, 1997; Miller and Lazcano, 1996; Maynard
Smith and Szathmáry, 1995; Chyba and McDonald, 1995; Elitzur, 1994; Eigen,
1992). However, even a cursory look at existing theories reveals a major concern.
Many of the evolutionary hypotheses are quite different to one another and often
based on quite different premises. For example, where on the planet did life emerge
– on the earth’s surface in a prebiotic soup (Oparin, 1957), or in hydrothermal
vents under the seas (Wächtershäuser, 1992)? Or did early life arrive here from
some extraterrestrial source (Melosh, 1988)? What was the composition of the
early atmosphere that led to life’s emergence – a reducing atmosphere of methane,
hydrogen, ammonia and water, or a neutral one containing mainly carbon dioxide,
nitrogen and water (Kasting, 1993)? Did life initially emerge from a prebiotic
aqueous environment, as was widely believed till quite recently, or did it derive
from two-dimensional metabolic organization – possibly on iron sulfide (Wächter-
shäuser, 1997), or clay surfaces (Cairns-Smith, 1985)? And given that life systems
Origins of Life and Evolution of the Biosphere 34: 307–321, 2004.
© 2004 Kluwer Academic Publishers. Printed in the Netherlands.
are primarily characterized by a metabolism and the ability to reproduce, which
came first – metabolism or replication? The issue of ‘metabolism first’ or ‘replic-
ation first’ is a particularly fundamental one since it bears on the very nature of
biological complexification. The uncertainty surrounding this issue effectively de-
clares that the physico-chemical principles that were responsible for the process of
complexification remain unclear. So it is not just historic questions such as, where
on earth did life originate, that await more definitive answers. Even the general
principles by which animate matter emerged remain unresolved and in dispute.
The purpose of this paper is therefore to address the issue of ‘metabolism first’
or ‘replication first’ in order to help clarify an ahistoric aspect of emergence – to
better understand the physico-chemical principles that enabled matter to undergo
a process of biological complexification, that eventually led to simple life forms.
By applying Wicken’s (1985) two-tier approach to causation, we believe a sound
causal argument can be made that life began with replication, not metabolism.
As noted above the current scientific confusion associated with the emergence of
life debate appears to stem from both historic and ahistoric uncertainties. Consider
historic uncertainty first. As the conditions that existed on the prebiotic earth are
very uncertain, there are enormous practical difficulties in assessing life’s mech-
anistic beginnings. The elucidation of a reaction mechanism is often problematical
evenwhen thereaction parameters and conditions (reactant concentrations, solvent,
pH, temperature, pressure, catalysts, etc.) are all specified. Yet in addressing the
question of life’s emergence, we have to acknowledge the fact that there is almost
nothing we know about that primordial process with any degree of certainty; we
are unable to identify the initial reactants, nor the reaction conditions for the pro-
cess that initiated that evolutionary transition from inanimate to animate. Clearly
then, attempts to formulate a reaction mechanism for a process whose key reagents
and reaction conditions are largely unknown, necessarily becomes a speculative
But we would argue that in a fundamental sense the precise evolutionary path-
way by which life emerged is actually a secondary one, and needs to be addressed
within the context of a more basic ahistoric question: what is the nature of the
physical and chemical laws that governed that evolutionary process of emergence?
Wächtershäuser (1997) termsthe science ofchemistry ‘an ahistoric science striving
for universal laws, independent of space and time or of geography and the calen-
dar’. Accordingly, in agreement with Wächtershäuser we believe the primary (and
more realistic) challenge regarding the origin of life remains to reduce the historic
process of evolution with all its inherent uncertainties, many of which are unlikely
to be ever resolved, to a universal chemical law of evolution, whose validity would
be independent of time and place. So, at least in the first instance, rather than ask:
CAUSATION AND THE ORIGIN OF LIFE
what was the mechanistic path that led inanimate matter to the earliest life forms,
we need to ask what general features would characterize the transition from in-
animate to animate. What characteristics would define that family of mechanisms
that could, in principle at least, lead to the emergence of life? Even though it is
generally acknowledged that all biological systems derived from a process of com-
plexification, the physico-chemical principles that would lead to the emergence
of such highly complex far-from-equilibrium systems remain unclear. A clearer
physico-chemical understanding of the principles underlying the emergence of life
would also go some way toward resolving the long-standing (and fractious) debate
regarding the definition of life. Asrecently discussed by Cleland and Chyba (2002),
the possibility of defining life in an unambiguous manner rests on our ability to
characterize life as a ‘natural kind’ (Putnam, 1973; Schwartz, 1977). However the
ability tocharacterize life asa‘natural kind’ in turn depends on amorefundamental
understanding of the physico-chemical principles that would have governed life’s
The above considerations lead us directly to the question of ‘metabolism first’
or ‘replication first’. Two of life’s most striking characteristics are: (a) its highly
complex metabolic system, and, (b) its replicating capability. So can we, as a first
step in describing the process of life’s emergence, at least in general terms, determ-
ine which of these two characteristics emerged first? In contrast to many aspects
of emergence this question is unlikely to be historical in character, as there may
well be a causal connection between these two key life characteristics, one that is
directly linked to the essence of life itself and indirectly to the nature of matter. In
order therefore to address this question from a causal perspective we make some
preliminary comments regarding causation.
