There
are now a small but growing number of biologists who are questioning one of the
cornerstones of neo-Darwinism. These doubts centre on the important concept of the copying error or
the random mutation. According to neo-Darwinism, copying errors supply the raw
material on which the well-understood mechanism of natural selection can act.
The reliance on the copying error as the source of evolutionary novelty or
variation has long formed the basis of the arguments against Darwin's Theory
and it allows groups such as Christian fundamentalists to increasingly voice
their objections to the currently accepted version of Evolutionary Theory.
For
some years Derek Hough has laid out the logical arguments in favour of
non-random mutations and has campaigned for an amendment to the definition of
the Theory of Evolution. Any new definition would exclude any reference to the
concept of the copying error and would be theoretically underpinned by the
important phenomenon of the algorithmic process, an understanding of which is
vital to an understanding of evolution.
Derek
Hough’s thesis can be explained by following the key stages through which
life’s genetic algorithm has passed since it first appeared, possibly
four billion years ago. Each of the following stages (except the first) can be
simulated and explored with the use of computerized genetic algorithms.
1. The origin of life. Inert chemicals happen upon self-replication.
2. Exponential growth in numbers soon leads to the ‘filling up’ of any initial space available to these primitive replicators.
3. Competition for survival leads to improved copying fidelity and differential fitness.
4. There will be a tendency for the early biosphere to automatically ‘condense’ into species. The creation of niches being driven by the elimination of ‘types’ or extinction in a knockout competition; a phenomenon well understood and explained by Darwin in On the Origin of Species.
5. Perfect copying fidelity would be a certain road to extinction. A species composed entirely of clones would soon be eliminated when faced with competition from any other species with the slightest advantage. Copying fidelity would now be selected against and regularly occurring copying errors that in any way create useful variety would be retained by natural selection and form the basis of a variety maintenance system. It is important at this stage to understand that Hough is talking about the retention of the mechanism for creating this useful variety and not the retention of the variety per se. Lynn Helena Caporale in her book Darwin in the Genome explores the possible genetics behind the evolution of the mechanism of variety-generation. [see note (i)]
6. The environment of each unit of inheritance is composed of all other units of inheritance. This environment is never stable. Each heritable unit is always faced with an endlessly varying or variable environment. With an environment that consists of a range of possibilities each species will maximise its fitness by maintaining an appropriate range of characteristics and skills. Genes that influence the outcome of reproduction will control the maintenance of this variety.
7. Design possibility is not unlimited and heritable units ‘learn’, in an algorithmic sense, that maintenance of variety around a specific, limited range of possibilities enhances the average survival chances of these units. There is an element of predictability in the large but essentially limited range of environmental variation, and life’s genetic algorithm ‘learns’ to predict some of this variation and each species can then cope better with the vicissitudes of a varying environment by maintaining an appropriate degree of variety, either in its gene pool or via regularly occurring relevant mutations. Our own immune system presents us with a model for understanding how a species can quickly adapt to an environmental challenge. The limit to design possibility is exemplified in the phenomenon of Convergence, which is clearly expounded in the book Life’s Solution by Simon Conway Morris. [see note (ii)]
8. At the same time as heritable units are maintaining variety, they are also co-operating with other heritable units. Working in teams has proved to be an advantage almost from day one, and these twin characteristics of co-operation and variety maintenance lead to the creation of unique combinations of sub-units. These new combinations of sub-units will lead to the creation of unique organisms, which, whilst not directly demanded by the environment, will survive if there is an available niche. Empty niches will rapidly be filled and convenient niches will become more rare. But the niche requiring sophistication and complexity beyond anything previously achieved will always be available and will always be sought out by this algorithmic search. Evolution can then be seen as a systematic search through the expanse of design possibility or Design Space. [see note (iii)]
Summary
The essence of Hough’s thesis is that it is not only organisms that are acted upon by natural selection but more importantly natural selection also ensures that the best systems for creating those organisms are also selected. One of the key points which investigators into evolutionary mechanisms must understand is that universal or partially universal inherited characteristics of life such as the functioning of cells, sexual reproduction, multicellularity and the genetic code itself have all been created by evolution. The system of maintaining variety from one generation to the next as outlined above is just another such universal characteristic and it is maintained because of its usefulness to life as a whole. (www.evolvability.org). The genes for such characteristics are not subject to the same environmental scrutiny as the genes for creating the physical characteristics of an organism. Derek Hough has previously described his thesis as the theory of the self-developing genome. The idea that certain genes control the reproductive success of other genes has profound implications for evolution. It explains phenomena that sit uncomfortably with neo-Darwinism such as sexual reproduction and group selection (see section on Group Selection below), both of which seem to break the rule of selfish survival. The new theory also accounts for the rapid evolution of antibiotic resistance.