2.1. TWO-TIER CAUSATION – DRIVING FORCE AND MECHANISM
The question as to what causes change and motion in the universe has been the
subject ofscientific enquiry since the timeofAristotle and before. Though Aristotle
outlined the framework for modern scientific thought, the teleological viewpoint
that he helped establish proved unable to adequately explain the design and order
so evident in the world, as manifested most strikingly in living systems. It was only
during the seventeenth century, two thousand years after Aristotle, that the modern
philosophical foundations of the natural sciences were laid down by Descartes,
Bacon and others. The essence of that change was that the deeply entrenched
teleological view of nature, in which purpose underlay natural phenomena (that
is, processes occurred as means to achieve particular ends) was replaced by a
mechanical-mechanistic world view, in which all of nature, including living sys-
tems, behaved according to well-defined laws of nature. The view attributed to van
Helmont (1648): ‘All life is chemistry’, characterised this new way of thinking,
and was reinforced 150 years later by Immanuel Kant (1952), though interestingly,
Kant’s view of living systems as a ‘natural purpose’ remained teleological, and led
to his famous comment that there could never be a Newton able to explain ‘a single
blade of grass.’
In line with modern scientific methodology, recent theories regarding the emer-
gence of life tend to be mechanistic in their approach. Most address the mechanistic
how question rather than the ostensibly inappropriate teleological why question.
The why question seems to take us back to that discredited methodological ap-
proach in which purpose played such a dominant role. But the why question can be
asked with a non-teleological intent. It can be asked in the sense of seeking out the
driving force for some phenomenon. As Wicken (1985) has argued, asking the why
question remains an essential component of scientific understanding; a two-tiered
approach tocausation isoften crucial inobtaining aproper understanding ofnatural
phenomena. Understanding whys is no less important to discovering hows and in
fact in many cases may be an important preliminary step before the mechanistic
how can be tackled. To make the point clear and to reaffirm the need for two-tier
causation let us consider two examples – one physical in orientation: how and why
does water flow on a given surface, and one chemical in orientation: how and why
do chemical reactions occur?
Water flow may manifest itself in different ways: as a leaking roof, a small
stream, a winding river, a waterfall, or there may be no flow at all, as in a desert,
a pond, or a lake. Inspection of a physical surface, its slope and composition, and
knowing the volume of water that is directed at that surface may well provide
a reasonable answer to the question: how will water flow? But the basis of any
attempt to answer the question of how water flows requires an understanding of
why water flows. Newton’s generalization regarding the existence of a universal
gravitational force provides a powerful framework for answering the why question
and this understanding then enables the how question to be usefully addressed.
Mayr (1988) has termed such causal explanations (e.g., gravity as the explanation
for water flow) as teleomatic, the term signifying that the consequence of a process
(in the above case a lowering of potential energy) is offered as the reason for its
Let us now move closer to home and consider the question of how and why
chemical reactions occur. In the present context the example is particularly per-
tinent given the modern view that living beings are ultimately nothing more than
complex chemical systems. In posing the above question the how enquires into
the mechanism of a particular reaction – that is, the detailed description of the
individual steps that lead from reactants to products. Forexample, does the reaction
take place in a concerted one-step process, or are intermediates formed along the
way? Is some form of energy input, such as irradiation, required for the reaction to
occur? Is a catalyst involved? Does the reaction take place in solution or on some
catalytic surface? Understanding how a reaction occurs - its mechanism, provides
us with a measure of control over that reaction, suggesting ways to speed it up or
slow it down, or possibly help eliminate unwanted side-reactions.
CAUSATION AND THE ORIGIN OF LIFE
But, as in the case of water flow, understanding why the reaction occurs is a
necessary preliminary step to understanding how it occurs, and it is the concep-
tual framework of thermodynamics that provides an answer to this question. The
science ofthermodynamics teaches usthat energy flowhas direction; closed macro-
scopic chemical systems react because there is a driving force that strives to lower
the free energy of the reacting system till it reaches the point of minimum free
energy – the so-called equilibrium state. Of course whether a particular reaction
actually takes place also depends on the potential mechanism of the reaction. Thus
for a macroscopic chemical system to react, both a thermodynamic driving force
and a kinetically feasible mechanism are required. Here again we see that for a
proper understanding of the factors controlling a chemical process, we need to
answer both themechanistic how andthe teleomatic whyquestions. Thermodynam-
ics is the science of the possible; kinetics and mechanism determine whether the
possible will or will not occur. As we will subsequently see, this two-tier approach
to causation may prove beneficial in our attempts to understand the process of life’s
2.2. METABOLISM FIRST OR REPLICATION FIRST
The debate between proponents of ‘metabolism first’ and ‘replication first’ con-
tinues unabated with both approaches subject to criticism. The ‘metabolism first’
approach has been criticized by some of the leading workers in the field (Maynard
Smith and Szathmáry, 1995; Orgel, 1992; Eigen, 1971; Lifson, 1997) based on the
assessment that key steps in the building up of such a metabolic system are highly
improbable. The ‘replication first’ approach is questioned based on the view that
the de novo appearance of oligonucleotides is improbable, and that there is no
clear path from an RNA world to the current dual world of proteins and nucleic
acids (Shapiro, 1984, 2000). We wish to address this key issue and in particular to
inquire how the ability ofliving things toboth replicate and metabolize relate toone
another. Within the context of life’s emergence, could there be a causal relation-
ship between these two life characteristics, and, if so, can that causal connection
assist us to deduce which came first? A reasonably definitive answer to the above
question would be a worthwhile step forward as it would significantly narrow the
range of existing mechanistic possibilities for the emergence of life.
In considering the emergence of life problem in the context of replication and
metabolism, one can in principle consider three alternative scenarios: (a) that meta-
bolism preceded replication, (b) that replication preceded metabolism, and (c) that
the process of emergence led to a primal system that was both metabolic and
replicative. Categories (a) and (c) actually belong to the same conceptual class
of ‘metabolism first’ mechanisms since it is the process by which a metabolic
system can emerge (with or without a replicative capability), that characterises
the ‘metabolism first’ grouping. In reality the more widely known ‘metabolism
first’ mechanisms belong to the (c) category, i.e., they propose that the emerging
primal system was both metabolic and replicative, though the primal replicative
capability is considered to have been compositional rather than genomic. Let us
now consider the two views – ‘metabolism first’ and ‘replication first’ – from
empirical, theoretical and causal standpoints.
2.2.1. Metabolism First
First, let us be clear as to what we mean by the term metabolism. Our usage con-
forms with conventional terminology and refers to that complex set of co-ordinated
and regulated chemical reactions present in all living beings, both autotrophic and
heterotrophic, whose primary role is to provide living entities with the necessary
energy to fuel and maintain the organism’s functions. The group of mechanisms for
life’s emergence that can be classified ‘metabolism first’, though differing greatly
in their other features, contend that the emergence of metabolism emerged prior to
(or simultaneously with) the emergence of a replicating capability. As a group this
school proposes that metabolism emerged either spontaneously or by a process of
random drift, and once established, may have also exhibited a crude, non-genomic
replicating capability (Dyson, 1985; Kauffman, 2000, 1993; Wächtershäuser, 1997,
Segre et al., 2000, New and Pohorille, 2000). Only at some subsequent evolution-
ary phase was a genomic replicating capability (RNA or DNA based) incorporated
into the existing system.
Let us begin by applying both a theoretical and an experimental test of this
hypothesis. If the emergence of life was associated with the spontaneous formation
of a metabolic system, two questions can be asked. Firstly, is there any experi-
mental evidence that supports the proposal that a metabolic system can emerge
spontaneously, and secondly, if we are unable to provide such experimental evid-
ence, is there some physical or chemical principle that would support or predict the
establishment of such a system?
At the current time there seems to be no experimental evidence to suggest that
metabolic systems can spontaneously form, for example, through the mixing of
their relevant components. The suggestion made by proponents of the ‘metabolism
first’ school, such as Dyson and Kauffman, that disordered and inactive molecular
systems can transform themselves, either by random drift or by sudden ‘crystallisa-
tion’, into actively metabolic ones (through whatKauffman termscatalytic closure)
remains without experimental support. In this context it is important to note that the
goal here would not be to generate the actual metabolic system that lead to life, but
toany system that could becharacterized asmetabolic, asademonstration that such
systems can exist. Till now however such evidence is lacking. Indeed, precisely
the reverse pattern is what we invariably see – metabolic systems are relatively
fragile and easily destroyed, resulting in the death of the living entity. Though it is
true that computer models have been developed that support the formation of non-
genomic replicating systems (New and Pohorille, 2000; Segre et al., 2000), such
theoretical models have yet to be validated by experiment. Ultimately experiments
must validate theoretical models, not the other way around.