Natural selection will now exert a two-fold influence in order to maintain an appropriate level of variety within each species. It acts on genes that maintain the variety-generating system and it also acts on ‘body-building’ genes to eliminate designs that sit uncomfortably at the edge of the fitness landscape. Occasionally, however, organisms of novel design escape from the straightjacket of the niche and jump into a new niche with their own fitness landscape, thereby facilitating speciation and evolution. The self-developing genome encourages pre-adaptive moves in Design Space whilst speciation separates out these new designs from the parental species and therefore allows the self-developing genome to drive further differentiation. [see note (iv)]
A wonderful emergent consequence of variety maintenance at the level of the sub-unit is the constant search for evolutionary novelty and complexity at the level of the organism.
Notes
(i)
Darwin
was very much pre-occupied with the mystery of the origin of new variation. (www.darwinvsdawkins.com). He considered the problem very
carefully. He knew nothing of genes or the arithmetic of genetic
algorithms. The closest he
came to the idea of variety generation was his theory of ‘use and
disuse’, which, in simple terms, stated that, for example, if you did a
lot of press-ups in your lifetime then your offspring would inherit bigger
biceps. The next generation would then possess a new range of bicep
potentiality (both in a physiological and a hereditary sense) and biceps could
go on getting bigger and bigger down the generations (or smaller if you
didn’t exercise them). Both Darwin’s and Hough’s theories
have similar outcomes although Hough’s theory dispenses with the need for
any information to pass from the soma (the body) to the germline (the genes).
(ii)
Organisms,
and indeed any heritable units, do not exist in isolation. Organisms only exist
in the context of a niche. An organism and its niche only ever exist together:
they define each other. Life on earth can then be seen to constitute an
integrated system, an interrelated network in which each individual unit has an
intimate knowledge (in an algorithmic sense) of their particular niche, which
is composed of other individual units. This interdependence, which stretches
back billions of years, determines the current characteristics of each
organism. These characteristics have been determined by the coordinated action
between the organism’s own genes and the genes belonging to the
constituents of its niche. This relationship leaves each organism with a
pre-programmed knowledge of its niche, a knowledge that takes account of any
environmental variation that can be tracked and prepared for. Biological
characteristics are limited by the possibilities of the genetic code and it is
this limitation, coupled with the length of time that the organism and its
niche have evolved together in unison that has driven the evolution of variety
maintaining mechanisms as a contingency against a continually varying or
variable niche. Computer simulations are similarly restricted by their code and
for this reason they are ideally suited to model evolution.
(iii) The emergence of complexity is
difficult to understand. Darwin, for example, was puzzled by the universal
phenomenon of ‘correlated variation’. He was convinced that there
must be some underlying biological mechanism that ensured that all the
different parts of an organism were correlated with each other. This
correlation is well explained by Hough’s theory. Let’s take the
evolution of the giraffe as a simplistic example. A giraffe needs long legs to
reach the higher branches. It then needs a long neck to enable it to feed at
ground level. And it then needs a strong heart to pump blood to a much greater
height. According to Hough, variety will be maintained in each of these three
characteristics. Each population of giraffes will contain a range of leg
length, neck length and heart strength. But the really interesting evolution
happens when a single individual, by chance, has each of these characteristics
at the high end of the range, i.e. long legs, a long neck and a strong
heart. This combination of characteristics will then be selected over other
combinations. Rapid evolution, which is possible as a result of a variety
maintenance mechanism, backs up Darwin’s description of an
organism’s characteristics as being ‘plastic’.
The emergence of non-trivial complexity is more
difficult to see but can be studied with the use of computerized genetic
algorithms. A good example of this approach can be found at http://myxo.css.msu.edu/papers/nature2003/Nature03_Complex.pdf
(iv)
Under
this system, differentiation (or evolution) is facilitated by an abundant
availability of unoccupied niches, which can occur, for example, following a mass
extinction. The self-developing genome will also take advantage of any
relaxation in natural selection following a sudden change of environment. For
example, domestication of animals for farming or as pets often allows genes
that were once subject to tight natural selection pressures to become available
for re-working to create other novel characteristics along with their
corresponding new niches. There was obviously a plentiful supply of empty
niches 600 million years ago when the evolution of large-scale life forms
initially took off: to attribute the creation of the variety required for the
pre-Cambrian explosion to a fortuitous series of copying errors is absurd.