CAUSATION AND THE ORIGIN OF LIFE
Of course the fact that an experimental metabolic system has not been observed
to date does not in itself preclude such a possibility; it may well be that the right
‘formula’ has not as yet been tried. However there does appear to be an inherent
theoretical problem with the concept of a spontaneously emerging metabolic sys-
tem. Orgel has recently analyzed two reaction schemes – the autocatalytic formose
reaction in which formaldehyde is converted to glycoaldehyde, and the reductive
(reverse) citric acid cycle on an FeS/FeS2surface and concluded that the likelihood
that such cycles can self-organize, whether in solution or on a mineral surface, is
remote (Orgel, 2000). If we apply Wicken’s two-tier approach to causation, the
question that needs to be asked is: what would the driving force for the appear-
ance of such a metabolic system be? Why (in a driving force sense) would a
disordered system spontaneously organise itself into an organized one displaying
functional coherence and maintaining itself far from equilibrium? Spontaneous
organization is, of course, a well-known phenomenon in chemistry. Micelle or
vesicle formation and crystallization are justtwoobvious examples. But, inanalogy
to other physical and chemical processes, micelle formation and crystallization
are driven by a thermodynamic force – the drive toward increased entropy. In
contrast, the spontaneous emergence of a far-from-equilibrium metabolic system,
as an isolated occurrence and unlinked to any other process, would be contrary
to the directive of the Second Law. In this case the thermodynamic consequence
would be the spontaneous conversion of a high entropy, disordered system into a
far-from-equilibrium, low entropy, organized one, and this is not what normally
takes place (we discuss the relevance of non-equilibrium ‘dissipative structure’
formation below). Note that the difficulty in explaining the spontaneous formation
of a metabolic system arises regardless of whether that system would have some
replicative capability or not.
Some comments regarding living systems as ‘dissipative structures’ now need
to be made. It is true that the spontaneous generation of ordered, non-equilibrium,
low entropy structures –termed ‘dissipative structures’, from an initially disordered
state is now a well-established phenomenon, and this may well have some implic-
ations for the possible emergence of a metabolic system in particular, and life in
general. However, the use of non-equilibrium ‘dissipative structures’ as a model
for enhancing our understanding of living systems, though supported by many
workers over recent years, has come under growing criticism (for recent reviews
see: Pross, 2003; Corning and Kline, 1998). The problem may be summarized as
follows: even if one dismisses the major objection that ‘dissipative structures’ do
not constitute reasonable models for biological systems, the ‘dissipative structure’
paradigm fails to provide insights into the nature of biological function, or into
the specific processes by which that function emerged. Accordingly, we are of
the view, enunciated by Collier some years ago, that there is no evidence that the
laws of non-equilibrium thermodynamics apply to biological systems in a non-
trivial fashion (Collier, 1988). Once the ‘dissipative structures’ paradigm for living
beings is put aside, we see no experimental or viable theoretical reason that sup-
ports the idea that a metabolic system, either replicative or non-replicative, can
spontaneously emerge. Consequently, we believe there is no sound basis for the
view that the emergence of life began with a chemical system that was: (i) ini-
tially both metabolic and replicative, or, (ii) initially metabolic, and subsequently
became replicative. This analysis is in contrast to the corresponding analysis for a
‘replication first’ mechanism.
2.2.2. Replication First
Let us now apply the same experimental and theoretical tests and procedures we
applied to the ‘metabolism first’ hypothesis to the ‘replication first’ hypothesis, and
inquire if there is any basis for believing a system with a replicative capability can
spontaneously emerge and subsequently become metabolic? Let us first consider
step-one of the two-stage process: Is there any unambiguous experimental evidence
for the existence of non-metabolic, replicating systems? The answer here is a defin-
itive yes. In contrast to the no answer for metabolic, non-replicating systems, the
existence of self-replicating molecular species is well established. Classical exper-
iments carried out by Spiegelman (1967) and Eigen (1992) and more recently by
Orgel (1992), Joyce, (1994), Sievers and von Kiedrowski (1994), Rebek (1994) and
others, all make clear that molecular replication is a natural chemical phenomenon.
One does not need to revert to theory to determine whether such processes occur
– their occurrence is established empirical fact. It is true that many questions re-
garding the prebiotic conditions (we discuss this aspect subsequently) that would
have enabled such a replication process to occur remain unanswered. It is also
true that non-enzymatic replication is a rather fragile and unsustainable process,
even in the hands of experienced chemists. But experiments beginning with those
of Spiegelman (1967) and extending through to the recent demonstrations of in
vitro evolution of polymerase ribozymes (Johnston et al., 2001; Joyce, 2002) have
clearly demonstrated the ability of nucleic acids to undergo cycles of replication,
mutation and evolution. Thus the fundamental point – that molecular replicat-
ing entities exist, and under appropriate circumstances can undergo a process of
replication, natural selection and Darwinian evolution, is beyond dispute.
Once the existence of evolving replicating entities is empirically established,
can we provide an in principle explanation for the incorporation of a metabolic
system into such a replicating system? We previously argued that the emergence
of a non-replicating metabolic entity seems to be forbidden by existing physico-
chemical principles. But would the emergence of a metabolic entity from a non-
metabolic one be possible within a replicating framework? Would such a conver-
sion be explicable in terms of established physico-chemical principles? In partic-
ular, could some causal connection between these two processes – replication and
metabolism – in that causal order, be identified?
In a recent publication (Pross, 2003), we pointed out that life’s processes, as
a whole, are under kinetic rather than thermodynamic control. By this we mean
that life’s processes, when viewed in chemical terms at both a molecular and a
CAUSATION AND THE ORIGIN OF LIFE
biological level, lead to the preferred formation of kinetic products that derive
from replication and are less stable, rather than to non-replicative products that
are thermodynamically more stable. The process of cell replication illustrates this
point well. An Escherichia Coli sample when placed in a solution of glucose and
essential mineral salts leads within hours to the production of billions of copies
of the bacterium. In chemical terms, the reaction that has taken place is one in
which 40% of the glucose has been converted to cellular material – the kinetic
pathway, while just 60% has been oxidized to carbon dioxide and water – the
thermodynamically preferred pathway (Monod, 1972, p. 19). But, significantly, the
component that follows the thermodynamic pathway does not do so in competition
withthekinetic pathway, but rather, asanecessary ancillary process that takes place
in order to facilitate the primary kinetic pathway. The thermodynamic pathway to
carbon dioxide and water only takes place so that the kinetic pathway to cellular
material becomes energetically feasible; just enough glucose is oxidized to cover
the energetic requirements of cell reproduction. Thus the effective driving force
for life’s processes is not the normal thermodynamic one that directs ‘regular’
chemical reactions, but a kinetic one – the enormous kinetic power of replication,
and it is this ‘force’ that channels thermodynamic forces into a kinetically support-
ing role. Thus according to this view biological complexification generally, and
the emergence of metabolic systems specifically, came about through a process of
The pathway that would describe the conversion of simple non-equilibrium rep-
licators to complex far-from-equilibrium metabolic replicators may be traced out,
at least in outline. If we presume that life began with some simple self-replicating
entity, initially non-metabolic, then kinetic selection would explore ‘complexity
space’ for more effective replicators, and one result of this search would be the
emergence of structurally more complex replicators (Pross, 2003). Simply put,
complex replicators turn out to be more effective (kinetically stable) than simple
ones (Pross and Khodorkovsky, 2004). However the trend toward greater com-
plexity and effective replicating ability has negative thermodynamic consequences
that have to be accommodated, and it is in this context that the driving force for the
emergence ofametabolic capability within simple replicating systems becomes ap-
parent. We believe that the emergence of metabolism is merely the causal outcome
of kinetic selection operating on an increasingly complex replicating entity. The
process of kinetic selection leads to the incorporation of an energy gathering cap-
ability that enables thermodynamic constraints associated with complexification to
An example may help to illustrate the principle. Consider the possibility that
a nucleic acid replicator undergoes self-assembly with a molecular photoreceptor
that is either present naturally in the replicator’s environment, or whose formation
is catalytically induced by the replicator. This specific act of self-assembly, result-
ing in the formation of a replicating assembly capable of absorbing, for example,
solar energy, would assist in enabling thermodynamic constraints associated with
ever-increasing complexification to be overcome. Photochemical activation of such
a replicating assembly would increase the free energy of the system, thereby open-
ing thermodynamic doors that would otherwise remain barred. Accordingly, this
most rudimentary metabolic system illustrates the means by which a process that
may appear to be locked into a thermodynamically unfavorable kinetic channel,
might nevertheless proceed in a thermodynamically-allowed manner. Complexific-
ation, whatever its role (metabolic or non-metabolic), is thus a direct consequence
of kinetic selection. Kinetic selection seeks out more effective replicators and since
complex, far-from-equilibrium replicators are more effective than simple, non-
equilibrium ones, that process of kinetic selection necessarily leads to a process
of metabolic complexification. The point we are making is that the emergence
of replication and metabolism can be causally related, but this causal connection
can only be understood if the emergence of replication preceded the emergence of
metabolism. The selection and incorporation of a thermodynamically-supportive
energy gathering capability is only possible within a replicative context.
Note that by postulating that replication preceded metabolism, weare proposing
a two-stage hypothesis for emergence, whose first step has experimental support,
and whose second step can be seen to be causally connected to the first step, and is
consistent with (or at least is not inconsistent with) established physico-chemical
principles. Let us recall that in considering a process in reverse order (i.e., in
which metabolism preceded replication) we need to postulate the emergence of
a metabolic system from a random system for which there is no empirical support,
and which, it has been argued (Lifson, 1997), may actually counter established
physico-chemical principles. Alternatively, it requires the spontaneous emergence
of a system with both metabolic and replicative capabilities, despite the fact that
there is no obvious reason that systems possessing both of these highly unusual
characteristics would spontaneously emerge. In comparing the two possible scen-
arios of emergence we believe the verdict is clear: the likelihood of the emergence
of a replicating entity that subsequently becomes metabolic exceeds the likeli-
hood of the spontaneous emergence of either a replicating metabolic system, or
a non-replicating metabolic system that subsequently becomes replicative.