Altruism and Group Selection
Arguments have always raged on the issue of whether
altruism can easily evolve and whether group selection can somehow facilitate
the evolution of this trait. The opinions of Richard Dawkins and the late John
Maynard Smith both lean towards the idea that selfish genes, within the group,
will always overcome any advantage that altruistic genes gain due to the
altruists’ contribution to group fitness. Indeed they argue strongly
against the importance of group selection as an evolutionary driving force
generally.
Wikipedia gives an example of how altruism might be
favoured. The article on Group Selection asks the reader to imagine a group
composed of four selfish organisms and one altruistic organism. When a selfish
organism meets the altruistic organism he wins 6 units of fitness whilst the
altruist gets only 4 units of fitness. But the selfish organisms avoid other
selfish organisms (because they know that they will do badly) and instead queue
up for a meeting with the sole altruistic gene. In this scenario each selfish
organism only gets 6 units of fitness but the altruist, because he has met
selfish organisms 4 times, actually does better with 4 times 4 units of
fitness. In the survival stakes, his 16 units easily outcompetes the 6 units of
each selfish individual and his genes for altruism comfortably get into the
future. This scenario, as outlined in the Wikipedia article, is not only highly
contrived but it is also not sustainable because sooner or later an even more
selfish gene will come along which, in the meeting with the altruistic gene, will take all the fitness units for
himself leaving the altruist with zero fitness units. Selfish genes will always eventually win
out; they will not be thinking of the survival of their species; sheer, instant
brutality will always give an individual an immediate advantage.
So far, advantage Dawkins and Maynard Smith.
It is easy to demonstrate with the use of computerised
genetic algorithms that any advantage that altruistic genes might give to
themselves because of their contribution to group fitness cannot overcome the
gene-level or organism-level brutality of selfish genes. Any advantage an
altruist is given because of group selection will eventually be eliminated by
the evolution of an even more vicious degree of brutality by selfish genes.
This advantage, which is due to the presence of the altruists, is eliminated
simply because the altruist himself is eliminated. Again, in this scenario,
selfish genes always win out. Selfish genes easily overcome the group-level
advantage given to altruistic genes because the altruists are eliminated at a
faster rate than group selection can increase their numbers.
It sounds like game, set and match to Dawkins and
Maynard Smith.
But surely we do observe altruism in nature? And
not only due to kin relatedness. How on earth can it evolve? The answer lies
with the self-developing genome. All that is needed is the appearance of a
mutator gene, a gene that works behind the scenes, to ensure that the outcome
of reproduction is, for example, always 50% selfish organisms and 50%
altruistic organisms. Unlike the scenario outlined above, where within-group
selection is more powerful than between-group selection, the odds in favour of the altruists are now improved because selfish
organisms within groups cannot eliminate mutator genes hidden inside other
selfish organisms. New altruistic genes can always appear because of the
presence of these mutator genes which are in effect acting as Trojan horses in
the gene pool of the group. In the organism versus organism competition natural
selection is blind to these mutator genes. As well as natural selection, the
presence of mutator genes ensures that the laws of probability now play a part
in the survival of specific characteristics. We can easily envisage the extreme
case of two groups, one with mutator genes (and therefore with some altruistic
organisms) and one without. Because of the units of fitness which the altruists
now contribute at the group level the groups containing mutator genes will now
be selected along with their potential for creating more altruistic genes. The groups without mutator genes (and
therefore without the accompanying altruistic genes) have no such advantage.
This differential advantage, at the group level, will continue to be created
and the fitness of the groups can be gradually ratcheted up via group
selection. What are really being selected are the mutator genes and groups
without them will be eliminated.
Mutator genes survive because of their presence in both winners and
losers in competitions between organisms; they do not get eliminated because
they are not exposed to competition in the same way as the genes that directly
create differentiated bodies and behaviours.
Even with the existence of these mutator genes selfish
genes will never abandon their attempt to eliminate the altruistic genes. In the example as outlined above where
the presence of the mutator gene ensures reproductive output of 50% selfish
organisms and 50% altruists the altruists will always remain as a minority
within each group. If there is an advantage at the group level leading to
selection for an even higher proportion of altruistic genes then these
selective pressures can ultimately lead to the selection of genetic systems of
the kind seen amongst social insects.
All the above scenarios can be explored with
computerised genetic algorithms. Anyone interested in evolution should write
their own simple programs for research. It is both rewarding and fun!