In the light of specific criticism that has been leveled at the ‘replication first’
approach, two questions should now be addressed. Firstly, how likely was it that
a replicating molecule was formed de novo on the prebiotic earth? Clearly the
‘replication first’ scenario crucially depends on the prebiotic formation of such
a molecule. Secondly, if such a molecule did in fact appear and resulted in the
evolution of the extravagant life theater, can that molecule be identified? Much of
the criticism of the ‘replication first’ school has been based on the idea that the
likelihood of de novo synthesis is implausible (Shapiro, 2000).
In attempting to answer the first question one might facetiously rephrase the
question as follows: given an effectively unknown reaction mixture, under effect-
ively unknown reaction conditions, reacting to give unknown products by unknown
mechanisms, could a particular product with a specific characteristic – namely,
CAUSATION AND THE ORIGIN OF LIFE
the ability to self-replicate, have been included amongst the reaction products?
The provocative manner the question has been phrased serves to indicate that it
is difficult to make reliable statements of either a positive nature (what did hap-
pen) or a negative nature (what did not happen) regarding events we know little
about, and all attempts to do so categorically seem to us unjustified. A priori,
and given our current state of knowledge, the possibility of de novo synthesis
of a molecular replicator cannot be summarily dismissed. But to the extent that
there is some information regarding conditions on the prebiotic earth, it seems to
us more supportive than dismissive of the thesis. In addition to the classic Urey-
Miller experiments, that showed the relative ease with which a range of organic
compounds could have been synthesized from a hypothesized (admittedly redu-
cing) prebiotic atmosphere of ammonia, methane, and hydrogen, subsequent work
has suggested that the synthesis of the four nitrogen bases (Oro, 1961; Sanchez
et al., 1967; Stoks and Schwartz, 1979; Robertson and Miller, 1995; Ferris et al.,
1968), the synthesis of sugars (Mizuno and Weiss, 1974) and sugar phosphates
(Pitsch et al., 1995) are also feasible under either abiotic or supposed prebiotic
conditions. Thus current indications are that a wide range of biologically important
molecules could well have been present on the prebiotic earth. On the other hand it
is true that no synthesis of a replicating entity, such as an RNA oligomer, has been
achieved to date under supposed prebiotic conditions. But that reservation must
be treated with caution. Our inability to do something cannot be taken to mean
that the particular thing cannot be done! Take flying for example - it took many
centuries (and many unsuccessful attempts) before man finally succeeded in that
particular endeavor. To summarise, given the enormous gaps in our knowledge of
the prebiotic environment, the lack of information concerning de novo synthesis of
some replicating entity certainly cannot be taken as definitive evidence that such a
possibility could not have occurred.
Lastly, assuming there was a primal self-replicating molecule, what was its
structure? The simple answer is that we currently do not know, and we may never
know. Asdiscussed previously, ifitis assumed that life began withsome replicating
molecule, then inquiring into the identity of that molecule is a historical question
that we can only answer by inspecting the historical record (for example, by ex-
amining fossil records and sequence analysis). Since there is no reason to believe
that only one path from prebiotic to biotic could in principle have existed, and if
the historical record is unable to supply us with the specific information needed
to identify that molecule (as has recently been argued by Miller and Lazcano
(1995)), we will have to satisfy ourselves with an in principle description of that
early molecule. At the present time scenarios such the one provided by the ‘RNA
world’ description, in which the first replicating molecule was either of an RNA-
type, empowered with both replicative and catalytic capabilities, or some simpler
precursor such as pRNA or PNA that evolved into an RNA type molecule, appear
attractive (Orgel, 1992, 1998, 2000). However, our current understanding and the
state of the historical record are such that any detailed model should be considered
indicative, rather than definitive.
3. Concluding Remarks
Enormous mechanistic confusion can derive from the fact that mechanisms alone
do not define a process – understanding driving forces is equally important. In
the absence of mechanistic information regarding a process, an understanding of
the driving force can provide insight into the kinds of mechanistic pathways that
are possible – mechanisms consistent with that driving force. If, as we believe, the
driving force for the emergence of life is the enormous kinetic power of replication,
then mechanisms for the emergence of life must reflect this kinetic imperative. Ap-
plying this rationale to the emergence of life problem, together with an examination
of available empirical data, leads us to conclude that replication would most likely
have preceded metabolism. We arrive at this conclusion based on the following
arguments: (1) there is irrefutable empirical evidence for replicating-only molecu-
lar entities that undergo chemical evolution and selection. (2) A causal relation-
ship between replication and metabolism can be demonstrated if replication would
have preceded metabolism, but not if metabolism would have preceded replication.
(3) The ‘metabolism first’ school of thought requires postulating the spontaneous
emergence of a highly complex and coordinated chemical system, despite there
being no empirical evidence that supports such an occurrence. Furthermore, such
a process is theoretically problematical and appears to lack any identifiable driving
force. (4) If, despite point 3, a metabolic system with a non-genomic replicat-
ive capability did manage to emerge, it is unclear how the subsequent transition
to a genomic replicative system could take place; the transition would require a
discontinuity in the very element that is central to life. A sudden change in the fun-
damental mechanism of storage and replication of biological information does not
seem consistent with the principle of incremental change that seems to be central
to the evolutionary process.
Finally, our arguments on the primacy of replication go to the very heart of the
life debate. In recent years the view that life is an emergent property of complex
systems has become more common. Our arguments, however, suggest that life’s
essence, rather than being a manifestation of complexity, is quintessentially simple.
Life, first and foremost, derives from the fundamental process of replication. We
argue that life’s emergence must have begun with chemical replication, and indeed
all of life chemistry should be classified as a distinct branch of chemistry that we
would term replicative chemistry – a branch that deals with those molecular sys-
tems with the unusual capacity to self-replicate. Thus despite life’s extraordinary
complexity, this complexity is not inherent, but stems from an evolutionary pro-
cess that involves continual complexification – a fundamental and ongoing process
associated with replicative chemistry. The complexification process from single-
CAUSATION AND THE ORIGIN OF LIFE
cell entities to multi-cell ones is now established evolutionary ideology, but we
believe, in line with the ‘replication first’ school of thought, that the evolutionary
process of complexification extends all the way down to a primal replicating en-
tity. Complexity in its various facets – structural, informational, and metabolic –
are all adaptations that kinetic selection (the equivalent of natural selection op-
erating at the molecular level) introduced to further the replicating ‘agenda’. Of
course, metabolism, as an adaptation, is more fundamental than other evolutionary
adaptations, such as the giraffe’s long neck, but it is an adaptation nevertheless.
Metabolism is the particular adaptation that was employed by kinetic selection to
keep the thermodynamic tiger at bay, and as such must have been preceded by
replication; kinetic selection can only operate once a replication-mutation cycle is
underway. Thus we believe that life began with a process of molecular replication
and that activity has remained central to all its subsequent evolutionary explora-
tions through ‘complexity space’. As we have noted previously, we believe life is a
manifestation of replication, not the other way around. Going forward we believe
the main challenges confronting the emergence of life field are in further clarifying
the nature of the specific chemical processes associated with the incorporation of
that structural, metabolic and informational capability into the primal replicating
The author is indebted to Professor Vladimir Khodorkovsky for stimulating and
most fruitful discussions.
Cairns-Smith, A. G.:1985, Seven clues to the origin of life, Cambridge, Cambridge U.P.
Chyba, C. F. and McDonald, G. D.: 1995, The Origin of Life in the Solar System: Current Issues,
Annu. Rev. Earth Planet Sci. 23, 215–49.
Cleland, C. E. and Chyba, C. F.: 2002, Defining ‘Life’, Orig. Life Evol. Biosph. 32, 387–393.
Collier, J.: 1988, Entropy, Information, and Evolution, in B. H. Weber, D. J. Depew and J. D. Smith
(eds), pp. 227–242, Cambridge, MIT Press.
Corning, P. A. and Kline, J. K.: 1998, Thermodynamics, Information, and Life Revisited, Part 1: To
Be Or Entropy, Syst. Res. 15, 273–295.
Dyson, F.: 1985, Origins of Life, Cambridge, Cambridge U.P.
Eigen, M.: 1971, Self-Organization of Matter and the Evolution of Biological Macromolecules,
Naturwissenschaften 58, 465–523.
Eigen, M.: 1992, Steps Toward Life: A Perspective on Evolution, Oxford, Oxford U.P.
Elitzur, A. C.: 1994, Let There Be Life, J. Theor. Biol. 168, 429–459.
Ferris, J. P. et al.: 1968, Studies in Prebiotic Synthesis. 3. Synthesis of Pyrimidines from Cyanoethyl-
ene and Cyanate, J. Mol. Biol. 33, 693–704.
Fry, I.: 2000, The Emergence of Life on Earth, New Brunswick, Rutgers U.P.
Helmont, J. B. van: 1648, Ortus Medicinae, Amsterdam, pp. 108–109; Translated in Great Exper-
iments in Biology, M. L. Gabriel and S. Fogel (eds) p. 155, Englewood Cliffs, N.J., Prentice
Johnston, W. K. et al.: 2001, RNA-catalyzed RNA Polymerization: Accurate and General RNA-
Template Primer Extension, Science 292, 1319–1325.
Joyce, G. F.: 1994, In Vitro Evolution of Nucleic Acids, Curr Opin. Struct. Biol. 4, 331–6.
Joyce, G. F.: 2002, The Antiquity of RNA-Based Evolution, Nature 418, 214–221.
Kant,I.:1952, in:Critiqueof Judgement, R. Hutchins(ed.),Ch. 65. Chicago, Encyclopedia Brittanica
Kasting, J. F.: 1993, Earth’s Early Atmosphere, Science 259, 920–926.
Kauffman, S. A.: 2000, Investigations, Oxford, Oxford U.P.
Kauffman, S. A.: 1993, The Origins of Order. Self-Organization and Selection in Evolution, Oxford,
Lahav, N.: 1999, Biogenisis. Theories of Life’s Origins, Oxford, Oxford U.P.
Lifson, S.: 1997, On the Crucial Stages in the Origin of Animate Matter, J. Mol. Evol. 44, 1–8.
Maynard Smith, J. and Szathmáry, E.: 1995, The Major Transitions in Evolution, Oxford, Freeman.
Mayr, E.: 1988, Toward a New Philosophy of Biology, p. 44, Cambridge, Harvard U.P.
Melosh, H. J.: 1988, The Rocky Road to Panspermia, Nature, 332, 687–8.
Miller, S. L. and Lazcano, A.: 1995, The Origin of Life – Did it Occur at hHgh Temperatures? J.
Mol. Evol. 41, 689–692.
Miller, S. L. and Lazcano, A.: 1996, The Origin and Early Evolution of Life: Prebiotic Chemistry,
the Pre-RNA World, and Time, Cell 85, 793–798.
Mizuno, T. and Weiss, A. H.: 1974, Synthesis and Utilisation of Formose Sugars, Adv. Carbohyd.
Chem. Biochem. 29, 173–227.
Monod, J.: 1972, Chance and Necessity, London, Collins.
New, M. H. and Pohorille, A.: 2000, An Inherited Efficiencies Model of Non-Genomic Evolution,
Simulation Practice Theory 8, 99–108.
Oparin, A. I.:1957, The Origin of Life on Earth, 3rd ed., Edinburgh, Oliver and Boyd.
Orgel, L. E.: 2000, Self-Organizing Biochemical Cycles, Proc. Nat. Acad. Sci. USA 97, 12503–
Orgel, L. E.: 1998, The Origin of Life – A Review of Facts and Speculations, TIBS 23, 491–5.
Orgel, L. E.: 1992, Molecular Replication, Nature 358, 203–209.
Oro, J.: 1961, Mechanism of Synthesis of Adenine from Hydrogen Cyanide under Possible Primitive
Earth Conditions, Nature 191, 1193–4.
Pitsch, S. et al.: 1995, Mineral Induced Formation of Sugar Phosphates, Orig. Life Evol. Biosph. 25,
Pross, A.: 2003, The Driving Force for Life’s Emergence. Kinetic and Thermodynamic Considera-
tions, J. Theor. Biol. 220, 393–406.
Pross, A. and Khodorkovsky, V.: 2004, Extending the Concept of Kinetic Stability: Toward a
Paradigm for Life, J. Phys. Org. Chem., in press.
Putnam, H.: 1973, Meaning and Reference, J. Philos. 70, 699–711.
Rebek, J., Jr.: 1994, A Template for Life, Chem. Ber. 30, 286–90.
Robertson, M. P. and Miller S. L.: 1995, An Efficient Prebiotic Synthesis of Cytosine and Uracil,
Nature 375, 772–4.
Sanchez, R. A. et al.: 1967, Studies in Prebiotic Synthesis. II. Synthesis of Purine Precursors and
Amino Acids from Aqueous Hydrogen Cyanide, J. Mol. Biol. 30, 223–53.
Schwartz, S. P.: 1977, Introduction, in Naming, Necessity and Natural Kinds, S. P. Schwartz (ed.),
Ithaca NY, Cornell U. P.
Segre, D. et al.: 2000, Compositional Genomes: PrebioticInformation Transfer inMutually Catalytic
Noncovalent Assemblies, Proc. Nat. Acad. Sci. USA 97, 4112–7.
CAUSATION AND THE ORIGIN OF LIFE
Shapiro, R.: 1984, The Improbability of Prebiotic Nucleic Acid Synthesis, Origins Life Evol.
Biosphere 14, 565–70.
Shapiro, R.: 2000, A Replicator was not Involved in the Origin of Life, IUBMB Life 49, 173–6.
Sievers, D. and von Kiedrowski, G.: 1994, Self-Replication of Complimentary Nucleotide-Based
Oligomers, Nature 369, 221–4.
Spiegelman, S.: 1967, An In Vitro Analysis of a Replicating Molecule, American Scientist 55, 221–
Stoks, P. G. and Schwartz, A. W.: 1979, Uracil in Carbonaceous Meteorites, Nature 282, 709–10.
Wächtershäuser, G.: 1992, Groundworks for an Evolutionary Biochemistry: The Iron Sulfur World,
Prog. Biophys. Mol. Biol. 58, 85–201.
Wächtershäuser G.: 1997, The Origin of Life and Its Methodological Challenge, J. Theor. Biol. 187,
Wicken, J.S.:1985, Thermodynamics andtheConceptual Structureof EvolutionaryTheory,J. Theor.
Biol. 117, 363–383.